In 11 days, on Earth Day, world leaders will assemble at the United Nations in New York to sign the Paris climate agreement. That document pledges to hold the planet’s warming to “well below” 2 degrees Celsius above pre-industrial levels, and even to aspire to a 1.5 degree C temperature limit.

The urgency of signing the agreement has been underscored by recent climate news and events, including devastating coral bleaching around the world, newly shattered temperature records and disturbing news about the vulnerability of Arctic permafrost and the Antarctic ice sheet.

But there’s a problem: It is far from clear that, even if governments sign on to the Paris agreement and start implementing it rapidly, they actually know how to limit warming to 2 or 1.5 degrees Celsius. There are a number of problems with thinking that anyone does, argues Glen Peters, a researcher with the Center for International Climate and Environmental Research in Oslo, Norway, in the latest installment of Nature Climate Change.

1. Knowing when we cross the threshold. The first problem is simply with knowing when the world is actually 1.5 degrees C or 2 degrees C above a pre-industrial baseline temperature, often taken to be the average between the years 1850 and 1900 (though this, too, is disputed). Indeed, some have noted that we breached the 1.5 degree threshold in February of 2016, albeit only for the space of a single month (which probably isn’t what scientists have in mind when they think of truly crossing a climate threshold).

The problem is both that it will be hard to define where the actual threshold lies, and also hard to be sure when we’ve crossed it, given differing baselines and periods of analysis, and the fact that temperatures will always fluctuate up and down. And there’s an even bigger problem, which involves so-called “overshoot” scenarios in which the world emits too much and presumably warms up more than 1.5 or 2 degrees Celsius, but then starts to mass-produce “negative emissions” technologies that pull some carbon dioxide back out of the air again.

If this happens, we also will not know, for some time, how long it will actually take to get back to 1.5 or 2 degrees C, moving in the opposite direction, once we’ve overshot. “It may not be known for many decades if 1.5C/2C has been exceeded or successfully avoided,” Peters notes.

2. Knowing how much we can emit. And then there’s the problem of the carbon “budget,” which the U.N.’s Intergovernmental Panel on Climate Change defined as no more than 1,000 billion tons, or gigatons, of carbon dioxide emissions (emitted from the year 2011) if we want a 66 percent chance or better of holding warming to 2 degrees C.

This number, too, is shot through with uncertainty. For instance, Peters notes that if you can tolerate less assurance of reaching 2 C, then you can suddenly allow for the possibility of vastly higher emissions: “The total quota for a 2C threshold increases by 800 GtCo2 for a decrease in the likelihood from 66% to 50%,” he notes. He also adds that there are big uncertainties related to the role of non-carbon dioxide gases, like methane, in shaping these budgets.

These uncertainties mean that, from the year 2016 and on, for a 66 percent likelihood of staying under 2C, the carbon dioxide budget is about 850 billion tons, Peters calculates — but that’s plus or minus 450 billion tons. That’s a pretty big uncertainty in a situation where we’re emitting well over 30 billion tons of carbon dioxide per year, and thus close to 100 billion tons every 3 years.

“It is likely that many of the uncertainties on the remaining quota will remain persistently large,” Peters writes.

3. Knowing the future. Finally, the so-called “integrated assessment models” that are employed to determine if and how we can actually stay under a given temperature threshold often include the assumption of new technologies that don’t exist yet, or new policies that aren’t in place, to help us get there. An example of the former would be massive carbon dioxide removal from the air — and an example of the latter would be setting a global price on carbon emissions.

These technologies and policies could develop in the future — but then, they also might not. “Nearly all the literature informing global climate policy uses these strong policy assumptions,” Peters writes, referring to global carbon taxes included in models. “There is an urgent need for scenarios based on more realistic policy assumptions, in addition to a broader range of technological pathways that capture political realities.”

The overall upshot, then, is pretty surprising, although couched in relatively dry academic language: “there are many key scientific knowledge gaps to be resolved before one can say, with confidence, whether 1.5 C or 2C are realistic temperature goals,” Peters writes. In other words, the whole world could be about to sign on to an international agreement that pledges targets that we may not be able to achieve.

It would be one thing if Peters was the only analyst making such observations. Unfortunately, he isn’t. Kevin Anderson and Oliver Geden, two other researchers with the University of Manchester and the German Institute for International and Security Affairs, have been making similar critiques.

Geden, writing last week in Nature Geoscience, actually argued that we should drop the 1.5 C and 2 C temperature targets in favor of a goal of achieving “net zero emissions” globally. And a recent paper in Nature Climate Change just proposed that the current remaining carbon budget could be lower than previously thought, lying in a range from 590 to 1,240 gigatons.

This whole discussion about carbon budgets and temperature thresholds may seem complex and arcane. But the reality is that these are the best metrics available when it comes to measuring how far we’ve already gone toward a place that nobody wants to reach. Yet, at the same time, it’s clear that they may lend a false air of precision — an incorrect sense that scientists know more than they actually do about how humans are steering the planet.

The U.N.’s Intergovernmental Panel on Climate Change was asked, as part of the Paris process, to undertake a “special report,” by 2018, on “the impacts of global warming of 1.5 °C above pre-industrial levels and related global greenhouse gas emission pathways.” But assuming it takes up this task, the report may simply find that there’s little or no way to get there.

“Basically the only way in these scenarios to get to 1.5 is immediate implementation of policies around the world and large negative emissions,” says Peters. “And I don’t think either of those things are feasible.”

Read more at Energy & Environment:

For more, you can sign up for our weekly newsletter here, and follow us on Twitter here.