An assessment of the differences between three satellite snow cover mapping techniques

Abstract The National Operational Hydrologic Remote Sensing Center (NOHRSC) of the National Oceanic and Atmospheric Administration's (NOAA's) National Weather Service (NWS) provides daily satellite‐derived snow cover maps to support the NWS Hydrologic Services Program covering the coterminous USA and Alaska. This study compared the NOHRSC snow cover maps with new automated snow cover maps produced by the National Environmental Satellite, Data, and Information Service (NESDIS) and the snow cover maps created from the Moderate Resolution Imaging Spectroradiometer (MODIS) imagery. The purpose of this paper is to demonstrate and account for the differences that occur between the three different snow cover mapping techniques. Because each of these snow cover products uses data from different sensors at different resolutions, the data were degraded to the coarsest relevant resolution. In both comparisons, forest canopy density was examined as a possible explanatory factor to account for those differences. NOHRSC snow cover maps were compared with NESDIS snow cover maps for 32 different dates from November 2000 to February 2001. NOHRSC snow cover maps were also compared with MODIS snow cover maps in the Pacific Northwest and the Great Plains for 18 days and 21 days, respectively, between March and June 2001. In the first comparison, where the NOHRSC product (∼1 km) was degraded to match the resolution of the NESDIS data (∼5 km), the two products showed an average agreement of 96%. Forest canopy density data provided only weak explanation for the differences between the NOHRSC and the NESDIS snow cover maps. In the second comparison, where the MODIS product (∼500 m) was degraded to match the resolution of the NOHRSC product for two sample areas, the agreement was 94% in the study area in the Pacific Northwest, and 95% in the study area in the Great Plains. Published in 2002 by John Wiley & Sons, Ltd.