Evaluation of wintertime precipitation estimates and forecasts in the Mountains of Colorado

Wintertime precipitation poses many observational and forecasting challenges, especially in the western US where mountainous terrain results in radar beam blockage and sparse distribution of in situ observations. Uncertainty in western US winter precipitation is known to be high, so much so that some studies have found model simulated precipitation to produce similar or better large-scale estimates of annual precipitation than gridded observation-based products during climatologically anomalous years. In this study we evaluate high resolution gridded precipitation estimates from multi-sensor products as well as forecasts from NOAA’s operational High Resolution Rapid Refresh (HRRR) model in the Colorado Rocky Mountains using in situ SNOTEL data for wintertime precipitation during water years 2022 and 2023. We particularly focus on the influence of forecast length, lead time, and model elevation on seasonal precipitation predictions from the HRRR. Additionally, we make use of the relatively dense network of observations deployed in the East River Watershed during the SPLASH and SAIL field campaigns to supplement the evaluation. The gridded products are found to largely underestimate cold season precipitation, with similar low biases compared to in-situ observations during both water years examined.