See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/265964253

Integrated Land and Water Resources Management Framework for Hirakud

Canal Subcommand (India) Using Gray Systems Analysis

Article  in  Journal of Water Resources Planning and Management · November 2013

DOI: 10.1061/(ASCE)WR.1943-5452.0000300

CITATIONS READS

4 910

3 authors:

Uday Mandal Anirban Dhar

ICAR-Indian Institute of Soil & Water Conservation Dehadun Indian Institute of Technology Kharagpur
17 PUBLICATIONS   79 CITATIONS    71 PUBLICATIONS   1,023 CITATIONS   

SEE PROFILE SEE PROFILE

Sudhindra N. Panda
Indian Institute of Technology Kharagpur
103 PUBLICATIONS   2,206 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Unified Physically based Regional Hydraulic and Hydrologic Modeling Framework View project

Academic excellence View project

All content following this page was uploaded by Anirban Dhar on 23 September 2014.

The user has requested enhancement of the downloaded file.

 Technical Note

Integrated Land and Water Resources Management

Framework for Hirakud Canal Subcommand (India)
Using Gray Systems Analysis
Downloaded from ascelibrary.org by The Indian Inst of Technology Kharagpur Librarian on 10/23/13. Copyright ASCE. For personal use only; all rights reserved.

Uday Mandal 1; Anirban Dhar 2; and Sudhindra Nath Panda, M.ASCE 3

Abstract: An integrated land and water management model incorporating inexact framework is developed for sectorwise and seasonwise
planning in the canal subcommand area. The gray systems theory is utilized for defining system parameters and decision variables within the
inexact framework. The model considers the objective of maximizing the net annual return subject to optimal allocation of land and water
resources for given imprecise information on net return of crop, available resources (land, surface water, and groundwater), and water demand
in the cultivable subcommand area. Gray linear programming (GLP) is utilized for solving the optimization problem. The developed model
is applied to the Hirakud canal subcommand area of western Odisha, India. The model is evaluated for different scenarios with inexact
bounds on available natural resources (land and water) for different crops. The model results show its usefulness in evolving crop-specific
land utilization levels and quantities of surface water and groundwater resources for deriving maximum benefit. The model also provides
season-specific crop-scheduling information. The evaluation results show potential applicability of the developed methodology. DOI: 10
.1061/(ASCE)WR.1943-5452.0000300. © 2013 American Society of Civil Engineers.
CE Database subject headings: Water resources; Optimization; Computer programming; Groundwater; Surface water; Water
management.
Author keywords: Optimization; Gray linear programming; Groundwater and surface water; Integrated land and water management.

Introduction considering the response of the systems is meaningless. Therefore,

Land and water are the two most important natural resources. These
are the key factors for sustainable agricultural development of a
nation. With the increasing trend of the population in India, avail-
ability of fresh natural resources is under threat. This necessitates
the need for development of an integrated land and water resources
management framework. In a dynamic system, this becomes a
challenging task for planners. Both the system parameters (e.g., net
irrigation requirement, irrigation water cost, and total cultivable
command area) and the decision variables (e.g., cropping pattern,
surface water, and groundwater) are susceptible to deviations (Dhar
and Datta 2009). An inexact framework-based integrated land and
water management model incorporating possible variation(s) in
parameters and decision variables is developed for resources allo-
cation of the command area.
Optimal irrigation planning is a prerequisite for policy decision(s).
Prescription of optimal strategy requires formulation and its solu-
tion. Optimization models have the capability of systematically se-
lecting optimal solutions. However, any decision making without
1
M.Tech. Student, Dept. of Agricultural and Food Engineering, Indian
Institute of Technology, Kharagpur 721302, India.
2 Most of these studies have incorporated an integrated land and water
Assistant Professor, Dept. of Civil Engineering, Indian Institute of
Technology, Kharagpur 721302, India (corresponding author). E-mail:
anirban.dhar@gmail.com; anirban@civil.iitkgp.ernet.in
3
Professor, Dept. of Agricultural and Food Engineering, Indian Institute
of Technology, Kharagpur 721302, India.
Note. This manuscript was submitted on August 15, 2011; approved on
July 3, 2012; published online on September 14, 2012. Discussion period
open until April 1, 2014; separate discussions must be submitted for indi-
vidual papers. This technical note is part of the Journal of Water Re-
sources Planning and Management, Vol. 139, No. 6, November 1,
2013. © ASCE, ISSN 0733-9496/2013/6-733-740/$25.00.
the need arises to simulate the physical processes occurring in the
system. To evolve an effective management strategy for macro-
scale region, a simplified water balance model can suffice the pur-
pose. Moreover, it becomes necessary to include imprecise system
parameter information within the model formulation.
Different irrigation water management studies have incorporated
components that can identify the optimal cropping patterns (Panda
et al. 1996; Jiracheewee et al. 1996; Mainuddin et al. 1997; Singh
et al. 2001; Nirala et al. 2005; Khare et al. 2007; Rejani et al. 2009),
determine the quantity of water to be delivered from different sour-
ces (Jiracheewee et al. 1996; Mainuddin et al. 1997; Kumar et al.
2006; Khare et al. 2007), and maximize the net annual return from
the agricultural production (Paul et al. 2000; Singh et al. 2001; Sethi
et al. 2006; Khare et al. 2007). In irrigation water management gen-
erally linear programming (Jiracheewee et al. 1996; Mainuddin et al.
1997; Singh et al. 2001; Nirala et al. 2005; Kumar et al. 2006), dy-
namic programming (Paul et al. 2000; Hamid and Mohammad
2011), goal programming (Latinopoulos and Mylopoulos 2005;
Vivekanandan et al. 2009), genetic algorithm (Nirala et al. 2005;
Kumar et al. 2006), chance constrained linear programming, and
deterministic linear programming (Sethi et al. 2006; Karamouz
and Zahraie 2010) are used depending on the nature of formulation.
management framework within the optimization formulation. How-
ever, only a few studies have incorporated an imprecise information
(both parameters and decision variables)-based formulation in irri-
gation planning (Lu et al. 2009; Li et al. 2010; Zhang and Huang
2011). Earlier studies using an inexact framework include water
quality management within an agricultural system (Huang 1998)
and a conjunctive water allocation problem (Lu et al. 2009).
Imprecise information can be handled using gray systems.
In-depth review of gray-number optimization can be found in

JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT © ASCE / NOVEMBER/DECEMBER 2013 / 733

J. Water Resour. Plann. Manage. 2013.139:733-740.

 Rosenberg (2009). Unlike fuzzy mathematics or a probabilistic ap- GW
ij = groundwater allocated in season j for sector i; Cij
GW

=
proach, a gray system relies on the interval numbers having equal unit cost of groundwater for all i and j; Cij = unit cost of surface
SW

weight for the whole interval. Fuzzy numbers are represented using
membership functions, whereas a probabilistic approach depends
on probability density function. These functions define preference
distribution for interval-valued numbers. Choice of membership
functions are subjective decisions. Thus, gray numbers are more
suitable for handling unbiased systems analysis problems. Gray
systems theory is proposed by Deng (1982). Gray system is in be-
Downloaded from ascelibrary.org by The Indian Inst of Technology Kharagpur Librarian on 10/23/13. Copyright ASCE. For personal use only; all rights reserved.
water for all i and j; x
= gray number (i.e., value within the
bounds [x− , xþ ]), where x is a gray number dummy variable.
The first term in Eq. (1) considers net return from different crops
for different sector/seasons. Moreover, the second term accounts
for the total cost incurred due to surface water and groundwater use.
Constraints

tween white (system with completely known information) and

black (system with completely unknown information) systems. Relevant constraints are discussed subsequently.
Thus, it has partially known and partially unknown characteristics. Water Allocation Constraints
“Gray number” is expressed as a known range (without any distri- Net irrigation requirement calculation for individual crop in different
butional information, generally uniform distribution). Interestingly, season is an essential component for the water allocation problem.
a deterministic number can be viewed as a gray number with same Depending on crop water demand, net irrigation requirement is to be
upper and lower value. satisfied by conjunctive use of surface and groundwater. This also
Parameters and control variables in an optimization model can depends on the available water resources within the system:
be represented as gray numbers to consider imprecise information.
ncrop
Thus, a physically meaningful integrated land and water manage- X
nsect X X
nsect

ment framework has been developed by using gray linear program- NIR


ijk Aijk − α




1 ðβ 1 SWij þ GWij Þ ≤ 0; ∀ j ð2Þ
ming (GLP). The model considers the objective of maximizing net i¼1 k¼1 i¼1

annual return subject to optimal allocation of land and water resour- where NIR
ijk = net irrigation requirement of crop k in season j
ces for given inexact information on net return of crop, available for sector i; α
1 = field water application efficiency, 60–70% ¼
land and water resources, and water demand in the cultivable com- ½0.60; 0.70; β

1 = conveyance efficiency, 65–80% ¼ ½0.65; 0.80.
mand area (CCA). The proposed model has been applied to the
Hirakud canal subcommand area of eastern India for interseasonal Land Area Constraint
integrated land and water resources planning. Summation of individual crop area should be less than or equal to
total cultivable command area for all seasons and sectors:
ncrop
X
Land and Water Management Model A


ijk ≤ TCAij ; ∀ i and j ð3Þ
k¼1
A single objective optimization models is developed for land and
water resources management in the canal command area. The man- where TCA

ij = total cultivable command area of sector i for all j.
agement framework considers uncertainties in both system param-
Water Availability Constraints
eters and decision variables. Thus, the developed management
Sectorwise and seasonwise surface water and groundwater use
model is robust in terms of optimality point of view and considers
should be less than or equal to total available surface water and
canal water, groundwater, rainfall, crop area, and cost components
groundwater in respective sectors and seasons:
within a single management framework. It consists of an objective
function and a set of constraints. The objective of this study is to SW


ij ≤ TSWij ; ∀ i and j ð4Þ
maximize the net annual benefit (sectorwise and seasonwise com-
ponents) with the optimum cropping pattern for the subcommand
area subject to geographic and water balance constraints. GW


ij ≤ TGWij ; ∀ i and j ð5Þ

where TSW
ij = total available surface water in sector i and

Model Formulation season j, and TGW
ij = total available groundwater in sector i and
season j.

