Aa Jamaica PDF

AVIATION ACCIDENT INVESTIGATION REPORT
RUNWAY OVERRUN ON LANDING
AMERICAN AIRLINES
FLIGHT AA331
BOEING 737-823
UNITED STATES REGISTRATION N977AN
NORMAN MANLEY INTERNATIONAL AIRPORT

KINGSTON, JAMAICA (MKJP)
22 DECEMBER 2009
REPORT NUMBER JA-2009-09
The Jamaica Civil Aviation Authority (JCAA) investigated this occurrence in accordance
with Annex 13 to the Convention on International Civil Aviation. It is not the purpose of
aircraft accident investigation to apportion blame or liability. The sole objective of the
investigation and the Final Report is the prevention of accidents and incidents.
Accredited Representative: The National Transportation Safety Board of the United States
of America.
Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

2
Synopsis
American Airlines Flight AA331, a Boeing 737-823 in United States registration N977AN,
carrying 148 passengers, including three infants, and a crew of six, was being operated under the
provisions of 14 Code of Federal Regulations (CFR) Part 121. The aircraft departed Miami
(KMIA) at 20:22 Eastern Standard Time (EST) on 22 December 2009 (01:22 Universal
Coordinated Time (UTC) on 23 December 2009) on an instrument flight rules (IFR) flight plan,
on a scheduled flight to Norman Manley International Airport (NMIA), ICAO identifier: MKJP,
Kingston, Jamaica.
The aircraft landed at NMIA on runway 12 in the hours of darkness at 22:22 EST (03:22 UTC)
in Instrument Meteorological Conditions (IMC) following an Instrument Landing System (ILS)
approach flown using the heads up display (HUD) and becoming visual at approximately two
miles from the runway. The aircraft touched down at approximately 4,100 feet on the 8,911 foot
long runway in heavy rain and with a 14 knot left quartering tailwind.
The crew was unable to stop the aircraft on the remaining 4,811 feet of runway and it overran the
end of the runway at 62 knots ground speed. The aircraft broke through a fence, crossed above a
road below the runway level and came to an abrupt stop on the sand dunes and rocks between the
road and the waterline of the Caribbean Sea.
There was no post-crash fire. The aircraft was destroyed, its fuselage broken into three sections,
w h i l e t h e left landing gear collapsed. The right engine and landing gear were torn off, t h e
left wingtip w a s badly damaged and the right wing fuel tanks were ruptured, leaking jet fuel onto
the beach sand.
One hundred and thirty four (134) passengers suffered minor or no injury, while 14 were seriously
injured, though there were no life-threatening injuries. None of the flight crew and cabin crew
was seriously injured, and they were able to assist the passengers during the evacuation.
__________________________________________________
Throughout the text, the male gender refers to both male and female.

3 TABLE OF CONTENTS
Table Of Contents
Synopsis ..................................................................................................................................................................2
Table Of Contents ...........................................................................................................................................3
ABBREVIATIONS ..............................................................................................................................................15
SECTION 1...........................................................................................................................................................19
FACTUAL INFORMATION ...............................................................................................................................19
1.0 Factual Information.........................................................................................................................................20
1.1 History of the Flight ....................................................................................................................................20
1.1.1 Departure..............................................................................................................................................20
1.1.2 Enroute .................................................................................................................................................21
1.1.3 Descent and Approach .........................................................................................................................24
1.1.4 Landing ................................................................................................................................................29
1.1.5 Information given to AA331 flight crew during approach ..................................................................31
1.2 Injuries To Persons .................................................................................................................................33
1.3 Damage To Aircraft ................................................................................................................................33
1.4 Other Damage .........................................................................................................................................33
1.5 Personnel Information .............................................................................................................................34
1.5.1 ATC Operational Personnel .................................................................................................................34
1.5.2 Flight Crew – General ..........................................................................................................................34
1.5.3 Flight Crew and Cabin Crew – Flight/Duty/Rest Times ......................................................................34
1.5.4 Captain .................................................................................................................................................35
1.5.5 First Officer ..........................................................................................................................................36
1.5.6 Cabin Crew ..........................................................................................................................................38
1.5.7 Crew Pairing ........................................................................................................................................38
1.6 Aircraft Information ................................................................................................................................38
1.6.1 General .................................................................................................................................................38
1.6.2 Maintenance .........................................................................................................................................39
1.6.3 Weight and Balance .............................................................................................................................39
1.6.4 Landing Speeds ....................................................................................................................................39
1.6.5 Automatic Flight System and HUD .....................................................................................................39
1.6.6 Weather Radar .....................................................................................................................................40
1.6.7 Wind Shear Detection and Prediction ..................................................................................................40
4 TABLE OF CONTENTS
1.6.8 Ground Spoilers ...................................................................................................................................40

1.6.9 Brake System .......................................................................................................................................40
1.6.10 Engine Controls .................................................................................................................................41
1.6.11 Rain Removal System ........................................................................................................................41
1.6.12 GPS Navigation Equipment ...............................................................................................................41
1.6.13 Aircraft description – Cabin, Seats and Exits ....................................................................................42
1.6.14 Aircraft Seats and Restraint Systems .................................................................................................43
1.6.15 Emergency Exits – Flight Deck .........................................................................................................43
1.6.16 Emergency Exits - Cabin ...................................................................................................................43
1.6.17 Evacuation Escape Slides .................................................................................................................44
1.6.18 Emergency Lighting...........................................................................................................................46
1.6.18.1 Emergency Lighting - General ........................................................................................................46
1.6.18.2 Interior Emergency Lighting ..........................................................................................................46
1.6.18.3 Exterior Emergency Lighting..........................................................................................................46
1.6.19 Public Address and Communication System .....................................................................................47
1.6.20 Emergency Equipment .......................................................................................................................47
1.7 Meteorological Information ........................................................................................................................47
1.7.1 METARS on AA331 Dispatch Document ...........................................................................................47
1.7.2 Graphic Area Forecasts ........................................................................................................................47
1.7.3 Weather over Jamaica ..........................................................................................................................48
1.7.4 En route Weather .................................................................................................................................48
1.7.5 Thunderstorms and Lightning ..............................................................................................................48
1.7.6 Hourly Weather Observations..............................................................................................................49
1.7.7 Aerodrome Forecasts ...........................................................................................................................50
1.7.7.1 Kingston/Norman Manley International Airport (NMIA) ................................................................50
1.7.7.2 Montego Bay - Sangster International Airport.................................................................................51
1.7.7.3 Grand Cayman - Owen Roberts International Airport - MWCR ......................................................51
1.7.8 Area Forecast and Flash Flood Warning..............................................................................................51
1.7.9 Kingston/Norman Manley International Airport, Wind Information...................................................52
1.7.10 Weather Information From Radar ......................................................................................................52
1.7.11 Recorded Rainfall Rates At MKJP Norman Manley International Airport .......................................53
1.7.11.1 Recorded Rainfall at MKJP Norman Manley International Airport ..............................................53
1.7.11.2 Definitions of Rainfall ....................................................................................................................54
5 TABLE OF CONTENTS
1.7.12 Weather Conditions on the Ground at Kingston ................................................................................54

1.7.13 AA Field Condition Report for Kingston ..........................................................................................54
1.7.14 Weather Information Provided by Air Traffic Control ......................................................................55
1.7.15 Rain Reports During Approach from Flight Crew .............................................................................56
1.8 Aids To Navigation .................................................................................................................................56
1.8.1 General .................................................................................................................................................56
1.8.2 Instrument Approaches at MKJP, Norman Manley International Airport ...........................................56
1.8.2.1 Runway 12 Instrument Approaches ..................................................................................................56
1.8.2.2 Runway 30 Instrument Approaches ..................................................................................................58
1.9 Communications .....................................................................................................................................58
1.10 Airport Information ...............................................................................................................................59
1.10.1 General ...............................................................................................................................................59
1.10.2 Runway Description...........................................................................................................................60
1.10.2.1 Runway 12 ......................................................................................................................................60
1.10.2.2 Runway 30 ......................................................................................................................................60
1.10.2.3 Runway 12/30, Transverse and Longitudinal Slopes ......................................................................61
1.10.2.4 Runway Survey ...............................................................................................................................62
1.10.2.5 Runway Drainage System ..............................................................................................................63
1.10.2.6 Runway 12 Lighting and Markings ................................................................................................63
1.10.2.7 Runway 30 Lighting and Markings ................................................................................................65
1.10.2.8 Runway End Safety Area (RESA) ..................................................................................................66
1.10.3 Runway Surface Friction Measurement .............................................................................................67
1.10.4 Runway Surface Condition Reports ...................................................................................................68
1.10.4.1. AIP Jamaica ...................................................................................................................................68
1.10.4.2 Air Traffic Control (ATC) .............................................................................................................68
1.10.4.3 AA Flight Manual Part II Jeppesen Page 10-7X.............................................................................69
1.10.4.4 American Airlines Field Condition Reports ..................................................................................69
1.10.4.5 Runway Condition Reports – ICAO Requirements ........................................................................71
1.10.4.6 Runway Condition Reports at Kingston .........................................................................................72
1.10.5 Airport Rescue and Fire Fighting (ARFF) .........................................................................................74
1.10.6 Meteorological Services and ATIS weather information..................................................................74
1.10.7 Airport Lighting .................................................................................................................................75
1.11 Flight Recorders ....................................................................................................................................75
6 TABLE OF CONTENTS
1.11.1 Cockpit Voice Recorder (CVR) .........................................................................................................75

1.11.2 Flight Data Recorder (FDR) ..............................................................................................................75
1.11.2.1 Flight Data Recorder .......................................................................................................................75
1.11.2.2 Information from Flight Data Recorder ..........................................................................................76
1.12 Wreckage And Impact Information ......................................................................................................82
1.12.1 Debris Path and Other Damage ..........................................................................................................82
1.12.2 Damage to Aircraft - General.............................................................................................................83
1.12.3. Damage to Engines ...........................................................................................................................86
1.12.4 Damage to Landing Gear ...................................................................................................................86
1.12.4.1 Nose Gear........................................................................................................................................86
1.12.4.2 Main Landing Gear .........................................................................................................................87
1.12.4.2.1 General .........................................................................................................................................87
1.12.4.2.2 Right Main Landing Gear ............................................................................................................87
1.12.4.2.3 Left Main Landing Gear ..............................................................................................................88
1.12.4.2.4 Evidence of Hydroplaning ...........................................................................................................90
1.12.5 Damage to Exits .................................................................................................................................90
1.12.5.1 Forward Left Door (L1) and Escape Slide ......................................................................................90
1.12.5.2 Forward Right Door (R1) and Escape Slide ...................................................................................91
1.12.5.3 Aft Left Door (L2) ..........................................................................................................................92
1.12.5.4 Aft Right Door (R2) ........................................................................................................................93
1.12.5.5 Automatic Over-wing Exits ............................................................................................................94
1.12.6 Damage to Flight Deck ......................................................................................................................94
1.12.6.1 General ............................................................................................................................................94
1.12.6.2 Flight Deck Seats ............................................................................................................................95
1.12.6.3 Flight Deck Floor ............................................................................................................................96
1.12.6.4 Flight Deck Sidewalls .....................................................................................................................96
1.12.6.5 Flight Deck Windows .....................................................................................................................96
1.12.6.6 Flight Deck Instrument Panel .........................................................................................................96
1.12.6.7 Flight Deck Door ............................................................................................................................97
1.12.6.8 Forward Entry, Cabin Crew Seats and Forward Galley Area .........................................................97
1.12.7 Main Cabin.........................................................................................................................................98
1.12.7.1 Forward Section - First Class Cabin – Row 3 through Row 6 ........................................................98
1.12.7.2 Centre Section – Economy Class Cabin – Row 7 through Row 22 ................................................99
7 TABLE OF CONTENTS
1.12.7.3 Aft Section - Economy Class Cabin – Row 23 through Row 28 ....................................................99
1.12.7.4 Aft Entry, Cabin Crew Seats and Aft Galley Area .......................................................................101
1.13 Medical And Pathological Information...............................................................................................102
1.13.1 Captain .............................................................................................................................................102
1.13.2 First Officer......................................................................................................................................102
1.14 Fire ......................................................................................................................................................102
1.15 Survival Aspects .................................................................................................................................103
1.15.1 Notification of Accident...................................................................................................................103
1.15.2 Emergency Lighting in aircraft ........................................................................................................103
1.15.3 Communications ..............................................................................................................................103
1.15.4 Evacuation........................................................................................................................................103
1.15.5 Emergency Exits ..............................................................................................................................105
1.16 Tests And Research.............................................................................................................................105
1.16.1 Escape Slide Tests...........................................................................................................................105
1.16.2 Emergency Lighting in aircraft ........................................................................................................106
1.16.3 Simulator Trials ...............................................................................................................................107
1.16.4 Display Guidance Computer ............................................................................................................107
1.16.5 Tests on Braking System..................................................................................................................108
1.16.5.1 Antiskid/Autobrake Control Unit..................................................................................................108
1.16.5.2 Antiskid Valves .............................................................................................................................109
1.16.5.3 Wheel Speed Transducers .............................................................................................................109
1.16.5.4 Autobrake Valve Module ..............................................................................................................109
1.17 Organization and Management Information .......................................................................................110
1.17.1 AA Manuals, Policies and Procedures .............................................................................................110
1.17.1.1 AA Procedures for Approach and Landing, and go-around .........................................................110
1.17.1.1.1 Stabilized Approach Requirement .............................................................................................110
1.17.1.1.2 CAT I ILS Approach Procedures ..............................................................................................111
1.17.1.1.3 Landing on a Slippery Runway/Hydroplaning ..........................................................................112
1.17.1.1.4 Factors Affecting Landing Distance ..........................................................................................114
1.17.1.1.5 AA Approach Preparation/Calculation of Landing Distance ....................................................114
1.17.1.1.6 AA Policy on the use of reverse thrust on landing ....................................................................115
1.17.1.1.7 AA Procedures for Landing with Tailwinds ..............................................................................116
1.17.1.1.8 AA Normal Procedures for Landing from ILS ..........................................................................117
8 TABLE OF CONTENTS
1.17.1.1.9 AA Information on Baulked Landings (Go Around) .................................................................118

1.17.1.1.10 AA No Fault Go-Around Policy and Missed Approach decision ............................................119
1.17.1.1.11 AA Bulletin 737-07 .................................................................................................................120
1.17.1.1.12 AA HUD Usage Policy and Briefing Guide ............................................................................120
1.17.1.1.13 AA B737 RRLL Tables, Wind Component and Landing Data Card .....................................121
1.17.1.2 AA– Other Information.................................................................................................................121
1.17.1.2.1. AA Crew Resource Management (CRM) Training ..................................................................121
1.17.1.2.2 AA Flight Manual Page 10-7X. .................................................................................................121
1.17.1.2.3 AA Boeing 737 Training ...........................................................................................................121
1.17.1.2.4 AA Dispatch Information .........................................................................................................122
1.17.1.2.5 AA Required Runway Landing Length (RRLL) table, and Landing Data Card. .....................125
1.17.1.2.6 AA Landing Distance Calculations............................................................................................127
1.17.1.2.6.1 Landing Data Card ..................................................................................................................127
1.17.1.2.6.2 AA Performance – Landing Table ..........................................................................................129
1.17.1.2.7 Advisory Circular No: 91-79 .....................................................................................................129
1.17.1.2.8 Crew Flight/Duty/Rest System .................................................................................................129
1.17.1.2.9 AA– Interviews with Staff .........................................................................................................129
1.17.1.2.10 AA– information regarding Bulletin 737-07 and AC 91-79 ...................................................130
1.17.1.2.11 AA Safety Initiatives Re: AA331 Landing Accident...............................................................132
1.17.2 Boeing Information ..........................................................................................................................133
1.17.2.1 Glide path, slippery runway, friction measurement .....................................................................133
1.17.2.2 Boeing Flight Crew Operations and Training Manuals, B737, Landing .....................................134
1.17.2.3 Boeing Information on Rejected/Baulked Landings ....................................................................136
1.17.2.4 Autopilot/Autothrottle Use During Approach ..............................................................................137
1.17.3 Air Traffic Control Procedures ........................................................................................................138
1.17.3.1 Assignment of Active Runway .....................................................................................................138
1.17.3.2 Weather Standby ...........................................................................................................................138
1.17.3.3 Information for arriving aircraft ....................................................................................................138
1.17.4 FAA Information .............................................................................................................................139
1.17.4.1 FAA Regulatory Requirements Regarding Landing Distance Calculation ..................................139
1.17.4.2 History of SAFO 06012, OpSpec/MSpec C082 and AA Bulletin 737-07 ....................................139
1.17.4.3 FAA Recommendations Regarding Landing Distance Calculation.............................................140
1.17.4.4 FAA Advisory Circular AC 91-79 ................................................................................................141
9 TABLE OF CONTENTS
1.17.4.5 SAFO 10005 .................................................................................................................................142

1.17.5 NTSB Recommendations to FAA....................................................................................................143
1.17.5.1 NTSB Recommendations, General ...............................................................................................143
1.17.5.2 NTSB Safety Recommendation, in reply to A-07-58 through -64 ...............................................143
1.18 Additional Information ......................................................................................................................144
1.18.1 Air Traffic Control ...........................................................................................................................144
1.18.1.1 Enroute Radar Service ..................................................................................................................144
1.18.1.2 Approach Radar Control Service .................................................................................................144
1.18.1.3 Aerodrome Control Service ..........................................................................................................144
1.18.2 JCAA ATS QA Audits, and Controller Training and Checking....................................................145
1.18.3 JCAA Automatic Terminal Information Service (ATIS) ................................................................145
1.18.4 JCAA Notices To Airmen (NOTAMS) ..........................................................................................145
1.18.5 JCAA Air Traffic Control Radar ...................................................................................................147
1.18.6 Estimated Water Depth and Braking Action ..................................................................................147
1.18.7 NTSB Review of Fatigue in Major U.S. Accidents .........................................................................149
1.18.8 AA use of FAA Safety Recommendations in AC No: 91-79 .........................................................149
1.18.9 GPS Approach Requirements and Information...............................................................................149
1.18.10 Stopping Performance Information ...............................................................................................150
1.18.11 Runway Water Depth Measurement .............................................................................................150
1.18.12 Hazards associated with tailwind landings. ..................................................................................151
1.18.13 The Hazards Associated With Go-Around After Touchdown .....................................................151
1.18.14 Flight Safety International Approach and Landing Accident Reduction Report ...........................152
1.18.15 Boeing 737-800 Tailwind Certification Process ...........................................................................152
1.18.16 Definition of Black Hole Approach ...............................................................................................152
APPENDICES TO FACTUAL INFORMATION .............................................................................................153
ANALYSIS .........................................................................................................................................................233
2.0 Analysis ........................................................................................................................................................234
2.1 Aircraft ..................................................................................................................................................234
2.2 Flight Crew Qualifications ....................................................................................................................234
2.2.a NOTAMS ...........................................................................................................................................234
2.3 Airport ...................................................................................................................................................235
2.4 Runway .................................................................................................................................................236
2.4.1 Runway Surface .................................................................................................................................236
10 TABLE OF CONTENTS
2.4.2 Runway Slope ....................................................................................................................................236

2.5 Air Traffic Services (ATS) ..................................................................................................................237
2.6 Meteorological Information .................................................................................................................240
2.7 Operations ............................................................................................................................................240
2.8 RNAV (GPS) Runway 30 Approach at Kingston ................................................................................241
2.9 Crew Resource Management (CRM) ...................................................................................................242
2.10 Go-Around .........................................................................................................................................245
2.11 Flight Crew Fatigue ...........................................................................................................................246
2.12 Pitch-Up Inputs and Visual Illusions .................................................................................................246
2.13 Simulator Trials .................................................................................................................................247
2.14 Runway Condition Report “Wet” ......................................................................................................247
2.15 Inconsistency between AC91-79 and SAFO 06012...........................................................................249
2.16 Kingston Field Report ........................................................................................................................249
2.17 Determination of Depth of Water on Runway, and Runway Braking Action ...................................251
2.18 Survival Aspects and Injuries ............................................................................................................251
2.19 AA Bulletin 737-07 and Landing Data Card .....................................................................................252
2.20 Bulletin 737-07 and “Advance Analysis” .........................................................................................252
2.21 AA Flight Crew Landing Distance Assessment..................................................................................253
2.22 Calculation of landing distances .......................................................................................................254
2.23 Flight crew compliance with Bulletin 737-07 ....................................................................................258
2.24 Discussion of “advance analysis” ......................................................................................................258
2.25 Placement of Bulletin 737-07 in AA Manuals ....................................................................................259
2.26 Issue of Autothrottle/Autopilot use ....................................................................................................260
2.27 FAA Safety Information SAFO 06012 and AC 91-79, and AA Bulletin 737-07 ...............................261
2.28 Landing Conditions and Landing Performance Assessment .............................................................262
2.29 Tailwind Landing Hazards and Tailwind Landing Training...............................................................266
2.30 AA Management And Operational Control .......................................................................................267
2.31 FAA Advisory Circular (AC) 91-79 ..................................................................................................267
2.32 Situational Awareness .........................................................................................................................268
2.32.1 Situational Awareness – Definition .................................................................................................268
2.32.2 Situational Awareness before Departure..........................................................................................268
2.32.3 Situational Awareness during Approach and Landing.....................................................................269
2.33 System Safety.....................................................................................................................................272
2.34 Conclusion .........................................................................................................................................273

2.35 Findings as to Causes and Contributing Factors ................................................................................275
FINDINGS .........................................................................................................................................................277
3.0 Findings.........................................................................................................................................................278
3.1 Findings as to Causes and Contributing Factors ...................................................................................278
3.2 Findings as to Risk ................................................................................................................................283
SAFETY ACTIONS ...........................................................................................................................................285
4.0 Safety Actions ...............................................................................................................................................286
4.1 Safety Actions Taken ............................................................................................................................286
4.1.1 AA- General .......................................................................................................................................286
4.1.2 AA- Go Around .................................................................................................................................287
4.1.3 AA – Reversion to 10 knot tailwind landing limit .............................................................................287
4.1.4 Jamaica Civil Aviation Authority ......................................................................................................288
4.1.5 The Boeing Company ........................................................................................................................288
4.1.6 Airports Authority of Jamaica............................................................................................................288
4.2 SAFETY ACTION REQUIRED ..........................................................................................................288
4.2.1 Operators of Transport Category Aircraft ..........................................................................................288
4.2.2 Advisory Circular 91-79 ....................................................................................................................289
4.2.3 AC 91-79 and SAFO 06012 ..............................................................................................................289
4.2.4 Training for Tailwind Landing ..........................................................................................................289
4.2.5 Go-around Callout .............................................................................................................................289
4.2.6 Tailwind Landings on Contaminated Runways .................................................................................289
4.2.7 Tailwind Landings on Wet (not contaminated) Runways..................................................................290
4.2.8 Deceleration Techniques on Contaminated Runways........................................................................290
4.2.9 Landing Distance Assessment ...........................................................................................................290
4.2.10 Runway End Safety Area Requirements ..........................................................................................290
4.2.11 Runway Surface Condition Reporting .............................................................................................290
4.2.12 Runway Surface Condition Inspection and Reporting at MKJP ......................................................291
4.2.13 Runway Markings at MKJP .............................................................................................................291
4.2.14 Distance-To-Go Markers at MKJP ................................................................................................291
4.2.15 Runway Lighting .............................................................................................................................291
4.2.16 Cabin Crew Jump Seat Emergency Equipment Stowage Area ........................................................291
4.2.17 American Airlines L1 Slide .............................................................................................................292
4.2.18 American Airlines automation mode matching re: autopilot/autothrottle use .................................292
4.2.19 Emergency Lighting.........................................................................................................................292
4.2.20 AA Runway Condition Reporting....................................................................................................292
4.2.21 AA “Advance Analysis” of runway condition .................................................................................292
4.2.22 American Airlines, use of the word “Wet” to describe runway conditions .....................................293
4.2.23 American Airlines, information from FAA Advisory Circular AC 91-79 .......................................293
4.2.24 Jamaica Air Traffic Services ............................................................................................................293
4.2.25 Boeing 737 overhead bins and PSUs ...............................................................................................293
4.2.26 Captain’s seat belt crotch bracket ....................................................................................................293
4.2.27 Captain’s uncertified seat cushion ...................................................................................................294
4.2.28 Runway Grooving ...........................................................................................................................294
4.2.29 Airports Group Recommendations .................................................................................................294
4.2.29.1 Aerodrome Certification ..............................................................................................................294
4.2.29.2 Aerodrome Maintenance ...............................................................................................................294
4.2.29.3 Aerodrome Emergency Plan .........................................................................................................295
4.2.29.4 Pavement .......................................................................................................................................295
4.2.29.5 Marking/lighting signs ..................................................................................................................295
4.2.29.6 Rescue Firefighting Service .........................................................................................................296
4.2.29.7 Declared Distances .......................................................................................................................296
4.2.29.8 Runway strip and runway end safety area ....................................................................................296
4.2.29.9 Maintenance of PAPI ...................................................................................................................296
4.2.30 CVR .................................................................................................................................................297
4.2.31 FAA Advisory Material ref: Wet Runways .....................................................................................297
4.2.32 Cancelling of NOTAMS ..................................................................................................................297
4.2.33 Recording of ATIS ...........................................................................................................................297
4.2.34 ATS Procedures ..............................................................................................................................297
4.2.35 ATS and NMIA MANOPS .............................................................................................................297

PHOTOGRAPHS
Photo 1 Aerial Photograph of Accident Site………………………………………………....59
Photo 2 Aircraft Wreckage…………………………………………………………………...82
Photo 3 Forward Fuselage …………………………………………………………………... 83

Photo 4 Centre Fuselage…………………………………………………………………….. 83
Photo 5 Aft Fuselage .……………………………………………………………………… 84
Photo 6 Left Wing.. ………………………………………………………………………......84
Photo 7 Right Wing………………………………………………………………………..... 85
Photo 8 Nose Landing Gear.. …………………………………………………………………86
Photo 9 Right Main Landing Gear ………………………………………………………… 87
Photo 10 Left Hand Main Landing Gear…………………………………………………… 88
Photo 11 Forward Left Door (L1) ………………………………………………………….. 90
Photo 12 Forward Right Door (R1).…………………………………………………….........91
Photo13 Aft Left Door ………………………...………….……….………..….…......... 92
Photo14 Aft Right Door…………………………………...………….……….…..…….. 93
Photo 15 View into Flight Deck .…………………………………...…………..………….... 94
Photo 16 The Captain’s Crotch Belt Bracket.…………………………...……………........... 96
Photo 17 Flight Deck Door…………………………………………...……………………... 97
Photo 18 First Class Cabin…………………………………………...…………………..... .98
Photo 19 Centre Section…………………………………………...………………………. .99
Photo 20 Aft Section…………………………………………..………………………… . 100

FIGURES
Figure 1 B737-823 Aircraft Cabin and Exits. …………………………………………................42

Figure 2 Escape Slide Pack. …………………………………………..……………....………… 44
Figure 3 Escape Slide Deployment…………………………………………..………………….. 45
Figure 4 Landing Sequence Key Events…………………………………………..……………. 80
TABLES
Table 1 Injuries to Persons……………………................................................................................33
Table 2 AA Personnel Information...................................................................................................34
Table 3 Aircraft Information.. ……..................................................................................................38
Table 4 Recorded Rainfall at MKJP…………………………………………...………................. 53
Table 5 ATC Surface Wind Reports………………………………………………...………........ 55
Table 6 Landing Data Card ………………………………………………...………................... 126
Table 7 Tailwind Component…………………………………………………………………... 256
Table 8 Autobrake Data………………………………………………………………………… 257
APPENDICES TO SECTION 1, FACTUAL INFORMATION
Appendix 1 Flight Data Recorder Plot by the Boeing Company................................................. .155

Appendix 2 Cockpit Voice Recorder Transcript........................................................................... 163
Appendix 3 Kingston ATC Transcripts........................................................................................ .183
Appendix 4 Jeppesen Approach Plates for Kingston MKJP/KIN..........………........................... 191
Appendix 5 AA Flight Manual Part 11........………............................................................. ........197
Appendix 6 AA 737 Operating Manual,.AA Bulletin 737-07..................................................... ..199
Appendix 7 AA Boeing 737-800 Required Runway Landing Length Tables.............................. 205
Appendix 8 AA Estimated Wet and Contaminated Runway Landing Distances......................... 209
Appendix 9 Boeing Landing Distance Charts ……...................................................................... .213
Appendix 10 Airport Group Report ..............................................................................................

15 ABBREVIATIONS
ABBREVIATIONS
AA American Airlines
ACARS Aircraft Communication Addressing and Reporting System
AC Advisory Circular (FAA)
ACC Area Control Center
ADIRU Air Data Inertial Reference Unit
AEP Airport Entry Pass
agl above ground level
AIP Aeronautical Information Publication
ALA Approach and Landing Accident
ALS Approach Lighting System
APS Airport Protection Services
AOC Airport Operations Centre
AQP Advanced Qualification Program
ARFF Airport Rescue and Fire Fighting
asl above sea level
ATC Air Traffic Control
ATIS Automatic Terminal Information Service
ATS Air Traffic Services
AWOS Automated Weather Observing System
C centigrade
CRM Crew Resource Management
CVR Cockpit Voice Recorder
DA Decision altitude
DH Decision height
DME Distance Measuring Equipment
EFIS Electronic Flight Instrument System
EMAS Engineered Materials Arresting System
EST Eastern Standard Time
FAA Federal Aviation Administration (U.S.)
FAR Federal Aviation Regulations
FCOM Flight Crew Operations Manual
FDR Flight Data Recorder
FL Flight Level
FMS Flight Management System
ft feet
fpm feet per minute
FPV Flight Path Vector

16 ABBREVIATIONS
GAIA Grantley Adams International Airport (Barbados)

GPS Global Positioning System
HIRL High Intensity Runway Lights
HMA Hot Mix Asphalt
HUD Heads Up Display
ICAO International Civil Aviation Organization
IFR Instrument Flight Rules
IFTSA International Fire Training Association
IMC Instrument Meteorological Conditions
ILS Instrument Landing System
JCAA Jamaica Civil Aviation Authority
JCARs Jamaica Civil Aviation Regulations
KIAS Knots indicated airspeed
kts knots
m metres
MANOPS Manual of Operations
mb millibars
MDA Minimum Descent Altitude
MEL Minimum Equipment List
METAR Meteorological Actual Report (Aviation Routine Weather Report)
MKJP Kingston Airport (Norman Manley International Airport)
MOU Memorandum of Understanding
NMIA Norman Manley International Airport
NOTAM Notice to Airmen
NTSB National Transportation Safety Board (United States)
NVM Non Volatile Memory
PA Public Address System
PAPI Precision Approach Path Indicator
PF Pilot Flying
PIREP Pilot Report
PM Pilot Monitoring
PSI Pounds Per Square Inch
PSU Passenger Service Unit
QRH Quick Reference Handbook
RA Radio Altitude (above ground)
REIL Runway End Identifier Lights
RESA Runway End Safety Area
RFF Rescue and Fire Fighting
RH Recorded History
RNAV Area Navigation
17 ABBREVIATIONS
RRLL Required Runway Landing Length

Rwy Runway
SAFO Safety Alert For Operators (FAA)
SIAL Simple Instrument Approach Light System
SOPs Standard Operating Procedures
SPECI Special meteorological report
TAF Terminal Aerodrome Forecast
TDZ Touchdown zone
UTC Coordinated Universal Time
VASI Visual Approach Slope Indicator
VFR Visual Flight Rules
VOR Very High Frequency Omni Range
VRef Reference landing speed
Z Zulu Time (same as UTC)

18
Intentionally Left Blank

19 Section 1, Factual Information
SECTION 1
FACTUAL INFORMATION

1.0 Factual Information

1.1 History of the Flight
1.1.1 Departure
American Airlines (AA) Flight AA331, a Boeing 737-823 with United States registration
N977AN, carrying 145 passengers and three infants, and operated by a crew of six,
departed Miami (KMIA) at 20:22 (Eastern Standard Time (EST), this being local time) on 22
December 2009 (01:22 Co-ordinated Universal Time (UTC) on 23 December 2009) on a
scheduled flight to the Norman Manley International Airport (NMIA: ICAO identifier MKJP),
Kingston, Jamaica. The flight landed on runway 12 at 22:22 EST, (03:22 UTC on 23 December
2009) following an ILS approach to the runway.
Flight dispatch was provided from the AA main base in Dallas, Texas. The flight crew, who had
flown two previous sectors, from Miami to Baltimore and Baltimore to Miami, changed aircraft in
Miami for the flight to Kingston.
The captain was designated as the pilot flying (PF) for the sector from Miami to Kingston, and
the first officer was designated as pilot monitoring (PM).
The flight crew checked the dispatch documents, which included the enroute and Kingston
weather and applicable Notices to Airmen (NOTAM), and proceeded with the preparation for the
flight. Forecast weather included turbulence enroute over Cuba and rain in Kingston.
The NOTAMs included notice of the closure of the Sangster International Airport (SIA; ICAO
identifier: MKJS) at Montego Bay for maintenance at 22:00 EST (03:00 UTC), close to the
original planned arrival time of AA331 at Kingston. Montego Bay was AA331’s primary
alternate, and AA Dispatch had filed for Owen Roberts International Airport (MWCR), Grand
Cayman, as a second alternate airport and so had ordered additional fuel. This brought the
calculated landing weight on arrival at Kingston very close to the aircraft’s maximum
landing weight of 144,000 pounds.
AA Dispatch added MWCR as a second, “nearest next” most suitable alternate to give the AA331
crew more options in case of adverse weather and other contingencies.
The Simple Instrument Approach Light (SIAL) system for runway 12 at Kingston was also
NOTAMed as being unserviceable.
The captain briefed the cabin crew regarding the turbulence expected enroute. The first officer
stated they did not conduct any special briefing prior to departure other than having a discussion
about weather over Cuba for forecast of light turbulence and general weather. They had no
detailed discussion about weather issues in Kingston.
The flight was initially delayed by the requirement to remove checked baggage belonging to a
passenger who had not boarded. A further short delay was caused during taxi as a result of an air
conditioning pack temperature controller fault warning. The crew paused during taxi and
discussed this with AA Dispatch and Maintenance, before using the Minimum Equipment List
(MEL) to defer the defect. There were no other deferred defects on the aircraft at this time. The
aircraft encountered some turbulence during the climb out of Miami.
1.1.2 Enroute
The flight plan was for departure from Miami, climbing initially to Flight Level (FL) 350 (that is,
35,000 feet above mean sea level) direct to reporting point EONNS, A509 to URSUS then by
UA301 to MKJP, Kingston.
The aircraft climbed to its first assigned cruising altitude of FL 350 and then to its final cruising
altitude of FL 370 and proceeded on to Kingston. While crossing Cuba, it flew into
turbulence which the flight crew described as “fairly rough” and “real bumpy”. Throughout the
flight there were several suspensions of the in-flight service. The captain ordered the cabin crew to
prepare early for landing.
The flight crew reported that the turbulence decreased as they got closer to Kingston. Just prior
to descent the flight crew received an Aircraft Communication Addressing and Reporting System
(ACARS) message from AA Dispatch advising of the closure of M K J S , the flight's primary
alternate airport, thus changing the primary alternate airport to MWCR. Because of the fuel
needed to fly to the alternate airport, M W C R , the crew discussed making one approach into
Kingston and, if this approach missed, they would proceed directly to their alternate.
The first officer stated that they received updates from MKJP Tower on the runway condition
and that it was reported as being “Wet”. He said they did not receive any advisory regarding
braking action being less than good and no report of any significant runway contamination.
He said, based on his experience flying 4,100 hours in this airplane and over a total of 10,000
hours, the weather was not abnormal and it was “just another day at the office”.
The first officer said he had been to Kingston many times before and had been there the prior
week on the same sequence. He had landed there at night and in the rain and had been there a
few times over the past year.
The flight crew reported that they briefed for arrival at Kingston prior to commencing descent.
During the briefing for the approach they decided that, based on the weather and the minimums
for the straight in approach to runway 12, it was a better option to do the straight in and land on
12 with a tailwind, rather than doing a circling approach to runway 30 with a low ceiling, and that
this had more likelihood of a successful outcome.
Minimum Descent Altitude (MDA) of 1,150 feet above sea level (asl)/1,140 feet above ground
level (agl) and visibility 3.7 kilometers for the circle-to-land procedure for runway 30.1 The most
1
See Appendix 4

recent Meteorological T e r m i n a l Air Report (METAR) for Kingston was ceiling 1,400 broken
agl.
In their interviews, both of the flight crew said there was no instrument approach to runway 30 at
Kingston, but they were both aware of the circle-to-land approach to runway 30 from the ILS
runway 12 approach.
The first officer said they did not have to calculate landing distance before each landing. They
had landing performance data and a tailwind was taken into consideration in that data. He said on
a runway like that at Kingston which was 9,000 feet long, with wet conditions, there was no
problem with the performance data for landing flaps 30 and they had both done this many times.
He also stated that for shorter than 8,000 feet, if wet, there were more factors like wet runway The
Jeppesen 11-1 Kingston, Jamaica ILS Runway 12 publication used by the crew indicated a,
tailwinds to consider; they were cumulative based on field elevation and landing weight. He said
he was very comfortable with the runway data but it was more critical when there was a shorter
runway and tailwind.
For the purposes of the investigation, this was taken to mean that the AA331 flight crew
believed that if the runway was more than 8,000 feet, and the tailwind was 15 knots or less, on a
wet runway it was not necessary to calculate landing distance before each landing. It was also
considered possible that the first officer meant a 9,000 foot. This was never determined by the
investigation, but the 8,000 foot runway scenario was considered to be most likely.
During the initial interviews with the flight crew immediately after the accident, the first officer
described his method of landing distance assessment, later called the "advance analysis", by the AA
Flight Safety Programs Manager (See 1.17.1.2.9) described in 1.1.2. No further interviews with the
flight crew were conducted, and no more exact details of the flight crew's concept of this landing
distance assessment were established, nor were there any details as to its origin and genesis.
The first officer said they sometimes landed at flaps 40 but the captain had briefed flaps 30;
which was normal for this situation. He said they had discussed that for these conditions and
for the go around, flaps 30 was the better choice. He said that flaps 40 would generally be used
for real short runways (sic). He said the landing performance charts were based on airplane
landing weight and that some airports had a special analysis for tailwind landing, but not
Kingston.
The first officer said circle-to-land approaches were not done much anymore, as most airports had
improved their facilities, but they did conduct circle-to-land approaches in training.
The first officer said they had both done the ILS runway 12 approach at Kingston many times
and were familiar with it, so it seemed more appropriate to fly the ILS runway 12 approach, and
land on runway 12. He said the weather in Kingston was rain at the field, and rain was also on
the radar, so the runway 12 tailwind was a consideration.
The flight crew also discussed the runway conditions based upon the reported weather, the
ceiling and wind. They decided to use Flap 30 and autobrakes 2 (the crew changed this to
autobrakes 3 on final approach). The first officer said he had done tailwind landings before in
weather.
Regarding the use of autobrake, the first officer said the flight crew had some discretion in its use
and, for example, on a long dry runway, they did not need to use it. As conditions
deteriorate, he said, one would want to give oneself the additional benefit of autobrake. He said if
autobrake did not work you could switch to manual braking but you would generally start with
autobrake. He said according to landing distance charts, manual braking provided the maximum
braking and shortest stopping distance; however it was not their standard operating procedure to
use manual braking.
The first officer said he thought the briefing was particularly thorough because of the conditions.
The briefing included the rain, visibility, landing considerations, runway options, circling options,
flaps, brakes, missed approach considerations, time and fuel to alternate, and the terrain and
minimum sector altitudes.
The briefing was not captured on the Cockpit Voice Recorder (CVR), as it took place more than
30 minutes before the CVR lost power at impact. The CVR only recorded the most recent 30
minutes.
There were no contradictions between the statements of the captain and the first officer.
The captain also stated that he called the Kingston Approach controller and requested the latest
conditions at Kingston, and reports from aircraft ahead of them. He said the controller told
him that the visibility at Kingston was good, and that other aircraft were landing there and
“no one was reporting anything out of the ordinary”.
AA331 first made contact with the Kingston Enroute controller at 21:47 EST, while level at FL
370 when the flight was eighty miles north of TOTON, a reporting point along the UA301
airway at the Kingston FIR northern boundary. At that time AA331 was cleared to maintain FL
370, proceed to TOTON, and thence direct to KEYNO, the intermediate fix for the ILS approach
runway 12.
The AA dispatcher sent an ACARS message to AA331 about 21:48 EST informing them
that the latest METAR for Kingston included thunderstorms and moderate rain shower activity,
that is, “SPECI MKJP 230228Z 31009KT 5000 TSRA BKN014 FEW016CB SCT030 BKN100
22/19 Q1013”.
De-coded, this is, “MKJP special weather observation at 21:28 EST (02:28 UTC 23 Dec), wind
310 degrees at 9 knots, visibility 5000 metres (m) (approximately 3 statute miles) in thunderstorms
and moderate rain, ceiling broken at 1,400 feet, few clouds at 1,600 feet in cumulonimbus clouds,
scattered clouds at 3,000 feet, broken at 10,000 feet, temperature 22° C, dew point 19° C, altimeter
setting 1013 millibars (mb)”.
The ACARS sent to AA331 at 21:28 EST (02:28 UTC) also reminded the captain that the
runway at Montego Bay would be closed from 03:00 UTC.
The CVR did not capture the pre-approach briefing, and did not record any Automatic Terminal
Information Service (ATIS) message. Also, the AA331 crew did not state whether or not they
received the ATIS in their post-accident interviews. The ATC transcripts showed that AA331
contacted the Approach controller before descent and requested the Kingston weather, and did
not state to the Approach controller that the flight had received the ATIS, nor did the Approach
controller request confirmation of its receipt.
The weather given to AA331 by the Enroute controller at 21:48 EST was, “wind at the station is
tree ( that is, three) one zero degrees at … seven and a half knots, the visibility is approximately
five miles … present weather there’s a moderate shower at the station … temperature is two one,
dew point two zero, the QNH is one zero one four … understand that there’s a broken clouds at
one thousand feet …” No runway condition report was included. This information was not the
same as that in the official weather reports (see above and 1.7.6).
1.1.3 Descent and Approach
The captain made a public address announcement for everyone to be seated, and for the cabin
crew to prepare the cabin early for landing due to the expected turbulence on descent. The
forward cabin crew members responsible for operation of cabin door exits L1 and R12 and who
were seated in the cabin crew seats at L1 during the landing, were able to conduct their pre
landing checks. The two aft cabin crew members seated at positions L2 and R2 (left and right
aft), who were responsible for cabin doors L2 and R2, were unable to complete their safety
checks due to the turbulence experienced in the aft of the aircraft.
At 21:51 EST the Enroute controller cleared AA331 to descend to 15,000 feet, at the pilot's
discretion, and again gave the altimeter setting of 1014 Mb. At 21:58 EST, before being officially
transferred from Enroute to Approach control, AA331 called ahead and asked the Approach
controller if any aircraft had reported turbulence on the approach and was advised that none had
been reported. At 21:58 EST AA331 then asked the Approach controller if any aircraft had
landed in the last hour and was told that one had arrived, but not from the north, and “didn’t have
any problems coming in.”3
When AA331 first contacted the Approach controller at 21:58 EST there was no mention by either
party of AA331 having received the current ATIS.
The flight crew described monitoring the fuel quantity and using the speed brakes and going to
lower altitudes to burn off sufficient fuel to ensure that the aircraft would be down to its maximum
landing weight at Kingston.
2
Pg. 42, Figure 1
3
Section 1.1.5

At 22:03:40 EST the Enroute controller transferred control of AA331 to the Approach controller
and AA331 reported on contact that the aircraft was descending out of FL 190 for one five
thousand feet, and at 22:04:15 EST reported through the transition level of FL 180 for one five
thousand feet. The Approach controller responded advising AA331 that they should expect an
ILS runway 12 approach, gave the altimeter setting as 1014 mb, and cleared the aircraft to
descend to and maintain one five thousand feet. The clearance was read back correctly by
AA331.
At 22:04:36 EST the Approach controller told AA331, “be advised the information given to you
by the Enroute controller is still the same, visibility five miles and there is moderate rain at the
station … wind three two zero at one zero knots …”. The ATC and CVR transcripts did not
contain any record indicating that a runway condition report and/or a braking action report were
requested by AA331, nor that this information was provided by the Approach controller, as was
required by ATS MANOPS.
The Approach controller then advised AA331 that they may have to circle from the ILS runway
12 approach and land on runway 30 as the wind at Kingston was now 320 degrees at one zero
knots.
AA331 acknowledged this and advised that they would go ahead and take a straight in on
runway 12 with the tailwind. At 22:14 EST the Approach controller cleared AA331 to maintain
four thousand feet, and on reaching KEYNO cleared for a straight in ILS approach runway 12, and
advised that the wind was now 320 degrees at 15 knots. AA331 acknowledged the clearance and
the Approach controller asked AA331 if they had understood that the wind was 320 degrees at
one four knots, asking them if they were still able to make a straight in approach for landing on
runway 12. AA331 responded that they had received the wind and could make the straight in
approach to runway 12.
The first officer said the aircraft had both a predictive and reactive wind shear system, and it was a
pictorial indication on the navigation display with an approximate location of the wind shear
along with one of several aural warnings. No wind shear warning appeared. He said they had
weather selected on the both radar displays.
At 22:17 EST the Approach controller handed control of AA331 to the Manley Tower and
AA331 contacted the Tower controller, reporting inbound on the ILS for runway 12, level at
2,800 feet. The Tower controller responded advising them of the wind as 320 at 12 knots, and
asked if they were still requesting runway 12. AA331 responded affirmatively and asked for the
wind again. The Tower controller responded with the wind at 320 at 14 knots, and AA331 stated
“ … that's affirmative … ” and repeated the intention to land on runway 12. The Tower
controller then cleared AA331 to land on runway 12, and stated “ … be advised runway wet”.
The CVR and ATC transcripts did not include any braking action report, or any other surface
condition description.

AA331 thanked t h e Tower controller. At 22:20 EST, AA331 advised the Tower c ont ro l l er
that they were three miles on final and the Tower controller responded that landing clearance
was still valid for runway 12. The CVR contained no discussion between the two flight crews
about the increased tailwind, the reported rain shower activity, the runway conditions or
calculation of landing distance.
The following information regarding the approach was provided by the flight crew during post-
accident interviews:
 They encountered heavy rain during approach and noted that the approach was noisy
due to the rain.
 During the descent, ATC provided them updates on weather, but none required them to
recalculate their landing numbers, and at no time was there a tailwind that exceeded their
limitations. AA company limitation for tailwind landing was 15 knots.
 As the aircraft approached Kingston there was light to moderate rain on the weather
radar, there were no significant storm cells to fly around, and the radar was indicating a
broad area of moderate rain.
 There were some terrain considerations when doing a missed approach. The
minimums were high for a circling approach, compared to the straight in ILS runway 12
approach minimums.
 They experienced light to moderate turbulence in the descent. There was terrain they
had to cross over on the descent into Kingston, and there was a published missed
approach and a single engine departure procedure to the south over the water, and most of
the weather was between them and the airfield. They were cleared direct to KEYNO,
which was normal, and nothing was rushed, high or fast, and the descent was normal.
 The magnetic track of the localizer, 117 degrees, was offset three degrees from that of
runway 12, which was 120 degrees magnetic4 .
 The first officer could not recall what the weather was on the ATIS but they were
receiving the latest weather from the Tower. He said there were no additional messages
from dispatch regarding weather changes.
 The first officer did not recall anyone immediately preceding them for landing at
Kingston, but the captain asked, and the Tower said, they had an aircraft land 30 minutes
prior. He did not specify what type of aircraft, and they did not report anything to the
4
Appendix (4)

Tower5. The captain queried the Tower controller several times to get the most accurate
information about the field.
 When they intercepted the ILS runway 12, the weather radar painted light to moderate
rain. The weather and rides were smooth, and they did not feel any shifting winds, and not
much turbulence on final. The captain had it on speed and on glide path early
 The flight crew started getting ground contact intermittently below 2,000 feet, and
consistently below 1,500 feet. The aircraft broke out from the clouds at about
1,000 feet and had the runway in sight, which was well above minimums for that
approach, and maintained visual contact with the runway throughout the
approach. They were still receiving rain, and t h e wipers were turned on. There were
no lower clouds restricting visibility.
 The aircraft was configured early at 1,000 feet, and was on the ILS glide slope. The first
officer was monitoring the tailwind on the navigation display and the tailwind component
was dropping as they descended to the final approach fix, where it was about an 8-9 knot
tailwind. The first officer said he did not look at it during the flare but he recalled that the
last time he saw it, it was showing 8-9 knots tailwind. He saw the visual approach slope
indicator (VASI) on final approach and they were “right on it”.
 Because of the weather, he said he gave a few additional call outs, and at 500 feet he
called, “On speed, sink 800”.
 The first officer stated that at no point was the aircraft high or above speed, and the
captain flew a nice approach.
 At about 550 feet, the captain disconnected the autopilot while leaving the auto
throttle engaged, and aligned the aircraft to the runway and at about 50 feet the aircraft
was on glide path and on speed. He said the approach was noisy, because of the heavy
rain. The windshield wipers were on.
 The aircraft was on speed, on the glide path and localizer and, after transitioning from
instrument cues to visual flight cues about three miles back, when the aircraft was
turned to line up with the runway, they both said the aircraft had crossed the threshold
"right in the slot".
The information from the captain agreed with that of the first officer, as stated above, and also
that of the CVR.
5
1.10.4.2

The Flight Data Recorder (FDR) data confirmed that the aircraft crossed the runway threshold at
70 feet radio altitude above ground (RA), that is, main gear height. This would have placed
the pilot's eye height at about 85 feet RA, about 14 feet above the PAPI slope, and the aircraft's
ILS antenna about 37 feet above the ILS glide slope. Thus, both the visual (PAPI) and ILS
indications to the flight crew over the threshold were that the aircraft was high on the approach.
The PAPI is a visual aid to the pilot for vertical guidance to the runway.
The localizer of the ILS serving runway 12 was offset three degrees north of the runway track,
and hence did not bring landing traffic to the touchdown point6 , so at Decision Altitude/Height
(DA/H) pilots using the ILS had to transition to approach and runway lighting cues and make a
very slight left turn to line up with the runway centerline7. The aircraft was equipped with a
Heads Up Display (HUD) above the left (that is, the captain’s) seat.
The captain, who was pilot flying and who was seated in the left seat, stated that this approach
was an offset localizer approach and in the HUD it actually gave the pilot too much data. The
captain stated in his post-accident interview t h a t t he HUD did not have a de-clutter mode.
The B737 HUD8 does have a de-clutter mode switch located on the side of the combiner.
The captain stated that, as the ILS was offset, the HUD indications did not match up with the
runway and when the aircraft broke out of the clouds he had the real runway, the Flight Path
Vector (FPV) and the aircraft, to get lined up. In this approach, he did not go to visual, but he
stayed on the localizer all the way down, so he said he eliminated a few of those issues. He
said with the HUD on A-2 mode, having the real runway, fake runway, and FPV was too
much information so he stayed on the localizer in A-1 mode until the last point where he
thought he could make a smooth turn to the runway and touchdown. The following information
from the HUD user guide describes the HUD Approach Modes:
AI Approach Mode
The AI mode is intended for CAT I ILS approaches. CDS flight director guidance
is used to drive the HUD flight director. The HSI symbol is displayed until the
localizer course is captured and the localizer deviation is less than 0.25 degrees for
seven seconds. At that time there is an automatic switch to the localizer deviation
scale. AI mode can be selected at any time, except in TO/GA or on the ground.
If AI mode is selected prior to capturing the glideslope and localizer, the HUD
remains in NP mode with AI mode armed.
6
Appendix 4 (4)
7
Appendix 4
8
AA B737 HUD Briefing Guide, Page 3

AII Approach Mode

The AII mode is intended for CAT II ILS approaches. It is similar to AI mode
except that CDS flight director guidance is used to drive the HUD flight director
until the glideslope and localizer are captured. At this time, the HUD switches to
HUD computer guidance.
The first officer indicated that, while visual, at about 500 feet, the captain aligned the aircraft with the
runway centerline. The FDR indicated that the aircraft made a shallow right turn from the ILS
localizer then a left turn to the inbound track to the runway, and continued the approach, crossing the
runway threshold at about 70 feet RA.
The FDR indicated that the captain made nose-up control inputs at approximately 70 feet RA and
manually disconnected the auto throttle at about 35 feet above the runway and closed the throttles to
idle. The auto throttle Flare Mode is designed to retard the power automatically when the aircraft
reaches 27 feet RA. Up to the time the auto throttle was manually disconnected by the captain, the
auto throttle was in speed mode and was maintaining the aircraft’s airspeed at VREF30 (reference
landing speed with flaps 30) + 5 knots, i.e., at 148 knots, which was a ground speed of 162 knots.
1.1.4 Landing
The FDR data showed the aircraft pitching slightly nose up as it passed the threshold, which
reduced the descent rate. The captain turned off the autothrottle manually as the nose pitched up,
about 4 seconds after threshold crossing, and the throttle levers were reduced to the flight idle stop
about 14 seconds after threshold crossing, which was 3,800 feet down the runway.
The float continued as the aircraft passed the PAPI lights at about 38 feet RA (where it should have
been on the ground in a normal landing). The shallow rate of descent was maintained for about ten
seconds until touchdown, which occurred at 4,100 feet down the runway, or 1,130 feet beyond the
touchdown zone, as defined by AA Flight Manual, Part I.
Both flight crews reported being “on speed and in the slot” as the aircraft crossed the threshold of
runway 12. The captain stated it was raining and the windshield wipers were turned on.
There was not much peripheral lighting around the runway because of its location across a peninsula
with sea at both ends, and the absence of nearby settlement. A commercial power outage had caused
the airport to be operating on its standby power generator. The outage resulted in even less than
normal peripheral lighting around the airport.
The aircraft was landed with flaps at 30 degrees.
In his interview, the captain stated that he was not sure what the speed difference was between flaps 30
and flaps 40 for landing. He said there was a different amount of thrust required and that flaps 40 was
good for a short or contaminated runway. He thought in this case, he was better off with flaps 30.
He said flaps 40 can cause you to float sometimes.

The first officer stated they sometimes landed at flaps 40 but the captain had briefed flaps 30, which
was normal for this situation. He said they had discussed that for these conditions and for the go
around, flaps 30 was the better choice. He said that flaps 40 would generally be used for really short
runways.
At least two pitch-up control wheel inputs were recorded by the FDR during the prolonged flare. The
aircraft touched down at 148 knots indicated airspeed (KIAS) and 162 knots ground speed, with a
14 knot tailwind component and a 7 knot crosswind component from the left.
The spoilers deployed at wheel spin up on the first touchdown, the aircraft bounced once and
then landed again 200 feet down the runway, then autobrake 3 activated at about 4,600 feet from
the threshold. The first officer called the ground spoiler, deployment and reverse thrust
engagement, and saw the green lights indicating that reversers were deployed.
The captain stated that the aircraft was not decelerating as expected using autobrake 3 and he
overrode the autobrake system, applying maximum manual braking on the brake pedals, and
selecting maximum reverse thrust with the thrust levers. He was joined by the first officer
simultaneously applying maximum manual braking on the brake pedals.
This commanded full braking pressure of 3,000 pounds per square inch (psi), as indicated by the
FDR, while the captain maintained directional control along the centerline. The wheel marks on
the runway indicated that the aircraft nose wheel drifted about twenty feet (six metres) left of
centerline as the aircraft approached the end of the runway. Both of the flight crew reported an
abnormal lack of deceleration during auto-braking and maximum manual braking, and both stated
that they soon realized the aircraft would leave the runway. The final radio transmission from
AA331 was at 22:22:16 EST.
The following information is from the statements of the flight crew:

 The first officer stated that the visibility at touchdown was adequate in
the rain.
 The first officer stated that everything was normal to touchdown, but with a little
float on the flare, and a smooth transition to landing and touchdown.
 The captain said he thought they touched down at about 1,500 feet.
 The first officer stated that he felt no yawing on landing, and there were no
controllability or directional control problems. He said he pressed the pedals
down for braking, and didn’t notice any differential or directional issues with
the pedals. He said he did not sense any antiskid cycling.
 The first officer stated that he did not recall any abnormal alerts or warnings on
landing. He thought the reference speed was about 148 knots and they were close
to max gross weight for landing.

