基于风云三号卫星组网微波湿度计观测资料的西南涡预报晴空同化初步分析

Impact assessment of MWHS-2 clear-sky data assimilation from joint Fengyun-3 satellite series on forecasting Southwest China vortex

摘要:

从2008年5月发射FY-3A星到2023年相继发射FY-3G星和FY-3F星,风云三号极轨系列卫星观测网越发完善,使得卫星资料在时间和空间上的分辨率都得到很大提高,促进了我国数值天气预报的发展。西南涡受复杂地形影响而生成于青藏高原东侧700～850 hPa高度,是影响我国夏季暴雨的主要天气系统。本研究基于WRF模式及WRFDA同化系统,利用FY-3C和FY-3D卫星组网微波湿度计观测资料进行晴空同化试验,并探究不同稀疏化设置对预报的影响。结果表明:与星下点分辨率相近的30 km稀疏化设置对于同化和预报的效果最好;同化卫星组网资料优于单个卫星资料,从而改善预报降水落区,降低大量级降水的虚报率,使降水预报准确率得到提高;单独同化FY-3C星微波湿度计资料对湿度场预报的积极影响更大,而单独同化FY-3D星微波湿度计资料对风场的预报改善更好。

Abstract:

Since the launch of the FY-3A satellite in May 2008 and subsequent missions,including the FY-3G for precipitation and the FY-3F until 2023,the Fengyun-3 (FY-3) second-generation polar-orbiting satellite series has developed into a comprehensive observation network.This network now operates with early-morning,mid-morning,afternoon,and precipitation-dedicated satellites simultaneously in orbit,significantly enhancing the spatial and temporal resolution of satellite data and supporting advances in numerical weather prediction (NWP) in China.The Southwest China vortex (SWCV),a major weather system contributing to intense summer rainfall in China,forms over the complex terrain east of the Qinghai-Xizang Plateau at 700—850 hPa.This study analyzes a SWCV event from July 21 to 23,2022,which caused heavy rainfall near the Sichuan-Shanxi border,where several stations recorded over 100 mm of precipitation within 24 hours.As the vortex moved eastward,it generated a rainfall belt extending from Chongqing through Hubei,Henan,and Shandong provinces.
The research investigates the impact of clear-sky assimilation of FY-3 satellite microwave humidity data on forecasting the SWCV using the WRF model and WRFDA system.Clear-sky assimilation experiments with MWHS-2 observations from FY-3C and FY-3D satellites were conducted to assess optimal thinning distances for data assimilation.The findings show that applying inappropriate thinning distances minimizes observation errors,reduces computational costs and improves forecast accuracy.Specifically,a thinning distance of 30 km,approximating the resolution of the sub-satellite point,optimizes forecast performance by maintaining an effective balance between minimizing spatial correlations in the data and ensuring sufficient observational density to constrain the analysis field.This configuration enhances predictions of precipitation location and intensity while reducing false alarms for heavy rainfall events.
Results demonstrate that assimilating data from both FY-3C and FY-3D satellites provides superior forecasts compared to single satellite assimilation,as network data assimilation increases the quantity of high-quality observations,thereby refining initial conditions and improving short-term weather forecast accuracy.FY-3C-only data assimilation particularly enhances relative humidity forecasts,improving intensity predictions and yielding higher threat scores (TSs) for extreme rainfall.Conversely,while FY-3D-only assimilation enhances wind field forecasts,improving the spatial accuracy of precipitation predictions.The combined FY-3CD satellite data assimilation adjusts lower-quality data points,producing more stable forecasts.Future work will extend these efforts by incorporating additional Fengyun satellites for comprehensive network assimilation,aiming to further improve NWP accuracy and optimize the use of satellite observations.