Objective Function Hydrologic Balance of Aquifer

To ascertain water availability, hydrologic balance in the canal
Objective function is for maximization of the net annual benefit. command system needs to be satisfied. Schematic line diagram
Mathematically the objective function can be expressed as in Fig. 1 shows the hydrologic balance in the aquifer system of
ncrop the canal command area. It is assumed that inflow into the aquifer
X
nsect X
nseas X
Maxf
ðA



ijk ; SWij ; GWij Þ ¼ a


ijk Aijk
is equal to outflow from the aquifer system. Mathematically, the
i¼1 j¼1 k¼1 hydrologic water balance constraint can be expressed as
X
nsect X
nseas X
nsect X
nseas
− ðCSW

ij SW

ij þ Cij
GW

GW

ijÞ ½GW






ij − fθ1 SWij þ θ3 EðRij Þ Ai
i¼1 j¼1 i¼1 j¼1

ð1Þ þ θ




2 ðβ 1 SWij þ GWij Þg ≤ PMA

ð6Þ

where i, j, and k = index for sector of command area, crop growing Taking all known terms in the right-hand side
season, and crops, respectively; j ¼ 1 for Kharif season and j ¼ 2
X
nsect X
nseas
for Rabi season; a
ijk = net return for crop k in season j of sector i; ð1 − θ






2 ÞGWij − ðθ1 þ θ2 β 1 ÞSWij
nsect = number of sectors; nseas = number of seasons; ncrop = number i¼1 j¼1
of crops; A
ijk = area allocated to crop k in season j of sector i;
SW
≤ PMA
þ θ



3 EðRij Þ Ai ð7Þ
ij = surface (canal) water allocated in season j for sector i;

734 / JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT © ASCE / NOVEMBER/DECEMBER 2013

J. Water Resour. Plann. Manage. 2013.139:733-740.

 Optimization Algorithm: GLP

An interactive solution algorithm has been proposed by Huang et al.

(1995) to solve gray mixed-integer linear problems through analy-
ses of the interrelationships between the parameters and the vari-
ables and between the objective function and the constraints. This
algorithm is adopted for framing the gray linear programming
algorithm (Huang et al. 1995; Huang 1998).
In a GLP model, parameters and control variables are expressed
as linear objective functions, equalities and/or inequalities with
Downloaded from ascelibrary.org by The Indian Inst of Technology Kharagpur Librarian on 10/23/13. Copyright ASCE. For personal use only; all rights reserved.

gray numbers. The GLP allows the interval information to be di-

rectly communicated into the optimization process and resulting
solution. In general, GLP model can be written as follows:
Maxf
¼ C
X
ð12Þ

Subject to
A
X
≤ B
ð13Þ

Fig. 1. Schematic diagram of groundwater balance for the study area x

j ¼ gray continuous variable; x


j ∈X ; ∀j ð14Þ

x

j ≥ 0; ∀j ð15Þ
where θ

1 = conveyance loss of surface water (fraction) = [0.20, where C
∈ fR
g1×n , A
∈ fR
gm×n , and B
∈ fR
gm×1 (R

0.35]; θ

2 = field water application loss (factor) = [0.30, 0.30]; denotes a set of gray numbers). For n gray coefficients, c
j in the
θ

3 = rainfall recharge (fraction) = [0.10, 0.18] (Sharma and Paul objective function (12), if k1 of them are positive, i.e., c
j ≥ 0
1999; Nirala et al. 2005); EðRij Þ
= expected rainfall of sector i (j ¼ 1; 2; : : : ; k1 ), and of model k2 are negative, i.e., c

j <0
and season j; A


i = total available rechargeable area; and PMA = (j ¼ k1 þ 1; k1 þ 2; : : : ; n), where k1 þ k2 ¼ n (not considering
permissible annual mining allowance of the aquifer. the case when two bounds of c
j have different signs).
Permissible annual mining allowance of the aquifer is deter- The stepwise outline of the algorithm is presented below:
mined by

PMA
¼ Δh
× A × S

y ð8Þ Algorithm
1. Formulate fþ submodel. P1 þ þ
where Δh
= annual average groundwater table fluctuations; S

y = a. Formulate
Pn fþ objective function: f þ ¼ kj¼1 cj xj
specific yield of the aquifer; and A = total command area. þ −
þ j¼k1 þ1 cj xj .
b. Formulate constraints corresponding to f þ objective
Maximum/Minimum Allowable Area function:
P1 Pn
Deviation from the existing cropping area has limitation due to • kj¼1 jaij j− Signða− þ
ij Þxj þ
þ
j¼k1 þ1 jaij j Signðaij Þ
þ
socioeconomic aspect of the farmers x−
j ≤ b

i ; ∀ i
• x

j ≥ 0; ∀j
Aþ
ijk ≤ μijk TAijk
max
ð9Þ 2. Solve fP þ
submodel using Pn conventional linear programming.
k1
fþ
opt ¼
þ þ
j¼1 cj xjopt þ
þ −
j¼k1 þ1 cj xjopt .
3. Solution of fþ submodel gives xþ jopt (j ¼ 1; 2; : : : ; k1 ) and
A− x−jopt (j ¼ k1 þ 1; k1 þ 2; : : : ; n).
ijk ≥ μijk TAijk ð10Þ
min
4. Formulate f− submodel. P1 − −
a. Formulate
Pn f − objective function: f − ¼ kj¼1 cj xj
where μmax
ijk = factor by which the existing area of crop k can be þ j¼k1 þ1 c− þ
j xj .
increased in season j for sector i; μmin
ijk = factor by which the b. Formulate constraints corresponding to fþ objective
existing area of crop k can be decreased in season j for sector i; function:
and TAijk = total cultivable command area for crop k in season j P1 Pn
• kj¼1 jaij jþ Signðaþ −
ij Þxj þ
−
j¼k1 þ1 jaij j Signðaij Þ
−
under sector i. þ

xj ≤ bi ; ∀ i
þ
• x−
j ≤ xjopt ðj ¼ 1; 2; : : : ; k1 Þ
Nonnegativity
• xþ
j ≥ x − ðj ¼ k þ 1; k þ 2; : : : ; nÞ
jopt 1 1
For gray linear programming all constraints must be greater or
• x

