Comment

Forests ยท Monitoring guide

Deforestation Monitoring Tools: 10 Methods Compared

The world lost 4.3 million hectares of tropical primary forest in 2025 (WRI). Here is what each monitoring tool sees, how often it updates, where it falls short, and how to combine satellite screening with field checks.

Authored by Farmonaut Research & Content Team
Reviewed and verified by Farmonaut Agricultural Experts Team
Updated 12 November 2026 ยท 15 min read
Sentinel-2 Satellite View Of Farmland And Boreal Woodlots Near Peace River, Alberta, Canada
Sentinel-2 view of Farmland and boreal woodlots near Peace River, Alberta, Canada. Contains modified Copernicus Sentinel data, processed in Google Earth Engine.

Key takeaways

  • Gross deforestation slowed from 17.6 million ha a year in 1990-2000 to 10.9 million ha a year in 2015-2025 (FAO FRA 2025).
  • Tropical primary forest loss fell 36% to 4.3 million ha in 2025 after a fire-driven record of 6.7 million ha in 2024 (WRI).
  • Global Forest Watch combines four alert systems: GLAD-L (30 m), GLAD-S2 (10 m), RADD radar (10 m, sees through cloud) and DIST-ALERT.
  • In Africa, forest loss was 18% lower in the first two years of GLAD alerts where managers subscribed (study of 22 countries, 2021).
  • The EU Deforestation Regulation applies from 30 December 2026 for large and medium operators (European Commission).
In this article
  1. Deforestation monitoring tools compared
  2. How much forest is still being lost?
  3. 1. Satellite remote sensing
  4. 2. Near-real-time deforestation alerts
  5. 3. AI and machine learning
  6. 4. National forest monitoring systems
  7. 5. Independent forest monitoring and audits
  8. 6. Online forest monitoring platforms
  9. 7. Commercial geospatial data providers
  10. 8. Drones, LiDAR and high carbon stock mapping
  11. 9. Ground sensors, field apps and GPS/GIS tools
  12. 10. Supply-chain traceability and due diligence records
  13. How to choose deforestation monitoring tools
  14. What farm and forest owners can do

The main deforestation monitoring tools today are satellite images, near-real-time alert systems, AI models that read those images, national forest inventories, independent audits, online platforms such as Global Forest Watch, commercial image providers, drones and LiDAR, ground sensors with GPS field apps, and supply-chain traceability records. Each answers a different question: where trees were lost, how fast you hear about it, who cleared them and whether a product is linked to that land.

The need is still large. The world lost 4.3 million hectares of tropical primary forest in 2025, 36% less than in 2024 but still 46% more than a decade earlier, according to the WRI Global Forest Review (2026). This guide compares ten tools and methods with dated, sourced numbers, shows what each one can and cannot see, and gives a calculator to compare field patrols with satellite screening for your own land.

Deforestation monitoring tools compared

The table sums up the ten tools. Resolution and timing come from the sources named in each section below. “Best for” is our plain summary of the uses each source describes.

Tool or method What it uses How often Best for Main limit
1. Satellite remote sensing Optical images (Landsat 30 m, Sentinel-2 10 m) and radar (Sentinel-1 10 m) Every few days, cloud permitting Seeing forest cover change over large areas Clouds and smoke block optical images
2. Near-real-time alerts GLAD-L, GLAD-S2, RADD and DIST-ALERT, combined on Global Forest Watch As new images arrive Early warning of new clearing Alerts need a field or image check
3. AI and machine learning Models trained on satellite images Depends on the image feed Sorting large image volumes, mapping risk Only as good as its training data
4. National forest monitoring systems Satellite maps, field plots, statistics Yearly to every few years Official national figures and climate reporting Slow; built for totals, not single sites
5. Independent forest monitoring Third-party field visits, documents, images Per audit Checking that rules are followed Costly; covers samples, not every hectare
6. Online monitoring platforms Global datasets, maps, dashboards Weekly to yearly layers Open access to loss data and alerts Global layers can miss very small clearings
7. Commercial image providers Very high-resolution optical and radar images On demand Close-up checks of a specific site Cost per image
8. Drones and LiDAR Aerial photos, laser scanning Per flight Tree height, canopy gaps, carbon mapping Small area per flight
9. Ground sensors and GPS/GIS tools Field apps, sensors, GPS points, permanent plots Continuous to every 5-10 years Proof on the ground, fire and soil data Local coverage only
10. Traceability records Plot coordinates, ledgers, due diligence files Per shipment Linking products to land that was not cleared Records are only as true as their inputs

How much forest is still being lost?

