Evaluation of NWS Probabilistic Snowfall Graphics and NBM Snow Percentiles in the Mid-Atlantic Region

As the National Weather Service (NWS) continues to develop products geared toward the needs of customers and core partners, verification of these products to assess forecast quality is needed. Probabilistic storm total snowfall graphics are one of many products that provide a range of snowfall possibilities that better communicates forecast uncertainties and enhances the quality of Impact-Based Decision Support Services (IDSS) during winter weather events. These graphics portray the reasonable best/worst case scenarios for core partners, which is commonly asked during winter briefings.

The National Blend of Models (NBM) is a suite of calibrated forecast guidance based on a blend of numerical weather prediction model data and post-processed model guidance. In recent upgrades to the NBM over the past few years, the NBM has started to provide snow percentile data (i.e. 10th, 25th, 50th, 75th, 90th, etc.) similar to what is provided in the current NWS Prob Snow product. As the NBM continues to improve in calibration over the next few years, it could become the basis for future probabilistic snow forecasting.

This project evaluates the probabilistic storm total snowfall graphics and NBM snow percentiles to determine how often different percentiles were exceeded in relation to the synoptic setup in the Mid-Atlantic region. The results of this project will be shared with forecasters and developers of the probabilistic storm total snowfall graphics, with the hopes of improving the recurrence intervals of exceedance to ensure they fall in line with the expected exceedance frequency, thereby improving the quality of forecasts.