 The captain said he brought the reversers back as much as he could, but he felt
like he was on ice as there was no deceleration, so he applied full manual braking.
 The first officer said he had landed in heavier rain before with less visibility and
had not had any problem with braking. He had never experienced a lack of braking
as they experienced in this event.
 The captain said that at touchdown he thought the runway was under water.
When he realized the aircraft was not decelerating, he thought about doing a go-
around, but did not, as he expected the aircraft would eventually slow down.
 The first officer said he did not recall seeing the runway end lights, and knew there
was a road beyond and below the runway. He felt that they crossed the road, and
that they hit and the landing gear failed, and that there was a secondary impact.
The flight crew was unable to stop the aircraft on the remaining runway and, according to the
FDR data, the aircraft exited the end of the runway at 62 knots ground speed. The aircraft then
passed through the chain link airport perimeter fence, crossed above the road located about 12
feet below the embankment at the end of the runway and came to rest on the sandy and rocky
shoreline area located there, east of the airport boundary. The FDR showed that forward motion
ceased and power was lost at 22:22:21 EST.
There was no warning of the impending impact, and no brace command was given. The flight
crew reported that after the aircraft came to a stop the cockpit was dark, and they completed the
emergency evacuation checklist.
1.1.5 Information given to AA331 flight crew during approach
The captain’s interview summary states: “He called approach control told them who he was, where
he was, and asked for what the airport conditions were and any reports from previous airplanes. He
said the controller told him the visibility was good and he asked if anyone had given pilot reports
for turbulence on the approach, to which the controller replied no one had reported anything out of
the ordinary.” This call was made at 21:58:47 EST, according to the ATC Transcript, while
AA331 was still under Enroute control, before transfer to Approach control.
The available evidence from ATC transcripts and controller interviews indicates the following
sequence of events:
1. At 21:47:17 EST, while the aircraft was under Enroute control, AA331 requested
Norman Manley International Airport (NMIA) weather from the Enroute controller.
2. The Enroute controller, who was seated nearby and within earshot of the Approach
controller, then asked the Approach controller for the NMIA weather.
3. The Approach controller reported that he told the Enroute controller that he had a recent
PIREP of ceiling 800 feet.

4. The Enroute controller reported that he gave AA331 the current field condition based on
the report on the AWOS as well as a cloud ceiling report from a departing aircraft and
tower observation.
5. At 21:48:33 EST the Enroute controller told AA331 “Wind at the station is tree (that is,
three) one zero degrees at … seven and a half knots, the visibility is approximately five
miles ah … there’s a moderate shower at station … the temperature is two one, dew point
two zero, the QNH is one zero one four … understand that there’s also a broken clouds at
one thousand feet … ”
6. At 21:51:56 EST the Enroute controller cleared AA331 for the approach, but did not
transfer AA331 to Approach control.
7. At 21:55:25 EST AA331 contacted Approach.
8. At 21:58:47 EST AA331 asked Approach if there are any turbulence reports, to which
Approach replied, “None reported”. AA331 then asked if there was any traffic landing in
the last hour.
9. At 21:59:15 EST Approach told AA331, “Affirm, we had one aircraft that landed in the
(last hour) … at zero two zero three and he came in from Montego Bay, came in from the
northwest and ah ... he didn’t have any problems coming in … didn’t have any body from
the north….”
10. At 21:59:37 EST AA331 contacted Enroute again.
11. At 22:03:43 EST Enroute transferred AA331 to Approach.
12. At 22:03:50 EST, in conversation by landline, Approach controller asked Tower controller
“ … still moderate rain at the station?”
13. At 22:03:55 EST Tower controller replied “Affirm … ”
14. At 22:03:56 EST Approach controller asked “And what’s the visibility on final … you still
can see Portmore?”
15. At 22:03:59 EST Tower controller replied “Ah … hmmm … well the light dem look
dimmer … so I’m … I’m I’m … gonna say less than ten …”
16. At 22:04:06 EST Approach controller said “Alright then … I’ll tell him five …”
17. At 22:04:10 EST Tower controller said “Yes, five is good …”
18. At 22:04:36 EST Approach told AA331 “ … be advised the information given to you by
the Enroute controller is still the same, visibility five miles and there is moderate rain at
the station.”
19. At 22:04:51 EST Approach reported wind 320 degrees at 10 knots.
22. At 22:17:28 EST Approach transferred AA331 to Tower frequency.
23. At 22:17:42 EST Tower reported wind 320 degrees at 12 knots.
24. At 22:17:52 EST Tower reported wind 320 degrees at 14 knots.
25. At 22:17:57 EST Tower said to AA331, “American three, three one cleared to
land runway one two … be advised runway wet”.
Until the report of “runway wet” by the Tower controller at 22:17:57 EST, less than five minutes
before landing, there was no mention of any runway condition report. There was no braking
action report given, and there was no mention of the ATIS by either AA331 crew or the Tower
controller.

The official METAR at the time of the final approach and landing of AA331, issued at 03:00 UTC
(22:00 EST), 22 minutes before the accident, was ceiling 1,400 feet broken, visibility 3,000 meters
(about 2 miles)9. Visual Flight Rules minima in controlled airspace are ceiling 1,000 feet, visibility
3 statute miles, if reported.10
1.2 Injuries To Persons
Crew Passengers Others Total
Fatal - - - -
Serious - 14 - 14
Minor/None 6 134 - 140
Total 6 148 - 154
Table 1: Injuries to Persons
1.3 Damage To Aircraft
The aircraft was substantially damaged during the overrun with the right main gear and right
engine torn off, the left main gear and the nose gear collapsed, the main fuselage split into three
sections and ruptured and deformed underneath, the wings and flaps damaged, and the fuel tanks
in the right wing ruptured.11
1.4 Other Damage
The airport boundary fence was damaged when the aircraft went through it. The jet fuel leaked
from the right wing tanks into the sandy and rocky area where the aircraft came to rest.
9
See 1.7.6
10
JCARs (2004), Tenth Schedule, sub-paragraph 10.665(a)
Section 1.12, ‘Wreckage and Impact Information’, details accounts of exterior and interior damage to aircraft
11

1.5 Personnel Information
1.5.1 ATC Operational Personnel
The records of the Enroute, Approach and Tower controllers involved were examined and it was
determined that, at the time of the accident, they all held valid ATC licenses and medical
certificates.
1.5.2 Flight Crew – General
Unless otherwise stated, AA personnel information was obtained through the Accredited
Representative, the NTSB.
Captain First Officer

Pilot License Airline Transport Airline Transport
Medical Expiry Date 01 January 2010 01 May 2010
Total Flying Hours 11,147 6,120
Hours on Type 2,727 5,027
Hours Last 24 hours 7 7
Hours Last 7 days 20 26
Hours on Type last 90 Days 177 163
Hours on Duty Prior to Landing 11:22 11:22
Hours off Duty Prior to Work Period 72+ 72+
Proficiency Check April 2009 April 2009
Last Route Check 14/04/2009 Not required
GPS Approach qualified Yes Yes
Table 2: AA Personnel Information
1.5.3 Flight Crew and Cabin Crew – Flight/Duty/Rest Times
The records indicated that the flight/duty/rest history of the crew was within the required limits.
At the time of the accident the pilots had been on continuous duty for the eleven and one half
hours immediately preceding the accident, and had been awake for more than 14 hours. In their
post-accident interviews, both pilots reported having had adequate rest during the three days
prior to the day of the accident and neither reported feeling fatigued.

1.5.4 Captain
The captain’s Federal Aviation Administration (FAA) medical records contained the following
information:
 Last medical: 6/16/09
 First Class
 Weight: 200 lbs.; Height: 74 inches.
 No limitations
 No meds (sic)
 Near and intermediate vision: 20/40; distant vision: 20/20.
 Color vision and hearing tests: passed.
The captain stated that he was 49 years old and employed by AA as a Boeing 737-800 captain. He
said he had about 13,000 hours total time and did not recall how much of it was as pilot in
command (PIC).
The captain said he graduated in 1982 with a degree in Aeronautics. He then worked for a flight
school and Part 91 charter company which had twin engine airplanes and six Learjets and ran
flights throughout the Bahamas for four years. He went from bei ng a flight instructor to a
Learjet captain in a couple of years. He said he became the director of operations and flight
training officer for this charter company, and he flew the Lear 25 and 35.
He then joined AA in 1986 as a Boeing 727 flight engineer and followed a typical airline profile
of progression to first officer on the Boeing 727, first officer on MD-11, then captain on the
Boeing 727. He progressed to check airman on the Boeing 727 where he conducted initial
operating experience, line checks and special qualifications training on other flight crew
operating into Central and South American cities.
When the Boeing 727 fleet was phased out he was assigned to the Boeing 737 as captain and
had flown on the Boeing 737 for several years. He was familiar with Caribbean routes, and had
landed at Kingston on runway 12 before during rainy weather.
He said he occasionally drank beer and thought the last time he had one may have been over the
last weekend. He was not a heavy drinker and did not use tobacco.
He stated that he had no significant changes to his financial status in the last twelve (12) months
and his personal life was great. He normally went to bed around 23:00 EST, and woke up
around 08:00 EST. He tried to get one hour of exercise each day. He said he did not have any
sleep disorders and liked to take a short 20 minute nap sometimes.
He had gone to bed the normal time on Monday night, about 23:00 EST. On Tuesday, the day of
the trip, he said he woke up around 08:00 EST. His first flight was scheduled to leave at about
12:00 EST, so he had some time to walk around the lake at his house to get some exercise.
He left his house about 09:30 EST to get to the Miami airport for the 11:00 EST show. He said he
was pilot flying on the first leg of the trip to Baltimore/Washington International Airport. He

knew that it had snowed there a lot lately so he was anxious to take a look. He said it was an
uneventful flight to Baltimore. They were a little late leaving Miami International Airport due to a
deferred auxiliary power unit (APU) problem.
The flight back to Miami International Airport with the first officer as pilot flying was uneventful.
While there they swapped airplanes and he said the only MEL item on the airplane was the right
air conditioning pack, which they placarded before leaving Miami International Airport.
He said that on any trip the most stressful part seemed to be just trying to get off the gate. He
said this particular flight was a little stressful having so many passengers on the flight and it
being so rough flying over Cuba – this was not a normal night coming across Cuba. He said they
did their best to ask for reports from other flight crews and from the controllers. He said once they
got to the Kingston area it was just rain, no turbulence, just heavy rain. He said that heavy rain
is noisy and when it is that noisy, it raises your stress level ensuring that you “dot the ‘i’s and
cross the ‘t’s.”
He said everything was going the way it should have been and that it was “standard operating
procedure”.
He said that he had flown with the first officer previously, also on the Boeing 727. He said
that the first officer was an excellent pilot and that, based on his experience as a check airman,
the first officer went above and beyond the call than most first officers. He said the first officer
was very personable and professional, and that they shared mutual interests, so he knew that
flying with him would mean that they would have interesting conversations. He said the
first officer’s greatest strength might be that he was a good communicator, but he did it in a way
to not overshadow the captain.
He said the AA Flight Manual, Page 10-7X, included a note about standing water at Kingston,
but did not state whether or not he was aware of this before the accident.
1.5.5 First Officer
The first officer’s FAA medical record had the following information:
 Last medical: 5/4/09

 First Class
 Weight: 203; Height; 72 inches
 No Limitations
 No meds (sic)
 Distant, near and intermediate vision: 20/20
 Color vision and hearing tests: passed
The first officer learned to fly at an aviation college in the United States of America, graduating in
1987 and worked as a flight instructor immediately following graduation. He then flew Part 135
scheduled operations in the Caribbean and Miami on smaller aircrafts, and joined American
Eagle in 1994 where he flew ATR aircraft based in Puerto Rico and Miami. He joined AA in
1998 and followed a typical career progression as flight engineer on the Boeing 727 for a year,
then first officer on the Boeing 727 for three years, transitioning to international flying on the
Boeing 737 as first officer from 2002, based in Miami. He had not flown as a captain and had
been a first officer at his previous company.
He said he normally went to bed around 23:00 EST, and, depending on his work schedule, would
wake up between 07:00 to 08:30 EST. He said he usually got six to eight hours of sleep to feel
rested. He stated he was off on Saturday, Sunday and Monday, and stayed at home within his
normal sleep-cycle.
He had gone to bed Monday evening about 23:00 EST, and he woke up on Tuesday around
07:00 EST. Usually it took 20 minutes to get to the Miami International Airport, but on that day it
took about 30 minutes and he arrived there at 11:05 EST.
He said his scheduling and workload on the day of the accident was typical and although it was
a long day, (this was the longest day of the three day trip) it was not atypical for his schedule.
He had breakfast at home, a meal on the flight to Baltimore and a snack on the way back to
Miami. The aircraft change in Miami took about one hour, and the first officer stated that he "felt
fine" after having flown the Baltimore to Miami leg. He said the weather did not affect his
workload significantly on the accident flight, except a few more minutes of talking about the
weather over Cuba.
He generally considered his health as excellent, and had no significant changes in his health or
financial or personal life in the last twelve months.
He had flown with the captain the first time on the Boeing 727, and once or twice over the last
year, but not immediately preceding 22 December, 2009. He considered him to be a very good
captain, and was very comfortable flying with him. He was open to input, with excellent flying
skills, and his greatest strengths were that he was very professional, had good management skills
handling others, was easy going and non-confrontational. He liked working for AA a lot and had
no pressures from outside influences to complete flights.
He said he would like to see something in the AA Flight Manual, Page 10-7X regarding
standing water at Kingston, to draw more attention to it. He said that in the conditions, they were
expecting to land on a wet runway, they had done so many times, but he said they had much less
braking than that. He did not recall anyone ever discussing among the pilot group any issues with
standing water at this airport.
Regarding the change from ten knots to fifteen knots allowable tailwind he said it was not a
runway limit but was a company limit and a 12,000 foot runway would not have made a
difference on that limitation. He said generally they did not want to land close to the ten knot
limit, even on a dry runway. He said if it was a runway limit it may have made them think of
doing something different.

1.5.6 Cabin Crew
The FAA minimum cabin crew for this Boeing 737-800 aircraft, with a configuration of 16 first
class seats and 132 economy class seats, was three, but four were assigned for this flight in
accordance with AA procedures for international B o e i n g 737 flights. All of the cabin crew
members were trained, certified and qualified for their assigned duties on the aircraft, in
accordance with the approved AA training program and FAA regulations.
1.5.7 Crew Pairing
For the flight crew this accident occurred on the third and final leg on day one of a three-day
pairing. The four cabin crew members, based in Miami, had been on reserve and were given this
flight assignment from Miami to Kingston as their first leg.
1.6 Aircraft Information
1.6.1 General
Manufacturer The Boeing Company
Type and Model B737-823
Year of Manufacture 2001
Serial Number 29550
Certificate of Airworthiness Issued 19 December 2001
Total Airframe Time and Cycles 24,610 hours and 10,402 cycles
Engine Type (Number Of) CFM 56-7B27 (2)
Maximum Allowable Take Off Weight 174,200 lbs.
Maximum Allowable Landing Weight 144,000 lbs.
Recommended Fuel Type JET A1
Fuel Type Used JET A1
Table 3: Aircraft Information

1.6.2 Maintenance
The maintenance records of the aircraft were examined, and it was determined that all required
inspections and scheduled maintenance had been performed; that all airworthiness directives
had been complied with and that the aircraft major repairs and alterations were satisfactorily
documented. The MEL had no open items except for the air-conditioning pack temperature
controller fault warning, which was deferred in accordance with the MEL in Miami before
departure. On November 17, 2009, the HUD system was placed on MEL due to a “System
Inop” display on the HUD. The system was returned to service on November 20, 2009, with no
later discrepancies identified. Otherwise, there were no deferred defects.
1.6.3 Weight and Balance
The zero fuel weight of the aircraft was 133,654 lbs. Based on the fuel on-board,
approximately 10,000 pounds, the actual landing weight of the aircraft was a b o u t 143,654
pounds, w h i c h w a s slightly below the maximum allowable landing weight of 144,000
pounds. The AA dispatch document for Flight AA331 indicated that the aircraft’s centre of
gravity was within limits for the duration of the flight.
1.6.4 Landing Speeds
For a landing weight of 144,000 lbs. and Flaps 30 the certified reference landing speed (VRef)
was 143 KIAS and the approach speed (VApp or target speed) was 148 KIAS.
1.6.5 Automatic Flight System and HUD
The aircraft was fitted with two separate autopilot systems either of which could be selected to
operate individually, one at a time to fly the aircraft, or which could be selected to operate in
tandem. It had two independent autothrottle systems to control engine power either of which
would be operated individually on its own to control one or both engines. Its navigation systems
consisted of VOR DME, Inertial Reference, and air data units which all fed a flight management
system computer that processed the data and fed it to the electronic flight instrument
displays on the instrument panel. The flight management system could be programmed by
the flight crew for flights by entering all required elements of a flight planned route via an input
device.
The aircraft had a tactical control panel above the main instrument panel on which manually
selected inputs to the FMS and autopilot and auto throttle systems could be made by the pilots,
that is, selection of heading or track, speed and altitude and of vertical speed. The electronic flight
instruments consisted of displays showing navigation and performance information to the flight
crew. Most information displayed on the Electronic Flight Information System (EFIS) displays
could also be displayed in the heads up display (HUD). Detailed information on all of this is found
in the AA FCOMs and is not duplicated in this report. The HUD system installed on this aircraft
was a BAE Systems HUD2020.

1.6.6 Weather Radar
The aircraft was fitted with a Rockwell Collins color digital radar that displayed radar returns
on the No.1 and/or No.2 Navigation Display on the left and right forward instrument panels,
when so selected by either flight crew member.
AA standard flight operating procedure for the operation of the color digital weather radar fitted
in the Boeing 737 aircraft required that flight crew operate it when there was a risk of severe
weather or reported thunderstorms in the route of flight. The weather radar was functioning
properly during the flight to Kingston, and, according to both flight crew members, the weather
radar was used during the flight.
1.6.7 Wind Shear Detection and Prediction
The aircraft was fitted with predictive and reactive wind shear warning devices and there was no
evidence from the crew interviews or MEL records that they were not functioning at the time of
the accident. The wind shear information was displayed pictorially on either of the navigation
displays in front of each flight crew member. Flight crew interviews and the CVR recording
indicated that no wind shear alert was activated in the cockpit during the flight or during the
approach and landing.
1.6.8 Ground Spoilers
The aircraft was fitted with electro-hydraulically operated ground spoilers, which were armed by
the first officer when the landing gear was lowered. The FDR did not record Spoiler Deployment
or Wheel Speed. The FDR plot of the Speedbrake Handle and the Air/Ground Switch, and the
first officer’s comment “Deployed” on the CVR at touchdown, indicated that the spoilers deployed
on the first touchdown, as the wheels spun up. The first officer also reported that he manually
checked the activation lever position during the landing roll. The evidence indicates that the
ground spoiler system was serviceable and operated normally during the landing.
1.6.9 Brake System
The braking system was comprised of four subsystems. The skid control compares the calculated
wheel speed with a velocity model to control wheel deceleration. If a wheel slows down too
quickly, the skid control releases brake pressure until the wheel speed increases. Skid control does
not operate at less than a speed of eight knots. During the normal antiskid operation, the skid
control operates for each wheel. During alternate antiskid operation, the skid control operates for
both wheels on one main landing gear.
Locked wheel protection compares the wheel speeds of the two outboard wheels to that of the
two inboard wheels. If the speed of the slower wheel decreases to less than 30% of the speed of
the faster wheel, the locked wheel protection releases brake pressure from the slower wheel.
Locked wheel protection does not operate at a speed less than 25 knots.

The touchdown protection releases brake pressure from wheels two and four while the airplane is
in the air and remains active until 0.7 seconds after the corresponding wheel spins up to 70 knots,
or when the ground mode has been sensed continuously for three seconds.
The number 1 and 3 wheels have similar protection, but it is accomplished via the hydroplane
protection logic/algorithm in the Antiskid-Autobrake Control Unit, since the difference between
the wheel speeds and the Air Data Inertial Reference Unit (ADIRU) measures w h e n airplane
ground speed is greater than 50 knots until wheel spin-up occurs at touchdown.
The touchdown/hydroplane protection compares wheel speed data to ADIRU ground speed data.
When the wheel speed decreases to 50 knots less than ground speed, the touchdown/hydroplane
protection releases pressure to the brake. The hydroplane function supplies protection to wheels
1 and 3 only.
This is a general description of how the Boeing 737-800 brake system is designed to work, not a
description of how the accident aircraft’s system worked during the accident sequence.12
1.6.10 Engine Controls
The two CFM 56-7B27 engines were controlled by an Electronic Engine Control (EEC) that was
inter-connected to the Flight Management System (FMS). The fuel pump controls, engine start
controls, and the engine anti-ice controls were located on the forward overhead panel. The auto
throttle controls were located on the glare shield panel and on the throttle levers. The auto
throttle control could be disengaged by pressing a button on the side of the throttle levers.
1.6.11 Rain Removal System
The aircraft was fitted with variable speed windshield wipers on the captain and first officer
windshields. The aircraft was also fitted with a rain repellent system, but this system was not
used in the USA due to environmental protection concerns about the toxicity of the repellent
fluid.
1.6.12 GPS Navigation Equipment
The aircraft was equipped with GPS navigation equipment Smiths Industries Multi-Sensor FMC,
Model No. 171596, Software 549849-006 and 549849-009, and was authorized for the conduct
of required navigation performance (RNP) instrument approach procedures (IAP). The Kingston
RNAV (GPS) Rwy 30 approach was confirmed to have been in the database of the aircraft’s
navigational equipment at the time of the accident, and was current.13
12
Sections 1.12.4 for information about the accident aircraft’s brake system.
13
Section 1.18.9

1.6.13 Aircraft description – Cabin, Seats and Exits
The Boeing 737-823 aircraft, registration N977AN, was a single aisle, passenger transport jet
aircraft. The flight deck was equipped with two pilot seats and two observer seats. The aircraft
passenger cabin was configured with 148 passenger seats. Passenger seats were placed two on
either side of the aisle in First Class (forward cabin) and three on either side of the aisle in
Economy (aft cabin). There were 16 seats in the forward cabin (rows 3 to 6) and 132 seats in the
aft cabin (rows 7 to 28). Overhead stowage bins were fastened to the cabin sidewalls and ceiling
throughout the cabin seating area. The overhead stowage bin doors were designed to latch in the
closed and open positions. The doors, once unlatched, would remain in the open position until
pulled closed.
The aircraft had four floor-level cabin doors, two forward doors between row three and the cockpit,
left and right, designated L1 and R1, and two aft doors in the aft galley, left and right, designated
L2 and R2. There were four over wing emergency exits, two on the left and two on the right, at
rows 13 and 14.14 The L1 door was used as the main entry and exit door for the passengers and
crew. The remaining floor-level cabin doors were utilized for aircraft service and as emergency
exits, and the over wing exits were for emergency use only. The aircraft was equipped with three
cabin crew stations at doors L1, L2 and R2.
Doors L1 and R1 are at the forward end of the passenger cabin, with a double, rearward facing
cabin crew seat installed on the bulkhead just forward of and adjacent to L1, and the forward
galley being adjacent to R1. The cabin attendants assigned to L1 and R1 sit in this seat when
they are required to occupy their seats.
R1
R2
2 Legend
L1 L2
G – Galley
L - Lavatory
Figure 1: B737-823 Aircraft Cabin and Exits
14
See Figure 1

Each cabin crew station was equipped with a double rearward facing cabin crew jump-seat and
emergency equipment, including flashlights and life vests, in a compartment below the jump-seat.
There were communication handsets for the cabin interphone and public address system at the
L1 and L2 stations. There were two megaphones, one at L1 and one at L2.
1.6.14 Aircraft Seats and Restraint Systems
The captain and first officer seats were certified to Title 14 CFR, Part 25, Paragraph 25.785
which includes meeting the requirements of Title 14 CFR, Part 25, Paragraphs 25.561 and 25.562.
Part of the requirements of these paragraphs was for the seats to withstand 9g static and 16g
dynamic forward decelerations with 10 degree yaw.
The seats of the captain and first officer were mounted on floor tracks. The first observer seat
folded to the flight compartment wall when not in use. The second observer seat was attached to the
flight compartment wall behind the captain seat. The captain, first officer and first observer
seats each had a five-point restraint system, including shoulder harnesses, crotch strap and lap belt,
with a rotary buckle release mechanism. The second observer seat had a shoulder harness and lap
belt.
The cabin crew seats and passenger seats were certified to Title 14 CFR, Part 25, Paragraph
25.785, which includes meeting the requirements of Title 14 CFR, Part 25, Paragraphs 25.561
and 25.562. Part of the requirements of these paragraphs is for the seats to withstand 9g static
and 16g dynamic forward decelerations with 10 degree yaw. The cabin crew seats were equipped
with shoulder harnesses and seat belt restraint system. Each passenger seat was equipped with a
lift-latch style lap belt.
1.6.15 Emergency Exits – Flight Deck
The flight deck had two emergency escape windows, one on each side by the captain and first
officer positions. There were two escape lanyards, one over each pilot seat, to enable the flight
crew to escape through the cockpit windows and lower themselves to the ground.
Due to floor deformation in the forward galley area, the m e m b e r s o f t h e flight crew were
unable to open the enhanced (that is, bullet-proof) flight deck door. However, the flight deck
door had an upper and a lower panel, both of which were designed to be removed from within the
flight deck, to provide emergency egress.
1.6.16 Emergency Exits - Cabin
There were four automatic over-wing emergency exits, two on the left side and two on the right
side, located adjacent to seats 13A, 13F, 14A and 14F (Fig. 1). Each over-wing exit door was
hinged at the top and opened outward and upward. There were two over-wing escape straps in
stowage tubes above each aft emergency exit. The emergency exit had to be deployed to gain
access to the over-wing escape strap. In an emergency, the hook end of the escape strap was

attached to a fitting on the wing enabling the passengers to move safely and maintain stability on
the wing in a ditching situation.
1.6.17 Evacuation Escape Slides
The aircraft was equipped with four single-lane slides at the L1, L2, R1 and R2 exit doors to
facilitate rapid evacuation in the event of an emergency (Fig.3). Each escape slide was packed
in a valise which was stowed inside the escape slide compartment on the lower half of the
inside of each exit door (Figs. 2 and 3). The slides were manufactured by the Goodrich
Corporation.
The deployment and inflation of each escape slide was designed to be automatically initiated
when the door was opened in the armed mode. The door was armed by removing the girt bar
from the stowage hooks located on the bottom of the slide compartment and securing it into the
floor brackets. There was a chain connecting the girt bar to the slide release latch on the slide
compartment.
When the door was opened about 60 degrees, tension on the chain would release the latch and the
hinged shelf holding the slide pack in the slide compartment would fall down. This allowed the
escape slide pack to fall out of the escape slide compartment.
Figure 2. Escape Slide Pack

Illustrates the escape slide compartment, and the escape slide pack that it contains.

An inflation cable attached the girt to the inflation valve. The outward movement of the door
would cause the slide pack to fall out of the escape slide compartment as described above.
As it fell, tension on the inflation cable would open the inflation valve of the inflation
cylinder. High pressure gas in the inflation cylinder would then be released to the escape
slide and it would begin to inflate. The escape slide would take approximately six seconds to fully
inflate. In the event that inflation did not start automatically, the slide could be inflated with a
manual inflation handle.
Figure 3. Escape Slide Deployment
The escape slide compartment, which contained the escape slide pack, was attached to the aircraft’s
door by inserting the lower centre support pin on the back of the escape slide compartment into the
bottom support bracket in the aircraft door, then rotating the top of the escape slide compartment
upwards flush to the aircraft door, and securing it with screws.
There were no escape slides at the over-wing exits, as the distance to the ground from the wings,
when the flaps were extended to Flap 1, which was the highest setting that was certified for the
Boeing 737-800 for take-off or landing, was such that FAA regulations did not require slides at the
over-wing exits.

1.6.18 Emergency Lighting
1.6.18.1 Emergency Lighting - General
The emergency lighting system was composed of nine independent systems, each one powered
by a battery pack, each one of which was continuously charged from the aircraft’s Number 1 DC
bus-bar. These packs were installed in the ceiling panels at the forward and aft entry areas and on
the side structure near the floor. If electrical power to Number 1 DC bus-bar failed, or if AC
power was turned off, the emergency exit lights illuminated automatically.
Exit lights were located throughout the passenger cabin to indicate the approved emergency exit
routes. The system was controlled by a switch on the cockpit overhead panel. The switch had
three positions, OFF, ARMED and ON and was guarded to the ARMED position.
The emergency exit lights could also be illuminated by a switch on the aft flight attendant control
panel. Lifting the guard and pressing the switch ON overrode the flight deck control and
illuminated the emergency exit lights. Control from this panel was available in the event of
failure of the automatic control. The flight deck aft dome light contained a separate bulb that was
powered by the emergency lighting system to provide for flight deck evacuation.
1.6.18.2 Interior Emergency Lighting
Interior emergency exit lights were located:
 In the lower inboard corner of stowage bins, to illuminate the sidewalls.

 Over the entry/service and over-wing emergency hatches to indicate the door and hatch
exits.
 In the ceiling to locate the exits and provide general illumination in the area of the exits.
 On the inboard leg of each aisle seat.
Self-illuminating exit locator signs were installed at the forward, middle, and aft end of the
passenger cabin, and on the bulkheads between cabin classes.
Floor proximity emergency escape path lighting consisted of locator lights spaced at regular
intervals down one side of the aisle. Lighted arrows pointed to over-wing exits and a lighted
EXIT indicator was near the floor by each door and over-wing exit. Escape path markings were
provided for visual guidance for emergency cabin evacuation when other sources of cabin
lighting were obscured. Colored lighting was located on the aisle side of the inboard
passenger seats at the over-wing emergency evacuation exit rows of seats 13C and 13D, and
14C and 14D.
1.6.18.3 Exterior Emergency Lighting
Exterior emergency lights illuminated the escape slides. The fuselage installed escape slide lights
were adjacent to the forward and aft service and entry doors. Lights were also installed on the
fuselage to illuminate the over-wing escape routes and ground contact area.

1.6.19 Public Address and Communication System
The aircraft was equipped with a public address system. Under normal circumstances, the flight
crew could make announcements to the passengers and communicate with the cabin crew from the
cockpit to the L1 and L2 cabin crew stations, and the cabin crew could also make announcements
to the passengers and communicate with the flight crew using the handsets at the L1 and L2
cabin crew stations. There was also a megaphone at the L1 and L2 cabin crew stations.
1.6.20 Emergency Equipment
The cabin was equipped with portable emergency equipment in accordance with the applicable
regulatory requirements.
1.7 Meteorological Information
1.7.1 METARS on AA331 Dispatch Document
The following information was on the finalized AA Dispatch document for

AA331, given to the investigation, and was marked “RH”, meaning recorded history.
This is a record of the Kingston METARS throughout the history of the flight, from the starting
of the dispatch document to the last ACARS message to the aircraft.
KIN
230317 230300Z 32008KT 3000 PLUSSHRA BKN014 FEW016CB SCT03
230317 230300Z 32008KT 33000 ?SHRA BKN014 FEW016CB SCT030 B
230315 230300Z 32008KT 3000 ?SHRA BKN014 FEW016CB SCT030 BK
230238 230228Z 31009KT 55000 TSRA BKN014 FEW016CB SCT030 BK
230235 230228Z 31009KT 5000 TSRA BKN014 FEW016CB SCT030 BKN
230217 230200Z 300122KT 5000 SHRA BKN014 SCT030 BKN100 22/2
230008 230000Z 32004KT 9999 FEW016 BKN030 BKN100 24/19 Q101
222316 222300Z 00000KT 9999 VCSH SCT016 SCT030 BKN100 24/20
222226 222200Z 35008KT 9999 FEW015 BKN032 BKN100 24/19 Q1
222116 222100Z 33005KT 9999 FEW015 BKN032 BKN100 23/19 Q1
222007 222000Z 31005KT 9999 -RA FEW012 SCT032 OVC100 23/19
221929 221900Z 10012KT 9000 -RA SCT012 OVC100 22/19
1.7.2 Graphic Area Forecasts
Graphic area forecasts were not included in the dispatch document given to AA331 flight crew,
nor were they required.

1.7.3 Weather over Jamaica
Weather information received after the accident from the Meteorological Service of Jamaica for
the period in which the accident happened was as follows:
“Weather data shows an area of convective activity moving northward over the island at the
time of the accident. Thunderstorm and rain showers reported immediately within the hour of
the accident with heavy rain showers at the time of the accident in cumulonimbus clouds.”
and,
“Synoptic conditions indicated a stalled stationary front over the Caribbean Sea in the vicinity of
Jamaica. GOES-12 infrared satellite imagery surrounding the period indicated an enhanced
area of clouds associated with convective activity moving northward across the island of
Jamaica and over Kingston at the time of the accident. The radiative cloud top temperatures
over Kingston at Satellite imagery from GOES-12 depicted at 0232Z, 0245Z, and 0315Z
temperatures of 235.4°, 235.6°, and 247.1° Kelvin, or -37° to -26° Centigrade (C) during the
period, which corresponded to cloud tops from 30,000 to 34,100 feet.”
1.7.4 En route Weather
From the AA331 Dispatch document, indicating the wind and temperature at various flight
levels:
ENRT WX FL 240 FL 300 FL 340 FL 390
I TD WIND WCP I TD WIND WCP I TD WIND WCP I TD WIND WCP
URS 0P09/22070M022 2P10/21103M060 2P09/21116M064 0P00/21118M061
UCA 0P10/21061M048 2P10/21093M081 2P08/21107M093 0P00/21111M094
1.7.5 Thunderstorms and Lightning
Recent thunderstorm was reported in the METAR for Kingston at 22:00 EST and in the special
report at 22:25 EST; however, the flight crew reported that on approach the aircraft radar returns
showed a broad area of light to moderate rain. The Doppler radar of the Meteorological Service
of Jamaica did not indicate the presence of severe weather, beyond a lot of rain activity.
Lightning was not reported by ground observers, nor by the control tower staff, nor was it
mentioned by the flight crew of AA331. The presence of thunderstorms typically results in the
observation or detection of lightning; however, lightning was not reported by ground observers,
nor by the control tower staff while the flight was on approach and at touchdown.

1.7.6 Hourly Weather Observations
The following official weather reports were based on observations recorded by the
Meteorological Service of Jamaica weather observer at the Norman Manley International
Airport.
METAR MKJP 230200Z 30012KT 5000 SHRA BKN014 SCT030 BKN100 22/20 Q1013
RERA
SPECI MKJP 230228Z 31009KT 5000 TSRA BKN014 FEW016CB SCT030 BKN100 22/19
Q1013
METAR MKJP 230300Z 32008KT 3000 +SHRA BKN014 FEW016CB SCT030 BKN100
21/20 Q1014 RETSRA
(Accident occurred at 22:22 EST (03:22 UTC, 23 Dec)
SPECI MKJP 230325Z 32011KT 2200 +SHRA BKN014 FEW016CB SCT030 BKN100 21/19
Q1014 RETSRA
METAR MKJP 230400Z 32014KT 1500 +SHRA BKN014 FEW016CB SCT028 BKN090
21/18 Q1013 RERA
Visibilities are in metres (converted to statute miles), cloud layers heights in feet above ground
level).
These are interpreted as follows:
Kingston (MKJP) weather observation at 21:00 EST (02:00 UTC, 23 Dec), wind 300 degrees at
12 knots, visibility 5,000 m (approximately 3 miles) in rain showers, ceiling broken at 1,400 feet,
scattered clouds at 3,000 feet, broken at 10,000 feet, temperature 22° C, dew point 20° C, altimeter
setting 1013 mb, recent rain.
Kingston (MKJP) special weather observation at 21:28 EST (02:28 UTC, 23 Dec), wind 310
degrees at 9 knots, visibility 5,000 m (approximately 3 miles) in thunderstorms and moderate rain,
ceiling broken at 1,400 feet, few cumulonimbus clouds based at 1,600 feet, scattered clouds at
3,000 feet, broken at 10,000 feet, temperature 22° C, dew point 19° C, altimeter setting 1013 mb.
8 knots, visibility 3,000 m (approximately 2 miles) in heavy rain showers, ceiling broken at
1,400 feet, few cumulonimbus clouds at 1,600 feet scattered clouds at 3,000 feet, broken at
10,000 feet, temperature 21° C, dew point 20° C, altimeter setting 1014 mb, recent thunderstorm
and rain.
(Accident occurred at 22:22 EST (03:22 UTC, 23 Dec.)

Kingston (MKJP) special weather observation at 22:25 EST (03:25 UTC, 23 Dec), wind 320
degrees at 11 knots, visibility 2,200 metres (approximately 1 1/2 miles) in heavy rain showers,
ceiling broken at 1,400 feet, few cumulonimbus clouds at 1,600 feet, scattered clouds at 3,000
feet, broken at 10,000 feet, temperature 21° C, dew point 19° C, altimeter setting 1014 mb,
recent thunderstorm and rain.
This observation was taken three minutes after the accident.
14 knots, visibility 1,500 metres (approximately 1 mile) in heavy rain showers, ceiling broken at
1,400 feet, few cumulonimbus clouds at 1,600 feet in, scattered clouds at 2,800 feet, broken at
9,000 feet, temperature 21° C, dew point 18° C, altimeter setting 1013 mb, recent rain.
1.7.7 Aerodrome Forecasts
1.7.7.1 Kingston/Norman Manley International Airport (NMIA)
A Terminal Aerodrome Forecast (TAF) was issued for MKJP at 16:00 EST (21:00 UTC) for the
period 00:00 UTC on 23 December to 24:00 UTC on 23 December (19:00 EST, 22 December to
19:00 EST 23 December). The TAF forecasted wind from 340° at 8 knots, visibility 10
kilometers or more (better than 6 miles), a few clouds at 1,800 feet, scattered clouds at 3,200
feet, and a ceiling at 9,000 feet with temporary conditions between 18:00 EST, 22 December and
09:00 EST, 23 December of visibility 8,000 metres (5 miles) in moderate rain showers, ceiling
broken at 1,600 feet, few clouds at 1,800 feet in cumulonimbus clouds, and scattered clouds at
3,200 feet. The rest of this forecast is not relevant to the investigation.
TAF MKJP 222100Z 2300/2324 34008KT 9999 FEW018 SCT032 BKN090

TEMPO 2300/2314 8000 SHRA BKN016 FEW018CB SCT032
BECMG 2314/2316 20010KT
TEMPO 2318/2324 8000 SHRA SCT018 SCT080
The next scheduled TAF for MKJP was issued at 22:00 EST (03:00 UTC, 23 Dec), immediately
prior to the landing, and forecast a wind from 330 degrees at 9 knots, visibility 10 kilometers or
more (better than 6 miles), ceilings at 1,600 feet agl, with temporary conditions of visibility
5,000 metres (3 miles) in moderate rain showers. The rest of this forecast is not relevant to the
investigation. It should be noted that neither of these forecasts predicted IFR conditions or heavy
rain at the time of the accident.
TAF MKJP 230300Z 2306/2406 33009KT 9999 BKN016 FEW018CB BKN070

TEMPO 2306/2324 5000 SHRA
BCMG 2315/2317 16010KT
BECMG 2401/2403 VRB03KT SCT018 BKN032

1.7.7.2 Montego Bay - Sangster International Airport
The following forecast for Montego Bay was provided to the flight crew by AA flight dispatch
prior to the departure from Miami.
MKJS 222100Z 2300/2324 14005KT 9999 BKN016 BKN070

BECMG 2308/2310 VRB03KT PROB40
TEMPO 2309/2314 8000 SHRA BKN016 BKN030
BECMG 2314/2316 04010KT FEW022 SCT032
1.7.7.3 Grand Cayman - Owen Roberts International Airport - MWCR
The following forecast was provided to the flight crew by AA D ispatch prior to departure
from Miami.
GCM TAF MWCR 222100Z 2300/2324 07018KT 8000 -SHRA SCT018 BKN080
TEMPO 2300/2306 4000 SHRA BKN018
BECMG 2313/2315 05013KT 9999 FEW020 SCT200
TEMPO 2318/2324 SHRA SCT018
1.7.8 Area Forecast and Flash Flood Warning
At 20:21 EST, the U.S. National Weather Service in San Juan, Puerto Rico, issued the following
Area Forecast:
FXCA62 TJSJ 230121

AFDSJU
AREA FORECAST DISCUSSION
NATIONAL WEATHER SERVICE SAN JUAN PR
921 PM AST TUE DEC 22 2009
UPDATE...DEEP LAYER TROUGH WITH INTENSIFYING SURFACE FEATURE

TRACKING NORTHEASTWARD OVER THE CENTRAL CARIBBEAN WILL BE
MAIN PLAYER OVER THE NEXT FEW DAYS. SATELLITE IMAGERY SHOWS
IMPRESSIVE DEEP CONVECTION NEAR THE SURFACE CENTER CURRENTLY
JUST SOUTHEAST OF JAMAICA AND RAPIDLY LIFTING NORTHEAST.
ACCOMPANYING U/L TROUGH AND RIGHT REAR QUAD OF IMPRESSIVE 250
POLAR/SUBTROPICAL JET PROVIDING STRONG SHEAR AND DYNAMICS
MORE TYPICAL OF A PROGRESSIVE MID LATITUDE PATTERN THAN THE
AVERAGE TROPICAL SURFACE LOW. AS THIS PATTERN PROGRESSES
NORTHEASTWARD...IT WILL ENCOUNTER SOME DEGREE OF WEAKENING
AND NORTHWARD DEFLECTION AROUND THE PERIPHERY OF THE
RESIDENT LOW MID TO MID LAYER RIDGE OVER THE E CARIBBEAN. THIS
SEEMS TO BE THE NUMBER ONE FACTOR PREVENTING A MORE
WIDESPREAD CONVECTIVE EVENT.

At 11:30 EST on December 22, the Meteorological Service of Jamaica released the following flash
flood warning to the general public of Jamaica. Note: this warning was not an official aviation
product and was not reflected in the TAFs, nor was it required to be included in the dispatch
release for AA331, nor required to be reviewed by flight crews.
A FLASH FLOOD WARNING means flooding has been reported or will occur
shortly. Motorists and pedestrians should not attempt to cross flooded roadways or
other low-lying areas as strong currents are likely. Residents in low-lying areas
should be on the alert for rising waters and be ready to move quickly to higher
ground.
A Frontal System and a Trough in the vicinity of Jamaica are expected to continue to
influence the weather across the island until Wednesday. Radar reports indicate that
light to moderate showers are affecting most parishes.
The forecast is for periods of showers, along with isolated thunderstorms to continue
across the island today and tomorrow. Cool conditions are also expected especially
over northern parishes.
1.7.9 Kingston/Norman Manley International Airport, Wind Information
The weather system over Jamaica on December 22nd, 2009 was a system known locally as a
"Norther", in which the wind blows from the northwest down-slope from the hills and mountains
that form the central ridge along the east-west axis of Jamaica. The northwest winds that have a
several mile reach from the mountains, blow across the Kingston Harbour toward the airport and
runway. The runway starts at the shoreline of the harbour and for about half its length is
bordered by its parallel taxiway with the open harbour just to the north side of the taxiway.
Typically the diurnal effect reverses at night causing an increase in these winds.
The runway wind direction and speed was measured by a remote anemometer located adjacent to
taxiway Alpha, about the midpoint of runway 12-30. Data was transmitted to the weather
observer office and to the airport control tower and relayed to the Approach controller. The
anemometer’s service record showed no evidence of erroneous readings. Wind data from the
anemometer was reported in hourly and special weather observations and was also provided by
voice radio to aircraft by the air traffic controllers working in the approach control and
aerodrome control units.
1.7.10 Weather Information From Radar
The flight crew used the aircraft weather radar to navigate around some weather build ups over
Cuba and the first officer reported checking the weather radar returns in the Kingston area,
noting that he saw a wide area of green with some yellow, indicating moderate rainfall, but saw no
red to indicate convective cloud buildups with heavy rain. This interpretation was
corroborated by the returns from the Jamaica Meteorological Service Doppler Radar recording
provided by the Jamaica Meteorological Service, which indicated much the same type of
weather. The aircraft weather radar returns were not recorded by the FDR, nor was there a
regulatory requirement for the FDR to record that data.

1.7.11 Recorded Rainfall Rates At MKJP Norman Manley International Airport
The Meteorological Watch Office at MKJP recorded 20.4 millimeters/0.804 inches of precipitation
in the one hour and 55 minutes preceding the accident, consistent with heavy rain (See
1.7.11.2), falling even-spaced during this period. This station also recorded a total of 62
millimeters/2.44 inches of precipitation between 1900 EST on 22 Dec 2009, and 0100 EST on 23
December 2009. While heavy rain continued to be reported at the time of the accident, an
increase in the rainfall was noted between 21:26 EST and 21:59 EST, prior to the accident.
1.7.11.1 Recorded Rainfall at MKJP Norman Manley International Airport
Date Time (EST) mm inch

22/12/2009 09:52:00 0.2 0.00788
22/12/2009 09:53:00 0.2 0.00788
22/12/2009 09:54:00 0.4 0.01576
22/12/2009 09:55:00 0.4 0.01576
22/12/2009 09:56:00 0.6 0.02364
22/12/2009 09:57:00 0.4 0.01576
22/12/2009 09:58:00 0.2 0.00788
22/12/2009 09:59:00 0.4 0.01576
22/12/2009 10:00:00 0.2 0.00788
22/12/2009 10:02:00 0.2 0.00788
22/12/2009 10:04:00 0.2 0.00788
22/12/2009 10:06:00 0.2 0.00788
22/12/2009 10:07:00 0.2 0.00788
22/12/2009 10:09:00 0.2 0.00788
22/12/2009 10:10:00 0.2 0.00788
22/12/2009 10:11:00 0.2 0.00788
22/12/2009 10:12:00 0.2 0.00788
22/12/2009 10:13:00 0.2 0.00788
22/12/2009 10:15:00 0.2 0.00788
22/12/2009 10:16:00 0.2 0.00788
22/12/2009 10:17:00 0.2 0.00788
22/12/2009 10:18:00 0.2 0.00788
22/12/2009 10:19:00 0.2 0.00788
22/12/2009 10:20:00 0.2 0.00788
22/12/2009 10:22:00 0.2 0.00788
22/12/2009 10:23:00 0.2 0.00788
Table 4 : Recorded Rainfall at MKJP

1.7.11.2 Definitions of Rainfall
Light - up to 0.10 inch/hour; maximum 0.01 inch in 6 minutes.

Moderate - 0.11 to 0.30 inch/hour; more than 0.01 to 0.03 inch in 6 minutes.
Heavy - more than 0.30 inch/hour; more than 0.03 inch in 6 minutes.
The rainfall rate in 1.17.11.1 above, where conversion from millimeters to inches, by a conversion
factor of 0.0394, averaged 0.049 inches per 6 minutes in the 30 minutes preceding the accident.
Therefore the rainfall rate at MKJP in the 30 minutes preceding the accident meets the definition
of “Heavy”.15
1.7.12 Weather Conditions on the Ground at Kingston
Rain had been falling steadily most of the day and moderate to heavy shower activity was
frequent. Visibility in rain was reported as three to five miles and the ceiling was around 1,000
feet, with wind from the northwest at 12 to 15 knots.
1.7.13 AA Field Condition Report for Kingston
The following field condition report for MKJP Norman Manley International Airport was included
in the AA331 dispatch release:
* KIN FIELD REPORT *

*******************************************************
* REPORT LAST UPDATED AT 1520 LOCAL TIME *
*******************************************************
--------------------------------------------------------------
DATE 22DEC09 TIME 1853 LOCAL
--------------------------------------------------------------
EXISTING TAA DRP(((((((
--------------------------------------------------------------
RUNWAY STATUS CONDITIONS BRAKING ACTION/RMKS
12 OPEN WET 0.10 IN WATER
RAMP/TXWY SURFACE WET 0.10 IN WATER
--------------------------------------------------------------* KIN FIELD
This report stated that it was last updated 15:20 Local Time (EST), i.e. 5 hours before AA331
departed Miami. No other Field Condition Reports for NMIA were found to have been
generated by AA or any other agency before the accident.
15
Definitions from the (US) Federal Meteorological Handbook, No. 1, Page 8-3, Table 8-1.
1.7.14 Weather Information Provided by Air Traffic Control
At 21:48 EST the Enroute controller told AA331 that the weather was approximately 5 miles
visibility in moderate rain (see 1.1.5). The current official weather at this time was the SPECI at
21:28 EST of 5,000 meters (that is, about 3 miles) in thunderstorms and moderate rain.16
At 22:04 EST on the Direct Line the Approach controller was informed by the Tower controller
that the weather was 5 miles visibility, and he then advised AA331 that the weather was the same
as given by the Enroute controller, that is, 5 miles visibility in moderate rain (see 1.1.5). The
current official weather at this time was the METAR at 22:00 EST of 3,000 metres, (about 2
miles) in heavy rain.
ATC passed the surface wind to the aircraft, as follows:
Surface Wind Reports from ATC
Time (UTC) Direction/Speed Source

02:48:33 From 310 degrees at 7.5 knots Enroute
03:04:51 From 320 degrees at 10 knots Approach
03:17:42 From 320 degrees at 12 knots Tower
03:17:52 From 320 degrees at 14 knots Tower
03:22:22 IMPACT
Table 5: ATC Surface Wind Reports
The flight crew acknowledged the wind information, and said they would continue for
runwa y12. The Enroute, Approach and Tower controllers were providing real time anemometer
readings from the anemometer positioned just north of the parallel taxiway adjacent to the
midpoint of runway 12.
At the time of the accident there was no runway Visual Range (RVR) and Ceilometer
equipment at NMIA, hence visibility information had to be provided by means of estimates and
reliance on official hourly METAR observations.
16
Section 1.7.6

1.7.15 Rain Reports During Approach from Flight Crew
In his interview, the captain stated:
 The first officer gave him extra call-outs “because the rain was so heavy”.
 When they got into the Kingston area there was heavy rain, making the approach noisy.
 When he touched down, he thought the runway was under water.
 After he got out of the aircraft “he had never been in rain that heavy: it was like
a monsoon”.
In his interview, the first officer stated:
 That there was rain, and the wipers were on.
 The radar was painting light to moderate rain.
1.8 Aids To Navigation
1.8.1 General
The Kingston Very High Frequency Omni-directional Range (VOR)/Distance Measuring

Equipment (DME) and the Kingston runway 12 ILS/DME systems were functioning normally at
the time of the accident. The runway 12 ILS and DME were flight tested by an FAA calibration
aircraft on 29th December 2009, shortly after the accident, and found to be within normal
performance limits, as per ICAO Annex 10. The PAPI lights were not checked by this calibration
aircraft.
1.8.2 Instrument Approaches at MKJP, Norman Manley International Airport
See approach charts at Appendix 4, “Jeppesen Approach Plates for Kingston MKJP/KIN”. These
approach plates were published by Jeppesen, and issued to, and used by, the AA331 flight crew.
1.8.2.1 Runway 12 Instrument Approaches
Runway 12 was served by a VOR/DME approach, an ILS approach and an RNAV Global
Positioning System (GPS) approach.
MKJP ILS 12 approach included a localizer and a glide path for lateral and vertical electronic
approach path guidance. ICAO Annex 10, Volume 1, Sixth edition (2006), permitted installation
of a laterally offset localizer antenna where there was insufficient land off the departure end of
the runway to site the antenna in the optimal location. There were limits as to the lateral offset
effect on the final approach course provided by that installation and the MKJP localizer 12 fell
within the limits contained in ICAO Doc 8168, Vol II PANS-OPS.
In practical terms it meant that the final approach course was offset to the right by 3 degrees
from the extended course of the runway centerline. It therefore intercepted the extended course
of the runway centerline approximately 0.75 NM from the threshold of the runway at a glide path
height of 280 feet above the runway 12 touchdown zone elevation.
The ILS localizer was on a magnetic track of 120 degrees, offset 3 degrees right of the runway
magnetic track, which was 117 degrees (See Appendix 4). The ILS glideslope intersected with
the runway surface 1,000 feet beyond the runway threshold.
The ILS was classified as an ILS Category One precision approach.
There was a PAPI system for runway 12, the glide slope of which intersected with the runway
surface 1,400 feet beyond the runway threshold.
The ILS runway 12 approach was designed based on International Civil Aviation Organization
(ICAO) Procedures for Air Navigation Services Operations (PANS OPS), based on criteria
contained in ICAO Doc 8168.
For the approach, the flight crew of AA331 used the Kingston, Jamaica ILS runway 12 Jeppesen
approach chart (22 AUG 08 (11-1)) (See Appendix 4), issued to them by American Airlines. This
chart stated “PANS OPS” in the lower left hand corner, indicating that these were the design
specifications used.
The ILS approach minima in an approach based on “PANS OPS” was established by specific
obstacle clearance criteria for the category of the ILS serving that particular runway, and was not
predicated on the availability of an Approach Lighting System (ALS).
The ILS runway 12 approach chart stated that 278 feet above sea level (asl) /270 feet above
ground level (agl) was the Decision Altitude/Height (DA/H).
The ILS runway 12 approach chart stated further, that the visibility of 1.9 kilometers was for
“FULL” and “ALS OUT” conditions.
ICAO Doc 8168, Vol. 2, Fifth Edition, amendment 4, did not contain any visibility requirements
as a basis of procedure design for straight-in ILS approaches. However, suggested (advisory)
visibility requirements were promulgated for approaches that involved visual maneuvering
(circling). ICAO Doc 8168, Part - I, Section 4, Chapter 7, 7.2.4 stated: “ … This information is
not required for the development of the procedure, but is included as a basis for the development
of operating minima.”

1.8.2.2 Runway 30 Instrument Approaches
Runway 30 was served by an RNAV (GPS) approach, with a PAPI system and a REIL (Runway
End Identification Light) system. The MDA was 390 feet asl /373 feet agl. Also, there was a
published circle-to-land approach for runway 30 from the ILS Runway 12 approach; the MDA for
Boeing 737 (Category C aircraft) was 1,150 feet asl/1,140 feet agl, with a visibility of 3.7
kilometers.
1.9 Communications
No malfunctions of ground or aircraft radio communications systems were reported between the
time AA331 first made contact with the Kingston Air Traffic Services, and the time the aircraft
lost power.
Communications in the aircraft during the flight were reasonable and effective. En route, the
captain advised the cabin crew and passengers about the weather and related turbulence.
According to the cabin crew interviews, the captain instructed the cabin crew to suspend the
cabin service, stow service items and take their seats due to the turbulence. Before commencing
descent, the captain requested that everyone remain seated due to the anticipated turbulence on the
approach into Kingston and for the cabin crew to prepare early for the landing.
No brace command was given prior to the accident by either the flight crew or cabin crew.
As soon as the aircraft came to a stop, the captain called out “Easy Victor” to notify the cabin
crew to evacuate the aircraft, but he was not sure that the PA system was working.
The flight attendant at R2 inboard reported that she thought she heard the captain yelling “Easy
Victor”.
The forward flight attendant used the megaphone to yell her commands.

1.10 Airport Information
Photo 1 Aerial Photograph of Accident Site
1.10.1 General
The accident occurred at Norman Manley International Airport (NMIA), Kingston, Jamaica,
ICAO identifier MKJP, position North 17 degrees 56.1 minutes, West 076 degrees 47.3 minutes,
magnetic variation 5.7 degrees west, elevation 10 feet asl.
MKJP was two miles south of the city of Kingston on the southeast coast of the island of
Jamaica, situated on a long spit of land leading to Port Royal, and oriented east/west.
MKJP was operated by the NMIA Airports Limited, and the Airports Authority of Jamaica. The
Jamaica Civil Aviation Authority ( JCAA) had regulatory oversight. At the time of the
accident MKJP was not certified by the JCAA, however the certification of the airport was in
progress.
MKJP was governed by the Civil Aviation Act of Jamaica and the Civil Aviation Regulations of
Jamaica (JCARs) 2004, as amended, Part XIII Aerodromes. As such, NMIA Airports Limited
was required to operate in compliance with these regulations. The regulations empowered the use
of the Aerodrome Standards and Recommended Practices contained in ICAO Annex 14 and
related documents.

