Impact of Cloud Microphysics Schemes and Boundary Conditions on Modeled Snowpack in the Central Idaho Rocky Mountains, USA

Abstract Hydrologic and land surface models require spatiotemporally complete and accurate hydrometeorological forcings. In mountainous regions, hydrometeorological forcings are often obtained as the output of coupled land‐atmosphere models, like the Weather Research and Forecasting (WRF) model, configured to run at spatial scales that permit orographic convection (e.g., 4 km). Models like WRF, however, require physical parameterizations, the selection of which significantly influences model predictions of precipitation, temperature, and radiant fluxes used as input to hydrologic and land surface models. Here we investigate the impact of two critical aspects of WRF configurations, namely the selection of the cloud microphysics parameterization and lateral boundary conditions, on modeled hydrometeorological forcings and associated snow conditions in a mountainous region of the western United States. We conducted eight experiments with WRF configured at convection‐permitting scales using two reanalysis data sets as lateral boundary conditions (ERA5 and CFSRv2) and four alternative cloud microphysics schemes. These experiments reveal that the choice of lateral boundary conditions and cloud microphysics schemes imposes substantial variability in simulated surface hydrometeorological conditions, with precipitation and radiation emerging as key factors. When compared to the accumulated precipitation average over the Snow Telemetry (SNOTEL) stations, the relative bias in precipitation across experiments ranges from −18.15% to +15.48%. These biases impact the land surface model, leading to discrepancies in modeled snow. The relative bias in snow water equivalent compared to the SNOTEL average ranges from −39.84% to 10.72%, while for snow depth, it falls between −37.72% and 0.32%. Further comparisons of annual snow fraction and snow disappearance date (SDD) with Moderate Resolution Imaging Spectroradiometer (MODIS) reveal a consistent overestimation at higher elevations, with snow persisting beyond the MODIS SDD. These findings highlight the critical role of model configuration in improving hydrometeorological forcings and enhancing hydrologic predictions in complex terrain.