Kelly A. Graham

and 11 more

As the Arctic climate rapidly warms, there is a critical need for understanding variability and change in the Arctic carbon cycle, but a lack of long-term observations has hindered progress. This work analyzes and interprets measurements of atmospheric carbon dioxide (CO2) mixing ratios from long-term on-ice measurements (the O-Buoy Network), as well as coastal observatories from 2009-2016. The on-ice measurements show smaller seasonal amplitudes when compared to the coastal observatories, in contrast to the general observation of poleward increases of seasonal cycle amplitude. Average on-ice mixing ratios were lower than their coastal counterparts during the winter and spring months, contradicting the expectation that wintertime presents a poleward increasing gradient of CO2. We compare the observations to CO2; simulated in an updated version of the GEOS-Chem 3-D chemical transport model, which includes new tracers of airmass history and CO2; sources and sinks. The model reproduces the observed features of the seasonal cycle and shows that terrestrial biosphere fluxes and synoptic transport explain most CO2; variability over the surface of the Arctic Ocean. Interannually, the coastal observations were more comparable in overall CO2; growth than concurrent measurements over sea ice. We find evidence indicating the presence of ocean gas exchange in and around sea ice during periods where this growth discrepancy occurs. Periods with large spatial gradients are examined, showing that release of CO2; from Arctic waters in years with low sea ice concentration could possibly contribute to the greater interannual increase of CO2; over sea ice compared to land.