Content uploaded by Rehana Shaik
Author content
All content in this area was uploaded by Rehana Shaik on Mar 16, 2018
Content may be subject to copyright.
Regional impacts of climate change on irrigation
water demands
S. Rehana
1
and P. P. Mujumdar
1,2
*
1
Department of Civil Engineering, Indian Institute of Science, Bangalore, Karnataka 560 012, India
2
Divecha Center for Climate Change, Indian Institute of Science, Bangalore, Karnataka 560 012, India
Abstract:
This paper presents an approach to model the expected impacts of climate change on irrigation water demand in a reservoir
command area. A statistical downscaling model and an evapotranspiration model are used with a general circulation model
(GCM) output to predict the anticipated change in the monthly irrigation water requirement of a crop. Specifically, we quantify
the likely changes in irrigation water demands at a location in the command area, as a response to the projected changes in
precipitation and evapotranspiration at that location. Statistical downscaling with a canonical correlation analysis is carried out to
develop the future scenarios of meteorological variables (rainfall, relative humidity (RH), wind speed (U
2
), radiation, maximum
(Tmax) and minimum (Tmin) temperatures) starting with simulations provided by a GCM for a specified emission scenario. The
medium resolution Model for Interdisciplinary Research on Climate GCM is used with the A1B scenario, to assess the likely
changes in irrigation demands for paddy, sugarcane, permanent garden and semidry crops over the command area of Bhadra
reservoir, India.
Results from the downscaling model suggest that the monthly rainfall is likely to increase in the reservoir command area. RH,
Tmax and Tmin are also projected to increase with small changes in U
2
. Consequently, the reference evapotranspiration,
modeled by the Penman–Monteith equation, is predicted to increase. The irrigation requirements are assessed on monthly scale at
nine selected locations encompassing the Bhadra reservoir command area. The irrigation requirements are projected to increase,
in most cases, suggesting that the effect of projected increase in rainfall on the irrigation demands is offset by the effect due to
projected increase/change in other meteorological variables (viz., Tmax and Tmin, solar radiation, RH and U
2
). The irrigation
demand assessment study carried out at a river basin will be useful for future irrigation management systems. Copyright © 2012
John Wiley & Sons, Ltd.
KEY WORDS climate change; statistical downscaling; GCM; irrigation demands; evapotranspiration
Received 9 November 2011; Accepted 20 April 2012
INTRODUCTION
The rising CO
2
and climate change due to global warming
directly affect both precipitation and evapotranspiration,
consequently the irrigation water demands. Moreover, the
irrigation water requirements of the crops change as a
function of climate change. Several authors have focused
on assessing the impacts of climate change on agriculture,
over the past decade. Most of these studies concentrated on
estimating the changes in crop productivity (e.g. Easterling
et al., 1993; Rosenzweig and Parry, 1994; Singh et al.,
1998; Brown and Rosenberg, 1999; Parry et al., 2004;
Harmsen et al., 2009; Liu et al., 2010). Assessment studies
focusing on the impacts of climate change on irrigation
demands using general circulation model (GCM) outputs
are becoming more accepted in recent years. GCMs are
excellent tools to study the climate change impact and have
been used in recent studies globally. Yano et al.(2007)
studied the effects of climate change on crop growth and
irrigation water demand for a wheat–maize cropping
sequence in a Mediterranean environment of Turkey.
The climate change scenarios of temperature and precipi-
tation were created by superimposing projected anomalies
of GCMs on observed climate data of the baseline period.
Elgaali et al.(2007)modeledtheregionalimpactof
climate change on irrigation water demand by considering
rainfall and evapotranspiration in the Arkansas River Basin
in southeastern Colorado. They assumed no change in crop
phenology and found an overall increase in irrigation water
demands due to climate change. The historical climate data
sets of historical and projections for the continental United
States are considered from Vegetation Ecosystem
Modeling and Analysis Project developed by Kittel et al.
(1995). Rodriguez Diaz et al. (2007) showed increase of
irrigation demand between 15% and 20% in seasonal
irrigation need by 2050 in the Guadalquivir river basin in
Spain with perturbed climate scenarios of temperature,
precipitation, solar radiation, wind speed (U
2
) and relative
humidity (RH). Shahid (2011) estimated the changes of
irrigation water demand in dry-season Boro rice field in
northwest Bangladesh in the context of global climate
change, with projected changes of rainfall and tempera-
tures estimated using the modeling software SCENario
GENerator (SCENGEN).
*Correspondence to: P. P. Mujumdar, Divecha Center for Climate Change,
Indian Institute of Science, Bangalore, Karnataka 560 012, India
E-mail: pradeep@civil.iisc.ernet.in
HYDROLOGICAL PROCESSES
Hydrol. Process. (2012)
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/hyp.9379
Copyright © 2012 John Wiley & Sons, Ltd.
de Silva et al. (2007) studied the impacts of climate
change on irrigation water requirements in the paddy field
of Sri Lanka and predicted an increase of 13% to 23% of
irrigation water demand depending on climate change
scenarios. The climate change scenarios of temperature,
radiation, U
2
and RH are developed by applying the
percentage changes of GCM to the baseline dataset. The
proportional (%) changes given by a selected GCM and
scenario are applied on an existing baseline climatological
dataset to develop the future scenarios of the variables
required for a water balance model to estimate the paddy
irrigation requirements for a single site.
Most of these studies focused on evaluation of crop
water requirements based on perturbed climate change
scenarios generated with GCM outputs or with available
downscaled data sets or using modeling softwares such as
SCENGEN. With the development of statistical down-
scaling models (SDSMs), the regional climate change
assessment studies are becoming more accepted. There-
fore, this study uses a SDSM, as the downscaling
methods are well accepted in the climate change impact
assessment studies in the recent years by the research
community. Therefore, this study emphasizes on adopt-
ing such sophisticated methods to quantify the future
projected irrigation demands. This forms the basic
difference between the present work and the work done
in de Silva et al. (2007). A multivariable downscaling
methodology is applied at each location to develop the
future scenarios of rainfall, temperature, RH and U
2
.
Further, the difference between the rainfall and the
potential evapotranspiration is considered as the irriga-
tion water requirement for a particular crop at a particular
location. This study stresses on climate change impact
assessment of irrigation demands at a reservoir command
area using a SDSM. To obtain the projected climate
change scenarios of rainfall as well as other meteoro-
logical variables which influence the evapotranspiration
(viz., RH, U
2
, radiation, maximum (Tmax) and minimum
(Tmin) temperatures) at the scale of command area, from
a GCM, a multivariable downscaling technique, canon-
ical correlation analysis (CCA) is adopted. The antici-
pated irrigation demands of the crops are examined for
the future scenarios by accounting for the changes in
rainfall and potential evapotranspiration.
