The Southeast Atmosphere Studies (SAS), which included the Southern Oxidant and Aerosol Study (SOAS); the Southeast Nexus (SENEX) study; and the Nitrogen, Oxidants, Mercury and Aerosols: Distributions, Sources and Sinks (NOMADSS) study, was deployed in the field from 1 June to 15 July 2013 in the central and eastern United States, and it overlapped with and was complemented by the Studies of Emissions, Atmospheric Composition, Clouds and Climate Coupling by Regional Surveys (SEAC4RS) campaign. SAS investigated atmospheric chemistry and the associated air quality and climate-relevant particle properties. Coordinated measurements from six ground sites, four aircraft, tall towers, balloon-borne sondes, existing surface networks, and satellites provide in situ and remotely sensed data on trace-gas composition, aerosol physicochemical properties, and local and synoptic meteorology. Selected SAS findings indicate 1) dramatically reduced NOx concentrations have altered ozone production regimes; 2) indicators of “biogenic” secondary organic aerosol (SOA), once considered part of the natural background, were positively correlated with one or more indicators of anthropogenic pollution; and 3) liquid water dramatically impacted particle scattering while biogenic SOA did not. SAS findings suggest that atmosphere–biosphere interactions modulate ambient pollutant concentrations through complex mechanisms and feedbacks not yet adequately captured in atmospheric models. The SAS dataset, now publicly available, is a powerful constraint to develop predictive capability that enhances model representation of the response and subsequent impacts of changes in atmospheric composition to changes in emissions, chemistry, and meteorology.

Underprediction of peak ambient pollution by air quality models hinders development of effective strategies to protect health and welfare. The U.S. Environmental Protection Agency's community multiscale air quality (CMAQ) model routinely underpredicts peak ozone and fine particulate matter (PM2.5) concentrations. Temporal misallocation of electricity sector emissions contributes to this modeling deficiency. Hourly emissions are created for CMAQ by use of temporal profiles applied to annual emission totals unless a source is matched to a continuous emissions monitor (CEM) in the National Emissions Inventory (NEI). More than 53% of CEMs in the Pennsylvania-New Jersey-Maryland (PJM) electricity market and 45% nationally are unmatched in the 2008 NEI. For July 2006, a United States heat wave with high electricity demand, peak electric sector emissions, and elevated ambient PM2.5 mass, we match hourly emissions for 267 CEM/NEI pairs in PJM (approximately 49% and 12% of unmatched CEMs in PJM and nationwide) using state permits, electricity dispatch modeling and CEMs. Hourly emissions for individual facilities can differ up to 154% during the simulation when measurement data is used rather than default temporalization values. Maximum CMAQ PM2.5 mass, sulfate, and elemental carbon predictions increase up to 83%, 103%, and 310%, at the surface and 51%, 75%, and 38% aloft (800 mb), respectively.

On high electricity demand days, when air quality is often poor, regional transmission organizations (RTOs), such as PJM Interconnection, ensure reliability of the grid by employing peak-use electric generating units (EGUs). These "peaking units" are exempt from some federal and state air quality rules. We identify RTO assignment and peaking unit classification for EGUs in the Eastern U.S. and estimate air quality for four emission scenarios with the Community Multiscale Air Quality (CMAQ) model during the July 2006 heat wave. Further, we population-weight ambient values as a surrogate for potential population exposure. Emissions from electricity reliability networks negatively impact air quality in their own region and in neighboring geographic areas. Monitored and controlled PJM peaking units are generally located in economically depressed areas and can contribute up to 87% of hourly maximum PM2.5 mass locally. Potential population exposure to peaking unit PM2.5 mass is highest in the model domain's most populated cities. Average daily temperature and national gross domestic product steer peaking unit heat input. Air quality planning that capitalizes on a priori knowledge of local electricity demand and economics may provide a more holistic approach to protect human health within the context of growing energy needs in a changing world.

No abstract