j ≥ 0; ∀j
equal to zero: −
5. Solve fP submodel using Pn conventional linear programming
fopt ¼ kj¼1
− 1
c− −
j xjopt þ
− þ
j¼k1 þ1 cj xjopt
A

ijk ≥ 0; SW

ij ≥ 0; GW

ij ≥ 0 ∀ i; j; and k ð11Þ 6. Solution of f− submodel gives x− jopt (j ¼ 1; 2; : : : ; k1 )
and xþ jopt (j ¼ k 1 þ 1; k 1 þ 2; : : : ; n).
þ þ
The single objective gray linear optimization model consists of 7. Finally, f
−
−
opt ¼ ½f opt ; f opt  and xopt ¼ ½xopt ; xopt ; ∀ j.
objective function (1) together with the constraint set (2)–(5), (7),
(9)–(11). It contains the decision variables A



ijk , SWij , and GWij . Where, Signðx
Þ ¼ 1, if x
≥ 0 and −1, if x
< 0; jxj− ¼ x− , if
The optimization model is solved using GLP. x ≥ 0 and −xþ , if x
< 0; jxjþ ¼ xþ , if x
≥ 0 and −x− , if

JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT © ASCE / NOVEMBER/DECEMBER 2013 / 735

J. Water Resour. Plann. Manage. 2013.139:733-740.

 x
< 0. More details about the gray optimization model formu-
lation can be found in Huang et al. (1995).
Application of GLP Model
The GLP optimization model is applied to the Hirakud subcom-
mand area in eastern India. The Hirakud subcommand area, situ-
ated in the western part of Odisha State (India) is selected as the
Downloaded from ascelibrary.org by The Indian Inst of Technology Kharagpur Librarian on 10/23/13. Copyright ASCE. For personal use only; all rights reserved.
study area. The study area (Fig. 2) is bounded by north latitudes
20° 15′ 53″ to 21° 35′ 10″ and east longitudes 83° 52′ 0.6″ to
84° 10′ 21″ and falls in two UTM zones, 44Q and 45Q. Topography
of the study area varies from plain to undulating. The general slope
(1–6%) is toward southeast direction. Central part of the study area
is more undulating in nature with high land slope (5–6%). The
whole area is bounded by river, canal, distributory, and the Hirakud
reservoir (constructed over the Mahanadi River). The Mahanadi
River, Sason main canal, and Huma tail distributaries are the boun-
daries of the eastern, western, and southern part of the study area,
respectively. The Mahanadi River is perennial, whereas the tribu-
taries (Harad and Multi Jhor River) are ephemeral. Around 15%
of the total canal length is lined. This area receives major portion
of the rainfall during southwest monsoon (from June to October).
Generally, July and August are the rainiest months of the season.
The average annual rainfall of this area has been found to be
1,495 mm. Expected rainfall [EðRij Þ
] is taken as [1364,
1963] mm. Temperature varies in between 8°C (in January) and
47°C (in May) with an average value of 28°C. Groundwater occurs
Fig. 2. Location map of the study area
736 / JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT © ASCE / NOVEMBER/DECEMBER 2013
J. Water Resour. Plann. Manage. 2013.139:733-740.
under phreatic condition in the weathered residuum and under
semiconfined to confined conditions in the fractures zone.
The study area faces waterlogging condition due to nonuniform
distribution of canal water that leads to seepage and field applica-
tion losses (i.e., excessive supply in head and middle reaches and
shortage in supply at the tail reach of the command area). It affects
crop productivity and total agricultural production. In this region,
farmers mostly cultivate rice during monsoon (Kharif) and winter
(Rabi) seasons. Moreover, other crops grown in Rabi season are
vegetables, pulses, and oilseeds. The study area has cropping in-
tensity of about 160% (100% in Kharif and 60% in Rabi seasons).
Preliminary analysis shows that both the system parameters (net
irrigation requirement, irrigation water cost, and total cultivable
command area) and the decision variables (e.g., cropping pattern,
surface water, and groundwater) are susceptible to deviations.
Thus, there is possibility of developing an integrated inexact
resources management model to determine seasonal optimum
cropping area and canal water/groundwater allocation policies.
Total cultivable subcommand area is 21,968 ha (1 ha ¼ 104 m2 ).
Growing season, soil characteristics and food habit of the farmers
dictate the existing cropping pattern of the study area. Therefore,
optimization model considers the lower and upper limit of area
under different crops. In the application, three different sectors (or
administrative blocks) namely Dhankauda, Maneswar, Jujumura
are considered. Sector/block and season-wise crop area, net irriga-
tion requirement (NIR), and net return (NR) are shown in Table 1.
Reference ET (evapotranspiration) values are calculated using the
method described in Hargreaves and Samani (1985). NIR values
 Downloaded from ascelibrary.org by The Indian Inst of Technology Kharagpur Librarian on 10/23/13. Copyright ASCE. For personal use only; all rights reserved.

