Hindawi Publishing Corporation

Mathematical Problems in Engineering

Volume 2015, Article ID 120873, 16 pages
[Link]

Research Article
A Bayesian Method for Water Resources
Vulnerability Assessment: A Case Study of
the Zhangjiakou Region, North China

Xuan Wang,1,2 Fangbing Ma,1,3 Chunhui Li,1,2 and Jie Zhu1,2

1
State Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, Beijing 100875, China
2
Key Laboratory of Water and Sediment Sciences of Ministry of Education, School of Environment, Beijing Normal University,
Beijing 100875, China
3
Shidu Town People’s Government, Beijing 102411, China

Correspondence should be addressed to Xuan Wang; wangx@[Link]

Received 18 June 2014; Accepted 23 August 2014

Academic Editor: Jian Li

Copyright © 2015 Xuan Wang et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Water resources vulnerability (WRV) assessment is an important basis for maintaining water resources security in a basin. In this
paper, considering the complexity of the water resources system and the uncertainty of the assessment information, a method
based on the Bayesian theory was developed for performing WRV assessments while using the constructed indicator system.
This system includes four subsystems, the hydrological subsystem, the socioeconomic subsystem, the ecoenvironmental subsystem
and the hydraulic engineering subsystem. The WRV degree for each subsystem and the integrated water resources system were
assessed. Finally, the assessment results and the characteristics of the Bayesian method were compared with those of the grey
relational analysis method and the parametric-system method. The results showed the following. (1) The WRV of the integrated
water resources system of the entire Zhangjiakou region was very high; Zhangjiakou City and Xuanhua County have tendencies
to belong to Extreme WRV, with probabilities of 26.8% and 25%, respectively, while the other seven administrative counties have
tendencies to belong to High WRV, with probabilities ranging from 24.6% to 27%. (2) Compared with the parametric-system
method and the grey relational analysis method, the Bayesian method is simple and can effectively address the uncertainty issues
with the reliable WRV assessment results.

1. Introduction for the system to recover to its former status. WRV is

generally reflected by such factors as the quantity and quality
Under the influence of climate change and human activ- of the surface water and groundwater and the carrying
ities, many countries in the world are faced with water capacity of the water resources [6]. The purpose of WRV
problems, such as water shortages, water pollution and assessment is to understand the water resources status in a
frequent occurrences of extreme hydrological events, which certain region, to find the main influencing factors of the
have put pressure on water availability [1, 2]. Thus, it is WRV, and then to propose appropriate countermeasures to
of importance to study the recovery capability of a water reduce the WRV. In recent years, following the continuous
resources system after being disturbed. Besides reliability, development of economic and social progress, the growth of
resilience, and sustainability, vulnerability is an important the population as well as the development of the urbanization
indicator to evaluate the performance of a water resources and industrialization, the demand for water resources is
system [3–5]. Water resources vulnerability (WRV) can be increasing, and the supply is tending toward shortages. At the
defined as the ease with which a water resources system same time, excessive production activities lead to a decline
can be threatened and damaged by both natural disasters in water’s ecological value. Having addressed water security
and human activities. Once this damage occurs, it is difficult issues that are associated with WRV and provided references
2 Mathematical Problems in Engineering

for the protection, rational development and utilization of counties of the Zhangjiakou region, and describe the prob-
regional water resources, WRV assessment has been drawing ability distribution characteristics at different counties; and
more attention. (c) compare the results with the existing methods, including
Since the late 1960s, many scholars and water managers the GRA method and the PS method. The proposed method
around the world have studied the WRV, but most of them will improve the existing method in addressing uncertain
concentrated on groundwater vulnerability assessment [7– information, and the assessment results can help to identify
10]. With respect to surface water vulnerability, most of the specific water resources problems of the Guanting Reservoir,
literature focused on the supply and demand balance of in such a way that appropriate measures can be taken for
water resources or selected several indicators and methods effective environmental management.
to evaluate the carrying capacity or sustainability of the
water resources [11]. Given the uneven distribution of water 2. Study Area
resources in space and time, it is necessary to pay close
attention to the WRV that is caused by climate change 2.1. Overview of the Study Area. The Zhangjiakou region,
and human activities. It is also important to enhance the which is located in the upper reaches of the Guanting Reser-
joint assessment of surface water and groundwater, water voir Basin, is selected as a case to assess the regional WRV
quantity, and water quality, to ensure a sustainable utilization (Figure 1). It is confronted with the dual task of developing the
of water resources of a given region. In recent years, many economy and protecting the natural ecological environment.
approaches have been developed to assess WRV, such as The concerned city and counties are Zhangjiakou City,
overlay and indicator methods [12–14], process-based sim- Huai’an County, Wanquan County, Xuanhua County, Yuxian
ulation methods [15, 16], statistical methods [9, 17], and County, Yangyuan County, Zhuolu County, Huailai County,
fuzzy comprehensive judgment methods [18]. For obtaining and Chongli County, with 17.8 thousand km2 accounting
more reliable assessment results, these methods required a for 40.5% of the area of Guanting Reservoir Basin. The
large amount of natural environmental and socioeconomic Zhangjiakou region has a continental monsoon climate, with
data with high accuracy. In [19], Li quantitatively analyzed an annual mean temperature ranging from 7∘ C to 10∘ C.
the regional WRV by using the fuzzy optimization (FO) The annual mean precipitation is rather low (approximately
method, the grey relational analysis (GRA) method, and the 400 mm), and 80% of the rainfall occurs between June
parametric-system (PS) method by constructing an indicator and September. The topography in the Zhangjiakou region
system with a comprehensive consideration of surface water is very complex; mountains, hills, and rivers account for
and groundwater, water quantity, and water quality. These approximately 30% of the area, respectively. The population of
methods, however, cannot address the uncertainty that is the region is 4.59 million, and the water resources per capita
inherent in all of the vulnerability assessments [20]. With the are 399 m3 , which is less than the internationally accepted
increase in the pressure exerted by socioeconomic develop- standard of extreme water scarcity of 500 m3 [28].
ment on a natural environment, the complexity and uncertain The region has a backward economic development.
characteristics of WRV that arise from both climate change Excessive consumption and pollution of water resources
and human activities has become even more obvious. To have exerted an enormous pressure on local water resources
improve the water resources management and protection, system and have directly influenced the security of the water
the regional WRV research that considers fuzziness and supply of Beijing [29]. Therefore, an effective WRV assess-
uncertainty of the assessment results is highly desired. ment is necessary and desired for regional water resources
With increasingly in-depth research on uncertain events, protection and management. The hydrological data came
the Bayesian method has received extensive attention [21]. from Zhangjiakou Water Conservancy Bureau, the water
It uses objective probability estimations and incomplete quality data came from Environmental Protection Agency of
information to estimate the probability of some unknown Zhangjiakou, and the ecoenvironmental and socioeconomic
states. This method can also describe uncertain information data were from the statistical yearbooks of the Zhangjiakou
flexibly and conduct uncertainty reasoning [22]. Compared region [19].
with the existing assessment methods, the Bayesian method’s
superiority lies in giving the probability or tendency to a 2.2. Key Factors Influencing WRV. Regional WRV is reflected
certain grade to which the system belongs, not just simply by the carrying capacity, recovery capacity, and sustainability
giving the grade. In addition, it has the advantages of a simple of the water resources [6, 30, 31]. The key factors that
principle and calculation process and can obtain reliable influence the WRV of the Zhangjiakou region include the
results for small samples. Thus so far, it has been widely used following aspects.
for water quality assessment [23–25], fault diagnosis, and risk
assessment [26, 27], but it has rarely been used for WRV (1) Hydrology. Influenced by the continental monsoon climate
assessment. The application of the Bayesian method in actual of the arid and semiarid area, the evaporation and runoff of
WRV assessment is expected for the effective management of this region have the characteristic of uneven distribution at
regional water resources. a spatial and temporal scale. The amount of water available
The objectives of this research were to (a) develop a is not enough. The influencing factors that are associated
Bayesian method for WRV assessment; (b) apply the pro- with these hydrological aspects mainly include the natural
posed method to assess the WRV condition of the nine conditions of the regional water resources, such as the
Mathematical Problems in Engineering 3

