The Sensitivity of the Predicted Hub-Height Winds to Land Surface and Boundary Layer Representation

Near surface cold air pools often persist for long periods of time during winter and spring over basins and valleys. The stable stratification due to the presence of surface inversion layer can impact the prediction of hub-height wind speed. We evaluate improved versions of High Resolution Rapid Refresh developed during the Wind Forecast Improvement Project 2 (WFIP2) and test the sensitivity of the forecasted low-level winds, particularly rotor-level winds, thermal stratification, and surface energy budget over the Pacific Northwest region to the use of different land surface model (LSM) as well as some planetary boundary layer (PBL) parameters identified from our previous uncertainty quantification work for the same region.  One winter and one spring cold pool case studies are conducted: the winter case is characterized with strong winds above the surface inversion layer that mixes out near noon, and in the spring case the cold pool dissipates after sunrise without the presence of strong synoptic forcings, such as upper level ridge. Both cases are accompanied with an increase of hub-height wind speed associated with the ‘mix-out’ of the cold air pools. The WFIP2 High Resolution Rapid Refresh (HRRR) simulates both events well in regards to the vertical and temporal variations of low-level wind and temperature at some key WFIP2 observational sites. Compared to the Common Land Model (CLM), the Rapid Update Cycle (RUC) LSM tends to simulate slightly stronger cold pool strength and stronger mixing down of warm air.  Improving the representation of Plant Function Type in CLM shows minor changes in the predicted temperature distribution and hub height winds, and the modification of PBL parameters also induced some variation in the predicted winds, particularly, decreasing the Prandtl number leads to minor increase in mix-down of warm and high momentum air. The sensitivity of the low-level wind prediction to vertical grid spacings and the representation of river stage will also be explored.