THE APPLICATION RESEARCH OF THE THEORY OF UNCERTAINTY-SET PAIR ANALYSIS IN ESTABLISHMENT OF WEATHER FORCAST MODELS

The forecast factors in weather forecast models have been selected strictly by means of the characteristics of forecasted objects, the physical meaning of forecasting factors and the experience in forecasting weather, and some technical ways. In general, these selected forecast factors have all better forecasting abilities, but the better abilities of these factors aren't always unchanged and sometimes, the badness ability of someone factor may produce error in forecast results. Set Pair Analysis (SPA) is a systemic theory and method used in researching and analyzing non-authenticity. Using the theory and method, this paper make judgment of station and analysis of same difference reverse about the factors before being used to forecast weather. In analyzing, the method weakens effect of those factors with badness abilities in forecast models, while those with better abilities in the models would play more important action. As a result the processes above have made the dynamic changes in the structure of the factors in the forecast models, which intensifies the rationality of forecasting mechanisms and the ability. The applications in interpret and use from the numerical forecast products indicate that the method has better effect.