water research 44 (2010) 24992516

Available at www.sciencedirect.com

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

Game theory based models to analyze water conflicts in the

Middle Route of the South-to-North Water Transfer Project
in China
Shouke Wei a,b,*, Hong Yang a,1, Karim Abbaspour a,1, Jamshid Mousavi a,c,1,
Albrecht Gnauck b
a

Department System Analysis, Integrated Assessment and Modelling, the Swiss Federal Institute of Aquatic Science and Technology
(EAWAG), Ueberlandstrasse 133, CH-8600 Dubendorf, Switzerland
b
Department of Ecosystem and Environmental Informatics, Brandenburg University of Technology, Konrad-Wachsman-Allee 1,
D 03046 Cottbus, Germany
c
Department of Civil Engineering, Amirkabir University of Technology (Tehran Polytechnic), 424 Hafez Avenue, Tehran, Iran

article info

abstract

Article history:

This study applied game theory based models to analyze and solve water conflicts con-

Received 13 August 2009

cerning water allocation and nitrogen reduction in the Middle Route of the South-to-North

Received in revised form

Water Transfer Project in China. The game simulation comprised two levels, including one

10 December 2009

main game with five players and four sub-games with each containing three sub-players. We

Accepted 24 January 2010

used statistical and econometric regression methods to formulate payoff functions of the

Available online 1 February 2010

players, economic valuation methods (EVMs) to transform non-monetary value into

economic one, cost-benefit Analysis (CBA) to compare the game outcomes, and scenario

Keywords:

analysis to investigate the future uncertainties. The validity of game simulation was evalu-

Game theory

ated by comparing predictions with observations. The main results proved that cooperation

Water conflicts

would make the players collectively better off, though some player would face losses.

Economic valuation

However, players were not willing to cooperate, which would result in a prisoners dilemma.

Scenario analysis

Scenarios simulation results displayed that players in water scare area could not solve its

The water transfer

severe water deficit problem without cooperation with other players even under an opti-

China

mistic scenario, while the uncertainty of cooperation would come from the main polluters.
The results suggest a need to design a mechanism to reduce the risk of losses of those players
by a side payment, which provides them with economic incentives to cooperate.
2010 Elsevier Ltd. All rights reserved.

1.

Introduction

From an economic perspective, water resources are composite

assets which provide a variety of services for consumptive and
productive activities. However, water quality degradation has

been an important cause of water scarcity in countries (Wang

et al., 2003; Wei and Gnauck, 2007a). Water resources
management on those problems is usually involved with
interactive and interdependent stakeholders with contradictory or conflicting interests (Fang et al., 1998, 2002; Van der

* Corresponding author at: Department System Analysis, Integrated Assessment and Modelling, the Swiss Federal Institute of Aquatic
Science and Technology (EAWAG), Ueberlandstrasse 133, CH-8600 Dubendorf, Switzerland. Tel.: 41 44 823 5568; fax: 41 44 823 5375.
E-mail addresses: shouke.wei@eawag.ch (S. Wei), hong.yang@eawag.ch (H. Yang), karim.abbaspour@eawag.ch (K. Abbaspour),
jmosavi@aut.ac.ir (J. Mousavi), Albrecht.gnauck@ti-cottbus.de (A. Gnauck).
1
Tel.: 41 44 823 5568; fax: 41 44 823 5375.
0043-1354/$ see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.watres.2010.01.021

2500

water research 44 (2010) 24992516

Veeren and Tol, 2003), goals and strategies (Wei and Gnauck,
2007a). Pollutant discharge is an essential but complex issue
in water resources management, and this complexity is not
only from intricate biochemical processes, but also from
different pollutant sources and multi-polluters with conflicting aims. Water quality and quantity conflicts are usually
caused by (1) water scarcity due to uneven precipitation, (2)
multiple users and pollutant sources discharging waste into
water, (3) different degrees of upstream pollutions restricting
the water use in downstream catchment, and (4) interbasin
water transfer breaking the long-established balance of water
quality and quantity in a basin.
To solve water conflicts cause by water scarcity, Donevska
et al. (2009) proposed some engineering solutions in terms of
reducing water losses, increasing water use efficiency and
waste water recycling, water conservation, and water transfer, and some other non-engineering measures. However,
methods of using water efficiency and waste water recycling
are not so sufficient to the regions facing extreme water
shortage. In addition, interbasin water diversion involves
a multidisciplinary problem (Yevjevich, 2001), which usually
brings about fundamental issues and conflicts concerning
socio-economical, environ-ecological, administrative and
legislative problems (Shao and Wang, 2003; Yang and
Zehnder, 2005). Besides, different economic and political
instruments have been widely used to solve water use
conflicts (Dinar and Howitt, 1997; Wang et al., 2003). Water
markets approach is one cited frequently in the literature
(Burness and Quirk, 1979; Howe et al., 1986; Colby, 1990; Green
and OConnor, 2001; Bhaduri and Barbier, 2003). Water market
methods can provide water users with incentives to allocate
water and reduce pollutants discharge efficiently, and such
market really exists in some countries, such as Australia
(Pegram et al., 1992), California (Howe and Goodman, 1995),
Chile (Hearne and Easter, 1995), India (Saleth, 1996), and Spain
(Reidinger, 1994), etc. However, water market requires
defining the original water rights, creating institutional and
legal mechanisms, and establishing basic infrastructures for
water trade (Holden and Thobani, 1996; Wang et al., 2003)
before it can operate well. Waste discharge is a public bad, and
every polluter can free-ride others achievement of treatment
(Wei and Gnauck, 2007b). Free-riding problem will cause
market failure. In the absent of market and property right,
conflicts between multi-stakeholders competing for water
uses are unavoidable (Pethig, 1992; Wei and Gnauck, 2007a).
There are rare water markets in reality and they are not real
free market (Dellapenna, 2000). Those economic and political
based water conflict solving methods can be summarized into
two classes, direct regulations and economic instruments
(OECD, 1989; Markandya and Recharddson, 1992; Wei and
Gnauck, 2007a). Direct regulation is also known as the
command and control strategies, which usually include
limitation quotas, standards, laws, etc. Economic and political
instruments make use of market mechanism, price incentives, water rights, subsidies, compensation, tradable permits,
green taxations, etc. However, environmental resource problems and its interrelationships with economic activities and
the dynamic ecosystem are very complex and cannot be
solved with simple policy tools (Carraro and Filar, 1995).
Command and control strategies usually lack incentive,

because it is mainly in virtue of legislation, power or force. Wei

and Gnauck (2007a) stated that the existing economic and
regulation instruments do not work so well in solving these
conflicts. From a technical strategy point of view, multiobjective optimization models have been used early to maximize the overall benefit in order to solve transboundary water
conflict in a river basin (Zeng et al., 2001; Yang and Zeng, 2004). In
recent years, more advanced and popular multi-objective
evolutionary algorithms have been used to solve conflicts
objectives in watershed (Bekele and Nicklow, 2005; Muleta and
Nicklow, 2005). In general, however, those optimization
measures neglect the real interests and benefits of the stakeholders in the basin, though they can capture the multiple
optimal solutions, sometimes called as Pareto optimal solutions.
Game theory is an appropriate approach to model and
solve such water conflicts. It was launched by John von Neumann, a great mathematician, and Oskar Morgenstern in
1944. Game theoretical modelling concepts and reasoning
have been widely applied in economic, commercial, social,
political, biological, and other sciences to help people analyze
social and behavioural phenomena. However, the applications of game theory to solve conflicts in water resources
management are comparatively few. As for this topic, it was
originally applied into the cost distribution in joint water
resource projects, i.e. waste water treatment and disposal
facilities (Giglio and Wrightington, 1972; Dinar and Howitt,
1997). Bogardi and Szidarovszky (1976) introduced possible
application of game theory, especially the oligopol game, in
four main areas of water management, and offered solutions
for the typical problems in decision analysis. Lewandowski
(1979) used a game-theoretic approach to model the behaviour of water users in a quality control problem, and he has
proposed a game-theoretic solution to different uses of
a water system. Coppola and Szidarovszky (2004) designed
a two-person conflicting game to analyze the optimal tradeoff between water supply and contamination risk for
a municipal wellfield. Salazar et al. (2007) described the
application of conflict resolution methods to a two-person
conflicts game in groundwater management. Besides, game
theory is regularly used to analyze equitable allocation of
waste loads to a common receiving medium (Kilgour et al.
1988; Okada and Mikami 1992; Wei and Gnauck, 2007b,c). It
was also applied to solve water allocation and pollution
problems in transboundary river, including inter-country river
(Frisvold and Caswell, 2000; Van der Veeren and Tol, 2003) and
intra-country river (Zeng and Yang, 2004; Yang and Zeng,
2004). In game theory, the development of conflict concepts
and methods has received increasing attention since the
pioneering work of Nash (1950). Based on the axiomatic
approaches of Nash, many modified and extended solutions
have been introduced. Among those solutions, four particular
methods have frequently citied and used in water management in the literature (Coppola and Szidarovszky, 2004; Salazar et al., 2007), including: (1) non-symmetric Nash solution of
Harsanyi and Selten (1972); (2) non-symmetric solution of
Kalai and Smorodinsky (1975); (3) non-symmetric area
monotonic solution of Anbarci (1993); and (4) non-symmetric
equal-loss solution of Chun (1988).
China possesses total water resources of 2812.4 billion m3,
ranking the 6th in the world (World Bank, 2002; Wei, 2007).

2501

water research 44 (2010) 24992516

Fig. 1 A sketch map of the South-to-North Water Transfer Project.

However, due to spatially uneven distribution of precipitation,

water shortage has been a prolonged and widespread problem
in Northern regions of China (Yang and Zehnder, 2001; Wei,
2007). In order to mitigate the existing water crisis, the engineers in China proposed the South-to-North Water Transfer
(SNWT) Projects, including the East Route Project, the Middle
Route Project and the West Route Project. The Middle Route
Project covers two municipalities and four provinces. It is
difficult to manage when water transfer involving such
different large regions. This study aimed to establish game
theoretical models to analyze water allocation and pollution
conflicts existing in the Middle Route of South-to-North Water
Transfer Project in China. The main goals of the study include:
 To forecast and analyze water supply, water demand and
water deficit in the different sectors in water scarce cities
(only Beijing municipality) in northern China,
 To predict and estimate pollutant production, discharge and
reduction from the different pollutant sources in the upper
basin of the Hanjiang River,
 To evaluate the economic benefits and losses of water diversion and pollutant reduction to the cities in the study area,
 To analyze future uncertainty of the game simulation under
different scenarios.

