Could – and should – we thin clouds to cool the planet?
On a clear day, the highest clouds can look like brushstrokes or feathers. These wispy streaks of ice drifting miles above the ground are cirrus clouds, and despite their gossamer appearance, they play a big role regulating Earth’s temperature.
In theory, modifying cirrus clouds could cool the planet. Some scientists are studying how intentionally thinning them would let more heat escape to space. It’s called cirrus cloud thinning (CCT), and it’s part of the broader, controversial family of proposals known as solar geoengineering or solar radiation modification.
“It’s important to remember, this is not a substitute for reducing greenhouse gas emissions. Many questions remain about the side effects of geoengineering,” says Harvard atmospheric scientist Zhiming Kuang.
Kuang is studying how cirrus clouds behave, how they form, and potential unintended consequences from changing them – like impacts on rainfall patterns. Major questions also remain, he adds, about how to govern any interventions: Who decides when everyone is potentially affected?
Clouds that both cool and heat
Low clouds mostly cool the planet. Cirrus clouds are different. Formed from ice crystals high in the atmosphere, they reflect some sunlight, but they also trap outgoing heat. In many cases, that blanket effect dominates, so cirrus can cause more warming than cooling.
What tips the balance is cirrus thickness, and thickness is shaped by how the ice crystals form. Up high, ice formation (nucleation) can happen with helpers or without. In heterogeneous nucleation, tiny “seed” particles like dust allow water vapor to latch on early; the result is fewer, bigger crystals that fall out faster, leaving a thinner, shorter-lived cloud. In homogeneous nucleation, the air has to get extremely saturated before ice forms from tiny water droplets ubiquitous in the atmosphere; this results in many more, smaller crystals that stay suspended longer, making the cloud thicker and longer-lived.
Kuang is investigating that boundary between the types of ice formation.
“If you reach the threshold, if you get to the regime of homogeneous nucleation, you will produce more numerous ice crystals. And that matters because if you have the same amount of water that’s condensing into more ice crystals, they will be smaller. Smaller ice crystals will fall [to Earth] more slowly, and the cirrus cloud persists for longer,” he says.
Would cirrus thinning work?
The idea is often discussed in terms of deliberately releasing tiny particles – such as mineral dust – into the upper troposphere (where cirrus clouds form) to trigger ice formation (heterogeneous nucleation).
But the same “seeds” can push cirrus clouds in opposite directions.
Adding ice-nucleating particles can sometimes prevent a burst of lots of tiny ice crystals, making the cloud thinner and shorter-lived. But in other situations – when the air is cooling slowly, for example, or in a cloud where the ice nucleation is already heterogeneous – adding particles can create more ice crystals and make cirrus clouds last longer or trap more heat, so the intervention could warm the planet instead of cooling it.
“Which one happens more often? That’s the question that needs to be addressed,” says Kuang.
Even if scientists understood the microphysics perfectly, the real atmosphere makes it tough to apply. Cirrus clouds form in different ways, and in fast-changing situations (like during a strong convective updraft of warm, moist air) the air can cool so quickly that seeding may not change what happens; while in other situations, the effect of seeding may depend on temperature changes over periods as short as minutes. Measuring cirrus is also tricky: Aircraft instruments can accidentally create fake small-crystal signals from shattered ice crystals, which can mislead researchers about what the clouds really contain.
“What happens when they fall back?”
The idea behind cirrus cloud thinning is a reminder of how easily a targeted intervention can collide with a system that is dynamic, messy, and not fully understood. Try to thin cirrus, and you might thicken it instead. And even if thinning is possible, it would require reliably predicting where the right kind of cirrus is forming and when it’s vulnerable to change.
Scientists also want to understand cirrus nucleation for reasons beyond cirrus thinning.
One is jet contrails, which can behave like cirrus and trap heat. A better understanding could help guide flight strategies that reduce persistent contrails. Another is tied to stratospheric aerosol injection (SAI), a solar geoengineering proposal that would add reflective aerosols to the stratosphere. Those particles may linger for about a year, Kuang notes, but eventually they drift back down through the troposphere, where they could affect cirrus formation. “The questions you want to ask is, what happens when they fall back? How do they modify the cirrus clouds? And what are the climate implications?”
