一次强降水模拟中土壤湿度初值的影响研究

Impact of Initial Soil Moisture on a Heavy Precipitation Simulation

为了分析WRFV3.3.1模式中土壤湿度初值对降水模拟的影响,在4个陆面方案的分析对比基础上,采用中国区域土壤湿度同化系统(CLSMDAS)输出的土壤湿度替换模式初始场中的土壤湿度,对2010年6月19日江西出现的一次强降水过程进行模拟.分析改进后的土壤湿度初值对模式模拟降水的影响。结果表明,耦合Noah方案的模拟结果要优于SLAB、RUC、PX方案;通过对比试验发现CLSMDAS输出的土壤湿度初始条件下的模拟结果比NCEP资料的模拟结果更接近于实况,能够很好地模拟出24 h累积降水的范围与强度,对应地表能量的响应也更为明显,表现出很好的区域波动性,能够捕捉到细节降水信息,且统计检验结果中各量级降水的TS评分基本都要高于NCEP土壤湿度初值条件下的结果,空报率、漏报率和预报偏差进一步减小,说明准确的土壤湿度初值能够提高模式的模拟预报能力。

In order to analyze the influence of soil moisture on the simulation of precipitation in the WRFV3.3.1 mode, we use the soil moisture output from the China Land Soil Moisture Data Assimilaton System （CLSMDAS） to replace the soil moisture in the WRF initial field, on the basis of comparative analysis of four land surface process schemes, the heavy precipitation which occurred in Jiangxi Province on 19 June 2010 was simulated to study the influence of improved soil moisture on the model simulation of precipitation. The results show that the Noah scheme gives more reasonable results than SLAB, RUC and PX schemes. By comparative test we find that the simulation from the CLSMDAS output soil moisture generally agree with the observations, being able to simulate the intensity and range of heavy precipitation in 24 h as well as the response of the surface energy. The statistical verification test also shows that the TS score is higher, the miss rate, the false alarm and the bias are lower than the results from NCEP initial soil moisture. All these tell us that the accurate soil moisture can improve the forecasting capability of model simulation.