2.

Material and methods

2.1.

Area description

The South-to-North Water Transfer (SNWT) projects in China

comprise the Western Route Project (WRP), the Middle Route
Project (MRP) and the Eastern Route Project (ERP) (Fig. 1). We
focus the study area on the MRP, including Beijing municipality, and the 6 cities in the provinces of Shaanxi, Henan and
Hubei in the upper basin of Hanjiang River (Table 1). The MRP
will divert water in 2010 from the Danjiangkou Reservoir in
the Hanjiang River Basin for 20 big cities and 100 counties in
Beijing and Tianjin Municipalities, Henan and Hubei provinces (CWRPI, 2005). It covers a total area of about 155,000 km2
and crosses about 200 river channels or canals with the total
cannel distance of 1246 km. Interbasin water transfer projects
to reduce water shortage are not new in China. However, WRP
covers two municipalities and four provinces. This project will
change the runoff and water level of the rivers, break the longestablished balances of benefits of different stakeholders, and
cause conflicts. Water transfer action within a region can be
effectively managed through the coordination of local
government and regional river administration, while water

Table 1 Description of the cities, their main interests and problems in the study.
Province

City or
municipality

ID

Interest

Main problems

Beijing (BJ)

Beijing

R1

Obtaining sufficient water for socio-economic

and environmental protection

Shaanxi (SX)

Hanzhong
Ankang
Shangluo
Shiyan
Xixia
Xichuan

C1
C2
C3
C4
C5
C6

Developing economy, improving living

standard, reducing pollution based on their
economic abilities

Developed region, severe water shortage, overusing

ground and ecological water, transferring water and
compensating to improve water quality in the basin
Undeveloped regions, reducing pollutants for water
transfer, imposing extra cost

Hubei (HUB)
Henan (HN)

2502

water research 44 (2010) 24992516

transfer involving different regions with such large areas is

usually difficult to manage.
The Hanjiang River basin lies in 30 080 40 110 N latitude,
106 120 114 140 E longitude. The river originates in the
southern part of Shaanxi Province, flows through Shaanxi and
Hubei provinces and joins the Yangtze River at Wuhan, the
capital city of Hubei province. It is about 1577 km long, the
longest tributary of Yangtze River; and the basin covers
a watershed area of 159,000 km2, the second largest river basin
in Yangtze River catchment. The Hanjiang River basin belongs
to subtropics monsoon area, and it is temperate and moist and
annual precipitation is 873 mm. According to the data series of
hydrology from 19561998, the river has total water resource
of 58.2 billion m3 and average annual natural runoff is
56.6 billion m3 (CWRPI, 2005). The river is traditionally divided
into three parts: upper, middle and lower rivers. The upper
river is from the river source to the Danjiangkou city, the
middle river from the Danjiangkou city to Zhongxiang City,
and the lower part from Zhongxiang City to the river mouth
(Zhang, et al., 2000). This paper only dealt with the upper basin
of the Hanjiang River. The upper river is 925 km long, and the
upper basin includes part of provinces of Shaanxi, Henan and
Hubei. This part possesses surface water of 36.796 billion m3,
groundwater of 10.647 billion m3, and overlap amount is
10.387 billion m3. In the upper basin, the u-shaped Danjiangkou Reservoir is the water source of MRP, covering an area of
1050 km2 with a total storage capacity of 17.45 billion m3. The
average annual inflow of the Reservoir is 38.78 billion m3
approximately occupying 70% of water volume of the entire
basin.
Beijing municipality, the capital of China, is located in the
north-eastern part of China and covers an area of 16,808 km.
Beijing belongs to semi-humid climate region dominated by
the Pacific monsoon, and the annual average precipitation is
about 601 mm. Water scarcity is one of the serious problems of
this city due to uneven distribution of precipitation and water
pollution (Wei and Gnauck, 2007a; Wei et al., in press). Beijing
has a population of 15.38 million (BJSB, 20012009); but its
current available water resource per capita is only 247 m3 per
year, which is much less than the standard for water shortage
(1000 m3 per capita) defined by the United Nation (UN), and is
also below the minimum (300 m3 per capita) that the United
Nations Educational, Scientific and Cultural Organization
(UNESCO) defines to ensure a modern social life and production. According to the internationally recognized standards of
extreme water deficit (500 m3/person), Beijing belongs to the
areas of extreme water deficit (Li and Xiu, 2004; Wei, 2007).

2.2.

Data sources

The data in this study includes climatological and hydrological data (19862005), water quality data (19952004) environecological data (19942005) and socio-economic data
(19782008). Climatological and hydrological data include
precipitation, evaporation, surface water, groundwater, and
water flows. Water quality data comprise pollutants concentrations, point pollution sources (industrial waste discharge
and urban domestic waste water discharge), waste water
reclaim amount, and non-point pollution sources (agricultural
fertilizer consumptions, soil erosions, rural domestics and

animal husbandry). Environ-ecological data contains ecological water use, urban water surface areas, public green areas,
and the numbers of newly planned trees. Those socioeconomic data mainly include urban and rural population,
water supply and water demand, gross industrial and agricultural products, urban and rural per capita net incomes.
Data in the studied area were collected mainly from the
following sources: (1) different monitoring stations and
numerous controlling sections along the Hanjiang River and
its tributaries, (2) Database of the Changjiang Water Resource
Protection Institute (DB-CWRPI), (3) Online Database of
National Bureau of Statistics of China (DB-NBSC), (4) Chinese
statistic yearbooks in related fields at different administration
levels (BJSB, 20012009; BJWB, 2005; NBSC and SEPAC,
20012005; CWRA, 19982004; HBEPB, 20042005; HBSB,
19962005; HNSB, 19942007; NBSC, 19852007; NBSC and
SEPAC, 20002005; SXSB, 19912006), (5) Official reports and
planning documents (CWRPI, 2005; BDRC, 2006), and (6)
Previous studies (Hu and Zhang, 2003; Zhang, et al., 2000; Wu
and Zhang, 2005).

2.3.
Analysis of water quality of the Danjiangkou
Reservoir
The Danjiangkou Reservoir has been deteriorated in recent
years, due to great amount of waste discharged into the River
without being treated. The water transfer project requires that
the reservoir water quality should conform to water class II of
Chinese Surface Water Standard II (GB 3838d2002) (SEPAC
and AQSIQC, 2002) before water transfer in 2010. We
selected annual average concentrations of BOD5 (Biology
Oxygen Demand after Five Days), DO (Dissolved Oxygen),
CODMn (Permanganate Index), NH3-N (Ammonia Nitrogen), TP
(Total Phosphorus) and TN (Total Nitrogen) during 19952004
from three water quality monitoring stations Dam, Taizishan (TZS), Taocha (TCA) for analysis of water quality in the
reservoir (Fig. 2).
The analyzing results illustrate that concentrations of
BOD5 (0.682.2 mg/L), DO (7.59.4 mg/L), CODMn (1.42.3 mg/L),
NH3-N (0.050.24 mg/L), all meet the Class II (4 mg/L) (Fig. 3a
d). The concentrations of TP reach 0.6 mg/L and 0.06 mg/L,
which cannot conform to the standard of Class II (0.025 mg/L)
in 2001 and 2003 in Taocha, but they meet the standard in
other years (Fig. 3f). However, the concentration of TN cannot
conform to the Class II, and it belonged to Classes IV and V
(Fig. 3e). The analysis suggests that the water quality deterioration of the Reservoir is mainly reflected by the increase of
concentration of TN, and thus this study focuses on TN
concentration reduction.

2.4.

Game-theoretic approach

Games exist in the situations where the actions of actors

(individuals or groups) are interacting and interdependent and
the choices of all actors affect the outcome (Scharpf, 1997). A
game is a metaphor of the rational behaviors of multi-actors
in an interacting or interdependent situation, such as cooperating or coalition, conflicting, competing, coexisting, etc.
(Wei and Gnauck, 2007a). A country, a region, a group, an
individual, organism, abiotic and biotic constituents or even

water research 44 (2010) 24992516

2503

Fig. 2 The Danjiangkou Reservoir and the water monitoring stations.

nature proper each can be an actor. The essence of this theory

is to study the interaction, strategies and equilibrium of
different actors. A game can be defined by set (1), but a normal
form game (or strategic game) can be generally described as
set (2).
GT bCN; A; V; I; O; ED
GbCN; S; VD

(1)
(2)

where GT a general symbol for all kind of games, G a normal

form game (or strategic game), N set of players, A actions
(moves), S strategy profile, V payoffs (or utility), I information set, O outcomes and E equilibrium or equilibria, i.e.
NAVI-OE. NAPI are collectively known as the rules of a game
and OE are the game results. The main aim of constructing
a game model is to define the rules (NAVI) in mathematical
language and get the solution from OE.
To analyze and solve the problems by means of noncooperative and cooperative games, we modelled and simulated the conflicts of pollution (i.e. TN) reduction and water
allocation in the study area as a set of games with two levels,
including one main game with five players at the first level and
four sub-games with each containing three sub-players at the
second level. In the game model, we used subscripts i and
i to stand generally for every player and every other
n  1 player (or player is opponent in some senses),
respectively. In game with sub-games, we also used m and
j to refer to every main player and sub-player, and mj to
point out which main player a sub-player belongs to, like 11
means sub-player belongs to main player 1 (see Nomenclature). Our game modelling and simulation process can be
generally expressed as defining a problem (or conflict) as
a game, analyzing the game, setting up game models,
analyzing the game models and solving the game. Noncooperative game modelling approach is used to find out
what the real payoff (or utility) of the players, and cooperative
game approach is to get the optimal solution. The main aim of
studying non-cooperative game is to find the solutions for
cooperation.

2.5.

