Assessment of the accuracy of snow water equivalent calculation with the use of global numerical weather prediction models and SnoWE snowpack model (by the example of the Kama River basin)

Abstract

The paper discusses the results of snow cover formation and snowmelt modeling in the Kama river basin (S = 507 km2) using two approaches previously developed by the authors. The first one is the SnoWE snowpack model developed at the Hydrometeorological Center of the Russian Federation and used in quasi-operational mode since 2015, and the second is GIS-based empirical technique which was previously implemented for the Kama river basin. Both methods are based on a combination of numerical weather prediction (NWP) models data with operational synoptic observations at the weather stations. The study was performed for the winter seasons 2018/19 and 2019/20. To assess the reliability of simulated snow water equivalent (SWE), we obtained in-situ data from 68 locations (snow survey routes) distributed over the entire area of ​​the river basin. As a result of the study, the main advantages and limitations of two methods for SWE calculation were identified. As for the maximum values of SWE, the root mean square error (RMSE) of simulated SWE ranges from 14% to 28% of the average observed SWE according to in-situ data. It was found, that the SnoWE model more reliably reproduces SWE in the lowland part of the river basin. Simultaneously, SWE was substantially underestimated according to the SnoWE model in the northern and mountainous parts of the basin,. The second method provides a more realistic estimate of the spatial distribution of SWE over the area, as well as a higher accuracy of calculation for its northern part of the river basin. The main drawback of the method is the substantial overestimation of the intensity of snowmelt and snow sublimation. Consequently, the accuracy of SWE calculations sharply decreases in the spring season. Wherein, SWE calculation accuracy in the winter season 2019/20 was substantially lower than in 2018/19 due to frequent thaws.