www.groupexcelindia.

com
EXCEL INDIA PUBLISHERS
NEW DELHI
No part of this publication may be reproduced or transmitted in any form by any means,
electronic or mechanical, including photocopy, recording, or any information storage and
retrieval system, without permission in writing from the copyright owners.

DISCLAIMER

The authors are solely responsible for the contents of the papers compiled in this volume.
The publishers or editors do not take any responsibility for the same in any manner.
Errors, if any, are purely unintentional and readers are requested to communicate such
errors to the editors or publishers to avoid discrepancies in future.

Published by
EXCEL INDIA PUBLISHERS
91 A, Ground Floor
Pratik Market, Munirka, New Delhi–110 067
Tel: +91-11-2671 1755/ 2755/ 3755/ 5755
Fax: +91-11-2671 6755
E-mail: [email protected]
Web: www.groupexcelindia.com

Typeset by
Excel Publishing Services, New Delhi–110 067
E-mail: [email protected]

Printed by
Excel Printing Universe, New Delhi–110 067
 Preface
World population is expected to approach more than 10 billion by 2050. Currently about 40% of
the Earth’s land is used for agricultural production. Future demands from the land for feed, food,
fibre and fuel will definitely put pressure for more and more land transformation. Also, as much as
70% human-appropriated water is used for agriculture making water the Earth’s most valuable
resource. Therefore, water along with other inputs must be judiciously utilized for higher
production and also for making agriculture sustainable. But while more production is important,
the ecosystem of the human-crop-environment must be protected from its degradation rendering
this model a highly complex situation for the researchers to deal with. This requires new
techniques, processes and methods to yield desired results.
This edited book on “Precision Farming and Resource Management” has attempted to address
the above issues in a limited way. The different chapters which are in fact research papers
submitted for presentation in the International Conference entitled “Emerging Technologies in
Agricultural and Food Engineering” organized at Indian Institute of Technology Kharagpur during
27–30 December 2016. These papers have been selected out of a large number of submissions and
have been subjected to critical reviews and thorough editing for high quality of publication by
several experts chosen on the basis of their knowledge and experience. This book has been divided
into three main sections, namely, Power and Machinery Systems, Natural Resource Management
and Agro-Environmental Engineering.
In the Power and Machinery Systems the emphasis today worldwide, is on precision agriculture,
i.e., precise application of crop inputs for maximizing crop production and enhanced productivity
of the system. Therefore, innovative techniques are being evolved and are being implemented with
sensors, transducers, micro controller etc. making all machines efficient and user friendly. This
section deals with all the above in a more subtle way. Conservation and appropriate management
of major natural resources such as soil and water under changing climate scenario are of
paramount importance for sustainable crop production. The second section, describes the topics
related to water management, micro irrigation, protected cultivation, surface and subsurface
hydrology and climate change. However, indicators show that renewable resources like soil, water,
forest and fisheries are under extreme pressure due to present practices, and their productivity is
on decline. Failure to manage these resources effectively in many cases is leading to conflict and
disaster. Therefore, the papers provided in the third section deal with environment friendly
methods and approaches for a congenial “crop-soil-human” system.
I am indebted to my faculty colleagues and students of different specializations for putting their
pains staking efforts in bringing out this publication.
I hope this book will provide first-hand information on new ideas, techniques and methods applied
in the fields of precision agriculture, agro-biotechnology, climate change and natural resources
management, and would be helpful to researchers and practicing engineers.
(V.K. Tewari)
 Contents

Preface v
Editors
V.K. Tewari
H. Raheman
E.V. Thomas
R. Machavaram
Kamlesh N. Tiwari
Madan Kumar Jha
Chandranath Chatterjee
Ashok Mishra
D.K. Swain
S. Dutta Gupta
B.S. Das
C.K. Mukherjee

