Inclusion of a Subgrid Orography Drag Parameterization for Improvement of Wind Speed Prediction over a Complex Terrain Region

The overprediction of near surface wind speed is frequently reported in mesoscale meteorological modeling over complex terrains. The poor representation of sub-grid orography drag has been attributed to the model discrepancy. Though many studies have tried to reduce the wind speed bias through a drag approach or roughness length approach, the problem still remains in meteorological simulations over complex terrains. This study implements a new sub-grid orographic drag parameterization into the Weather Research and Forecast (WRF) model (version 3.9.1) to improve the wind speed prediction over the South Korea region and local circulations over the Seoul metropolitan area. The newly implemented parameterization is based on a roughness length approach obtained from the large-eddy simulations. In order to evaluate the new ingredient and existing option, a series of the WRF simulations has been conducted for one year period in 2016 over the Korean peninsula with a horizontal grid resolution of 3 km. The simulation results are compared against surface meteorological measurements over South Korea. The results showed that the new parameterization improved significantly the model performance by showing the mean bias error of 0.5 m s^-1 compared to approximately 1.5 m s^-1 in the reference simulation and 1.2 m s^-1 in the simulation made with an existing option. The improvement was more apparent in the nighttime than the daytime. The enhanced surface roughness over the domain also improved sea breeze prediction over the Seoul metropolitan area in terms of occurrence timing and inland penetration length. The results of this study suggest that the new implemented parameterization will be useful in predicting local circulations and associated air quality.