Predicting Snow-To-Rain Transitions Across The Western U.S.: When Is Daily Air Temperature Sufficient?

Abstract:

The phase of precipitation at the land surface is critical for determining the timing and amount of water available for hydrological and ecological systems. Natural variability in precipitation phase due to elevation, micro-climate, and storm characteristics make it a challenge to predict phase. In addition, regional warming is expected to move the snow-rain elevation higher in the future, which has the potential to alter water availability. Despite this, there are few techniques for direct observation of precipitation phase and many predictive techniques apply simple temperature thresholds (i.e. 0 degree Celsius) to determine spatiotemporal patterns. In this paper, we asked two questions: 1) what is the optimum daily temperature for predicting snow-rain transitions in the mountains of the Western U.S.? and 2) what errors in precipitation phase estimation are associated with common temperature thresholds? We use 502 Snow Telemetry (SNOTEL) stations with data from 2004 to 2014 to determine rain versus snow using a combination of precipitation, snow depth, and SWE observations. From the observations, we determined that daily maximum temperature is a better predictor of rain and snow events than average temperature. The optimum temperature varied from -2.0 to 3 C, with an average of 0.3 C across ecoregions. The Northern Basin and Northern Cascades with lower average elevations had higher temperature thresholds and the Southern Rockies with highest elevations had the lowest thresholds. Developing a relationship based on station elevation improved the RMSE by 12%, whereas using an optimum temperature developed for each station improved the RMSE by 34% on average. While using optimum temperature thresholds reduce error in prediction, they do not eliminate misclassification of rain-show transitions. These results highlight a current weakness in our ability to predict the effects of regional warming that could have uneven impacts on water and ecological resource management across the Western U.S. Future efforts should consider including additional atmospheric information to improve the fidelity of precipitation phase prediction.