多部多普勒天气雷达资料同化对暴雨个例模拟的影响分析

The effects of assimilation of multiple Doppler radar data on the simulation of a heavy rain event

利用中尺度模式WRF(Weather Research and Forecasting Model)及其资料同化系统3DVAR,将国内多普勒天气雷达(CINRAD)反射率因子及径向风资料直接用于中尺度数值模拟;对安徽梅雨期一次暴雨过程进行模拟试验,经过质量控制后,分析同时同化安徽省内六部多普勒天气雷达观测资料对模拟结果的影响。结果表明:(1)经过质量控制后,同化雷达径向风和反射率因子对初始场的风场和湿度场均有较为明显的改善,说明用该雷达资料质量控制方案同化雷达资料是可行的;(2)同化多部雷达径向速度资料能使风场气旋性增强,同化反射率资料能调整初始水汽场,使对流层中下层水汽含量增加;(3)同化多部雷达径向风和反射率因子资料,能提前模拟出强降水回波结构,且其中尺度特征更清晰,降水落区和强度预报更接近实况,同化雷达资料对降水预报能持续影响到12 h,并提高了12 h降水预报准确率。

Using Weather Research and Forecasting Model (WRF V3.1) and its three-dimensional variational data assimilation system (WRF-3DVAR), we have directly applied basic reflectivity factor and radial velocity data from CINRAD at home to a mesoscale numerical simulation. We have performed a simulation experiment for a heavy rain event in Meiyu period in Anhui Province. After passing the quality control for the experiment results, we have analyzed the effect of assimilation of data simultaneously from the 6 radars in Anhui Province on the modeling outcomes. The results are as follows. (1) After passing the quality control, assimilating basic reflectivity factor and radial velocity can clearly improve the wind and humidity distribution in the initial field, which is expatiated that assimilating radar data used by the quality control scheme mentioned above is feasible. (2) Assimilating radial velocity from the several Doppler radars can increase the cyclonic of wind field, while assimilating basic reflectivity factor can adjust initial water vapor field and increase the water vapor content in the lower and mid-dle troposphere. (3) Assimilating radial velocity and basic reflectivity factor from the several Doppler radars can better simulate echo structure of severe precipitation in advance, and make the echo structure display clear mesoscale characteristics, as well as the distribution and intensi-ty forecast of rain close to observations. Additionally, the effect of radar data assimilation on precipitation forecast can last for 12 hours, so the assimilation improves the accuracy of 12-hourly precipitation forecast.