Proceedings of the Eastern Asia Society for Transportation Studies, Vol.

9, 2013

Black Spot Cluster Analysis of Motorcycle Accidents

Mark Richmund DE LEONa, Nelson DOROYb, Hussein LIDASANc, Jun CASTROd
a,b,c,d

School of Urban and Regional Planning, University of the Philippines,

Diliman, Quezon City, Philippines
a
E-mail: [email protected]
b
E-mail: [email protected]
c
E-mail: [email protected]
d
E-mail: [email protected]

Abstract: Road accident is now considered as a global problem. It is a cause of concern in the
Philippines since road accidents are at present the major causes of fatalities, injuries and
property damages. In Metro Manila, majority of the road accidents are incidents involving
motorcycles. It is in this context that this study seeks to promote motorcycle safety awareness
through the spatial analysis of motorcycle accident locations in Quezon City. The study will
show that the clustering of accidents can be inferred from a Kernel Density Map, generated
through the use of GIS, after manually plotting its location based on existing traffic accident
records. It was found out that the higher density of accidents involving motorcycles in Quezon
City are along major corridors such as Commonwealth Avenue, Elliptical Road, Quirino
Highway, Mindanao Avenue and EDSA.
Keywords: motorcycle, tricycle, accidents, blackspot analysis, GIS, Kernel Density Map

1.

INTRODUCTION

Road accidents are a problem worldwide, wherein in 2004 around 1.2 million people were
killed (2.2% of all deaths) and 50 million more were injured in motor vehicle accidents. This
translates to 2 lost lives per minute. Developing countries are particularly at a disadvantage as
70% of the fatalities occurring in these states. In 1990, road accident is the 9th leading cause
of death. By 2020, it will be the 3rd leading cause of death. Road accident is also the leading
cause of injury, with road accident injuries higher than occupational injuries.
The Philippines, being a developing country, unfortunately is an example of rising
implications of road accidents. Figure 1 shows that the increase in registration of vehicles in
the Philippines follows the trend of its growth in Gross Domestic Product (GDP) and
population. This simply means that more Filipino people can now purchase vehicles due to
rising disposable incomes. However, the lack of comprehensive road safety program likely
will result to the continued increase in road accidents.

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

Figure 1. Comparison of Growth Rates (Vehicle Registration, Population and GDP)

Amongst vehicles that are involved in road accidents, motorcycles are the most vulnerable as
both the driver and the passenger are highly exposed to the environment. Contributing to the
vulnerability is the tremendous increase in motorcycle registration (which includes tricycles)
as illustrated in Figure 2. It is noteworthy that the Land Transportation Office do not classify
motorcycles and tricycles separately, because the registration of tricycles was delegated to
local government units.
Motorcycles, now outnumber other vehicles in a given year starting 2005 (except in absolute
total numbers). This increase can be attributed to the fact that motorcycles have become an
important mechanism to promote social mobility because of price affordability. Lower income
people now have improved mobility and yet it increased their vulnerability to injuries.

Source: Land Transportation Office, 2011

Figure 2. Motor Vehicle Registration by Vehicle Classification

The relative mobility of motorcycles has become the danger for the passengers and riders.
Being nimble and having the ability to turn faster, motorcycles can weave in and out of traffic.
This makes the likelihood of collision against other vehicles higher. Based on hospital account,
motorized two-wheeler users tend to sustain multiple injuries, including to the head, chest and

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

legs. The majority of the fatal injuries are to the head. Lower-leg injuries either from direct
contact with the impacting vehicle or as a result of being crushed contribute substantially to
more fatal injuries. Motorcycles now have the highest rate of accidents in Metro Manila as
demonstrated in Figure 3.