STUDY AREA
The command area of the Bhadra reservoir is considered
for the assessment of impacts of climate change on
irrigation demands. Bhadra is a tributary of Krishna
River, originating from Gangamula in the Western Ghats
of Chikamagalur District in Karnataka state, India. The
river flows through nearly 190 km from its origin and
joins River Tunga to form the River Tunga-Bhadra. The
Bhadra reservoir intercepts the river flow and provides
water for irrigation. The reservoir project also generates
hydropower to a minor extent. The gross command area
under the Bhadra Canal System is 162,818 ha with a
culturable command area of 121,500 ha out of which
105,570 ha have been earmarked for irrigation. The
irrigated area of 105, 570 ha is considered for impact
assessment in this study. The irrigated area predomin-
antly consists of red loamy soil except in some portion
of the right canal area, which has black cotton soil. The
assessment of irrigation demands is carried out on
paddy, sugarcane, permanent garden and semi dry crops,
which are the typical crops grown in the Bhadra
command area.
The meteorological variables (Tmax and Tmin, U
2
and RH)
from 1969 to 2005 at Shimoga and high-resolution
gridded daily precipitation data from1971 to 2005 at a
0.5
0
0.5
0
grid interpolated from station data are
obtained from the India Meteorological Department
(IMD), Pune. The command area of Bhadra river spreads
over the districts of Chitradurga, Shimoga, Chickmagalur
and Bellary. Nine IMD locations are selected to evaluate
the irrigation demands in the command area. The total
irrigated area of each crop in the command area is
distributed equally among these selected nine locations.
Thus, each downscaling location represents an area
consisting of all the crops. The 0.5
0
0.5
0
IMD grid
points falling in the districts of Chitradurga, Shimoga,
Chickmagalur and Bellary are considered as rainfall
downscaling locations as shown in Figure 1. The latitudes
and longitudes of each of the nine downscaling locations
are given in Table I.
STATISTICAL DOWNSCALING
The statistical downscaling techniques are generally used to
bridge the spatial and temporal resolution gaps between the
coarser resolution of the GCMs and the finer resolution
required in the impact assessment studies. Generally, these
methods involve deriving empirical relationships that
transform large-scale simulations provided by a GCM
(climate variables as predictors) to regional-scale variables
(surface variables as predictands). As a first step in the impact
studies, the predictands to be downscaled must be selected.
The hydro-meteorological variables that have a major
influence on crop water requirements are the rainfall and
evapotranspiration (Elgaali et al., 2007; Rodriguez Diaz
et al., 2007). Evapotranspiration is mainly influenced by the
air temperature, U
2
, RH, and solar radiation. Many impact
assessment studies on reference evapotranspiration have
dealt with only temperature variables of Tmax and Tmin (e.g.
Harmsen et al., 2009; Lovelli et al., 2010; Maeda et al., 2011;
Torres et al., 2011). However, the present study uses
temperature variables as well as RH, U
2
and radiation. The
temperature variables (Tmax and Tmin), RH and U
2
are
modeled (as predictands) with a statistical downscaling
technique using GCM outputs. The data used and down-
scaling methodology are described in the following section.
Data extraction and statistical downscaling
The first step in statistical downscaling is the selection
of atmospheric predictor variables to model the selected
predictand variables. Following the literature (Table II)
S. REHANA AND P. P. MUJUMDAR
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
and the availability of the predictors from the GCM, 13
large-scale atmospheric predictors (precipitation flux,
precipitable water, surface air temperature at 2 m, mean
sea level pressure, geopotential height at 500 mb, surface
U-wind, surface V-wind, specific humidity at 2 m, surface
RH, surface latent heat flux, sensible heat flux, surface
short wave radiation flux, surface long wave radiation
flux) are selected. Five predictand variables are chosen to
be modeled by the selected predictors. These are rainfall,
Tmax and Tmin, RH and U
2
.
An area from 10
0
–20
0
Nto70
0
–80
0
E, encompassing
the region where meteorological variables are to be
downscaled, is chosen for the large-scale predictors. Data
on the predictors at monthly time scale are obtained from
the National Centers for Environmental Prediction/
National Center for Atmospheric Research (NCEP/
NCAR) reanalysis data (Kalnay et al., 1996) (available
at http://www.cdc.noaa.gov/cdc/data.ncep.reanalysis.
html) and are used for training the downscaling model.
The medium resolution Model for Interdisciplinary
Research on Climate version 3.2 (MIROC 3.2) GCM
(medium-resolution of 1.125 1.125 deg, GCM from the
Center for Climate System Research, Japan) is used with
the A1B scenario (IPCC, 2007), for the impact assess-
ment. The particular GCM is used keeping in view the
availability of the projections on the predictors at the
monthly scale. The A1B scenario represents a balanced
emission scenario with medium emission trajectories, and
is used here as a possible future scenario.
Large-scale monthly atmospheric variables output from
the MIROC 3.2 GCM for the A1B scenario (720 ppm
Downscaling Location
s
Bhadra Reservoir
Command Area
Figure 1. Downscaling locations in the Bhadra Command Area
Table I. Locations for downscaling precipitation
Location Latitude Longitude
1 13.5
0
N 75.5
0
E
2 13.5
0
N 76.0
0
E
3 14.0
0
N 75.0
0
E
4 14.0
0
N 75.5
0
E
5 14.0
0
N 76.0
0
E
6 14.0
0
N 76.5
0
E
7 14.5
0
N 76.0
0
E
8 14.5
0
N 76.5
0
E
9 15.0
0
N 76.0
0
E
CLIMATE CHANGE AND IRRIGATION DEMANDS INTEGRATION
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
CO
2
stabilization experiment) is extracted from the multi-
model data set of the World Climate Research
Programme’s Coupled Model Inter Comparison Project
(available at https://esg.llnl.gov:8443/about/ftp.do). The
dimension of the predictor variables set is 25/30/42
(number of NCEP grid points for surface flux, surface/
pressure and radiation flux variables, respectively)
13
(number of predictors), which is very large, and working
out the model with this large number would be
computationally cumbersome. Principal component
analysis (PCA) is applied on the large data set to reduce
the dimensionality and to effectively summarize the
spatial information from the 25/30/42 grid points. It was
found that 95% of the variability of original set is
explained by the first 12 PCs. The eigen vectors or
coefficients obtained from NCEP data were applied to the
standardized MIROC3.2 data to get the projections in the
principal directions. Standard procedure of statistical
downscaling (e.g. Raje and Mujumdar, 2009) involving
standardization, interpolation, PCA and developing a
statistical relationship between predicands and predictors
is followed in this study. Interpolation is performed
before standardization to obtain the GCM output at NCEP
grid points as the location of NCEP/NCAR grid points
and MIROC grid points do not match. A Mercator
projection (conformal cylindrical map projection), suit-
able for tropical regions (Mulcahy and Clarke, 1995),
is first performed, and then a linear interpolation is
performed between the projected points. Standardization
(Wilby et al., 2004) is performed prior to PCA
and downscaling to remove systematic bias in mean
and standard deviation of the GCM simulated
climate variables.
