Development of a response surface model of aviation's air quality impacts in the United States

Abstract

The air quality impacts of aviation are becoming increasingly important given their impact on human health and the projected growth of aviation. In the United States, the government has set targets to manage and reduce the environmental impacts of aviation. In an environmental policy assessment context it is often necessary to rapidly evaluate many possible scenarios, and quantification of uncertainty is important. This makes direct application of comprehensive air quality models such as the Community Multiscale Air Quality (CMAQ) modeling system impractical due to computational cost. Here we develop a response surface model (RSM) – a form of rapid surrogate model – of the impact of aviation emissions on air quality in the United States. We develop an RSM design space and populate it with results from 46 CMAQ simulations, and perform cross-validation of the resultant RSM. The RSM models present-day as well as future impacts amid changing population and non-aviation emissions sources. This enables rapid estimates of the (particulate matter) air quality and human health impacts of aviation emissions scenarios. Cross-couplings between precursor gaseous emissions and PM2.5 species are found, consistent with competition for atmospheric ammonia. We apply the RSM to quantify the human health benefits of emissions reductions in 2018. Using the RSM we estimate that in 2005, aviation landing and takeoff emissions cause ∼195 [90% CI: 80–340] early deaths, while the same emissions cause ∼350 [90% CI: 145–610] mortalities in 2018. An emissions tradespace between aviation NOx and SOx emissions is constructed. It is found that with fleet-wide desulfurization of jet fuel, a 35% reduction in aviation NOx emissions would result in maintaining the same level of aviation-attributable early deaths in 2018 relative to 2005 levels, while an 80% reduction in NOx emissions would half aviation-attributable early deaths.