Earth is taking in more energy from sunlight than it sends back out to space, and that gap has roughly doubled since the early 2000s. Part of that comes from feedback loops that are making things worse (less sea ice and snow, fewer bright clouds), and part of it, perhaps counterintuitively, comes from cleaning up air pollution that had been reflecting some sunlight back to space. Most of the speed-up shows up as Earth getting darker and reflecting less sunlight. This is increasingly becoming a frightening problem, because the energy imbalance gap is growing faster than climate models expected and the measurements are making the gentlest climate outcomes look less likely.

It’s also why I think we should be putting more resources into monitoring and observation and trying to understand it. According to the CVF bulletin, the satellite record that measures this is aging, with one of NASA’s CERES instruments switched off last year and the Terra satellite’s science mission expected to end early next year. The successor instrument, Libera, doesn’t launch until 2027, and there are currently no plans for anything after it. Meanwhile, the US budget request for next year proposes cutting NASA Earth Science by more than half. The CVF is recommending that its members make continuity of these measurements a formal position going into COP31 in November, pointing out that the countries least responsible for the imbalance are “the most exposed to losing the ability to measure it.”

I wanted to put together a simple microsite for this so that people could understand what is going on in a more accessible way, and it’s now live at earthenergyimbalance.com. At its center is a 90-second animated video, and the rest of the page walks through the latest reading, why the planet is getting darker, why it’s a risk and what can be done about it. Spark Climate, Probable Futures, and Carbon Brief all have good written explainers, and the site links to them, but I wanted something short and visual that you could send to someone who hasn’t heard of EEI yet. (I built the animation with Opus 5.5, after seeing a lot of examples on social media of what people were doing with it.)

One thing I’ve heard over and over again from people who work with policymakers is how critical neutrality is in describing problems. So the site starts with the problem, gets you to understanding what it is and what its implications and consequences are, and only afterwards gets to the things we might do to respond.

The site is really meant for people who are climate aware but not necessarily familiar with this level of nuance. I suspect most of you know a few of them, so please send it their way. And if you spot anything wrong, or anything that confused you, reply to this email or leave a comment and I’ll fix it (if you work on this science, I especially want to hear from you).

Visit earthenergyimbalance.com

I walked into the Explorers Club last Thursday night eager and excited to join the panel for the session “Plan B For the Planet?” (hosted by Marianne Hagen of the Seabed Curtain Project—go listen to our podcast together here) because I knew it would be a room of people extremely interested in understanding the realities of where the climate change situation actually is. The Explorers Club was founded in 1904 to promote scientific research in the field, and its members are a concentrated group of people who value natural systems, conservation, and experiencing the wild diversity of life Earth has to offer. In other words, I could not imagine a group more ready to hear about what we have to do to stabilize the Earth systems most at risk.

And so, to a fully packed room in a building whose walls are lined with the famous Explorers Club flag, elephant tusks, wooden sleds from polar expeditions, and photos and memorabilia from 120+ years of adventurers, I was asked, what does Plan B mean to me?

The crowd was very climate conscious but not necessarily aware of things like tipping point risks or climate interventions. When I mentioned stratospheric aerosol injection (SAI), I literally heard people gasp and groan, and when I said that fewer than 1,000 people in the world work on this stuff, there were more gasps. It’s been a while since I’ve written any sort of top-level explainer, so for new subscribers, this article is a summary of how I view the situation.

We are currently emitting about 40 billion tonnes of carbon dioxide every year, and that number has not yet peaked. If you include the warming impacts of superpollutants like methane (a significantly worse greenhouse gas), then we are emitting the equivalent of approximately 55 billion tonnes annually. This number has not peaked in the data yet. Mitigation (the word the international policy community uses to mean “decarbonize”) has slowed down the rate of growth (which is a victory), but it has most certainly not reversed the trend yet.

Back in 2015, I thought, “Hey, it’s weird that I don’t see any significant activity working on pulling the CO2 back out of the air,” and so I embarked on building a business model around creating financial incentives for removing that legacy atmospheric CO2. For complicated reasons (and I have explained the longer background context here), the carbon dioxide removal (CDR) market has not reached scale. In 2015, I assumed that a combination of technological improvements, a shift in attitudes, a change in the regulatory environment, and financial incentives would result in a situation where CDR was reaching over 1 billion tonnes removed per year by 2025, and we were well on our way to scaling to tens of billions. The sad fact of the situation is that CDR has done nothing of the sort, and while amazing progress has been made in building an industry that has seen billions of dollars invested in hundreds of companies, our actual atmospheric impact has been effectively negligible.

Meanwhile, warming has accelerated. Earth’s energy imbalance (meaning, energy in exceeds energy out) has roughly doubled since the early 2000s, due to several factors, including feedback loops in Earth systems that are making things worse (decline in reflectivity, fewer clouds, less sea ice, etc.), and, perhaps counterintuitively, a decline in air pollution because of regulatory changes.

Temperatures exceeded 1.5°C above the pre-industrial baseline for the first time in 2024, and we are going to remain in overshoot for the rest of our lives.

The thing that really worries me is that if we merely rely on Plan A (cutting emissions and removing some amount of legacy atmospheric emissions), it is going to take too long, because there are feedback loops, tipping point risks, and Earth systems and human systems at risk of collapse, all of which could cause catastrophic damage to the planet and to our civilization.

If we do not take extraordinary action to protect coral reefs, prevent permafrost from melting, ensure food systems produce enough food for billions of people, restore sea ice, and prevent glacier and ice sheet collapse, then I worry that we will face what is known as derailment risk. That is, if we are already paying trillions of dollars for adaptation measures because of sea level rise, extreme storms, increased fires, etc., then what happens to our insurance markets? To our disaster response budgets? And then if we are getting punched in the mouth repeatedly by the effects of climate change, will we still be able to afford to finish the job of mitigation and decarbonize our society? Will our democratic governments be able to maintain legitimacy in the face of increased migration and citizens upset with all of the damages they experience?

The fact of the matter is, we have less time than we think to take action. These systems could collapse well before we even hit something like 3°C of warming. Fortunately, there is a field of people including scientists, policy experts, philanthropic and government funders, insurance and finance professionals, startup founders, and engineers working on the response options. We call this field Earth systems stabilization, or just stabilization for short.

The major risk areas that most people working in this field are concerned about are:

• Arctic cluster risks generally, because the Arctic is warming four times faster than the rest of the planet. This includes sea ice loss, AMOC collapse (the Atlantic Ocean current system that includes the Gulf Stream), melting permafrost, increased boreal fires, and more.

• The Amazon turning into savanna because of drought and the shutdown of the hydrological flows that renew the forest.

• Coral reef collapse (we are unfortunately already facing this).

• Sea level rise from both thermal expansion (water takes up more volume as it gets warmer) and ice sheet melt.

• And probably the most important of all, the darkening of the Earth as we lose reflectivity because of the effects of climate change.

We have options!

There are proposals for how we might respond to almost all of these risks. Some are better researched than others, but in all cases we as humanity would certainly benefit from an increase in the amount of research and experimentation. These include:

• Refreezing Arctic sea ice by pumping water on top of the ice in winter.

• Blocking warm water from melting the underside of the Thwaites Glacier in Antarctica.

• Thinning high-altitude cirrus clouds so that infrared energy radiating from the Earth can escape to space.

• Spraying salt aerosols and/or upwelling cold water to dissipate the energy in large hurricanes.

• Spraying salt aerosols to increase the reflectivity of certain clouds, both as a cooling method regionally and potentially as a way to shade coral reefs.

• Rewilding the Arctic with bison so they trample down the snow that insulates the ground and prevent permafrost from melting.

• Refilling inland basins and/or seas such as the Caspian Sea in order to put the water from sea level rise somewhere that lowers the overall level.

• The big one: injecting aerosols into the stratosphere (SAI) so as to reflect 1% of incoming sunlight and potentially reduce the overall temperature of the planet by 1°C.

• And there’s an even farther-out concept for putting a sunshade in space to function the same way (block incoming sunlight), without the atmospheric changes. Here’s a podcast I recorded with that team.

This newsletter and podcast exist to discuss these ideas with the people working on them, and I have a large back catalog you can dive into to learn more. But perhaps start here: How to Learn Everything You Need to Know About Climate Cooling.

I co-led a research study in early 2026 that found the single biggest barrier to researching these climate interventions is a lack of philanthropic funding. The second largest problem is that there are simply not enough people. We estimated that fewer than 1,000 people worldwide work on these issues in a full-time capacity, and those who do work on them are operating on a grand total of about $25 million of funding per year. This is woefully inadequate for responding to these risks, and it is incredibly important that we grow this field with as much haste as possible. If you are reading this and have philanthropic capital to deploy, the single highest-leverage thing you could do right now is fund research and development into climate interventions and Earth systems stabilization.

Climate Week NYC 2026

The event at the Explorers Club was the highlight of my week, and it was also the conclusion for me before heading home to Seattle. What I saw this year in NYC was heartening, though. By rough count, there were nearly twice as many events as last year that focused on stabilization. And at those events, I heard much more openness and bold statements about the need for stabilizing these Earth and human systems. I heard more voices saying that we must consider SAI as a response to temperature overshoot. The taboo around this is not yet broken, but it has definitely weakened. I wrote a note summarizing my observations from the week:

Back to the question, what does Plan B mean to me? A hypothetical future deployment of SAI or any other climate intervention, in my view, is absolutely not giving up on mitigation or carbon removal. Those are not optional, and interventions are not a replacement solution. Keep in mind that interventions come with their own risks and downsides and costs, and they only make sense to consider when comparing the risk of inaction to the risk of action.

But in my view, we simply must figure out how to stabilize these systems safely. It is a problem to be solved, and if in the course of research or discussions around governance we find some roadblock that can’t be overcome, then we must try another path until we find a more viable response option. If we don’t find ways to preserve our capacity for mitigation of climate change, then we are in for a very bad time. Plan B is not really “Plan B,” then. Plan B is how we make it possible to complete Plan A.

In what was pitch perfect timing, on the same day I announced changing the name of this Substack to The Overshoot Era, the United Nations Environment Programme (UNEP) released a report titled “Limiting Overshoot - Navigating exceedance of 1.5°C and pathways towards return”.

Some have asked me if I’d write about it, but to be honest, I thought a few other Substackers covered it much better than I could have, so I want to share those with you here.

First was Security Takeaways from the UNEP Overshoot Report, by Erin Sikorsky:

Hot Fronts: Security In a Warming World

Security Takeaways from the UNEP Overshoot Report

Today, the UN Environment Program released its new report, Limiting Overshoot, acknowledging that global temperatures will rise beyond 1.5 degrees above pre-industrial levels, “but it is still possible to bring temperatures down and achieve global goals in the Paris Agreement.” Doing so requires an “oversh…

Read more

a month ago · 7 likes · Erin Sikorsky

Erin makes the smart point here that from a security perspective (i.e. militaries) they simply must start preparing now for a reality that is seeming ever more likely:

Greater exploration of the diplomatic/security practicalities of carbon removal and solar radiation modification: I differ from the UNEP report here, which wants to keep the former as minimal as possible and dismisses the latter as unproven. I understand the political inclination – we cannot pretend there is an ‘easy’ way out of the conundrum we’ve built ourselves and give countries a free pass on mitigation. However, security actors in particular have to be realistic – these technologies are almost certain to be a part of our future, and we must prepare in such a way that ensures we maximize their utility in enhancing our security, not threatening and undermining it. We cannot wish away the conversation.

I’m currently reading Erin’s book, Climate Change on the Battlefield: International Military Responses to the Climate Crisis, and I keep thinking this stabilization field is underestimating the role of security actors in future interventions. Don’t get me wrong, discussions about the intersection of security and catastrophic climate and tipping-point risks are happening, but as Erin points out in her book, when you consider what resources countries will lean on (are already leaning on!) for massive infrastructure or disaster response, it is hard not to conclude militaries are going to play a strong role. And when I think about the future of climate interventions, it seems painfully obvious to me that this is going to be governments as customers, contracting out to private actors, in coordination with both civilian and military institutions. That’s how everything else works, so why wouldn’t it here?

The second article was from Noah Deich, who published UNEP Report Review: “Limiting Overshoot”:

Noah makes a point I very much agree with, which is that they are still not addressing the reality of what we must consider for our response options:

Now, why the report made me grumpy: despite all of these clear and useful conclusions in the report, it quickly becomes clear that the authors really want you to conclude that we have to ramp up our efforts on emissions reductions way faster than we currently are doing. In particular, they don’t want you to think that we should slow down investment in reductions today because we can instead rely more on removals in the future — and they reeeeeaaaaly don’t want you to make a similar conclusion on solar radiation management (SRM).

He goes on to point out that not talking about CDR and SRM honestly undermines the credibility of the report:

Hiding the ball on SRM and misrepresenting claims about the limits of carbon removal in a report on overshoot is an unnecessary risk to the entire report’s credibility. Why not instead just lay out the facts on carbon removal and SRM, and let policymakers come to their own conclusions about how to proceed with each?

To be clear, I agree with the UNEP’s conclusion that the focus of tackling overshoot should center on accelerating emissions reductions. But it comes across as patronizing — and reveals the report authors’ fairly bleak assessment of policymakers’ decision making ability — when they try to shape the narrative in this direction in such a heavy-handed way.

I agree strongly with this as well. There’s a very obvious elephant in the room here, which is, what happens if we don’t slam the brakes on emissions like it calls for? The report simultaneously points out that we have not begun decreasing emissions, and the attendant risks of not decreasing emissions are potentially catastrophic. But then it just leaves dangling the issue of what we should actually consider doing about that.

SRM isn’t the only thing missing, either. The report says warm-water coral reefs are crossing their thermal threshold right now at 1.4°C, that the carbon coming out of thawing permafrost may never go back in, and that the West Antarctic and Greenland ice sheets and the Atlantic circulation are among the tipping elements it expects to be at growing risk the longer warming stays above 1.5°C, and then it says that even optimistic carbon removal would bring temperatures back down by only a few tenths of a degree this century. So by its own numbers those systems are on their own for decades, and there is nothing in it about the work already underway to protect specific systems, like brightening clouds over the Great Barrier Reef, thickening Arctic sea ice, or the seabed curtain proposed for Thwaites Glacier.

Large reports like this are very difficult to put together, and they often end up being a lowest common denominator sort of affair, where only the most sanitized statements can make it through all of the veto points. But come on people, the problem is worsening quite rapidly, and we’re still not even mentioning SRM, while the CDR chapters read like a list of reasons it won’t work rather than a plan for making it work.

Lastly we have a slightly more sanguine take from Quico Toro with One Cheer for UNEP’s Limiting Overshoot Report:

One Percent Brighter

One Cheer for UNEP’s Limiting Overshoot Report

The UN Environment Programme (UNEP) put out a fat report this week and I read it with the mix of trepidation and forlorn hope these reports always elicit. As always, I read it looking for signs that the Overton Window on climate is moving, that previously unsayable things are becoming sayable, that “respectable opinion” looks different than how it looke…

Read more

a month ago · 33 likes · 2 comments · Quico Toro

Quico refers to the whole international policy apparatus as the “UN Climate Blob” because it is a slow-moving beast. That name incorporates some of what I meant by the lowest common denominator above:

Look, a blue ribbon report like this can’t, obviously, say anything interesting: co-authored by dozens of researchers in universities and government labs spread over six continents, the UN machine will make sure to squeeze out any sign of originality long before a beast like this comes anywhere near publication. Many very respectable people at eminent institutions have their fingerprints on this product, none of them want to face the blowback that would come from upsetting orthodoxies. Reports like this one exist not to be provocative, but to capture the consensus position in the UN Climate Blob. By design, Limiting Overshoot was never going to be anything other than a compendium of things very cautious people can all agree on.

But Quico is cautiously optimistic because this report represents a small step towards acknowledging the truth of where we are right now. Warming is accelerating, and holding on to 1.5 degrees is no longer feasible. A major UN agency saying as much means a lot of people who made commitments in 2015 to never let this happen are eating crow, and that’s a difficult thing to do.

Quico also points out the obvious truth that I have been shouting into the ether forever, that expensive CDR is simply not going to happen. The only way we can actually solve this problem long-term is if CDR can be done massively at scale and for very low price per tonne:

There is simply no reason to believe that a world that has failed to mobilize $800 to prevent the emission of one ton of CO₂ from a steel mill will somehow, magically, find the will to mobilize $800 to capture a ton of CO₂ via DAC. (If anything, the opposite is true: the steel-maker has a good reason to make steel —so they can sell steel!— the DAC plant has costs but no revenues.)

What a report like Limiting Overshoot was never going to do is acknowledge something that, to me at least, is fully obvious: that there’s no point in CDR if it’s going to cost as much, ton per ton, as mitigation. That gigaton-scale CDR is only ever going to be a thing if it’s an order of magnitude cheaper than the alternative.

All in all, this report is a mixed bag. It’s a very good thing that the “UN Climate Blob” is finally admitting that we’re going to blow right past 1.5, and that even in their most optimistic scenario getting back under it is a decades-long project. That’s a good first step. But it is of utmost importance that the next steps (talking seriously about SRM, about carbon removal that is cheap enough to actually scale, and about the regional interventions that could protect reefs and ice sheets and the rest while the slower solutions catch up, all alongside rapid emissions cuts) be taken with great haste. This current El Niño is going to give us a taste of what is in our very near future, and no one is going to like it. We must find ways to move faster in preparing for that future.

One answer, and as far as I can tell the only program studying it as a system, is to put some of that water somewhere else: into the desert basins and shrinking inland seas that sit below sea level around the world. It sounds like science fiction until you hear the numbers, the sites, and the list of questions that have to be answered first. This conversation is a tour of all three, with the person building the research program to answer them.

