1Laboratoire des Sciences du Climat et de l'Environnement, UMR8212, 91191 Gif-sur-Yvette, France
2Laboratoire de Météorologie Dynamique/CNRS/IPSL, Ecole Polytechnique, Palaiseau, France
3Climate Monitoring and Diagnostics Laboratory, NOAA, Boulder, Colorado, USA
4Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA
5Earth Observation Science, Space Research Centre, University of Leicester, Leicester, UK
6SRON Netherlands Institute for Space Research, Utrecht, the Netherlands
7Centre for Australian Weather and Climate Research, CSIRO Marine and Atmospheric Research, Aspendale, Victoria, Australia
Abstract. Satellite retrievals of methane weighted atmospheric columns are studied within a Bayesian inversion system to infer the global and regional methane emissions and sinks. 19-month inversions from June 2009 to December 2010 are independently computed from three different space-borne observing systems under various hypotheses for prior-flux and observation errors. Posterior methane emissions are inter-compared and evaluated with surface mole fraction measurements, via a chemistry-transport model. Sensitivity tests show that refining the assigned error statistics has a larger impact on the quality of the inverted fluxes than correcting for residual airmass-factor-dependent biases in the satellite retrievals. Improved configurations using TANSO-FTS, SCIAMACHY, IASI and surface measurements induce posterior methane global budgets of respectively, 568 ± 17 Tg yr−1, 603 ± 28 yr−1, 524 ± 16 yr−1 and 538 ± 20 yr−1 over the one-year period August 2009–July 2010. This consistency between some of these satellite retrievals and surface measurements is promising for future improvement of CH4 emission estimates by inversions.