Performance Assessment of Dynamic Downscaling of WRF to Simulate Convective Conditions during Sagebrush Phase 1 Tracer Experiments

Large-Eddy Simulations (LES) corresponding to four convective intensive observation periods of Sagebrush Phase 1 tracer experiment were conducted with realistic boundary conditions using Weather Research and Forecast model (WRF). Multiple nested domains were used to dynamically downscale the conditions from domain with grid size of 24 km to local scales with grid size of 150 m. Sensitivity analysis of mesoscale model was conducted using three boundary layer, three surface layer and two micro-physics schemes. Model performance was evaluated by comparing the surface meteorological variables and boundary layer height from the mesoscale runs and observed values during tracer experiment. Output from mesoscale simulations was used to drive the LES domains. Effect of vertical resolution and sub-grid scale parameterizations were studied by comparing the wind speed and direction profiles along with turbulent kinetic energy at two different heights. Atmospheric stability estimated using the Richardson number and shear exponent evaluated between 8- and 60-m levels was found to vary between weakly unstable to unstable. Comparing the wind direction standard deviations coupled with the wind speeds showed that the WRF-LES underestimated the wind direction fluctuations for wind speeds smaller than 3-ms(-1). Based on the strengths of convection and shear, WRF-LES was able to simulate horizontal convection roll and convective cell type features.