Data assimilation of NOAA surface and MIPAS satellite observations for improvements towards the global budget of carbonyl sulfide

Carbonyl sulfide (COS) is a long-lived trace gas with an average tropospheric mixing ratio of approximately 485 pmol mol^-1 and a lifetime of about 2 years. In the absence of a significant atmospheric trend, its budget is considered balanced, but significant uncertainties remain on individual sources and sinks. Although challenging, accurate quantification of the COS budget is important because COS uptake by terrestrial biosphere has the potential to be used for improved assessment of Gross Primary Productivity (GPP). Here, we will report inversion progress of optimizing COS global budgets by using TM5-4DVAR. To this end, we assimilate data from the MIPAS limb sounder onboard the ENVISAT satellite that measured atmospheric emission profiles down to the upper troposphere from 2002 – 2012. Tropospheric COS retrievals are assimilated together with NOAA COS surface observations, and a bias correction scheme is employed to correct for potential calibration differences. We will show that: 1) inversion experiments close the budgets and are in favor of reduced biosphere uptake; 2) co-assimilation and bias correction scheme improve fitting ground and satellite retrievals and validation against HIAPER Pole-to-Pole Observations (HIPPO); 3) the prior flux errors are also important for achieving more realistic inversions.