Refining methane emission estimates in the Amazon basin: Addressing spatiotemporal variability and habitat diversity

Recent studies highlight the critical role of methane emissions from tropical wetlands in driving the accelerated atmospheric CH₄ growth rate observed in the last decade. The Amazon lowland region, where up to 30% of the area can be seasonally flooded, is one of the largest natural methane sources. The total methane flux estimates for the Amazon basin from top-down and bottom-up approaches converge at 31–46 TgCH₄/year. However, understanding methane emission trends and interannual variability—such as inundation extent and seasonality—requires improved attribution of emissions to specific wetland types and habitats. In this study, we present a refined bottom-up estimate of methane fluxes for the lowland Amazon that addresses key challenges to regionalizing fluxes in the basin: i) the large seasonal variation in inundated areas and habitats, ii) the diversity of aquatic ecosystems across the Amazon, and iii) the spatiotemporal variability of methane fluxes. 

We link local methane flux measurements collected during more than 20 years of field campaigns to specific river and wetland types and incorporate seasonal variability in inundation extent using dynamic remote sensing products (i.e. open water data from the Global Surface Water for lakes, Global River Width from Landsat (GRWL) for rivers, and wetland inundation extent from the High-Resolution Surface WAterFraction (SWAF-HR, based on SMOS L-band imagery) for the Amazon basin, and (4) GIEMS-D15 (merge of multiple satellites) for the remaining portions of South America). Wetland types (herbaceous and woody vegetation) were obtained from the JERS-1 L-band based classification of Hess et al. (2015) for the Amazon Basin and ESA-CCI land cover for the rest of South America. The magnitude and seasonal variability of our bottom-up fluxes are evaluated against fluxes derived from atmospheric CH₄ mole fraction measurements at two Amazonian sites, whose footprints go beyond the Amazon Basin. While our product successfully captures the seasonal variability at both sites, it underestimates the overall magnitude of emissions compared to other estimates, even when accounting for emissions from flooded forest tree stems. Our findings represent an important improvement of bottom-up estimates representing the diversity of wetland habitats and processes driving methane emissions, but further work is needed to understand the mismatch with other methane emissions products.