Abstract
This thesis focuses on how variations in snow water equivalent (SWE) and snow covered area (SCA) influence the process
of snowmelt on macro- and localscale and how this could be simulated with proper results in a hydrological model. Variations in SWE were observed by manual measurements in the catchment area of Lake Grønsjøen in the middle of Norway during the spring 2004. Variations in SCA were found by analyzing remote sensing images covering the same area.
Two types of hydrological models were used in the analysis. The HBVmodel represents a semi-distributed model with a simple terrain classification (hypsographic curve), while DEMLab represents a distributed model where the terrain is represented by a digital elevation model. Observation of snow (SWE and SCA) and discharge were used in validation of the models.
Comparing the different remote sensing images shows major differences between the images, even between images from the same day. The differences between the images made it difficult to point out one true distribution of SCA. The differences also affected the precision in the simulation results. After the first calibration-process the semi-distributed HBV-model produced better simulationresults than the distributed DEMLab, but after recalibration, updating and distribution of the parameters in the snowroutine in DEMLab, the simulation results from DEMLab showed to give a better representation
of runoff.
i