Development of simple snow density model for wide area

At first, a model is constructed to estimate snow depth, which consists of heat balance and consolidation process model, using the Meteorological Agency data. This model needs precipitation, temperature and wind velocity as input data and calculates hourly snow depth, runoff from snow and snow density. Secondly, this model is applied to 783 observing points in Japan. There are 203 points where the correlation coefficient of snow depth between the observation and the calculation is over 0.75. It was assumed that the calculation of snow density is excellent result when the model can reappear the snow depth. As a third step, the relationship between the change of snow depth and the density is investigated for every region. It is found that the density increases slowly in the beginning of snowfall season, stays stable in the middle season of snowfall. It increases rapidly in snowmelt season where there is heavy snow and the maximum depth is over 1.0 m. In addition, the density in the little snow region with a maximum snow depth below 1.0 m is about uniform and the value is about 0.15(g/cm³) throughout winter. Wide area information of snow depth from satellite is given roughly once a month. Therefore the gradient of snow depth is compared for a month with the density. As the result, It is shown that snow density can be estimated using the gradient and the time to estimate the same in wide area is very short.