Synoptic climatology of Rocky Mountain snowfall in the second phase of the Atmospheric Model Intercomparison Project

As rapid population growth continues to impact urban areas of the western United States, seasonal snowfall in the Rocky Mountains is studied with increasing frequency as the seasonal snowpack provides much needed runoff to support the high water demand in the arid and semi-arid regions common to western North America. In this paper, the synoptic climatology of snowfall (snow water equivalent-SWE and snow covered area-SCA) in the Rocky Mountains during the period 1979-1995 is examined with the use of a suite of models from the second phase of the Atmospheric Model Intercomparison Project (AMIP). The global circulation models (GCMs) used for this study vary in their resolution, topography and model characteristics. Model performance in simulating Rocky Mountain snowfall is stratified according to model properties to evaluate the potential impact of varying model dynamics. The models' ability to capture the interannual variability and timing of snowfall in the mountains as well as the synoptic conditions favorable to snowfall in this region is evaluated. The observed rate and timing of snowmelt in this region, being of great importance to hydrologists and water resource managers is compared to the modelled values.

Snowfall in the mountains of western North America is favored with a 20-30° westward shift of the mean atmospheric ridge located over the western continent. The models' ability to reconstruct this pattern during high snow years is evaluated. Similarly, the models' ability to reconstruct patterns of unusually low snowfall in the mountains of western North America is also examined. An analysis of sea-surface temperature (SST) patterns and the association with snowfall in this region is discussed. Standard deviations of snowfall in the observed record are obtained from SNOTEL and NWS stations and are compared to the modelled variability in the 17-year record. The quality of the modelled variability is carefully examined as the models' ability to capture variability is important to evaluating the overall model dynamics and performance.