Multi-Instrument Uncertainty Quantification for Atmospheric Carbon Dioxide Retrievals

An expanding constellation of greenhouse gas monitoring satellites is enabling scientific investigations at multiple spatial and temporal scales. Challenges remain in integrating data products from multiple platforms due to instrument and data processing differences. This work illustrates the application of a common framework for uncertainty quantification (UQ) to the CO2 retrievals from the Orbiting Carbon Observatory-2 and -3 (OCO-2/3) instruments. The probabilistic behavior of these retrievals is investigated through Monte Carlo experiments. The approach can propagate instrument-specific parameters in combination with shared geophysical conditions when the satellites make nearly coincident observations. The procedure yields a statistical model for XCO2 given the retrieved state that can be applied to operational products. We present comparisons of the OCO-2 and OCO-3 conditional distributions for coincident observations as well as global summaries over a recent one-year period.