IN THE SPRING, the fields of flowers at Harborview Farms cause drivers to pull off the road to take pictures. "It's unbelievable the amount of people who will compliment me on my flowers," Trey Hill says as we walk through a field of rapeseed. A few weeks earlier, Hill tells me, this was all corn.

Like his father before him, Hill is a commodity farmer. He grows corn, soy, and wheat on 10,000 acres in northern Maryland, chasing yields with genetically engineered seeds and herbicides. Unlike his father, Hill also has one foot in an entirely different world: regenerative agriculture.

Hill's introduction to the practice came in the early 2000s (though he didn't hear the phrase regenerative agriculture until a decade later). A group of environmentalists approached him and asked whether he would be willing to experiment with planting cover crops after the harvest to help stabilize the soil and keep fertilizer out of Chesapeake Bay. Hill thought it was a stupid idea. The Chesapeake was polluted—but by factories and industrial-scale poultry operations, not by Harborview. He was already measuring the levels of phosphorus and nitrogen in his soils and applying fertilizer only where and when he was supposed to. But the State of Maryland was willing to pay for the experiment, and Hill figured he had nothing to lose.

Two decades later, Hill is still putting in cover crops—the state pays farmers as much as $80 per acre to plant them. Much to his surprise, the practice has also brought him a lot of joy. "It's kind of bringing farming for me . . ." Hill pauses for a second here. "Not bringing it back," he continues, hesitating. Harborview is still a conventional farm. But there's just something about cover cropping compared with the rigid science of growing for the commodity market. Hill experiments with different seed mixtures in different areas, checks in with fellow growers and researchers for advice, and loves the trial and error of it. "When you start doing this, it's, 'Oh shit, not sure what's gonna happen,' right? 'But it's gonna be fun.'"

Hill and other cover croppers may soon have company. Regenerative agriculture—also called regenerative farming and now carbon farming in more climate-focused circles—has become very buzzy in recent years. In their most recent report, scientists with the Intergovernmental Panel on Climate Change estimated that "soil carbon sequestration in croplands and grasslands" worldwide could store as much as 8.6 billion tons of CO2 annually—an amount equivalent to roughly 1.5 times the United States' annual emissions. In the early days of the Biden administration, Democratic Georgia representative David Scott, the chair of the House Agriculture Committee, opened a hearing on climate change with a clip from Kiss the Ground, a documentary in which Gisele Bündchen, Woody Harrelson, and soil scientist Ray Archuleta tell viewers that they can stop the climate crisis by supporting regenerative agriculture. "There's so much bad news about our planet," Harrelson intones over footage of a desiccated landscape. "This is a story of a simple solution. A way to heal our planet. The solution is right under our feet, and it's as old as dirt."

A few months later, in his first address to Congress, President Joe Biden not only promised to tackle climate change but also enumerated exactly how the administration might do it. On the list was "farmers planting cover crops so they can reduce the carbon dioxide in the air and get paid for doing it."

Politically, regenerative agriculture is a rare creature—one of the few forms of carbon sequestration that have attracted significant bipartisan support. But while dirt may be old, the science of how to persuade it to store carbon isn't.

 The term regenerative agriculture was popularized by publisher and organic-farming advocate Robert Rodale in the 1980s. Rodale did not get into specifics, merely saying that regenerative-farming methods enhance the "land and soil biological production base." Definitions have only proliferated since then. Some characterize regenerative agriculture by result (amount of carbon sequestered or level of soil fertility). Others see it as a set of processes, like planting cover crops, eliminating tillage, rotating crops, and integrating livestock to fertilize fields. When researchers at the University of Colorado surveyed 229 scientific papers that mentioned regenerative agriculture, they found that while use of the phrase had gained momentum since 2018, there was very little agreement as to what it meant, beyond vague goals like "improve soil health," "reduce or eliminate tillage," and "integrate livestock."

The study authors saw this lack of a common definition as a challenge for several reasons. It's tough for researchers to study something and test its proponents' claims if they can't agree on what it is first. Consumers get confused about what they're buying, and it's easier for a term with a vague definition to become watered down or corrupted, despite the ideals of the earliest practitioners. And finally, it's difficult to develop and advocate for laws and policies supporting something whose definition is so mutable.