Guest

Charlotte Love is Program Lead for the Coastal Futures Resilience Initiative (CFRI) inside ARC at Renaissance Philanthropy. She studied earth sciences at Scripps Institution of Oceanography, completed a PhD in hydrology working with the Army Corps of Engineers on flood-frequency estimation, built transboundary water-security networks with the National Academies across North America, MENA, and Central Asia, and served as a AAAS Science & Technology Policy Fellow at the Federal Railroad Administration. CFRI is currently a team of one; she’s hiring and raising.

Key topics

• 00:00 — Cold open: sea level rise is permanent on human timescales

• 01:55 — Charlotte’s path: Scripps, hydrology, the Army Corps, Central Asian water diplomacy, and why the executive branch needs answers rather than papers

• 06:35 — Why the rise is locked in: deep-ocean heat storage and thermal expansion

• 09:34 — What SAI and carbon removal don’t fix, and why every intervention is maintenance

• 11:14 — The CFRI idea: storing seawater in endorheic basins to replace the storage land ice once provided

• 13:11 — The numbers: eleven sites, 51 cm over a century, and the 20 cm target for 2050

• 16:14 — Phase one: proof of concept for the Qattara Depression and Dead Sea, scoping for the Caspian

• 20:04 — Groundwater, salinity, and the questions a water scientist asks first

• 24:24 — Why Egypt has wanted to fill the Qattara for decades: hydropower, desalination, dust

• 29:05 — The Dead Sea and Caspian as restoration projects, and the Caspian’s research void

• 35:26 — The Salton Sea as an early learning site

• 40:18 — Phase two, phase three, and the handoff question

• 45:14 — Is there enough talent? Gravity, topography, and why this scares people less than SAI

• 47:36 — Closing

Links and resources

• CFRI’s launch in 2025: “New ARC Report on Refilling Inland Seas: Harnessing ancient basins to confront rising seas”

• Coastal Futures Resilience Initiative (CFRI), ARC at Renaissance Philanthropy

• CFRI’s 2025 basin assessment report (the eleven sites)

• Previous episode: Marianne Hagen on the Seabed Curtain Project

The Overshoot Era is about what we do now that we’re past 1.5 degrees, and every episode is a conversation with someone building one piece of the response — seabed curtains, seawater storage, atmospheric cooling, the governance that has to hold all of it together. Subscribers get every episode, the full show notes, and the written essays where I connect the pieces. Free at overshootera.com.

Subscribe now

Last year I went to NYCW in search of geoengineering-related events, and had to ask around the few contacts I had in the field in the weeks leading up to find the Luma events for things like Reflective hosting a happy hour, and EDF hosting a panel on SRM. I was fortunate to run into some people who then got me invites to other events, and the week was a smashing success. The most useful event was a Climate Emergenices Forum workshop hosted by Renaissance Philanthropy, where I met Ben Lachman, who I’ve been working with since this February. (I wrote about that workshop in The Architecture for Cooling Earth Is Finally Taking Shape.)

There are many more events this year that relate to stabilization, and Ben had the good idea to pull them all together in one place before the week starts, so nobody has to ask around. The list has over forty events on it as I write this and it keeps growing.

These are all events that are related or adjacent to the field of Earth systems stabilization, from climate tipping points to sunlight reflection to the Arctic and ocean systems. Each event is tagged by date, topic, and location. Here’s what I intend to be at:

• Monday evening, The Degrees Initiative Looks Back: 15 Years of SRM Research and Engagement. Degrees has spent those years building SRM research capacity in the Global South, which is a good vantage point for asking what has actually changed.

• Tuesday morning, Ocean Visions’ The State of Ocean-Climate Action in a World of Accelerating Change.

• Wednesday morning, Degrees again with A New Era for SRM Research, launching what they describe as the first global study of SRM and agriculture.

• Wednesday midday, Ocean Visions’ Navigating Non-Linear Climate Risk and Response: From the Arctic to the Global Climate System, whose description reads like it was written for readers of this newsletter.

• Wednesday afternoon, the SRM Research Hub Crawl that the Chesapeake Climate Action Network is running, a move-at-your-own-pace set of conversations with researchers and advocates, and probably the best bet on the list for meeting people.

• Wednesday evening, C2ES’s Is the United Nations Ready to Deal With Climate Tipping Points?, on whether the UN system can handle tipping-point shocks or needs new machinery. There’s a livestream too.

• Thursday evening, A Plan B for the Planet? Environmental Observations and Technologies for the Future at The Explorers Club, put on with the Seabed Curtain Project and UArctic. Marianne Hagen, who co-leads the Seabed Curtain Project and who I had on the podcast in August for Can We Save the Doomsday Glacier?, is on the panel. The evening is built around sea level rise as a security and economic-risk question, which is how Marianne framed it on the podcast too. This is one is paid.

• Thursday night, The Original Climate Week Karaoke with Renaissance Philanthropy.

A couple of these need host approval to register, so don’t wait on them. I’ll be in New York Monday the 21st through Thursday the 24th. If you’re going to be there and want to meet up, reply to this email and tell me which days you’re around. And if you know of an event that belongs on the list, submit it here and Ben will add it. With any luck this saves a few of you the asking-around I did last year. I hope to see some of you there.

Subscribe now

At first, I wrote about startup stuff: my experience at Nori and winding things down there, board governance, and lessons for early-stage founders. Then I wrote a piece with Nick van Osdol about why carbon removal won’t reach gigatonne scale, out of frustration that the industry had grown while its actual atmospheric impact stayed negligible. The response was strong! I got asked to speak during San Francisco Climate Week, and I kept writing about climate since that turned out to be what people wanted to read from me. Soon I was writing about climate interventions and my own path into the field. All the while I kept the branding because it was just easier.

But after some time writing about this space, I’m no longer just the new guy showing up and asking beginner questions. I’ve been sharing my opinions on how I think this field should evolve, and that invites feedback from people who have been working on this much longer than I have and who have more expertise in it, and some of that feedback has been that the name Inevitable & Obvious isn’t really a good look here. To be more specific, the inevitability framing is a known problem in the governance conversation around solar geoengineering: when someone declares that deployment is coming, they’re foreclosing the public’s choice before any deliberation has actually happened, and a guy with a tech background putting the word Inevitable on the masthead reads to a lot of people as exactly that. The name predates any of my climate writing, but that doesn’t really matter, because nobody reading it for the first time knows that. I think this is valid and useful feedback. I would like to continue learning from and collaborating with the experts in this field, and changing a name is a pretty small thing to do in exchange for that.

Now, I want to be clear: I do believe we have to respond to the catastrophic risks that come with unabated warming. And I think one of the better options on the table for that is planetary-scale cooling. That frankly does seem obvious to me. However, the way in which we do it, the timing of it, who actually deploys it, and how those decisions get made, none of that is obvious, and it most certainly is not inevitable, because we don’t really know how we’re going to get to that point. There are a lot of unknowns, and that invites a lot of risk.

There’s also a more practical reason. When people come across this newsletter for the first time, Inevitable & Obvious doesn’t tell them anything about what it’s about, and I would really like to make this subject matter easier to understand. My objective is to bring more people into the field, get them educated on this, and find ways for new skill sets and experiences to contribute to solving these very thorny problems.

So putting all that together, I decided it was time for a new name that better reflects what this newsletter is about, and that name is The Overshoot Era. There's disagreement in the literature over the details, but overshoot means exceeding a temperature threshold. In this case, it means global warming going past 1.5℃, the more ambitious of the Paris Agreement's two targets.1

We've already exceeded 1.5℃ at least temporarily, and forecasters expect we'll cross that threshold again this year and next with the extra-strong El Niño that is now upon us. We're at the start of a new era in climate change, and you're going to hear the word overshoot a lot more as the planet heats up. I just don't see a path where we hold 1.5 anymore, and this newsletter is about what we do next.

The Overshoot Era is me working through what I can contribute to a field that exists to buy time while decarbonization and carbon removal finish the job. I share what I'm learning from the people researching and building this stuff, my opinions on how the field could and should evolve, and enough synthesis to show what's actually going on. The questions I keep coming back to are how we bring the temperature down as quickly as is safely possible, how we keep important Earth systems from crossing their tipping points, and how we preserve society's capacity to solve climate change. Those are the things I read about and ask my podcast guests about, and then I tell you what I think.

Some housekeeping: The Overshoot Era is the new name for both the newsletter and the podcast. The website is https://overshootera.com/. You don’t need to do anything on your end. Emails will still show up in your inbox the same, the podcast feed remains the same, old links will redirect to the new domain, and the whole archive is still here. Lastly, a hat tip to Charlotte DeWald at Arctic Stabilization Initiative for the name suggestion, thank you!

I hope you like this direction. If you have thoughts or feedback, leave them in the comments or email me. All posts here are free, no matter what. If this is useful to you, please subscribe, share it with your colleagues, and consider a paid subscription. The money helps, and paid subscriptions also push this publication up Substack’s Rising list, which is where new readers find it. If you value my work, that’s the best way to put it in front of more people. I appreciate you coming along for this ride.

Subscribe now

If you’re new, How to Learn Everything You Need to Know About Climate Cooling is the place to start.

When it comes to climate interventions, we spend a lot of time asking whether we can do this as well as whether we should. Asking “can we?” leads to lab research and modeling work. Asking “should we?” leads to conversations about social legitimacy and governance. Both are important, for concrete reasons. Social legitimacy is a critical component of any large-scale intervention project, because to deploy these interventions durably you need the support of policymakers, and policymakers need the support of the people affected. Governance matters because these will be multi-decade megaprojects with hard decisions and tradeoffs to work through, now and for decades after deployment starts, and that requires institutions capable of working through them well. Any large intervention project will depend on those things as much as it depends on the finance and the engineering and the ongoing operations and maintenance.

What’s interesting to me is how little time we spend, by comparison, on a third question, which is how we would actually do any of this at scale. By this I mean, it’s one thing to ask will marine cloud brightening work to shade coral reefs or even provide larger-scale cooling benefits. It’s an entirely different question to ask, ok then how many MCB sprayers does that require, on the back of how many ships, where should they be located, how much will this cost, who will support this, and more.

I think this current state of unknown scalability is a natural result of how the field has evolved. For a long time, can-we was a science question, and scientists worked on it. Over the last decade plus, as interventions started to become more real, the should-we people arrived, from policy shops and think tanks and the social sciences. The people who work on operational scaling mostly haven’t arrived yet, and why would they? There’s been no money in it, and we’re only just now seeing the first funded companies. It does mean that when these conversations happen, hardly anyone in the room is thinking about what it would take to run one of these projects at scale.

If this question is interesting to you, there will be an entire session to discuss how to operationally scale interventions at the Global Risk and Stabilization Summit in Reykjavik, October 12-13.

Apply for registration at https://globalstabilization.org/.

My background is in engineering, and I spent years building a company in the neighboring field of carbon removal. A lot of what I thought about then, and still think about now, is how you scale something like this in the way most likely to succeed, and what you would have to start doing today if you worked backwards from that.

So this piece is about the how. I’m going to make the case that stabilizing climate interventions comprise a reference class. However different these projects look from each other, they’re going to run into similar challenges around social legitimacy and establishing governance. Who decides, and where does their authority to decide come from? How do the projects get funded, and who pays? What technology goes into them, who manages and runs them, and how does that work get contracted out? How does the engineering and operational scaling actually happen, and how are these things maintained over years and decades? How do they evolve as our understanding of the Earth systems they touch evolves, and how do we monitor their impacts and make decisions about what we learn? Every one of those is incredibly complex, but the interventions will be thought about similarly and in relationship to each other. And the best place to start on the how is the track record of extraordinarily large projects, which is the subject of How Big Things Get Done by Bent Flyvbjerg and Dan Gardner.

Flyvbjerg is an Oxford megaproject researcher with a database of more than 16,000 large projects across 136 countries, from home renovations to nuclear power plants to the Olympics. And what his database shows is that only 8.5 percent of the projects in it came in on time and on budget. Only 0.5 percent(!), one project in two hundred, came in on time, on budget, and delivered the benefits they promised. That is a horrifically poor record of performance. The pattern is consistent enough that Flyvbjerg calls it “the iron law of megaprojects.”

So why does this keep happening? A big part of Flyvbjerg’s answer is that projects get committed to before they get planned well, and the people who want them have incentives to commit early. Politicians want to be first, or to build the biggest, because that is what gets remembered, and they know how hard funding is to win. He summarizes this as a tactic where politicians want to just start “digging the hole,” because once the hole exists, people susceptible to the sunk cost fallacy will keep committing money rather than admit the project was a mistake.

Where I live in Seattle, we lived through this in the 2010s. When we replaced the Alaskan Way Viaduct, the design called for a double-deck tunnel with two lanes in each direction, running underneath downtown to connect the south end of the city with the north end. To bore it, the state commissioned Bertha, the largest tunnel boring machine ever built at the time. Five months into the dig, Bertha hit an abandoned steel pipe and broke, and it sat under the city for about two years while the contractor dug a rescue shaft, lifted the four-million-pound front end to the surface, and rebuilt it. Once that machine was in the ground there was never any question of stopping. The tunnel is completed and done now and it’s actually fantastic to have, but it was much more expensive than planned and took years longer than planned, and the stall alone accounted for a couple hundred million dollars in overruns.

Bertha’s problem wasn’t tunneling, because that’s a mature discipline with a century of accumulated practice behind it. Bertha’s problem was that no machine like it had ever been built, so all the myriad ways in which it could break or run into problems hadn’t been put to real-world testing. The same dynamic shows up at small scale too. I learned it personally by buying a Fisker Ocean, a first-of-a-kind car from a brand-new EV company. Fisker went bankrupt shortly after delivering the first ~5,000 vehicles (of which mine was one), and so I’ve spent a lot of money on repairs that should have been covered under warranty had the company survived. One of the problems with the company was an inexperienced COO (who was married to the CEO) who took charge for selecting parts and components, invariably choosing the cheapest option. Look in the photo below—do you see the white inserts in the door handles? They’re purely decorative, but it turns out they were made of a plastic material that happens to crumble into dust when exposed to UV light. I don’t know about you, but I drive my car outside, and so after about 18 months, I had to replace the door handles entirely. I’ve totally learned my lesson. The next EV I buy will be from a major manufacturer who knows how to build cars and, more importantly, run a business that builds cars. This is a common pattern in climate tech startups too. When I’ve talked with climate VCs over the last year or so, they’ll typically say: we don’t want to fund first-of-a-kind tech. First-of-a-kind applications are great, but use off-the-shelf components. Components that have been through years of use have had their failure modes found and fixed. New components haven’t.

Big projects also go better when they’re delivered by teams whose experience compounds, because so much of what an experienced team knows is tacit, the kind of knowledge you get from doing and can’t fully write down. Sort of how if you try to explain to a child learning to ride a bicycle how to balance, you can’t really. They just have to figure it out for themselves based on feel. That feeling is tacit knowledge.

Flyvbjerg showcases the Olympics as the archetypal example where this problem arises. The games are chronically over budget and behind schedule, partly because they move from city to city, so each host starts over completely. Maybe you’d think cities would hire vendors with experience from previous Olympics, and that does happen to a degree, but Flyvbjerg found that in most cases politicians promise contracts to local vendor companies (read: inexperienced vendors) in order to win public support for funding the games. This is another reason to think in reference classes. A field that treats its projects as one class can carry teams and their accumulated experience from project to project, while a field that treats every project as unique rebuilds that experience from zero each time, at each project’s expense.

Flyvbjerg calls this underlying error the uniqueness bias. Everyone believes their project is special, and in one sense every project is, the same way every human being is an individual. But humans share a set of common characteristics, and projects do too. The Sydney Opera House is unique in its location and design, and it is also an opera house that needs good acoustics, a certain number of seats, entryways, and different classes of tickets to sell. Daniel Kahneman calls the two ways of seeing a project the inside view, where my project is special and I forecast how it will go based on its particulars, and the outside view, where my project is one of a class and I forecast it based on how that class has actually performed. Almost everyone takes the inside view, and Kahneman and Flyvbjerg both consider that a fatal forecasting error.

Imagine renovating your kitchen. You measure everything, price out the materials, and add it up to get a total budget. And then when the work begins, the countertop breaks on delivery and you find mold under the floorboards. Dealing with those setbacks weren’t in your itemized plan, but they are in the average cost of kitchen renovations, because other people hit surprises too. So that’s actually how the book recommends forecasting what a project will take to do. Start from that average for the reference class (i.e. other kitchen renovations in your city) and adjust for what’s genuinely different about your project. That might seem too simple to be correct, but Flyvbjerg has empirical evidence that shows it reliably produces better forecasts. He turned this into a formal method called reference class forecasting.

Now apply this to climate interventions. When I’ve suggested that these projects have a lot to learn from each other, the pushback from inside the field has usually been some version of: “they’re too unique between them,” “SAI has nothing to teach an Antarctic seabed curtain,” and “a curtain has nothing to teach a reef-shading program.” I think that’s the inside view, and it’s precisely the error that the forecasting research warns against. Whatever the intervention, you have to get people and equipment to remote, difficult places, power what they’re doing, and keep them supplied. You have to measure whether the thing worked. You have to earn and keep the support of the people affected. Those are common characteristics of the class. And when you view this from the perspective of a government contending with catastrophic climate risks, these projects are plainly related. Tornadoes are different from hurricanes are different from earthquakes are different from extreme wildfires, but FEMA is the same agency in the US that responds to all of them. If you view these Earth systems from a risk-first perspective, then stabilizing interventions are related and intertwined.

Let’s go back to that 0.5 percent success rate for very large projects. If only one project in two hundred delivers on time, on budget, and with the promised benefits, then we should assume that any stabilizing intervention will run over time, over budget, and/or fail to deliver what it promised. This is what history shows us to be true, and it is a terrible track record to start from!

