Impact of Horizontal Resolution on the Robustness of Radiation Emulators in a Numerical Weather Prediction Model

Abstract

Developing a machine-learning-based radiative transfer emulator in a weather forecasting model is valuable because it can significantly improve the computational speed of forecasting severe weather events. To replace the radiative transfer parameterization in the weather forecasting model, the universal applicability of the radiation emulator is essential, indicating a transition from the research to the operational level. This study investigates the degradation of the forecast accuracy of the radiation emulator for the Korea peninsula when it is tested at different horizontal resolutions (100–0.25 km) concerning the accuracy attained at the training resolution (5 km) for universal applications. In real-case simulations (100–5 km), the forecast errors of radiative fluxes and precipitation were reduced at coarse resolutions. Ideal-case simulations (5–0.25 km) showed larger errors in heating rates and fluxes at fine resolutions, implying the difficulty in predicting heating rates and fluxes at cloud-resolving scales. However, all simulations maintained an appropriate accuracy range compared with observations in real-case simulations or the infrequent use of radiative transfer parameterization in ideal-case simulations. These findings demonstrate the feasibility of a universal radiation emulator associated with different resolutions/models and emphasize the importance of emulating high-resolution modeling in the future.