Ground-level ozone is a secondary atmospheric pollutant that damages human and vegetation health. The chemical production of ground-level ozone involves the photochemical reactions between nitrogen oxides (NOX = NO + NO2) and volatile organic compounds (VOCs). China is experiencing high levels of ozone due to high precursor emissions in association with rapid urbanization and industrialization in past decades. In this fast-changing environment, a better understanding of the ozone formation process, as well as accurate estimations of the ground-level ozone and its precursors, are sorely needed for effective pollutant control measures.
In this thesis, I focused on extending the current knowledge of the relationship between ozone production and precursor emissions, using modeling simulations coupled with extensive in situ observations in China. I extended the current relationship between peak ozone concentration and precursor emissions to peak time and used the ozone measurements in China cities as constraints to diagnose the NOX and the VOC emissions as well as the nighttime mixing height. I also used box and 1-D models to investigate the local radical activities in a small town in North China Plain based on observations from a field campaign with extended chlorine chemistry and find that the current 3-D model with simplified mechanism underestimate the RO2 production from the intermedia oxygenated volatile organic compounds (OVOCs), leading to an underestimation of the effects of VOCs and NOX emissions on ozone production. I tested the uncertainty of the current empirical fitting method of NOX emissions from satellite observations, using NO2 columns from both a generated ideal plume and plumes from 3-D model simulations. I find that the accuracy of the fitting method largely depends on the cross-section integral distance and need to be considered. Future works will involve validation of the ozone peak time as constraints for precursor emissions by observations under varying emission strength, modification on the current simplified mechanism to correct the OVOC and RO2 production, and quantifying the error in fitted emissions due to the cross-section integral distance.