Agricultural Water Management 259 (2022) 107235

Contents lists available at ScienceDirect

Agricultural Water Management

journal homepage: www.elsevier.com/locate/agwat

Sustainable management of agricultural water and land resources under

changing climate and socio-economic conditions: A multi-dimensional
optimization approach
Mo Li a, b, Xiaoxu Cao a, Dong Liu a, b, *, Qiang Fu a, b, *, Tianxiao Li a, b, Ruochen Shang a
a
School of Water Conservancy & Civil Engineering, Northeast Agricultural University, Harbin, Heilongjiang 150030, China
b
Key Laboratory of Effective Utilization of Agricultural Water Resources of Ministry of Agriculture, Northeast Agricultural University, Harbin, Heilongjiang 150030,
China

A R T I C L E I N F O A B S T R A C T

Handling Editor J.E. Fernández Conflict between limited water supply and the ever-increasing water demand poses the challenge of synergetic
management of agricultural water and land resources (AWLR). Sustainable development strategy and changing
Keywords: environment increase the multi-dimensional characteristic and complexity of the management of AWLR. This
Agricultural water and land resources paper establishes a model framework for the multi-dimensional optimization of AWLR in a changing environ­
Multiple dimensions
ment. The model framework is advantageous of: (1) Comprehensively allocating water and land resources on the
Collaborative optimization
basis of clarifying their interactions; (2) Balancing incompatible goals from multiple dimensions including re­
RCPs-SSPs model
Sustainability sources, society, economy, ecology, and environment; (3) proposing alternative allocation schemes of AWLR that
Water security can response to the changing environment of both natural and socio-economic changes. Allocation schemes of
AWLR based on the model framework are generated, analyzed and evaluated. The comprehensiveness, equi­
librium, and security of multi-dimensional targets help obtain the optimum adaptation allocation plans of AWLR
to cope with changing environment. The real-world case study in Songhua River Basin in Northeast China verifies
the feasibility and practicality of the model framework. The study found that the model framework can manage
AWLR in a sustainable way and meanwhile provide decision makers alternatives plans of AWLR for different
natural and social changing environments, which will further contribute to the alleviation of agricultural water
scarcity and the promotion of agricultural sustainable development.

1. Introduction which in turn affect the agricultural production and sustainable devel­
opment of the region. This leads to the necessity to optimal allocate
Water resources shortage has aggravated the contradictions and AWLR simultaneously to realize SDGs.
conflicts among natural resources, economy, ecology, society and Many methods, such as linear programming, nonlinear program­
environment, and is becoming a severe global challenge. Most Sustain­ ming, dynamic programming, multi-objective programming, genetic
able Development Goals (SDGs), which are released in 2015, are directly optimization algorithm, game-based theory and synergetic-based the­
or indirectly affected by water shortage (Vanham et al., 2018). Agri­ ory, have been used for optimal allocation of agricultural water and/or
cultural irrigation water accounts for 70% of global freshwater con­ land resources. Recently, the comprehensive allocation of AWLR on the
sumption (Dai and Li, 2013; Li et al., 2018), and agricultural water basis of considering their relationship are reported. For example, Xie
shortage will seriously restrict the sustained economic and social et al. (2018) developed an inexact stochastic-fuzzy programming model
development. Agricultural water allocation has an impact on the allo­ for optimal allocation of land resources and irrigation water resources to
cation of crop cropping structures, as water and land resources interact achieve the goal of system economic benefit maximization. Linker
with each other. The allocation of AWLR is a prerequisite for food pro­ (2020) proposed a model-based optimization method for allocation of
duction and is related to food security, economic development, and cropping areas and water, which could be used to make the optimal
environmental effects (Fan et al., 2020; Li et al., 2020; Zeng et al., 2018), irrigation plans for different crops in different regions. Tang et al. (2020)

* Corresponding authors at: School of Water Conservancy & Civil Engineering, Northeast Agricultural University, Harbin, Heilongjiang 150030, China.
E-mail addresses: [email protected] (D. Liu), [email protected] (Q. Fu).

https://doi.org/10.1016/j.agwat.2021.107235
Received 23 January 2021; Received in revised form 20 August 2021; Accepted 2 October 2021
Available online 19 October 2021
0378-3774/© 2021 Elsevier B.V. All rights reserved.
M. Li et al. Agricultural Water Management 259 (2022) 107235

developed a remote-sensing-based optimization approach to obtain system’s comprehensiveness, equilibrium, security to select the optimal
sustainable management strategies of AWLR at grid-scale. Water AWLR allocation plans that contribute to promote to the sustainable use
shortage and the discordant development of social-economy and of AWLR. The model framework is applied to Songhua River Basin in
ecological-environment have exacerbated the need for sustainable Heilongjiang Province, northeast China.
management of AWLR. The sustainable allocation of AWLR requires that
the focus should not be solely on economic benefits, the comprehensive 2. Modeling
effects of natural resources use, social economy development and
ecological environment protection are desired to consider simulta­ This section develops a model-based framework for simultaneous
neously. In response to requirements of sustainable development, deci­ allocation of AWLR. The framework is mainly composed of two parts.
sion makers need to comprehensively consider the cooperative effect of The first part focuses on the construction of a multi-dimensional multi-
natural resources, economic, social, ecological and environmental di­ objective optimization model, including the quantification of multiple
mensions, when managing AWLR. Multi-objective programming (MOP) objective functions of resources, economic, social, ecological and envi­
technique is an effective way to balance conflict goals, and some studies ronmental dimensions and the corresponding constraints, and the model
have been reported to allocate agricultural water and/or land resources solution method. The second part emphasizes the generation of com­
using MOP. For example, Tang et al. (2019) established a MOP model for bined changing environment patterns including both climate change
the agricultural water allocation based on the spatial and temporal and social-economic development changes. The framework is illustrated
variations of precipitation and evaporation, with the goal of maximizing in Fig. 1.
economic benefits and, minimizing water shortage. Ren et al. (2019)
proposed an improved multi-objective stochastic fuzzy programming 2.1. Multi-dimensional modeling for agricultural water and land
method for the simultaneous optimization of irrigation water quantity allocation
and irrigation areas, aiming at maximizing net income, maximizing
agricultural water productivity and minimizing irrigated area. Yang 2.1.1. Notations
et al. (2020) proposed a fuzzy multi-objective linear fractional pro­ This part develops a multi-dimensional multi-objective program­
gramming (FMOLFP) algorithm based on superiority and inferiority ming model for the simultaneous allocation of AWLR. The following is
measures method to optimize agricultural planting structure in order to the list of notations to present the model clearly. (Table 1).
achieve the goals of maximum economic benefit per unit irrigation
water and maximum grain production efficiency. However, few studies
simultaneously optimized both agricultural water and land resources
with the trade-off of multi-dimensional goals of natural resources use,
economic development, social stability, ecological and environment 2.1.2. Objective function
protection in a model framework.
Since the 20th century, the increase in water demand has caused 2.1.2.1. Resource dimension. Agricultural production consumes a
severe water shortages in many parts of the world (Schewe et al., 2014; certain amount of water and land resources. This paper selects two in­
Wada et al., 2011, 2013). Crop water demand is a sensitive factor in the dicators, i.e. irrigation water use efficiency (IWUE) and water footprint
allocation of AWLR (Gong et al., 2020). Climate change and to quantitatively characterize how to make full and effective use of
socio-economic development together affect crop water demand, which limited water and land resources to produce grain.
further influence water shortages (Greve and Seneviratne, 2015; Gud­ Maximum IWUE. This goal is to maximize the yield of food crops per
mundsson et al., 2017; Huang et al., 2017). Climate change affects crop unit of irrigation water. It takes into account the relationship between
water requirements by meteorological elements such as temperature, crop yield and water consumption, and it can be expressed as follows:
duration of sunshine and wind speed, while socio-economic develop­
ment that is driven by human activities leads to changes in hydrological ∑
I ∑
J
Yij Aij
elements, which accordingly affects water demand through social water
(1)
i=1 j=1
max IWUE =
cycle processes and crop planting conditions. Therefore, the in-depth ∑
I ∑
J ( )
Aij Xij sur + X gro
ij + Pei
study of the joint effect of climate change and social and economic i=1 j=1
development on crop water demand, and how does it drive the
multi-dimensional allocation of AWLR, is necessary and significant. This The concept of water footprint can help reflect the utilization effi­
will be conducive to improve the coping capacity of comprehensive use ciency of different forms of water resources throughout the growth
of agricultural water and land resources on complex changes in the period of crops, including green, blue, and grey water footprint. Green
environment. Nevertheless, relevant studies are limited. water footprint refers to the green water consumption (i.e. the total
Therefore, this study establishes a comprehensive allocation model evapotranspiration of rainwater) during the growth period of crops. Blue
framework of AWLR that simultaneously considers conflict goals from water footprint refers to the consumption of blue water resources (i.e.
multiple dimensions of resource use efficiency, economic benefits, social the total evapotranspiration of irrigation water) during the growth
equity, ecological and environmental protection, on the basis of quan­ period of the crop. Grey water footprint refers to the grey water demand
tifying dual changing environment modes of climate change and social (i.e. the amount of polluted dilution water produced by pesticides and
economic development change. The model framework is capable of: (1) fertilizers) during the growth period of crops. The utilization efficiency
comprehensively allocating water (surface water and groundwater) and of blue water, green water, and grey water is based on the virtual water
land resources for different kinds of crops in different regions to alleviate theory, which can be obtained by the method of (Hoekstra and Hung,
the imbalance of water supply-demand, considering the correlation 2002):
between water and land resources within a model framework; (2) / /
lgp
∑
optimally coordinating contradictories between natural resources use VWC = CWR Y = 10 × ETc Y (2)
efficiency, resources allocation equity, economic benefit, biological d=1