MKJP operated 24 hours a day and was open for traffic operating under visual flight rules and
instrument flight rules (VFR and IFR).
Runways 12 and 30 were closed immediately after the accident, and reopened on the morning
following the accident with modified Declared Distances to compensate for the protrusion of the
wreckage of AA331 into the airspace at the end of runway 12.
MKJP had a maintenance program. This maintenance program was not approved as required by
Regulation 108 and the Twenty First Schedule, paragraph 21.135,
An inspection of the aerodrome was made subsequent to the accident, by the Airports Group
investigators. The resulting report is in Appendix 10.
1.10.2 Runway Description
1.10.2.1 Runway 12
The Aeronautical Information Publication, Jamaica, Amendment 01/07, valid at time of the
occurrence, stated the following information regarding MKJP runway 12:
 Runway 12 was 8,911 feet long by 151 feet wide (2,716 metres by 46 metres). The
orientation of runway 12 was 112 degrees True/117 degrees Magnetic.
 The take-off run available (TORA) of runway 12 was 8,911 feet (2,716 metres), take-off
distance available (TODA) was 13,367 feet (4,073 metres), accelerate stop distance
available (ASDA) was 8,911 feet (2,716 metres), the landing distance available (LDA)
was 8,911 feet (2,716 metres).
 The slope of runway 12 was 0.02/0.0.
There was a short up-slope starting at about 6,911 feet from the runway threshold. The gradient
of the up slope at the departure end of the runway did not fall within the recommendations of
ICAO Annex 14. 17
The elevation change down the centerline of the runway was 9 feet.
1.10.2.2 Runway 30
The Aeronautical Information Publication, Jamaica, Amendment 01/07, valid at time of the
occurrence, stated the following information regarding MKJP runway 30:
 Runway 30 was 8,911 feet long by 151 feet wide (2,716 metres by 46 metres). The
orientation of runway 30 was 292 degrees True/297 degrees Magnetic.
17
Section 1.10.3

 The take-off run available (TORA) of runway 30 was 8,911 feet (2,716 metres), take-off
distance available (TODA) was 13,367 feet (4,073 metres), accelerate stop distance
available (ASDA) was 8,911 feet (2,716 metres), the landing distance available (LDA)
was 8,911 feet (2,716 metres).
 The slope of runway 30 was 0.02/0.0
The elevation change down the centerline of the runway was 9 feet.
1.10.2.3 Runway 12/30, Transverse and Longitudinal Slopes
The drawings from an engineering survey commissioned for this investigation indicated that a
transverse slope of 0.9% away from the centerline on both sides was typical along the full length
of the runway. Runway construction also included drainage. A survey of the transverse slope did
not identify any deviations from engineering drawings.
The meaning of the slope information for both runways in the AIP, which stated “The slope of
runway 12 (and 30) was 0.02/0.0”, was not clear, and no explanation for this was found.
A longitudinal survey of runway 12/30, prepared by EDM Consultants, dated February 1997,
titled: “Rehabilitation of Runway, Taxiway and Pavement Works, Runway Profile” was provided
to the investigation by the Airports Authority of Jamaica. This provided the following
information:
1. The slope for runway 30 from 0 m to 260 m (853 ft) was -1.233%
2. For the portion of runway 30 from 0 m to 260 m (853 ft) it stated “This portion of runway
longitudinal slope exceeds 0.8%”.
3. The slope for runway 30 from 260 m (853 ft) to 790 m (2,592 ft) was -0.098%.
4. The slope for runway 30 from 790 m (2,592 ft) to 1,200 m (3937 ft) was -0.052%.
5. The slope for runway 30 from 1,200 m (3,937 ft) to 2,335 m (7,661 ft) was 0.004%.
6. The slope for runway 30 from 2335 m (7,661 ft) to 2,720 m (8,924 ft), was -0.034%.
ICAO Annex 14, Vol 1, Fifth Edition – Aerodromes, 1.7.4, Table 1-1, states that an aerodrome
with a field length of
1,800 m and over is a Code number 4 Aerodrome reference code.
NMIA is a Code number 4 aerodrome.
ICAO Annex 14, Vol 1 states: “Longitudinal slopes, Recommendation – the slope computed by
dividing the difference between the maximum and minimum elevation along the runway centre
line by the runway length should not exceed … 1 per cent where the code number is … 4.” 18
18
ICAO Annex 14 – Aerodromes, 3.1.13

On the subject runway survey the elevation above sea level at the west end (threshold of runway
12) was 2.476 m, and the elevation at the east end (threshold of runway 30) was 5.676 m,
(difference is 3.2 m), and the runway length was 2,720 m, giving a slope of 1.18%.
ICAO Annex 14 – Aerodromes states: “Longitudinal slopes, Recommendation – Along no portion

of a runway should the longitudinal slope exceed … 1.25 per cent where the code number
is 4, except that for the first and last quarter of the length of the runway the longitudinal slope
should not exceed 0.8 %.” 19
Along no portion of the runway did the longitudinal slope exceed 1.25%.
The longitudinal slope of the first 260 m (853 ft) of the first quarter of runway 30 exceeded
0.8%.
ICAO Annex 14 – Aerodromes, states: Declared distances.
a) Take-off run available (TORA). The length of runway declared available and suitable for
the ground run of an aeroplane taking off.
b) Take-off distance available (TODA). The length of the take-off run available plus
the length of the clearway, if provided,
and recommends: “The length of the clearway should not exceed half the length of the
runway.”
Both runway 12 and runway 30 ended at the ocean, which provided an unobstructed clearway,
therefore the clearway f or bot h runwa y 12 a nd runwa y 30 was d ecl ar ed to be half the
length of the runway, that is, the TODA was 8,911 feet x 1.5 = 13,367 feet.
1.10.2.4 Runway Survey
A visual survey of the runway was conducted. This showed:
1. There was evidence of edge damming on both sides of the runway.

2. There was evidence of water flowing off both sides of the runway, then soaking away.
3. Maintenance of drainage system was poor.
4. There were obstacles on runway strip. (rocks, old concrete blocks and uncovered
drainage cisterns).
5. Runway surface had polishing in touchdown zone of runway 12.
6. Surface was not deformed, and had positive transverse slopes along entire length.
19
ICAO Annex 14 – Aerodromes, 3.1.14

7. There was no cracking, apart from longitudinal joint cracks on entire length, and lateral
cracks associated with construction, plus some block cracking in touchdown zone of
runway 12.
8. Runway was in good condition overall.20
1.10.2.5 Runway Drainage System
The runway drainage was achieved by the transverse slope on either side of the runway so that
the water would flow away to the sandy areas to the south and the centerfield to the north and
then enter into the drainage system and then soak away into the drainage ducts.
Examination of the surveyor’s drawings showed no evidence of the drainage system in the
runway strip being linked to the main drainage system or drainage ducts routed under the apron.
1.10.2.6 Runway 12 Lighting and Markings
The following information was gathered from the Aeronautical Information Publication,
Jamaica, Amendment 01/07, valid at the time of the accident, and from the operator, NMIA
Airports Limited:
 The approach lighting for runway 12 was an SIAL system, with lead-in lights.
 Runway 12 was equipped with a PAPI. The design angle of the beam projection was 3º
and was required to conform as closely as possible to the angle of the ILS glide path. The
PAPI consisted of four light units on each side of the runway, each being in the form of a
horizontal bar. These were located 1,400 feet from the runway threshold. The aircraft was
on slope if the two units furthest from the runway showed white, too high if all units
showed white and too low if all units showed red.
 Runway 12 was equipped with high-intensity runway edge lights (HIRL), uniformly
spaced at 60 m (196 feet) intervals. There were five variable intensity settings. At the
time of the occurrence, the runway edge lights were on setting 5, the maximum setting.
 The HIRL lights were white for the first 5,911 feet of runway, alternate red and
white for the next 2,000 feet, and red for the last 1,000 feet. Runway 12 was also
equipped with high speed exit taxiway lights.
Runway 12 was not equipped with touchdown zone (TDZ) lighting, nor was this required or
recommended by ICAO Annex 14 for Category 1 systems.
Runway 12 was not equipped with distance remaining markers nor was it required to be by
Jamaica Civil Aviation Regulations, 2004, as amended, nor was this recommended by ICAO,
Annex 14.
20
Appendix 10, Airports Group Report

At the time of the accident there was an active NOTAM for the SIAL system for Runway 12 that
indicated that the system was not operational - see NOTAM in section 1.18.3 informing aircraft
operators of this. This NOTAM had been in effect since 30 November 2009. This did not affect
the use of the runway for the landing of AA331.21
All other parts of the lighting system for runway 12 were serviceable. Runway 12 had white
runway markings consisting of the following:
• Threshold markings - a series of vertical bars marking the threshold;

• Runway designation markings, consisting of the runway number;
• Touchdown zone markings, consisting of repeating series of vertical bars either side of
the centreline, every 500 feet within the first 3,000 feet of the runway;
• Aiming point markings at 1,500 feet from the threshold, and
• Centreline markings - a dashed line indicating the centreline of the runway.
There was no lighting embedded in the runway surface, and the runway markings did not contain
any reflective material.
ICAO Annex 14 (valid at the time of the accident) recommended that, “At aerodromes where
operations take place at night, pavement markings should be made with reflective materials
designed to enhance the visibility of the markings.” 22 (bolding added).
All markings and lighting on runway 12 met the standards for runway lighting and markings of
ICAO Annex 14.23 Though the Norman Manley International Airport was not certified in
accordance with Part XIII of the JCARs 2004, NMIA Airports Ltd., as an applicant for an
aerodrome certificate, was required by Schedule 21, 21.135 to detail the particulars of the
procedures for the inspection and maintenance of aeronautical lights (including obstacle
lighting), signs, markers and aerodrome electrical systems.
Jamaica Civil Aviation Regulations, 2004, as amended, stated:
113. An aerodrome operator shall ensure that his aerodrome is operated and maintained
with a reasonable degree of care and diligence and compliance with the standards and
practices specified in Regulation 91(2).
118. - (1) Subject to any directions issued under maintenance of paragraph (2), an
aerodrome operator shall operate a n aerodrome. and maintain an aerodrome in
accordance with the procedures set out in the approved Aerodrome Operator’s Manual.
21
Section 1.8.2.1
22
ICAO Annex 14, 5.2.1.7, Doc. 9157 Part 4, “Aerodrome Design”
23
ICAO Annex 14, Aerodromes Design and Operations, Volume 1, Chapter 3

The NMIA had a draft manual, but did not have any approved manual at the time of the
accident. In the absence of any approved manual and in the absence of an Aerodrome
Certificate, the aerodrome operators of the NMIA were being held to the standards contained in
ICAO annex 14, Volume 1, Fifth Edition.
Annex 14 stated:
“The system of preventive maintenance employed for a precision approach runway category I
shall have as its objective that, during any period of category I operations, all approach and
runway lights are serviceable and that, in any event, at least 85 per cent of the lights are
serviceable …” 24
1.10.2.7 Runway 30 Lighting and Markings
The following information was gathered from the Aeronautical Information Publication,
Jamaica, Amendment 01/07 (valid at the time of the accident), and the NMIA operator:
 The approach lighting for runway 30 was a Simple Instrument Approach Light (SIAL)
system. This did not have lead-in lights.
 Runway 30 was equipped with a precision approach path indicator (PAPI). The design
angle of the beam projection was 3º. The PAPI was located 1,400 feet from the runway
threshold, and consisted of four light units on the each side of the runway in the form of a
horizontal bar. The aircraft was on slope if the two units nearest the runway showed red
and the two units furthest from the runway showed white; the aircraft was too high if all
units showed white, and too low if all units showed red.
 Runway 30 was equipped with high-intensity runway edge lights (HIRL), uniformly
spaced at 60 m (196 feet) intervals. There were five variable intensity settings.
 Runway 30 was equipped with runway End Identification Lights (REIL), which were
white strobe lights identifying the threshold of the runway.
 Runway 30 was not equipped with touchdown zone (TDZ) lighting, nor was this
recommended by ICAO standards or required by JCARs.
At the time of the accident, all parts of the lighting system of runway 30 were operational.
Runway 30 had white runway markings consisting of the following:
• Threshold markings - a series of vertical bars marking the threshold;

• Runway indication markings, consisting of the runway number;
• Touchdown zone markings, consisting of repeating series of vertical bars either side of the
centreline, every 500 feet within the first 3,000 feet of the runway;
• Aiming point markings at 1,500 feet from the threshold; and
24
ICAO Annex 14, Chapter 10, 10.4.10

• Centreline markings - a dashed line indicating the centreline of the runway. The centreline was
not lighted.
The runway was not equipped with distance remaining markers nor was it required to be by
Jamaica Civil Aviation Regulations 2004, as amended, nor was this recommended by ICAO,
Annex14.
There was no lighting embedded in the runway surface, and the runway markings did not contain
any reflective material.
ICAO Annex 14 states: “At aerodromes where operations take place at night, pavement markings
should be made with reflective materials designed to enhance the visibility of the markings”
(bolding added).25
Apart from the lack of reflective material in the runway markings, all the relevant markings and
lighting on runway 30 met the recommendations for runway lighting and markings of ICAO
Annex 14, Aerodromes Design and Operations, Volume 1.
There was a maintenance program for the PAPIs, as per NMIA Maintenance Manual, 5.7.
The PAPIs for both runway 12 and runway 30 were examined and found to meet the
requirements of ICAO Annex 14 (5.3.5.1). All units were operational and appeared to be well
maintained.
As determined by the maintenance records the previous calibration/maintenance was performed

on 08 August 2008. The systems were found to be within calibration standards when checked on
26 December 2009. Periodicity of maintenance was not specified in ICAO Annex 14 and was
normally subject to the manufacturer’s specifications.
The Aerodrome Group Report recommended that MKJP institute a specific

maintenance/calibration periodicity and a periodic maintenance schedule to service the PAPIs.
1.10.2.8 Runway End Safety Area (RESA)
Neither runway 12 nor runway 30 had a RESA. Runway 12 ended with a drop of about 12 feet
to a road running north/south, then a rocky berm which rose to the shore of the Caribbean Sea.
Runway 30 ended in the waters of Kingston Harbour. There were no obstructions at either end
of the runway.
25
ICAO Annex 14, 5.2.1.7, Doc. 9157 part 4, “Aerodrome Design”

1.10.3 Runway Surface Friction Measurement
The runway surface was hot mix asphalt (HMA), hard paved and was not grooved. The
maintenance department of NMIA Airports Limited had a program in place to remove
rubber deposits and take runway friction measurements. NMIA Airports Limited did not have its
own friction measuring equipment but rented it when measurements were to be taken.
The previous friction test was conducted five years before the accident. ICAO Annex 14 states:
“Measurements of the friction characteristics of a runway surface shall be made periodically with a
continuous friction measuring device using self-wetting features.” 26
Thus, as the ICAO recommendation did not specify a maximum period between tests, the
frequency of friction testing did meet the ICAO recommendation. The relatively long period
between tests, in this case five years, was due to the NMIA not being in possession of the testing
equipment.
For runway 12, the most recent runway friction measurement before the accident was
performed on 17 July 2004 by technical personnel from Grantley Adams International Airport
(GAIA), Barbados, using a Findley Irving Mark 2 Grip Tester, Serial Number GT 289, registered
to the GAIA.
The results of this measurement indicated that, for runway 12, the average reading for the first
third of the runway was 0.65, the second third was 0.62, and the last third was 0.58. The
friction levels measured by the Grip Tester exceeded the values listed in Table A-1 of Section
7.9 of ICAO Annex 14 volume 1 for maintenance planning level and the minimum friction level.
The friction levels given were absolute and were intended to be applied without any
tolerance when the average coefficient of friction for the entire runway fell below 0.43 at a speed
of 65 Km. There were recommendations for rubber removal in the touchdown areas from 275 to
650 metres along the runway three metres south of the centreline, and this work was
subsequently done.
As part of the accident investigation, a similar friction measurement procedure for runway 12
was performed by GAIA technical personnel, using the same equipment, on 05 January 2010,
fourteen days after the accident. Standard measurements were made in the wheel path area of the
runway, approximately three metres left and right of centreline. Additional full-length friction
measurement tests were conducted at 6 metres left and right of centreline to further define the
surface characteristics of the runway.
The results indicated that, at three metres left and right of centreline, the average reading for
the first third of the runway was 0.66, the second third was 0.60, and the last third was 0.54. In
26
ICAO Annex 14, Vol. 1, Ch. 10, Paragraph 10.2.3

the first third, the lowest reading for 100 metres was 0.51, and in the last third, there were two
readings of 0.48. These readings were above levels requiring maintenance action.
The report from GIAI stated “All of the frictional averages obtained are well over the
recommended ICAO minimum friction value of 0.42”.
The runway macro texture was measured and calculated for Runway 12 using the Langley
Technique for Measuring Surface Roughness, also known as the “Grease Test”.
Five separate areas of the runway were examined. The texture measurements were not taken in
areas where the wheel braking occurred. Qualitative observation indicated the area where the main
gear wheel braking occurred had less texture or was more “polished” than the locations where
the Grease Test was performed.
The runway was observed following rain activity in February 2010. The whole runway was
wet, with about 20% of the surface covered with shallow water patches.
1.10.4 Runway Surface Condition Reports
1.10.4.1. AIP Jamaica
The A e ron a ut i c al I n f o r m a t i o n P u b l i c a t i o n (AIP) states:
“Information on standing water at Kingston/Norman Manley and Montego Bay/Sangster is

transmitted by Air Traffic Services (ATS) to landing and departing aircraft. Drainage of the
runways at these aerodromes is generally good, however standing water occurs during heavy
rains and runways at times become slippery during these periods. No facilities exist for the
measurement of standing water at these aerodromes”. 27
The JCAA staff responsible for the AIP stated that the subject statement has been in the Jamaica
AIP since 01 January 2000, and was in the Jamaica AIP at the time of the accident, as AIP
Amendment 02/07. No documented history of, or background to, this statement was available from
the JCAA or from any other source.
1.10.4.2 Air Traffic Control (ATC)
There was no braking action report requested from, or given by, the arriving aircraft preceding
AA331, an Airbus 320, which landed at 21:07 EST, approximately one hour and 15 minutes
prior to AA331 landing.
The only runway condition report given to AA331 was given by the Tower controller at 03:17:57
27
AIP for Jamaica, AD 1.1-1, 30 Sep 07, “3. Dissemination of information of runways affected by water”

UTC, less than 5 minutes before impact, was, “Be advised runway wet”. ATC did not inform
AA331 that no braking action report was available. 28
Heavy precipitation was recorded prior to and during the landing of AA331 by the
Meteorological Watch Office 29 and reported in the 03:00 UTC METAR and the 03:25 UTC
SPECI (See 1.7.6). The evidence indicated that the 03:00 UTC METAR was transmitted to the
aircraft by ACARS 30 at 22:15 EST and 22:17 EST. ATC did not inform the AA331 flight crew
of “heavy” rain.
In a post-accident interview the Tower controller stated that the wet runway condition report was
based on the controller’s observation that it was raining.
The ATS Manual of Operations (MANOPS) stated that the Approach controller, on initial
contact, should give the current runway condition, and the latest braking action report, or state
that none had been received. The Approach controller did not do this.
The Meteorological Services of Jamaica provides aeronautical weather services in Jamaica through
an aeronautical meteorological watch office at NMIA and an aeronautical meteorological station at
MBJ.
1.10.4.3 AA Flight Manual Part II Jeppesen Page 10-7X
The AA Flight Manual Part II Jeppesen Page 10-7X for Kingston, Jamaica, dated 11 JUL 08,
stated “Runway is uneven and subject to pools of standing water after heavy rain.” 31
1.10.4.4 American Airlines Field Condition Reports
The AA Dispatch Procedures Manual, states:
“Field Condition Reports are updated by station personnel at the station opening, every
three hours after, when conditions have changed or when pilot braking action reports are
received.” 32
28
Section 1.1.3
29
Section1.7.11
30
Section 1.7.1
31
Appendix 5 to Section 1
32
AA Dispatch Procedures Manual, Section 10, Page 14, 8/4/08, Paragraph 5. Field Condition Reports, 5.1

Also, the AA Station Manual for Kingston, 11.2 A stated:
“Airport Inspections will be conducted at least weekly by each General Manager or a

designated representative. Inspections will increase in frequency during periods of
inclement weather. Inspections should provide detailed information concerning …
condition of runways … amount of standing water on runways … ”
Also, in 11.4, it stated:
“During periods of severe or inclement weather, local station management is responsible

for keeping Dispatch informed and for updating the local field condition report.”
The following runway condition report was in the AA331 dispatch document:
*******************************************************
*******************************************************
--------------------------------------------------------------
--------------------------------------------------------------
--------------------------------------------------------------
--------------------------------------------------------------
(bolding added)
This report stated it was last updated at 15:20 Local Time (EST), that is, 5 hours before AA331
departed Miami.
When the AA NMIA Station Manager was asked if AA procedures required the obtaining of a
runway Condition Report when the weather was inclement, it was stated that weather reports go
from Meteorological Services to AA dispatch in Dallas and that the AA station staff would have to
get that information from Meteorological Services or the Control Tower.
It was also stated that the AA station staff do not have airside passes for the maneuvering area and
do not do runway condition reports. When asked if extremely heavy rainfall would prompt a
runway inspection by AA, the response was “No, NMIA does that.”
AA dispatch was asked where this runway condition report came from, and responded: “The AA
operations agent in Kingston entered the information into the KIN field report. The ops agent had
put this in their field condition report since they could see water on the tarmac after many hours of
torrential rain on that date. It was not an accurate measurement as it was not obtained from an
airport or government source. The operations agents in KIN, at that time, thought if they called the
field “wet” they needed to add a numerical value to the description.” AA also stated, “The
numerical value was based on standard practices, and not a measurement that was obtained from
an airport or government source.” There was no evidence that any such field condition reports
were conducted at NMIA in the period before the accident by any agency, or disseminated to AA
personnel.
There was no evidence that the AA331 flight crew was aware that this field report was not based
on any observation or measurement.
It should be noted that it is impossible to accurately measure the water depth over an entire
runway, and qualitative reports from visual inspections are the best that can be done. ICAO does
have a format or standard to describe such a process at this time. 33
1.10.4.5 Runway Condition Reports – ICAO Requirements
ICAO Document 4444 PANS-ATM, stated:
Information that water is present on a runway shall be transmitted to each aircraft

concerned, on the initiative of the controller, using the following terms:
DAMP — the surface shows a change of colour due to moisture.

WET — the surface is soaked but there is no standing water.
WATER PATCHES — patches of standing water are visible.
FLOODED — extensive standing water is visible. 34
ICAO Document 9137, stated:
“Standing water checks should be carried out on request from air traffic control or
airport operations. A verbal assessment for the centre half of the width of the runway is
required (see Annex 14, Chapter 2). On completion of the check the results should be
passed to air traffic control and recorded for reference purposes.” 35
ICAO Annex 14, describes the obligation of the aerodrome operator:
“Information on the condition of the movement area and the operational status of related
facilities shall be provided to the appropriate aeronautical information services
units, and similar information of operational significance to the air traffic services
units, to enable those units to provide the necessary information to arriving and
33
Section 1.10.4..5
34
ICAO Document 4444 PANS-ATM, fifteenth edition (2007), Chapter 11, “Air Traffic Services
Messages,” 11.4.3.4.2
35
ICAO Document 9137, Part 8, Chapter 6 “Rain,” 6.5.2

departing aircraft. The information shall be kept up to date and changes in conditions
reported without delay.” 36
“Airport Operational Services”, the ICAO “Airport Services Manual” stated:
“During adverse weather, airport operations will advise air traffic control of relevant
surface conditions and should carry out such various checks that the weather may
dictate.” 37
Such surface conditions, as stated in Section 6.2.2, include ‘rain – giving reduced runway
friction’.
There was no provision in Part 8 for a methodology to be used by which an airport operator
could provide quantifiable data related to standing water on the landing surface of the runway.
1.10.4.6 Runway Condition Reports at Kingston
The JCAA ATS MANOPs, Second Edition, Effective September 1992, included phraseologies
which made use of the term “Wet”, however, the conditions under which a runway was reported
as being “Wet” were not defined in the JCAA ATS MANOPs.
ICAO Annex 11, “Information on aerodrome conditions and the operational status of associated
facilities” stated:
Aerodrome control towers and units providing approach control service shall be kept
currently informed of the operationally significant conditions of the movement area,
including the existence of temporary hazards, and the operational status of any associated
facilities at the aerodrome(s) with which they are concerned. 38
There were no ATS Unit Specific instructions that required aerodrome controllers to obtain runway
reports from NMIA during periods of inclement weather, nor was there evidence of any
Memorandum of Understanding (MOU) between KIN Tower and MKJP regarding such a process.
In the absence of a formal agreement between the parties, there was no standardized, mutually
agreed-upon procedure to obtain and report runway surface conditions.
36
ICAO Annex 14, Volume 1, Fifth Edition, Chapter 2, 2.9.1
37
Part 8 “Airport Operational Services”, Section 6.2 of the ICAO “Airport Services Manual”
38
ICAO Annex 11, 7.2

P art 8 “Airport Operational Services”, Section 6.2 of the ICAO “Airport Services Manual”,
stated:
“During adverse weather, airport operations will advise air traffic control of relevant
surface conditions and should carry out such various checks that the weather may
dictate.”
NMIA was not in compliance with this.
The NMIA Operations Manual, which had not been approved by the JCAA, did not indicate any
changes to the pavement inspection frequency if inclement weather was being experienced at the
aerodrome. “Operating Procedures and Safety Measures,” indicated the following:
1. The pavement of the maneuvering area, (movement area, excluding the apron), was
inspected every morning at daybreak by the Airport Protection Services.
2. An airside surface inspection was conducted by the Operations Coordinator at the
beginning of every shift. 39
The manual did not appear to indicate the number of shifts available or when they began,
therefore it was not possible to determine the frequency with which airside inspections were
conducted.
The management of NMIA Airports Ltd. was questioned as to the runway inspections that were
conducted during 22 December 2009. Two completed “Daily Airside Inspection Report” forms
for that date were produced, one indicating that Airport Protection Services (APS) did an
inspection at 06:20 Local, and the other that APS did an inspection at 06:35 Local. It was stated
that each Operations Coordinator conducts two inspections, one around the beginning of his/her 8-
hour shift, and the other around the end of the shift, and that only if there was a noteworthy finding
was it reported to Airport Operations Centre (AOC) and recorded in the AOC log.
It was stated that inspections were done on that day, but there were no adverse findings, so no
report was made. Apart from this statement, there was no other evidence, documented or
otherwise, available to confirm that these inspections were done.
There was no evidence that the runway was examined by the personnel of AA, ATC, NMIA or
any other organization at the NMIA airport during the heavy precipitation in the period preceding
the time of this accident. There was no indication that any runway condition reports were
generated, or disseminated to AA personnel from ATC, NMIA or any other organization in the
period before the accident.
39
NMIA Operations Manual, Original Issue (September 2008), Chapter 4 “Operating Procedures and Safety Measures,” Paragraph
4.5, sub-paragraph 4.5.5.

1.10.5 Airport Rescue and Fire Fighting (ARFF)
In accordance with the Civil Aviation Regulations of Jamaica, as amended 2004, Paragraph 134,
NMIA Airports Limited, the operator of MKJP, provided Airport Rescue and Fire Fighting
Services Category 8 ARFF, as specified by ICAO Annex 14, Section 9.2.
The airport was a single runway with a parallel taxiway and four connecting taxiways. The fire
station was located on the north side of the taxiway at approximately two thirds of the distance of
the total runway length from the threshold of runway 12.
According to ICAO Annex 14 the operational objective of the rescue and fire-fighting services
shall be to achieve a response time not exceeding three minutes to any point of each operational
runway, in optimum visibility and surface conditions. There was no direct access from the fire
station to the runway; however the routes utilizing the taxiway to either threshold allowed the
services to meet the required response time to either threshold.
The aircraft left the runway at 22:22:21 EST. One ARFF fire tender was at that time deployed
on the main ramp providing standby fire protection during the refueling of a Boeing 747 aircraft
which was about to depart. The driver of the fire tender on seeing the aircraft departing the end
of the runway, went immediately to the end of runway 12, across from the crash site, reaching it
in a position from which it could project foam fire suppressant over the entire accident site, at
22:25 EST.
The crew of this fire tender gave a situation report to the other two ARFF fire tenders and
advised them that they would have to access the site by driving off airport and by the Port Royal
road around to the accident site. These two fire tenders reached the crash site at
22:31 EST.
As there was no fire, the ARFF fire-fighting capability was not needed, but they remained in
attendance due to the continued risk from the spilt fuel and electrical hazards. The investigation
found that the ARFF services at MKJP met the regulatory requirements, and that the response was
adequate.
1.10.6 Meteorological Services and ATIS weather information
The Jamaica Meteorological Services (JMS) provided weather observations and forecasts at
MKJP throughout the 24 hour period. Landing Trends could be provided with the landing
forecast upon request.
The JMS also provided weather data for the Airport ATIS system. This was achieved by an
Automated Weather Observing System (AWOS) located 1,333 metres from the threshold of
runway 12 and 208 metres from the centerline of the runway, on the north side. This site
contained the equipment to provide certain elements of the ATIS information, that is, a wind vane
anemometer and tipping bucket rain gauge. Elements such as visibility were not obtained by
automated readings since NMIA did not have a transmissometer.

1.10.7 Airport Lighting
The ambient lighting around the buildings at the airport was less than usual, as there had been a
power cut, and the airport was powered by standby generators.
1.11 Flight Recorders
1.11.1 Cockpit Voice Recorder (CVR)
Recorder Manufacturer/Model: L-3 Communications FA 2100-1010

Recorder Serial Number: 000142599
For the purposes of the investigation of this occurrence, a CVR Group was formed, whose
members represented the JCAA, National Transportation Safety Board (NTSB), FAA, Allied
Pilots Association, AA and the Boeing Company,
The CVR Group meeting convened on 28 December 2009, at the premises of the NTSB,
Washington DC. The CVR was examined at the CVR laboratory of the NTSB, and a complete
transcript o f the 30-minute, 38-second digital recording was prepared.
This model CVR recorded 30 minutes of digital audio stored in solid-state memory modules.
Four channels of audio information were retained: one channel for each flight crew, one channel
for the cockpit area microphone (CAM) and one channel for the flight observer/third crewmember.
The CVR had not sustained any heat or structural damage. The audio information was extracted
from the recorder without difficulty and was of good quality.
The CVR data was correlated to local time using the air traffic control transcript provided by the
JCAA. The recording and transcript began at 21:51:44 EST as the aircraft was descending
toward Kingston. The recording contained events from the descent, approach, and accident
sequence. The recording ended immediately following the accident at 22:22:21 EST.
The flight crew was invited to review the CVR transcript and suggest corrections or additions. The
suggested technical corrections were made to the transcript with the concurrence of the CVR
group. Information from the CVR is included in Section 1.1 of this report. The complete CVR
transcript appears in Appendix 2.
1.11.2 Flight Data Recorder (FDR)
1.11.2.1 Flight Data Recorder
Recorder Manufacturer/Model: L-3 Communications FA 2100

Recorder Serial Number: 00788
This recorded 1,176 individual parameters, and could record a minimum of 25 hours of flight
data.

The Flight Data Recorder (FDR) was retrieved from the aircraft at the accident site and
forwarded to the NTSB laboratory, where data was download and transcription was completed.
The FDR was received in an undamaged state. The FDR was serviceable at the time of the
accident, performed as required and contained data relevant to the accident. The relevant data was
of good quality, and was analyzed by the NTSB, Boeing flight safety personnel and the accident
investigation FDR and Performance groups.
The Flight Data was plotted by the NTSB and by the Boeing Aircraft Company.
The FDR was correlated to local time using the air traffic control transcript provided by the
JCAA. Appendix 1 contains a copy of the Boeing FDR plot.
1.11.2.2 Information from Flight Data Recorder
Examination of the FDR revealed the following:
Note 1: “ - ” , as in “ - 1,000 feet” means 1,000 feet before the threshold of runway 12. “ + ”, as
in “ + 1,000 feet” means 1,000 feet after the runway threshold. Distances and times are
approximations.
Note 2: The electronic approach guidance on runway 12 was from an ILS transmitter with a 3
degree slope which intersected with the runway 1,000 feet from the threshold. The visual
approach guidance on runway 12 was from a PAPI which had a 3 degree slope and intersected
with the runway 1,400 feet from the threshold.
Note 3: The distance between the glideslopes of the ILS and the PAPI was designed so that a
Boeing 747, the largest aircraft using the runway at the time of the installation of the PAPI and
ILS, could be flown with the aircraft on both glideslopes. As the Boeing 737-800 had a shorter
pilot eye/ILS antenna distance than the Boeing 747, the Boeing 737-800 could not be on both
glideslopes at the same time.
Note 4: Information from the Boeing 737 NG Flight Crew Training Manual, 5.28 and 6.5,
indicated that if the pilot of a Boeing 737-800 aircraft flew a 3 degree PAPI glideslope that met
the ground at 1,400 feet from the threshold, the main gear would meet with the runway (no
flare) at about 1,044 feet from the runway threshold. Also, that if the pilot of a Boeing 737-800
aircraft flew a 3 degree ILS glideslope that met the ground at 1,000 feet from the threshold, the
main gear would meet with the runway (no flare) at about 471 feet from the runway threshold.
Note 5: Localizer track was offset 3 degrees north of the runway track (117 degrees Vs. 120
degrees). 40
40
Appendix 4 (4)

 At three miles from the threshold the aircraft was on an ILS coupled approach, gear
down, flaps 30 degrees, with autopilot and auto-throttle engaged, and with glideslope
and localizer captured.
 Autopilot was disengaged 41 seconds before threshold crossing, at about 1.8 miles from
the threshold.
 ILS glideslope deviation started 10 seconds before threshold crossing, at -3,000 feet, to 1
dot high at -1,000 feet, about 4 seconds before reaching the threshold.
 Pitch was increasing from +2 degrees (up) to +3 degrees as the aircraft

crossed the runway threshold at about 70 feet RA.
 Pitch increased from +2 degrees to +4 degrees in 9 seconds, from -400 feet to

+2,300 feet.
 The aircraft crossed the runway threshold at 70 feet RA (that is, main gear height),
which placed the pilot's eye height at about 85 feet RA, about 14 feet above the PAPI
slope, and the aircraft's ILS antenna about 37 feet above the ILS glide slope. Thus both
the visual (PAPI) and instrument (ILS) indications to the flight crew when over the
threshold were that the aircraft was high on the approach.
 Vertical speed decreased from about 660 feet per minute (fpm) down at the threshold, to
0 fpm at +2,600 feet, in 10 seconds.
 Auto throttle was disengaged 4 seconds after threshold crossing, at +1,200 feet, at about
40 feet RA.
 At +1,400 feet, about where the PAPI glide path met the ground, the aircraft was
well above the PAPI glide path, descending through 38 feet RA, at a vertical speed of
about 3 feet per second down, and at an airspeed of 148 KIAS.
 The Throttle Resolver Angles were reduced from about 56 degrees, where they had been
during the latter stages of the approach, at +2,000 feet, and were at the flight idle stop at
+3,700 feet.
 N1 (engine speed, both engines) was at 60% until +2,000 feet, 9 seconds after threshold
crossing, then fell to about 30% (idle) at +4,300 feet.
 15 seconds elapsed between threshold crossing and first touchdown at +4,100 feet.
 At touchdown, airspeed was 148 KIAS, and ground speed was 162 knots.

 The first touchdown was at +4,100 feet, followed by a bounce, and the second touchdown
was at +4,300 feet.
 Brake pressure was at zero until +4,600 feet.
 Brake pressure was at 500 psi from +4,800 to +6,300 feet. (autobrake 3 design
deceleration was achieved)
 The Throttle Resolver Angles were at Max Reverse Thrust at +4,800 feet.
 N1, for reverse thrust, started to increase at +4,500 feet.
 N1, for reverse thrust, was at 82% (maximum) at +5,500 feet.
 Brake pressure reached 3,000 psi (maximum manual brake pressure) at +6,900 feet, and
remained there until runway departure.
 Boeing FDR graphs showed aircraft ground speed at end of runway was 62 knots, in
accordance with runway condition between Wet Smooth and Standing Water.
 Boeing FDR graphs showed actual rate of deceleration from braking on the wet runway
to be less than defined by FAR 25.109 for a Wet-Smooth runway, but more than
defined by the NASA Standing Water Runway model.
All of the information from the FDR, the CVR, ATC transcripts, radar tapes and the flight crew
interviews was compatible, and showed that the aircraft flew a stabilized final approach, on
course, on glide slope and on target air speed until 600 feet agl. It recorded that the captain
disengaged the autopilot at about -11,200 feet (1.8 Nautical miles) prior to the runway threshold,
and turned the aircraft slightly to the right at about 600 feet above ground level, then turned left
and flew the visual approach to runway 12. This resulted in the aircraft being slightly above the
ILS electronic glideslope at the end of the turn, at about 1,500 feet from the threshold. The
aircraft then continued the approach in a stable configuration and crossed the threshold of
runway 12 at about 70 feet RA. At this point the aircraft was very slightly to the right of the
extended runway centerline, on speed, more than one dot above the ILS glideslope and with the
autothrottle still engaged.
The FDR indicated that the aircraft had started to deviate above the ILS glideslope about 3,000
feet before the threshold, and exceeded one dot above the glideslope at 1,000 feet before the
threshold. The FDR showed at this point a slight upward elevator deflection, and the aircraft’s
pitch attitude increasing from 2 degrees nose up to 4 degrees nose up, and the downward vertical
speed decreasing, at threshold crossing. This manual pitch up control input was followed by
several more until touchdown at 4,100 feet down the runway about 15 seconds later. Some
aileron and rudder control inputs were also recorded.

The FDR recorded the manual disconnection of the auto-throttle at 35 feet RA, and the N1
being reduced coincident with the throttles being retarded manually to the forward idle stop at
27 feet RA. The aircraft was maintained at the target airspeed of Vref plus 5 = 148 KIAS,
giving a ground speed of 162 knots, until touchdown. The aircraft bounced on the first
touchdown, touched down 200 feet down the runway, the speed brakes deployed, and on wheel
spin up the autobrake engaged to the autobrake 3 setting, with a brake system pressure of
approximately 500 psi. The FDR revealed that the aircraft achieved a deceleration rate of
0.22g, which is consistent with the designed deceleration rate for an autobrake 3 setting.
Throttles were initially set manually to idle and then the N1s of both engines were recorded as
advancing quickly to full reverse at +5,500 feet. Full manual braking, as reported by the flight
crew, was recorded by the FDR indicating maximum brake system pressure of 3,000 psi at
+6,900 feet. Full reverse and full manual braking were maintained until impact.
The aircraft remained close to centre-line, however the rate of deceleration remained fairly
constant, that is, did not increase, and was not what the flight crew expected from full manual
braking. The aircraft went off the end of the runway at a ground speed of 62 knots, with the
FDR recording ceasing at impact.
The FDR indicated that at 3 miles from landing the wind was from 358 degrees at 8 knots, and at
touchdown increased to 323 degrees at 16 knots. At touchdown the calculated tailwind
component was 14 knots, and the calculated crosswind component was 7 knots from the left.
The "typical landing flare time” is described in the Boeing 737NG Flight Crew Training
Manual, 41 as four to eight seconds, and is a function of approach speed. The FDR of AA331
indicated that the flare time was about ten seconds. Flap 30 was used for the landing, and the
aircraft was landed on speed plus five knots, however, the landing was at about 4,100 feet from
the runway threshold, 1,130 feet past the touchdown zone limits.
The touchdown zone is defined in A A Flight Manual (FM) 42 as the “first 3,000 feet of the
runway beginning at the threshold”, or the first third of the runway, in this case 8,911/3 = 2,970
feet. The AA Flight Manual (FM), 43 defines the desired touchdown point as the first 800 to 1,500
feet beyond the threshold.
The AA B737 Wind Component and Landing Data Card 44 stated that the Required Runway
Landing Length for Dry or Wet/Good Runway Conditions was based, and included,
41
Boeing 737NG Flight Crew Training Manual, Page 6.10 See 1.17.2.2
42
A A Flight Manual (FM), Part I, Section 10, paragraph 7.2, page 32
43
AA Flight Manual (FM), Part I, Section 10, paragraph 7.3, page 32
44
Appendix 7

“demonstrated ‘air distance’ from runway threshold to touchdown”, and for Medium/Fair or Poor
runway Conditions “includes 1,000 feet ‘air distance’ from threshold to touchdown”. AA was
asked to define “demonstrated ‘air distance’”, but they were not able to provide this information.
The length of runway 12 was 8,911 feet.

KIAS KNOTS INDICATED AIRSPEED
KGS KNOTS GROUND SPEED
VS VERTICAL SPEED
FT FEET
FPS FEET PER SECOND
Figure 4: Landing Sequence Key Events
AGL ABOVE GROUND LEVEL
PSI POUNDS PER SQUARE INCH
Not To Scale A/T AUTOTHROTTLE
PAPI
R12
ROAD TO
PORT
ROYAL

AIRCRAFT
WRECKAGE
OVER THRESHOLD R12 At 1,200 FT At 2,000 FT At 4,100 FT At 4,800 FT At 5,500 FT At 6,900 FT At 8,911 FT
148 KIAS, 162 KGS A/T OFF THROTTLE RESOLVER FIRST TOUCHDOWN BRAKE PRESSURE TO MAX REVERSE BRAKE PRESSURE TO AIRCRAFT DEPARTS
VS -11 FPS 162 KGS ANGLE REDUCED
Jamaica Civil Aviation Authority – 500 PSI
American (AUTOBRAKE
Airlines 331 3)Accident
THRUST
Report 3,000 PSI (MAX END OF RUNWAY
70 FT AGL VS -3 FPS N1 DECREASING FROM 60% At 4,300 FT 160 KGS 140 KGS MANUAL) 62 KGS
AUTOPILOT OFF A/T ON 40 FT AGL 162 KGS SECOND TOUCHDOWN 130 KGS
N1 64% VS -4 FPS 148 KIAS, 162 KGS
GLIDESLOPE >+1 DOT 30 FT AGL N1 30% (IDLE)
1.12 Wreckage And Impact Information
1.12.1 Debris Path and Other Damage
After leaving the runway the airplane impacted the airport perimeter fence about 12 feet beyond
the threshold, and three runway fence posts were broken. It then travelled airborne about 74 feet
horizontally and dropped about 9.5 feet vertically before its initial impact with the sand berm.
The initial impact point was characterized by disturbed sand, rocks, and vegetation consistent
with the locations of the engines and landing gear. After impacting the berm, the airplane travelled
about 100 feet along the up-sloping berm before coming to rest on the beach, a short distance
from the water line, and narrowly missed colliding with the concrete pedestals for the approach
lights for runway 30.
During the accident there was some slight pavement lip damage at the end of the runway, two
threshold end lights were damaged, the airport boundary fence was broken, and some
contamination resulted on the beach sand area due to the rupture of the fuel tank in the right
wing.
The only markings on the runway conclusively identified as being from the accident airplane
were some scrub marks on the white threshold markings and some light tire marks adjacent to
the threshold. The scrub marks were relatively cleaner areas of paint with dimensions and tread
patterns consistent with the tires of the nose landing gear and those of both main landing gears.
There was a noticeable upslope at the end of runway 12 that began about 670 feet before the end
of the runway.
The scrub marks were centred about 27 feet, 10 inches to the left of the runway centreline at the
west end of the runway 30 threshold marks. The light tire marks exited the runway threshold
about 39 feet, 3 inches left of the runway centreline.
The 4th runway end light from the left side was broken from its mount and the 5th runway end
light was turned clockwise. The 6th runway end light was missing from its mount and the runway
mount was severely corroded.
The forward section of the aircraft came to rest on a track of 088 degrees magnetic and the centre
and aft sections were on a track of 128 degrees magnetic.

Photo 2 Aircraft Wreckage
1.12.2 Damage to Aircraft - General
The aircraft was substantially damaged during the overrun with the right main gear and right
engine torn off, the left main gear and the nose gear collapsed, the fuselage split into three
sections and ruptured and deformed underneath, and the wings and flaps were damaged. The
fuel tanks in the right wing were ruptured, and the fuel had leaked out.
Of the three sections into which the aircraft broke, the forward section consisted of the flight
deck, the forward cabin crew seats, and passenger seats from rows three through six. The
centre section contained passenger seat rows 7 through 22, including the four over wing
emergency exits. The aft section consisted of the aft fuselage of the aircraft from passenger seat
row 23 through to the tail of the aircraft.
There was no evidence of pre-impact damage to, or pre-impact failure of, any part of the
aircraft’s fuselage, wings or control surfaces.

Photo 3 Forward Fuselage
Photo 4 Centre Fuselage

. Photo 5 Aft Fuselage
Photo 6 Left Wing

Photo 7 Right Wing
1.12.3. Damage to Engines
The two CFM 56-7B27 engines were examined on site. Both had suffered extensive impact
damage. Both had the thrust reverser actuators extended, and the translating cowls of both were in
the aft position, indicating that the thrust reversers were extended at impact. This information
agreed with the FDR data. There was no evidence of pre-impact malfunction of, damage to, or
pre-impact failure of, any part of either engine or associated systems.
1.12.4 Damage to Landing Gear
1.12.4.1 Nose Gear
The nose gear was folded back by impact forces, but the FDR indicated that it was down and
locked before impact. The wear on the tires was typical of in-service tires. The left tire pressure
was 200 psi, and the right was 0 psi.

Photo 8 Nose Landing Gear
1.12.4.2 Main Landing Gear
1.12.4.2.1 General
As the sequence of events brought into question the functioning of the aircraft’s braking
system, a detailed examination of the left and right main landing gears was conducted.
1.12.4.2.2 Right Main Landing Gear
The right main landing gear was separated from the wing and came to rest about 30 feet aft of
the right wing. The damage was consistent with the impact pattern. The FDR indicated that the
right main landing gear was down and locked before impact.
The brakes were intact and appeared to be undamaged. No leaks were observed at the brake
pistons. The inboard brake wear pin protrusion beyond the guide was approximately 13/8 inches.
The outboard brake wear pin protrusion beyond the guide was approximately 9/16 inch. The
wheels were intact and appeared to be undamaged.
The inboard tire had retread level R1 dated 10/2009. The remaining tread depth was measured
from the base of the groove to the rib. From the outside to the inside of the tire the tread depth of
each groove measured 3/32 inch, 6/32 inch, 7/32 inch and 1/32 inch, respectively. The measured
tire pressure was 215 psi. Both side shoulder ribs had heavy abrasion and large amounts of
melted rubber balls 360° around the tire. The centre tread rib exhibited typical abrasion for a
rolling tire. Both shoulders exhibited abrasion wear with some heavy abrasion similar to that
identified on the shoulder ribs, but with very little evidence of melted rubber balls.
The outboard tire had retread level R4 dated 9/2009. The following measurements of the
remaining tread depths were recorded (measured outboard to inboard): 9/32 inch, 12/32 inch,
13/32 inch and 7/32 inch, respectively. The measured tire pressure was 210 psi. The serial side
shoulder rib had heavy abrasion and melted rubber balls 360° around the tire. The non-serial side
shoulder rib also exhibited heavy abrasion and melted rubber balls 360° around the tire, however,
there was not as much abrasion as on the serial side shoulder rib. The centre tread rib
demonstrated typical abrasion wear for a rolling tire. There was some heavy abrasion on the
shoulders similar to that identified on the shoulder ribs but very little evidence of melted rubber
balls.
There was no evidence of pre-impact damage to, or pre-impact failure of, any part of the right
main landing gear system, including the tires and the brakes.
Photo 9 Right Main Landing Gear
1.12.4.2.3 Left Main Landing Gear
The left main landing gear was folded aft under the left wing and impacted the left inboard flap.
The damage was consistent with the impact pattern. The FDR indicated that the left main
landing gear was down and locked before impact.
The brakes were intact and appeared to be undamaged. No leaks were observed at the brake
pistons. The inboard brake wear pin protrusion beyond the guide was about 7/8 inch.
The outboard wear pin protrusion beyond the guide was about ¾ inch. The wheels were intact and
appeared to be undamaged.

The in-board tire had retread level R1 dated 9/2009. The following measurements of the
remaining tread depths were recorded (measured outboard to inboard): 4/32 inch, 6/32 inch, 5/32
inch and 2/32 inch. The measured tire pressure was 199 psi. Both side shoulder ribs exhibited
some abrasive wear damage 360° around the tire. There was no evidence of melted rubber balls.
The centre tread rib exhibited typical abrasion wear for a rolling tire.
The outboard tire had retread level R2 dated 9/2009. The following measurements of the
remaining tread depths were recorded (measured outboard to inboard): 6/32 inch, 12/32 inch,
12/32 inch and 2/32 inch. The measured tire pressure was 204 psi. Both side shoulder ribs had
heavy abrasion 360° around the tire, but there was no evidence of melted rubber. The centre
tread rib exhibited typical abrasion wear for a rolling tire.
There was no evidence of pre-impact damage to, or pre-impact failure of, any part of the left
main landing gear system, including the tires and the brakes.
Photo 10 Left Hand Main Landing Gear

1.12.4.2.4 Evidence of Hydroplaning
The tires were examined at the accident site and it was observed that the tires on the right main
landing gear had melted rubber balls and the tires on the left main landing gear did not.
The rubber balls on the right main tires could possibly be an indication that the tires were
turning and the brakes were working to slow them, but there was reduced friction such that the
rubber balls were not removed during the braking process. With good friction, the rubber balls
would normally be scrubbed off by the runway during braking. There was no ready explanation
why the tires on the right gear should have melted rubber balls, and the tires on the left gear did
not.
There was no evidence that the wheels locked up during the landing roll, or of any hydroplaning
on any of the tires, as there were no signs of rubber reversion on the tires or marks on the
runway.
1.12.5 Damage to Exits
1.12.5.1 Forward Left Door (L1) and Escape Slide
The forward left door, designated L1, was found in a partially open position, with the door
handle in the “open” position. The caution strap was found snapped in position above the
viewing window.
The escape slide pack was jammed into the doorway on the aft corner of the escape slide and was
not able to fall free from the door sill 45. Once the escape slide pack had been removed, the door
was tested for operation. It was found to operate to the open position, but could not be fully closed
in the locked position
45
Photo 11

Photo 11 Forward Left Door (L1)
1.12.5.2 Forward Right Door (R1) and Escape Slide
The forward right door, designated R1, was found opened and the door handle in the “open”
position. The caution strap was found snapped across the viewing window. The door was
tested for operation and found to be difficult to close due to the buckled galley floor, but could
be fully closed in the locked position.
The escape slide pack had been released from the escape slide compartment. The escape slide
was found deflated, but there was evidence that the slide had been inflated post impact. 46
46
See Photo 12

Photo 12 Forward Right Door (R1)
1.12.5.3 Aft Left Door (L2)
The L2 door was found in the fully open position. The door handle was positioned between the
open and locked positions. The caution strap was found snapped across the viewing area. The
operation of the door was checked and found to open normally.
The L2 escape slide pack was found on the ground, released from the slide pack compartment,
but the slide was not inflated.
After the accident the aft fuselage of the airplane was too close to the ground for the slide to
fall far enough to activate the inflation mechanism.

Photo 13 Aft Left Door (L2)
1.12.5.4 Aft Right Door (R2)
The R2 door was found in the fully open position with the handle positioned between the open
and locked positions. The caution strap was found snapped across the viewing area. The
operation of the door was checked and was found to open normally.
The R2 escape slide pack was found on the ground, released from the slide pack compartment,
but the slide was not inflated. After the accident the aft fuselage of the airplane was too close to
the ground for the slide to fall far enough to activate the inflation mechanism.

Photo 14 Aft Right Door (R2)
1.12.5.5 Automatic Over-wing Exits
All four automatic over-wing exits were found open and all four operated normally when tested.
The escape straps for the aft left and aft right hand over-wing exits were found in their stowed
positions.
1.12.6 Damage to Flight Deck
1.12.6.1 General
The forward section, which included the flight deck, came to rest in a longitudinal direction
about 40 degrees left of that of the centre section.

1.12.6.2 Flight Deck Seats
Both the captain and first officer’s seats were subjected to high horizontal forces during the
impact. A mathematical calculation was conducted by IPECO, manufacturer of the flight deck
seats, and it was estimated that the G-loads during the impact sequence were between 10 and
12.7 G for a forward deceleration.
The captain’s seat crotch belt was found broken loose from the seat structure because the captain’s
crotch belt bracket was broken. 47 Both the crotch belt and the outboard shoulder harness were
found buckled. The captain’s outboard shoulder harness was locked in the extended position and
the inboard shoulder harness retracted normally.
Photo 15 View into Flight Deck
47
Photo 16

Both the captain and first officer’s seats translated freely in the tracks, but had restricted travel in
the forward most position due to contact with the damaged side wall panels. The first observer’s
seat was found jammed in the stowed position. The second observer’s seat floor structure
fittings were found sheared.
The Cabin Safety Investigator’s Factual Report indicated that the bottom seat cushion of the
captain’s seat was not an approved part for that seat. It had the wrong part number, and was
not a certified seat bottom cushion. This seat cushion invalidated the dynamic qualification of
the seat to TSO C127a and in particular the occupant injury criteria for spinal compression load.
1.12.6.3 Flight Deck Floor
The flight deck floor forward of the flight deck door was buckled upwards. The floor under both
pilot seats was continuous up to the rudder pedals. Forward of the rudder pedals, the floor was
articulated forward and downward away from the captain and first officer positions.
1.12.6.4 Flight Deck Sidewalls
Both of the sidewalls adjacent to the captain and first officer seats were pushed inward towards
the seats. The sidewalls came in as far as the outboard rudder pedal on each side.
1.12.6.5 Flight Deck Windows
All flight deck windows remained intact. The two emergency egress flight deck windows were
found closed. Investigators unlatched the actuation handles on site and attempted to open the
windows, however, damage to the window tracks prevented the windows from opening.
1.12.6.6 Flight Deck Instrument Panel
The instrument panels separated from the fuselage just below the windscreen and articulated
forward, away from the pilots. The whole instrument panel was rotated forward (outboard) and
downward.

Photo 16 The captain’s crotch belt bracket
1.12.6.7 Flight Deck Door
The floor both inside and outside the flight deck had buckled and the flight deck door could only
be opened 2 inches at the top and 1.5 inches at the middle and floor levels. The upper and
lower blowout panels were not in the door. The lower panel was found, and was undamaged.
The upper panel was not found.
1.12.6.8 Forward Entry, Cabin Crew Seats and Forward Galley Area
The forward entry area floor buckled laterally about midway through the R1 doorway. The
buckling continued clear across the cabin flooring from left to right. Most of the buckling was in
the R1 entry area and in the aisle to the flight deck door.
In the galley area, forward of R1, it was evident that the structure was buckled, causing some
items to be jammed in position. Both of the cabin crew members seated at L1 reported that
they were unable to retrieve their flashlights because the doors to the emergency equipment
compartments under their jump seats were jammed due to the buckling of the floor.

Photo 17 Flight Deck Door
1.12.7 Main Cabin
1.12.7.1 Forward Section - First Class Cabin – Row 3 through Row 6
The interior of the forward section sustained the heaviest damage compared to the centre and aft
sections. This section of the aircraft was angled about 40 degrees to the left of the centre section,
leaving an opening on the right side about 6 feet wide, and no opening on the left side. The
ceiling panels, overhead stowage bins, passenger service units (PSUs), and spacer panels were
found displaced. The forward video monitor was stowed. The aft ceiling mounted video
monitor remained attached to the stringer and the monitor shroud was displaced downward. The
life raft was found in its stowage compartment in the ceiling above Rows 3 and 4. The
attachments on the left side of the life raft compartment remained attached to the stringers, but
the right side was detached from the stringers.
The majority of the seats in this area were in good condition, with exceptions throughout the
section.

Photo 18 First Class Cabin
1.12.7.2 Centre Section – Economy Class Cabin – Row 7 through Row 22
The damage to this section was the least of the three sections. The majority of the seats in this
area were in good condition, with exceptions throughout the section. The overhead stowage bins
remained in place, with the exception of the bins located at the fuselage breaks. All of the PSUs
were dislodged from under the overhead stowage bins and were found either hanging in the seat
rows, lying on passenger seats or on the floor.
1.12.7.3 Aft Section - Economy Class Cabin – Row 23 through Row 28
In the aft section, row 23 overhead stowage bins were dislodged on both the ABC and DEF sides
of the aisle, and resting on the seats. All PSUs had become dislodged from under the overhead
bins and were found either hanging in the seat rows, lying on passenger seats or on the floor.

Photo 19 Centre Section
The aft-most portion of floor between the lavatories had buckled exposing a gap between cabin
floor panels and the aft galley floor. The majority of the seats in this area were in good condition,
with exceptions throughout the section.