Regression model

In order to formulate functions of the input variables and

payoffs of the players, statistical and econometric regression
methods were used. The linear regression model can be
simply expressed by the following equations:
Yp b0

n X
m
X



bk Xkp 3p

(3)

k1 p1

where Yp values of dependent variables in observation kp;

Xkp independent (or explanatory) variables; bk parameters
of the equation; 3p disturb (or error) term. The equation
includes two components:
(1) b0

n P
m
P
k1 p1

bk Xkp , the non-random component,

(2) 3p, the random component.

To use ordinary least squares (OLS) estimation methods,
some nonlinear models were transformed into linear ones by
logarithm conversions at one side or both sides of the
modelling equations. Besides, polynomial regression and
vector auto-regression were also applied. Autoregressive (AR)
and/or Moving average (MA) terms were included in some
model equations to account for serial correlation. In addition,
balanced panel data and its related modelling approaches
were employed to establish model of the gross agricultural
products and nitrogen fertilizer consumptions. The validity of
the models was evaluated by comparing predictions with
observations.

2.6.

Transport process of pollutants

) total nitrogen
The transporting process of pollutants (W
(TN) in this study into the reservoir during a period of time
can be classified as (1) producing, (2) entering the rivers, (3)
reaching into the reservoir, (4) nitrification/denitrification
processing and forming the final concentration in reservoir.
Part of pollutants will be decayed due to biochemical and

2504

water research 44 (2010) 24992516

3.0

b
Dam
Taizishan

2.0

Taocha
ClassII

DO (mg/L)

BOD5 (mg/L)

2.5

10

1.5
1.0

8
Dam
Taizishan

Taocha
ClassII

0.5
6

0.0
95

96

97

98

99

00

01

02

03

04

4.5

95

d
Dam
Taizishan

3.5
3.0
2.5

1.0

99

00

01

02

03

04

Dam
Taizishan

0.3

Taocha
ClassII

0.2
0.1
0.0
89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04

89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04

1.6

0.07

1.5

0.06

1.4

0.05

TP (mg/L)

TN (mg/L)

98

2.0
1.5

97

0.5
0.4

Taocha
ClassII

NH3-N (mg/L)

CODMn (mg/L)

4.0

96

1.3
1.2
1.1

Dam
Taizishan
ClassIV

1.0
0.9

95

96

97

98

99

00

02

03

0.04

Taoc ha
Cl assI I

0.03
0.02

Taocha
ClassIII

01

Dam
Taizis han

0.01
04

0.00

95

96

97

98

99

00

01

02

03

04

Time (year)

Fig. 3 Water quality of the Danjiangkou Reservoir in three monitoring stations: Dam, Taizishan and Taocha (a) BOD5, (b)
Do, (c) CODMn, (d) NH3-N, (e) TN, and (f) TP from 1995 to 2004.

ecological processes. This process and the annual mean

concentration of TN reached in the reservoir can be expressed
by Eqs. (4) and (5), respectively:
t [ra lra kra fra
Ltra W
ra

(4)

Ctra Ltra =Qft

(5)

where ra subscript presenting human activity in a region; Ltra

load of TN into the reservoir from a certain human activity in
t pollutant TN production from
a region during time t; W
ra
a certain human activity in a region during time t;
[ra ; lra ; kra ; fra generally called pollutant transport coefficients, i.e. rate of TN loss, coefficient of TN into the river, rate
of TN into the reservoir, as well as rate of TN finally maintaining in the reservoir, respectively; Ctra average TN

concentration into the reservoir during time t; Qft mean

water inflow into the reservoir during time t.
Based on the previous studies (Yang et al. 2006, Cheng et al.
2006, Song et al. 2006), the values of nitrogen transport coefficients were defined in Table 2. Urban domestic sewage and
industry waste water are transported by pipelines, and they
are emitting directly into the local river surface. Therefore,
nearly 100% of nitrogen enters regional rivers, and thus rate of
loss ([) from production resource and rate of entering river (l)
are defined as 1. We also defined that the rate of TN maintaining in the reservoir is 1, i.e. no loss in the reservoir. A unit
of Pig equivalence was used to measure the livestock and
poultry by pig unit based on their annual average nitrogen
production. According to the study on the spatial and
temporal change of nitrogen and phosphorus produced by
livestock and poultry in China (Wu, 2005), it defined that 1 pig

2505

water research 44 (2010) 24992516

Table 2 Different transportation coefficients of nitrogen.

Nitrogen source

Nitrogen fertilizer
Soil erosion
Urban domestic sewage
Industry waste water
Animal husbandry
Rural domestic life

Shaanxi

Hubei

0.10
0.21
1.00
1.00
0.10
0.10

0.96
0.81
1.00
1.00
0.96
0.96

0.80
0.80
0.80
0.80
0.80
0.80

1.00
1.00
1.00
1.00
1.00
1.00

0.10
0.21
1.00
1.00
0.10
0.10

0.96
0.81
1.00
1.00
0.96
0.96

0.90
0.90
0.90
0.90
0.90
0.90

1.00
1.00
1.00
1.00
1.00
1.00

0.10
0.21
1.00
1.00
0.10
0.10

0.96
0.81
1.00
1.00
0.96
0.96

0.90
0.90
0.90
0.90
0.90
0.90

1.00
1.00
1.00
1.00
1.00
1.00

Therefore, the payoff values of the players are not at the same
time level. In details, the benefits of Beijing obtaining from
water diversion will be produced after 2010, while the losses of
the cities in the Hanjiang River basin due to reduction pollution for water transfer will be generated before 2010. In this
study, we start our pollution reduction from the base 2005,
and we only calculate the benefits of Beijing from 2010 to 2015
in order to compare those 6-year benefits to the 6-year losses.
In this sense, the future values should be discounted and
transformed into the current values. The future values are
termed as nominal values and the present values as
comparable or real values. In economics, Consumer Price
Index (CPI) is one of widely used deflator to kick out the price
inflation and change the nominal values into comparable
values. The CPI observation values of Beijing used for the
value discount in this study are listed in Table 3, and the
discount formula can be expressed as:

is equal to 1/5 of large animals, 2 goals or sheep, and 30

poultry, respectively, in nitrogen production. Those
researches also stated that the average annual nitrogen
amounts from a persons manure and liquid, a pigs manure
and liquid are 1.32 kg a1, 3.07 kg a1, 7.58 kg a1 and
3.93 kg a1, respectively.

2.7.

Henan

Total nitrogen reduction

Total nitrogen (TN) concentration reduction in the Danjiangkou Reservoir was planned to follow a linear trend to reach the
Chinese water quality standard of Class II (0.2 mg/
L  TN  0.5 mg/L) by 2010, and the two main reasons for this
consideration are: (1) a straight line is the shortest distance
between two points in geometric and mathematic principle;
(2) a straight line trend to reduce TN means time-cost saving.
The linear trend of upper threshold (Cmax) and lower threshold
(Cmin) of TN concentrations during different years (t), are
expressed by Eqs. (6) and (7).

DtR DtN ptI =ptI

(8)

t0

Cmax 0:127t 255:1

(6)

Cmin 0:177t 355:3

(7)

2.8.

where DR comparable or real value of payoff (V0 and U0 ) in

0
year t0 , DtN nominal value of payoff (V and U ) in year t, ptI
0
t
Consumer Price Index in year t , pI Consumer Price Index in
year t.

Nominal and real values

2.9.

In the case area, there is a clear time value included in the

benefits and losses of players, because pollution reduction
(cost) will be processed before water transfer (benefit).

Other methods

We used demand-supply principle (DSP), cost-benefit analysis

(CBA) and economic valuation methods (EVMs) to compare the

Table 3 Consumer Price Index of Beijing used for the value transformation.
t
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
a

CPI

CPI

CPI

CPI

100.0
101.8
107.9
109.2
111.2
111.8
114.2
134.4
143.5
155.8

1988
1989
1990
1991
1992
1993
1994
1995
1996
1997

187.6
219.9
231.8
259.4
285.1
339.3
423.8
497.1
554.8
584.2

1998
1999
2000
2001
2002
2003
2004
2005
2006
2007

598.2
601.8
622.9
642.2
630.6
631.9
638.2
647.8
653.6 (666.9)
669.3 (672.3)

2008
2009
2010
2011
2012
2013
2014
2015

703.4 (702.2)
740.4
783.8
829.0
873.4
914.8
952.1
984.6

Data from 1978 to 2008 (BJSB, 20012009).

Values from 2009 to 2015 are predictions.
c
The values in parenthesis from 2006 to 2008 are predictions used to test the accuracy of the prediction.
d
The prediction model is: PI 6521.67  0.79 PI (2)** 1.6 PI (1)** 3.30t* with R2 0.99, Adj-R2 0.99, F-statistic 2148.56,
Prob(F-statistic) < 0.000001, *significant at p <0.01, **significant at p < 0.0001.
b

2506

water research 44 (2010) 24992516

Table 4 Descriptions of the four scenarios.

Scenario
SN1

SN2

SN3

SN4

Description
Demographic changes, economic growth, waster water discharge and reclaiming rate, environ-ecology protection,
water resource exploitation, hydro-climatology condition were just as usual; and water resource was in the condition
of normal years (P 50%).
Compared with SN1, demographic changes largely decreased, economic growth rate greatly increased, waste water
discharge significantly reduced, waste water reclaiming rate rapidly enhanced, environ-ecology and water resource
exploitation greatly protected, and it met wet years (P 20% in water resources and P 10% of water flow).
Compared with SN1, demographic changes decreased, economic growth rate increased, waster water discharge reduced,
waste water reclaiming rate enhanced, environ-ecology and water resource exploitation well protected, and it was in
moderate dry years (P 75% of both water resources and water inflow).
Compared with other 3 scenarios, demographic changes increased, economic growth rate decreased, waster water discharge
increased, waste water reclaiming rate declined, environ-ecology and water resource exploitation not well protected, and it
met high dry years (P 95% of both water resources and water inflow).

outcomes and results of the game modelling. EVMs were also

applied to estimate the benefit and loss in monetary term.
Various economic valuation methods can be used to quantify
economic value of water resources and losses of water
pollution, such as Direct market value method (DMVM),
Shadow engineering Method (SEM) or Replacement cost
approach (RCA), and Opportunity cost or benefits method
(OCM/OBM), Cost analysis method (CAM), Hedonic Pricing, etc.
(Feng and Wang, 2003; Li and Xiu, 2003, 2004; Wei, 2005). We
adopted DMVM to measure the benefits of water transfer,
CAM to calculate the losses of nitrogen pollutant reduction,
and OCM and OBM to evaluate the losses of Beijing and
benefits of cities in the upper river basin of Hanjiang without
water transfer.

