Convective-scale ensemble forecasts of the heavy precipitation of Typhoon Lekima (2019) in Zhejiang Province

Abstract

Ensemble forecasting plays an important role in studying the precipitation forecast errors of landfalling tropical cyclones. In this study, convective-scale ensemble forecasts based on the WRF model are conducted, focusing on the heavy precipitation of typhoon Lekima (2019) in Zhejiang, China. The perturbations of the initial/boundary conditions (ICs/BCs) and the microphysical schemes are constructed to represent the uncertainties of ensemble forecasts. The sensitivity of different perturbations on ensemble forecasts and the precipitation forecast errors are discussed. The results show the IC/BC perturbations have a greater impact on ensemble precipitation forecasts. An advantaged group and two disadvantaged groups are selected from 63 ensemble members based on the typhoon track errors, and the precipitation forecasts of the advantaged group have the highest forecast skill scores. The forecast errors are highly correlated with the intensity of the subtropical high (SH) and the steering flow of typhoon. The significantly weak SH in the ICs/BCs leads to a weaker SH in the forecasts, resulting in an easterly and slower typhoon motion and more accumulated precipitation. Also, the precipitation forecast errors are highly associated with the simulation errors of typhoon structure in different landfall stages, while the vertical wind shear may cause these errors. In addition, the orographic convergence and uplift promote the precipitation production, but the model tends to overestimate the orographic effects. About the simulation of hydrometeors, large snow content leads to increased precipitation in the Thompson scheme, while in the WDM6 scheme, the strong evaporation of low-level raindrops results in decreased precipitation.