Diplomats from 197 countries agreed earlier this month to new rules governing how they can buy and sell credits to neutralize carbon emissions. But while they were deliberating, some of the biggest names in climate science, who defined "net zero” in 2009, found something wrong with the math underlying those debates.

"Achieving ‘net zero’ no longer means what we meant by it,” said Myles Allen, professor of geosystem science at University of Oxford, one of the authors of a new paper published last month in the journal Nature.

Their new analysis skewers an assumption at the heart of how countries and companies track emissions — that a ton of carbon dioxide is the same everywhere, whether it’s dispersed in the atmosphere, embedded in forest wood or pulled from the air and pumped deep underground forever. That fungibility is the foundation of carbon markets. It lets a ton of CO2 in a forest stand as a fair trade for a ton put in the atmosphere.

That rule-of-thumb turns out to be a vast oversimplification that could render many well-meaning net-zero efforts meaningless.

The confusion stems from a basic fact about how the Earth’s carbon cycle works: Scientists know what humanity emits into the atmosphere doesn’t entirely stay in the atmosphere. Less than half of that total stays in the atmosphere on average. The rest flows into the land and ocean. To keep track of all that carbon — and how they assign responsibility for removing it — scientists keep two ledgers, one for nature and one for humanity. All the CO2 absorbed every year into land, trees and water is a service the planet offers to wash humanity’s past CO2 emissions out of the air. So, these carbon drawdowns go into the nature ledger.

It’s important to emphasize that land and oceans are drawing down past emissions. That means they cannot be relied on to also neutralize future emissions. This is where the revelation comes in: Countries may have been double counting.

In other words, it’s redundant for countries to claim credit for CO2 for work already being done by land and oceans. Those emissions are already spoken for.

"We can't count on them (emissions) to do two jobs at once. That's the point,” Allen said. "If we're going to count on them to mop up our historical emissions ... we can't at the same time use them to offset future fossil fuel emissions.”

These differences between natural and industrial bookkeeping add up. For example, Allen said, consider a situation in which — using current carbon accounting — the world was expected to stay below 1.5 degrees Celsius. The flaws in accounting are so significant that they could be concealing another 0.5 C rise. (Allen is also chair of the advisory board of Puro.earth, a carbon registry.)

There are consequences of this accounting mismatch. The first is, it increases the urgency to stop burning fossil fuels, the authors write, or to capture and bury pollution with emerging methods. The climate that humanity grew up in relied on millions of years of coal, oil and gas sitting underground. The main solution therefore is to leave it there, capture the carbon from smokestacks and permanently bury it, or clean it out of the open air.

Returning carbon underground is "geological net zero,” and it’s what the authors originally had in mind in 2009. No countries are currently pursuing it.

Separate from fossil fuel burning and carbon capturing, they write, nature must be left alone, to passively soak up history’s CO2. And all that land needs to be conserved, undeveloped, to keep the carbon out of the atmosphere and pull down even more. Rich countries bear historic responsibility for ensuring that happens, they write.

As if this weren’t complicated enough, there is more to the story than two ledgers, with past carbon falling into nature and future carbon captured and stashed underground. That’s because there is value to human management of land that reduces atmospheric CO2. In other words, if "managed land” is proven to take down CO2 then those tons can be counted against emissions, the scientists say.

What "managed land” means is a headache to pin down. Countries have no uniform standard, and often claim all of their land as managed. In fact, so much land is claimed that their combined pledges are virtually impossible to foresee happening. They may be taking credit for emissions already in nature’s ledger.

There are other reasons why storing carbon in the biosphere is inferior to geological storage, they write. As wildfires continue to show every year, there’s nothing permanent about living things. In 2023, the hottest year on record, trees and land absorbed virtually no carbon. Any potential slowdown in the land and ocean carbon sponges would leave a greater amount sitting in the atmosphere, further aggravating warming.

Scrutiny of private carbon markets has led noteworthy players to exit and others to focus on cutting emissions directly. Those speed bumps mean that voluntary carbon markets are ahead of governments in thinking about the problem in some ways, according to Sassan Saatchi, co-founder and CEO of CTrees, a scientific nonprofit that wants to "track carbon in every tree on the planet.”

Saatchi called the paper "a timely warning” even as fixing the problem "is a difficult thing to ask countries to really abide by. The scientific community has to have much better recommendations.”

The paper pulls together into one place a number of concerns scientists have accumulated about land use and carbon accounting, said Pamela McElwee, a Rutgers University professor and contributor to the U.N. Intergovernmental Panel on Climate Change. That includes separating how everyone accounts for natural and industrial CO2 drawdown. Countries should be able deduct CO2 that’s drawn down permanently, back into the Earth, from their gross fossil-fuel emissions. But carbon absorbed by land and oceans doesn’t count as "geological net zero,” and should not be credited against emissions.

"It really needs to be apples to apples and so let's treat it that way,” she said.

It’s difficult to contemplate any reforms that might take already limited funding away from forest management. "If I could be assured that we could do both those things simultaneously” — reach geological net zero and preserve natural drawdown — "that would be ideal.”