productivity, air and water environmental effects, corresponding re­

ETc = Kc × ET0 (3)
sources, social, economic, ecological and environmental dimensions,
respectively; (3) obtaining a series of alternative allocation plans of 3
The virtual water content (VWC, m /kg) refers to the ratio of the
AWLR that response to dual change environment of both climate change crop water requirements (CWR, m3/ha) during the growth period to the
and socio-economic development change; (4) gaining in-sight into the total crop yield Y (kg/ha). 10 is the conversion coefficient of water depth

2
M. Li et al. Agricultural Water Management 259 (2022) 107235

Fig. 1. Model-based framework for simultaneous allocation of AWLR.

3
M. Li et al. Agricultural Water Management 259 (2022) 107235

Table 1 Table 1 (continued )

Definitions of symbols. Indices Definition
Indices Definition
CF Compound fertilizer application cost (yuan/kg)
i Index of subarea (i = 1,2,…,I) CM Phosphate fertilizer application cost (yuan/kg)
j Index of crop (j = 1,2,…,J) CO Potash application cost (yuan/kg)
sur Superscript for surface water NM Nitrogen content in compound fertilizer
gro Superscript for groundwater PM Phosphorus content in compound fertilizer
max Superscript for maximum KN Nitrogen inflow coefficient
min Superscript for minimum KP Phosphorus inflow coefficient
Variables ECO2 Nitrous oxide emissions (kg)
Aij Planting area of crop j in subarea i (ha) EN2O Nitrous oxide emissions (kg)
Xijsur Surface water consumption of crop j in subarea i (m3/ha) ECH4 Methane emissions (kg)
Xijgro Ground water consumption of crop j in subarea i (m3/ha) CS Fixed carbon in crop production (kgCO2eq)
Parameters FD Per capita food demand (kg/ people)
Yij Yield of crop j in subarea i (kg/ha) TK Maximum arable area (ha)
VWCij Virtual water content of crop j in subarea i (m3/kg) TN Maximum allowable total nitrogen loss (kg/ha)
VWGij Green water content in the virtual water content of crop j in subarea i SL Maximum allowable soil loss (kg/ha)
(m3/kg) qh Carbon dioxide emission coefficient of nitrogen fertilizer (kg CO2/kg)
VWBij Blue water content in the virtual water content of crop j in subarea i Qf Carbon dioxide emission coefficient of compound fertilizer (kg CO2/
(m3/kg) kg)
VWRij Grey water content in the virtual water content of crop j in subarea i qo Carbon dioxide emission coefficient of potash fertilizer (kg CO2/kg)
(m3/kg) qm Carbon dioxide emission coefficient of phosphorus fertilizer (kg CO2/
Cij Planting cost of crop j in subarea i (yuan/ha) kg)
Hij Amount of nitrogen fertilizer applied of crop j in subarea i (kg/ha) qe Carbon dioxide emission coefficient of electricity (kg CO2/kg)
Fij Application amount of compound fertilizer of crop j in subarea i (kg/ qd Carbon dioxide emission coefficient of fuel (kg CO2/kg)
ha) qp Carbon dioxide emission coefficient of pesticide (kg CO2/kg)
Mij Application amount of phosphate fertilizer of crop j in subarea i (kg/ qs Carbon dioxide emission coefficient of agricultural film (kg CO2/kg)
ha) Qr Methane emission coefficient from paddy field (kg CH4/kg)
Oij Potassium fertilizer application rate of crop j in subarea i (kg/ha)
Note: yuan is the monetary unit of China.
Lij Soil loss of crop j in subarea i (kg/ha)
Aijmax Maximum planting area of crop j in subarea i (ha)
Aijmin Minimum planting area of crop j in subarea i (ha)
IRijmax Maximum irrigation water volume of crop j in subarea i (m3/ha) (mm) to water volume per unit area (m3/ha); The ETc is calculated from
IRij Irrigation water demand of crop j in subarea i (m3/ha) the day of planting (the first day) to the day of harvest (lgp is the length
NRij Nitrogen demand of crop j in subarea i (kg/ha) of the growth cycle, in days). Based on the concept of virtual water and
αij The ratio of the irrigated area to the planted area of crop j in subarea i water footprint, the following three objective functions are used to
Pei Effective rainfall in subarea i (m3/ha)
POi Population in subarea i (people)
reflect the utilization efficiency of AWLR.
φi Water production factor in subarea i (m3/ha) Maximum green water utilization rate. This objective is to maximize
Qi Total water resources supply in subarea i (m3) the consumption of green water resources, and the purpose is to make
Nsi Soil nitrogen content in subarea i (kg/ha) full use of effective rainfall to meet crop water requirements.
Psi Soil phosphorus content in subarea i (kg/ha)
CEi Power consumption in subarea i (kw⋅h/ha) ∑
I ∑
J
CDi Consumption of diesel fuel used in subarea i (kg/ha) Yij Aij VWGij
(4)
i=1 j=1
CPi Application amount of pesticide in subarea i (kg/ha) max Green Water =
Si Consumption of agricultural film used in subarea i (kg/ha) ∑
I ∑
J
Yij Aij VWCij
etaisur Irrigation utilization coefficient of surface water in subarea i i=1 j=1
etaigro Irrigation utilization coefficient of groundwater in subarea i
SWi Available surface water supply in subarea i (m3) / /
VWG = CWRgreen Y = 10 × ETgreen Y (5)
GWi Available groundwater supply in subarea i (m3)
EBi Minimum economic benefit of planting crops in subarea i (yuan)
Pj Price of crop j (yuan/kg) ETgreen = min(ETc , Pe) (6)
NTj Background coefficient of N2O emission of crop j (kg/ ha)
QHj Emission coefficient of N2O in nitrogen fertilizer of crop j Minimum blue water footprint. This objective is to minimize the
QFj Emission coefficient of N2O in compound fertilizer of crop j consumption of blue water, with the goal of minimizing irrigation water
λj Grass to grain ratio of crop j consumption to save water.
γj Root carbon content of crop j (%)
τj Crown-root ratio of crop j ∑
I ∑
J
σj Return rate of crop j (%) Yij Aij VWBij
(7)
i=1 j=1
Coefficient min Blue Water =
CWRgreen Green water requirements for crops (m3/ha) ∑
I ∑
J
Yij Aij VWLij
CWRblue Blue water requirements for crops (m3/ha) i=1 j=1
ET0 Reference crop evapotranspiration (mm/day)
ETC Crop water requirement (mm/day)
VWB = CWRblue /Y = 10 × ETblue /Y (8)
KC Crop coefficient
δ Nitrogen leaching rate (%)
AR Nitrogen scalar (kg) ETblue = max(0 , ETc − Pe) (9)
Cmax Allowable maximum pollution concentration (mg/L)
Cnat Background concentration (mg/L)
Minimal grey water footprint. This goal is to minimize the grey water
PW Global average production capacity of water resources (m3/ha) demand, and the purpose is to reduce the amount of pollutant dilution
r Global water balance factor water within the allowable range to reduce environmental pollution and
TB Total crop benefit (yuan) save water.
TC Total crop cost (yuan)
Psur Surface water price (yuan/m3) ∑
I ∑
J
Pgro Groundwater price (yuan/ m3) Yij Aij VWRij
CH Nitrogen application cost (yuan/kg) (10)
i=1 j=1
min Grey water =
∑
I ∑
J
Yij Aij VWLij
i=1 j=1

