forest climate change trees ecosystem carbon storage
forest climate change trees ecosystem carbon storage

Researchers Find Flaws in High-Profile Study on Trees and Climate

Four independent groups say the work overestimates the carbon-absorbing benefits of global forest restoration, but the authors insist their original estimates are accurate.

katya katarina zimmer
Katarina Zimmer

After a year teaching an algorithm to differentiate between the echolocation calls of different bat species, Katarina decided she was simply too greedy to focus on one field. Following an internship with The Scientist in 2017, she has been happily freelancing for a number of publications, covering everything from climate change to oncology.

View full profile.

Learn about our editorial policies.

Oct 17, 2019

ABOVE: A study suggested that filling in treeless spaces with forest could store considerable amounts of carbon—and mitigate climate change.

In July, a high-profile study in Science estimated that Earth has space for another 0.9 billion hectares’ worth of trees—an area the size of the continental US. Simply allowing forests to recover in those areas would suck more than 200 gigatons of carbon out of the atmosphere, a significant chunk of what humans have emitted in the last century. “Global tree restoration is our most effective climate change solution to date,” the authors claimed in the paper.

Now, four independent groups take issue with the study’s methodology. Although reforestation remains a powerful tool in tackling climate change, the authors overstated the number of trees that could feasibly grow under Earth’s current climate, and how much carbon they could pull out of the air, according to the critiques published today (October 17) in Science. The authors respond to each criticism in an accompanying response, insisting that their estimates are accurate.

“I think one of the reasons that so many people in the scientific community got a little bit fired up in response to this article was that in our field it’s kind of well accepted that we don’t know how much carbon is currently stored in Earth’s forests. . . . Therefore, it’s even harder to estimate how much carbon could be stored in the future,” says Laura Duncanson, a remote sensing scientist at the University of Maryland whose research focuses on forest carbon mapping.

See “Earth Has Room for 1 Trillion More Trees

In the original study, ecologist Thomas Crowther of the Swiss Institute of Technology in Zurich and his colleagues first used a machine learning algorithm to predict where additional trees could naturally grow, based on climatic conditions under which existing forests are known to exist. Then, his team used a handful of published estimates on the carbon stored in existing forests to estimate how much carbon those additional trees could lock in once they reach maturity. After taking into account the carbon that would be trapped in the soil, leaf litter, and dead wood associated with the trees, they arrived at their 205 gigaton estimate.

One of the researchers to take issue with the research is Eike Luedeling, an agricultural scientist at the University of Bonn in Germany. “There are technical issues [that] I think pile up a bit to the point where I have some serious concerns,” he tells The Scientist. “I think they were probably about an order of magnitude off, probably a factor of ten or possibly more,” although it’s impossible to say exactly, he adds.

Unavailable carbon storage estimates

One issue Luedeling points out in a letter in Science is the figures Crowther’s team used to convert hypothetical canopy cover of additional forests into a concrete carbon value. Such figures for specific forest types are hard to come by, so they resorted to using a single estimate of carbon storage—which Luedeling suggests may originate from a tropical savannah—for a variety of ecosystems, including mangrove and Mediterranean forests and for potential forests in tundra regions, savannahs, and montane grasslands—distinct biomes in which trees grow differently and likely build up different amounts of carbon. While researchers have to take some shortcuts in in global modeling studies, oversimplifying the potential carbon storage of different forests in this way likely introduced many uncertainties into the team’s predictions, Luedeling says. “There are a lot of places in this study where these numbers always end up on the higher end,” resulting in inflated estimates.

Duncanson, who has collaborated with one of the original study’s coauthors in the past but wasn’t involved in this project, says such critiques are valid. The problem is that there simply aren’t any good estimates on carbon storage for different forest types, so, ultimately, “[Crowther’s] estimates cannot be verified or disproven.”