Canonical correlation analysis
In the procedure for statistical downscaling followed
in this study, a mathematical transfer function is to be
adopted to derive predictor–predictand relationship
which can account for the multivariate predictands.
The most commonly used statistical technique with
multivariate data sets is CCA. CCA can be used as a
downscaling technique for relating surface-based
observations and free-atmosphere variables when sim-
ultaneous projection of predictands is of interest (e.g.
Barnett and Preisendorfer, 1987; Graham et al., 1987;
Karl et al., 1990; Barnston, 1994; Mpelasoka et al.,
2001; Juneng and Tangang, 2008). CCA has found wide
application in modeling precipitation and meteorological
variables (e.g. Von Storch et al., 1993; Gyalistras et al.,
1994; Busuioc and von Storch, 1996). An advantage of
the CCA in the context of downscaling is that the
relationships between climate variables and the surface
hydrologic variables are simultaneously expressed, as
they in fact occur in nature, by retaining the explained
variance between the two sets. CCA finds pairs of linear
combinations between the N-dimensional climate variables,
X, (predictors, in this case) and M-dimensional surface
variables, Y, (predictands, in this case) which can be
expressed as follows:
Um¼aTX;m¼1; ::::::: min N;MðÞ (1)
Vm¼bTY;m¼1; ::::::: min N;MðÞ (2)
where U
m
and V
m
are called predictor and predictand
canonical variables respectively, a = [a
1
,a
2
,.....a
N
]and
b=[b
1
,b
2
,.....b
M
] are called the canonical loadings. The
objective of canonical correlation is to identify msets of
canonical variables such that the correlation, r, between the
predictor canonical variable, U
m
, and the predictand
canonical variable, V
m
, is maximum. This way N-dimensional
predictor set and M-dimensional predictand set is reduced
to m-dimensional canonical variables which will be
further useful in developing the regression equations for
each predictand. After the estimation of canonical variables,
regression relation is established for each of the predictand
as discussed in the following section.
Linear regression using CCA
The methodology involves training the surface observed
predictands and NCEP atmospheric predictor data with the
CCA analysis after data preprocessing with standardization
and PCA. The PCs obtained based on NCEP data are used as
reference to develop the GCM PCs. A separate regression
Table II. Predictors selected for the statistical downscaling
Predictand Predictors
Rainfall Mean sea level pressure, geopotential
height at 500 mb(Ghosh and Mujumdar,
2006); specific humidity at 500 hPa,
precipitation flux, surface air temperature
at 2 m, maximum surface air temperature
at 2 m, minimum surface air temperature
at 2 m, surface U-wind and surface
V-wind (Raje and Mujumdar, 2009).
Maximum and
minimum
temperatures
Air temperature, zonal and meridional
wind velocities at 925 mb, surface flux
variables such as latent heat, sensible
heat, shortwave radiation and long
wave radiation fluxes (Anandhi et al.,
2009).
Wind variables Geopotential height, air temperature,
U-wind and V-wind speed, relative
humidity, vertical velocity, absolute
vorticity as multilevel quantities
evaluated at 1000 hpa height (Davy
et al., 2010)
relative humidity,
water vapor pressure,
dew-point
temperature, and
dew-point deficit
Geopotential height at 500, 850 and
1000 hpa, wind speed and vorticity at
500, 850 hpa, temperature at 850 hpa,
humidity variables (relative humidity,
specific humidity, water vapor pressure,
dew-point temperature, dew-point
deficit at 850 hpa) (Huth, 2005)
The references cited in the table indicate the earlier studies in which the
predictors are used for the specified predictands
S. REHANA AND P. P. MUJUMDAR
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
equation is derived for each meteorological predictand
variable from the canonical variable coefficients and
correlations computed from the observed data. First few
PCs are extracted based on the percentage variance
explained by them. The selected PCs from the NCEP data
are considered as predictor set to perform CCA to fitthe
regression relation between the climate variables and
surface-based observations. The observed predictor
canonical variable, U
obs,q
, is computed from Equation (1)
with the NCEP PCs as follows:
Uobs;q¼aTXNCEP;PCs (3)
In Equation (3), qrepresents the minimum among the
number of PCs considered and the number of predictands
considered. As the number of PCs considered is 12 in this
case, to account for 95% variability, and the number of
predictands considered is five, CCA will yield five
predictor and predictand canonical variables and five
canonical correlations between them. The predictand
canonical variable, V
predicted,q
, can be evaluated from the
predictor canonical variable, U
obs,q
, obtained from
Equation (3) as follows:
Vpredicted;q¼rCq Uobs;q(4)
In Equation (4), r
Cq
is the canonical correlation coeffi-
cient and represents the percent of variance in the predictand
canonical variable explained by the predictor canonical
variable. It is a diagonal matrix of size qxq. The regression
equations (Equation (4)) are applied to the interpolated
NCEP gridded GCM output to model future projections of
hydro-climate predictands. The downscaled scenario for
each of the predictand can be derived according to:
Ypredicted;q¼b1
Vpredicted;q(5)
where Y
predicted,q
is the qnumber of predictand variables to
be evaluated from the predictand canonical variables
V
predicted,q
and the predictand canonical loadings b.
Prediction of future scenario is made using the PCs of
monthly outputs of the atmospheric variables (predictors)
from the GCM in place of NCEP PCs in Equation (3). The
canonical correlations and the loadings are computed using
statistical toolbox of MATLAB (2004). This downscaling
methodology is applied to downscale the rainfall and other
meteorological variables at nine downscaling locations.
Shimoga station meteorological parameters are used for
other downscaling locations due to the availability of
observed data only at Shimoga station. A monthly time
period is considered for all variables. The SDSM is trained
using the past records of atmospheric and surface meteoro-
logical data of 25 years (1971 to 1995) to estimate the
canonical scores, and the model is tested with the remaining
data, for the period 1996 to 2004. Once the model
performance is found satisfactory in the testing period, it
can be applied for obtaining the future predictions. Table III
gives the details of the statistics such as mean, standard
deviation of observed and CCA downscaled results for the
testing period of 1996 to 2004. The R-value in Table III
indicates the correlation coefficient between the observed
and CCA modeled results for various variables. The results
of CCA downscaling model are used as model input
variables to simulate the impact of climate change on
irrigation demands for each crop at each downscaling
location.
ESTIMATION OF IRRIGATION DEMANDS
The total irrigation demand in the command area is
computed based on the potential evapotranspiration of a
crop and the rainfall contribution. The total demand in
period t, for a particular crop, c, at a downscaling station,
s, is given by:
Dt;c;s¼ETc
tRt;s
Ac;sif Rt;s<ETc
t(6)
Dt;c;s¼0ifRt;s>ETc
t(7)
where ETc
tis the potential evapotranspiration of a crop, c
in period t;R
t,s
is the rainfall contribution in period t,ata
downscaling station, s;A
c,s
is the area over which the crop
cis grown at station s.
