Accessibility Settings

color options

monochrome muted color dark

reading tools

isolation ruler
Aerial View of Deforestation Area For Agricultural Land Drone.
Aerial View of Deforestation Area For Agricultural Land Drone.

Image: Shutterstock



Harnessing the Power of Global Forest Watch for Data-Driven Reporting on Land Cover Change

Picture it: 2013. Macklemore’s “Thrift Shop” was No. 1 on the Billboard Hot 100; travel between the US and Cuba was still heavily restricted; the Star Wars sequel trilogy was but a glimmer in Disney’s eye.

And if you wanted to include forest loss numbers in your reporting, your options were limited to annual reports that used oft-dubious data self-reported by governments and the occasional peer-reviewed unicorn. Regardless of source, this data was a year old at best.

But in 2014 the fog began to lift when the World Resources Institute released Global Forest Watch, an interactive, free-to-use online platform that visualizes and analyzes land cover change datasets around the world.

Debuting with its flagship tree cover loss dataset and a smattering of context layers, the platform has blossomed into a comprehensive portal that connects the public to more than four dozen global, national and regional datasets.

As an editor who specializes in data-driven coverage of land cover change, I’ve been using Global Forest Watch in my work at Mongabay since its debut 10 years ago.

I am not speaking hyperbolically when I say that it is the most powerful tool in my reporting and editing kit. I use it every day, and most of the hundreds of stories I’ve produced over the past decade would not have been possible without it.

While all of the platform’s layers have their strengths, I’d like to share the ones that I’ve found to be most helpful to me as an editor and journalist.

Tree Cover Loss, Deforestation Alerts

Global Forest Watch’s forest change datasets are its bread and butter.

The “historical” tree cover loss layer has been around since the platform’s inception and currently includes data for 2001–2022 that can be analyzed for the area affected (i.e., how many hectares were deforested in a given area over a given period).

For more recent data, toggle the integrated deforestation alerts layer, which spans the past two years and is updated (drumroll please) daily. This layer is currently limited to the tropics and is broadcast as “alerts” that aren’t user-analyzable for area affected.

Tree cover loss data on a mosaic satellite image of Brazil by Global Forest Watch

Tree cover loss data on a mosaic satellite image of Brazil, captured in June 2024 by Planet Labs. Image: Screenshot, Global Forest Watch

Both datasets come from the Global Land Analysis and Discovery lab at the University of Maryland, which uses a flock of satellites to detect clearance events in forest canopies around the world. This data is standardized, objective, and transparent. Every year, GLAD lab analysts and researchers crunch the integrated deforestation data, which they then use to update the tree cover loss layer.

The platform’s “analysis” tool allows users to analyze areas of the map by uploading geospatial files or drawing shapes on the map itself.

Several layers also allow you to analyze them by clicking on their shapes (e.g., protected areas such as national parks). You can do this for administrative areas that are baked into the tool’s base map as well.

While the integrated deforestation layer’s alerts can’t be analyzed for area affected, they are very useful for finding and keeping tabs on areas of ongoing forest loss. Their resolution is currently 10 square meters, allowing detection of even small-scale deforestation events such as logging roads.

Satellite Imagery

Now, no large dataset is immune to error. Occasionally satellites will record false positives (i.e., forest loss where no forest loss actually occurred). This is more likely to happen with recent integrated deforestation alerts that are so new only one satellite pass has detected them.

Both tree cover loss and integrated deforestation layers may also pick up natural forest loss events such as landslides, as well as the harvesting of tree plantations, which isn’t particularly useful for those of us whose reporting interests lie in anthropogenic deforestation.

Luckily, Global Forest Watch has another powerful tool to help “skytruth” its land cover change data: satellite imagery.

This imagery consists of several layers. My favorites are Landsat 8/Sentinel 2, which updates daily, and the monthly mosaics captured by Planet Labs.

Why would I use monthly sat images if daily ones are available? Often there is too much cloud cover for the dailies to reveal much — particularly during rainy seasons. Planet Labs filters this out and stitches together cloud-free (or nearly cloud-free) images that show a clearer picture of what happened on the ground in a given month.

