Snow hydrology: The parameterization of subgrid processes within a physically based snow energy and mass balance model

The objective of this research was to develop techniques for the representation and parameterization of subgrid and distributed snow processes within snowmelt models. Snowmelt is driven by energy exchanges at the snow surface that have horizontal variability down to scales of 1 to 10 m. When areas with large extent are modeled, it is impractical to apply models distributed on a 1 to 10 m grid. Large model elements, either grid squares or topographically delineated, need to be used. Therefore, it is necessary to develop modeling approaches that can parameterize the variability within these elements, referred to as subgrid variability. The role of subgrid variability increases as the model element size is increased, so as scale increases the representation of subgrid variability becomes more important. Computational tools are needed to explore the scale dependence of the subelement representation. Since subgrid variability is closely related to the topography, the parameterization of subgrid variability from the topography or topographical features was investigated.; This dissertation is a collection of four papers that address some of these challenges. The goal is to combine physically based modeling emphasizing physical understanding of the reasons for and processes involved in spatial variability of snow and snowmelt with the analysis of extensive existing remotely sensed and field-based data from a Colorado Front Range watershed - Green Lakes Valley (GLV) watershed. A small-scale distributed model was used to quantify and refine the representation of the spatial snow accumulation and melt processes. This then formed the basis for parameterization of subgrid variability through the use of depletion curves and the derivation of these depletion curves from digital elevation data. As a final step the scale dependence of depletion curves was explored and to some extent quantified.