2.10.

Scenario design

Four scenarios were designed to analyze the uncertainties of

the simulation. The normal modelling and simulation were
established based on past data under an assumption of
business as usual, and they were regarded as the first

scenario (SN1). The other three main scenarios were designed

according to the possible changes of constrains and input
variables in SN1. The second scenario (SN2) is very optimistic,
in which the situation is much better than that in SN1 from an
economic and environmental perspective. By contrast, the
fourth scenario (SN4) is much more pessimistic. The third
scenario (SN3) is a coordinated one, whose situations
approximately lied between SN2 and SN4. The main descriptions of those scenarios are showed in Table 4. Based on those
descriptions of the main scenarios, the assumptions of
scenarios were quantified in the Table 5.

3.

Model

Water conflicts in the study can be briefly stated that R1 will

transfer water from the Danjiangkou Reservoir in the Hanjiang River. Water transfer requires the cities (C1, C2, C3, C4, C5,
C6) reducing their pollutant (TN) discharge in order to improve
the water quality in the Reservoir, while TN reduction will
raise cost to those cities (Fig. 4). In this connection, the
conflicts concerning water allocation and water pollutant

Table 5 Assumption of the main scenarios for all

players.
Main force (average annual change
rate % on base of baseline SN1)

SN2

SN3

SN4

Demographic changes
Gross industrial products
Net income
Gross agricultural products
Livestock and poultry
Fertilizer consumptions
Soil erosion
Industry waste water discharge
Urban domestic sewage discharge
Reclaim water
Industry waste water treatment
Urban and rural sewage treatment
Ecological water demand
Ecological water use
Water resource (probability)
Water inflow (probability)

1.0
3.0
3.0
6.0
6.0
6.0
6.0
6.0
6.0
3.0
2.0
12
5.0
4.0
20
10

0.3
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
2.0
1.0
8.0
2.0
3.0
75
75

0.3
6.0
6.0
3.0
3.0
3.0
3.0
3.0
3.0
1.0
0.5
5.0
1.0
0.5
95
95

Fig. 4 Sketch of the regions and cities included in the

game simulation.

water research 44 (2010) 24992516

reduction in this study area are unavoidable if the interests

and benefits are not balanced well.

3.1.

Formulating the game model

The water conflicting problem in the study area is simulated

as normal (or strategic) form games with two levels, including
one big game and 4 sub-games. The mathematic definitions of
those games can be expressed generally as follows:
GT JCG1 ; G0D
G0JCG01 ; G02 ; .; G0m D

3.1.3.

Vit ; Uti
(10)

G0m bCNmj ; Smj ; Vmj D

m 1; 2; 3; 4;

(11)

j 1; 2; 3;

and m; ji; i 1; 2; 3; .; n

(12)

where i, m, j, and mj subscripts refer to every player, main

player, sub-player, and which main player a sub-player belong
to, respectively; GT a game set; G1 the first level game; G0 , G0m
set of sub-games, and a sub-game of main player m,
respectively; Nm, Nmj set of main players m, and sub-players
mj, respectively; Sm, Smj strategy profile of every main player
and sub-player, respectively; Vm, Vmj payoff of main player m
and sub-player mj, respectively.

3.1.1.

Definition of the players

In the game at the first level, R1 is defined as main player 1 (P1),

C1, C2 and C3 main player 2 (P2), C6 main player 3 (P3), and C4
and C5 main player 4 (P4). In each sub-games, industry,
household and agriculture of every main player are defined as
sub-players 1, 2 and 3, denoted by (P11, P12, P13,.Pmj). They can
be expressed as follows:
Nm fP1 ; P2 ; .Pm g;

m 1; 2; 3; 4



Nmj P11 ; P12 ; .Pmj ;

m 1; 2; 3; 4;

(13)
j 1; 2; 3

(14)

Definition of the payoff functions

The payoffs of the main player 1 (P1) and his sub-players (P11,
P12 and P13) are the benefits obtained by using water, and
therefore their payoff functions can be formulated by their
available water and the economic values produced by water
use. For other 3 main players and their sub-players, their
payoffs are the cost to reduce TN discharge, and thus their
payoff functions can be formulated by the TN reduction and
the cost to reduce TN. Those payoff functions can be generally
expressed by Eq. (17).

(9)

G1 bCNm ; Sm ; Vm D

2507

 f Q t ; m; ji; m 1; j 1; 2; 3
 i t
; m; ji; m 2; 3; 4; j 1; 2; 3
g W
i

(17)

where i, m, j subscripts refer to every player, main player and

sub-player, respectively; t superscript stands for time, V, U
payoff of every player i non-cooperative and cooperative
game, respectively, Qit available water of every player i
t pollutant TN reduction by every player.
during time t, W
i

3.2.

Assumptions

The games are static and finite with incomplete

information;
- All the players are rational, and their aims are to maximize
their welfares;
- There is no administrative intervention during game processing, but the game processing is influenced by the
current policies;
- The cities in the same administrative regions is willing to
cooperate with each other due to the similar interests, i.e.
C1, C2, and C3 cooperation with each other; the same for C4
and C5;
- Cooperation or non-cooperation of other players excluded
from this example will depend on whether players 1, 2, 3
and 4 cooperate or non-cooperate;
- Water demand of each player will keep the same in
different hydrological conditions.
-

Among them:

3.3.
P1 fR1 g; P2 fC1 ; C2 ; C3 g; P3 fC4 g; and P4 fC5 ; C6 g
Pm1 fIndustryg; Pm2 fhouseholdg; and Pm3
fagricultureg

3.1.2.

Game simulation processes

(15)

(16)

Definition of the strategies

To simplify the problems, we defined that every player has

only two strategies. Rather, in the main game, P1 has either
strategies of transferring water from other players (SWT) or not
transferring water (SNT), i.e. solving water shortage internally;
and P2, P3 and P4 have similar two strategies: reducing TN
pollution for water transfer (SRP) and not doing so (SNR). In the
sub-games 1, the sub-players of main player 1 will choose
strategies of struggling for water without considering too
much of environ-ecology (SWS) and sharing their imitated
water resources considering environ-ecology and increasing
water use efficiency (SNS). In the sub-games 2, 3 and 4, the two
strategies of the sub-players are (1) to discharge pollutant into
reservoir freely (SPF); (2) to reduce the pollutant TN discharge
according to their economic abilities (SPA).

The game simulation flow can be illustrated by Fig. 5, which

includes 5 games. These five games are divided in two levels.
Game 1 is the main game at first level and games 25 are the
four sub-games at the second level. The main game is named
as a benefit-loss game, in which we will compare the benefits
and losses of different players under non-cooperation and
cooperation. The first sub-game (game 2) is a water obtaining
game, where sub-players make decision how to get their
water. The other three sub-games (games 25) are classed as
TN reduction game, where sub-players will decide if they should
reduce pollutant TN discharge into the river (Fig. 5). We start
the games from the main game, and then game 2, game 3, and
so on. In game one, P1 starts first, and he uses either strategy 1
(SWT) or strategy 2 (SNT). Then P2 moves and he knows P1 has
two strategies, but does not know which strategy P1 will use
due to incomplete information. Thus he will use his strategy 1
or strategy 2 according to his real situations. P3 moves next,
and then P4 moves. For each sub-game, we also start from subplayer 1, then sub-player 2 and sub-player 3. In game 1, when

2508

water research 44 (2010) 24992516

Fig. 5 A sketch of game simulation process, (a) the main game at first level, (b) sub-game 1, (c) sub-game 2, (d) sub-game 3,
and (e) sub-game 4.

P1 uses strategy of transferring water, and others players

reducing pollution for water transfer, the game enters
a cooperative game. In game 2, it also enters a cooperative
game if sub-players 1113 all adopt the strategy 2 (SNS). Similarly, games 35 each will also enter a cooperative game when
sub-players employ their strategy 2 (SPA). All problems in
those situations need cooperative methods to solve and get
the results. The results in those cooperative situations and
reverse situations have more practical values than those in
other mixed situations. Those results are the outcome (V1, V2,
0
0
0
; Q12
; Q13
,
V3, V4) and (U1, U2, U3, U4), (Q11, Q12, Q13) and Q11
0
0
0
;W
and W
;W
;W
, W
;W
;W
and
;W
W
21
22
23
21
22
23
31
32
33
0 ;W
0 , as well as W
;W
;W
and W
0 ;W
0 ;W
0 .
0 ;W
W
31
32
33
41
42
43
41
42
43
Therefore, we just calculate those results in the study.

3.4.

Game solution

3.4.1.

A non-cooperative game model

A non-cooperative game solution model for water resources

management presents that every player i maximizes the net
benefits, i.e. differences between benefits produced from
water usage and the costs charged for waste water reduction
or treatment. The model is expressed by Eq. (18).

MaxVit

Q;W;t

Zn

 
dt
e dt
Bti Q  Kti W

(18)

where i subscript general refers to every player, t superscript stands for time (year), Vit payoff of every player i
pollutant
during time t, Q available water for player i, W
TN reduction by every player i, edt discount factor, Bti Q
benefit function of every player i to use available water
cost of every player i to abate pollutant
during time t, Kti W
TN during time t. Unlike other study to use interest rate, we
used Eq. (8) to make discount or value transformation in this
study.