Table 1. Sector/Blockwise and Seasonwise Net Irrigation Requirement, Net Return, and Area under Different Crops (Deputy Director of Agriculture, Modipara, Sambalpur, Odisha; Personal Communication with
the Farmers)
Area for sector/block (ha)
Net return
Dhankauda Maneswar Jujumura Net irrigation requirement (mm) (Rs:=haa)
Kharif Rabi Kharif Rabi Kharif Rabi Kharif Rabi Kharif Rabi
Crops (k) A−
11k Aþ
11k A−
12k Aþ
12k A−
21k Aþ
21k A−
22k Aþ
22k A−
31k Aþ
31k A−
32k Aþ
32k NIR−
i1k NIRþ
i1k NIR−
i2k NIRþ
i2k NR

i1k NR

i2k

Rice 5,560.48 5,564.38 2,165.66 2,169.56 6,247.90 6,252.11 1,912.24 1,912.24 2,716.75 2,720.53 400.80 401.75 647 788 766 859 12,000 34,000
Maize 7.80 11.71 7.80 7.80 10.52 16.83 6.31 8.42 17.02 21.74 2.84 2.84 507 616 521 581 5,970 11,200
Millet 7.80 11.71 — — 10.52 16.83 — — 17.96 22.69 — — 346 433 — — 7,570 —
Pigeon pea 58.53 66.34 — — 44.18 50.49 — — 28.36 28.36 — — 630 755 — — 20,969 —
Green gram 444.84 487.76 509.22 567.75 567.99 635.31 715.25 757.32 308.16 362.99 156.92 175.82 233 309 217 256 5,550 5,550
Black gram 241.93 280.95 136.57 156.08 357.62 538.54 21.04 42.07 160.70 266.57 102.09 130.45 336 414 318 365 8,950 10,950
Horse gram 85.85 124.87 206.81 265.34 58.90 67.32 126.22 180.92 47.26 56.72 56.72 83.19 325 400 359 406 8,700 10,100
Other pulses 66.34 70.24 15.61 19.51 48.38 149.36 21.04 25.24 69.01 74.68 73.73 75.62 235 298 258 300 6,950 6,500
Groundnut 15.61 31.22 58.53 62.43 16.83 29.45 46.28 50.49 9.45 30.25 20.80 22.69 563 682 393 442 15,000 12,000
Til 247.78 316.07 468.25 624.33 273.48 378.66 218.78 302.93 161.64 199.46 230.65 287.37 400 477 407 453 8,739 6,500
Mesta 1.95 11.71 — — 10.52 75.73 — — 6.62 19.85 — — 388 483 — — 4,575 —
Sweet potato 3.90 3.90 11.71 19.51 4.21 6.31 16.83 25.24 9.45 17.96 5.67 9.45 496 612 450 498 33,750 45,750
Vegetables 648.53 732.42 803.83 803.83 1,030.38 1,214.66 525.92 610.07 627.29 683.06 268.46 306.27 388 483 517 576 37,000 35,000
Chilli 46.83 54.63 117.06 124.87 33.66 56.80 126.22 130.43 17.96 36.87 62.39 64.28 708 880 540 680 20,000 16,100
Ginger 37.07 39.02 — — 88.35 88.35 — — 18.91 28.36 — — 960 1,150 — — 170,000 —
Turmeric 0.00 1.95 — — — — — — 1.89 2.84 — — 1,163 1,308 — — 36,500 —
Sugarcane 1.95 3.90 0.00 3.90 2.10 3.37 0.00 4.21 0.00 1.89 0.00 1.89 424 473 1,059 1,183 32,000 —
Wheat — — 13.66 15.61 — — 16.83 18.93 — — 15.13 17.02 — — 431 475 — 9,600
Gram — — 5.85 7.80 — — 2.10 4.21 — — 9.45 11.34 — — 355 412 — 8,300
Field pea — — 27.32 46.83 — — 25.24 42.07 — — 26.47 37.81 — — 356 393 — 16,000
Cow pea — — 39.02 42.92 — — 33.66 37.87 — — 28.36 28.36 — — 536 604 — 12,000
Mustard — — 202.91 234.13 — — 147.26 155.67 — — 105.87 109.65 — — 405 451 — 16,425
Sunflower — — 11.71 15.61 — — 8.42 12.62 — — 3.78 5.67 — — 366 403 — 14,500
Potato — — 15.61 35.12 — — 8.42 29.45 — — 30.25 39.70 — — 381 420 — 18,000
Onion — — 89.75 109.26 — — 96.77 117.81 — — 51.05 60.50 — — 636 706 — 56,250
Garlic — — 23.41 28.10 — — 8.42 12.62 — — 13.23 15.13 — — 552 613 — 63,600

J. Water Resour. Plann. Manage. 2013.139:733-740.

Coriander — — 70.24 70.24 — — 61.01 61.01 — — 24.58 24.58 — — 766 579 — 7,750
a
Indian rupees.

JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT © ASCE / NOVEMBER/DECEMBER 2013 / 737
 are calculated based on Doorenbos and Pruitt (1977) and Allen Rs: 4,500=ha-m in Rabi season, whereas the cost of groundwater
et al. (1998). varies between Rs: 3,700=ha-m (Kharif) to Rs: 4,000=ha-m (Rabi).
Minimum cultivable command area is taken as 95% of the total
command area. Rainfall is considered to be varying between the
Results and Discussion 50th percentile (1,364 mm) and the 90th percentile (1,963 mm)
(calculated based on the available historical data). Minimum net
In order to evaluate the GLP formulation, two different scenarios return is taken as 80% of the maximum net return (based on market
are considered. analysis). Considering the socioeconomic aspect, the lower limit of
the cropping area is taken as 50% of the prescribed value given in
Downloaded from ascelibrary.org by The Indian Inst of Technology Kharagpur Librarian on 10/23/13. Copyright ASCE. For personal use only; all rights reserved.