For this field, that default is untenable. For one, with temperatures accelerating higher, coral reefs reaching a tipping point, and AMOC collapse risk looming, we don’t have the time to waste. But also the money and the mandate will eventually have to come from governments and international agencies, which requires social legitimacy. Public support is what lets a government fund a decades-long project, and it erodes fast when a project is late, over budget, and visibly failing to deliver what it promised. These projects will also outlast the governments that start them. Support has to survive elections, and given the risk of derailment (the risk that society’s capacity to even respond to climate change will degrade because of the damages caused by climate change), the political environment these projects live in will get harder over time.

A stuck megaproject ends in one of two ways: either you keep going and absorb the cost and time or you pull the plug—which might also cost you. Once the Bertha machine was under the city, there was no way they weren’t going to finish the job, whatever it cost. The book’s other example is Japan’s Monju fast-breeder reactor, which took about thirty years and more than $9 billion dollars to build, operated for roughly 250 days in total, and never delivered meaningful power before the government decided in 2016 to decommission it. An intervention project that languishes is not neutral. It consumes money, attention, and public patience that the rest of the field needs.

Policymakers and the public are going to view these projects as a class whether or not the field does, the same way FEMA’s disasters are a class. If one intervention fails badly in public, it damages the ability to do all the others. (Nuclear power already demonstrated this. After its most public failures, the entire technology class froze for a generation.) So I think we have to hold these projects to incredibly high standards of operational excellence—on time, on budget, delivering what was promised—because the whole portfolio’s permission to proceed depends on it.

The reference classes for costing this work already exist. The International Thwaites Glacier Collaboration has been running science expeditions to one of the least accessible places on the planet on a five-year, fifty-million-dollar budget, so we know roughly what it costs to get scientists and equipment there and back. But an intervention would need more than visits. If we’re going to attempt a seabed curtain to slow the collapse of Thwaites, then phase one is a permanent remote sensing hub at the glacier, collecting data continuously instead of expedition by expedition. Marianne Hagen and I talked about this on the most recent episode of my podcast. That hub is a bigger ask than what the science teams do today, and reference class forecasting would handle it the normal way, by starting from what expeditions cost and adjusting upward for continuous presence. The subsea cable industry lays and maintains fiber across ocean basins as a routine commercial service, and that industry is the nearest class for putting a curtain on the seabed. Militaries have spent a century learning how to supply remote bases. When a piece of an intervention really is first of a kind, the move is still to find the nearest class, take its average, and adjust, instead of pricing the project from its particulars and finding the surprises later.

The book’s answer to the iron law problem is modularity. Flyvbjerg uses Lego as the model and asks you to figure out what the small repeatable piece of your project is, your Lego block, because the projects in his database that reliably come in on time and on budget are the ones built by repeating a small unit many times. He compares solar and wind farms, which are assembled from repeated panels and turbines and reliably deliver, against nuclear plants, which are enormous custom builds and sit near the bottom of his rankings. The Empire State Building went up fast for the same reason, one repeated floor at a time, with the crews getting better at each one. Repetition is how experience accumulates. For stabilization, I think the Lego blocks mostly won’t belong to any single intervention. When you decompose these systems into components, the same technologies keep showing up: underwater drones and autonomous platforms, cold-rated sensors and communications, remote power generation, the logistics of moving equipment to hard places, etc. They’re shared across sea ice, ice sheets, coral reefs, the stratosphere, and more. If you build those as repeatable products instead of expedition one-offs, then every intervention that uses them inherits the learning from all the others. That’s one more reason the reference class matters. The Lego blocks only become visible when you look at the projects together.

So here’s the paradigm I think this field should adopt while it’s still early enough to adopt one. Treat stabilizing interventions as a reference class with shared characteristics (African Futures Tech Lab already put out their own climate interventions taxonomy, which sort of gets at this), and take the outside view when forecasting them. Cost interventions from the classes that already exist (polar expeditions, cable ships, remote bases, stratospheric planes), then adjust for what’s genuinely new. Prefer proven components underneath novel applications, and treat any component that has never existed before as a schedule and budget risk. Structure the work as repeatable programs built from shared modules, so experience compounds across interventions instead of being relearned by every project the way the Olympics relearn hosting every four years. None of this slows anything down. The iron law is what punishes moving fast without planning, and given the urgency and the stakes, we cannot afford to languish our way through this. If we’re going to deliver on these projects, and we have to, this is what delivering takes.

Key topics

• 0:00 — Cold open: the warm water reaching Thwaites

• 2:58 — From Norwegian diplomacy and Aker Solutions to glacier intervention, and the moment the Ukraine invasion changed her math

• 18:39 — What sea level rise actually costs: coastlines, ports, agriculture, real estate, financial stability

• 26:18 — Why West Antarctica: the sill, the channel, and the soft underbelly of Thwaites

• 31:55 — The cork in the champagne bottle, and why Antarctica is a bowl of ice

• 34:44 — The curtain: fixing a leak in your basement, 150 meters tall and 80 kilometers wide

• 41:22 — The expedition: a borehole through 1,000 meters of ice, a cable stuck at 700, and no ship until 2028

• 47:37 — A permanent underwater observatory, the white paper, and what happens next

Links and resources

• Seabed Curtain Project

• John Moore’s original 2018 proposal for seabed-anchored curtains (Nature comment, “Geoengineer polar glaciers to slow sea-level rise”)

• New York Times coverage of this year’s Thwaites expedition

This episode is part of an ongoing series on the interventions that could hold Earth systems stable long enough for mitigation and carbon removal to finish the job, covering everything from stratospheric aerosols to the ice itself. Subscribing free at inevitableandobvious.com gets you every episode, the full show notes, and the essays that connect them.

Subscribe now

Stabilization is a term I thought I was making up. I started suggesting it last fall because I wanted a way to talk about this work that could move past geoengineering, and I think about naming more than most people in climate do because I spent ten years doing category creation in carbon removal (CDR). I was just a guy proposing a word, and I proposed it to solve some specific problems.

My intent with this article is to do two things:

1. Steelman the arguments we’ve heard against the term

2. Clarify a more specific definition of stabilization and invite further discussion, refinement, and improvement

Here are the parameters Ben and I discussed (starting last fall) as we developed the “Stabilization Plan” concept and the term that came from it:

• If important Earth systems destabilize to the point of impairing the ability of governments and markets to respond to the root causes of climate change at all, then we are in serious future danger. We simply must find ways to avoid this outcome.

• To avoid it, the field needs to grow, which means more philanthropic funding and more people coming to the space. Only a few foundations have been willing to fund research, and most of that goes to SAI. This leaves other stabilizing interventions severely underfunded, all from the same small set of funders we came to call “the usual suspects.”

• Everyone (even the interconnected usual suspects) refers to these interventions differently, so the field is fragmented in how it organizes, funds, researches, and acts on any of them.

• The fragmentation happens in part because “geoengineering” is a poisoned word that has come to mean many things, and it carries an innate hubris that turns many people off. “Climate engineering” carries most of the same turnoffs.

• To have the potential for success, this work should eventually be funded and carried out by nation states, which means politicians need to be able to talk about it in a way that garners support from their constituents.

• Emergency framing is a nonstarter.

• “Climate interventions” covers what many of these things are, but it is unlikely to galvanize anyone beyond the most involved researchers or practitioners. As a field name it sounds dangerous and risky. It’s not “for” something the way “renewable energy” or “carbon removal” or even “adaptation” are “for” solutions. It doesn’t describe a desired end state or set a goal. It doesn’t invite people to join the field and contribute. It’s not aspiring or inspiring.

The esteemed systems expert Donella Meadows wrote that her number two most effective way to intervene in a system is to change “The mindset or paradigm out of which the system — its goals, power structure, rules, its culture — arises.” From Meadows:

So how do you change paradigms? Thomas Kuhn, who wrote the seminal book about the great paradigm shifts of science, has a lot to say about that. In a nutshell, you keep pointing at the anomalies and failures in the old paradigm, you keep coming yourself, and loudly and with assurance from the new one, you insert people with the new paradigm in places of public visibility and power. You don’t waste time with reactionaries; rather you work with active change agents and with the vast middle ground of people who are open-minded.

So last fall I suggested we needed a new paradigm for thinking about these activities. I proposed that the thing we all want is stability, because stability is a necessary precondition for finishing the job of decarbonization and legacy carbon removal, which will take decades if not centuries. To those of us who have adopted the term, stabilization feels like it could be the new paradigm that helps set the “goals, power structure, rules, and culture” for striving to prevent Earth systems from destabilizing further, and the many systems downstream of them from following. I took her advice, pointed at the anomalies, and wrote a bunch of articles about it.

And the word spread further than I expected! Several groups working in this space converged on it as shared language, papers have mentioned it, we co-hosted an event in San Francisco built around it, at least two organizations (that I know of) rebranded to include Stabilization in their names, and the Global Risk and Stabilization Summit convenes in Iceland this October.

It turns out I wasn’t making it up, though. Stabilization carries a long history in climate circles, which I didn’t know when I started using it and have since learned. As more people adopted the framing, we started hearing objections from scientists, researchers, and policy folks we interacted and shared the report with. Many of these objections have changed how we use it. Other points we disagree with, and I’ll explain why. All of it is good, however, because it forces us all to discuss more directly the underlying problems we’re trying to solve together.

Most of what’s below is a composite of feedback from interviews, report responses, and calls, left unattributed since we promised interview participants confidentiality. The exception is Holly Buck, who published her opinion last week, and I respond to her thoughts directly further down.

Our initial framing on stabilization

Stabilization is work that attempts to intervene directly in Earth systems destabilized by climate change that are at risk of runaway or self-reinforcing outcomes. This is done to buy the time now required to complete implementation of slower, more permanent climate solutions.

Nobody trying to shade coral reefs by brightening clouds is purporting to stabilize the entire climate. Neither is someone who attempts to refreeze Arctic sea ice. In both cases, they are acting on systems that, if their destabilization proceeds unabated, will threaten the stability of other systems as well—from human systems like food production to ecosystems and weather. The object of a stabilization intervention is an Earth system, and the beneficiary is every system downstream of it.

An analogy we (meaning Ben and I) use frequently comes from healthcare. When a trauma team stabilizes a patient, nobody in the room believes the patient has been cured; stabilizing is the interstitial step that keeps additional treatment possible. In the same way, stabilization measures are meant to be temporary, wound down as emissions cuts and carbon removal bring the climate back toward safety.

This is also why we’ve been moving toward using stabilization by itself rather than prepending it with climate. What we’re trying to stabilize is a set of Earth systems. Calling it climate stabilization has led some people to see an all-encompassing term for all parts of climate response, while we see a need to create space alongside mitigation and adaptation.

The third pillar

The two established pillars of climate change response, reflected in the IPCC’s working group structure, are mitigation and adaptation. When we say mitigation, we mean managing the stocks and flows of greenhouse gases, cutting new emissions (decarbonization) and drawing down legacy ones (carbon dioxide removal). This is a paraphrase of the IPCC, which defines mitigation as “a human intervention to reduce emissions or enhance the sinks of greenhouse gases”.

Adaptation absorbs consequences of climate change in human systems, e.g. with seawalls and heat plans and changed agricultural practices; in the IPCC’s words, it is “the process of adjustment to actual or expected climate and its effects” so as to lessen harm.

Stabilization acts on destabilizing Earth systems directly. These three pillars sort the work clearly. Halting deforestation or sequestering carbon in biochar both influence carbon stocks/flows, so they go with mitigation. Building seawalls or planting species of crops that tolerate heat better are clearly adaptation measures. Preserving an ice sheet doesn’t target greenhouse gas concentrations, so it is clearly not mitigation, and as I wrote in What’s the Difference Between Adaptation and Geoengineering?, it is also quite different from adaptation. Where, then, should it go? Our proposed answer is stabilization.

The IPCC glossary does have a stabilization entry, and while this isn’t commonly used today, it means something different: a state where the atmospheric concentration of greenhouse gases holds constant over time. That is mitigation’s finish line—an end state rather than a pillar of work. I address this collision between the older sense of the word and ours in the objections below.

Adding a pillar has happened before. Adaptation was not always considered a reasonable and important aspect of climate change response. Ian Burton, the geographer who did more than anyone to push adaptation onto the climate agenda from the first international climate meetings onward, spent years being told by mitigation’s most fervent advocates that even talking about adaptation was counterproductive. If we could just focus on mitigation, then adaptation wouldn’t be necessary. Here’s Al Gore in 1989:

The advice to adapt rather than prevent these [climate] changes is unwise counsel. The changes are occurring so rapidly that adaptation will not be possible without major disruptions in our civilization. The advice to adapt is an obstacle to the correct political response, which is prevention.

Burton’s answer was that both would be needed, because you “wouldn’t be able to mitigate fast enough” to escape adapting. He kept at it until the IPCC gave adaptation a working group home in the mid-1990s, and in 2021 he shared a Frontiers of Knowledge Award for the paradigm shift. He spent years doing what Meadows describes, pointing at the anomalies until the paradigm moved.

So why does stabilization need a pillar at all? The answer is the one Burton gave in the 1990s, except now we have even more data. Atmospheric CO2 reached 427 ppm in 2025. We emit around 40 billion tonnes of CO2 every year (with no sign of that going down), and the entire carbon removal industry has durably removed about one and a half million tonnes, cumulatively, ever. Even the most optimistic scaling scenarios put cumulative removal by 2050 at roughly 100 billion tonnes, which works out to only two or three years of current emissions. And CO2 does not clean itself up on any timescale that helps us. A large share of every tonne we emit stays in the atmosphere for centuries, so ending all emissions tomorrow would only stop the problem from growing without shrinking it. The Earth systems at risk will keep responding to the heat already locked in while we spend decades (potentially centuries!) scaling carbon removal. Burton said we wouldn’t mitigate fast enough to escape adapting. The numbers today say we won’t mitigate or remove fast enough to escape stabilizing.

The objections

This is just geoengineering with a new coat of paint.

I said at the top that I wanted to move past geoengineering, and so have many others, so the suspicion of a “rebrand” is fair. We argue that the renaming is necessary to change the conditions preventing the field from growing. “Geoengineering” and “Stabilization” point in different directions. Geoengineering is solutions-first. It names a set of technologies and leads with what we could do to the planet, and one of the clearest findings from our interviews was that solution-first framing turns people away while risk-first framing brings them along. Stabilization is risk-first. It names the state we want, and the only reason to pursue it is that specific Earth systems are being destabilized. The actions stay what they are—call them climate interventions or Earth-system interventions or stabilizing interventions—undertaken in service of the goal and only when justified by the risk. The contents of the term has also changed. The original geoengineering definition bundled in carbon removal, while stabilization leaves CDR as part of mitigation and includes work nobody calls geoengineering, like Arctic rewilding or localized reef cooling. In our Building Permission Space report, we made the case that this renaming is intentional and useful for the field’s goals.

Stabilization already means something else.

It does, and this was the objection we heard most from people close to the UNFCCC/IPCC processes. The Framework Convention’s stated objective is to stabilize greenhouse gas concentrations at a safe level, so for more than thirty years “climate stabilization” has meant managing concentrations, or holding temperature at a target, or sometimes just solving climate change in general. The general take has been something like: the whole climate era began as a stabilization project, so why reach for a word the field already used up? Our response is that as we define it, stabilization is neither the stabilization of concentrations, which is mitigation’s objective, nor a fixed temperature target. The objects are particular Earth systems. And it is a very useful word that accurately describes the attempted work of this field. Our shift away from climate stabilization and toward stabilization alone attempts to integrate this feedback.

If you’re going to use it, make it the umbrella.

Some people flipped the objection around and suggested the opposite, keeping the word but making it the overarching category, a stabilization that contains mitigation and adaptation along with anything else as the means, since the point of every climate response is a stable Earth. I understand the logic, but believe the strategy is backwards. Asking the entire climate world to reorganize its language under a term we floated a year ago would be both naive and ill-advised. But adding stabilization alongside the existing pillars asks nothing of anyone except consideration and tolerance. A new umbrella also wouldn’t help create what this field needs. Large climate philanthropies and government policy agendas allocate resources to line items for mitigation, for adaptation or resilience, sometimes for carbon removal and superpollutant destruction, and whatever remains to a scattering of uncategorized projects around the edges. A new umbrella over the whole table restructures their work without creating the line item that so many uncategorized and underfunded research efforts sorely need.

Can you stabilize a system while the planet keeps warming?

Yes, partially, and partial is already how the second pillar works. If warming is the destabilizing force and warming continues, then stabilizing anything is incomplete by definition. The same is true of adaptation, which is by definition something we do while the Earth keeps warming; a sea wall does not stop the sea from rising, and nobody concludes that adaptation is incoherent. Stabilization operates the same way, by attempting to hold specific systems stable, as long as required, with success measured against the system’s existing trajectory rather than the planet’s temperature.

It over-promises.

Stabilization names an outcome, and this objection is correct that the name doesn’t answer whether the outcome is deliverable. We know a lot about the coupling between cutting emissions and atmospheric concentrations. We know much less about the coupling between attempting to stabilize an ice sheet and the prevention of ice sheet collapse. Additionally, any proposed stabilizing intervention could act and fail, or worse, act and cause new harm. This objection would state that a humbler word like interventions is safer because it promises nothing beyond the action, and whoever intervenes still has to make the case that it will work.

That asymmetry is real, and it’s precisely why the field needs more funding: funding for people studying the potential risks of interventions, funding for people developing governance mechanisms around deploying them, and funding for people developing new or alternative interventions so that we have options when some interventions turn out to be harmful. But mitigation and adaptation also name outcomes we haven’t finished delivering. Both are partially done and pointed in a generally good direction, neither is moving fast enough, and the global conversation about each is whether we’re delivering. Nobody reads those words as guarantees. A pillar’s name states the intention the work gets organized around and judged against.