Gross deforestation, meaning forest turned into farms, roads, mines or towns, is slowing but has not stopped. The FAO Global Forest Resources Assessment 2025 estimates that 489 million hectares of forest were lost to deforestation between 1990 and 2025. The yearly rate fell from 17.6 million hectares in 1990-2000 to 13.6 million in 2000-2015 and 10.9 million in 2015-2025.

Chart

World deforestation rate by period

Million hectares of forest converted to other land uses per year (gross deforestation)

0.05.010.015.020.01990-20002000-20152015-2025Deforestation 10.9Source: FAO, Global Forest Resources Assessment 2025, chapter 2 (forest extent and change).
0.05.010.015.020.01990-20002000-20152015-2025Deforestation10.9Source: FAO, Global Forest Resources Assessment 2025, chapter2 (forest extent and change).
The yearly rate fell by about 38% from the 1990s to 2015-2025, but almost 11 million hectares a year are still cleared.

Net loss is a different number. It subtracts new forest from planting and natural regrowth. FAO puts net loss at 4.12 million hectares a year in 2015-2025, up from 3.68 million in 2000-2015, because forest gain slowed in countries such as China and Canada. Good deforestation monitoring tools keep these two numbers apart: a new plantation does not replace a cleared old-growth stand.

Fire is now a large part of the picture. In 2024, tropical primary forest loss reached 6.7 million hectares, nearly twice the 2023 figure, and fires caused nearly half of it, according to a WRI release of May 2025. In 2025, fires caused 42% of the 25.5 million hectares of global tree cover loss, including 5.3 million hectares in Canada, the WRI review states. Note that tree cover loss is not the same as deforestation: burned forest often grows back.

1. Satellite remote sensing

Satellite images are the base layer for almost every other tool on this list. NASA’s Landsat satellites map tree cover loss at 30-meter resolution, while the European Sentinel-2 satellites give 10-meter optical images and Sentinel-1 gives 10-meter radar images, according to Global Forest Watch’s guide to its integrated alerts (updated January 2026).

Chart

Pixel size of the main forest alert systems

Meters per pixel; smaller pixels catch smaller clearings

0 m8 m15 m22 m30 m30 mGLAD-L (Landsat)10 mGLAD-S2(Sentinel-2)10 mRADD (Sentinel-1radar)Source: Global Forest Watch, Explore GFW’s Updated Integrated Disturbance Alerts (updated January 2026).
0 m8 m15 m22 m30 mGLAD-L (Landsat)30 mGLAD-S2 (Sentinel-2)10 mRADD (Sentinel-1 radar)10 mSource: Global Forest Watch, Explore GFW’s Updated IntegratedDisturbance Alerts (updated January 2026).
The Sentinel-based systems see changes nine times smaller in area than the Landsat-based GLAD-L.

Optical images show colour and plant vigour, much like a photo. They are blocked by cloud and smoke. Radar sends its own signal and sees the ground through cloud, which matters in the wet tropics and in fire season. Using both gives a more complete record.

  • Good for: mapping forest cover, clear-cuts, new roads and land conversion across large areas.
  • Watch out for: 30-meter pixels can miss narrow logging gaps; 10-meter data captures smaller changes.
  • Cost: Landsat and Sentinel data are open to the public; processing and analysis cost time or money.

Accuracy depends on the method and the place. Guyana’s national monitoring system, which uses mainly Sentinel-2 images supplemented by Landsat, reported 99.16% overall accuracy for its forest/non-forest change mapping over 1990-2020 in an independent check, with forest loss missed in 5.56% of the non-forest reference samples (Guyana MRVS accuracy assessment, year 2020). That is one country’s result, not a general figure for all tools.

A walk-through of a web app that shows satellite images and vegetation indices for mapped fields.

โ–ถ A walk-through of a web app that shows satellite images and vegetation indices for mapped fields. ยท 8:02

2. Near-real-time deforestation alerts

Alert systems scan each new image and flag pixels where vegetation has just been lost. Global Forest Watch now combines four systems in one “integrated disturbance alerts” layer: GLAD-L (Landsat, 30 m), GLAD-S2 (Sentinel-2, 10 m, since May 2021), RADD (Sentinel-1 radar, 10 m, run by Wageningen University) and DIST-ALERT, added in January 2026, which covers disturbance in all vegetation, inside and outside forests.

Alerts do not stop clearing by themselves. Someone has to act on them. A study of 22 tropical countries found that, in Africa, forest loss was 18% lower in the first two years of GLAD alerts where forest managers subscribed to them, compared with the five years before, according to an Oregon State University summary of the study (January 2021). The same study did not find a clear effect in South America or Asia, which shows that alerts work best when they are tied to a response.

