For most of the industrial era, spreading crushed limestone across farmland has been treated as a carbon problem. The rock contains carbon; it ends up in contact with air and water, and by the most conservative accounting conventions, carbon is assumed to escape as carbon dioxide. Regulators and major climate assessments have baked that assumption into their default rules for decades.
A Yale-led study published in Nature in September 2026 now argues that the opposite is true and that the error came from asking the wrong question. Across more than 120 years of records from the Mississippi River Basin, the researchers conclude that agricultural liming has removed somewhere between 300 and 400 million metric tons of carbon dioxide from the atmosphere since 1900.
The finding matters well beyond the Corn Belt, because the substance doing the work is not the trendy carbon-crushing mineral of the moment. It is ordinary agricultural lime, a soil amendment farmers in the basin have used for generations to raise pH and improve yields. And the study quietly undercuts a large slice of the market that has grown up selling the alternative.
What agricultural liming actually is
Most people associate soil pH management with gardening, but on a commercial scale it is one of the largest single material movements in agriculture. Every year, more than 30 million tons of agricultural lime, made up mainly of calcium and magnesium carbonates, are applied to US croplands to hold soil pH at the level plant growth needs.
Acidification is worth understanding on its own terms, because it is the reason the practice has survived so long. Modern agriculture acidifies soil in two ways. Nitrogen fertilizer releases hydrogen ions as plants take up nutrients. And since the industrial revolution, acid deposition from air pollution has added a second, largely unacknowledged load. Without intervention, soil pH drifts downward, nutrients lock up, and yields fall.
Limestone, or ag-lime, is the standard fix. It is quarried, crushed, sometimes pelletized, and spread at rates commonly running from under one to three tons per acre. Rain and irrigation water react with it, the hydrogen ions get neutralized, and the pH rises. Crops respond within the same or the following growing season. The practice is old, cheap in relative terms, and completely unglamorous.
The counterfactual that reverses the answer
Here is where the Yale study departs from a press-release summary. The finding does not come from measuring how much carbon limestone releases. It comes from measuring what would have happened without it.
This distinction is the entire argument. In the absence of liming, the acidity generated by fertilizer use and by acid deposition would still drive chemical reactions in soil and water that release carbon dioxide. There is a baseline emission, and it is not zero.
“A key advance of the new study is to account for what would have happened without lime,” the researchers wrote. When lime is evaluated against that more realistic scenario rather than against a perfect no-lime fantasy, the long-term balance comes out as net carbon removal.
Stated plainly, the comparison is between two active worlds:
- The no-lime world: a field that received no lime, stayed acidic, acidified further under the same fertilizer regime, and lost the carbon-holding capacity it would otherwise have retained.
- The limed world: a field where hydrogen ions are neutralized, pH rises, and reacted carbon forms bicarbonate that travels through soil, groundwater, and rivers to the ocean.
Limestone is being compared against the soil it prevents, not against silence. That is why the headline number can be so large. It is a difference between two active worlds, not a gross tally of one.
What 120 years of records show
The study is unusual in that it is a retrospective analysis rather than a new experiment. Writing in Nature, researchers G. Philip Robertson of Michigan State University and Stephen K. Hamilton, also of Michigan State, described the work by Suhrhoff and colleagues as a comprehensive analysis of water chemistry in the Mississippi River Basin. Drawing on a trove of historical liming records across an area covering 41% of the contiguous United States and roughly 65% of US croplands, the team assessed how much carbon the region’s rivers and soils actually retain.
Their estimate puts net removal since 1900 at 300 to 400 million metric tons of carbon dioxide. The researchers describe this as indicating an opportunity to scale up carbon removal by helping more farmers carry out liming, noting that many cannot afford optimal soil pH management.
Noah Planavsky, a professor of Earth and planetary science at Yale and a corresponding author, framed the significance in terms of data availability rather than novelty of chemistry.
This is a once-in-a-lifetime opportunity to trace more than 120 years of historical records that can tell us whether enhanced weathering works. It’s a unique look at a process that we already know has a benefit to farmers, and now we know is also beneficial to the climate by removing CO2 from the atmosphere.
Tim Jesper Suhrhoff, the first author and a geochemist at Yale, put the policy appeal plainly.
