JES-00826; No of Pages 6

J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX

Available online at www.sciencedirect.com

ScienceDirect
www.elsevier.com/locate/jes

1 Water quality assessment in Qu River based on fuzzy water

2 pollution index method

F
Ranran Li, Zhihong Zou⁎, Yan An

O
3Q1

4 School of Economics and Management, Beihang University, Beijing 100191, China. E-mail: liranran1101@163.com

O
5

87 AR TIC LE I NFO ABSTR ACT

R
9 Article history: A fuzzy improved water pollution index was proposed based on fuzzy inference system and 14

P
10 Received 13 January 2016 water pollution index. This method can not only give a comprehensive water quality rank, 15
11 Revised 11 March 2016 but also describe the water quality situation with a quantitative value, which is convenient 16
12 Accepted 15 March 2016 for the water quality comparison between the same ranks. This proposed method is used 17
D
13 Available online xxxx to assess water quality of Qu River in Sichuan, China. Data used in the assessment were 18
collected from four monitoring stations from 2006 to 2010. The assessment results show 19 Q2
E
27 Keywords: that Qu River water quality presents a downward trend and the overall water quality in 2010 20
28 Water quality assessment is the worst. The spatial variation indicates that water quality of Nanbashequ section is the 21
29 Fuzzy inference
T

pessimal. For the sake of comparison, fuzzy comprehensive evaluation and grey relational 22
30 Water pollution index method were also employed to assess water quality of Qu River. The comparisons of these 23
31
C

Fuzzy comprehensive evaluation three approaches' assessment results show that the proposed method is reliable. 24
32 © 2016 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. 25
Published by Elsevier B.V. 26
E

34
35
33
36
R

38
37 Introduction quality index method (Huang, 2001; Yi and Yu, 2003; Qiu 56
et al., 2013) is used by many provinces and cities hydrology 57
R

39 Water shortage problem in developing countries is more and department for water quality assessment of drinking water 58
40 more serious in recent years (Ongley, 1998). Especially in source, which is proposed by Ministry of Water Resources, 59
41 China, severe water shortage problems and a large amount of Monitoring and Evaluation Center of Water Environment. 60
O

42 sewage make the situation of surface water pollution exceed- However, both these two methods have limitations. Single 61
43 ingly serious (Lindberg et al., 2014; Zhou et al., 2014). River factor evaluation method gives the water quality rank, but is 62
C

44 water quality is becoming an important factor to hinder social unable to quantify the water quality changes; water quality 63
45 and economic development. Therefore, how to analyze and index method can not only give the water quality rank, but 64
N

46 assess water quality accurately is of great significance both to also reflect the spatial and temporal variations of water qual- 65
47 society and economy. ity condition. But it needs to consider too many parameters. 66
48 A good water quality assessment method should not only In recent years, many new methods have been continually 67
U

49 provide the water quality rank, but also accurately reflect the applied in the evaluation of water quality, such as multivar- 68
50 spatial and temporal variations of water quality condition. At iate statistical techniques (Shrestha and Karama, 2007; Huang 69
51 the same time, a water quality assessment method which can et al., 2010), artificial neural network (Ip et al., 2009; Yan et al., 70
52 be widely used in environmental management should be easy 2010), and grey evaluation method (Shi et al., 2012). Never- 71
53 to calculate and master, in addition to the scientificity and theless, most of these methods' calculation processes are 72
54 accuracy. Single factor evaluation method is widely adopted trivial. Then, water pollution index (WPI) method (Liu et al., 73
55 by environmental protection department in China. Water 2013; Liu and Wu, 2014) is proposed. This method simplifies 74

⁎ Corresponding author. E-mail: zouzhihong@buaa.edu.cn (Zhihong Zou).

http://dx.doi.org/10.1016/j.jes.2016.03.030
1001-0742/© 2016 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. Published by Elsevier B.V.

