Interannual Methane Sources Inferred by Inversion of Atmospheric Transport

A primary goal of developing the methane (CH4) atmospheric measurement network is to better characterise the sources of the second anthropogenic greenhouse gas in space and time. Chemical transport models (CTM) can be used to analyse atmospheric measurements in terms of surface fluxes using inverse methodology. The main sink of methane, chemical removal by OH radicals, is also an unknown of the inverse problem that can hardly be optimized using surface CH4 observations alone. We have developed a double 3D atmospheric inversion of both methyl Chloroform (CH3CCl3) and methane atmospheric observations in order to optimize interannual OH concentrations and CH4 surface sources successively, for the 1984-1997 period. The inversion scheme infers CH4 monthly sources and sinks from GLOBALVIEW-CH4 observations and GAGE-AGAGE CH3CCl3 observations using INCA-LMDZ CTM in forward and backward (self-adjoint) modes. Natural tropical ecosystems are found to explain most of the interannual variability (IAV) of CH4 sources over the period. In this presentation, We illustrate this result and present contribution to IAV of all types of emissions, at a continental scale. Several sensitivity inversions are also presented to quantify the influence inversion set up on inferred sources : OH optimization, stations used, observation error, flux aggregation error, ....