Verification of Prediction Capability of AREM on Low Temperature,Rain,Snow and Frost over Southwestern China in 2008

By the objective method, this paper was to verify the prediction of AREM, a regional mesoscale model, during the historically rare low temperature period with rain, snow and frost disaster occurred over the southern China in early 2008. The results show that AREM made a satisfactory prediction of air temperature at 2 meter above the ground. For Chengdu, Chongqing and Guiyang where the temperature declined remarkably, AREM predicted correctly not only the beginning and ending time of temperature drop but also the temperature evolvement, except the small prediction error; as for Kunming and Lasa where the temperature fall were not as remarkable as the other three stations, AREM predicted the waves of the temperature during this period in despite of the relatively larger prediction error, especially Lasa. AREM also correctly predicted the sites with the maximum falling temperature. According to the process-averaged Ts score, AREM was inferior to T213 and Japanese model in precipitation prediction.