An investigation of a simulated low-level jet produced by different PBL schemes in the WRF-ARW as verified against tower data

The prediction of winds between 50 and 150 m above ground level (AGL) is crucial for the design, operation, and maintenance of wind farms. The lack of standard observations in this layer makes the development and verification of numerical weather prediction models difficult and has contributed to a lack of progress in planetary boundary layer (PBL) scheme development towards renewable energy applications. Furthermore, the design of operational data assimilation systems, which have a strong potential benefit towards the forecasting of low-level winds, is also hindered without a dense and accessible quantity of wind tower data to ingest.

A new high-frequency data assimilation system, the Rapid Refresh (RR), is planned for operational implementation at the National Center for Environmental Prediction (NCEP) late in 2010, replacing the current operational Rapid Update Cycle (RUC). Further, a much higher- resolution nest (HRRR) is likely to become available through NCEP within 3-5 years. Both are strong candidates for use in renewable energy applications, either directly or as a source of initial and lateral boundary conditions for higher resolution WRF forecasts from private vendors. The RR and HRRR forecast model component is the Advanced Research version of the Weather Research and Forecasting model (WRF-ARW). As part of our development and testing, new PBL schemes are being examined and compared to current operational schemes. Model errors within the PBL, immediately above the surface, are relatively unknown and need to be assessed in order to further system development.

Acquisition of data from over 90 towers throughout the Great Plains region allows an investigation of the performance of different PBL schemes within the WRF-ARW in simulating low-level jet (LLJ) cases occurring 18-20 August 2007. This study evaluates the performance of the WRF-ARW, with focus on the vertical structure of the winds and turbulent kinetic energy (TKE) within a LLJ. Model simulations are compared with wind tower measurements throughout the region of the LLJ. Notable differences are observed for different PBL schemes, such as LLJ maximum wind speeds, vertical wind shear in the lowest 100 m of the PBL, and predicted TKE. Sensitivity experiments are performed with enhanced vertical resolution and modifications of surface layer mixing lengths. Results are shown for LLJ onset, mature phase, and decay.