Amazon deforestation pressure

What this driver is

It is a mechanism, not a global forest-cover mood, not a COP30 brochure, and not a welfare score.

It is not biodiversity loss, not ecosystem restoration, not wildfire risk, and not the warming bundle on climate crisis. Those measure other biomes, species, restoration finance, fire weather, and GMST. This node is Amazon clear-cut and the enforcement or non-enforcement that moves the PRODES rate.

It applies from the PRODES year (1 August–31 July) through the 2021 peak of 13,038 km², the 2022 consolidated 11,594 km², the 2023 consolidated 9,064 km², and the March 2026 INPE note that consolidated 2024 at 6,518 km² and 2025 at 5,731 km². It is not ukraine-war and not territorial-revisionism. A Congo logging concession does not hang here.

The observational core is INPE PRODES for the nine Legal Amazon states (https://terrabrasilis.dpi.inpe.br/ (opens in a new tab)). DETER is the alert system for enforcement, not the annual rate. A ministerial press line that “50% below 2022” is a comparison of two INPE rates; it is not a unique Factrail hectare and not Contribution.size.

Mechanism

The pressure moves welfare through named channels.

First, clear-cut. PRODES maps suppression of native vegetation ≥6.25 ha in the Legal Amazon year. INPE’s 2022 technical note set the consolidated 2021 rate at 13,038 km² and 2022 at 11,594 km² (https://www.gov.br/inpe/pt-br/assuntos/ultimas-noticias/a-taxa-consolidada-de-desmatamento-para-os-nove-estados-da-amazonia-legal-em-2022-foi-de-11-594-km2-1 (opens in a new tab)). INPE’s 8 May 2024 note set 2023 at 9,064 km². The 2 March 2026 INPE PDF set 2024 at 6,518 km² and 2025 at 5,731 km², a 12.07% decline versus 2024, with Mato Grosso up 26.73% while eight other states fell (https://data.inpe.br/ (opens in a new tab)). Those km² are INPE’s. This page does not invent a second area.

Second, command-and-control. PPCDAm was relaunched in 2023 after years of weakened Ibama/ICMBio capacity. Embargoes, fines, and credit restrictions are the instruments.

Third, fire and degradation. PRODES is not a fire inventory. INPE and independent work recorded rising degradation and fire-scar shares even as clear-cut fell. Fire-weather belongs on wildfire risk. Treating a lower PRODES year as “the forest is safe” is a category error.

Fourth, commodity and land-grab channels. Cattle, soy, land speculation, and undesignated public forest are how a satellite polygon becomes a pasture.

Fifth, diplomacy. COP30 in Belém (November 2025) and the Tropical Forest Forever Facility pitch are host-city and finance instruments. They do not rewrite 2021–2022. A 2030 “zero deforestation” pledge is a target, not a 2025 rate of zero.

Timing is a dry-season fortnight for a clear-cut and a PRODES year for the official rate. A later PRODES year above 2021, or a genuine near-zero year in the same table, would be a new fact.

Evidence base

The best evidence is INPE’s own consolidated notes.

2021 = 13,038 km² and 2022 = 11,594 km²: INPE 2022 consolidated note. 2023 = 9,064 km²: INPE/MCTI 8 May 2024. 2024 = 6,518 km² and 2025 = 5,731 km²: INPE PDF, 2 March 2026. TerraBrasilis is the public dashboard. G1 and other independent outlets reported the same 2021–2023 figures in the week of each release.

The meter is the Legal Amazon PRODES rate scaled by the 2021 peak of 13,038 km², so every point is a sourced km² divided by a sourced km² from the same official series. 2021 is 1.0, 2022 is 0.8892, 2023 is 0.6952, 2024 is 0.5000, 2025 is 0.4396. That is a documented transformation of one INPE table, not an invented hectare coefficient. Government claims of “avoided MtCO2e since 2022” stay on their MMA edge instrument; they are not copied here as Contribution.size.

Evidence is strongest for a consolidated PRODES note. It is weaker for DETER alerts as an annual rate. It is weakest where a COP30 host speech is asked to rebate 13,038 km² in 2021, or where a unique indigenous-territory death integer is asked for. This page does not invent that integer.

Effect on welfare indicators

This driver, when clear-cut actually runs, is a land-use emissions and smoke channel.

• Global CO2 emissions per capita: direction higher residual carbon when Amazon clear-cut and subsequent fire run. Horizon: same PRODES year to a few years. Confidence: medium — the live series is production-based fossil-heavy; LULUCF is the Amazon channel and is not a unique Factrail GtCO2 here.
• Population-weighted PM2.5 air pollution exposure (global): direction higher particulate in fire-and-clear years. Horizon: days to a dry season. Confidence: medium for basin smoke; wildfire risk remains the fire-weather companion.
• Under-five mortality: direction higher deaths only where smoke, mercury from mining, or displacement is documented in the same population. Horizon: months to years. Confidence: low as a global coefficient. No child-death size is invented.

An electricity-access or PISA path does not follow. Missing coefficients stay missing.

The welfare sign is negative: strengthening Amazon clear-cut increases lower-is-better carbon and particulate readings. A 2025 rate below 2021 is a lower intensity of the same driver, not a green-transition miracle that rebates 2021.

Interactions and contested points

Reinforcement is obvious when a high PRODES year and a fire season sit together. Offsetting is the claim that PPCDAm, a Belém COP, or a TFFF pledge rebates the 2021 peak. Those instruments can be stored. They do not cancel 13,038 km².

The strongest competing view is that 2023–2025 declines show the driver is over. That objection is serious. This node stores the decline (5,731 km² in 2025) and the remaining annual clear-cut, plus Mato Grosso’s 2025 increase in the same INPE table. A lower rate is not a zero rate.

It does not. Global pressure is the LIVE generic companion. Mixing them would be the twin the roster forbids.

A third competing view is that PRODES misses degradation, so the decline is cosmetic. Degradation and fire-scar series are neighbouring evidence.

Not established: a unique Factrail km², a unique GtCO2, a unique death total, or that COP30 zeroed the basin. A later PRODES year is a new point on this meter.