IN 2018, HILL was approached by a Seattle-based carbon-trading company called Nori. Two years earlier, he had converted all of his acreage to no-till farming. Nori crunched some numbers using a model based on COMET (a tool developed by the US Department of Agriculture and Colorado State University to calculate greenhouse gas emissions), gauged Harborview's carbon-sequestration potential, and came up with an offer. If Hill signed up with the company and planted more cover crops, Nori would pay him $15 per ton of carbon sequestered and credit him retroactively for the carbon stored owing to his switch to no-till.

That first year, Hill made $115,000 for storing an estimated 8,000 metric tons of carbon. That's not a huge amount considering the extra work involved—it's less than he gets from the state for cover cropping. But it's enough of an incentive that Hill has been enrolling more acreage in Nori to increase the payout.

Nori is one of several outfits that are paying farmers to store carbon and then selling it as offsets to individuals and companies that want to brand themselves as sustainable. Some of these offset sellers are run by heavy hitters in the ag world: Bayer, Corteva, Farmers Business Network, and Truterra (the last is an offshoot of Land O'Lakes). The carbon-offset market tends to draw large farms like Harborview rather than small ones, since there's more of a payoff for the upfront paperwork and meticulous crop tracking when you have thousands of acres as opposed to a couple hundred. The price for carbon credits in the United States fluctuates, but $15 to $20 per ton of carbon sequestered is common.

Carbon markets first came to global attention during talks leading up to the 1997 Kyoto Protocol. The international treaty to lower greenhouse gas emissions allowed countries to invest in projects in developing countries to avert or sequester carbon emissions, gaining offsets in return. Kyoto didn't specify what the carbon market would look like, and trading never happened under the treaty. But in its wake, multiple private and public carbon markets appeared. In some of these, a private organization served as the intermediary, much like a stock exchange.

The credits bought and sold by Kyoto participants were verified by one of a handful of third-party organizations. Two requirements were particularly important to the verification process: permanence (a carbon offset needs to keep the removed carbon out of the atmosphere for a specific period of time, usually 100 years) and additionality (the carbon credits need to represent emissions reductions that would not have happened otherwise).

One of the privately operated carbon markets to emerge in the wake of Kyoto was the Chicago Climate Exchange, which launched in 2003 with offsets that included no-till farms. It folded in 2010, after it became clear that most investors were not interested in a voluntary market and were waiting for the federal government to step in. When climate legislation failed to go through the Senate, the market folded. But that hasn't stopped a new crop of carbon brokers from emerging in recent years. Last year, Microsoft purchased offsets for 100,000 tons of its carbon emissions from Truterra. PepsiCo announced that it would adopt regenerative-farming practices "across 7 million acres"—including farms growing potatoes for Lay's and oats for Quaker—resulting in "a net-reduction of at least 3 million tons of GHG emissions by 2030." Cargill, which controls 18 percent of the domestic meat market, announced that it will reduce emissions from its beef supply chain by persuading its wheat, corn, and soy growers to adopt no-till and cover cropping. It estimated that this would sequester 1.7 million tons of carbon by 2030.

In May 2021, USDA leadership announced $10 million in funding to improve the way the department measures soil carbon sequestration on farms. One goal is "strengthening" COMET (the acronym stands for CarbOn Management and Emissions Tool), which was first developed for farmers as a way to qualify for state conservation grants.

The announcement came a few months before I spoke with Adam Chambers, the USDA climate scientist who designed COMET in partnership with Colorado State University. Chambers walked me through how it works, selecting a state, then a county, then a series of cover crops. "If that's a nonleguminous, seasonal cover crop on an irrigated system, and then we put in 180 acres of cover crops, we sequester 29 tons in a given year," he told me. "It's that simple."

It did sound simple. But how did he know that 29 tons was an accurate estimate? The projections are extrapolated from a database of peer-reviewed long-range studies measuring carbon-sequestration efforts, Chambers told me.

But here's the catch: The science behind the studies that the database is built on is incredibly contentious.