1.12.7.4 Aft Entry, Cabin Crew Seats and Aft Galley Area
The aft galley complex compartments were all closed and locked. All cart compartments were
stowed, latched, locked, operable, and intact. The work light was set to “DIM”.
Photo 20 Aft Section
The handset was found stowed. The entry light switch was in the “DIM” position, ceiling lights
switch in “DIM” position, work light switch in “ON” position, emergency light switch in “OFF”
position, and window lights switch in the “OFF” position.
The cabin crew member seated at L2 and the one seated at R2 reported that they were unable to
retrieve their flashlights because the doors to the emergency equipment compartments under
their jump seats were jammed due to the buckling of the floor.

1.13 Medical and Pathological Information
1.13.1 Captain
The captain was examined by a medical doctor at the University of the West Indies, Kingston a
few hours after the accident. Drug tests for Cocaine, Cannabis and Ethanol were conducted, and
were negative. Chest and abdominal examination revealed transverse abrasions to the anterior
aspect abdomen consistent with his seat restraint. There was mild abdominal tenderness in this
region, and some mild tenderness over cervical spine. The resulting report stated that he had
sustained some trauma to head, chest and abdomen, and abdominal abrasions consistent with seat
restraint, but was otherwise essentially normal. There was no medical finding that could have
contributed to causing this accident.
1.13.2 First Officer
The first officer was examined by a medical doctor at the University of the West Indies,
Kingston a few hours after the accident. Drug tests for Cocaine, Cannabis and Ethanol were
conducted, and were negative. The resulting report stated that he had sustained some trauma to
chest and upper abdomen consistent with seat restraint; also right knee and left hand had pain
and swelling. Otherwise he was essentially normal. There was no medical finding that could
have contributed to causing this accident.
1.14 Fire
Although the conditions were ideal for combustion, that is, electrical sparks, spilt fuel, fuel
fumes, hot mechanical parts etc., there was no evidence of any fire. The presence of
heavy rain probably impeded the chance of fire. The first ARFF unit arrived within 3
minutes, and remained on standby in case of fire.

1.15 Survival Aspects
1.15.1 Notification of Accident
ATC was immediately aware of the accident, and informed the NMIA emergency services.
ARFF had one unit already deployed on the main ramp providing standby fire protection during
the refueling of a Virgin Atlantic aircraft about to depart, and that unit went immediately to the
end of runway 12, across from the crash site, reaching it in a position from which it could project
foam fire suppressant over the entire site, at 22:25 EST, according to interview information from
ARFF personnel. The ARFF crew advised other ARFF personnel by radio that they would have
to access the crash site by driving off-airport to the Port Royal road, which they did, at the east end
of runway 12. They arrived at the aircraft at about 22:31 EST
1.15.2 Emergency Lighting in aircraft
Please refer to 1.16.2 of this report.
1.15.3 Communications
Because of the sudden and unexpected nature of the accident, there was no opportunity for the
flight crew to alert the cabin crew and passengers to the impending impact.
The cabin crew interviews indicated that the inter-phone/public address system was not
functioning after the accident. It was not determined as to whether or not the flight crew’s public
address system was functioning after the accident, but there were no reports that any message
from it was audible in the passenger cabin. It was reported that the forward flight attendant used
the megaphone to yell her commands. In all other respects, communications, both normal and
emergency, worked as planned and without interruption.
1.15.4 Evacuation
The following information is from the statements of the flight crew and the cabin crew:
 As the aircraft went over the lip at the end of the runway, the captain hit his head on the
cockpit ceiling, then his face hit the HUD, causing him to be dazed.
 No brace command was given prior to the accident by either the flight crew or cabin
crew. As soon as the aircraft came to a stop, the captain called out “Easy Victor” to
notify the cabin crew to evacuate the aircraft.
 The captain was not sure that the PA system had worked. One of the cabin crew at the aft
of the aircraft thought she heard the captain calling “Easy Victor”,
 After the impact, when the aircraft came to a rest, it was quiet and dark, and the flight crew
tried to open the cockpit window exits but they were jammed closed.

 The captain saw sparks, and knew they had to get out of the aircraft quickly. He had
difficulty in getting out of his seat. It was obvious at that point that the aircraft was
damaged, and they could see the water on the rocks through the broken cockpit.
 The captain called for the emergency evacuation checklist, and the flight crew
completed the items on the checklist, not knowing if the aircraft had power, and even
though the engines were probably not on the aircraft.
 The first officer said he didn’t remember seeing the instrument panel in front of him, and
could see the emergency light on.
 The cabin crew ordered evacuation almost immediately after the aircraft stopped.
 The cabin crew could not open door L1, and had difficulty opening door R1.
 The flight crew lost track of time, but believed they were in the cockpit for about five to
ten minutes before they exited.
There was a large hole in the front of the aircraft and the first officer reported seeing the sea water
in front of the aircraft, smelling fuel and seeing some sparks and said he considered exiting via the
hole. The first officer tried to open the cockpit door, but it was jammed, however he was able to
unlatch and remove the upper escape panel in the flight deck door and he left the flight deck
through there, followed immediately by the captain.
Some emergency vehicles were already there when the flight crew exited, arriving promptly.
They tried to organize the passengers, but it was very chaotic. They noticed other people
coming up to the airplane as the passengers scattered. The cabin crew all acted in a timely and
appropriate manner, helping to reduce passenger injuries or possible loss of life.
Passengers evacuated from L2, R1, R2, the left and right over-wing exits and the right side of
the forward break in the aircraft. The L1 door was jammed and could not be opened fully. The
slide at the R1 door was the only slide to deploy. The captain stated that the slide was lying on
the rocks, and people jumped out into it and hit the rocks. The L2 and R2 slides were attached
to the airplane, but, due to the close proximity of the door sills to the ground, they did not
deploy and each was lying on the ground outside of its respective door.
The cabin crew had already got everyone off the airplane when the captain told the first officer to
go out and assemble the passengers. Some of the passengers alighted the aircraft within seconds,
while others took several minutes. Some passengers helped each other, while some stepped over
others in their efforts to get out of the aircraft.
The captain tried to go through first class, but saw that it was completely blocked by collapsed
overhead bins, and the damage from the break in the fuselage. He saw one of the cabin crew
carrying a large woman who didn’t want to leave, and he believed she was the last passenger to
leave. The captain told the first officer to verify that the passengers were off, and he went into the
back of the airplane. He said it was chaos, with rain and jet fuel on the ground. He said he had
never seen rain that heavy and described it as being “like a monsoon”.

The flight crew exited the aircraft at R1, and entered the rear section of the aircraft through
R2. They then rechecked the front of the aircraft and ensured there was no one remaining inside.
They then proceeded to the road behind the aircraft to try to consolidate the passengers.
The captain said he thought the response was prompt, but once all passengers were finally
gathered it was very chaotic. He said there were sirens and people around, and he thought there
was a delay in getting the area secure. There was no one from ARFF securing the area or working
on getting access to the airplane. He was the only one going into the airplane and he said he felt
like he needed some help searching the airplane.
There was no post-crash fire. The 148 passengers, and the six crew members evacuated the
aircraft with some passengers being assisted by the cabin crew or other passengers. Fourteen
passengers received serious injuries.
A passing Jamaica Urban Transit Corporation bus, travelling from Port Royal to Kingston on its
last run of the evening, stopped at the crash site. The driver had witnessed the crash, and she
picked up most of the passengers and drove them to the NMIA Arrivals area, then returned to the
crash site for another load of people.
The passengers were ferried to the main terminal in various vehicles and some of them
departed with those meeting them, in their own vehicles or by taxi. Transportation to the
hospital was arranged for the injured. The evidence indicated that the flight crew and the cabin
crew performed as trained in response to the emergency circumstances.
1.15.5 Emergency Exits
The damage to the emergency exits and evacuation slides is described in Section 1.12.5 of this
report.
1.16 Tests And Research
1.16.1 Escape Slide Tests
It was apparent that the L1 escape slide pack had fallen prematurely from the escape slide
compartment, and tests were conducted to determine the reason. During testing, it was found
that when the escape slide compartment was installed on the door the lower center support pin on
the back of the slide compartment was not properly inserted into the bottom support bracket in the
aircraft door. No other problems were identified with the L1 slide or slide compartments.
Although the improper installation of the slide compartment may have contributed to
the premature slide release, the exact cause of the slide release jamming could not be determined.

The Boeing Company has since produced a proposed revision to the Boeing 737 maintenance
manual to further clarify proper installation procedures. 48
1.16.2 Emergency Lighting in aircraft
The Emergency Lighting system was tested, and the following was found:
The right forward emergency lighting battery pack was found to be charged. The right forward
pack connector was found to be damaged. On-aircraft testing showed that the wiring to the R1
door lights (internal and external) and exit signs was intact. The wiring to the centre, overhead
bin and seat lights did not have continuity and was not connected to the lights.
The left forward emergency lighting battery pack was found to be charged. The left forward pack
connector did not have any visible damage. On-aircraft testing showed that wiring to the
overhead bin and L1 door lights did not have continuity and was not connected to any of the
lights, because the wiring was severed between all the lights and the battery pack.
The five centre section emergency lighting battery packs were all found to be at very low voltage
and, according to Boeing, would be considered to be depleted of charge. On-aircraft testing
showed that wiring to the left and right over-wing emergency exit lights and exit signs, including
the over-wing exterior lights, was intact. The left forward wing exit light was missing and the
right forward over-wing light and centre over-head exit light did not illuminate. The wiring to all
of the seat lights was intact and illuminated the lights when voltage was applied to the connector
wiring, except for the last row of main cabin lights. This was due to the severed wiring at the
right fuselage break at rows 22 and 23. The bin lights at the left over-wing exits located at rows
13 and 14, did illuminate when voltage was applied for the specific light. The rest of the
overhead bin lights did not have proper continuity at the pack connectors.
The right aft emergency lighting battery pack was found to be at very low voltage. However, on-
aircraft testing showed that wiring to the R2 door lights (internal and external) and exit signs was
intact and illuminated the lights when voltage was applied for the specific light. The wiring to
the centre overhead bin and seat lights did not have continuity and was not connected to the
lights. As a result of the break in the sidewall and cabin floor on the right side in this section the
seat light wiring was severed and was only visibly connected to the last two seat rows, but was
not connected to the battery pack connector.
The left aft emergency lighting battery pack was found to be at very low voltage and, according
to Boeing, would be considered to be depleted of charge. On aircraft testing showed that wiring
to the L2 door lights (internal and external) and exit signs was intact and the lights illuminated
when voltage was applied to the lights. The wiring to the overhead bin lights did not have
continuity and was not connected to the lights.
48
Section 4.1.5

The emergency light switch in the aft flight attendant control panel was in the off position.
Three of the four cabin crew members, according to their interviews, and 15 of the 17 passengers
who answered questionnaires, reported seeing no lights. However, as detailed above, testing
showed that some of the overhead bin and floor lighting was operational; and all but the two
forward exit and door lights and possibly the forward over wing exits, displayed signs of
having been illuminated. One passenger reported that he saw the exterior lights illuminated.
1.16.3 Simulator Trials
Full replication of the flight in a flight simulator using the actual flight data was not considered
to be necessary, since sufficient data existed for the investigation team to complete the required
analysis.
Members of the operations investigation team observed several instrument approaches flown in
an AA B737-800 full flight simulator at the AA Dallas training center. These flights were set up
to simulate the Kingston ILS runway 12 approach using wind and weather conditions set to
simulate those on the accident flight, as well as introducing wet runway and flooded runway
simulated deceleration effects. It was observed that whenever the runway condition was set to that
of a flooded runway the aircraft did not stop on the runway.
This was not considered to be relevant for the investigation as the simulator did not replicate the
actual runway, or the actual braking conditions and this was for demonstration purposes only. It
was also noticed that on almost all simulator transitions to visual cues from the offset localizer to
line up with the runway for landing, the aircraft climbed a little, going slightly high during the left
turn, placing it about ¼ dot above the ILS glide-path. On all approaches the autothrottle was left
on after the autopilot was disconnected.
1.16.4 Display Guidance Computer
The Display Guidance Computer was tested at the BAE Systems facility in Redmond WA in the
presence of JCAA and NTSB investigators.
The Non Volatile Memory (NVM) was downloaded from the unit prior to testing. There were
multiple primary glideslope deviation flagged invalid and secondary glideslope deviation flagged
invalid events recorded just prior to landing for the accident flight. All of these events were
recorded at an altitude of 60 ft, or below while the HUD was in mode “A1”. There was also one
Pitch Angle #1 and #2 Miscompare event recorded for the accident flight. This event occurred
after the aircraft had touched down, at an airspeed indication of 47 knots.
The unit was tested in the presence of NTSB and JCAA investigators, in accordance with BAE
Systems Test Procedures Document 560/62745 Rev. P dated February 2002.

The results were as follows:
1. Bonding Resistance Check – Seven of the nine pins were within limits; however the
resistance checks for pins 2 & 3 of connector J1b were not within limits (300 milli-Ohms,
test limit 30 milli-Ohms).
2. Isolation Resistance Check – All checks for this process were within the prescribed
parameters.
3. Initialization – The unit successfully completed this function.
4. HUD Test Mode Check – All tests and displays for this process were as designed.
5. Configuration Check – All hardware and software configurations were per specifications
for the dash number of the unit.
6. Overhead Data Check – The data and status displayed were as designed.
7. Discrete Input Tests – All tests and displays for this process were as designed.
8. Cockpit Discrete Input Tests – All status shown were as designed.
9. Static HUD Display – Ten of the thirteen displays for this process were as designed.
A. The second HUD Display was incorrect; it revealed ‘No AIII Miscompare’.
The ‘AIII’ did not illuminate but ‘No AIII’ did illuminate on the Annunciator
panel.
B. The decluttered HUD display was incorrect; it revealed a ‘No AIII

Miscompare’.
C. The Aeronautical Radio, Inc. (Specification) Input status did not receive an
input from FMC Bus#1 as indicated on the Multipurpose Control and Display
Unit (MCDU).
The NVM was downloaded from the unit after testing. The same events from the accident flight
were displayed on this download as on the previous download. As supported by these test results
and analysis of the NVM, and the fact that no malfunction was reported by the captain, the
HUD was believed to be functioning correctly at the time of the accident.
1.16.5 Tests on Braking System
1.16.5.1 Antiskid/Autobrake Control Unit
The Antiskid/Autobrake Control Unit was tested at the Crane Aerospace and Electronics facility
in Burbank, CA, observed by investigators from the JCAA and the NTSB. The manufacturer’s
Test Procedure TP-42-935-2 Rev. C dated October 20, 1999 was run on the unit. The unit passed
the automated and manual tests with no failures noted. The unit was considered to have been
serviceable at the time of the accident.
1.16.5.2 Antiskid Valves
The four antiskid valves were tested at the Crane Aerospace and Electronics facility in Burbank,
CA in the presence of JCAA and NTSB investigators, in accordance with Crane Test Procedures
TP 39-353 Rev. G dated August 12, 1991.
All units passed the electrical continuity test.
All the units passed the internal leak test except for one, which had a negligible leakage rate.
All units passed the step response tests and the signal tracking response tests.
All units passed the hysteresis tests.
The proof test portion of the acceptance test procedure was performed last and all units passed.
The gear retraction test was found to be acceptable for all units.
A graph of brake pressure versus current was created for each unit and compared to gated
parameters. All units were slightly outside of the required envelope for a new or overhauled
unit. The antiskid units were considered to have been serviceable at the time of the accident.
1.16.5.3 Wheel Speed Transducers
The Wheel Speed Transducers were tested at the Crane Aerospace and Electronics facility in
Burbank, CA in the presence of JCAA and NTSB investigators.
The adaptor and canon plugs were removed from the transducers and all units were tested in
accordance with Crane Test Procedure TP 140-025 Rev. J dated May 15, 1995, used to test new
and overhauled units.
All Wheel Speed Transducer units successfully passed all aspects of the test, and were
considered to have been serviceable at the time of the accident.
1.16.5.4 Autobrake Valve Module
The Autobrake Valve Module was tested at the Woodward HRT facility in Santa Clarita, CA in
the presence of JCAA and NTSB investigators, in accordance with the manufacturer’s test
procedure Document No. HR 72700405 Rev. F, dated June 18, 1997.
The unit tested within the Autobrake Production Steady State Pressure Gain Envelope and the
Autobrake Production Steady State Pressure Gain Expanded Scale Envelope. The threshold, step
off and hysteresis were within limits.
The Brake Pressure Variation Test and the Pressure Threshold Tests were within limits.

The Proof Pressure Test with + 8mA applied and the low pressure test were successful, and there
was no leakage.
The Flow Test @ 1000 psi & 3000 psi was successful.
The Return Flow Test @ 1800 psi was within limits.
The Leakage Test @ 1000 psi and the de-energized Leakage tests were within limits.
The status light for the pressure switch (solenoid) did not illuminate within the specified pressure
band of 900 to 1100 psi on the pressure supply increase or the minimum pressure value of
over 750 psi on the pressure supply reduction as expected. An impulse test of this pressure
switch was accomplished and the unit performed as expected. This test simulated the on-
airplane response when autobrakes were activated. The Autobrake Valve Module was considered
to have been serviceable at the time of the accident.
1.17 Organization and Management Information
At the time of the accident, American Airlines, Inc., operating as American Airlines (AA), was a
major U.S. airline and a subsidiary of AMR Corporation. It was headquartered in Fort Worth,
Texas, adjacent to its largest hub at Dallas/Fort Worth International Airport. American Airlines
operated an extensive international and domestic network, with scheduled flights throughout
North America, the Caribbean, South America, Europe, and Asia/Pacific.
1.17.1 AA Manuals, Policies and Procedures
1.17.1.1 AA Procedures for Approach and Landing, and go-around
1.17.1.1.1 Stabilized Approach Requirement
The requirements for a stabilized approach are described in AA B737 Operating

Manual, 49 as follows:
Significant speed and configuration changes during an approach can complicate aircraft
control, increase the difficulty of evaluating an approach as it progresses, and
complicate the decision at the decision point; i.e., DA, DH, MDA. A pilot must assess the
probable success of an approach before reaching the decision point. This requires the pilot to
determine that requirements for a stabilized approach have been met and maintained.
To limit configuration changes at low altitude, the airplane must be in landing configuration
by 1,000 feet agl (gear down and landing flaps). A stabilized approach must be
established before descending below the following minimum stabilized approach heights:
• IMC – 1000 feet agl
• VMC – 500 feet agl
49
AAB737 Operating Manual, Vol 1, Section “Approach – Landing – Go-Around”, Page 15.8
A stabilized approach means the airplane must be:

• At Approach Speed (VREF + additives),
• On the proper flight path, and at the proper sink rate,
• At stabilized thrust.
These requirements must be maintained throughout the rest of the approach for it to be
considered a stabilized approach.
If the stabilized approach requirements cannot be satisfied by the minimum stabilized
approach heights or maintained throughout the rest of the approach, a go-around is
required.
1.17.1.1.2 CAT I ILS Approach Procedures
Included in the normal procedures for an ILS approach, as described for flight crews in the
AA 737 Operating Manual: 50
Wind
• Tailwind: Max 10 knots (may be further modified by Performance section requirements)
At Decision Height
■ If requirements for descent below DA are satisfied:
Disengage autopilot no later than 50 feet agl and maintain stabilized approach to
touchdown.
■ If requirements for descent below DA are not satisfied:
Execute missed approach
Conditions Requiring a Go-Around:

• Anytime:
Localizer deviation of two and one half dots (magenta localizer diamond goes hollow at two
and one half dots).
• Prior to DA:
Any of the required airplane or ground equipment fails.
• At DA:
Requirements to continue approach below DA (FAR 121.651) have not been met.
• Below DA:
A reduction in visual references occurs which prevents the PF from safely continuing the
approach and landing.
• The Captain determines that a landing cannot be safely accomplished within the
touchdown zone.
50
AA 737 Operating Manual, Vol. 1, Section “Approach – Landing - Go- Around”, Page 25.3

1.17.1.1.3 Landing on a Slippery Runway/Hydroplaning
AA B737 Operating Manual, provides the following information to its flight crews for landing on
slippery runways:
Landing is a lot more complicated on a runway with very little or no friction with the
tires. This reduction in friction (braking action) is usually caused by either ice or water
(hydroplaning).
On short or slippery runways, braking may be least effective at the departure end
because of rubber deposits, snow or ice. In these conditions, since the middle of the runway
offers the best friction for wheel braking, brakes should be aggressively applied by the use
of MAX autobrakes or manual braking immediately after.
Icy runway: Icy runways are treated with sand, salt and/or chemical deicers to provide
adequate traction. Use your best judgment and factors such as the extent of the runway
contamination, runway length and braking action reports to decide if a landing should be
attempted at all.
Hydroplaning: Hydroplaning occurs when the tire(s) ride on a film of water and don’t touch
the runway. The speed at which a tire is prone to hydroplane is primarily a function
of tire pressure. To complicate matters, if the water is contaminated by ice, dust, oil, or
grease, it makes the water more viscous and more difficult for the tires to displace. In this
case the tire hydroplanes more readily and at slower speeds. For example, an aircraft
landing on a thin layer of slush may hydroplane more easily than on a thicker layer of
standing water. 51
If a tire hydroplanes long enough to stop spinning, then all braking action is lost. The
effects of contamination on braking action / landing distance can be:
• Snow: Twice as slippery as dry pavement
• Ice: 4-16 times as slippery as dry pavement
• Wet runway: 60% increase in landing distance without hydroplaning
• Hydroplaning: Double or triple the landing distance without reverse thrust.
Rubber Reversion: Rubber reversion is a form of hydroplaning where heat under the tire
produces high pressure, superheated steam. The extreme heat causes the rubber to revert to
its uncured state and form a seal around the footprint area, trapping the high-pressure steam
51
AA 737 Operating Manual, Vol. 1, Section “Approach - Landing - Go- Around”, page 35.10

and causing the tire to “float”. Rubber reversion can occur if the runway touchdown zone is
damp and the wheels don’t spin-up.
Factors Affecting Hydroplaning:
Standing Water Depth.
Landing on a runway during or immediately after heavy rain is the condition most favorable
for hydroplaning. Even with no wind, most crowned runways have adequate drainage even
in heavy rain. Drainage can be seriously affected in crosswinds above 10 knots. When
landing on a wet runway with a crosswind, the Upwind side of the runway may have more
water on it than the downwind side. A 15 to 20 minute waiting period after a heavy
downpour is usually sufficient for water to drain.
Tire Condition
If the tire tread depth is greater than the depth of the water on the runway, then hydroplaning
usually won’t occur. Knowing the general condition of the tires is helpful when landing on a
contaminated runway.
Tire Pressure
Hydroplaning can occur at speeds above 9 times the square root of the tire pressure.
Once started, however, hydroplaning can persist down to 8.1 times the square root of the tire
pressure. So, the range of speeds most conducive for hydroplaning at 200 psi is 127 to 115
knots.
Runway Condition
Grooved runways reduce the effective water depth compared to un-grooved runways.
Also, heavy rubber deposits in the touchdown zone can adversely affect the runways landing
characteristics.
Aircraft Touchdown
Cross the threshold at VREF (plus wind additives). Touchdown should be firm and at the
desired aim-point – don’t float. Eliminate any side drift prior to touchdown and ensure
spoilers deploy (manually if necessary).

1.17.1.1.4 Factors Affecting Landing Distance
Advisory information for normal and non-normal configuration landing distances is contained in
the PI Section of the AA Quick Reference Handbook (QRH).
Actual stopping distances for a maximum effort stop are approximately 60% of the dry
runway field length requirement. Factors that affect stopping distance include: height and
speed over the threshold, glide slope angle, landing flare, lowering the nose to the runway,
use of reverse thrust, speed-brakes, wheel brakes and surface conditions of the runway.
Note: Reverse thrust and speed-brake drag are most effective during the high speed
portion of the landing. Deploy the speed-brake lever and activate reverse thrust with as
little time delay as possible.
Speed-brakes fully deployed, in conjunction with maximum reverse thrust and maximum
manual antiskid braking, provides the minimum stopping distance.
Floating above the runway before touchdown must be avoided because it uses a large portion
of the available runway. The airplane should be landed as near the normal touchdown point
as possible. Deceleration rate on the runway is approximately three times greater than in the
air.
1.17.1.1.5 AA Approach Preparation/Calculation of Landing Distance
The AA B737 Aircraft Operating Manual Vol. 1, Section “Climb – Cruise – Descent”, Page
15.7 states the following for approach preparation: (51)
Considerations Prior to Descent

To optimize situation awareness, the Captain should ensure that significant terrain and
obstacles affecting arrival or approach are identified. Review charted MSA, Grid MORA,
MEA, contour or spot elevation, EGPWS, and TRR indicated on the flight plan (highest actual
terrain height 5 left and right of course between route waypoints on the planned route).
Both flight crew should review field conditions and special procedures for the arrival airport,
including Ops Advisory pages. When the ATIS indicates that SMGCS (Surface Movement
Guidance Control System) procedures are in effect, brief the SMGCS page.”
Approach Briefing
To optimize situational awareness, planning for the approach and landing should begin
before departure. The arrival weather forecast, NOTAMS, field conditions, and MEL
items should be considered during preflight planning.
The approach briefing is completed prior to top of descent, to the extent possible, to minimize
distractions in high-density operations at lower altitudes. The Captain will conduct whatever
briefing is appropriate to the situation (e.g., poor weather, inexperienced crewmember,
special qualification airports, etc.). The Captain may delegate the briefing to the F/O.

Each pilot is responsible for reviewing the applicable approach chart. Set-up for the
instrument approach, if available. Crosscheck the Jeppesen page against the FMS data for
the arrival, approach and missed approach. 52
This contains no specific reference to the circumstances under which a Required Runway
Landing Length calculation should be made. 53
However, AA Performance Manual, Section R14, Bulletin 737-07, dated 11-26-

06 (see Section 1, Appendix 6) stated:
“The new FAA recommendation is to confirm landing performance limits just prior to
landing, using the actual runway conditions at time of landing. If the landing conditions,
from the time of dispatch do not change, there is no need to do this assessment, because the
requirements for dispatch are sufficient to assure adequate performance at time of landing.
“However, if conditions change, or deteriorate, the flight crew should use the charts on the
revised Wind Component and Landing Data Card (attached to this bulletin) to confirm
adequate runway length for landing.”
The “new FAA recommendation” referred to was Safety Alert For Operators (SAFO) 06012.
1.17.1.1.6 AA Policy on the use of reverse thrust on landing
AA B737 Operating Manual, states:
Apply reverse thrust as soon as possible after nose wheel touchdown.

• When reversing, if directional control is lost, reduce reverse until control is regained.
Use forward idle thrust if necessary.
• Do not come out of reverse at a high RPM. Sudden transition of reversers before
engines spool down will cause a forward acceleration.
• Use as much of the runway for roll-out as needed to slow airplane to a safe taxi speed
before turning off a wet/slippery runway.
CAUTION
In an emergency, use maximum reverse thrust, if required, to stop in the remaining
runway. 54
This information is found in the Cold Weather Operations portion of the AA 737
Operating Manual.
52
AAB737 Aircraft Operating Manual Vol 1, Section “Climb – Cruise – Descent”, Page 15.7
53
Section 2.20
54
AA737 Operating Manual, Vol. 1, Section “General”, page 30.27

1.17.1.1.7 AA Procedures for Landing with Tailwinds
AA B737 Operating Manual, states:
Landing Flaps
• Flaps 15, 30 or 40 are the normal landing flap settings.
• Use of flaps 40 is recommended when landing with:
– Braking action reported less than good,
– Tailwind,
– Wet / contaminated runway, or
– When deemed prudent by the Captain.
• Flaps 15 may be required at high density altitude airports when climb
limited maximum landing weight is exceeded for landing flaps 30 or 40.
Refer to the Performance Manual – LANDING. 55
The fight dispatch document for the accident flight contained the following “Special Message” to
flight crew members and was applicable to AA331:
SPC MSG NBR 9482

SUBJECT- 15 KNOT TAILWIND AUTHORIZATION
REFERENCE- 737 OPERATING MANUAL VOLUME 1
EFFECTIVE IMMEDIATELY, AMERICAN AIRLINES 737 AIRCRAFT ARE AUTHORIZED TAKEOFFS
AND LANDINGS WITH UP TO AND INCLUDING 15 KNOTS OF TAILWIND
COMPONENT FOR VISUAL, CIRCLING, NON-ILS, AND CAT I ILS
APPROACHES ONLY. THE RESTRICTION OF 10 KNOTS TAILWIND MAXIMUM
STILL APPLIES FOR HUD LOW VISIBILITY TAKEOFFS AND CAT II OR III
APPROACHES. RESTRICTED CAPTAINS EXERCISING AMERICAN AIRLINES EXEMPTION 5549
MUST COMPLY WITH THE WIND LIMITATIONS IN FM I, SECTION 10,
PARAGRAPH 2.2. AS ALWAYS, FLIGHT CREW MUST ENSURE THE REPORTED
TAILWIND COMPONENT COMPLIES WITH AIRPLANE PERFORMANCE
REQUIREMENTS FOR THE RUNWAY IN USE.
AA confirmed to the investigation by e-mail on 02 December 2011 that the intention of this
message was to make it mandatory for AA B737 flight crews to perform a Landing Performance
Assessment before landing with a tailwind. With relation to this authorization, the captain was not
restricted.
Neither of the flight crew of AA331 had received specific training from AA on landing with
tailwinds of up to 15 knots, nor did AA conduct specific training in simulators on tailwind
landings. The captain said he had had some training on tail wind landings, but did not specify
when or on which type of aircraft. When interviewed for this investigation, s o m e AA B737
training captains said they did not train pilots for tailwind landings.
The AA Aircraft Operating Manual and the Flight Crew Training Manual for the B737-800 did
not contain any information regarding landing with tailwinds other than in the Operating Manual
55
AA737 Operating Manual, Vol 1, Section “Approach - Landing - Go-Around”, page 15.2

excerpts quoted in Section 1 of this report, and the Limitations section, where the maximum
tailwind component is stated to be 15 knots.
1.17.1.1.8 AA Normal Procedures for Landing from ILS
The AA Boeing 737-800 Flight Manual Part 1, Section 10, “Approach and Landing”, page 32
defines the desired touchdown point as the first 800 to 1,500 feet beyond the threshold, and the
touchdown zone as the first 3,000 feet of the runway, beginning at the threshold.
The AAB737 Operating Manual recommended the following procedures for landing on slippery
runways:
Landing
Obtain current runway condition reports for both destination and alternate.
Request runway surface friction information from Approach Control or tower (refer to
Flight Manual Part I and II).
Flaps 40 is recommended for landing on slippery runways (refer to “Landing on a
Slippery runway” in the APPROACH – LANDING – GO-AROUND section).
Falling or blowing snow can create visual illusions or depth perception problems. When
flying the HUD, the Taxi Light may be left off if it causes visual illusions.
If landing on a wet / slippery runway, the recommended technique is:
• Land on speed.
• Touchdown at the planned point. A firm landing is better than a “grease job”.
• Keep nose wheel firmly on runway with elevator.
CAUTION
An excessive amount of down elevator will download the main gear and reduce braking
efficiency.
• Use maximum autobrakes or aggressive manual braking and auto spoilers.
• Maintaining directional control is the highest priority.
• Apply reverse thrust as soon as possible after nose wheel touchdown.
• When reversing, if directional control is lost, reduce reverse until control
is regained. Use forward idle thrust if necessary.
• Do not come out of reverse at a high RPM. Sudden transition of reversers before
engines spool down will cause a forward acceleration.
• Use as much of the runway for roll-out as needed to slow airplane to a safe taxi speed
before turning off a wet/slippery runway. 56
CAUTION
In an emergency, use maximum reverse thrust, if required, to stop in the remaining
runway.
56
AA737 Operating Manual, Vol. 1, Section “General”, page 30.27

This information is found in the Cold Weather Operations portion of the AA 737 Operating Manual
and is located in the Boeing FCOM. AA crews are required to use the guidance in the FAA
approved AA Operating Manual. – Vol. 1.
1.17.1.1.9 AA Information on Baulked Landings (Go Around)
AA did not practice rejected landings following touchdown as a regular simulator training
procedure, although go-around procedures were practiced on a regular basis as well as baulked
landings.
The captain reported that he had been trained in the simulator on go-around procedures after
touchdown.
There was no statement in the AAB737 Operating Manual to the effect that a first officer who
was the pilot monitoring could call for a go-around, with the pilot flying (that is, the captain)
being obliged to follow through with this. It was not possible to determine exactly what was the
company’s procedure regarding this.
The AA Flight Manual, stated the following:
Missed-approach decision: The Captain has full control and authority in the operation
of the aircraft (FAR 121.533, 121.535), however the Captain should give every
consideration to a recommendation by another cockpit crewmember that a missed approach
be executed. 57
AA noted a need for clarification on go-arounds in their Flight Operations Information Bulletin
Number 2010-06, dated April 2010 (post-accident) when they said:
Feedback from our pilots and an analysis of our current stable approach and go-around
guidance indicates that there are several key areas which require attention or need
clarification in order to reduce and perhaps eliminate the reluctance of pilots to execute
a go-around when an approach is unstable:
• Specific go-around flight parameters are not very clearly defined in the Operating Manual
which makes it nebulous and left to wide interpretation as to when a go-around is
mandatory.....
57
AA Flight Manual, Section 10, Page 2

AA has subsequently revised the AA Boeing 737 Operating Manual, with the following
language:
Go-Around Requirements
On final, the Pilot-Flying is responsible for executing a go-around if any of the

parameters listed below are exceeded without Pilot-Flying correction. If the Pilot-
Monitoring observes that the Pilot-Flying is not executing a go-around, he or she is
responsible for directing a go-around by calling – “GO-AROUND”. The directed go-
around will be executed unless an emergency situation overrides this requirement. 58
1.17.1.1.10 AA No Fault Go-Around Policy and Missed Approach decision
AA Flight Manual states:
American Airlines has a no-fault go-around policy, recognizing that a successful approach
can end in a missed approach. If the stabilized approach requirements cannot be satisfied by
the minimum stabilized approach height or maintained throughout the rest of the approach, a
go-around is required. If in the pilot’s judgment a safe landing cannot be accomplished
within the touchdown zone, or the aircraft cannot be stopped within the confines of the
runway, a go-around is also required (bolding added). 59
A. American Airlines has a no-fault go-around policy. Pilots should execute every
approach with the presumption that a missed approach is a successful outcome. Plan
each approach through t h e m i s s ed -approach p r o c e d u r e and make the decision to
land only when all criteria are safely satisfied.
B. Missed-approach decision: The Captain has full control and authority in the
operation of the aircraft (FAR 121.533, 121.535), however the Captain should give every
consideration to a recommendation by another cockpit crewmember that a missed approach
be executed. 60
The First Officer is required to immediately advise the Captain of any

deviation, from applicable regulations, policies or procedures, or any unsafe
condition which may place the aircraft, passengers or crewmembers in jeopardy. The
Captain may choose to disregard this a d v i c e , but r e g a r d l e s s o f t h e d e g r e e of
58
AA Boeing 737 Operating Manual, page 15.8
59
AA Flight Manual Part I, Section 10, Paragraph 6.0, “Missed approach”
60
AA Flight Manual Part I, Section 10, Paragraph 1.2, “Missed Approach Decision Making

frequency with which advice may go unheeded, flight crewmembers will be held responsible
for continuing to offer advice for the Captain's consideration. 61
1.17.1.1.11 AA Bulletin 737-07
AA Bulletin 737-07, dated 11-27-06, 62 was concerned with expanding runway surface condition
reports, and stated flight crews must use “the most adverse, reliable and appropriate braking
action report, or the most adverse expected conditions … when assessing the required landing
distance prior to landing.” It includes a Braking Action Chart which states that Standing
Water depth of 1/8 inches or less is “Good”, and Standing Water depth of 1/8 inches or more
is “Poor”.
It should be noted that AA Bulletin 737-07 stated “… if conditions change, or deteriorate, the
flight crew should use the charts on the revised Wind Component and Landing Data Card
(attached to this bulletin) to confirm adequate runway length for landing”. For Dry and
Wet/Good conditions these charts provided more than the minimum recommended 15% safety
margin and provided a 15% buffer for Fair/Medium and Poor braking action.
The investigation determined that the flight crew had this information on board, but was unable to
determine exactly how the information was used in the cockpit prior to landing, in conjunction
with the Required Runway Landing Length table on the Landing Data Card. 63
1.17.1.1.12 AA HUD Usage Policy and Briefing Guide
The AA737 Operating Manual, “HUD Usage Policy” states:
HUD use is mandatory for all take-offs and all approaches and landings to assist in tail-
strike prevention, unless it is placarded. … On approach, the HUD must be used no later
than 1,000 feet AFL. HUD will be used regardless as to whether the captain is pilot flying
or pilot monitoring. The HUD provides multiple tools to enhance situational awareness and
safety, including tail-strike prevention for both take-off and landing … 64
According to the AA B737 HUD Briefing Guide 65 the HUD will provide a flare cue below
approximately 90 feet agl. In A1 mode, the flare cue provides pitch guidance only; aim point and
61
AA Flight Manual Part 1, Chapter “Crew Qualification and Responsibility‟, Section 3, page 5, part 1.8, “First Officer
Responsibility”
62
Appendix 6
63
Appendix 7
64
AA737 Operating Manual, Section “General”, Page 15.1 “HUD Usage Policy”
65
AA B737 HUD Briefing Guide (Section XI, Flare Cue, Page 5

lateral placement is the Captain’s responsibility (e.g. – flare cue will not attempt to land the
aircraft in the touchdown zone.)
The artificial runway is not displayed in A1 mode. 66 The flight director is removed 50
ft. below minimums but reappears at 90 feet as the flare cue. 67 Final Runway alignment would
be via visual reference to the runway centerline, and glide path aligned to the PAPI.
1.17.1.1.13 AA B737 RRLL Tables, Wind Component and Landing Data Card
Estimated Wet and Contaminated Runway Landing Distance tables. These are in Appendices 7 and
8.
1.17.1.2 AA – Other Information
1.17.1.2.1. AA Crew Resource Management (CRM) Training
AA had an approved CRM training program which all flight crew attended. In interviews
with training personnel, they stated that CRM training was part of the initial, upgrade and
new-hire training, and in the Advanced Qualification Training program (AQP), every
nine months. The training included how the crew interacts, and they used examples of
good and poor interactions, with incidents being recreated by crews in a video made by
the human factors department. Also, during check rides, the tone of the cockpit was
evaluated to ensure it was conducive to efficient flight safety, and to ensure all pilots
were willing to speak up.
1.17.1.2.2 AA Flight Manual Page 10-7X.
This included information about standing water at MKJP Kingston. It stated “Runway is uneven
and subject to pools of standing water after heavy rain.” See Appendix 5.
1.17.1.2.3 AA Boeing 737 Training
AA followed a standard program of pilot training for the Boeing 737 that was approved by the
FAA as meeting FAR part 121 standards, and, for recurrent training, used an annual
qualification program of ground school, simulator training and checks, and an annual line check.
Both flight crews were current on the aircraft for 90 day landings, both had valid proficiency
checks and line checks, and no issues were noted from their training records.
Neither had been specifically trained in landing with tailwinds, nor in the use of HUD with an
offset ILS. The B737 Training Manager advised that this was not normally done in flight
66
AA HUD Briefing Guide, Section II, HUD Symbols and Display Overview, page 1
67
AA HUD Briefing Guide, Section VIII, NP Mode, page 3

simulators and that crosswind landings and landing on a short runway (Orange County) and at
high altitude runways were all practiced in the simulator training. The captain reported that he
had r e c e i v e d t r a i n i n g r e g a r d i n g g o - a r o u n d after touchdown.
1.17.1.2.4 AA Dispatch Information
In accordance with AA procedures, Flight AA331 was dispatched by the AA Dispatch in

Dallas, Texas, by a qualified flight dispatcher.
The AA Dispatcher who dispatched this flight was interviewed, and he made the following
statements:
1. The AA station manager was responsible for providing field condition reports.
2. The captain would know he was legal to fly to Kingston because a dispatch release was
sent.
3. Zero winds were considered for the release.
4. The captain was expected to select the best runway for landing.
5. Dispatch did not play any part in runway selection.
6. Flights were dispatched either dry or wet.
7. The captain could not tell whether the flight was dispatched wet or dry – there was no
place on the flight plan to indicate that.
8. As long as the braking action was better than Nil, the flight would be released.
9. The captain was responsible for determining the braking action on arrival.
10. He sent AA331 an updated METAR about 21:48 EST, which included weather going
from rain showers to thunderstorms.
11. He did not know of any Caribbean airport being other than good/wet, and had never
received a report of standing water at Kingston.
The METAR mentioned in item 10 was:

SPECI MKJP 230228Z 31009KT 5000 TSRA BKN014 FEW016CB SCT030 BKN100 22/19
Q1013
“Kingston (MKJP) special weather observation at 21:28 EST (02:28 UTC, 23 Dec), wind 310
degrees at 9 knots, visibility 5,000 m (approximately 3 miles) in thunderstorms and moderate
rain, ceiling broken at 1,400 feet, few clouds at 1,600 feet in cumulonimbus clouds, scattered
clouds at 3,000 feet, broken cloud at 10,000 feet, temperature 22° C, dew point 19° C, altimeter
setting 1013 mb.” AA331 was dispatched “Wet”.

METARS on AA331 Dispatch Document
The following information was on the finalized American Airlines dispatch document for
AA331, given to the investigation, and was marked RH, meaning recorded history.
KIN
230317 230300Z 32008KT 3000 PLUSSHRA BKN014 FEW016CB SCT03
230317 230300Z 32008KT 33000 ?SHRA BKN014 FEW016CB SCT030 B
230315 230300Z 32008KT 3000 ?SHRA BKN014 FEW016CB SCT030 BK
230238 230228Z 31009KT 55000 TSRA BKN014 FEW016CB SCT030 BK
230235 230228Z 31009KT 5000 TSRA BKN014 FEW016CB SCT030 BKN
230008 230000Z 32004KT 9999 FEW016 BKN030 BKN100 24/19 Q101
222316 222300Z 00000KT 9999 VCSH SCT016 SCT030 BKN100 24/20
222226 222200Z 35008KT 9999 FEW015 BKN032 BKN100 24/19 Q1
222116 222100Z 33005KT 9999 FEW015 BKN032 BKN100 23/19 Q1
222007 222000Z 31005KT 9999 -RA FEW012 SCT032 OVC100 23/19
221929 221900Z 10012KT 9000 -RA SCT012 OVC100 22/19
The latest field condition reports for KIN were included in the AA331 dispatch release and
showed the following:

*******************************************************
*******************************************************
--------------------------------------------------------------
--------------------------------------------------------------
--------------------------------------------------------------
--------------------------------------------------------------* KIN FIELD
This report was last updated 15:20 Local Time, i.e. five hours before the flight departed
Miami.
By FAA regulation, the “Field Report” was not required for dispatch purposes. However, AA
Bulletin 737-07, “Landing Performance Check”, stated “If the landing conditions from the time
of dispatch do not change, there is no need to do this assessment … ”
Therefore it would be necessary for the flight crew to have the “Field Report”, that is, “ … the
landing conditions from the time of dispatch” to compare them with the “ … actual runway
conditions at the time of landing…” and then decide whether or not it was necessary to do a
landing performance confirmation.

The following forecasts were used for planning for the AA331 flight:
*** FORECAST USED FOR PLANNING ***
MIA TAF/
TAF KMIA 222103Z 2221/2321 06008KT P6SM SCT040 SCT150 OVC250
FM230100 04006KT P6SM SCT030 BKN250
FM231400 07012KT P6SM SCT025 SCT250
FM231800 08015G22KT P6SM SCT030 BKN050
WSI/JB
KIN 221500Z 2218/2318 12016KT 9999 -RA FEW016 BKN032 TEMPO

2218/2302 6000 TSRA BKN016 SCT018CB BECMG 2302/2304 35008KT
TEMPO 2304/2312 8000 SHRA BKN016 FEW018CB SCT032 BECMG
2314/2316 20006KT
Relevant part decoded is: Kingston, issued on 22 December at 15:00 UTC for period 18:00 UTC
on 22 December to 18:00 UTC on 23 December, wind 120 at 16 knots, visibility more than 10
kilometers (6 miles), light rain showers, few clouds at 1,600 feet, broken cloud at 3,200 feet.
Temporarily, between 18:00 UTC on 22 December and 02:00 UTC on 23 December, visibility
6,000 meters, thunderstorms and rain showers, broken cloud at 1,600 feet, scattered cloud at
1,800 feet with cumulonimbus cloud, becoming, from 02:00 to 04:00 UTC on 23 December
wind 350 at 8 knots. (Estimated time of arrival was about 03:21 UTC, so the balance of
forecast is not relevant).
MBJ 221500Z 2218/2318 09012KT 9999 -RA FEW018 BKN030 OVC080
TEMPO 2218/2302 7000 SHRA BKN017 FEW020CB BKN030 BECMG
2302/2304
15005KT BECMG 2308/2310 24010KT SCT020 BKN030 BECMG
2316/2318
32010KT FEW020 SCT034
The following forecast information for Kingston was provided to the flight crew by AA flight
dispatch prior to departure from Miami.
TAF MKJP 222100Z 2300/2324 34008KT 9999 FEW018 SCT032 BKN090

TEMPO 2300/2314 8000 SHRA BKN016 FEW018CB SCT032
BECMG 2314/2316 20010KT
TEMPO 2318/2324 8000 SHRA SCT018 SCT080
De- coded, this is: a Terminal Aerodrome Forecast (TAF) was issued for MKJP at 16:00 EST
(21:00 UTC) for the period 00:00 UTC on 23 December to 24:00 UTC on 23 December (19:00
EST, 22 December to 19:00 EST 23 December).
The TAF forecasted wind from 340 degrees at eight knots, visibility 10 kilometers or more
(better than six miles), a few clouds at 1,800 feet, scattered clouds at 3,200 feet, and a
ceiling at 9,000 feet with temporary conditions between 18:00 EST, 22 December and 09:00
EST, 23 December of visibility 8,000 metres (five miles) in moderate rain showers, ceiling broken
at 1,600 feet, few clouds at 1,800 feet in cumulonimbus clouds, and scattered clouds at 3,200 feet.
(Balance not relevant to the investigation).
The graphic area forecast was not included in the dispatch document given to AA331 flight crew,
nor was it required to be.
The NOTAM referred to in 1.18.4 was included in the dispatch document received by the
AA331 flight crew.
The following Special Message was included in the dispatch document:
/// SPECIAL INFO MESSAGES ///

SPC MSG NBR 9482
EFFECTIVE IMMEDIATELY, AA 737 AIRCRAFT ARE AUTHORIZED TAKEOFFS
AND LANDINGS WITH UP TO AND INCLUDING 15 KNOTS OF TAILWIND
COMPONENT FOR VISUAL, CIRCLING, NON-ILS, AND CAT I ILS
APPROACHES ONLY. THE RESTRICTION OF 10 KNOTS TAILWIND MAXIMUM
STILL APPLIES FOR HUD LOW VISIBILITY TAKEOFFS AND CAT II OR III
APPROACHES. RESTRICTED CAPTAINS EXERCISING FAA EXEMPTION 5549
PARAGRAPH 2.2. AS ALWAYS, PILOTS MUST ENSURE THE REPORTED
The last sentence of this, “AS ALWAYS, PILOTS MUST ENSURE THE REPORTED
REQUIREMENTS FOR THE RUNWAY IN USE” meant that it was mandatory for flight crew to
complete a landing performance assessment before landing. This was confirmed by AA Quality
Assurance Management in an Email to the NTSB on 01 December 2011.
Based upon the weather forecasts prior to departure and the available runway length at Kingston,
AA331 was a legally dispatched flight. 68
1.17.1.2.5 AA Required Runway Landing Length (RRLL) table, and Landing Data Card.
The AA B737 Wind Component and Landing Data Card 69, which stated that the Required
Runway Landing Length for Dry or Wet/Good runway Conditions was based, and included,
“demonstrated ‘air distance’” from runway threshold to touchdown”, and for Medium/Fair or
Poor Runway Conditions “includes 1,000 feet ‘air distance’ from threshold to touchdown,
contrasts with other AA information, as follows: The touchdown zone is defined in AA Flight
Crew Operations Manual (FCOM) 70, as the “first 3,000 feet of the runway beginning at the
threshold”, or the first third of the runway, in this case 8,911/3 = 2,970 feet. The Flight Crew
Operations Manual (FCOM) 71, defines the desired touchdown point as the first 800 to 1,500
feet beyond the threshold.
68
Section 1.17.4.1
69
Appendix 7
70
AA Flight Crew Operations Manual (FCOM), Part I, Section 10, paragraph 7.2, page 32
71
Flight Crew Operations Manual (FCOM), Part I, Section 10, paragraphm7.3, page 32,
AA was asked to define “demonstrated ‘air distance’”, and responded:
The manufacturer has never listed the “air distance” used in the certified landing distances.
The AFM only contains the total distance. The Fair/Medium and Poor columns, comes
from the manufacturer, but is not certified. In this case, the manufacturer specifies
that 1,000 feet is used for the “air distance”.
The investigation assumed that the “air distance” for the Wet/Good landing figures was about
1,000 feet.

1.17.1.2.6 AA Landing Distance Calculations
1.17.1.2.6.1 Landing Data Card
The following Required Runway Landing Length table, from the AAB737-800 Aircraft
Operating Manual, Appendix 4 for Landing Performance, Landing Data Card, is for a Flap 30,
sea level landing. 72
Aircraft Runway Runway Reported Reported

Gross condition: condition/ braking braking
weight reported action: Action:
1000’s DRY
braking action:
pounds FAIR/MED POOR
WET/GOOD
100 4,122 4,740 5,360 6,790
110 4,431 5,096 5,770 7,350
120 4,787 5,505 6,180 7,910
130 5,175 5,952 6,590 8,480
140 5,566 6,401 7,010 9,070
144 5,714 6,571 7,180 9,300
150 5,937 6,827 7,430 9,660
160 6,293 7,237 7,850 10,250
170 6,647 7,644 8,270 10,840
174.2 6,795 7,814 8,440 11,090
Headwind per -21 feet -21 feet -33 feet -49 feet
knot
Tailwind per +101 feet +103 feet +121 feet +191 feet
knot
No Reverse 0 0 +1,633 feet +3,766 feet
Thrust
Table 6: Landing Data Card
72
Appendix 7

WET/GOOD braking action
Given:
Runway Elevation 10 Ft asl

Runway condition Less than 1/8 inch (3mm) water.
Landing Weight 144,000 LB
Flap setting 30°
Touchdown point 1,000 Ft
Reverse Thrust None
Braking Maximum Manual
With tail wind 0, Required Runway Landing Length 6,571 Ft
With tail wind 8KT, Required Runway Landing Length 7,395 Ft (6,571 Ft + 824 Ft)
With tail wind 14KT, Required Runway Landing Length 8,013 Ft (6,571 Ft + 1,442 Ft)
Note, these figures are based on the 14 Code of Federal Regulations (CFR) Part 25.125 demonstrated
landing distance on a dry runway, and include a 92% safety margin.
FAIR/MEDIUM braking action
Given:
Runway Elevation 10 Ft asl

Runway condition More than 1/8 inch (3mm) water.
Landing Weight 144,000 LB
Flap setting 30°
Touchdown point 1,000 FT
Reverse Thrust Normal
Braking Max autobrakes
With tail wind 0, Required Runway Landing Length 7,180 Ft
With tail wind 8 KT, Required Runway Landing Length 8,148 Ft (7,180 Ft + 968 Ft)
With tail wind 14 KT, Required Runway Landing Length 8,874 Ft (7,180 Ft + 1,694 Ft)
Note: these figures are the manufacturer’s estimated landing distance, and include a 15% safety
margin.
For both WET/GOOD and FAIR/MEDIUM the calculated required runway landing lengths are
based on crossing the runway threshold at 50 Ft agl, at the normal reference landing speed, VRef,
with the aircraft in the correct landing configuration and no system malfunction that would affect
the landing distance.

1.17.1.2.6.2 AA Performance – Landing Table
The information in the AA 737 Operating Manual, 16 Performance – Landing, “Estimated Wet
and Contaminated Runway Landing Distances” 73, gives the following information for landing
at 144,000 pounds, sea level, flap 40, maximum manual braking, maximum reverse thrust, landing
1,000 feet from threshold with 14 knot tail wind, braking action FAIR, no additional margin. In
this case the absolute distance for stopping was 7,240 feet. This distance plus 15% margin equals
8,326 feet.
1.17.1.2.7 Advisory Circular No: 91-79
FAA Advisory Circular (AC) 91-79, Appendix 4, stated:
“For example, if there is no clear report of runway condition but if the pilot knows rain has
been in the area, that pilot should assume that the runway is at least wet. If there is rain
actively falling on the runway, standing water should be assumed”,
That is, the runway condition should be considered as contaminated. AC 91-79 was provided to
the AA Safety Officer by the FAA and the Safety Officer in turn provided AC 91-79 to the
various fleet captains. It was not tracked by the Safety Officer to see if any action was taken in
relation to its recommendations, nor was the material in AC 91-79 considered to be mandatory in
the way that the material in SAFO 06012 was put into Bulletin 737-07, and expected to be
followed by AA crews. It should be noted that AC91-79 did not carry regulatory force, and was
only advisory.
1.17.1.2.8 Crew Flight/Duty/Rest System
AA had a comprehensive system of flight and duty time limitations and maintained good
records of flight crew flight/duty/rest times. It had a scheduling alert system to prevent assignment
of crew members to flights when they had not had the minimum required rest periods prior to those
flights. The system was documented in the Company Operations Manual.
1.17.1.2.9 AA Interviews with Staff
Evidence from interviews with AA training and operational staff indicated that flights were
dispatched either Dry or Wet/Good. In the Kingston Field Report the measurement “0.10 IN
WATER” was not the result of an actual measurement, but was more of a term used to describe
wet, as opposed to dry, runways.
73
Appendix 8

When one of the senior captains was asked whether crews should re-evaluate runway conditions
from Wet/Good as stated by a controller to Medium or Poor when flying an approach in heavy rain,
he responded that people on the ground have better information than pilots in the air.
Interviews with AA training management and training staff indicated the following:
1. Since 2001 AA had operated B737 with a 15 knot landing tailwind limit to four
destinations, otherwise the limit was 10 knots.
2. Problems with operating into San Jose airport had resulted in the increase of the tailwind
landing limit for the whole AA B737 fleet from 10 to 15 knots.
3. No simulator or other training on tailwind landing was in place or was introduced when
the AA B737 tailwind landing limit was increased from 10 to 15 knots.
4. No changes in landing procedures were made when the AA B737 tailwind landing limit
was increased from 10 to 15 knots.
5. The AA flight crews were informed of the B737 tailwind landing limit increase from 10
to 15 knots by company bulletin.
6. The decision making process to increase the B737 tailwind landing limit from 10 to 15
knots included Ops Engineering, Flight Ops and Boeing.
7. The decision making process to increase the B737 tailwind landing limit from 10 to 15
knots did not include the AA B737 Lead Check Airman, the Managing Director of Flight
Training or the Manager, Flight Safety.
8. AA’s B737 AQP did not include tailwind landing.
9. An AA B737 Check Airman stated he did not use tailwind landings in checks.
Also, the following statements were made:
1. Tailwind has no effect on touchdown point.

