An evaluation of a severe smog episode in the Eastern U.S. using regional modeling and satellite measurements

Abstract

An ensemble of regional chemical modeling (WRF/Chem with RADM2) simulations, satellite, ozonesonde, and surface observations during July 7-11, 2007 was used to examine the horizontal and vertical signature of one of the worst smog events in the eastern U.S. in the past decade. The general features of this event -- a broad area of high pressure, weak winds and heavy pollution, terminated by the passage of a cold front -- were well simulated by the model. Average 8-hr maximum O3 has a mean (±Σ) bias of 0.59 (±11.0) ppbv and a root mean square error of 11.0 ppbv. WRF/Chem performed the best on poor air quality days, simulating correctly the spatial pattern of surface O3. Yet the model underpredicted O3 maxima by 5-7 ppbv in the Northeast and overpredicted by 8-11 ppbv in the Southeast. High O3 biases in the Southeast are explained by overpredicted temperatures in the model (>1.5°C). Sensitivity simulations with 1) accelerated O3 dry deposition velocity and 2) suppressed multiphase nitric acid formation pushed the model closer to observations. Simulated O3 vertical profiles over Beltsville, MD showed good agreement with ozonesonde measurements, but the modeled boundary layer depth was overpredicted on July 9, contributing to the low bias over this region. During this severe smog episode, space-borne TES detected high total tropospheric column ozone (TCO) over the Western Atlantic Ocean off the coast near North and South Carolina. The standard product (OMI/MLS) missed the magnitude of these local maxima, but the level-2 ozone profile (OMI) confirmed the TES observations. HYSPLIT back trajectories from these O3 maxima intersected regions of strong convection over the Southeast and Great Lakes regions. When lightning NO emissions were implemented in WRF/Chem, the high concentrations of NOx and O3 off the coast were well reproduced, showing that the exported O3 was produced by a combination of natural NO and pollutants lofted from the lower atmosphere. Lastly, WINTER MONEX O3 data from 1978 are presented for the first time here in discussion of open cell convection over Indonesia.