实时天气背景依赖的反射率因子间接同化及多暴雨个例试验

为避免直接同化时反射率非线性观测算子线性化带来的线性近似误差问题，目前许多研究和业务中还常采用间接同化方式来同化雷达反射率因子，其通过背景场温度判定水凝物类型及比例。基于一种实时天气背景依赖的雷达反射率因子间接同化方案，进行了4次暴雨过程（2次强对流，2次锋面）的循环同化及预报试验。结果表明:对于强对流暴雨个例，相对于传统温度判定方案，天气背景依赖方案的温度预报误差更小、降水预报评分更高，而对于锋面过程区别不明显；进一步机理分析表明，对于强对流暴雨个例，由于背景依赖方案在同化反射率因子时引入了实时天气背景信息，使得分析场水凝物结构能够更好表征实际对流特征且与其它模式变量更为协调，进而改善了模式预报的热、动力及水汽条件，从而改善了降雨预报效果；而锋面暴雨由浅对流过程占主导，水凝物以低层的雨水为主导，冰相水凝物对于该过程的影响较小，由于两种方案反演的雨水结构和量级均相似，因此降雨预报差异较小。

Real-Time Background-Dependent Indirect Assimilation of Radar Reflectivity Factor and Experiments for Multi Heavy Rainfall Cases

In present operating systems, indirect assimilation is frequently used to assimilate the radar reflectivity factor to avoid the problems caused by the linearization of the observation operator. Based on a real-time background-dependent radar reflectivity factor indirect assimilation scheme, cycling assimilation and forecasting experiments of four heavy rainfall processes (two convective and two frontal) were carried out. The results show that compared with the traditional temperature-determination scheme, the background-dependent scheme has smaller temperature forecast errors and higher precipitation forecast scores for the severe convective rainfall cases, but the difference in frontal process is not obvious. Further analysis shows that for severe convective rainfall, the background-dependent scheme introduced real-time background information when assimilating the reflectivity factor, allowing the hydrometeor structure of the analysis field to characterize the actual convective characteristics better and be more coordinated with other model variables, thereby improving the thermal, dynamic, and humidity conditions of the model forecast, thus improving precipitation forecasting. For heavy frontal rainfall, the hydrometeor structural difference in the analysis field of the two schemes is not obvious; thus, the difference in precipitation forecast is small.