One of the exciting things about this result is that it aligns climate action with something that is already good for farmers.
Christopher Reinhard, a professor at the Georgia Institute of Technology and co-corresponding author, added the access point.
Better soil pH management can improve yields and soil health, and our work shows that it can also be good for the climate. That gives us another reason to expand limited access to it where it’s sensible.
The part that should worry the enhanced weathering industry
Carbon removal has a dominant strategy called enhanced rock weathering, which IL covered in 2020 when a University of Sheffield team estimated it could pull up to two billion tons of carbon dioxide a year from global croplands. The pitch is simple and appealing: crush silicate rock such as basalt into powder, spread it on fields, and let a natural geological process run faster. Because silicate minerals contain no carbon at all, every bit of carbon that ends up locked into the resulting bicarbonate is drawn from the atmosphere or the soil.
That zero-carbon-content detail is precisely what limestone does not have, and it is the source of the two-decade argument the Yale paper has just entered.
According to the Carbon Removal Standards Institute’s account of the chemistry, one unit of ag-lime can weather to produce up to two units of inorganic carbon, but only one of those units comes from the atmosphere. The other comes from the rock itself, which had been storing that carbon since the limestone formed. When a silicate or quicklime weathers instead, both units are atmosphere-derived.
So on a strict chemical reading, limestone can be a carbon source and basalt cannot. That is why the 2006 IPCC Guidelines for National Greenhouse Gas Inventories treat liming as an emission source, with default emission factors of 0.12 tons of carbon per ton of limestone and 0.13 tons of carbon per ton of dolomite. Those numbers are simply the carbonate carbon content of each rock, 12% for calcium carbonate and 13% for dolomite, and inventories convert them to carbon dioxide using a factor of 44/12, which works out to roughly 0.44 and 0.48 tons of carbon dioxide per ton of material applied. On paper, spreading lime is treated as releasing carbon.
What the Yale study does is challenge that convention from the other end. Rather than arguing about which carbon atom came from where, it measures the net against reality and finds removal.
The uncomfortable consequence is that the more fashionable option may not be the better one. Modeling published by CarbonPlan using a reactive transport model called SCEPTER, and examining conditions similar to the US Great Plains, found that current liming practices can remove more carbon than basalt-enhanced weathering projects, particularly in less acidic conditions. Calcite and basalt both work, but the trade-off depends heavily on where you draw the accounting boundaries in space and time.
That is an uncomfortable result for a sector that has raised capital on the promise of permanence and additionality in silicate rock. It is also a result that favors the version of the technology that already exists, already ships by the truckload, and already has infrastructure behind it.
Why the evidence base for weathering is weaker than the enthusiasm
It is worth being blunt about how immature the evidence remains in the wider enhanced weathering field, because the Yale result is sometimes being read as validating it.
CarbonPlan’s own review of the published literature found that removal estimates span four orders of magnitude and that in 12 of 116 cases the reported estimate or its uncertainty range was consistent with zero removal, or with net emissions. The organization notes that there is still no widespread agreement on how removal should even be defined, and important parts of the carbon cycle remain poorly understood.
A separate critique in Frontiers in Climate argues that accelerated silicate weathering rates may be systematically overestimated when accessory carbonates within rock powders dissolve and dominate the dissolved inorganic carbon flux, which can be misread as silicate weathering. It also flags that measuring cations alone is prone to misinterpretation, because cations are often not balanced with dissolved inorganic carbon.
Enhanced weathering is also early in its maturity as a commercial method. A 2026 Perspective in the Oxford-based journal Carbon Removal notes that the IPCC lists the technology readiness level of enhanced rock weathering as low to medium, or 3 to 4, because long-term impacts, co-benefits, costs and spillover effects remain poorly understood. The same assessment warns that quantifying net carbon removal in multi-year field trials remains challenging.
A separate formal expert elicitation published in Nature Communications Earth & Environment estimated that in the American Midwest, for soils at pH 5.5 to 6, the share of theoretical potential carbon removal actually realized ranges from 27% to 39%, with global potential estimates varying by feedstock from a modest source to more than 5 billion tons a year.
None of this makes enhanced weathering wrong. It makes the 120-year liming record genuinely more robust evidence than most of what surrounds it, precisely because it measures an activity that already happened rather than a rate that must be modeled.
