A Copernicus Sentinel-2B satellite map of South Sudan shows the tropical forests, swamps and grassland that comprise the majority of the country's terrain.
Photo Credit: European Space Agency
(CC BY-SA 4.0)
Scientific Frontline: "At a Glance" Summary
- Global Dataset Discrepancy: A comparative analysis of eight major global forest datasets reveals that they concur on the identification of forest locations only 26% of the time, highlighting severe inconsistencies in digital baselines.
- Methodological Divergence: The study attributes these variations to differing technical definitions of "forest"—specifically regarding canopy cover thresholds (e.g., 10% vs. 50%)—and the specific remote sensing technologies employed to interpret land use.
- Socioeconomic Impact Data: In a specific case study of India, estimates of the population living in poverty near forests ranged dramatically from 23 million to 252 million, depending solely on the forest map utilized.
- Scale of Uncertainty: Definitional variances result in uncertainty factors of up to 10, capable of instantly reclassifying millions of hectares between "forest" and "non-forest" status in global inventories.
- Implications for Climate Policy: These discrepancies undermine the reliability of carbon storage estimates and nature-based markets, posing risks to the accurate allocation of climate finance and the validation of conservation policies.
- Proposed Resolution: The researchers introduced a decision-support flowchart to assist stakeholders in dataset selection and advocated for hybrid models that validate satellite imagery with ground-level data to improve accuracy.
For decades, global efforts to combat climate change and protect biodiversity have relied on a high-tech promise: that satellite-derived maps can tell us exactly where the world's forests are.
But a new study from the University of Notre Dame reveals that these digital baselines are often in sharp disagreement, creating confusion that threatens to undermine effective climate funding and international development efforts. Because these maps determine everything from carbon storage estimates to the enactment of conservation policies, even small discrepancies can have serious consequences for both people and the planet.
The study, co-authored by Daniel C. Miller, the Coyle Mission Collegiate Associate Professor at Notre Dame’s Keough School of Global Affairs, reveals major differences among the world’s most widely used forest datasets. When comparing eight of the most popular datasets, the researchers found that their identification of forest locations concurred only 26 percent of the time.
The problem stems from how different researchers define “forest” and the digital technology they use to view forests, said Miller, a core faculty affiliate of the Keough School’s Pulte Institute for Global Development.
“When land is viewed from the sky, it’s difficult to know at a global scale whether something is a forest or not,” he said. “Some might consider a small patch of trees to be a forest, but for others only a large, dense area of trees will count.”
The study found that the discrepancy among datasets creates major uncertainty, sometimes by a factor of 10. For example, some maps might count a savanna interspersed with trees as forest based on a 10 percent canopy cover threshold, while others require 50 percent. These small definitional differences can flip millions of hectares from “forest” to “non-forest” in an instant.
The researchers used case studies from Brazil, India and Kenya to show how these digital maps affect human lives and global policy challenges. In India, for example, the estimated number of people living in poverty near forests fluctuated from 23 million to 252 million depending solely on which map was used.
A framework for the future
To help policymakers, journalists and others navigate this “digital wilderness,” the researchers created a flowchart to help non-experts determine which data sets are most appropriate for their specific region or goal. Miller said that future work should integrate hybrid data that combines on-the-ground views with satellite data.
“By bridging the gap between satellite technology and on-the-ground reality, we can provide more accurate, inclusive data that truly supports both the planet and the people who protect it,” he said.
Real-world stakes
Miller warned that if forest definitions continue to vary, countries could overestimate — or dangerously underestimate — their carbon sequestration potential.
“It could make a real difference in climate finance commitments and how much vulnerable communities gain,” he said.
Co-author Sarah Castle of the University of Wisconsin–Madison emphasized that accurate data is a prerequisite for trust in global markets for environmental goods like carbon and timber and underscored the need for better standardization and transparency so the global community can agree on common reporting metrics.
“If we cannot establish a reliable baseline for forest area, it undermines trust in nature-based markets and makes it nearly impossible to accurately measure the role forests play in supporting people’s lives and livelihoods,” Castle said.
Funding: Funding for the study was provided by Notre Dame Research and the Keough School’s Pulte Institute for Global Development through the University of Notre Dame Poverty Initiative.
Published in journal: OneEarth
Authors: Sarah E. Castle, Peter Newton, Johan A. Oldekop, Kathy Baylis, and Daniel C. Miller
Source/Credit: University of Notre Dame | Renée LaReau
Reference Number: env011226_01
.jpg)