A Comparison of Parameter Estimation Methods Using Typical Precipitation Data in Plain River Networks: Case Study in Taihu Basin, China

摘要

As one of the most developed regions in China, precipitation frequencyanalysis is of great necessity to Taihu Basin for flood disasterprediction and prevention. In this study, four parameter estimationmethods of probability distribution functions, namely conventionalmoments method, curve fitting method, maximum likelihood methodand L-moments method are used to estimate parameters of Pearson Ⅲdistribution models to fit these sets of data collected from four stationsin Taihu Basin. The performance of the four parameter estimationmethods is tested by K-S test and is also compared by weighted leastsquares (WLS) criterion. This study can provide a reference for theselection of parameter estimation methods while using Pearson Ⅲdistribution model in Taihu Basin.