عنوان مقاله [English]
Monitoring snow cover areas and estimating the resulting snowmelt runoff in mountain areas is highly important.
There are mainly two approaches to improve the performance of snowmelt runoff in rainfall-runoff models, including improvement of snow parameters through observations and modifying their snowmelt module. This study investigates the impact of these approaches on the accuracy of snowmelt runoff using the SWAT model in the mountain Basin of Zarineh River, northwest of Iran. Firstly, the snow parameters of SWAT were extracted through the MODIS daily snow cover products and the in-situ snow depth data and were replaced in the model. Secondly, the UBCWM energy balance model was substituted for the original SWAT snowmelt module, which is based on the degree-day method. Results show that the baseline SWAT model underestimates runoff at the upstream and downstream hydrometric stations. However, the performance of the model with the modified snow parameters improves at the two stations by 6% and 14% for Nash–Sutcliffe Efficiency (NSE) and 35% and 14% for Kling-Gupta Efficiency (KGE), respectively. On the contrary, using UBCWM in SWAT deteriorates the simulated runoffs at the two stations due to underestimating the peaks and overestimating the snowmelt duration. This study's findings can help achieve more accurate runoff estimations in mountain basins, especially during peak runoffs.