Evaluation of a cloudy cold-air pool in the Columbia River Basin in different versions of the HRRR model

Abstract. The accurate forecast of persistent orographic cold-air pools in numerical weather prediction models is essential for the optimal integration of wind energy into the electrical grid during these events. Model development efforts during the Second Wind Forecast Improvement Project (WFIP2) aimed to address the challenges also related to this. We evaluated three different versions of NOAA's High-Resolution Rapid Refresh model with two different horizontal grid spacings against in situ and remote sensing observations to investigate how developments in physical parameterizations and numerical methods targeted during WFIP2 impacted the simulation of a persistent cold-air pool in the Columbia River Basin. Differences between the different model versions were in particular visible in the simulated temperature and low-level cloud fields. The model developments led to an enhanced low-level cloud cover in the cold pool, resulting in better agreement with the observations. This removed a diurnal cycle in the near-surface temperature bias at stations throughout the basin by reducing a cold bias during the night and a warm bias during the day. However, low-level clouds did not clear sufficiently during daytime in the newest model version, which led to a warm bias near-the surface during the second night of the reforecasts and leaves room for further model developments. Besides the improvements during the persistent phase of the cold pool, the model developments also led to a better representation of its decay by slowing down its erosion.