2025 年 21 巻 p. 24-33
Over the Northwest Pacific off Hokkaido, low clouds such as stratocumulus, stratus, and fog form frequently in summer. Despite the scientific and socioeconomic importance of these low clouds, our understanding of their reproducibility in weather prediction models under different synoptic circumstances is still lacking. This study assesses the ability of low-cloud representation and prediction in the Japan Meteorological Agency Meso-Scale Model (MSM) mainly in June and July 2020-2022, focusing particularly on synoptic fluctuations of near-surface temperature advection and low-cloud fraction. The low-cloud fraction predicted from the preceding day corresponds quite well with that in the analysis. However, the low-cloud fraction in the analysis is not highly correlated with satellite observations on daily timescales, while the mean low-cloud fraction agrees with the observations. The MSM sometimes simulates unrealistic fog within the near-surface stable layer associated with warm advection, resulting in a positive cloud fraction bias and a negative downward shortwave radiation bias. Under both warm and cold advection, the MSM may also underestimate the low-cloud fraction, although it still reproduces the horizontal distribution of low clouds rather consistently with satellite observations.