Please cite this article as: Li, R., et al., Water quality assessment in Qu River based on fuzzy water pollution index method, J.
Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.030
 2 J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

75 the water quality index method and is extended to general Table 1 – Environmental Quality Standards for Surface t1:1
76 water quality assessment. The method is able to quantify Water of China (GB3838-2002). t1:2
77 t1:3
the water quality condition; however, it continues the theory Rank DO CODMn BOD5 NH3-N TP t1:4
78 of single factor evaluation method and uses the most polluted (mg/L) (mg/L) (mg/L) (mg/L) (mg/L)
79 indicator as the basis of water quality rank. So the assess-
I 7.5 2 3 0.15 0.02 t1:5
80 ment results are relatively conservative. After that, an im- II 6.0 4 3 0.50 0.10 t1:6
81 proved water pollution index method (IWPI) (Li et al., 2014) III 5.0 6 4 1.00 0.20 t1:7
82 is presented. In this method, each indicator is given a weight IV 3.0 10 6 1.50 0.30 t1:8
83 by entropy, and the section IWPI is the weighted sum of all V 2.0 15 10 2.00 0.40 t1:9
84 indicators' WPIs. According to the quantitative results, the
85 water quality can not only identify the water quality rank south, then enters Guang'an County, Huaying City, Hechuan 132
86 intuitively, but also can reflect the spatial and temporal County and finally empties into Jialing River. The upstream 133

F
87 variations of the water quality. is called Nanjiang River, next to Enyang River in Bazhong City 134
88 Methods mentioned above are all deterministic methods. is called Ba River, and next to Zhou River in Qu County is called 135

O
89 As we all know, it may be not accurate to use deterministic Qu River. It runs for 720 km, and covers a watershed area 136
Q5
90 methods deal with complex and changeable environment of 39,200 km2, with the annual average flow of 663 m3/sec. Qu 137

O
91 problems. In view of the limitations and complexities of County locates in the southwest of Dazhou City, and is adjacent 138
92 deterministic models, fuzzy logic, capable of integrating and to Guang'an, Nanchong and Bazhong City. It covers 2013 km2 139
93 accounting for the inaccurate, vague, qualitative and fuzzy with 60 towns. And the population is about 145 million. 140

R
94 information, has been increasingly applied to environmental Four monitoring stations in Qu County were selected in 141
95 issues in recent years (Chau, 2006). Fuzzy logic was first this study. They are Xipingcun (Station A) in Sanhui Town, 142

P
96 introduced by Zadeh (1965) and then has been widely used in Lianhuacun (Station B) in Huinan Town, Jinjikou (Station C) 143
97 many fields. It is appropriate for developing environmental in Qujiang Town and Nanbashequ (Station D) in Qujiang 144
98 indices, due to its ability to reflect human thoughts and Town. Dissolved oxygen (DO), permanganate index (CODMn),
D 145
99 expertise, and its capacity to deal with uncertain, ambiguous five day biochemical oxygen demand (BOD5), ammonia nitro- 146
100 and subjective information. Furthermore, it is also a reliable gen (NH3-N) and total phosphorus (TP) were selected to assess 147
E
101 method to report the assessment results in linguistic terms for the water quality of Qu River. Environmental Quality Stan- 148
102 decision-makers with no expertise. Therefore, development of dards for Surface Water of China (GB3838-2002) is shown in 149
103 Table 1, and the data of the five parameters at four stations 150
T

environmental indices based on fuzzy logic have drawn much

104 attention (Karmakar and Mujumdar, 2006; Sowlat et al., 2011), from 2006 to 2010 are shown in Table 2 (Shi et al., 2012). 151
C

105 especially in water quality (Chang et al., 2001; Ocampo-Duque

106 et al., 2006, 2007; Zou et al., 2006; Icaga, 2007; Lermontov et al., 1.2. Water pollution index method 152

107 2009). For example, Ocampo-Duque et al. proposed a method-

E

108 ology based on fuzzy inference systems to assess water quality Water pollution index method is based on single factor 153
109 (Ocampo-Duque et al., 2006). Gharibi et al. developed a novel evaluation method. According to water quality ranks and 154
R

110 water quality index based on fuzzy logic, which is a compre-

111 hensive artificial intelligence approach to the development Table 2 – Concentrations of five indicators monitored from t2:1
t2:2
R

112 of environmental indices for routine assessment of surface four stations from 2006 to 2010.
t2:3
113 water quality (Gharibi et al., 2012). Year Station DO CODMn BOD5 NH3-N TP t2:4
(mg/L) (mg/L) (mg/L) (mg/L) (mg/L)
O

114 Fuzzy water pollution index (FWPI) method is proposed

115 in this article. This method modified the improved water 2006 A 7.63 1.60 1.52 0.16 0.067 t2:5
116 pollution index method using fuzzy inference system. In this B 6.96 4.31 3.58 0.62 0.142 t2:6
C