  • Set up area alerts for your land so you hear about changes within days, not months.
  • Check each alert on a recent image before sending a crew.
  • Keep a log of alerts and the action taken; buyers and certifiers may ask for it.

3. AI and machine learning

Satellites now produce far more images than people can read. Machine-learning models sort them: they separate forest from cropland, tell a clear-cut from a seasonal change, and estimate where clearing is likely next, based on factors such as distance to roads and past loss.

AI is behind most modern alert systems and global maps, including the tree cover loss data that the University of Maryland’s GLAD lab produces for Global Forest Watch. WRI now also offers Global Nature Watch, an AI-powered platform that answers questions about this forest data in plain language, according to the 2026 Global Forest Review.

Two cautions. First, a model trained in one forest type can do worse in another, so check results against images or field visits. Second, “AI” on a product page does not tell you its accuracy. Ask for a published accuracy test for your region.

How satellites, AI and drones are used together to watch land and crops at scale.

โ–ถ How satellites, AI and drones are used together to watch land and crops at scale. ยท 13:27

4. National forest monitoring systems

National forest monitoring systems combine satellite maps, field plots and statistics to give official figures. Governments use them for policy, law enforcement and climate reporting, including REDD+ results-based payments. Their strength is a careful, repeatable method; their weakness is speed, as most report once a year or less.

In the United States, the Forest Inventory and Analysis program of the USDA Forest Service runs a national network of permanent plots remeasured every five to ten years depending on location, and adds remote sensing to its field work, according to the USDA FIA programme page. Canada reports deforestation by cause each year in its State of Canada’s Forests report.

The Canadian data show that clearing in a temperate or boreal country is mostly not about logging. Natural Resources Canada estimates 49,248 hectares of deforestation in 2023, led by agriculture, then mining, oil and gas, then built-up land. Areas that are harvested and replanted count as forest, not deforestation.

Chart

Deforestation in Canada by cause, 2023

Hectares of forest converted to other uses, four largest causes; total for all causes 49,248 ha

06,25012,50018,75025,000Agriculture22,460Mining, oil and gas14,757Built-up9,537Hydroelectric1,462Source: Natural Resources Canada, The State of Canada’s Forests: Annual Report 2025.
06,25012,50018,75025,000Agriculture22,460Mining, oil and gas14,757Built-up9,537Hydroelectric1,462Source: Natural Resources Canada, The State of Canada’sForests: Annual Report 2025.
Farming caused the most clearing; harvested forest that is replanted is not counted as deforestation.

5. Independent forest monitoring and audits

Independent forest monitoring means a third party, such as a civil-society group, auditor or certification body, checks whether forest laws and permits are being followed. Monitors visit sites, compare permits with what is on the ground and review satellite images.

  • It adds trust, because the checker is not the company or agency being checked.
  • It finds problems a map cannot show, such as logging outside permit dates or wrong species.
  • It is costly and covers samples, so it works best when satellite alerts point monitors to the right places.

For farm and timber businesses, the practical version is a documented audit trail: maps of each plot, dated images, permits and the response to every alert.

6. Online forest monitoring platforms

Online platforms put global data in a browser. Global Forest Watch, run by the World Resources Institute, shows yearly tree cover loss from the University of Maryland, near-real-time alerts and drivers of loss. Users can draw an area, subscribe to alerts and download data. WRI also hosts the yearly Global Forest Review, which explains each year’s trends.

Other platforms mix the same open satellite data with field records, farm boundaries and reports for a company’s own sites. When you compare platforms, ask:

  1. Which satellites and resolution does it use, and how often does it update?
  2. Does it separate natural forest from plantations and crops?
  3. Can you upload your own plot boundaries and export evidence for audits?
  4. Does it state its accuracy and the date of its forest baseline?

7. Commercial geospatial data providers

Commercial providers sell very high-resolution optical images, often under one meter, and radar images on demand. A buyer can order a fresh image of one site to confirm an alert, check a dispute or document conditions before and after work.

These images are most useful as a second step. Use open 10-30 m data to screen large areas, then buy close-up images only where something has changed. That keeps costs down and gives you clear evidence where it matters.

8. Drones, LiDAR and high carbon stock mapping

Drones fly below the clouds and capture centimetre-level photos of a plot. They are good for checking an alert, measuring a small clearing, counting replanted trees or showing canopy damage after a storm. Their limit is area: one flight covers a small block compared with a satellite scene.

