Deforestation and Land-Use Change
Forests are both a slow-moving carbon sink and a major source of CO₂ when they are cleared. Land-use-change emissions are the most uncertain part of the global carbon budget, and the difference between two reasonable accounting choices can be larger than the headline trend they are reporting.
Last reviewed on 2026-04-27.
Two ways to count a vanishing forest
FAO country reporting
The FAO Forest Resources Assessment is the long-running international source on forest area. It rests on definitions agreed by member countries: a forest is land at least 0.5 hectares with tree canopy cover above 10% and trees that can reach at least 5 metres at maturity. National statistical offices report changes every five years using a mix of inventories, remote sensing and expert review. The FRA series goes back to the early 1990s and is the only source for a consistent country-by-country forest-area trend.
Satellite tree-cover loss
Global Forest Watch publishes a different object: annual tree-cover loss derived from Landsat satellite imagery. It does not distinguish a permanent conversion to cropland from a logging coupe that will regrow, or even a fire that will recover; it counts pixels in which the canopy has been removed. The series is consistent across countries and updated quickly, but the headline numbers are systematically larger than FAO “deforestation” figures because tree-cover loss includes more events than land-use change does.
Net change versus gross loss
Net forest-area change subtracts forest gain — natural regrowth and plantations — from forest loss. Gross loss reports only the negative side. A country can have substantial deforestation in one region while net forest area is rising nationally because of plantations elsewhere. Both figures are real, but they answer different questions: the gross loss tells you what is being cleared; the net change tells you whether the country’s overall forest stock is shrinking. Headline charts can use either; check which.
Primary versus all forest
Primary forest — undisturbed natural forest — has different ecological and carbon characteristics from secondary or plantation forest. A country whose total forest area is stable can still be losing primary forest, especially in the tropics. Reports that aggregate “forest” without distinguishing primary, secondary and planted forest blur a distinction that matters for biodiversity and for long-term carbon storage.
From hectares to a CO₂ flux
Why land-use emissions exist as a separate line
The Global Carbon Budget tracks three main fluxes of CO₂: emissions from fossil fuels and industry, emissions from land-use change, and the net uptake by oceans and the land biosphere. Land-use-change emissions cover deforestation, forest degradation, peatland drainage, and the slow decomposition of cleared biomass; they are partly offset by reforestation and the sink in remaining forests. Because the same forest can be both a source and a sink simultaneously, the net land-use-change number is small relative to the gross fluxes that compose it — and the uncertainty bar is large.
Bookkeeping versus dynamic vegetation models
There are two main methods. Bookkeeping models multiply observed land-use changes by per-hectare carbon factors derived from inventories. Dynamic global vegetation models simulate how vegetation responds to land-use change and to changes in CO₂ and climate. The two methods give different answers — the dynamic models tend to assign more of the residual sink to managed land, while the bookkeeping approach is closer to what national greenhouse-gas inventories report. The Intergovernmental Panel on Climate Change discusses this gap explicitly in its Working Group I assessment reports.
National inventories and the “managed-land” convention
Country reports under the UNFCCC use a managed-land convention: emissions and removals are counted only on land considered managed. This excludes much of the natural land sink in countries with large unmanaged forests, but includes the full forest sink in countries that classify all forest as managed. Subtracting national inventory totals from the global atmospheric budget therefore leaves a residual that is mostly an artefact of the convention, not a missing source. For consistent international comparisons, use the same data product for every country.
Tropical hotspots
The bulk of recent gross deforestation is concentrated in a small number of tropical biomes: the Amazon, the Congo Basin, the forests of insular Southeast Asia, and parts of West Africa and Mesoamerica. Outside these regions, total forest area is roughly stable or rising. Year-to-year variation in the global headline is therefore dominated by what happens in those hotspots, especially in fire years.
Common pitfalls when reading deforestation charts
Treating tree-cover loss as deforestation
A pixel of canopy lost in a managed timber rotation will regrow within a decade and is not deforestation in any standard definition. Tree-cover loss aggregates rotations, fires, hurricanes and conversions; only a fraction is irreversible.
Comparing FAO and GFW directly
FAO uses a forest-land definition; GFW uses a tree-cover definition. Pasted side by side they imply different country rankings. If you need a country comparison, pick one product and use it for every country.
Single-year headlines
Tropical deforestation has large interannual variation driven by El Niño, fire seasons and policy enforcement cycles. A single “X% increase from last year” rarely captures the underlying trend; multi-year means smooth out the noise.
Confusing gross emissions with net climate impact
Gross deforestation emissions are large; the net land-use flux is smaller because forests elsewhere are absorbing carbon. Both numbers are interesting but they answer different questions. Reports that quote gross loss alongside net atmospheric impact can sound contradictory when they are simply about different definitions.
Sources
The standard references on forest area and deforestation are the FAO Forest Resources Assessment, Global Forest Watch (using University of Maryland Hansen et al. tree-cover-loss data), the European Commission JRC Tropical Moist Forest dataset, and country forest inventories where they exist. For the carbon side, the Global Carbon Project’s annual Global Carbon Budget gives the central estimates and uncertainty ranges; the IPCC AR6 Working Group I chapters on the global carbon cycle explain how the bookkeeping and dynamic-model strands relate.
Land-use-change emissions sit alongside the fossil-fuel side of the carbon ledger discussed in emissions by country and historical emissions; a complete picture of national responsibility usually requires reading both kinds of accounting together.