Analysis On Random Characteristics Of Runoff In Seasonal Frozen Soil Area

Influenced by climate and human activities,hydrological systems often exhibit more complex features that cannot yet be described mathematically.Hydrological stochastic simulation refers to the establishment of a stochastic model based on the statistical characteristics of observation data and its random change law,through which a large amount of simulation data is obtained,and a reasonable prediction of the future state of the hydrological system is analyzed and calculated,so as to solve the design,operation or management planning problem of the system.Due to the limited length of the runoff data series in the basin,the uneven distribution of the size of the runoff during the year,and the seasonal permafrost basin climate and lower cushion surface conditions have certain particularities,making the study of runoff in the cold area quite difficult,if the runoff data of the limited year is used as a sample,combined with the characteristics of the seasonal permafrost area,a sufficiently long sequence is generated through random simulation,and its characteristics and change trend are analyzed,which will help improve the scientific level of basin planning and design and management application,and save precious and limited water resources.It is also important to fully realize the sustainable use of water resources.Based on the measured data of rainfall runoff from 1964 to 2014 in the Yongcui River Basin,four common marginal distribution functions were selected as candidate distributions,and the optimal edge distribution was selected by the goodness of fit test,and the Copula function with simple parameter calculation,strong applicability and simple method was selected to construct the monthly rainfall-runoff joint distribution model and the adjacent monthly runoff joint distribution model,and the monthly runoff of multiple years was generated by random simulation.The main conclusions are:(1)The marginal distribution of rainfall follows gamma distribution in January,February,August,September and November,P-? distribution in March,July and December,and Weibull distribution in April,may,June and October;The marginal distribution of runoff follows gamma distribution in January,February,March,July,October and December,P-? distribution in April,may,June,August and November,and Weibull distribution in September.The fitting degree test is carried out for the edge distribution.The K-S test statistics of the edge distribution of monthly rainfall and runoff are within the 95% confidence interval,and the fitting effect is good.(2)The parameters of three Copula functions are determined,and the difference between the theoretical rank correlation coefficient and the measured rank correlation coefficient is compared to determine the advantages and disadvantages of the Copula function.The Gumbel Copula function was selected as the combined rain-runoff distribution function for February,May,July,September,and November,Frank Copula as the combined distribution function for rain-runoff in January,March,April,June,August,October,and December,and the Gumbel Copula function for the adjacent monthly runoff joint distribution function for December-January,February-March,October-November,and November-December,Clayton The Copula function is the joint distribution function of adjacent month runoff from January to February,April to May,and September to October,and the Frank Copula function is the joint distribution function of adjacent month runoff from March to April,May to June,June to July,July to August,and August to September.The goodness-of-fit evaluation of the Copula function is performed,and the results show that the selected Copula function meets the fitting conditions.(3)Based on the copula joint distribution model,the random simulation of watershed runoff in 1000 years is carried out according to the measured samples.The results show that the relative errors of the mean value,mean square deviation and variation coefficient of the model simulation value in winter are large,but they are not more than 15%,and the change trend of the statistical parameters of the measured value and the simulation value is roughly the same in each month.The simulation effect is good,which verifies the applicability of Copula function in Yongcui River Basin.