4
M. Li et al. Agricultural Water Management 259 (2022) 107235

δ × AR trol of total nitrogen and total phosphorus emissions. This goal is to

VWRij = (11) minimize the loss of nitrogen and phosphorus flowing into the soil and
(Cmax − Cnat)Yij
rivers, and thus reduce pollution load and non-point source pollution in
farmland. This objective function can be expressed as follows:
2.1.2.2. Economic dimension. Economic benefit from food production
has an important and directly impact on social and economic develop­ I ∑
∑ J

ment. This paper uses the maximum profit to quantify the objective min NPSPsn = Lij Nsi Aij (17)
function of economic dimension. Profit equals to the total revenue minus
i=1 j=1

the total cost (including planting costs and water supply costs). This I ∑
∑ J
objective function can be expressed as follows: min NPSPsp = Lij Psi Aij (18)
i=1 j=1
max Profit = TB − TC (12)
I ∑
∑ J
( )
∑
I ∑
J min NPSPrn = KN Hij + NM × Fij Aij (19)
TB = Yij Pj (13) i=1 j=1
i=1 j=1

( I ∑
∑ J
( )
TC = Aij Cij + CH × Hij + CF × Fij + CM × Mij + CO × Oij + Psur × X sur min NPSPrp = KP Mij + PM × Fij Aij (20)
ij
i=1 j=1
)
gro gro
+ P × X ij Minimum carbon footprint. This goal is to minimize the net green­
(14) house gas emission by subtracting carbon fixation amount from the total
greenhouse gas emissions during the entire growth period of crops
2.1.2.3. Social dimension. Sustainable allocation of AWLR needs to (Zhang et al., 2017). This objective function can be expressed as follows:
consider the fairness and rationality of resources allocation, which can min Carbon footprin t = ECO2 + 298 × EN2 O + 25 × ECH4 − CS (21)
help ensure social stability. This paper uses Gini coefficient to describe
the fairness of resources allocation among sub-regions (Hu et al., 2016). I ∑
∑ J
(
The value range of the Gini coefficient is 0–1, and smaller value of Gini ECO2 = Aij Hij × qh + Fij × qf + Oij × qo + Mij × qm + CEi × qe
coefficient indicates more balanced distribution of AWLR. Therefore, the i=1 j=1
)
objective function of the social dimension is to minimize Gini coefficient
+ CDi × qd + CPi × qp + Si × qs
and it can be expressed as follows:
⃒ (
⃒∑
⃒
J ) (22)
X sur + X gro Alj
1 ∑ ∑⃒⃒j=1 lj
I I lj
( )
min Gini = ∑ sur gro × ⃒ I ∑
∑ J
J
⃒ POl
∑I (Xij +Xij )Aij l=1 k=1 ⃒ EN2 O = Aij NTj + Hij × QHj + Fij × QFj (23)
j=1 ⃒
2I i=1 j=1
(15)
POi
i=1
∑ J ( ) ⃒⃒
sur gro ⃒ ∑
I

j=1
X kj + X kj Akj ⃒
⃒ ECH4 = Ai1 × qr (24)
− ⃒ i=1
POk ⃒
⃒ [ ( ) ]
⃒ ( )
I ∑
∑ J
Aij Yij 1 + λj γj I ∑
∑ J
CS = + c Aij Yij λj σj +d (25)
τj
2.1.2.4. Ecological dimension. Agricultural production tends to change
i=1 j=1 i=1 j=1

ecological condition while consuming AWLR. This paper uses ecological 298 and 25 are the carbon dioxide conversion coefficients of N2O and
pressure index (EPI) to quantify the objective of ecological dimensions. CH4 respectively; c and d are empirical coefficients.
The EPI can be expressed as the ratio of ecological footprint to ecological
carrying capacity (Xiu-Juan and Qin-Yu, 2009), which reflects whether 2.1.3. Constraints
the consumption of AWLR exceeds the carrying capacity of resources The objective function of the above five dimensions is subject to the
and reflects the supply-demand relationship of AWLR. Among them, the following conditions.
ecological footprint of water resources reflects the consumption of water
resources by planting, and the ecological carrying capacity of water 2.1.3.1. Resource dimension. Water constraints: The amount of surface
resources reflects the water resources that the water body can provide. water and groundwater used should be less than the corresponding
The smaller the EPI, the more abundant water resources can be pro­ amount of water availability. This constraint can be expressed as
vided, which is more conducive to ecological protection. Therefore, the follows:
objective function of the ecological dimension is to minimize the EPI,
∑
J
which can be expressed as follows: Aij × ∂ij × X sur sur
(26)
ij ≤ SWi × etai ∀i
∑
I ∑
J j=1
gro
(
Aij X sur
ij
+X ij +Pei )
i=1 j=1
×r ∑
J
min EPI = Pw
(16) Aij × ∂ij × X gro gro
(27)
∑
I
Qi ij ≤ GWi × etai ∀i
0.4r × φi × Pw j=1
i=1
Land constraints: The total planting area should be less than the
2.1.2.5. Environmental dimension. Agricultural production generates maximum arable area. The planting area of each crop in each region
considerable effect on the environment. Two aspects of environment should be larger than the minimum planting area to guarantee food
effects were taken into account: non-point source pollution and green­ production and smaller than the maximum allowed planting area
house gas emissions. considering land carrying capacity. This constraint can be expressed as
Minimum non-point source pollution. The control of non-point follows:
source pollution (NPSP) is embodied in the application amount con­

