Atmospheric observations consistent with reported decline in the UK's methane emissions (2013-2020)

Atmospheric measurements can be used as a tool to evaluate national greenhouse gas inventories through inverse modelling. Using 8 years of continuous methane (CH4) concentration data, this work assesses the United Kingdom's (UK) CH4 emissions over the period 2013-2020. Using two different inversion methods, we find mean emissions of 2.10±0.09 and 2.12±0.26Tgyr-1 between 2013 and 2020, an overall trend of -0.05±0.01 and -0.06±0.04Tgyr-2 and a 2%-3% decrease each year. This compares with the mean emissions of 2.23Tgyr-1 and the trend of -0.03Tgyr-2 (1% annual decrease) reported in the UK's 2021 inventory between 2013 and 2019. We examine how sensitive these estimates are to various components of the inversion set-up, such as the measurement network configuration, the prior emissions estimate, the inversion method and the atmospheric transport model used. We find the decreasing trend to be due, primarily, to a reduction in emissions from England, which accounts for 70% of the UK CH4 emissions. Comparisons during 2015 demonstrate consistency when different atmospheric transport models are used to map the relationship between sources and atmospheric observations at the aggregation level of the UK. The posterior annual national means and negative trend are found to be consistent across changes in network configuration. We show, using only two monitoring sites, that the same conclusions on mean UK emissions and negative trend would be reached as using the full six-site network, albeit with larger posterior uncertainties. However, emissions estimates from Scotland fail to converge on the same posterior under different inversion set-ups, highlighting a shortcoming of the current observation network in monitoring all of the UK. Although CH4 emissions in 2020 are estimated to have declined relative to previous years, this decrease is in line with the longer-term emissions trend and is not necessarily a response to national lockdowns.