The 2019 Methane Budget And Uncertainties At 1 Degree Resolution And Each Country Through Bayesian Integration Of GOSAT Total Column Methane Data And A Priori Inventory Estimates
Abstract. We present 2019 global methane (CH4) emissions and uncertainties, by sector, at 1-degree and country-scale resolution based on a Bayesian integration of satellite data and inventories. Globally, we find that agricultural and fire emissions are 227 +/− 19 Tg CH4/yr, waste is 50 +/− 7 Tg CH4/yr , anthropogenic fossil emissions are 82 +/− 12 Tg CH4/yr, and natural wetland/aquatic emissions are 180 +/− 10 Tg CH4/yr. These estimates are intended as a pilot dataset for the Global Stock Take in support of the Paris Agreement. However, differences between the emissions reported here and widely-used bottom-up inventories should be used as a starting point for further research because of potential systematic errors of these satellite based emissions estimates. Calculation of emissions and uncertainties: We first apply a standard optimal estimation (OE) approach to quantify CH4 fluxes using Greenhouse Gases Observing Satellite (GOSAT) total column CH4 concentrations and the GEOS-Chem global chemistry transport model. Second, we use a new Bayesian algorithm that projects these posterior fluxes to emissions by sector to 1 degree and country-scale resolution. This algorithm can also quantify uncertainties from measurement as well as smoothing error, which is due to the spatial resolution of the top-down estimate combined with the assumed structure in the prior emission uncertainties. Detailed Results: We find that total emissions for approximately 58 countries can be resolved with this observing system based on the degrees-of-freedom for signal (DOFS) metric that can be calculated with our Bayesian flux estimation approach. We find the top five emitting countries (Brazil, China, India, Russia, USA) emit about half of the global anthropogenic budget, similar to our choice of prior emissions. However, posterior emissions for these countries are mostly from agriculture, waste and fires (~129 Tg CH4/yr) with ~45 Tg CH4/yr from fossil emissions, as compared to prior inventory estimates of ~88 and 60 Tg CH4/yr respectively, primarily because the satellite observed concentrations are larger than expected in regions with substantive livestock activity. Differences are outside of 1-sigma uncertainties between prior and posterior for Brazil, India, and Russia but are consistent for China and the USA. The new Bayesian algorithm to quantify emissions from fluxes also allows us to “swap priors” if better informed or alternative priors and/or their covariances are available for testing. For example, recent bottom-up literature supposes greatly increased values for wetland/aquatic as well as fossil emissions. Swapping in priors that reflect these increased emissions results in posterior wetland emissions or fossil emissions that are inconsistent (differences greater than calculated uncertainties) with these increased bottom-up estimates, primarily because constraints related to the methane sink only allow total emissions across all sectors of ~560 Tg CH4/yr and because the satellite based estimate well constrains the spatially distinct fossil and wetland emissions. Given that this observing system consisting of GOSAT data and the GEOS-Chem model can resolve much of the different sectoral and country-wide emissions, with ~402 DOFS for the whole globe, our results indicate additional research is needed to identify the causes of discrepancies between these top-down and bottom-up results for many of the emission sectors reported here. In particular, the impact of systematic errors in the methane retrievals and transport model employed should be assessed where differences exist. However, our results also suggest that significant attention must be provided to the location and magnitude of emissions used for priors in top-down inversions; for example, poorly characterized prior emissions in one region and/or sector can affect top-down estimates in another because of the limited spatial resolution of these top-down estimates. Satellites such as the Tropospheric Monitoring Instrument (TROPOMI) and those in formulation such as the Copernicus CO2M, Methane-Sat, or Carbon Mapper offer the promise of much higher resolution fluxes relative to GOSAT assuming they can provide data with comparable or better accuracy, thus potentially reducing this uncertainty from poorly characterized emissions. These higher resolution estimates can therefore greatly improve the accuracy of emissions by reducing smoothing error. Fluxes calculated from other sources can also in principal be incorporated in the Bayesian estimation framework demonstrated here for the purpose of reducing uncertainty and improving the spatial resolution and sectoral attribution of subsequent methane emissions estimates.
