Abstract:Numerical forecasts, especially ensemble forecast techniques, greatly improve the ability to predict extreme weather. At present, for the temperature, wind, precipitation and other factors, the European Center's extreme index products based on ensemble forecast products provide a quantitative basis for their extremity. However, there is currently no strong convective weather extreme index product for operational forecast. Therefore, this paper counts the physical quantities closely related to the strong convective weather, calculates its extreme weather index, and counts the threshold distribution of the extreme weather index in different severe convective weather. The results show that the extreme weather index is closely related to the strong convective weather, and the distribution and threshold of different types of strong convective weather extreme indices have their own characteristics. Based on the above results, using the extreme exponential and model precipitation data, using the support vector machine method, this paper establishes an objective forecasting method for different types of severe convective weather, and provides objective support products for operational forecast of extremely strong convective weather.