1
3 1
1
1
2
4
8

7
6

Guanting Reservoir Basin Beijing

Other province

Figure 1: Location of the Zhangjiakou region, North China (1. Zhangjiakou City; 2. Huai’an County; 3. Wanquan County; 4. Xuanhua County;
5. Yuxian County; 6. Yangyuan County; 7. Zhuolu County; 8. Huailai County; 9. Chongli County).

stability of the water resources system, the water abundance, has brought severe industrial point source pollution to the
the ability to renew the underground water, and the regional region. At the same time, because of the low coverage of the
climate conditions. vegetation, the water losses and soil erosion here are very
serious and the ecological environment is vulnerable. The
(2) Socioeconomy. With economic development and popu- influencing factors that are associated with the ecoenviron-
lation increase, the water resources per capita (399 m3 ) of ment mainly include the impact of land exploitation and
the Zhangjiakou are much lower than the national average utilization on water resources, potential pollution intensity
(2100 m3 ); thus, the region is suffering from serious water of agricultural activities on water quality of regional surface
shortage issues. At the same time, the economy of Hebei water, and groundwater, water loss, and soil erosion.
Province is backward, as well as its science and technol-
ogy. Extensive pattern of economic growth leads to a low (4) Hydraulic Engineering and Technical Management. A large
utilization rate of the water resources and the increasing number of water conservancy systems have been constructed
demand for water resources to meet the requirements of both in this region since the late 1950s, including water stor-
human life and industrial production, which has imposed age structures, abstraction systems, and other water supply
great pressure on the water resources. The influencing factors projects. Hydraulic engineering can regulate and allocate
that are associated with the socioeconomic aspects mainly water resources and prevent flood disasters, to meet the needs
include the population pressure on the water resources, the of the water resources of the people and the production. Thus,
water management level, the water supply and water demand making use of hydraulic engineering can enhance the anti-
status, the water exploitation and utilization conditions, and interference ability of the water resources system. In addition,
the water reuse rate. considering that agriculture is the largest supporting industry
of this region, it is necessary to increase the water-saving
(3) Ecoenvironment. The area lies in a fragile belt in terms irrigation area and improve the water use efficiency through
of the ecoenvironment. The agricultural application of pes- technological advances and consciousness enhancement, to
ticides is approximately 2000 tons a year, and fertilizer is reduce its influence on WRV. The influencing factors that
approximately 60,000 tons a year, which leads to serious are associated with the hydraulic engineering and technical
agricultural nonpoint source pollution of the region. Since management mainly include the agricultural water-saving
the late 1980s, rapid development of industrial and mining capability, as well as the regulation and storage capacity of the
enterprises together with low treatment of effluent sewages reservoirs.
4 Mathematical Problems in Engineering

Table 1: Assessment standard of the indicators for different WRV types (Li, 2007) [19].

Subsystem Indicator Description/reflection No Low Medium High Extreme

WRV WRV WRV WRV WRV
a1 Regional water resources income 1500 1200 800 400 100
a2 System stability 10 20 30 40 50
A
a3 Water abundance 40 30 20 10 5
a4 Ground water recharge capacity 80 50 20 10 5
a5 Regional climate 0.5 1 3 5 7
b1 Population pressure on water resources 2500 2000 1500 1000 500
b2 Water utilization efficiency 100 350 600 850 1100
b3 Water utilization level 100 300 500 700 900
B
b4 Water manage level 5 15 25 35 45
b5 Water supply and demand balance 1 0.9 0.8 0.7 0.6
b6 Status of water resources development 10 30 50 70 90
b7 Repetitive utilization degree of industrial water 80 65 50 35 20
c1 Impacts of land development on water 90
30 350 65 80
resources
c2 Agricultural pollution intensity on water 1500
10 25 280 1000
C quality
c3 Surface water quality 0 0.1 0.2 0.5 1
c4 Groundwater quality 0 5 35 60 85
c5 Area changes of soil and water loss 0 10 20 30 40
c6 Degree of sewage disposal 90 75 60 45 30
d1 Water saving capacity of agriculture 80 60 40 30 20
D d2 Regulating capacity of hydraulic engineering 50 40 30 20 10
d3 Storage and release capacity of reservoirs 80 65 50 40 30
Note: A: hydrological subsystem; B: socioeconomic subsystem; C: ecoenvironmental subsystem; D: hydraulic engineering subsystem; a1 : mean annual
precipitation (mm); a2 : annual relative variability of precipitation (%); a3 : mean water resources amount of unit area (104 m3 /km2 ); a4 : groundwater recharge
of unit area (104 m3 /km2 ); a5 : drought index; b1 : water resources per capita (m3 /capita); b2 : water consumption per unit GDP (m3 /104 yuan); b3 : water
consumption per capita (m3 /capita); b4 : difference between planned and actual water utilization amount (%); b5 : water supply and demand ratio; b6 : water
resources development and utilization ratio (%); b7 : repetitive utilization ratio of industrial water (%); c1 : land development ratio (%); c2 : fertilizer use intensity
(kg/ha); c3 : pollution index of surface water; c4 : ratio of polluted groundwater exceeding the standard (%); c5 : ratio of soil and water loss (%); c6 : disposal ratio
of sewage (%); d1 : water-saving irrigation ratio (%); d2 : ratio of design utilizable storage of reservoir to surface water quantity (%); d3 : ratio of effective storage
to total storage of reservoirs (%).

The four aspects interact with each other, for example, 3. Methodology
the increase in the population and water demand and the
decrease in the water resources per capita can make the water 3.1. Conceptual Model Diagram for WRV Assessment Based
shortage even more serious. At the same time, the increase in on Bayesian Theory. In [19], Li divided WRV into four sub-
the pollutant emissions leads to a significant deterioration in systems, the hydrological subsystem, the socioeconomic sub-
the water environment. Over disafforestation, farming, and system, the ecoenvironmental subsystem, and the hydraulic
husbandry lead to a low coverage of the vegetation, thus engineering subsystem, and then constructed an indicator
aggravating the water losses and soil erosion. The disharmo- system for the WRV assessment. The vulnerable degree of
nious development of the natural resources and environment every indicator was divided into five levels, which were No
has reduced the carrying capacity, the recovery capacity, and WRV, Low WRV, Medium WRV, High WRV, and Extreme
the sustainability of the water resources system and increased WRV (see Table 1).
the WRV. Therefore, it is necessary to consider the above Based on Bayesian theory, we calculated the posterior
four aspects comprehensively, to establish an indicator system probability of every indicator, and then, we calculated the
that considers the supply and demand balance of the water posterior probability of the multi-indicator comprehensive
resources and that combines the water quantity with the water WRV assessment of each subsystem. For the multilayered
quality, to fully reflect the influence of the climate change and WRV assessment, the subsystem in one layer can be seen as an
human activities. indicator of the next layer; then, we recalculated the posterior
Mathematical Problems in Engineering 5

a1 ··· a5 b1 ··· b7 c1 ··· c6 d1 ··· d3

Hydrological Socioeconomic Ecoenvironmental Hydraulic engineering

subsystem subsystem subsystem subsystem

Integrated water
resources system

Figure 2: The relationship among the indicators, subsystems, and integrated system.

