J3.11
Fractional snow cover in the Colorado River and Rio Grande basins, 1995–2002
Roger C. Bales, University of Arizona, Tucson, AZ; and B. Imam, D. Lampkin, S. R. Helfrich, and S. R. Fassnacht
Estimating snowcover properties at a basin scale, particularly snow water equivalent (SWE), remains a challenge. We have used a combination of remote sensing and ground-based data to estimate SWE for portions of the Southwestern U.S. for an eight-year period. Two immediate applications for this product were: i) to provide snowpack information for evaluating hydrologic models of snowmelt runoff and other components of the water balance, and ii) to evaluate the seasonal and interannual variability in snowcover patterns. Snow covered area (SCA) maps with a 1-km2 grid were developed from AVHRR scenes using a three-part cloud masking procedure and spectral unmixing algorithm. Using this approach fractional SCA in each pixel was estimates, representing a potentially significant improvement from traditional binary (snow/no-snow) mapping. A 1-km2 SWE product was developed for the same area using interpolation of ground-based SNOTEL data, followed by masking with the SCA scenes. In this way the interpolated SWE maps were adjusted on a pixel-by-pixel basis for the fraction of area actually snow covered. This fractional product gives significantly different snow coverage than do binary products, resulting in 20-50% differences in basin-wide SWE estimates. Even larger differences result in comparing interpolated SWE from SNOTEL with versus without masking using the SCA images. Areas with persistent snowcover are relatively reproducible from year to year, and correspond to higher elevations. However, the annual maximum snow extent, or area with any snowcover during the year, exhibited significant interannual variability, and was not well correlated with maximum SWE. While the current SCA and combined SCA-SWE product are a clear step toward improved spatial snow estimates, there are several areas for possible future improvement, including: i) using vegetation information to improve snow mapping in forested areas, ii) developing more-representative ground-based measurements and iii) using data with more spectral properties, e.g. MODIS, to improve separating snow from other landcover. The current products are available on a set of CD’s (see resac.hwr.arizona.edu).
Joint Session 3, Instrumentation and Remote Sensing to Observe Water in all its Phases (Joint with the Symposium on Observing and Understanding the Variability of Water in Weather and Climate and the 17th Conference on Hydrology)
Tuesday, 11 February 2003, 8:30 AM-5:30 PM
Previous paper Next paper