Modeling Interannual Variations and Spatial Gradients of Atmospheric CO2 Over the Arctic Ocean

Abstract:

Understanding ocean-atmosphere CO₂ fluxes in the Arctic Ocean has been limited by the difficulty of making long-term measurements there. With the Arctic warming at alarming rates, understanding CO₂ exchange in and around sea ice is imperative for long-term projections and carbon cycle science. In this work, we analyzed and modeled CO₂ observations from long-term on-ice measurements (the O-Buoy Chemical Observing Network), as well as coastal observatories in the NOAA GLOBALVIEWplus network. The GEOS-Chem 3-D global chemical transport model (version 12.7.2 at 2º x 2.5º) CO₂ simulation was updated to use recent surface fluxes: NOAA’s CarbonTracker for terrestrial and ocean fluxes, CEDS for fossil fuel emissions, GFED 4.1s for biomass burning, and archived chemical production of CO₂. Overall, GEOS-Chem replicated the observed long-term growth rate and nearly matched the magnitude of the seasonal cycle. The largest monthly averaged model-observation discrepancies occurred during the months of June through September, which may indicate incorrect magnitude of terrestrial uptake, or incorrect atmospheric transport. High frequency differences, particularly in autumn, may result from a lack of flux parameterization over the Arctic Ocean (where sea ice has been traditionally assumed to have no flux). It has been long understood that sea ice does not act as a capping lid to gas exchange, but rather plays an integral part through multiple physical, chemical, and biological mechanisms, and these lessons should be incorporated into models. In addition, we identified periods where nearby buoys and coastal land stations differed in observed CO_2, for which we used surface contact tracers and tagged CO₂ tracers to interpret possible causes for these differences.