SECTION I: POWER AND MACHINERY SYSTEMS

1. Precision Agriculture: An Overview
V.K. Tewari 3

2. Combination Tillage Implement for Improving the Performance

of Tractor Implement Combination
H. Raheman 20

3. Draft Prediction of Commonly used Tillage Implements for

Sandy Clay Loam Soil in India
Ajay K. Roul and Hifjur Raheman 31

4. Behavior of Deflection and Contact Characteristics of

Agricultural Tyre with Zero Sinkage
Sanjeev Kumar and K.P. Pandey 46

5. Development and Performance Evaluation of a Powered Disc

Harrow
Ganesh Upadhyay and Hifjur Raheman 54

6. Impact of Promotional Activities of Tractor Sales Market in

West Bengal
V.K. Tewari and K.R. Jha 69

7. Water Emulsified Diesel and its Performance in IC Engine

Neeraj Kumar, Rajendra Machavaram and Hifjur Raheman 81

8. Posture Assessment as a Measure of Work Stress during Tractor

Operation
P. Vidhu Kampurath and V.K. Tewari 92
viii Contents

SECTION-II: NATURAL RESOURCE ENGINEERING AND MANAGEMENT

9. Precision Farming: Context of Irrigation Water Management
Kamlesh N. Tiwari 105

10. Strategies for Micro Irrigation in Guava and Mango for

Improving Quality Production
V.K. Singh, Manoj K. Soni, Anurag Singh and Ashutosh K. Pandey 122

11. Effect of Irrigation and N-Fertigation Levels on Greenhouse Grown

Tomato (Lycopersicon esculentum L.) in Semi-arid Tropical Climate
K.S. Kumar, B. Ramanjaneya and G. Shwetha 128

12. Canopy Spectral Modification for Enhancing Water and

Fertilizer Use Efficiency through Use of Color Shading Nets: A
Study for Cucumber Crop
S.B. Gadge, S.D. Gorantiwar and A.R. Jadhao 138

13. Development of a Greenhouse for Year Round Cultivation in

Sub-Humid Sub-Tropical Climatic Regions and Evaluate Its
Suitability
Vikas Kumar Singh, K.N. Tiwari and Santosh D.T. 147

14. Multi-storey Multi-Layer Stacks Protected Crop Production

Farms: A Review
Amaresh Sarkar and Mrinmoy Majumder 156

15. Watermelon Cultivation under Mulch and No-Mulch Conditions

R.M. Satasiya, D.K. Antala, V.D. Satasiya and P.M. Chauhan 168

16. Perspective of Precision Agricultural Technology in Lateritic

Zones of West Bengal: Role of KVKs and Farmers’ Clubs
Pradipta Chandra, Titas Bhattacharjee and Bhaskar Bhowmick 177

17. Modeling Evapotranspiration using Bootstrap based Quadratic

Synaptic Neural Networks
Sirisha Adamala N.S. Raghuwanshi and Ashok Mishra 195

18. Crop Evapotranspiration Estimation from Landsat 8 Data over

Kangsabati Command Area
Subhajit Roy, Chandranath Chatterjee and Narendra Singh Raghuwanshi 204

19. Evaluation of Indirect Methods for Estimating Water

Requirements of Salient Crops
Ramadhar Singh and Karan Singh 213

20. Field Verification of AquaCrop Results for Irrigated Tomato

A.S. Talokar, M.U. Kale and S.B. Wadatkar 224
 Contents ix

21. Enhancement of Agricultural Water Productivity in Irrigation

Canal Command
H.V. Parmar and N.K. Gontia 230

22. Water and Nitrogen Balance Studies of Rice Crop Grown Under
Drainage Lysimeters
Ashish Patil and K.N. Tiwari 243
23. Effect of Hydrogel Application on Water Holding Capacity of Soil
Saba Praveen, Ram Kumar Sahu and Sudarshan Prasad 254

24. Evaluating Hydrological Uncertainties Arising due to DEM

Resolution: Study for the Himalayan River Basin Gandak
Brijesh Kumar, Kanhu Charan Patra and Venkat Lakshmi 261
25. Estimation of Drought Return Periods using Bivariate Copula
Pawan S. Wable, Madan K. Jha and Ankit Shekhar 271

26. Altitudinal Variations of Soil Physico-Chemical Properties at a

High Altitude Catchment in Arunachal Himalayan Region
P. Mishra, L. Nirin, T. Golom, N. Bole, A. Bhadra and A. Bandyopadhyay 281
27. Effect of Capillary Radius Distribution on the Hydraulic
Properties of Porous Media
Junichiro Takeuchi, Hidetaka Tsuji and Masayuki Fujihara 292
28. Groundwater Potential Assessment in Tiruchirappalli District of
Tamil Nadu using Catastrophe Method
M. Annie Jenifer and Madan K. Jha 305

