Concurrent data assimilation of methane concentrations and fluxes

We use atmospheric inverse modelling to provide observation-based estimates of methane emissions at the national scale in Europe. We apply the numerical weather prediction model ICON-ART to obtain an ensemble of methane concentrations by varying the meteorology, the lateral boundary conditions and emission fields. By comparing to ground based observations of the ICOS network, we employ a 4D LETKF to assimilate both the concentrations and emissions concurrently. We create an ensemble of emissions in two ways: We can perturb the underlying emission field with a gaussian random field, or we can separate it into regions and economic sectors and scale these. We compare the two approaches and the resulting emission estimates to national greenhouse gas inventories and synthesis inversion results with a focus on Germany. The first results are presented for 2021 and we identify a considerable mismatch with the reported emissions in central Europe.