Procedures Methods

Likelihood Normal distribution sampling method

Posterior probability Bayesian formula

Weight AHP method

WRV assessment of each subsystem of the Zhangjiakou region

AHP method Bayesian formula

WRV assessment of the integrated water resources system of

Zhangjiakou region

Figure 3: Procedures for WRV assessment by using the Bayesian method.

probability of the multisubsystem WRV assessment of the of the multi-indicator comprehensive WRV assessment of
integrated water resources system with the Bayesian formula each subsystem. Next, we used the AHP method again to
(Figure 2). calculate the weight of every subsystem, and we calculated the
posterior probability of the multisubsystem WRV assessment
of the integrated water resources system with the Bayesian
3.2. Procedures for the WRV Assessment Model Based on formula. The calculation steps that were taken with the
the Bayesian Theory. We used the Bayesian formula [32] to Bayesian method were as follows.
calculate the probability of a system belonging to a certain
WRV type, to identify the vulnerable degree of the water (1) Calculate the prior probability of every indicator that
resources system. The WRV assessment procedures with belongs to a certain standard type. Due to insuffi-
the Bayesian formula are shown in Figure 3. First, we used cient information about the indicators, this research
the normal distribution sampling method to calculate the assumed that the probability of each type was equal.
likelihood and used the Bayesian formula to calculate the Specifically, we had the following:
posterior probability of every indicator. Second, we adopted
the analytic hierarchy process (AHP) method to obtain the 1
𝑃 (𝑦𝑗1 ) = 𝑃 (𝑦𝑗2 ) = 𝑃 (𝑦𝑗3 ) = 𝑃 (𝑦𝑗4 ) = 𝑃 (𝑦𝑗5 ) = , (1)
indicator weights and calculated the posterior probability 5
6 Mathematical Problems in Engineering

where 𝑃(𝑦𝑗𝑖 )𝑖=1,2,...,5 is the prior probability of the 𝑗th (3) Calculate the posterior probability with (7):
indicator that belongs to the 𝑖th standard type.
(2) Calculate the likelihood of every indicator. Here, we 𝑃 (𝑦𝑗𝑖 ) 𝑃 (𝑥𝑗𝑘 | 𝑦𝑗𝑖 )
𝑃 (𝑦𝑗𝑖 | 𝑥𝑗𝑘 ) = , (7)
adopted the normal distribution sampling method ∑5𝑖=1 𝑃 (𝑦𝑗𝑖 ) 𝑃 (𝑥𝑗𝑘 | 𝑦𝑗𝑖 )
to estimate the likelihoods. The calculation processes
of the normal distribution sampling method were as
where 𝑃(𝑦𝑗𝑖 | 𝑥𝑗𝑘 ) is the likelihood of every indicator
follows.
that belongs to 𝑖th type.
(i) Calculate the variation coefficient 𝐶V𝑗 for the 𝑗th (4) Calculate the comprehensive probability of the multi-
indicator with (2): indicator WRV assessment of each subsystem with
(8):
(∑5𝑖=1 𝑦𝑗𝑖 ) 𝑚
𝑎𝑗 = ,
5 𝑃𝑖 = ∑ 𝜔𝑗 𝑃 (𝑦𝑗𝑖 | 𝑥𝑗𝑘 ) , (8)
𝑗=1
5 2
√ ∑𝑖=1 (𝑦𝑗𝑖 − 𝑎𝑗 ) (2)
𝜎𝑗 = , where 𝑃𝑖 is the weighted sum of the posterior
4 probability of the multi-indicator and represents the
𝜎𝑗 probability that WRV of the 𝑘th subsystem belongs
𝐶V𝑗 = , to the 𝑖th standard type, 𝜔𝑗 is the weight of the 𝑗th
𝑎𝑗
indicators, and 0 ≤ 𝜔𝑗 ≤ 1, ∑𝑚 𝑗=1 𝜔𝑗 = 1. Here, we
where 𝑖 is the number of standard types (𝑖 = used the AHP method to determine the weight 𝜔𝑗 of
1, 2, . . . , 5), 𝑗 is the number of indicators of a every indicator of each subsystem.
subsystem (𝑗 = 1, 2, . . . , 𝑛), 𝑦𝑗𝑖 is the standard (5) Calculate the posterior probability of the multisub-
value of the 𝑗th indicator of the 𝑖th standard system WRV assessment of the integrated water
type, 𝑎𝑗 is the mean value of the normal distri- resources system using (8) again. Here, 𝑃𝑖 is the
bution, and 𝜎𝑗 is the standard deviation. weighted sum of the posterior probability of four
(ii) Use 𝐶V𝑗 to obtain the variation coefficient of the subsystems and represents the probability that WRV
sampling values by assuming the following: of the integrated water resources system belongs to
the 𝑖th standard type, 𝜔𝑗 is the weight of the 𝑗th
𝐶V𝑗𝑖 = 𝐶V𝑗 , (3) subsystem, and 0 ≤ 𝜔𝑗 ≤ 1; ∑𝑚 𝑗=1 𝜔𝑗 = 1, 𝑚 = 4,
𝑘 = 1.
where 𝐶V𝑗𝑖 is the variation coefficient of the
sampling values of the 𝑖th type. 3.3. Other WRV Assessment Methods for Comparison. In
(iii) Calculate the standard deviation with (4), with order to verify the effectiveness of the proposed method
taking 𝑦𝑗𝑖 as the mean value of the 𝑖th type and compare the results among different methods, we chose
the other two WRV assessment methods for simultaneous
𝜎𝑗𝑖 = 𝐶V𝑗𝑖 ⋅ 𝑦𝑗𝑖 , (4) calculation, as follows.