In the demand equations given above (Equations (6)
and (7)), the soil moisture contribution to meeting crop
water demand is neglected. Further, the rainfall amount
considered in the evaluation of irrigation demands is the
total rainfall measured from rain-gauges at each down-
scaling location instead of effective rainfall. The compu-
tation of effective rainfall involves measured rainfall,
surface runoff losses, percolation losses beyond root zone
and soil moisture details.
Evapotranspiration model
The reference evapotranspiration is estimated by
Penman–Monteith (Allen et al., 1998) equation, given
as follows:
ETt;R¼0:408ΔRnGðÞþg900=Tþ273ðÞðÞU2esea
ðÞ
Δþg1þ0:34U2
ðÞ
(8)
where ET
t,R
is the reference evapotranspiration of each
month (mm/month), Δis the slope of the vapor pressure
curve, R
n
is net radiation at the surface (w/m
2
), gis
psychrometric constant, Tis the average air temperature
at 2-m height, U
2
is wind speed at 2-m height, e
s
is the
saturated vapor pressure and e
a
is the actual vapor
pressure (kpa).
The future projections of meteorological variables
downscaled from the GCM outputs, including RH, U
2
,
R
n
,Tmax and Tmin, are used as input to the evapotrans-
piration model (Penman–Monteith equation (Equation (8))
to evaluate the anticipated changes in the reference
evapotranspiration. Among these meteorological variables,
solar radiation could not be directly downscaled in this
study due to the nonexistence of observed solar radiation
data for the study region. Most of the methods to estimate
solar radiation (e.g. Angstrom, 1924; Hargreaves, 1994)
CLIMATE CHANGE AND IRRIGATION DEMANDS INTEGRATION
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
include the information of cloud cover, Tmax and Tmin,
sunshine hours, RH and site-specificcoefficients. However,
Hargreaves and Samani (1982) recommended a simple
equation to estimate the solar radiation based on Tmax and
Tmin. As the observations of Tmax and Tmin are available
for the study region, these variables can be downscaled, and
the future projections of solar radiation can be computed
based on the downscaled variables of Tmax and Tmin. The
R
n
in the Equation (8) is estimated using Hargreaves’s
radiation formula (Hargreaves and Samani, 1982):
Rn¼krs Tmax Tmin
ðÞ
1=2Ra(9)
where k
rs
is an adjustment factor equal to 0.16 for interior
locations and 0.19 for coastal locations; T
max
and T
min
are
the mean monthly maximum and minimum air temperatures
respectively in
0
C;R
a
is extraterrestrial radiation (w/m
2
)and
is computed from expressions given in Allen et al. (1998).
The reference evapotranspiration (ET
t,R
) obtained
(Equation (8)) needs to be adjusted to obtain the potential
crop evapotranspiration (ETc
t;p) with crop coefficients for
each period, tfor a crop c (k
t,c
) Thus, ETc
t;pis given by:
ETc
t;p¼ETt;RXkt;c(10)
The potential evapotranspiration for each crop (Equation
(10)) and the rainfall in each period, tdownscaled from
CCA downscaling, are used to compute future projections
of irrigation demands for each crop in each period, t. The
irrigated area for different crops under left and right bank
canal commands (Table IV) and duration of the crops with
their sowing dates (Table V) are used in the computation of
irrigation demands. The crop factors used for paddy,
sugarcane, permanent garden and semidry crops corre-
sponds to Rice, Sugarcane, Group E crops (Citrus) and
Maize, respectively, from Michael (1978) as given in
Table VI. The total irrigation requirement (including
left bank and right bank canal) at the field level for each
crop in each month is estimated as per the cropping pattern
in Table V.
RESULTS AND DISCUSSION
Impact of climate change on rainfall and
reference evapotranspiration
Simulated rainfall refers to the rainfall obtained from the
NCEP data and the predicted rainfall results from use of
CCA downscaling model with MIROC 3.2 GCM for the
A1B scenario. The CCA model is able to well simulate
the observed data (Figure 2(a) for Locations 1 to 9) for
the training period of 1971 to 1995 with both NCEP and
GCM. The GCM predicted rainfall as shown in Figure 2 (a)
for Locations 1 to 9 for the training period of 1971–1995
are modeled with the monthly predictors in the MIROC
3.2 GCM for the current climate with 20c3m experiment.
All future projections are for the A1B scenario for
25 years time slices of 2020–2044, 2045–2069 and
2070–2095 (Figure 2 (b) for Locations 1 to 9). The green
box plots are for the period of 2020 to 2044, the blue box
plots are for the period of 2045 to 2069 and the red box
plots are for the period of 2070 to 2095. The projected
monthly rainfall shows an increasing trend in all months
at all nine downscaling locations. The expected rainfall
increase is determined by the change in the large-scale
atmospheric variables (air temperature, mean sea level
pressure, geopotential height, humidity and wind
variables) considered as predictors (Table II) in the study
region. Such an increase in rainfall is also observed in
the study of Meenu et al. (2011) for the same case
study of Bhadra command area with SDSM and also
with support vector machine.
Table V. Crop duration and sowing dates
Crop Duration (days) Sowing date
Paddy 120 June 15
Sugarcane 365 July 01
Permanent Garden 365 June 01
Semidry Crops 123 July 01
Table IV. Crop distribution in the command area
Canal
Paddy
(ha)
Sugarcane
(ha)
Permanent
garden (ha)
Semidry
Crops (ha)
Total area
(ha)
LBC 3484 1713 303 867 6367
RBC 34 720 24 800 18 849 20 834 99 203
Total 38 204 26513 19 152 21 701 105 570
RBC: Right Bank Canal; LBC: Left Bank Canal
Table III. Comparison of observed versus computed statistics (Testing period, 1996 to 2004)
Statistic
Rainfall (mm) Downscaling Locations Maximum
Temperature
(C)
Minimum
Temperature
(C)
Relative
Humidity
%
Wind
Speed
kmph123456789
Observed Mean 174.93 59.10 130.75 73.33 75.18 55.3 44.97 40.16 42.22 31.25 19.44 70.78 3.73
Computed Mean 171.96 55.09 79.28 69.41 75.41 53.05 38.38 38.99 31.68 31.48 19.57 69.95 3.74
Observed
Standard Deviation
230.63 68.50 306.22 87.34 86.92 64.60 51.25 55.72 52.95 2.77 2.32 10.03 1.26
Computed
Standard Deviation
181.92 47.18 189.88 65.71 62.06 43.76 36.73 38.51 36.99 2.40 1.82 7.72 1.17
R-Value 0.87 0.74 0.58 0.84 0.82 0.73 0.78 0.72 0.77 0.93 0.89 0.88 0.96
The relative humidity, wind speed, maximum and minimum temperatures in the table are at station Shimoga.
S. REHANA AND P. P. MUJUMDAR
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
Figure 3 shows similar results of other meteorological
variables, RH, U
2
, Tmax and Tmin. All the meteoro-
logical variables are well simulated by CCA downscaling
(Figure 3 (a)) for the training period of 1971 to 1995. The
projections of Tmax and Tmin and RH also show an
increasing trend for all the months. The U
2
projections do
not show any particular trend.