IF YOU WANT to measure how much carbon is sequestered in a way that is considered authoritative for soil carbon markets, one method is to take several pounds of dirt (sometimes hundreds of pounds of dirt) out of a field and mail it to a lab that will blow it up to detect the carbon. Those lab numbers won't tell you about an entire field, though, says Hill, because carbon doesn't distribute itself evenly through a landscape. "You could take a sample here," he says, pointing to an area of the field where the rapeseed is particularly dense. "And then you see my cover crop didn't come up there, and you take a sample there. It's going to be completely different."

Carbon levels in soil vary depending on where you sample horizontally. They also vary depending on how deep you sample vertically. Soil scientists once believed that carbon could become permanently locked into soils in the form of large, stable, carbon-rich molecules, but the consensus has shifted: All soil carbon breaks down eventually, just at different rates depending on the depth.

Carbon sequestered at lower depths won't stay there forever, but it can stick around for thousands of years. That persistence is why unplowed wild savannas and rainforests serve as massive carbon reserves. Scientists don't fully understand how these deeper carbon stores form. One explanation is that they're less subject to decomposition by the microbes that feed on carbon and breathe it back into the atmosphere as CO2.

Carbon in the top 30 centimeters (about 12 inches) of soil is far more fickle. Massive amounts of carbon go into and out of agricultural soils every year for all sorts of reasons. (Most no-till farmers still plow every one to three years, for example.) There's no consensus about whether it's possible to get that highly cyclical carbon to stay where it is. And in recent years, some research has cautioned that increasing global temperatures might reduce soil carbon sequestration. As temperatures increase, microbes in the soil tend to release carbon. The closer these microbes are to the soil's surface, the more carbon that escapes into the atmosphere.

Over a decade ago, a team of scientists at the University of Minnesota and the USDA Agricultural Research Service in St. Paul published an article in the journal Agriculture, Ecosystems & Environment that pointed out that a significant number of studies making carbon-sequestration claims about no-till farming were based on samples collected from the top 30 centimeters. One 30-year comparison of no-till and tilled plots run by CSIRO, Australia's national science agency, found that shifting a plot from tillage to no-till did not increase the carbon sequestered by that soil significantly. John Kirkegaard, an agronomist at CSIRO, has a working theory that microbes near the soil's surface are nitrogen-hungry, just like plants. If there isn't an abundance of nitrogen for them to devour, they'll break down existing organic matter in the soil instead, releasing some of the carbon stored there in the process.

Applying nutrients in the form of manure, often by adding livestock, is a big part of regenerative agriculture. Hill uses manure from local poultry farms at Harborview but is limited in how much he can apply by state laws. Add too much nitrogen and it will leach into nearby waterways, throwing aquatic ecosystems out of balance. Not only that, but applying manure is associated with nitrous oxide emissions, another greenhouse gas pollutant.

While the science behind carbon farming is speculative, the emissions it aims to offset are far more certain. The US food system is responsible for about 10 percent of the country's greenhouse gas emissions, according to the EPA, and that percentage would be much larger if it weren't for the outsize energy use of other sectors.

Agriculture also uses a lot of land. All landscapes have a "carbon opportunity cost," says Tim Searchinger, a researcher who studies food systems and climate change at Princeton University. Even the best managed cover crop can't match the carbon-storage capacity of a forest or prairie or wetland ecosystem, and so any area that is used for growing crops sacrifices some higher amount of potential carbon storage. That trade-off is necessary to make sure people are fed. But, he argues, we need to limit further farmland expansion to keep emissions from getting even higher.

THE REFRAIN THROUGHOUT Kiss the Ground is that regenerative agriculture is a solution to climate change because it's bringing livestock back onto the farm, where their manure can fertilize the soil and help it store carbon instead of polluting a waterway somewhere else. Or, as Gabe Brown, one farmer featured in the film, puts it, "The problem isn't the animal. The problem is where the animals are at." Brown raises livestock using a method called mob grazing—popular with many regenerative farmers—whereby animals are moved daily from one area to the next via a system of gates and fences to keep the fields from becoming overgrazed.

But for every farm and livestock operation to run this way, we'd have to do something extremely unpopular—something that Kiss the Ground never mentions. We'd have to eat less beef. A lot less. Even less than environmental groups like the World Resources Institute and Project Drawdown already recommend.