where 𝜎𝑗𝑖 is the standard deviation of the 𝑖th (1) The GRA Method. The GRA method is an impact
type. evaluation model that measures the degree of similarity or
difference between two sequences based on the level of
(iv) Standardize the sampling value with (5) relation [33]. The calculation steps are as follows.
𝑥𝑗𝑘 − 𝑦𝑗𝑖 (i) Compose a reference sequence (i.e., (9)) and an
𝑡𝑗𝑘 = , (5)
𝜎𝑗𝑖 assessment sequence (i.e., (10)):

where 𝑥𝑗𝑘 is the value of the 𝑗th indicator of the 𝑥𝑖 (𝑘) = {𝑥𝑖 (1) , 𝑥𝑖 (2) , . . . , 𝑥𝑖 (𝑛)} (𝑖 = 1, 2, . . . , 𝑡) , (9)
𝑘th subsystem, and 𝑡𝑗𝑘 is the standardized value
of 𝑥𝑗𝑘 . 𝑦𝑗 (𝑘) = {𝑦𝑗 (1) , 𝑦𝑗 (2) , . . . , 𝑦𝑗 (𝑛)} (𝑗 = 1, 2, . . . , 𝑚) ,
(v) Calculate the likelihood 𝑃(𝑥𝑗𝑘 | 𝑦𝑗𝑖 ) of every (10)
indicator with (6)
where 𝑛 is the number of parameters, 𝑚 is the number
󵄨 󵄨
𝑃 (𝑥𝑗𝑘 | 𝑦𝑗𝑖 ) = 2 (1 − Φ (󵄨󵄨󵄨󵄨𝑡𝑗𝑘 󵄨󵄨󵄨󵄨)) , (6) of regions to be assessed, 𝑡 is the number of ranges
of parameters, 𝑥𝑖 (𝑘) is the observed value of the
where 𝑘 is the number of subsystems (𝑘 = 𝑘th parameter of the 𝑖th range, and 𝑦𝑗 (𝑘) is the
𝑡𝑗𝑘 representative value of the 𝑘th parameter of the 𝑗th
1, 2, . . . , 4), and Φ(|𝑡𝑗𝑘 |) = ∫−∞ (1/√2𝜋)𝑒−𝑢/2 𝑑𝑢. region.
Mathematical Problems in Engineering 7

(ii) Normalize the reference sequence with (11) and the (2) The PS Method. The PS method can be used to calculate
assessment sequence with (12) and (13): the vulnerability indicator, to identify the degree of the WRV.
The steps that are taken with this method are as follows.
{ 𝑥𝑖 (𝑘) − 𝑥1 (𝑘)
{
{ (i) Establish a reference system 𝐶𝑖 = (𝐶1 , 𝐶2 , . . . , 𝐶𝑛 ) that
{ 𝑥 (𝑘) − 𝑥1 (𝑘)
𝑥𝑖󸀠 (𝑘) = { 𝑡 (11) can be represented by the median of every indicator,
{
{ 𝑥 (𝑘) − 𝑥𝑖 (𝑘)
{ 1 . and an assessment system 𝑃𝑖 = (𝑃1 , 𝑃2 , . . . , 𝑃𝑛 ) that
{ 𝑥1 (𝑘) − 𝑥𝑡 (𝑘) uses the observed indicator values in the study region.

For an indicator for which the greater value means a (ii) Normalize the values of the assessment system by (17)
higher vulnerability, normalization can be performed with and (18).
(12): For the indicator that the greater the value, the more
vulnerable the system, normalization can be performed as
{ 0, 𝑦𝑗 (𝑘) ≤ 𝑥1 (𝑘)
{
{ follows:
{ 𝑦𝑗 (𝑘) − 𝑥1 (𝑘)
𝑦𝑗󸀠 (𝑘) = { , 𝑥1 (𝑘) < 𝑦𝑗 (𝑘) < 𝑥𝑡 (𝑘) (12) 𝑃𝑖
{
{ 𝑥𝑡 (𝑘) − 𝑥1 (𝑘)
{ 𝑋𝑖 =
𝐶𝑖
. (17)
{1, 𝑦𝑗 (𝑘) ≥ 𝑥𝑡 (𝑘) .

For an indicator for which the smaller value means a For the indicator that the smaller the value, the more
higher vulnerability, normalization can be performed with vulnerable the system, normalization is performed as follows:
(13): 𝐶𝑖
𝑋𝑖 = . (18)
0, 𝑦𝑗 (𝑘) ≥ 𝑥1 (𝑘) 𝑃𝑖
{
{
{
{ 𝑥1 (𝑘) − 𝑦𝑗 (𝑘)
󸀠
𝑦𝑗 (𝑘) = { , 𝑥1 (𝑘) > 𝑦𝑗 (𝑘) > 𝑥𝑡 (𝑘) (13)
{
{ 𝑥1 (𝑘) − 𝑥𝑡 (𝑘)
{ (iii) Calculate the WRV degree by (19):
{1, 𝑦𝑗 (𝑘) ≤ 𝑥𝑡 (𝑘) .
𝑉𝑖 = 𝜆 𝑖 × 𝑋𝑖 , (19)

(iii) Identify Δ 𝑗𝑖 (𝑘), Δ max , and Δ min with (14): where 𝑉𝑖 is the WRV value of a region and 𝜆 𝑖 is the
󵄨 󵄨 weight of indicator.
Δ 𝑗𝑖 (𝑘) = 󵄨󵄨󵄨󵄨𝑦𝑗󸀠 (𝑘) − 𝑥𝑖󸀠 (𝑘)󵄨󵄨󵄨󵄨 ,
󵄨 󵄨 For the multilayer indicator system, 𝑉𝑖 obtained from one
Δ max = max max 󵄨󵄨󵄨󵄨𝑦𝑗󸀠 (𝑘) − 𝑥𝑖󸀠 (𝑘)󵄨󵄨󵄨󵄨 , (14) layer appears as 𝑋𝑖 of its next layer; then we recalculate 𝑉𝑖
by using (19). Finally, we can obtain the WRV degree of the
󵄨 󵄨
Δ min = min min 󵄨󵄨󵄨󵄨𝑦𝑗󸀠 (𝑘) − 𝑥𝑖󸀠 (𝑘)󵄨󵄨󵄨󵄨 , integrated water resource system. The greater the value of 𝑉𝑖
is, the more vulnerable the region is.
where Δ 𝑗𝑖 (𝑘) is the difference between 𝑦𝑗󸀠 (𝑘) and
𝑥𝑖󸀠 (𝑘), and Δ max and Δ min represent the maximum 4. Results and Discussion
and minimum values among all the Δ 𝑗𝑖 (𝑘), respec-
4.1. WRV Assessment Results of Subsystems and the Integrated
tively.
Water Resources System Based on the Bayesian Method. To
(iv) Calculate the grey relational coefficient 𝜁𝑗𝑖 (𝑘): compare the calculation results of the Bayesian method with
those of the PS method and the GRA method [19], we
Δ min + 0.5Δ max chose the corresponding data in 2003 (see Table 2). With
𝜁𝑗𝑖 (𝑘) = . (15)
Δ 𝑗𝑖 (𝑘) + 0.5Δ max (7), the probability that each indicator of the subsystems
belonged to a certain WRV level were obtained. Then, the
(v) Calculate the grey relational grade 𝑟𝑗𝑖 : WRV probabilities of the subsystems were obtained with (8).
For the indicators of the hydrological subsystem (Figure 4),
𝑛
the lower the indicator value was, the larger the tendency
𝑟𝑗𝑖 = ∑ 𝜆 𝑖 𝜁𝑗𝑖 (𝑘) , (16) of a county to belong to a higher WRV was. Some WRV
𝑘=1
levels to which indicators of the subsystems belonged were
the same for most indicators of all nine counties (i.e., High
where 𝜆 𝑖 is the weight of the 𝑖th parameter.
WRV for the mean annual precipitation, Extreme WRV
(vi) Determine the maximum value 𝑟max of the grey for the groundwater recharge of unit area, and Medium
relational grades. The range of 𝑟max represents the WRV for the drought indicator), with the exception of two
vulnerability type which a region belongs to. For the indicators (i.e., the annual relative variability of precipitation,
multilayer indicator system, 𝑟𝑗𝑖 for one layer should and the mean water resources amount of unit area); see
be regarded as 𝜁𝑗𝑖 for its next layer and then be Figures 4(a1 ), 4(a2 ), 4(a3 ), 4(a4 ), and 4(a5 ). However, the
recalculated with (16). probabilities that each indicator belonged to some WRV
8 Mathematical Problems in Engineering

Table 2: Actual values of indicators for study area.

