Impact of variable air-sea O2 and CO2 fluxes on atmospheric potential oxygen (APO) and land-ocean carbon sink partitioning

A three dimensional, time-evolving field of atmospheric potential oxygen (APO similar to O-2/N-2+CO2) was estimated using surface O-2, N-2 and CO2 fluxes from the WHOI ocean ecosystem model to force the MATCH atmospheric transport model. Land and fossil carbon fluxes were also run in MATCH and translated into O-2 tracers using assumed O-2: CO2 stoichiometries. The modeled seasonal cycles in APO agree well with the observed cycles at 13 global monitoring stations, with agreement helped by including oceanic CO2 in the APO calculation. The modeled latitudinal gradient in APO is strongly influenced by seasonal rectifier effects in atmospheric transport. An analysis of the APO-vs.-CO2 mass-balance method for partitioning land and ocean carbon sinks was performed in the controlled context of the MATCH simulation, in which the true surface carbon and oxygen fluxes were known exactly. This analysis suggests uncertainty of up to +/- 0.2 PgC in the inferred sinks due to variability associated with sparse atmospheric sampling. It also shows that interannual variability in oceanic O-2 fluxes can cause large errors in the sink partitioning when the method is applied over short timescales. However, when decadal or longer averages are used, the variability in the oceanic O-2 flux is relatively small, allowing carbon sinks to be partitioned to within a standard deviation of 0.1 Pg C/yr of the true values, provided one has an accurate estimate of long-term mean O-2 outgassing.