117 method, all indicators' weights are incorporated in setting C 6.42 3.95 3.07 0.62 0.150 t2:7
118 rules, rather than simply using the heaviest pollution indica- D 6.98 4.32 3.64 0.39 0.173 t2:8
N

119
Q3 tor represents the overall water quality condition. The assess- 2007 A 8.41 1.52 0.98 0.15 0.018 t2:9
B 7.16 1.49 1.41 0.17 0.038 t2:10
120 ment results can be obtained directly by the fuzzy inference.
C 6.18 2.56 2.33 0.36 0.024 t2:11
121
U

Then this proposed method is used to assess water quality of

D 5.34 2.87 3.24 0.37 0.125 t2:12
122 Qu River in Sichuan, China. The assessment results are com- 2008 A 6.86 1.95 1.76 0.07 0.095 t2:13
123 pared with those obtained by fuzzy comprehensive evaluation B 7.15 1.43 1.39 0.11 0.074 t2:14
124 method and grey relational method. C 5.47 3.47 3.02 0.18 0.132 t2:15
D 5.90 2.58 3.81 0.17 0.127 t2:16
2009 A 6.51 2.86 1.62 0.09 0.085 t2:17
B 6.06 1.98 1.35 0.42 0.064 t2:18
126
125 1. Materials and methods
C 5.65 4.46 3.76 0.66 0.045 t2:19
D 5.97 4.39 3.10 0.53 0.062 t2:20
127 1.1. Study area 2010 A 6.52 2.80 1.33 0.18 0.039 t2:21
B 5.36 3.88 3.41 0.57 0.107 t2:22
128 Qu River is a tributary of the Jialing River. It originates from C 5.09 4.12 3.72 0.46 0.115 t2:23
129
Q4 the south of Micang Maintain, which locates at the junction D 4.38 3.57 4.26 0.73 0.254 t2:24
130 of Sichuan province and Shaanxi provinces. Qu River flows Data sources are from the report by Shi et al., 2012. t2:25
t2:26
131 through eight counties and twenty-one towns from north to

Please cite this article as: Li, R., et al., Water quality assessment in Qu River based on fuzzy water pollution index method, J.
Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.030
 J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX 3

155 their corresponding WPI limiting values (the corresponding X

m
200
199
f ij ¼ bij = bij ; k ¼ 1= lnm ð8Þ
156 WPI limiting values of Rank I, II, III, IV and V are 20, 40, 60, 80, j¼¼1
157 and 100, respectively) (Huang, 2001; Yi and Yu, 2003; Liu and
158 Wu, 2014), all evaluation indicators' WPIs of each section are If fij = 0 , fij ln fij = 0. 202
201
203
159 obtained by interpolation method, then the maximum value Lastly, the weights are computed. The ith evaluation 204
160 of all the indicators' WPIs is selected as the section WPI. indicator's weight wi is 205
161 As mentioned by Liu et al. (2013), the concrete calculation
!
162 process of WPI is as follows. Xn
wi ¼ ð1−Hi Þ= n− Hi ð9Þ
163 Each indicator's WPI:
i¼1

CðiÞ−Cl ðiÞ n
WPIðiÞ ¼ WPIl ðiÞ þ  20 ð1Þ where, 0 ≤ wi ≤ 1; ∑ wi ¼ 1. 207
206
Cu ðiÞ−Cl ðiÞ
i¼1

F
1.3. Fuzzy water pollution index method 208
165
164 where, C(i) is the monitoring value of the ith indicator, Cl(i)

O
166 and Cu(i) are the lower and upper limiting values of the ith
Next fuzzy inference system will be used to improve this 209
167 indicator's rank in Table 1 respectively, and Cl(i) ≤ C(i) ≤ Cu(i)
method. In a fuzzy inference system, a quantitative numerical 210

O
168 (for DO, Cu(i) ≤ C(i) ≤ Cl(i)), WPIl(i) is the corresponding lower WPI
value is fuzzified into a qualitative state and processed by 211
169 value of the ith indicator, i = 1 , ⋯ , n.
an inference engine, through sets and operators, rules in a 212
170 Besides, when the standard values of two ranks are the

R
qualitative sphere, allowing the use of information such as 213
171 same in Table 1, interpolation on low score value range is
individual knowledge and experience, and permitting quali- 214
172 used.
215

