Satellite imagery makes deforestation before and after comparisons possible at a scale that ground surveys never could. A single Sentinel-2 satellite pass captures 290 km of terrain at 10-meter resolution, and the archive now stretches back over a decade. Comparing images from two dates reveals exactly where forests stood and where they have been cleared.
This article explains how satellites detect deforestation, which regions show the most dramatic changes, and why the December 31, 2020 cutoff date in the EU Deforestation Regulation makes before-and-after evidence a compliance requirement.
What Is Deforestation?
Deforestation is the permanent conversion of forest land to a non-forest use. It differs from forest degradation, where tree cover thins but the land remains classified as forest. The distinction matters for regulation and measurement: degradation is harder to detect from space and is not always captured in deforestation statistics.
The Food and Agriculture Organization defines forest as land spanning more than 0.5 hectares with trees taller than 5 meters and canopy cover above 10%. When that land is converted to agriculture, pasture, or urban development, it counts as deforestation. Selective logging that maintains canopy cover above 10% does not.
Global Forest Watch, run by the World Resources Institute, reported that the tropics lost 6.7 million hectares of primary forest in 2024, the highest annual figure on record. Fire accounted for a larger share than in previous years, driven by drought conditions across the Amazon, Southeast Asia, and Central Africa.
How Satellites Detect Deforestation
Satellite-based deforestation detection relies on comparing land surface reflectance across time. Healthy vegetation reflects strongly in the near-infrared (NIR) spectrum and absorbs visible red light. When trees are removed, NIR reflectance drops sharply and red reflectance increases. Algorithms flag pixels where this spectral signature shifts beyond a threshold.
Key Satellite Programs
Landsat (NASA/USGS) has provided continuous 30-meter resolution imagery since 1972, making it the longest satellite record available. The Hansen Global Forest Change dataset, built from Landsat data, maps tree cover loss annually at 30-meter resolution for the entire planet from 2000 onward.
Sentinel-2 (ESA) captures 10-meter resolution imagery every 5 days, a significant improvement over Landsat’s 16-day revisit cycle. Higher resolution and faster revisits allow earlier detection of small-scale clearing. Sentinel-2 data feeds Google’s Dynamic World land cover classification, which maps land cover types at near-real-time frequency.
MODIS (NASA) covers the entire Earth daily at 250-meter resolution. While too coarse for farm-level analysis, MODIS data powers fire detection systems like FIRMS (Fire Information for Resource Management System) and provides the earliest alerts for large-scale clearing events.
Synthetic Aperture Radar (SAR) from satellites like Sentinel-1 penetrates cloud cover, which is a persistent problem in tropical forests. SAR detects structural changes in the forest canopy regardless of weather conditions, making it essential for monitoring cloud-heavy regions like the Congo Basin and Borneo.
Change Detection Methods
The simplest deforestation before and after analysis compares two images from different dates. Analysts overlay a recent image on an older baseline and visually inspect areas where green forest has been replaced by bare soil, crops, or pasture. Automated systems do this at scale using spectral indices like NDVI (Normalized Difference Vegetation Index), which quantifies vegetation density as a number between -1 and +1. A drop from 0.7 to 0.2 at a given pixel indicates likely clearing.
More advanced methods use time-series analysis across hundreds of images to distinguish deforestation from seasonal variation. Algorithms like BFAST (Breaks For Additive Season and Trend) detect abrupt changes in the vegetation signal while filtering out normal seasonal fluctuations in leaf cover.
Deforestation Before and After: Key Regions
Satellite comparisons reveal the scale of forest loss most clearly in three tropical hotspots.
Brazilian Amazon
The Brazilian Amazon has lost roughly 17% of its original forest cover since systematic monitoring began. Brazil’s national space agency INPE recorded 5,796 km2 of clearing in the 2024-25 monitoring period, down from a peak of 13,235 km2 in 2020-21. Before-and-after imagery from the Xingu basin shows large rectangular clearings where cattle ranches replaced continuous forest canopy. The regularity of the clearing patterns, often visible as straight-edged blocks, distinguishes human-driven deforestation from natural forest dynamics.
Southeast Asia
Indonesia and Malaysia have experienced decades of forest conversion for oil palm plantations. Between 2001 and 2023, Indonesia lost over 28 million hectares of tree cover according to Global Forest Watch. Borneo before-and-after comparisons from Landsat show intact peat swamp forest in 2000 replaced by gridded oil palm plantations by 2020. The geometric pattern of plantation rows is unmistakable in satellite imagery.