The IPCC accounting problem
Underneath the chemistry is a bureaucratic problem, and it is the part most likely to determine whether this finding changes anything.
The Yale team is explicit that it supports the IPCC’s broader effort to document and measure emissions. Their argument is narrower: that the default methodology, which treats the carbon in applied lime as emitted carbon dioxide, is incomplete, and that a more complete framework could recognize situations where liming benefits both farmers and the climate.
Rather than treating good soil-pH management as an additional source of CO2, a more complete framework could recognize situations where liming benefits both farmers and the climate.
There is a legitimate defense of the default method here, and it should be stated rather than glossed over. Default emission factors exist because they can be applied uniformly and cheaply, and because a conservative assumption cannot be accused of inflating climate action. Rewriting a default is a political act with distributional consequences, particularly for soil types where the chemistry does not favor removal.
That is also the reason this is a story about a category of method rather than a policy victory. Nobody changed a rule in September 2026.
What could go wrong
The researchers themselves are careful, and the caveats deserve to travel with the headline number.
It is local. Suhrhoff stated that the climate impact of liming depends on local factors such as existing soil acidity, soil buffering, and hydrology. Key dependencies include:
- Existing soil acidity: Already near-neutral soil will not respond like an acidic Midwestern field.
- Soil buffering capacity: Heavily buffered soils resist the pH shift that removal depends on.
- Hydrology: The pathway to the ocean depends on how water moves through the profile.
- Time horizon: How long until removal emerges, and whether it outlasts the up-front emissions.
- Material composition: Dolomite and impure limestones behave differently from pure carbonate.
Because of all this, the Mississippi River Basin result will not transfer directly to every agricultural setting.
Per hectare, the number is modest. Spread across the roughly 100 million acres of limed land in the basin, net removal works out at approximately 0.01 to 0.03 tons of carbon dioxide per hectare per year. That is a low figure next to the enhanced weathering median. The case for liming does not rest on any single acre being spectacular; it rests on the fact that a low rate applied across a very large existing area, without new machinery or new supply chains, adds up to more total carbon removal than a higher rate that is still largely theoretical or confined to trial plots.
There is an up-front cost. Quarrying, grinding, and transporting 30 million tons of material a year is not free. Suhrhoff’s phrasing explicitly includes the magnitude and duration of any up-front carbon dioxide emissions. The study nets these out over a long timeframe; a short-horizon accounting might look very different.
Not all liming materials are equal. Dolomite, which contains magnesium as well as calcium carbonate, behaves differently. So do impure limestones, and so do sourcing and transport distances. A ton of carbonate rock is not a universal unit.
The carbon may be slow. The mechanism is that reacted carbon forms bicarbonate ions that travel through soil, groundwater, and rivers and eventually reach the ocean, where the associated carbon can be stored for long periods. Long is not the same as permanent, and ocean residence times are themselves a subject of active research.
It is not a substitute for emission reduction. A slow, local, soil-pH-driven process does not compete with the scale of direct decarbonization, and framing it as a solution to the climate crisis rather than a useful complement would invite the same skepticism that has dogged carbon removal markets for a decade.
Affordability may matter more than chemistry
There is a line in the Yale release that is easy to skim past and that arguably contains the most important consequence.
We know that many farmers can’t afford optimal soil pH management. We need to consider how we can fund this process by taking into account the carbon removals.
That is a proposal to finance a soil amendment by monetizing its climate effect, and it is the mechanism by which a real finding in geochemistry would translate into lower input costs for growers and better yields on the land.
It is also a mechanism with a long and unpromising history. Ever-Green Energy’s plan to tap heat from the 172 million gallons of wastewater flowing daily out of St. Paul’s treatment plant, a $150 million scheme that would displace the natural gas powering half of the largest hot-water district heating system in the United States, applied for an EPA climate grant and was not selected. The St. Paul project remains unfunded.
Soil pH correction is far cheaper than infrastructure, but it is not free, and the farmers who would benefit most from liming are frequently the least able to fund it. Any honest version of this story has to hold both facts at once: the geochemistry is stronger than the marketing suggested, and the delivery problem is unsolved.
What this means for gardeners and growers
Nothing here is a reason for a home gardener to start buying rock dust by the ton. The scale economics that make sense across the Mississippi River Basin do not exist on a domestic plot, and the local-dependence caveats apply with full force.