Ben put it more simply on one of our calls: nobody aspires to intervene. Intervening is a cost, a thing you would rather not have to do, while stability is what people want, and a field should be named for what it is for. The outcome-name also does something useful for these technologies that carry the most hubris risk. A field named stabilization hands the world the standard to judge it by, because for any proposed intervention the question becomes whether the system is measurably more stable, on what evidence, and at what cost. A field named for its activity can “succeed” just by being busy. I would rather have a name that makes it easier to understand what success means.

Science needs rigor, and this term doesn’t have it yet.

Correct. The term is ambiguous partly because this usage is new. Scientists parse every word, want rigorous definitions, and have watched loose terminology cause damage in this field before. Stabilization is not yet a rigorously defined scientific term, and the definition below is a field-building definition, proposed to satisfy the parameters at the beginning of this article—where politicians, funders, and newcomers were the constraints—rather than built to survive peer review. Those audiences pull in different directions than science, and right now this objection is valid, as there is no real scientific rigor around it. What I’d ask of scientist readers is participation. Our own usage has already changed in response to feedback. The move away from the climate modifier came from these conversations, and the pillar framing in this piece is itself a second or third iteration. Our hope is that the word earns rigor the way words do, by entering scientific and academic discussion and getting argued over. We started this discussion around stabilization, but we don’t claim to own the word.

The public case against the word

Last week Holly Buck published the fullest public version of the case against the term, Selling Stability: The Problem with Pitching “Climate Stabilization” as a New Category of Climate Response, at African Tech Futures Lab. Holly has studied geoengineering and its politics for the better part of two decades and wrote After Geoengineering (which I loved, and you should read), and I learn something from her writing every time. She made her case in public, so I’ll respond to it the same way.

By my reading, we agree on many underlying premises. The tipping-element risks are real and potentially urgent and severe, a reflexive loop in which climate change erodes our capacity to respond to climate change is a real risk, the research landscape for response measures is broken, and a container for these interventions would help people think about them and would probably motivate additional research funding. The remaining disagreement seems to me to be over whether this container, with this name, costs more than it delivers. Her history section also complicates the prior-usage objection by identifying that the earliest use of climate stabilization in the scientific literature, a 1974 paper in Science, meant climate control and intervention, and she describes today’s usage as the term coming full circle.

Two of her arguments moved me. The first is about baselines:

Stabilization elides questions about whose baseline is being stabilized—and whose is being destabilized by stabilization itself. This is the same question that was asked in 1974, though from many more voices. For much of the majority of the world, things have not been stable at all. Indeed, for communities who have lived through colonialism, extraction, and displacement, “stability” was rarely on offer to begin with. Stabilization language casts political acts in terms of neutral systems maintenance, obscuring whose reality and power is actually being protected.

I partly agree. The objects in our definition are physical systems, and the destabilization of those particular systems lands first and hardest on the people for whom stability was never on offer. A disrupted Atlantic circulation would impact monsoons and agriculture, collapsing coral reefs would impact local food systems, and sea level rise will reach every low-lying coast on Earth. That is an argument about who has the most at stake. It is not an argument about what anyone wants, and we don’t pretend the two are the same. High stakes don’t produce consent. Deciding which systems get stabilized, at what baseline, and for whose benefit is the rightful job of politics. Our field definition refuses to presuppose which systems ought to be stabilized or what a stable future should look like, and Holly’s critique clearly shows why those questions deserve real deliberation. We are explicitly not trying to hide political choices behind a fig leaf of technology or technical descriptions. All of that social legitimacy and governance work can and should fit within the pillar.

The second is about what the word implies.

Stabilization presents an inaccurate picture of what is actually happening to the planet. Solar radiation modification is likely to stabilize global mean temperature, but create novel biospheric conditions, because the amount and type of sunlight is being changed and CO2 emissions would be decoupled from temperature. This would be a new situation for ecosystems, in ways that have not been seriously researched. As a recent piece by ATFL’s Climate Intervention Futures Initiative noted, stabilization describes an aspiration, not a guarantee. Stabilizing temperatures is not the same thing as holding the climate steady. Ultimately, it’s the planet itself that does the stabilizing.

I agree without reservation. Her line about it being an aspiration rather than guarantee is close to how I’d describe all three pillars. The medical sense of the word carries some of this already, since a stabilized patient is not a healed one. This is an example of where a critique pushed us to improve the clarity of the definition, and the definition below now says clearly that a stabilized system is not a restored one.

Where I disagree with the article is her conclusion that the container problem dissolves if we just call things what they are. She asks where the desire to taxonomize comes from, notes that a name changes none of the work, and asks who the sales pitch is for. They are fair questions, but they fold two things together, because the need for a class term exists before anyone tries to “sell” anything.

The first reason for the term is pragmatic. The potential interventions are so varied that on the surface they almost seem unrelated. What do wildfire management policy that keeps abrupt permafrost thaw in check, reflective coverings on glaciers, marine cloud brightening over reefs, cirrus cloud thinning, sea ice thickening, and stratospheric aerosols all have in common? Anyone who needs to discuss the class as a class, in a research agenda or a governance framework or a funder’s memo, either recites that list or borrows a word that fails somewhere, because geoengineering doesn’t fit the wildfire policy, cloud thinning is not sunlight reflection, and none of it is mitigation or adaptation. The nearest candidate is “climate interventions”, and we use it ourselves for naming the actions. But the field is more than its actions. The field includes the monitoring that tells us whether a system is approaching a threshold and it includes the governance work that decides whether anyone should act. None of that is “intervening”. And part of the field’s job is deciding whether to act at all, so naming the whole field for the act would prejudge the very question the governance work exists to ask. The definition below says we don’t presuppose that intervening is the right answer, and the field’s name shouldn’t presuppose it either. An umbrella term is useful when people keep needing to refer to a class of work the existing words don’t cover, and that need would exist even if nobody were trying to grow the field.

The second reason is the one the parameters at the top of this piece state outright. We see the need far more money and many more people in this work, and names and aspirations are, in part, how that happens. Right now one foundation might file this work under climate interventions, another under tipping point risks, another under extreme adaptation, and most file it nowhere at all, which is the fragmentation problem seen from a funder’s perspective. A single coherent term reduces that friction, and it gives the entrepreneurs and researchers who aren’t here yet a field they can find. A program doing marine cloud brightening can and should keep saying MCB, plainly; the umbrella term is for talking about the class and for growing it. A name doesn’t change any of the work to be done, but it does change who shows up to contribute to the work.

Holly raises more than I can take up here, on agency and on what the frame leaves out, and I hope you read her piece in full. I believe we agree on the deepest point, that the work matters more than the taxonomy, and the taxonomy is only worth arguing about because it affects how fast the work gets people and money. And ultimately, this term is a proposal, not dogma. If the field finds it useful, we’ll see it get adopted more widely. And if the field doesn’t, then we’ll all collectively figure something else out.

The definition, as it stands

Here is the full definition as we are now using it, revised by everything above. It will keep changing, and the feedback form in the observations report is the best channel for changing it.

Stabilization is the work of keeping critical Earth systems from crossing into runaway, self-reinforcing destabilization, the kind that could foreclose our collective capacity to respond to climate change at all. It is an emerging pillar of climate response that sits alongside mitigation and adaptation. Mitigation, which includes carbon removal, addresses the cause by reducing and drawing down greenhouse gases. Adaptation absorbs the consequences in human systems reducing harm. Stabilization, if it is feasible and desirable, intervenes directly in the Earth systems most at risk of nonlinear, possibly irreversible change, to buy the time that mitigation requires.

This is not the stabilization of greenhouse gas concentrations, and it is not holding global temperature at a fixed target. Those are objectives served by mitigation and removal. Stabilization is a near-term response to a risk profile that has developed faster than we have implemented mitigation efforts. Possible tipping elements including the Greenland and West Antarctic ice sheets, the Atlantic overturning circulation, permafrost, and warm-water coral systems are approaching limits on timescales that decarbonization and carbon removal simply cannot address in time.

Stabilization measures are meant to be temporary, wound down as mitigation and removal bring the climate back to safety. We are honest that the warming already committed to the system through past emissions means that some measures may need to hold for decades or longer. They remain a bridge, not a destination. A stabilized system is not a restored one. Interventions that keep a threshold from being crossed still leave the planet in a changed state, and what it means for ecosystems to hold in novel conditions is an area this field must research.

When we call stabilization a field, we mean something specific and modest. A field is a loose-knit body of people working across many disciplines, from climate science and engineering to finance, governance, and frontline engagement, in pursuit of a shared set of goals. It is not an organization, it has no central authority, and it does not speak with one voice. We are not declaring a field into existence. Much of this work is decades old, carried on by researchers and governance experts through years when pursuing it cost people funding and reputation. What we are observing is a field that is beginning to cohere. We can see this in the new research, new funding, and new initiatives emerging across these areas, and we are naming it so the people doing the work can find each other. Some of them will disagree about the name, and reasonable disagreement among contributors is a feature of a healthy field rather than a flaw in it. Even so, we assert that the moment has come for this work to organize and grow.

Avoiding catastrophic, irreversible destabilization is worth preparing to act on now, before the conditions for a considered response are gone. That commitment is to building tools and preserving options, not to deploying them. We do not assume that intervention is the right answer in any specific case. Nor do we presuppose which systems ought to be stabilized, or what a stable future should ultimately look like. Those are the scientific, governance, and legitimacy questions this work exists to consider. What we assert is narrower: building the capacity to choose well, in time, beats being forced to choose without it.

I believe stabilization is the right name, and none of the feedback has felt like a deal-breaker. It has felt like people pointing out where the term is unclear and could become clearer, and to me that means the process is working. Meadows wrote that paradigms change when people point at the anomalies of the old one and speak, with assurance, from the new one. That is what this piece has tried to do. And, helpfully, the critics were doing the same thing to us. They pointed at our anomalies, and the definition above is better for it. Perhaps Earth-system stabilization is a better term than climate stabilization. Or perhaps it’s better on its own as just stabilization. The formal definition should end up written by many more hands than ours. If the name helps you explain your work, use it. If it gets in your way, tell me how, because the definition is only going to improve by people pushing back on it.

Related: the Global Risk and Stabilization Summit happens October 12-13 in Iceland, and you can apply to attend at globalstabilization.org.

A little over a year ago, in July 2025, I spent four days at the Global Tipping Points conference in Exeter, listening to scientists lay out what happens if the Amazon tips into savannah or the Atlantic circulation shuts down, and then watching much of that same community treat the question of directly intervening in those systems as unfit for polite discussion. I wrote about that experience at the time, and I came home feeling radicalized. If these risks are as serious as the science says, then preparing to respond to them requires an enormous amount of work that has barely started: governance structures, social legitimacy, community engagement, and public narratives, in addition to all the engineering, field research, funding pipelines, and talent pipelines that any technical field needs.

The situation was worse than just an early start, though, because many of the people best positioned to argue for that preparation were instead arguing against it. I’ve written about the resulting dynamics before: the taboo blocks the funding, the missing funding blocks the people, and the missing people guarantee the work stays undone, which keeps the whole subject easy to dismiss.

So in the fall of last year I went out and pitched a project plan. I was calling it the Stabilization Plan, modeled loosely on Leopold Aschenbrenner’s Situational Awareness, and the idea was a targeted informational campaign making the case for climate stabilization to the audiences whose permission matters most. I eventually laid out the thinking here in this newsletter in The Stabilization Framework. I pitched it to philanthropies the way I had pitched Nori to venture capitalists, because that was the fundraising I knew how to do, and it turns out that’s not how you raise nonprofit money (oops). The feedback I got was remarkably consistent. Why start a new organization when the work would make more sense housed inside an existing one? And could the scope come down to something milestone-based rather than one grand campaign? Both pieces of feedback turned out to be right.

That’s when I had the good fortune of talking with Joshua Elliott, the chief scientist at Renaissance Philanthropy and leader of the Advanced Research for Climate Emergencies (ARC) program, who was interested in exactly this narrative dimension of the climate intervention space. I had also met Ben Lachman at a Climate Emergencies Forum workshop during New York Climate Week, a longtime entrepreneur whose background runs from software and hardware through fifteen years of climate mitigation work, and he became my partner in all of this. Together with Joshua we scoped a three-month research pilot, which kicked off in February.

If you come from the for-profit world, you would recognize what we did as customer discovery. We conducted 29 formal interviews with funders, researchers, practitioners, and policy actors around the world who work on climate interventions, asking where people get blocked, how fundraising actually goes, and what the politics look like, spoken and unspoken. We deliberately scoped the research to three risk domains: coral reefs, the cluster of Arctic risks, and permafrost. We chose those domains because when mainstream media writes about geoengineering it is almost always writing about stratospheric aerosol injection, and we wanted to test whether other interventions could be disaggregated from SAI. You do need planetary-scale governance for SAI, but you don’t need planetary-scale governance in place to start experimenting with marine cloud brightening to shade a specific local coral reef, for example. The prerequisites are fewer, and the work could start sooner.

We also ran a live test of the framing. In April we co-hosted an event during SF Climate Week called Stabilize Earth, which I wrote about at the time, and which turned out to be one of the field’s largest semi-public gatherings to date. All of this fed into a short observations report, where we acknowledged the biases plainly. We were very much talking to the converted, so the observations reflect converging views within the field rather than the views of the public or of skeptics.

Permission Space And Narrative Framing Report

148KB ∙ PDF file

Download

Download

(Download the report above or click here to open in a browser.)

The example that best captures the state of this field is permafrost. The frozen ground of the boreal north holds roughly twice as much carbon as the entire atmosphere, and under continued high warming, published estimates of what could escape as CO2 and methane by 2100 run to hundreds of gigatons of CO2-equivalent. When you ask people how many researchers worldwide are working directly on keeping permafrost frozen (not counting generic GHG mitigation efforts), even people who work in climate, almost nobody guesses anywhere near the real answer, which is a handful at most. That contrast between the scale of a risk and the number of people working on it repeats across the whole field.

By our estimate, all of climate stabilization runs on fewer than a thousand full-time people worldwide and on the order of tens of millions of philanthropic dollars per year. Climate mitigation, for comparison, has tens of thousands of organizations, millions of people, and roughly two trillion dollars a year in capital behind it. None of this takes anything away from mitigation, which as Ben likes to put it is the boulder already rolling downhill, just not quite fast enough. But the comparison is still valid, because this shows what happens to a category of climate response when the subject stays taboo long enough that people and funding never arrive.

The full report contains nine observations, and I would rather you read them than my summary of them, but a few are worth pulling forward here. Risk-first framing opens doors that solution-first framing closes. Latent support is larger than visible support, because speaking up in this field carries asymmetric career risk, and as we put it in the report, silence in this field is not opposition. Philanthropic capital follows institutional cover, not just conviction. Majority World institutional engagement with intervention research, in India and in Ghana among others, is moving faster than many Western observers realize. And nearly half of our interviewees independently named some version of the same missing piece: connective infrastructure that the field does not yet have.

That last observation is the one Ben and I are now acting on. Following the pilot, we joined Renaissance Philanthropy to lead the Climate Emergencies Forum and evolve it from the convening body it has been since 2024 into a proactive field-building organization. Our logic is that the destabilization of these Earth systems will eventually force governments to fund stabilization research and readiness at billions of dollars per year, and whether that capital arrives in time, and whether it lands well, depends on what exists to receive it. Absorbing investment at that scale takes people, institutions, de-risked science and engineering, and public legitimacy, and today the field is short on every one of those dimensions. Closing that gap within roughly a decade means working on three problems now.

1. The field needs more philanthropic capital. Tens of millions of dollars per year is the current reality, and most of it has gone to global sunlight-reflection approaches rather than the broader portfolio of system-specific interventions. We think this needs to grow by two orders of magnitude within about ten years, and in the near term that growth mostly has to come from philanthropy, because philanthropic capital is what grows a field to the point where governments can fund it seriously. (I plan to write separately about the role of for-profit companies in stabilization, which I think is real but different.)

2. The field needs more people. There are hundreds of full-time people today, and we would like to see tens of thousands within ten years. Permafrost is the extreme case, but the pattern repeats across the risk domains where there are whole categories of problems that essentially no one is working on. Ben and I have already been fielding calls from seasoned entrepreneurs who attended Stabilize Earth and want to contribute without knowing where to start, which tells us the on-ramps matter as much as the recruiting.

3. The field needs common groundwork. The most consistent finding across our interviews was that the field lacks shared infrastructure that no single organization can build for itself, so each one rebuilds it alone. That groundwork starts with shared definitions and frameworks that let a program officer or a newcomer place stabilization in context, and a shared map of the risks, the interventions, the funding, and the gaps, similar in ambition to what Frontier’s gap database did for carbon removal. It extends to the community-engagement protocols, governance frameworks, and regulatory strategy guides that every new organization currently reinvents from scratch. All of that duplicated effort slows everyone down and keeps the field hard for incoming talent and capital to find.

Offering stabilization as a name is part of the same work. This field has been forming whether anyone names it or not. We’re suggesting a name so that the people already doing the work can find each other, and so that funders and newcomers can find them, because you can’t recruit someone into a field that officially doesn’t exist.

Which brings me to the summit. In October, CEF is convening the Global Risk and Stabilization Summit in Iceland, October 12 and 13, intended as the field-wide gathering. There will be several hundred people in attendance, with the goals of building a shared picture of the field’s current state and creating the collaborations it needs next. If you are working in or want to work in this field, you can apply to join us directly at globalstabilization.org. I hope to see many of you there.

