Ensemble Air Quality Prediction at Convection-Resolving Resolutions for the Continental United States

Air quality simulations and predictions are subject to many unavoidable sources of uncertainties. Such uncertainties come from initial conditions of both meteorological and chemical fields, and from model treatments of physical and chemical processes. The source and sink terms constitute additional uncertainties. A single deterministic air quality prediction will not be able to provide estimates of the forecast uncertainties, which are equally important for decision making. Ensemble simulations and predictions, which incorporate as many sources of error as possible, can provide guidance on both the most likely pollutant evolutions and the range of possibilities. For the forecast of meteorological conditions, ensemble forecasting at resolutions that can treat convective storms and precipitation explicitly represent a major advance in terms of forecast accuracy, and the much better resolved three-dimensional atmospheric flows, including the stability conditions and turbulence, will also greatly improve the transport and dispersion of atmospheric pollutants. The Center for Analysis and Prediction of Storms (CAPS) at the University of Oklahoma has been running a realtime continental US (CONUS) domain storm-scale ensemble forecasting (SSEF) system for a number of years during the spring storm season at 4 or 3 km grid spacing. The benefit of coupling an air quality model with a convection-resolving ensemble forecasting system has so far been little explored, however. As a first step towards establishing such a system, we will introduce an ozone transport and dispersion model into CAPS's SSEF system, and test the system for a severe ozone episode in the South Great Plains during August 25-29, 2011 when good measurements are available. The performance of the system, including its ability to properly estimate forecast uncertainties will be examined. The work will lay a foundation for realtime ensemble transport and dispersion forecast experiments for greenhouse gases, as part of the ACT-America Project.