An Analysis of Snow Processes Within a Regional Climate Model

The Penn Sate-National Center for Atmospheric Research fifth-generation Mesoscale Model (MM5) driven by a 6-hourly reanalysis dataset from the National Centers for Environmental Prediction has been used to study the impact of snowpack on climate variability and the mechanisms of snowmelt over the Sierra Nevada region. The analyses of a one way nested 48km to 12km model run during the 1998 snowmelt season (April - June) showed that the underestimated snowpack resulted when there was stronger precipitation and higher temperature. An observed daily snowpack dataset collected from the automated Snowpack Telemetry system was assimilated in the model to improve its performance. The results showed that the assimilation processes greatly reduced the warm bias because the energy used to increase the temperature in the original model run was consumed by snowmelt. The cooled surface led to a more stable simulated atmospheric structure, reduced the intensity of spring storms, and therefore, suppressed the exaggerated precipitation. In the meantime, both atmospheric and land-surface conditions affected the snowmelt rate in the region studied. The faster snowmelt positively correlated with high pressure, which resulted in stronger downward solar radiation due to fewer clouds and brought warm air from the tropical Pacific Ocean to the region. The variation of vegetation fraction also strongly affected the snowmelt because it modified surface albedo and changed the radiative forcings on the snow surface. However, the exaggerated snowmelt was produced by the land-surface model in the version of MM5 used in this study because of the lower simulated surface albedo and the neglect of the effects of vegetation fraction on snowmelt. Thus, a realistic physically based land-surface scheme in a meso-scale model is crucial to predictions of snowmelt, which is essential to the climate variability and water resources in the Sierra Nevada region.