Assessment of NWP forecast models in simulating offshore winds through the lower boundary layer by measurements from a ship-based scanning Doppler lidar
Evaluation of model skill in predicting winds over the ocean was performed by comparing retrospective runs of numerical weather prediction (NWP) forecast models to ship-borne Doppler lidar measurements in the Gulf of Maine – a potential region for U.S. coastal wind farm development. Deployed on board the NOAA research ship Ronald H. Brown during a 2004 field campaign, the High-Resolution Doppler Lidar (HRDL) provided accurate, motion compensated wind measurements from the water surface up through several hundred meters of the atmospheric marine boundary layer (MABL). The quality and resolution of the HRDL data allow detailed analysis of wind flow at heights within the rotor layer of modern wind turbines and data on other critical variables to be obtained, such as wind speed and direction shear, turbulence, low-level jet properties, ramp events, and many other wind energy relevant aspects of the flow. Here we will focus on quantitative validation of NWP models wind forecasts within the lower MABL by comparing to HRDL measurements.
This study presents validation of two modeling systems rerun in special configurations for these 2004 cases; the hourly-updated Rapid Refresh (RAP) system and a special, hourly-updated version of the North America Mesoscale forecast system, known as NAM Rapid Refresh (NAMRR). These models were run at both normal resolution (RAP, 13 km; NAMRR, 12 km) and high-resolution versions: the High-Resolution Rapid Refresh (HRRR, 3 km) and the NAMRR-CONUS-nest (4 km). Each model was run twice: with (experimental runs) and without (control runs) assimilation of data from 11 Wind Profiling Radars located along the U.S. East Coast. The impact of the additional assimilation of the 11 profilers was estimated by comparing HRDL data to modeled winds from both runs.
The results obtained demonstrate the importance of high-resolution lidar measurements to validate NWP models, and to better understand what atmospheric conditions may impact the accuracy of wind forecasts in the marine atmospheric boundary layer. Results of this research will also provide a first guess as to the uncertainties of wind resource assessment using NWP models in one of the U.S. offshore areas projected for wind plant development.