Research on resilience capacity assessment of urban water supply system in China under ood and drought disaster

The evaluation index system of urban water supply system is further improved, and the indexes are qualitatively 9 classified and quantitatively analyzed. The cloud model is used to simulate the four indexes of the study case to ensure the reliability of the simulation. ⚫ The whole water supply system is added as one of the toughness evaluation indexes to ensure the integrity of 12 evaluation. ⚫ The toughness of urban water supply system was evaluated and compared by two methods. 14 ⚫ According to the four system dimensions of urban water supply system, the paper puts forward the targeted promotion 15 strategy. Abstract 17 Under the influence of global climate change, urban flood and drought disasters occur frequently, so that is 18 extremely important to construct the resilience capacity of urban water supply system. Based on the framework of 19 system toughness and capability analysis, using correlation analysis and factor analysis to construct the index 20 system of resilience capacity assessment of urban water supply system, which reflects the four dimensions of 21 water source, water plants, water supply and distribution network and users, and five dimensions of social, natural 22 environment, economy, physics and organization, and the weights of all indexes are given. The multi - level comprehensive evaluation model based on cloud model and the index comprehensive evaluation index method 24 based on entropy weight were used to evaluate the resilience of the water supply system in Qingdao under flood 25 and drought disasters, and the evaluation results of the two methods were compared. Finally, based on the 26 evaluation results, the influencing factors of water supply system resilience were analyzed, and the corresponding 27 strategies to improve the resilience of water supply