29. Loss of Soil and Nutrients from Complex Slopes

A.A. Atre, S.A. Patil, A.D. Kadlag and M.G. Shinde 317
30. Water Resource Development for Improving Livelihood of
Tribal Farmers in Plateau Region of Odisha (India)
R.K. Panda, Ranu Rani Sethi, S.K. Rautaray,
R.K. Mohanty, P. Panigrahi and S.K. Ambast 328

SECTION III: AGRO-ENVIRONMENTAL ENGINEERING

31. Batch Biodegradation of Floral Waste under Aerobic Conditions
Ved Prakash Ranjan and Sudha Goel 337

32. Fertilizer Inputs and Corresponding Changes in Plant Nutrient

Availability in a Rice Cultivated Soil of Assam, India
Mridusmita Debnath, Chandan Mahanpta and Arup K. Sarma 351
33. Isolation and Characterization of L-methioninase Producing
Bacteria from Marine Sources
Abhishek Das, Pradyut Kundu and Sunita Adhikari 360
 Estimation of Drought Return Periods
using Bivariate Copula

Pawan S. Wable*, Madan K. Jha and Ankit Shekhar

AgFE Department, Indian Institute of Technology Kharagpur,
West Bengal, India
E-mail: *[email protected]

Abstract: This study presents the copula-based multivariate probabilistic approach for drought
characterization by modeling severity-duration relationship of drought events. Chinchondipatil
rain-gauge station located in Ahmednagar district of Maharashtra, India was taken as a case study
and the rainfall data of 1985-2009 were used. The drought conditions were analyzed using
'standardized precipitation index' computed based on monthly rainfall accumulated over a time
scale of six months (SPI-6). Drought duration was taken as the number of consecutive months for
which SPI-6 remains below a threshold value (<-1) and drought severity as the cumulative values
of SPI values within the drought duration. The drought severity and duration were fitted to the
gamma distribution and the exponential distribution, respectively. Clayton copula, the most
commonly used copula in hydrologic studies, was employed to construct the joint distribution
function of drought severity and duration. The return periods in terms of drought severity and
duration were determined by univariate and bivariate frequency analyses. In the bivariate analysis
using copula, the return periods of drought events were studied by considering two cases:
(i) drought duration OR severity exceeding a specific value, and (ii) drought duration AND
severity exceeding a specific value. It was revealed that the bivariate analysis provides more
information about drought characteristics than the univariate analysis. The findings of this study
are useful for the risk assessment of drought, which in turn can ensure effective planning and
management of water resources in the study area.
Keywords: Standardized precipitation index (SPI), Drought characterization, Clayton copula,
Return period, Bivariate analysis

1. INTRODUCTION
Droughts are undesirable and often sudden events cause environmental, economic and
social impacts. Huschke (1959) defined drought: “A period of abnormally dry weather
sufficiently prolonged for the lack of water to cause serious hydrologic imbalance in the
affected area”. Drought is a recurring phenomenon, which may virtually occur in any
climatic zone. It is different for different people, e.g., for hydrologist it means lack of
streamflow, for meteorologist it means lack of rainfall and for agriculturist it is lack of
soil moisture. The hydrologic event is usually expressed by return period by which one
may quantify their statistical severity and frequency. In terms of drought, the return
272 Precision Farming and Resource Management