P
tative environmental parameters and factors to be integrated
173 Section WPI:
and processed. Relevant knowledge of fuzzy inference can be 216
WPI ¼ maxðWPIðiÞÞ ð2Þ found in many literatures (Yen and Langari, 1998; Ross, 2004; 217
218
D
Ocampo-Duque et al., 2006; Gharibi et al., 2012; Turksen, 1991;
174
175
176 After that, an IWPI method (Li et al., 2014) is presented. IWPI Karr and Gentry, 1993; Jang and Sun, 1995; Yager and Filev, 219
177 method improves the section WPI, that is, each indicator's 1994; Ross, 2004), which will not be described here in detail. 220
E
178 WPI is also calculated according to (1), but the section IWPI is A fuzzy inference system can be divided into three parts: 221
fuzzification, fuzzy inference, and defuzzification (Li, 2006). 222
T

X
n
The fuzzification process involves the transformation of inputs 223
IWPI ¼ wi  WPIðiÞ ð3Þ
from a numerical value of a variable into a membership grade, 224
C

i¼1
which describes a property of the variable. The fuzzy inference 225
180
179 where wi is the weight of the ith indicator, which is calculated by includes the fuzzy operations of multiple-part antecedents, 226
E

181 entropy method (Zou et al., 2006). The steps of using entropy the implication methods from the antecedent to the conse- 227
Q6
182 weight method to determine weight are as follows (Qiu, 2002): quent for every rule, and an aggregation method to join the 228
R

183 Firstly, normalize the original data matrix. Suppose there consequents across all the rules. Finally, defuzzification con- 229
184 are m evaluation sections for the evaluation river, and each sists in transforming the fuzzy output into a non-fuzzy numer- 230
185 ical value which can be used in non-fuzzy contexts (Silvert, 231
R

section has n evaluation indicators, then the judgment matrix

186 is 2000; Ocampo-Duque et al., 2006). 232
  Fig. 1 shows the evaluation process of FWPI method. The 233
O

R ¼ rij nm ð4Þ antecedent sets (DO, CODMn, BOD5, NH3-N and TP) and the 234
consequent set (FWPI) were created by trapezoid (I and V sets) 235
188
187 where rij is the ith evaluation indicator's measured value of
C

and triangular pertinence (all others) functions; the five 236

189 the jth evaluation section. Normalize the judgment matrix parameters ranks were as Table 1 and the corresponding five 237
190 and the matrix B is obtained. bij is the element of B, if the ranks, I, II, III, IV and V were obtained; the FWPI ranks referred 238
N

191 indicator value is the larger the better, bij is defined to the WPI ranks. That is, the FWPI value ranges from 0 to 239
Q7
  100 and water quality can be classified as I(0–20), II(20–40), 240
rij − min rij
U

j
bij ¼     ð5Þ
max rij − min rij
j j

193
192
194 otherwise,
 
max rij −rij
j
bij ¼    : ð6Þ
max rij − min rij
j j

196
195
197 Second, define the entropy. The ith evaluation indicator's
198 entropy Hi is

X
m
Hi ¼ −k f ij ln f ij ; i ¼ 1; 2; ⋯; n ð7Þ Fig. 1 – Evaluation process diagram of fuzzy water quality
j¼1 index method.

Please cite this article as: Li, R., et al., Water quality assessment in Qu River based on fuzzy water pollution index method, J.
Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.030
 4 J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

Table 3 – Evaluation results of FWPI method. t3:1

t3:3
t3:2
Year Station FWPI Rank t3:4

2006 A 14.05 I t3:5

B 37.17 II t3:6
C 29.75 II t3:7
D 36.48 II t3:8
2007 A 11.58 I t3:9
B 18.03 I t3:10
C 14.17 I t3:11
D 36.80 II t3:12
2008 A 15.19 I t3:13
B 15.02 I t3:14
Fig. 2 – Membership functions of DO. C t3:15

F
31.70 II
D 33.19 II t3:16
2009 A 14.04 I t3:17

O
B 25.08 II t3:18
241 III(40–60), IV(60–80) and V(80–100). Only membership func-
C 21.08 II t3:19
242 tions of DO (Fig. 2) was given as follows, and for the sake of
D 24.72 II t3:20

O
243 brevity, membership functions of CODMn, BOD5, NH3-N and TP 2010 A 18.49 I t3:21
244 were no longer listed here. The output fuzzy sets for inference B 33.10 II t3:22
245

R
and FWPI ranks were shown in Fig. 3. C 34.99 II t3:23
246 This fuzzy inference system has 125 rules. Weights are D 42.42 III t3:24
247 embodied in the fuzzy rules. Being impossible to write them