The reference evapotranspiration estimated from the
projections of Tmax and Tmin, RH and U
2
using the
evapotranspiration model (Equation (8)) is shown in
Table VI. Monthly crop coefficients (Source: Michael, 1978)
Crop
Months
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Paddy (Rice) 0.85 1.00 1.15 1.30 1.25 1.10 0.90
Sugarcane 0.75 0.80 0.85 0.85 0.90 0.95 1.00 1.00 0.95 0.90 0.85 0.75
Permanent Garden (Citrus) 0.50 0.55 0.55 0.60 0.60 0.65 0.70 0.70 0.65 0.60 0.60 0.55
Semidry crops (Maize) 0.85 1.00 1.15 1.30 1.25 1.10 0.90
1971 1975 1979 1983 1987 1991 1995
0
500
1000
Location 1
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
200
400
600
Monthly Rainfall (mm)
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
1971 1975 1979 1983 1987 1991 1995
0
200
Location 2
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
50
100
150
200
Monthly Rainfall (mm)
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
(a)
(b)
(a)
(b)
Figure 2. Downscaling results of rainfall from the CCA model from Locations 1 to 9. In above figures, (a) shows the observed, simulated from NCEP
data and predicted from MIROC 3.2 GCM with 20c3m experiment for the training period of 1971 to 1995, (b) represents the future projections from
MIROC 3.2 GCM with A1B scenario for each month with green box plots for period 2020–2044, blue box plots are for period 2045–2069 and the red
box plots are for period 2070–2095
CLIMATE CHANGE AND IRRIGATION DEMANDS INTEGRATION
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
Figure 4. The observed evapotranspiration for each month
shown in the Figure 4 is computed from the evapotrans-
piration model (Equation (8)) with observed meteoro-
logical data for the period 1971 to 1995. The future
projections of reference evapotranspiration predicted to
increase for all months. Particularly, the change of
evapotranspiration is more in the months of April and
May due to the large projected changes of Tmax and
Tmin variables.
Impact of climate change on irrigation water demands
The irrigation water requirements are computed for
paddy, sugarcane, permanent garden and semidry crops at
Locations 1 to 9. The monthly reference evapotranspir-
ation is corrected with crop coefficients for each crop to
compute the potential evapotranspiration which in turn
can be used to compute the irrigation water demand of the
crop. The monthly irrigation water demands are estimated
from the projections of rainfall at each of the location
downscaled from CCA model and potential evapotrans-
piration projections from Equation (10). The monthly
projected variation of irrigation water requirements for
Locations 1 to 9 are shown in Figures 5–7 and 8,
respectively, for paddy, sugarcane, permanent garden and
semidry crops. The annual irrigarion demands for the
crops at the nine locations are shown in Figure 9. The
predicted change of irrigation water demands at each
location is a function of rainfall at that location and the
reference evapotranspiration.
Irrigation water requirement - paddy
The crop growing period of paddy spans from April to
October. The irrigation demands of paddy are computed
for these months as shown in Figure 5. However, at
Locations 1 and 3, paddy demands are only in the months
of April and May, while for the other months, the rainfall
is sufficient to fulfill the water requirements of paddy. The
months showing the demands as zero indicates the water
needed for optimal growth of the crop is provided by
rainfall and irrigation is not required in those particular
months. For remaining locations, the demands are present
for all the months starting from April to September except
in the month of October (Figure 5). At Locations 7, 8 and
9 in September month, where the current demands are
zero, significant increase in the projected irrigation water
requirements are observed due to the increase in the
1971 1975 1979 1983 1987 1991 1995
0
1000
2000
Location 3
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
500
1000
Monthly Rainfall (mm)Monthly Rainfall (mm)
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
1971 1975 1979 1983 1987 1991 1995
0
200
400
Location 4
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
100
200
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
(a)
(b)
(a)
(b)
S. REHANA AND P. P. MUJUMDAR
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
evapotranspiration demand of crops. For example, the
monthly mean rainfall of May is increasing at Location 7
from 28.75 mm to 32.64 mm for the period of 2020–2044,
to 38.37 mm for the period of 2045–2069 and to 42.61 mm
for the period of 2070–2095. At the same time, the increase
in Tmax and Tmin are also increasing. For example,
monthly Tmax temperature for May is increasing from
observed 33.64 Cto36.26C for 2020–2044, to 37.51
C for 2045–2069, to 38.31 C for 2070–2095. Similarly,
monthly minimum temperature of May is also increasing
with observed 21.49 Cto21.46C for 2020–2044, to
22.33 C for 2045–2069 and 23.01 C for 2070–2095. A
significant increase in RH from observed 67.02% to
70.97% for 2020–2044, to 71.73% for 2045–2069, to
72.40% for 2070–2095 is also seen from the results. The
minor changes in U
2
are from observed 4.089 m/s to 4.25
m/s for 2020–2044, to 4.26 m/s for 2045–2069 and to 4.38
m/s for 2070–2095. Such increase in RH, U
2
, temperature
variables results in net increase in evapotranspiration, for
example, at Location 7 in the month of May. That is, the
increase in evapotranspiration offsets the increasing effect
of rainfall at Location 7 indicating increased irrigation
demand in future for paddy (Figure 5). However, at some
locations, paddy demands are predicted to decrease at
monthly scale, e.g. at Location 2 in August month
(Figure 5) due to the relative increase in rainfall compared
to the evapotranspiration at that location. Overall irrigation
requirements of paddy are predicted to increase at all nine
locations at monthly scale (Figure 5) and at annual scale
(Figure 9). The maximum annual paddy demand is
predicted to occur at Location 8 (Figure 1) with current
demand as 14.00 Mm
3
with increasing demands as 26.97
Mm
3
for the period of 2020–2044, with 27.35 Mm
3
for the
period of 2045–2069, with 27.8 Mm
3
for the period of
2070–2095.
Irrigation water requirement - sugarcane
Sugarcane crop is growing in all 365 days of a year,
and the crop water demand exists in all 12 months.
Sugarcane demands are more in the months of April and
May for all nine locations (Figure 6) due to lower rainfall
and higher temperatures in these months. For the month
of January, the demand is predicted to decrease at
Locations 1, 2, 4, 5 and 6 compared to the current
demands depending on the projections of rainfall and
1971 1975 1979 1983 1987 1991 1995
0
200
400
Location 5
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
50
100
150
200
Monthly Rainfall (mm)Monthly Rainfall (mm)
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
1971 1975 1979 1983 1987 1991 1995
0
200
Location 6
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
50
100
150
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
(a)
(b)
(a)
(b)
CLIMATE CHANGE AND IRRIGATION DEMANDS INTEGRATION
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
evapotranspiration. Even though small reductions of
demands in the monthly scale are observed, the annual
irrigation water demands are predicted to increase for
sugarcane over the Bhadra command area (Figure 9). The
maximum annual irrigation demands occur at Location 8
(Figure 1) with current demand being 15.29 Mm
3
and
projected demands of 23.12 Mm
3
for 2020–2044, 23.16
Mm
3
for 2045–2069, 23.5 Mm
3
for 2070–2095.
