Item

Abstract

Hydrologic processes are often measured and understood at the point scale. However, there is significant spatial variability in rates of snowmelt, sublimation, and snowfall over larger landscapes. The role of atmospheric and local terrain characteristics in controlling the variations between the small-scale and the bulk response to spatially heterogeneous mass and energy inputs is poorly understood. The application and identification of appropriate upscaling rules are necessary to translate small-scale descriptions of snow processes into constitutive relationships that are applicable at larger scales. As a proof of concept, this study examines the temporal and spatial variability of sublimation and snowmelt fluxes in a drainage basin in the Canadian Rockies using machine learning methods. In the approach, Raven, a fine-resolution hydrologic model, generates high-resolution fluxes and state variables used in training and testing machine learning models. This study involves estimating spatially averaged results from discretized fine-scaled models without explicit knowledge of detailed local response and with and without low-order statistics of state (e.g., the standard deviation of snow water equivalent). A series of experiments progressively increasing in complexity are used to test and validate that the upscaling methodology can successfully represent the impact of heterogeneity within the system. Weights from the machine learning models are then incorporated into a mass balance equation at a coarser resolution to demonstrate the efficacy of this upscaling methodology for practical hydrologic modeling.