Figure 3. Percentage Share of Fatal Accidents by Vehicle Involvement

Given the scarcity of government resources particularly funding for road infrastructure, it is
therefore rational that more research in road safety is needed particularly for motorcycle
accidents. However, most researches in the country have focused on statistical analysis of
accident rates. Moreover, previous road accident mapping through GIS has general
assessment across vehicle types or classifications. A study on motorcycle accidents is thus an
additional dimension to the limited literature of road safety in the country.
The main objective of this study is to characterize the spatial profile of motorcycle
accidents through the use of GIS. The expected outputs of this research is the (i) Identification
of motorcycle accident black spot clusters; and (ii) Spatial profiles of motorcycle accidents.
2.

SCOPE AND STUDY AREA

The scope of this study involves the spatial analysis of motorcycle and tricycle accident
location, being the highest incidence of road accidents by vehicle involvement.
Quezon City is the largest city in Metro Manila in terms of land area and population.
And more importantly, Quezon City major arterial corridors pass through its jurisdiction. The
study area is Quezon City since it is has the highest incidence of motorcycle and tricycle
accidents across Metro Manila. The total road accidents in Quezon City is 25,519, the highest
amongst local government units in Metro Manila. And out of these road accidents, there were
4,485 motorcycle and tricycle accidents in the same city as shown in Table 1.
Table 1. Road Accidents in Quezon City
Category
All Vehicle Type Motorcycles/Tricycles
Fatal
144
58 (40%)
Non-Fatal
4,797
2,422 (50%)
Property Damage Only
20,578
2,005 (10%)
TOTAL
25,519
4,485

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

3.

REVIEW OF RELATED LITERATURE

Many highway agencies have been using Geographic Information System (GIS) for analyzing
accident data. Identification of problematic locations is one of the most important aspects in
accident studies. Although there is no accepted international standard definition of an accident
blackspot, for the purpose of this study, a blackspot or hazardous road section will be defined
as a road section wherein the expected number of accidents is higher than other similar
locations (Sorensen and Elvik 2007). The GIS based application combines the information
collection capabilities with the visualization. In 2004, The University Putra Malaysia
developed a GIS and Road Accident View System and a set of applications developed for
managing accident database entries. The developed system was designed based on the
Universiti Putra Malaysia (UPM) community area and may be adapted very easily to any
other places.
The system was developed using Microsoft Visual Basic 6.0 in Windows XP platform.
The database was designed in terms of textual format. The Accident Report from UPM
security unit was used as the source for information needed in the database development
wherein the location of the accident was recorded on a map. And by browsing on it, the user
can perform queries on a particular condition to get the number of accidents.
The research paper of UPM has shown that traffic accidents data can be collected,
analyzed and presented in a systematic way through a GIS program. With the proper setup of
Accident GIS System, the user or decision maker can retrieve, analysis, and display database
on its spatial characteristics. The GIS Software can provide the core functions through
entering, searching and retrieving information from database; main supportive functions
through Node Analysis and Distribution Plot; and Minor functions through Searching Engine
and Accident Ranking. Thus the GIS based application was concluded to improve the
accuracy and timeliness in priorities accident location.
In another research paper, Kernel Estimation cluster analysis technique was used to
automatically identify road traffic accident black spots and black areas in Christchurch,
New Zealand. GIS and Python scripting was used to implement the technique, using spatial
data of recorded accident locations. Kernel Estimation was able to quickly identify the
statistically significant accident clusters.
In Flahaut et al (2002) focused on spatial data analysis problem: defining the location
and the length of black zones. Two methods are compared: the use of local spatial
autocorrelation indices (a decomposition of the global Moran index) and kernel estimation.
Both methods differentiate local dangerousness and generate a smoothing of the empirical
process. Although each method are different in its conceptual approaches, both provided
quite similar results under a specific parameters.
Furthermore, Lu (2000) compared kernel density to other spatial analysis techniques
and found out that Kernel density is more reliable and desirable for hot spot analysis. Lu has
the following observations: First, kernel density uses more information about point
distribution than virtually all distance view cluster algorithms. A continuous density surface is
generated over all locations to let data pattern show itself. Users can visually inspect the
variability of density over the whole surface and identify hot spot depending on his/her
observation point of view. Second, since density is a measurement of magnitude, in addition
to showing spatial clustering, hot spots are comparable over space as well as between studies.
Algorithms based on spatial separation of observations are shy to support the quantitative
comparison of hot spots. Third, kernel density technique is more arbitrariness free and
provides relatively stable analysis results to user.
It is for this reason that Kernel density is adopted in this study.