Congo Basin
Central Africa’s Congo Basin contains the world’s second-largest tropical forest, but deforestation has accelerated in recent years. The Democratic Republic of Congo lost 1.3 million hectares of primary forest in 2023. Much of this clearing is smallholder-driven, creating a fragmented pattern of small clearings that gradually expand outward from roads and rivers. Cloud cover makes monitoring difficult, which is why SAR-based detection is essential for this region.
Explore Global Deforestation Hotspots
The satellite images above capture deforestation before and after at specific sites. The interactive map below shows where commodity-driven deforestation concentrates globally. Filter by commodity or region, and use the timeline slider to track how forest loss shifted between 2001 and 2022. Zoom in on the regions covered above to see how individual case studies fit into broader deforestation patterns.
The geographic concentration of forest loss along agricultural frontiers shaped the regulatory response. The European Union designated December 31, 2020 as the deforestation cutoff date under the EUDR, the subject of the next section.
The EUDR Cutoff: December 31, 2020
The EU Deforestation Regulation gives the deforestation before and after comparison a specific legal meaning. Under the EUDR, any company placing palm oil, soy, cocoa, coffee, rubber, cattle products, or wood on the EU market must prove that the product was not grown on land deforested after December 31, 2020.
This means operators need satellite evidence showing what a specific parcel of land looked like before and after that date. If land classified as forest on December 31, 2020 shows as agricultural land in a recent image, that parcel fails the EUDR test. Products sourced from it cannot legally enter the EU market.
The regulation requires GPS coordinates for every plot of origin. Farms larger than four hectares need polygon boundary mapping. These coordinates are cross-referenced against satellite imagery and land cover classification data to confirm that no deforestation occurred after the cutoff.
Platforms like Continuuiti’s LULC+ automate this comparison by pulling Sentinel-2 imagery and Dynamic World land cover data for any coordinate, then generating a year-by-year land cover timeline that shows whether forest cover existed on the cutoff date and whether it persists today.
Can Deforestation Be Reversed?
Reforestation and natural forest regrowth can restore tree cover on previously cleared land. Secondary forests now cover roughly 28% of formerly deforested tropical land, according to NASA research. However, secondary forests take 20 to 100 years to approach the biodiversity and carbon storage capacity of primary forest.
For EUDR purposes, reforestation does not retroactively make a parcel compliant. If forest was cleared after December 31, 2020, planting new trees on the same land does not satisfy the regulation. The rule tests whether deforestation occurred, not whether the land was subsequently replanted. This distinction makes the before-and-after satellite record a permanent compliance reference point.
EUDR compliance systems must therefore maintain historical satellite data for every sourcing location, creating a baseline record that regulators and auditors can verify independently.
Frequently Asked Questions
What does deforestation before and after look like from space?
Satellite imagery shows deforestation as a shift from dense green canopy (high near-infrared reflectance) to bare soil or agricultural land (low near-infrared reflectance). Before-and-after comparisons from programs like Landsat and Sentinel-2 reveal rectangular or fragmented clearing patterns where continuous forest previously stood.
Which satellites are used to detect deforestation?
Four main satellite systems detect deforestation: Landsat (30m resolution, since 1972), Sentinel-2 (10m resolution, 5-day revisit), MODIS (250m, daily coverage for large-scale alerts), and SAR satellites like Sentinel-1 that penetrate cloud cover in tropical regions.
How much deforestation happened in 2024?
The tropics lost 6.7 million hectares of primary forest in 2024, the highest annual total on record according to Global Forest Watch. Fire played a larger role than in previous years, driven by drought conditions across the Amazon, Southeast Asia, and Central Africa.
What is the EUDR December 2020 cutoff date?
The EU Deforestation Regulation uses December 31, 2020 as its cutoff date. Any product sourced from land that was deforested after this date cannot be placed on the EU market. Companies must use satellite imagery and GPS coordinates to prove their supply is deforestation-free relative to this date.
Can reforestation make deforested land EUDR-compliant?
No. Under the EUDR, reforestation does not retroactively make a parcel compliant. If forest was cleared after December 31, 2020, planting new trees on the same land does not satisfy the regulation. The rule tests whether deforestation occurred, not whether the land was later replanted.
What is the difference between deforestation and forest degradation?
Deforestation is the permanent conversion of forest land to a non-forest use, such as agriculture or urban development. Forest degradation is a reduction in tree cover density where the land still meets the definition of forest (above 10% canopy cover). Degradation is harder to detect from satellites and is not always captured in deforestation statistics.
Deforestation before and after satellite comparisons have moved from academic research to regulatory evidence. For companies sourcing commodities covered by the EUDR, maintaining a satellite-verified record of land cover at every sourcing location is now a compliance obligation backed by penalties of up to 4% of EU turnover.
