Generation of Turbulent Inflow Boundary Conditions for WRF LES

In a coupled mesoscale-microscale simulation using the Weather Research and Forecasting (WRF) model, turbulence may not fully develop near the inflow boundary due to a sudden jump in the spatio-temporal scale between the two domains. The development of turbulence is especially challenging in stable conditions as the negative buoyancy suppresses the generation of turbulence. The present work adds turbulence energy associated with spectral energies derived from either velocity fields from precursor simulations with finer grid spacing or observations to the inflow lateral boundary obtained for the coarser grid. In this study, the spectral content of higher frequency range is extracted from either high-resolution simulations or measurements from tower-mounted sonic anemometers as higher spatial modes using the proper orthogonal decomposition (POD) method. This method allows correlated components of wind velocity to be added in the inflow boundary condition in a way that is physically consistent with observed PODs. As a first step, temporally varying, but horizontally homogeneous higher spatial mode content is added at the inflow boundary. The initial results obtained using the highly resolved simulated data for the first 400 m show an area of increased turbulence in the flow field for the convective conditions compared to standard simulations. As a second step, a stochastic method that uses the simulated or observed data to estimate the turbulent inflow condition will be used to generate the inflow lateral boundary condition rich in higher spectral energy both in space and time.