Investigation of boundary-layer wind predictions during nocturnal low-level jet events using the Weather Research and Forecasting model

Abstract

The accuracy of boundary-layer wind profiles occurring during nocturnal low-level jet (LLJ) events, and their sensitivities to variations of user-specifiable model configuration parameters within the Weather Research and Forecasting model, was investigated. Simulations were compared against data from a wind-profiling lidar, deployed to the Northern Great Plains during the U.S. Department of Energy-supported Weather Forecast Improvement Project. Two periods during the autumn of 2011 featuring LLJs of similar magnitudes and durations occurring during several consecutive nights were selected for analysis. Simulated wind speed and direction at 80 and 180 m above the surface, the former a typical wind turbine hub height, bulk vertical gradients between 40 and 120 m, a typical rotor span, and the maximum wind speeds occurring at 80 and 180 m, and their times of occurrence, were compared with the observations. Sensitivities of these parameters to the horizontal and vertical grid spacing, planetary boundary layer and land surface model physics options, and atmospheric forcing dataset, were assessed using ensembles encompassing changes of each of these configuration parameters. Each simulation captured the diurnal cycle of wind speed and stratification, producing LLJs during each overnight period; however, large discrepancies in relation to the observations were frequently observed, with each ensemble producing a wide range of distributions, reflecting highly variable representations of stratification during the weakly stable overnight conditions. Root mean square error and bias values computed over the LLJ cycle (late evening through the following morning) revealed that, while some configurations performed better or worse in different aspects and at different times, none exhibited definitively superior performance. The considerable root mean square error and bias values, even among the ‘best’ performing simulations, underscore the need for improved simulation capabilities for the prediction of near-surface winds during LLJ conditions. Copyright © 2015 John Wiley & Sons, Ltd.