Downscaled Probabilistic Snowfall Forecast Products for the Western United States

Snowfall amount forecasting typically involves the use of a quantitative precipitation forecast (QPF), determination of precipitation type, and application of a snow-to-liquid ratio (SLR). Significant spatiotemporal variability in SLR is known to exist during individual storms (i.e., from 4:1 to 40:1 or more in rare cases). Snowfall forecasts based on overly simplistic SLR estimates such as the climatological mean or near-surface temperature may yield large errors as a result. Forecast methods well-adapted for complex terrain are desirable, particularly across the western contiguous United States (CONUS).

In this poster we present a suite of high-resolution (800-m grid spacing) snowfall forecast products derived from lower-resolution operational modeling systems including ensembles. These products are publicly available at weather.utah.edu and include downscaled QPF, downscaled snowfall accumulation, snow-to-liquid ratio, and the height of the 0.5°C wet bulb temperature (a proxy for snow level). In addition to spatial visualizations of mean and upper-percentile QPF and snowfall amounts, plumes and violin plots are available to aid in the interpretation of probability and uncertainty in precipitation rate, SLR, and highlight the potential for extreme outcomes.

These products aim to enhance the quality of snowfall forecasts available to a wide variety of user groups. Avalanche professionals and private-sector forecasters across the western CONUS frequently use these products and benefit from the spatial detail and visualizations of probabilistic information. These products are experimental and currently under development with significant upgrades to the snow-to-liquid algorithm slated for prior to the 2020-2021 winter precipitation season.