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

4. METHODOLOGICAL FRAMEWORK
Given the experience of previous countries in plotting accident locations, the study team
developed its own methodology to present GIS Blackspot mapping in Metro Manila,
specifically in Quezon City. Motorcycle and tricycle accidents were extracted from the Metro
Manila Accident Reporting and Analysis System (MMARAS) database of road accident
reports in Metro Manila, which is being maintained by the MMDA. The MMARAS is
operated by the Road Safety Unit (RSU) of the MMDA-Traffic Operations Center (TOC),
with the cooperation and assistance of the Traffic Enforcement Group under National Capital
Regional Police Office (TEG-NCRPO) Philippine National Police (PNP). The accident data
are based from individual report forms for each accident, gathered by Data Researchers Group
of the RSU from different stations and Districts Offices of the Traffic Enforcement Group.
However, it is important to note that there is 60 to 70% under-reporting of road
accidents in the Philippines (Vibal, 2003). Road accidents reported in hospitals are examined
to be higher compared to accidents reported in police stations. Whether there is inherent
distrust to the police community or general avoidance of perceived further administration
costs, a good reporting system for road accidents is essential to fully understand the
implications of road safety.
Upon collection of the data from MMDA, the individual locations of these motorcycle
and tricycle accidents are then plotted or place-marked in Google Earth to have
geo-referenced accident locations. GIS technicians were hired to facilitate the expeditious
plotting of numerous accident locations. The location files from Google Earth were then
imported into ArcGIS 9.3, for which these were converted into shape files (point feature).
Other attributes of the motorcycle and tricycle accidents from the MMARAS database were
spatially-joined with the location shape files.
As a further analysis, accident valuation was also done to determine the total costs
arising property damages, injuries and fatalities. Each accident was given corresponding
socio-economic cost values using the results of the previous study of one of this papers
authors made. However, in the previous study, the author had four (4) classifications of
accidents, namely fatal, serious injury, minor injury, and property damage only, which is
consistent with internationally used accident classification systems. And yet in MMDAs
MMARAS database, there are only 3 classifications; fatal, non-fatal injury, and property
damage only accidents. As such, for injury accidents, the cost for minor injury accidents in
the previous study was adopted for non-fatal injuries of the MMDA database.
Attributing costs to each accident will enable planners and decision makers to have a
grab of the social and economic magnitude of accidents, where each live and property damage
are given monetary values to consider loss economic opportunities, medical costs, etc.
The accident locations are then mapped out, for which Kernel Density Mapping is
undertaken. Kernel Density calculates a magnitude per unit area from point or polyline
features using a kernel function to fit a smoothly tapered surface to each point or polyline.
Other thematic maps are derived including spatial grid indexing of attributes. Figure 4 shows
this methodological framework adopted for this study.

