Modelling of snow processes in catchment hydrology by means of downscaled WRF meteorological data fields

Abstract

Abstract. Detailed physically based snowmelt models require a complete set of meteorological forcing data at the model's scale. Besides precipitation and temperature, time series of humidity, wind speed, and radiation have to be provided. The availability of these time series is in many cases restricted to a few meteorological stations and consequently, snowmelt modelling is often highly uncertain. To overcome this dilemma, the suitability of downscaled atmospheric analysis data for physically based snowmelt simulations in hydrological modelling is studied. We used the Weather Research and Forecast model (WRF) to derive spatial and temporal fields of meteorological surface variables as boundary conditions for four different snowmelt models. The simulations were carried out at the point scale and at the catchment scale for the Sieber catchment (44.4 km2), Harz Mountains, Germany. For the latter, all snowmelt models were integrated into the hydrological modelling system PANTA RHEI. All models performed well at both scales. In conclusion, the presented approach is suitable to derive reliable estimates of snowpack and snowmelt processes as part of water balance and flood simulations for catchments exposed to snow.