Improving Forecasts of Low-Level Winds Induced by Thermal Troughs in the Columbia River Basin

Many wind profiling and scanning instruments have been deployed in the Columbia River Gorge and Basin, in support the Wind Forecast Improvement Project 2 (WFIP 2), which spanned 01 October 2015 to 31 March 2017. Measurements from this instrumentation, which included over 20 profiling stations, were compared against the High-Resolution Rapid Refresh (HRRR) and an experimental very high-resolution nest (∆x = 750 m) model to diagnose systematic wind-speed biases within the turbine rotor layer, the associated errors aloft, as well as errors in temperature, radiation, and surface fluxes. This analysis focuses on hub-height wind-speed events between 4 and 12 m s^-1, where wind speed forecast errors result in large power-generation forecast errors due to the slope of the wind speed-power relationship.

The dominant forecast challenge during the warm season was found to be the wind ramps associated with westerly gap flows within the Gorge, forced by the diurnal variation of the cross-Cascade pressure gradient. Many important physical processes are concurrent within these events, such as the differential surface heating associated with the large Bowen ratio and cloud cover differences across the Cascades, the development of the convective and stable boundary layer, and the interaction of the synoptic/mesoscale flows with the complex topography. Efforts to improve the representation of these processes have been focused on the column-based turbulent mixing and subgrid-scale drag schemes, but investigating the effects of horizontal (and 3D) mixing has also helped to improve the representation of some flow features. This presentation will highlight the testing and development of various model components, showing examples of case studies and retrospective periods to illustrate the improvements. We will demonstrate that the improvements made in WFIP 2 will be extendable to other regions, complex or flat terrain. Ongoing and future challenges in RAP/HRRR physics development will be noted.