Case Studies of Improved HRRR Low-Level Wind Forecasts from the Wind Forecast Improvement Project II

The Wind Forecast Improvement Project II (WFIP2) is a multiagency venture aimed at improving low-level wind forecasts in regions of complex terrain for the benefit of wind-energy applications. To support this objective, experimental modifications to the Mellor–Yamada–Nakanishi–Niino (MYNN) turbulence parameterization are being developed within the WRF–ARW modeling framework. A provisional version of NOAA's High-Resolution Rapid Refresh (HRRR) model, the HRRR–WFIP2, provides a platform for real-time and retrospective forecasting using experimental MYNN scheme modifications, and serves as a pathway for eventual operational implementation of improved model parameterizations developed during WFIP2.

Measurements from WFIP2 profiling instruments in the Columbia River Basin indicate a propensity of the HRRR model to prematurely erode cold, stably stratified low-level air, often contributing to significant errors in turbine-height wind speed forecasts. To help mitigate this problem, the MYNN scheme in the HRRR–WFIP2 contains experimental modifications designed to improve the representation of local mixing within the planetary boundary layer and free atmosphere. These modifications include (1) a revision of the MYNN buoyancy length scale, (2) elimination of the influence of the surface-layer length scale at heights away from the ground, and (3) a reformulation of the MYNN master-length-scale blending approach. Drawing from wind-ramp events sampled during the WFIP2 field campaign, we examine case studies that demonstrate improved wind-forecast skill stemming from these MYNN scheme modifications in stably stratified flow. Results from reforecasts in the 3-km HRRR–WFIP2 and its concurrent 750-m nest will be presented, along with long-term verification statistics.