Methane budget, seasonality and interannual variability of the three major river basins in Tropical South America

Tropical wetlands are one of the largest natural methane sources but lack of in-situ observations and uncertainty in wetland extent leads to large uncertainly. In this study we analyze the methane budget from three major river basins in South America: the Orinoco, the Amazon, and the Pantanal basins using two atmospheric inversions:  the CAMS-CH₄inversion, which assimilates satellite and in-situ data and the CarboScope methane inversion system constrained by in-situ data only. We make a comparative analysis focusing on the seasonal cycle, interannual variability, and the total methane budget from 2000 to 2019.

The budget difference in posterior estimates between CAMS-CH₄ and CarboScope for these basins are as follows: Amazon Basin: -18.03 TgCH₄/yr, Pantanal Basin: -11.65 TgCH₄/yr, Orinoco Basin: -0.96 TgCH_4/yr.  All together the total flux difference is -30.56 TgCH₄/yr, indicating that CarboScope estimates larger total methane fluxes than the CAMS-CH₄ inversion. Note that a similar difference (30.98 TgCH₄/year) is also seen in the prior fluxes, suggesting that the optimization does not reduce the prior difference in the regions of interest.  While the Amazon Basin emits largest amount of methane, the Orinoco Basin exhibits the highest emissions per unit area, with 21.2 mgCH₄/m²/day. In comparison, Amazon and Pantanal basins have emission of 19.26 mgCH₄/m²/day and 13.36 mgCH₄/m²/day. This shows the significant contribution of the smallest basin, in terms of methane flux density. Not surprisingly, both models indicate that wetlands are the primary methane source in the Amazon and Orinoco basins (~80%). In the Pantanal, CAMS-CH4 shows equal contributions from wetlands and anthropogenic sources, whereas CarboScope attributes dominance to anthropogenic emissions. Interestingly, seasonal patterns differ between the two models. In CAMS-CH₄ there is a strong seasonality, with maximum methane emissions occurring during the wet season across all basins, in CarboScope, there is a double-peak in the Amazon Basin during March (wet) and August (dry). Finally, we investigate the inundation patterns and their relationship to methane emissions trends in these basins, as well as the factors influencing interannual variability to enhance our understanding of the processes driving these emissions.