Research On The Stochastic Modeling For Regional Aridity

In process of researching aridity,a important way is to research regional aridity characteristic using the method of stochasic hydrology.The key to the stochastic modeling is to build a suitable stochastic model of precipitation.In resent years,most models of precipitation process have been developed under some hypothesis or using advaced mathematical tools,but the applicability of these models is not strong. Under consideration of the observed precipitation sequence containing much important infomation such as the temporal and regional characteristic of precipitation,this paper deals with statistical characteristic of the annual precipitation probabilistic process using the stochastical analysis theories,some methods and the technique of computer numerical simulation. And then two stochastic models of the annual precipitation sequence and monthly precipitation sequence are proposed for the practical use. They are the first order autoregressive model and the seasonal first order autoregressive model with Pearson Type III distribution. The diagnostic test of the models have been passed. The calculation results show that the method designed is quite satisfactory. The models can reproduce the annual precipitation sequence and monthly precipitation sequence.And then the regional aridity index can be estimated statistically,in addition,they can reproduce the characteristic of history precipitation series.This paper introduced the detaild formulas of numerical simulation for annual precipitation sequence and monthly precipitation sequence,and provided a whole computer program for each system model,which makes the results of the paper used more conveniently.