3.1.1.6 - Tropical rainforest water and carbon
The Amazon illustrates how water cycling supports a large carbon store and how environmental change and human activity can alter both. Use its regional contrasts and dated evidence rather than treating every part of the rainforest as one uniform system.
Locating and defining the Amazon case
The Amazon is a South American tropical rainforest system, but the river basin, forest biome and Brazilian administrative region are different boundaries.
The river basin includes land draining towards the Amazon and tributaries, from Andean headwaters in the west towards the Atlantic in the east. The forest biome describes an ecological area and extends across national boundaries. Brazilian Legal Amazon is an administrative region used for the clearance evidence below; it does not mean the whole basin or all Amazon rainforest. A statement about that reporting unit must retain its name.
This case examines basin-scale water links and contrasts the more disturbed eastern and southeastern Brazilian Amazon with the western-central region in carbon evidence. Moisture comes from oceanic air transport and is repeatedly returned to the atmosphere by forest evapotranspiration. Rainfall, seasonality, soils and land-use pressures vary within the region; southern and eastern areas generally have a more pronounced dry season than wetter western areas. These differences affect resilience to drought and fire.
Keep three scales visible: a cleared plot changes interception and soil conditions locally; extensive forest loss can change moisture supplied downwind; the combined carbon flux contributes to global atmospheric change. A study's estimated source over one region or period does not prove that every intact tree or every hectare is a source.
Water cycling through the forest
Rainforest vegetation both receives rainfall and returns water that can contribute to later rainfall elsewhere.
The layered canopy intercepts rain; some evaporates directly, while throughfall and stemflow deliver water to the ground. Leaf litter, roots and soil structure can favour infiltration. Water is held in soil, percolates towards groundwater, or reaches streams through subsurface and surface routes. Deep roots can access water unavailable to shallow-rooted replacement vegetation during some dry periods. These capacities vary with soil depth, water table and prior drought, so forest does not mean unlimited storage.
Transpiration moves root-accessible water through plants into the atmosphere. Combined with evaporation, it maintains evapotranspiration and atmospheric moisture supply. Air can carry that moisture downwind, where cooling and condensation contribute to precipitation. Spracklen and colleagues' 2012 study combined satellite precipitation and vegetation evidence with atmospheric transport estimates; its observed forest–rainfall association is consistent with this mechanism. It is not a measurement that every local storm was generated by the trees underneath it.
Clearance commonly reduces canopy interception and evapotranspiration. Compaction or soil disturbance can increase fast runoff locally, yet a broader reduction in recycled moisture can reduce rainfall supply downwind. More local runoff immediately after clearing and less regional rainfall over a longer period are compatible outcomes at different scales. Dry-season effects are particularly consequential where vegetation relies on water carried over from the wet season.
Carbon stocks and regional exchange
A large standing carbon stock does not establish the sign of a rainforest's annual net exchange.
Photosynthesis puts carbon into trees and other vegetation, while respiration and decomposition return it. Warm moist conditions support rapid biological transfers, and a substantial share of ecosystem carbon is held in living biomass. Litter adds carbon to soil; soil storage depends on the balance between inputs, microbial processing and losses. Drought can restrict uptake, increase mortality and leave combustible material; clearance and fire can release stored carbon and reduce future uptake.
INPE's 2021 account of Gatti and colleagues' study describes aircraft measurements from 2010–2018 across four Amazon regions and stronger emissions in more disturbed eastern areas. The regional method integrates the influence of a wider area than a single forest plot. Fire-related emissions must be distinguished from the remaining net biological exchange; neither is the amount of standing biomass.
A 2023 analysis by Basso and colleagues, combining atmospheric observations with an inverse transport model, estimates a small net source for its Amazon study area over 2010–2018: 0.13 ±0.17 petagrams of carbon per year. It finds larger emissions in the east and greater uptake in the western-central region, with fire important to the total. The uncertainty is large relative to the net estimate, and this study does not establish a significant trend in the net flux over those years. Region definitions and methods differ between studies. Use the spatial contrast and fire mechanism without converting these findings into a claim that all intact Amazon forest has permanently stopped absorbing carbon.
Drought, clearance and coupled change
Environmental stress and human activity can reinforce one another, but their contributions need separate evidence.
During drought, lower water availability can reduce photosynthesis and evapotranspiration, while heat and dry fuel can increase fire risk. Human clearance for farming and infrastructure removes canopy, disturbs soils and fragments remaining forest. Edges can be more exposed to drying and fire. Fire releases carbon; losing forest moisture recycling can intensify water stress under suitable regional conditions. The return from water stress to further forest loss creates a possible amplifying loop. Its strength varies by location, and it is not proof of one certain basin-wide tipping date.
INPE's consolidated PRODES result reports 5,731 square kilometres of clearance in Brazilian Legal Amazon for 1 August 2024–31 July 2025, compared with 6,518 for the previous reporting year. The reduction is . Annual loss slowed; it did not become zero, and lost forest did not necessarily regrow. The figures measure mapped removal of primary forest cover under the programme's definition, not every form of degradation or a direct carbon-emissions total.
The consolidated INPE release, published 18 August 2026 uses satellite analysis. The earlier provisional 2025 figure is superseded here. These recent administrative-region totals must not be merged as though they were the 2010–2018 carbon study's flux observations. They provide different evidence about ongoing pressure on the system.
Combining satellite and qualitative evidence
Different evidence sources answer different questions about forest change and its management.
In GIS, mapped clearance polygons can be compared with forest cover, catchment boundaries and dry-season climate layers. Match dates, coordinate systems and definitions before calculating overlap; a pixel count needs an appropriate area conversion. Cloud cover, resolution and detection rules affect what is recorded. Comparing mapped clearance with river response or atmospheric measurements can test connections, but spatial coincidence alone does not isolate a cause. A satellite clearance map cannot reveal household motives or consent.
Two primary institutional perspectives show why qualitative evidence matters. In its 27 November 2025 COP30 assessment, the Articulation of Indigenous Peoples of Brazil emphasises territorial rights, Indigenous-led management and effective consent. Brazil's Rural Sustentável programme presents improved agricultural and forest management as a route combining poverty reduction, biodiversity protection and climate benefit in the Amazon and Atlantic Forest biomes. These are paraphrased institutional positions, not invented interviews or the view of every resident. Both can favour conservation while assigning different priorities to governance and livelihoods.
Code each account's stated objectives and compare them with physical mechanisms: retaining canopy protects stocks and moisture recycling; improving production on existing land may reduce pressure for further clearance only if displacement and incentives are controlled. Institutional purpose affects emphasis, but does not automatically invalidate an account. A sound judgement combines dated physical evidence, uncertainty, local rights and livelihoods, and the scale at which the proposed action changes water and carbon transfers.