Subsystem Indicator 1 2 3 4 5 6 7 8 9
a1 387.20 406.40 406.40 409.10 405.70 373.60 456.50 385.80 446.80
a2 20.83 27.75 22.71 22.65 14.12 21.90 22.77 18.83 18.30
A a3 4.52 6.84 5.60 6.56 6.50 4.06 7.90 8.27 3.45
a4 2.58 4.76 3.09 4.80 2.43 2.94 2.95 5.68 1.65
a5 3.00 2.42 2.35 2.79 2.05 2.67 2.21 2.19 1.82
b1 68.54 474.45 294.77 458.45 438.82 271.44 667.83 438.90 652.50
b2 198.26 1158.04 1317.14 736.19 412.31 311.41 570.93 464.04 523.04
b3 194.44 336.77 422.60 485.15 193.68 138.14 296.07 372.02 250.80
B b4 1.49 1.96 9.25 7.81 3.81 12.30 1.85 17.19 6.44
b5 0.87 0.98 0.94 0.93 0.96 1.09 1.37 0.96 0.99
b6 174.41 67.61 106.64 100.58 48.33 45.46 48.19 86.01 33.69
b7 70.00 65.00 68.00 70.00 62.00 60.00 45.00 43.00 41.00
c1 76.24 50.94 47.59 59.18 76.83 58.36 71.69 51.00 56.37
c2 180.86 191.40 129.31 90.61 280.19 82.88 362.39 505.88 48.66
c3 3.46 1.13 0.32 2.56 0.40 0.42 0.23 1.47 0.62
C c4 49.82 50.03 71.39 63.65 61.54 89.99 42.86 0 66.64
c5 65.74 63.32 60.37 70.78 63.57 63.82 64.51 62.18 50.38
c6 70.00 50.00 65.00 68.00 50.00 55.00 35.00 35.00 30.00
d1 61.74 60.77 32.52 58.32 23.83 32.98 47.16 58.28 36.29
D d2 7.10 52.48 13.79 56.81 39.41 19.77 1.07 159.78 1.06
d3 89.64 82.62 76.07 81.47 56.2 78.58 88.00 60.20 75.00
Note: The symbolic descriptions for the 1st column of A∼D and the 2nd column of a1∼d3 are the same as those in Table 1; administrative divisions 1–9 are
shown in Figure 1.

level were different for the nine counties, for example, for Zhuolu County, which tended to belong to High WRV
a probability of 40.08%–46.51% belonging to High WRV (30.80%) (see Figure 5(b1 )). This finding meant that the
for the mean annual precipitation, while 26.56%–36.49% regional population increase had put great pressure on the
belonging to Extreme WRV for the groundwater recharge water resources system. For the other indicators except
of unit area. For the indicator of the mean water resources for the water resources per capita, there were different
amount of unit area, six counties belonged to Extreme WRV, WRV levels for each county. In general, the repetitive water
with the exception of Huai’an, Zhuolu, and Huailai Counties, utilization ratio of industry in this region tended to be Low
which belonged to High WRV (see Figure 4(a3 )). The WRV WRV or Medium WRV (89% altogether), which meant that
assessment results of the hydrological subsystem are shown the industrial water resources consumption was not serious.
in Figure 4(A). We can see that all of the nine counties had It is partly because the economic development mode of the
a high probability of belonging to High WRV (ranging from region was mainly based on agriculture. Although the water
27.07% to 33.48%), followed by Medium WRV (22.3%–25.4%) supply and demand ratio for counties had a tendency to
and Extreme WRV (14.4%–24.5%). This finding meant that belong to Low WRV (11.95%–41.20%) or No WRV (43.06%–
the WRV of the hydrological subsystem of the Zhangjiakou 87.89%) from Figure 5(b5 ), the water demand had been met
region was high. The hydrological subsystem reflected the by overextracting groundwater for a long time by the analysis
natural situations of regional water resources. The abundance, combined with Figure 4(a4 ). In this way, the water resources
stability, and hydrographic and meteorological conditions use is not sustainable. Through the probability calculation
had a comprehensive impact on the WRV degree of the with (7), while Yuxian and Chongli counties tended to belong
hydrological subsystem. Being located in arid and semiarid to Low WRV, Yangyuan and Chongli counties belonged to
areas, water supply available of the Zhangjiakou region No WRV while the other counties belonged to Extreme
was insufficient, although the stability of water resources WRV (see Figure 5(B)). It can be concluded that the rapid
system was not too bad. Specifically, water abundance and increase of population and low water utilization efficiency
groundwater recharge capacity were very weak, resulting in resulted in a high WRV of socioeconomic subsystem. As
a high WRV probability of the hydrological subsystem. The a basic water resources security concern, the groundwater
results were consistent with those of the GRA method and overdraft needs more attention for the purpose of sustainable
the actual situations [19]. water use in the future, although it temporarily alleviated the
For indicators of the socioeconomic subsystem shortage of water resources.
(Figure 5), water resources per capita of eight counties were For the indicators of the ecoenvironmental subsystem
low (approximately 399 m3 ), which had a high probabil- (Figure 6), the most possible two WRV levels for the indicator
ity of belonging to Extreme WRV (26.67%–54.02%), except of the ratio of soil and water loss were the same for all nine
Mathematical Problems in Engineering 9

0.5 0.5

0.4 0.4

Probability
0.3 0.3
Probability

0.2 0.2

0.1 0.1

0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division
(a1) Mean annual precipitation (a2 ) Annual relative variability
of precipitation
0.5 0.5

0.4 0.4

0.3
Probability

Probability

0.3

0.2 0.2

0.1 0.1

0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division
(a3 ) Mean water resources (a4 ) Groundwater recharge
amount of unit area of unit area
0.5
0.35

0.4 0.30

0.25
Probability
Probability

0.3
0.20

0.2 0.15

0.10
0.1
0.05

0.0 0.00
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division

No WRV High WRV No WRV High WRV

Low WRV Extreme WRV Low WRV Extreme WRV
Medium WRV Medium WRV
(a5 ) Drought index (A) Hydrological subsystem

Figure 4: Each indicator’s WRV and integrated WRV of the hydrological subsystem in the Zhangjiakou region (administrative divisions 1–9
are shown in Figure 1).
10 Mathematical Problems in Engineering

0.6
0.6
0.5
0.5
0.4
0.4
Probability

Probability
0.3
0.3

0.2
0.2

0.1 0.1

0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division
(b1) Water resources per capita (b2 ) Water consumption per unit GDP
0.5 0.5

0.4 0.4

0.3 0.3
Probability

Probability

0.2 0.2

0.1 0.1

0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division
(b3 ) Water consumption per capita (b4 ) Difference between planned and
actual water utilization amount
0.9 0.9

0.8 0.8

0.7 0.7

0.6 0.6
Probability
Probability

0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division

No WRV High WRV No WRV High WRV

Low WRV Extreme WRV Low WRV Extreme WRV
Medium WRV Medium WRV
(b5 ) Water supply and demand ratio (b6 ) Water resources development
and utilization ratio

