Improved understanding of anthropogenic and biogenic carbonyl sulfide (COS) fluxes in western Europe from long-term continuous mixing ratio measurements

Abstract. Lack of knowledge still remains on many processes leading to carbonyl sulfide (COS) atmospheric fluxes, either natural, such as the oceanic sources or the vegetation and soil uptakes, or anthropogenic, with emissions from industrial activities and power generation. Moreover, COS atmospheric mixing ratio data are still too sparse to evaluate the estimations of these sources and sinks at the regional scale; in this context, regional estimates are very challenging. This study assesses the anthropogenic emissions and biogenic COS uptakes at the regional scale, in the footprint of a measurement site in western Europe, at a seasonal to diurnal time resolution over half a decade. The continuous time series of COS mixing ratios obtained at the monitoring site of Gif-sur-Yvette (GIF; in the Paris region) from August 2014 to December 2019 are compared to simulations with the Lagrangian model FLEXPART (FLEXible PARTicle), transporting oceanic sources, biogenic land fluxes from the land surface models ORCHIDEE and SiB4 (Simple Biosphere Model), and anthropogenic emissions by two different inventories. At GIF, the seasonal variations in COS mixing ratios are dominated by the contributions of the background and ocean, the weekly to daily variations are driven by the biogenic land contribution and anthropogenic emissions may dominate for short episodes of high concentrations. The anthropogenic emission inventory based on reported industrial emissions and the characteristics of coal power plants in Europe is consistent with the observations. The main limitation of this inventory is the flat temporal variability applied to anthropogenic fluxes due to the lack of information on industrial and power-generation activities in viscose factories and in coal power plants. As a consequence, there are potential mismatches in the simulated plumes emitted from these hot spots. We find that the net ecosystem COS uptake simulated by both ORCHIDEE and SiB4 is underestimated in winter at night, which suggests improvements in the parameterization of the nighttime uptake by plants for COS. In spring, SiB4 simulates persistent nighttime uptake by vegetation, which is different than ORCHIDEE, which leads to more realistic simulations with SiB4 than with ORCHIDEE. In summer, both models represent fluxes sufficiently well, with better agreement from ORCHIDEE in terms of magnitudes.