period is the average time period between the occurrences of events of equal or below a
certain magnitude (Haan, 2002).
Traditionally, the drought characteristics (duration, magnitude, and intensity) used in
estimating return period are defined in terms of univariate form. However, univariate
analysis is important only when the random variable is significant in the design process.
Gonzalez and Valdes (2003) showed that discrepancies occur due to the use of univariate
distributions and it is to be corrected by employing bivariate analysis. In multivariate
analysis of drought, due to simplicity bivariate distributions have been commonly applied
based on drought duration and severity (Shiau and Shen, 2001; Gonzalez and Valdes,
2003; Mishra et al., 2009). However, traditional multivariate technique requires the
marginal distributions of the considered random variable to follow same family of
distribution, which is difficult to achieve in most cases and if achieved, the mathematical
formulation becomes very complicated to solve when number of variables are increased
(Chen et al., 2010). To overcome this difficulty, recently copula functions have been
employed (Mishra and Singh, 2011).
A copula is a function that connects univariate distribution function to a multivariate
distribution function describing the dependence among correlated variables (Nelson,
2006). Copulas have been widely used in financial studies, but nowadays it has been
increasingly used in hydrology for multivariate analysis of hydrological random variables
(Klein et al., 2011). Salvadori and De Michele (2007) provided detailed theory and
methodology for the application of the copula in hydrology. In recent studies, copulas
have been used for the bivariate analysis of flood events (Karmakar and Simonovic,
2009; Klein et al., 2011; Sraj et al., 2015) and for the analysis of droughts (Shiau et al.,
2007; Ganguli and Reddy, 2012). Xu et al. (2010) applied copula for uncertainty analysis
in statistical modeling of extreme events. Bárdossy and Li (2008) applied copula for
geostatistical interpolation for groundwater quality parameters. Reddy and Ganguli
(2012) modeled the dependence between groundwater levels, precipitation and ENSO
index for risk assessment of groundwater levels using copula.
The basic approach for assessing the risk in hydrological studies is performed by
quantifying the return period of any extreme hydrologic event. As the return period
provides an efficient way of performing risk analysis, therefore for this study it was
estimated by using a well-known copula function. Many studies relating to frequency of
drought have been performed in the world, but in India only one such study has been
reported till now using copulas (Ganguli and Reddy, 2012). Therefore, it calls for more
and more studies in this direction in different climatic regions of India. Hence, for this
study a semi-arid region was chosen to perform risk analysis using copulas.
2. MATERIALS AND METHODS
2.1 Calculation of Standardized Precipitation Index (SPI)
The area selected for this study was upper Bhima River basin, western Maharashtra,
India. It is in semi-arid regions with an average annual rainfall of 650 mm. The daily data
 Estimation of Drought Return Periods using Bivariate Copula 273

of a Chinchondipatil raingauge station located at 18° 58’ N latitude and 74° 57’ E
longitude with an elevation of 652 m above MSL for 25 years (1985-2009) was used for
drought analysis. The rainfall data were procured for State Data Storage Centre,
Hydrology Project (HP), Nashik. For calculation of SPI values, a long-term rainfall series
was fitted to gamma distribution and then their cumulative distribution was transformed
to a standard normal distribution (McKee et al., 1993). An important feature of SPI is that
it can quantify precipitation deficit for multiple time scales such as one month (SPI-1),
three months (SPI-3), and six months (SPI-6) and so on. In this study, SPI-6 was
considered for drought analysis as it is generally used for seasonal drought identification
and has more effect on water supplies than on agriculture (Ganguli and Reddy, 2012;
Masud et al., 2015). The SPI drought indices for this study have been calculated based on
rainfall data using SPI Calculator provided by the National Drought Mitigation Center,
University of Nebraska Lincoln (http://drought.unl.edu/).
The theory of runs proposed by Yevjevich (1967) was used to characterize the
drought. The interpretation of drought characteristics calculated using SPI-6 for one
drought incidence is shown in Fig. 1. Drought onset and termination are starting and
ending of the water deficit period, respectively. Drought duration is the time for which
the SPI-6 values remain below a threshold value ‘-1’, whereas drought severity is the
cumulative value of SPI-6 within the drought duration (shaded area), and drought inter-
arrival time is the time interval between two consecutive drought events (Fig. 1).

Fig. 1: Calculation of Drought Duration, Drought Severity

and Inter-arrival Time based on Theory of Runs
274 Precision Farming and Resource Management

2.2 Overview of Copula

Copula is a tool for modeling and capturing the dependence of two or more random
variables. They are joint distribution in which random variables are of standard uniform
distribution. A bivariate distribution function FX ,Y ( x , y ) of two correlated random
variables X and Y with continuous marginal or cumulative distributions functions
(CDFs), and v F y y .

According to Sklar’s theorem (Sklar, 1959) the joint distribution can be expressed as
(Nelson, 2006):

FX ,Y (x, y) C FX (x), FY ( y) C(u, v) x, y (1)

where, C = copula function, ‘C’ is unique if Fx x and Fy y are continuous.

