Characteristics of systematic biases in the winter Northern Hemisphere of the Korean Integrated Model

The Korean Integrated Model (KIM), operated by Korean Meteorological Administration (KMA) since April 2020, has been improved through continuous updates on model physical processes and data assimilation. As the model version was updated, performance was substantially improved during the summer season, but the improvement was not clear in the winter. Therefore, in this study, we examine a dropout case degrading performance of KIM in the northern hemisphere on 2022 winter in order to investigate the causes of systematic biases related with the KIM performance during the winter season. The worst performance in terms of 5-day forecast of 500 hPa geopotential height occurred at 12UTC on the 20th of January 2022 with cold biases in the lower Arctic and Siberia regions. Especially, the cold bias in Siberia is strong at the surface and is amplified as the forecast time increases from the initial time. The surface temperature cold bias in Siberia appears robustly in winter as well as this case. This is contrast with the summer season with warm bias, which mostly seems to be due to the underestimated soil moisture and surface latent heat flux. In winter, the latent heat and soil moisture seem to have little effect on the cold bias, and it is analyzed that lower tropospheric cloud may one of main factors that affect the cold bias in Siberia. The cloud and snow contents associated with the lower tropospheric cloud in Siberia are less compared with the ERA5 reanalysis, which in turn contributes to the underestimation in surface downward longwave flux through the cloud-radiation interaction. The reduction in radiative surface heating seems to be one of main causes of the lower tropospheric cold bias in Siberia.