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

Process

Tasks/Commands

Outputs

Accident database
mining

Filter, Pivot Table, Cross-Tab

List of Motorcycle Accidents

Plot accidents in
Google Earth

Place Mark Insertion

Accident Locations in KMZ

Import KMZ Files

into ArcGIS

Quick Import

Accident Location Shape file

Join shape files and

accident database

Join Attributes

Accident Attributes Shape files

Accident valuation

Field Calculator

Accident Costs

Spatial analysis

Kernel Density, Grid Spatial Index

Accident Hotspots, Thematic Maps

Conclusion

Road Safety Analysis, Site Verification

Measures Prioritization

Figure 4. Methodological Framework for the Study

5. SUMMARY OF FINDINGS
The accident data for this study was extracted from MMARAS which is an Accident GIS
Database System operated and maintained by the MMDA. However, there are inherent
downsides to the MMDA-MMARAS, which are mostly budget related. There is no GIS
software. Data collection is manually done by researchers collecting filled-up accident
investigation forms from PNP-TEUs. Sometimes, accident investigation forms are not used
and instead, the researcher would need to extract information from narrative police blotter
reports from logbooks. In the process, important data may have been overlooked either by the
police or by the researcher. Furthermore, accident data from accident reports are manually
encoded through MS Excel. Based on interviews from MMDA personnel and PNP-TEU
police investigators, there is a general inadequacy of training for accident investigation and
analysis. In spite of these constraints, a total of 3,697 data points were utilized, which
comprised of motorcycle and tricycle accidents in Quezon City for the year 2010. These data
points were the usable records that have been imported from the accident plots in Google
Earth. Some of the accident data have been identified as invalid because of inconsistent
location (i.e. EDSA corner Commonwealth Avenue) or locations which cannot be found using
Google Earth, Wikimapia, and or OpenStreetMap (i.e. in front of a sari-sari store).
After plotting in Google Earth, the KMl file was imported into ArcGIS for which these
were converted into shape files (point feature). After joining corresponding attributes, the
resulting GIS database can be sub-classified according to subject attributes or themes. Figures
5 to 8 show the accident points of motorcycles and tricycles in Quezon City, for which the
accident plots were translated into maps generated through ArcGIS.

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

Figure 5. Plot of Motorcycle and Tricycle

Accidents in Quezon City

Figure 6. Plot of Fatal Accidents in Quezon

City

Figure 7. Plot of Non-fatal Injury Accidents in

Quezon City

Figure 8. Plot of Property Damage Accidents

in Quezon City

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

The accident plot maps do not show the spatial significance or clustering of accidents in
certain road sections. It seems that every road is plotted with accident sites. By using the
Kernel Density function of ArcGIS, the density of features is a calculated in a given area
around the accident sites. Spatial analysis of motorcycle and tricycle accidents through the
Kernel Density Method would show that the spatial density of accidents involving
motorcycles and tricycles are heaviest along Commonwealth Avenue, Elliptical Road, EDSA,
Quirino Highway and Mindanao Avenue (see Figure 9). Meanwhile, the spatial density of
motorcycle and tricycle accidents is lightest in local roads within Quezon City (i.e. city and
barangay roads).
Figures 11 and 12 show the densities or clustering of the accident counts in terms of
fatal accidents and night time accidents. The Fatal Accident cluster map shows that the most
fatal motorcycle accidents occur in Quirino Highway corner Mindanao Avenue and Tandang
Sora corner Mindanao Avenue. The stretch of Commonwealth from QC Circle to North
Fairview is also littered with fatal accident sites specifically near Philcoa and UP Technohub.
And when compared with night time fatal accidents, this map shows little difference as most
fatal motorcycle accidents happened at night at a rate of 32 out of total 47 fatal accidents.
Further analysis of the fatal accident statistics shows that there are 32 drivers killed, 4 drivers
injured; 8 passengers killed, 8 passengers injured; 10 pedestrians killed, 10 pedestrians
injured; and Average age of fatalities is 31 years old. Based on these statistics, we can deduct
that of the 47 fatal accidents, 32 drivers were killed. This shows that for motorcycle accidents,
the most vulnerable victim will always be the driver. The average age of the fatalities is
31years old, which can be considered as productive members of the society. These early
deaths can also be treated as an economic loss if we take into account the economic
productivity of the fatality if he could have lived much longer.

Figure 9.

