Global Modeling and Analysis of Anthropogenic Combustion and Associated Emissions

Abstract

Anthropogenic combustion and associated emissions have significant impacts on air quality and climate. However, current estimates of emissions from anthropogenic combustion are still subject to large uncertainties, especially in rapidly-developing regions. This hinders accurate assessments of their regional and global impacts on air quality and climate, which presents an urgent need to understand, assess, monitor, and predict anthropogenic combustion and associated emissions particularly at city-to-national scales. Combustion products co-emitted to the atmosphere and their relationships are typically related to characteristics of combustion processes. Thus, in order to understand anthropogenic combustion and associated emissions, my PhD study seeks to answer three major scientific questions: (1) To what extent could current observations of trace gases co-emitted from combustion be used to understand anthropogenic combustion, emissions, and related driving factors? (2) How well do present global climate-chemistry models simulate trace gases from combustion activities and could those models be used to study anthropogenic emissions? (3) To what extent could the current understanding of anthropogenic combustion and emissions be improved by jointly analyzing satellite, ground-based, aircraft measurements, and model simulations of trace gases co-emitted from combustion? To address the first scientific question, I combine air pollution measurements from multiple satellite instruments across 2005-2014 to characterize emergent features of the ratios of carbon monoxide (CO) and sulfate dioxide (SO2) to nitrogen dioxide (NO2) enhancements from anthropogenic emissions over 36 cities in China. The resulting emission pattern is well-correlated with economic development and traces a common emission pathway that resembles the evolution of air pollution in more developed cities. The absence of this progression in the current IPCC Representative Concentration Pathway emission inventory is most likely due to its deficient representation of the shift towards cleaner combustion in more developed cities. The results highlight the usefulness of augmenting observational capabilities by exploiting relationships of combustion tracers in constraining the temporal variation of emissions for gaseous pollutants. In addition, it is also desired to monitor and assess anthropogenic combustion and its impacts through modeling. Thus, to address the second scientific question, I evaluate simulations of two important anthropogenic combustion products (carbon dioxide (CO2) and CO) from a state-of-the-art high-resolution global prediction system, the Copernicus Atmosphere Monitoring Service (CAMS), by comparing with the Korea-United States Air Quality (KORUS-AQ) field measurements (May to June 2016) that aims to understand the factors controlling air quality over East Asia. The results show a slight overestimation for CAMS CO2 and a moderate underestimation for CAMS CO. CAMS also captures the observed more efficient combustion over Seoul compared to China outflows. Furthermore, to address both the second and third scientific questions, I combine observations and model simulations to uncover important combustion sources over East Asia, using the Community Atmosphere Model with chemistry (CAM-chem) with a CO tagging mechanism, where artificial CO tracers (i.e., tags) from specific sources are tracked as standard CO. With 17 CAM-chem tagged CO simulations using various model configurations, I quantify key regional sources of CO during KORUS-AQ. The results show that emissions from middle East Asia dominate continental outflows to Korea, while Korean emissions play an overall more important role for ground sites and plumes within the boundary layer in Korea. The CAM-chem tagging results are generally consistent with other source contribution approaches. Following the CO modeling, together with newly developed CO2 modeling and tagging mechanism in CAM-chem, I demonstrate the use of joint analysis of CO and CO2 towards a multi-species inversion. I simulate atmospheric CO2 as well as CO in CAM-chem using optimized carbon fluxes for CO2. The model results generally agree with observations from satellite, aircraft, and ground-based observations during KORUS-AQ. Then, I implement a CO2 tagging mechanism into the model. The modeled fossil fuel CO2 tags agree well with fossil fuel CO2 derived from radiocarbon samples during the field campaign. I also show that signatures of plume transport and sectoral emissions of CO2 are enhanced in CO analyses. Overall, this work elucidates the use of jointly analyzing CO2 and CO in tracking fossil fuel CO2, quantifying regional sources, and understanding combustion efficiency of sources. In my future work, I will (1) combine observations and model simulations of atmospheric gases to obtain improved estimates of their emissions from anthropogenic combustion based on inverse modeling techniques, and (2) use the improved emission estimates to quantify the impact of trace gases on air quality and climate.