Abstract:In natural dry land vegetation system, soil water is closely related to atmospheric water, surface water and evapotranspiration, and there is a water balance relationship. According to the relationship of water balance and the measured hydrological data, the relationship between soil water content and continuous no rain days is created with continuous no rain days as the influence factor. The relation is the prediction model of single station without rain, and the relationship between soil water content increment and water reduction is established by precipitation as the influence factor. The prediction model of rainfall increase soil moisture can predict the dynamic change of soil water content in short term. The prediction model has been applied to the soil moisture analysis and evaluation of the central and Western Jilin Province in 2016. The application results show that the model is full of theoretical basis and meets the water balance relationship in the natural dryland vegetation system. The modeling is simple and the application is convenient. It can carry on the short-term soil moisture content and the prediction of soil moisture, and the forecast results can be scientifically reversed. The dynamic change law of soil water content was shown.