Calculating of snow cover characteristics on a plain territory using the model SPONSOR and data of reanalyses (by the example of Moscow region)

Abstract

The technique for calculating the snow cover characteristics (a water equivalent and a snow cover thickness) with high spatial and time resolution on spacious plains is proposed. The model SPONSOR of local heat- and moisture exchange (Land-Surface Model, LSM) and data of reanalyses NCEP/DOE and ECMWF ERA-Interim were used for calculations. The above characteristics of the snow cover on the test area of the Moscow region were calculated using this method over the period 1979–1996. The results were compared with actual data of the snow gauge stations and with data on snow cover, derived directly from reanalysis. The data from the NCEP/DOE reanalysis did not show satisfactory agreement with data of the observations for both the water equivalent and the thickness (Fig.  1,  б and Fig. 2, б): deviations reached 60–70%. Monthly mean values of snow water equivalent from the ERA-Interim reanalysis were in a good agreement with the observations, but the snow thicknesses were reproduced much worse. At the same time, using the LSM SPONSOR with input meteorological data from the reanalyses allowed obtaining the snow cover characteristics which were in a good agreement with data of the observations for both the monthly means and individual daily values. The correlation coefficients with the data of snow gauge surveys increased, on the average, up to 0.83–0.89 for the water equivalent, and up to 0.85–0.91 for the snow depth (see the Тable in the text). Especially good results were obtained when meteorological data from the ERA-Interim reanalysis were used together with the LSM SPONSOR (Fig. 1, д and Fig. 2, д). It allows us to conclude that meteorological data from the ERA-Interim reanalysis together with data of regular observational network can be used as an additional source of information for calculations of the snow characteristics. This conclusion is especially important for areas with sparse network of regular observations.