Denmark, Finland, and Iceland are updating their nutrition recommendations to encourage a more climate-friendly diet—one that includes less meat. The last time the US updated its dietary guidelines, a plan to add similar language was scuttled, thanks to meat and dairy industry influence. North Americans eat three times more beef than the global average. The US beef industry as a whole manages an inventory of 93.6 million cows and calves, and land for animal farming in the US already takes up more than 400 million acres. The US could not convert to regenerative grazing practices and produce the amount of meat we eat now without using more than twice as much land.

Even if you take livestock out of the equation, there's no question that cover crops, no-till, and crop rotation are good for soil health, says Deb O'Dell, a soil scientist and adjunct assistant professor at the University of Tennessee. The USDA already pays farmers to restore wetlands and to not farm landscapes that are important habitat for certain plant and animal species. But these programs don't always have the funding for everyone who wants to participate. The Environmental Quality Incentives Program and the Conservation Stewardship Program, for example, can fund only about 25 percent of applicants. Expanding these programs would not be a stretch. Call it "ecosystem services," O'Dell says. Consider it economic aid to rural areas struggling with poor water quality. Even if there are no climate benefits at all, holding on to your topsoil makes good sense.

Checking up on farmers to make sure they're following through on ecosystem services is also easier than monitoring soils. Satellite imagery, drones, and other remote sensing tools can identify farmers who burn or till their fields or fail to plant a cover crop.

Searchinger sees carbon farming as a distraction. "If you're actually serious about solving climate change, the first thing you do is you start trying to restore your peatlands," he says. At the Pocosin Lakes National Wildlife Refuge, in North Carolina, one restored-peatlands project is storing 70 million pounds of carbon and 2.2 million pounds of nitrogen per year. "There's no technical uncertainty," Searchinger says. "So why aren't we doing it?"

IN FEBRUARY, THE USDA launched Partnerships for Climate-Smart Commodities—offering $1 billion in funding to organizations that develop systems to help farmers "quantify, monitor, and verify" reduced emissions from carbon sequestration.

In March, the House Agriculture Committee held another hearing on climate change, this one in anticipation of the 2023 farm bill, which will set agricultural policy for the next five years. There was the usual debate over whether agricultural emissions are a thing—basically an "Is Climate Change Real 2.0." (Spoiler: Yes.) Republican Pennsylvania representative Glenn Thompson and Joe Outlaw, an economist at Texas A&M University, worried that developing climate programs would create "winners and losers," as though USDA policies don't already do that.

Alma Adams, a Democratic representative from North Carolina, was the only elected official present who mentioned the Paris climate accord. "Unless there's a real systemic change," she said, "we now know that emissions from agriculture will make it impossible to achieve the Paris Agreement of limiting global warming to no more than 2 degrees Celsius." It was hard to say which possibility was more ominous—that carbon farming would progress to become a tool for greenwashing on a national scale or that the United States would continue to do nothing to rein in the sector that generates a 10th of its greenhouse gas emissions.

That greenwashing is already a reality, according to research by Sonali McDermid, Jennifer Jacquet, and Oliver Lazarus, three researchers at New York University. Out of the 35 largest global meat and dairy companies in the world, only six have made zero-emissions commitments. None of those companies are making substantial changes in the way they do business. Meat companies are feeling pressure to show some kind of action on climate, Jacquet says. Regenerative agriculture is a way to draw attention away from climate policies that would cut into profits, like eliminating USDA subsidies for livestock and the crops that are grown to feed them.

To Matthew Hayek, an environmental scientist at NYU, the way we talk about the agriculture industry is similar to conversations about the oil industry 25 years ago—directing attention away from the pollution it is causing by focusing on a new technology instead of old problems. "We were talking about coal with carbon capture, and there was really amazing and promising research into carbon capture and sequestration," Hayek says of those early discussions about cutting carbon emissions. It wasn't enough. "That research was critical. But it was never a silver bullet."

The same could be said for regenerative agriculture. Cover cropping, crop rotation, and reduced tillage do store some carbon. That's why Project Drawdown includes it in the mix of its encyclopedic rundown of climate solutions. But it's important not to overhype it, says Chad Frischmann, Project Drawdown's head of research. "We should be constructively critical of these solutions," he says. "Always. Because some of these solutions may not be solutions tomorrow, right? And there may be new solutions that come along that we find that are even better."

This article appeared in the Summer 2022 quarterly edition with the headline "Dirt First."