Hydrologic modelling of the Zackenberg river basin : an applied study using the Soil and Water Assessment Tool

Sammanfattning: Since the 1980s, the Arctic has experienced an amplified warming of more than twice the rate of the global-mean, leading to large-scale changes in the Arctic hydrologic system, ultimately having cascading feedbacks on the global climate. However, few of today’s distributed models manage to capture the complex processes in Arctic hydrology, and therefore, the aim of this thesis was to evaluate the usage of the distributed Soil and Water Assessment Tool (SWAT) model, to see whether it could capture the different surface water paths and processes characterizing the Arctic water cycle. The model was applied on the Zackenberg river basin, situated in Northeastern Greenland, using spatial data of topography, soil, and vegetation cover, together with observed meteorological data from Zackenberg climate station. The model was evaluated by comparing simulated river discharge in Zackenberg river to observations, using Nash-Sutcliffe Efficiency (NSE) and Coefficient of determination (R2), and by comparing simulated water flows to previously modelled fluxes. The modelled river discharge had a NSE and R2 of 0.62, indicating good agreement. Glacier melt was estimated to a mean of 930 mm w.e./year, within the range of previous estimates, while snowmelt was largely underestimated with a value of 96 mm w.e/year, possibly caused by topographic influences on snow distribution and precipitation data input. The model successfully captured seasonal freezing- and thawing cycles, but largely simplified active layer dynamics. The results indicate that the null hypothesis may be rejected. For future improvements, the methodology should include the usage of elevation bands and coupling to heat transfer algorithms, to fully capture snow distribution and seasonal permafrost dynamics.