The report remains open for feedback, and the form linked inside it still works. We have already received a lot, especially on the word stabilization itself, where the objections have been good enough that they deserve their own article. That piece will come next.

Ted Nordhaus is a hard person to dismiss on this subject. By his own account he has been supportive of geoengineering research over the years: he argued in Foreign Affairs in 2018 that the two degree target created its own moral hazard by making adaptation and geoengineering research look unnecessary, he suggested in his review of Elizabeth Kolbert's Under a White Sky that future societies might engineer the climate to preserve biodiversity, and in 2024 he proposed using marine cloud brightening to moderate the natural climate swings of the El Niño cycle. When someone with that record announces that he now doubts the wisdom of even the research, it's worth finding out where the argument goes wrong.

And he gets a lot right. The break-glass-in-case-of-emergency framing deserves to die. There is no observable threshold where the world will agree that the emergency has arrived, and if we wait for one, we will have waited too long, because interventions at meaningful scale have lead times measured in years while the outcomes worth preventing become unpreventable quietly, well before they become undeniable. I used the framing myself when I first started writing about climate interventions; one of my early pieces is literally titled Break Glass, Cool Planet. What moved me off it was Oliver Morton’s The Planet Remade, which makes the best case against the emergency framing I’ve read. A technology you only reach for once the catastrophe is obvious is a technology you reach for too late.

He is also right that humanity has, so far and in aggregate, adapted faster than the climate has changed. Climate-related mortality has fallen dramatically over the past century, and disaster losses have stayed roughly flat as a share of the wealth exposed to them. Those trends are real, and parts of the climate movement do themselves no favors by pretending otherwise. The aggregate hides a great deal, since the adaptive capacity sits mostly in rich countries while the residual risk sits mostly everywhere else, but the direction of the trend is what he says it is. And he is right that catastrophism runs through parts of the climate conversation, the strain that gave us deadlines like twelve years to save the world, and that some advocates of cooling research have picked up the habit, describing the present as already catastrophic partly because catastrophe justifies the tool. That drift is real, I have watched it happen, and it deserves the criticism.

So I agree with more of his premises than I expected to when I opened the piece. The argument still fails, and it fails at one specific step. Ted claims that rapid climate change, the kind that outruns adaptation, is “almost singularly associated with geoengineering termination shock.” Most of the essay leans on that sentence, and the sentence is wrong.

Where the argument turns

His strongest argument runs something like this: human societies handle slow change well and fast change badly, so the rate of change matters as much as the magnitude; anthropogenic warming has been fast on geological timescales but slow on human ones, which is why adaptation has kept pace; termination shock would compress decades of warming into a few years, which adaptation cannot absorb; therefore the technology built to manage climate risk is the one thing that could turn a manageable problem into a catastrophe. As far as the mechanism goes, I think he is correct: speed, more than size, is what breaks adaptation.

Where he goes wrong is in believing the Earth system only moves fast when people push it. The Greenland ice cores record more than two dozen episodes during the last ice age, the Dansgaard-Oeschger events, in which temperatures over Greenland jumped by 8 to 16 degrees Celsius within a few decades. The driver was the Atlantic ocean circulation abruptly reorganizing itself. No one was flying stratospheric aircraft in the ice age. The climate did that on its own, repeatedly, and the same circulation is the one making headlines today as the AMOC, whose stability is an open research question: one much-debated study put a possible collapse near the middle of this century, another found it on course to tip without saying when, and others argue the observational record is too short to predict anything at all. That disagreement is the uncomfortable part, because it means we are leaning our weight on a system with known fast failure modes without knowing how close the edge is. The same concern applies to the West Antarctic ice sheet and to the carbon starting to leak from thawing permafrost.

This matters for his adaptation argument too, because the adaptive capacity he celebrates is a product of the stable conditions it grew up in. Wealth, infrastructure, insurance, supply chains, and functioning states are what stand between climate hazards and human suffering, and every one of them assumes a stable background against which to keep compounding. A world absorbing an Atlantic circulation disruption or simultaneous crop failures across major growing regions is a world with less capacity to adapt to everything else, a dynamic Laurie Laybourn, who leads the Strategic Climate Risks Initiative, has named derailment risk: the consequences of destabilization eat away at the capacity societies need to address its causes. You cannot keep absorbing systemic shocks forever, because the shocks consume the capacity you absorb them with.

The experiment already ran

The strangest omission in Ted’s essay is an experiment that already ran. On January 1, 2020, an International Maritime Organization regulation cut the permitted sulfur content of shipping fuel from 3.5 percent to 0.5 percent, and sulfur dioxide emissions from the global fleet fell by roughly 80 percent within a year. Shipping exhaust had been brightening marine clouds for decades, visibly, in the satellite record. The regulation switched a large share of that off almost at once.

The lead paper on what happened next has the phrase “inadvertent geoengineering termination shock” in its title, and that is peer-reviewed language, in Communications Earth & Environment, describing a shipping regulation. A modeling study in Earth’s Future estimated the regulation pulled global warming forward by roughly two to three years, and estimates of the added heat imbalance run from about 0.06 to 0.2 watts over every square meter of the planet, depending on method, which is significant for a single rule change. The 2023 and 2024 temperature spikes had other contributors too, including El Niño and the Hunga Tonga eruption, and scientists are still arguing over how much of the jump came from the shipping rule. The direction is not in dispute.

I take two things from this. First, humanity has been running an uncontrolled aerosol cooling program for decades and terminated a large piece of it abruptly, through an air-quality rule made with no climate deliberation and no plan for measuring what would follow, which means the ungoverned path is already a matter of historical record. Second, the event confirms that termination effects are physically real, and that is the argument for treating termination as a design constraint rather than a verdict. Andy Parker and Peter Irvine worked through the failure pathways back in 2018 and concluded that a deliberate program built with backup deployment hardware and operated by agreement among several capable states would be resilient against all but the most extreme catastrophes, because sustaining a termination shock requires every capable actor on Earth to stop, simultaneously, for years, while temperatures climb around them and each has an overwhelming incentive to restart. Termination risk is a reason to engineer and govern deployment carefully, and a strange reason to know less about the system we would be governing.

The march is underway

Ted writes as though stopping the research would stop the march toward deployment. The companies he links to are evidence against that. He cites Stardust, the Israeli-American startup that raised 60 million dollars last October, the largest known funding round in the field’s history, with a business model of selling deployment capability to national governments and outdoor experiments already planned. Stardust took heavy criticism early on for keeping its particle chemistry secret, and in May it started publishing: white papers and preprints with university collaborators, disclosing particles built from amorphous silica and calcium carbonate. People I talk to in the research community were glad to see it, and I was too, though the safety questions are far from settled. Make Sunsets, meanwhile, has been lofting balloons and selling cooling credits for years and recently made its patent available royalty-free. Whatever you think of these companies, the capability is being built regardless of what academic researchers and their funders decide. And notice what pushed Stardust toward openness: a research community that existed to demand it and scientific norms that gave it something to publish into, which is the infrastructure Ted proposes to strip away.

So run his scenario forward under his own proposal. Suppose the research stops: the university programs wind down, the philanthropy dries up, the monitoring and attribution science never matures, the governance conversations never happen. The political incentives he describes remain, and so do the startups. The next accidental climate jolt arrives with no one equipped to measure it. If you wanted to guarantee the outcome Ted fears—deployment arriving through commercial secrecy or political panic with no public science and no institution capable of stopping or steering it—this is how you would arrange the world. Opposing research is a recipe for the exact worst outcome he believes might happen. The choice in front of us has never included a world where nobody builds this. The choice is between preparing in a way that mitigates the risks as much as possible and just allowing it to happen.

The position the horseshoe can’t see

The most interesting claim in the essay is sociological: that advocates and opponents of geoengineering have converged on identical, catastrophist views of climate risk, distinguishable only by the solution attached. There are people who fit that description. But the taxonomy has a hole in it, and I know because I live there.

I spent a decade on the other side of this problem. I co-founded Nori, one of the first carbon removal marketplaces, and spent years convincing investors and researchers and policymakers that removing carbon dioxide from the atmosphere was necessary and possible. I believed we could scale our way out. What changed my mind was the arithmetic: carbon removal is decades away from moving global temperature, the tipping-point risks arrive on a faster clock, and the gap between those two timelines has to be covered by something. I am not a catastrophist about the present. I don’t believe we need to overhaul the global economy to respond to climate change, and I came to this work through market-based solutions. Ted’s horseshoe has no place for that position, and a growing part of this field lives in it.

His use of the word geoengineering deserves a look too, because in the essay it means one thing: global stratospheric aerosol injection, with its globe-spanning politics and its worst-case termination profile. Most of the intervention portfolio looks nothing like that. Marine cloud brightening is regional, and its clouds stop responding within days of stopping. Proposed interventions to slow ice sheet loss are anchored to particular glaciers and can be removed. Ted knows the difference, because the MCB program he proposed in 2024 to moderate the El Niño cycle was (even offered half in jest as he says it was) a bounded, regional intervention whose logic depends on the distinction his new essay erases. The question his essay never answers is what separates the intervention he proposed from the ones he now opposes. The way I hold the whole picture now: mitigation and carbon removal address the causes of the problem, adaptation shores up the human systems that absorb the consequences, and stabilization protects the climate systems that both of the others need time to work. Naming the pillar for its outcome also sets the standard it has to be judged against, something the word geoengineering never did.

Where does that leave the disagreement? Ted’s real question, underneath the essay, is whether a world this bad at telling real emergencies from manufactured ones can be trusted with a lever this large. That is a good question. It is also a governance research question, and governance research is part of what he is proposing to stop. The monsoon impacts, the termination pathways, the attribution science, the institutional designs that could make starting and stopping legitimate: these are the research agenda, and the people working on them are, in my experience, more sober about the risks than either pole of his horseshoe. He is asking one of the right questions, and I would welcome the exchange, because this field is better off with its sharpest critics inside the conversation. Readers who want to see what the field contains before picking a side can start with my standing guide, How to Learn Everything You Need to Know About Climate Cooling.

Subscribe now

Every plan to cool the climate or hold an ice sheet together counts on capable governments to pay for the work and run it, and enough cooperation between them to act together. That’s a bigger assumption than it looks, and right now it’s coming apart.

Start with the trust between governments. In January, Mark Carney told the World Economic Forum that the international order is in a rupture, not a transition, and he named the institutions he sees under threat: the World Trade Organization, the United Nations, and the COP—the climate negotiations themselves. Carney ran two central banks before he became Canada’s prime minister, so he isn’t a man given to drama. You can see what he means in how this year’s two big wars are being handled. The US and Israeli war on Iran that resulted in the shutdown of the Strait of Hormuz, and Russia’s grinding war on Ukraine, are being managed, to the extent they’re managed at all, by leaders cutting private deals with each other rather than through any institution built for the job. Trump ended the Iran war over dinner with Macron. The talks his team have been brokering on Ukraine have gone nowhere. Either way, the shared machinery for handling big problems is being bypassed in real time. It’s the era of “the art of the deal.”

Now look at the governments themselves. Just this week we saw Keir Starmer forced out as British prime minister by his own party, the sixth British PM to go in seven years, with Nigel Farage’s Reform UK pushing for an early election. The populist churn behind that isn’t unique to Britain; it’s running through most of the democratic world. So just as more weight would fall on individual governments to act, those governments are less steady than they’ve been in a generation.

Put those two things side by side. The shared institutions are fraying, which pushes more onto individual governments, and those governments are wobbling like a glass vase during an earthquake. Both halves of what a climate intervention would need—capable governments and enough cooperation to act together—are eroding at once. That’s the world any serious plan to cool the planet now has to survive in.

If we want a field that can put serious money and real operations behind interventions in, say, ten years from now, the governments we’re counting on have to be capable ten years from now too, and that capability doesn’t build itself. Watching where the world is heading matters just as much as watching the science on tipping points and intervention options. Stabilization already has a lot of moving parts that people are working hard on: the research, the funding, the field itself, and the genuinely difficult job of running an intervention in the real world. State capacity is one more, and I think it’s the one we’re paying the least attention to.

What stabilization is

Stabilization, the way I use the word, is the near-term work of keeping critical Earth systems from tipping into runaway change, the kind that would wreck our ability to deal with climate change at all. (I’ve written about it at more length in The Climate Stabilization Framework.) Cutting emissions and removing carbon dioxide go after the cause. Adaptation deals with the damage. Stabilization does something else. It steps in directly to protect the systems most at risk of a sudden, irreversible tip: the ice sheets, the Atlantic overturning circulation, the permafrost, the coral reefs, the Amazon, etc. The point is to buy time for the slower tools to catch up, because if we don’t, then our society will be hit by famines, droughts, migration, populism, and more, things that society not only can't withstand but that would eliminate our ability to address climate change at the root.

Stabilization is meant to be temporary, with interventions safely wound down as cutting emissions and removing carbon bring the climate back toward safe temperatures. And it works directly on shared systems that no company can own or sell, which makes it the one part of the climate response you can’t build a market around. And even if you wanted to try, what’s the business case? You can sell carbon removal today (with a lot of difficulty) and it hasn’t come close to affecting the total amount of CO2 in the air. You can’t “sell” a stable ice sheet. So more than any other part of the climate response, stabilization depends on governments that work.

What a capable government needs

So what does state capacity actually mean? The plain version is whether a government can get done the things it’s supposed to get done. Some governments are good at it, plenty aren’t, and the worrying part, which Francis Fukuyama has written about for years, is that even rich, long-stable democracies can lose the knack over time. There’s a useful distinction, from the sociologist Michael Mann (funnily enough, not the climate scientist Michael Mann, who hates this stuff and said so again in the Guardian this month), between a government’s power to give orders and its power to actually carry them out on the ground. The second kind is what stabilization needs, and it’s the harder kind to build.

What the climate-scientist Mann and his co-authors argued, as it happens, makes my case for me. Their warning is that cooling schemes would be wildly expensive, and that once you start one you can’t stop without risking a sudden snap-back of all the warming you’d been holding off. They mean it as a reason not to try. It’s also a near-perfect description of why this would take a strong, steady government rather than a market. The best argument against doing it doubles as the argument for being able to. And that’s where the skeptics lose me. A dependency this big is a map of what we’d have to build, not a reason to walk away from it.

The way I usually get a handle on a problem like this is to start from the goal and work backwards. What are we actually trying to make possible? Then, what has to be true for that to happen, what are the pieces, is there an order to them, what does each one depend on, and so on until you can see what has to get done now. So: the goal is to be able to carry out a stabilization intervention well and responsibly when we need to (and from where I sit, I feel confident we’re going to need to). That means de-risking these enough that doing one is a real option rather than a desperate gamble. What would a government have to be able to do to get there?

I count six things, and they aren’t a checklist you work through top to bottom. Think of them as links in a chain of dependencies. Some are genuinely sequential, where one has to be in place before the next is even possible. Others build each other up at the same time. The reason the chain is the right picture anyway is that all six have to hold. A weak link stops everything that depends on it, however strong the rest are.

It starts with seeing the threat coming, which means setting up the monitoring and sensing to track these systems and understand what’s happening to them. Underneath even that is the most basic thing of all, the will to want to see the threat, which plenty of governments don’t have. The United States right now is choosing not to look. Nothing else on the list matters if a government won’t let itself notice the problem in the first place.

Before a government can do anything this big and this contested, it needs permission, and that comes in two forms, one deeper than the other. The deeper one is legitimacy: enough trust at home, and enough consent from other countries, that people actually accept the intervention instead of fighting it. It’s the slowest of the six to build and the easiest to lose, and for anything at planetary scale it has to come early, because nothing downstream gets authorized or funded without it. Up until now, I’ve been referring to this legitimacy as the “permission space” necessary to work on these things. The legal kind follows from the legitimacy, and it is the formal authority to permit the work, oversee it, and enforce the rules. That authority, or the lack of it, stops a small outdoor experiment far more often than money or science ever does, and it tends to follow public consent rather than lead it.

None of the cross-border parts of that—the shared consent and the authority that spans jurisdictions—hold together without coordination, both between agencies intra-government, as well as between governments (that might agree on very little). Coordination is the connective link running underneath the others, and it’s the one the current breakdown is fraying fastest.

Then there’s paying for it, and, just as important, continuing to pay. That’s a different thing from deciding to spend once, because stabilization will play out over decades while budgets run year to year and election to election. The hard part is going to be the second and the third and the tenth and the hundredth payment, many years after political movements enabled the first payment.

Only with all of that in place can a government actually do the work (or contract it out), which takes the people and the institutions to run something big and complicated for years on end. That muscle can’t be assumed. Anyone who watched California try and fail to build high-speed rail knows how hard it is for a capable, wealthy government to just build a thing.

Some of this is being built right now. The clearest example is the pay-for-it link in the United Kingdom. A government agency there called ARIA runs a research program on climate cooling worth around fifty-seven million pounds, the biggest pot of public money for this kind of work anywhere in the world. It’s also a good reminder of how fragile these links are. ARIA was set up by Boris Johnson’s Conservative government, which was three (soon to be four!) governments ago. The cooling money is committed, but whether the next government keeps funding a speculative cooling program, with budgets tight and politics turning, is anyone’s guess.

Other links are being torn down just as fast. This is most obvious in the United States, where the Trump administration has been pulling apart the climate spending in the Inflation Reduction Act and cutting climate programs, including some of the monitoring that the see-it-coming link depends on. Put it together and the picture is sobering. No government has anywhere close to all six links, though plenty have a strong one or two. None has the whole chain, and that will need to improve rapidly, because a strong funding link or a strong research link doesn’t add up to being ready on its own.

It’s getting built unevenly

That’s the sobering version. The hopeful version is that these links do get built, country by country, and how it happens is something we can actually learn from. It mostly comes down to framing. How a government describes a risk decides which part of the government wakes up to it, and that matters more than you’d think.

