AREM数值模式对2005年汛期四川的降水预报

应用中国科学院大气物理研究所的AREM模式，对2005年汛期四川进行了实时降水预报。结果表明：（1）AREM预报性能略好于成都区域中心业务运行模式ETA模式，对大雨和暴雨的TS评分，AREM略高于T213预报。（2）从梯度评分看，高度、温度、涡度等要素均为可预报，而地面温度、整层水汽含量和整层水汽通量散度可预报性较低。（3）AREM对5次区域性暴雨有较好的反映，但与实况还存在一定的差异，AREM降水强度预报较实况偏弱。

Precipitation Prediction with AREM Numerical Model in Sichuan Flood Season in 2005

Using AREM model, a real-time precipitation prediction in Sichuan flood season in 2005 is made. The results are as follows: 1.the performance of AREM model is better than ETA model which is operation model in Chendu region meteorological center and TS score values of AREM model is higher than that of T213 model. 2.The elements such as height, temperature and vortex are predictable. But the predictability of the elements such as ground temperature, vapor contents and vapor flux divergence of the whole atmosphere level is lower. 3. AREM model has prediction ability to five local heavy rain happed in Sichuan flood season in 2005, but still are there the differences between the precipitation prediction of the model and observation. The intensity of precipitation of the model is weaker than the observation.