LiDAR, a laser scanner on a plane or drone, measures tree height and canopy structure. That makes it useful for estimating forest carbon and for “high carbon stock” mapping, which separates dense forest that should be kept from degraded land that could be used. Many projects use LiDAR samples to train satellite models over a wider area.

9. Ground sensors, field apps and GPS/GIS tools

Ground tools give the proof that remote tools cannot. They include GPS field apps for marking plot corners and photos, permanent inventory plots, and sensors for heat, smoke or soil moisture that can warn of fire. GIS software brings all of these together with satellite maps.

i-Tree, a set of tools from the USDA Forest Service and partners, estimates the benefits of trees, mostly in towns and around buildings rather than wild forest. Its website states that a USDA Forest Service policy change has halted its traditional funding indefinitely, so check its current support status before you rely on it.

  • Use GPS points and photos to confirm satellite alerts.
  • Use permanent plots to measure regrowth and tree health over years.
  • Use sensors where fire risk is high and response time matters.

10. Supply-chain traceability and due diligence records

Traceability tools link a product, such as timber, beef, soy, cocoa, coffee, palm oil or rubber, to the plot where it was grown. They store plot coordinates, dates and the satellite checks that show the land was not cleared after a cut-off date. Some use digital ledgers or blockchain so records cannot be changed later.

Rules are driving demand. The EU Deforestation Regulation (EUDR) applies from 30 December 2026 to products such as cattle, cocoa, coffee, palm oil, rubber, soy and wood placed on or exported from the EU market, according to the European Commission’s EUDR implementation page. The Commission lists 30 June 2027 for micro and small operators, except those already covered by the EU Timber Regulation, and says the EUDR replaces that older timber rule. Check the current rules with the Commission or a trade adviser before you ship, as dates have changed before.

How to choose deforestation monitoring tools

Start with the question you need to answer, then pick the cheapest tool that answers it well:

  1. “Has anything changed on my land?” Open satellite data plus alerts.
  2. “What exactly happened here?” A recent high-resolution image, a drone flight or a field visit.
  3. “Can I prove my product is clean?” Plot maps, a dated forest baseline and an alert log, kept as due diligence records.
  4. “How much carbon or timber is there?” Field plots and LiDAR, scaled up with satellite models.

Most owners combine screening from space with field checks only where something changed. The calculator below compares that mix with patrol-only monitoring. All costs are your inputs.

Try the numbers for your own farm below.

Calculator

Patrols only, or satellite screening plus spot checks?

Compare yearly monitoring costs for your own land.

ha

US$

rounds

US$

%
1,000US$ a year saved by screening first (negative means patrols alone cost less)
  • Patrols only: 12,000 US$/yr
  • Satellite screening: 8,000 US$/yr
  • Spot checks on the ground: 3,000 US$/yr
  • Screening plus spot checks: 11,000 US$/yr

Example values only (prices are examples, use today’s price in US$), near break-even, not a quote or a benchmark; every number is your input. Patrol cost is your full cost per hectare per visit (labour, fuel, vehicle). Satellite cost is your yearly cost per hectare for a monitoring service or staff time. The spot-check share is the part of your land you still visit each round after screening.

What farm and forest owners can do

  1. Map every field and woodlot boundary once, with GPS or on a satellite map.
  2. Record a dated forest baseline, such as a satellite image from before any cut-off date that applies to you.
  3. Watch your land through the season so you notice clearing, fire, storm damage or encroachment early.
  4. Check each change on the ground and note what you did.
  5. Keep permits, maps and alert logs together for buyers, lenders and insurers.

Farmonaut is one option for step 3: its satellite monitoring shows vegetation maps of the fields and plantations you draw, through the web and mobile app, on Android and on iOS. Teams that manage many sites can use the admin dashboard, and developers can add the same data to their own systems through the satellite API and its developer documentation.

Related pages on our site: sustainable forestry benefits ยท California forestry management ยท product traceability ยท carbon footprinting ยท fleet management ยท crop loans and insurance ยท agricultural extension

Step-by-step tutorial: draw a field and read its satellite health maps in the web app.

โ–ถ Step-by-step tutorial: draw a field and read its satellite health maps in the web app. ยท 8:43

Limits of remote forest monitoring

No single tool sees everything. Clouds and smoke delay optical images: WRI notes that some fire loss in its 2025 data happened in late 2024 but was hidden by smoke and haze at the time. Coarse pixels miss small or selective logging. Global maps may count a harvested plantation as loss and a new plantation as gain, which is why FAO reports deforestation and net change separately.

