Modelling spatial variability of snowmelt in an arctic catchment
Pohl, Stefan Hans Gustav
Spring snowmelt in open environments is characterized by a high degree of spatial variability due the combination of a highly variable end of winter snow cover and spatially variable snowmelt energy fluxes. This often leads to the quick development of a mosaic pattern of coexisting snow covered and snow free patches. Snow cover and melt energy variabilities and the resulting melt patterns greatly affect timing, location, and rate of meltwater release, as well as the surface energy balance of the composite landscape. Although spatially variable snow covers and melt energy fluxes have been considered for mountainous regions, the importance of the various controlling factors for snowmelt in low relief regions is not well known. As a result of the lack of previous studies, it has not been possible to properly address these processes in applicable hydrologic or land-surface models. The goal of this study is to provide a better understanding of the relative magnitude of the small-scale variabilities in snowmelt of open environments, and if important, to make recommendations on how to include these processes in both hydrologic and land surface models. The present dissertation specifically considers the small-scale variability in snowmelt over arctic tundra surfaces, although the methods used could be applied to a wide variety of open environments. A "state of the art" coupled hydrologic model - land surface scheme, WATCLASS, was employed to simulate snowmelt in the study basin. The study shows that while the timing of snowmelt and meltwater runoff was fairly well predicted by the model, the spatial variability of the snowmelt processes was not well captured. The study indicates that the omission of topographical effects on end of winter snow cover and snowmelt energy fluxes limited the models capability to simulate snowmelt patterns of snow covered and snow free areas. The topographical influences on two major factors of the snowmelt energy balance, incoming solar radiation and turbulent fluxes of sensible and latent heat, were, therefore, studied in detail with small-scale (resolution = 40 m) model simulations. The results show that small-scale variabilities in both energy fluxes play an important role for determining melt rates, meltwater runoff, and surface energy balance values even in the relatively gentle terrain of the study area. Finally, the obtained energy fluxes were used to compute a spatially distributed, full snowmelt energy balance. The results show that the overall variability depended strongly on cloud cover and dominant wind directions in relation to incoming solar radiation angles. The energy balance was subsequently used in combination with a variable end of winter snow cover to simulate the progress of melt throughout the research basin. The study shows that in order to accurately predict the first snow free areas and areas with late lying snow drifts, small scale variabilities in end of winter snow cover and snowmelt energy fluxes need to be considered. Little inter - annual differences were found in the distribution of snow covers and energy fluxes suggesting that it might be possible to statistically link small-scale variabilities in snowmelt processes to certain key terrain properties for use in larger scale models. However, more studies in different topographical settings are needed to test this approach.