采用中尺度数值模式WRF（Weather Research and Forecasting）中FY-3C卫星MWHS-2（Micro-Wave Humidity Sounder 2）卫星资料的直接同化模块， 利用WRF-3DVAR（Three Dimensional Variation）方法对三江源地区的两次降水过程进行同化对比试验， 详细对比分析了FY-3C搭载的MWHS-2微波湿度仪和NOAA-18（National Oceanic and Atmospheric Administration-18）搭载的MHS（Microwave Humidity Sounder）微波湿度仪两种同化资料对模拟结果的影响。结果表明： MWHS-2和MHS资料同化的模拟结果基本一致， 对于30 mm以下的降水， 无论同化与否， 模拟的降水范围均会偏大； 但对于30 mm以上的降水， 模拟范围和量级均偏小。资料同化增强了500 hPa西南风场， 从而加大了水汽输送的强度， 使得高空槽偏向西南， 同时加强了300 hPa风场扰动， 在共同作用下从而导致了30 mm以上的降水范围的增大， 对降水量级结果能有一定改善， 但降水落区相较于实际偏南。卫星资料的同化会使得水汽通量场和比湿场的落区范围明显改善， 但对强度的改善作用并没有对落区的改善作用明显， 水汽通量场和比湿场的改变造成了降水预报落区和强度的改变。在温度场方面， WRF模式模拟的误差较大， 卫星资料的同化对温度场有一定的改善作用， 但不如对湿度场的改善作用明显。因而， MWHS-2或MHS卫星资料的同化对三江源降水模拟有一定的改善作用， 该研究结果对三江源降水的预报改进能有重要指导意义。

Based on the mid-scale WRF （Weather Research and Forecasting） numerical model， this study uses the direct assimilation module data from the FY-3C satellite MWHS-2 （Micro-Wave Humidity Sounder 2）， and the WRF-3DVAR （Three Dimensional Variation） method to conduct assimilation comparison experiments on two precipitation processes in the Three Rivers Source area.Moreover， this study extensively compares the effects of two assimilation data sets on simulation results： data from the MWHS-2 （microwave humidity instrument carried by FY-3C） and data from the MHS （Microwave Humidity Sounder） carried by NOAA-18 （National Oceanic and Atmospheric Administration-18）.Results of the study show consistent simulation results for MWHS-2 and MHS data assimilation.For instance， regardless of whether assimilation is conducted or not， the simulated precipitation range is overestimated for precipitations less than 30 mm.However， the simulated range and amount are underestimated for precipitations higher than 30 mm.Furthermore， data assimilation strengthens the 500 hPa southwest wind field， increases the intensity of water vapour transport， shifts the high-altitude trough to the southwest， and strengthens the 300 hPa wind field disturbance， all of which increase the precipitation range above 30 mm and， to a certain extent， improve the precipitation level results.However， compared to the actual situation， the precipitation area is biased southward.Satellite data assimilation significantly improves the range of the water vapour flux field and specific humidity field precipitation areas， but its effects on the intensity improvement are not obvious.At the same time， changes in the water vapour flux field and specific humidity field change the precipitation forecast area and intensity.As for the temperature field， the WRF simulation includes significant errors.Satellite data assimilation shows a certain improvement effect on the temperature field and a more significant improvement effect on the humidity field.In other words， MWHS-2 or MHS satellite data assimilation improves the precipitation forecast in the Three Rivers Source area.These conclusions provide significant guidance for improving the precision of precipitation forecasting in the Three Rivers Source area.