1971 1975 1979 1983 1987 1991 1995
0
100
200
Location 7
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
50
100
150
Monthly Rainfall (mm)Monthly Rainfall (mm)
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
1971 1975 1979 1983 1987 1991 1995
0
100
200 Location 8
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
50
100
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
(a)
(b)
(a)
(b)
1971 1975 1979 1983 1987 1991 1995
0
200
Location 9
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
50
100
150
Monthly Rainfall (mm)
Observed NCEP Simulated Predicted from Miroc 3.2 GCM (20c3m)
(a)
(b)
S. REHANA AND P. P. MUJUMDAR
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
Observed NCEP GCM
65
70
75
80
Relative Humidity (%)
(a)
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
62
64
66
68
70
72
74
76
78
80
(b)
(i)
Observed NCEP GCM
3.4
3.6
3.8
4
4.2
Wind Speed (kmph)
(a)
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2
3
4
5
6
(b)
(ii)
Observed NCEP GCM
29.5
30
30.5
31
31.5
32
32.5
Maximum Temperature (Deg C)
(a)
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
29
30
31
32
33
34
35
36
37
38
39
40
(b)
(iii)
Observed NCEP GCM
18
18.5
19
19.5
20
20.5
Minimum Temperature (Deg C)
(a)
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
17
18
19
20
21
22
23
24
(b)
(iv)
Figure 3. Downscaling results of (i) relative humidity, (ii) wind speed, (iii) maximum temperature and (iv) minimum temperature from the CCA model.
In above figures, (a) denote annual scale observed, simulated from NCEP and simulated from MIROC 3.2 GCM with 20c3m experiment for the training
period of 1971 to 1995. (b) denotes monthly scale projections with the green box plots are for 2020–2044, blue box plots are for 2045–2065 and red box
plots are for 2070–2095
CLIMATE CHANGE AND IRRIGATION DEMANDS INTEGRATION
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
Irrigation water requirement - permanent garden
The crop water requirement of permanent garden spans
for the entire year, and the irrigation demands are
estimated for the all 12 months. The annual demands of
permanent garden are predicted to increase (Figure 9)
even though the decreases in demands are small for the
monthly scale (Figure 7). The maximum annual demand
occur at Location 3 with 2.89 Mm
3
of current demand
increasing to 6.95 Mm
3
for period of 2020–2044, 8.79
Mm
3
for a period of 2045–2069, 10.26 Mm
3
for a period
of 2070–2095.
Irrigation water requirement - semidry crops
The growing period for semidry crops spans from April
to October and the demands for the corresponding months
are quantified as shown in Figure 8. Water requirements
for the semidry crops are predicted to increase at monthly
scale (Figure 8) as well as at annual scale (Figure 9). At
most of the locations, the estimated current irrigation
demands are zero, but the projected demands are
increasing. The maximum increase in annual demand
occurs at Location 7 with current demand being 2.64
Mm
3
and increasing to 15.26 Mm
3
for the period of
2020–2044, 17.12 Mm
3
for the period of 2045–2069,
19.68 Mm
3
for the period of 2070–2095. Annual
irrigation demands are less for semidry crops compared
to the other crops as the command area is small and also
the crop growing period is small, being restricted to the
months of April to October only.
Due to their cropping pattern and the command area,
water requirements of Paddy and Sugarcane are higher
compared to those of permanent garden and semi dry
crops. For all crops at all nine locations, the projected
irrigation demands are higher compared to the current
demands. Even though the projected demands are higher
compared to observed ones, the relative difference in the
future demands for the periods of 2020–2044, 2045–2069
and 2070–2095 are small, due to the projected increase
in the rainfall in the Bhadra command area. The annual
irrigation demand assessment carried out in this
study will give an overall idea about the changes in
demands for each particular crop at each downscaling
location. Moreover, the monthly analysis of demands
for each crop at a particular location will be useful
for the decision makers for better management of
irrigation systems.
Figure 4. Monthly reference evapotranspiration for Bhadra Command
area estimated from MIROC 3.2 GCM output with A1B scenario
A M J J A S O
0
2
4
6
8
Paddy Irrigation Water
Requirement (Mm3)
Location 1
A M J J A S O
0
2
4
6
8
Location 2
A M J J A S O
0
2
4
6
8
Location 3
A M J J A S O
0
2
4
6
8
Location 4
A M J J A S O
0
2
4
6
8
Location 5
A M J J A S O
0
2
4
6
8
Location 6
A M J J A S O
0
2
4
6
8
Location 7
A M J J A S O
0
2
4
6
8
Location 8
A M J J A S O
0
2
4
6
8
Location 9
Present 2020-2044 2045-2069 2070-2095
Figure 5. Monthly (April to October) irrigation water requirement for paddy at Locations 1–9 for Bhadra Command Area
S. REHANA AND P. P. MUJUMDAR
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
CONCLUSIONS
A methodology is developed in the present study for
predicting the future irrigation water demands in the
command area of a river. The expected changes of rainfall,
RH, U
2
, Tmax and Tmin are modeled by using a SDSM,
CCA, with MIROC 3.2 GCM output for the A1B scenario.
The potential evapotranspiration projections are modeled
with an evapotranspiration model (Penman–Monteith
equation) accounting for the projected changes in
temperature, RH, solar radiation and U
2
. The need to
calculate the evapotranspiration using the temperature
J F M A M J J A S O N D
0
2
4
6
Sugarcane Irrigation
Water Requirement (Mm3)
Location 1
J F M A M J J A S O N D
0
2
4
6
Location 2
J F M A M J J A S O N D
0
2
4
6
Location 3
J F M A M J J A S O N D
0
2
4
6
Location 4
J F M A M J J A S O N D
0
2
4
6
Location 5
J F M A M J J A S O N D
0
2
4
6
Location 6
J F M A M J J A S O N D
0
2
4
6
Location 7
J F M A M J J A S O N D
0
2
4
6
Location 8
J F M A M J J A S O N D
0
2
4
6
Location 9
Present 2020-2044 2045-2069 2070-2095
Figure 6. Monthly irrigation water requirement for sugarcane at Locations 1–9 for Bhadra Command Area
J F M A M J J A S O N D
0
1
2
3
Permanent Garden Irrigation
Water Requirement (Mm3)
Location 1
J F M A M J J A S O N D
0
1
2
3
Location 2
J F M A M J J A S O N D
0
1
2
3
Location 3
J F M A M J J A S O N D
0
1
2
3
Location 4
J F M A M J J A S O N D
0
1
2
3
Location 5
J F M A M J J A S O N D
0
1
2
3
Location 6
J F M A M J J A S O N D
0
1
2
3
Location 7
J F M A M J J A S O N D
0
1
2
3
Location 8
J F M A M J J A S O N D
0
1
2
3
Location 9
Present 2020-2044 2045-2069 2070-2095
Figure 7. Monthly irrigation water requirement for permanent garden at Locations 1–9 for Bhadra Command Area
CLIMATE CHANGE AND IRRIGATION DEMANDS INTEGRATION
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
variables, humidity, solar radiation and U
2
rather than only
temperature variables has therefore been stressed. The
irrigation water requirements are quantified by accounting
for projected rainfall and potential evapotranspiration. The
monthly irrigation water demands of paddy, sugarcane,
permanent garden and semidry crops are quantified at nine
downscaling locations covering the entire command area of
Bhadra river basin. The annual irrigation water requirements
for paddy, sugarcane, permanent garden and semidry crops
are predicted to increase in the Bhadra command area. The
projected changes in irrigation demands will be helpful in
developing adaptive policies for reservoir operations.
