Photo: Jenny Laird (Flickr)

Solar geoengineering—altering the atmosphere to reflect more sunlight back into space and cool the planet—has moved from the scientific fringes into mainstream climate mitigation discourse. The idea is untested, and may prove catastrophic. But a pair of researchers are now pushing back against one of its biggest criticisms: that, if a solar geoengineering scheme was aborted, it’d lead to a sudden and rapid spike in global temperatures.

Not necessarily, they say. Writing this week in Earth’s Future, Andy Parker of the Institute for Advanced Sustainability Studies in Postdam, Germany, and Peter Irvine of Harvard’s engineering school note that there’s been little consideration of how likely a so-called “termination shock” really is. They argue that any future solar geoengineering projects need to be geared toward reducing the risk of such a catastrophe as much as possible.

The most oft-discussed solar geoengineering idea is something called stratospheric aerosol injection (SAI), where we spray a bunch of sunlight-reflecting particles into the upper atmosphere to reduce the amount of incoming radiation hitting the Earth’s surface. Trouble is, if we went whole-hog on SAI without reducing carbon emissions, greenhouse gases would continue to accumulate in our atmosphere, meaning we’d need to keep pumping particles skyward forever to keep global warming at bay.

If we stopped, the planet’s temperature would start to rise again, possibly at a rate much faster than humans or natural ecosystems could adapt. Widespread ecological chaos and destruction would ensue. That’s the termination shock.

It’s also a worst-case scenario. “We noticed that in the analysis on the termination shock, people were looking at modeled worst case scenarios,” Parker told Earther. That makes sense. “People are familiar with the idea that if you stop [solar geoengineering], it could be bad, so they’ve tried to model how bad.”


Parker and Irvine’s new paper is a first stab at how to make it less bad. They explain that termination shock is most likely to occur if a solar geoengineering effort aimed at suppressing a large amount of warming—say, the 0.5 degrees Celsius per decade expected in the high-end (RPC 8.5) carbon emissions scenario—was phased out suddenly and completely. But if we used solar geoengineering to offset a smaller amount of warming, perhaps just a few tenths of a degree, they say it might be phased down over half a century without catastrophic impacts. Even smaller amounts of SAI might be halted suddenly with few impacts at all.

Perhaps humanity would take that tack if we were trying to limit global warming—say, we had already passed the Paris Agreement’s aspirational 1.5 degree warming threshold and were trying not to broach the more dangerous 2 degree mark—while at the same time working toward rapid decarbonization.


Even if a large-scale SAI project were aborted suddenly, the termination shock wouldn’t be immediate. It could take months to a year for enough of those sun-scattering particles to fall out of the stratosphere for temperatures to begin rising again, meaning if the project was aborted accidentally—by a terrorist attack, for instance—we could restart. We’d just need to be smart enough to keep some backup systems around.

The paper identifies some scarier scenarios that could cause an SAI effort to be terminated inadvertently, including a large-scale economic collapse, or global warfare. Again, the authors argue that backup systems scattered across multiple nations would reduce the likelihood of termination shock, although the risk can’t be eliminated entirely.

Parker emphasized that while there’s still a great deal of uncertainty, the safest SAI approach would almost certainly involve broad international support. “If one country races off and does [solar radiation management] unilaterally, and they turn it up high straight away, chances [of termination shock] are much much higher than if it was turned on slowly,” he told Earther.


Of course, a shift in the political winds could cause an SAI project supported by many nations to be suddenly abandoned, especially if some nations began suffering damages from its use. While Parker and Irvine still consider sudden termination an unlikely outcome in this scenario, they note that it could have “significant justice implications,” especially if “those suffering were poor and vulnerable.”

Parker hopes the new paper sparks follow-up research aimed at refining how much solar geoengineering could be done without risking a damaging bounceback effect. He also emphasized that bringing down global carbon emissions must continue to be our number one climate priority.

At the same time, he said, the world is talking about geoengineering. We need better modeling of the risks in order to prevent people’s worst fears from being realized.


“The genie’s out of the bottle,” he said. “And I don’t believe not talking about it is gonna make the idea go away.”