You could say the story of our collective future will be the story of how well, or how poorly, we manage carbon.
Humanity evolved in the Holocene, a warm nest of atmospheric stability that over the past 10,000 years has supported the web of life. That time is ending. We’re quickly entering the Anthropocene, the age in which humans are changing the life support systems of the Earth. Much has been written about how we’ve altered the planet through our carbon emissions—burning fossil fuels, cutting down forests, over-tilling the soil—but the narrative doesn’t stop there.
Humans are endlessly adaptive and creative. We as a species are learning to intervene in positive ways, and a more hopeful story is now emerging, one in which we not only begin to sequester carbon and reverse the climate crisis, but we also bring back biodiversity and regenerative livelihoods. It’s a story that has yet to be written, and some of the key heroes in that story are scientists, like the team at Earthshot Labs.
“One of the holy grails in ecology is forecasting the effects of active forest replanting versus natural regeneration,” says Steve Klosterman, Earthshot Lab’s Principal Scientist. “It’s a data problem. We don’t have as much growth data as we would like for actively planted trees.”
“A great deal of work has gone into assembling global datasets of tree growth in regenerating forests,” Steve says. These datasets have been key to creating predictive models, but thus far they are all geared toward predicting carbon accumulation under an ENR (Enabled Natural Regeneration) scenario. Under an ENR management scenario, land stewards might add a cattle break or fire break, but otherwise no trees are planted. Nature is left to its own devices.
If you’re a land steward in Panama or the Amazon or Burkina Faso and you want to reforest your land, where can you find accurate data that will give you the highest yield of sequestered carbon and ensure that your forest will survive?
Thus far there are no great data sets for planted trees. How much does tree planting accelerate carbon accumulation? “No one has collected this data together into one big synthesis study,” says Steve. “That’s one thing I’m excited about over the next year or so.”
We know that replanting forests also produces a host of benefits in addition to carbon. What we don’t know is how and where to plant trees for the greatest benefit, how those trees might respond to a changing climate, and what risks those future forests will confront. If you’re a land steward in Panama or the Amazon or Burkina Faso and you want to reforest your land, where can you find accurate data that will give you the highest yield of sequestered carbon and ensure that your forest will survive? How to design the most cost-effective projects that restore biodiversity, repair habitat connectivity, and create regenerative livelihoods? These are some of the key problems the team of scientists at Earthshot Labs have set out to solve.
“The core work of the Earthshot science team is to predict carbon accumulation, in addition to avoiding CO2 emissions from deforestation,” says Steve. “All our predictive modeling right now is for forests. We have developed a simple predictive model of ENR (in forests) based on data assembled for Latin America, but we're now working on a global ENR model. The goal is to predict carbon accumulation anywhere we have projects.” Which means all over the world. In less than a year, Earthshot has quickly scaled up a pipeline of forty-two projects. The science team is developing a two-stream approach, using both empirical and process-based models. Both streams will generate predictions of carbon accumulation and biodiversity potential under climate and management scenarios, and by comparing the differences they will see where current understanding is lacking and where more data is needed.
“We’re entering a new climate space, and we don’t have the data,” says Polly Buotte, one of Earthshot’s ecologists and a Contributing Author to the 6th IPCC report. “My role is to develop the process models to predict future reforestation and the potential for associated benefits like biodiversity. The models we have were not specifically built for this purpose.”
The IPCC gives us forecasted carbon accumulation rates, known as Tier 1 estimates, for the entire world, but those models don’t target individual ecosystems. Steve describes the IPCC estimates as “widely used, but simplistic default values for forest growth. They’re spatially coarse and not very accurate.” Tier 1 estimates exist for both natural regrowth and plantation scenarios. They’re recognized in the reforestation community as one way to predict reforestation rates, but they are blunt instruments. They’re not designed for the spatial scale of individual reforestation projects. To make use of this data requires looking at a PDF document and then translating that to computer code. It’s tedious and time consuming. Nor can those Tier 1 estimates account for a changing climate.
A much better model is needed, a process-based model that builds off first principles instead of climate-growth relationships and one that targets localized needs: how do we bring back and sustain biodiversity in the Azuero peninsula of Panama? How do we ensure profitable carbon stock for the Yawanawá people in the Brazilian Amazon that will allow them to maintain control of their land and support their cultural heritage?
One of the models Polly is employing for Earthshot is CLM-FATES—Functionally Assembled Terrestrial Ecosystem Simulator. The “Functionally Assembled” is the really important part. It allows you to say that this tree is shade tolerant, and that tree is not; this tree is drought resistant, but not that one. When Polly codes all those strategies, the model can identify a combination of tree types well-suited for a specific area, in the Madre de Dios region of the Peruvian Amazon, for example. “With this model you can do cool things like including projected changes to climate. How will these trees live or die in 30 years with certain climate changes?”
