Comparison of Process-Based and Temperature-Index Snowmelt Modeling in SWAT

Snowmelt hydrology is an important part of hydrological analyses where significant proportion of precipitation is expected to fall in a snow form. Many models have long been introduced to enable the simulation of snowmelt processes in the watershed ranging from simple temperature based equations to complex and sophisticated process-based equations. Usually, mixed results have been reported whether or not the difference between results achieved by incorporating data demanding models vis-A -vis simple temperature-index models is justifiable. In this study, we compared the performances of physically based energy budget and simpler temperature-index based snowmelt calculation approaches within the SWAT model at three sites in two different continents. The results indicate insignificant differences between the two approaches. The temperature-index based snowmelt computation method had the overall model efficiency coefficients ranging from 0.49 to 0.73 while the energy budget based approach had efficiency coefficients ranging from 0.33 to 0.59 only. The magnitude of the differences varied based on where the models were applied. However, comparison between two process-based snowmelt estimation procedures (with and without the inclusion of aspect and slope as factors dictating the incoming solar energy) indicate that accounting for ground surface slope and aspect in the snowmelt model slightly improved the results. We conclude that for most practical applications where net solar radiation, not turbulent heat flux, dominates the snowmelt dynamics, a simpler temperature-index snowmelt estimation model is sufficient.