Comparative assessment of 1-month temperature forecasts in South Korea based on dynamical downscaling of CFSv2 and FuXi global predictions

Despite the growing availability of global ensemble forecasts, their application at regional scales remains challenging due to the coarse spatial resolution. Meanwhile, machine learning (ML)-based forecasts are emerging as promising alternatives to traditional physics-based models, yet they are mostly deterministic, and their potential has not been fully explored beyond the medium-range timeframe. This study presents a comparative assessment of one-month temperature forecasts over South Korea based on dynamical downscaling of the physics-based CFSv2 and the ML-based FuXi-ENS. Selected ensemble members from each system are dynamically downscaled using WRF, and their forecasting performance is evaluated in terms of temporal and spatial patterns. Results indicate that temporal predictability is largely inherited from the driving global forecasts, whereas dynamical downscaling primarily enhances the spatial distribution of temperature. While the downscaled CFSv2 exhibits larger relative improvements in spatial correlation due to its poor performance, FuXi-ENS provides comparatively more skillful temperature patterns initially and benefits from a more refined representation of spatial variability after downscaling. Overall, this study offers a comprehensive assessment of extended-range temperature prediction over South Korea, illustrating the operational potential of hybrid approaches that combine global seasonal forecasts with regional dynamical downscaling and providing insights for the future development of hybrid forecasting systems.