What is transferable is the diagnostic habit. If you have never tested your soil pH, the most actionable carbon-related action available to you is very likely not a geological one. Healthy soil pH supports plant growth, microbial activity, and water retention, and getting it right is measurable in a single afternoon with a cheap test. Soil is one of the largest available carbon stores, and pH governs how much of that carbon a soil can hold.
For larger landholders, the practical takeaway is that a soil test is a carbon measurement. The practices that raise pH and those that build organic matter are unusually well aligned, and the emerging accounting frameworks may eventually reward both from the same balance sheet.
The bigger question
The most interesting thing about the Yale study is not the tonnage. It is the method. It took a practice that is old, cheap, unglamorous, and already fully deployed, and measured it properly against reality instead of against an idealized baseline.
That is the opposite of how most of the carbon removal industry operates, and the sector’s preference for novel mineral feedstocks delivered by new logistics networks is precisely what makes the result uncomfortable. The cheapest, fastest, and best-evidenced version of enhanced weathering may turn out to be the one already sitting in a quarry truck.
Whether that translates into anything depends less on further geochemistry than on two administrative questions: whether climate accounting bodies revise a default emission factor set in 2006 and whether anyone can work out how to fund lime for the farmers who need it most.
Frequently asked questions about agricultural liming and carbon
Does spreading limestone on farmland release carbon dioxide?
Partly, and this is the crux of the debate. Limestone contains carbon that was locked into the rock when it formed, and when it weathers some of that carbon is released. CarbonPlan’s account of the chemistry notes that one unit of ag-lime can produce up to two units of inorganic carbon, only one of which is drawn from the atmosphere. On that strict reading, liming can be a net source. This is why the IPCC default treats applied lime as an emission, and why the Yale study’s counterfactual approach, rather than a chemical argument, is what changes the result.
Is agricultural liming the same as enhanced rock weathering?
Mechanistically, very similar. Both spread crushed rock on farmland, both raise soil pH, and both can move carbon into long-lived storage. The difference is the rock. Farmers usually apply carbonate, which is carbon-bearing. Enhanced weathering companies usually promote silicate rock such as basalt or olivine, which contains no carbon. CarbonPlan argues that at the technical level they are the same activity pursued for different goals.
Does lime count as a carbon sink in current climate accounting?
Generally no. The 2006 IPCC Guidelines assign liming a default emission factor of 0.12 tons of carbon per ton of limestone and 0.13 tons of carbon per ton of dolomite, which converts to roughly 0.44 and 0.48 tons of carbon dioxide per ton of material. Those are Tier 1 defaults, chosen because they are the carbonate’s carbon content and can be applied uniformly without site-specific data, and they mean spreading lime is booked as a carbon cost. The Yale study argues a more complete framework could credit situations where liming removes carbon in net terms, but no default has been revised.
Is enhanced rock weathering as effective as limestone?
Possibly not, and this is the finding most likely to cause friction. Modeling by CarbonPlan using the SCEPTER reactive transport model found that current liming practices can remove more carbon than basalt projects, particularly in less acidic conditions. Separately, the evidence base for enhanced weathering is wide and soft: CarbonPlan found removal estimates spanning four orders of magnitude, with 12 of 116 cases consistent with zero removal or net emissions.
How much carbon did liming remove in the Mississippi River Basin?
The study estimates between 300 and 400 million metric tons of carbon dioxide since 1900, across a region covering 41% of the contiguous United States and roughly 65% of US croplands. This is a cumulative net figure measured against a realistic no-lime counterfactual, not a gross tally, and the researchers stress that results depend on local soil acidity, buffering, and hydrology.
Could I use this technique on a small farm or allotment?
You already can, if your soil needs it. Correcting soil pH is standard practice and genuinely beneficial to plant health regardless of any carbon consideration. What does not transfer is the carbon accounting. The scale economics, the century-long record, and the funding models all depend on agricultural volumes, and the researchers were explicit that the Mississippi River Basin findings will not transfer directly to every setting.
For the broader context on how soil can be used as a climate tool, see how researchers recommend returning farmland to its natural state and how microbe-coated seeds aim to boost plant carbon capture.
