Global 3-D model analysis of the seasonal cycle of atmospheric carbonyl sulfide: Implications for terrestrial vegetation uptake

We use a global 3-D simulation of atmospheric carbonyl sulfide (COS) to interpret observations at a network of surface sites. We aim to identify the primary factors underlying observed seasonal variations and to constrain COS uptake by terrestrial vegetation. Model simulations are based on a recent estimate of global COS fluxes, with closure between sources and sinks. We find that the dominant influences on seasonal variation of COS are terrestrial vegetation uptake in the northern extratropics, and ocean fluxes in the southern extratropics. Simulations underestimate the amplitude of the observed seasonal cycle in the northern hemisphere, particularly at terrestrial sites, indicating that COS uptake by terrestrial vegetation has been underestimated in recent budgets. Fitting the observed seasonal variation at northern hemisphere sites in the model requires a doubling of the global vegetation sink to similar to 490 Gg S y(-1), while fitting the southern hemisphere data suggests a reduction of similar to 50 Gg S y(-1) in the southern extratropical ocean source. Balancing these changes in COS fluxes requires an additional source (similar to 235 Gg S y(-1), equivalent to 40% of identified sources) missing from present budget estimates. Discrepancies between annual mean observations and simulated concentrations, derived from our best estimates of seasonal fluxes, are largest in the tropics, suggesting an underestimate of COS sources at these latitudes.