31
For the past few years, the global climatic variation is severe, and extreme weather events occur frequently. The frequency 32 and impact degree of urban flood and drought disasters is gradually increasing, and the urban water supply system is facing 33 more severe challenges [1] . Especially in China, on August 16, 2011, high temperatures occurred in eastern Sichuan, 34 Chongqing, most of Hubei, northeastern Guizhou and most of the south of the Yangtze River. Eighty-seven counties (cities 35 and districts) in Guizhou Province were affected by the drought to varying degrees, and more than 20 million people 36 affected by the drought were facing drinking water difficulties. On July 21, 2012, heavy rain hit most parts of China, with 37 Beijing and its surrounding areas hit by the heaviest rainstorms in 61 years. As a result of the torrential rains, 62 counties 38 region was 30~50% less than that in the same period of the year. In addition, affected by the drought in the previous three 44 years, the underground water level continued to decline, and the reservoirs generally lacked water. As of September 30, the 45 total volume of large and medium-sized reservoirs in the four cities of Jiaodong was 653 million cubic meters, 54.6% less 46 than that in the same period of the year. There were 21 large and medium-sized reservoirs below the dry or dead water level, 47 accounting for 95% of the total number of dry reservoirs in the province. In July, Qingdao and Weifang were once 107,000 48 cubic meters short of daily water supply, and 150 million cubic meters short of daily water supply, which seriously affected 49 normal urban water use. In late August 2018, affected by typhoons "Capricorn" and "Vimbiya", many places in Weifang. 50 Shandong Province suffered from torrential rain and flood disasters rarely seen in history, causing serious damage to water 51 supply and power facilities. Thus, in the event of sudden disaster, the urban water supply system becomes extremely fragile. 52 The failure of the urban water supply system not only affects the normal production and life of the city, but also may lead to 53 the collapse of the entire city operation [2] . With the continuous acceleration of China's urbanization, how to effectively 54 ensure the safety of urban water supply and improve the resilience of the water supply system has become an instant 55 problem to be solved.Therefore, this article from the perspective of urban water supply system under inundation and 56 drought resilience, for urban water supply system, based on the correlation analysis and factor analysis to build resilience 57 ability evaluation system for urban water supply system, the multi-level evaluation model based on cloud model of urban 58 water supply system resilience ability are analyzed in the simulation, It is expected to provide theoretical support for 59 toughness evaluation of water supply and distribution system in different cities. The toughness capacity assessment 60 framework of urban water supply system constructed of this research is shown in Figure 1. 61 Literature analysis correlation analysis factor analysis Setting up index system of resilience capacity assessment of urban water supply system Select the index original data of resilience capacity assessment of urban water supply system  Figure 1. Toughness capacity assessment framework of urban water supply system(From top to bottom, the figure  63 shows the logical process and methods of toughness evaluation of urban water supply system in four parts) 64 Research status of resilience capacity assessment of urban water supply system standards. 101 • Literature Visualization Analysis 102 CiteSpace 5.7.R2 software was used to conduct a visual analysis of the literature related to urban water supply system in the 103 recent 10 years, and the results were shown in Figure 2. The study found that,urban water supply system usually includes 104 subsystems such as water source, water plants, water supply and distribution pipe network and users. However, at present, 105 there are many researches on toughness evaluation of water source system and water supply and distribution pipe network 106 system in urban water supply system, and a series of research achievements have been made to enrich the connotation of 107 toughness of water supply and distribution system. However, research often focuses on a certain subsystem in the water 108 supply system, and there are few studies on the construction of the resilience index system and the resilience assessment of 109 the entire water supply system under flood and drought disasters. Index system of resilience capacity assessment of urban water supply system 114 This paper improves the resilience assessment framework of urban water supply systems constructed by Balaei [18] and 115 Lukuba [19] et al., and applies it to the assessment of resilience capacity of urban water supply systems in China(This is 116 shown in Figure 1). On the basis of considering the whole process management of urban water supply system (water source, 117 water plants, water supply and distribution pipe network system and user), fully consider the minimum necessary associated 118 urban management systems closely related to urban water supply system (as shown in Figure 3), such as social, natural 119 environment, economic, physical, organizational and other systems.The toughness capacity evaluation index system of 120 urban water supply system has been further improved. 121 Therefore, this paper argues that, the toughness capacity of water supply and distribution system is defined as the ability of 122 the water supply and distribution system to withstand disasters, reduce disaster losses, and reasonably allocate resources to 123 quickly recover normal water supply from disasters. 124 According to the principles of comparability, representativeness, feasibility, relevance, non-reproducibility and conformance 129 to toughness implication, the toughness evaluation indexes of urban water supply system were selected [20] . Based on the 130 above principles, an open "cylinder index selection model" was established in this paper, as showed in Figure 4. After layers 131 of filtration, toughness indexes of urban water supply system were selected. Basic data preparation --audition 137 According to the above selection model and selection principles for toughness indicators, all the indicators related to the 138 toughness of the water supply system, including the water source, water supply and water consumption, can be included in 139 the alternative index system. However, due to various reasons, some index data of the selection are unavailable or 140 discontinuous. For example, the degree coefficient of process equipment at the water plant is not available. The net value 141 and original value data of fixed assets of Tianjin, Xinjiang and Gansu are not available in China Urban Water Supply 142 Yearbook, so it is impossible to calculate the degree coefficient of process equipment.Therefore, some of the indicators in the audition should be selected first, and those indicators whose data are unavailable or discontinuous should be eliminated. By calculating the Pearson correlation coefficient between each index and water resources per capita, comprehensive 154 production capacity of water supply, water consumption of urban residents, water consumption of industrial enterprises and 155 the quantity of people affected by flood and drought, the correlation degree and significance test value can be obtained. The 156 threshold value of the significance level test was set at 0.01 to carry out strict screening. Delete by the analysis of 157 population density, population and 6 years old above college degree and above 6 years old,, water supply pipe network 158 leakage rate loss modulus, water conservancy facilities as a result of small, regional GDP index correlation index, finally be 159 related to urban water supply system resilience has 42 indicators (five kinds of index overlap, overlapping indicators 160 calculated only once). 161

162
Since the two indexes of total water supply and total water use are highly correlated, it is not appropriate to conduct a factor 163 analysis. After eliminating the total water use index after comprehensive consideration, factor analysis is conducted for the 164 remaining 41 indexes. Before factor analysis of the data, KMO and Bartlett tests should be carried out, and the final results 165 are shown in Table 1 According to the analysis in the above table, 9 common factors with feature roots greater than 1 were extracted through 172 principal component analysis, and the single contribution rate of these 9 common factors changed slightly after the factor 173 rotation, but the cumulative value of the contribution rate was 82.77%, which did not change before and after the factor 174 rotation. Combined with the factor load matrix, the correlation between these 9 common factors and specific indicators is 175 further analyzed, so as to further classify and reduce the dimension of indicators. According to the impact load coefficient, 176 the load coefficient is set to be above 0.5. Finally, 14 indexes are deleted and 27 indexes are retained, this is shown in Table  177 3. 178