A M J J A S O
0
1.5
2.5
Semidry Irrigation
Water Requirement (Mm3)
Location 1
A M J J A S O
0
1.5
2.5
Location 2
A M J J A S O
0
1.5
2.5
Location 3
A M J J A S O
0
1.5
2.5
Location 4
A M J J A S O
0
1.5
2.5
Location 5
A M J J A S O
0
1.5
2.5
Location 6
A M J J A S O
0
1.5
2.5
Location 7
A M J J A S O
0
1.5
2.5
Location 8
A M J J A S O
0
1.5
2.5
Location 9
Present 2020-2044 2045-2069 2070-2095
Figure 8. Monthly semidry irrigation water requirements for Locations 1–9 for Bhadra Command Area
1 2 3 4 5 6 7 8 9
0
10
20
30
Location
Irrigation Water
Requirement (Mm3)
Irrigation Water
Requirement (Mm3)
Irrigation Water
Requirement (Mm3)
Irrigation Water
Requirement (Mm3)
Paddy
1 2 3 4 5 6 7 8 9
0
5
10
15
20
25
30
Location
Sugarcane
1 2 3 4 5 6 7 8 9
0
5
10
15
20
25
30
Location
Permanent Garden
1 2 3 4 5 6 7 8 9
0
5
10
15
20
25
30
Location
Semidry Crops
Present 2020-2044 2045-2069 2070-2095
Figure 9. Projected annual irrigation water requirements at each location for each crop for Bhadra Command Area
S. REHANA AND P. P. MUJUMDAR
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
In this study, the soil moisture contribution to meeting
crop water demand is neglected. However, for an accurate
representation of the crop water demands, the soil
moisture dynamics of individual crops must be consid-
ered in the impact assessment studies. Further, the rainfall
amount considered in the estimation of irrigation
demands is the actual rainfall instead of effective rainfall.
The effective rainfall is the fraction of actual amount of
rainwater useful for meeting the water need of the crops.
The effective rainfall calculation includes the soil water
retention and percolation, the key aspects which should
be included in further studies in order to develop more
useful projected demands accounting for climate change.
Further, the future projected irrigation demands are due
to a single GCM using a single scenario. It is widely
acknowledged that the mismatch between different GCMs
over regional climate change projections represents a
significant source of uncertainty (e.g. New and Hulme,
2000; Simonovic and Li, 2003; Simonovic and Davies,
2006; Wilby and Harris, 2006; Ghosh and Mujumdar,
2007). Further studies are necessary to evaluate the future
irrigation demands for different GCMs with scenarios to
model the underlying GCM and scenario uncertainty. The
results will serve as guidelines for the decision makers to
accommodate sufficient water in those months where
rainfall only will not be sufficient to fulfill the crop water
requirements. Further, the results will be useful in
examining different cropping patterns in the command
area keeping in view the increased crop water demands
and possible decrease in streamflow.
REFERENCES
Allen RG, Pereira LS, Raes D, Smith M. 1998. Crop Evapotranspiration
Guidelines for Computing Crop Water Requirements. FAO Irrigation
and Drainage Paper 56, ISBN 92-5-104219-5, Food and Agriculture
Organization of the United Nations, Rome.
Anandhi A, Srinivas VV, Kumar DN, Nanjundiah RS. 2009. Role of
predictors in downscaling surface temperature to river basin in India for
IPCC SRES scenarios using support vector machine. International
Journal of Climatology 29(4): 583–603.
Angstrom A. 1924. Solar and terrestrial radiation. Quarterly Journal of the
Royal Meteorological Society 50: 121–126.
Barnett TP, Preisendorfer RW. 1987. Origins and levels of monthly and
seasonal forecast skill for United States air temperature determined
by canonical correlation analysis. Monthly Weather Review 115:
1825–1850.
Barnston AG. 1994. Linear statistical short-term climate predictive skill in
the Northern Hemisphere. Journal of Climate 7: 1513–1564.
Brown RA, Rosenberg NJ. 1999. Climate change impacts on the potential
productivity of corn and winter wheat in their primary United States
growing regions. Climate Change 41:73–107.
Busuioc A, Von Storch H. 1996. Changes in the winter precipitation in
Romania and its relation to the large-scale circulation. Tellus, Series A:
Dynamic Meteorology and Oceanography 48(4): 538–552.
Davy RJ, Woods MJ, Russell CJ, Coppin PA. 2010. Statistical
Downscaling of Wind Variability from Meteorological Fields.
Boundary-Layer Meteorology. DOI: 10.1007/s10546-009-9462-7.
De Silva CS, Weatherhead EK, Knox JW, Rodriguez-Diaz JA. 2007.
Predicting the impacts of climate change—a case study on paddy
irrigation water requirements in Sri Lanka. Agricultural Water
Management 93(1–2): 19–29.
Easterling WE, Crosson PR, Rosenberg NJ, McKenney MS, Katz LA,
Lemon KM. 1993. Agricultural impacts of and response to climate
change in the Missouri-Iowa-Nebraska- Kansas (MINK) region.
Climate Change 24:23–61.
Elgaali E, Garcia LA, Ojima DS. 2007. High resolution modeling of the
regional impacts of climate change on irrigation water demand. Climate
Change 84: 441–461.
Ghosh S, Mujumdar PP. 2006. Future Rainfall Scenario over Orissa with
GCM Projections by Statistical Downscaling. Current Science 90(3):
396–404.
Ghosh S, Mujumdar PP. 2007. Nonparametric methods for modeling
GCM and scenario uncertainty in drought assessment. Water Resources
Research 43: W07405. DOI: 10.1029/2006WR005351.
Graham NE, Michaelsen J, Barnett TP. 1987. An investigation of the El
Nino-Southern Oscillation cycle with statistical models. 1. Predictor field
characteristics. Journal of Geophysical Research 92:14251–14 270.
Gyalistras D, von Storch H, Fischlin A, Beniston M. 1994. Linking GCM-
simulated climatic changes to ecosystem models: case studies of
statistical downscaling in the Alps. Climate Research 4(3): 167–189.
