The objective of this paper is to interpret microwave scattering signatures over snow cover as observed by the Advanced Microwave Sounding Unit (AMSU) for the retrievals of snow water equivalent and snow depth. A case study involving seasonal snow cover over the U.S. Great Plains was analyzed in detail. Area-wide analysis of the relationship between snow depth and the AMSU scattering signatures in the 23--150 GHz window region showed weak correlation, deteriorated by the dependence of these signatures on snow metamorphism. The lower frequency scattering index, computed as the difference in the brightness temperature between 23 and 31 GHz channels, was low and insensitive to fresh snow predominant in December, but increased later in the season, and thus was more sensitive to snow depth for older snow cover. However, this seasonal increase in microwave scattering was observed for every snow depth range, suggesting a strong dependence on snow metamorphism. In contrast, the 89 GHz scattering index responded to relatively shallow snow cover in December, but was less sensitive to snow depth variability later in the season. A snow hydrology model was applied at specific locations to estimate snow water equivalent (other than snow depth) for comparisons with the AMSU measurements. Overall, the lower frequency index was the best predictor of snow water equivalent and snow depth. However, correlation was higher for snow density and snow water equivalent. This was attributed to the response of this scattering index to the grain size evolution with time, which correlated better with the snow density and water equivalent changes in the snow cover than snow depth. Correlation between the snow water equivalent and the lower frequency index for fresh snow cover was significantly improved by switching to the higher frequency index at 89 GHz as predictor when the lower frequency index at 31 GHz was less than the 5 K threshold. Correlation further improved for fresh snow cover associated with a positive 150 GHz scattering index. These results suggest that the utility of AMSU measurements for the retrievals of snow depth and snow water equivalent could be better optimized by proper selection of the scattering thresholds that minimize the variability of important snow parameters such as grain size.