Figure 5: Continued.
Mathematical Problems in Engineering 11

0.5 0.35

0.30
0.4
0.25
0.3
Probability

Probability
0.20

0.2 0.15

0.10
0.1
0.05

0.0 0.00
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division

No WRV High WRV No WRV High WRV

Low WRV Extreme WRV Low WRV Extreme WRV
Medium WRV Medium WRV
(b7 ) Repetitive utilization ratio of industrial water (B) Socioeconomic subsystem

Figure 5: Each indicator’s WRV and integrated WRV of the socioeconomic subsystem in the Zhangjiakou region (administrative divisions
1–9 are shown in Figure 1).

counties (i.e., Extreme WRV ranked first with a high prob- increased, leading to the deterioration of water environment
ability of 62.54%–79.18%, and High WRV ranked second). in Zhangjiakou region. In addition, polluted groundwater
The WRV levels were different from each other for other would cause the deterioration in ecoenvironment. All these
indicators, including the land development ratio, the fertilizer factors had resulted in WRV of ecoenvironmental subsystem.
use intensity, the pollution index of surface water, the ratio To decrease the WRV, an extensive mode of economic
of polluted groundwater exceeding the standard, and the dis- development should be changed.
posal ratio of sewage, due to their respective ecoenvironment For the indicators of the hydraulic engineering subsystem
conditions. Because most of the large counties were located (Figure 7), the water-saving irrigation ratio was usually high,
in remote mountainous areas, their fertilizer use intensity with 67% of the counties belonging to Medium WRV or Low
was not excessively large. Therefore, as for the indicator of WRV and 22% belonging to High WRV (Wanquan County
the fertilizer use intensity, except for the fact that Huailai and Yangyuan County) (see Figure 7(d1 )). Because the sit-
County tended to belong to High WRV, others belonged to uations of water resources and hydraulic engineering were
Medium WRV (Figure 6(c2 )). In addition, the pressure of different for the nine counties, the WRV for the indicators
economic development had greatly aggravated the pollution of the regulating capacity of the surface water for hydraulic
of the surface water, leading to a high probability for the engineering and storage was different (see Figure 7(d2 )). For
indicator of the pollution of the surface water belonging to the indicator of the storage and release capacity of reservoirs,
Medium WRV (39.52%), High WRV (38.68%–51.08%), and except for Huailai County belonging to Low WRV and Yuxian
Extreme WRV (77.68%–99.99%) (Figure 6(c3 )). For example, County belonging to Medium WRV, the other counties
all of the three most prosperous counties in this region, that is, belonged to No WRV (see Figure 7(d3 )). Overall, hydraulic
Zhangjiakou City, Xuanhua County, and Huailai County, had engineering played an important role in the decline of the
Extreme WRV. Among the nine counties, Zhangjiakou City WRV; more than 50% of the counties, including Zhangjiakou
had the largest population and the largest economic output. City, Huai’an County, Xuanhua County, Zhuolu County, and
Except for the fact that the indicator of the disposal ratio Huailai County, had Low or No WRV in this region due to the
of sewage for Zhangjiakou City had a Low WRV level, its reallocation and adjustment of the flood by means of water
WRV levels for the other indicators were usually higher by conservancy facilities (see Figure 7(D)).
more than 1-2 levels compared with other counties, leading With (8), the WRV of the integrated water resources
to an Extreme WRV for the comprehensive assessment system of the Zhangjiakou region was obtained (see Figure 8).
of the ecoenvironmental subsystem. Overall, the ecological We can see the following: (1) WRV of the ecoenvironmental
environment of the Zhangjiakou region was very vulnerable; subsystem of Zhangjiakou City and XuanHua County was
except for Zhuolu County belonging to the Medium WRV, more vulnerable than other counties, which led to a high
all of the other counties belonged to the Extreme WRV or tendency to belong to Extreme WRV (26.8% and 25.0%,
High WRV (Figure 6(C)). With the rapid development of resp.); other counties belonged to High WRV ranging from
industry and agriculture, waste water discharge and extensive 24.6% to 27.0%; (2) there were no areas belonging to No WRV,
use of fertilizers and pesticides in agricultural production had Low WRV, and Medium WRV, indicating that the WRV of the
12 Mathematical Problems in Engineering

0.5 0.5

0.4 0.4

0.3
Probability

Probability
0.3

0.2 0.2

0.1 0.1

0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division
(c1) Land development ratio (c2 ) Fertilizer use intensity

1.0 0.6

0.9
0.5
0.8
0.7
0.4
0.6
Probability
Probability

0.5 0.3
0.4
0.2
0.3
0.2 0.1
0.1
0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division
(c3 ) Pollution index of surface water (c4 ) Ratio of polluted groundwater
exceeding the standard

0.8 0.5

0.7
0.4
0.6

0.5
Probability

0.3
Probability

0.4
0.2
0.3

0.2
0.1
0.1

0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division

No WRV High WRV No WRV High WRV

Low WRV Extreme WRV Low WRV Extreme WRV
Medium WRV Medium WRV
(c5 ) Ratio of soil and water loss (c6 ) Disposal ratio of sewage

Figure 6: Continued.
Mathematical Problems in Engineering 13

0.5

0.4

0.3

Probability
0.2

0.1

0.0
1 2 3 4 5 6 7 8 9
Administrative division

No WRV High WRV

Low WRV Extreme WRV
Medium WRV
(C) Ecoenvironmental subsystem

Figure 6: Each indicator’s WRV and integrated WRV of the ecoenvironmental subsystem in the Zhangjiakou region (administrative divisions
1–9 are shown in Figure 1).

Table 3: Results comparison between the Bayesian method, the GRA method, and the PS method for the integrated water resources system.

Administrative division Bayesian method GRA method PS method

Zhangjiakou City Extreme WRV Extreme WRV Extreme WRV
Huai’an County High WRV High WRV Extreme WRV
Wanquan County High WRV High WRV Extreme WRV
Xuanhua County Extreme WRV High WRV Extreme WRV
Yuxian County High WRV High WRV Extreme WRV
Yangyuan County High WRV High WRV Extreme WRV
Zhuolu County High WRV High WRV High WRV
Huailai County High WRV High WRV Extreme WRV
Chongli County High WRV High WRV Extreme WRV

integrated water resources system of the Zhangjiakou region some indicators, and the resulting WRV levels were higher
was very serious. WRV was an objective attribute of the water than those use the other two methods.
resources system, whose vulnerability came from its internal The results of the Bayesian method and the GRA method
characteristics and external driving forces. The assessment were almost the same, with the exception of the assessment
results above were consistent with the actual situation of the for Xuanhua County. The two methods divided the sample
water resources in the Zhangjiakou region, which showed data into several grades, to determine the vulnerable degree
that the Bayesian method can be well used for regional WRV of the water resources in a large range, which can fully reflect
assessment. the uncertainties in the water resources system. However,
the GRA method had a high requirement of assessment
4.2. Results Comparison with the PS Method and the GRA and standard data. The assessment results for WRV were
Method. A comparison of the assessment results among the the possibility or tendency of the adverse impact of natural
Bayesian method, the GRA method, and the PS method disasters and human activities on the water resources system,
was conducted (Table 3). It can be seen that the WRV of 6 which can reflect the uncertain and fuzzy characteristics
regions with the PS method was higher than those with the of the WRV reasonably. At the same time, it was very
Bayesian method, and 7 regions were higher than those with simple and quick, which made it more convenient and
the GRA method. The PS method had the advantages of a reliable for the WRV assessment. In the practical calculation,
simple principle and simple linear calculation procedure for multiple methods should be used comprehensively for WRV
both the assessment and reference systems, but it can easily assessment, to provide a more reliable decision-making basis
generate distorted results due to magnifying the influence of for water resources management.
14 Mathematical Problems in Engineering