Kernel Density Map of Motorcycle and Tricycle Accidents in Quezon City

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

Figure 10. Fatal Motorcycle and Tricycle

Figure 11. Nightime Fatal Motorcycle and
Accidents Density Map
Tricycle Accidents Density Map
In this study, clustering can be viewed as densities or levels of incidences (low or hig). For the
high occurrence of injury accidents, there seem to have no single prevailing accident cluster
site for this type of accident. The entire stretch of Commonwealth and Quirino Highway and
portions of EDSA are considered as hotspots for injury accidents (see Figure 12). Meanwhile,
as shown in Figure 13, Property Damage Only Accidents have high incidence rates at Quirino
Highway corner Mindanao Avenue; Quirino Highway coner Tandang Sora; Quirino Highway
corner Baesa Road; Quirino Highway (Bayan and Regalado, SM Fairview); Stretch of EDSA
from North Avenue to Roosevelt; EDSA corners Kamias and Aurora Boulevard; Mindanao
Avenue coner Tandang Sora; Quezon Circle; Katipunan Avenue; and Commonwealth
(Batasan and North Fairview).

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

Figure 12. Motorcycle and Tricycle Injury

Accident Density Map

Figure 13. Motorcycle and Tricycle

Property Damage Density Map

It can be further examined in the points from the GIS database that these road sections where
PDO accidents are dominant are also installed with u-turn slots. Also, there were at least
3,225 vehicles involved in motorcycle accidents. These comprise of 1382 motorcycles, 359
Tricycles, 990 cars among others. Using the results of the study conducted by de Leon (2005),
all of these property damage accidents in QC already cost PhP 69 Million pesos.
Figures 14 and 15 show the clustering or density of fatal and injury pedestrian accidents.
Figure 14 shows the Kernel Density map of fatal accidents occurring mostly in the QC Circle
and Commonwealth Avenue. On the other hand, Figure 15 highlights the large occurrence of
pedestrian injury casualties in Quirino Highway (Novaliches Bayan); Quirino Highway
(Baesa Road); Quirino Highway (Tandang Sora); Batasan Complex; Commonwealth (Luzon
Avenue); and SM Fairview. Commercial establishments at the northern part of QC (SM
Fairview) also attract pedestrian and commuter activity thereby exposing more people to
motorcycle accidents.

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

Figure 14. Motorcycle and Tricycle resulting

to Fatal Pedestrian Accident Density Map

Figure 15. Motorcycle and Tricycle in conflict

with Pedestrian Injury Density Map

Alarmingly, based on the statistics of MMDA, pedestrian casualties have an average age of 51
years old. These show a disregard of motorcycle users to the pedestrians especially the elderly.
It can be surmised that the driving behaviour of motorcycle riders weaving in and out of
traffic and sometimes invading the sidewalks is the cause of these accidents.
Although based on the database, daytime motorcycle accidents (2,466) outnumber night
accidents (1,231), it can be said that most of these night accidents are fatal in nature and are
mostly resulting to injury accidents. Figure 17 shows the cluster or density of accidents during
night time. It shows accidents are numerous at night at: Quirino Highway corner Mindanao
Avenue; Quirino Highway coner Tandang Sora; Quirino Highway corner Baesa Road;
Quirino Highway (Bayan and Regalado, SM Fairview); Stretch of EDSA from North Avenue
to Roosevelt; EDSA corners Kamias and Aurora Boulevard; Mindanao Avenue coner Tandang
Sora; Quezon Circle; and Commonwealth (Batasan and North Fairview).
Assigning costs for each accident enables planners and decision-makers to grasp the
magnitude of the accident in a given site in terms of pesos and centavos. Although this may
seem inhumane, where we assign values to the lives and limbs lost, albeit this is the most
straight forward way of evaluating the cost of accidents to the society. This is also useful in
determining benefit savings from lives saved because of road safety improvements which are
otherwise difficult to justify and difficult to fund with road safety not within the list of priority
programs of the government.