P
FWPI: fuzzy water pollution index. Rank I, II, III refer to Table 1. t3:25
t3:26
248 all in this paper, some examples are given as below:

249 If DO is “V” and CODMn is “I” then Gr.1 is “III”. D

250 If BOD5 is “I” and NH3-N is “I” then Gr.2 is “I”. downward trend. The water quality of Station A retained Rank 267
251 If Gr.1 is “III” and Gr.2 is “I” then Gr.4 is “II”. I in five years, without obvious change. The water quality of 268
252 If Gr.4 is “II” and Gr.4 is “II” then FWPI is “II”.
E
Stations B and C have some fluctuation between Rank I and 269
Rank II. The water quality of Station D declines from Rank II to 270
253 Defuzzification of the outputs was carried out by using
271
T

254 the center of gravity method, which is the most convention- Rank III.
According to the temporal and spatial variations of the 272
255 ally and physically applicable method for defuzzification. Its
C

four stations from 2006 to 2010, the overall water quality 273
256 derivation is based on the following equation (Ross, 2004):
in 2010 and the water quality of Station D were the worst, 274
Z
respectively. That attributes to rapid population increase 275
μðzÞzdz
E

Z¼ Z ð10Þ and fast industrial development, which causes the increase 276
μðzÞdz of wastewater discharge. Station D is the place where the 277
R

population and industrial factories are most intensive of Qu 278

258
257
259 All the computations were processed using the “fuzzy logic County. A large number of domestic sewage and industrial 279
R

260 toolbox” in MATLAB2012 (Li, 2006). wastewater are discharged into the river. Therefore the water 280
quality of Station D is the worst. 281
O

In order to validate the effectiveness of the FWPI method, 282

262
261 2. Results and discussion the results by fuzzy comprehensive evaluation and grey rela- 283
tional model were compared with those by the FWPI (shown 284
C

263 Table 3 shows the evaluation results of Qu River by FWPI in Table 4). The results of grey relational model method were 285

264 method. According to the FWPI value and Fig. 3, the not computed, but transcribed from reference (Shi et al., 2012). 286
N

265 corresponding ranks can be obtained. The results indicated The outcome indicates that the evaluation results get from 287

266 that from 2006 to 2010, Qu River water quality presents a this method are generally in agreement with the other two 288
289
U

methods. However, some differences exist, for example, at

Station A in 2009, the results by the proposed method and M3 290
are both Rank I while that by M2 is Rank II; at Station C in 291
2009, the results by this method and M2 are both Rank II while 292
that by M3 are Rank I. From the monitoring data of all five 293
indicators at Station A in 2009, it is both reasonable whether 294
the water quality is identified as Rank I or Rank II. At Station C 295
in 2009, according to the principle of fuzzy membership degree, 296
BOD5 is in Rank III, TP is in Rank I, and other indicators are all 297
in Rank II, so the water quality should be Rank II. By contrast 298
and careful analysis, the results of FWPI method are almost 299
consistent with other two common water quality assessment 300
Fig. 3 – Output fuzzy sets for inference and FWPI ranks. methods, so this method is reliable and can provide effective 301
FWPI: fuzzy water pollution index. assessment for decision-making in water quality. 302

Please cite this article as: Li, R., et al., Water quality assessment in Qu River based on fuzzy water pollution index method, J.
Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.030
 J O U RN A L OF E N V I RO N ME N TA L S CI EN CE S X X (2 0 1 6 ) XX X–XXX 5

t4:1 Table 4 – Results and comparison of three methods.

t4:3
t4:2
t4:4 2006 2007 2008 2009 2010

t4:5 A B C D A B C D A B C D A B C D A B C D

t4:6 M1 I II II II I I I II I I II II I II II II I II II III
t4:7 M2 I II II II I I I II I I II II II II II II I II II III
t4:8 M3 I II II II I I I II I I II II I II I II I II II III

t4:9 M1 represents fuzzy water pollution index method; M2 represents fuzzy comprehensive evaluation method; M3 represents grey relational
t4:10 model method.
t4:13
t4:12
t4:11 Rank I, II, III refer to Table 1.