2. Emphasis was placed on landing in the touch down zone.
3. Landing ground speed with a headwind at high altitude is similar to, and may be faster
than, landing at another airport with a tailwind.
4. There is little difference between a tailwind landing and a high altitude landing.
5. There is no benefit to training tailwind landing.
6. There is no difference between a 10 knot and a 15 knot tailwind landing.
1.17.1.2.10 AA– information regarding Bulletin 737-07 and AC 91-79
The following information was given to the investigation by the AA Flight Safety
Programs Manager:
1. Re: AA flight crew training regarding Bulletin 737-07:
AA flight crews are provided comprehensive training on 737 Aircraft Performance

during their initial training on the airplane and are subsequently required to complete a
Performance Manual review with a Ground School instructor during recurrent training. In
addition, the 737 Check Airmen review performance issues during the initial and recurrent
training simulator sessions when we simulate landing on contaminated runways.
2. Re: statement in Bulletin 737-07 “to comply with new FAA recommendations”:
This (wording) is taken directly from an AMR Certificates Management Office letter
dated September 8, 2006 requesting AA assure their office we would comply with the
recommended actions of SAFO 06-012. Bulletin 737-07 was issued to the 737 fleet in
response to this SAFO and CMO letter. Other fleets issued similar Bulletins to their Operating
Manual Performance Sections as well.
3. Re: question as to why compliance with Bulletin 737-07 by AA flight crews not made
mandatory by AA (that is, AA used “should”, which means “recommended”, therefore it was
not mandatory):
AA used the same language as the FAA Advisory Circulars and Safety Alerts for
Operators on which the Bulletin was based.
4. Re: question “Did AA management expect AA flight crews to follow the instructions
(“recommendations”) in Bulletin 737-07, or assume that they would?”
AA management both expects and assumes flight crews will follow the instructions
(recommendations) in Bulletin 737-07.
5. Re: question as to why AA put Bulletin 737-07 in the Performance Section, when it was an
issue relating to operational procedures:
The Performance Section is issued as part of Operating Manual Volume 1 which is the flight
crew’s operational procedures manual. The Performance Section is what the pilots refer to, if
needed, in order to determine aircraft performance for different phases of flight. The
Performance Section is also where the FAR Part 25 performance requirements are
presented. Also, as an attachment to Bulletin 737-07, a Landing Data Card was provided for
the pilots’ reference prior to landing. This additional card is typically carried by the pilots in
their “trip book”, a small binder used to contain documents for ready reference by the pilots.
6. Re: question “What would a runway condition report of “Wet” from a controller at
Kingston airport (Jamaica) mean to an AA pilot?”
A simple report of a runway condition of “Wet” would normally indicate Good braking action
to an AA pilot. We would expect the tower to provide any amplifying information, e.g. braking
action reports from other flights or standing water on the runway such as required by the
Jamaica AIP.
7. Re: question “Why did the AA331 flight crew use an assumption that if the runway was more
than 8,000 feet, was “wet”, the tailwind was less than 15 knots, aircraft at max landing weight,
then it was safe to land? Is this method approved by AA, taught by AA, commonly used by
AA flight crews?”

A runway condition report of “Wet” with no other modifying information, e.g. “braking
action Poor”, would indicate braking action Good. An acceptable technique for a flight
crew who flies into a certain airport frequently is to conduct an “advance analysis” of the
worst case scenario for the landing runway, that is, for a known landing length, braking
action, wind component, landing weight, etc. the crew could determine in advance that as
long as they landed below the maximum weight for these worst case conditions the runway
length was acceptable. A review of this “advance analysis” prior to landing using the
actual conditions at time of landing is acceptable.
American Airlines stated that no AA documents or training materials specifically define the
“advance analysis” concept that the JCAA cites in the draft report. AA flight crews are not
trained to use, required to use, or discouraged from using this method.
8. Re: question “Did AA management make SAFO 06012 and AC 91-79 available to AA
flight crews?”
AA typically produces their own revisions, Operating Manual Bulletins, Info Bulletins,
etc. after receiving relevant documents from outside sources, e.g. Bulletin 737-07 is
based on information contained in SAFOs and Advisory Circulars. Our documents are
succinct representations of the FAA/NTSB documentation. AA also has a link on their
AAPILOTS website to the FAA and NTSB web sites where pilots can download and read
any safety documents they desire.
9. Re: question “Does AA have some system to verify that AA flight crews have read and
understood Bulletins such as 737-07?”
Flight Manual Part One instructs the crews to read and insert Operating Manual revisions
and bulletins as soon as possible after receipt. AA maintains a list in our flight planning
system software that provides the pilots with the latest status of manual revisions,
bulletins, etc. AA pilots are required to ensure they have all the current manual
revisions, etc. before flying. The current list of manual status is also available on the
AAPILOTS web site.
1.17.1.2.11 AA Safety Initiatives Re: AA331 Landing Accident
See Section 4.0, Safety Action, Paragraph 4.1.1.

1.17.2 Boeing Information
1.17.2.1 Glide path, slippery runway, friction measurement
Boeing 737-800 Flight Crew Training Manual, states:
Height of the airplane over the runway threshold also has a significant effect on total landing
distance. For example, on a 3° glide path, passing over the runway threshold at
100 feet altitude rather than 50 feet could increase the total landing distance by approximately
950 feet. This is due to the length of runway used up before the airplane actually touches down.
Glide path angle also affects total landing distance. As the approach path becomes flatter,
even while maintaining proper height over the end of the runway, total landing distance is
increased.
Slippery Runway Landing Performance

Appendix A.2.7
When landing on slippery runways contaminated with ice, snow, slush or standing water, the
reported braking action must be considered. Advisory information for reported braking actions
of good, medium and poor is contained in the PI section of the QRH. The performance level
associated with good is representative of a wet runway. The performance level associated
with poor is representative of a wet ice covered runway. Also provided in the QRH are
stopping distances for the various autobrake settings and for non-normal configurations.
Flight crew should use extreme caution to ensure adequate runway length is available
when poor braking action is reported.
Pilots should keep in mind slippery/contaminated runway advisory information is based on an

assumption of uniform conditions over the entire runway. This means a uniform depth for
slush/standing water for a contaminated runway or a fixed braking coefficient for a slippery
runway. The data cannot cover all possible slippery/contaminated runway combinations and
does not consider factors such as rubber deposits or heavily painted surfaces near the end of
most runways.
One of the commonly used runway descriptors is coefficient of friction. Ground friction
measuring vehicles typically measure this coefficient of friction. Much work has been done
in the aviation industry to correlate the friction reading from these ground friction measuring
vehicles to airplane performance. Use of ground friction vehicles raises the following
concerns:
• the measured coefficient of friction depends on the type of ground friction measuring vehicle
used. There is not a method, accepted worldwide, for correlating the friction measurements
from the different friction measuring vehicles to each other, or to the airplane's braking
capability.

• most testing to date, which compares ground friction vehicle performance to airplane
performance, has been done at relatively low speeds (100 knots or less). The critical part of the
airplane's deceleration characteristics is typically at higher speeds (120 to 150 knots).October
31, 2008
Boeing 737 NG Flight Crew Training Manual
Landing
FCT 737 NG (TM) 6.33
• ground friction vehicles often provide unreliable readings when measurements are
taken with standing water, slush or snow on the runway. Ground friction vehicles might not
hydroplane (aquaplane) when taking a measurement while the airplane may hydroplane
(aquaplane). In this case, the ground friction vehicles would provide an optimistic reading of
the runway's friction capability. The other possibility is the ground friction vehicles might
hydroplane (aquaplane) when the airplane would not, this would provide an overly pessimistic
reading of the runway's friction capability. Accordingly, friction readings from the ground
friction vehicles may not be representative of the airplane's capability in hydroplaning
conditions.
• ground friction vehicles measure the friction of the runway at a specific time and location.
The actual runway coefficient of friction may change with changing atmospheric
conditions such as temperature variations, precipitation etc. Also, the runway condition
changes as more operations are performed.
The friction readings from ground friction measuring vehicles do supply an additional piece of
information for the pilot to evaluate when considering runway conditions for landing. Crews
should evaluate these readings in conjunction with the PIREPS (pilot reports) and the physical
description of the runway (snow, slush, ice etc.) when planning the landing. Special care
should be taken in evaluating all the information available when braking action is
reported as POOR or if slush/standing water is present on the runway (bolding added).
October 31, 2008. 74
1.17.2.2 Boeing Flight Crew Operations and Training Manuals, B737, Landing
The Boeing 7 3 7 - 8 2 3 Flight Crew Operations Manual, provides the following normal
procedure for landing from an ILS approach, with regard to autopilot and auto throttle handling:
If suitable visual reference is established at DA(H), MDA(H) or the missed approach point,
disengage the autopilot and autothrottle (bolding added). Maintain the glide path to
landing. 75
74
Boeing 737-800 Flight Crew Training Manual, Revision 8, Page 6.32 "Factors Affecting Landing Distance."
75
The Boeing 7 3 7 - 8 2 3 Flight Crew Operations Manual, AA Inc., Section “Normal Procedures – Amplified Procedures”,
page NP.21.55

The Boeing Flight Crew Training Manual, Page 6.9, contains the following instructions for
landing the B737:
Flare and Touchdown
The techniques discussed here are applicable to all landings including one engine inoperative
landings, crosswind landings and landings on slippery runways. Unless an unexpected or
sudden event occurs, such as wind shear or collision avoidance situation, it is not appropriate
to use sudden, violent or abrupt control inputs during landing. Begin with a stabilized
approach on speed, in trim and on glide path.
When the threshold passes under the airplane nose and out of sight, shift the visual sighting
point to the far end of the runway. Shifting the visual sighting point assists in controlling
the pitch attitude during the flare. Maintaining a constant airspeed and descent rate assists in
determining the flare point. Initiate the flare when the main gear is approximately 20 feet
above the runway by increasing pitch attitude approximately 2° -
3°. This slows the rate of descent.
After the flare is initiated, smoothly retard the thrust levers to idle, and make small pitch
attitude adjustments to maintain the desired descent rate to the runway.
Ideally, main gear touchdown should occur simultaneously with thrust levers reaching idle. A
smooth thrust reduction to idle also assists in controlling the natural nose-down pitch change
associated with thrust reduction.
Hold sufficient back pressure on the control column to keep the pitch attitude constant. A
touchdown attitude as depicted in the figure below is normal with an airspeed of
approximately VREF plus any gust correction.
Note: Do not trim during the flare or after touchdown. Trimming in the flare increases the
possibility of a tail strike.
For airplanes equipped with HUD, flare guidance is provided in the AIII mode. Follow the
guidance cue and perform the flare and landing using HUD guidance and visual cues.
Monitor the roll out annunciation (as installed) and transition to rollout guidance. Use normal
procedures to decelerate to taxi speed.
The Boeing Flight Crew Training Manual, contains the following information regarding the
flare profile:
• 3° approach glide path

• flare distance is approximately 1,000 to 2,000 feet beyond the threshold
• typical landing flare times range from 4 to 8 seconds and are a function of approach
speed

• airplane body attitudes are based upon typical landing weights, flaps 30, VREF 30 + 5
(approach) and VREF 30 + 0 (landing), and should be reduced by 1° for each 5 knots above
this speed.
Typically, the pitch attitude increases slightly during the actual landing, but avoid over-
rotating. Do not increase the pitch attitude after touchdown; this could lead to a tail strike.
Shifting the visual sighting point down the runway assists in controlling the pitch attitude
during the flare. A smooth thrust reduction to idle also assists in controlling the natural nose
down pitch change associated with thrust reduction.
Hold sufficient back pressure on the control column to keep the pitch attitude constant. Avoid
rapid control column movements during the flare. If the flare is too abrupt and thrust is
excessive near touchdown, the airplane tends to float in ground effect.
Do not allow the airplane to float; fly the airplane onto the runway. Do not extend the flare by
increasing pitch attitude in an attempt to achieve a perfectly smooth touchdown. Do not
attempt to hold the nose wheels off the runway. 76
1.17.2.3 Boeing Information on Rejected/Baulked Landings
The Boeing 737 NG Flight Crew Training Manual describes the referenced procedure for
rejected (baulked) landings as follows:
Rejected Landing
A rejected landing maneuver is trained and evaluated by some operators and regulatory
agencies. Although the FCOM/QRH does not contain a procedure or maneuver titled Rejected
Landing, the requirements of this maneuver can be accomplished by doing the Go-Around
Procedure if it is initiated prior to touchdown. ( Refer to Chapter 5, Go- Around after
Touchdown, for more information on this subject.)
The Boeing B737 Flight Crew Training Manual, Section “Approach and Missed Approach”,
Page 5.76, provides the following instructions regarding a go around after touchdown:
Go-Around after Touchdown:
If a go-around is initiated before touchdown and touchdown occurs, continue with

normal go-around procedures. The F/D go-around mode will continue to provide go- around
guidance commands throughout the maneuver.
If a go-around is initiated after touchdown but before thrust reverser selection, auto
speed-brakes retract and autobrakes disarm as thrust levers are advanced. The F/D go-
76
Boeing Flight Crew Training Manual, Page 6.10

around mode will not be available until go-around is selected after becoming airborne. Once
reverse thrust is initiated following touchdown, a full stop landing must be made. If an engine
stays in reverse, safe flight is not possible.
1.17.2.4 Autopilot/Autothrottle Use During Approach
In an article by Bill McKenzie of Flight Crew Operations, Boeing Commercial Airplanes, May
2004, entitled “Auto throttle Use with Autopilot Off” and used for B757 training, it is stated that
the following are some disadvantages of auto throttle use with autopilot off:
 Excessive airspeed landing in gusts and turbulence,

 Potential pitch coupling close to the ground,
 Additional MCP coordination,
 Excessive unexpected throttle movement,
 Less thrust awareness, and
 Airspeed crosscheck skills not exercised.
The writer recommended:
 Use Of Manual Thrust Control In Manual Flight,

 Auto Throttle Is Disconnected No Lower Than 300 Feet,
 Operator Establishes A Clear Policy.
It is stated in the Boeing 737 NG Flight Crew Training Manual:
Autothrottle Use:
Auto throttle use is recommended during takeoff and climb in either automatic or
manual flight. During all other phases of flight, auto throttle use is recommended only
when the autopilot is engaged in CMD. 77
This is not an FAA approved AA manual; it is the manufacturer’s guidance. This information was not
contained in the FAA approved manuals for AA B737 operations, and was not presented to AA
crews in AA publications. AA flight crews are required to use the guidance in the FAA approved AA
Operating Manual.
77
Boeing 737 NG Flight Crew Training Manual, page 1.34

1.17.3 Air Traffic Control Procedures
1.17.3.1 Assignment of Active Runway
The JCAA Air Traffic Services Manual of Operations (ATS MANOPS), Amendment No. 15 to Part
7, sub-paragraph 710.1 A., valid at the time of the accident, stated:
“(The controller shall) Assign the operationally suitable runway most

nearly aligned into the wind if the wind speed is 5 knots or more.”
The Tower controller on duty at the time of the accident reported being aware of this procedure,
but that frequently pilots chose to land with a tailwind. The Tower controller reported having
coordinated with the Approach controller for AA331 to use runway 12, and that the ATIS was
stating that runway 12 was the active runway.
1.17.3.2 Weather Standby
The JCAA Air Traffic Services Manual of Operations (ATS MANOPS), Amendment No. 15 to
Part 7, sub-paragraph 703.5 states:
“(The controller shall) Place the local Airport Emergency Services on WEATHER
STANDBY if conditions are of such that landings are difficult, or difficult to
observe.”
The Airport Emergency Services were not placed on Weather Standby during the period of
AA331’s approach and landing.
1.17.3.3 Information for arriving aircraft
JCAA ATS MANOPS, Amendment 25, sub-paragraph 807.1 C stated that, in addition to
meteorological information, Approach Control shall transmit to the arriving aircraft “Current
runway surface conditions, in case of precipitants or other temporary hazards.”
Sub-paragraph 1102.2 of this MANOPS stated “If no report has been received for the runway of
intended use, issue an advisory to that effect”, and that the phraseology to be used is “Runway is
wet, no braking action reports received.” No such runway surface condition or braking action
reports were transmitted by ATC to AA331, except for the Tower controller saying “ … be advised
runway wet.” when giving landing clearance less than 5 minutes before the aircraft landed.

1.17.4 FAA Information
1.17.4.1 FAA Regulatory Requirements Regarding Landing Distance Calculation
14 Code of Federal Regulations (CFR) part 121, section 121.195(b), part 135 section 135.385(b),
and part 91, section 91.1037(b) and (c) require operators to comply with certain landing distance
requirements at the time of takeoff. These requirements limit the allowable takeoff weight to
that which would allow the airplane to land within a specified percentage of the landing distance
available on: (1) the most favorable runway at the destination airport under still air conditions; and
(2) the most suitable runway in the expected wind conditions. The performance effects of a
tailwind on landing are not necessarily taken into account for dispatch purposes, depending on
the airplane dispatch basis.
Dispatch, or factored, landing distance calculations are used during flight planning to ensure that
dispatched airplanes will be able to land safely at the intended destination airport or a planned
alternate and are based on estimated landing weights and forecast conditions. Factored landing
distances, including preflight landing safety margins, are required and standardized by U.S. and
international aviation authorities. Specifically, in accordance with 14 CFR 121.195, “Airplanes:
Turbine Engine Powered: Landing Limitations: Destination Airport,” “no person operating a turbine
engine powered large transport category airplane may take off unless the weight of the airplane on
arrival…would allow a full stop landing at the intended destination airport within 60 percent of
the effective length of each runway.”
The effective runway length (factored) is further extended by 15% if “the runways at the
destination airport may be wet or slippery at the estimated time of arrival.
1.17.4.2 History of SAFO 06012, OpSpec/MSpec C082 and AA Bulletin 737-07
The June 7, 2006 Federal Register publication provided advance notice of the FAA intent to issue
mandatory Operation Specification/Management Specification (OpSpec/MSpec) C082, N
8400.C082. This would have required all turbojet operators under Title 14 of the Code of Federal
Regulations (14 CFR) parts 121, 135, 125, and 91 subpart K to conduct landing performance
assessments (not necessarily a specific calculation) before every arrival based, in part, on planned
touchdown point, procedures and data at least as conservative as the manufacturer’s, updated wind
and runway conditions, and an additional 15 percent safety margin. However, the FAA subsequently
decided not to issue the mandatory OpSpec/MSpec C082 at that time and, in August 2006, published
Safety Alert For Operators (SAFO) 06012 as an interim guidance measure. SAFO 06012 addressed
similar issues to the proposed mandatory OpSpec/MSpec, but operator compliance with the SAFO
was, by definition, voluntary.
It is significant to note that the summary of this publication states:
“The following advance notice of policy and information would provide clarification and
guidance for all operators of turbojet aircraft for establishing operators’ methods of
ensuring that sufficient landing distance exists for safely making a full stop landing with an

acceptable safety margin, on the runway to be used, in the conditions existing at the time of
arrival, and with the deceleration means and airplane configuration to be used.” (bolding
added)
This publication stated that one example of a means of compliance would be:
“Establishment of a minimum runway length required under the worst case meteorological
and runway conditions for operator’s total fleet or fleet type that will provide runway lengths
that comply with this notice and OpSpec/MSpec C082.”
The FAA did provide a mechanism for operators to formally document full “compliance” with the
FAA recommendations in SAFO 06012. See SAFO 06012, page 1, item 2, Discussion, which states
that SAFO 06012 is based on the FAA’s policy statement published in the Federal Register on June
7, 2006, and incorporates revisions based on public comments received by the FAA. It also states
that operators may use Operation/Management specification paragraph C382 to record their
voluntary commitment to this practice, pending rulemaking.
There was evidence that AA issued Bulletin 737-07 to comply with SAFO 06012, in response to a
request from the FAA AMR Certificates Office, dated 08 September 2006. There was no evidence
of AA using OpSpec/MSpec C382 to record their voluntary commitment to this practice.
There was no evidence presented to the investigation that the "advance analysis" used by the AA331
flight crew on the accident flight had received any such "coordination" with the FAA.
Furthermore, there was no evidence as to whether or not the "advance analysis" used by the AA331
flight crew had been modified to account for the increased landing distance that resulted when the
AA B737 tailwind landing limit was increased from 10 knots to 15 knots.
1.17.4.3 FAA Recommendations Regarding Landing Distance Calculation
FAA SAFO 06012, 8/31/06, and AC No: 91-79, 11/06/07, were sent to, and were received by,
American Airlines.
The expressed purpose of SAFO 06012 was as follows:
“This SAFO urgently recommends that operators of turbojet airplanes develop

procedures for flight crews to assess landing performance based on conditions
actually existing at time of arrival, as distinct from conditions presumed at time of
dispatch. Those conditions include weather, runway conditions, the airplane’s weight,
and braking systems to be used. Once the actual landing distance is determined an
additional safety margin of at least 15% should be added to that distance.”
The expressed purpose of AC 91-79 was:
“This AC provides ways for pilots and operators of turbine-powered airplanes to

identify, understand, and mitigate risks associated with runway overruns during the

landing phase of flight. It also provides operators with detailed information that may be
used to develop company standard operating procedures to mitigate those risks.”
AC 91-79, 78 states that: “If there is rain actively falling on the runway, standing water should be
assumed”, and SAFO 06012 states that standing water constitutes a runway condition that should
be considered as contaminated.
1.17.4.4 FAA Advisory Circular AC 91-79
The following is taken from FAA Advisory Circular 91-79, 11/06/07, Appendix 4.
10. Water on the runway is the most common contaminant dealt with by the pilot, but it still
provides several challenges. Is the runway wet? Is there standing water on the runway? For
that matter, how wet is “wet”? These are questions for which no clear answers are available.
Some manufacturer’s wet and contaminated supplemental data provide the following or
similar definitions (taken from European standards) for wet and contaminated runways:
a. Wet Runway. A runway is considered as wet when there is sufficient moisture on the surface
to cause it to appear reflective, but without significant areas of standing water.
b. Runway Contaminated by Standing Water, Slush, Dry Snow, or Wet Snow.
(1) A runway is considered to be contaminated when more than 25 percent of the runway
surface area (whether in isolated areas or not) within the required length and width
being used, is covered by surface water more than 3 mm (0.125 inch) deep, or by slush or loose
snow equivalent to more than 3 mm (0.125 inch) of water.
(2) The “reflective surface” in the wet runway definition and the 3-millimeter depth over 25
percent of the runway require that the pilot be able to see the runway, since these types of
conditions are rarely reported to flight crew.
NOTE: The FAA has taken the position that a runway does not need to be reflective to be
considered wet. If a runway is contaminated or not dry, that runway is considered wet
(bolding added).
11. If there is any sort of restriction to visibility at the field, or even just low clouds, the pilot
will not have sufficient time, once the runway is in view, to ascertain the runway condition and
select the appropriate landing distance chart. This requires that the pilot gather the available
information prior to approach and make a conservative choice. For example, if there is no
clear report of runway condition, but the pilot knows rain has been in the area, that pilot
should assume the runway is at least wet. If there is rain actively falling on the runway,
standing water should be assumed. If there is any doubt, assume the most conservative
condition that requires the longest landing distance.
78
Section 1.17.1.2.7

12. Once committed to an operation on a wet or contaminated runway, the pilot should expect
a lower level of deceleration than routinely experienced on a dry runway. As an example,
consider the landing distances for wet and contaminated surfaces, assuming zero wind and
a zero runway gradient. The wind and runway gradient conditions do not require additional
corrections. Under these conditions, for the example airplane, a wet runway increases the
landing distance over a dry runway by approximately 26 percent. If standing water is present,
the landing distance increases approximately 52 percent. Remember that the contaminant
only affects the ground roll and braking. It has no impact on the air distance from 50
feet to touchdown, which is included in the landing distance. For this example, the air
distance is almost half of the total dry runway landing distance. In the presence of standing
water, for the total landing distance to increase 52 percent, the ground braking distance
increases 100 percent. This situation might surprise a pilot the first time it is encountered
because the deceleration rate on this surface will be only one-half of what the pilot might be
accustomed to. For any contaminant, the pilot should expect a relatively low deceleration rate
in the initial phase of braking. A wet runway may be less severe than other contaminants, but
the pilot must remember that any increase in total landing distance will occur entirely in the
landing ground roll braking segment."
The investigation found no evidence that any of the related information in AC91-79 was
incorporated in AA Manuals.
It should be noted that AC91-79, Appendix 4, 10 b. (2), as above, stated “If a runway is
contaminated or not dry, that runway is considered wet”, and that SAFO 06012, item 4. Definitions,
stated “i. A wet runway is one that is neither dry nor contaminated”.
1.17.4.5 SAFO 10005
SAFO 10005 was published by the FAA 3/1/10 (shortly post-accident).
The stated purpose of the SAFO was: “To recommend that all operators should provide written
policy to flight crews emphasizing that either pilot may make a go-around callout and that the
response to a go-around callout is an immediate missed approach”.
The Recommended Action was that operators should publish or reinforce existing written
policy emphasizing that:
1. Either the pilot flying or the pilot monitoring may make a go-around callout, and
2. The flying pilot’s immediate response to a go-around callout by the non-flying pilot is
execution of a missed approach.

1.17.5 NTSB Recommendations to FAA
1.17.5.1 NTSB Recommendations, General
The NTSB has an extended history of recommendations to the FAA regarding the question of
requiring flight crews to calculate landing distance for actual conditions at time of arrival, but to the
date of the publication of this report no such requirement has been put in place by the FAA. 79
1.17.5.2 NTSB Safety Recommendation, in reply to A-07-58 through -64
NTSB Safety Recommendation to the FAA, in reply to A-07-58 through -64, dated 16 October
2007, stated:
Because the FAA has not required actions to address the Board’s urgent safety
recommendation, flight crews of transport-category airplanes may still be permitted to land
in wet, slippery, or contaminated runway conditions, without performing arrival landing
distance assessments that incorporate adequate safety margins. As another winter season
approaches, the urgent need for such margins becomes more critical. The Safety Board
concludes that because landing conditions may change during a flight, preflight landing
assessments alone may not be sufficient to ensure safe stopping margins at the time of
arrival; arrival landing distance assessments would provide pilots with more accurate
information regarding the safety of landings under arrival conditions. Further, the Safety
Board concludes that although landing distance assessments incorporating a landing
distance safety margin are not required by regulation, they are critical to safe operation of
transport-category airplanes on contaminated runways. Therefore, the Safety Board
believes that the FAA should require all 14 CFR Part 121, 135, and 91 subpart K operators
to accomplish arrival landing distance assessments before every landing based on a
standardized methodology involving approved performance data, actual arrival conditions,
a means of correlating the airplane’s braking ability with runway surface conditions
using the most conservative interpretation available, and including a minimum safety
margin of 15 percent. (bolding added).
79
Section 2.21

1.18 Additional Information
1.18.1 Air Traffic Control
1.18.1.1 Enroute Radar Service
Enroute and Approach control services were co-located in the Kingston Air Traffic Control Centre
(KATCC). The evidence indicated that all of the Enroute radar and communication equipment was
serviceable at the time of the accident.
The Enroute controller was a graduate of the CAA Training Institute (CAATI) and held an ATC
license with an Area Radar Control rating. Controllers are required to complete the ATC basic
and advanced courses at the CAATI to qualify for the ATC license and then to undergo on the-
job-training for rating validation. This had been done in accordance with the JCARS Eighth
Schedule.
The Enroute controller held a valid air traffic controller license, medical certificate and a valid
radar rating. The controller’s work schedule did not reveal a pattern of fatigue.
1.18.1.2 Approach Radar Control Service
The Approach radar and the communication equipment were serviceable. The Approach controller
held a valid air traffic controller license, medical certificate and a valid Approach Radar rating. The
controller’s work schedule did not reveal a pattern of fatigue.
Runway assignment for IFR traffic required coordination between the Approach controller and the
Aerodrome controller. Runway 12 was often used by incoming traffic in light tailwinds due to its
having an ILS approach, and when AA331 was assigned to runway 12 the wind was from the north
at seven knots.
There was a letter of agreement between Approach and Aerodrome control units for operational co-
ordination and the exchange of information. This agreement also required the Aerodrome controller
to keep Approach units informed of aerodrome conditions.
1.18.1.3 Aerodrome Control Service
The evidence indicates that all control tower equipment was serviceable including the
communication and runway lighting controls. The Tower controller held a valid air traffic
controller license, medical certificate and a valid Tower controller rating. The controller’s work
schedule did not reveal a pattern of fatigue.
None of the controllers issued a runway surface condition report or braking action report to
AA331. There were no reports available to the controllers to pass since none had been reported to
ATC. However, the Tower controller did advise AA331 at 21:17 EST, “Be advised runway wet”.

On initial contact with an aircraft the Enroute controller and/or the Tower controller was required to
pass information to arriving aircraft concerning runway surface condition reports and braking action
reports or had to advise the aircraft that none are available. No reports existed and the controllers did
not advise AA331 that no reports were available.
1.18.2 JCAA ATS QA Audits, and Controller Training and Checking
At the time of the accident JCAA ATS had no consistent system of QA audits, no proficiency
checks and no recurrent training for controllers.
1.18.3 JCAA Automatic Terminal Information Service (ATIS)
There was no record to show that the computer in the tower that d i s s e m i n a t e d the ATIS
message was unserviceable. There was no record that an ATIS message was generated or
broadcast. The CVR of AA331 also did not capture an ATIS broadcast. The ATIS message was not
recorded by ATS.
There was no requirement in ICAO Annex 11, Chapter 6, 6.2.2 “Communications between air
traffic services units and other units,” for ATIS messages to be recorded, nor was there any
requirement in the Jamaican Civil Aviation Regulations (JCARs) nor any mention of this in the
Air Traffic Services Planning Manual (Doc 9426).
The first officer in his interview reported that he did not recall what the weather was on the ATIS.
The captain stated that he did not onl y rely on ATIS reports, and for this flight he called the
Approach controller for the airport conditions. The flight crew did not make any other mention of
the ATIS in their interviews, and there was no mention of the ATIS in the recordings of the CVR
and the ATC transmissions.
The Tower controller, w h e n interviewed, stated that the ATIS broadcast in effect at the time
of the accident was announcing that runway 12 was active.
1.18.4 JCAA Notices To Airmen (NOTAMS)
The following information was included in NOTAMS which were in effect for the
Norman Manley International Airport at Kingston, and the Sangster International Airport at
Montego Bay on 22 December 2009, at the time of the accident.

Norman Manley International Airport

------------------------------------------------------
 - RUNWAY -
 -- APCH LIGHTS RUNWAY 12 U/S. CAUTION ADZ.
 30NOV09/1201 31JAN10/1200 MKJP A0176/09
 - FACILITIES -
 -- WORKMEN AND EQPT WILL OPR AT A DISTANCE OF 60M FROM THR RUNWAY
 12,TO A DISTANCE OF 420M IN THE KINGSTON HARBOUR. CTN ADZ.
 01DEC09/1201 31JAN10/1200 MKJP A0184/09
 MBJ NO/
This indicated that the approach lighting system serving runway 12 at Kingston was
unserviceable, and that work was being carried out on the system.
Sangster International Airport
1. The NOTAM announcing the planned runway closure at Montego Bay was as follows:
(A0208/09 NOTAMN
Q)MKJK/QMRLC/IV/NBO/AE/000/999/182956.21N0775527.3W
A)MKJS
B)0912230300
C)0912231100
E)RUNWAY 07/25 CLSD TO FACILITATE REPAIRS.)
Sent: Dec. 19, 2009
This announced the runway was to be closed from 03:00 UTC to 1100 UTC on 23 December 2009
(Accident occurred at 03:22 UTC, on 23 December 2009).
This NOTAM was sent to Aeronautical Fixed Telecommunication Network (AFTN) at 23:42
UTC on 19 December 2009, and received at 2358 UTC.
2. The NOTAM cancelling the planned runway closure was as follows:
(A0215/09 NOTAMC A0208/09

Q)MKJK/QMRXX//////
A)MKJS
E)SCHEDULED RUNWAY REPAIRS CANCELLED.)
Sent: Dec. 22, 2009
This announced that the scheduled runway repairs were cancelled.
This NOTAM was sent to AFTN at 21:23 UTC on 22 December 2009, and received at 21:30 UTC.
The evidence collected during the investigation indicates that the NOTAM “AO208/09 NOTAMN”,
which announced the planned closure of runway 07/25 at Montego Bay, was received by AA
Dispatch.
The evidence also indicated that the NOTAM “AO215/09 NOTAMC AO208/09”, announcing the
cancellation of this runway closure, was never delivered to the flight crew of AA331, even though it
was issued several hours before the flight’s departure.
AA researched the matter, and reported that it was highly likely that NOTAM AO215/09
successfully reached AA Dispatch, and that this would have resulted in the original NOTAM
AO208/09 being removed from the system at approximately 21:41 UTC on 12/22/09.
AA stated,
“Note - The AA NOTAM system 'NIMS" does not alert on NOTAM cancellations. It just
removed the NOTAM from the flight planning system in response to a cancellation message.
Thus there would not be an easy way for the dispatcher to alert the pilot that a NOTAM has
been removed.”
1.18.5 JCAA Air Traffic Control Radar
The evidence indicated that the Kingston air traffic Area (Enroute) and Approach control radars
were functioning normally during the approach and landing of AA331.
1.18.6 Estimated Water Depth and Braking Action
The Performance section of the investigation determined that the depth of water on MKJP
runway 12 during the flight 331 ground roll can be reliably estimated using the following
independent methods:
1. Empirically-based models for pavement water drainage rates that rely on factual data
documented by the investigation, including MKJP precipitation rates and runway 12/30
pavement properties.
2. Existing B737NG wet runway flight test data.
3. Available AA flight 331 FDR data, B737-800 aerodynamics and engine simulation models,
and models for airplane/ runway braking action or runway surface contamination required to
match the event stopping performance during the ground roll segment on an improved surface.
The depth of water on MKJP runway 12 during the flight 331 ground roll can be estimated using
empirically-based models to calculate water depth as a function of rainfall intensity and the runway
12/30 width, transverse slope, and pavement macro-texture characteristics. Based on the NASA
rainfall rate/flooding exposure model and the Texas Transportation Institute (TTI)
accumulated water depth model data presented to the JCAA at the Aircraft Performance
Group Briefing in August 2010, no evidence supports a runway 12/30 water depth accumulation
close to 3 millimeters (mm) of water.
These models indicate that, depending on the pavement macro-texture and airplane lateral
position, there was at most a 0.5 to 1.0 mm water depth in the calculated main gear braked

wheel path. A similar conclusion holds (at most 0.75 to 1.5 mm water depth) for all possible
pavement macro-textures, even if the measured rainfall rate at the time of the accident is doubled to
1.0 inch/hour and the calculated airplane lateral position is artificially and significantly biased toward
the runway shoulder.
An independent Boeing analysis of the existing rainfall precipitation and runway 12/30 factual data is
consistent with the TTI model results described here.
The NASA, TTI, and Boeing pavement drainage model results do not incorporate potential runway
edge water damming, surface wind influence on water drainage, and/or airport infield water-ponding
effects. However, based on the qualitative FAA inspection of runway 12/30 (during the runway 12/30
visit in August 2010 in light rain conditions) and the absence (to date) of compelling JCAA or U.S.
Team data to the contrary, runway edge water damming and/or airport infield water ponding effects on
water depth in the main gear braked wheel paths for a B737- 800 are believed to be negligible The
runway 12/30 transverse slope data indicate that the outboard runway edge lines and the runway
shoulder edges lie, on average, about 8 inches and at least 11 inches below the runway centerline
elevation, respectively.
The estimated depth of water in the airplane braked wheel paths during the runway 12 ground
roll was less than 3 mm (less than .125 in.), which corresponds to an equivalent stopping
performance level better than that expected for either a runway covered with standing water or a
flooded runway. From an aircraft performance perspective, standing water and flooded are
synonymous terms used to describe surface water depth greater than 3 mm (.125 in) on more than
25 percent of the runway surface area (whether in isolated areas or not) within the required
runway length and width being used.
Independent of the above models for estimating water depth, the airplane deceleration observed
during the flight 331 ground roll is consistent with B 737NG (-700/900) flight test data on an
artificially wet (no active rainfall; water applied to pavement prior to airplane touchdown using a
fleet of tanker trucks), un-grooved runway with pavement macro-texture characteristics similar to
or better than MKJP runway 12. These Boeing flight test data benefit from explicit knowledge that
less than 25 percent of the runway surface was subject to standing water 1/8 inch deep or greater.
In other words, the AA 331 braking performance level has been achieved on a wet, un-grooved
runway with no active rainfall and less than 1/8 inch depth of water.
Analysis of the AA flight 331 event indicates the achieved airplane braking coefficient value
was less than that traditionally considered to be associated with a wet runway (0.2), but better than
what Boeing associates with a flooded runway (0.05). The factual evidence indicates that the
airplane/ runway stopping performance interaction for AA flight 331 was essentially consistent
with AA Fair/Medium braking action.
In other words, the AA flight 331 stopping performance level is inconsistent with the Boeing
model for standing water or flooded runway.
Quoted performance i s b a s e d o n Wet/Good braking action, maximum manual braking, without

reverse thrust and landing at about 1,000 feet from the threshold.

1.18.7 NTSB Review of Fatigue in Major U.S. Accidents
The following information was obtained from the NTSB Accident Report NTSB/AAR-08/02,
PB2008-910402, Runway Overrun During Landing, Pinnacle Airlines Flight 4712, Traverse
City, Michigan, April 12, 2007:
“Research has shown that long duty days can be associated with pilot fatigue and
degraded performance. Aviation accident data show that human-performance-related
airline accidents are more likely to happen when pilots work long days.
The Board’s 1994 study of flight-crew-related major aviation accidents found that
captains who had been awake for more than about 12 hours made significantly more
errors than those who had been awake for fewer than 12 hours. Such errors included
failing to recognize and discontinue a flawed approach; pilots often exhibited a
tendency to continue the approach despite increasing evidence that it should be
discontinued … Research and accident history also show that fatigue can cause
pilots to make risky, impulsive decisions; become fixated on one aspect of a
situation; and react slowly to warnings or signs—any of which can result in an
approach being continued despite evidence that it should be discontinued.
Additionally, research shows that people who are fatigued become less able to
consider options and are more likely to become fixated on a course of action or a
desired outcome.”
1.18.8 AA use of FAA Safety Recommendations in AC No: 91-79
The investigation was advised that FAA Advisory Circular AC No: 91-79, dated 11/06/07 was
provided to the AA Safety Officer by the FAA, who, in turn, provided the AC to the various AA
fleet captains. The investigation was unable to determine to what degree the guidan ce
in the AC had been incorporated into AA manuals.
1.18.9 GPS Approach Requirements and Information
AA held Operations Specifications C300 and C384, which authorized the conduct of required
navigation performance (RNP) instrument approach procedures (IAP). Both flight crew were
qualified to conduct RNAV approaches. The Jeppesen chart (12-2) for MKJP/KIN, Kingston,
Jamaica, RNAV (GPS) Rwy 30, dated 22 August 08, was in the possession of the AA331 flight
crew during the flight.
The AA331 flight plan indicated that the aircraft was RNAV capable. The dispatch paperwork
showed no MEL restrictions on the aircraft’s RNAV capability. The Kingston RNAV (GPS) Rwy 30
approach was in the database of the aircraft’s navigation equipment. 80
80
See also Section 1.6.12

1.18.10 Stopping Performance Information
During type certification of the 737-800, Boeing demonstrated landing and stopping performance
tests on dry runways in tailwind conditions up to 31 knots, using maximum wheel braking and no
reverse thrust. No data was demonstrated for wet or contaminated runways, and the FAR 25 safety
factor method (121.195) was used to certify the aircraft flight manual performance for wet and
contaminated runways.
The wet or slippery runway landing distance requirements of 14 CFR 121.195(d) are based on
applying a safety factor to demonstrated dry runway distances rather than demonstrating wet
runway landing distances. Stopping performance data for the AA Boeing 737-800 was issued on a
“Landing Data Card” 81 for Dry, Wet/Good, Fair/Medium and Poor runways, with corrections for
aircraft weight, flap setting, airport elevation, headwinds and tailwinds, and reverse thrust.
The Boeing 737-800 aircraft was certificated per AC 25-7A Change 1 (FAA Flight Test Guide) for
15 knot tailwind landings and a bulletin (SPC MSG NBR 9482, see 1.17.1.2.4) had been issued by
AA to all flight crews removing the company’s previous 10 knot tailwind limitation and
increasing it to 15 knots for all aerodromes.
1.18.11 Runway Water Depth Measurement
The following information was obtained from the ICAO Airport Services Manual (Doc 9137),
Part 2, Pavement Surface Conditions:-
2.1.10 There has been some speculation on whether measuring water depth could
perhaps replace measuring runway friction. To this end, a study was undertaken to ascertain a
list of requirements to be met by such measuring devices. [. . .] Although possible [to design],
it would not be practical to develop a device that could meet all of the [. . .] requirements; it is
preferable to develop programmes aimed at improving the surface texture and drainage of
runway rather than measuring the water depth. [. . .] Even assuming that a device meeting [all
of the] requirements [was to be] developed, another big difficulty appears to be the number
and location of devices needed for a runway. [At the time of writing], it has been concluded
that standardization of water depth measuring devices with the object of measuring runway
friction is not practical. Work continues in this area.”
81
Appendix 7

1.18.12 Hazards associated with tailwind landings.
A paper entitled “Safety aspects of tailwind landings” was published by G.W.H van Es and A.K.
Karwal, National Aerospace Laboratory of the Netherlands, in January 2001.
This study discussed the hazards associated with tailwind landings, as follows:
1. On glide slope, excessive rates of descent (that is, more than 1,000 fpm) can be
required when conducting a descent in a strong tailwind. Flap speed limits may be
exceeded.
2. Increased ground speed may result in approach becoming de-stabilized or rushed, and
increases pilot workload.
3. On glide-slope, low engine power settings resulting from high rates of descent increase
spool-up time if go-around is necessary.
4. Increased landing groundspeed results in pilots tending to bleed off speed by floating the
aircraft before touchdown, thus using up runway available.
5. Close to ground, tailwind tends to decrease causing increase in True Air Speed (inertial
effect), thus amplifying floating.
6. FMS is unreliable indicator of tailwind, has 2 – 3 second delay, and is measured only
where aircraft actually is.
7. Combination of tailwind and wet runway results in high risk of runway over run.
8. Wake vortices patterns from a tailwind landing differ from those in calm or headwind
conditions.
1.18.13 Hazards Associated With Go-Around After Touchdown
Although go around after landing was a hazardous procedure, the option for the flight crew to do so
was open until such time as reverse thrust was selected, and the captain reported that he considered
this when the aircraft was on the runway.
The hazards associated with go-around after touch-down are:
1. Asymmetric engine spool-up,

2. Speed bleed-off,
3. Time for re-set of flap and trim,
4. Possible insufficient runway remaining,
5. Loss of directional control.
See also 1.17.4.5, SAFO 10005, regarding go-around call out procedures.

1.18.14 Flight Safety International Approach and Landing Accident Reduction Report
Approach and landing accident reduction (ALAR) has long been among the primary goals of the
Flight Safety Foundation (FSF). When the international FSF ALAR Task Force published its report
in 1998, it cited data showing that an average of 17 fatal ALAs had occurred each year from 1980
through 1998 in passenger and cargo operations involving aircraft weighing 5,700 kg/12,500 lb. or
more.
The task force’s work, and the subsequent safety products and international workshops on the
subject, have helped reduce the risk of ALAs, but the accidents still occur. In 2009, of 17 major
accidents, nine were ALAs, compared with 19 and eight the previous year.
Since the ALAR campaign began, members of the FSF Controlled Flight Into Terrain and Approach
and Landing Action Group (CAAG) have conducted numerous ALAR workshops around the world,
and the Foundation has distributed more than 40,000 copies of the FSF ALAR Tool Kit — a unique
set of pilot briefing notes, videos, presentations, risk-awareness checklists and other products
designed to prevent approach and landing accidents.
A major update of the FSF ALAR Tool Kit — featuring the findings of analyses of recent
accident data, as well as the data-driven findings of the FSF Runway Safety Initiative — was
issued in 2010. It is the Foundation’s intention to periodically update the ALAR Tool Kit to include
new information aimed at reducing the risk of approach and landing accidents.
1.18.15 Boeing 737-800 Tailwind Certification Process
The FAA certified the Boeing 737-800 series aircraft for operations in tailwinds of 10 knots or less,
and 15 knots or less, and operators could purchase the aircraft with a performance package for either
a 10 or 15 knot tailwind limitation. AC 25-7A, Change 1, which was active guidance material at the
time of the certification described an acceptable method for the manufacturer to comply with the
requirements, including flight testing. U.S. operators wishing to use either package had to be
approved by the FAA.
1.18.16 Definition of Black Hole Approach
In the article by A. Howard Hasbrook, “The Black Hole Approach: Don’t Get Sucked In!”,
published in Business and Commercial Aviation Magazine, August 1971, there is the following
definition:
“ … a black hole approach is a long, straight-in approach at night to a brightly lit runway
over featureless and unlit terrain”.

153
Appendices to Factual Information
APPENDICES TO FACTUAL INFORMATION

154

155
APPENDIX 1
AA331 Flight Data Recorder Plot by The Boeing Company

156

157

158

159

160

161

162

163
APPENDIX 2
Cockpit Voice Recorder Transcript

164
The following is a transcript of the L-3 Communications FA 2100-1010 solid-state cockpit voice recorder
(CVR), serial number 000142599, installed on an American Airlines Boeing 737-800 (N977AN), which
overran runway 12 at Norman Manley International Airport in Kingston, Jamaica, on 22 December 2009.
LEGEND
CAM Cockpit area microphone voice or sound source
HOT Flight crew audio panel voice or sound source
RDO Radio transmissions from N977AN
APR Radio transmission from the Kingston approach controller
TWR Radio transmission from the Manley airport tower controller
TAWS Terrain avoidance and warning system sound source
PA Aircraft public address system sound source
-1 Voice identified as the captain
-2 Voice identified as the first officer
-3 Voice identified as the flight attendant
-? Voice unidentified
-A Identified as first Kingston approach frequency
-B Identified as second Kingston approach frequency
* Unintelligible word
# Expletive
@ Non-pertinent word
( ) Questionable insertion
[ ] Editorial insertion
Note 1: Times are expressed in local EST
Note 2: Generally, only radio transmissions to and from the accident aircraft were transcribed.
Note 3: Words shown with excess vowels, letters, or drawn out syllables are a phonetic representation of the words as
spoken.
Note 4: A non-pertinent word, where noted, refers to a word not directly related to the operation, control or condition of the
aircraft.

165
21:51:44.0
START OF RECORDING
START OF TRANSCRIPT
21:51:54.5
HOT-1 so now uh if we have to go to an alternate it's going
to be Grand Cayman instead of Montego Bay. if we go
to Montego Bay...
21:52:10.6
CAM-2 Grand Cayman it's gonna be a little bit less. it's gonna
be lot more fuel.
21:52:11.2
HOT-1 ...yeah to go to Grand Cayman it's gonna be a lot
more gas. we need to do that at ten thousand two
hundred pounds. so missed approach right to Grand
Cayman.
21:52:22.3
HOT-2 alright.
21:52:24.1
HOT-1 here we go. startin' on down.
21:52:36.1
HOT-1 be off for just a second.
21:52:40.5
PA-1 well ladies and gentlemen we've started our initial
descent into the Kingston area. uh it's gonna be quite
a bumpy ride on the descent here. other aircraft up
ahead of us have been telling us they've been uh
shouldn't be much worse than this but it's gonna be a
little bit too hard for the flight attendant's to
continue their service so I've asked them to take their
seats and buckle in for the remainder of the flight. we
are gonna touch down about twenty past the hour.
and the current weather they're having uh rain
shower activity at the airport at this time. and uh it's
still uh the temperature's right around uh seventy
three degrees. like to take this time once again and
thank you for choosing American and we hope to see
you again on future American flights.
21:53:23.5
HOT-1 anytime that # window's open and you hit that it goes
right for speed.
21:53:36.4
CAM-2 here goes the recall. watch your eyes there.

166
21:53:38.7
HOT-1 alright.
21:53:40.7
CAM-2 still working fine.
21:53:47.4
CAM-2 oh we talked about the approach. approach brief.
21:53:50.4
HOT-1 complete for now?
21:53:52.1
CAM-2 complete for now.
21:54:17.0
HOT-1 why don't you pull up the Grand Cayman weather.
21:54:19.2
CAM-2 check on that.
21:54:36.4
HOT-1 it's probably going to go by the identifier which is...
21:54:39.6
CAM-2 I've got it up here somewhere...M-W-C-R.
21:55:11.3
RDO-1 Kingston Approach American three three one.
21:55:27.3
RDO-1 Kingston Approach American three three one.
21:55:38.0
HOT-1 and I heard after TOTON direct KEYNO correct?
21:55:40.2
CAM-2 that's correct.
21:55:52.4
HOT-1 alright there's TOTON direct KEYNO.
21:55:55.8
CAM-2 that looks good.
21:56:06.8
HOT-2 unlimited visibility scattered at one thousand six
hundred scattered at nine thousand no sig one
twenty at nine. weather there looks nice.

167
21:56:22.1
HOT-1 alright.
21:56:28.2
HOT-1 grid MORA's coming in here. are five thousand to the
uh west and ninety eight hundred to the east.
21:56:40.8
HOT-2 alright
21:56:41.5
HOT-1 if we have to deviate for weather. so anything to the
east ten thousand feet is good.
21:56:48.3
HOT-2 sounds good.
21:57:03.3
HOT-1 bring the heat on for me.
21:57:14.6
HOT-1 # been doing this # for the last three trips man.
21:57:23.0
HOT-1 kill those strobes.
21:57:25.6
CAM-2 say again?
21:57:27.0
HOT-1 kill the strobes. just put it on steady or something.
21:58:14.8
HOT-1 trying to get a hold of approach to see uh if they're
having any trouble on the approach but there might
not be anybody in there.
21:58:23.1
RDO-1 Kingston Approach this is American three three one.
check
21:58:30.8
APP-A American three three one Manley Radar.
21:58:33.2
RDO-1 yes sir we're still presently one hundred and twelve
nautical miles from the airport understand that you
have moderate rain shower on the approach. just
wondering if any aircraft have been uh experiencing
turbulence on the approach.

168
21:58:44.3
APP-A American three three one none reported.
21:58:47.5
RDO-1 none reported. have you had anybody land in the last
hour?
21:58:52.3
APP-A okay stand by. let me just tell you the last aircraft that
landed at the time...affirm we had one aircraft that
landed in the la— at uh zero two zero three and he
came in from Montego Bay. he came in from the the
northwest and uh he didn't have any problems
coming in. didn't have anybody from the north.
21:59:18.0
RDO-1 okay thank you and uh we'll talk to you in a few
minutes. we're about a hundred uh and ten miles
from the airport.
21:59:22.4
APP-A uh roger.
21:59:25.5
HOT-1 heh. he said none reported but then I said 'have you
had any airplanes land in the last hour' and he goes
'uh let me check.' [sound of laughter]
21:59:33.8
HOT-1 [sound of laughter]
21:59:35.8
HOT-1 no wonder there's none reported. they had one
airplane landed at thirty past. came in from Montego
Bay.
21:59:44.5
HOT-2 so we got that going for us.
21:59:45.8
HOT-2 yeah.
21:59:48.3
HOT-1 one night I was doing— I was doing three approaches
over top of Managua. Taca's up ahead of me you
know. he— Taca finally lands. I'm coming in I say
'Taca when did you break out' and the guy goes 'ooy
the weather's no good it— it's okay though it's no
good oh it's okay.' I went 'no what altitude you break
out at?' he goes 'oh it's okay.' [sound of laughter]
[sound of gutteral noise] third go around you know.

169
22:00:12.8
HOT-2 I uh had the same thing on my I-O-E on the seven two
going into uh San Pedro Sula with one of these Dallas
guys he had no idea what was going on out there.
22:00:22.4
HOT-1 yeah he was—.
22:00:24.0
HOT-2 he was lost.
22:00:24.4
HOT-1 working hard.
22:00:25.6
HOT-2 so we were like— I was like you know I saw there was
like a low pressure system moving into the area
before we left Miami. figured we'd get there ahead of
it. so I asked the guy. said 'you know what uh find out
what the field condition. we're getting bounced all
over the place.'
22:00:43.9
HOT-1 yeah.
22:00:44.5
HOT-2 you know it's one of those things where you got to do
a teardrop entry alignment pattern.
22:00:48.7
HOT-1 oh yeah you got to come back all around. nasty.
22:00:53.0
HOT-2 yeah bounced around in the seven two and it was like
uh so he says 'what are your field conditions?' he says
'well it's some rain. it's some clouds. it's okay.' [sound
of laughter]
22:01:01.0
HOT-1 [sound of laughter] some rain. [sound of laughter]
weather's okay.
22:01:08.0
HOT-2 so we did two approaches in there and ended up in
San Salvador.
22:01:12.8
HOT-1 nice.
22:01:14.0
PA-3 ladies and gentlemen please remain in your—.

170
22:02:24.3
HOT-1 that's some good rain shower man sitting down here.
22:02:30.4
HOT-2 yeah go figure.
22:02:42.6
INT-1 hey I just want to make sure everybody stays seated
for the rest of the flight okay...okay thank
you...alrighty alrighty...bye.
22:03:26.1
APP-A American three three one call approach one twenty
point six.
22:03:29.8
RDO-2 good day American three three one.
22:03:33.2
HOT-1 I'm gonna keep it slow as we get into this stuff.
22:03:35.0
HOT-2 yeah.
22:03:41.4
RDO-2 uh Kingston good evening American three three one.
we're out of one niner zero for one five thousand.
22:03:59.0
APP-B American three three one Manley Radar.
22:04:01.9
RDO-2 Manley American three three one one eight zero for
one five thousand.
22:04:06.5
APP-B American three three one Manley Radar. expect I-L-S
approach runway one two. Q-N-H one zero one four
descend and maintain five thousand.
22:04:17.5
RDO-2 descend and maintain uh five thousand American uh
three three one.
22:04:21.1
HOT-1 down to five thousand.
22:04:21.6
APP-B American three three one affirm. be advised the
information that was given to you by the onward
controller is still the same. visibility five miles and
there is moderate rain at the station.

171
22:04:34.0
RDO-2 five miles with rain American uh three three one.
22:04:37.7
APP-B American three three one you may have to circle to
land. the wind uh three two zero degrees at one zero
knots.
22:04:44.2
HOT-1 that's still good. we'll still land straight in.
22:04:47.4
RDO-2 understand that uh that— we can uh go ahead and uh
take a straight in with that American three three one.
22:04:53.5
APP-B American three three one roger copy.
22:05:07.9
HOT-1 alright once you get above ten you can secure the
heat.
22:05:13.3
HOT-2 alrighty.
22:05:14.1
HOT-1 and down to five thousand feet now.
22:05:18.0
HOT-2 five thousand set...got twelve degrees engine heat is
off.
22:05:23.5
HOT-1 thank you.
22:05:32.7
HOT-2 and I don't know if we said this. transition level
altimetres?
22:05:35.7
HOT-1 yes sir I got ten fourteen ten fourteen ten fourteen
and lights are on.
22:05:39.3
HOT-2 alright.
22:05:41.3
PA-1 flight attendants prepare for landing.
22:05:44.2
HOT-2 P-A?

172
22:05:45.4
HOT-1 is complete.
22:05:46.7
HOT-2 before landing checklist.
22:05:47.9
HOT-? *.
22:06:02.5
HOT-1 okay I see yellow but I don't see any red.
22:06:04.9
HOT-1 how's that temperature look? I mean the uh one forty
forty four?
22:06:08.1
HOT-2 okay we're looking good.
22:08:05.3
HOT-2 how about radios and displays?
22:08:08.9
HOT-1 I-M-L-Y one two zero inbound got uh one forty eight
for ref. keep an eye on that...seven hundred pounds
here. we should be alright. two seventy eight and ten
fourteen.
22:08:21.4
HOT-2 very close.
22:08:24.8
HOT-2 alright set and checked.
22:08:26.1
HOT-1 reset and crosschecked.
22:08:28.8
HOT-2 set and crosschecked.
22:08:41.4
HOT-1 a little more drag.
22:09:40.6
HOT-1 having fun Peter?
22:09:43.2
HOT-2 enjoying it. it's been— like you said it's been one of
those um— figure they paid for an airplane ride.
22:09:52.5
HOT-1 well they're gettin' one tonight.

173
22:09:52.9
HOT-2 they're complaining about all this stuff before. when
they get there they'll be happy now.
22:09:57.2
HOT-1 yeah.
22:09:58.1
HOT-2 out of ten.
22:10:00.4
HOT-1 I just wish I was making about two hundred and fifty
thousand dollars a year to do this every time I come
to work you know.
22:10:05.1
HOT-2 exactly.
22:10:44.0
HOT-2 HUD?
22:10:45.6
HOT-1 HUD is set...think it is.
22:11:22.0
HOT-1 when we get down here to five thousand we're gonna
be limited on our—.
22:11:26.0
HOT-2 I'm sorry?
22:11:26.3
HOT-1 when we get down here to five thousand we're gonna
be limited on our turns.
22:11:27.7
HOT-2 yeah...right.
22:11:30.7
HOT-1 gonna have to keep it nice and slow.
22:12:08.9
HOT-1 that thing's gonna come— bring us in offset and I'm
gonna have to fly the instruments all the way down.
so when we break out the runway's gonna be to the
left.
22:12:18.6
HOT-2 okay.
22:12:25.5
HOT-1 one to go.

