Calculation Of Non-stationarity Rainfall Frequency In Xiangjiang River Basin Based On Maximum Theory

Extreme rainfall events and floods caused by extreme rainfall will pose a huge threat to human life and property.Especially in recent years,the frequency and precipitation of heavy rainfall events in China have increased significantly.Research on the analysis of extreme rainfall frequency It has become more and more important in the process of engineering hydrology and flood control calculation,and the premise of the current hydrological frequency research is that the hydrological sequence needs to be consistent.Due to significant climate change and human activities,more and more hydrological sequences show non-stationarity characteristics,so traditional hydrological frequency analysis methods are no longer applicable.This paper selects the Xiangjiang River Basin as the research object,and studies the impact of climate change on extreme rainfall events through the non-stationarity analysis of the extreme rainfall sequence in the basin.The main research contents and achievements of this paper are as follows:(1)Cluster analysis of extreme rainfall in the Xiangjiang River Basin.For the annual maximum 24-hour rainfall series of 24 rainfall stations in the Xiangjiang River Basin,a combination of multivariate extreme value theory(F-madogram),which is more suitable for extreme value time series clustering,and the segmentation algorithm around the center point(PAM algorithm)are used.method for clustering.The Xiangjiang River Basin is divided into 3 clusters,and each cluster represents the central stations of Yanling Station(representing cluster I area),Zhuzhou Station(representing cluster Ⅱ area),and Langli Station(representing cluster III area),and the clustering results of the K-means method are used to analyze the results.verify.The results show that the clustering results of maximum rainfall are reasonable and the trend of extreme rainfall sequences of rainfall stations within each cluster is relatively consistent.(2)Screening of climatic factors representing central sites.Based on the results of cluster analysis,the climatic driving factor with the most significant correlation with the extreme rainfall sequence at the representative station was selected from a large number of climate factors by mathematical statistics method.The climatic factor of Zhuzhou Station is JJA SLPa-PC1(the first principal component of the average value of sea level pressure anomalies from June to August);the climatic factor of Langli Station For AMJ Nino 12(April-June average of Nino12)and AMJ AO(April-June average of Arctic Oscillation AO).(3)Frequency analysis of non-stationarity extremum rainfall representing a central site.Three different hypothetical models for calculating the non-stationarity rainfall frequency were constructed for the three representative central sites:a consistent model,a time trend model,and a time-varying moment model with climate influencing factors as covariates.Through the goodness-of-fit test of the model and the comparison of the effect of simulated rainfall quantile changes,the time-varying moment models with climate impact factors as covariates are all optimal at the three sites,and this model can fully describe the changing trend of extreme rainfall events,to quantitatively estimate the impact of climatic drivers on extreme rainfall events.Based on the changes of the extreme rainfall quantile values of the three stations with the climate driving factors,combined with the cluster analysis results,the influence and prediction of the climate driving factors on the occurrence probability of extreme rainfall events in each region of the Xiangjiang River Basin can be obtained.Provide a scientific basis for formulating countermeasures for flood control in the Xiangjiang River Basin affected by the Pacific Ocean.