Study of the forecasting method for the classified severe convection weather based on a meso-scale numerical model

For severe convective weather short-term forecasting of thunderstorm-gale, short-time severe rain, hail and tornado, the hundreds of convective parameters (physical quantities) and their 15 days moving average values were calculated based on the 0.25°×0.2° reanalysis data of the East Asia region from Japan Meteorological Agency (JMA), These parameters were adopted to gain the featured convective parameter sequence reflecting both the obvious difference between severe convection and climatic mean state and their own relative stability, after the time and site matching of the severe convective weather events during February to September 2001-2009 in Jiangsu Province with the proximity principle as well as the weight distribution and successive screening of the above parameters by the fuzzy relative deviation matrix evaluation technology. Meanwhile, according to the spectrum distribution of the featured convective parameters in the historical cases of classified severe convection, the piecewise function of the frequency distribution of the convective parameters was obtained. Then, the convective parameters forecasts were calculated with a meso-scale numerical model, and comprehensively considering the historical frequency distribution and weight allocation of the above parameters, the classified severe convective weather forecast probability was constructed. Taking the advantage probability as a classification criterion, the classified severe convective weather forecast was made. Finally, the operational system was established to provide objective forecast products of classified severe convection in a fully automatic mode to the daily business and meteorological service work of the Youth Olympic Games in Nanjing.