The Distribution Models And Methods Of Estimating Parameters In Precipitation Frequency Analysis Of Notrh Region Of Shaanxi

By using the principle of statistical hydrology, the precipitation frequency analysis canprovide scientific basis for water resourses rational utilization and ecological constructionwork.The thesis reviewed and organized the relevant literatures of precipitation frequencyanalysis at home and abroad, elaborated the present analysis methods and latest research trendof precipitation frequency, summarized and analysised the present progresses, including theprecipitation frequency distribution and parameter calculation methods. Based On this, thethesis took the precipitation sequences in Shannxi north region for example, studied theoptimum choice for frequency distribution and parameter estimation. The main conclusion areas follows.(1) We take12kinds of probability distribution function, which belong to normaldistribution class, Г distribution class, extreme value distribution class, Wakeby distributionclass and Logistic distribution class; The main parameter estimation methods are the methodof moment (MOM), the method of maximum likelihood (ML) and the method of probabilityweighted moment (PWM).(2) Based on the common probability distribution models, the precipitation frequencyanalysis and parameter statistics of the annual common precipitation frequency distributionmodels pitation frequency analysis includes kinds of probability distribution precipitation dataof Shaanxi north region are conducted by using three kinds of parameter estimation methods.The most suitable theoretical frequency distribution models, the best parameter estimationmethods and the annual precipitation at different frequency have been obtained.