5
M. Li et al. Agricultural Water Management 259 (2022) 107235

∑
I ∑
J Decision-making is often difficult due to conflicting objectives. The
Aij ≤ TK (28) application of this method makes multivariate optimization feasible to
form a reasonable system that can eliminate the conflict of multiple
i=1 j=1

objectives while satisfying a given budget condition, and is therefore

Amin ≤ Aij ≤ Amax ∀i, j (29)
ij ij
widely used as a multi-objective solution method (de Sousa Xavier et al.,
2015; Fattahi and Fayyaz, 2010; Miao et al., 2021).
2.1.3.2. Social dimension. Food security constraints: The yield of food For a MOP model containing n objective functions, the following
crops should meet the food needs of the regional population. This formula can be used to convert the MOP model into a single-objective
constraint can be expressed as follows: programming model:
∑
J [ ( )k ]1/k
Aij Yij ≥ FD × POi ∀i (30) ∑
N
Z opt
n − Zn (x)
min D(x) = wkn (38)
j=1
n=1 Z opt
n − Zn
inf

2.1.3.3. Ecological dimension. Crop water demand constraint: Irrigation x is the vector of decision variables; n is the subscript of the objective
water should meet the maximum and minimum requirements of crop function; Wnk is the weight of the objective function of n. This study uses
water demand to avoid water waste and to ensure crop growth. This the analytic hierarchy process to define the weight of the objective
constraint can be expressed as follows: function (Srdjevic and Medeiros, 2008); Zn(x) is the objective function
IRij ≤ X sur gro max
(31) value; Znopt, Zninf are respectively the maximum and minimum values of
ij + X ij + Pei ≤ IRij ∀i, j
objective function n; k is a measurement parameter with a value of (1,2,
⋅⋅⋅∞), which reflects the significance of the maximum deviation from the
2.1.3.4. Economic dimension. Minimum economic benefit constraint: ideal point.
The output value of food crops should be greater than economic benefit
index to ensure the economic development of regional agriculture. This 2.3. Changing environment module
constraint can be expressed as follows:
∑
J 2.3.1. Climate change
Aij Yij Pj ≥ EBi ∀i (32) Climate change has aggravated the contradiction between water
j=1
supply and demand. It directly affects water cycle process of farmland,
and thus changes allocation patterns of AWLR. In 2014, the Fifth
2.1.3.5. Environmental dimension. Non-point source pollution con­ Assessment Report (AR5) issued by the Intergovernmental Panel on
straints: Non-point source pollution amount should be less than the Climate Change (IPCC) proposed a new emission scenario model RCPs
maximum allowable total nitrogen loss amount. This constraint can be (Representative Concentration Pathways). RCPs integrate simple
expressed as follows: climate models, atmospheric chemistry, assessment models and global
∑ I ∑ J
) carbon cycle models, and are divided into four emission scenarios ac­
(NM × Fij + Hij − NRij Aij ≤ TN × TK (33) cording to the concentration of different greenhouse gases: Low forcing
i=1 j=1 level (RCP2.6), Moderately stable scene (RCP4.5 and RCP6.0), and High
baseline emission scenario (RCP8.5). This study selects two emission
Maximum allowable soil loss constraints: The amount of farmland
scenarios, RCP4.5 and RCP8.5, which are moderately stable and high
soil loss should be less than the maximum allowable soil loss. This
baseline emission scenarios.
constraint can be expressed as follows:
General Circulation Models (GCMs) are considered to be the best
∑
I ∑
J
available tool for simulating future climate and are widely used to
Aij Lij ≤ SL × TK (34)
predict future scenarios of regional water resources (Malik et al., 2020).
i=1 j=1
The combination of multiple models gives better predictions than one
model alone (Pierce et al., 2008), and the ability of each GCM to
2.1.3.6. Non-negative constraints. All decision variables are positive.
simulate different variables is different. Selection criteria based on the
This constraint can be expressed as follows:
assessment of GCMs’ performance in China and the relative indepen­
X sur
ij ≥ 0 ∀i, j (35) dence and good performance of precipitation temperature simulations
(Knutti et al., 2013; Yan et al., 2015), this paper selects four GCMs that
X gro
ij ≥ 0 ∀i, j (36) are suitable for the study area and are among the more commonly used
in Asian climate change and hydrology studies for climate prediction. As
Aij ≥ 0 ∀i, j (37) shown in Table 2, based on meteorological forecast data (temperature,
wind speed, sunshine duration, etc.) under different representative
concentration pathways (RCPs) from 2020 to 2050, crop water demand
2.2. Solving method can be obtained using Penmen Monteith formula.

There are various solution methods for the multi-objective pro­

gramming model constructed in this paper, such as: the minimum de­
viation method, the affiliation function method, the weight method, the Table 2
constraint method, and the intelligent algorithm. In this paper, a CMIP5 GCMs used in this paper.
compromise programming approach (CP) is used to solve the model
Model Resolution Institution Country
(Zeleny, 1973). In contrast to other decision making methods, CP aims
BCC-CSM1- 128 × 64 Beijing Climate Center, China China
to minimize the distance between the optimal value and the standard
1 Meteorological Administration
value. This distance concept can be used as a measure of human pref­ CCSM4 288 × 192 National Center for Atmospheric Research USA
erence to define the ideal point, and therefore CP is able to determine the IPSL- 96 × 96 Institute Pierre Simon Laplace France
solution that is closest to the ideal solution (Khodaei et al., 2018). The CM5A-LR
main advantage of CP is its ability to determine the closest ideal solution MRI- 320 × 160 Meteorological Research Institute Japan
CGCM3
without any interference from decision makers (Salman et al., 2019).

