The ABC's of Snowmelt: A Topographically Factorized Energy Component Snowmelt Model

Document Type


Journal/Book Title/Conference

Hydrological Processes, Special Issue: Snow Hydrology




12, 13


John Wiley & Sons Ltd


Brownsville, Vermont

Publication Date


First Page


Last Page



Because of the crucial role snowmelt plays in many watersheds around the world, it is important to understand and accurately quantify the melt process. As such, numerous mathematical models attempting to describe and predict snowmelt have arisen. There are two main categories of models: conceptual index models and more intricate energy balance models. The index models, like the degree-day or radiation index models, are practical enough for use in large basins for operational purposes; while the energy balance models, though they are complicated and require large amounts of data, can represent the physics behind melt and give more accurate representations of the spatial distribution of melt within small research basins. The ABC model presented here attempts to bridge the gap between these two extremes by providing a simple yet physically justifiable method that uses elevation and radiation indices together with some measurements to distribute melt over a watershed. This new model separates the energy that causes snowmelt into three components: a spatially uniform component, a component that is proportional to elevation, and one that is proportional to solar illumination (which is determined by topography). Measurements of snowmelt at several topographically unique points (called ‘index points’) in a watershed are related to elevation and solar illumination through regression in order to factor the melt energy into the three separate components at each time step. The model is driven using inputs from snowmelt measurements at the index locations used to calibrate the regression at each time step. Then the spatial patterns of solar illumination and elevation are used to predict the spatial distribution of melt over the whole watershed. Field data supplemented with synthetically generated data is used to test the model.

This document is currently not available here.