Table 1. Distribution of available and optimal water resources (both

Scenario-I
In Scenario-I, minimum and maximum available surface water in
both the cropping seasons are taken as 85 and 100% of the design
surface water availability (based on design discharge of Sason main
canal), respectively. Moreover, seasonwise variation in the cost of
surface water and groundwater are considered. Cost of surface
water is Rs: 2,700=ha-m (1 ha-m ¼ 104 m3 ) in Kharif and
surface water and groundwater) along with profit for Scenario-I are
given in Table 2. The net annual return from the subcommand area
ranges between Rs: 3.45 × 108 and Rs: 5.20 × 108 . Table 3 shows
that unlike other crops allocated area for rice cultivation signifi-
cantly differ in terms of maximum and minimum allocatable area.
Rice needs more water, and it fetches less profit as compared to
other crops. Thus, the optimization model suggests reduction of

Table 2. Distribution of Available and Optimal Water Resources along with Profit for Scenario-I
Available water resources Optimal allocation for Scenario-I
Surface water Groundwater Surface water Groundwater Profit for Scenario-I
Seasons
(j) Blocks (i) SW−
ij SWþ
ij GW−
ij GWþ
ij SW−
ij SWþ
ij GW−
ij GWþ
ij Min Max
Kharif Dhankauda 7,268.24 8,550.87 1,653.73 2,126.23 7,268.24 8,550.87 377.44 2,126.23 5.56 × 107 9.14 × 107
Maneswar 7,836.81 9,219.78 1,783.10 2,292.56 7,836.82 9,219.78 387.88 2,292.56 8.22 × 107 1.27 × 108
Jujumura 3,521.47 4,142.91 846.96 1,088.94 3,521.48 4,142.91 251.15 1,088.94 4.04 × 107 6.14 × 107
Rabi Dhankauda 6,891.00 8,107.06 1,173.62 1,508.93 2,565.93 3,684.34 1,173.62 1,508.93 8.29 × 107 1.16 × 108
Maneswar 7,430.06 8,741.25 1,265.43 1,626.98 2,627.19 3,762.86 1,265.43 1,626.98 6.56 × 107 9.45 × 107
Jujumura 3,338.70 3,927.88 601.07 772.80 1,714.15 2,434.24 601.07 772.80 1.80 × 107 2.99 × 107
Total profit in Rs. 3.45 × 108 5.20 × 108

Table 3. Optimal Allocation of Land Resources for Scenario-I

Blocks
Dhankauda Maneswar Jujumura
Kharif Rabi Kharif Rabi Kharif Rabi
Crops A−
11k Aþ
11k A−
12k Aþ
12k A−
21k Aþ
21k A−
22k Aþ
22k A−
31k Aþ
31k A−
32k Aþ
32k

Rice 5,230.85 5,564.38 2,169.56 2,169.56 5,838.38 6,252.11 1,912.24 1,912.24 2,553.09 2,720.53 401.75 401.75
Maize 0.00 0.00 7.80 7.80 0.00 0.00 8.41 8.42 0.00 0.00 2.84 2.84
Millet 11.70 11.71 — — 16.83 16.83 — — 22.69 22.69 — —
Pigeon pea 66.34 66.34 — — 50.49 50.49 — — 28.36 28.36 — —
Green gram 263.17 310.22 567.75 567.75 281.69 326.91 757.32 757.32 171.75 224.01 175.82 175.82
Black gram 280.95 280.95 156.08 156.08 538.54 538.54 42.07 42.07 266.57 266.57 130.45 130.45
Horse gram 124.87 124.87 265.34 265.34 67.32 67.32 180.92 180.92 56.72 56.72 83.18 83.19
Other pulses 70.24 70.24 19.51 19.51 149.36 149.36 25.24 25.24 74.68 74.68 75.62 75.62
Groundnut 31.22 31.22 62.43 62.43 29.45 29.45 50.49 50.49 30.25 30.25 22.69 22.69
Til 316.07 316.07 624.33 624.33 378.66 378.66 302.93 302.93 199.46 199.46 287.36 287.37
Mesta 0.00 0.00 — — 0.00 0.00 — — 0.00 0.00 — —
Sweet potato 3.90 3.90 19.51 19.51 6.31 6.31 25.24 25.24 17.96 17.96 9.45 9.45
Vegetables 732.42 732.42 803.83 803.83 1214.66 1214.66 610.07 610.07 683.06 683.06 306.27 306.27
Chilli 54.63 54.63 124.87 124.87 56.80 56.80 130.43 130.43 36.87 36.87 64.28 64.28
Ginger 39.02 39.02 — — 88.35 88.35 — — 28.36 28.36 — —
Turmeric 1.95 1.95 — — 0.00 0.00 — — 2.84 2.84 — —
Sugarcane 3.90 3.90 3.90 3.90 3.37 3.37 3.37 3.37 1.89 1.89 1.89 1.89
Wheat — — 15.61 15.61 — — 18.93 18.93 — — 17.01 17.02
Gram — — 7.80 7.80 — — 4.21 4.21 — — 11.34 11.34
Field pea — — 46.82 46.83 — — 42.07 42.07 — — 37.81 37.81
Cow pea — — 42.92 42.92 — — 37.87 37.87 — — 28.36 28.36
Mustard — — 234.12 234.13 — — 155.67 155.67 — — 109.65 109.65
Sunflower — — 15.61 15.61 — — 12.62 12.62 — — 5.67 5.67
Potato — — 35.12 35.12 — — 29.45 29.45 — — 39.70 39.70
Onion — — 109.26 109.26 — — 117.81 117.81 — — 60.50 60.50
Garlic — — 28.10 28.10 — — 12.62 12.62 — — 15.13 15.13
Coriander — — 70.24 70.24 — — 61.00 61.01 — — 24.58 24.58

738 / JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT © ASCE / NOVEMBER/DECEMBER 2013

