Evaluation of Satellite Images of Snow Cover Areas for Improving Spring Flood in the Hbv-model

For the hydro power companies it is of major importance to know the magnitude of the spring flood. Hence, the amount of snow is crucial for forecast modelling. In mountainous areas with plenty of snow, the meteorological network is often sparse, and they lack direct observations of the snow pack. Therefore, if it is possible to combine satellite snow cover area information with a hydrologic model, it would enhance the quality of the forecasts. In this project, financed by ELFORSK, a comparison is made between snow covered area (SCA) from satellite processed images and the HBV-model. The satellite processed images are based on data from Terra/Modis and Envisat/SAR. The images are produced in near real-time by NORUT and Kongsberg Satellite Services AS, for a project called Nordic Mountain Snow Hydrology (GSE) supported by European Space Agency (ESA). The HBV-model was calibrated without the use of snow observations. The evaluation has been made for mountainous catchments with less than 30 % forest cover. In most catchments there is a consistency between the difference in observed and simulated SCA and the error in simulated runoff. During spring 2008 updates of the HBV-model will be made in real-time.