非降水回波同化在台风"山竹"数值预报中的应用OACSTPCD

China is one of the countries most affected by typhoons in the world.The intensity of typhoons and the detailed precipitation forecast before and after landing have always been challenging issues in operational forecasting.Therefore,numerical models have become important forecasting tools.However,high-resolution models are prone to false convection due to various factors,leading to false reports of precipitation.This paper addressed the issue of insufficient assimilation of no-rain echoes in large areas of low reflectivity in typhoon radar data assimilation.Utilizing the high temporal and spatial observation characteristics of Doppler shore-based weather radar for detailed wind fields and hydrometeor distribution,based on the radial wind speed and reflectivity factor data of two S-band Doppler radars in Guangzhou and Yangjiang,through the WRF and its 3DVAR system,no-rain echo assimilation experiments were conducted to research whether this assimilation method can effectively suppress false precipitation in the model during typhoon weather processes.For the strong Typhoon"Mangkhut"in 2018,data assimilation and numerical model forecasting were carried out.Results show that the assimilation of no-rain echoes reduces the cyclonic wind field increment and water vapor increment,reduces the combined reflectivity of the typhoon's no-rthern spiral rain belt and Hainan area,suppresses the model's false convection,and makes the typhoon's intensity and structure closer to radar observations.At the same time,the assimilation of no rain echoes also increases the minimum sea level pressure of the typhoon center,significantly improves the 12 h forecast of the typhoon's minimum sea level pressure,reduces the excessive precipitation forecast in the northern part of Guangdong and the northern part of Hainan Island,improves the TS score and FAR score of the 6 h precipitation forecast,and effectively reduces the false precipitation report situation of the conventional reflectivity assimilation experiment.