Let’s look at Iceland for an example of this. Late in 2025, its government became the first to officially treat a possible collapse of the Atlantic overturning circulation as a national security threat, not just an environmental one. That sounds like wordplay until you look at the effects of that. Calling AMOC a security threat sent it to Iceland’s National Security Council, the one body that can put every ministry in the same room at once. Suddenly the people who handle the country’s energy, food, transport, and infrastructure were planning around it together. If you frame it as an environmental issue, I have a hard time believing that any of that happens. I talked with Páll Gunnarsson of the Reykjavík Institute, who was close to all of this, about how a country of four hundred thousand people moved faster than far bigger ones. Some of that is particular to Iceland, where the Gulf Stream is common knowledge and the government is already built around hazard preparedness. But the lever, the part that could be duplicated by other states, was the framing: security, not environment.

Finland did a quieter version of the same thing. While it held the rotating presidency of the Nordic Council of Ministers, it pushed tipping points onto the agenda, and the result was a Nordic report on AMOC earlier this year that treats the risk as a governance problem for the whole North Atlantic. The response to this report is what I found interesting. The same science landed in front of several governments and got very different responses. Norway is still deciding whether to call AMOC a security risk. The UK noted the reports but figures a sudden collapse is unlikely this century, and Ireland briefed its prime minister and a parliamentary committee but has not taken it further. You could imagine tracking this properly, every country scored on all six links, so you could see at a glance who has actually built what.

The derailment risk

Now put the two halves together, the links being cut and the links being built, and you get a loop other people have already named. Laurie Laybourn and his colleagues call it the doom loop, or derailment risk: climate damage chips away at a society’s ability to deal with climate change, which lets the damage get worse, which chips away further, and around it goes. The same tipping points that would make us want to intervene are the ones that would wreck the funding and the institutions and the coordination we’d need to intervene at all. Stabilization needs state capacity, and state capacity needs stabilization. The crisis in the news is that loop running in real time. A world that can freeze its own energy markets in a couple of weeks, run through six UK prime ministers in seven years, and hear another one of its leaders (Carney) say the climate talks are part of an order that’s falling apart is a world quietly losing the ability to do the kind of slow, patient work that takes decades.

This is the part that changed my own mind. I got into climate work in 2015 believing markets could do most of the heavy lifting, and I still think that’s true for a lot of it, because markets are how we’re actually decarbonizing and how a lot of adaptation will get done. There will be companies and markets in stabilization too, which is something I’ll write about another time. But follow the logic of wanting stabilization to actually succeed, and you end up somewhere that this pro-market classical liberal/libertarian did not expect ten years ago. The specific things stabilization requires, the planet-scale interventions nobody can sell, can only be done by governments that work. Wanting those interventions means caring whether governments are capable, which is not how I used to think. There’s even a name for roughly where I’ve landed, state capacity libertarianism: what’s worth asking of a government is just that it actually be able to do what it sets out to do, whether that’s a lot or a little. Wanting a government that works is not a small thing to want anymore. It might be one of the most important things this whole field comes to want.

Which gets me to the one thing I’d ask of this field. We watch the science on these interventions closely, and the engineering, and we should watch the state of the world with the same attention, as a real input and not background noise. If you aren’t paying attention to where things are headed, you won’t be ready to plan, or argue, or act when it counts. I ended the framework piece by saying the goal is to build the ability to choose well, in time, instead of being forced to choose with no good options left. Here’s what I’d add. That ability was never only about the science or the technology. It’s about whether institutions and governments can actually function, and that capacity is eroding right in front of us. Helping build it, helping make those institutions capable and trusted enough to do this work, is part of the job too.

Subscribe now

Link to previous article I wrote on this:

Guest introduction

Ross Centers founded what became the Planetary Sunshade Institute in 2018 and is the founder and CEO of Ethos, where he’s developed electrochemical metallurgy techniques on simulated lunar regolith. Morgan Goodwin is PSI’s executive director and has spent twenty years working in climate. The Institute recently convened the first planetary sunshade workshop in Nottingham, drawing more than eighty researchers, and is funded in part by the UK’s ARIA to baseline sunshade architectures. Learn more at planetarysunshade.org.

Key topics with timestamps (also use as YouTube chapter markers)

• 04:37 — A field comes into being. How a single paper from the 1980s and a workshop of eighty-plus people in Nottingham turned the sunshade from an idea into a research field.

• 07:52 — Welcome to climate. Morgan on what it’s like to enter climate work from the space world, and why the people closest to you often won’t engage.

• 22:13 — What a Dyson Swarm actually is. Solar-powered data centers orbiting the Sun, and the Kardashev-scale logic that points there.

• 27:09 — The planetary heat limit. Why even a fully renewable civilization eventually warms the planet through waste heat alone.

• 28:59 — SpaceX as a Dyson Swarm company. Ross’s read on what the largest IPO in history is actually for.

• 34:58 — Mining the Moon. Morgan’s Hoover Dam analogy and the economics of building from lunar material.

• 47:06 — The Heliogyro. A forty-kilometer spinning solar sail made almost entirely from aluminum, walked through step by step.

• 57:41 — Who governs the climate. Why only governments can deploy SRM, and what “the climate government” means.

• 1:01:29 — Sunshade and SAI as complementary backups. Why having two cooling technologies beats having one.

• 1:16:01 — Does a sunshade ever come down? The generational-memory question, and whether planetary management becomes permanent.

Inevitable & Obvious is where I work through the technologies and the politics of cooling the planet, with the people actually building the field. If conversations like this one are useful to you, subscribing for free at inevitableandobvious.com gets you every episode plus the writing that goes deeper than the audio.

Subscribe now

A planetary sunshade, if you haven’t encountered the concept, is exactly what it sounds like: reflective structures positioned in space between Earth and the Sun to reduce the amount of solar energy reaching us. The idea has been around since at least 1989, when James Early at Lawrence Livermore National Laboratory wrote a short paper laying out the basic physics of placing a reflector at the L1 Lagrange point, a gravitationally stable spot about 1.5 million kilometers from Earth, roughly 1% of the distance to the Sun. Early’s paper was concise and elegant, but it stayed almost entirely theoretical for 35 years.

That makes this workshop a seminal moment. Ross Centers, founder of the Planetary Sunshade Institute, opened by framing it as the transition from “a collection of papers and interesting ideas” into a field. A field has contributing thinkers, conferences, shared vocabulary, competing design proposals, and institutional credibility. The IAA co-sponsorship matters here because it anchors the sunshade conversation in aerospace engineering rather than atmospheric science, which is a different credibility base and a different audience entry point than the one SAI research has used.

Within the SRM world, the sunshade has long been the idea people acknowledge politely and then move past. Stratospheric aerosol injection is cheaper, faster, more studied, and deployable on a timeline that matches the urgency of the Earth’s energy imbalance. The sunshade, by comparison, requires building things in space at a scale humanity has never attempted. When I’ve talked with people in the SAI community about sunshades, the general take has been something like: sure, interesting, but we have more pressing problems. And the sunshade community itself largely agrees. Ross has said explicitly that he won’t support sunshade development without SAI proceeding in parallel, because if there’s a gap in coverage while the sunshade scales up, you risk exactly the kind of abrupt warming you’re trying to prevent.

So I went to Nottingham with genuine curiosity but also with the zero-sum question in mind: if funding is limited, why should any resources go here instead of toward SAI?

The answer I came away with surprised me.

What you’re actually building

Among all the presentations I saw, there were two main design families that dominated the technical discussions.

The first is solar sails: very thin reflective foils, massively large, positioned to block or redirect incoming sunlight. Solar sails have real flight heritage. NASA has tested several, from the small NanoSail-D experiment in 2010 through the Advanced Composite Solar Sail System launched in 2024, which demonstrated deployment and controlled orbital maneuvering using photon pressure. The engineering challenge with solar sails at sunshade scale is control. Photons exert pressure on the sail surface, and at the dimensions you’d need, maintaining precise positioning is a hard, unsolved problem.

The second design is the heliogyro: imagine a wheel with long blades instead of spokes, where each blade rotates on its own axis while the entire structure spins. The spin keeps the blades taut through centrifugal force, which may be better for the control problem. This is the design the Planetary Sunshade Institute favors for further research.

The diameter of a single heliogyro unit would be roughly 40 kilometers long. To shade about ~1% of incoming sunlight, you’d need over 5,000 of these units, with a total system mass of about 67 megatons. That is just an insane amount of mass to get into space. Which means boosting all of it from Earth probably isn’t the right pathway. According to an estimate presented at the workshop, 99.96% of the mass would need to come from the moon, with only about 23 kilotons launched from Earth (roughly 230 Starship flights for the sunshade components alone, with more needed for the lunar mining and factory infrastructure). If all that is true, then it follows that you can’t have a sunshade without industrializing the moon first.

The engineering is early. People at this workshop were presenting design concepts for the first time, and nobody is building hardware. The research projects represented here were largely preliminary, and there have been no major collaborations between space agencies on this. But there is now an organized technical community doing real engineering work, not just writing speculative papers, and the IAA affiliation gives that work institutional grounding.

SAI now, sunshade later

The sequencing logic I kept hearing at the workshop runs like this: SAI is the better overall near-term intervention. It’s deployable within a decade, relatively inexpensive (I’ve written about this before, but it’s estimated to cost roughly $10 billion per year for significant cooling), and the physics are (relatively) well understood. The sunshade is the long-term replacement, coming online in the second half of this century as SAI is gradually phased out.

Ross framed the two as “harmonious and complementary approaches” and described the relationship using a cycling metaphor: the sunshade rides in SAI’s peloton until conditions allow it to break away as the lead vehicle. His target is what he called “parity-of-readiness,” meaning that by the time SAI deployment begins, the sunshade community should already have demonstrators flying and a buildable plan on the shelf.

Given the hypothetical choice (with all costs and political constraints momentarily ignored) between injecting sulfate particles into the stratosphere and putting a reflector in space, I’d likely lean toward the reflector. SAI is clearly physically and politically plausible, but it introduces atmospheric pollutants, carries real risks of disrupting regional precipitation, and would likely make the sky noticeably whiter (which I think is both minor and maybe the most sad aspect of SAI). Ross compared SAI to chemotherapy and the sunshade to precision immunotherapy; I think that overplays how precise a sunshade would actually be, but the directional intuition feels correct. Reflection from space avoids stratospheric chemistry and aerosol side effects, can be adjusted or even shut off, and doesn’t add new particulates into the atmosphere. It introduces different risks rooted in large-scale space operations, dual-use and security concerns, and the politics of who controls a planet-scale piece of infrastructure in orbit. But on the overall physics and just straight vibes of the whole thing, the sunshade feels far more elegant to me.

Why the economics changed

One of the biggest obstacles to taking the sunshade seriously has always been cost. We’re talking about an estimated trillion+ dollars to build. In a world where climate philanthropy struggles to raise even tens of millions for SAI or climate stabilization research, the idea that anyone would fund building thousands of 40-kilometer structures in space for climate reasons alone has been, to put it generously, aspirational.

What shifted my thinking at this workshop was what I’d call the convergence argument: the economic forces pushing toward space industrialization have nothing to do with climate, and as it happens, they’re building exactly the infrastructure a sunshade would require.

If you look at the growth curves for compute and energy demand, I just don’t see any reason they’re going to flatten. And it’s not only compute. We’re electrifying everything as we decarbonize1, which means energy demand goes up even as emissions go down. This is the story of human civilization: energy demand constantly increases. If you play those curves out over time, generating energy off-planet starts to become cost-competitive, and eventually necessary. And though it may seem ludicrous to mention today, there is an actual thermodynamic ceiling on how much energy you can produce on Earth’s surface before waste heat alone becomes a climate problem. Current global energy use is around 18 to 20 terawatts. At roughly 25 times that level, waste heat adds about 1 watt per square meter of radiative forcing, which is significant. At 75 times current use, you’re at 3 watts per square meter, enough on its own to substantially warm the planet even in a fully decarbonized world. That ceiling is still far off, but the market forces pushing us toward off-planet energy generation will kick in well before we reach it.

The early stages of this transformation off-planet are already underway. Overview Energy is developing space-based solar power, with satellites in geosynchronous orbit collecting continuous sunlight and beaming it back to Earth. They’ve completed an airborne demonstration and are targeting a first orbital demo in 2028. Starcloud is building orbital data centers with large solar arrays and radiative cooling for GPU-heavy AI workloads, betting that abundant solar energy and the vacuum of space as a heat sink make orbital compute cost-competitive with terrestrial facilities. These are early companies working on early technology, but they represent a capability we’re going to have to develop regardless: generating and consuming energy off of Earth. And building that capability is also building the industrial base you’d need for a sunshade.

And then there’s SpaceX, which is raising what will likely be the largest IPO in history next month. At a recent keynote, Elon Musk discussed long-term plans for lunar industrialization, including a concept for a “terafab” and a mass driver(!) on the moon2. SpaceX isn’t the only company in this domain, but their ambitions set the scale for what the industry is heading toward. Just this week, NASA held a conference laying out their own phased moon base timeline: first crewed mission by 2029, semi-annual missions through 2032, and sustained human presence beyond that, with cargo capacity scaling from 4,000 kilograms to 150,000 kilograms per delivery.

The moon has aluminum and silicon in abundance. If you’re already harvesting lunar resources for commercial purposes, building orbital infrastructure for energy and compute, and launching at the frequency that SpaceX’s cost curves suggest, the sunshade stops being a standalone climate megaproject. It becomes a climate application built on commercial infrastructure that’s coming regardless of whether anyone cares about Earth’s energy balance.

To be clear, I don’t want to overstate this. The trend lines point in this direction, and you can see how the pieces fit together, but any of a dozen things could slow or redirect the trajectory of space industrialization. What I can say is that when I arrived at Nottingham, I assumed we’d probably build a sunshade eventually, but I couldn’t tell whether that meant the 2040s or the 22nd century. I didn’t yet see where it fit within our current trajectory. I left the workshop thinking the overlap between space industrialization and climate need gives it a much more concrete path than the pure-climate-funding frame ever could.

The zero-sum question, revisited

I posed the zero-sum funding question to many people at the workshop. If five million dollars could go to SAI research, or marine cloud brightening, or ice sheet protection, or sunshade research, where should it go? The honest answer is almost certainly one of the nearer-term stabilization interventions, where even single-digit millions can make meaningful impact now. Most sunshade researchers would probably agree.

Two responses changed how I think about the question, though. The first was about different money. The SpaceX IPO is going to create enormous new wealth among people who are interested in space for space reasons, not climate reasons. Those newly wealthy people represent a different potential funding pool than climate philanthropy, and they might be more naturally drawn to the sunshade than to atmospheric particle injection.

The second was the technology-tree argument, and it resonated with the part of me that spent years playing strategy games as a kid. If you’ve ever played a game like Civilization, you know that certain technologies require prerequisites, and you can’t skip the early steps just because you need the later ones urgently. Solar sail control systems, heliogyro dynamics, and lunar resource extraction techniques all have their own development timelines. Getting that work started now, even modestly, is how you have capability ready when the commercial space infrastructure matures enough to make full-scale deployment practical.

The governance question

On the third day of the workshop, there were breakout sessions covering different aspects of the field, from technical requirements to financing to governance. I participated in the governance breakout, with about a dozen people for 80 minutes. I’ve sat in on, read, and watched a lot of governance discussions about SRM over the past year, and this one was a genuinely useful exercise that surfaced some of the more interesting ideas I’ve encountered. The group ran a backcasting exercise from the year 2100: imagine a sunshade is operating successfully, and describe the governance system that made it possible. Then work backward.

The model that drew the most energy in the room was that of an independent central bank: an institution given a mandate (temperature stabilization for a thriving civilization), operating with its own resource mobilization capacity, technology-agnostic across a defined toolkit, making evidence-based decisions independently of direct political control while remaining accountable to a representative framework. The Bank of England was the reference case. The room explicitly rejected a UN-agency model as lacking the autonomy and enforcement power the task requires.

The group was already grappling with the hardest problems. Weaponization came up repeatedly: a sunshade is dual-use technology, and whoever controls it controls Earth’s thermal management from orbit. Then there’s what someone in the room called the “Sunshade Industrial Complex” risk, the concern that at trillion-dollar scale, the institution maintaining the sunshade develops its own incentive to keep expanding it. The group’s response to that was interesting: the capability itself persisting isn’t the problem, because climate management is going to be a permanent need. What you have to guard against is the logic of “we need an ever-larger sunshade” becoming self-justifying.

I’ll be recording a podcast with Morgan Goodwin, the executive director of the Planetary Sunshade Institute, and Ross Centers next week, where we’ll dig deeper into both the technical and governance questions. More on that soon.

Where I landed

I went to Nottingham thinking the sunshade was genuinely interesting but pretty far out. I couldn’t see the path to it becoming real in the way I can see that path for SAI or marine cloud brightening or even ice sheet interventions. I came away thinking the trajectory of space industrialization makes this far more plausible than I’d appreciated, in ways that have nothing to do with the usual arguments for or against space-based reflection.

SAI research remains the right near-term priority. Nobody at the workshop disagreed with that. The sunshade isn’t going to help with the warming we’ll face in the 2030s and 2040s. But if you zoom out to the rest of this century, to the question of what long-term temperature management actually looks like at the scale and duration we’re going to need, the sunshade is where the economic trend lines and the climate need meet. The world is unlikely to fund a trillion-dollar climate project any time soon. But it may build the infrastructure for other reasons, and a planetary sunshade becomes the highest-value climate application you can run on it.

