Forests and Carbon
World forests remove about 15.6 Gt CO₂/year as they grow, while deforestation, fire and degradation release about 8.1 Gt CO₂/year (Harris et al. 2021, 2001–2019). The net forest flux is a sink of about 7.6 Gt CO₂/year, roughly 20% of annual fossil CO₂ emissions. Parts of the tropics, notably the eastern Amazon, have flipped from net sink to source; temperate and boreal forests remain sinks.
Key insights
Tropical forests are the largest gross sink
Tropical forests make the largest gross removal of any biome, but tropical deforestation returns much of it, leaving them a much smaller net sink than their size implies. Amazon eastern regions are now established carbon sources due to combined deforestation and degradation. Temperate forests in Europe, the US and China are reliable sinks. Boreal forests (Russia, Canada) also sequester carbon but face increasing fire and pest risk.
Deforestation rates have fallen but remain high
FAO's Forest Resources Assessment 2025 puts gross deforestation at 17.6 million hectares a year in 1990–2000 and 10.9 million in 2015–2025 — largely in Brazil, DRC, Indonesia, Bolivia. Brazil saw deforestation surge 2019-2022 under Bolsonaro; Lula government reduced rates ~50% in 2023-24. Indonesia deforestation has fallen substantially since 2016 peak. DRC remains a major frontier as forest carbon escapes attention compared to Amazon.
REDD+ has produced mixed results
Reducing Emissions from Deforestation and Forest Degradation framework operating since COP13 (2007). Countries can receive performance-based payments for verified emission reductions. Funding has reached billions of dollars (Brazil's Amazon Fund, Norway's funding to Indonesia, World Bank Forest Carbon Partnership Facility). Effectiveness debate: payments have aligned with reduced deforestation in some cases, but causality is hard to establish.
World forest area 1990–2020 (FRA 2020 series)
Billion hectares
Show data table
| Forest area (Gha) | |
|---|---|
| 1990 | 4.236 |
| 2000 | 4.158 |
| 2010 | 4.106 |
| 2015 | 4.075 |
| 2020 | 4.058 |
Key Finding: Net loss of about 178 million ha (4.2%) from 1990 to 2020 in the FRA 2020 series. FRA 2025 revised the series, with 4.14 billion ha in 2025 and a 324 million ha net loss since 1990. Loss has slowed but continues.
Annual primary tropical forest loss — top countries (2023)
Million hectares
Show data table
| Mha primary forest lost | |
|---|---|
| Brazil | 1.15 |
| DRC | 0.55 |
| Bolivia | 0.49 |
| Indonesia | 0.21 |
| Peru | 0.18 |
| Colombia | 0.11 |
| Cameroon | 0.1 |
| Madagascar | 0.06 |
| Mexico | 0.05 |
| Paraguay | 0.05 |
| Malaysia | 0.05 |
Key Finding: Brazil dominates absolute primary forest loss; DRC and Bolivia second tier.
Methodology & caveats
Gross vs net forest emissions
Gross emissions: total CO₂ released from forest loss. Gross absorption: total CO₂ taken up by growing forests. Net flux: difference. Headlines often quote the net flux (about -7.6 Gt), which obscures the large gross fluxes in both directions. Both numbers matter for policy.
Forest definitions
FAO defines forest as land with tree canopy cover above 10% (or natural ability to reach this). Different definitions include or exclude: tree plantations, savanna woodland, agroforestry. 'Primary forest' (intact, never logged) is the most-protected category — about 29% of total forest area (FRA 2025). Most 'forest' is secondary or managed.
Why tropical forests are turning to sources
Combination of: direct deforestation (clearing for cattle, soy, palm oil), degradation from selective logging, increased fire frequency, drought stress reducing photosynthesis. East Amazon is already net carbon source. West Amazon still net sink. Tipping-point models suggest larger fractions could flip with continued warming/deforestation pressure.
Why forest carbon estimates disagree
The net forest flux depends on what is counted as forest and as human-caused. Harris et al. (2021) use satellite tree-cover change, counting a wide range of woody land. National inventories instead report on 'managed' land, which can include natural regrowth, and so yield a larger reported sink. Global Carbon Budget figures split the flux differently again, treating regrowth under land-use change and the CO₂-driven uptake under the land sink. None of these can simply be added together.
Saturation and permanence
A forest sink is a loan, not a removal for good. Carbon returns when trees burn, die or are logged, and fire seasons such as Canada's in 2023 can turn a region from sink to source within a year. Uptake also weakens as forests age and as drought limits growth. The figure that matters most for the future is therefore not today's sink but how long it lasts under further warming.
Cite this page
GlobeChart. “Forests and Carbon.” Last updated 2026-10-31. https://globechart.com/co2-emissions/forests-carbon/
Please also credit the original publisher named under each chart: FAO Global Forest Resources Assessment 2020, 2025; WRI Global Forest Watch, Hansen et al. data. Spotted an error? Tell us.