Evaluating Snowmelt Modeling Methods of Different Complexities at a Watershed in the Canadian Rockies

Three snowmelt models of different degrees of model complexity and data requirement were used to simulate snow accumulation and ablation processes for a watershed in the Canadian Prairies. Results show that modifying the popular temperature index method by incorporating near-surface soil temperature as an additional predictor to air temperature could considerably improve the model performance. In addition, the modified method could achieve hourly simulation results that are comparable to a physically based energy balance method. Majority of the improvement in the modified temperature index method occurs when the melt rate is varied as a function of near-surface soil temperature.