Regional Climate Simulations of Cold Season Precipitation Over the Great Basin

Cold season simulations and predictions of precipitation in the Great Basin are crucial for water supply outlooks and hydro-climate assessments, as snow accumulations provide a natural reservoir that is released in the spring and summer as runoff. The utility of regional climate models (RCMs) to simulate cold season precipitation has been studied more rigorously in other regions of the western United States, such as the Sierra Nevada and Cascade Ranges, but little work has focused on the arid to semi-arid interior Great Basin. In this study we use the Weather Research and Forecasting (WRF) model in climate mode forced with NCEP/NCAR reanalysis data to simulate precipitation during two contrasting cold seasons: (1) 2010-2011 (wet), and (2) 2011-2012 (dry). A triple nested domain (12-km, 4-km, and 1.33-km) setup is used to examine the effects of grid spacing on precipitation spatial distributions. With computer science technology advancing rapidly there is a growing need to assess super high-resolution RCM configurations, which are likely to become more commonplace in the future. For each cold season, four different microphysics (MP) parameterizations are tested (two single-moment and two double-moment). Prediction of mass variable number concentrations through double-moments schemes are more costly than single-moment, and there is a great need to test if a more complex MP scheme is necessary to identify high-resolution precipitation features associated with orography, such as maximum accumulation zones with regards to aspect (west-east) and elevation. In this study, we show the sensitivity of MP and grid spacing configuration over the Great Basin by comparing to in situ observations (valley and mountain locations) from a number of networks throughout the region. Altitudinal gradients of simulated precipitation and temperature were further compared to three statistically derived gridded climate products which are commonly used when observations are scarce: (1) PRISM (4-km and 800-m), (2) Daymet (1-km), and (3) a NLDAS/PRISM hybrid (4-km). A striking result highlights that the more complex double-moments schemes tend to under predict precipitation accumulations, particularly in high elevation zones.