2019 年 128 巻 1 号 p. 77-92
The Japanese Northern Alps, facing the Sea of Japan and composed of several high mountains with peaks over 3,000 m above sea level, are known to form one of the heaviest snowfall areas in the world. However, the horizontal distribution of snow cover and inter-annual and seasonal variations of snow cover have not been understood completely because of a lack of observations. Therefore, numerical simulations are conducted using the Nonhydrostatic Regional Climate Model (NHRCM) with 2-km and 5-km grid spacings (hereafter, NHRCM02 and NHRCM05, respectively), and the simulations are compared to snow cover observations at Murododaira and along the Tateyama-Kurobe Alpine Route. Both NHRCM02 and NHRCM05 simulate interannual variations of snow depth observed at Murododaira in mid-March and late April well, although the snow depth is slightly overestimated. Compared to snow cover observations in the winter of 2014/15, NHRCM05 has remarkably large biases of snow depth at Midagahara and Daikandai. The overestimations of snow depth are improved by NHRCM02 at these stations. The 17-year-mean maximum snow depth shows that NHRCM05 calculates a greater snow depth at the windward side of the Japanese Northern Alps than does NHRCM02, which also influences the altitudinal dependence of snow depth. The horizontal resolution of NHRCM05 is not high enough to resolve convective precipitation over the Sea of Japan in winter. Therefore, NHRCM05 simulates less precipitation over the Sea of Japan and more over mountainous areas than NHRCM02, which results in large snow cover biases at the windward side of the Japanese Northern Alps.