Generally, drought duration and severity can be fitted to the gamma and exponential
distributions, respectively (Shiau and Modarres, 2009). The Kolmogorov-Smirnov (K-S)
goodness-of-fit test was used to detect the efficacy of the proposed models to represent
the observed data.
In this study, from the Archimedean copula family, the Clayton copula was employed
to model the dependence between drought duration and severity as it is simple and
commonly used in hydrology (Shiau and Modarres, 2009). The expression of Clayton
copula-based joint cumulative distribution function (JCDF) of drought duration and
severity is given as follows:
1
C FS , FD FS FD 1 , 0 (2)

where, FS and FD are cumulative distribution functions of severity and duration,

respectively.
For Clayton copula, the relation between parameter, and Kendall’s coefficient of
correlation, , is given by the following equation (Nelson, 2006):

, 0, (3)
2
Kendall’s coefficient of correlation, can be obtained by the following equation:
c d
(4)
c d
where, c and d = number of concordant and discordant pairs, respectively.
 Estimation of Drought Return Periods using Bivariate Copula 275

2.3 Estimation of Return Periods using Univariate

Distribution and Bivariate Copula
In univariate context, the return period of droughts in terms of drought severity and
duration can be expressed as (Ganguli and Reddy, 2012):
E(I d )
TD (5)
1 FD (d)

E( I d )
TS (6)
1 FS (s)
where, E(Id) is the expected or mean inter-arrival time of drought events which can
be estimated from observed drought data.
The expected or mean inter-arrival time E( I d ) can be calculated using the following
equation:
N (x y)
E( Id ) (7)
Dinc 1
where, N= Total number of months, x = Number of non-drought months before the
start of the first drought, y = Number of non-months after the end of the last drought,
Dinc = Number of drought incidences.
The two cases of bivariate return periods can be computed: either drought duration
OR severity exceeds a specific value, TDS (i.e., D d or S s); drought duration AND
severity exceeds a specific value, T’DS (i.e., D d and S s). These two types of return
periods can be computed by using a copula-based approach, which is given by Eqns. (8)
and (9) (Ganguli and Reddy, 2012).
E(Id ) E(I d ) E(Id )
TDS (8)
P D or S s 1 FDS d , s 1 C FD d , FS s

E( I d ) E(I d )
T 'DS
P D d and S s 1 FD d SS s FDS d , s

E(Id )
(9)
1 FD d FS s C FD d , FS s

These two cases of bivariate return periods were considered in this study and were
compared with univariate return periods. In addition, contour lines or iso-lines for both
TDS and T’DS were constructed in MATLAB software using the contour function.
276 Precision Farming and Resource Management

3. RESULTS AND DISCUSSION

3.1 Evaluation of Drought using SPI-6
The time series of SPI-6 at Chinchondipatil station for the 1985-2009 period is shown in
Fig. 2. By applying the theory of runs to this time series, the number of drought
incidences, i.e., SPI-6 values less than -1 are found to be 17. The maximum value of
drought severity (15.45) and duration (8 months) is observed in the year 1994 (Fig. 2).
The mean inter-arrival time E(Id) calculated by using Eqn. (7) is found to be 16.9 months.

Fig. 2: SPI-6 Time Series for Chinchondipatil Station during 1985–2009 Period, in this Figure, Red
Circle Indicate Drought with Maximum Severity and Duration

3.2 Copula-based Joint CDF of Drought Severity and Duration

Based on the K-S goodness-of-fit test at the 5% level of significance, it is confirmed that
proposed gamma distribution fits drought severity and drought duration represents the
exponential distribution, as the calculated value of K-S is less than the critical value. The
estimated Kendall’s tau ( ) value between drought duration and severity is 0.89. This
obtained value indicates that there is high dependence between drought duration and
severity which further verifies the use of joint probability distribution for frequency
analysis. Using this Kendall’s tau ( ) value, the parameter ( = 15.58) was obtained using
Eqn. (3). Further, the joint probability distribution using Clayton copula (Eqn. 2) is
estimated for drought duration and drought severity and its surface plot is shown in
Fig. 3. It can be inferred that the value of copula CDF increases with the increase of FS (s)
and FD (d) values and vice-versa.
3.3 Univariate and Bivariate Drought Analysis
The univariate and bivariate drought return periods were estimated using available
analytical equations. The univariate return period for drought duration and severity were
estimated using Eqns. (5) and (6). It is found in most cases that the univariate return
periods in terms of severity (TS) are always smaller than the duration return periods (TD)
(Table 1). For the year 1994, having a maximum drought severity and duration, the
 Estimation of Drought Return Periods using Bivariate Copula 277

univariate return period for drought duration is 25.43 years and that for drought severity
is 66.29 years. However, in case of bivariate return periods TDS (Eqn. 8) and T’DS (Eqn. 9)
are 21.83 and 116.83 years, respectively.

