Improving Retrievals of Air Quality Parameters from Geostationary Measurements

Satellite-based measurements of backscattered sunlight have made essential contributions to knowledge of atmospheric trace gas composition, and subsequently to the understanding of atmospheric chemistry and climate, since the Backscattered Ultraviolet Experiment (BUV) was launched in 1970. The modern generation of solar backscatter instruments have unprecedented spatial resolution, and instruments placed in geostationary orbits can capture diurnal variations previously inaccessible from space. This dissertation investigates improvements to algorithms for retrieving air quality parameters from measurements made by the solar backscatter instrument TEMPO (Tropospheric Emissions: Monitoring of Pollution) over Canada, with a focus on uncertainties created by snowy surfaces and measurement configurations at high latitudes. Rotational Raman scattering is implemented using the Monte Carlo method within SASKTRAN, a radiative transfer framework developed at the University of Saskatchewan. This work is motivated by the Ring effect, an atmospheric phenomenon resulting from rotational Raman scattering which interferes with TEMPO-like measurements. Air mass factors, which quantify the path history of the backscattered light captured by an instrument, are computed and compared across multiple radiative transfer methods, demonstrating their effectiveness for a range of measurement configurations. Horizontally resolved air mass factors are computed and used to quantify the error introduced by common approaches which neglect the horizontal variation of the surface and atmosphere. Two approaches to cloud characterization for correcting trace gas measurements are combined with the goal of reducing difficulties introduced by snowy surfaces. A proof of concept of the combined retrieval demonstrates the potential for extracting additional information content, and the applicability to TEMPO spectra is investigated.