6
M. Li et al. Agricultural Water Management 259 (2022) 107235

2.3.2. Social and economic development path meteorological elements and population on water demand, so as to
Human activities such as application of chemical fertilizer, pesticide select the best AWLR management policies that are suitable for regions,
and agricultural films, will produce large volumes of greenhouse gases, have practical significant for the sustainable use of AWLR.
which will affect global climate (Hansen et al., 2010; Trenberth et al., In order to assess the uncertainty of water supply and demand in a
2015). In the meanwhile, the path and direction of social and economic changing environment, this study used the popular tool, General Cir­
development will also increase the uncertainty of water resources supply culation Model (GCMs), to simulate future climate (Wang et al., 2017a,
and demand. The Shared Socio-economic Path (SSPs) are reliable tool 2017b), which are proposed in the fifth assessment report of the Inter­
for the Intergovernmental Panel on Climate Change (IPCC) to describe governmental Panel on Climate Change (IPCC AR5). The RCP scenarios
future social development in the context of climate change (Fujimori are combined to predict future climate change. Organically integrate the
et al., 2014). It is proposed by IPCC with a comprehensive consideration climate scenario and the future economic development path of the so­
of population, economy, technological progress, resource utilization and ciety, obtain the water and land resource planning plan under the
other factors. SSPs are social-economic development scenario related to changing environment model, establish a decision-making framework,
greenhouse gas emissions, which can organically combine social and and obtain the optimal socio-economic development policy suitable for
economic development with climate change, environmental change, the authorities.
resource utilization change, etc. Therefore, this study generates 12 scenarios of water supply and
SSPs include five shared social-economic paths: (1) Sustainability demand (4 General Circulation Models (GCM) and 3 changing envi­
path-SSP1 (Sustainability), that is, green development path. This path ronmental models) considering the comprehensive effect of climate
takes sustainable development as the goal, independent of fossil fuels change and socio-economic development paths. The generation of the
supply and has low resource intensity requirements. (2) Middle of the three changing environmental models is based on the combination of
Road-SSP2 (Middle of the Road), which is facing challenges of moderate SSPs and RCPs, including sustainable development plan (SSP1-RCP4.5),
climate change. This path maintains the current social and economic middle path plan (SSP2-RCP4.5), and regional competition plan (SSP3-
development trend, and gradually reduces the dependence on fossil fuels RCP8.5), as shown in Fig. 2. This part aims to make in-depth analysis of
by achieving factional developing goals. (3) Regional competition path- future water shortage based on optimal allocation of AWLR, and thus
SSP3 (Regional Rivalry). This path represents the condition of incon­ provide decision making for sustainable AWLR, especially for vulnerable
sistent development regional development. There are big differences regions with resource shortage.
between regions, and the development is depended heavily on fossil
fuels. (4) The unbalanced path-SSP4 (Inequality), which focuses on 3. Case study
challenges, and the development within and between regions or coun­
tries is highly uneven; (5) Development path that is dominated by fossil 3.1. Study area
fuels-SSP5 (Fossil-fueled Development). This path emphasizes the heavy
use of fossil energy, attaches importance to the improvement of eco­ The study area is selected in the coastal cities of the Songhua River
nomic strength, and faces high social challenges. Among them, SSP1 can basin that flows through Heilongjiang Province. The Songhua River
depict a sustainable future, SSP2 is a development path that is more Basin is one of the seven major river basins in China. It is located in the
consistent with the development patterns observed in the past, and SSP3 central and southern Heilongjiang Province. It has a continental
can provide a reference for the path of pursuing economic development. monsoon climate, with the annual average rainfall is approximately
Paths of SSP4 and SSP5 attach great importance to economic develop­ 500 mm. It is a major food production base of rice, corn and soybeans,
ment so that the social and economic development is extremely unbal­ which were selected as the studied crops in this study. There are 8 cities
anced, which are extreme economic development modules. Excluding in the Songhua River Basin that flow through Heilongjiang Province,
the more extremely unbalanced two economic development models, the namely Harbin, Hegang, Shuangyashan, Yichun, Jiamusi, Qitaihe,
three development paths SSP1, SSP2 and SSP3 can provide a qualitative Mudanjiang, Suihua. The area is 235,300 km2. The main inflow of water
description of the development of a rational globalization of socioeco­ in this area is concentrated in the period from June to September. The
nomic conditions and are widely used in many studies (Fricko et al., study area experiences with a long frost period and faces water shortage
2017; Greve et al., 2018). Combining with representative concentration problem. Therefore, it is necessary to rationally allocate AWRL in the
pathways (Kriegler et al., 2012), this study selects three economic area to alleviate water shortage and thus promote agricultural sustain­
development models of SSP1, SSP2 and SSP3. Based on the three eco­ ability. Fig. 3 describes the study area.
nomic development models, the population-development-environment
analysis (PDE) model is used to predict the population of the study area. 3.2. Data collection

2.3.3. Coupled module In order to predict future supply-demand ratio of regional water
Climate change will affect meteorological elements such as temper­ resources, population and water demand in 2020–2050 were firstly
ature, duration of sunshine, and wind speed, and further affect water forecast. Population and GDP data come from the official SSP database.
demand of crops. The path of socio-economic development affects Water demand prediction was obtained based on crop evapotranspira­
population, GDP, etc., and hence affects regional water demand. tion which can be calculated by multiplying reference evapotranspira­
Combining climate change with socio-economic development paths, and tion ET0 with crop coefficient Kc. The acquisition of ET0 relied on
comprehensively considering and predicting the impact of meteorological data (average minimum temperature, average maximum

Fig. 2. The establishment of three change environment model schemes.

7
M. Li et al. Agricultural Water Management 259 (2022) 107235

Fig. 3. Study area.

temperature, average precipitation, average wind speed and sunshine distribution (Gu et al., 2008). According to the Class III standard in the
hours) which were from China Meteorological Data Service Center Chinese Groundwater Quality Standard (GB/T 14848-93), the allowable
(CMDC). ET0 was calculated by the Penman-Monteith formula recom­ maximum pollution concentration was set as Cmax = 20 mg/L. The
mended by the Food and Agriculture Organization (FAO) (Allan et al., background concentration was set Cnat = 0. The leaching rate δ of ni­
1998), and it was expressed as follows: trogen fertilizer was 10%. ETgreen is calculated by the CROPWAT model
of the Food and Agriculture Organization of the United Nations (Sun
0.408Δ(Rn − G) + γ 900(e a − ed )
et al., 2013).
ET0 = T+273
(39)
Δ + γ(1 + 0.34U2 ) The water balance factor r is equal to the global average ecological
productivity of land for water divided by the global average ecological
Δ is the slope related to the temperature curve and saturated water productivity of all types of ecologically productive land area, taken as
vapor pressure (kPa ◦ /C); Rn is the net radiation amount of the plant 5.19. The global average production capacity of water resources Pw is
surface (MJm− 2 day− 1); G is the heat flux of the soil (MJ/m2.d); γ is the taken as 3140 m3/ha. The 0.4 in the denominator of the ecological
common quantity of dry humidity (kPa ◦ C− 1); ea is the saturated vapor dimension objective function is the maximum use factor of the regional
pressure (KPa); ed is the actual observed vapor pressure (KPa); T is the water resources set for sustainable development.
average air temperature (◦ C); U2 is 2 m wind speed at height (m/s). Effective rainfall Pe was estimated based on FAO CROPWAT model
Table S1 in the supplementary material and Fig. 4 below show the that was expressed as follows:
forecast data of population and crop water demand. Table 3 shows the ( )
⎧
relevant parameters of each partition in the model. ⎨ P × 4.17 − 0.02 × P , P < 83
Regional water supply data and regional planting areas were all Pe = 4.17 (40)
⎩
derived from the Heilongjiang Province Water Resources Bulletin (41.7 + 0.1 × P) , P ≥ 83
(2000–2017), the Heilongjiang statistical almanacks, and field in­
vestigations. Table 3 shows relevant basic parameters of the model. This P is the amount of rainfall (mm).
study simulate runoff using selection method based on Person-III

Fig. 4. Crop water demand forecast data. Note: The numbers 1, 2, and 3 after the short form of region’s name represent the three time periods of 2020–2030,
2031–2040 and 2041–2050, respectively;.

8
M. Li et al. Agricultural Water Management 259 (2022) 107235

Table 3
Parameters for different subareas and crops.
Area Surface water Groundwater supply Utilization coefficient of Utilization coefficient of Water production Yield (kg/ha)
supply (104 m3) (104 m3) surface water irrigation groundwater irrigation factor (m3/ha)
Rice Corn Soybean

HRB 417,441 170,376 0.43 0.63 212,044 8186 8770 2204

HG 48,949 121,929 0.45 0.73 241,022 5910 5583 1352
SYS 91,285 132,338 0.42 0.69 149,533 7440 7913 1908
YC 28,060 24,964 0.43 0.73 253,233 7568 6103 1837
JMS 187,305 402,345 0.43 0.72 138,000 7539 7861 1943
QTH 34,703 13,425 0.43 0.73 115,650 6748 6299 1874
MDJ 138,345 26,346 0.40 0.73 228,750 7274 7183 2019
SH 173,726 144,234 0.41 0.75 131,072 8135 8660 2159

Note: HRB-Haibin; HG-Hegang; SYS-Shuangyashan; YC-Yichun; JMS-Jiamusi; QTH-Qitaihe; MDJ-Mudanjiang; SH-Suihua.

3.3. Results analysis

3.3.1. Agricultural water resource allocation schemes

According to the predicted water supply and demand in different
regions, the optimal water allocation under the three changing envi­
ronmental modes were obtained, as shown in Fig. 5. It can be seen from
the figure that as the development path was changed, the changes of
water allocation follows the law: regional competition path> interme­
diate development path> sustainable development path. This indicates
that the demand for water for regional economic development
increased, thus, the corresponding water allocation would increase.
From the results of use proportion of different sources of water re­
sources, surface water was given priority. The total allocation amount of
groundwater under sustainable, intermediate path and regional
competition modules were 2758 million m3, 3565 million m3, and 3801
million m3, respectively. It can be seen that in the path of sustainable
development, surface water supply could already meet most water re­
quirements in different regions. With the requirements of development,
the increase in water demand requires an increase in the distribution of
groundwater.

