Using geostatistical methods to estimate snow water equivalence distribution in a mountain watershed

TitleUsing geostatistical methods to estimate snow water equivalence distribution in a mountain watershed
Publication TypeConference Proceedings
Year of Conference1998
AuthorsBalk, B., Elder K., and Baron J.
Conference Name66th Annual Western Snow Conference
Series TitleProceedings of the 66th Annual Western Snow Conference
Date PublishedApril 1998
PublisherWestern Snow Conference
Conference LocationSnowbird, Utah
KeywordsSnow accumulation, Snow covered area, Snow density, Snow depth, Statistical analysis
Abstract

Knowledge of the spatial distribution of snow water equivalence (SWE) is necessary to adequately forecastthe volume and timing of snowmelt runoff. In April 1997, peak accumulation snow depth and density measurements were independently taken in the Loch Vale watershed (6.6 km2), Rocky Mountain National Park, Colorado. Geostatistics and classical statistics were used to estimate SWE distribution across the watershed. Snow depths were spatially distributed across the watershed through kriging interpolation methods which provide unbiased estimates that have minimum variances. Snow densities were spatially modeled through regression analysis. Combining the modeled depth and density with snow-covered area (SCA) produced an estimate of the spatial distribution of SWE. The kriged estimates of snow depth explained 37-68% of the observed variance in the measured depths. Steep slopes, variably strong winds, and complex energy balance in the watershed contribute to a large degree of heterogeneity in snow depth.

URLsites/westernsnowconference.org/PDFs/1998Balk.pdf