0.5 1.0
0.9

0.4 0.8
0.7
0.6

Probability
0.3
Probability

0.5
0.2 0.4
0.3
0.1 0.2
0.1
0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division
(d1) Water-saving irrigation ratio (d2 ) Ratio of design utilizable storage of
reservoir to surface water quantity

0.7 0.5

0.6
0.4
0.5
Probability

0.3
Probability

0.4

0.3
0.2
0.2
0.1
0.1

0.0 0.0
1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9
Administrative division Administrative division

No WRV High WRV No WRV High WRV

Low WRV Extreme WRV Low WRV Extreme WRV
Medium WRV Medium WRV
(d3 ) Ratio of effective storage to total storage of reservoirs (D) Hydraulic engineering subsystem

Figure 7: Each indicator’s WRV and integrated WRV of the hydraulic engineering subsystem in the Zhangjiakou region (administrative
divisions 1–9 are shown in Figure 1).

5. Conclusions Bayesian method were consistent with the actual situation.

This finding indicated that the Bayesian method was feasible
A Bayesian method that used the sampling normal distribu- for the WRV assessment with a simple calculation process.
tion method and the AHP method was developed to assess The Zhangjiakou region undertook the important mis-
the WRV of the Zhangjiakou region. The results showed that sion of providing a water supply for Beijing City, but the
(1) the vulnerability of the region was very high; Zhangjiakou water resources condition was not good. Water scarcity and
City and Xuanhua County belonged to Extreme WRV with water pollution had seriously affected the sustainable devel-
the probabilities of 26.8% and 25.0%, respectively, while other opment of the regional agricultural production and economic
counties belonged to High WRV (24.6%–27.0%). This finding development. To reduce the WRV and enhance the ability
indicated that the socioeconomic development had put a of self-recovery of the water environment and to improve
large amount of pressure on the water resources. Under sustainable utilization of water resources of the Zhangjiakou
the approximately same natural conditions for these nine region, effective countermeasures must be taken from the
adjacent areas, the economy developed more rapidly, and the following aspects: (1) developing a water-saving industry and
WRV will become more serious. (2) Additionally, the results technology for the sewage purification, to improve the reuse
of WRV assessment of the Zhangjiakou region by using the level of the waste water and to construct a water-saving
Mathematical Problems in Engineering 15

0.35 3 1
1
0.30 1
1
0.25 2 4
8
Probability

0.20

0.15
7
6
0.10

0.05
5
0.00
1 2 3 4 5 6 7 8 9
Administrative division

No WRV High WRV

Low WRV Extreme WRV Extreme WRV
Medium WRV High WRV
(a) (b)

Figure 8: WRV of the integrated water resources system in the Zhangjiakou region: (a) WRV probabilities; (b) WRV spatial distribution
(administrative divisions 1–9 are shown in Figure 1).

society, which can reduce the water environment deterio- Acknowledgments

ration and shortage of water resources in the Zhangjiakou
region; (2) increasing the vegetation coverage, which would The research was funded by the National Science Founda-
enhance the capacity for soil and water preservation, and tion for Innovative Research Group (no. 51121003), and the
strengthening controls on nonpoint pollution and population National Natural Science Foundation of China (50509001).
growth, to promote the sustainable development of a social The authors would like to extend special thanks to the editor
economy and ecoenvironment; (3) raising public environ- and the anonymous reviewers for their valuable comments in
mental awareness for the effective protection and sustainable greatly improving the quality of this paper.
utilization of water resources.
Due to insufficient information, we assumed that the References
prior probability of an indicator that belonged to the different
vulnerability grades was equal. However, their prior proba- [1] Z. J. Li, Q. W. Chen, Q. Xu, and K. Blanckaert, “Generalized like-
bilities were different in practice. The quantity of indicator lihood uncertainty estimation method in uncertainty analysis of
numerical eutrophication models: take bloom as an example,”
data that was available largely limits the accuracy of the
Mathematical Problems in Engineering, vol. 2013, Article ID
assessment. Thus, acquiring more indicator data could help 701923, 9 pages, 2013.
to improve the precision of the WRV assessment. In addition,
[2] X. Wang, F. B. Ma, and J. Y. Li, “Water resources vulnerability
there is a threshold of the damage to the water resources
assessment based on the parametric-system method: a case
system; the WRV assessment can provide only a level range study of the Zhangjiakou region of guanting reservoir Basin,
(such as No WRV and Extreme WRV), and it is a future task North China,” Procedia Environmental Sciences, vol. 13, pp.
to determine the damage threshold through a mechanism 1204–1212, 2012.
study. [3] K. Jinno, Z. X. Xu, A. Kawamura, and K. Tajiri, “Risk assessment
of water supply system during drought,” International Journal of
Conflict of Interests Water Resources Development, vol. 11, no. 2, pp. 185–204, 1995.
[4] Z. Xu, K. Jinno, A. Kawamura, S. Takesaki, and K. Ito, “Perfor-
The authors declare that there is no conflict of interests mance risk analysis for Fukuoka water supply system,” Water
regarding the publication of this paper. Resources Management, vol. 12, no. 1, pp. 13–30, 1998.
16 Mathematical Problems in Engineering