3.3.2. Equilibrium analysis

Considering the interaction of multi-objective of resources, eco­
nomic, social, ecological and environmental dimensions, the equilib­
rium and merits of allocation schemes of AWLR under different
changing environment modes were obtained, as shown in Fig. 6. The
area size of the graphs under each boundary reflected the equilibrium of
the optimal allocation plans of AWLR, while the satisfaction degree of
objectives in each dimension reflected the metrics of the allocation
plans. It can be seen from the figure that during 2020–2030, the equi­
librium and merits of AWLR allocation plans under the middle path were
better than the other two paths. From 2031–2040, the equilibrium and
merits of AWLR allocation plans under middle path mode and under
regional competition mode were similar, and the plans under sustain­
able development mode was better than the other modes. From
2041–2050, the equilibrium of the sustainable development mode was
better than that of the other two modes. However, in comparison, AWLR
allocation plans under sustainable development mode corresponded to
the smallest economic benefit, while the economic benefit under
regional competition path was the largest. However, the ecological and
environmental effect under regional competition were too small to meet
the goal of balanced development. Therefore, AWLR allocation plans
under the middle path was more optimal plans that could be adopted.

3.3.3. Economic benefit Fig. 5. Regional water distributions under different changing environmental
Ecological benefit was usually the most notable targets. The value modes. Note: SR, SC and SS represent -surface water allocation for rice, corn
and changing trend of economic benefit were a key indicator to deter­ and soybean, respectively (106 m3); soybean (106 m3); GR, GC and GS represent
mine regional usage of AWLR. Fig. 7 shows the changing trends of -ground water allocation for rice, corn and soybean, respectively (106 m3).
economic benefit under changing environment mode of IPSL-CM5A-LR.
Based on the figure, it can be seen that in 2020–2030 and 2031–2040,
the economic benefits in the period of 2020–2030 and 2031–2040 were

9
M. Li et al. Agricultural Water Management 259 (2022) 107235

Fig. 6. Equilibrium of multi-dimensional objectives. Note: The coordinate values represent the satisfaction degree of each dimension.

Fig. 7. Changing trend of economic benefit under changing environment modes.

the largest under the regional competition mode. The economic benefit
under the intermediate development mode during 2031–2040 was
slightly larger than that under the regional competition mode. The
economic benefit under the sustainable development mode during
2020–2030 was greater than that under the middle path mode, while the
corresponding values under the remaining periods were all the smallest.
Therefore, from an overall point of view, the economic target value
experienced a changing trend law of regional competition path> inter­
mediate development path> sustainable development path. In Fig. 7
(changing trends of economic benefit under CCSM4), the economic
benefit during 2041–2050 had the following rules of intermediate
development path> regional competition path> sustainable develop­
ment path. Hence, it can be found that the economic benefit might not
achieve its optimum value for the sustainable use of AWLR, although
economic benefit was prioritized. From the perspective of sustainable
development, it was particularly important to ensure sustainable and
comprehensive economic development when making strategic decisions
on the use of AWLR. In addition, the economic target values under the
three changing environmental modes in the three periods did not reach Fig. 8. Correlation coefficient of target values under different dimensions of
the corresponding ideal value. This attributed to the balance of the other resources, economy, society, ecology and environment.
four dimensions, as shown in Fig. 8. The figure represents the correlation
coefficients and provides a qualitative and quantitative analysis of the 3.3.4. Water security
interactions between different dimensions. The figure shows that the The optimal allocation plans of AWLR should be based on regional
environmental, resource, social and ecological dimensions were nega­ water security. The water supply-demand ratio is an important indicator
tively correlated with the economic dimension. Among them, the envi­ that can reflect water security, which was adopted in this study to
ronment, resource and ecological dimensions had the largest negative evaluate the adaptation and metrics of different allocation schemes of
correlation with the economic dimension, which indicates that the AWLR, as shown in Fig. 9. The shades of the colors in the picture indicate
guarantee of the effect of the other dimensions would constrain the the security of the allocation schemes of AWLR.
economic development to some extent. The other dimensions, except the In the period 2020–2030, for some regions where the initial water
economic dimension, were positively correlated with each other. These supply-demand ratio was poor, such as Hegang City and Mudanjiang
show that environmental, resource, social and ecological dimensions City, the optimal water supply-demand ratio under the sustainable
had a mutually reinforcing and co-fulfilling relationship. This study took development model was improved corresponding to the optimal water
sustainability as the goal to make trade-offs between different objectives security condition. Some regions with good initial water supply-demand
in different dimensions to obtain the optimum allocation plans of AWLR. ratios, such as Harbin, Shuangyashan, Yichun and Jiamusi, were rec­
The economic benefit tended to be as large as possible at the premise of ommended to choose AWLR allocation plans under regional competition
seeking the balance of other targets in the rest four dimensions. mode at the initial stage to attach importance on economic

10
M. Li et al. Agricultural Water Management 259 (2022) 107235

Fig. 9. Water supply-demand conditions under different change scenarios. Note: SD-sustainable development mode; MR-middle path mode; RC-regional competition
mode; The ordinate value is the value of water supply-demand ratio.

development. In the middle period, i.e. from 2031 to 2040, most regions security condition of Harbin would present the best under the scenario
performed well in water security under the middle path mode and mode of the middle path.
regional competition mode, and Hegang City could still maintain a
sustainable development mode. During the period 2041–2050, most
regions had the optimum water security condition under the middle 3.4. Results options
path mode, and water security in Hegang City was the best.
Take Harbin (the provincial capital of Heilongjiang Province) as an This study used the analytic hierarchy process to determine the
example, the changes of water supply-demand ratio was shown in objective weights of the five dimensions of resources, ecology, society,
Fig. 10. From the initial state to 2031, Harbin had the best water supply economy, and environment under different changing environmental
and demand situation under the scenario of regional competition, but modes. Table 4 can inform decision makers on the determination of
afterwards it shows a downward trend until the lowest under the three dimensional focus in a changing environment.
scenarios. Water supply and demand condition under sustainable sce­ Based on objective weights, the optimal solutions of the multi-
nario was always in a relatively stable state, while the water supply and dimensional model under different changing environment modes
demand condition under the middle path presents a state of gradual could be solved. Table 5 provides the AWLR adaptation plans for each
ascent and then remaining stable. Therefore, after 2031, the water region at different future time periods based on the balance and
comprehensiveness of each dimension and the combined consideration
of water security.
Table 6 shows the ratio of water supply to water demand at this stage
in each region, which gives an indication of the initial supply and de­
mand situation in each region. The table shows that the areas with good
initial supply and demand are: Jiamusi City, and Yichun City. In
contrast, Hegang City and Mudanjiang City had a poor initial supply to
demand situation, with the value of supply-demand ratio both being less
than 0.5. It can be seen from the table that based on the overall equi­
librium, metrics of economic benefits and water security, allocation
schemes of AWLR of Hegang and Mudanjiang cities where the initial
water supply-demand situations were poor, were recommended to adopt
sustainable development mode in the initial time period. For cities with
Fig. 10. Water supply and demand ratio of Harbin. good initial water supply-demand conditions, this study suggests the

11
M. Li et al. Agricultural Water Management 259 (2022) 107235

Table 4
Objective weights.
Development path Resource dimension Economic dimension Social dimension Ecological dimension Environmental dimension

