Abstract
A practical model has been developed to estimate the daily snow water equivalent. The model was designed to fit the mesh Agricultural meteorological data system that was developed by NARO (National Agriculture and Food Research Organization). The input data for this model are the daily value of mean/maximum/minimum air temperature, mean wind speed, mean relative humidity, global solar irradiance, downward long-wave radiation, and precipitation. These are supplied from the mesh Agricultural meteorological data system. In order to accurately estimate, reduce the computational load on the server, this model calculates the heat balance at the snow surface in half-day intervals. The features of the model are as follows: 1) It individually calculates the daytime and nighttime heat flux that would have been canceled in daily data accumulation. 2) By using a distribution map of the wind speed ratios of the daytime and nighttime, average wind speeds of daytime and nighttime are estimated from the daily mean wind speed. Therefore, the amounts of sensible and latent heat exchange are estimated accurately. Verification results of the nationwide snow data of AMeDAS from 2010 to 2014 indicate that the RMSE of the snow cover period are between 10.2 and 15.6 days, RMSE of snow disappearance date are between 7.8 and 13.3 days. It was suggested that the main source of error is in the estimation of snowfall water.
