Applying satellite observations to improve bottom-up national emission inventories for methane: application to Colombia

Countries report inventories of methane emissions under the Paris Agreement to the United Nations Framework Convention on Climate Change (UNFCCC), but these estimates can have large uncertainties in activity data and emission factors. Top-down information from inversion of satellite observations can provide constraints to improve these estimates, but has been limited by coarse resolution and the quality of bottom-up inventories used both as prior estimates and to attribute inversion results to specific sectors. Here, we address these issues through partnership between bottom-up and top-down methods to quantify 2023 emissions in Colombia. We use satellite observations from TROPOMI and point source imagers (GHGSat, EMIT, aircraft AVIRIS-NG) in an analytical Bayesian inversion at ≈12×12 km2 resolution with the Integrated Methane Inversion v2.0 framework, and evaluate inversion results with independent satellite observations from GOSAT. We construct a spatially resolved version of the emission inventory from Colombia's Biennial Update Report (BUR) to the UNFCCC for use as prior estimate in the inversion, and combine it with high-resolution wetland extent data (GLWDv2) to separate anthropogenic from wetland emissions. The inversion yields posterior methane emissions for Colombia of 8.9 (8.7–9.1) Tg a−1, contributed by wetlands (5.7 (5.5–6.0) Tg a−1) and anthropogenic emissions (3.2 (3.1–3.2) Tg a−1), mainly livestock (2.0 (1.9–2.1) Tg a−1) and waste (0.76 (0.74–0.76) Tg a−1). Adjustments relative to the BUR are +18 % for livestock, +19 % for waste, +60 % for oil/gas, and +92 % for coal. We provide recommendations to improve the BUR, offering a blueprint for combining satellite observations with national inventory data to produce sectorally and spatially resolved emissions estimates that can inform policy.