Fig. 3: Surface Plot of Joint CDF (C(FS(s), FD(d)), where FS(s)

and FD(d) Denotes CDF of Severity and Duration
The iso-lines of the bivariate return period for specific severity (TDS and T’DS) are
shown in Fig. 4(a) and Fig. 4(b). The iso-lines or contours of bivariate return periods
revealed that various combinations of severity and duration lead to same return period. It
can be shown from these figures that the contour lines for specific joint return periods in
which either severity or duration exceeded TDS have no bounds [Fig. 4(a)]; whereas the
bivariate return periods in which both severity and duration exceed (T’DS) are described
by horizontal and vertical axes [Fig. 4(b)], i.e., have bounds. An important observation
from calculation of return period TDS and T’DS (Table 1) is that the difference between
them is more expressive for maximum values compared to median values of drought
duration and severity (Sraj et al., 2015).
The contours of the bivariate return period for OR and AND conditions can be
explained in a more specific way as for instance if we consider drought duration and
severity equal to 4 and 10 months, respectively, then in that case the corresponding return
period for OR and AND condition is equal to or nearly 5 years. However, on comparison
of T’DS corresponding to AND with OR condition (TDS), the former provides better
drought return period estimation as it is bounded by the maximum severity of value 10
against any value greater than 10 (unbounded) for latter condition. These estimated return
periods indicate how often the drought is likely to repeat itself, which further can help to
assure the water availability during the severe drought periods. In addition, such
278 Precision Farming and Resource Management

information can serve vital for water policy and decision makers to develop management
strategies to mitigate the impacts of droughts in the study area.
Table 1: Univariate and Bivariate Drought Return Periods
Univariate Return Period Bivariate Return Period
TD TS TDS TDS'
17.71 25.04 8.48 23.08
12.34 9.36 4.95 8.92
2.02 1.90 1.13 1.21
4.17 3.24 1.92 2.51
2.02 1.84 1.10 1.21
2.02 2.24 1.21 1.34
4.17 4.46 2.35 2.84
2.90 2.67 1.58 1.76
2.02 1.86 1.11 1.21
25.43 66.29 21.83 116.33
2.90 2.63 1.56 1.76
2.90 2.46 1.47 1.74
2.90 2.49 1.48 1.74
5.98 4.65 2.70 3.72
2.02 1.90 1.13 1.21
2.02 2.04 1.19 1.24
2.90 3.29 1.71 2.00
Note: Bold values indicate the return period values corresponding to maximum drought duration and
severity.

(a) (b)

Fig. 4: Bivariate Return Period of Drought Severity and Duration:

(a) Drought Duration OR Severity Exceeding a Specific Value TDS;
(b) Drought Duration AND Severity Exceeding a Specific Value T’DS
 Estimation of Drought Return Periods using Bivariate Copula 279

