Exploring Improvements to the Aerosol Parameterization in the OCO-2 XCO2 Retrieval Algorithm

The NASA Orbiting Carbon Observatory-2 (OCO-2) retrieval algorithm, Atmospheric Carbon Observations from Space (ACOS), has been significantly improved in terms of both precision and accuracy since its inception. However, one of the primary sources of uncertainty when measuring the column-averaged dry-air mole fraction of carbon dioxide (X_CO2) remains its cloud and aerosol parameterization. This is because near-infrared measurements are highly sensitive to even small levels of cloud or aerosol contamination. Therefore, the OCO-2 retrieval algorithm must include cloud and aerosol properties in its state vector. The current retrieval algorithm, version 8, uses a monthly climatology of aerosol from MERRA-2 to derive many of its aerosol priors. Here, we present recently updated work on using instantaneous aerosol priors from the Goddard Earth Observing System Model Version 5 Forward Processing for Instrument Teams (GEOS-5 FP-IT) aerosol model and evaluating the corresponding impact on X_CO2. We find that using more realistic aerosol a priori information results in a small reduction in scatter compared to TCCON as well as regional bias differences compared to an ensemble model validation source. We also present work on reducing the complexity of the ACOS algorithm by solving for fewer but more intelligently selected cloud and aerosol parameters in order to be in better agreement with the true number of degrees of freedom in the measurements. This reduction of complexity has the potential benefits of a more accurate and precise X_CO2, less nonlinearity, and increased computational speed.