[5] Z. X. Xu, K. Takeuchi, H. Ishidaira, and X. W. Zhang, “Sus- [20] F. B. Ma, X. Wang, and C. H. Li, “Research progress in water
tainability analysis for Yellow River water resources using the resources vulnerability assessment,” Journal of Water Resources
system dynamics approach,” Water Resources Management, vol. and Water Engineering, vol. 23, no. 1, pp. 30–37, 2012 (Chinese).
16, no. 3, pp. 239–261, 2002. [21] F. Zong, H. G. Xu, and H. Y. Zhang, “Prediction for traffic
[6] L. L. Liu, “Concept and quantitative assessment of vulnerability accident severity: comparing the Bayesian network and regres-
of water resource,” Bulletin of Soil and Water Conservation, vol. sion models,” Mathematical Problems in Engineering, vol. 2013,
22, no. 2, pp. 41–44, 2002 (Chinese). Article ID 475194, 9 pages, 2013.
[7] J. D. Arthur, H. A. R. Wood, A. E. Baker, J. R. Cichon, and G. L. [22] T. Chan, H. Ross, S. Hoverman, and B. Powell, “Participatory
Raines, “Development and implementation of a Bayesian-based development of a Bayesian network model for catchment-based
aquifer vulnerability assessment in Florida,” Natural Resources water resource management,” Water Resources Research, vol. 46,
Research, vol. 16, no. 2, pp. 93–107, 2007. no. 7, Article ID W07544, 2010.
[8] B. Kattaa, W. Al-Fares, and A. R. Al Charideh, “Groundwater [23] S. S. Qian, A. Schulman, J. Koplos, A. Kotros, and P. Kellar,
vulnerability assessment for the Banyas Catchment of the Syrian “A hierarchical modeling approach for estimating national
coastal area using GIS and the RISKE method,” Journal of distributions of c hemicals in public drinking water systems,”
Environmental Management, vol. 91, no. 5, pp. 1103–1110, 2010. Environmental Science and Technology, vol. 38, no. 4, pp. 1176–
[9] A. Sorichetta, M. Masetti, C. Ballabio, S. Sterlacchini, and 1182, 2004.
G. P. Beretta, “Reliability of groundwater vulnerability maps [24] R. Wu, S. S. Qian, F. Hao, H. Cheng, D. Zhu, and J. Zhang,
obtained through statistical methods,” Journal of Environmental “Modeling contaminant concentration distributions in China’s
Management, vol. 92, no. 4, pp. 1215–1224, 2011. centralized source waters,” Environmental Science & Technology,
[10] A. Sorichetta, M. Masetti, C. Ballabio, and S. Sterlacchini, vol. 45, no. 14, pp. 6041–6048, 2011.
“Aquifer nitrate vulnerability assessment using positive and [25] F. B. Ma, C. H. Li, X. Wang, Z. F. Yang, C. C. Sun, and P. Y. Liang,
negative weights of evidence methods, Milan, Italy,” Computers “A Bayesian method for comprehensive water quality evaluation
and Geosciences, vol. 48, pp. 199–210, 2012. of the Danjiangkou Reservoir water source area, for the middle
[11] D. Mirauda and M. Ostoich, “Surface water vulnerability assess- route of the South-to-North Water Diversion Project in China,”
ment applying the integrity model as a decision support system Frontiers of Earth Science, vol. 8, no. 2, pp. 242–250, 2014.
for quality improvement,” Environmental Impact Assessment [26] M. Kalantarnia, F. Khan, and K. Hawboldt, “Dynamic risk
Review, vol. 31, no. 3, pp. 161–171, 2011. assessment using failure assessment and Bayesian theory,”
[12] C. Yu, Y. Yao, G. Hayes, B. Zhang, and C. Zheng, “Quantitative Journal of Loss Prevention in the Process Industries, vol. 22, no.
assessment of groundwater vulnerability using index system 5, pp. 600–606, 2009.
and transport simulation, Huangshuihe catchment, China,” [27] D. L. Kelly and C. L. Smith, “Bayesian inference in probabilistic
Science of the Total Environment, vol. 408, no. 24, pp. 6108–6116, risk assessment-the current state of the art,” Reliability Engineer-
2010. ing and System Safety, vol. 94, no. 2, pp. 628–643, 2009.
[13] J. J. Martı́nez-Bastida, M. Arauzo, and M. Valladolid, “Intrinsic [28] “Zhangjiakou bureau of hydrology and water resources sur-
and specific vulnerability of groundwater in central Spain: the vey,” Second Water Resources Assessment Report of Zhangji-
risk of nitrate pollution,” Hydrogeology Journal, vol. 18, no. 3, pp. akou City, Hebei Institute of Architectural Engineering, Hebei
681–698, 2010. Province, 2004, (chinese).
[14] J. Zhou, G. Li, F. Liu, Y. Wang, and X. Guo, “DRAV model and its [29] X. Wang, Q. Cui, and S. Y. Li, “An optimal water allo-
application in assessing groundwater vulnerability in arid area: cation model based on water resources security assessment
a case study of pore phreatic water in Tarim Basin, Xinjiang, and its application in Zhangjiakou Region, Northern China,”
Northwest China,” Environmental Earth Sciences, vol. 60, no. 5, Resources, Conservation and Recycling, vol. 69, pp. 57–65, 2012.
pp. 1055–1063, 2010.
[30] M. A. Hamouda, M. M. Nour El-Din, and F. I. Moursy,
[15] P. Martı́nez-Santos, M. R. Llamas, and P. E. Martı́nez-Alfaro,
“Vulnerability assessment of water resources systems in the
“Vulnerability assessment of groundwater resources: a model-
Eastern Nile Basin,” Water Resources Management, vol. 23, no.
ling-based approach to the Mancha Occidental aquifer, Spain,”
13, pp. 2697–2725, 2009.
Environmental Modelling and Software, vol. 23, no. 9, pp. 1145–
1162, 2008. [31] P. E. Waggoner, Climate Change and US Water Resources, John
Wiley & Sons, New York, NY, USA, 1990.
[16] C. Butscher and P. Huggenberger, “Enhanced vulnerability
assessment in karst areas by combining mapping with modeling [32] F. V. Jensen, An Introduction to Bayesian Networks, vol. 210, UCL
approaches,” Science of the Total Environment, vol. 407, no. 3, pp. Press, London, UK, 1996.
1153–1163, 2009. [33] J. W. K. Chan and T. K. L. Tong, “Multi-criteria material selec-
[17] A. K. Antonakos and N. J. Lambrakis, “Development and tions and end-of-life product strategy: grey relational analysis
testing of three hybrid methods for the assessment of aquifer approach,” Materials and Design, vol. 28, no. 5, pp. 1539–1546,
vulnerability to nitrates, based on the drastic model, an example 2007.
from NE Korinthia, Greece,” Journal of Hydrology, vol. 333, pp.
288–304, 2007.
[18] H.-N. Zhu, X.-Z. Yuan, J. Liang, G.-M. Zeng, and H.-W. Jiang,
“An integrated model for assessing the risk of water environ-
mental pollution based on fuzziness,” China Environmental
Science, vol. 31, no. 3, pp. 516–521, 2011 (Chinese).
[19] J. Y. Li, Study on vulnerability assessment of water resources of
guanting reservoir basin [M.S. thesis], Beijing Normal Univer-
sity, Beijing, China, 2007, (chinese).
Advances in Advances in Journal of Journal of
Operations Research
Hindawi Publishing Corporation
Decision Sciences
Hindawi Publishing Corporation
Applied Mathematics
Hindawi Publishing Corporation
Algebra
Hindawi Publishing Corporation
Probability and Statistics
Hindawi Publishing Corporation
[Link] Volume 2014 [Link] Volume 2014 [Link] Volume 2014 [Link] Volume 2014 [Link] Volume 2014

The Scientific International Journal of

World Journal
Hindawi Publishing Corporation
Differential Equations
Hindawi Publishing Corporation
[Link] Volume 2014 [Link] Volume 2014

Submit your manuscripts at

[Link]

International Journal of Advances in

Combinatorics
Hindawi Publishing Corporation
Mathematical Physics
Hindawi Publishing Corporation
[Link] Volume 2014 [Link] Volume 2014

Journal of Journal of Mathematical Problems Abstract and Discrete Dynamics in

Complex Analysis
Hindawi Publishing Corporation
Mathematics
Hindawi Publishing Corporation
in Engineering
Hindawi Publishing Corporation
Applied Analysis
Hindawi Publishing Corporation
Nature and Society
Hindawi Publishing Corporation
[Link] Volume 2014 [Link] Volume 2014 [Link] Volume 2014 [Link] Volume 2014 [Link] Volume 2014

International
Journal of Journal of
Mathematics and
Mathematical
Discrete Mathematics
Sciences

Journal of International Journal of Journal of

Hindawi Publishing Corporation Hindawi Publishing Corporation Volume 2014

Function Spaces
Hindawi Publishing Corporation
Stochastic Analysis
Hindawi Publishing Corporation
Optimization
Hindawi Publishing Corporation
[Link] Volume 2014 [Link] [Link] Volume 2014 [Link] Volume 2014 [Link] Volume 2014