Inverse Modeling of Surface CH4 and δ13C-CH4 Measurements to Understand Recent Trends in Global Methane Emissions

Methane (CH4) is the second most important greenhouse gas with a radiative forcing of 0.97 W/m2 including both direct and indirect effects and a global warming potential of 28 over a 100-year time horizon. Unlike CO2 whose rate of growth in the atmosphere has remained positive and increased in recent decades, the behavior of atmospheric methane is considerably more complex and is much less understood on account of the spatiotemporal variability of its emissions which include biogenic (e.g. wetlands, ruminants, rice agriculture), thermogenic (fossil fuels), and pyrogenic (i.e. biomass burning) sources. After sustained growth during most of the 20th century, the CH4 growth rate declined falling from 15 ppbv/yr during the 1980s to 6 ppbv/yr in the 1990s to near-zero and even negative values in the early 2000s. With some surprise however, the growth rate rebounded in 2007 and has been on average 6 ppbv/yr during the past 10 years. During this same period the 13CH4/12CH4 ratio of atmospheric CH4 also declined suggesting the recent CH4 growth was caused by an increase in 13CH4-depleted biogenic emissions. Here, we provide additional insight into the recent behavior of atmospheric methane by performing a global three-dimensional Bayesian inversion of surface CH4 and 13CH4/12CH4 ratios over the period 1985-2015 using NOAA Global Monitoring Division (GMD) CH4 measurements and the GEOS-Chem chemical-transport model (CTM) at a horizontal grid resolution of 2ox2.5o. The use of the 3-D model allows us to exploit spatial patterns in the global CH4 and 13CH4/12CH4 fields that provide additional constraints on the retrieval of the time-dependent CH4 fluxes. This work follows up on our previous CH4 inversion where we used a 4ox5o horizontal grid for GEOS-Chem to retrieve fluxes from 1985 to 2009. At higher resolution more information is extracted from the observations due to improved model skill and a smaller number of stations aggregated within model grid cells. This increases the weights of the measurements relative to the a priori fluxes in the inversion producing stronger observational constraints on the optimized fluxes. This work assesses the contribution of spatial heterogeneities in the observed CH4 record to the retrieval of global CH4 fluxes and provides a new look into the causes of the recent growth in atmospheric methane.