Modeling METREX Tracer Releases Using High-Resolution WRF Meteorology

Abstract:

The Metropolitan Tracer Experiment (METREX) was conducted throughout 1984 over the Washington DC area and consisted of hundreds of passive tracer releases with air samples collected at many downwind locations. The experiment included intensive 4-hour tracer releases conducted once per month starting in April of 1984. We simulate several of these intensive releases that had good air sampling coverage using the HYSPLIT Lagrangian particle dispersion model driven by meteorological fields generated by WRF on grids as fine as 1 km and utilizing an urban canopy model (UCM) to parameterize the bulk impacts of buildings within the DC urban airshed. Results indicate that 1 km meteorology generated with the UCM dramatically improve the dispersion simulation compared to those using coarser meteorological inputs generated without the UCM for the November 8^th intensive release. This improvement is due to better representation of the vertical mixing at night in WRF. However, for other cases the impact of resolution and the UCM is not as significant as the dispersion was influenced by other factors. We also investigate the sensitivity of the dispersion calculations to the topographic wind speed correction available in WRF 3.6 and several WRF planetary boundary layer (PBL) parameterizations.