Together, the platform’s forest loss datasets and satellite imagery layers are a potent combination for finding and confirming deforestation events.

Where before we were beholden to potentially unreliable info and had no way to ground truth other than by traveling to remote sites in person, we can now detect and confirm potential stories from our laptops and phones.

And while we at Mongabay do prefer to conduct investigations in the field when possible, Global Forest Watch helps us winnow out false leads, while also allowing us to remotely and quickly cover stories for which on-the-ground reporting isn’t feasible.

Many Layers of Context

In addition to using Global Forest Watch to detect and verify deforestation events, it’s also a useful source for context.

It includes many layers that shed light on why a place is important and what’s at stake if it continues to lose its forests.

The context layer I use most often is probably primary forests. That’s because other land cover change datasets don’t discern between planted forest — e.g., tree fiber or oil palm plantations — and natural forest; the satellites simply detect changes in canopy coverage, regardless of what kind of canopy it is, and flag them as tree loss.

In places such as Southeast Asia where tree crops are common, simply using land cover change data alone can result in false positives and inflated, inaccurate deforestation figures if they’re picking up plantation clearing and not actual forest loss.

This is where the primary forests layer comes in. It delineates areas that were covered in primary forest as of 2001 — so that, when toggled on in concert with other layers, you can be sure those pink pixels mean deforestation.

Next up is intact forest landscapes, which are areas of native vegetation that are extensive and undisturbed enough to retain their original biodiversity levels. Basically, integrated forest landscapes are extra-primary forest.

Also useful is the protected areas layer. Deforestation is not allowed in many protected areas, so using this layer is a great way to identify potentially illegal activity. (Caution: Protected area geospatial data is outdated for some locations, such as Madagascar, so I recommend confirming boundary extent with other sources.)

The Alliance for Zero Extinction sites layer, while somewhat of an awkward mouthful, is a real boon for quickly identifying areas where endemic, endangered species live. In other words, a good hook.

Finally, I’d like to give a shoutout to fire alerts. Global Forest Watch uses a powerful satellite dataset from NASA’s Visible Infrared Imaging Radiometer Suite that is updated daily and is crucial for Mongabay’s fire reporting. The fire dataset detects heat, which means it isn’t hindered by smoke or clouds, and allows us to watch fires start and spread in near-real time.

All signs are pointing to a long, early, destructive fire season in 2024, so I expect we’ll be leaning on this dataset extra hard this year.

As misinformation proliferates and productivity expectations rise, it can be hard to know what info sources to trust as a journalist.

But the standardized, transparent datasets of Global Forest Watch offer a reliable — and dare I say enjoyable? — path to good reporting.

And your editor will thank you for the easy fact-checking.

This article was originally published by the Society of Environmental Journalists online journal and is reprinted here with permission.  

Morgan Erickson-Davis is a senior editor at Mongabay. She focuses on data-driven coverage of tropical deforestation, which allows her to combine her interests in maps, science, and writing. Before Mongabay, Erickson-Davis studied biology and environmental science and took the path of most adventure, from working in river restoration and harvesting donor corneas in Montana to monitoring bycatch on commercial fishing vessels out of Hawaii and Tahiti. She currently lives in Minneapolis with a menagerie of pets and her partner, Frank.

Republish our articles for free, online or in print, under a Creative Commons license.

Republish this article

Material from GIJN’s website is generally available for republication under a Creative Commons Attribution-NonCommercial 4.0 International license. Images usually are published under a different license, so we advise you to use alternatives or contact us regarding permission. Here are our full terms for republication. You must credit the author, link to the original story, and name GIJN as the first publisher. For any queries or to send us a courtesy republication note, write to

Read Next

Reporting Tools & Tips

Protecting Forests with FOI Laws

Every year the world loses 13 million hectares (32 million acres) of forests, an area about the size of Greece. A critical way to stem this forest loss is to make concessions data more transparent. This information is critical as it gives journalists, civil society groups and the public the information they need to hold the government and the private sector accountable for deforestation.