2013 年 69 巻 4 号 p. I_175-I_180
A physically based distributed biosphere hydrological model with three layered energy balance snow melt module (WEB-DHM-S) has been implemented at point scale at Yagisawa dam site to evaluate the long-term simulation of snow depth from 1948 to 2006 using high resolution JP10 reanalysis data. The time series of snow depth, its averaged value and anomaly and the snow cover durations are evaluated. The model is able to capture the seasonal and inter-annual variability of snow depth well with average bias at 0.013 m and root mean square error at 0.29 m. A continuous negative anomaly was well simulated from 1985 to 1994. Long-term analysis showed that the decadal change was more apparent for snow cover days with snow depth above 50 cm and the number of snowfall days had a decreasing trend. In addition, the snow depth values from 1948 to 1963, an outcome of this research, can be used as the reference dataset by the scientific community.