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

Figure 16. Motorcycle and Tricycle Night

Accidents Density Map

Figure 17. Total Cost of Accidents within a

500 meter Grid

Given this accident cost mapping as shown in Figure 17, it can be seen that motorcycle
accidents in QC are most costly in: QC Circle (Philcoa and East Avenue side); Balintawak
Interchange; and Quirino Highway corner Mindanao Avenue. The cost of accidents in the
above-mentioned sites amount to PhP 5.5 Million to PhP 15 Million for the year 2010. Given
this historical cost, it is only just to allot funding for road safety improvements in these areas
to avoid future loss of lives and limbs.
In summary, spatial analysis of motorcycle accidents through the Kernel Density
Method would show that the spatial density of accidents involving motorcycles and tricycles
are heaviest along Commonwealth Avenue, Elliptical Road, EDSA, Quirino Highway and
Mindanao Avenue (see succeeding figures) Meanwhile, the spatial density of motorcycle and
tricycle accidents is lightest in local roads within Quezon City (i.e. city and barangay roads).
A kernel density map of accident costs would also illustrate that costs are highest in
major intersections of the arterial corridors previously identified. The spatial densities of other
accident attributes (e.g. accidents by fatalities, injuries, property damage; day-time and
night-time accidents; accidents by victims, pedestrians) follow a trend similar to the spatial
density of accident counts.
6. CONCLUSIONS
Based from the Black Spot Cluster Analysis, the arterial corridors in Quezon City that
are critical for motorcycles are Commonwealth Avenue, Elliptical Road, EDSA, Quirino
Highway and Mindanao Avenue. The Quezon City Government, MMDA and the Department
of Public Works and Highways (DPWH) might want to prioritize these roads for a general
program of road safety measures. There is another enormous set of road accident data points
outside of Quezon City that can be plotted and analyzed through GIS. The pattern of accident
clustering, however, can still be the same. And these corridors are the abovementioned arterial

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

roads, which traditionally have higher traffic volumes and pass-through traffic.
This study is a good case for GIS as an important tool for the spatial characterization of road
accidents. Despite the need to improvement robustness, MMDA database is a good source of
information from which road accidents in Metro Manila can be spatially profiled through GIS.
GIS is a good tool in this road safety audit activity as GIS can provide a rapid appraisal of
accidents in a site given the adequacy of data.
Furthermore, majority of the cost of accidents are resulting from property damages as
these occur most often. The wide occurrence of underreporting especially for property
damage accidents should be taken seriously. Accurate reporting could have given the
authorities and policymakers factual values on the damages involved thus they have tools to
evaluate existing policies and programs on road safety. Without accurate data, more
under-reporting and inaccuracy of data for accident will be expected, thus giving the false
impressions of roadworthiness and road user safety.
7. RECOMMENDATIONS
Enhancement of MMDA MMARAS. Assessing the data system, the collection of data is
hampered by internal administration problems, which are more related to budget constraints.
As such, the use of an extensive database system has been downgraded continuously and the
personnel of the RSU are now only utilizing Microsoft Excel spread sheets to compile road
accident data. This was in sharp contrast to the original intention of coming up with a
modern and updated database system that can be used in not only for coming up with
measures to address road safety but also for research and analyses.
Further to this, the intention of encoding these data into a GIS database system has been
deferred. Moreover, it is perceived that the process of data collection from the individual
TEGs of the PNP scattered in the Metropolis is hampered by inconsistencies and omission of
important data. This is due to the fact that the researchers of the MMDA-RSU will have to
retrieve accident data logged by traffic investigators of the PNP from their logbook and input
them to MMDAs database. Likewise, the accident logs of the PNP are perceived to have
some degrees of unreliability because of its narrative nature, which relies heavily on the
aptitude of the police officer to write a comprehensive accident report. And with the
absence of a template, necessary data to be collected are sometimes overlooked.
Given the above shortcomings of the database system of MMARAS, it is imperative
that improvement and updating of the system and its collection and management process be
proposed. There is a need to further develop the database system so as to ensure a reliable
and usable road accident database for Metro Manila will be available.
Implementation and Prioritization of Road Safety Measures. The density maps will
show the spatial clustering of black spots. This means that these density maps show the
clustering of high incidence of motorcycle accidents in Quezon City. With resources always
an issue, government agencies can prioritize the implementation of road safety measures in
these black spots or clustering of high incidence of subject accidents. There is a wide
spectrum of road safety measures from motorcycle lanes to road safety signs, which are
options that government agencies can choose from. However, these measures should be
subject for further studies and evaluated for its applicability in the subject road sections.
Use of Mobile GPS Device. Accident reporting can be enhanced by equipping the
traffic accident investigators with Mobile GPS devices that can plot the location of road
accidents and directly record accident investigation data via a mobile device enabled with
accident investigation software. This will ensure faster accident investigation process and will
facilitate data processing and retrieval. Having a simple accident investigation software may

