This research will compare a high-impact, false alarm snow forecast with a more accurate prediction that occurred 10 days later within a similar weather pattern. On Friday, 13 April 2007, eastern Colorado and most of the central Rocky Mountain Front Range was preparing for the arrival of a substantial spring snowstorm. As the storm developed, the track of the low-pressure system moved considerably farther to the south than predicted just 24 hours earlier. A similar storm developed along the same track on 23 April 2007, accompanied by mountain snow and severe thunderstorms on the eastern Colorado plains. In this latter case, forecasters expressed high confidence and correctly forecasted the evolution of the storm and its varied weather.

In this work, differences and similarities between the two events are identified and reported from meteorological, operational forecasting, and media perspectives. The influence of forecast confidence is key aspect common to all perspectives. One focus of this research examines differences in National Weather Service Area Forecast Discussions (AFDs) from offices in New Mexico, Colorado, Kansas, and Wyoming. Comparing the AFDs helps characterize the general confidence with which predictions are made regarding storm track and timing as the events developed. The concept of forecast continuity appears relevant to this case study since the AFDs reveal notions of when and why forecasts were changed. Another focus area in this work is the role that media played in portraying this significant weather event. The media's role in forecasting is relevant due to this region's heightened sensitivity to social and economic impacts of snow following one of the largest blizzards on record during December 2006. Forecast accuracy was already questioned among the communities included in our case study. This was exacerbated by extensive media promotion of a significant “Friday the 13th” upslope snowstorm. A final component of this work involves sensitivity simulations using WRF model simulations. An ensemble of WRF hindcasts is launched using available analysis products from the time prior to the storm at which AFDs began to express strong uncertainties in storm evolution. It is intended that results from this work will help increase the forecasting accuracy and presentation of similar Colorado cyclones in the future.