Analysis Of Multivariate Wetness-Dryness Encountering Probability Of Hotan Rivers In Xin Jiang

The Hotan River Basin is located in the south of Xinjiang,is not only one of the important water conveyance areas of Tarim River,but also an important hub area for the "The development of the western region in China" and "The Belt and Road".However,with the current global temperature rise and the impact of human activities,the runoff and climate factors in Hotan River Basin have changed to a certain extent.Therefore,it is necessary to analyze the Hotan river runoff and its influencing factors in order to understand the evolution law of Hotan river runoff from 1960 to 2016 and determine the influence degree of different factors.The runoff of Hotan River is affected by different factors,and its change is full of uncertainty,which leads to the unreasonable distribution of water resources.It is an important way to identify the uncertainty and improve the distribution of water resources to evaluate the annual runoff of Hotan River.Due to the limitation of the traditional single variable description of hydrological change,based on the evaluation results of flood and drought,the probability of flood and drought in Hotan River is calculated by statistical method.In order to ensure the accuracy of the calculation results,the copula function is used to analyze the probability of the two tributaries of Hotan River(Yulongkashi River and karakashi River)in the same time,and the optimal condition of runoff complementary is explored by comparing the results of statistical method,so as to provide scientific guidance for further improving the utilization efficiency of water resources and the application of water resources scheduling in Hotan area.The main research contents and achievements are as follows:(1)Using Mann Kendall test,Pettitt mutation test and Morlet wavelet to analysis the variation trend,mutation characteristics and periodic law of runoff and influencing factors.The multivariate linear equation and neural network model of influencing factors and runoff were established to quantitatively analyze the contribution level of climate factors and human activities to runoff change.The results show that the runoff,precipitation and temperature of Yulongkashi River and Karakashi River are increasing.The abrupt change nodes of runoff and precipitation are located at the beginning of the 20 th century,and the abrupt change of temperature occurred in 1984,which indicates that the abrupt change of runoff and precipitation has a lag,and the abrupt change time of temperature is consistent with the climate change in Northwest China.The runoff experienced two periods of dry wet transition in a large scale,the precipitation showed three times of more less alternation,and the temperature experienced two periods of cold warm transition.Through the analysis of neural network,it is considered that climate change is the main factor affecting runoff change,while human activities are the secondary factor.Temperature is the main influencing factor of runoff recharge,and the effect of precipitation recharge is lower than that of temperature recharge.(2)In view of the uncertainty of the classification of runoff grade,an appropriate evaluation model is used to judge the runoff,so as to ensure the accuracy of runoff grade evaluation.Based on the monthly runoff data of Yulongkashi River and Karakashi River from1960 to 2016,the mean standard deviation method was used to distinguish the high and low flow state of the two tributaries.The results show that the Yulongkashi river has the most times in the normal year,and the karakashi river has the largest proportion in the low flow year.According to the set pair method,the proportion of two runoff is more than 21% in wet years and more than 49% in dry years.Under the coupling analysis of fuzzy set pair,the ratio of wet and dry years of Yulongkashi River and karakashi river is 17.54% and 15.79% respectively in wet years,and 45.61% and 31.58% respectively in dry years.Based on the three detection results,Yulongkashi River and karakashi River have the most frequency in dry year and the least frequency in wet year.Compared with the annual runoff in different years,it is considered that the fuzzy set pair can more clearly and objectively describe the wet and dry state of Hetian river runoff.(3)Based on the runoff data of Yulongkashi River and Karakashi River from 1960 to 2016,the empirical frequency of the series was calculated according to the single variable empirical method,and fitted with the four marginal distributions.Two error test methods are used to test the fitting effect,and the optimal single variable distribution is substituted into the copula function,and the error test method is used to optimize the joint variables.According to the frequency of 37.5% and 62.5% as the boundary of the abundant and withered types,the probability of 9 kinds of abundant and withered combination types is obtained through the function calculation formula.The copula function calculation shows that the probability of the two tributaries with the same abundance,the same level and the same dry is 8.41%,18.48%and 21.41% respectively,and the asynchronous probability is 51.7%.Based on the fuzzy set pair evaluation results,the probability of asynchronous encountering is 64.17%,which is higher than that of synchronous encountering by 35.83%.The results show that the probability of asynchronous high and low flow is greater than that of synchronous high and low flow,which indicates that there is good complementarity between the two tributaries.