F
Chau, K.W., 2006. A review on integration of artificial intelligence 345
304
303 3. Conclusions into water quality modelling. Mar. Pollut. Bull. 52 (7), 726–733. 346

O
Gharibi, H., Mahvi, A.H., Nabizadeh, R., Arabalibeik, H., Yunesian, 347
305 The fuzzy water pollution index (FWPI) method is developed. M., Sowlat, M.H., 2012. A novel approach in water quality 348
306 This method corrects perceived deficiencies of water quality assessment based on fuzzy logic. J. Environ. Manag. 112 (24), 349

O
307 assessment and water resources management when the con- 87–95. 350
Huang, D.L., 2001. New evaluation method for drinking water 351
308 ventional, deterministic methods can be inaccurate or concep-
quality in China. Hydrology 21 (z1), 62–64. 352

R
309 tually limited. In addition to giving the overall water quality Huang, F., Wang, X.Q., Lou, L.P., 2010. Spatial variation and source 353
310 rank, this method also describes the overall water quality apportionment of water pollution in Qiantang River (China) 354

P
311 condition with a quantitative value, which is convenient for using statistical techniques. Water Res. 44 (5), 1562–1572. 355
312 comparison between different sections of the same ranks and Icaga, Y., 2007. Fuzzy evaluation of water quality classification. 356
313 making decisions. The proposed method is applied to assess Ecol. Indic. 7 (3), 710–718. 357
Ip, W.C., Hu, B.Q., Wong, H., Xia, J., 2009. Applications of grey 358
314 water quality of Qu River in Sichuan, China. Fuzzy comprehen-
D relational method to river environment quality evaluation in 359
315 sive evaluation and grey relational model are also employed to
China. J. Hydrol. 379 (3), 284–290. 360
316 assess the water quality. It can be seen that the results of this
E
Jang, J.S.R., Sun, C.T., 1995. Neuro-fuzzy modeling and control. 361
317 proposed FWPI is almost consistent with those of other two Proc. IEEE 83 (3), 378–406. 362
318 methods, and can be used for water quality assessment in China. Karmakar, S., Mujumdar, P.P., 2006. Grey fuzzy optimization 363
T

319 Analysis on the water quality of Qu River suggested that model for water quality management of a river system. Adv. 364
320 improving the sewage treatment rate, optimizing the city life Water Resour. 29 (7), 1088–1105. 365
C

Karr, C.L., Gentry, E.J., 1993. Fuzzy control of pH using genetic 366
321 and rural life layout, and controlling the pollution of industry,
algorithms. IEEE Trans. Fuzzy Syst. 1 (1), 46–53. 367
322 and agriculture, should be carried out to prevent and control
Lermontov, A., Yokoyama, L., Lermontov, M., Machado, M.A.S., 368
E

323 water pollution, and then the ecological environment of Qu 2009. River quality analysis using fuzzy water quality index: 369
324
Q8 River can be gradually improved. Ribeira do Iguape river watershed, Brazil. Ecol. Indic. 9 (6), 370
R

325 The fuzzy water pollution index method achieves assess- 1188–1197. 371
326 ment by setting the fuzzy rules. However a large number of Li, G.Y., 2006. Intelligent Control and Matlab. Publishing House of 372
Electronics Industry, Beijing. 373
R

327 parameters will lead to tremendous increase of inference

Li, R.R., Zou, Z.H., An, Y., 2014. Water quality evaluation of Yanhe 374
328 rules. To solve this problem, it is applicable to reduce the
River based on the improved water pollution index. The 375
329 dimension of parameters firstly using some methods such as
O

Twelfth International Conference on Industrial Management, 376

330 principal component analysis. The authors are still working in pp. 275–278. 377
331 the development of an index with more parameters for a more Lindberg, R.H., Östman, M., Olofsson, U., Grabic, R., Fick, J., 2014. 378
C

332 realistic evaluation of water body. Occurrence and behaviour of 105 active pharmaceutical 379
ingredients in sewage waters of a municipal sewer collection 380
system. Water Res. 58 (3), 221–229. 381
N

Liu, H., Wu, J.J., 2014. Water pollution index method applied in the 382
334
Q9
333 Uncited reference
evaluation of river water quality. J. Anhui Agric. Sci. 42 (21), 383
384
U

