Abstract
We developed a snowfall/snowmelt model consisting of approximately 1km-square gridded meshes, in which daily snow accumulation and snowmelt are calculated based on the simplified energy balance with precipitation, temperature, wind speed and sunshine duration data. The model was applied to the Seki River Basin, a representative snowy basin amidst a warm climate in Japan. By utilizing satellite images of snow-capped areas, we estimated the spatial distribution of a model parameter of sensible/latent heat transfer in regions where observed meteorological data are sparse. Comparison of the calculated and observed snow water equivalents revealed that the model successfully represents the spatial distributions of snowfall/snowmelt within the ranges of 200mm, except for areas where locally intensive snowfalls occur. In addition, calculated discharges of snow melting periods, obtained by applying the estimated snow melt in each mesh to a distributed water circulation model, showed high accuracy in representation of flood peaks and regressions.
