Estimating SWE distribution with a combination of ground-based radar measurements, modeling and remote sensing

English Abstract:

Estimating the spatial distribution of snow water equivalent is often complicated by large spatial variability due to the effects of wind, vegetation, topography, and microclimate. In order to accurately capture important features of the snow distribution at local scale, measurements and models must be performed at resolutions that are often not practical over hydrologically significant scales. However, with recent increases in measurement technology and computational power, spatially distributed estimates of SWE at high resolution from both measurements and models are becoming feasible at the basin scale. We compare SWE estimates using traditional techniques (depth, core measurements) with estimates using ground-based radar, airborne LiDAR, and iSnobal model results for the Reynolds Creek Experimental Watershed, near peak SWE in March 2009. The combination of coincident high-resolution ground-based measurements, remote sensing, and modeling allows accurate quantification of the uncertainties in each technique, and this unique dataset is used to develop efficient data fusion strategies for accurate spatial estimates of SWE with quantified uncertainty.