Sustainable development 0.2 0.2 0.2 0.2 0.2

Middle 0.1867 0.3301 0.1097 0.1867 0.1867
Regional competition 0.1771 0.4001 0.0951 0.1639 0.1639

Table 5
Water and land resources management plan.
Region 2020–2030 2031–2040 2041–2050

Sus Mid Com Sus Mid Com Sus Mid Com

Harbin √ √ √
Hegang √ √ √
Shuangyashan √ √ √
Yichun √ √ √
Jiamusi √ √ √
Qitaihe √ √ √
Mudanjiang √ √ √
Suihua √ √ √

Note: Sus-Sustainable; Mid-Middle; Com-Compete.

balanced. Also for each dimension, there were more than one objectives
Table 6 to reflect characteristics of each dimension more comprehensive.
Initial water supply-demand situation.
Therefore, targets of irrigation water use efficiency, water saving, eco­
Region Supply-demand ratio nomic benefit, resources allocation equity, ecological stresses, non-point
Harbin 0.5269 pollution and greenhouse gas emission were optimized synergistically,
Hegang 0.4454 which would contribute to the management of agricultural water and
Shuangyashan 0.5444 land resources from a holistic view. (2) Agricultural water and land
Yichun 0.5691
Jiamusi 0.6640
allocation plans were generated under complex changing environment.
Qitaihe 0.5145 More specifically, the effect of both natural changes and human activ­
Mudanjiang 0.4919 ities on agricultural water and land use were considered. Natural
Suihua 0.5163 changes were quantified by climate changes modes (combination of
different GCM and RCP scenarios), and human activities were quantified
by socio-economic development paths. This will help provide decision-
AWLR allocation plans to adopt a competition-competition-intermediate
makers crisis response to changing environment when planning agri­
mode within 2020–2050. During 2041–2050, water security in all the
cultural water and land resources.
studied eight areas presented the optimum condition in the middle path
The feasibility and applicability of the developed model framework
scenario. For Qitaihe City and Suihua City, although their initial supply-
was demonstrated by applying it in the Songhua River Basin in Hei­
demand conditions were not poor, there was still a certain gap between
longjiang Province, northeast China. Agricultural water and land re­
them and the optimum water supply-demand situation. Therefore, this
sources allocation schemes of different crops in different regions under
study suggested Qitaihe City and Suihua City to adopt a middle-
different climate change and socio-economic development path changes
competition-middle management plan. Accordingly, considering initial
scenarios were obtained. The corresponding changing trend of resources
water supply-demand conditions, decision makers could choose the
use efficiency, economic benefit, social stability, ecological presses and
optimum AWLR allocation plans based on the developed multi-
pollution were in-depth analyzed. The optimum agricultural water and
dimensional optimization model framework.
land resources allocation plans were also generated based on the
comprehensive consideration of system’s comprehensiveness, equilib­
4. Conclusion
rium and security. The model framework can be applied to other regions
with water scarcity. Future work will focus on the dynamics of the
This paper constructed a model framework for the comprehensive
model.
allocation of agricultural water and land resources in a changing envi­
ronment. The model framework includes: (1) development of a multi-
dimensional optimization model for simultaneous allocation of agri­ Declaration of Competing Interest
cultural water and land resources, which contains objectives involving
natural resources, economic, social, ecological and environmental di­ The authors declare that they have no known competing financial
mensions, and constrains corresponding to each dimension. The interests or personal relationships that could have appeared to influence
compromised planning method was adopted to solve the multi-objective the work reported in this paper.
programming model; (2) quantification of changing environment
modes, combing climate change scenarios and socio-economic path
Acknowledgements
scenarios.
Compared with previous studies that were focused on the manage­
This research was supported by funds from the National Natural
ment of agricultural water and land resources, the main innovations and
Science Foundation of China (52079029), the National Science Fund for
contributions of this study mainly reflected in: (1) Agricultural water
Distinguished Yong Scholars of China (51825901), Yong Innovative
and land resources can be optimized simultaneously, considering their
Talents Training Programming for Universities of Heilongjiang Province
interactions. In the meanwhile, conflicts between different dimensions
(UNPYSCT-2020108), and the Key Laboratory Program of Ministry of
of resources, economy, society, ecology and environment were
Agriculture and Rural Affairs of the People′ s Republic of China

