Study On Spatial And Temporal Variation Of Extreme Precipitation And Frequency Distribution In Guanzhong Area

With the continuous increase in the frequency and intensity of global extreme climate events,floods and floods are frequent,severely affecting people's normal lives and constraining the sustainable development of regional social economy.Studying the temporal and spatial patterns of change and frequency distribution characteristics of extreme precipitation events is of great significance to regional flood control and drought relief planning,water resources management,disaster prevention and reduction work,and sustainable socio-economic development.In this paper,Guanzhong area in Shaanxi Province was selected as the study area,and 11 different extreme precipitation indexes were selected as the research object,and so the temporal and spatial changes and periodic characteristics of extreme precipitation events were revealed.At the same time,univariate,multivariate and regional frequency analysis methods were used to study the frequency characteristics of extreme precipitation events.The main content and conclusions of the paper are as follows:(1)Mann-Kendall trend test,sliding t test,and wavelet transform analysis methods were used to analyze the time course and periodic characteristics of annual Maximum Daily Precipitation RX1 DAY,Maximum Daily Precipitation RX5 DAY for the Last Five Days,Simple Precipitation Intensity SDII,Annual Precipitation PRCPTOT,Strong Precipitation Day R10,Strong Precipitation Day R20,Continuous Dry Day CDD,and Continuous Wet Day CWD in Guanzhong area.The results show: on the intergenerational changes,the extreme precipitation indices RX1 DAY,RX5DAY,SDII,and R20 showed an overall increase with time,the average increase was 1.5%~ 5.8%.And the average values of PRCPTOT,R10,CDD,and CWD showed a decreasing trend as a whole.The average reduction was 0.5%~2.9%.However,the coefficient of variation and extreme ratio of extreme precipitation indices showed an increasing trend as a whole,indicating that the degree of inhomogeneity and the extent of change of extreme precipitation indices in the years have gradually increased.The average increase range of deviation coefficient is 1.4%~20.3%,and the average increase range of extreme value ratio is 4.0%~21.7% On the interannual changes,the extreme precipitation indices RX1 DAY,RX5DAY,SDII,PRCPTOT,R10,R20,CDD,and CWD did not show significant trends and abrupt changes.(2)Radial basis function interpolation was used to study the spatial distribution characteristics of extreme precipitation indices.The results show: the spatial variations of the mean values of the extreme precipitation indices are relatively consistent.On the contrary,the spatial variation of the coefficient of variation of the extreme precipitation indexes is quite different.There is no obvious spatial distribution rule.(3)Taking the generalized extreme distribution(GEV),generalized Pareto distribution(GP)and Pearson type III distribution(P-III)as alternative distributions,the linear moment method was used to calculate the parameters of each alternative distribution and adopted Kolmogorov-Sminrov test to evaluates the goodness of fit of the candidate distribution.The results show: both the P-III distribution and the GEV distribution can be used as the theoretical frequency distribution of annual maximum daily precipitation RX1 DAY,annual maximum fiveday precipitation RX5 DAY,simple precipitation intensity SDII,and annual precipitation total PRCPTOT.Spatially,the optimal distribution of extreme precipitation indexes RX1 DAY,RX5DAY,SDII,and PRCPTOT in the central area of the study area is the GEV distribution and the area with the P-III distribution as the optimal distribution is mainly concentrated in the surrounding area of the study area.(4)Two-dimensional joint distribution of extreme precipitation indices was constructed using the link function of Archimedean Copula function.The above-mentioned optimal distribution was used as the marginal distribution.The least square method of deviation sum(OLS)and graph evaluation method were used to select the best fitting effect.Copula function.The results show: there is a good correlation between the annual maximum daily precipitation RX1 DAY,the annual maximum five-day precipitation RX5 DAY,and the simple precipitation intensity SDII.G-H Copula function is the optimal Copula function for joint distribution of extreme precipitation index RX1DAY-RX5 DAY and RX1DAY-SDII in Guanzhong area.The FK Copula function is the optimal Copula function for the joint distribution of the extreme precipitation index RX1DAY-RX5 DAY and RX1DAY-SDII in the Guanzhong region.(5)Hierarchical clustering analysis was used to classify similar precipitation regions in the Guanzhong region.The consistency and homogeneity of annual maximum daily precipitation RX1 DAY,annual maximum five-day precipitation RX5 DAY,simple precipitation intensity SDII,and annual precipitation total PRCPTOT in similar areas were tested.On this basis,the optimal regional distribution of extreme precipitation indices was selected.The results show: The results of the extreme precipitation subarea are basically the same as those of Shaanxi Province,and the extreme precipitation indexes at all stations in the subdistrict satisfy the consistency and uniformity requirements.The optimal distributions of extreme precipitation indexes RX1 DAY,RX5DAY,SDII and PRCPTOT in the eastern Guanzhong region are respectively GEV distribution,P-III distribution,GEV distribution and P-III distribution;the optimal distributions of extreme precipitation indexes RX1 DAY,RX5DAY,SDII and PRCPTOT in the Central Shaanxi Plain are all GEV distributions;the optimal distributions of extreme precipitation indexes RX1 DAY,RX5DAY,SDII,and PRCPTOT in Guanzhong area are GEV distribution,GEV distribution,P-III distribution,and GEV distribution respectively.