Two decades of fires in the Brazilian Amazon and the differences in patterns between open and forest vegetation

The Amazon is facing unprecedented disturbance. Determining trends in Amazonia’s carbon balance and its sensitivity to disturbance requires reliable vegetation models that adequately capture how its ecosystems exchange carbon with the atmosphere. Our work presents novel estimates of vegetation carbon exchange with the atmosphere across the Amazon, estimated using ground- and satellite-based ecosystem measurements. Our model agrees with independent aircraft observations from discrete locations.

Multi-inverse modeling inter-comparison projects offer a chance to assess uncertainties in inversion estimates arising from various sources. This study proposes a method to quantify errors of regional terrestrial biosphere CO₂ flux estimates from an inverse model ensemble by using airborne CO₂ measurements. Our observation-based error estimates exceed the ensemble spread of flux estimates in regions with high anthropogenic emission regions, suggesting systematic biases in inversion estimates.

Atmospheric methane (CH₄) has been growing steadily since 2007 for reasons that are not well understood. Here we determine sources of methane using a technique informed by atmospheric measurements of CH₄ and its isotopologue ¹³CH₄. Measurements of ¹³CH₄ provide for better separation of microbial, fossil, and fire sources of methane than CH₄ measurements alone. Compared to previous assessments such as the Global Carbon Project, we find a larger microbial contribution to the post-2007 increase.

We assimilate MOPITT CO satellite data in the TM5-4D-Var inverse modelling framework to estimate Amazon fire CO emissions for 2003–2018. We show that fire emissions have decreased over the analysis period, coincident with a decrease in deforestation rates. However, interannual variations in fire emissions are large, and they correlate strongly with soil moisture. Our results reveal an important role for robust, top-down fire CO emissions in quantifying and attributing Amazon fire intensity.

Methane (CH₄) is an important greenhouse gas emitted from wetlands like those found in the basin of the Amazon River. Using an atmospheric model and observations from GOSAT, we quantified CH₄ emissions from Amazonia during the previous decade. We found that the largest emissions came from a region in the eastern basin and that emissions there were rising faster than in other areas of South America. This finding was supported by CH₄ observations made on aircraft within the basin.

On average, the terrestrial biosphere carbon sink is equivalent to ~ 20 % of fossil fuel emissions. Understanding where and why the terrestrial biosphere absorbs carbon from the atmosphere is pivotal to any mitigation policy. Here we present a regionally resolved satellite-constrained net biosphere exchange (NBE) dataset with corresponding uncertainties between 2010–2018: CMS-Flux NBE 2020. The dataset provides a unique perspective on monitoring regional contributions to the CO₂ growth rate.

For reliable measurements of CO₂ mole fractions and its stable isotope composition in air samples, one needs to carefully dry them during collection. Here we describe evaluation of a portable, consumable-free and power-free Nafion-based drying system that is currently being used for sample collection over the Amazon. Laboratory tests indicate that this Nafion-based system does not influence the mole fraction measurements of CH₄, CO, N₂O, SF₆, and CO₂ and the stable isotope composition of CO₂.