Water source B1
Reservoir capacity at the year-end C1 Quantity of permanent residents at the year-end C2

Urbanization rate C3
Water resources per capita C4 Water consumption per 10,000 RMB of industrial added value C5 Water consumption per 10,000 RMB GDP C6

Water plants B2
Domestic water consumption of urban residents C7 Comprehensive production capacity of water supply C8 Personnel employed in urban units in the management of water conservancy, environment and public facilities C9 Urban sewage treatment rate C10 Total water supply C11 Investment in waste water treatment project has been completed C12 Water supply and distribution network B3 Length of water supply pipe C13 Density of water supply pipeline in built-up area C14 Investment in fixed assets of water conservancy, environment and public facilities management industry C15 Users B4 The quantity of people affected by floods and droughts C16 Percentage of urban basic medical insurance coverage at year-end C17 Quantity of people enrolled in unemployment insurance C18 Quantity of community health service centers C19 State funds for education C20

GDP per capita C21
Per capita disposable income of urban residents C22 More old population dependency ratio C23 Urban registered unemployment rate C24 Natural population growth rate C25 Economize water consumption C26 Water consumption exceeding the planned quota C27 Table 3. Urban water supply system index classification 179 By referring to relevant toughness theoretical literature and combining with the above factor analysis results, the research 180 team discussed together and regarded the above 9 common factors as the 9 factors affecting the toughness of the water 181 supply and distribution system and classified and named them. Combined with the above toughness capability analysis 182 framework, the following 5 index system dimensions were finally formed: Organization--Factor No. • Construction of toughness ability index system of urban water supply system 187 Based on the above composition of urban water supply system, as showed in Figure 3, and combined with the result of 188 factor analysis, construct an index system for the resilience of urban water supply systems, as showed in Table 5   data each year. Therefore, the comprehensive evaluation value lacks comparability in the longitudinal comparison, which 220 affects the final evaluation result [22] . Therefore, this article put forward considering the index data of different areas more 221 than one year averaged them, based on this data to get the weight of a unified, and the weight as a fixed value, when to stay 222 also adopt the unified when the appraisal object weight to calculate the toughness value, so quickly to evaluate an area 223 water supply and distribution system resilience, And make the calculated results more comparable. 224 Therefore, when calculating the unified weight, this paper adopts the original data of toughness index of the water supply 225   Let u be a quantitative domain expressed by numerical value, and C be a qualitative concept on U. If the quantitative value 235 x∈U is a random realization of the qualitative concept C, the certainty of x for C, u(x)∈[0,1] is a random number with a 236 stable tendency： 237 Then the distribution of x in the domain U is called the cloud model (cloud for short), decided as C(x); Each x is called a 239 cloud droplet. 240 When representing a concept as a whole, three digital features are used to realize it, namely, expectation Ex, entropy En and 241 hypermetropy He, as showed in Figure 5. When determining the membership degree, the traditional fuzzy membership degree is a fixed value. However, when the 246 cloud model is used to calculate the membership degree of the index in the cloud, the membership degree of the index for 247 the evaluation set is not accurate and unique, thus reducing the subjectivity and difficult [24] . 248 Establishment steps of multi-level comprehensive evaluation model based on cloud model [25][26] : 249 (1) Factor field U and comment field V are established for the evaluation objects. 250 (2) The index weight W calculated in Section 2.3 is adopted. 251 (3) A single factor evaluation was conducted between U and V, and a fuzzy relational matrix R was established. Let the 252 factor i(i=1, 2, …, n) corresponding grade j(j=1, 2, …, m)the upper boundary value is xij and the lower boundary value 253 is x'ij, then the qualitative concept of level j corresponding to factor I can be represented by the normal cloud model, where: 254 Since the boundary value is the transition value of two adjacent levels, and the membership degree of the two levels is equal, 256 there are: 257 In the formula: 268 Qingdao has more people and less water, and the spatial and temporal distribution of precipitation is uneven. Especially in 278 recent years, severe water supply crisis caused by extreme weather has appeared. With the rapid economic growth and the 279 continuous improvement of urbanization, the safety of the water supply has become an important factor restricting the 280 sustainable economic and social development of Qingdao. Therefore, Qingdao has built a large water supply and 281 distribution system of "three sources" raw water supply, "four vertical and three horizontal" pipe network transmission and 282 distribution in the main urban area, and "one ring and three lines" unified allocation in Qingdao. This paper takes Qingdao 283 as an example. The data are from 2010-2019 "Qingdao Water Resources Bulletin", "Qingdao Statistical Yearbook", 284 "Qingdao Statistical Bulletin", "Shandong Province Statistical Yearbook" and so on. At present, the distribution of water 285 resources in Qingdao is shown in Figure 6. The status quo of the main water supply projects in Qingdao is shown in Table 7.    According to the established toughness capability index system and index standard, the normal cloud model is used to 293 represent the grade standard of each index in Equations (9) and (11). 294 Indicators with higher weights were successively selected, such as per capita water resource C4, completed investment of 295 wastewater treatment project C12, length of water supply pipeline C13, and water consumption saving C26. Equation (1) and 296 cloud matrix R were used to establish the normal cloud membership function of evaluation index standard, as showed in 297  Finally, according to steps (5) and (6) of the cloud model, the comprehensive evaluation results are obtained, as showed in 307 Table 8 and Figure 8. In order to ensure the integrity of the evaluation, the whole water supply system (B0) is added as one 308 of the toughness evaluation indexes. 309 Year  The toughness of the water plant system has been at a low level for ten years. The toughness of the water supply and 318 distribution pipe network system has been stable and at a medium level from 2010 to 2018, and has risen to a higher level in 319 2019. Since 2012, toughness of all subsystems has been ranked in order. The user system and the water supply and 320 distribution pipe network system have the same toughness and rank the highest, followed by the water plant system and the 321 water source system. 322