The next workshop is in two years and I plan to go back. My favorite thing about doing this work is meeting the people who are directly building toward these solutions, and this was one of those rooms. There were 70 people from around the world who showed up to a university in the English Midlands to work on something that most of the climate world still treats as science fiction. I think they might be early, but I don’t think they’re wrong.

Sunlight reflection has spent decades as an idea scientists batted around at academic conferences. Now venture-backed companies are racing to build deployment systems, state legislatures are passing bans on weather modification research, and the public still mostly thinks the contrails overhead are a government plot. Ben Kalina has been pointing a camera at this entire evolution for fifteen years. Plan C for Civilization, his documentary on the field’s emergence as serious research, is the closest thing we have to a real historical record of how this is unfolding. We talked about what changed Ben’s mind about the timeline, why average audiences leave his film more curious not less, the strange dynamics of venture capital entering a space that runs on trust, and what of the Stardust Solutions optics problem can and cannot be engineered away.

Ben Kalina is a documentary filmmaker and a professor at Drexel University. His new film Plan C for Civilization follows the emergence of solar geoengineering research, centered on David Keith, the SCoPEx experiment at Harvard, and the late entrance of Make Sunsets. The film is currently in festival distribution and screening tour. More at plancforcivilization.com.

Key topics with timestamps

• (02:00) The fifteen-year project — How a film begun in 2009 ended up documenting a field that took most of that time to become real

• (05:30) David Keith, Ken Caldeira, and the 1998 Aspen meeting — The origin moment that drew two researchers in by trying to prove the idea ridiculous

• (07:40) Why SCoPEx mattered symbolically — A small experiment that did the work of declaring solar geoengineering research a legitimate field

• (09:20) The Make Sunsets entrance — Move-fast culture meets a field that desperately needs trust

• (13:30) The choice to go to Sweden — What the headline missed about the actual procedural reality

• (16:30) Time and the climate conversation — Why the carbon curve has changed how Ben thinks about decisions

• (19:50) From “scare them straight” to “help them grapple” — How a fifteen-year project changed its own purpose

• (23:00) What the focus groups show — Why average viewers come away wanting more research, not less

• (28:30) The fall tour — Taking the film to states with proposed weather modification bans

• (36:00) Storytelling and fiction — Why direct air capture, ocean iron fertilization, and SAI all need better cultural artifacts

• (46:30) Slow science versus move fast — The dichotomy the field is currently navigating

• (54:30) Stardust and the trust problem — Why isotopic markers cannot do all the work

• (58:00) Who is actually in this field — Fifteen years with the researchers

Notable quotes

“This stuff is coming at us, solar geo, much, much faster than most people realize. There’s probably some major decisions that are going to start getting made in the next few years. And the public has no clue. And if they do, they think they’re talking about chemtrails.” — Ben Kalina

“When you start talking about making the planet more reflective, that’s a ‘could possibly go wrong’ thought. That’s a classic ‘why the hell would we do that.’” — Ben Kalina

“We could have the science down to a T with solar geoengineering. You could have isotopic markers on the particles you’re putting up there, and there’s still going to be a big piece of this that is just trust.” — Ben Kalina

“Nobody in this field is a climate denier. They all come to this kind of as a career last resort, after having sort of figured out that there’s nothing else where they can contribute.” — Ben Kalina

“Stardust Solutions, in this moment, being an Israeli company of former nuclear scientists funded by Silicon Valley tech billionaires whose stated goal is to change the planet’s temperature, and accidentally, by the way, name themselves after the empire’s plan to build the Death Star.” — Ben Kalina

Páll Gunnarsson is the founder of the Reykjavík Institute, an Icelandic policy organization focused on climate tipping point response. Before founding the institute, he spent over a year working on AMOC advocacy as an independent activist, a track that began when he realized no one in the Icelandic governance system was actually focused on the threat. He came to climate work after a career in software engineering.

Key topics

• (00:00) Enforced vulnerability and the moral argument against the moral hazard framing of climate interventions

• (02:35) Iceland’s National Security Council process and Páll’s New York Climate Week trip in late 2025

• (10:30) Why Iceland intuits AMOC risk: the Gulf Stream as cultural common knowledge and a governance system attuned to natural hazard preparedness

• (15:00) Iceland’s new government and the youngest cabinet acting fastest

• (22:30) The pattern Páll keeps finding himself repeating: assume someone is working on the problem, discover no one is, decide to do it yourself

• (37:00) Parallel timelines: why intervention capability research can’t wait for climate science to settle

• (44:00) Backs against the river: the moral hazard inside the moral hazard argument, and the case against using the vulnerable as motivation

• (50:19) Why multilateral SRM governance might be abandoned in favor of a coalition of the willing, and what’s coming at the September OceanEye pledging event in the EU

Notable quotes

“It’s actually fine if we overreact, but let’s react.” — Páll on Icelandic risk culture

“I have a need to be globally useful. And if I’m not working along those lines, then I become very uncomfortable.” — Páll on motivation

“It’s ethically problematic to say that we need the most vulnerable people to continue to be vulnerable in order to drive sufficient decarbonization ambition.” — Páll on the moral hazard argument

Links and resources

• Reykjavík Institute

• Páll Gunnarsson on LinkedIn

• The Iceland Minister for Environment, Energy and Climate / Johan Rockström op-ed (Iceland and Sweden, January 2026)

• Nordic Council of Ministers report: A Nordic Perspective on AMOC Tipping (February 2026)

• Páll’s SRM360 op-ed on enforced vulnerability

• EU OceanEye initiative (~€50M fund, announced March 2026; pledging event September 2026)

• Solar Geoengineering Research Governance Platform (SGRG)

This conversation is part of what I am trying to do at Inevitable & Obvious: track what is actually happening in climate stabilization at the policy and research level, not just the discourse level. Páll’s work is one of the clearer cases of a single national process opening real political space for the rest of the field, and we will be following the September OceanEye pledging event closely. Subscribe here for the full conversation, weekly episodes, and the writing that goes with them.

Subscribe now

This was the building where, on Tuesday, we hosted the first major public event dedicated to climate stabilization.

The event was called Stabilize Earth, organized with our friends at Devonian Systems as part of SF Climate Week. Over 350 people registered, and somewhere between 250 and 300 were in the room throughout the day. They were journalists, foundation leaders, policy people, climate tech founders, scientists, researchers, venture capitalists, and people looking to break into this space for the first time.

I've been arguing for months that climate stabilization deserves to be treated as a distinct field, the fourth pillar of climate response alongside mitigation, adaptation, and carbon removal. What's been thrilling in recent weeks is watching that framing take hold across the ecosystem independently, with organizations and researchers arriving at the same language on their own. Stabilize Earth was a chance to make it official.

I’ve been here before. In early 2018, before the IPCC had formally acknowledged that carbon removal was required, we held a conference called Reversapalooza. We stood up and said: carbon removal is needed, it’s possible, and here’s how we start building a field around it. At the time, “carbon removal” barely existed as a category. We didn’t have any sort of community of practice, let alone dedicated funding, or groups of people working in a way that could be called an “industry.” We helped change that.

Tuesday felt like the start of the same thing, but for a different and even more urgent set of challenges. The climate interventions field today faces many of the same obstacles carbon removal faced eight years ago: taboo, insufficient funding, the need for massive R&D, and an overwhelming sense of urgency. The night before the event, at a dinner for speakers, I said: I don’t know if you’ve all considered yourselves part of the same field, but I’m here to tell you that you are. You’re all working on things that stabilize Earth systems so we have a fighting chance with the slower solutions. So let’s draw a circle around this and acknowledge it.

The day opened with Phil Duffy, chief scientist at Spark Climate Solutions and the former climate science advisor to President Biden, who co-authored the 2023 White House report on solar radiation management. He was followed by Joshua Elliott, chief scientist at Renaissance Philanthropy, who laid out the state of tipping point risks and the energy imbalance problem with total clarity.

From there the day moved through panels, lightning talks, and conversations that covered an enormous range of work: from methane capture and destruction to rewilding the Arctic, to SAI and cloud thinning and weather modification, and of course, scalable carbon removal. The interventions panel, moderated by my partner in this work, Ben Lachman, brought together Dakota Gruener of Reflective, Ross Centers from the Planetary Sunshade Institute, Sasha Post from Outlier Projects, and Charlotte DeWald from the Arctic Stabilization Initiative, which essentially used this event to launch itself publicly.

We had lightning talks from Frost Methane Labs, Undercurrent, Spark Climate Solutions (on both atmospheric methane destruction and livestock enteric methane), Supercool Earth, Make Sunsets, Ocean Visions, OceanTherm, the Alaska Future Ecology Institute, Bright Ice Initiative, and more. Rob Jackson, one of the world’s leading climate scientists, presented on the state of greenhouse gas concentrations. On the carbon removal side, we heard from Isometric, CUR8, Mast Reforestation, GainForest, Rainbow Registry, and Devonian itself. Ryan Orbuch from Lowercarbon Capital, Ira Ehrenpreis from DBL Partners, and Johanna Wolfson from Azolla Ventures represented the investor perspective. And we closed the day with a fireside chat with Neal Stephenson, the author of Termination Shock, who joined virtually to discuss the cultural and imaginative dimensions of climate interventions.

But my favorite part of the day was seeing how excited everyone was in their lunch conversations. The tables were packed and the conversations were loud, with the room just buzzing. People from all sorts of backgrounds and experiences were talking with others who were experts in various domains of stabilization. I had the palpable sense that people had been waiting for this room to exist.

I recognize that feeling because I felt it at Reversapalooza. It’s the feeling of people discovering they’re not alone, and that the thing they’ve been working on in relative isolation is part of something larger that’s starting to take shape.

I said in my closing remarks that I see two priorities for the climate stabilization field right now. First, we need to grow the field itself, which means more people, more organizations, more coordination, and more visibility. Second, we need dramatically more philanthropic capital flowing into this space. We’re currently operating on the scale of tens of millions of dollars. We need hundreds of millions, and eventually governments need to be directing billions toward large-scale climate interventions.

We recorded the entire day, and I’ll be sharing videos as soon as they’re back from editing. In the meantime, you can see the full agenda and speaker lineup at stabilize.earth.

This was the first Stabilize Earth, and it won’t be the last. The field is just getting started, and I will have a lot more to say about what I’ve been working on behind the scenes to support it very soon.

I hope to see you at the next one!

Subscribe now

Coincidentally, on the same day that the DSG article went live, Heatmap News reported that Microsoft is pausing future carbon removal purchases. If you’re a CDR veteran like me, then your LinkedIn feed since then has been nonstop wall-to-wall think pieces about what this means for the CDR industry. And I get why! Microsoft was responsible for 90% of all CDR purchases worldwide last year. But I thought I’d share a different angle on this. Because the timing of these two articles tells you something about what carbon removal’s actual problem is, and it isn’t a lack of governance.

In 2018, my co-founders and I at Nori held a conference called Reversapalooza. The whole purpose was to convince people that pulling carbon out of the atmosphere was possible and necessary, that decarbonization alone was like slamming the brakes on a car hurtling toward a cliff when in fact you also need to turn the car around in time. We had to make that case because almost nobody in climate wanted to hear it. Carbon removal was widely seen as a dangerous distraction, a moral hazard that would let emitters off the hook. Getting investors and researchers and policymakers to take it seriously required years of work from a small community of people who believed the math was already pointing at the inescapable need to remove atmospheric greenhouse gases.

So when I read the governance-deficit version of CDR’s history, I’m reading something that doesn’t match what I lived through across nearly a decade of building this industry. And because that narrative is now shaping how people think about cooling interventions, I want to share what it looked like to me from the inside.

The legitimate governance concerns

First I want to acknowledge that I do see two fairly legitimate governance concerns in CDR’s history. The first is the Russ George ocean iron fertilization (OIF) incident in 2012, which raised real questions about oversight for open-system interventions in complex environments like the ocean. When dealing with such complex systems, there very likely needs to be more robust governance around that, or at the very least serious consideration of what the consequences would be if something goes wrong. In this case, the consequences were much more political than ecological, but the damage to the field was real.

The second is the moral hazard concern about assuming future carbon removal in climate models. Baking CDR into 1.5°C pathways before the technology existed at scale does create a risk that present-day emissions budgets are built on promises nobody can keep. Given that we are effectively in 1.5 overshoot now, that is plausible, and I take that challenge seriously.

But the governance-deficit reading of what happened after 2018 gets the story backwards in ways that matter for the conclusions being drawn.

What the IPCC moment actually was

From inside the nascent carbon removal community, the IPCC’s 2018 special report on 1.5°C looked very different than the governance-deficit narrative describes. Before that report, carbon removal was treated almost exclusively as moral hazard, a dangerous distraction from the “real work” of cutting emissions. Researchers who worked on it faced professional risk. Entrepreneurs who tried to build companies around it couldn’t get investors to take them seriously. The field existed, but barely, and under a cloud of suspicion.

The IPCC report changed that by establishing, through the most authoritative scientific process we have, that limiting warming to 1.5°C would require removing carbon from the atmosphere at scale. That conclusion went through the full gauntlet of IPCC review, and it told investors, governments, and entrepreneurs that carbon removal was a legitimate problem to work on.

Without that legitimization, the carbon removal industry as we know it wouldn’t exist. There would be no wave of venture investment, no corporate advance market commitments like Frontier, no Microsoft pledge to go carbon negative, no federal programs funding direct air capture hubs, and no ecosystem of startups and researchers and standards bodies learning in real time how to pull carbon out of the atmosphere and prove that they did it. We’d still just have ongoing academic discussions about whether we should be allowed to try.

Robert Höglund recently reframed CDR’s current phase as being about “proving and learning” rather than rapid scale. I truly wish we had rapidly scaled the industry around the fastest and cheapest methods of removal, but that’s not what happened. And since we don’t have any clear plan on how to pay for the trillion tonnes of CO2 that need to be removed, that framing is probably the most honest description of where we are: build the industry know-how and capability now so that as financial mechanisms hopefully develop over time, we’re ready to grow when the moment comes. The value of the CDR industry right now can’t be measured primarily in tonnes removed. It has to be measured in the knowledge being generated. We continue to learn which methods work best, how to measure what we’ve removed, what financing mechanisms can sustain the work, how to engage communities, and how to drive down costs.

If the implication of the governance-deficit narrative is that this process of proving and learning should have waited for more governance infrastructure before being allowed to develop, then I’d ask what, specifically, we should have been waiting for. We can’t skip that learning phase. We can’t wait until 2040 and then decide to build an industry from scratch. It will take time to march down cost curves and uncover the unknown hurdles that invariably arise when moving from the lab to the field. Every year of development compounds, and every year of delay is a year we don’t get back.

What actually happened inside CDR

The crux of the governance-deficit narrative rests on an implication that private actors outpaced oversight, leading to verification problems and erosion of trust. It invokes the structural incentive pressures of voluntary frameworks in which verifiers are paid by project developers, and suggests this produced the appearance of oversight rather than real accountability.

Those concerns are real, but in the carbon avoidance offset market, not in carbon removal. The scandals that most people associate with carbon credits (e.g. the Verra/REDD+ collapse, the South Pole implosion, and the ProPublica investigation into phantom credits from forest protection projects) were all about carbon avoidance, where you’re trying to prove that emissions would have occurred in a counterfactual scenario. That’s where the overcrediting happened, because the counterfactuals were inherently unfalsifiable and the incentives to be generous were strong.

Carbon removal credits represent a fundamentally different kind of claim. When Climeworks captures CO2 from ambient air and injects it into basalt in Iceland, the CO2 captured can be measured by a physical process. When a biochar company pyrolyzes biomass and applies the char to soil, the carbon content is testable. The measurement challenges in CDR are real (e.g. enhanced rock weathering relies partly on modeling, ocean alkalinity enhancement also relies on modeling and involves complex systems) but the failure mode is uncertainty, not fraud. I’m not aware of a single documented case of a carbon removal company fabricating or deliberately inflating its removal claims. The closest anyone has found is a methodological disagreement about how Climeworks amortizes embodied emissions in its credit accounting, and a 2009 biochar investment fraud that had nothing to do with CDR claims themselves. And methodological disagreement is actually healthy for iterating towards a better overall monitoring, reporting, and verification (MRV) ecosystem.

The culture inside the carbon removal industry has been the opposite of what the governance-deficit narrative implies. When we were designing Nori’s verification system, we spent months wrestling with the verifier incentive problem, which is the exact concern raised in DSG’s article. We wanted buyers to fund verification, because they’re the ones who benefit from knowing a credit is real. In practice, we couldn’t make that work for our initial market. We were doing soil carbon with farmers in the US, and verification costs ran $4,000 to $5,000 per project. We used a third-party system built on a US government-funded public model called DayCent for the actual measurement (I’m not running Nori anymore, so I don’t have a lawyer in my ear saying “it’s not measurement, it’s estimation!”), with simple audits on top. Our long-term vision, which we described in Nori’s white paper (found here), was to make verification as transparent and replicable as possible, essentially open-source, so that anyone could inspect and duplicate the process the way scientific results are supposed to be reproducible.

We weren’t unique in this. The entire CDR industry has been obsessively focused on proving that its removals are real. If anything, the industry has over-indexed on measurement precision in ways that structurally limit growth and chew up valuable human and financial capital. But fraud, corner-cutting, and deliberately inflated claims? That hasn’t been the CDR industry’s failure mode.

And the timeline doesn’t add up, because the Russ George incident happened in 2012, long before the CDR industry really started to grow. I didn’t enter the space until 2015, and Nori was founded in 2017. The IPCC report came out in 2018. These were different eras and largely different communities. The CDR industry, as far as I could tell, wasn’t even really aware of the Russ George ocean iron experiment, and we focused on building rigorous verification systems anyway, because if your removal claims don’t hold up, your business is dead. We genuinely wanted to solve climate change, which means you need real removals, not paper ones. And the practical realities of working with farmers and buyers and regulators forced rigor on us anyway. The governance-deficit narrative implies a reckless industry that ignored warning signs. What actually happened is that the work itself produced the rigor.