3.4.1.1. Water quantity optimization. Water quantity optimization means that per unit economic value will be produced
by consuming minimum unit of water. It also means that
consumption per unit of water will produce maximum unit of
economic values. It can be expressed by:

MaxBti Q

n
X
t0

bti Qit

(19)

2509

water research 44 (2010) 24992516

t
 Qit
Wst Wgt Rt  Qed

(20)

from player is controlling section y  1 in upstream during

time t; Cw;y
c controlling concentration (i.e. standard) of
in player is controlling section y.
pollutant W

t
t
t
Qga
Qtrt  Qed
Qws

(21)

3.4.2.

t
t
aQit  Qeu
 Etws  Qit
Qyt  Qy1

(22)

0  Rt  Qit

(23)

subject to

Q ti  Qit  Q i

(24)

where i subscript refer to every player including the main

player and sub-player, t superscript stands for time (year),
Bti Q benefit function of available water of every player i, bti
benefit coefficient of water use, i.e. value produced by every
player i using per unit of water during time t, Qit available water
t
environ-ecological water
for every player i during time t, Qed
t
available water used for
demand of player i during time t, Qeu
environ-ecology of player i during time t, Wst surface water
t
0 resource amount; Qws

resource amount during time t, Wgt W

water demand to keep certain amount of urban water surface
t
water demand to maintain certain area of
during time t, Qga
public green area during time t, Qtrt water demand of newly
planed trees during time t, Rt reclaimed waste water during
time t, a coefficient of waste water and sewage back into water
t
water inflow from upstream controlling (i.e.
during time t, Qy1
player is) section y  1 in during time t, Qyt water flow in the
observed (player is) section y during time t, Etws evaporation of
t
water surface during time t, Q ti and Q i minimum and
maximum of water demand of player i during time t.

A cooperative game model

The cooperative game model means that all the players

cooperate with each other to maximize the overall net benefits. It is expressed by Eq. (30). Every player in cooperative
game is to maximize the net benefits which he can obtain
from cooperation. It is expressed by Eq. (31).
t

Max U

Zn

Q;W;t

t
 
dt
e dt
B Q  Kt W

(30)

Max Uti Vit max

n
Y


UtB =j i

(31)

subject to
Ut 

n
X

Vit UtB

(32)

i1

UtB  0

(33)

where i subscript refer to every player, t superscript stands

for time, Ut the total benefit obtained from cooperative game
during time t, Bt Q the benefit function of water use in
the cost to abate
cooperative game during time t, Kt W
t
pollutant TN in cooperative game, Ui payoff of each player i
in cooperative game, Vit payoff of every player i noncooperative game, UtB total net benefit obtained from cooperative game, j distribution factor of cooperative benefit.

3.4.1.2. Water quality optimization. Water quality optimization means that every player i minimizes the costs to reduce
pollutant (TN) discharge into the water body. It can be
expressed as follows:
m h


i
X
 
nX
gw
y  1t 1  hw;y  Lw;y
1
Lw;y
w
Min Kti W
c
t L w;
y1

subject to

(25)

t
L
t
t
w; y  1 Qy1
Cw;y1

(26)

L t
w; y Qyt Cw;y
t

(27)

Lw;
yc Qyt Cw;y
c

(28)

Q > 0; L > 0; C > 0; h  0

(29)

cost of every player i to abate pollutant W

(TN)
where Kti W
(TN)
during time t, gw cost coefficient to reduce pollutant W

of every player i; Lw;y

t load of pollutant W in controlling (i.e.

t load of pollutant W
player is) section y during time t, Lw;y1

from upstream controlling (i.e. player is) section y  1 during

in player is
time t, Lw;y
c controlling load of pollutant W
decomposition (assimilation)
controlling section y, hw;y

in player is controlling section y,

coefficient of pollutant W

Cw;y
t concentration of pollutant W in player is controlling

concentration
of pollutant W
section y during time t; Cw;y1

4.

Results and discussion

4.1.

Evaluation of game simulation

The validity of the simulation was evaluated by comparing

predictions with observations (data) during 2000 and 2006. To
display the compared results of all the values at different sizes
in one diagram, scientific notion method was applied to
transform all numbers into the form a  10b, the significant
a was defined as any real number between 1 and 10, and three
decimal places was set to a to keep all residuals (differences
between observations and predictions) from zero after the
transformation. The evaluation results illustrate that forecasts and observation are very close (Fig. 6a) with an average
error of 3.6% except one error is 16.51% and another 10.48%
(Fig. 6b). Through this evaluation, it is clear that the simulation has a very good ability to reflect reality and can be used
for future prediction.

4.2.

Water deficit

The simulation results of water deficits of sub-players 11, 12

and 13 in game 2 are displayed in Table 6. In the noncooperative game, available water for sub-player 11 (Q11) and
sub-player 13 (Q13) will decrease from 5.38  108 m3 to
4.10  108 m3, and 10.46  108 m3 to 7.99  108 m3, respectively

2510

water research 44 (2010) 24992516

Fig. 6 Game simulation evaluation (a) comparison of predictions with observation, (b) forecasting errors.

Table 6 Simulation results of available water (Q) and water deficit (Qd) of sub-players 1113 in game 2 (3108 m3).
t

2010
2011
2012
2013
2014
2015

Non-cooperation

Cooperation

Q11

Q12

Q13

Q1

0
Q11

5.38
5.10
4.83
4.58
4.33
4.10

15.92
16.31
16.71
17.1
17.49
17.88

10.46
9.94
9.43
8.93
8.45
7.99

31.76
31.35
30.97
30.61
30.27
29.97

4.02
3.82
3.62
3.43
3.24
3.06

Comparison (water deficit)

0
Q12

0
Q13

Q10

Q11d

Q12d

Q13d

Q10

11.89
12.21
12.52
12.8
13.08
13.33

7.81
7.44
7.06
6.69
6.32
5.96

23.73
23.46
23.20
22.92
22.63
22.34

1.36
1.28
1.21
1.15
1.09
1.04

4.03
4.1
4.19
4.3
4.41
4.55

2.65
2.5
2.37
2.24
2.13
2.03

8.03
7.89
7.77
7.69
7.64
7.63

,W
0 ) and reduction (W
R) by players 2123 in game 3 (tons/a).
Table 7 Simulation results of total nitrogen production (W
t

2005
2006
2007
2008
2009
2010

No-cooperation

Cooperation

21
W

22
W

23
W

2
W

0
W
21

694.0
634.2
601.7
570.9
541.7
514.0

40,131.7
40,139.5
40,195.2
40,275.3
40,367.7
40,466.3

273,586.4
275,665.6
279,338.8
283,116.7
286,994.6
290,772.4

314,412.1
316,439.3
320,135.8
323,962.9
32,7904.0
331,752.7

530.8
451.9
315.2
301.5
270.2
157.8

0
W
23

0
W
2

21R
W

22R
W

23R
W

2R
W

30,692.3
28,603.8
21,055.7
21,271.8
20,135.9
12,426.3

250,061.8
237,215.5
187,125.1
190,377.2
184,065.9
130,270.4

240,459.1
225,497.5
167,699.1
171,104.4
163,562.6
101,874.3

163.2
182.3
286.5
269.4
271.5
356.2

9439.4
11,535.7
19,139.5
19,003.5
20,231.8
28,040.0

64,350.3
79,223.8
133,010.7
133,585.7
143,838.1
201,482.3

73,953.0
90,941.8
152,436.7
152,858.5
164,341.4
229,878.4

from 2010 to 2015. In contract, available water for sub-player

12 showed an increasing trend, from 15.92  108 m3 to
17.88  108 m3 during the same period of time. In cooperation
game, the water amounts obtained by them are less than that
in non-cooperation, mainly because the players stop overusing underground and ecological water and share their
limited water resources. By comparing the available water in
cooperation to that in non-cooperation, it was found that
those players will face serious water deficits in the condition
of cooperation.

4.3.

Comparison (reduction)

0
W
22

Nitrogen reduction

The simulation results of nitrogen reductions in games 35 are

revealed from in Tables 79. In game 3, players 21, 22 and 23
can produce total nitrogen (TN) of 694.0514 tons, 40,131.7

40,466.3 tons, and 273,586.4290,772.4 tons, respectively from

2005 to 2010 in non-cooperative game. In cooperative game,
they have to reduce the nitrogen production in order to
improve the water quality for water transfer. Comparing the
TN productions in non-cooperative to that in cooperative
games, it found that players 21, 22 and 23 will reduce,
respectively, TN of 163.2356.2 tons, 9439.428,040.0 tons, and
64,350.3201,482.3 tons from 2005 to 2010 (Table 7). Based on
the similar analysis, it revealed that in game 4 sub-players 31,
32 and 33 should reduce, respectively, nitrogen of 89.2
506.2 tons, 3695.011,581.8 tons and 15,672.551,276.9 tons,
during the same period of time (Table 8). In game 5, the results
conformed that players 41, 42 and 43 should reduce, respectively, TN of 45.6165.8 tons, 1120.73247.4 tons and 13,553.3
52,755.2 tons to meet the water quality standard from 2005 to
2010 (Table 9).

2511

water research 44 (2010) 24992516

,W
0 ) and reduction (W
R) by players 3133 in game 4 (tons/a).
Table 8 Simulation results of total nitrogen production (W
t

2005
2006
2007
2008
2009
2010

No-cooperation
31
W

32
W

33
W

379.2
467.7
543.8
606.1
668.3
730.6

15,709.5
15,873.6
16,055.1
16,255.1
16,474.5
16,714.4

66,632.0
68,232.0
69,751.5
71,198.2
72,602.2
74,001.1

Cooperation

Comparison (Reduction)

Total

0
W
31

0
W
32

0
W
33

0
W
3

31R
W

32R
W

33R
W

3R
W

82,720.7
84,573.3
86,350.4
88,059.4
89,744.9
91,446.0

290.0
333.3
284.8
320.1
333.4
224.4

12,014.5
11,311.7
8410.3
8585.3
8217.7
5132.6

67,048.2
64,964.0
53,137.3
54,465.3
53,357.6
40,169.1

63,263.9
60,267.7
45,233.5
46,509.5
44,765.9
28,081.1

89.2
134.4
259.0
286.0
334.9
506.2

3695
4561.9
7644.8
7669.8
8256.8
11,581.8

15,672.5
19,609.3
33,213.1
33,594.1
36,387.3
51,276.9

19,456.8
24,305.6
41,116.9
41,549.9
44,979.0
63,364.9

,W
0 ) and reduction (W
R) by players 4143 in game 5 (tons/a).
Table 9 Simulation results of total nitrogen production (W
t

2005
2006
2007
2008
2009
2010

4.4.

