Detecting contributions of ocean fluxes to atmospheric CO2 variability

Quantifying ocean flux variability in response to interannual variations in climate is important to understand the global carbon cycle. The ocean carbon sink arises from physical, biological, and chemical mechanisms, and the balance of these processes are spatially heterogeneous. Sparse ocean surface measurements complicate our ability to diagnose the net air-sea flux, especially at interannual timescales. Here, we combine atmospheric CO2 observations and simulated CO2 concentrations to identify fingerprints of air-sea gas exchange. We use NASA's Orbiting Carbon Observatory-2 (OCO-2), which directly measures the total column carbon dioxide mole fraction over the ocean. We simulate corresponding CO2 observations using the GEOS-Chem atmospheric transport model driven by temporally and spatially resolved oceanic CO2 flux estimation products based on the SOCAT pCO2 observations. We use these simulations to quantify the ocean contribution to variations in 3-dimensional atmospheric CO2 mole fraction. We separately tag fluxes from ocean biomes so as to track the atmospheric CO2 resulting from each region. We compare the coherence of the estimated annual cycle and interannual variability phase and amplitude against OCO-2 XCO2 data, especially during and after the strong El Nino in 2015-2016. Finally, we explore the model-data mismatches for several given locations. These comparisons offer us new information on ocean carbon cycling and a better understanding of how to leverage ocean CO2 observed from satellite.