Sure, scientists know plenty about the classic greenhouse drivers of climate change, CO2 and methane. But humanity has also been pumping particulates into the system, and these tend to cool things down. Power plants that burn fossil fuels, for instance, release sulfur dioxide, which can lead to the formation of particles in the atmosphere that bounce the sun’s energy back into space. (Which, as it happens, may be a way to geoengineer the planet to counteract climate change. Not by burning more fossil fuels, of course, but by
adding particulates in the atmosphere.)
The researchers’ approach to this study was to combine models, and more models, and then some more—16 total—not with warming trends, but how temperature fluctuated from 1880 to 2016. “Essentially, the models tell us the relationship between temperature variations and climate sensitivity, and the observations tell us the temperature variations in the world," says Cox. "Together they allow us to get better estimates of climate sensitivity for our planet.”
So, the numbers. What the researchers landed on was an ECS range of 2.2 to 3.4°C, compared to the commonly accepted range of 1.5 and 4.5°C. Admittedly, 2.2 on the low end isn’t ideal for the future of our planet. (For each degree of warming, for example, you might expect up to a
400 percent increase in area burned by wildfires in parts of the western US. Very not ideal.) And the researchers say this means the probability of the ECS being less than 1.5°C—the Paris Climate Agreement’s super optimistic goal beyond the 2°C goal—is less than 3 percent. The upside, though, is they say this new estimate means the probability of the ECS passing 4.5°C is less than 1 percent.
But hold up, says Swiss Federal Institute of Technology climate scientist
Reto Knutti, who wasn’t involved in the research. “What's the chance of something fundamentally being wrong in our models?” he asks. “Is that really less than 1 percent? I would argue there's more than a one in a hundred chance that something has been forgotten in all of the models, just because our understanding is incomplete.”
Not that what these researchers have done is bad science. It’s just that global climate change is an exceedingly complex problem. There’s no way any scientist can dig down into all the granular details—changes in vegetation, small-scale hydrology, every single weather event like a hurricane or tornado. So what scientists do is find simplified descriptions of these small-scale events. “For clouds, for instance, you say, 'OK, the more humidity the more likely it is to rain, and if you have more than 95 percent saturation, then you rain,'” says Knutti. “It's an ad hoc way of describing rain without properly describing the process of rain formation, because you can't.”