174
22:12:27.9
HOT [sound of altitude alert]
22:12:30.6
HOT-2 six for five.
22:13:08.8
HOT-1 looks like it's just on the other end of it. on the other
side of it it's clear.
22:13:11.6
HOT-2 yeah...exactly.
22:13:14.1
HOT-1 that's # up.
22:13:50.7
HOT-1 how are we looking on that weight?
22:13:56.0
HOT-1 *.
22:13:56.1
HOT-2 five hundred pounds to go.
22:14:00.1
HOT-2 looks like you're gonna— you're gonna make it but uh
showing we're gonna burn uh six hundred so we're
not going to make it by much.
22:14:10.3
HOT-1 yeah.
22:14:16.6
HOT-1 wanna ask this guy after KEYNO if we're cleared for
the approach. he's never said anything to us.
22:14:22.5
HOT-2 alright.
22:14:34.7
APP-B American three three one descend and maintain four
thousand. at KEYNO cleared straight in I-L-S approach
runway one two. be advised wind now three two zero
degrees at one five knots.
22:14:44.8
HOT-1 okay.

175
22:14:47.4
RDO-2 okay after uh— descend down to four thousand now
and uh after KEYNO cleared for the I-L-S runway one
two American three three one.
22:14:56.7
APP-B American three three one uh affirm and uh did you
copy the wind three two zero degrees at one four
knots. are you able to still land uh make a straight in
approach runway one two?
22:15:07.0
HOT-1 yes sir. flaps one.
22:15:07.7
RDO-2 we copy the wind and we can go straight in to one
two.
22:15:11.4
22:15:12.5
APP-B roger.
22:15:13.6
HOT-2 below two forty flaps one. let me uh—.
22:15:18.5
CAM [sound similar to trim-in-motion]
22:15:22.7
HOT-1 let's bring it flaps five.
22:15:26.0
HOT-2 still below two forty going to flaps five.
22:15:26.7
CAM [sound similar to flap handle movement]
22:15:28.3
22:15:35.9
HOT-1 can't even see # with that— all that radar in there.
22:15:37.7
22:15:58.8

176
22:16:10.3
HOT-2 four thousand feet set uh...two thousand if you want
to do a V-NAV now.
22:16:17.2
HOT-1 yeah you can set two thousand in there
22:16:33.4
22:16:41.5
HOT-1 go flaps ten.
22:16:44.2
HOT-2 below two ten flaps ten. CIBUG at twenty eight—
above twenty eight hundred.
22:16:45.4
22:16:48.9
HOT-1 twenty eight hundred?
22:16:49.9
HOT-2 yeah.
22:16:50.1
HOT-1 uh get on em.
22:16:50.4
HOT-2 uh af— if you want to go L-NAV I—.
22:16:52.1
HOT-1 let's go twenty eight hundred until I get to it.
22:16:54.0
HOT-2 yeah.
22:16:56.8
HOT-2 alright.
22:16:56.8
HOT-1 it's not gonna pick up the approach just yet.
22:16:57.7
22:17:14.5
APP-B American three three one contact the tower one one
eight decimal six five good day.

177
22:17:14.8
22:17:19.2
RDO-2 eighteen sixty five good night American three three
one.
22:17:25.0
TAWS twenty five hundred.
22:17:31.6
RDO-2 Manley Tower good evening American three three
one's with you. we are level inbound on the ILS
runway one two. level twenty eight hundred.
22:17:41.6
TWR American three three one wind three two zero
degrees one two knots. confirm still requesting
runway one two.
22:17:42.7
22:17:47.0
HOT-1 yes ma'am.
22:17:48.0
RDO-2 uh that's affirmative. uh say again the wind.
22:17:48.2
22:17:50.8
HOT-1 what'd she say this last time?
22:17:52.0
TWR three two zero now at one four knots.
22:17:54.3
HOT-1 that's good.
22:17:55.2
RDO-2 that's affirmative uh runway one two.
22:17:57.5
TWR American three three one clear to land runway one
two. be advised runway wet.
22:18:01.2
RDO-2 thank you.

178
22:18:03.2
HOT-2 runway's wet you want to go to brakes three
perhaps?
22:18:06.2
HOT-1 yeah let's do that. brakes three.
22:18:07.4
HOT-2 little tailwind.
22:18:09.1
HOT-1 yup localizer capture.
22:18:25.8
HOT-2 alright we're two hundred pounds above— one
hundred pounds above the uh landing weight. so
we're looking not too bad.
22:18:32.5
HOT-1 s'go gear down. flaps fifteen.
22:18:35.3
22:18:36.4
CAM [sound similar to landing gear deployment]
22:18:37.5
22:18:43.4
HOT-2 below two hundred knots flaps fifteen. speedbrakes?
22:18:46.3
HOT-1 speedbrake is armed with a green light.
22:18:47.5
HOT-2 armed green light.
22:18:48.9
22:18:54.0
TAWS twenty five hundred.
22:18:55.4
HOT-1 let's go flaps twenty five.
22:18:57.0

179
22:18:58.1
HOT-1 trying to burn up a little bit of it.
22:18:58.4
22:18:58.9
HOT-2 below one ninety flaps twenty five.
22:19:02.6
HOT-1 and flaps thirty.
22:19:04.5
HOT-2 below one seventy five.
22:19:04.8
HOT-1 set missed approach altitude.
22:19:06.0
22:19:07.9
HOT-2 missed approach altitude is set three thousand.
22:19:11.9
22:19:12.6
HOT-1 landing gear down and green.
22:19:15.8
HOT-2 down and green. flaps?
22:19:20.0
HOT-1 thirty thirty and green light.
22:19:21.8
HOT-2 thirty thirty green light.
22:19:32.8
HOT-2 alright...before landing checklist is complete.
22:19:40.5
HOT-2 I can give you the wipers if you like.
22:19:50.1
HOT-2 I've got ground contact.
22:19:53.6
HOT-1 no runway though.

180
22:19:54.9
HOT-2 no runway yet.
22:19:56.5
22:20:00.6
CAM [sound similar to windscreen wipers]
22:20:26.4
22:20:32.1
TAWS one thousand.
22:20:32.8
HOT-1 let me know when you see it Pedro.
22:20:34.0
HOT-2 alright um before landing checklist is complete.
22:20:37.6
HOT-1 alright.
22:20:39.8
RDO-2 and American uh three three one we're on a three
mile final.
22:20:46.0
TWR American three three one landing clearance still valid.
cleared to land runway one two.
22:20:49.9
RDO-2 thank you.
22:20:50.5
HOT-2 just wanted to check. cleared to land she says.
22:20:55.5
HOT-2 runway's in sight.
22:20:56.7
HOT-1 alright.
22:21:00.2
HOT-1 autopilot's comin' off.
22:21:00.8
HOT [sound similar to autopilot disconnect warning]

181
22:21:07.5
TAWS five hundred.
22:21:08.3
HOT-2 you're on speed sinking eight.
22:21:16.4
TAWS plus hundred.
22:21:24.8
TAWS minimums.
22:21:26.6
HOT-2 on speed sinking eight.
22:21:27.1
22:21:38.8
TAWS one hundred.
22:21:43.0
TAWS fifty.
22:21:44.6
TAWS forty.
22:21:46.0
CAM [sound of click]
22:21:46.5
TAWS thirty.
22:21:48.9
TAWS twenty.
22:21:52.4
TAWS ten.
22:21:56.0
CAM [sound similar to speedbrake motor]
22:21:56.1
CAM [sound of three clicks]
22:21:57.2
HOT-2 deployed.
22:21:58.2
CAM [sound similar to nose gear touchdown]

182
22:22:00.4
CAM [sound similar to thrust reversers]
22:22:06.8
HOT-2 autobrakes are off.
22:22:11.7
HOT-1 we're not stopping man.
22:22:15.3
HOT-1 come on baby.
22:22:17.7
HOT-1 aw #.
22:22:18.3
HOT-2 #.
22:22:19.3
HOT-? [sound of grunt]
22:22:20.4
HOT-? # no.
22:22:21.1
CAM [sounds of impact]
22:22:22.0
END OF TRANSCRIPT
END OF RECORDING

183
APPENDIX 3
Kingston ATC Transcripts: En route, Approach and Tower

Kingston Air Traffic Control Centre
Transcript of Voice Recordings

Accident AA331, 23 December 2009

184
Part 1 Of 3 En route Control
TIME CHANNEL STATION TRANSMISSON

02:47:00 128.1 AAL331 Ah….Kingston Control, good evening
American Tree Tree One..
:03 “ Kingston Enr American Tree Tree One Kingston

Radar….
:05 “ AAL331 Kingston American Tree Tree

One we’re level tree seven
zero, we are eight zero miles
north of position TOTON…..
:13 “ Kingston Enr American Tree Tree One Kingston

Radar, radar contact maintain
flight level tree seven zero
proceed TOTON Direct to
KEYNO….
:19 “ AAL331 TOTON direct KEYNO, American

Tree Tree One and ah… it…is it
possible for you to give us the
aah…weather at Kings….
Manley….
:27 “ Kingston Enr Roger sir, give me one minute, okay…
:30 “ AAL331 Thank you….
02:48:28 “ Kingston Enr American Tree Tree One…copy

aah.. Manley’s field conditions…
:32 “ AAL331 Go ahead please……
KATCC Transcript of Voice Recordings – Accident AAL331 Part 1- Enroute Control Page 1 of 3

185

02:48:33 128.1 Kingston Enr Wind at the station is tree one zero
degrees at aah..seven and a half knots
the visibility is approximately five
miles ah….present weather there’s a
moderate shower at station… the
temperature is two one, dew point two
zero the QNH is one zero one four…
understand that there’s also a broken
clouds at one thousand feet…..
02:49:03 “ AAL331 Okay, understand that ..understand

the wind you say was aah.. Tree one
zero seven.. Thank you….
:08 “ Kingston Enr And aah.. Tree Tree One affirm…
And Kingston American Tree Tree One

we’re back with you tree seven zero
02:51:13 aah.. Havana has aah..authorized
lower, we would like to start our descent
now if it’s okay with you…..
Station calling Kingston say again

aah… your request…..
:31 “ AAL331 Aah….Kingston it’s American aah…Tree

Tree One we’re level tree seven zero
aah… Havana’s ... switched us over to
:36 you and aah... they’ve authorized lower
and we’d like to request aah… start to
Kingston Enr
descend now….
American Tree Tree One descend your

discretion one five thousand QNH at
Manley One Zero One Four…
KATCC Transcript of Voice Recordings – Accident AAL331 Part 1- Enroute Control

Page 2 of 3

186

128.1 AA331 Okay one…aah…pilot’s discretion one five
02:51:56 thousand, American aah.. Tree Tree
One…..
“ Kingston Enr American Tree Tree One affirm…
02:52:01
“ Kingston Enr American Tree Tree One call

03:03:40 Approach One Twenty point
six…
:43 “ AA331 Good day, American Tree

Tree One…
KATCC Transcript of Voice Recordings – Accident AAL331 Part 1- En route Control

Page 3 of 3

187
Kingston Air Traffic Control Centre Transcript Of Voice Recordings Accident Aal331,
December 23, 2009 (UTC)
Part 2 Of 3 Approach Control
TIME CHAN STATION TRANSMISSON

NEL
02:55:25 “ AAL331 Kingston Approach, American Three Three One….
:41 Kingston Approach, American Three

AAL331 Three One.
Kingston Approach, this is American

AAL331 Three Three One….
American Three Three One, Manley

Manley APP Radar…..
Yes Sir, we’re still presently a hundred and

AAL331 twelve nautical miles from the airport.
Understand, that you have moderate rain
showers on the approach Just wondering if
any aircraft have been experiencing
turbulence on the
approach….
American Tree Tree One, none reported.

None reported. Have you had anybody
Kingston Enr land in the last hour?
Ok, standby, let me just tell you the last

aircraft that landed and the time ..
Affirm we had one aircraft that landed in

the la.. at zero two zero three and he came
in from Montego Bay, came in from the the
northwest and ah.. he didn’t have any
problems coming in ..
didn’t have any body from the north….
KATCC Transcript of Voice Recordings – Accident AAL331 Part 2 – Approach Control 1 of 3

188
02:59:31 120.6 AAL331 OK thank you and ah we’ll talk to

you in few minutes. We’re about a
hundred ah and ten miles from the
airport ….
:37 “ Manley APP
Ah roger…..
03:03:55 AAL331
Kingston good evening American
Tree Tree One we’re out of one
“ niner zero for one five thousand…
03:04:13 Manley APP American Tree Tree One Manley

“
:15 AAL331 Radar... Manley, American Three
Three One One

:20 Manley APP eight zero for one five thousand…
“
American Tree Tree One..aah..Manley
Radar expect ILS Approach runway
one two QNH one zero one four descend
:31 AAL331 and maintain five thousand…
“ Descend maintain ah…five thousand

:36 Manley APP American ah… Tree Tree One…
American Tree Tree One affirm be

advised the information that was
“
given to you by the enroute
controller is still
the same visibility five miles and there is
:48 AAL331 moderate rain at the station…
Five miles of rain…American ah… Tree

:51 Manley APP Tree One
American Tree Tree One you may

have to circle to land the wind is
03:05:01 AAL331 tree two
zero degrees at one zero knots…
Understand that, aah… that we can ah…

go ahead and ah… take a straight-in
with that American Tree Tree One…

189
03:05:07 120.6 Manley APP American Tree Tree One roger

copied…
03:14:48 “ Manley APP
American Tree Tree One descend
and maintain four thousand at
KEYNO cleared straight-in ILS
Approach runway one two…be
advised the wind’s now
tree two zero degrees at one five knots..
03:15:01 “ AAL331
Okay after ah… descend down

to four thousand now and
ah...after KEYNO cleared for
the ILS runway one two…
:11 “ Manley APP American…. Three Three
One…
American Tree Tree One…affirm

and ah.. did you copy the wind tree
two zero degrees at one four knots.
:21 “ AAL331 Are you able
to still land ah.. to make a
“ straight-in approach runway
:26 Manley APP one two?....
03:17:28 Manley APP Er….we copied the wind and we can go

straight-in to
“
one two…
:33 AAL331
Roger…..
American Tree Tree One

contact the Tower one one
eight decimal six five good
day…
Eighteen sixty five good

night…American
Tree Tree One…..

190
TRANSCRIPT: 118.65 MHZ (Manley Tower Frequency)

Record Playback: Start Time – 12/23/09 02:38:24Z & Stop
Time – 12/23/09 03:48:14Z
RE: ACCIDENT: AAL 331 – DECEMBER 23, 2009
TIME FREQUENCY STATION TRANSMISSON
03:17:32 118.65 AAL331 Manley Tower good evening American three three one is with you… we
are… level… inbound on the ILS runway one two….level twenty eight
:42 “ MLY TWR hundred feet.
American three three one wind three two zero degrees at one two (fades
:48 “ out) confirm still requesting runway one two (fades out
AAL331
That’s affirmative; ah… say again the wind?
“ MLY TWR
:52 Three two zero…now at one four knots.
“ AAL 331
That’s affirmative… ahh runway one two.
:54
“ MLY TWR American three three one cleared to land runway One two… be advised
:57 runway wet.
AAL331
“ Thank you
03:18:00 “ “
And American three three one… we are on three mile final.
MLY TWR
03:20:40 “ American three three one landing clearance still valid, cleared to land
runway one two.
“ AAL 331
(Noise in background) (expletive) nnnnoo… (expletive)…aah… (expletive) nnno!
:46 ACFT 628
“ Manley Tower aircraft six two eight
ACFT 628 Manley Tower aircraft six two eight?

03:22:16
“ MLY TWR Six two eight standby
:40
“
03:23:01
:04

191
APPENDIX 4
Jeppesen Approach Plates for Kingston MKJP/KIN
1. MKJP: KEYNO ONE ARRIVAL.
2. MKJP: Airport Information.
3. MKJP: Page 10-9 Runway Information.

4. MKJP: Page 11-1 ILS Runway 12.
5. MKJP: Page 12-2 RNAV (GPS) Runway 30.

192

193

194

195

196

197
APPENDIX 5
AA Flight Manual Part II, Kingston MKJP Page 10-7X

198

199
APPENDIX 6
American Airlines 737 Operating Manual – Performance Section
Bulletin No. 737-07, Date 11-27-06

Landing Performance Check

200
AA737 Operating Manual

201

202

203

204
Intentionally left blank

205
APPENDIX 7
American Airlines Boeing 737-800 Required Runway Landing Length Tables
American Airlines Boeing 737-800 Wind Component and Landing Data Card

206

207

208

209
APPENDIX 8
American Airlines Estimated Wet And Contaminated Runway

Landing Distances

210

211

212

213
APPENDIX 9
Boeing Landing Distances Charts

214

215
APPENDIX 10
Airports Group Report

216
Jamaica Civil Aviation Authority

4 Winchester Road
Kingston Jamaica
Report Date: December 31, 2009
Accident:
Operator: American Airlines
Airplane: 737-823,
Location: Kingston, Jamaica
Date: December 22, 2009
Time: 2222 EST (approximate)
Aerodromes Group
Group Chairman Jamaica Civil Aviation Authority

Member U.S. Federal Aviation Administration
Member Norman Manley International Airports Limited
Member National Transportation Safety Board
Observer Air Jamaica Flight Safety Manager
Summary: On December 22, 2009, about 2222 Eastern Standard Time, N977AN, a Boeing
737-823 airplane, registered to Wells Fargo Bank Northwest N.A. Trustee, and operated by
American Airlines Inc., as a Title 14 CFR Part 121 international passenger flight from Miami,
Florida, to Kingston Jamaica, overran the runway while landing at NMIA, Kingston Jamaica.
Instrument meteorological conditions prevailed in the area with reportedly heavy rain at the time,
and an instrument flight rules flight plan was filed. Of the 154 persons on onboard the flight, the
pilot, copilot, four flight attendants, and 134 passengers were not injured. Fourteen passengers
received injuries, of which six passengers received serious injuries. The flight originated at
Miami International Airport, Miami, Florida, about 2022.
Details of the Investigation:
1. Aerodrome Configuration: Norman Manley Airport is located on the Palisadoes

Peninsular on 532 acres, some 10miles from the major City of Kingston and 2 miles
northeast of the town of Port Royal at the end of the peninsular, on the island of
Jamaica. The reference point for this aerodrome is N17°56.0’ W076°47.3’
(WGS84).The aerodrome elevation is 3.04 meters above sea level with a geoid undulation
of 19.27 meters.

217
The airport is described as an Airport of Entry and is the major airport serving the south side of
the island. For navigation purposes and air traffic control all international flights are controlled
by the Kingston Flight Information Region prior to handing off to Kingston Area Control Center
for further rerouting to the Airport. The airport is open 24 hours. The airport is not certified in
accordance with ICAO Annex 14 Standards.
The airport is served by a single Runway (12/30) designated as a Code 4E with an

asphaltic concrete surface. Runway End Safety Areas and Stopways are not provided. The
reference temperature for this runway is 31° C. A runway strip 2730 meters long and 300
meters wide is provided to ICAO Annex 14 Standard with regards dimensions.
A Code 4E parallel taxiway is provided with four service taxiways with entry to various points
to the runway. A Taxiway strip is provided which meets the physical dimensions required by
ICAO Annex 14.
2. Aerodrome Certification: The ICAO requirement for aerodrome certification is specified in

ICAO Annex 14 (effective November, 2003). It specifies that every aerodrome used for
international operations be certificated by the state civil aviation authorities in that particular
state. The authority for aerodrome certification in Jamaica has been promulgated in the Civil
Aviation Regulation, as modified in 2004. Although the NMIA has initiated the certification
process in conjunction with the JCAA, currently the state has not issued an operating certificate
to NMIA, and thus the aerodrome according to ICAO standards is not certificated. The airport
operator has developed the requisite aerodrome manual, aerodrome programs, aerodrome
emergency plan, and has submitted them (along with the appropriate application) to the JCAA.
However, certification is still pending.
3. Aerodrome Manual: The Norman Manley International Airport (MKJP) has developed a
detailed aerodrome manual that was reviewed in the course of the investigation. The manual as
reviewed would meet the basic requirements of ICAO Annex 14 (aerodrome certification) and
ICAO Document 9774 (Manual for the Certification of Aerodromes) in terms of content and
specificity. It included an emergency plan as required. It also contained the minimum essential
information on site, facilities, services, equipment, operating procedures, organization and
management as specified by Doc 9774. However, the manual did not appear to include
detailed information on Safety Management Systems as required by Annex 14 and JCARs
regulation 142. The manual has been submitted to JCAA for review and is pending approval at
this time.

218
4. Aerodrome Emergency Plan:
4.1 General Information: The NMIA has developed an emergency plan that has been
submitted to JCAA for consideration of approval. The plan as reviewed as part of this
investigation addressed the minimum elements required of ICAO Annex 14 and ICAO Doc
9774. Those subsections included aircraft emergencies, sabotage, unlawfully seized aircraft,
dangerous goods, structural fires and response, natural disasters, and emergencies that are site
specific (that is, water rescue in this particular case). The manual has not currently been
approved by JCAA. A copy of the plan was found available to RFF staff during the
inspection/review of RFF facilities.
4.2 Water rescue plan (specialty rescue services): As MKJP and particularly Runway
12/30 is surrounded by water, ICAO Annex 14 (9.1.14) requires development of a site specific
water rescue plan. The MKJP aerodrome emergency plan had information on water rescue but
the information provided in the manual was sparse at best. The response procedures alluded to a
aerodrome owned boat (which did not physically exist) operated by trained RFF firefighters
(who are not currently trained) using flotation devices (which are currently not on hand) to
aid victims in the water. The plan also relied heavily on resources to be provided from mutual-
aid agencies (specifically the Jamaican Defense Force Coast Guard and the Port Authority of
Jamaica). The JDF Coast Guard is currently adequately staffed and does operate from an
operating base that sits within 5 miles of the aerodrome. Response from the JDF-CG would most
likely be adequate and would occur within a reasonable time (assuming that JDF-CG resources
are not deployed on other missions at the time of an aerodrome accident). Currently, MKJP has
initiated Memorandum of Understanding (MOU) agreements with both the JDF-CG and the Ports
Authority. These documents are currently in legal review and when completed will enhance the
water rescue capability of the aerodrome. In addition, aerodrome management stated
that they are currently in final negotiations of the purchase of a 27 foot water rescue boat that
would be stored with RFF and would be operated by that service. There is an existing boat
ramp in the vicinity of the RFF station that could be used by the rescue boat.
5. Runway Condition
5.1 Existing Runway Data
The runway is 2716 meters long and 46 meters wide. The runway slope for Runway
12 is 1% U while the slope for Runway 30 is 1% D. The runway is designed with
0.1% transverse slope along the full length. The geometric centre of the runway
is 1357.80 meters west of the threshold at runway 30 on the center line. The threshold
elevations for runway 12/30 are 2.34 / 5.31 meters respectively.

219
Threshold elevations/geoid undulation for runway 12 is 2.34 meters /

18.41meters and that of runway 30 is 5.31 meters/ 21.09 meters, respectively.
Touchdown zone elevations for Runways 12/ 30 are 2.40 / 3.02 meters respectively.
The bearing strength of runway 12/30 is adequate to allow continuous operations by

the aircraft for which designed. (PCN 68/A/W/T). Taxiways and holding bays are of
the same bearing strength as the runways.
5.2 Visual runway condition assessment
The runway was visually inspected. The macro surface appeared in good condition
with the runway surface rough to the touch. There was evidence of runway edge
damming beginning to form along the entire length of the south side. There was also
evidence of significant runway edge damming forming on the north side, at the
runway shoulder edge from taxiway Echo to Taxiway Charlie.
The water flow had built dams from 6 inches to almost 12 inches in height.
Examination of Surveyors drawings and measurements constructed in 1995,
demonstrated that the edge of the threshold of runway 30 was 5.3 meters while a
low point of 2.10 meters was measured at a point 800 meters from the runway 30
end. This was borne out by the physical evidence observed on the north side of
the runway between taxiway Echo and Charlie. The runway from this point eastward
appear to rise slightly showing evidence on the north side between Taxiway Echo and
Delta of a counter flow of water, borne out by the building of a debris mound just off
the shoulder appearing to separate the two water current flows. It is apparent from the
surveyor’s drawings that it is possible to have significant water flow. From the
physical evidence observed, there is edge damming for a significant distance along
the edge of the runway north side between Taxiways “Echo” and “Charlie” from the
runway 30 threshold. The evidence of water flow in this area indicates that further
observations would be required.
Due to the positives transverse slope of the runway water flows to the runway strip on
both sides of the runway and then soaks away. In the area on the north side of the
runway between taxiways “Echo” and “Delta” the water flows to a swale for further
distribution to drainage ducts.
The maintenance of the swale and the areas immediately surrounding the ducts
is poor, in fact over time the surface has receded below the level of the ducts causing
them to be ineffective in moving the water to the drainage system. In examination of
the surveyor’s drawings there was no evidence of the drainage system in the runway
strip being linked to the main drainage system or sewer lines routed under the apron.

220
Throughout the runway strip there were obstacles, such as blocks of concrete which
need to be removed.
It is recommended that a proper maintenance plan be developed for the

Runway strip and that a survey of the drainage system be conducted. A further visual
check was conducted for distress on the runway surfaces in the following four major
areas:
 Surface defects
There were no visual signs of raveling or flushing; however in the touchdown
zone area of Runway 12 along the center line there were signs of polishing
and minor pock marks where small pieces of the surface area had popped out.
 Surface deformation
There was no sign of rutting, distortion or rippling of any kind. In determining
the transverse slope there was a continuous positive transverse slope along the
entire length of the runway. A further informal engineering survey of several
stations along the runway also demonstrated a positive transverse slope
along the length of the runway however in one area within the landing zone
of Runway 12 the transverse slope demonstrated less than 0.1% and a
reading of approximately 0.45% was measured.
 Cracks
There were no thermal, reflection or slippage cracks along the length of the
Runway. There were however, several longitudinal joint cracks along the
entire length of the runway on either side of the centerline; there are lateral
cracks as well which could be associated with the methodology utilized while
laying down the runway surface. These cracks showed no signs of lifting,
demonstrating that water was not at this time seeping below the surface.
In the landing zone area of Runway 12, particularly, between 1500 to 2000
feet from the threshold, there were signs of block cracking.
There were signs of lightning strikes in the first 5000 feet of the runway from
the Runway 12 end. This was demonstrated by small pop outs. These were of
no significance and there was no debris on the runway.

221
5.3 Runway condition inventory
The runway surface overall was in good condition, however a crack sealing
program would be appropriate to prevent further widening of the cracks by water
penetration.
6. Taxiway Condition
All taxiway surfaces were inspected and found to be in good condition; however
Taxiway “Alpha” “Echo” and “Delta” needs further discussion.
Taxiway “Alpha” is parallel to Runway 12/30 and is the main taxiway entering the runway
at the threshold of Runway 12. Subsequently all heavy jet traffic are routed by this taxiway to
Runway 12. There is evidence of deterioration of the taxiway surface in the turn leading to the
Hold line. The surface area for approximately 200 feet back from the hold line on either side of
the taxiway center line displays areas of surface distortion – rippling.
The shoulder entering the runway on the east side at Taxiway Delta shows sign of water erosion
with a sinkhole in the peak of the curve. Although no threat to the taxiway surface at this
time it is beginning to show signs of substrata erosion.
The shoulders entering the runway on the east side at taxiway “Echo” showed signs of water
erosion. There is a sinkhole with the substrata deteriorating beneath the surface and in the peak of
the curve, causing the surface to break away. At this time the taxiway surface is not affected.
Appropriate maintenance action needs to be taken to correct the anomalies.
7. Runway strip/Runway end safety areas
7.1 Runway strip: MKJP Runway 12/30 is designed and operated as a Code 4 Runway.
Annex 14 requires Code 4 Runways with precision instrument approaches to maintain a Runway
strip of 150M on each side of the runway centerline. Due to existence of an off-set ILS localizer
on Runway 12 there is some question at MKJP regarding the applicability of Cat I precision
approach criteria. Regardless of that question regarding precision approach criteria for 12, the
aerodrome manual for MKJP as currently written states that the aerodrome will maintain Code 4
precision runway strip standards thus the runway was evaluated using 150M as the standard.
The investigation conducted found that MKJP meets the runway strip criteria along the full
length of 12/30 on both the North and South sides with the following notable exceptions:

222
- At both runway ends (12 And 30) the 60M runway strip as required by Annex
14 (3.4.2) does not exist. Physical examination of the Runways indicated a runway strip on
Runway 12 of less than 15M and less than 10M on runway 30.
- Along the entire length of 12/30 on both sides of the runway there were a
number of objects (such as rocks, concrete blocks as large as 1M wide, 1M long and
.3M in depth) and manhole/junction box covers that protruded above the existing grade and
would thus be in violation of the requirement of Annex 14 3.4.7 requiring that the only
objects in the graded portion of the strip be limited to those items that are frangible and required
by function to be there. In addition there was one sinkhole (approximately 4M long, 3M wide
and 1-3M deep) that sat approximately 75M from the runway edge on the North side of the
Runway between taxiways B and C. In addition to this sinkhole, there were sinkholes forming
along taxiways D and E that were as deep as .5M and approximately .5M wide, running as long
as 5M along taxiway D.
- The localizer antenna array on the South side of the 30 end of the runway
currently sits on a concrete pad that is approximately .3M in height above grade.
Although the aerodrome group was provided with documentation regarding the frangibility of the
localizer antenna, the pad upon which the antenna currently rests would appear to be contrary to
the recommended practice of Annex 14 (3.4.6) which states that objects which may endanger
airplanes in a runway strip should be removed as far as practicable. In this case feathering of
the existing grade to the base of the localizer antenna would remove this object and danger
from the runway strip.
- The aerodrome group during inspection of 12/30 found numerous electrical
junction box covers in the runway strip. These junction box covers measured approximately 1M
x 1M. These covers were constructed of fiberglass and covered holes that appeared to be as
deep as 2M. Although not specifically mentioned in Annex 14 the investigative team noted that
these covers would not hold the weight of an aircraft during a runway excursion and thus we
recommend that these covers be replaced with metal covers sufficiently strong to sustain the
weight of a passing aircraft.
- Issues associated with runway edge ‘damming’ that is occurring along the
entire runway length (both sides), but especially along the North side of the runway from the 30
end to Taxiway C is more specifically addressed in the runway pavement section of this
report.
7.2 Runway End Safety Area (RESA): Currently no RESA is provided on either the 12 or
30 end of the sole MKJP Runway. To address this issue the aerodrome Master Plan currently
contains plans for the extension of Runway 12 into the water combined with the displacement
of the runway 30" threshold to provide a RESA compliant with Annex 14 requirements. The
exact time of completion of this project is currently unspecified.
8. Markings: A physical examination of the apron, taxiway and runway markings at MKJP
found the markings to be in good condition and compliant with the requirements of ICAO
Annex 14. There were a few taxiway markings (particularly the taxiway edge markings on the
West end of the airport near taxiway E and D) that had faded slightly and thus we would
recommend a paint project be considered to enhance those markings.
223
In addition, we did not physically see, nor did the paint specifications provided to us by MJKP
management indicate, that the aerodrome uses any reflective enhancement (glass beads) in the
paint used on either the runway or the taxiways. Although only an Annex 14 recommended
practice (5.2.1.7), the use of glass beads is discussed in ICAO Doc 9157 Parts 2 and 4 as
having not only the ability to visually enhance markings but having the secondary benefit of
providing a modicum of friction enhancement on the surface of the marking.
9. Sign Systems: A physical examination of the apron, taxiway and runway sign systems at
MKJP found the signs to be in good condition and essentially compliant with the requirements of
ICAO Annex 14. All signs were internally illuminated and appeared to properly maintained. The
exceptions to full Annex 14 compliance are as noted:
- ICAO Annex 14 standards (5.4.2.8) require that a runway designation sign be installed
on each side of the runway holding position marking at each taxiway/runway intersection
IAW the standards and distances specified in the Annex. MKJP currently has only one
designation sign per runway /taxiway intersection.
- Annex 14 (5.4.7.1) requires a road-hold position sign at all road entrances to a Runway.
We noted two locations (in the vicinity of the ILS localizer antenna and at the approach end of
Runway 12) where these signs were missing.
- The runway exit sign from runway 30 onto taxiway Bravo(high speed exit) was missing.
- The taxiway Charlie direction sign at taxiway Alpha was out of service during the night
lighting inspection.
10. Lighting Systems: The runway and taxiway lighting systems at MKJP consisted of
medium intensity taxiway lighting and high-intensity runway lighting on Runway 12/30.
Systems are controlled from a Mimic board in the Air Traffic Control Tower. The Mimic board
control currently operates all runway lighting with the exception of taxiway lighting on the
East and West ends of taxiway Alpha. That lighting is currently controlled from the electrical
vault and was on and operational during the night inspection. With minor exceptions Runway
and taxiway lighting was found operational, well-maintained, and designed/installed in
accordance with the requirements of Annex 14. The minor exceptions noted during the night
inspection were:
- 1 light on the Runway 12 threshold lighting system was out of service.
- Four lens-cap fixtures on Runway 12/30 were found installed backwards and were
thus not providing the correct light angle and luminosity. This can give a distorted sight picture to
landing aircraft.
- 3 lights of the runway 30 threshold light system were missing.
- At the taxiway Delta high-speed exit from Runway 12/30 one in-pavement edge light
was out of service.
- One amber/white split-lens runway edge light on Runway 12/30 at taxiway Echo was
misaligned.
- 1 taxiway edge light at taxiway Alpha near taxiway Echo was out of service.

224
11. Visual Aids

11.1 Visual guidance slope indicators. Annex 14 (5.3.5.1) requires visual approach
slope indicator systems for runways used by turbojet aircraft, regardless of other approach
aids provided. Currently MKJP has installed and provided a Precision Approach Slope Indicator
(PAPI) on both ends of 12/30. The normal PAPI configuration as specified in the Annex consists
of 4 lights located to the left of centerline as viewed by the pilot. MKJP currently provides a 4-
box PAPI on each side of centerline (8 boxes total) for each runway end. The systems were
operational and appeared to be well-maintained. The last PAPI calibration (as determined by
maintenance record) was December 26, 2009. The previous calibration/maintenance was
performed August 08, 2008. Systems were found on December 26, 2009 to be within calibration
standards. Although periodicity of maintenance is not specified in Annex 14 and is thus
normally subject to manufacturers’ specifications, we recommend that MKJP institute a specific
maintenance/calibration periodicity and a periodic maintenance schedule to service the VGSI’s.
11.2 Wind cones. Lighted wind-cones are provided at both runway ends of 12/30. The
wind-cones were operational, the lighting was operational, and they met the installation/design
and maintenance requirements of Annex 14.
11.3 Approach lighting systems. Currently MKJP provides a Simple Approach

Lighting System (SALS) on both runway ends of 12 and 30. The 30 end system is modified due
to geography of the terrain and does provide the standards specified in Annex 14. During the
time of the inspection the SALS on 30 was operational, while the SALS on 12 was out of
service. NOTAM records indicate that the SALS on 12 was out of service on December 22,
2009.
12. Aerodrome programs
12.1. Wildlife control programs: Annex 14 currently requires that states develop and
promulgate a national procedure for recording and reporting bird strikes. Currently that
requirement has not been established in Jamaica and all reporting (although it occurs and there
are concerted efforts to collect bird strike data) tends to be local rather than nationally
coordinated. At MKJP the aerodrome manual covered bird strike reporting and wildlife control in
a cursory fashion, yet the aerodrome maintained a wildlife log was aware of the dangers
associated with wildlife. During the assessment there were no significant wildlife issues noted.
Some pelicans were observed, some cattle egrets were noted and there was colloquial evidence
of feral dogs that would make their way onto airport property. In summary, although there is no
national procedure for bird strike reporting, the issue of wildlife control does not appear to be a
pressing issue for MKJP. However, there is always value in having a trained wildlife biologist
conduct an assessment of the wildlife situation of MKJP and the surrounding area and we would
recommend such.

225
12.2. Ground vehicle control: Annex 14 requires that each certificated aerodrome develop
a Surface Movement and Guidance Control System (9.8.1) in addition to an appropriate Apron
Management Service (9.5.1 where traffic conditions dictate). MKJP traffic control appeared
adequate for the volume of traffic at the aerodrome. The aerodrome provides a driver training
program and controlled driver access through a ‘badging /licensing’ system that differentiates
between movement and maneuvering area driver authorization. In addition, with nine Operations
Coordinators assigned to the Operations Department, the control of vehicles and runway access
appeared to be in compliance with Annex 14 and met local condition requirements.
12.3. Self-inspection; Daily self-inspections are conducted by both the Operations

Coordinators and RFF staff. These daily inspection are organized through use of a checklist and
findings are consolidated in the electronic maintenance management system. We found
however that even though the departments make every attempt to conduct adequate self-
inspections and condition reporting there is evidence that additional effort and training
may be warranted. Currently there is little to no standards training provided to staff who conduct
self-inspections and little evidence that inspections are more than simple FOD and light checks.
We would recommend that MKJP institute a more robust self-inspection training program for
staff who conduct self-inspections, consisting of standards (Annex 14 and ICAO Doc) review,
testing on standards, and management review of self-inspection findings.
12.4. Aerodrome maintenance: Currently the guidance for maintenance standards found
in either the JCAR (2004), in JCAA guidance, or in the MKJP Aerodrome Manual is minimal at
best. Standards instead seem to refer exclusively to ICAO standards as specified in Annex 14 and
the associated ICAO documents. In many cases (particularly in the case of maintenance of
runway strips and RESAs) this may be inadequate. We feel that the situation found along the
runway edge associated with accumulation of debris and ‘edge-damming’ along the runway
may be the result of insufficient guidance provided to self-inspection and aerodrome
maintenance staff. In addition, we feel that the situation associated with the accumulation of
debris associated with edge-damming is also the result of insufficient training being provided
to maintenance staff. If staff had been properly trained in the ICAO requirements associated
with provision and maintenance of runway strips, the situation that developed along the runway
edges may not have developed. This is the one area where we feel that MKJP could do a better,
more thorough job, and JCAA could provide more oversight. In our opinion this is the one
specific area where the airport may not be in compliance with the recommended practice of
Annex 14, chapter 10, and the requirements of ICAO 9774 for the provision of qualified
staff.
12.5. Foreign object debris control. Annex 14.10.2.1 requires aerodromes maintain
surfaces of pavements free from foreign object debris (FOD). The FOD control program
appeared to be adequate and there was little to no FOD observed on surfaces during the
investigation.

226
13. Rescue firefighting services
13.1 RFF program status: MKJP is designated a RFF Category 8 aerodrome and the
aerodrome does meet the requirements of Annex 14, Chapter 9 in that regard. We found the RFF
service at MKJP to be very professional and capable.
13.2 Equipment
13.2.1. RFF vehicles: To meet RFF Category 8 requirements MKJP would have
to meet the Annex 14 requirements of Table 9.2. During the period of the assessment
there were three operational RFF vehicles capable of discharging 24996 liters of water for foam,
and capable of discharging 677 kg of dry chemical. This is in excess of the Annex 14 Table 9.2
requirements. The vehicles were found in good condition and were well-maintained. A
preventive maintenance and daily inspection program was in place for the vehicles.
13.2.2 Rescue equipment. The RFF services maintained minimum rescue

equipment ranging from simple hand-tools (which were available on all vehicles) to hydraulic
spreaders (maintained on RFF14). We did find evidence of training having been completed
on this equipment and firefighters seemed to aware and capable of using the equipment. We were
told that a powered rescue saw was not available and we would recommend that RFF services
acquire such. In spite of this we do believe that RFF services met the intent of Annex 14
recommended practice 9.2.20.
13.2.3. Personal protective equipment (PPE). The PPE requirements of Annex

14.9.2.38 (and as specified in ICAO Doc 9137, part 1) were determined to be met. Each
firefighter had assigned to them a structural firefighter coat and pants, a helmet with shield,
gloves, protective boots, a hood and a personal SCBA mask. There were 9 Scott Self-Contained
Breathing units found and all were determined to be serviceable and full with air. In addition
there were 15 additional SCBA bottles that were also found full. Re- supply of air comes from
an agreement with JFB. We recommend that the aerodrome consider acquisition of their own
re-supply system to refill SCBA bottles.
13.2.4. Medical. Emergency medical qualifications for RFF personnel are not
clearly defined in Annex 14. MKJP RFF has elected meet emergency medical requirements
through the training of all firefighters in basic first responder medical training and by
maintaining a number of staff with Emergency Medical Technician (EMT) capability. This
is commendable and as the RFF service also provides ambulance service, required to a degree by
Jamaican governmental requirement. There are no Annex 14 issues associated with the
emergency medical capability of MKJP. We do however recommend that the aerodrome
(through the aerodrome manual and the emergency plan) more formally state their position. As
there is currently no written minimum emergency medical standard for RFF services this current
posture could evaporate. We recommend this be addressed through aerodrome standards
coordinated with JCAA.

227
13.3. Staffing. Current staffing minimums require a minimum shift of 9 persons, of

which would be three APS Officers and 6 firefighters. This staffing is maintained 24 hours/day
and exceeds the requirements of Annex 14, Chapter 9 (and the guidance of ICAO Doc 9137, part
1).
13.4. Training. Annex 14.9.2.34 requires that all RFF staff be properly trained with the
recommended training spelled out in Doc 9137, part 1. Our investigation found that the MKJP
training program did address training detailed in Doc 9137 for both initial and recurrent training
and we found the training program met the requirements of Annex 14. We found a training
syllabus (conducted on an annual basis) and evidence that firefighters were being adequately
trained IAW ICAO guidance. We also found that firefighters are rotated through the US South
Carolina Fire Academy, Columbia, SC, on a recurring basis of approximately once every 3 years.
This is an IFTSA certified 48 hour program that provides quality, professional RFF training. We
do understand that this is a considerable expense for MKJP and we commend them for their
professional consideration. We would recommend two items regarding training:
- RFF services should make greater effort to acquire rescue diagrams and
detailed training material for each type commercial service aircraft that uses MKJP (that is,
737, A320, 747 etc.) In addition, we recommend continued effort to coordinate live ‘hands-on’
training on these aircraft as these aircraft may become available at the airport.
- We recommend development of detailed lesson plans of each required
training subject on the annual training syllabus. This would include detailed training
requirements, goals, objectives, outcomes and references required. Additional guidance can be
obtained from NFPA 1003.
13.5. Agent available: MKJP RFF services currently use Aqueous Film Forming Foam
(AFFF) as the primary agent, with protein-foam and PKP Dry Chemical as complementary
agents. Each operational RFF vehicle had full AFFF tanks that were capable of discharging at
least 2 operational loads of water for foam. In addition, on hand at the RFF station was an ample
supply of AFFF, Protein-Foam and dry chemical. All agent stocks were found to be within
serviceability dates and were stored properly. The RFF service also had arrangements for water
re-supply during an emergency event with water tenders coming from the Jamaica Fire Brigade
(JFB). We would recommend that these arrangements be made more formal through negotiation
of a MOU and by inclusion of detail in the AEP.
13.6. Access roads. Previous audits of MKJP indicated an issue with the fact that RFF
vehicle access to the runway is not direct to within 1000M from runway threshold. This access
routing has not changed since previous audits. However, as there is direct access to taxiway
Alpha and as this is an Annex 14 recommended practice (9.2.26) this does not appear to be a
major issue. In addition, there was ample evidence of testing done on each watch shift (verified
by investigators over the past 12 month period) to verify response times meeting the Annex
14.9.2.21 requirement.

228
13.7. Alerting
13.7.1. Station alerting process and equipment: Current staffing levels for RFF
duty shifts require a minimum staffing of 9 persons. At all times (24 hours/day) one person is
assigned to the watch-room to monitor and direct emergency communications. Besides radio
communications between ATCT and RFF there was a direct line ring-down telephone that is
tested daily. This ring-down system was utilized during the incident on December 22nd. During
the investigation this phone was also tested. Communications meet the recommendations of
Annex 14 (9.2.31).
13.7.2. Vehicle and on-scene communications. RFF vehicles are all equipped with
VHF radios capable of transmitting on ATC frequencies. RFF vehicle 11 radio was out of service
during the investigation. In addition, each vehicle has an additional radio used for internal
APS/aerodrome communications.
13.8. Emergency medical services capacity: MKJP RFF services, currently maintains a
‘mass-casualty’ trailer that is stocked with a minimum basic load of medical supplies. It appeared
that these medical supplies would be sufficient for a mass-casualty situation for an aircraft of
the most demanding nature (in this case the 747). However, even though there appeared to be
sufficient medical equipment we were not provided with an inventory and there was no
evidence of periodic inventories and assessment of serviceability dates having been conducted.
Although, there is little guidance in Annex 14 and the RFF service seemed to be proactive in
the acquisition and maintenance of medical equipment, we would recommend a
reassessment of need and more formal record keeping and inventory of this equipment.
14. Electrical power systems.

The power supply for this airport is normally routed from the national power grid; however in
case of supply interruption, the airport administration has put in place auxiliary power
systems that will automatically switch on to take up the load of all necessary power requirements
needed. The main auxiliary power system is located at a remote site on Airport property with
fuel quantity capable of supplying continuous operation for four days or longer depending on
usage. The site is well constructed clean and appears well maintained. There is no fire
suppression system. There is however a 20 lb fire extinguisher on site in case of minor fire
events.
The power systems have a fifteen second switch over time, which meets the secondary power
supply system requirement in ICAO Chapter 8 Table 8-1.
In case of failure of the main auxiliary supply system the each runway has its own system
which will power the runway edge lights , taxiway lights, threshold lights runway end guard
lights PAPI, ILS and VOR approach systems.

229
The following systems are in place:

Main substation
i.) Two 2MVA Caterpillar Diesels (automatic Switch over)

ii.) One 938 kVA caterpillar Diesel as additional backup (automatic Switch over)
Runways
i.) 12 (West Substation One 50 kVA standby generator

ii.) 30 East B Substation One 100 kVA standby generator
Administration
i.) One 300 kVA Standby generator located in the area of Stand 10 to operate
Control Tower and Administrative facilities.
Stand by Lamps
Battery powered lamps are utilized for lighting the runway and taxiway in case of
auxiliary power failure.
15. Recommendations.
Aerodrome Certification:
- Continue to move forward toward the goal of meeting the JCAA and
Annex 14 requirement of full aerodrome certification.
- Consider coordination of a certification training program for both JCAA and
MKJP staff on certification standards, conduct and development of a certification program, and
self-inspection/condition-reporting.
Aerodrome maintenance:
- JCAA and MKJP develop specific maintenance standards and publish them,
particularly on the subject of pavement and runway strip maintenance.
- Provide detailed maintenance training to those responsible for self- inspections
and for aerodrome maintenance.
Aerodrome Emergency Plan:

- Coordinate additional detail in the emergency plan, particularly in the water
rescue section of the plan.

230
Pavement:
- Coordinate a full runway assessment to determine current longitudinal slope,
current transverse slope (along runway full-length), and the existing grade in first third of
Runway 30 (from approximately taxiway D to the 30 threshold).
- Coordinate a hydrologic engineering assessment to determine water flow on the
first third of runway 30 from approximately taxiway D to the 30 threshold.
- Coordinate completion of a full friction assessment of the runway to determine
compliance with established friction maintenance standards and deviation from the findings of
the last friction assessment (2004).
Marking/lighting signs:
- Add additional runway designation signs at every runway hold-position
- Consider revising current PAPI configuration to the ICAO standard

Marking.
(PAPI 4L) by removing the additional 4 boxes on the right side of 12 and 30. This would reduce
maintenance, minimize the potential confusion to aircrews that may occur due to the separate
systems being out of calibration, and would still confirm to ICAO standards.
Rescue Firefighting Service:

-Develop detailed training lesson plans for required training subjects of initial and
recurrent training.
-Consider acquisition of a simple SCBA bottle refill system for aerodrome use.
-Acquire powered rescue saw(s) for aerodrome use.
-Enhance existing training procedures used for aircraft familiarization…to include
acquisition of current training aids for each commercial service aircraft currently
using MKJP.
Declared Distances: Currently, the aerodrome operator complies with the requirement of
Annex 14.2.8, for declared distance computations. The figures are published in the Jamaica AIP.
Currently however, the only declared distance figures for MKJP 12/30 that differ from the
distance representing 12 threshold to 30 threshold (2716M) distance, are the TODA figures for
both 12 and 30 (currently published as 4074M). This represents acknowledgement of existing
clearways on both runway ends. We recommend consideration be given to review of all declared
distance figures to accommodate the fact that there is no existing runway strip or runway end
safety area at either runway end.
Runway strip and Runway end safety area.

- Remove existing debris and vegetation from the existing drains in the
runway strip between taxiway D to E. Verify that these drains are operational
and that they are tied into the drain system and box-culvert under the aerodrome
apron.
231
- Conduct periodic surveys of the runway strip and RESA (as provided) to
determine safety issues and correct found deficiencies on a more frequent basis.
- Remove all current deviations (objects not required) from runway strip.
- Feather existing localizer base to grade.
Summary Statement:
In spite of the fact that MKJP (Norman Manley International Airport) has not fully met the
JCAA-JCAR/Annex 14, Volume I, requirements of aerodrome certification, it is the opinion of
the aerodromes group that the aerodrome as currently operated and managed meets or exceeds
the requirements of what could be called ‘safety critical items’ associated with Annex 14
aerodrome certification compliance, except in the notable area of provision and maintenance of
runway strip and runway end safety areas.
Attachments:
- Part 3, Aerodrome Particular to be reported to AIS (AIP Aerodromes
Section).
- 12/30/09 Memo from Mr. Roy Williams re. Runway Cross Section
Check MKJP
- APS Watch room Log Entries 12/22/09 (2225h) to 12/23/09 (0434h).
- MOU between NMIA and Port Authority of Jamaica
- MOA between NMIA and Jamaica Defense Force – Coast Guard
- Page 4-81 to page 4-83 of MKJP Aerodrome Manual regarding wildlife control
procedures.
- Page 4-43 to page 4-56 of MKJP Aerodrome Manual regarding self-
inspection, maintenance management, secondary power, and aerodrome works
during construction.
- PAPI calibration check records 08/08/08 and 12/26/09.
- Localizer antenna Runway 12 statement of frangibility.
- Material Safety Data Sheet (paint specifications) for runway and taxiway paint
used by aerodrome operator.
- Results of runway friction test conducted 07/17/04 of MKJP Runways 12 and
30 using a Grip-Tester Model GT289 system.
- Copy of daily preventive maintenance RFF vehicle systems checklist
utilized by MKJP APS (RFF).
- 01/02/09 copy of Memo detailing RFF APS fire service training syllabus for the
calendar year 2009.
- Page 6-1 to page 6-3 of MKJP Aerodrome Manual regarding reported airfield
deficiencies (RESA, runway-taxiway separation, runway strip, runway
longitudinal slope, aircraft holding position placement, and RFF services access
to Runway 12/30).