Alerts are signals, not proof. A flagged pixel can be a harvest under permit, a storm, a fire or a mapping error. Treat every alert as a reason to look, then confirm it with an image or a visit. For legal or compliance decisions, check the local rules and get professional advice.

How we checked the numbers

We used the FAO Global Forest Resources Assessment 2025 for deforestation and net-change rates and kept its periods. Tree cover loss figures come from WRI’s Global Forest Review and press release, based on University of Maryland data; tree cover loss includes fire and harvest and is not the same as deforestation. Satellite resolutions come from Global Forest Watch. Canada’s data come from Natural Resources Canada’s 2025 report, US inventory details from the USDA Forest Service, and EUDR dates from the European Commission. Accuracy figures are reported for the country and year each source covers. Every source is listed at the end of the article.

FAQ

What are the best tools for monitoring deforestation in near real time?

Satellite alert systems such as GLAD-L, GLAD-S2 and RADD, now combined with DIST-ALERT in Global Forest Watch’s integrated disturbance alerts. Radar-based RADD can see through cloud. Confirm each alert with a recent image or a field check.

What is the most reliable technology for tracking responsible forest practices?

No single technology is enough. The most dependable approach combines satellite monitoring for coverage, field checks or independent audits for proof, and dated records that link products to mapped plots.

Where can organizations access trusted forest monitoring tools?

Global Forest Watch (WRI) offers open maps, alerts and downloads. National agencies, such as the USDA Forest Service and Natural Resources Canada, publish official inventory and deforestation data. Commercial platforms and image providers add site-level detail.

Can satellites see through clouds?

Optical satellites such as Landsat and Sentinel-2 cannot. Radar satellites such as Sentinel-1 can, which is why radar alerts are useful in cloudy tropical regions and during smoky fire seasons.

Is tree cover loss the same as deforestation?

No. Tree cover loss counts any removal of tree canopy, including fire, storms and harvests that regrow. Deforestation, in FAO’s definition, is forest turned into another land use, such as farmland or roads.

Authored by: Farmonaut Research & Content Team

Reviewed and verified by: Farmonaut Agricultural Experts Team

Reviewed November 2026 against the FAO Global Forest Resources Assessment 2025, the WRI Global Forest Review (2026) and May 2025 release, Global Forest Watch’s integrated alerts guide (January 2026), Natural Resources Canada’s 2025 forest report, the USDA Forest Service FIA program, Guyana’s MRVS accuracy report and the European Commission’s EUDR pages.

Farmonaut satellite monitoring

Watch your fields and woodlots from space

Draw your fields, plantations and tree blocks once and Farmonaut shows satellite vegetation maps for them through the season, on web, Android or iOS.

Open The Farmonaut Crop Monitoring Web AppGet The Farmonaut App On Google PlayGet The Farmonaut App On The App Store
Farmonaut Web App With A Satellite Crop Health Map

Sources

  1. WRI, Global Forest Review: Tropical rainforest loss slowed in 2025, but fire is a growing threat (2026): gfr.wri.org
  2. WRI, Global forest loss shatters records in 2024, fueled by massive fires (press release, May 2025): wri.org
  3. FAO, Global Forest Resources Assessment 2025, chapter 2: Forest extent and change: openknowledge.fao.org
  4. Global Forest Watch, Explore GFW’s updated integrated disturbance alerts (updated January 2026): globalforestwatch.org
  5. Oregon State University, Alert system shows potential for reducing deforestation (January 2021): news.oregonstate.edu
  6. Guyana Forestry Commission, MRVS and accuracy assessment report, year 2020 (UNFCCC REDD+ platform): redd.unfccc.int
  7. USDA Forest Service, Forest Inventory and Analysis program: research.fs.usda.gov
  8. Natural Resources Canada, The State of Canada’s Forests: Annual Report 2025: natural-resources.canada.ca
  9. i-Tree Tools, home page and funding notice (USDA Forest Service and partners): itreetools.org
  10. European Commission Green Forum, Implementing the EU Deforestation Regulation (EUDR): green-forum.ec.europa.eu
  11. European Commission, EU Deforestation Regulation: entry into application dates: environment.ec.europa.eu
  12. FAO, Sustainable forest management: fao.org
  13. Android: play.google.com
  14. iOS: apps.apple.com


Discussion

Questions, corrections or field experience? Share them below. First-time comments are reviewed before they appear, and your email is never published.

No comments yet. Be the first to add a question or your experience.

Leave a comment

Farmonaut Farmonaut Trusted by 200,000+ users and 100+ businesses 200,000+ users trust us Get started