Jean-François Bastin, an ecologist at Belgium’s Ghent University and the first author of the original study, stresses that he and his colleagues took measures to account for the differences in carbon storage between forests, for instance, by devising a metric based on the average tree cover known to occur in specific types of forests, he says. If anything, the 205 gigaton carbon estimate is conservative, for several reasons. One is that the team only used carbon storage estimates from protected forests, many of which are known to be degraded. Using these lands as a proxy of potential carbon storage “actually underestimates what can be achieved through restoration,” he tells The Scientist.

Luedeling also takes issue with the map that Crowther’s team ultimately produced that shows where additional trees could grow globally. Many of those areas aren’t available for tree regrowth because they’re already in use.” Those regions include land used for livestock grazing, as well as populated areas such as Kinshasa, the capital of Democratic Republic of Congo, and rural settlements globally inhabited by more than 2.5 billion people, Luedeling writes in his letter. Some of this may be due to poor land use data for those regions, he notes.

The three other critics point out a number of other problems they see in the team’s methodology. For instance, two groups note that forests not only influence the climate by sucking carbon out of the air, but also by altering the surface albedo—the reflectivity of the Earth’s surface, for instance. If trees recolonize high-latitude areas, they may contribute to climate warming by virtue of their color, which absorbs sunlight rather than reflecting it away from Earth, as ice would. Accounting for such considerations would have been out of scope of their study, Bastin and his colleagues write in their response. Other critiques are based on flawed interpretations of their study, they add.

“Dangerously misleading” or raising awareness?

According to Bastin, the biggest misunderstanding lies in the fundamental goal of their study. Many perceived the research as offering directive guidance on where policymakers should initiate tree-planting projects. However, “our paper was not about telling people what they should do,” he stresses. The team only tried to assess the potential of what would happen “if we just removed human activity [from the equation],” he says.

Several groups of scientists took particular issue with the paper’s original statement that global tree restoration is “our most effective climate change solution to date,” an assertion one of the critics called “dangerously misleading” as it implies trees are the unique solution to climate change. Land, and how we use it, can be a big part of the solution to climate change, as outlined highlighted in a recent report by the Intergovernmental Panel on Climate Change. But those strategies only “buy us time” while people cut greenhouse gas emissions, which is arguably the most powerful climate change mitigation strategy, says Luedeling.

Although the study had considerable simplifications, Duncanson says—many of which the authors concede in their original paper—the fundamental message that trees can help us mitigate climate change caught people’s attention around the world, and that may be a positive thing, she notes. “The bottom line is that everybody agrees with that. And now not only is the scientific community ready for this next generation of datasets and ecosystem models, but now the rest of the world is also paying attention to this.”

And, Duncanson adds, better data to more accurately predict the effects of climate mitigation endeavors using trees are on the way. She’s currently working as part of NASA’s GEDI project, which aims to use laser-based satellite data from the international space station to generate a 3D reconstruction of Earth’s forests. This will open the path to assess forest carbon stocks more accurately than present field studies, she says.

E. Luedeling et al., “Forest restoration: overlooked constraints,” Science, doi:10.1126/science.aay7988, 2019. 

P. Friedlingstein et al., “Comment on ‘the global tree restoration potential,” Science, doi:10.1126/science.aay8060, 2019. 

J.W. Veldman et al., “Comment on ‘the global tree restoration potential,” Science, doi:10.1126/science.aay7976, 2019. 

S.L. Lewis et al., “Comment on ‘the global tree restoration potential,” Science, doi:10.1126/science.aaz0388, 2019. 

J. Bastin et al., “Response to Comments (Friedlingstein, Veldman, Lewis) on ‘The globaltree restoration potential,’ Science, doi:10.1126/science.aay8108, 2019. 

Correction (October 17): A previous version of this article incorrectly stated that the regions considered suitable for forest growth in the original paper included rural settlements in Africa inhabited by more than 2.5 billion people. Those settlements are not confined to Africa but are instead distributed globally. The Scientist regrets the error.

Katarina Zimmer is a New York–based freelance journalist. Find her on Twitter @katarinazimmer.