232
Section 2, Analysis

233
Section 2, Analysis
SECTION 2
ANALYSIS

234
Section 2, Analysis
2.0 Analysis
2.1 Aircraft
The aircraft, including all systems designed to decelerate the aircraft on landing, was serviceable
and airworthy at the time of the occurrence, except for the air-conditioning pack temperature
controller fault warning, which was deferred in accordance with the MEL in Miami before
departure. As lack of deceleration was a major factor in the accident, the aircraft’s braking
system was examined in great detail, and no fault was found (see 1.12.4 and 1.16.5).
The investigation revealed no mechanical aspect of the aircraft that contributed to the accident. 82
2.2 Flight Crew Qualifications
The captain and first officer were properly certified and qualified under Federal Aviation
Regulations and company requirements. The investigation revealed no pre-existing medical or
behavioral conditions which might have adversely affected the flight crew’s performance during
the accident flight. 83
2.2.a NOTAMS
As described in 1.18.4, NOTAM AO208/09 announcing the planned closure of the runway at
Montego Bay was given to the AA331 flight crew in their original briefing. However, it seems that
NOTAM AO215/09 cancelling that closure was not a NOTAM per se, but simply resulted in the
original NOTAM being cancelled and deleted from the list of active NOTAMS. Thus it is quite
likely that there was no mechanism in the system whereby the dispatcher would be alerted of this
change, and, as the dispatcher was busy with other flights and had no reason to re-check the
NOTAMS for AA331, he would not have realized the NOTAM was cancelled, and hence would
not have alerted the AA331 flight crew to this.
The investigation does not consider that the crew’s decision making process during the planning of
the approach and landing would have been any different if they had been informed that the
planned closure of Montego Bay was cancelled, but nevertheless the investigation recognized that
this characteristic of the NOTAM system could result in a hazardous situation in other
circumstances. For this reason, the investigation recommends that NOTAMs cancelling other
NOTAMs should be brought to the attention of flight crews using these NOTAMs. 84
82
Section see 1.6
83
Section 1.13
84
Section 4.2.32, Safety Action Required

235
Section 2, Analysis
2.3 Airport
NMIA Airports Limited stated that the approach lights for runway 12 had been unserviceable
since 30 November 2009, due to an underwater electrical fault. There was an active NOTAM
regarding this included in the AA331 dispatch document. The approach lights for runway 12
were not an operational requirement for the use of runway 12 in Instrument Meteorological
Conditions (IMC). 85
The absence of the approach lights for runway 12 may have reduced depth perception cues for
the captain during the final stages of the approach, but these were lead-in lights, and were there to
assist the flight crew with acquiring and lining up with the runway during the final stages of
approach, when transitioning to visual flight.
Although NMIA Airports Limited did not have its own friction measuring device, this was not
an issue as friction measurements were needed less frequently for lightly used runways such as
MKJP runway 12, and NMIA Airports Limited contracted the services from the Grantley Adams
International Airport in Barbados when needed. It would have been desirable, but it was not
essential, for the airport to have its own friction measuring equipment.
TDZ lights are not a standard for CAT 1 Precision Approach runways. Reflective markings are
not a standard, but are an ICAO recommended practice. Therefore MKJP was in compliance with
ICAO in these regards. However, the investigation believes that "black hole" approaches should
be enhanced by runways having TDZ lighting and/or reflective markings. The lack of these may
have been a factor as they would have assisted the pilot to establish with more precision where the
aircraft was in relation to the touchdown zone during the landing. On a dark, horizon-less,
rainy night such as it was, these would have been of great value to the AA331 flight crew.
Since NMIA was still undergoing certification at the time of the accident, the
investigation carried out an inspection of the airport by subject matter experts. A list of some of
the deficiencies identified during this inspection is in 1.10.2.4. 86
85
Section 1, 1.8.2.1 and 1.10.2.6
86
Appendix 10, Airports Group Report

236
Section 2, Analysis
The Summary Statement of this examination was as follows:
In spite of the fact that MKJP (Norman Manley International Airport) has not fully met the
JCAA-JCAR/Annex 14, Volume I, requirements of aerodrome certification, it is the opinion of
the aerodromes group that the aerodrome as currently operated and managed meets or
exceeds the requirements of what could be called ‘safety critical items’ associated with
Annex 14 aerodrome certification compliance, except in the notable area of provision and
maintenance of runway strip and runway end safety areas. 87
NMIA Airports Limited lacked operational procedures for the conduct of runway surface
inspections during inclement weather and the lack of agreements between the airport, air traffic
service and other users, for the furnishing and distribution of inspection results, precluded flight
crews from being apprised of the most recent runway conditions prior to arrival. The investigation
recommends that these measures should be implemented.
The lack of documentation of runway inspections, as described in 1.10.4.6, indicated a deficiency

in the runway inspection procedures at NMIA.
The meteorological reports (See 1.7), the flight crew reports 88 and the evidence of the aircraft’s
performance from the FDR indicated that runway 12 was probably in a Fair/Medium braking
action performance condition, not in a Wet/Good condition, as per the AA Landing Data Card. 89
2.4 Runway
2.4.1 Runway Surface
As described in 1.10.3, the runway surface was examined and tested by GAIA and the frictional
averages were found to meet and exceed the ICAO recommended minimum values. It was thus
determined that the frictional characteristics of the runway surface did not contribute to the
accident.
2.4.2 Runway Slope
The investigation accepted the information from the longitudinal survey of runway 12/30,
prepared by EDM Consultants, dated February 1997, titled “Rehabilitation of Runway, Taxiway
and Pavement Works, Runway Profile”, which was provided to the investigation by
87
Appendix 10
88
See 1.1
89
Appendix 7

237
Section 2, Analysis
the Airports Authority of Jamaica, as being a fair representation of the runway longitudinal profile
at the time of the accident. 90
In accordance with this information, the longitudinal slope of the runway was within the quoted
ICAO recommendations, except that the slope of -1.233% in the first 260 m (853 ft.) of the first
quarter of runway 30 exceeded the ICAO recommendation of a maximum slope of 0.8% in the
first and last quarter of the runway. The slight upslope of runway 12 would have had a small
decelerating effect on the aircraft during the landing roll.
Although the slope of -1.233% in the first 260 m (853 ft.) of the first quarter of runway 30 (that is,
the upward slope at the end of runway 12) exceeded the ICAO recommendation of a maximum
slope of 0.8%, ironically this upslope had the effect of giving the aircraft some upward propulsion
just before it left the runway, thus resulting in the aircraft clearing the depression in which the
road to Port Royal lay and having a softer and more level landing site on the sand berm on the
east side of the road rather than the rock-faced slope beside the roadway, and also it reduced the
speed of the aircraft at impact. These factors resulted in the cockpit and cabin being in a more
survivable condition due to the absorption of energy; they also stopped the aircraft before it
collided with the concrete pillars supporting the approach lights of runway 30, and before the
aircraft entered the sea (see Photo 2). Thus, more serious injuries and/or fatalities to crew and
passengers from G forces and damage to the aircraft structure, and possibly from drowning,
were avoided.
2.5 Air Traffic Services (ATS)
Although the J CAA ATS Manual of Operations (MANOPS) stated that the controller, on
initial contact, must give the current runway conditions and also the latest braking action
report to the aircraft, or state that none had been received, RSC reports based on actual
observations could not be obtained by ATC since there was no formal agreement with NMIA to
provide this information to ATC. 91
Flight crews rely on ATS to give them current and accurate runway conditions reports. MKJP
ATS was not in compliance with this requirement with regard to AA331. The controller should
have informed the AA331 flight crew that no braking action report had been received, in
accordance with the MANOPs.
Evidence from the Tower controller, and local knowledge, indicated that, even with a tailwind,
flight crews probably preferred to land using the Runway 12 ILS approach and landing when
arriving from the north, as it afforded a quicker and more convenient procedure than the
90
See 1.10.2.3
91
Section 1.10.4.6

238
Section 2, Analysis
circling for runway 30 or the RNAV (GPS) Runway 30 approach, and required less taxi time to
reach the terminal building. 92
JCAA ATS MANOPS specified that the active runway should be the one most in line with the
wind, if the wind was more than five knots. On initial contact with AA331, the Approach
controller advised the flight crew to expect an ILS runway 12, while advising them the wind was
310 degrees at 10 knots and that they might have to conduct a circling procedure for runway 30.
Based on the weather reported by the Enroute controller being ceiling 1,000 feet broken, the
circling approach would not have been possible since the ceiling was below the circling
minimums.
For AA331, a Boeing 737-800 aircraft, that is, Category C, the circling approach limits were 1,150
feet above sea level ground level, and 3.7 kilometers visibility. 93 The flight crew immediately
replied that they could take a straight in approach to ILS runway 12. The option to conduct an
RNAV (GPS) Runway 30 instrument approach to runway 30 was not suggested to the flight
crew. F u r t h e r m o r e , the flight crew stated that they were not aware of the option of the
RNAV (GPS) Rwy 30 approach, as further discussed in 2.7.
At 22:04 EST Manley Approach Radar informed AA331 that “… the information that was given
to you by the Enroute controller is still the same, visibility five miles and there’s moderate rain
at the station”. It appears that this weather given to AA331 by the Approach controller, which
was not from official weather reports, may have been derived from the AWOS information and
the Tower controller’s observations (See 1.1.2). The Approach controller should have stated to
AA331 that the weather he was giving was from controller observations. This was a
procedural error, and the reason for it was not determined. Also, it was not determined whether
or not the Approach controller had received the 22:00 EST (03:00 UTC) METAR by 22:04 EST.
The official METAR at the time of the final approach and landing of AA331, issued at 03:00
UTC, 22 minutes before the accident, was ceiling 1,400 feet broken, visibility 3,000 meters
(about two miles) 94 JCARs Tenth Schedule, sub-paragraph 10.665(a), states Visual Flight
Rules minima in controlled airspace are ceiling 1,000 feet and visibility 3 statute miles, if
reported. Thus the weather was officially Instrument Meteorological Conditions (IMC),
however the visibility given to AA331 by the Approach controller, based on a visual
observation by the Tower controller (See 1.1.5), was Visual Meteorological Conditions (VMC).
This difference may be ascribed to rain-shower activity, which can result in inconsistent visibility.
92
Appendix 4 (3)
93
Appendix 4(4)
94
Section 1.7.6

239
Section 2, Analysis
The 03:00 UTC (22:00 EST) METAR was “ … visibility 3,000 metres (about two miles) in heavy
rain showers, ceiling broken at 1,400 feet, few cumulonimbus clouds at 1,600 feet … ” (see 1.7.1
and 1.7.6). There is no evidence that this information was passed to the flight crew by the Tower
controller, nor is there any evidence to confirm that the flight crew received the 03:00 UTC
METAR on the ATIS. The RH (recorded history) on the flight dispatch document indicated that
the 03:00 UTC METAR was sent to AA331 at 03:15 and 03:17 UTC
(See 1.7.1) by ACARS, but there was no evidence, from the CVR or elsewhere, that the flight
crew took this information into their situational awareness (see definition of situational
awareness,), nor was it likely that the flight crew would look at an ACARS message at this busy
stage of the approach.
In accordance with JCAA ATS MANOPS the controller should have advised AA331 of the latest
runway condition report and braking action report. If no braking action report had been
received, the controller was required to state this to arriving aircraft. In the case of AA331,
none was available, and the controller did not inform the AA331 fli ght crew of this,
nor did the AA331 flight crew request this information. The reason for this was not determined.
This did not contribute to the accident, but it did show the flight crew’s low level of situational
awareness. This was another instance of an air traffic controller not following established
procedures in accordance with JCAA ATS MANOPS.
The investigation considered that the circumstances of the landing of AA331, in 1½ miles
visibility in heavy rain, at night, with a 14 knot tailwind, were difficult, and difficult to observe.
However, the Weather Standby system, as described in 1.17.3.2, was not activated by ATS. The
reason for this was not determined, but it may have been due to the fact that the weather
conditions were described only qualitatively, not quantitatively, and were not clearly defined.
Some of the discrepancies with ATS, as described above, may have contributed to the accident.
It was apparent that there were some instances where JCAA ATS procedures did not fully
conform with ICAO guidance material.
The investigation was unable to conduct an in-depth examination of the ATS organization, but
from what was revealed, as described in this section, the possibility of some systemic
weaknesses was indicated.

240
Section 2, Analysis
2.6 Meteorological Information
There was evidence that heavy rain was falling at the airport, before, during and following the
accident, with visibility as low as 1½ miles and ceiling broken at 1,400 feet, and the FDR data
indicated that the surface wind on runway 12 had a 14 knot tailwind component and a seven knot
crosswind component from the left, at the time of landing. 95
The AA331 flight crew was informed of the wind conditions by ATC several times, and
remarked in their post-accident interviews on the noise of the rain striking the aircraft. However,
at no time was the AA331 flight crew advised by ATC that the rain was “heavy”.
2.7 Operations
There was evidence from the AA dispatch document for flight AA331 that the aircraft’s weight
and centre of gravity were within the required limits for the duration of the flight, and the dispatch
of the flight was within the regulatory requirements. 96
The investigation determined that AA331 was legally dispatched. The interview with the
dispatcher (see 1.17.2.4) indicated that AA dispatched only to Dry or Wet runway conditions
The dispatcher stated that he had never had occasion to report braking action at Kingston and
all other Caribbean airports, to be other than good/wet. Also he stated that he had never
received a report of standing water at Kingston.
AA information provided to flight crews about the Kingston runways included a caution of the
possibility of pools of standing water on runway 12/30 after heavy rain, but neither the AA331
flight crew nor the dispatcher was aware of this. 97
The dispatcher also stated that “the captain was expected to select the best runway for
landing”. The investigation determined that, under the prevailing conditions, the best runway for
landing was runway 30, using the RNAV (GPS) Rwy 30 approach, but the captain did not select
this runway. 98
95
Section 1.11.2.2
96
Section 1.6.3 and 1.17.1.2.4
97
Appendix 5, Page 10-7X of AA Flight Manual, Part II
98
Section 1.17.1.2.4

241
Section 2, Analysis
2.8 RNAV (GPS) Runway 30 Approach at Kingston
The statements of both of the flight crew indicated that they were not aware of the RNAV (GPS)
Rwy 30 approach at Kingston (See 1.1.2) prior to the accident. The Jeppesen publications in the
aircraft’s cockpit library contained the plates for this approach and it was confirmed to have been
in the aircraft’s FMS database (See 1.6.12). It is possible the flight crew did not notice the
availability of this approach as they had been focused on programming runway 12 ILS into the
FMS. AA SOPs required that FMS approach programming be performed prior to the approach
briefing, and normally this happened prior to descent. At this time the most recent wind report
was 310 degrees at 9 knots, that is, suitable for runway 12.
At 22:04 EST, the Approach controller told AA331 to expect an ILS Runway 12 approach,
wind from 320 degrees at 10 knots and suggested a circling approach for runway 30, however the
flight crew chose to do a straight in approach for runway 12. At 22:14:34.7 EST the Approach
controller cleared the aircraft for the ILS Runway 12 approach and advised the wind was from
330 degrees at 15 knots. This wind information, eight minutes prior to landing, was the second
t i m e t h e A A 3 3 1 f l i g h t c r e w w a s i n f o r m e d t h a t the tailwind had increased. At no
time did the controller offer or ask the crew if the aircraft was capable of flying the RNAV
(GPS) Rwy 30 approach; instead, at 22:14:56.7 EST the Approach controller asked if AA331
was still able to do the straight in approach to runway 12, to which AA331 responded
affirmatively.
Although the AA331 flight crew was aware of the circle-to-land procedure to runway 30 from
the ILS Runway 12 approach and that option was suggested to them by the Approach
controller, they decided not to conduct that approach as it required a higher ceiling than was
reported and they decided (as stated in the first officer’s post-accident interview) that the straight
in to runway 12 was more appropriate than doing a circling approach, and had more chance of a
successful outcome. Given the weather conditions, and the fact that the flight crew was not aware
of the RNAV (GPS) Rwy 30 approach, this was an appropriate decision. However, the straight in
RNAV (GPS) Runway 30 approach would have offered a better option than the circle-to-land
procedure, with a minimum descent altitude of 373 feet agl, and the advantage of an into-wind
landing. All of the necessary elements for this flight to conduct the RNAV (GPS) Rwy 30
approach were available, except that the flight crew was not aware of it and the controller did not
offer it as an option, which resulted in a continuing lack of awareness of it by the flight crew.
If the flight crew had used the RNAV (GPS) Rwy 30 approach and landed on Runway 30, the
accident would probably have been avoided. At the time of handover to the Tower controller, it
was evident from the transcripts that the Tower controller thought AA331 had requested the
Runway 12 approach and so likely did not think to ask whether the AA331 flight crew would
like an RNAV (GPS) Rwy 30 approach.

242
Section 2, Analysis
2.9 Crew Resource Management (CRM)
At 22:17 EST the 03:00 UTC METAR was sent to the aircraft by ACARS, but there was no
evidence, from the CVR, or elsewhere, that the flight crew took this information into their
situational awareness.
The training required to be completed by the crew in the AA Advanced Qualification Program
(AQP) included CRM and the crew reported that they had completed it. AA AQP training
programs were approved by the FAA.
The flight crew stated they had discussed Kingston weather in Miami, and had discussed the issue
of enroute turbulence. At that time the available information to the flight crew indicated that the
Kingston weather forecast was within wind and weather limits for approach and landing on
runway 12 and offered no significant challenges. They knew there was an ILS Runway 12
approach, and were not concerned about the arrival. Scattered thunderstorms were in the
weather forecast, as were rain showers, normal for the Caribbean, and by using the aircraft
weather radar they could avoid these, so thunderstorms, while certainly of interest, would not
have been a major concern. In the interviews the flight crew stated that a thorough approach
briefing was conducted, although this was not captured on the CVR. The flight crew had prepared
for the ILS runway 12 ILS approach and flew direct to KEYNO as cleared by ATC.
It was clear from the CVR record that throughout the approach the flight crew members
were aware of weather, engine icing prevention, rain and cloud, and they turned the strobe lights
off to avoid flicker distractions in cloud. They were alert concerning the weather shown on the
aircraft radar during the approach and to the terrain constraints in the event of a missed approach.
They were also aware of, and focusing on, their fuel quantity limitation that allowed them only
one approach to land at Kingston. The fuel quantity was such that the aircraft was still slightly
above maximum landing weight approaching KEYNO, and they had made efforts during descent
and transition to burn off enough fuel to reduce the aircraft to maximum landing weight for the
landing at MKJP, by slowing the aircraft and using flaps early to increase drag and burn more
fuel. They were still focused on the landing weight consideration during the late stage of the
intermediate approach.
In the latter stages of the transition to the final approach fix, some eight minutes before landing,
AA331 was cleared for the ILS Rwy 12 and, in the same transmission, was advised by the
Approach controller that the wind was from 320 degrees at 15 knots. The first officer read the
clearance back and then the Approach controller asked if AA331 was still able to make a straight
in to runway 12 and gave the wind from 320 degrees at 14 knots. The AA331 flight crew
replied in the affirmative.
The CVR contained no spontaneous discussion between the flight crew in response to this weather
information that was given by the Approach controller, which was the second indication to the
flight crew that the tailwind had increased. This did not prompt further discussion between the
flight crew regarding adjusting the landing technique (using additional flap or braking) for the
increased tailwind, which would be expected. This lack of discussion may indicate that their
243
Section 2, Analysis
CRM was not adequate. It is possible that the flight crew may have been experiencing some
attention tunneling due to distraction by the heavy rain they were flying through, and
concentration on the weather picture on the radar between the aircraft and the runway. The
early flap selections and speed reductions for fuel still to be burned off for landing, and
announcing HUD settings and presentation may have also diverted their attention. As far as they
were conc ern ed , the wind was still within the company limit for landing. Less than five
minutes prior to landing, the runway was reported as “Wet” by the Tower controller. This
resulted in the first officer suggesting a change from autobrakes setting 2 to setting 3, to which the
captain agreed.
The crew’s CRM could have been better if they had conducted more discussion regarding the
increased tailwind, had adhered exactly to AA Bulletin 737-07, and performed a last minute
landing performance check when they knew the tailwind had increased. However, the evidence
indicates that they were of the belief that the l a n d i n g d i s t a n c e a s s e s s m e n t ( “advance
analysis”) the first officer described in his interview was sufficient to meet this requirement, and
it appeared possible that use of this was normal practice in the company, so no concerns were
raised. Until they were made aware of the runway surface condition as “Wet” by the Tower
controller they were proceeding with some level of complacency, landing in rain with a tailwind
(as they stated in interviews they had done in the past), maybe not the best condition, but certainly
not an unusual situation for them. Until the Tower controller’s report indicated to them that the
runway was “Wet”, they were proceeding as if the runway was dry to wet-good condition and
using autobrakes 2 setting. The report of runway “Wet” did not raise any concerns to them since
it had no modifying information, such as a braking action report.
Better CRM might have led them to a d i s c u s s i o n of the risks associated with a tailwind
landing on a wet runway with rain falling, but at this stage of the flight there was little or no time
for such discussion as they were almost on final approach. The only change in their plan was the
adjustment of the autobrake setting from 2 to 3. Any other change at this point, even a change of
approach to RNAV (GPS) Rwy 30, would have meant a go-around, with reprogramming the
FMS and considerable maneuvering. The crew had already decided against circling for runway
30 due to the weather conditions, and there was insufficient fuel for flying more than one
approach, then diverting to Cayman. The crew was therefore committed to the approach for
runway 12, without the option to fly the RNAV (GPS) Rwy 30 approach, even if it had been
offered by ATC.
The ILS Rwy 12 approach was flown accurately and on-speed and was therefore unremarkable.
The first officer gave extra callouts to assist the captain on short final, calling altitude at 500 feet,
on-speed and sink rate 800 fpm, all normal parameters. The aircraft crossed the threshold of
runway 12 at 70 feet RA, which was 20 feet higher than the ideal crossing height, and was on-
descent to the touchdown point or very slightly beyond it at that moment. The captain had
switched his attention from the HUD offset localizer to visual cues once the runway was visual,
and was using outside references for landing. Both pilots described the aircraft as being “in the
slot” as it crossed the threshold although this was confirmed by the FDR as being 20 feet high.
Just as the aircraft was crossing the runway threshold the FDR recorded a nose up pitch motion

244
Section 2, Analysis
on the controls and the rate of descent, which had been normal for that speed, was reduced. As
the autothrottle was still connected and operating in speed mode the aircraft stayed on speed and a
float ensued. Neither the captain nor first officer remarked on this.
It was raining heavily as the aircraft approached the runway and floated, before flaring and
touching down. The first officer did not alert the captain to the fact that the aircraft was still
airborne beyond the touchdown zone. Due to the heavy rain, lack of peripheral lights on the right
side of the aircraft and the first officer’s duties that required his attention in the cockpit, this may
not have been readily apparent to him. The captain disconnected the auto throttles manually at
about 35 feet RA, and retarded the throttles to idle. The FDR indications show pitch control
motions indicative of the captain “feeling for the runway”, and this prolonged the flare. The first
officer's duties during the flare and landing phase were to monitor the flight instruments,
observe the HUD pitch monitor and warn the captain if the pitch angle became too high, monitor
the spoiler deployment, engine thrust reverser deployment and autobrake activation and call if
there were deviations.
The RA auto voice call-out function was calling out the main wheel height above the runway as
the aircraft descended. The first officer's attention during this phase would have been divided
between the instrument panel and watching the flare and landing; also, during this time, he
reached over to confirm that the ground spoiler lever was in the armed position.
These duties required his attention during those few seconds, consequently he could not monitor
everything the aircraft was doing related to the landing. The first officer had a lot of
confidence in the captain's flying skills, as he stated in his interview, and it was possible that he
was complacent for a few seconds, coupled with the distractions of his normal duties.
The first officer described the aircraft as floating for a short while, as did the captain, but neither
of them reported a prolonged float, or of being aware of the aircraft passing the PAPI lights, or
realizing the aircraft had passed the touchdown zone while still airborne, so neither called for a go-
around.
Better CRM might have been displayed if there had been more discussion of the t a i l wind and
its effect on landing distance, or even discussion related to the choice of landing on runway 12.
However, it is probable that, since the crew was unaware of the option of the RNAV (GPS) Rwy
30 approach and believed that the landing distance assessment was covered by the “advance
analysis”, they did not enter any such discussion. Overall, the CRM interaction was about what
would be expected in the circumstances, and the only deficiency was that there appeared to be
some complacency regarding the severity of the landing conditions.

245
Section 2, Analysis
2.10 Go-Around
There was no instruction in the AAB737 Operating Manual to the effect that a first officer who
was the pilot monitoring could call for a go-around, with the pilot flying (that is, the captain)
being obliged to follow through. 99 The instructions in the AA Flight Manual indicate clearly that
the captain could disregard a first officer’s go-around call. 100
The investigation did not determine exactly what was the philosophy of AA regarding this,
but it was considered to be possible, despite the company procedures in AA Flight Manual Part
1, Chapter “Crew Qualification and Responsibility”, “First Officer Responsibility”, that a first
officer who was pilot monitoring might be reluctant to call for a go-around. 101
Although go-around after landing can be a hazardous procedure for several reasons, the option for
the flight crew to do so was open until such time as reverse thrust was selected, and the captain
reported that he considered this when the aircraft was on the runway and not decelerating as
expected (see 1.1.4), but chose not to do so. It is possible that, if the first officer had called for go-
around after threshold crossing and before the captain selected reverse thrust, then the accident
might have been avoided.
However, it was not guaranteed that a go-around would have been suc c essful i f at t em pt ed
during the few seconds after touchdown, before the reversers deployed, since the aircraft was
already on the ground decelerating, the engines were spooled down to idle, the flaps were at 30,
the runway might have had pools of water on it in places, there was less than half of the runway
remaining for takeoff and there was a strong tailwind.
FAA SAFO 1005, issued 3/1/10 (shortly after the accident), which recommended:
1. Either the pilot flying or the pilot monitoring may make a go-around callout, and
2. The flying pilot’s immediate response to a go-around callout by the non-flying pilot is
execution of a missed approach,
showed that the FAA was concerned about the standard operating procedures for go-around in the
industry, and the reluctance of first officers and/or monitoring pilots to call for go-around.
99
Section 1.17.1.1.9
100
Section 1.17.1.1.9
101
AA Flight Manual Part 1, Chapter “Crew Qualification and Responsibility”, Section 3, page 5, part 1.8, “First Officer
Responsibility”

246
Section 2, Analysis
2.11 Flight Crew Fatigue
Although the flight crew had just had three days' rest, and their flight/duty/rest times were within
the required limits, at the time of the accident they were at the end of the third flight of a long
duty day. The captain’s statement indicated that getting off the gate was usually the most
stressful part of a flight, and having so many passengers, the turbulence over Cuba and the noise
of the rain striking the aircraft would have added to the flight crew’s stress on this flight. The
flight crew had been on duty for nearly 12 hours, and awake for more than 14 hours, and it was
almost “bedtime” in their recent diurnal cycle. Although the flight crew had just had three days
off duty, they were at the end of a long duty day, and were possibly fatigued: however, the extent
to which this affected their performance could not be determined.
The information in 1.18.7, NTSB Review of Fatigue in Major US Accidents, includes findings
from this NTSB study regarding fatigue resulting in deteriorating decision making, task fixation
and reluctance to discontinue a flawed approach.
2.12 Pitch-Up Inputs and Visual Illusions
Just as the aircraft flew over the runway threshold at about 70 feet RA, the DFDR indicates that
the captain made several slight pitch up inputs. At this point the first officer made no
comment, and the captain continued with the landing. The following night visual illusion factors
and depth perception impediments may have contributed to the pitch up inputs and the subsequent
long landing:
1. The landing was conducted in the hours of darkness, with visibility as low as
2,200 metres in heavy rain, as recorded in the SPECI taken three minutes after the
accident.
2. Due to the absence of touch down zone and centre line lighting, and the absence
of reflective material in the runway marking paint on runway 12, most of the light from
the landing lights was reflected away from the runway surface and away from the aircraft
instead of back to the cockpit; this, together with the halo effect caused by the diffusion
of light by water when viewing runway edge lights through a wet windshield, may have
limited the captain’s depth perception cues, and made it difficult for him to judge exactly
the position of the touchdown zone.
3. The captain’s visibility may have been impeded by the heavy rain on the windshield, and
the rapid movement of the windshield wipers.
4. The captain did not have a clear far visual horizon for horizontal and vertical reference
due to the absence of lights beyond the runway and the reduced visibility in the
rain.

247
Section 2, Analysis
5. The possibility that the captain experienced a visual illusion that the aircraft was lower
than it actually was because its higher groundspeed, caused by the tailwind, resulted in
the runway lights going by faster, may have caused the captain to prematurely make the
pitch up inputs to flare for landing.
The combination of these factors may have made the captain uncertain of the position of the
aircraft in relation to the touchdown zone while descending for landing, and caused him to
inadvertently pitch the nose of the aircraft up.
2.13 Simulator Trials
The ¼ dot deviation above the glideslope noted during the simulator trials 102 happened as each
pilot disconnected the autopilot and manually turned to the right (southward) about ten to fifteen
degrees off the localizer track to intercept and then turn left (northward) again to line up
visually with the runway extended centre line, at about 300 to 500 feet agl. In some cases a
slight second correcting turn either right or left was required to establish the drift correction. All
of these four or five examples resulted in the aircraft being about a quarter dot high on the ILS
glide-path after lining up with the runway centre-line, perhaps due to the aerodynamic
characteristics of the aircraft in the turn, or due to pilot technique in pulling nose up slightly on the
elevators due to the visual illusion of descending in the turns.
It was concluded that this slight pitch up event may have been because the pilots had disengaged
the autopilot and were manually flying the simulator.
As the simulator did not accurately replicate the wet runway conditions, no conclusion could be
reached from the simulator trials regarding the aircraft stopping performance.
2.14 Runway Condition Report “Wet”
There was no record on the CVR or the ATC tapes or any other evidence of the AA331 crew
having received a runway condition report from any ATC controller or elsewhere from the time of
the AA331’s initial contact with the Enroute controller at 02:47:00 UTC until informed “Be
advised runway wet … ” by the Tower controller, given with the landing clearance at 03:17:57
UTC, 4 minutes and 44 seconds before the accident. The information given to the crew by the
Approach controller at 02:59:15 UTC, “he didn’t have any problem coming in”, was not a
runway condition report. 103
It was apparent that there were some differences in the use of the term “Wet” to describe
102
Section 1.16.3
103
Appendix 3

248
Section 2, Analysis
the runway condition. The evidence indicated that to the AA331 flight crew an unqualified “Wet”
report meant AA “Wet/Good”, that is, less than 1/8 inch of water, 104 whereas to the Tower
controller “Wet” meant simply that the runway was wet, based on the fact it was raining, but did
not include any measurement of depth or any description of the runway condition. In ICAO terms
“Wet” meant “the surface is soaked, but there is no standing water.” 105
Although the flight crew had not requested or received a runway condition report until told by
ATC “Be advised, runway wet” when they got their landing clearance, it is likely that the flight
crew felt confident that the aircraft could land safely on the runway.
However, the flight crew:
 Had some evidence that there could be moderate to heavy rain at the airport,
indicating that the worst-case braking action might be worse than Wet/Good (in the
captain’s interview he mentioned heavy rain, ATC reported moderate rain, aircraft radar
was giving yellow return, the captain stated “That’s some good rain shower sitting down
here” from observation of the weather radar return, a n d t h e aircraft was in what was
recorded as “heavy rain”, (see 1.7.11, and as per the 03:00 UTC Metar), during the final
approach and landing stage).
 Did not request a runway condition report before completing the approach
briefing, or subsequent to this.
 Did not receive the (unsolicited) runway condition report from ATC of “Be
advised, runway wet.” until less than 5 minutes before landing.
 Did not request a braking action report, and ATC did not inform them that none had
been reported.
Furthermore, the flight crew’s briefings appeared to be incomplete and almost certainly did not
include a review of Page 10-7X, which cautioned regarding the potential for standing water after
heavy rain. Although the captain did not indicate in his interview whether or not he was aware
of Page 10-7X before the accident, if he was aware of it before the accident then he should have
anticipated pools of standing water, and done a Landing Distance Assessment for Fair/Medium
braking action conditions, accordingly. In the reported weather conditions present (rain and
thunderstorms) this warning should at least have been part of the Approach/Descent briefing, in
which case the first officer would have been aware of it, but the first officer stated in his
interview that he was not aware of it. Therefore it is justified to say that the captain was not
aware of this warning before the accident. If the crew had been aware of this, it might have
prom pt ed t hem t o change thei r plan t o land in heavy rain conditions, and the accident
might have been avoided.
104
Section1.17.1.2.10(6)
105
Section 1.10.4.5

249
Section 2, Analysis
ATC should have given AA331 the runway condition report and braking action report (or lack of
such) on initial contact. 106 It would have been better airmanship on the part of the flight crew if
they had requested this information when it was not offered by ATC, as this information is a
normal part of a flight crew’s landing planning, and situational awareness. However, the flight
crew’s passivity in this regard was likely due to an expectation that ATC would have advised
them of any significant information without being prompted. Also, in accordance with AA
Bulletin 737-07, the flight crew was directed to confirm whether or not landing conditions had
changed from the time of dispatch, hence it was necessary for the flight crew to obtain this
information for approach and landing planning.
When the AA331 flight crew received this report of “Runway wet”, the first officer said to the
captain “Runway’s wet. You want to go to brakes three?” (that is, change autobrake setting from
2 to 3). The captain agreed to changing to autobrake 3.
There was no further discussion of applying AA recommended guidance for tailwind landing on a
wet runway, which the investigation considers would have been prudent.
The different interpretations regarding the meaning of the word “wet” probably were confusing. If
the flight crew had been aware that the runway condition could have been less than Wet/Good, they
might not have attempted to land on runway 12 with a 14 knot tail wind at that time. The lack of
discussion on the part of the flight crew might have been due to a possible expectation that ATC
would advise them of any significant conditions.
2.15 Inconsistency between AC91-79 and SAFO 06012
As stated in 1.17.4.3, it should be noted that AC91-79, Appendix 4, 10 b. (2), as above, stated “If a
runway is contaminated or not dry, that runway is considered wet”, and that SAFO 06012, item 4.
Definitions, stated “i. A wet runway is one that is neither dry nor contaminated”. These two
statements are inconsistent. As they are both FAA advisory material, these statements should be
amended so that they are consistent.
2.16 Kingston Field Report
As described in 1.10.4 of this report, there was some confusion regarding responsibility for
obtaining runway condition reports and how the information was disseminated to operators and
flight crew.
The evidence indicated that the “WET 0.10 IN WATER” report on the runway was a
standard one given by AA personnel at Kingston to AA dispatch in Dallas when it rained, and
106
Section 1.17.3.3

250
Section 2, Analysis
that actually no observation or measurement was made, or obtained from any other source, by
personnel of American Airlines.
This was contrary to American Airlines procedures, as described in the AA Kingston

Station Manual. 107 The reason that these procedures were not followed was not determined. AA
management should have ensured some means of obtaining a runway surface condition report.
ICAO guidance suggested that the airport operator, in this case NMIA Airports Limited,
should inspect the runway and advise ATC of runway conditions during periods of inclement
weather, but this guidance was not adopted and included in the MKJP Operations Manual as an
instruction to staff, and was not done by MKJP during the approach of AA331. 108
If a runway surface condition inspection had been conducted and passed on to ATC, it might have
alerted the flight crew to the possibility the runway braking action was not AA Wet/Good, and
prompted them to abandon their plan to land on runway 12 in heavy rain with a 14 knot tail wind.
ICAO guidance in Doc 4444 Para 7.5 required ATC to transmit the runway condition to incoming
traffic during periods of inclement weather, but the NMIA ATC Unit Specific instructions did not
require ATC to obtain runway condition reports from NMIA, and the ICAO definitions for
reporting water on a runway (See 1.10.4.5) were not in JCAA ATS MANOPS nor in the NMIA
Operations Manual. ATC simply reported “runway wet” when it rained. This may have been
the result of the lack of any formal arrangement between the parties for the provision of runway
condition reports.
The JCAA ATS MANOPS and the NMIA Manual should be amended to include these
definitions.
In the case of AA331, the evidence indicates that AA331 was not given any runway condition
report by ATC until the Manley Tower said “ … be advised runway wet …” less than 5
minutes before AA331 landed. In any event, the statement “ … be advised runway wet …”
was not a runway condition report, as per ICAO recommendations, since the runway had not been
physically inspected.
In addition to this, and as stated in the Jamaica AIP 109, there was no technical means for
measuring standing water at NMIA, although the ICAO guidance could easily have been
followed by NMIA (See 1.10.4.5).
107
Section 1.10.4.4
108
Section 1.10.4.6
109
Section 1.10.4.1

251
Section 2, Analysis
2.17 Determination of Depth of Water on Runway, and Runway Braking Action
As described in 1.18.6, the investigation determined that the stopping performance of

AA331 on the runway was equivalent to AA RRLD Fair/Medium braking action.
This finding was supported by the:
 Captain’s description of feeling “like he was on ice” after landing,

 Captain’s statement that he thought the runway was “under water”,
 Lack of deceleration with maximum manual braking,
 Evidence on some of the main landing gear tires that the braking was friction limited,
 Reports of heavy rain.
It should be noted that the water depth over an entire runway cannot be accurately measured, and
qualitative reports made from visual inspections are needed in lieu of quantitative reports. ICAO
does have guidance for description of such conditions. 110
2.18 Survival Aspects and Injuries
The L1 door was jammed because the escape slide pack contacted the floor when it was
prematurely released from the escape slide compartment. This could have had serious
consequences and have resulted in serious injuries or fatalities if there had been a post-impact
fire, or the aircraft had been submerged in the sea.
Apart from the L1 door, all the other exits worked, and the evidence indicated that all the
passengers were evacuated from the aircraft within about 15 minutes of impact.
It was not determined why the R1 slide deflated. The L2 and R2 slide packs were released
from the slide compartments, but did not inflate because the fuselage was too close to the
ground for the slide packs to fall far enough to activate the inflation mechanisms.
Three of the four cabin crew members, according to their interviews, and 15 of the 17
passengers who answered questionnaires reported seeing no lights. However, many of the
emergency lighting batteries were found discharged post-accident, indicating that some of the
emergency lighting systems may have been working. There was insufficient evidence to
determine conclusively whether or not the emergency lighting system worked as designed after the
impact.
The evidence indicated that some of the passengers may have been injured by falling overhead
bins and passenger service units.
110
Section 1.10.4.5

252
Section 2, Analysis
The captain’s reported back injuries were possibly attributable to the failure of the captain’s seat
crotch belt bracket, the uncertified seat bottom cushion, or a combination of these two
conditions. 111.
2.19 AA Bulletin 737-07 and Landing Data Card
AA Flight Safety Programs Management stated that American Airlines Bulletin 737-07 was
generated for American Airlines to comply with FAA SAFO 06012.
Despite use of the word “should” in Bulletin 737-07, the AA Flight Safety Programs Manager
stated: “AA management both expects and assumes flight crews will follow the instructions
(recommendations) in Bulletin 737-07”. 112 This left open the question as to whether or not it was
mandatory for AA flight crews to conform to Bulletin 737-07.
AA stated that Bulletin 737-07 was part of the AA331 flight crew’s Performance training,
therefore it was assumed that the flight crew was fully aware of the information in this bulletin.
However, the investigation was unable to determine why AA had not made it clear to the AA331
flight crew that following the directions in Bulletin 737-07 was mandatory (if that was their
intention), nor why the AA331 flight crew did not follow these directions.
2.20 Bulletin 737-07 and “Advance Analysis”
AA complied with SAFO 06012 with Bulletin 737-07, 113 which stated “ … the flight crew should
use the charts …”; however the AA331 flight crew were using the “advance analysis”, which was a
“worst case scenario”, although AA stated:
“no AA documents or training materials specifically define the “advance analysis” concept
that the JCAA cites in the draft report. AA flight crews are not trained to use, required to
use, or discouraged from using this method”.
The reason for this contradiction was not determined.
It should be noted that the “advance analysis” did not specify aircraft configuration at landing, 114
and the initial configuration planned by the flight crew, that is, autobrakes 2 and flap 30 would not
111
Section 1.12.6.2
112
Section1.17.1.2.10 (4)
113
Section 1.17.4.2 and Appendix 6
114
Section 1.1.2

253
Section 2, Analysis
have enabled the aircraft to stop on a runway with AA Fair/Medium braking action with a 14 knot
tailwind, with a 15% safety margin, even if it had landed at 1,000 feet from the threshold.
Furthermore, there was no evidence as to whether or not the "advance analysis" used by the AA331
flight crew had been modified to account for the increased landing distance that resulted when the
AA B737 tailwind landing limit was increased from 10 knots to 15 knots.
2.21 AA Flight Crew Landing Distance Assessment
During his interview, the first officer described using a process for landing distance assessment
that essentially constituted a worst case scenario analysis. The investigation was unable to
determine exactly what the source of the data was used on this flight.
The first officer said they did not have to calculate landing distance before each landing. They had
landing performance data and a tailwind was taken into consideration in that data. He said on a
runway like at Kingston where it was 9,000 feet long, with wet conditions, there was no
problem with the performance data for landing flaps 30 and they had both done this many times.
He also stated that for runways shorter than 8,000 feet, if it was wet, the more factors like wet
runway, tailwinds to consider; they were cumulative based on field elevation and landing weight.
He said he was very comfortable with the runway data but it was more critical when there was
a shorter runway and tailwind. 115
For the purposes of the investigation, this was taken to mean that the AA331 flight crew believed
that if the runway was more than 8,000 feet, and the tailwind was 15 knots or less, on a wet
runway, with the aircraft at maximum landing weight of 144,000 pounds, with flap 30, then a
safe landing was assured, and it was not necessary to calculate landing distance before each
landing. The investigation also considered that it might be possible that the above assumption was
based upon a 9,000 foot runway, rather than an 8,000 foot runway.
This method of estimation was later referred to as “advance analysis” by the American
Airlines Flight Safety Programs Manager (see 1.17.1.2.10, #7).
115
Section 1.1.2

254
Section 2, Analysis
When the AA Flight Safety Programs Manager was asked:
“Why did the AA331 flight crew use an assumption that if the runway was more than
8,000 feet, was “wet”, the tailwind was less than 15 knots, aircraft at max landing
weight, then it was safe to land? Is this method approved by AA, taught by AA,
commonly used by AA flight crews?”
he responded:
A runway condition report of “Wet” with no other modifying information, e.g. “braking
action Poor”, would indicate braking action ‘Good’. An acceptable technique for a flight
crew who flies into a certain airport frequently is to conduct an “advance analysis” of the
worst case scenario for the landing runway, that is, for a known landing length, braking
action, wind component, landing weight, etc. the flight crew could determine in advance
that as long as they landed below the maximum weight for these worst case conditions the
runway length was acceptable. A review of this “advance analysis” prior to landing using
the actual conditions at time of landing is acceptable (Emphasis added).
The AA manuals contained all the required data to make such a worse-case calculation; however
they did not describe or recommend this technique. Guidance to AA flight crews was contained in
AA Bulletin 737-07, which recommended confirming landing performance limits just prior to
landing using the actual runway conditions at the time of landing, noting that if landing conditions
had not changed from time of dispatch there was no need to do this assessment again.
The “advance analysis” used by the f l i g h t crew of AA331, and as a l l u d e d t o by the first
officer in his interview, did not specify the following aspects:
1. Method of braking,
2. Use of reverse thrust,
3. Distance of touchdown point from the runway threshold.
Also, it assumed Wet/Good braking action at worst, without confirmation of this.
The question remains, how did the AA331 flight crew arrive at the “advance analysis” of, “if the
runway was more than 8,000 feet, and the tailwind was 15 knots or less, on a wet runway, with the
aircraft at maximum landing weight of 144,000 pounds, with flap 30, it was not necessary to
calculate landing distance before each landing”?
2.22 Calculation of landing distances
There was no evidence that the flight crew had followed the exact recommendations of Bulletin
737-07 to perform a landing distance assessment, but instead, as described by the first officer in
his interview, had used a landing distance assessment worst case scenario, that is, “advance
analysis”, which one AA manager considered to be an acceptable means of compliance. The exact
figures used to make this “advance analysis” were not stated during the flight crew interviews and
255
Section 2, Analysis
the investigation later used the AA B-737 Landing Data Card tables to reconstruct a worst case
scenario based on the following assumptions:
 Aircraft at maximum landing weight

 Tailwind component 15 knot limit
 Flaps 30 (as used)
 Auto spoiler deployment
 No reverse thrust for AA Dry Runway and AA Wet/Good braking action: two engine,
detent two, reverse thrust for AA Fair/Medium braking action.
 Use of maximum manual braking or maximum autobrakes
 Sea level
 On speed
 On path and slope
 Anti-skid wheel braking operative
The Dry runway figures are from demonstrated certification flight tests, and include the
demonstrated air distance to touchdown (assumed to be about 1,000 feet), and a 67% safety factor.
The Wet/Good braking action data include the demonstrated dry runway air distance (assumed to
be about 1,000 feet) and safety factor of 92%, while the figures for Fair /Medium braking action
are derived using a constant aircraft braking coefficient associated with the braking action/runway
surface condition, air distance of 1,000 feet and considerations for the aircraft configuration,
spoiler use, reverse thrust, autobrake MAX setting or maximum manual braking, plus a 15%
margin.

256
Section 2, Analysis
The table below compares the worst case scenario at 15 knot tailwind component with the actual
tailwind component and the reported tailwind component that could have been used in the
calculation of landing distance required at the time of the approach briefing at top of descent.
Tailwind Component Dry Wet/Good Fair/Medium

Runway braking braking action
action
67% 15 % safety factor
safety 92 % safety
factor factor
(feet) (feet) (feet)
15knots (maximum tailwind landing limit) 7,230 8,120 9,000

(1,000 foot air distance included).
Actual, at landing, 14 knots (1,000 foot air 7,130 8,010 8,870

distance included).
Reported at 02:48 UTC, about the time of the 6,420 7,290 8,030
approach and landing briefing, 310/7.5,
equivalent to a 7 knot tailwind (1,000 foot air
distance included).
Received at 02:35 UTC was 02:28 UTC 6,620 7,500 8,270

SPECI via ACARS, 310/09 knots equivalent
to 9 knot tailwind (1,000 foot air distance
included).
Table 7 : Tailwind Component
Note 1: The numbers in this table are derived from the American Airlines B737-800 Required
Runway Landing Length tables in Appendix 7.
Note 2: The numbers in this table do not include the effect of 5 knots above Vref.

257
Section 2, Analysis
Tailwind Component Dry Wet/Good Fair/Medium

Runway braking action braking action
(feet) (feet)
Autobrake 3 un-factored (1,000 foot air 7,110 7,140 8,250

distance included and 14 knot tail wind).)
Autobrake 3 factored (+ 15%)(1,000 foot 8,170 8,210 9,490

air distance included, and 14 knot tail
wind).)
Autobrake Max un-factored (1,000 foot air 4,840 5,930 8,000

distance included and 14 kt tail wind)
Autobrake Max factored (+ 15%)(1,000 5,570 6,820 9,200

foot air distance included, and 14 knot
tailwind)
Table 8: Autobrake Data
Note 1: The Autobrake 3 data for 30 flap comes from the Boeing B737-800 Landing Distance
tables in Appendix 9. This information would not have been available to the AA331 flight crew
on the accident flight.
Note 2: The numbers in this table include the effect of 5 knots above Vref, temperature correction
for ISA +5° C and two engine, detent two reverse thrust.
1. For Runway length 8,000 feet or more.
If an 8,000 foot runway is accepted for the “advance analysis”, then the “advance analysis” was
either incomplete or incorrect, because it did not accommodate a 15 knot tail wind with 30 flap.
The following scenarios fall within an 8,000 foot runway “advance analysis” for a wet runway
surface condition, or a runway with Wet/Good braking action:
(a) 10 knot tail wind, with 30 flap (former tail wind landing limitation).
(b) 15 knot tail wind, with 40 flap (tail wind limitation at time of accident).
(c) 13 knot tail wind, with 30 flap (maximum tail wind for 30 flap landing).

258
Section 2, Analysis
2. For Runway length 9,000 ft. or more.
If a 9,000 foot runway is accepted for the “advance analysis”,
(a) Covers all scenarios for runway Wet.

(b) Covers accident scenario, including Fair/Medium.
(c) Covers all scenarios including Fair/Medium, but yields slightly less than 15% margin for
Kingston.
The above data indicates that a landing within the runway, in Fair/Medium conditions, could have
been achieved, but not with a 15% safety margin, had the flight crew landed the aircraft at 1,000
feet and used the AA recommended maximum deceleration techniques for the circumstances.
The CVR indicated that the flight crew originally planned to use autobrake 2 setting for
deceleration and later, on being informed that the runway was wet, elected to use autobrake 3
setting. Post accident analysis of the FDR data indicated that the aircraft could have been stopped
on the available runway using autobrake 3 setting, with no safety margin, if it had touched down
within the first 1,600 feet, and reverse thrust was immediately applied.
The captain stated that he thought for this landing he was better off with flaps 30, and he said
flaps 40 can cause the aircraft to float sometimes. The aircraft was landed with flap 30.
2.23 Flight crew compliance with Bulletin 737-07
Although Bulletin 737-07 stated “ … the flight crew should … ”, it is evident from the statement
of the Flight Safety Programs Manager (see 1.17.1.2.10, Item 4) that AA expected the flight crews
to comply with this bulletin. 116 The flight crew, in using the “advance analysis”, was therefore
not strictly following the guidance of Bulletin 737-07. However, the AA Flights Safety Programs
Manager indicated for the investigation that he considered the "advance analysis" method to be an
acceptable means of complying. The reason for this contradiction was not determined.
2.24 Discussion of “advance analysis”
The investigation was informed that the FAA advocated and briefed the equivalent concept of
“advance analysis” within the aviation industry in 2006, prior to the release of SAFO 06012. It
was stated that operator time of arrival landing distance assessments can be effectively
implemented in multiple ways, including, but not limited to, a pre-calculated matrix of
conditions/thresholds that flight crews may be able to quickly reference or recall to ensure that
landing performance requirements are met for many landing conditions (without repeated or
116
Section 1.17.1.2.10, Item 4

259
Section 2, Analysis
detailed calculations prior to every landing), and that an “advance analysis” may take into
consideration the most adverse conditions likely to be encountered.
The discussion in 1.17.4.2 refers to voluntary compliance with OpSpec/MSpecC382, but there
was no evidence that AA had used this option.
The investigation was unable to further research or substantiate this information.
However, the investigation believes that there may be risk inherent in entrusting critical
performance information to memory, and believes that when the advance analysis technique is
utilized that the operator should ensure the flight crews employ the recommended aircraft
configuration and deceleration techniques to match the runway conditions.
AA stated that no such concept as the “advance analysis” was defined, approved or trained by
American Airlines. The investigation determined that a flight crew using an "advance analysis"
needs to know what conditions (that is, assumptions, technique, deceleration devices, runway
surface condition) the analysis is based on. The "before landing" assessment is then to
confirm whether or not the time of arrival conditions are within those used in the
"advance analysis."
2.25 Placement of Bulletin 737-07 in AA Manuals
AA Bulletin 737-07 stated:
“ … if (landing) conditions change, or deteriorate (as they had, as described in

2.23), the flight crew should use the charts on the revised Wind Component landing Data
Card … to confirm adequate runway length for landing (taking into account)
meteorological conditions, surface condition of the runway … (etc.).
Bulletin 737-07 was specifically addressing an action, that is landing performance assessment,
required by AA flight crews to complete before landing. This information should have been
included or referenced in the AAB737 Aircraft Operating Manual, Vol 1, P. 5 – 7,
“Considerations prior to descent”, but was instead placed in the Performance Section. This
may have resulted in it not being brought adequately to the attention of AA flight crews.
There was no mention of, or requirement to, perform a landing performance assessment
per se in the above mentioned “Considerations prior to descent”. 117
117
Section 1.17.1.1.5

260
Section 2, Analysis
2.26 Issue of Autothrottle/Autopilot use
The investigation considered that the captain’s leaving the auto-throttle engaged (after
disengaging the autopilot) until the aircraft was about 1,200 feet past the threshold, at about 40
feet RA 118 was a factor contributing to the accident, as the aircraft then touched down at 148
knots, that is, Vref + 5, instead of the desired Vref of 143 knots 119, making the landing ground
speed faster, the flare longer, and the required stopping distance longer.
As described in 1.17.2.4, “Autopilot/Auto-throttle Use During Approach”, it is stated in the

Boeing 737 NG Flight Crew Training Manual, page 1.34, that auto-throttle use is recommended
only when autopilot is engaged in CMD.
Furthermore, although it was not established whether or not AA management was aware of the
article by Bill McKenzie of Boeing 120, and this article referred to the Boeing 757, there was no
evidence that the article was not valid for the Boeing 737-800, and it presented strong
arguments for not using auto-throttle when auto pilot was disengaged.
The investigation determined that:
AA did not follow Boeing’s recommendation in this regard.

 The FAA approved the AA B737 Flight Crew Training Manual without this
recommendation being included.
 AA flight crews were not required to follow the Boeing guidance on use of the auto-
throttles.
The investigation considers that the Boeing guidance and the McKenzie article provide
justifiable reasons for following the Boeing guidance on the auto throttle and that there is
sufficient evidence for the FAA to require that all Boeing 737-800 operators follow the Boeing
guidance in this regard.
118
See 1.11.2.2
119
Section 1.1.3 and 1.6.4
120
Section 1.17.2.4

261
Section 2, Analysis
2.27 FAA Safety Information SAFO 06012 and AC 91-79, and AA Bulletin 737-07
The investigation reviewed NTSB recommendation A-07-61, FAA AC-91-79 and AA Bulleting
737-07 guidance and safety recommendations in relation to this accident and determined that
there is:
 No evidence that AA incorporated the guidance in AC 91-79 into AA guidance material,

nor was it mandatory for AA to have done so;
 Evidence of conflict within AC 91-79 in relation to runway surface condition assumptions

(See 1.17.4.3 last sentence in paragraph 10) and between that conflict and the contents of
paragraph 11 of AC 91-79; and,
 Evidence of conflict between AC 91-79 and SAFO 06012 with relation to the conflict
within AC 91-79 (cited above).
With reference to the above:
The statements “If a runway is contaminated or not dry, that runway is considered wet” from AC
91-79, Appendix 4, Page 3, has two possible literal interpretations:
1. If a runway is contaminated, that runway is considered wet.

2. If a runway is not dry, that runway is considered wet.
Interpretation 1 conflicts with guidance below which states “If there is rain actively falling
on the runway, standing water should be assumed”.
Interpretation 2 is incomplete as it does not account for any contaminated runway

performance levels (when contaminants other than water are reported).
Interpretations 1 and 2 conflict with Guidance from SAFO 06012 in Item 4i,
Definitions, Runway Surface Conditions (Page 2).
The investigation believes that there is a safety risk if guidance material provided to the industry
contains information that may not be correct or that may conflict with other pertinent guidance
material, and that clarity, accuracy and consistency are needed to avoid any ambiguity or
misunderstanding on the part of the target audience of flight crew and operators.

262
Section 2, Analysis
FAA SAFO 06012, dated 31/08/06 and FAA AC 91-79, dated 06/11/07 121
were available to American Airlines.
SAFO 06012, 6 e 2, stated that standing water is equivalent to Poor Braking Action, and this was
passed on to the flight crews in Bulletin 737-07.
AC 91-79, Appendix 4, Paragraph 11, stated “If there is rain actively falling on the runway,
standing water should be assumed” (Note: this is NOT qualified by the preceding
statement “ ... if there is no clear report of runway condition … ”) (see 1.17.4.3.). Thus, if
this recommendation was followed, AA flight crews should have been using Fair/Med or Poor
Required Runway Landing Length tables for landing distance calculations when rain was
actively falling, rather than Wet/Good. There was no evidence that AA had passed this
statement on to their flight crews in the form of a Bulletin or manual amendment. It is possible
that the flight crew of AA331 might not have attempted the landing under the assumption that the
runway braking action was AA Wet/Good if they had been instructed to assume standing water
when rain was falling on the runway.
The AA Performance Manual, Section R14, Bulletin 737-07 (dated 11-26-06), as quoted in
1.17.1.1.5, stated: “The new FAA recommendation is to confirm landing performance limits
just prior to landing, using the actual runway conditions at time of landing.” This information
was not appropriately placed as a mandatory part of the approach briefing to be performed by
the flight crews, but was more of a reported recommendation in the Performance section, and
the wording does not appear as a clear AA instruction to the flight crews. It appears that the
flight crew of AA331 were not familiar with this information, as they continued with the
approach for landing in heavy rain and a tailwind close to the limits, assuming the runway was in
a Wet/Good condition without confirming the landing performance limits just prior to landing.
2.28 Landing Conditions and Landing Performance Assessment
AA Bulletin 737-07 stated “If the landing conditions, from the time of dispatch do not change,
there is no need to do this assessment.”, and “ … if landing conditions change, the flight crew
should … confirm adequate runway length for landing. This assessment must take into account
meteorological conditions …” Thus it should be noted that, in accordance with AA Bulletin 737-
07, “landing conditions” included meteorological conditions as well as runway conditions.
The “KIN FIELD REPORT” received by the flight crew at the time of dispatch was “WET
0.10 IN WATER”. There was no evidence that the flight crew received any updated runway
condition report which would have been used in the approach briefing, and the runway condition
report of “Wet” from ATC was a very general term differentiating Wet from Dry, and was not
121
Section 1.17.4.3

263
Section 2, Analysis
referring to less than 1/8 inch of water, as was the AA Wet/Good. The report was received less
than five minutes before landing, long after the pre-approach briefing.
The forecast landing conditions at Kingston at the time of dispatch were:
KIN 221500Z 2218/2318 12016KT 9999 -RA FEW016 BKN032

TEMPO 2218/2302 6000 TSRA BKN016 SCT018CB BECMG 2302/2304 35008KT
TEMPO 2304/2312 8000 SHRA BKN016 FEW018CB SCT032 BECMG
2314/2316 20006KT
that is, Kingston, issued 22 December (2009) at 15:00 (all times UTC), forecast from 18:00 on
22 Dec to 18:00 on 23 Dec, wind from 120 degrees at 16 knots, visibility more than 10
kilometres, light rain showers, few clouds at 1,600 feet agl, broken cloud at 3,200 feet agl,
temporarily from 18:00 on 22 Dec to 02:00 on Dec 23 visibility 6000 metres, thunderstorms and
moderate rain, (ceiling) broken cloud at 1600 feet agl, with scattered cloud with cumulonimbus
cloud at 1800 feet agl, becoming from 02:00 to 04:00 wind 350 degrees at 8 knots (balance not
relevant to investigation as the aircraft was estimated to land at Kingston before 0400).
The Special Weather Observation (SPECI) received by AA331 from dispatch at about
21:48 EST, around the time the flight crew did the approach briefing, was:
SPECI MKJP 230228Z 31009KT 5000 TSRA BKN014 FEW016CB SCT030 BKN100
22/19 Q1013
That is, Kingston (MKJP) special weather observation at 21:28 EST (02:28 UTC, 23 Dec), wind
310 degrees at 9 knots, visibility 5,000 metres in thunderstorms and moderate rain, ceiling broken
at 1,400 feet, few clouds at 1,600 feet in cumulonimbus clouds, scattered clouds at 3,000 feet,
broken cloud at 10,000 feet, temperature 22° C, dew point 19° C, altimeter setting 1013 mb.”
This was sent to AA331 by dispatch at 21:48 EST, because thunderstorms were not in the
previous METAR.
Comparison of the presumed weather at the time of dispatch to “the actual landing conditions at
the time of landing”, that is, the weather given by the Approach controller at 03:04 UTC, shows
that the wind had changed from forecast 350 degrees at eight knots to reported 320 degrees at 10
knots, visibility had changed from forecast three and a half miles (6,000 metres) to reported 5
miles, and ceiling had changed from forecast 1,600 feet to reported 1,000 feet. Strictly speaking,
the “actual landing conditions at the time of landing” included wind from 320 degrees at 14 knots.
Thus, in accordance with the requirements of Bulletin 737-07, as the weather had not significantly
changed when the landing briefing was probably completed at top of descent, a landing distance
assessment was not required at that time; however, when the wind reports increased to a 15 knot
tailwind at about 03:15 UTC, about seven minutes before landing, and the increased wind speed
would have required a landing distance assessment, it was too late in the approach for the flight
crew to do so. Nevertheless, there was no obligation for the flight crew to continue with the
landing under these circumstances, and it would have been prudent for them to have performed a
go-around.

