Sharp rises in the atmospheric abundance of ethane (C 2 H 6 ) have been detected from 2009 onwards in the Northern Hemisphere as a result of the unprecedented growth in the exploitation of shale gas and tight oil reservoirs in North America. Using time series of C 2 H 6 total columns derived from groundbased Fourier transform infrared (FTIR) observations made at five selected Network for the Detection of Atmospheric Composition Change sites, we characterize the recent C 2 H 6 evolution and determine growth rates of ∼5% yr −1 at mid-latitudes and of ∼3% yr −1 at remote sites. Results from CAM-chem simulations with the Hemispheric Transport of Air Pollutants, Phase II bottom-up inventory for anthropogenic emissions are found to greatly underestimate the current C 2 H 6 abundances. Doubling global emissions is required to reconcile the simulations and the observations prior to 2009. We further estimate that North American anthropogenic C 2 H 6 emissions have increased from 1.6 Tg yr −1 in 2008 to 2.8 Tg yr −1 in 2014, i.e. by 75% over these six years. We also completed a second simulation with new top-down emissions of C 2 H 6 from North American oil and gas activities, biofuel consumption and biomass burning, inferred from space-borne observations of methane (CH 4 ) from Greenhouse Gases Observing SATellite. In this simulation, GEOS-Chem is able to reproduce FTIR measurements at the mid-latitudinal sites, underscoring the impact of the North American oil and gas development on the current C 2 H 6 abundance. Finally we estimate that the North American oil and gas emissions of CH 4 , a major greenhouse gas, grew from 20 to 35 Tg yr −1 over the period 2008-2014, in association with the recent C 2 H 6 rise.

Abstract. In support of the global stocktake of the Paris Agreement on climate change, this study presents a comprehensive framework to process the results of an ensemble of atmospheric inversions in order to make their net ecosystem exchange (NEE) carbon dioxide (CO2) flux suitable for evaluating national greenhouse gas inventories (NGHGIs) submitted by countries to the United Nations Framework Convention on Climate Change (UNFCCC). From inversions we also deduced anthropogenic methane (CH4) emissions regrouped into fossil and agriculture and waste emissions, as well as anthropogenic nitrous oxide (N2O) emissions. To compare inversion results with national reports, we compiled a new global harmonized database of emissions and removals from periodical UNFCCC inventories by Annex I countries, and from sporadic and less detailed emissions reports by non-Annex I countries, given by national communications and biennial update reports. No gap filling was applied. The method to reconcile inversions with inventories is applied to selected large countries covering ∼90 % of the global land carbon uptake for CO2 and top emitters of CH4 and N2O. Our method uses results from an ensemble of global inversions produced by the Global Carbon Project for the three greenhouse gases, with ancillary data. We examine the role of CO2 fluxes caused by lateral transfer processes from rivers and from trade in crop and wood products and the role of carbon uptake in unmanaged lands, both not accounted for by NGHGIs. Here we show that, despite a large spread across the inversions, the median of available inversion models points to a larger terrestrial carbon sink than inventories over temperate countries or groups of countries of the Northern Hemisphere like Russia, Canada and the European Union. For CH4, we find good consistency between the inversions assimilating only data from the global in situ network and those using satellite CH4 retrievals and a tendency for inversions to diagnose higher CH4 emission estimates than reported by NGHGIs. In particular, oil- and gas-extracting countries in central Asia and the Persian Gulf region tend to systematically report lower emissions compared to those estimated by inversions. For N2O, inversions tend to produce higher anthropogenic emissions than inventories for tropical countries, even when attempting to consider only managed land emissions. In the inventories of many non-Annex I countries, this can be tentatively attributed to a lack of reporting indirect N2O emissions from atmospheric deposition and from leaching to rivers, to the existence of natural sources intertwined with managed lands, or to an underestimation of N2O emission factors for direct agricultural soil emissions. Inversions provide insights into seasonal and interannual greenhouse gas fluxes anomalies, e.g., during extreme events such as drought or abnormal fire episodes, whereas inventory methods are established to estimate trends and multi-annual changes. As a much denser sampling of atmospheric CO2 and CH4 concentrations by different satellites coordinated into a global constellation is expected in the coming years, the methodology proposed here to compare inversion results with inventory reports (e.g., NGHGIs) could be applied regularly for monitoring the effectiveness of mitigation policy and progress by countries to meet the objective of their pledges. The dataset constructed by this study is publicly available at https://doi.org/10.5281/zenodo.5089799 (Deng et al., 2021).