Effect of Dynamic Downscaling from Mesoscale to Microscale using Weather Research & Forecasting model on Plume Dispersion for Project Sagebrush Tracer Experiment-Phase 1

Atmospheric boundary layer (ABL) conditions during the convective intensive observation periods of Sagebrush tracer experiment phase-1 (PSB) were modeled using the dynamic-downscaling methodology in Weather Research & Forecast (WRF) model. Multiple nests were created with grid sizes ranging from 24-km (mesoscale) to 150-m (microscale). Parameterization schemes were used in WRF for modeling boundary layer physics in domains of grid size greater than 1-km. Large-eddy simulations (LES) were run for domains of grid size less than 1-km. Output from mesoscale domains was used to drive the inner fine resolution large-eddy domains. Surface meteorological fields and vertical profiles of wind speed, direction and temperature were validated against the observations during the experiment period. Boundary layer heights modeled were compared against the radiosonde observations. Turbulence profiles near the ground were well captured by the LES when compared against the sonic anemometer readings from the PSB dataset. Fluctuations in the wind direction were underestimated by WRF model for wind speeds less than 3 ms^-1. WRF modeled ABL has convective cell type features for the days with low sensible heat flux and wind shear. For days with moderate to stronger winds and surface heat fluxes, the convective features resembled that of rolls with their roll-axis aligned parallel to the wind direction.

LES modeled meteorological (MET) fields were used to simulate the tracer dispersion using HYSPLIT Lagrangian plume model in "offline" mode. The metrics of total ground concentration, plume width were compared among the plume model results ran with (i) LES data simulated by WRF model, (ii) re-analysis data obtained from Air Resource Laboratory archives and (iii) local MET data from observations during PSB. The effect and advantages of dynamic downscaling were discussed using the above-mentioned metrics.