Spatial snowfall distribution in mountainous areas estimated with a snow model and satellite remote sensing

We describe an approach for reconstructing snowfall that combines satellite observations of the snow disappearance date (SDD) with a snow model for two mountainous areas in Japan having distinct snow climatology. This approach allows assessment of the distribution of snow and topographical effects on snowfall within a catchment. We also evaluated how the reconstructed snowfall affects the catchment snow hydrology. Validation at observation sites demonstrated that a combination of the snow model and snowfall reconstructions successfully estimated the seasonal changes of snow water equivalent (SWE). In Japan, the dependence of snowfall on elevation is stronger in mountainous areas along the coast on the Sea of Japan side of Japan's central mountain spine (where snowfall triples with every 1000 m increase in elevation; C-sf = 0.002 m(-1)) compared with inland locations of the same region (where snowfall doubles with every 1000 m increase; Csf = 0.001 m(-1)). Moreover, the reconstructed snowfall improved the estimation of maximum catchment SWE. Maximum total SWE, estimated with reconstructed snowfall, was 3.8 x 10(8) m(3) in Kurobe catchment (along the coast on the Sea of Japan side of Japan's central mountain spine), while that, estimated by convectional method with the spatially-constant Csf = 0.001 m(-1), was 2.0 x 10(8) m(3). As a result, estimations of the snow disappearance date and of the catchment snowmelt were also improved. These results suggest that it is useful to estimate the spatial snow distribution, especially where steep topography causes large gradients of snowfall amount with respect to elevation.