Four of the world's most credentialed climate scientists just published a warning that solar geoengineering, the increasingly fashionable idea of blasting reflective particles into the stratosphere to cool the planet, could trigger a temperature swing of up to 30 degrees Celsius if it ever gets switched off. That's not a typo. Thirty degrees. For context, the entire difference between an ice age and today is about 5 degrees.
What Is Solar Geoengineering and Why Is Everyone Suddenly Talking About It
The pitch sounds simple enough. We're not cutting carbon emissions fast enough, so why not buy ourselves some time by dimming the sun a little? Inject sulfur dioxide or other reflective aerosols into the stratosphere, bounce some sunlight back into space, and voila, cooler planet. Problem managed.
The Guardian recently ran a series declaring it's time to seriously talk about geoengineering. So a group of four scientists, Raymond Pierrehumbert, Julia Slingo, Michael Mann, and Valerie Masson-Delmotte, who between them have spent over a century studying climate physics, took the invitation seriously. Their response, also published in the Guardian, is essentially a very polite way of saying: please stop before you get us all killed.
Beyond the stratospheric aerosol injection proposals getting the most attention, there's a whole carnival of other schemes being floated. Proposals to interfere with polar environments. Wildly expensive plans to dam the Bering Strait. The ambition is genuinely staggering. The scientific foundations, according to these four, are not.
The Termination Shock Problem Nobody Wants to Talk About
Here's the core problem with solar geoengineering and it's a genuinely terrifying one. Carbon dioxide, once you put it in the atmosphere, sticks around for millennia. A meaningful chunk of what we're emitting today will still be warming the planet thousands of years from now. That warming is locked in, building up like pressure behind a dam.
Solar geoengineering masks that pressure rather than releasing it. The aerosols you inject decay in years, not centuries. So if you start a solar geoengineering program and then, for any reason, you have to stop it, all that pent-up warming gets released at once. The scientists writing in the Guardian call this "termination shock," and their modelling suggests the temperature swing could be anywhere from less than 1 degree Celsius to as much as 30 degrees Celsius within just a decade of shutdown.
Let that range sink in. After 10 years of the same stratospheric aerosol injection program, models cannot agree on whether you get a minor correction or a climate catastrophe that dwarfs anything industrial humanity has produced so far. As the scientists put it: "We are essentially flying blind."
And here's the kicker on the "we can just turn it off if it goes wrong" argument. According to their analysis, it would take up to two decades just to build the infrastructure to run one of these programs. By the time you've built it, you're completely dependent on maintaining it, indefinitely, across generations, through wars, economic collapses, and political upheavals. Forever, basically. Or face the shock.
The Funding Is Going to Deployment, Not Safety
You might assume that given these stakes, the research money flowing into geoengineering is focused on understanding whether it's safe. You would be wrong.
The four scientists are explicit about this. What the current wave of funding is actually targeting, they write in the Guardian, is engineering technology for deployment. Not the scientific foundations. Not rigorous modelling of what actually happens when you start injecting things into the stratosphere at scale. Deployment. The juggernaut, as they put it, is rolling regardless of what the damage might be.
There has been no formal intercomparison study examining how sensitive the climate is to these interventions. No rigorous modelling assessment exploring different solar geoengineering scenarios. The models that have been used don't even agree on basic questions like how much intervention would be required or what the regional effects on rainfall would be. Aerosols, clouds, and regional precipitation are among the least well-understood parts of the entire climate system, and that's exactly what these proposals would be messing with.
Small 'Safe' Experiments Won't Tell You What You Need to Know
A common reassurance from geoengineering proponents is that we can just run small-scale experiments first, see what happens, and scale up carefully. The scientists writing in the Guardian call this "fundamentally naive," which in the understated world of academic climate science is roughly equivalent to screaming into a pillow.
The reason small experiments can't answer the big questions is simple physics. The global climate system operates through massive forces: ocean circulation currents that redistribute heat across entire hemispheres, year-to-year fluctuations in cloud patterns that dwarf any signal a small experiment could generate. Any experiment would be drowned out by natural variability. You'd learn essentially nothing about what large-scale deployment would actually do.
The scientists draw a direct comparison to how carefully the IPCC process worked through the science of greenhouse gas emissions. Over six assessment reports since 1990, tens of thousands of scientists worked to understand regional consequences before any serious policy conclusions were drawn. It took more than a century of emissions before the signal was clear enough to detect and attribute with confidence. Now geoengineering proponents want to introduce a new and far more uncertain forcing into that same system, with no comparable scientific foundation in place. The asymmetry is staggering.
This Is a Political Problem as Much as a Scientific One
Imagine the geopolitics of a world where one or more countries are running a stratospheric aerosol injection program that, if disrupted, triggers catastrophic temperature swings across every nation on Earth. Who controls the dial? What happens when a program run by one government produces drought in another country's breadbasket? What happens when the country running it collapses, or decides it no longer wants to?
The scientists raise this directly, warning that geoengineering would introduce a major new destabilizing technology to an already turbulent political climate. There is currently no governance framework for any of this. None. The engineering is being developed faster than any international agreement to manage it could possibly be constructed.
This isn't a hypothetical future problem. The funding is happening now. The technology development is happening now. The governance framework is not.
The Dingo Take
Look, there is a version of the geoengineering conversation worth having. The climate situation is genuinely dire, the emissions trajectory is genuinely terrifying, and it's not unreasonable to want every option on the table. The scientists writing in the Guardian don't even fully close the door. What they're saying is: if we're going to consider this seriously, we need the same scientific rigor we'd apply to anything else with the potential to alter global temperature by 30 degrees. That's not a crazy ask.
What's crazy is that the money is going to deployment technology anyway. What's crazy is that the models don't agree on outcomes by a factor of thirty and the answer from the techno-optimist crowd is apparently "sure, let's build the infrastructure and see." We spent decades being told that climate scientists were being alarmist, that the models were too uncertain, that we shouldn't make trillion-dollar decisions based on imperfect science. Funny how that standard evaporates when the proposal involves shooting aerosols into the stratosphere and crossing our fingers.
Four scientists with over a century of combined climate expertise just published a piece describing the current geoengineering funding wave as a juggernaut rolling forward with complete disregard for potential damage to the planet. That's not a fringe position. That's some of the most credentialed voices in the field telling us, plainly, that we're making a serious mistake. The question is whether anyone with a checkbook is listening.
