This undated handout photo provided by NASA shows the Thwaites Glacier in West Antarctica. (NASA via AP)

For some time now, fears of climate disaster have been at least partly assuaged by the thought that if the planet really begins to heat up, well, at least we may have a backup plan.

That backup plan is so-called “geoengineering” — artificially altering the planet still further so as to offset warming temperatures. One leading idea in the space is to fill the Earth’s stratosphere with sulfate aerosol particles, which would have a cooling effect by reflecting sunlight back to space. We know this would work because we know that large volcanic eruptions cool the planet, and that they do so by a similar mechanism — firing sulfur high into the skies.

But it’s also very risky — there are many possible unintended consequences of geoengineering. Thus, the only reason to really consider it is if you’re on the verge of climate impacts so severe — impacts like, say, the potential collapse of the West Antarctic ice sheet, leading to 10 or more feet of global sea level rise — that it becomes the lesser evil.

That’s why a new study recently accepted in Geophysical Research Letters could be so significant. For it calls into question whether geoengineering — at least using sulfate aerosols — can actually save this ice sheet, which is already beginning to be destabilized in our warming world.

The new paper, by Kelly McCusker of the University of Victoria in British Columbia, Canada, and two colleagues from the University of Washington in Seattle, uses a climate and ocean model simulation to examine the fate of West Antarctica in a world in which humans pump huge volumes of sulfate aerosols into the atmosphere to curb global warming. And it finds that while the planet would indeed cool in such a scenario, West Antarctica would continue to melt, especially in the region of the vulnerable Pine Island glacier.

Thus, the authors conclude, the notion that this form of geoengineering can serve as “a ‘backstop’ measure that could be rapidly deployed to avoid so-called climate emergencies, such as destabilization of marine ice sheets” is “not supported in this study.” Indeed, the article’s title puts it even more bluntly: “Stratospheric sulfate aerosol injections cannot preserve the West Antarctic Ice Sheet.”

But why wouldn’t a cooler, geoengineered Earth let us preserve West Antarctica? To understand this result, you first need to know a little more about what’s going on with this part of the vast Antarctic ice sheet, which is smaller than East Antarctica but still large enough to fundamentally remake coasts the world over if its ice falls into the sea.

West Antarctica is losing ice because warmer water is getting underneath its oceanfront ice shelves, and melting them from below. These ice shelves hold back huge volumes of landlocked ice, but the more they retreat, the faster ice can slide into the sea and drive sea level rise.

But this isn’t a simple story about the globe’s warming surface oceans. Rather, the new study says, West Antarctica’s huge Thwaites and Pine Island glaciers are melting due to shifting ocean currents. In particular, so-called circumpolar deep water — which is actually warmer than the surface seas above it, since it arrives in the Antarctic region from elsewhere, comprising “a mixture of deep water from all oceans” — is pushing up from below to reach the base of these glaciers, and causing them to melt.

That, in turn, raises the question of what’s driving these changed ocean currents — and the answer lies in the complex dynamics of the atmosphere. According to McCusker and her colleagues, these currents are driven by shifts in atmospheric polar winds that are, in turn, shaped by how global warming affects temperatures in the atmosphere. In essence, the southern polar jet stream shifts further poleward due to global warming, subsequently changing ocean currents in the area and driving responses in circumpolar deep water that lead to more melting of West Antarctic glaciers.

It is precisely this mechanism, the study found, that geoengineering fails to adequately address.

In the climate simulation employed by McCusker and her colleagues, in the year 2035, humans begin pumping 8 million metric tons of sulfate per year into the atmosphere for three years, and then increase it further annually after that. This definitely works to cool down the planet, ramping temperatures down to late 20th century levels within a decade. But the glaciers of West Antarctica continue to melt.

The study contrasts this outcome with an admittedly impossible but still instructive scenario — one in which, in the year 2035, concentrations of carbon dioxide suddenly and inexplicably go back to their 1850 levels. This, too, cools the planet rapidly. But it would also help stabilize the glaciers.

The reason for the contrasting outcomes is that in the geoengineering scenario, elevated levels of atmospheric carbon dioxide still affect the southern hemisphere polar jet stream in such a way as to drive warm water towards the Antarctic glaciers. Thus, greenhouse gas removal “is a more effective means for cooling subsurface ocean temperatures than stratospheric aerosol injection fundamentally because of changes in ocean circulation that are brought about by changes in atmospheric circulation,” the authors write.

“If you allow the world to warm, and then you put in a lot of stratospheric sulfate aerosols with the intention of avoiding some very severe climate impacts,” says McCusker, “we find that because you can’t totally counteract the circulation changes, you can’t counteract the availability of warm water to be near the ice shelf outlets.”

Granted, all of this is happening in a single climate modeling study, so you probably need to take it with a grain of salt, suggests Ken Caldeira, a climate researcher at the Carnegie Institution for Science in Stanford, Calif., who has published widely on geoengineering. Caldeira suggests that different climate model simulations, using different amounts or deployments of sulfate aerosol, might show different results. He also suggests that other particles than sulfate might be used in geoengineering endeavors, potentially leading to different outcomes.

“It is very good that the authors have raised the question of the ability of solar geoengineering to preserve the Antarctic ice sheet,” said Caldeira by e-mail. “The answer to this question is far from settled.”

Still, the new research serves as a reminder that a world in which the stratosphere is filled with sulfate aerosols is not simply one in which overall temperatures decrease. Other things happen, too. Previous research, for instance, has suggested that there may be less rainfall in such a world — another downside. Moreover, once you start geoengineering, you can’t stop again, or else temperatures could bounce back super quickly.

All of which reinforces a point that even scientists studying geoengineering, just in case we might someday need it, repeatedly stress — there’s just no substitute for emitting less carbon dioxide, or even trying to find ways to pull it back out of the atmosphere.