The impact of grid and spectral nudging on the variance of the near-surface wind speed

Abstract

Grid and spectral nudging are effective ways of preventing drift from large-scale weather patterns in regional climate models. However, the effect of nudging on the wind speed variance is unclear. In this study, the impact of grid and spectral nudging on near-surface and upper boundary layer wind variance in the Weather Research and Forecasting Model is analyzed. Simulations are run on nested domains with horizontal grid spacing of 15 and 5 km over the Baltic Sea region. For the 15-km domain, 36-h simulations initialized each day are compared with 11-day simulations with either grid or spectral nudging at and above 1150 m above ground level (AGL). Nested 5-km simulations are not nudged directly but inherit boundary conditions from the 15-km experiments. Spatial and temporal spectra show that grid nudging causes smoothing of the wind in the 15-km domain at all wavenumbers, both at 1150 m AGL and near the surface where nudging is not applied directly, while spectral nudging mainly affects longer wavenumbers. Maps of mesoscale variance show spatial smoothing for both grid and spectral nudging, although the effect is less pronounced for spectral nudging. On the inner, 5-km domain, an indirect smoothing impact of nudging is seen up to 200 km inward from the dominant inflow boundary at 1150 m AGL, but there is minimal smoothing from the nudging near the surface, indicating that nudging an outer domain is an appropriate configuration for wind-resource modeling.