12
M. Li et al. Agricultural Water Management 259 (2022) 107235

(AWR2021001). Malik, S., Pal, S.C., Sattar, A., Singh, S.K., Das, B., Chakrabortty, R., Mohammad, P.,
2020. Trend of extreme rainfall events using suitable Global Circulation Model to
combat the water logging condition in Kolkata metropolitan area. Urban Clim. 32,
Appendix A. Supporting information 100599.
Miao, P., Yue, Z., Niu, T., Alizadeh, Aa, Jermsittiparsert, K., 2021. Optimal emission
Supplementary data associated with this article can be found in the management of photovoltaic and wind generation based energy hub system using
compromise programming. J. Clean. Prod. 281, 124333.
online version at doi:10.1016/j.agwat.2021.107235. Pierce, D.W., Barnett, T.P., Hidalgo, H.G., Das, T., Bonfils, C., Santer, B.D., Bala, G.,
Dettinger, M.D., Cayan, D.R., Mirin, A.J.J.o.C., 2008. Attribution of Declining
References Western US Snowpack to Human Effects. 21(23), 6425–6444.
Ren, C., Li, Z., Zhang, H., 2019. Integrated multi-objective stochastic fuzzy programming
and AHP method for agricultural water and land optimization allocation under
Allan, R., Pereira, L., Smith, M., 1998. Crop Evapotranspiration-Guidelines for multiple uncertainties. J. Clean. Prod. 210, 12–24.
Computing Crop Water Requirements-FAO Irrigation and Drainage Paper 56, 56. Salman, S.A., Shahid, S., Ismail, T., Al-Abadi, A.M., Wang, X.-j, Chung, E.-S., 2019.
Dai, Z.Y., Li, Y.P., 2013. A multistage irrigation water allocation model for agricultural Selection of gridded precipitation data for Iraq using compromise programming.
land-use planning under uncertainty. Agric. Water Manag. 129, 69–79. Measurement 132, 87–98.
de Sousa Xavier, A.M., Costa Freitas, Md.B., de Sousa Fragoso, R.M., 2015. Management Schewe, J., Heinke, J., Gerten, D., Haddeland, I., Arnell, N.W., Clark, D.B., Dankers, R.,
of Mediterranean forests — a compromise programming approach considering Eisner, S., Fekete, B.M., Colón-González, F.J., Gosling, S.N., Kim, H., Liu, X.,
different stakeholders and different objectives. For. Policy Econ. 57, 38–46. Masaki, Y., Portmann, F.T., Satoh, Y., Stacke, T., Tang, Q., Wada, Y., Wisser, D.,
Fan, X., Zhang, W., Chen, W., Chen, B., 2020. Land–water–energy nexus in agricultural Albrecht, T., Frieler, K., Piontek, F., Warszawski, L., Kabat, P., 2014. Multimodel
management for greenhouse gas mitigation. Appl. Energy 265, 114796. assessment of water scarcity under climate change. Proc. Natl. Acad. Sci. USA 111
Fattahi, P., Fayyaz, S., 2010. A compromise programming model to integrated urban (9), 3245–3250.
water management. Water Resour. Manag. 24 (6), 1211–1227. Srdjevic, B., Medeiros, Y., 2008. Fuzzy AHP assessment of water management plans.
Fricko, O., Havlik, P., Rogelj, J., Klimont, Z., Gusti, M., Johnson, N., Kolp, P., Water Resour. Manag. 22, 877–894.
Strubegger, M., Valin, H., Amann, M., Ermolieva, T., Forsell, N., Herrero, M., Sun, S.K., Wu, P.T., Wang, Y.B., Zhao, X.N., 2013. The virtual water content of major
Heyes, C., Kindermann, G., Krey, V., McCollum, D.L., Obersteiner, M., Pachauri, S., grain crops and virtual water flows between regions in China. J. Sci. Food Agric. 93,
Rao, S., Schmid, E., Schoepp, W., Riahi, K., 2017. The marker quantification of the 1427–1437.
Shared Socioeconomic Pathway 2: a middle-of-the-road scenario for the 21st Tang, Y., Zhang, F., Engel, B.A., Liu, X., Yue, Q., Guo, P., 2020. Grid-scale agricultural
century. Glob. Environ. Change 42, 251–267. land and water management: a remote-sensing-based multiobjective approach.
Fujimori, S., Kainuma, M., Masui, T., Hasegawa, T., Dai, H., 2014. The effectiveness of J. Clean. Prod. 265, 121792.
energy service demand reduction: a scenario analysis of global climate change Tang, Y., Zhang, F., Wang, S., Zhang, X., Guo, S., Guo, P., 2019. A distributed interval
mitigation. Energy Policy 75, 379–391. nonlinear multiobjective programming approach for optimal irrigation water
Gong, X., Zhang, H., Ren, C., Sun, D., Yang, J., 2020. Optimization allocation of management in an arid area. Agric. Water Manag. 220, 13–26.
irrigation water resources based on crop water requirement under considering Trenberth, K., Fasullo, J., Shepherd, T., 2015. Attribution of climate extreme events. Nat.
effective precipitation and uncertainty. Agric. Water Manag. 239, 106264. Clim. Change 5, 725–730.
Greve, P., Kahil, T., Mochizuki, J., Schinko, T., Satoh, Y., Burek, P., Fischer, G., Vanham, D., Hoekstra, A.Y., Wada, Y., Bouraoui, F., de Roo, A., Mekonnen, M.M., van de
Tramberend, S., Burtscher, R., Langan, S., Wada, Y., 2018. Global assessment of Bund, W.J., Batelaan, O., Pavelic, P., Bastiaanssen, W., Kummu, M., Rockström, J.,
water challenges under uncertainty in water scarcity projections. Nat. Sustain. 1 (9), Liu, J., Bisselink, B., Ronco, P., Pistocchi, A., Bidoglio, G., 2018. Physical water
486–494. scarcity metrics for monitoring progress towards SDG target 6.4: an evaluation of
Greve, P., Seneviratne, S.I., 2015. Assessment of future changes in water availability and indicator 6.4.2 “Level of water stress”. Sci. Total Environ. 613, 218–232 (in Chinese
aridity. Geophys. Res. Lett. 42 (13), 5493–5499. with English abstract).
Gu, W., Shao, D.G., Huang, X.F., Dai, T., 2008. Multi-objective risk assessment on water Wada, Y., van Beek, L.P.H., Bierkens, M.F.P., 2011. Modelling global water stress of the
resources optimal deployment. J. Hydraul. Eng. 3, 339–345. recent past: on the relative importance of trends in water demand and climate
Gudmundsson, L., Seneviratne, S.I., Zhang, X., 2017. Anthropogenic climate change variability. Hydrol. Earth Syst. Sci. 15 (12), 3785–3808.
detected in European renewable freshwater resources. Nat. Clim. Change 7 (11), Wada, Y., van Beek, L.P.H., Wanders, N., Bierkens, M.F.P., 2013. Human water
813–816, 813-+. consumption intensifies hydrological drought worldwide. Environ. Res. Lett. 8 (3),
Hansen, J., Ruedy, R., Sato, M., Lo, K., 2010. Global surface temperature change. Rev. 034036.
Geophys. 48 (4), RG4004. Wang, G.-q, Zhang, J.-y, Xu, Y.-p, Bao, Z.-x, Yang, X.-y, 2017a. Estimation of future water
Hoekstra AY, Hung PQ, 2002. Virtual water trade: a quantification of virtual water flows resources of Xiangjiang River Basin with VIC model under multiple climate
between nations in relation to international crop trade. Value of Water Research scenarios. Water Sci. Eng. 10 (2), 87–96.
Report Series No. 11. UNESCO-IHE, Delft. Wang, Y.Y., Huang, G.H., Wang, S., 2017b. CVaR-based factorial stochastic optimization
Hu, Z., Wei, C., Yao, L., Li, L., Li, C., 2016. A multi-objective optimization model with of water resources systems with correlated uncertainties. Stoch. Environ. Res. Risk
conditional value-at-risk constraints for water allocation equality. J. Hydrol. 542, Assess. 31 (6), 1543–1553.
330–342. Xie, Y.L., Xia, D.X., Ji, L., Huang, G.H., 2018. An inexact stochastic-fuzzy optimization
Huang, J., Yu, H., Dai, A., Wei, Y., Kang, L., 2017. Drylands face potential threat under model for agricultural water allocation and land resources utilization management
2 ◦ C global warming target. Nat. Clim. Change 7, 417–422. under considering effective rainfall. Ecol. Indic. 92, 301–311.
Khodaei, H., Hajiali, M., Darvishan, A., Sepehr, M., Ghadimi, N., 2018. Fuzzy-based heat Xiu-Juan, T.A.N., Qin-Yu, Z., 2009. Dynamic analysis and forecast of water resources
and power hub models for cost-emission operation of an industrial consumer using ecological footprint in China. Acta Ecol. Sin. 29 (7), 3559–3568.
compromise programming. Appl. Therm. Eng. 137, 395–405. Yan, D., Werners, S.E., Ludwig, F., Huang, H.Q.J.J.o.H.R.S., 2015. Hydrological response
Knutti, R., Masson, D., Gettelman, A., 2013. Climate model genealogy: generation CMIP5 to climate change: the Pearl River. China under different RCP scenarios, 4, 228–245.
and how we got there. Geophys. Res. Lett. 40 (6), 1194–1199. Yang, G., Li, X., Huo, L., Liu, Q., 2020. A solving approach for fuzzy multi-objective
Kriegler, E., O’Neill, B.C., Hallegatte, S., Kram, T., Lempert, R.J., Moss, R.H., linear fractional programming and application to an agricultural planting structure
Wilbanks, T., 2012. The need for and use of socio-economic scenarios for climate optimization problem. Chaos Solitons Fractals 141, 110352.
change analysis: A new approach based on shared socio-economic pathways. Glob. Zeleny, M., 1973. Compromise programming. Multiple Criteria Decision Making.
Environ. Change 22 (4), 807–822. University of South Carolina Press, Columbia, pp. 263–301.
Li, M., Fu, Q., Singh, V.P., Liu, D., 2018. An interval multi-objective programming model Zeng, X., Chen, C., Liu, A., Wei, H., Zhang, H., Huang, G., Wu, Y., 2018. Planning a
for irrigation water allocation under uncertainty. Agric. Water Manag. 196, 24–36. sustainable regional irrigated production and forest protection under land and water
Li, M., Fu, Q., Singh, V.P., Liu, D., Li, T., Zhou, Y., 2020. Managing agricultural water and stresses with multiple uncertainties. J. Clean. Prod. 188, 751–762.
land resources with tradeoff between economic, environmental, and social Zhang, D., Shen, J., Zhang, F., Li, Ye, Zhang, W., 2017. Carbon footprint of grain
considerations: A multi-objective non-linear optimization model under uncertainty. production in China. Sci. Rep. 7, 4126.
Agric. Syst. 178, 102685.
Linker, R., 2020. Unified framework for model-based optimal allocation of crop areas
and water. Agric. Water Manag. 228, 105859.

13