UC Berkeley

Satellite observations of NO2 provide information about the spatial and temporal variability of NO2 column densities that can be used to infer the processes controlling NO2 concentrations. However, satellite observations require a priori information to properly transform the observed radiances into vertical column densities. Previous work has shown that using a priori data at equal or better spatial resolution to the satellite pixels significantly improves the retrieved NO2. Of particular importance are the a priori vertical profiles that represent the vertical distribution of NO2 within each satellite pixel.

In this dissertation, I show that the temporal resolution of the a priori NO2 profiles, as well as the spatial resolution, is important to accurately retrieve NO2. I show that using profiles with high spatial resolution, but coarse temporal resolution, overestimates the rate at which NO2 is lost in the outflow from an urban source, and that high spatial and temporal resolution of these profiles is necessary to simultaneously retrieve the NO2 column density and lifetime in an urban plume. To account for this, I design and validate an upgrade to the Berkeley High Resolution (BEHR) NO2 retrieval that incorporates daily, high resolution profiles for over 7 years. With this retrieval, I show direct observations of the relationship between NOx concentration and lifetime by examining the weekend-weekday changes in column density and lifetime from several US cities between 2005 and 2014. Specifically, I show that Chicago, IL and Dallas, TX are undergoing a transition from NOx-suppressed to NOx-limited chemistry, which may indicate that future reductions in NOx emissions will be more effective at controlling O3 production, but have less direct effect on NOx concentrations.