7164.
335 Wang, 2008 Liu, Y., Zheng, B.H., Fu, Q., Luo, Y.P., Wang, M., 2013. Application of 385
water pollution index in water quality assessment of rivers. 386
Environ. Monit. China 29 (3), 49–55. 387
337
336 Acknowledgments Ocampo-Duque, W., Ferre-Huguet, N., Domingo, J.L., 388
Schuhmacher, M., 2006. Assessing water quality in rivers with 389
338 This work was supported by the National Natural Science fuzzy inference systems: a case study. Environ. Int. 32 (6), 390
339 Foundation of China (No. 51478025). 733–742. 391
Ocampo-Duque, W., Schuhmacher, M., Domingo, J.L., 2007. A 392
neural-fuzzy approach to classify the ecological status in 393
34 0 REFERENCES surface waters. Environ. Pollut. 148 (2), 634–641. 394
Ongley, E., 1998. Modernization of water quality programs in 395
developing countries: issues of relevancy and cost efficiency. 396
341
342 Chang, N.B., Chen, H.W., Ning, S.K., 2001. Identification of river Water Qual. Int. 37–42 (Sep/Oct). 397
343 water quality using the fuzzy synthetic evaluation approach. Qiu, W.H., 2002. Management Decision and Applied Entropy. 398
344 J. Environ. Manag. 63 (3), 293–305. China Machine Press, Beijing, pp. 193–196. 399

Please cite this article as: Li, R., et al., Water quality assessment in Qu River based on fuzzy water pollution index method, J.
Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.030
 6 J O U RN A L OF E N V I RO N ME N TA L S CIE N CE S X X (2 0 1 6 ) XXX –XXX

400 Qiu, M.L., Liu, L.H., Zou, X.W., Wu, L.X., 2013. Comparison of water Wang, P.P., 2008. Water quality index method using for Quanzhou 421
401 quality evaluation standards and evaluation methods between drinking water source evaluation. Environ. Sustain. Dev. (2), 48–51. 422
402 at home and abroad. J. China Inst. Water Resour. Hydropower Yager, R., Filev, D., 1994. Essentials of Fuzzy Modeling and Control. 423
403 Res. 11 (3), 176–182. John Wiley & Sons Inc., New York. 424
404 Ross, T.J., 2004. Fuzzy Logic With Engineering Applications. John Yan, H., Zou, Z., Wang, H., 2010. Adaptive neuro fuzzy inference 425
405 Wiley & Sons, New York. system for classification of water quality status. J. Environ. Sci. 426
406 Shi, J.P., Li, X., Wang, W., 2012. Study on space–time variety of 22 (12), 1891–1896. 427
407 water environment quality based on gray relational model. Yen, J., Langari, R., 1998. Fuzzy Logic: Intelligence, Control, and 428
408 Guangdong Agric. Sci. 39 (4), 111–117. Information. Prentice Hall, United States. 429
409 Shrestha, S., Karama, F., 2007. Assessment of surface water Yi, W., Yu, Q., 2003. Discussion about water quality evaluation 430
410 quality using multivariate statistical techniques: a case study index method in drinking water source. Environ. Monit. China 431
411 of the Fuji river basin, Japan. Environ. Model. Softw. 22 (4), 19 (5), 43–47. 432
412 464–475. Zadeh, L.A., 1965. Fuzzy Sets Inform. Control 8 (65), 338–353. 433
413 Silvert, W., 2000. Fuzzy indices of environmental conditions. Ecol. Zhou, H.B., Chen, T.B., Gao, D., Zheng, G.D., Chen, J., Pan, T.H., 434

F
414 Model. 130 (1–3), 111–119. 2014. Simulation of water removal process and optimization of 435
415 Sowlat, M.H., Gharibi, H., Yunesian, M., Tayefeh Mahmoudi, M., aeration strategy in sewage sludge composting. Bioresour. 436

O
416 Lotfi, S., 2011. A novel, fuzzy-based air quality index Technol. 171C, 452–460. 437
417 (FAQI) for air quality assessment. Atmos. Environ. 45 (12), Zou, Z.H., Yun, Y., Sun, J.N., 2006. Entropy method for 438
418 2050–2059. determination of weight of evaluating indicators in fuzzy 439

O
419 Turksen, I.B., 1991. Measurement of membership functions and synthetic evaluation for water quality assessment. J. Environ. 440
420 their acquisition. Fuzzy Sets Syst. 40 (1), 5–38. Sci. 18 (5), 1020–1023. 441

R
442

P
D
E
T
C
E
R
R
O
C
N
U

Please cite this article as: Li, R., et al., Water quality assessment in Qu River based on fuzzy water pollution index method, J.
Environ. Sci. (2016), http://dx.doi.org/10.1016/j.jes.2016.03.030