J. Water Resour. Plann. Manage. 2013.139:733-740.

 Table 4. Optimal Allocation of Water Resources along with Profit for Scenario-II
Optimal allocation for Scenario—II
Surface water Groundwater Profit for Scenario-II
Seasons (j) Blocks (i) SW−
ij SWþ
ij GW−
ij GWþ
ij Min Max
7
Kharif Dhankauda 7,268.24 8,550.87 385.54 2,126.23 5.56 × 10 9.14 × 107
Maneswar 7,836.82 9,219.78 402.43 2,292.56 8.19 × 107 1.27 × 108
Jujumura 3,521.48 4,142.91 228.49 1,088.94 4.06 × 107 6.15 × 107
Rabi Dhankauda 4,131.48 4,131.48 0.00 1,328.85 8.95 × 107 1.21 × 108
Downloaded from ascelibrary.org by The Indian Inst of Technology Kharagpur Librarian on 10/23/13. Copyright ASCE. For personal use only; all rights reserved.

Maneswar 4,530.21 4,530.21 0.00 1,433.80 7.15 × 107 9.91 × 107

Jujumura 2,045.74 2,045.74 0.00 609.18 2.43 × 107 3.45 × 107
Total profit in Rs. 3.63 × 108 5.34 × 108

area under rice cultivation. However, allocated areas attain the re-
spective maximum values for other crops due to availability of sur-
face water and groundwater in sufficient quantity. Based on
available land and water resources, optimum crop area allocation,
water utilization and minimum-maximum profit corresponding to
different blocks/sectors in various seasons are given in Tables 2 and
3. Model results indicate that for Scenario-I the prescribed amount
of surface water use is more than the groundwater in Kharif season.
However, prescribed groundwater use is more in the case of the
Rabi season.
Scenario-II
Scenario-II is similar to Scenario-I except for the cost of surface
water and groundwater. The costs of surface water and groundwater
are taken as Rs: 2,700=ha-m and Rs: 3,700=ha-m (lower value for
Scenario-I), respectively, for both the seasons. In Scenario-II, profit
(net return ranges between Rs: 3.63 × 108 and Rs: 5.34 × 108 ,
Table 4) is increased as compared to Scenario-I. Maximum amount
of surface water is allocated due to its lower unit cost as compared
to the unit cost of groundwater that leads to increase in total crop-
ping area. Results also indicate that in Rabi season groundwater
utilization is significantly reduced because of high groundwater
pumping cost. Scenario-II related results for optimum water utiliza-
tion and maximum profit corresponding to different blocks/sectors
and seasons are given in Table 4. In Scenario-II, optimal crop area
allocation shows minor change. Model result prescribes reduction
in rice cultivation area by 1.72% and increase in Green Gram cul-
tivation area by 16% during Kharif season in Maneswar block as
compared to Scenario-I. It is evident that prescribed results are
deterministic numbers (the same value for lower and upper limits)
for land area allocation in Rabi season. This is due to the fact that
both surface water and groundwater resources are available in
sufficient quantity. Moreover, cropping intensity for Rabi season
is only 60%. With the same area allocation Scenario-II shows in-
crease in profit for Rabi season due to less surface water and
groundwater cost. In Scenario-II, result shows increase in SW− ij by
55% and in SWþ ij by 8.4%. Moreover, result indicates zero use for
þ
GW− ij and reduction in use for GWij by 12% (Table 4). Interest-
ingly, both the scenarios prescribe zero crop area for Maize during
Kharif season. This is due to the fact that Maize fetches less profit
with high NIR as compared to others crop. In Rabi season crop area
for Maize shows maximum upper limit.
Conclusions
An integrated land and water management framework is developed
to determine maximum benefit, optimum conjunctive use (canal
water and groundwater), planning and cropping pattern in different
blocks (sectors), and seasons under uncertainty of hydrological
events (e.g., rainfall, canal water availability). A generic formu-
lation is developed for optimal land and water resources allocation.
The formulation is solved utilizing gray linear programming for
different scenarios. The developed methodology is applied to the
Hirakud subcommand area in eastern India.
Two different scenarios are considered for the evaluation of
developed methodology. In Scenario-I, the surface water and
groundwater used are maximum for the Kharif and Rabi seasons,
respectively. Rice cultivation significantly differs from maximum
and minimum allowable area as compared to other crops. In
Scenario-II, profit is increased in comparison to Scenario-I. Maxi-
mum surface water is used for both the seasons. Also, cropping area
under rice cultivation increases significantly compared to other
crops. In Rabi season, groundwater utilization is significantly re-
duced because of high pumping costs. In Rabi season, optimum
land area allocation shows the maximum area allocated for maxi-
mum profit and values are deterministic in nature. Results show that
there is scope for increasing the cropping intensity in Rabi season.
Thus, some encouragement to the farmers (may be in terms of
water cost reduction/subsidy) can help in improving the situation.
This limited evaluation results shows the potential applicability of
the developed methodology for command area systems. However,
more rigorous (inclusion of detailed physics-based description of
the system in terms of complex equations) evaluations are needed
for actual application to the field.
Acknowledgments
The authors are thankful to Mr. Susant Kumar Rout, Mr. Trilochon
Dash, and Mr. R. K. Pradhan of the Hirakud Canal Command Area,
Burla Subdivision, Sambalpur, Odisha for providing relevant data
for this study.
References
Allen, R. G., Pereira, L. S., Raes, D., and Smith, M. (1998). “Guideline for
computing crop water requirement.” Irrigation and Drainage Paper
No. 56, FAO, Rome.
Deng, J. L. (1982). “Control problems of grey systems.” Syst. Control Lett.,
1(5), 288–294.
Dhar, A., and Datta, B. (2009). “Saltwater intrusion management of coastal
aquifers. II: Operation uncertainty and monitoring.” J. Hydrol. Eng.,
14(12), 1273–1282.
Doorenbos, J., and Pruitt, W. J. (1977). “Guideline for predicting crop
water requirement.” Irrigation and Drainage Paper No. 24, FAO,
Rome.
Hamid, R. S., and Mohammad, A. A. (2011). “Optimal crop planning and
conjunctive use of surface water and groundwater resources using fuzzy
dynamic programming.” J. Irrig. Drain. Eng., 137(6), 383–397.

JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT © ASCE / NOVEMBER/DECEMBER 2013 / 739

J. Water Resour. Plann. Manage. 2013.139:733-740.

 Hargreaves, G. H., and Samani, Z. A. (1985). “Reference crop evapotran-
spiration from temperature.” Appl. Eng. Agric., 1(2), 96–99.
Huang, G. H. (1998). “Theory and methodology—a hybrid inexact—
stochastic water management model.” Eur. J. Oper. Res., 107(1),
137–158.
Huang, G. H., Baetz, B. W., and Patry, G. G. (1995). “Grey integer
programming: An application to waste management planning under
uncertainty.” Eur. J. Oper. Res., 83(3), 594–620.
Jiracheewee, N., Oron, G. B., Murty, V. V. N., and Wuwongse, V. (1996).
“Computerized database for optimal management of community irriga-
Downloaded from ascelibrary.org by The Indian Inst of Technology Kharagpur Librarian on 10/23/13. Copyright ASCE. For personal use only; all rights reserved.
tion systems in Thailand.” Agric. Water Manage., 31(3), 237–251.
Karamouz, M., Zahraie, B., Kerachian, R., Eslami, A. (2010). “Crop
pattern and conjunctive use management: A case study.” Irrig. Drain.,
59(2), 161–173.
Khare, D., Jat, M. K., and Sunder, J. D. (2007). “Assessment of water
resources allocation options: Conjunctive use planning in a link canal
command.” Resour. Conserv. Recycl., 51(2), 487–506.
Kumar, D. N., Srinivasa, K. R., and Ashok, B. (2006). “Optimal reservoir
operation for irrigation of multiple crops using genetic algorithms.”
J. Irrig. Drain. Eng., 132(2), 123–129.
Latinopoulos, D., and Mylopoulos, Y. (2005). “Optimal allocation of
land and water resources in irrigated agriculture by means of goal
programming: Application in Loudias river basin.” Global NEST J.,
7(3), 264–273.
Li, W., Li, Y. P., Li, C. H., and Huang, G. H. (2010). “An inexact two-stage
water management model for planning agricultural irrigation under
uncertainty.” Agric. Water Manage., 97(11), 1905–1914.
Lu, H., Huang, G. H., and He, L. (2009). “Inexact rough-interval two-stage
stochastic programming for conjunctive water allocation problems.”
J. Environ. Manage., 91(1), 261–269.
740 / JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT © ASCE / NOVEMBER/DECEMBER 2013
View publication stats J. Water Resour. Plann. Manage. 2013.139:733-740.
Mainuddin, M., Gupta, A. D., and Onta, P. R. (1997). “Optimal crop plan-
ning model for an existing groundwater irrigation project in Thailand.”
Agric. Water Manage., 33(1), 43–62.
Nirala, A. K., Panda, S. N., Billib, M., and Hollender, H. M. (2005). “Sec-
torial planning of land and water resources in a large canal command in
the sub-humid region of Eastern India.” Zeitschrift fur Bewasserungs-
wirtschaft, 40(1), 115–134.
Panda, S. N., Khepar, S. D., and Kaushal, M. P. (1996). “Interseasonal
irrigation system planning for waterlogged sodic soils.” J. Irrig. Drain.
Eng., 122(3), 135–144.
Paul, A., Panda, S. N., and Kumar, D. N. (2000). “Optimal irrigation allo-
cation: A multilevel approach.” J. Irrig. Drain. Eng., 126(3), 149–155.
Rejani, R., Jha, M. K., and Panda, S. N. (2009). “Simulation-optimization
modeling for sustainable groundwater management in a costal basin of
Orissa, India.” Water Res. Manage., 23(2), 235–263.
Rosenberg, D. E. (2009). “Shades of grey: A critical review of grey-number
optimization.” Eng. Optim., 41(6), 573–592.
Sethi, L. N., Panda, S. N., and Nayak, M. K. (2006). “Optimal crop plan-
ning and water resources allocation in a coastal groundwater basin,
Orissa, India.” Agric. Water Manage., 83(3), 209–220.
Sharma, B. R., and Paul, D. K. (1999). “Water resources of India.” 50 years of
Natural Resources Management Research, ICAR, New Delhi, 31–48.
Singh, D. K., Jaiswal, C. S., Reddy, K. S., Singh, R. M., and Bhandarkar,
D. M. (2001). “Optimal cropping pattern in canal command area.”
Agric. Water Manage., 50(1), 1–8.
Vivekanandan, N., Viswanathan, K., and Gupta, S. (2009). “Optimization
of cropping pattern using goal programming approach.” OPSEARCH,
46(3), 259–274.
Zhang, Y. M., and Huang, G. H. (2011). “Inexact credibility constrained
programming for environmental system management.” Resour.
Conserv. Recycl., 55(4), 441–447.