No-cooperation

Cooperation

41
W

42
W

43
W

4
W

0
W
41

193.7
203.8
213.7
223.5
233.4
243.2

4764.5
4750.4
4734.0
4718.3
4702.3
4686.5

57,621.9
60,562.8
64,212.5
67,959.0
71,948.2
76,135.0

62,580.1
65,517.0
69,160.2
72,900.8
76,883.9
81,064.8

148.1
145.2
112.0
118.0
116.4
74.7

0
W
43

0
W
4

41R
W

42R
W

43R
W

4R
W

3643.8
3385.2
2479.8
2492.0
2345.6
1439.1

49,026.8
48,111.8
38,584.6
40,835.1
40,824.4
28,309.3

47,860.6
46,688.0
36,228.7
38,503.3
38,350.7
24,893.3

45.6
58.6
101.7
105.5
117.0
168.5

1120.7
1365.2
2254.2
2226.3
2356.7
3247.4

13,553.3
17,405.2
30,575.6
32,065.7
36,059.5
52,755.5

14,719.5
18,829.0
32,931.5
34,397.5
38,533.2
56,171.5

Payoffs

Results of payoffs at nominal prices in game 1 are presented in

the Matrix 1. In the matrix, the first column and the second
column refer to the payoffs resulting from the simulations of
non-cooperative and the cooperative games, respectively. In
each column, the first, second, third and fourth numbers refer
to the payoffs of players 1, 2, 3 and 4 respectively. The zeros
are used to (1) keep the matrix symmetric, (2) state no game
played there. These results proved that the non-cooperative
game will cost player 1 a total loss of 17.3  1011 yuan from
year 2010 to 2015, but it would yield players 2, 3 and 4 a benefit
of 1.1  1011 yuan. However, compared the overall costs to
benefits, there was an overall loss of 16.2  1011 yuan when
each player does not cooperate with the others. In contract,
the cooperative game results showed that there is an overall
benefit of 16.2  1011 yuan, though players 24 lose
1.1  1011 yuan. Those nominal values have been transformed

Matrix 1 Payoff matrix of players 14 in game 5

(3108 yuan at nominal prices).

Comparison (reduction)

0
W
42

into comparative value (real value) (Matrix 2) based on Eq. (8)

to make reasonable comparison. Based on those results, the
payoffs of players 1 and his sub-players in the years of 2010,
2011, 2012, 2013, 2014 to 2015 have been transferred into the
values in years of 2005, 2006, 2007, 2008, 2009 and 2010,
respectively. These results calculated at real prices proved
that the non-cooperative game will cost player 1 a total loss of
13.6  1011 yuan during 20102015, but it yields players 24
a benefit of 1.1  1011 yuan from 2005 to 2010. However,
comparing the overall costs and benefits, there is an overall
loss of 12.5  1011 yuan when each player does not cooperate
with the others. In contract, the cooperative game result
showed that there is an overall benefit of 12.5  1011 yuan,
though players 24 lose 1.1  1011 yuan (Matrix 2).
Matrix 3 explained the real losses of all sub-players under
the game 1. In the matrix, the rows above the line present noncooperative results and down are cooperative results. From
those results, it showed that non-cooperation among players
1, 2, 3 and 4 will cost sub-players 11, 12 and 13 losses of
662.83  1081222.25  108 yuan, 1230.70  1082614.94 
108 yuan and 24.51  10827.74  108 yuan, respectively, due to
water deficits during 20102015. However, sub-players of

Matrix 2 Payoff matrix of players 14 in the noncooperative and cooperative game (3108 yuan at
comparable prices).

2512

water research 44 (2010) 24992516

Matrix 3 Payoff matrix of all the sub-players in the non-cooperative and cooperative game (3108 yuan at comparable
prices).

1113 can avoid those losses if players 14 are cooperative, but

cooperation impose cost to sub-players of 2123, 3133, 4143.
The losses of sub-players of 21, 22, 23, 31, 32, 33, 41, 42 and 43
are 0.15  1080.32  108 yuan B, 39.07  10840.18  108 yuan,
38.89  10898.50  108 yuan, 0.59  1083.36  108 yuan,
11.46  10816.98  108 yuan, 6.96  10826.04  108 yuan,
0.08  1080.29  108 yuan, 38.21  10838.41  108 yuan and
3.47  10817.83  108 yuan, respectively from 2005 to 2010. On

SN1

SN2

SN3

Sn4

41.5
41.0
40.5
40.0
39.5
39.0
2010

b 40

42.0

Available water resources (108m3)

Water demand (108m3)

the contrary, the sub-players of 1113 will have no such losses

if players 14 are cooperative, but cooperation also imposes
cost to the sub-players of 2123, 3133, 4143. For example, the
sub-players 21, 22 and 23 will lose 0.15  1080.32  108 yuan,
39.07  10840.18  108 yuan and 38.89  10898.50  108 yuan,
respectively from 2005 to 2010. Therefore, all the players will
be better off if a side payment is made between them at the
end of the cooperative game.

2011

SN1

2012

2013

2014

SN2

SN3

SN4

2015

d
Water deficit (108m3)

Water deficit (108m3)

-4
-8
-12
-16
-20
-24
2010

2011

2012

2013

2014

2015

SN1

SN2

SN3

SN4

2012

2013

2014

36
32
28
24
20
16
2010

2011

2015

0
-2
-4
-6
-8
-10
2010

SN1P11
SN2P11
SN3P11
SN4P11

2011

2012

SN1P12
SN2P12
SN3P12
SN4P12

2013

SN1P13
SN2P13
SN3P13
SN4P13

2014

2015

Time (year)

Fig. 7 Scenarios (SN) of (a) water demand, (b) available water resources, (c) water deficit of player 1 (P1), (d) water deficits of
sub-players 11, 12 and 13 (P11, P12, P13) without cooperation with outside players.

2513

water research 44 (2010) 24992516

Total nitrogen discharge (tons)

25,000
20,000
15,000
10,000
5,000
0
2005

c 20,000
Total nitrogen discharge (tons )

b 24,000

2006
SN1P2
SN2P3
SN3P4

2007

2008

SN1P3
SN2P4
SN4P2

2009

SN1P4
SN3P2
SN4P3

8,000

SN1P31
SN2P31
SN3P31
SN4P31

SN1P32
SN2P32
SN3P32
SN4P32

2006

2007

2008

12,000
8,000
4,000
0
2005

SN1P33
SN2P33
SN3P33
SN4P33

2009

2006

SN1P21
SN2P22
SN3P23

SN2P2
SN3P3
SN4P4

4,000
0
2005

16,000

2010

16,000
12,000

20,000

8,000

Total nitrogen discharge (tons )

Total nitrogen reduction (tons)

a 30,000

7,000

2007

2008

SN1P22
SN2P23
SN4P21

2009

SN1P23
SN3P21
SNP22

2010
SN2P21
SN3P22
SN4P23

6,000
5,000
SN1P41
SN2P41
SN3P41
SN4P41

4,000
3,000

SN1P42
SN2P42
SN3P42
SN4P42

SN1P43
SN2P43
SN3P43
SN4P43

2,000
1,000
0
2005

2010

2006

2007

2008

2009

2010

Time /year

Fig. 8 Scenarios (SN) of (a) nitrogen reduction by players 24 (P2P4), (b) nitrogen discharged into the reservoir, by subplayers 2123 (P21P23), (c) nitrogen discharged into the reservoir by sub-players 3133 (P31P33), and (d), nitrogen discharged
into the reservoir by sub-players 4143 (P41P43).

Form those comparing results, it is clear that the players

should cooperate with each other so as to maximize the
overall benefits. However, every player is usually afraid of
cooperation, because they face risks and uncertainties of
losses when they are not sure if others really want to cooperate. Furthermore, every player can be better off by free
riding in non-cooperation. In water scarce area, players can
get their water by free riding in terms of overusing groundwater and ecological water. In the Hanjiang River basin, every
player can also be better off by free riding others achievement
of pollution reduction. In a long run, non-cooperation will
deteriorate the environ-ecology and water quality. Therefore,
non-cooperation results in a game of Prisoners dilemma.
The methods to solve the dilemma are usually to design
a mechanism to change the rules and drive the players to
reach collective rationality. The driving forces usually refer to
something like laws, regulations, contracts and other binding
agreement. In contract with those legislation methods,
economic methods such as tax, fine, compensation and so on,
are also such kinds of driving forces. In this study, reducing
waste water and increasing water quality will impose cost to
players in the reservoir catchment, but they can create a large
benefit to the players in water receiving area. In this sense, all
the players will have incentives to cooperate if a mechanism
could guarantee to transfer part of the benefits obtained from
cooperation to cover the losses of players.

4.5.

Scenario simulation

The main comparison results from the water demand game

simulation under the four scenarios are illustrated in Fig. 7.
The scenario results revealed that, from 2008 to 2015, water
demand of player 1 (Fig. 7a) will increase under each of those
four scenarios, though the efficiency of water consumption
will be highly increased in those scenarios. Player 1 and his
sub-players would face shortage problems (Fig. 7c and d) in
each of the four scenarios (Fig. 7b), mainly due to increase of
ecological water demand. Comparing to other sub-players,
player 12 will face most serious water deficits in each of the
four scenarios (Fig. 7d). It also found that, due to extremely
severe water scarce situation, those players cannot solve their
water deficits without cooperation with other players, even
under the optimistic scenario 4 (S(4)), where it is in the wet
years (P 20%), high waste water recycling amount, etc.
The comparing simulation results of TN reduction under
the four scenarios are illustrated in Fig. 8. The results showed
that, in each of the scenarios, player 2 should take more
responsibility to reduce nitrogen production (Fig. 8a), because
he is the main polluter discharging more pollutant TN into the
reservoir than that of other players. The uncertainty of noncooperation probably come from this player and his subplayers because they face big loss to reduce their TN
discharge based on the payoffs results in scenario 1 (Matrix 2).

2514

water research 44 (2010) 24992516

From the scenario results of sub-games, it is also clear that

sub-players 23, 33 and 43 are the main polluters in games 3, 4
and 5, respectively, because they discharge more nitrogen
than that other players do in each of those gamed under the
four scenarios (Fig. 8c and d).Those sub-players will be the
uncertainty sources of non-cooperation in those games.

5.

