If we can’t reduce greenhouse-gas emissions fast enough to ward off catastrophic climate change, geoengineering may offer a fallback plan. Serious researchers are increasingly exploring measures, such as spraying tiny particles into the air to reflect more sunlight back into space, that wouldn’t reduce emissions but could offset the rise in global temperatures.

A study published January 22 in Nature Ecology & Evolution suggests this is a bad idea. If the world ever starts doing geoengineering, the study warns, it may be too dangerous to ever stop.

The problem, explains the paper, is that deliberately cooling the planet would mask any additional warming produced by greenhouse gases. This means that if the world decided to stop geoengineering, say, 50 years later, the greenhouse effect that had built up during that time would warm the planet very rapidly. 

In many areas, temperatures would rise two to four times faster than historical averages, the study found. Change would occur too fast for many plants and animals to migrate to new areas, fragmenting ecosystems and driving species extinct. It could also reduce rain across the Amazon, Northern Europe, and Asia and increase the number and severity of tropical forest fires.

Some scientists have argued that this risk—which has been pointed out before—is a “sword of Damocles” that should deter us from seriously considering the idea (see “A Cheap and Easy Plan to Stop Global Warming”).

But several factors would reduce the danger of “sudden termination.”

First, the threat arises only if we conduct geoengineering on a large scale and for a long time.

Second, most proponents of geoengineering research say it has to happen in addition to, not instead of, reducing greenhouse gases. They see it as a stopgap measure that gives the world more time to move to a cleaner energy system. Fifty years from now, it’s likely that we will have made far bigger strides in doing so, as well as developing better ways of removing carbon dioxide from the sky.

Finally, if the world does embark on large-scale geoengineering, it’s unlikely to shut down suddenly.

In an op-ed accompanying the study, Phil Williamson of the University of East Anglia argued that this could happen: “Political decision-making is inherently inconsistent over time and frequently affected by regime change,” he wrote. But there’s wide agreement that any decision to move forward with efforts that could affect the entire globe should be made through an open, transparent process conducted by an international body such as the United Nations. That would at least provide a buffer against rash decisions based on shifting sentiments or new leadership in any single nation.

It’s possible some rogue actor might go it alone: a Pacific island nation desperate to avoid sinking below the waves, for example, or an equatorial country struggling through a devastating drought. But again, this would create a big “termination risk” only if the rest of the world allowed the geoengineering to go on for decades.

Ken Caldeira, a senior scientist at the Carnegie Institution who has long studied the potential effects of geoengineering, says the mere fact that halting it would have disastrous consequences would be a pretty good reason not to halt it.

“Usually, if something catastrophic were to happen if we stopped doing something, that is taken as a reason to continue doing that thing,” he said in an e-mail. Clearly the world does make poor decisions sometimes, but usually it needs a good short-term reason to do so, like the threat of war.

David Keith, a Harvard professor who has closely studied solar geoengineering, has long argued that if we ever decide to use these tools, we should start slowly and ramp up gradually over time. That would allow scientists to closely monitor changes, including the unexpected kind, and adjust accordingly. Geoengineering could have environmental side effects and uneven impacts in different regions, but those risks would probably become clear early on. Climate engineering could also be tapered off slowly in a way that would allow us to manage the impact on ecosystems.

Caldeira adds that most climate simulations suggest biodiversity will be under greater threat from unchecked climate change than from geoengineering. “Solar [geoengineering] in models offsets most of the climate change in most places most of the time,” he said.