Ha et al. (2026) Comparative assessment of 1-month temperature forecasts in South Korea based on dynamical downscaling of CFSv2 and FuXi global predictions

Identification

• Journal: Scientific Reports
• Year: 2026
• Date: 2026-09-08
• Authors: Subin Ha, Xiaohui Zhong, Lei Chen, Hao Li, Jina Hur, Hyun‐Han Kwon, Eun‐Soon Im
• DOI: 10.1038/s41598-026-69209-8

Research Groups

• Department of Atmospheric Sciences, Fudan University
• National Centers for Environmental Prediction (NCEP)
• European Centre for Medium-Range Weather Forecasts (ECMWF)

Short Summary

This study compares the performance of three global seasonal forecast systems - CFSv2, FuXi-ENS, and SEAS5 - in predicting 1-month temperature forecasts over South Korea. The results show that dynamical downscaling can enhance spatial detail and improve regional representation, but its effectiveness depends on the quality of the input forecasts.

Objective

• Evaluate the performance of global seasonal forecast systems (CFSv2, FuXi-ENS, SEAS5) in predicting 1-month temperature forecasts over South Korea.
• Assess the added value of dynamical downscaling in improving spatial representation and regional usability.

Study Configuration

• Spatial Scale: Regional scale (South Korea)
• Temporal Scale: Monthly timescale (July)

Methodology and Data

• Models used:
  • CFSv2 (Climate Forecast System version 2)
  • FuXi-ENS (Machine learning-based seasonal forecasting system developed by Fudan University)
  • SEAS5 (Seasonal Forecasting System version 5 from ECMWF)
• Data sources:
  • Korean Meteorological Administration observations
  • ERA5 reanalysis data

Main Results

• Dynamical downscaling enhanced forecast performance by adding spatial detail and better capturing topographically driven variability.
• Downscaled FuXi-ENS successfully reproduced colder conditions over mountains and warmer conditions over plains, reducing the systematic cold bias of the global forecast.
• CFSv2 also benefited from downscaling, with substantial improvements in spatial correlation with observations at station locations.

Contributions

• This study provides a comprehensive assessment of extended-range temperature prediction over South Korea using hybrid approaches that combine global seasonal forecasts with regional dynamical downscaling.
• The results highlight the importance of considering both temporal and spatial variability when evaluating forecast performance.

Funding

• National Natural Science Foundation of China (Grant No. 42175040)
• National Key Research and Development Program of China (Grant No. 2018YFC1504402)

Citation

@article{Ha2026Comparative,
  author = {Ha, Subin and Zhong, Xiaohui and Chen, Lei and Li, Hao and Hur, Jina and Kwon, Hyun‐Han and Im, Eun‐Soon},
  title = {Comparative assessment of 1-month temperature forecasts in South Korea based on dynamical downscaling of CFSv2 and FuXi global predictions},
  journal = {Scientific Reports},
  year = {2026},
  doi = {10.1038/s41598-026-69209-8},
  url = {https://doi.org/10.1038/s41598-026-69209-8}
}