Conclusions

This study established game-theoretic simulation models to

analyze the problems of water scarcity and nitrogen reduction
in the South-to-North Water Transfer Project. The simulation
is consisted of two levels, 1 main game with 4 players and 4 subgames with 12 sub-players. Beijing municipality, Shaanxi
(Hanzhong, Ankang and Shangluo cities), Hubei (Shiyan city)
and Henan (Xixia and Xichuan cities) were defined as players 1,
2, 3 and 4; and industry, household and agriculture of those
four players as the sub-players 11, 12 and 13, 21, 22 and 23, 31, 32
and 33, and 41, 42 and 43, respectively. The main results
revealed that player 1 and its sub-players cannot solve their
water deficit problem without cooperation with other players
even under an optimistic scenario. Sub-player 12 will face most
serious water deficit based on the simulation results of four
scenarios. Cooperation with other players is the dominant
strategy of those players. Players 24 and their sub-players will
face costs to reduce pollutant total nitrogen for the water
diversion. The uncertainty of non-cooperation might come
from player 2 in game 1, and sub-players 3 in the games 25.
This study also proofed that non-cooperation will cause whole
society a loss although some players can get benefits. In
contract, cooperation brings some players losses, but it will
produce much more collective benefits. However, players
usually are not willing to cooperate, because they will face risks
of losses. This usually results in a game of Prisoners
dilemma. The players are willing to cooperate if a mechanism
can guarantee to transfer part of the benefits obtained from
cooperation to cover the losses of players. Suggestions on the
mechanisms can include: (1) to sign a binding agreement on
the beneficial players funding losers to build necessary pollution treatment plan; (2) to transfer water using and controlling
right to the losing players; and (3) to include the losses of losers
into the water prices for winners. These game simulation
results will not only benefit the water users to be better off, but
also benefit water administration for decision support on water
distribution, water pricing and ecological compensation, etc.

references

Anbarci, N., 1993. Noncooperative foundations of the area

monotonic solution. Quarterly Journal of Economics 108, 245258.
Bekele, E.G., Nicklow, J.W., 2005. Multiobjective management of
ecosystem services by integrative watershed modeling and
evolutionary algorithms. Water Resources Research 41,
W10406. doi:10.1029/2005WR004090.
BDRC (Beijing Development and Reform Committee), 2006. Water
Resources Protection and Use Planning for No. 11 Five-year
Plan of Beijing. The Beijing Development and Reform
Committee, Beijing (in Chinese).

Bhaduri, A., Barbier, E.B., 2003. Water Transfer and International

River Basin Cooperative Management: The Case of the Ganges.
Department of Economics and Finance, University of
Wyoming, PO Box 3985, Laramie, WY 82071, USA.
BJSB (Beijing Statistic Bureau), 20012009. Statistic Yearbooks of
Beijing (20012009). China Statistics Press, Beijing (in Chinese).
BJWB (Beijing Water Resources Bureau), 2005. Beijing Water
Resources Bulletin. Beijing Water Resources Bureau, Beijing (in
Chinese).
Bogardi, L., Szidarovszky, F., 1976. Application of game theory in
water management. Applied Mathematical Modelling 1, 1620.
Burness, H.S., Quirk, J.P., 1979. Appropriative water rights and the
efficient allocation of resources. American Economic Review
69 (1), 2537.
Carraro, C., Filar, J.A. (Eds.), 1995. Control and Game-Theoretic
Models of the Environment. Birkhauser, Boston.
Cheng, H.G., Hao, F.H., Ren, X.Y., Yang, S.T., Xiong, W., Lei, S.P.,
2006. The study of the rate loss of nitrogenous non-point
source pollution loads in different precipitation levels. Acta
Scientiae Circumstantiae 26 (3), 392397 (in Chinese).
Chun, Y., 1988. The equal-loss principle for bargaining problems.
Economics Letters 26, 103106.
Coppola, E., Szidarovszky, F., 2004. Conflict between water supply
and environmental health risk: a computational neural network
approach. International Game Theory Review 6 (4), 475492.
Colby, B.G., 1990. Transaction costs and efficiency in western
water allocation. American Journal of Agricultural Economics
72 (5), 921182.
CWRA (China Water Resources Administration), 19982004. China
Water Resource Bulletin (19982004). China Water Resources
Administration, Beijing (in Chinese).
CWRPI (Changjiang Water Resources Protection Institute), 2005.
Report on the Environmental Impacts of Middle Route of Southto-North Water Transfer Project. Wuhan, China (in Chinese).
Dellapenna, J.W., 2000. The importance of getting names right:
the myth of markets for water. William and Mary
Environmental Law Policy Review 25, 317377.
Dinar, A., Howitt, R.E., 1997. Mechanisms for allocation of
environmental control cost: empirical test of acceptability and
stability. Journal of Environmental Management 49, 183203.
Donevska, K., Dodeva, S., Tadeva, J., 2003. Urban and agricultural
competition for water in the Republic of Macedonia. Internet
resources. Available at: http://afeid.montpellier.cemagref.fr/
mpl2003/Conf/Donevska.pdf (accessed 27.11.09).
Fang, L., Hipel, K.W., Kilgour, D.M., 1998. The graph model
approach to environmental conflict resolution. Journal of
Environmental Management 27 (2), 195212.
Fang, L., Hipel, K.W., Wang, L.Z., 2002. Gisborne water export conflict
study. In: Schmitz, G.H. (Ed.), Proceedings of the Third
International Conference on Water Resources and
Environmental Research, vol. 1. Dresden, Germany, pp. 432436.
Feng, Y.L., Wang, H.J., 2003. Study on resources value of water.
Journal of Hydraulic Engineering 34 (8), 111117 (in Chinese).
Frisvold, G.B., Caswell, M.F., 2000. Transboundary water
management: game theoretic lessons for projects on USMexico. Agricultural Economics 24, 101111.
Giglio, R.J., Wrightington, R., 1972. Methods for apportioning the
costs of a water resource project. Water Resources Research 8,
11331144.
Green, G.P., OConnor, J.P., 2001. Water banking and restoration of
endangered species habitat: an application to the Snake River.
Contemporary Economic Policy 19 (2), 225237.
Harsanyi, J.C., Selten, R., 1972. A generalized Nash solution for
two-person bargaining games with incomplete information.
Management Science 18, 80106.
HBEPB (Hubei Environmental Protection Bureau), 20042005. The
Hanjiang River Bulletin (20042005). Hubei Environmental
Protection Bureau, Wuhan (in Chinese).

water research 44 (2010) 24992516

HBSB (Hubei Statistic Bureau), 19962005. Statistic Yearbooks of

Hubei Province (19962005). China Statistics Press, Beijing (in
Chinese).
Hearne, R.R., Easter, K.W., 1995. Water allocation and water
markets: an analysis of gains-from-trade in Chile. In: Technical
Paper Series Number 315. World Bank, Washington, DC, USA.
HNSB (Henan Statistic Bureau), 19942007. Henan Statistic
Yearbooks (19942007). China Statistics Press, Beijing (in
Chinese).
Holden, P., Thobani, M., 1996. Tradable Water Rights: A Property
Rights Approach to Resolving Water Shortages and Promoting
Investment. World Bank Policy Research Working Paper No.
1627. http://www-wds.worldbank.org/servlet/
WDSContentServer/WDSP/IB/1996/07/01/000009265_
3961214131318/Rendered/PDF/multi_page.pdf Available at:
(accessed 3.12.09).
Howe, C.W., Goodman, D.J., 1995. Resolving water transfer
conflicts through changes in water market process. In:
Dinar, A., Loehman, E.T. (Eds.), Water Quantity/Quality
Management and Conflict Resolution: Institutions, Processed
and Economic Analysis. Praeger Publishers, Westport,
Connecticut, pp. 119129.
Howe, C.W., Schurmeier, D.R., Douglas Shaw, W., 1986.
Innovative approaches to water allocation: the potential for
water markets. Water Resources Research 22 (4), 439448.
Hu, J.J., Zhang, Y.H., 2003. The current soil erosion situations in
the areas of the Danjiangkou Reservoir and its upstream and
the countermeasures. China Water Resources A, 4749 (in
Chinese).
Kalai, E., Smorodinsky, M., 1975. Other solutions to Nashs
Bargaining problem. Econometrica 43, 513518.
Kilgour, D.M., Okada, N., Nishikori, A., 1988. Load control
regulation of water pollution: an analysis using game theory.
Journal of Environmental Management 27, 179194.
Lewandowski, A., 1979. A game-theoretical approach to
modelling the behaviour of water users in a quality control
problem. In: Findeisen, W. (Ed.), Studies in Control
Methodology for Water resource Systems. Techn. Dept., Inst.
Automat. Control, Techn. Uni. Warsaw, Warsaw, pp. 8397.
Li, J.X., Xiu, S.L., 2003. Calculation model of water pollution
induced economic loss for river basin. Journal of Hydraulic
Engineering 10, 6874 (in Chinese).
Li, S.T., Xiu, X.Y. (Eds.), 2004. South-to-North Water Transfer and
China Development. Economic Science Press, Beijing (in
Chinese).
Markandya, A., Recharddson, J. (Eds.), 1992. Environment
Economics: A Reader. St. Martins Press. Inc., New York.
Muleta, M.K., Nicklow, J.W., 2005. Decision support for watershed
management using evolutionary algorithms. Journal of Water
Resources Planning and Management 131 (1), 3544.
Nash, J.F., 1950. The bargaining problem. Econometrica 18,
155162.
NBSC (National Bureau of Statistics of China), 19852007. China
Statistic Yearbooks (19852007). China Statistics Press, Beijing.
NBSC (National Bureau of Statistics of China), SEPAC (State
Environmental Protection Administration of China), 2000
2005. China Environmental Statistic Yearbook. China Statistics
Press, Beijing (in Chinese).
OECD (Organisation for Economic Co-operation and
Development), 1989. Economic Instruments for Environmental
Protection, Paris.
Okada, N., Mikami, Y., 1992. A game-theoretic approach to acid
rain abatement: conflict analysis of environmental load
allocation. Water Resources Bulletin 28, 155162.
Pegram, J.J., Delforce, R.J., Cowell, M.L., Norris, V., Anthony, G.,
Anderson, R.L., Musgrave, W.F., 1992. Transferable Water
Entitlements in Austria. Centre for Water Policy Research,
University of Armidate, Austrila.

2515

Pethig, R. (Ed.), 1992. Conflicts and Cooperation in Managing

Environmental Resources. Springer-Verlag, Berlin, Heidelberg,
New York, Lodon, Paris, etc.
Reidinger, R., 1994. Observations on water markets for irrigation
systems. In: Le Moigne, G., Easter, K.W., Ochs, W.J., Giltner, S.
(Eds.), Water Policy and Water Markets, Technical Paper 249.
World Bank, Washington, DC, USA.
Salazar, R., Szidarovszky, F., Coppola, E., Rojano, A., 2007.
Application of game theory for a groundwater conflict in
Mexico. Journal of Environmental Management 84, 560571.
Saleth, R.M., 1996. Water Institutions in India. Commonwealth
Publishers, New Delhi.
Scharpf, F.W., 1997. Games Real Actors Play: Actor-Centered
Institutionalism in Policy Research. Westview Press,
Boulder, CO.
SEPAC (State Environmental Protection Administration), AQSIQC
(Administration of Quality Supervision, Inspection and
Quarantine of China), 2002. Environmental Quality Standards
for Surface Water of the Peoples Republic of China (GB
3838d2002) (in Chinese).
Shao, X.J., Wang, H., 2003. Interbasin transfer projects and their
implications: a China case study. International Journal of River
Basin Management 1 (1), 514.
Song, G.Q., Yan, M., Zhang, W.M., Zhang, Y., 2006. Preliminary
estimation of TN load of Shiyan to TN concentration of
Danjiangkou Reservoir. Environmental Science and
Technology Supplement(SUM 26), 164165 (in Chinese).
SXSB (Shaanxi Statistic Bureau), 19912006. Shaanxi Statistic
Yearbooks (19912006). China Statistics Press, Beijing.
(in Chinese).
Van der Veeren, R.J.H.M., Tol, R.S.J., 2003. Game theoretic
analyses of nitrate emission reduction strategies in the Rhine
river basin. International Journal of Global Environmental
Issues 3 (1), 74103.
Wang, L.Z., Fang, L., Hipel, K.W., 2003. Water resource allocation:
a cooperative game approach. Journal of Environmental
Informatics 2 (1), 1122.
Wei, S.K., 2005. Managing World Heritage Sites in China from an
Economic Perspective. Dissertation. de-Verlag, Berlin.
Wei, S.K., 2007. A survey on the situation of water resource and
water management in China. In: Gnauck, A. (Ed.), Modeling
and Simulation of Ecosystems. Shaker Verlag, Aachen, pp.
171186.
Wei, S.K., Gnauck, A., 2007a. Water supply and water demand of
Beijing a game theoretic approach for modeling. In:
Gomez, J.M., Sonnenschein, M., Muller, M., Welsch, H.,
Rautenstrauch, C. (Eds.), Information Technologies in
Environmental Engineering. Springer Verlag, Berlin
Heidelberg, pp. 525536.
Wei, S.K., Gnauck, A., 2007b. Game theoretic approaches to model
water conflicts on a river basin scale. In: Gnauck, A. (Ed.),
kosystemen. Shaker
Modellierung und Simulation von O
Verlag, Aachen, pp. 2240.
Wei, S.K., Gnauck, A., 2007c. Simulating water conflicts using
game theoretical models for water resources management. In:
Tiezzi, E., Marques, J.C., Brebbia, C.A., Jrgensen, S.E. (Eds.),
Ecosystems and Sustainable Development VI. WIT Press,
Southampton, Boston, pp. 312.
Wei, S.K., Lei, A.L., Islam, S.N. Modelling and simulation of
industrial water demand in Beijing. Frontiers of
Environmental Science & Engineering in China, in press. doi:
10.1007/s11783-010-0007-6.
World Bank, 2002. World Development Indicators 2002. Available
via World Bank website www.worldbank.org. Accessed on 28
Jan 2007.
Wu, S.X., 2005. The Spatial and Temporal Change of Nitrogen and
Phosphorus Produced by Livestock and Poultry and their
Effects on Agricultural Non-Point Pollution in China. Ph.D.

2516

water research 44 (2010) 24992516

dissertation. The Graduate School of Chinese Academy of

Agricultural Sciences, Agricultural Resources and Regional
Planning Institute of Chinese Academy of Agricultural
Sciences, Beijing (in Chinese).
Wu, P.L., Zhang, W., 2005. Water crisis and sustainable water
resource utilization in Beijing. Journal of Liaoning Technical
University 24 (3), 436439 (in Chinese).
Yang, H., Zehnder, A., 2001. Chinas regional water scarcity and
implications for grain supply and trade. Environment and
Planning A 33, 7995.
Yang, H., Zehnder, A., 2005. The South-North Water Transfer
Project in China: an analysis of water demand uncertainty and
environmental objective in decision making. Water
International 30 (3), 339349.
Yang, S.T., Cheng, H.G., Bu, Q.S., Zhang, J.Y., Shi, X.X., 2006.
Estimation of soil erosion and nitrogen and phosphorus load
in its application in assessment of the absorbed China. Acta
Scientiae Circumstantiae 26 (3), 366374 (in Chinese).
Yang, Z.F., Zeng, Y., 2004. Mathematical model for water conflict
and coordination in transboundary regions. Acta Scientiae
Circumstantiae 24 (1), 7176.
Yevjevich, V., 2001. Water diversions and interbasin transfers.
Water International 26 (3), 342348.
Zeng, W.H., Cheng, S.T., Yang, Z.F., 2001. Integrated management
of river basin. China Environmental Science 21 (2), 173176 (in
Chinese).
Zeng, Y., Yang, Z.F., 2004. Policy conflict analysis of water
quality improving for the transboundary regions of Guanting
reservoir. Advances in Water Science 15 (1), 4044.
(in Chinese).
Zhang, J.Y., Luo, L., Li, C.S., Ma, R.H., Wang, D.B., Tan, Y.Y.,
Kuang, Q.J., 2000. Study on the ecological impacts of middle
route of south-to-north water transfer project on middlelower streams of Hanjiang River. Environmental Science and
Technology Supplement SUM 90, 189 (in Chinese).

Nomenclature
A: action (or moves) in a game
B(Q): water benefit function in cooperative game
Bi(Q): water benefit function of player i in non-cooperative game
BOD5: biology oxygen demand after 5 days (mg/L)
Cra: TN concentration into reservoir from one human activity in
a region (mg/L)
Cw;y
c : limiting concentration of pollutant (TN) in controlling
section y (mg/L)
Cw;y
: concentration of pollutant (TN) in the controlling section y
(mg/L)
Cw=;y1
: concentration of pollutant (TN) from upstream controlling

section y  1 (mg/L)
CODMn: permanganate index (mg/L)
Cmax: upper threshold of TN concentrations (mg/L)
Cmin: lower threshold of TN concentrations (mg/L)
DN: nominal value of payoff (V and U ) (108 yuan)
DR: real value of payoff (V0 and U0 ) (108 yuan)
DO: dissolved oxygen (mg/L)
edt : discount factor
E: game equilibrium
Ews: evaporation of water surface (mm)
GT, G: game, normal form (or strategic)game
G1, G0 : first level game, sub-game respectively
G0m : main player ms sub-game
I: information set of a game
): cost function to abate pollutant (TN) in cooperative game
K(W
(108 yuan)

): cost function of player i to abate pollutant in nonKi(W

cooperative game (108 yuan)
Lw;y
c : limiting TN load in section y (tons)
Lra: load of TN into the reservoir from one human activity in
a region (mg/L)
LW
,y: load of pollutant (TN) in section y (tons)
from upstream controlling section y  1
LW
,y1: load of pollutant W
(tons)
N: set of players
Nm, Nmj: set of main players, sub-players
NH3-N: ammonia nitrogen (mg/L)
O: game outcome
Pm, Pmj: main player, sub-player
0
ptI ; ptI ,: consumer price index (CPI) in time t, t0
Qeu: water used for environ-ecology (108 m3)
Qf: water inflow into the reservoir (108 m3)
Qi,Qi0 : available water of every player in non-cooperative, cooperative game (108 m3)
Qed: environ-ecological water demand (108 m3)
Qid: water deficit of player i (108 m3)
Q i : maximum water demand of player i (108 m3)
Q i : minimum water demand of player i (108 m3)
Qy1: water flow from upstream section y  1 (108 m3)
Qy: water flow in the section y (108 m3)
Qws: water demand to keep certain water surface (108 m3)
Qga: water demand of public green area (108 m3)
Qtr: water demand of newly planed trees (108 m3)
R: reclaimed water (108 m3)
SN: scenarios
S: strategy profile of a game
Sm,Smj: strategy profile of main player, sub-player
TN: total nitrogen (mg/L)
UB: net benefit in cooperative game (108 yuan)
Ui, U0i : payoff, real payoff of player i in cooperative game (108 yuan)
V: payoff (or utility) in a game
Vi, Vi0 : payoff, real payoff of player i in non-cooperative game
(108 yuan)
Vm,Vmj: payoff of main player, sub-player (108 yuan)
Wg: groundwater resources (108 m3)
Ws: surface water resources (108 m3)
0 : pollutant TN production of player i in non-cooperative,
i,W
W
m
cooperative game (tons)
iR: pollutant TN reduction of player i (tons)
W
ra: pollutant TN produced from a certain activity in a region
W
(tons)
Xkp: independent (or explanatory) variables
Yp: dependent variablesGreek symbols
a: coefficient of waste water back into water
b: parameter in linear equation
b: benefit coefficient
g: cost coefficient to reduce pollutant (TN)
j: distribution factor of cooperative benefit
[: loss coefficient of TN from production source
l: coefficient of TN into river
k: coefficient of TN into reservoir
4: coefficient TN finally maintaining in reservoir
h: assimilation coefficient of pollutant
3p: disturb (or error) termSubscripts and superscripts
ra: one certain human activity in a region
k, p: observation numbers
t, t0 : time (year)
i, i: every player, other n  1 player
m, j: every main player, sub-player in sub-game
mj: which main player a sub-player belongs to
: referring to pollutant (TN in this study)
W
y, y  1: lower, upper stream controlling sections