Representation of Snow in Urban Drainage Models

Modeling guidelines of parsimony (simplicity of assumptions), accuracy, modesty, and testability are discussed in terms of urban snowmelt induced runoff. Heterogeneity and variability at the 10–100-m spatial and <1 day timescales of snow distribution, energy availability, and surface permeability in urban environments are outlined. The snow accumulation and melt routines of three drainage models that have been applied to urban settings are reviewed; two of these, MouseNAM and SWMM, were designed for urbanized catchments; the third, HBV, is a regional-scale model for rural catchments. The surface hydrology of the models have varying degrees of complexity. All contain a temperature index for melt—this method is shown to be theoretically unsound without modification for urban simulations. There are discordant scales between processes important for snowmelt runoff generation and those modeled. Literature on model development, validation, and application is lacking. Urban snow data are also sparse, which limits the possibility of assessing or improving model performance.