Spatial And Temporal Evolution Characteristics And Recurrence Periods Of Extreme Precipitation In The Huaihe River Basin

In recent years,the frequency of global extreme weather events has been accelerating,and flood and drought events often occur,which have caused bad effects on people's daily production and life,and played a certain hindering role in the sustainable development of economy and society.The study of the temporal and spatial evolution of the extreme precipitation index,the analysis of the joint distribution and recurrence periods of the characteristic variables of rainstorm and drought,and the identification of the possible risk areas of the basin in case of disasters can provide a theoretical basis for the planning of the basin water resources system,and also serve as an important guide for the construction of the regional flood and drought prevention and mitigation management.In this paper,eight extreme precipitation indices were selected based on the daily precipitation data of nineteen meteorological stations in the middle and upper reaches of the Huaihe River basin from 1960 to 2020,and the spatial and temporal evolution of extreme precipitation indices in the basin was analyzed by referring to relevant literature.Then,the spatial distribution of heavy precipitation,heavy precipitation days,drought duration and intensity recurrence periods are analyzed according to the optimal distribution of each station,and finally,a two-dimensional joint distribution model of heavy rainfall and drought is constructed using the joint distribution function and their recurrence periods are estimated.The regional surface precipitation of the middle and upper reaches of the Huaihe River basin was calculated by the Tyson polygon area weighting method,and then the Mann-Kendall test and wavelet analysis were used to analyze the cyclical changes of extreme precipitation trends.The study showed that the extreme precipitation indices had no abrupt change points during the study period,except for the strong precipitation(R95p).The oscillatory cycle of total precipitation is most pronounced around 27-32 years,with the center of the oscillation being around 28 years,making 28 years the first principal cycle,and the first principal cycles of the remaining extreme precipitation indices are all around28 years.From a spatial perspective,the distribution of the extreme precipitation indices is more or less the same,following a stepwise decrease from south-east to north-west,with the northern region of Henan falling at the lower end of the threshold in the spatial distribution of several indices.The distribution functions of heavy precipitation,heavy precipitation days,drought duration and intensity were then fitted,and K-S tests were conducted to select the optimal distribution functions for each index,and finally the recurrence periods were calculated.The results show that among the surface-averaged indicators,the P-III distribution is the optimal probability model for the number of days of intense precipitation and drought duration,the Weibull distribution is the optimal distribution model for fitting the intense precipitation indicator,and the optimal fitted distribution for drought intensity is the LN distribution.It can be concluded from the univariate recurrence period estimates that if the size of the recurrence period is increased,the interval of variation in the values widens.Based on the two-dimensional Copula function distribution to construct the two-dimensional joint distribution of the two variables of drought and the two variables of heavy rainfall,and using the three goodness-of-fit evaluation indexes to select the best-fit model,and then plotting the relevant probability graph and the graph of the return period,the final results show that: the optimal distribution function to construct the number of days of intense precipitation and intense precipitation is the Clayton Copula function,the Frank Copula function is more suitable for frequency analysis of drought duration and intensity distribution.Based on the graphs,it can be concluded that the co-occurrence return period of the two variables is greater than their joint return period,and the difference between the three return periods increases as the value of the variables increases.