Answers are grounded in this paper and cite their sources.
One of the key factors contributing to uncertainty in numerical weather prediction (NWP) models is cloud-related feedback mechanisms. This study represents the first attempt to couple the radiative transfer model named Dayu with the Weather Research and Forecasting (WRF) Model to produce Advanced Himawari Imager (AHI) brightness temperatures (BTs) based on WRF-simulated cloud profiles. These simulated BTs were then compared with actual AHI observations. This study focuses on Typhoon Hato, which caused extensive economic damage and casualties in 2017. The results demonstrate that though the WRF Model effectively simulates the typhoon’s track and the general spatial distribution and characteristics of clouds, significant negative biases exist between the simulated and observed BTs in the cloudy regions. To better evaluate the simulated cloud properties, a machine learning–based cloud classification algorithm was developed to differentiate between clouds at varying altitudes for the AHI imager onboard Himawari-8 satellite. The analysis reveals that the WRF Model performs well in simulating clear skies and high-level clouds, though it slightly overestimates high-level cloud cover, with the ratio of WRF-simulated to satellite-observed high-level cloud pixels ranging from 1.06 to 1.32 during the forecast period. In contrast, it significantly underestimates the presence of low-level clouds, with the ratio falling below 0.1. Furthermore, the overestimation of high-level cloud cover by the WRF Model contributes to an overestimation of precipitation. This study contributes to a deeper understanding of cloud-related uncertainties in typhoon simulations using the WRF Model.
No takes yet. Share an insight, caveat, or question.
Wu et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: