A devastating snowstorm affected much of the central Rockies and adjacent plains during the period 16-20 March 2003. The greatest impact was over the Front Range of Colorado and southern Wyoming where heavy snow accumulations crippled travel for several days. Significant snowfall also occurred over the Grand Junction National Weather Forecast Office (GJT) county warning area (CWA) which includes western Colorado and eastern Utah. Surprisingly, a great deal of the precipitation which fell over the GJT CWA occurred in fairly deep easterly flow which is typically not conducive for widespread heavy snowfall for this region. Total snow accumulations up to several feet fell over the area.

A closer investigation of the storm showed that widespread precipitation did occur over much of the CWA. However, the enhanced snowfall totals occurred, primarily, in a strong easterly component on 17-18 March along the Continental Divide of Colorado and as far west as the eastern Uintah mountains of northeastern Utah. NCEP model guidance did quite well with storm evolution and strength but did not handle the fine scale detail of the precipitation fields over this complex topographic region. To exasperate the forecast process, WSR-88D beam blockage over the higher terrain made precipitation detection and strength quite difficult to ascertain over mountain locations. To help resolve fine scale features across the region, the GJT forecasters had a "local mesoscale model" available for operational use, the Regional Atmospheric Modeling System (RAMS). RAMS was part of a research effort between the NWS and Colorado State University (CSU) through the Cooperative Program for Operational Meteorology, Education and Training (COMET). RAMS, which was developed at CSU, has been run in a prototype forecasting mode over Colorado for winter seasons since 1991. The RAMS forecasts did a fairly good job detailing the spatial and temporal distribution of precipitation, however, it tended to over-predict precipitation amounts, especially near the Continental Divide.

This study will examine the primary forcing mechanisms which focused heavy snowfall across portions of the GJT CWA. It will detail the benefits of augmenting the NCEP model suite with a local mesoscale model especially in complex terrain. Further examination in a research mode with 1-3 km grid spacing version of the workstation Eta, MM5 and/or RAMS will also be conducted.