Proceedings of the Eastern Asia Society for Transportation Studies, Vol.9, 2013

also reduce perceived discrepancies derived from encoding of accident investigation forms to
the MMARAS database.
More quantitative analysis on Accidents/ Accident Factors. As a further step,
research is needed in the statistical risk, spatial, and land use relationship of road accidents.
Modelling of the different contributory factors (or attributes) of road accidents can also be a
good material for future study. More detailed spatial analysis, can help in coming up with
physical mitigating measures and schemes so as to minimize road accidents, notably
motorcycle-related ones. Furthermore, spatial analysis may be able to provide any link
between land use and accidents and as such, appropriate and responsive land use-related
measures can be recommended.
More quantitative data about accident analysis should be pursued for future researches
like severity indices and accident exposure rates which would require analysis of MMDA
against LTO data i.e. motorcycle-km, vehicle-km traveled, etc.
8.

ACKNOWLEDGEMENTS

The authors would like to acknowledge Asian Development Bank (Manila) for supporting this
research on motorcycle accidents. Specifically, the authors would like to thank. Engr. Nilo F.
Farrofo of ADB for his inputs to the study.
9.

REFERENCES
Deepthi, J.K. and Ganeshkumar, B. (2010) Identification of Accident Hot Spots: A GIS
Based Implementation for Kannur District, Kerala, India. International Journal of
Geomatics and Geosciences, Volume 1, No. 1
de Leon, M.R.M and Cal, P.C. (2005) Estimation Of Socio-Economic Cost Of Road
Accidents In Metro Manila . Journal of the Eastern Asia Society for Transport Studies
(EASTS). Vol. 6, pp. 3183 3198
Flahaut, B., Mouchart, M., San Martin, E. and I. Thomas (2002). The local spatial
autocorrelation and the kernel method for identifying black zones. A comparative
approach. Accident Analysis and Prevention.
Elvik, R., and Sorensen, M. (2009) Handbook of Road Safety Measures, 2nd
edition.Emerald Group Publishing Limited, United Kingdom.
Liang, L.Y., Masoem, D.M., Hua, L.T. (2005) Traffic Accident Application using
Geographic Information System, Journal of the Eastern Asia Society for Transport
Studies (EASTS).
Lu, Y. (2000). Spatial Cluster Analysis of Point Data: Location Quotients Versus Kernel
Density. University Consortium of Geographic Information Science (UCGIS)
Summer Assembly Graduate Papers. Portland, Oregon.
Sabel, C.E., Kigham, S., et.al. (2005) Road Traffic Accident Simulation Modelling - A
Kernel Estimation Approach. 17th Annual Colloquium of the Spatial Information
Research Centre, New Zealand
Wedagama, D.M.P., Bird, R.N., Dissanayake, D. (2008) The Influence of Urban Land
Use on Pedestrians Casualties - Case Study Area: Newcastle upon Tyne , IATSS
Research Vol 32(1)
Vibal, T.H. (2003) Traffic Accident Analysis Through Hospital Records, Thesis
(M.Sc.C.Engr.) - University of the Philippines, Diliman, Quezon City.