Wind lidar profile measurements in the coastal boundary layer: comparison with WRF modelling

We use measurements from a pulsed wind lidar to study the wind speed profile in the planetary boundary layer (PBL) up to 600 m above the surface at a coastal site. Due to the high availability and quality of wind lidar data and the high vertical range of the measurements, it is possible to study the sensitivity of PBL schemes of mesoscale models to both lower and upper boundary conditions. We therefore run the mesoscale weather research and forecasting (WRF) model using two different roughness descriptions, two different synoptic forcings and two different PBL schemes at two vertical resolutions. When WRF is compared to the wind lidar data combined with measurements from a tall meteorological mast, it is found that the model surface layer fluxes are largely overestimated and that the vertical wind speed profile does not have enough shear in the lower part of the PBL, partly as a consequence of the smooth-to-rough transition at the coastline. When using a more representative
roughness than the default, the biases in the surface friction velocity and heat flux are reduced and the wind speed is slightly improved. Both PBL schemes show too much mixing during stable conditions and an underestimation in the amount of observed low level jet. The wind speed predicted by WRF does not improve when a higher resolution is used. Therefore, both the inhomogeneous (westerly) and homogeneous (easterly) flow contribute to a large negative bias in the mean wind speed profile at heights between 100 and 200 m.