The Wind Forecast Improvement Prohect 2 (WFIP2): Overview and Preliminary Model Improvements

Abstract

The Wind Forecast Improvement Project 2 (WFIP2) is a public-private partnership effort led by the U.S. Department of Energy (DOE) and National Oceanic and Atmospheric Administration (NOAA). The WFIP 2 aims to increase understanding of atmospheric processes that affect wind power forecasts in complex terrain and to make corresponding improvements to the skill of foundational weather forecast models, which are used as inputs to wind power forecasts. The project team includes Vaisala, who is developing decision-support tools to provide uncertainty information to industry agents and operators. This presentation will provide an overview of the project. The project includes a field campaign in the Columbia River Gorge, an area of complex terrain with some 5 GW of installed wind capacity. The field campaign was conducted October 2015 through March 2017, during which period many vertical profiles from wind profiling radars, sodars, and lidars were collected. Instruments were deployed in a nested configuration to investigate phenomena across a range of spatial scales. This data set is available for public use from the Department of Energy’s Data Archive and Portal. The NOAA hourly updating, 13-km Rapid Refresh (RAP) and the 3-km High Resolution Rapid Refresh (HRRR) numerical weather prediction models, and aspects of these models that have been targeted for improvements, will be discussed. These areas include the boundary layer scheme’s treatment of local and non-local vertical mixing, the representation of clouds, and a wind farm parameterization. Special effort has been made to introduce scale-aware adaptive physics, which can be applied to any model resolution. This presentation will discuss preliminary results including changes in RAP and HRRR forecast skill for turbine-height wind speeds and direction under specific weather regimes and as a function of diurnal cycle. Effects of mountain wakes, marine pushes, cold pools, and gravity waves are among the phenomena targeted by model verification and improvement efforts.