The governance-deficit narrative treats ocean iron fertilization as representative of carbon removal broadly, but OIF is a specific approach with specific risks around ecological disruption, unpredictable nutrient cycling, and uncertain permanence in a complex ocean system. Those risks legitimately warrant careful research governance before large-scale deployment, and they resemble the kinds of risks that a deployment of stratospheric aerosol injection would carry. But the bulk of the CDR industry works in domains where the risks are fundamentally different. Direct air capture doesn’t disrupt ecosystems, nor does enhanced rock weathering trigger cascading biological effects. Soil carbon sequestration doesn’t carry planetary-scale risks. The governance challenges for these approaches rest on measurement accuracy, science, and market design, not on intervening in complex natural systems with unpredictable consequences. Generalizing from OIF to all of carbon removal, and then from all of carbon removal to SAI, compresses distinctions that actually matter.

The real problem

Microsoft has been, by every measure, the carbon removal industry’s anchor buyer, responsible for 90% of all purchases last year, with 45 million tonnes contracted in total. The next-largest buyer, the Frontier coalition led by Stripe, has purchased 1.8 million tonnes. When a single company accounts for that much of an entire market’s demand, you don’t have a market so much as a dependency, and the company on which everything depends just stepped back.

I wrote about this structural fragility a year ago in “Carbon Removal’s Wrong Turn.” The industry designed itself as a monopsony (i.e. a market dominated by a single buyer class) and then built an elaborate supply-side infrastructure of standards, registries, and verification protocols to serve that buyer class. Corporate sustainability departments with limited budgets and high risk aversion became the target customer. The result was a system that prioritized credit quality and measurement precision over atmospheric impact and scaling, because that’s what those buyers needed to protect their reputations.

Carbon removal doesn’t lack governance. If anything, it is drowning in it. Registries like Puro and Isometric certify projects as having actually removed CO2 and issue the credits. Ratings agencies like Sylvera, BeZero, and Calyx Global provide ratings on both projects and measurement methodologies. For a while in the first half of the 2020s, it felt like everyone in climate tech wanted to start an MRV company. The industry has built more accountability infrastructure than almost any nascent sector I can think of, and the result hasn’t been trust and order. It’s been confusion and fragmentation that makes it even harder for new buyers to enter the market, because nobody can tell which standard to trust or which rating to believe.

And the costs to support this measurement and verification ecosystem are high on both sides! CDR providers have to fill out lengthy, detailed applications for every ratings agency and registry they want to work with, and though the content is essentially the same, every organization uses a different format, so companies end up burning enormous amounts of time on what is basically bureaucratic red tape instead of removing carbon. On the buyer side, offtake contracts have grown extraordinarily complex as every conceivable risk factor and reversal scenario gets incorporated into the legal frameworks, creating yet another barrier to participation for anyone who isn’t a major tech company with dedicated legal teams.

People inside the industry have been wrestling with these structural problems from the very beginning. At Nori, we tried to sidestep some of this complexity by only selling ex-post credits—carbon dioxide that had already been removed—backed by a warranty, rather than trying to build the entire apparatus of forward contracting and risk allocation that made other models so cumbersome. Other teams took different approaches. The governance challenges this narrative describes weren’t being ignored by an industry racing ahead recklessly. They were being actively worked on by founders and researchers navigating an impossibly difficult environment, building companies through a pandemic, inflation and interest rate increases, regulatory whiplash, and a market that isn’t really a market, while simultaneously trying to invent the measurement, verification, and market infrastructure that the field needed to function.

Meanwhile, the actual question of who pays for gigatonne-scale carbon removal beyond a handful of tech companies’ sustainability budgets has never been answered. That’s the problem we saw when we started Nori. We tried to build the financial infrastructure that made it as easy as possible for people to pay for carbon removal, initially through crypto-based mechanisms that could create broader market participation. Regulatory and practical barriers kept us from executing that vision, but we understood from the beginning that the supply side without the demand side was going to produce exactly the crisis the industry is now experiencing.

None of this is to say that the people building standards and verification infrastructure were wrong to do so. My whole argument here is that it’s a good thing that the CDR industry cared about getting verification right! The problem for the industry, though, is structural, not individual, and it’s not about missing governance. The system oriented itself around a tiny buyer class and optimized for their needs, and the result is an industry that's incredibly good at proving what it removes but isn’t sure about where to go from here.

What this means for the conversation ahead

I share the concern about private actors moving ahead of governance in the sunlight reflection space. I’ve written about the structural problems of venture-backed companies developing proprietary intellectual property for planetary-scale interventions, and I pushed back on the notion of acting now without accounting for bigger risks in a recent podcast conversation with Luke Iseman of Make Sunsets. I want robust governance for cooling technologies. I want it to be fair and inclusive and oriented toward managing real risks.

But I also want governance that prepares us for decisions we will need to make, not governance that substitutes for making them. And I want something CDR has not yet built: the operational capacity to actually scale up when the time comes.

Governance frameworks for something like stratospheric aerosol injection aren’t just about deciding whether and when to deploy. They need to address how we build international monitoring infrastructure, how we coordinate deployment logistics across nations, and how we will manage a program that will need to run for decades or centuries. The governance question and the operational question are inseparable. If funders and policymakers look at CDR and conclude that the main lesson is to build governance infrastructure before doing anything else, they risk repeating exactly what CDR did, building elaborate process while leaving the harder questions of financing, operational capacity, and practical readiness unanswered. The climate stabilization field doesn’t need to repeat that pattern, spending years building governance process while the operational capacity to actually do the work goes unbuilt. We need both to happen simultaneously.

This and every other article I publish is free because I want these ideas to reach as many people as possible. Paid subscriptions are how I keep doing this work independently. They allow me to follow the research on climate interventions and meet the researchers, practitioners, founders, and policymakers shaping how this landscape evolves.

Paid members get access to our community chat, where we discuss the latest developments in climate interventions and make sense of them together. I’m sharing all the really interesting videos, papers, stories, and other links I’m coming across in there. If you’ve found value in this newsletter, I’d appreciate your support.

Subscribe now

In this conversation, he walks through what those private meetings actually sounded like, why a landmark UN event was killed by COVID days before launch, how a Pakistani minister's first question revealed what actually drives national policy, and why the biggest gap right now isn't research or technology but the societal conversations that still aren't happening.

Chapter Timestamps

• [00:00] Opening: “We can’t talk about this publicly” — what Janos heard behind closed doors

• [02:21] What was C2G, and what does “governance” actually mean?

• [06:59] Breaking down SRM governance into manageable pieces — research, decision-making, and the question of what happens if the answer is no

• [11:44] The case against unilateral deployment — counter-geoengineering and the IPCC’s warning

• [19:13] “What planet are you coming from?” — how reception shifted from bewilderment to engagement

• [23:51] The Belgium UN event that COVID killed, and what it would have meant

• [29:19] Why politicians are afraid to go beyond the 1.5°C frame

• [35:02] How countries actually develop positions — the multi-layered feedback loops behind government policy

• [41:47] “What do Pakistani scientists say?” — the global south, Degrees Initiative, and building local capacity

• [47:57] What C2G learned about trust, impartiality, and opening doors

• [53:12] The biggest gap right now: societal conversations that aren’t happening

• [56:33] What keeps Janos awake — geopolitics, bandwidth, and the timing problem

• [1:01:42] If C2G were active today, what would it do differently?

Links and Resources

• Carnegie Climate Governance Initiative (C2G): c2g2.net

If this kind of inside view conversation where we talk about how governance actually works, what's happening behind closed doors, and what isn't, then subscribe at inevitableandobvious.com. Every episode brings you into a direct conversation with the people navigating the hardest questions in climate intervention. Next up: we're continuing to build this picture of the emerging ecosystem with more voices you won't hear anywhere else.

Subscribe now

A week or so ago my mom, who lives in Phoenix, told me the forecasted high temperature was 105 degrees, while it was still technically winter. I grew up there, and that absolutely smashes the record for the earliest hundred-degree day the city has ever seen. When I describe the the latest book on climate risks that I’m reading to my girlfriend, I can see the discomfort settle across her face in real-time. From AMOC weakening in the next decade or two, to coral reefs already crossing tipping points, to the cascading feedbacks that could start slipping beyond our ability to respond, this is indeed kind of overwhelming to think about! It all melts into this fog of compounding bad news where everything destabilizing blurs together and it becomes hard to feel anything other than paralyzed.

I’ve written a lot about that dread and the specific risks behind it. I’ve argued that carbon removal won’t scale fast enough, that a narrative vacuum around cooling interventions is being filled by conspiracy theorists, and that we need a climate stabilization framework to make sense of the full range of interventions that could keep Earth systems from crossing points of no return. If you’re new to this newsletter, those links will catch you up.

What I want to add to that picture is that I think the climate stabilization field’s approach to public engagement is missing a dimension. The work I’ve been doing on permission space is real and important, especially among the funders and policymakers who control the resources this work needs. But permission is not sufficient on its own. We also have an imagination problem, and I don’t think enough of us are taking it seriously.

When people in this field talk publicly about climate interventions, the most positive it ever gets is “this is a grim necessity." The most optimistic framing available is something like: these tools are terrifying, but if we’re careful and we build the right governance, they might be less terrifying than the alternative. I’ve more or less said the same thing myself, and it’s at least correct on the facts. But I’ve started to think it’s insufficient for what we actually need, which is sustained democratic support for some of the most expensive and ambitious engineering projects in human history. We need the kind of support that doesn’t evaporate the first time a political coalition shifts or a budget fight gets ugly.

It’s a bit cliche, but I want to draw a comparison to the Apollo program. Some people might assume the Moon landings were carried along by a wave of popular enthusiasm, but that’s a myth. The Harris Poll surveyed Americans throughout the 1960s, and a majority never thought Apollo was worth the cost. In 1967, 54% said the four billion dollar price tag wasn’t justified. 57% would have rather spent the money on water desalinization. When asked in the same year whether they’d support the program if the Russians weren’t in space, 61% said no. Apollo ran on Cold War geopolitics, not democratic enthusiasm.

And yet the cultural penetration of the space age was real and enormous. Ninety-three percent of American households watched the Apollo 11 landing. Fashion designers like André Courrèges, Paco Rabanne and Rudi Gernreich built entire fashion lines around space-age aesthetics. Star Trek, The Jetsons, 2001: A Space Odyssey, and all sorts of science fiction that imagined humanity reaching for the stars were all published then. People thought space was cool even while telling pollsters they’d rather spend the money on something else. The imagination and the poll numbers operated somewhat independently of each other.

What does that mean for climate interventions? It means we can’t rely on a geopolitical forcing function like the Cold War to carry the politics. There is no Soviet Union in this race. Which means the cultural imagination piece might matter even more than it did for Apollo, because without that captivation, without people being able to see a future they actually want to live in, the political will for sustained, expensive, multi-decade engineering programs is going to be incredibly fragile. And yes, I understand how ridiculous it sounds to talk about sustained international cooperation and multi-decade political commitment in this particular political moment. And yet we still have to figure out how to do it, because the physics doesn’t care about our election cycles.

I believe that the things we are going to have to build to stabilize this planet are some of the most extraordinary engineering challenges in human history, and almost nobody is talking about them that way.

I mean, just look at glacier and ice sheet stabilization. I’ve written about the blurry line between adaptation and geoengineering before, and ice is where that distinction breaks down most obviously. The Thwaites Glacier in West Antarctica, which glaciologists call the Doomsday Glacier because its collapse could eventually raise sea levels by over two feet on its own(!), is grounded on bedrock below sea level in a configuration that makes it inherently unstable as warm water eats away at the grounding line. Proposals to intervene involve building structures on the seafloor in some of the most hostile conditions on Earth. We’re talking about working at depth, in near-freezing water, under ice shelves, deploying a massive underwater civil engineering project to redirect warm ocean currents in the most environmentally harsh location on the planet. And this kind of thing isn’t purely hypothetical anymore, at least in the Arctic north. Groups like Real Ice are building autonomous underwater drones designed to help refreeze Arctic sea ice, and Arctic Reflections is running experiments pumping seawater onto existing ice to thicken it. These are early-stage efforts, but they’re real hardware being tested in real conditions. The engineering challenges are immense and they’re also, honestly, really cool.

Or let’s look at coral reef protection. The Australian Reef Restoration and Adaptation Program is already testing marine cloud brightening as a shading mechanism for the Great Barrier Reef, trying to figure out whether you can generate enough localized cooling to reduce bleaching stress during heat events. Other researchers at Undercurrent are experimenting with nanobubble cooling. Groups are designing physical shading structures. This is an entirely new category of ecosystem engineering, and it’s happening now because the reefs can’t wait for emissions reductions to save them, the thermal stress is already here. Scaling any of these approaches means inventing monitoring systems and deployment methods and feedback loops that don’t exist yet, combining atmospheric science and marine biology and engineering into something nobody has a name for.

And then there are the ideas that were literally science fiction. These are space-based sunshades, reflectors positioned in orbit to deflect a fraction of incoming sunlight before it reaches the atmosphere. Arthur C. Clarke and Isaac Asimov wrote about solar sails and orbital construction, not because they were imagining climate interventions (nobody was worried about warming in the 1950s) but because they understood that humanity’s relationship with space would eventually become an engineering relationship, not just an exploratory one. The Planetary Sunshade Foundation is working on this today. And one of the things that makes it so fascinating is the mass problem: you very likely can’t launch enough material from Earth to build reflectors at the scale that would matter (too expensive), which means you’d need to harvest aluminum and silicon from the Moon and manufacture components in space. As someone who has read most of Clarke and Heinlein and who devoured The Expanse series, I find it genuinely remarkable that some version of the industrial space economy those authors imagined might become necessary in my lifetime not as exploration or commerce but as planetary self-defense.

And for anyone reading this who thinks that sounds ludicrous and impossibly far from where we are today: we went from basic machine learning algorithms to a serious global debate about whether we’ve achieved artificial general intelligence (AGI) in roughly a decade. Things move fast. Exponential technological growth is not a metaphor anymore, it’s the actual trajectory we’re on, and the exponential part is really starting to kick in.

I should explain why I look at all of this and feel what I feel, because I realize that looking at a list of planetary emergencies and getting excited is not the default response. I grew up building computers with my dad, spent my teenage years reading Heinlein and Clarke and Crichton and watching Star Trek and Stargate SG-1. I built a Hackintosh (a desktop PC running a hacked version of Mac OS X) when I was sixteen and felt like I’d unlocked something fundamental about what was possible if you were willing to just figure things out. I’m an engineer at heart. When I look at the portfolio of interventions we’re going to need to stabilize the climate system, I feel something I wouldn’t expect to feel after years of staring at increasingly alarming data: I feel excited. The problems are enormous and the things we're going to have to build to address them are correspondingly wild. I think it matters that some of us feel that way and say so.

The field right now is so worried about triggering moral hazard objections that it has overcorrected into a posture where the only acceptable register is solemnity and reluctance. I even find myself doing this too. And the crazy thing is, most of us working in this space don't actually think the moral hazard concern is a strong enough reason not to pursue the research. The risks of inaction are far worse than the risks of someone misusing cooling as an excuse to keep polluting. But we still talk about it like we're apologizing for being in the room. The result is that we've collectively made it very difficult to generate enthusiasm about the work, and in doing so we've cut ourselves off from one of the most powerful forces available for building the kind of broad, durable public support that these projects are actually going to require.

And this is not techno-optimism in the lazy sense. It drives me a little bit crazy when people like Sam Altman wave their hands about how artificial superintelligence will solve climate change. Bits are not atoms. You cannot prompt-engineer your way to a glacier stabilization system. We have to actually build these things, in the real world, with real materials and real institutions and real international cooperation. The physical infrastructure for climate stabilization is going to require the kind of sustained collective effort that software startups have never had to contemplate. To paraphrase Stewart Brand, we are the planetary stewards now, whether we feel ready for it or not, and we can’t hand that responsibility off to an AI model and go get coffee.

But I am saying that we are capable of building enormously complicated things when we decide to, and that the challenges of climate stabilization represent an opportunity to rebuild something we’ve lost. People want to trust their institutions. People want to believe that large-scale collective action is possible. The cynicism of the current moment is real, but underneath it there’s a hunger for something worth doing together, something that matters, something where the engineering and the governance and the international coordination are all hard problems that we get to solve because solving them literally saves lives. When I get asked “how do you keep working on this?” the honest answer is that I love solving problems, and I can’t imagine a bigger or more interesting set of problems than figuring out how humanity stabilizes its own climate system. That includes the social and political challenges too. How do we build governance for planetary-scale interventions? How do we get institutional buy-in across countries with fundamentally different interests? Those are fascinating problems. Hard, yes. But fascinating!

We’re the generation that gets to attempt these things. I don’t know yet whether we’ll succeed, but I know that treating the attempt only as a grim obligation, and never allowing ourselves or anyone else to notice that it’s also one of the most extraordinary collective endeavors humans have ever undertaken, seems like a mistake we can’t afford. More of us should be willing to say, out loud, that this work is amazing, and the world is going to need to hear that if we’re actually going to build any of it.

This and every other article I publish is free because I want these ideas to reach as many people as possible. Paid subscriptions are how I keep doing this work independently. They allow me to follow the research on climate interventions and meet the researchers, practitioners, founders, and policymakers shaping how this landscape evolves.

Paid members get access to our community chat, where we discuss the latest developments in climate interventions and make sense of them together. I’m sharing all the really interesting videos, papers, stories, and other links I’m coming across in there. If you’ve found value in this newsletter, I’d appreciate your support.

Subscribe now