Radiance-based assessment of bulk microphysics models with seven hydrometeor species in forecasting Super-typhoon Lekima (2019) near landfall

Abstract

Extreme precipitation concerning typhoons brings great losses to coastal cities every year. How to accurately describe the natural cloud and precipitation processes in model remains a major challenge in typhoon forecasts. Observations found rain and hail can both exist under severe typhoon convections. But current operational forecasting models fail to take hail into consideration in super-typhoon prediction. This study assesses the performances of five typical bulk microphysical schemes (BMSs) of the Weather Research and Forecasting (WRF) model in the forecast of Super-typhoon Lekima (2019) near landfall. Unlike other microphysical schemes, all the selected BMSs in this study incorporate seven hydrometeor species including hail, which can represent the hydrometeor distribution within natural super-storm more realistically as supported by field observational facts. Through statistical-physical evaluation processes and using radar-satellite joint observations, the goal is to identify the best scheme to accurately forecast Super-typhoon Lekima and to find the future avenue to possible microphysical scheme improvement. It is found that the WRF double-moment 7-class (WDM7) scheme exhibits the smallest root mean square error (RMSE) of approximately 1.5 for both composite reflectivity and brightness temperature signatures, and predicts the closest macro-physical hydrometeors structure, amplitude and location to observations. Moreover, the azimuthal mean brightness temperature distributions suggest large forecasting errors over typhoon inner rainband, with WDM7 scheme performing best. The vertical reflectivity profiles further suggest that WDM7 scheme could well characterize the updraft, water content, and cold-cloud processes, but needs further improvement in the representation of warm-cloud processes. The performances of BMSs are found to be closely related to the moments used to describe the evolution of the drop size distribution. Among all the schemes, the double-moment schemes generally perform better in simulating warm-cloud processes, while the single-moment schemes show better performance in simulating cold-cloud processes. As a partially double-moment and partially single-moment scheme, WDM7 scheme exhibits unique advantages over the others. The optimum choice of moments is an important step toward the future improvement of BMS.