Comparison of Northern Hemisphere Snow Cover Data Sets

Abstract

Four Northern Hemisphere snow cover data sets are compared on a weekly basis for the 25-month period, July 1981 through July 1983. The data sets are the NOAA/NESDIS Weekly Snow and Ice Chart, the Composite Minimum Brightness (CMB) Chart, the U.S. Weekly Weather and Crop Bulletin (data only for North America), and Air Form data. The NOAA/NESDIS chart is produced through the use of photo-interpretation of visible satellite imagery and ground observations. The U.S. Crop Bulletin is also done manually, using only ground observations. The CMB chart and the Air Force data are both produced using automated processes, the first by way of visible satellite imagery and the second by way of ground observations, climatology, satellite observations and persistence. Since the NOAA/NESDIS chart is the only standard and complete data set dating back to the mid 1960s, it is used as the basis for the study. The main emphasis of this paper is a comparison of the CMB and the NOAA/NESDIS chart. The CMB frequently overestimated snow cover, especially the southward extent of the main snow boundary and area far from the snow boundary which were not present on the NOAA/NESDIS chart. On numerous occasions, the outline of mountain ranges was either distorted or totally missed by the CMB. The CMB also underestimated snow cover, especially in densely forested areas. Other regions of underestimation by the CMB can be attributed to the bias factor of the NOAA/NESDIS chart. (The NOAA/NESDIS chart uses the latest snow cover information while the CMB is composited over a week.) The U.S. Crop Bulletin agreed fairly well with the NOAA/NESDIS chart cast of the Rockies, but often differed to the west. The Air Force data set, an undocumented operational product, differed quite a bit from the NOAA/NESDIS chart.