Hargreaves GH. 1994. Simplified coefficients for estimating monthly solar
radiation in North America and Europe. Dept. Paper, Dept. Biol. and
Img. Engrg. Utah State Univ.: Logan, Utah.
Hargreaves GH, Samani ZA. 1982. Estimating potential evapotranspiration.
Journal of Irrigation Drainage Engineering, ASCE 108(3): 25–230.
Harmsen EW, Miller NL, Schlegel NJ, Gonzalez JE. 2009. Seasonal climate
change impacts on evapotranspiration, precipitation deficit and crop yield
in Puerto Rico. Agricultural Water Management 96:1085–1095.
Huth R. 2005. Downscaling humidity variables. International Journal of
Climatology 25: 243–250.
Intergovernmental Panel on Climate Change (IPCC). 2007. Climate Change.
The Physical Science Basis—Contribution of Working Group I to the
Fourth Assessment Report of the Intergovernmental Panel on Climate
Change,SolomonSet al. (ed). Cambridge Univ. Press: New York; 2007.
Juneng L,Tangang FT. 2008. Level and source of predictability of
seasonal rainfall anomalies in Malaysia using canonical correlation
analysis. International Journal of Climatology 28: 1255–1267.
Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin L,
Iredell M, Saha S, White G, Woollen J, Zhu Y, Leetmaa A, Reynolds R,
Chelliah M, Ebisuzaki W, Higgins W, Janowiak J, Mo KC, Ropelewski C,
Wang J, Jenne R, Joseph D. 1996. The NCEP/NCAR 40-year reanalysis
project. Bulletin of the American Meteorological Society 77(3): 437–471.
Karl TR, Wang WC, Schlesinger ME, Knight RW, Portman D. 1990. A
method of relating general circulation model simulation climate to
the observed local climate. Part I: Seasonal Statistics. Journal of
Climate 3: 1053–1079.
Kittel TGF, Rosenbloom NA, Painter TH, Schimel DS. 1995. VEMAP
Modeling Participants, The VEMAP integrated database for modeling
United States ecosystem/vegetation sensitivity to climate change.
Journal of Biogeography 22: 857–862.
Liu S, Mo X, Lin Z, Xu Y, Ji J, Wen G, Richey J. 2010. Crop yield
response to climate change in the Huang-Huai-Hai plain of China.
Agricultural Water Management 97(8): 1195–1209.
Lovelli S, Perniola M, Di Tommaso T, Ventrella D, Moriondo M,
Amato M. 2010. Effects of raising atmospheric CO
2
on crop
evapotranspiration in a Mediterranean area. Agricultural Water
Management 97(9): 1287–1292.
Maeda EE, Pellikka PKE, Clark BJF, Siljander M. 2011. Prospective
changes in irrigation water requirements caused by agricultural
expansion and climate changes in the eastern arc mountains of Kenya.
Journal of Environmental Management 92(3): 982–993.
MATLAB. 2004. Statistics Toolbox. The Math Works Inc. http://www.
mathworks.in
Meenu R, Rehana S, Mujumdar PP. 2011. Assessment of hydrologic
impacts of climate change in Tunga-Bhadra basin, India with HEC-
HMS and SDSM. Hydrological Processes, Accepted Article. DOI:
10.1002/hyp.9220.
Michael AM. 1978. Irrigation Theory and Practice. Vikas Publishing
House Pvt Ltd: New Delhi.
Mpelasoka FS, Mullan AB, Heerdegen RG. 2001. New Zealand climate
change information derived by multivariate statistical and artificial
neural network approaches. International Journal of Climatology
21: 1415–1433.
Mulcahy KA, Clarke KC. 1995. What shape are we in? The display of
map projection distortion for global change research. In Proceedings of
GIS/LIS ’95. American Society of Photogrammetry and Remote
Sensing: Bethesda, Md; 175–181.
New M, Hulme M. 2000. Representing uncertainty in climate change
scenarios: A Monte Carlo approach. Integrated Assessment 1: 203–213.
Parry ML, Rosenzweig C, Iglesias A, Livermore M, Fischer G. 2004.
Effects of climate change on global food production under SRES
emissions and socio-economic scenarios. Global Environmental
Change 14(1): 53–67.
CLIMATE CHANGE AND IRRIGATION DEMANDS INTEGRATION
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
Raje D, Mujumdar PP. 2009. A conditional random field based
downscaling method for assessment of climate change impact on
multisite daily precipitation in the Mahanadi basin. Water Resources
Research 45(10): W10404. DOI: 10.1029/2008WR007487.
Rodriguez Diaz JA, Weatherhead EK, Knox JW, Camacho E. 2007.
Climate change impacts on irrigation water requirements in the
Guadalquivir river basin in Spain. Regional Environmental Change 7:
149–159.
Rosenzweig C, Parry ML. 1994. Potential impact of climate change on
world food supply. Nature 367: 133–138.
Shahid S. 2011. Impact of climate change on irrigation water demand
of dry season Boro rice in northwest Bangladesh. Climate Change
105: 433–453.
Simonovic SP, Davies EGR. 2006. Are we modeling impacts of climate
change properly? Hydrological Processes 20: 431–433.
Simonovic SP, Li L. 2003. Methodology for assessment of climate change
impacts on large-scale flood protection system. Journal of Water
Resources Planning and Management 129(5): 361–371.
Singh B, Maayar ME, André P, Bryant CR, Thouez JP. 1998. Impacts of a
GHG-induced climate change on crop yields: Effects of acceleration in
maturation, moisture stress, and optimal temperature. Climate Change
38:51–86.
Torres AF, Walker WR, McKee M. 2011. Forecasting daily potential
evapotranspiration using machine learning and limited climatic data.
Agricultural Water Management 98(4): 553–562.
Von Storch H, Zorita E, Cubasch U. 1993. Downscaling of global climate
change estimates to regional scales: an application to Iberian rainfall in
wintertime. Journal of Climate 6: 1161–1171.
Wilby RL, Harris IA. 2006. A framework for assessing uncertainties in
climate change impacts: low-flow scenarios for the river Thames, UK.
Water Resources Research 42: W02419. DOI: 10.1029/2005WR004065.
Wilby RL, Charles SP, Zorita E, Timbal B, Whetton P, Mearns LO. 2004.
The guidelines for use of climate scenarios developed from statistical
downscaling methods. Supporting material of the Intergovernmental
Panel on Climate Change (IPCC), prepared on behalf of Task Group on
Data and Scenario Support for Impacts and Climate Analysis (TGICA)
(<http://ipccddc.cru.uea.ac.uk/guidelines/StatDown Guide.pdf>).
Yano T, Aydin M, Haraguchi T. 2007. Impact of climate change on
irrigation demand and crop growth in a Mediterranean environment of
Turkey. Sensors 7: 2297–2315.
S. REHANA AND P. P. MUJUMDAR
Copyright © 2012 John Wiley & Sons, Ltd. Hydrol. Process. (2012)
DOI: 10.1002/hyp
