This study compares two satellite-derived Northern Hemisphere snowcover estimates with daily climate station data from the United States and Canada. First, weekly NOAA/NESDIS snow charts, constructed from AVHRR and GOES visible wavelength images are analyzed. Second, various snowcover algorithms employing daily passive microwave data from the SSM/I are investigated. Data spanning four winter seasons from 1987/1988 through 1990/1991 are used. The analysis shows where and when each satellite method compares well with the ground-based
data, and where and when they differ. These differences are subdivided into errors of commission and errors of omission. Interpretation of these errors reveals the relative strengths and weaknesses of each satellite method in estimating continental snowcover. Using this information, a snowcover model which combines the strengths of both satellite methods is developed and tested