The model of thunderstorms forecast based on rough set

Thunderstorm is one of the worst natural disasters around the world. Currently, it still lacks of the model that can forecast the high-resolution and short-term approaching thunderstorms. Rough set method can analyze different incomplete information effectively, and then do the effective reasoning. Rough set based two techniques of attribute reduction and rule extraction were used to choose reasonable combination of forecasting factors and extract the effective decision rules, and a forecast solution model was established for the high-resolution(forecast range: 5km×5km) and the short-term (next 3 hours) thunderstorm forecast. Finally, the proposed model was tested on the given real dataset, and compared with the traditional numerical forecast solution model. The results show that it is a better in the accuracy of thunderstorm forecasting.