When the Greenland ice sheet went into a record meltdown in the summer of 2019, it raised a very terrifying specter of the future. Here was a 12.5-billion-ton mass of ice—one that’s been melting at a quickening pace since the 1980s—melting in a way scientists didn’t expect to happen for decades.

While the ice sheet won’t completely disappear for centuries, any further increase in its melt will put coastal communities at risk of inundation. There’s an argument to be made that we should do everything possible to save the ice, and a new study explores a very controversial idea to that end: cooling the planet.

The findings, published last month in Earth’s Future, explore what would happen if the world pumped particles high into the atmosphere that would reflect sunlight back into space. This high-altitude air conditioning scheme, known as solar radiation management or SRM, would bring down the global average temperature. The paper’s results show that cooling would help slow—though not stop—the melting of the ice sheet. That could buy coastal regions time but also change the climate in other ways that may end up hurting other regions around the world.

The threats to Greenland ice are coming from seemingly all directions. Relatively hot water has undercut the glaciers tumbling down from the ice sheet, while rising air temperatures have melted it from above. Recent wildfires have also left a cake of dark soot that absorbs more sunlight, enhancing surface melt. And then there are surface pools of water that funnel into cracks in the ice, further destabilizing it. In short, climate change is engaged in an all-out war on the ice sheet that will only intensify in the coming century. Projections indicate the ice sheet melting could lead to anywhere from 10 to 20 centimeters (4 to 8 inches) of sea level rise this century, with much higher rates of rise in the Pacific.

To understand the impact geoengineering could have on Greenland, scientists ran models that simulate the climate throughout the century under a few different scenarios. In some scenarios, greenhouse gas pollution increased throughout the century or peaked by mid-century. They then took those same scenarios and added in a splash of geoengineering. The models they ran basically shot particles into the stratosphere around the equator that then dispersed around the globe. The impacts are based in part on what happens after volcanic eruptions, and the study used the equivalent a quarter of a Pinatubo eruption every year.

The results show that the planet-cooling properties of the particles would extend to Greenland, lowering air temperatures there by about 1.1 degrees Celsius (2 degrees Fahrenheit) compared to the scenarios without geoengineering. Runoff from melt would drop anywhere from 20 to 32 percent, with the biggest difference coming under the high greenhouse gas emissions scenario. All that means sea level rise would slow.

Doug MacMartin, a Cornell scientist who has studied geoengineering, told Earther the study’s findings were in line with you’d expect from solar radiation management but that the findings used older models. Future work updated with new models could go a long way to building on the findings, and the study itself notes that some key aspects of atmospheric circulation are simplified.

But it’s not all puppies and unicorns. Geoengineering alone wouldn’t be able to stop ice from melting. It could also wreak havoc with weather in other parts of the globe, potentially pitting nations against each other. And then there’s the most concerning aspect of geoengineering, which is that once we start in on cooling the planet, stopping it could risk sudden, catastrophic climate change. So any plan to do this requires a long, hard think and making sure everyone has a say, not just rich countries with tons of coastal real estate.

“Could geoengineering affect the Arctic? How effective would it be at offsetting projected polar changes due to climate change? Would there be any side effects or unintended consequences?” Ben Kravitz, a geoengineering researcher at Indiana University, told Earther in an email. “Answering these questions is critical, and this study is a welcome step in addressing such questions.”

Indeed, as the world continues to put off the best climate change solution (which would be cutting emissions), it moves closer to considering more risky ways to address the changes set in motion by decades of unbridled fossil fuel use. That makes exploring these ideas in the scientific literature terrifying and important all at once.

“We don’t know enough about [solar radiation management] today, but we don’t want to be in a position of making hasty choices some point in the future—which is exactly why studies like this one are so important,” MacMartin told Earther in an email.

Equally important, and something that MacMartin emphasized as well, is actually doing the hard work of cutting carbon emissions. As the results make clear, blocking out sunlight is nowhere near enough to protect the coasts.