4. CONCLUSIONS
The study reveals the application of the copula technique in semi-arid region of western
India for estimation of drought return periods and its comparison with univariate return
periods. Based on the results of this study, the following conclusions can be drawn:
It is found that the univariate return periods in terms of severity (TS) are always
smaller than the univariate duration return periods (TD).
The return periods in which either drought duration OR severity exceeds a
specific value (TDS) are smaller than that of the return periods in which both
drought duration AND severity exceeds a specific value (T’DS).
On comparing the univariate return periods (TD and TS) estimated using drought
severity and duration separately with the bivariate return periods (TDS and T’DS),
out of all TDS resulted in smaller return period values, whereas T’DS is found to be
higher as one additional condition is used for estimating the return periods.
The iso-lines for specific bivariate return periods in which either severity OR
duration exceeds has no bounds, whereas in case of bivariate return periods in
which either severity AND duration exceeds has bounds.
Overall, it can be concluded that bivariate return periods provide risk reliable results,
which are more reasonable for practical realm and thus, can be useful for the risk
assessment associated with droughts and hydrologic design in the study area. For
modeling the dependence between drought duration and severity, a rigorous evaluation of
other copula families together with consideration of other drought characteristics is
recommended in the future as a follow-up to this study. This study as a whole is useful
for drought-affected regions around the globe.
REFERENCES
[1] Bárdossy, A. and Li, J. (2008). Geostatistical interpolation using copulas. Water Resources Research,
44, W07412, doi: 10.1029/2007WR006115.
[2] Chen, L., Guo, S.L., Yan, B.W., Liu, P. and Fang, B. (2010). A new seasonal design flood method based
on bivariate joint distribution of flood magnitude and date of occurrence. Hydrological Sciences Journal,
55(8): 1264–1280.
[3] Ganguli, P. and Reddy, M.J. (2012). Risk assessment of droughts in Gujarat using bivariate copulas.
Water Resources Management, 26(11): 3301-3327.
[4] Gonzalez, J. and Valdes, J.B. (2003). Bivariate drought recurrence analysis using tree ring
reconstructions. Journal of Hydrologic Engineering, 8(5): 247-258.
[5] Haan, C.T. (2002). Statistical Methods in Hydrology. The Iowa State University Press, Ames, Iowa,
USA.
[6] Huschke, R.E. (1959). Glossary of Meteorology. American Meteorological Society, Boston.
[7] Karmakar, S. and Simonovic, S.P. (2009). Bivariate flood frequency analysis. Part 2: A copula-based
approach with mixed marginal distributions. Journal of Flood Risk Management, 2(1): 32-44.
[8] Klein, B., Schumann, A.H. and Pahlow, M. (2011). Copulas–new risk assessment methodology for dam
safety. In: A.H. Schumann (editor), Flood Risk Assessment and Management, Springer, The
Netherlands, pp. 149-185.
280 Precision Farming and Resource Management

[9] Masud, M.B., Khaliq, M.N. and Wheater, H.S. (2015). Analysis of meteorological droughts for the
Saskatchewan River Basin using univariate and bivariate approaches. Journal of Hydrology, 522: 452-466.
[10] McKee, T.B., Doesken, N.J. and Kleist, J. (1993). The relationship of drought frequency and duration to
time scales. In: 8th Conference on Applied Climatology. American Meteorological Society, Boston, MA,
pp. 179–184.
[11] Mishra, A.K. and Singh, V.P. (2011). Drought modeling: A review. Journal of Hydrology, 403(1):
157-175.
[12] Mishra, A.K., Singh, V.P. and Desai, V.R. (2009). Drought characterization: a probabilistic approach.
Stochastic Environmental Research and Risk Assessment, 23(1): 41-55.
[13] Nelsen, R.B. (2006). An Introduction to Copulas. Springer, New York, USA.
[14] Reddy, M.J. and Ganguli, P. (2012). Risk assessment of hydroclimatic variability on groundwater levels
in the Manjara basin aquifer in India using Archimedean-copulas. Journal of Hydrologic Engineering,
ASCE, 17(12): 1345-1357.
[15] Salvadori, G. and De Michele, C. (2007). On the use of copulas in hydrology: theory and practice.
Journal of Hydrologic Engineering, ASCE, 12(4): 369-380.
[16] Shiau J.T., Feng S. and Nadaraiah S. (2007). Assessment of hydrological droughts for the Yellow River,
China, using copulas. Hydrological Processes, 21(16): 2157–2163.
[17] Shiau, J.T. and Modarres, R. (2009). Copula based drought severity duration frequency analysis in Iran.
Meteorological Applications, 16(4): 481-489.
[18] Shiau, J.T. and Shen, H.W. (2001). Recurrence analysis of hydrologic droughts of differing severity.
Journal of Water Resources Planning and Management, ASCE, 127(1): 30-40.
[19] Sklar A. (1959). Fonctions de répartition à n dimensions et leurs marges. Publications de l’Institut de
Statistique de l’Universit´e de Paris, 8: 229–231.
[20] Sraj, M., Bezak, N. and Brilly, M. (2015). Bivariate flood frequency analysis using the copula function:
a case study of the Litija station on the Sava River. Hydrological Processes, 29(2): 225-238.
[21] Xu, Y.P., Booij, M.J. and Tong, Y.B. (2010). Uncertainty analysis in statistical modeling of extreme
hydrological events. Stochastic Environmental Research and Risk Assessment, 24(5): 567-578.
[22] Yevjevich, V. (1967). An Objective Approach to Definitions and Investigations of Continental
Hydrologic Droughts. Colorado State University, Fort Collins, Hydrology Paper 23.