Evaluating snow models with varying process representations for hydrological applications

摘要: Much effort has been invested in developing snow models over several decades, resulting a wide variety of empirical and physically based models. For the most part, these are built on similar principles. The greatest differences found how each model parameterizes individual processes (e.g., surface albedo compaction). Parameterization choices naturally span range complexities. In this study, we evaluate performance different parameterizations for hydrological applications using an existing multimodel energy-balance framework data from two well-instrumented alpine sites with seasonal cover. We also include temperature-index intensive, multilayer our analyses. Our results show that mass observations provide useful information evaluating ability to predict snowpack runoff, whereas depth alone not. appear transferable between study sites, behavior which is not observed temperature predictions due site-specificity turbulent heat transfer formulations. Errors input validation data, rather than formulation, seem be factor affecting performance. three types reproduce daily runoff when appropriate structures chosen. Model complexity was determinant predicting reliably. shows usefulness identifying under given constraints such as availability, properties interest computational cost.