264
Section 2, Analysis
The information regarding the new 15 knot tailwind limit, SPC MSG NBR 9482, was in the
AA331 dispatch document, Page 18, and it required pilots to determine landing performance
when landing with tailwinds.
It read:
/// SPECIAL INFO MESSAGES ///

SPC MSG NBR 9482
EFFECTIVE IMMEDIATELY, AA 737 AIRCRAFT ARE AUTHORIZED TAKEOFFS AND
LANDINGS WITH UP TO AND INCLUDING 15 KNOTS OF TAILWIND COMPONENT FOR
VISUAL, CIRCLING, NON-ILS, AND CAT I ILS APPROACHES ONLY. THE
RESTRICTION OF 10 KNOTS TAILWIND MAXIMUM STILL APPLIES FOR HUD LOW
VISIBILITY TAKEOFFS AND CAT II OR III APPROACHES. RESTRICTED CAPTAINS
EXERCISING FAA EXEMPTION 5549
PARAGRAPH 2.2. AS ALWAYS, PILOTS MUST ENSURE THE REPORTED
As confirmed by American Airlines, the words "must ensure" indicated that it was not optional
and it was mandatory for it to be done for all tailwind landings. “Performance requirements”
are not the same as “tailwind limitations.”
The evidence indicated that AA informed the Boeing 737-800 f l i g h t crews of this new
increased tailwind limitation without providing them with any additional training, or making any
restrictions for wet or contaminated runway operations, or un-grooved runway operations. It also
indicated that the persons in AA who had responsibility for training may have been complacent
regarding the hazards of tail wind landings. 122 The same concern applies regarding the FAA
personnel who provided oversight for American Airlines, and who approved the increase in
tailwind landing limits from 10 to 15 knots, without requiring a training program for tailwind
landing, although this may have been because the FARs did not require it to be trained or tested.
There was no evidence of any discussion between the two flight crew members of AA331
regarding these requirements for landing performance assessment, that is, Bulletin 737-07 and
SMN 9482, and they were not mentioned by either pilot in the interviews. In his interview
the first officer, however, alluded to using an “advance analysis” 123, which was interpreted by the
investigation to mean that it was safe to land on a wet runway of more than 8,000 feet with a
tailwind of up to 15 knots. An examination of this “advance analysis” revealed that it
depended on the runway being at the worst Wet/Good, as per the AA RRLL tables (See
2.22), and that landing with 30 degrees of flap, maximum braking at touchdown, at about 1,000
from the threshold was required to achieve the published performance.
122
Section 1.17.1.2.9
123
Sections 1.1.2

265
Section 2, Analysis
Although the first officer stated (See 1.1.2) that during the flight they received updates from
Manley Tower on the runway condition and that it was reported as “Wet”, and also that they did
not receive any advisory regarding braking action being less than good and no report of any
significant runway contamination, in fact the evidence from the CVR and ATC transcripts (see
Appendices 2 and 3) indicated that the AA331 flight crew did not request or receive any runway
surface condition report from the time of first contact with Enroute control at 21:47:00 EST until
told “Runway wet” at 22:17:57 EST with landing clearance, less than 5 minutes before impact,
and about 26 minutes after the approach briefing, and they received no braking action report, nor
were they advised that none was available. The only exception to this is the evidence from the
ATC transcripts that at approximately 21:59 EST, just before being transferred from Enroute to
Approach, AA331 contacted Approach, asked if any arriving aircraft had reported turbulence on
approach, and was told by the Approach controller that none of the aircraft landing were reporting
“anything out of the ordinary” (See 1.1.2 and 1.1.5 and Appendices 2 and 3). However, this was
not a runway condition report, and it should have been apparent to the flight crew that the runway
was wet, from the weather reports, the aircraft radar and their actual observations. They had
landed at MKJP before in rain and with a tailwind, so they were complacent, or at least
comfortable with the situation, having landed successfully in similar conditions many times
before.
At 22:17:57 EST (03:17:57 UTC) the Tower controller said “American three three one, cleared to
land runway one two. Be advised runway wet.” Note that no braking action was included in this
report. The first officer’s comment to the captain immediately after this, that is, “Runway’s wet.
You want to go to brakes three, perhaps?”, indicated surprise, as if this was new information
necessitating a change of plan (that is, go from autobrake 2 to autobrake 3), and yet the aircraft
was in rain on the approach, and rain was in the Forecast and t h e latest SPECI and
METARS for MKJP, and rain was displayed on the aircraft radar. As stated, the condition
Wet from ATC did not necessaril y specify Wet/Good, as in the AA RRLL tables, therefore
the flight crew could not determine from this whether or not the runway condition had changed
from the dispatch report of “WET 0.10 IN WATER”.
Thus, it is evident that, in accordance with Bulletin 737-07, the flight crew should have
gathered the relevant information, and assessed whether or not it was necessary to “check runway
length before landing, using actual reported conditions”, then performed the necessary check.
However, as the Bulletin 737-07 stated “the flight crew should use the charts … ” it was not
mandatory for them to do so. As stated by AA Flight Safety Programs Management, the
company expected and assumed that flight crews would follow the instructions, but the
evidence did not indicate that this was made clear to the AA flight crews.
The AA331 flight crew would not get a worse runway condition report from ATC than Wet,
because the ATS MANOPs did not contain any other descriptive term for precipitant
contamination; it was either Dry or Wet. The evidence indicated that the flight crew was not
aware of this. Thus, the runway surface condition could have been worse than Wet.

266
Section 2, Analysis
Note that Bulletin 737-07 stated “The flight crew must use … the most adverse expected
conditions” (Emphasis added).
The SPECI at 02:28 UTC, and the METAR at 03:00 UTC included thunderstorms and moderate
rain, and heavy rain showers, respectively, the controller reported moderate rain to the flight
crew, and the captain reported in his interview that the approach was noisy because of the heavy
rain showers. Thus, in the absence of runway condition report and braking action report, the
f l i g h t crew should have planned for “the most adverse expected (braking action)
condition”, which was not Wet/Good.
In accordance with the investigation’s interpretation of the flight crew’s “advance analysis”
regarding Required Runway Landing Length, they were safe to land on an 8,000 foot runway with
Wet/Good conditions, and a 15 knot tailwind, so landing on a runway of 8,911 feet with a 14 knot
tailwind and unknown braking action (not verified as Wet/Good as per the AA Landing Data
Card), but quite possibly worse than Wet/Good, was not in accordance with AA procedures,
and clearly showed the flight crew’s reduced Situational Awareness (see definition 2.32.1)
regarding the landing conditions.
Thus, the AA331 flight crew did not follow the direction of Bulletin 737-07, nor that of SPC MSG
NBR 9482, to do a landing performance assessment, although the circumstances indicated they
should have considered this.
AC 91-79 recommended that flight crews should assume standing water if rain is actively falling
on the runway.
2.29 Tailwind Landing Hazards and Tailwind Landing Training
Paragraph 1.18.12, “Hazards associated with tailwind landings” demonstrates clearly what these
hazards are. This is in contrast to the lack of awareness demonstrated by AA Training
Management and Training staff (See 1.17.1.2.9) and the resulting lack of this training for AA
flight crews. The investigation was unable to determine the reason for this gap in knowledge and
training in American Airlines. This gap created a great potential threat in the AA operations and it
is likely that this accident may have been avoided if the AA331 flight crew had been adequately
trained in this respect. Furthermore, there was no FAR requirement for tailwind landing training,
nor was there in ICAO Annex 6.

267
Section 2, Analysis
2.30 AA Management And Operational Control
The following factors, which were shown to have possibly contributed to the accident,
indicated that there was some weakness in parts of the AA Management and Operational
Control system:
1. AA dispatch use of a Field Report for Kingston which included an unmeasured

“measurement” of water depth, contrary to company procedures (see 1.10.4.4).
2 AA331 flight crew not following some prescribed and recommended operating
procedures.
4. AA lack of a clear definition of the term “Wet” in runway condition reporting.
5. The lack of understanding by AA flight training staff of the hazards of tail wind
landing.
6. The lack of AA flight crew training on the hazards of tail wind landing.
The investigation considered that factors such as these should have been corrected during training,
Line Checks, internal audits or FAA Inspections.
2.31 FAA Advisory Circular (AC) 91-79
It should be noted that, in the sentence sequence from FAA AC 91-79:
“For example, if there is no clear report of runway condition, but the pilot knows rain
has been in the area, that pilot should assume the runway is at least wet. If there is rain
actively falling on the runway, standing water should be assumed. If there is any
doubt, assume the most conservative condition that requires the longest landing
distance.”
the sentence “If there is rain actively falling on the runway, standing water should be
assumed.” is not qualified by “if there is no clear report of runway condition”.
The AA331 flight crew did not ask for a runway condition report, and were not given one by
ATC until less than 5 minutes before landing, when the tower controller said “Be advised
Runway wet”. The first officer then said to the captain “Runway’s wet. You want to go to
brakes three, perhaps?” as if this was new information, and required a change of plan.
However, AC 91-79 was not mandatory and it is not known whether or not the flight crew on
AA331 was ever provided of a copy of this document, nor was it mandatory that they follow the
caution in this circular. Also, the report of “Be advised runway wet” was received less than 5
minutes before landing, therefore was not used in the flight crew’s initial planning of the landing.
The flight crew landed with no braking action report, and the evidence indicates it was highly
probable that they would have completed the landing without any runway condition report at all.

268
Section 2, Analysis
The investigation considers that the information in AC 91-79 would enhance safety awareness if
included in safety promotion programs aimed at prevention of runway overruns by Part 121
operators.
2.32 Situational Awareness
2.32.1 Situational Awareness – Definition
Situational Awareness in Aviation (Endsley, 2005)
Situational awareness is a term used to describe a person’s awareness of their

surroundings, the meaning of these surroundings, a prediction of what these
surroundings will mean in the future, and then using this information to act.
For the purposes of this investigation, “Situational Awareness” includes information available to
the AA331 flight crew.
2.32.2 Situational Awareness before Departure
From their training the flight crew should have followed the instructions in the AA B737
Operating Manual, which stated “To optimize situational awareness, planning for the approach
and landing should begin before departure.” 124
The evidence indicated that neither of the flight crew was aware of the Page 10-7X standing water
warning, nor of the RNAV (GPS) Rwy 30 approach. If the flight crew had been aware of Page
10-7X, they would have had to consider the possibility of standing water on the runway.
Thus, at this point, before departure, with the information in their possession, even though
ILS Rwy 12 was the most convenient for approach and landing, the flight crew should have been
aware of other landing options, the obvious one being RNAV (GPS) Rwy 30, which offered low
minimums.
However, the flight crew reported that their pre-flight briefing did not include any discussion
about planning the approach and landing, contrary to 1.17.1.1.5. (Note: the investigation
considered that the captain may not have attempted to land with a strong tailwind in heavy
rain on a runway of limited length for which there was published a warning of “pools of
standing water after heavy rain”).
124
See 1.17.1.1.5

269
Section 2, Analysis
Before departure the crew checked the NOTAMS and weather for destination enroute and
alternate and would have been busy with pre-departure preparation.
They were distracted by a baggage removal before start and later during taxi by an MEL item
which also delayed the flight. I t m a y b e t h a t t hey did not check all Jeppesen information
for Kingston b e c a u s e they had both flown there before. The airport briefing page for NMIA
that indicated items of interest and operational significance, Page 10-7X (see Appendix 5),
included a warning of standing water on runways after heavy rain. In their post-accident
interviews, neither pilot mentioned using this page in discussion of the arrival during the flight
preparation or pre-landing stages.
Thus, the flight crew did not make themselves familiar with all the available information before
departure. Beyond seeing that the forecast was above landing limits, they did not give any
consideration to the expected landing conditions at Kingston before departing from Miami, nor
is there any evidence that they had any concerns about the runway conditions or braking action at
Kingston until just before landing.
This shows that the flight crew’s Situational Awareness before departure was incomplete, partly
due to the inaccurate information given to them, and partly because they did not make themselves
aware of the resources available to them, predict more accurately what the possible landing
conditions at Kingston could be (most adverse), and make an appropriate landing plan.
2.32.3 Situational Awareness during Approach and Landing
The evidence indicates that the flight crew probably did not receive the ATIS for Kingston
while still at cruise (see 1.18.3) as they did not report having received the ATIS to the Approach
controller on first contact as they normally would, and they requested the latest weather from the
Enroute controller. Also, the only mention of the ATIS in the flight crew interviews was
the first officer’s statement that he did not recall what the weather was on the ATIS. However,
the flight crew could request whichever runway they deemed suitable. In this case, during the
approach, the Approach controller queried the flight crew twice, and the Tower controller did so
once, on their intention to land on runway 12, and alerted the crew to the unfavourable wind
direction and strength, but the captain still decided to land on runway 12.
The evidence indicated that the flight crew was probably not aware of the ACARS transmitted at
03:00 UTC containing the METAR (22:00 EST) for Kingston, and there was no discussion
regarding this on the CVR. It is highly unlikely that a crew in the later stage of an IFR approach
in poor weather would have the opportunity to read an ACARS message.
The flight crew reported that they performed the approach briefing before commencing descent.
They described doing a “particularly thorough” briefing, and they discussed the preference to do
an ILS Rwy 12 approach rather than attempting the circle-to-land approach for runway 30 with
a low ceiling, which showed they were aware of the tailwind factor. However, there is no
evidence of the flight crew having done the landing performance assessment as per Bulletin 737-
270
Section 2, Analysis
07 and SPC MSG NBR 9482, nor that they looked “thoroughly” at the Jeppesen information
for MKJP, which included the Page 10-7X warning of standing water after heavy rain, and the
RNAV (GPS) Rwy 30 approach at MKJP.
Contrary to AA “Considerations Prior to Descent” (see 1.17.1.1.5), which states “Both flight crew
should review field conditions and special procedures for the arrival airport, including Ops
Advisory pages.”, there was no evidence that the flight crew requested or received any runway
condition or braking action report, nor that they were concerned about runway conditions on
runway 12 at Kingston during the approach, until less than 5 minutes before landing, when the
first officer suggested to the captain changing from autobrake 2 setting to autobrake 3.
The probability that Fair/Medium braking action conditions might exist was indicated when
dispatch sent the flight crew a Special Weather Observation just before descent, reporting 3 miles
in thunderstorms and moderate rain, with wind from 310 degrees at 9 knots; however, there was
no evidence that the flight crew incorporated this new and less favourable information
into their Situational Awareness.
During the approach briefing the flight crew had decided to use flaps 30 and autobrake 2 for
landing. This was contrary to AA procedures (See 1.17.1.1.8) which recommended flap 40 for
tailwind and wet runway landings, and the use of maximum manual braking on wet
runways. This configuration decision demonstrated that the flight crew’s Situational
Awareness was inadequate, considering the landing conditions and the necessity to decelerate
and stop as soon as possible after landing.
At the commencement of descent, the crew’s Situational Awareness was that there was moderate
rain in thunderstorms at Kingston, that there was a 9 knot tailwind, and that the braking action
condition was Wet/Good. They seemed to have omitted to take into account any consideration of
runway braking action being less than Wet/Good, despite all of the indications, and they did not
plan to appropriately configure the aircraft for landing on a wet runway with a tailwind close to
the recently increased tailwind limitation.
The decision-making process of the flight crew indicated that their thinking was narrowed
at this time, as evidenced by the lack of discussion about the worsening conditions, and the lack
of any further discussion about a possible go-around. The conversation between the flight crew
and the controllers indicated that the flight crew was concerned mainly with the tail wind being
15 knots or less, and the controllers appeared more concerned with the tailwind conditions than
was the flight crew.
Throughout the 30 minutes of the CVR recording, the flight crew had no discussion or extra
briefing about the hazards of the tailwind landing (See 1.18.12). This may have been due to the
lack of training by AA regarding tailwind landings, and the resultant lack of awareness by the
flight crew. If the flight crew had been aware of these hazards, they might not have attempted to
land on runway 12 with a 14 knot tailwind in heavy rain, and the accident might have been
avoided.

271
Section 2, Analysis
It appeared to the investigation that, as the flight crew continued the approach, their Situational
Awareness was diminishing as the situation changed, and they were embarking on a difficult and
hazardous landing without any concerns. The plan upon which they were acting was inadequate,
and did not include consideration of the AA recommended techniques for a tailwind landing on a
wet runway.
Heavy rain showers were reported in the 03:00 METAR, 22 minutes before landing, and at the
time of landing, and the c a p t a i n reported they were flying in heavy rain on the approach.
22:14 EST, eight minutes before landing, the flight crew started receiving tailwind reports
of up to 15 knots from ATC. Even within their incomplete state of Situational Awareness, this
should have triggered a serious warning to the flight crew. They should have known that with
the rainfall reported and what they observed through the cockpit windows (described by the
captain as “heavy rain”) that a safe landing with a 15 knot tailwind would be marginal, as the
braking action was probably not Wet/Good. However, they continued the approach without
any discussion or extra briefing, and, at each wind check from ATC of less than 15 knots, they
immediately responded they would land with the tailwind, despite the heavy rain and the
controllers’ querying of the intention to land on runway 12.
Less than five minutes before landing, when the flight crew received the first runway condition
report, which was simply “ … be advised, runway wet … ”, the first officer said to the captain
“ … runway’s wet, you want to go to brakes three perhaps?”, which showed he took the landing
distance into consideration. It also indicated that the crew had not given the runway
condition much consideration. Even though the controller’s report of “wet” was not equivalent
to the Wet/Good braking action of American Airlines, the crew continued on this “advance
analysis” assumption of Wet/Good.
At this point, if their Situational Awareness had been adequate, with the increased tailwind
and the observed heavy rain, they should have known that the safety margin was probably
greatly reduced, and a go-around was advisable.
When the aircraft passed over the threshold of runway 12, then pitched up slightly and went
above the ILS and the PAPI glide paths, and the vertical speed flattened out, as signified by the
Vertical Speed Indicator and the slow Radar Altimeter voice count-down, the captain did not
appear to realize that the landing was going to be long, and it would not be safe to land
when the runway length available was already so limited. However, the captain continued
with the landing, and the first officer made no comment.
As discussed, the lack of touch down zone lighting or reflective runway markings to give the
captain cues to know exactly where the touchdown zone was, might have contributed to causing
the captain to inadvertently pitch up slightly over the threshold. The subsequent float did not
appear to have alerted the flight crew to the fact that the aircraft was going to land long, and that a
go-around should be conducted immediately.

272
Section 2, Analysis
The crew did not describe any of the specific assumptions regarding their “advance analysis”
during the crew interviews, nor was the crew questioned or queried about the specific configuration
requirements. It should be emphasized that the “advance analysis” upon which the crew were
relying to ensure a safe landing did not specify runway condition/braking action, flap setting, use
of maximum brakes, use of reverse thrust or distance of touchdown point from the
threshold.
In fact, as shown in 2.19.3, landing at 1,500 feet with autobrakes 2 (as originally selected)
would require more than the available runway, and with autobrakes 3, the landing distance
available would be marginal. Also, the flight crew had missed the signs of landing long, telling
them that the landing would probably not be successful. Had they realized this, it should have
dictated that a go-around was the only acceptable action at this point; however, the captain
continued with the landing and attempted to stop the aircraft before the end of the runway. The
aircraft then landed long, did not decelerate as the crew expected and ran off the end of the
runway.
2.33 System Safety
As described in Section 1 of this report, the NTSB strongly urged the FAA to make Landing
Distance Assessments (LDA) mandatory before every landing, stating (see 1.17.5.2)
“As another winter season approaches, the urgent need for safety margin becomes more
critical”
It also stated:
“Further, the Safety Board concludes that although landing distance assessments
incorporating a landing distance safety margin are not required by regulation, they are
critical to safe operation of transport-category airplanes on contaminated runways.
Therefore, the Safety Board believes that the FAA should require all 14 CFR Part 121,
135, and 91 subpart K operators to accomplish arrival landing distance assessments before
every landing based on a standardized methodology involving approved performance data,
actual arrival conditions, a means of correlating the airplane’s braking ability with
runway surface conditions using the most conservative interpretation available, and including
a minimum safety margin of 15 percent.”
The FAA responded with SAFO 06012, which recommended LDAs be performed, but only if
landing conditions had changed from those presumed before the flight. AA responded to SAFO
06012 by issuing Bulletin 737-07 to AA flight crews, reflecting SAFO 06012, but, as is evident in
the investigation, did not make it compulsory for AA flight crews to follow this. The AA331 flight
crew, in turn, used an unapproved method of landing distance calculation.

273
Section 2, Analysis
Thus, this system of NTSB-to-FAA-to-Operators-to-Flight Crew, which should have ensured that
an extremely urgent safety issue within the industry be resolved, as evidenced by the NTSB’s
strongly worded appeal to the FAA, failed to have the desired effect, and this failure can be said to
have contributed to this accident.
2.34 Conclusion
Section 2.32 describes how the flight crew started off with an incomplete Situational Awareness,
and how this Situational Awareness became gradually degraded as the flight progressed, and
eventually resulted in this accident.
A degradation of Situational Awareness, as described above, provides a description of

what happened, but does not satisfactorily answer why it happened.
This was a very experienced flight crew flying in a familiar aircraft type, familiar with the
Caribbean and in circumstances which they must have encountered numerous times in their years
of flying, so they were not dealing with conditions with which they were not familiar.
There was no obvious reason for the flight crew not performing the pre-flight and approach
briefings completely, and thus not being aware of the Page 10-7X standing water warning
and the RNAV (GPS) Rwy 30 approach. These omissions did not reflect the expected
actions of an experienced captain (who had also been a check airman), nor of an experienced first
officer, and there was no evidence that either of the flight crew was unwilling to conform to
company directives and training. These omissions cost the flight crew awareness that landing
on runway 12 in heavy rain with a 14 knot tailwind was marginal, and that there was a safer
option to land on runway 30 using the RNAV (GPS) Rwy 30 approach. This contributed to
causing the accident.
The fact that AA had not made mandatory the instructions in Bulletin 737-07, as the NTSB had
recommended, may be related to the accident. If AA had made it mandatory and ensured their
flight crews understood and followed this, the accident might have been avoided. Nevertheless
the flight crew should have been aware of Bulletin 737-07, as it was part of their training, and
it was, in any case, the flight crew’s duty to ensure the aircraft could land safely.
Furthermore, the statement in SPC MSG NBR (SMN) 9482 meant that under the circumstances of
this tailwind landing it was mandatory for the crew to check the landing performance
requirements by direct reference to the AA RRLL card, or AA Bulletin 737-07, which they did not
do.
It is evident that the training received by the flight crew regarding completing a landing
performance assessment, with respect to Bulletin 737-07 and SMN 9482, was inadequate, and did
not ensure that the flight crew fully understood what was expected of them.

274
Section 2, Analysis
The deficiencies in the information provided to the flight crew regarding the runway condition
could be said to have caused the flight crew to consider runway 12 braking action to be Wet/Good
and led them to land on it, but the heavy rain during the approach and landing should have
prompted the flight crew to consider the possibility that the braking action was worse than
Wet/Good, as indeed turned out to be the case. Also, ATC could have assigned runway 30 as the
active runway, and suggested the RNAV (GPS) Rwy 30 to the flight crew.
The first officer was the “pilot monitoring” during the flight, but his CRM was
inadequate, and he should have called for go-around when there were indications that the aircraft
would land long. It is possible that he had been inadequately trained, and also possible that he,
as first officer and pilot monitoring, was reluctant to call go-around to the captain, pilot
flying (See 1.17.1.1.9). If he had called for a go-around when there were indications that the
aircraft was landing long, the accident could have been avoided. 125
The lack of tailwind landing training by AA was certainly significant, as the flight crew seemed
to have no concerns about this. If they had been appropriately trained and better informed,
they probably would not have elected to land in these tailwind conditions on a wet runway, and
the accident could have been avoided.
The circumstances indicated that, even before departure, the flight crew had formed the opinion
that it would be safe to land on the wet runway 12 at Kingston as long as the tailwind did not
exceed 15 knots, and that they maintained this belief and stuck to this plan right until the
aircraft left the runway. This was partly due to not following company guidance, inadequate
planning and awareness, lack of training, poor judgment and because they did not take in new
information as the flight progressed and modify their plan accordingly; in other words, their
Situational Awareness was not adapted to the changing circumstances as the flight progressed.
The flight crew should have realized that the worst case scenario in the heavy rain was
Fair/Medium runway condition, consulted the RRLL tables, and found with an 8 knot tail wind
they needed 8,148 feet to land on the 8,911 foot runway (including touchdown at 1,000 feet,
maximum braking, normal reverse thrust) and with only a 15% safety margin. With this
knowledge the crew would have realized that, when the tail wind increased to 14 knots, a safe
landing was almost certainly not possible. As it was, the crew used the “advance analysis”, landed
the aircraft at 4,100 feet, and did not apply maximum braking until 6,900 feet down the runway.
The long landing itself could be said to be a result of deficient Situational Awareness, in as far as
the lack of reflective markings and touch down zone lighting would have made it difficult for the
captain to know exactly where the aircraft was, in relation to the touch down zone. Nevertheless,
if either of the crew was uncertain as to whether or not the aircraft landed in the touchdown zone,
he should have called go around, as per AA procedures.
125
See 2.9

275
Section 2, Analysis
Before the aircraft flew over the runway threshold the flight crew had plenty of
indications that the runway available would be minimal, even if the aircraft landed in the
prescribed touch down zone. Therefore the investigation questioned why, when t h e r e w e r e
i n d i c a t i o n s t h a t the aircraft was going to land long, the flight crew did not initiate a go-
around.
In answer to this, it is probable that the flight crew, as is typical of persons doing a complex and
repetitive task, had become complacent and narrowed their performance down to a minimal level,
which had served them well for many years; however, in this case, it resulted in an accident.
The system within which they were operating this flight, that is, that of American Airlines, the
FAA and the JCAA ATC, appeared on the surface to be adequate and had multiple safeguards and
redundancies, which should have ensured the safety of the flight, despite there being some
deficiencies within this system, as described in this report. However, the major deficiencies in the
circumstances of this flight, and the ones which probably precipitated the accident, were the
flight crew’s lack of awareness of the dangers of a tailwind landing on a wet runway and
the Approach controller not offering the RNAV (GPS) Rwy 30 approach as an option when the
wind exceeded the limit for designating 12 as the active runway. The AA331 flight crew did
not ask for runway 12, it was assigned to them. If the flight crew had been aware of the dangers
of a tailwind landing on a wet runway, they may have abandoned the landing attempt when
they got close to the runway, and diverted safely to Grand Cayman.
Another significant factor was the crew’s use of the “advance analysis” with an incomplete
recollection of the conditions associated with this, that is, use of maximum manual braking and
landing at 1,000 feet from the threshold.
It is also possible that the flight crew was by that time tired, and in a state of narrowed attention,
focused upon the landing and convinced that a safe landing was possible; thus, awareness of the
constraints of the landing they were attempting may have been no longer part of their thinking,
and they continued the landing attempt even though the landing was long.126
2.35 Findings as to Causes and Contributing Factors
The investigation determined that the most probable cause of this accident was that the aircraft
touched down 4,100 feet beyond the threshold, and could not be stopped on the remaining runway.
The flight crew’s decision to land on a wet runway in a 14 knot tailwind, their reduced situational
awareness and failure to conduct a go-around after the aircraft floated longer than usual contributed
to the accident.
126
See 1.18.12

276
Findings
Intentionally left
blank

277
Findings
SECTION 3
FINDINGS

278
Findings
3.0 Findings
3.1 Findings as to Causes and Contributing Factors
(Definition: “Each Finding identifies an element that has been shown, through the results
of thorough analysis, to have operated in the occurrence or to have almost certainly have
operated in the occurrence. These Findings are related to the unsafe acts, unsafe
conditions or safety deficiencies which are associated with the safety significant events that
played a major role in causing or contributing to the occurrence”).
1) In the dispatch document, the flight crew was not provided with an accurate and current
report on the runway condition at Kingston, nor was it required.
2) The flight crew did not r e v i e w t h e P a g e 1 0 - 7 X o r t h e a p p r o a c h

o p t i o n s , and consequently was not aware of the standing water warning at Kingston on
Page 10-7X, and was not aware of the RNAV (GPS) Rwy 30 approach at Kingston.
3) ATC did not offer AA331 the option of the RNAV (GPS) Rwy 30 approach (the flight
plan showed the aircraft was RNAV capable).
4) The flight crew’s Situational Awareness was incomplete in that they did not realize that
the standing water warning of Page 10-7X, the heavy rain, the weather reports they were
receiving and the lack of runway condition reports or braking action reports indicated
that a Medium/Fair braking action condition was a possibility, and hence was the worst
case scenario.
5) The flight crew decided to land in heavy rain on a wet runway in a tailwind close to the
tailwind landing limit.
6) The flight crew did not perform an adequate landing distance assessment.
7) The flight crew did not use the RNAV (GPS) Rwy 30 approach, and land into wind on
runway 30.
8) The flight crew initially briefed to land with autobrake 2, then changed this to autobrake 3
on final approach, whereas “MAX autobrakes or manual braking” was the recommended
American Airlines procedure for the conditions.
9) The flight crew did not plan for “the most adverse conditions”, as instructed in the
American Airlines B737 Aircraft Operating Manual.
10) The flight crew elected to land with flap 30, rather than the flap 40 recommended for short
field and tailwind wind landing in the AA B737 Operating Manual (See 1.17.1.1.8).

279
Findings
11) The flight crew did not adjust their landing plan to the rapidly changing weather conditions.
12) The flight crew did not select the most suitable runway for landing.
13) The captain did not disengage the autothrottle when he disengaged the autopilot, nor was
this required per AA SOPs, although it was part of Boeing recommended procedures (see
1.17.2.2).
14) The aircraft crossed the runway threshold 20 feet above the ideal height, and landed long.
15) The captain did not follow the company SOPs for landing technique and go-around.
16) The captain did not follow company recommendations for landing configuration, for
landing with a tailwind and on a w e t runway, or for the landing profile to be
flown.
17) The captain pitched the nose of the aircraft up when passing over the threshold,
resulting in the aircraft floating in the flare, then landing long.
18) The first officer did not provide all the necessary and appropriate monitoring and CRM
input during the flight, especially during the final stages of the landing.
19) The aircraft touched down at 4,100 feet from the runway threshold.
20) The flight crew planned to land the aircraft without determining the runway surface
condition and the braking action.
21) The aircraft did not land within the desired touchdown point or within the touchdown zone.
22) There was a 14 knot tailwind component when the aircraft landed.
23) The aircraft touched down at Vref plus 5 knots, thus increasing the landing distance
required.
24) The flight crew did not conduct a go-around when the long landing made this
necessary.
25) There was evidence of heavy rainfall and reduced visibility at NMIA before and during
the landing of AA331.
26) There was reduced friction on the runway, as evidenced by the longitudinal
deceleration rate recorded on the FDR. The presence of melted rubber balls on some of
the aircraft’s main landing gear tires could also be an indication of reduced friction.

280
Findings
27) There was evidence of water on the runway at NMIA, from the measured rainfall, ATC
runway report, weather reports, FDR data and the captain’s statement.
28) The flight crew did not apply maximum manual braking until the aircraft was more than
6,800 feet from the threshold.
29) The application of speed brakes, maximum manual braking and full reverse thrust was
not sufficient to stop the aircraft before the runway end.
30) The flight crew’s situational awareness became degraded as the flight progressed.
31) The flight crew members were possibly fatigued after being on duty for nearly 12
hours, and awake for more than 14 hours.
32) The CRM in the cockpit was not adequate, and the first officer, as “pilot monitoring” did
not call for go-around when the aircraft was landing long.
33) The American Airlines staff at Kingston did not follow the American Airlines, Kingston
Station Manual procedures regarding runway condition reporting to AA Dispatch during
inclement weather.
34) The “Field Report” of “0.10 IN WATER” by AA staff at Kingston was not the result of
any measurement or inspection, but only indicated that there was water on the runway.
35) The AA dispatcher was unaware of the Page 10-7X standing water warning at
MKJP.
36) AA had not made mandatory to B737 flight crews in Bulletin 737-07 the FAA
recommendation in SAFO 06012 that flight crews conduct landing performance
assessments before landing, although AA Flight Safety Programs management expected
that all AAB737 flight crews would conform to Bulletin 737-07.
37) The use of the “advance analysis” in place of conformance to Bulletin 737-07, was
considered to be acceptable by AA Flight Safety Programs management.
38) The “advance analysis” used by the AA331 flight crew was not adequate as it did not
stipulate flap setting, braking technique, planned touchdown point, use of reverse thrust or
runway condition/braking action.
39) The “advance analysis” used by the AA331 flight crew did not meet the
requirements of a landing performance assessment.
40) AA stated “Note: no AA documents or training materials specifically define the

‘advance analysis’ concept that the JCAA cites in the draft report. AA flight crews are not

281
Findings
trained to use, required to use, or discouraged from using this method.”
41) The advice in FAA AC 91-79, to assume standing water when rain was falling on the
runway, was not included in the American Airlines operating procedures, nor was it
required to be included.
42) There was no evidence that AA B737 flight crews had received ground or flight
training related to the techniques and considerations for landing in tailwind conditions, and
the attendant hazards, even when the B737 tailwind landing limit was recently
increased from 10 knots to 15 knots.
43) The flight crew did not follow the requirements of SPC MSG NBR 9482, which
stated “As always, pilots must ensure the reported tailwind component complies with
airplane performance requirements for the runway in use.”
44) ATC runway surface condition reporting did not fully conform to the ICAO
recommendations.
45) The flight crew did not request a runway condition report or a braking action report from
ATC.
46) ATC did not alert the crew that no braking report had been received, as
47) ATC did not inform AA331 that the runway was wet until less than five minutes
before the aircraft landed.
48) ATC did not inform AA331 of the reported “heavy rain”.
49) ATC did not assign runway 30, the into-wind runway, as the active runway, as
50) ATC did not follow the ATS MANOPS in terms of active runway assignment,
placing of Weather Standby, reporting of weather, and giving the arriving traffic a
braking action report.
51) The Enroute and Approach controllers gave the AA331 flight crew estimated weather
reports, and did not state that this was ATC observed weather, not official weather reports.
52) Neither NMIA nor ATC had any specific procedures for conducting runway
condition inspections during inclement weather, and disseminating this information to
landing traffic, contrary to ICAO recommendations.
53) The recommended ICAO terminology for describing water on a runway was not used by
ATC, NMIA or American Airlines.
282
Findings
54) There was a lack of consistency in the terminology used for runway condition
reporting by ICAO, ATC, NMIA and American Airlines.
55) Runways 30 and 12 at NMIA did not have embedded centre-line or touchdown zone
lighting, and the painted runway markings did not contain reflective material, as
recommended by ICAO.
56) NMIA did not perform a runway surface condition inspection before AA331 landed, nor
was this part of their procedures.
57) There was no Runway End Safety Area (RESA) at the end of runway 12.
58) The investigation indicated that there may have been some weaknesses in the FAA
oversight of the AA Boeing 737 operations related to tailwind landing training, and
approval of an increased tailwind landing limit.
59) The investigation indicated that there were some gaps in the management and operational
control of American Airlines, specifically runway condition reporting, and flight crews
following recommended procedures.
60) There was no MOU between NMIA and ATS that bound NMIA to keep ATS currently
informed of runway conditions, in accordance with ICAO Annex 11, Chapter 7, 7.2, and
ICAO Annex 14 2.9.2.

283
Findings
3.2 Findings as to Risk
1) The L1 door slide deployed prematurely, causing the door to be jammed and unusable for
evacuation.
2) The captain’s seat crotch belt bracket broke during the impact sequence.
3) Although most of the overhead bins remained in place, the overhead bins in the first
class section through the forward break, and at the aft break, dislodged.
4) All of the passenger service units throughout the airplane became dislodged.
5) The AA KIN FIELD REPORT of “0.10 IN WATER” was based only on the observation
of water on the apron, and not of water on the runway
6) The AA landing performance tables were predicated on landing at 1,000 feet from the
threshold, whereas AA Standard Operating Procedures spoke to the “desired touchdown
point” as being 800 – 1,500 feet, and the “touchdown zone” as being 1000 – 3000 feet, or
first third of the runway.
7) Both of the cabin crew members seated at L1 reported that they were unable to
retrieve their flashlights because the doors to the emergency equipment compartments
under the jump seats were jammed due to the buckling of the floor.
8) Some parts of the aircraft’s emergency lighting system did not function after the accident.
9) There was a conflict between the recommendations of the NTSB and FAA for flight
crews to perform landing distance assessments according to specific parameters, and the
much less specific “advance analysis” used by the AA331 flight crew.
10) There was no written or recorded record of the ATIS, and none was required
by ICAO, JCARs, or the Air Traffic Services Planning Manual (Doc 9426).
11) JCAA ATS lacked recurrent training and proficiency checks for controllers.
12) There was inconsistency between AC 91-79 and SAFO 06012 related to runway surface
condition reporting.
13) There was no clearly defined policy in the American Airlines B737 Operating
Manual to the effect that the first officer could call for a go-around, with it being
compulsory for the captain to follow through with this.

284

285
Section 4, Safety Actions
SECTION 4
SAFETY ACTIONS

286
4.0 Safety Actions
4.1 Safety Actions Taken
4.1.1 AA- General
As a result of the accident involving AA331, American Airlines reported that the
company has taken the following safety initiatives:
Administrative
 A letter dated February 2, 2010 to all the pilots from the Managing Directors
Of Flight on adherence to Standard Operating Procedures.

FOT Bulletins
 Flight Operational Informational Bulletins distributed to all pilots on
issues of importance.
 737 Approach and Landing Overview- 2010-03
 Revised Stabilized Approach and Go-Around-2010-06
 Landing Data Card-2010-13
Posters
 Posters that were placed in all briefing rooms as a result of a Safe Ops
Committee recommendation.
 Go-Around
 SOP Flight Manual PT I
 Pink Bulletins generated to address important procedural changes.
 Stabilized Approach Requirements-FM-009 4-27-10
 Landing Touchdown Point-FM-017 8-03-10
Human Factors
 Presentation to pilots at all Recurrent Human
Factors classes
737 Operating Manual-VOL I

 Manual changes as a direct result of lessons learned or renewed
emphasis in procedures resulting from AA331.
 Special policy for runways 6,000 feet or less available landing length
 (APP/LNG/GA 15.2)
 Standard Callouts- All Descents, Approaches and Landings (15.5)
 Deviation Callouts (15.7)
 Go-Around Requirements (15.8)

287
 Stabilized Approach Requirement (15.10)

 737 Approach and Landing Summary (37.1/2)
 Go-Around After Touchdown/Rejected Landing (40.4)
Recurrent Training (R-9/R-18)

 Additions and/or renewed emphasis in the 737 briefing and simulator recurrent training
as a direct result of AA331. Similar changes have been made in initial/transition,
requalification and upgrade 737 training programs.
 Landing Performance Check (Slides 25-26)
 Landing Performance Check Problem including contaminated runway and
tailwind corrections (Slides 27-31)
 Stabilized Approach (Slides 35-37)
 RAD Mandatory and Variable Events including Tailwind Landing andRejected Landing
(Slides 44 and 53)
4.1.2 AA- Go Around
Subsequent to the accident, AA has revised their B737 OM, page 15.8, with the
following language:
"Go-Around Requirements”
On final, the Pilot-Flying is responsible for executing a go-around if any of the parameters
listed below are exceeded without Pilot-Flying correction. If the Pilot- Monitoring observes
that the Pilot-Flying is not executing a go-around, he or she is responsible for directing a go-
around by calling – “GO-AROUND”. The directed go-around will be executed unless an
emergency situation overrides this requirement."
4.1.3 AA – Reversion to 10 knot tailwind landing limit
The following message was appended to all AA737 flight plans from 08
February 2012.
SUBJECT- 737 TEMPORARY 10 KNOT TAILWIND LANDING LIMIT
REFERENCE - 737 OM VOL 1, REV 61, LIMITATIONS PAGE 10.2
FLIGHT OPS AND FLIGHT SAFETY WILL BE CONDUCTING A SAFETY RISK

ASSESSMENT ON THE USE OF THE 15 KNOT TAILWIND LIMITATION FOR
LANDING.
288
IN THE INTERIM, THE TAILWIND LIMITATION FOR LANDING IS

NOT TO EXCEED 10 KTS. THIS LIMIT MAY BE FURTHER RESTRICTED BY
ANY LANDING PERFORMANCE CALCULATIONS BASED ON THE CURRENT
RUNWAY CONDITIONS AT THE TIME OF LANDING.
CAPTAIN BOB JOHNSON, MANAGING DIRECTOR OF FLIGHT
4.1.4 Jamaica Civil Aviation Authority
Following this occurrence the JCAA has required controllers to be stricter in assignment of active
runway based on wind considerations and give pilots less choice to land with a tailwind.
4.1.5 The Boeing Company
As a result of the malfunction with the escape slide pack on the L1 door the Boeing Company
reported that they have produced a proposed revision to the Boeing 737 Maintenance Manual to
avoid the installation error from recurring.
4.1.6 Airports Authority of Jamaica
The Airports Authority of Jamaica has included RESA in the runway extension project.
4.2 SAFETY ACTION REQUIRED
4.2.1 Operators of Transport Category Aircraft
Operators of Transport Category Aircraft should be required to conduct landing performance

assessments before every landing, based on a standardized methodology involving approved
performance data, actual arrival conditions, a means of correlating the airplane’s braking ability
with runway surface conditions using the most conservative interpretation available, and
incorporating a minimum safety margin of 15 per cent.

289
4.2.2 Advisory Circular 91-79
Following previous runway excursion and overrun accidents the FAA issued Advisory
Circular AC No. 91-79, “runway Overrun Prevention”, which was sent to all U.S.
airlines. It was, by its nature, advisory, a recommendation for a particular action or conduct, and
was not mandatory by regulation. In Paragraph 11 of Appendix 4 of this AC, it was stated: “If
there is rain actively falling on the runway, standing water should be assumed”. This part of
the Advisory Circular should be made mandatory for operations of Transport Category
Aircraft, such that when active moderate to heavy rain is falling on the runway, the runway
surface condition shall be considered to be in a Wet/Poor or Contaminated state for the purposes
of landing performance assessments, that is, water depth more than 1/8 inch.
4.2.3 AC 91-79 and SAFO 06012
The FAA should ensure that the guidance related to runway surface descriptors in AC 91-79 and
SAFO 06012 is unambiguous and compatible, so that operators and flight crews have ready access
to consistent FAA guidance regarding runway surface condition definitions and their proper use in
FAA acceptable landing distance assessments.
4.2.4 Training for Tailwind Landing
At the time of the accident there was no FAR requirement for tailwind landing training, nor was
there such a requirement in ICAO Annex 6. Training in tailwind landing, and the hazards
involved, should be made a mandatory part of Transport Category Aircraft flight and ground
training programs.
4.2.5 Go-around Callout
Operators of Transport Category Aircraft should include in their Standard Operating Procedures a
requirement that either pilot, whether acting as pilot flying or pilot monitoring, should be required
to call for go-around if he/she sees that the aircraft will not land in the touchdown zone, and that
the other pilot will follow through with the go around procedure without question or hesitation.
4.2.6 Tailwind Landings on Contaminated Runways
The landing o f T r a n s p o r t Category Aircraft in t ai l wi nd conditions on contami nat ed

runways and on runways where heavy rain is actively falling should be firmly discouraged.

290
4.2.7 Tailwind Landings on Wet (not contaminated) Runways
(i) Operators of Transport Category Aircraft should caution their flight crews against conducting
tailwind landings on wet (not contaminated) runways, when an alternative and suitable into-
wind runway is available.
(ii) Operators of Transport Category Aircraft should inform their flight crews of the hazards
related to tailwind landings on wet (not contaminated) runways when the aircraft is at or near its
maximum landing weight.
4.2.8 Deceleration Techniques on Contaminated Runways
Air operators should ensure their flight crews follow the company’s recommended procedures for
landing on wet or contaminated runways.
4.2.9 Landing Distance Assessment
When heavy rain is actively falling on the runway of intended landing, flight crews of
Transport Category Aircraft should assume the runway is contaminated and complete a landing
performance assessment based on contaminated runway surface performance data.
4.2.10 Runway End Safety Area Requirements
Runway End Safety Areas that meet the standards of ICAO Annex 14 for a Code 4D
precision approach runway should be established and/or Engineered Materials Arrestor Systems
(EMAS) of an appropriate length for the airport's design aircraft should be installed at NMIA.
NMIA has a plan to extend the runway 12-30 and RESA is included in this plan. EMAS or
RESA is the recommended solution since the runway abuts the sea at both ends.
4.2.11 Runway Surface Condition Reporting
(i) ICAO guidance for runway condition inspection and reporting procedures should include a
description of the recommended training for airport operations personnel carrying out these
procedures. The training syllabus should cover the criteria for conducting routine and non-routine
inspections and provide guidance on reporting the results of those inspections.
(ii) Guidance in (i) above should include criteria for determining the frequency for performing
special runway surface condition inspections when there is, or has been, heavy rainfall on the
active runway.

291
4.2.12 Runway Surface Condition Inspection and Reporting at MKJP
i) Until ICAO publishes the guidance material, it is recommended that NMIA Airports Limited
should conduct regular inspections of the runway surface during or following inclement weather,
prior to aircraft landings, and should report runway standing water coverage as a percentage of
runway covered to the aerodrome control tower for inclusion in ATC voice advisories and/or
ATIS at MKJP.
ii) An inter-unit agreement between NMIA Limited and ATC should be developed to structure the
responsibilities and procedures each organization will follow for runway surface condition
inspection and reporting at MKJP.
4.2.13 Runway Markings at MKJP
It is recommended that the runway surface guidance markings at MKJP be replaced as soon as
practicable using retro-reflective glass bead paint markings. Runways 12 and 30 both have a
“black hole” approach, beginning and ending at the sea, with very little settlement or lighting on
either side of the runway.
At night, when the runway is wet, pilots cannot see the runway markings, as the light reflects
from the wet surface away from the aircraft. The installation of reflective paint would make night
landings a lot safer.
4.2.14 Distance-To-Go Markers at MKJP
Although it is not an ICAO standard, safety would be enhanced by the installation of reflective
distance-to-go markers for the last 4,000 feet of each runway at MKJP.
4.2.15 Runway Lighting
It is recommended that NMIA International Airports Limited install touch down zone lighting and
centerline lighting on runways 12 and 30 to enhance the safety of aircraft landing in inclement
weather.
4.2.16 Cabin Crew Jump Seat Emergency Equipment Stowage Area
All the cabin crew members at each of the cabin crew stations reported that they were unable
to retrieve their flashlights from under their jump seats as the doors to the emergency equipment
stowage compartments were jammed due to the buckling of the floor. Therefore, it is

292
recommended that these compartments are relocated to a position, and/or the doors to those
compartments are made of a softer cover, to allow for easy access in emergency conditions.
4.2.17 American Airlines L1 Slide
American Airlines should ensure that the proposed revision to the Boeing 737
Maintenance Manual regarding installation of the slide pack is followed in the company’s
maintenance procedures.
4.2.18 American Airlines automation mode matching re: autopilot/autothrottle use
The FAA approved manuals for the American Airlines Boeing 737 should reflect the instruction
of the Boeing 737 NG Flight Crew Training Manual, page 1.34, which states:
Autothrottle use is recommended during takeoff and climb in either automatic or manual
flight. During all other phases of flight, auto throttle use is recommended only when the
autopilot is engaged in CMD.
4.2.19 Emergency Lighting
Boeing should design an emergency cabin lighting system for the B738 aircraft which will
continue to operate when there are breaks in the fuselage after an accident.
4.2.20 AA Runway Condition Reporting
AA should develop a means of verifying the field reports of runway surface conditions reports
sent by its outstations.
4.2.21 AA “Advance Analysis” of runway condition
AA should immediately instruct their flight crews that if they use an “advance analysis”
technique to assess landing performance, they should take all factors into account, including
runway surface condition, recommended aircraft configuration, and use of recommended
deceleration techniques. If the flight crews do not use an “advance analysis”, they should be
instructed to perform, when required, the Landing Performance Assessment prescribed in AA
Bulletin 737-07, prior to landing.

293
4.2.22 American Airlines, use of the word “Wet” to describe runway conditions
AA should immediately ensure that their flight crews understand that a runway described as “wet”
by ATC in overseas locations does not necessarily signify that the runway surface is in a condition
equivalent to AA Landing Data Card “Wet/Good”.
4.2.23 American Airlines, information from FAA Advisory Circular AC 91-79
AA should instruct all their flight crews that, in the absence of a runway condition report, where
there is rain actively falling on a runway, standing water should be assumed, as per FAA AC 91-
79.
4.2.24 Jamaica Air Traffic Services
 The JCAA ATS MANOPS should clearly define the circumstances under which Air
Traffic controllers informs Aerodrome authorities of conditions associated with
inclement weather, so that the Aerodrome authorities conduct runway surface
inspections and provide the results of these inspections to ATS units.
 JCAA Air Traffic Controller ab initio, advanced and recurrent training should
include sensitization to the hazards to Aerodrome and ATS operations which are
amplified by the presence of inclement weather.
 The JCAA should have a consistent system of QA Audits.
 JCAA Air Traffic controllers should undergo awareness training of performance

based navigation procedures, including RNAV.
 JCAA Air Traffic controllers should have recurrent training and proficiency checks.
4.2.25 Boeing 737 overhead bins and PSUs
It is recommended that further investigation of the means of securing the overhead bins and PSUs
in Boeing 737 aircraft should be conducted with a view to modifying these means to avoid injuries
to passengers in similar accidents.
4.2.26 Captain’s seat belt crotch bracket
It is recommended that the requirements for captain’s seat belt crotch bracket be further
investigated, and modifications made to prevent failure of this in similar accidents.
294
4.2.27 Captain’s uncertified seat cushion
AA should take action to prevent the use of uncertified replacement seat bottom cushions, as
described in 1 . 1 2 . 6 . 2 . U s e o f u n c e r t i f i e d s e a t c u s h i o n s i n v a l i d a t e s t h e T S O C 1 2
7 a certification basis of the seat, and compromises the occupant injury performance and
decreases the protection against spinal compression loading to flight crew in the event of an
accident.
4.2.28 Runway Grooving
It is recommended that runway 12/30 at NMIA is grooved. It has been demonstrated that runway
grooving is an effective means for improving tire traction during aircraft ground operations under
adverse weather conditions. A number of airport runways, both military and civil, have been
transversely grooved in an effort to improve all-weather airplane ground performance.
Grooving the runway improves the drainage of some runways, provides skid resistance and
prevents hydroplaning during wet weather. Test results demonstrated that, on similarly wetted
grooved runways, the transverse runway grooves produced substantially greater aircraft braking
friction levels than were shown by the wetted un-grooved surface data. The data also suggest that
the effects of tire tread wear are secondary to the greatly enhanced tire/pavement water drainage
capability available on grooved runways.
4.2.29 Airports Group Recommendations
4.2.29.1 Aerodrome Certification
- Continue to move forward toward the goal of meeting the JCAA and
Annex 14 requirement of full aerodrome certification.
- Consider coordination of a certification training program for both JCAA and
MKJP staff on certification standards, conduct and development of a certification program, and
self-inspection/condition-reporting.
4.2.29.2 Aerodrome Maintenance
 JCAA and MKJP develop specific maintenance standards and publish them,
particularly on the subject of pavement and runway strip maintenance.
 Provide detailed maintenance training to those responsible for self-

inspections and for aerodrome maintenance.

295
4.2.29.3 Aerodrome Emergency Plan
- Coordinate additional detail in the emergency plan, particularly in the water

rescue section of the plan.
4.2.29.4 Pavement
 Coordinate a full runway assessment to determine current longitudinal slope,

current transverse slope (along runway full-length), and the existing grade in first
third of runway 30 (from approximately taxiway D to the 30 threshold).
 Coordinate a hydrologic engineering assessment to determine water flow on the
first third of runway 30 from approximately taxiway D to the 30 threshold.
 Coordinate completion of a full friction assessment of the runway to determine
compliance with established friction maintenance standards and deviation from the
findings of the last friction assessment (2004).
4.2.29.5 Marking/lighting signs
Add additional runway designation signs at every runway hold-position
 Consider revising current PAPI configuration to the ICAO standard(PAPI 4L) by

removing the additional 4 boxes on the right side of 12 and 30. This would reduce
maintenance, minimize the potential confusion to aircrews that may occur due to the
separate systems being out of calibration, and would still confirm to ICAO standards.

296
4.2.29.6 Rescue Firefighting Service
 Develop detailed training lesson plans for required training subjects of initial and recurrent
training.
 Consider acquisition of a simple SCBA bottle refill system for aerodrome use.
 Acquire powered rescue saw(s) for aerodrome use.
 Enhance existing training procedures used for aircraft familiarization…to include
acquisition of current training aids for each commercial service aircraft currently using
MKJP.
4.2.29.7 Declared Distances
Currently, the aerodrome operator complies with the requirement of Annex 14.2.8, for
declared distance computations. The figures are published in the Jamaica AIP. Currently however,
the only declared distance figures for MKJP 12/30 that differ from the distance representing 12
threshold to 30 threshold (2716M) distance, are the TODA figures for both 12 and 30
(currently published as 4074M). This represents acknowledgement of existing clearways on
both runway ends. We recommend consideration be given to review of all declared distance figures
to accommodate the fact that there is no existing runway strip or runway end safety area at either
runway end.
4.2.29.8 Runway strip and runway end safety area
Remove existing debris and vegetation from the existing drains in the runway
strip between taxiway D to E. Verify that these drains are operational and that they
are tied into the drain system and box-culvert under the aerodrome recording pron.
 Conduct periodic surveys of the runway strip and RESA (as provided) to
determine safety issues and correct found deficiencies on a more frequent basis.
 Remove all current deviations (objects not required) from runway strip.
 Feather existing localizer base to grade.
4.2.29.9 Maintenance of PAPI
It is recommended that MKJP institute a specific maintenance/calibration periodicity and a

periodic maintenance schedule to service the PAPIs.

297
4.2.30 CVR
It is recommended that all aircraft manufactured after 2002 should be equipped with CVRs
capable of recording 120 minutes. This would enhance the CVR’s usefulness in
accident/incident investigation.
4.2.31 FAA Advisory Material ref: Wet Runways
The statements in AC91-79, Appendix 4, 10 b. (2), “If a runway is contaminated or not dry, that
runway is considered wet”, and in SAFO 06012, item 4. Definitions, “i. A wet runway is one that
is neither dry nor contaminated”, both being FAA Advisory material, are inconsistent. It is
recommended that they are amended to be consistent.
The need to review the guidance and to make it consistent is to eliminate the opportunity for
operator and/or flight crewmember confusion or non-conservative interpretation that could
negatively affect the arrival landing distance assessment advocated by the FAA, the NTSB and
other stakeholders.
4.2.32 Cancelling of NOTAMS
This should be like normal NOTAMS, not just result in a cancelled NOTAM disappearing, as this
could be hazardous to crew not having the necessary information.
4.2.33 Recording of ATIS
The ATIS should be recorded, as the record might be necessary for occurrence investigation.
4.2.34 ATS Procedures
Given that the investigation revealed that the JCAA ATS MANOPs was partially, but not fully,
aligned to the provisions in Doc 4444 in some cases, it is recommended that regulations be
promulgated to require the JCAA, as the ATS provider, to align procedures with ICAO provisions
and that the ATS provider should comply with those requirements.
4.2.35 ATS and NMIA MANOPS

It is recommended that the JCAA ATS MANOPS and the NMIA MANOPS should be amended to
include the ICAO definitions for reporting water on the runway.

Aa Jamaica PDF

Uploaded by

Copyright:

Available Formats

Aa Jamaica PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Aa Jamaica PDF

Uploaded by

Copyright:

Available Formats

AVIATION ACCIDENT INVESTIGATION REPORT

RUNWAY OVERRUN ON LANDING

UNITED STATES REGISTRATION N977AN

NORMAN MANLEY INTERNATIONAL AIRPORT

REPORT NUMBER JA-2009-09

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

1.6.8 Ground Spoilers ...................................................................................................................................40

1.7.12 Weather Conditions on the Ground at Kingston ................................................................................54

1.11.1 Cockpit Voice Recorder (CVR) .........................................................................................................75

1.17.1.1.9 AA Information on Baulked Landings (Go Around) .................................................................118

1.17.4.5 SAFO 10005 .................................................................................................................................142

2.4.2 Runway Slope ....................................................................................................................................236

2.34 Conclusion .........................................................................................................................................273

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Photo 1 Aerial Photograph of Accident Site………………………………………………....59

Photo 2 Aircraft Wreckage…………………………………………………………………...82

Photo 3 Forward Fuselage …………………………………………………………………... 83

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Figure 1 B737-823 Aircraft Cabin and Exits. …………………………………………................42

APPENDICES TO SECTION 1, FACTUAL INFORMATION

Appendix 1 Flight Data Recorder Plot by the Boeing Company................................................. .155

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

GAIA Grantley Adams International Airport (Barbados)

RRLL Required Runway Landing Length

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Intentionally Left Blank

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

1.0 Factual Information

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

1.1.3 Descent and Approach

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

AII Approach Mode

The aircraft was landed with flaps at 30 degrees.

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

The following information is from the statements of the flight crew:

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

1.1.5 Information given to AA331 flight crew during approach

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

1.2 Injuries To Persons

Crew Passengers Others Total

Minor/None 6 134 - 140

Total 6 148 - 154

Table 1: Injuries to Persons

1.3 Damage To Aircraft

1.4 Other Damage

Jamaica Civil Aviation Authority – American Airlines 331 Accident Report

1.5 Personnel Information

1.5.1 ATC Operational Personnel

1.5.2 Flight Crew – General

Captain First Officer

Table 2: AA Personnel Information