This winter storm, as it progressed across the southern and central Rocky Mountains during 16-20 March 2003, was characterized by tremendous forcing both synoptically and orographically. In addition to the 1-in-50-year nature of the snowfall, the possibly 1-in-100 year total volume of precipitation, and the most economic damage in Colorado history of any winter storm, three significant mesoscale features were particularly unusual, difficult to predict, and worthy of further investigation.

First, a synoptic- and meso-alpha-scale, high theta-e, deep inflow developed and persisted for an unusually long period of time (more than 36 hours). This feature, at times only 100-200 km in width, extended directly from the Gulf of Mexico to central Colorado, with an approximate fetch length of 1500 km. In combination with the orographic-forcing of the Front Range of the Rocky Mountains, this inflow produced a barrier jet that was very strong and critical to the development of the snow distribution.

Second, a surprising lee-side precipitation maximum was generated by this storm. Typical so-called upslope snowstorms generate snowfall that is quite heavy on the upwind side of the barrier, but only 10-20% of the upwind-side snow depth falls downwind. In this case, while 1.0 - 2.2 m of snowfall were reported upwind, approx. 2.0 m fell in some locations on the lee side, within ~30 km west of the top of the barrier (the Continental Divide). In this case, we investigate the role of orographically-forced gravity waves, the unusually deep moisture profile and easterly flow that extended to above 150 hPa in lee-side snowfall.

Finally, on the meso-gamma scale, a local thermodynamic effect, potentially related to terrain-induced flow, created a surprising snow minimum in a location along the mountain-plains interface whose elevation alone suggests that orographic enhancement of snowfall should have been significant. This snow "gap" was characterized by temperatures that remained 1-2 C warmer than surrounding areas during the storm, resulting in snow depths (as opposed to liquid precipitation) of less than 0.1 m during the storm. Within 10-20 km of this location, however, snow depth amounts were 0.5-1.0 m. This snow depth gap was clearly evident on satellite images after the storm as a triangular region measuring ~200 square km, oriented east-west. Other snow minima will be discussed as time permits.

In this work we develop hypotheses for the generation of these three features through observational and model analyses of thermodynamics, microphysics, and dynamical mechanisms.