Submitted to: Journal of Hydrometeorology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/14/2002
Publication Date: 4/20/2002
Citation: Winstral, A.H., Elder, K., and Davis, R.E. 2002. Spatial snow modeling of wing redistribution snow using terrain-based parameters. Journal of Hydrometeorology 3:524-538. Interpretive Summary: In the arid Western U.S., over 70% of the annual water supply results from snowmelt. In this region, water supply and river streamflow forecasts are heavily reliant upon models of snowmelt runoff capable of predicting melt rates on an hourly to seasonal basis. The streamflow forecasts are used by reservoir managers to regulate water levels that maximize energy production and water supply while minimizing potential flooding, and by downstream irrigators to enhance crop production. Snowmelt runoff models are based on inputs on the distribution of snow (i.e. how much water is available for runoff at each particular location) and the energy components that control the rate of melt at that location. Wind is one of the strongest influences on snow distribution in high alpine regions causing large snow accumulations to form downwind of mountain peaks with little or no accumulation on the windward sides. The complexity of wind fields over alpine terrain, however, has hindered the ability of complex, physics-based modeling techniques to accurately and efficiently provide the wind induced snow redistribution information required by forecasters. Using a computerized representation of terrain, we devised measures that quantify the degree to which the upwind terrain shelters or exposes a site. The shelter/exposure factors that we developed used a minimum of computer time and were found to effectively characterize the effects of wind upon the observed snow distribution. These parameters provide an efficient means of including wind effects within snowmelt runoff models suitable for forecasting purposes.
Technical Abstract: Although wind is widely recognized as a dominant control on snow accumulation in exposed alpine regions, the complexity of wind fields in these typically rugged locales has hindered inclusion of wind redistribution effects in spatial snow models. This study avoided the difficulties associated with physically exhaustive wind field modeling and took a terrain-based approach to simply characterize the effects of wind. One parameter, , was based on maximum upwind slopes relative to seasonally averaged winds to characterize the wind scalar at each pixel location in an alpine basin. A second parameter, , measured upwind breaks in slope from a given location and was combined with an upwind application of to create a drift delineator parameter, D0. D0 was used to delineate sites of intense redeposition on lee slopes. Based on 504 snow depth samples from a May 1999 survey of the upper Green Lakes Valley, CO, USA, the correlation of the developed parameters to the observed snow distribution and the effect of their inclusion in a spatial snow distribution model were quantified. was found to be a significant predictor, accounting for more of the variance in the observed snow depth than could be explained by processes represented by elevation, solar radiation, or slope. Samples located in D0-delineated drift zones were shown to have significantly greater depths than samples located in non-drift zones. A regression tree model of snow distribution based on a predictor variable set of , D0, elevation, solar radiation, and slope explained 8-23% more variance in the observed snow distribution, and performed noticeably better in unsampled areas of the basin compared to a regression tree model based on only the latter three predictors.