324
The weight of the regional water supply and distribution system adopts the unified weight value Wj calculated by the 325 Entropy Weight, and the toughness evaluation index system of the regional urban water supply system constructed in 326 Section 3.2. The evaluation indexes go through the process of data acquisition, input and data standardization. 327 Here the original index data is standardized by the extremum method (the Min-max normalization method) to eliminate the 328 dimensionality effect. 329 The normalized index rij of j evaluation indexes in the i year was calculated, and then the comprehensive evaluation index V 330 =Wj * rij was synthesized.In the case of 2019, the calculation process is similar for the other years. The final toughness 331   Table 9. Toughness capability evaluation indexes are comprehensive evaluation index 333 It can be seen from Figure 9 that, in general, toughness of Qingdao's water supply and distribution system during the period 336 from 2010 to 2019 showed a wavy upward trend, although some years showed a decline.Each subsystem toughness index 337 change trend of water supply, water supply subsystem comprehensive index of the whole year of a downward trend and the 338 final year of the initial drop apparently, user subsystem comprehensive index is on the rise and growth of apparent, 339 waterworks subsystem have been falling and rising trend, for water distribution network system is slightly rising trend, the 340 growth is not obvious. 341 • Comparison of the two evaluation results 342 By comparing the multi-level comprehensive evaluation results of the cloud model with those of the comprehensive 343 evaluation index method of entropy weight index, it can be seen that the latter is to calculate the fixed comprehensive index 344 value, the index value and index weight into positive correlation, while the reflect of the subsystems of the water supply and 345 distribution system and the overall dynamic change trend, However, it is difficult to describe in detail the degree to which 346 each evaluation unit belongs to a certain level.For example, in the example, the index per capita water resource has a large 347 weight. If the index comprehensive evaluation index method based on Entropy Weight is adopted to calculate, it will have a 348 large and positive impact on the toughness of the system. If the multi-level evaluation model of the cloud model is used, the 349 toughness level is at a low level, which has a great but negative impact on the toughness of the system.Moreover, the 350 multi-level evaluation result based on cloud model is more flexible, which examines the extent to which the evaluation unit 351 belongs to a certain level, and the boundary of the level also changes within a certain acceptable range. All subsystems in 352 the water supply and distribution system determine the toughness of a water supply and distribution system. However, the 353 overall toughness of the water supply and distribution system is not simply composed of the superposition of all subsystems, 354 but the interaction and mutual influence of all subsystems. Therefore, toughness capability is more in line with the fuzzy 355 connotation of the cloud model theory.