Solar Geoengineering Risks
The Dimmer Switch We Can't Turn Off
Scientists have a plan to cool the planet that works almost too well — and that's exactly the problem.
The Idea
Solar geoengineering, specifically stratospheric aerosol injection (SAI), involves releasing reflective particles — typically sulfur dioxide — into the upper atmosphere to scatter a fraction of incoming sunlight back into space. It mimics what large volcanic eruptions do naturally: after Mount Pinatubo erupted in 1991, global temperatures dropped by roughly half a degree Celsius for about two years. The logic is seductive. If we can replicate that effect on demand, we buy time while the hard work of decarbonisation catches up. But here's the twist that keeps climate scientists up at night: SAI doesn't fix the underlying problem. It masks it. Carbon dioxide levels keep rising; the oceans keep acidifying. What you get instead is a kind of planetary thermostat held in place by continuous human intervention — one that requires perpetual, globally coordinated maintenance to keep running. The deeper risk is what researchers call 'termination shock.' If SAI were ever abruptly stopped — due to geopolitical breakdown, funding failure, or conflict — temperatures could rebound violently within years rather than decades, hitting ecosystems and agricultural systems with a rate of change they'd have no time to adapt to. You'd have traded a slow-moving crisis for a sudden one. And because SAI affects precipitation patterns unevenly, some regions might face drought while others benefit — turning climate intervention into a geopolitical flashpoint before it's even begun.
In the World
In 2021, a small research team at Harvard attempted to launch a test balloon from Swedish territory under a project called SCoPEx — Stratospheric Controlled Perturbation Experiment. The plan was modest: lift a gondola carrying instruments into the stratosphere to study how aerosol particles disperse, without actually releasing anything. No sulfur, no cooling, just observation. The launch was cancelled anyway. The Swedish Space Corporation withdrew its support after objections from Indigenous Saami groups and environmental organisations, who argued that even a preparatory experiment normalised a technology whose global consequences no one had consented to. It was a striking moment: a balloon carrying no chemicals, conducting no intervention, grounded by the weight of what it represented. The episode crystallised a governance problem that the technology hasn't solved and possibly can't. Who decides whether to dim the sun? A coalition of wealthy nations? A UN body? The country most immediately threatened by sea-level rise? These are not hypothetical questions — a small group of determined actors could, in principle, begin low-level SAI unilaterally. The cost and technical barrier are lower than almost any other planet-scale intervention. Researcher David Keith, one of the field's most prominent voices and a Harvard professor, has spent years arguing that the risks of not researching SAI are greater than the risks of studying it — while simultaneously acknowledging that the governance frameworks needed to deploy it responsibly don't yet exist, and may be harder to build than the technology itself.
Why It Matters
Solar geoengineering sits at the intersection of two things humans are genuinely bad at: long-term coordination under pressure, and resisting the appeal of a quick fix when the slow fix feels impossibly hard. The climate crisis is already producing exactly the conditions — urgency, fear, asymmetric vulnerability — that make unilateral action tempting. Understanding SAI isn't just for scientists or policymakers. It reframes a question most of us thought we understood: 'Can technology save the planet?' The honest answer is that technology can change the planet, dramatically and quickly, in ways that may solve one problem while creating three others. The question of whether that counts as saving depends entirely on who you ask and where they live. It also sharpens something more personal. Every time a complex system is propped up by a technical patch rather than a structural fix — in infrastructure, in finance, in health — the same termination shock logic applies. The patch becomes load-bearing. Knowing that pattern makes you a more clear-eyed reader of any proposal that promises to manage a problem rather than resolve it.
A Question to Ponder
If a technology could genuinely prevent mass suffering but required permanent global cooperation to keep running safely, would the attempt be worth the risk of what happens when that cooperation eventually breaks down?
Get a new one of these every morning.
Start learning with Thinkable