While each member of the Earthshot team is working on a unique problem, the exciting work is seeing it all come together into one online platform. The result is LandOS: a global ecological simulator that will allow land stewards anywhere in the world to type in their location and receive a carbon projection for their land parcel.
Using LandOS, land stewards can discover how much carbon their land could grow, what risks those trees will face from drought or fire, and how best they can secure funding from carbon credit buyers. Once appropriate data becomes available, LandOS will be able to predict which tree species are best to plant in a specific bioregion. With carbon markets still in their nascency and ecologically-accurate data difficult for non-scientists to access, the LandOS platform will give the world’s land stewards the best information available, in real time.
Scientifically rigorous and ethically grounded, the aim of LandOS is to provide high level due diligence for the carbon market.
The platform is already being used by Earthshot projects around the world. Once LandOS is rolled out for the public, land stewards will have a free online tool to make a projection of forest growth anywhere in the world under different climate and management scenarios.
Working with big data sets is central to Earthshot’s work, which means that Earthshot’s ecologists are also data scientists. For his work on LandOS, ecologist Joe Hughes uses Landsat imagery, combines it with abiotic factors like elevation, slope, aspect, water potential, and climate variables, and creates high resolution maps of where biomass is located. Joe also maps where biomass has been historically. “Using historical Landsat images lets us look into the past to see where forest regrowth has happened in environments similar to a place we’re interested in reforesting now, which helps us predict how forests might grow in our project area and how it might look in 30 years,” he says.
The point of this work is to make estimates of how much forest we can grow, as well as determining the trajectory over time. This allows a user to calculate how much carbon can be captured in any given locale. The goal is to take places similar to an Earthshot project site, see how they’ve grown biomass in the past, and use that past knowledge to predict the future, especially when factoring in projections around climate change.
"Carbon markets are a tool we can use to grow forests that are rich in biological diversity, that are structurally complex”
Joe views biomass as a means to an end. The larger goal is planetary regeneration. “Carbon markets are a tool we can use to grow forests that are rich in biological diversity, that are structurally complex,” Joe says. “We don’t want a plantation. We don’t want rows of pine trees.” There may be a lot of biomass in a pine plantation, Joe explains, but in the end it’s just a monoculture, like a cornfield. “It’s fundamentally just agriculture, it’s not a natural system. You’re dumping fertilizers on it, it causes eutrophication downstream, you’re degrading the soil. That’s not what we want to do.”
One of Earthshot’s unique offerings via LandOS is to model biodiversity and structural complexity in forests. Carbon capture is important because people need to make money. But equally important is increasing biodiversity for its own sake. A structurally complex forest is important for a high functioning ecosystem that supports other life beyond the human.
In the coming years, the Earthshot team plans to incorporate cultural preferences into their forest modeling. There might be a particular sacred mountain that needs to be left alone, or a need to reforest along rivers. The team also wants to incorporate indigenous views of the land, and include species selection and reforestation techniques that are not yet understood by science, or have thus far been too complex to model. Conceptually, each of these pieces of cultural wisdom are little dials, Joe explains, “and you can dial up each function. How much of this do we want, how much of that?” LandOS will find those locations and match the selected preferences.
As these scientists’ work makes clear, carbon is the basis of Earthshot’s financial model, but that’s not all. The Earthshot team views the carbon market as the means to a larger end: creating livelihoods for people most affected by climate change, bringing back biodiversity in forest ecosystems around the world, and restoring humanity’s broken relationship with nature. How to encode such ambitious goals into scientific modeling is the creative challenge Earthshot scientists have set out to solve.
You could say the story of our collective future is the story of carbon, perhaps, but that’s really just the beginning. Once humanity begins replanting the world’s great forests–the lungs of the planet–we will sequester carbon, yes, but we will do so much more. We will create regenerative livelihoods for multigenerational land stewards. Restore biodiversity. Repair watersheds. We will bring nature back.
The mission of Earthshot Labs operates from a conviction that we must restore nature for its own sake, that things like a structurally complex forest, healthy soil, and abundant ecosystems have intrinsic value apart from whatever use case they provide for people.
The human economy is a subset of nature’s economy, but since the beginning of the Industrial Revolution we’ve pretended otherwise. Earthshot Labs is on a mission to reverse the script. Human impact need not be extractive. We can participate in the healing of nature.
It’s a holistic, even audacious, vision, one that Earthshot’s scientists are encoding into their models.
The holy grail for us all is an ancient one: how to live in right relationship with the natural world. By replanting the world’s great forests, we can bring nature back. We can also reclaim those parts of ourselves we’ve allowed to atrophy: a feeling of kinship with the web of life, a fierce desire to protect our common home.