| Harbin is the capital of Heilongjiang province named “Commodity Grain Base of China” and plays an important role in China's social stability and economic development.With the acceleration of urbanization and the increasing intensity of agricultural development,the water resources system of Harbin has undergone significant changes in recent years.The deterioration of surface water quality,over-exploitation of groundwater,and frequent droughts and floods have seriously hindered the steady development of Harbin.Aiming at a series of problems existing at present,this article has conducted in-depth research on the hydrological factors of Harbin using multiple methods,revealed the temporal and spatial characteristics of the complexity changes of hydrological factors in Harbin.The risk of drought in Harbin was analyzed from the perspective of complexity.Harbin's water resources allocation model was constructed from the perspective of complexity,and the model was solved using intelligent optimization algorithms.In order to improve the trend fitting process in the traditional multifractal detrended fluctuation analysis(MF-DFA),the complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)and the wavelet packet transform(WPT)were combined to construct the CEEMDAN-WPT-MFDFA algorithm and analyzes the temporal and spatial characteristics of monthly precipitation,monthly evaporation,and the monthly runoff complexity in twelve counties of Harbin.The results show that there are significant differences in the monthly precipitation complexity of each sub-region in Harbin.The complexity of the northwest is the lowest and its predictability is the best.The complexity and predictability of the middle and Midwest are about average.The complexity of the southeast is higher than that of the northwest,middle and Midwest and its predictability is worse.The complexity of Shuangcheng is the highest and its predictability is the worst of all the studied sub-regions.In addition,the evaporation complexity in the northwest of Harbin is the highest.The evaporation complexity is higher in the central and southern regions and the central region.The evaporation complexity is lowest in the north-central and northeastern parts of Harbin.Overall,the monthly evaporation sequence in the southeastern region is less complex than in the northwest.The monthly runoff sequence complexity of each river station in Harbin has obvious differences.In terms of rivers,the monthly runoff sequence complexity of the three stations of the Songhua River is the lowest,and the runoff complexity of the three stations gradually decreases from Harbin station,Tonghe station and Yilan station;The complexity of runoffs vary greatly between the two stations of Mangniu river;The monthly runoff complexity of the four rivers including the Ashi,Balan,Lalin,and Mayi Rivers is relatively low;the monthly runoff complexity of the Chalin River and the Xibei River is the highest.In order to evaluate the performance of the CEEMDAN-WPT-MFDFA algorithm,we chose the equal probability coarse graining LZC algorithm,fuzzy entropy,and wavelet entropy as comparisons.We compared the complexity of the 3 algorithms.The result shows that the equal probability coarse graining LZC algorithm and the fuzzy entropy algorithm appear in the situation where the complexity measurement of some sub-regions are equal,indicating that these 2 algorithms do not extract the information of the sequence completely or the extraction accuracy is not sufficient.The complexity measurement of the fuzzy entropy and wavelet entropy are in short intervals and the distribution is dense;the complexity of each sub-region cannot be distinguished well.However,the complexity measurement of the CEEMDAN-WPT-MFDFA algorithm is in a wide interval,the distribution is uniform and reasonable,and the complexity characteristics of each sub-region can be expressed clearly.The CEEMDAN-WPT-MFDFA algorithm is still stable after adding white noise and colored noise,compared to the other algorithms.Fuzzy entropy and wavelet entropy are still stable under the influence of white noise.However,after we added colored noise,these 2 algorithms experienced a large deviation.The equal probability coarse graining LZC algorithm has a large deviation under the influence of both white noise and colored noise.The result of these contrasts shows that the CEEMDAN-WPT-MFDFA has a stronger robustness under the influence of noise.We analyzed the influencing factors of hydrologiy complexity in Harbin from two aspects of topography.The results show that precipitation and evaporation are closely related to topographic conditions,and the complexity of precipitation and evaporation in the hilly terrain with more complex terrain is higher.Compared with human activities,the complexity of rainfall and evaporation and human beings have great relations with the transformation of underlying surfaces,such as cultivated land and construction area.The complexity of runoff changes is mainly related to the drainage area controlled by river stations.The larger the controlled drainage area is,the lower the complexity of runoff is.and human activity.Based on the complexity research,a drought risk BP neural network evaluation model was constructed from the perspective of complexity,and evaluated the risk of drought in Harbin.The result shows that the drought risk in Harbin was overall serious and showed an upward trend year by year;the main factor that caused severe drought risk in Harbin was social factors,high-intensity agricultural development was the main factor leading to drought risk.and technology peacekeeping and economic dimension The weakening of the danger has caused the drought risk in Harbin to be somewhat inhibited.Overall,the drought risk in Harbin is still very serious and there are still many issues that need to be addressed.Reasonable agricultural development,improving water infrastructure construction,accelerating rural economic development reform,increasing the introduction of water and water-saving scientific and technological achievements,and promoting the level of water management are important measures to reduce the risk of drought.Finally,this paper constructs a Harbin optimal allocation model of water resources based on ant colony optimization algorithm,constructs objective function and constraints,and predicts water demand and water supply on the basis of existing water resources supply and demand,and uses ant colony algorithm to solve the model.The results of Harbin's water resources optimization configuration show that in 2020,the water supply capacity of some parts of Harbin could not meet the water demand,among which the shortage of water in Harbin City and Mulan County was particularly serious;in some areas,supply and demand balance could be achieved;only Wuchang,Shangzhi,Shuangcheng and other regions Can meet the conditions of sufficient water.By 2030,the water supply capacity of Harbin has increased,but the residents' living and industrial and agricultural output of water demand rose faster.Therefore,by 2030,the pressure on water supply in Harbin will be even greater.The research results provide theoretical support for realizing the management and efficient use of the water resources system in Harbin,and provide scientific guarantees for the steady and sustainable development of Harbin.At the same time,it provides a reference for other scholars to creseach urban hydrological system. |
PDF Full Text Request