In this episode, Dakota Gruener of Reflective walks me through her organization’s new project, which maps the gaps in our scientific understanding of stratospheric aerosol injection — currently the leading candidate for directly cooling the planet. We get into what we don’t know (including a factor-of-two disagreement on basic aerosol physics), who’s already doing this without oversight, and the unsettling governance question of who controls the Earth’s thermostat once humanity has grabbed it.
(PDF transcript)
(Active transcript)
Text transcript:
David Roberts
Hello everyone. Greetings and salutations. This is Volts for March 27, 2026: “Should we block some sunlight to cool the planet?” I’m your host, David Roberts.
For decades, human beings have been running a gigantic, unintentional geoengineering experiment. The combustion of fossil fuels has pumped millions of tons of aerosols (mostly sulfur particles) into the lower atmosphere, a layer called the troposphere.
The particles don’t stay up there long — they react with water and fall as acid rain within days to weeks — but their sheer quantity has served to block quite a bit of sunlight from the surface of the Earth. Though estimates vary widely, our best understanding is that human-generated aerosol particles have suppressed around 0.5°C of warming. In other words, without them, we might already be experiencing the dreaded two degrees of warming.
As fossil fuels are replaced with clean energy in coming years, aerosol particles will quickly fall out of the troposphere and the warming they’ve been “hiding” could come back with a vengeance, portending what could be a rapid leap in global temperatures.
It’s pretty grim. But what if we put those sulfur particles into the stratosphere, the layer of atmosphere above the troposphere? There they stay longer, typically 12 to 18 months, and reflect considerably more solar radiation. We could put that sheltering blanket back in place, on purpose.
That idea — that huge amounts of short-term global warming and all its ill effects could be avoided with reasonably inexpensive technology — is what animates the field of solar radiation management (SRM).
My guest today, Dakota Gruener, runs a non-profit research organization called Reflective that aims to accelerate research into solar radiation management and ensure that it is conducted responsibly and transparently. Her group has just released a new database revealing gaps and uncertainties in the research — today we will discuss that research, those gaps, whether all of this can be done safely, who will coordinate it, and much more.
With no further ado, Dakota Gruener, welcome to Volts. Thank you so much for coming.
Dakota Gruener
Thanks for having me.
David Roberts
Dakota, let’s start. Give us a little background. What is your professional training and background, and why did you feel the need to start this organization?
Dakota Gruener
Yeah, I spent most of my career in global health and development doing vaccine policy with the Gates Foundation, then worked in the CEO’s office at Gavi, the Global Alliance for Vaccines and Immunizations, then switched gears and worked on digital privacy and digital rights.
In that work, we were doing a lot of work with UNHCR, the UN’s refugee agency, and with the government of Bangladesh, and these two projects were happening simultaneously. The work with UNHCR was sobering because it brought me face to face with how many people have already been displaced and how overstretched the international system already is with that displacement. In the work with the government of Bangladesh, we were working with them on forms of digital credentials that would work across institutional and international borders.
The Bangladeshis were, frankly, quite unusual amongst governments in saying that that was a critical requirement. Usually most governments were focused on “How do you prove who someone is within their own borders?” When we pushed them on why, what they would say behind closed doors was, “We expect potentially hundreds of millions of Bangladeshis displaced in the coming century because of warming.”
David Roberts
They’re thinking already about how to just track them?
Dakota Gruener
It was a bit of “How do we ensure that they have continued access to the services that they’re going to need?” It wasn’t even a sense of tracking them, but “We expect these people to be displaced. If you can’t prove who you are, you’re really at a loss to access necessary services.” I just couldn’t square those two conversations. We are already beyond what the limits of the international system could handle, and people are planning for significantly more displacement.
I started waking up every day thinking, “If I’m not working on climate, what am I doing with my time?”
When I shifted then into climate, what really struck me was timelines. Even if we do everything we can and we need to decarbonize with maximum effort, it’s not going to be fast enough to address this. Even if we scale CDR as fast as we possibly can, that’s also going to take a really long time.
David Roberts
That’s carbon dioxide removal.
Dakota Gruener
Yes. With each of those trajectories, the question became, what happens in the near term? Of course, there is important stuff around adaptation. But SAI kept coming up — that stratospheric aerosol injection, one of the most researched forms of sunlight reflection. What was striking there was people were discussing it as a potential tool to buy time, but we do not have the evidence to evaluate that and we are not on a trajectory to learn enough, fast enough. That, to me, felt like it was something worth pursuing.
David Roberts
You looked around. I’m sure you are aware there are all kinds of geoengineering schemes, throwing stuff in the ocean and crushing up rocks. We’ve covered a few of them on the pod. Is the reason you went for stratospheric SAI — shooting the particles up into the sky — that the one that struck you as most promising or just most near to hand, closest to being ready? Most promising. Is it still, do you think, the leader of the pack in terms of the geoengineering possibilities?
Dakota Gruener
Major scientific assessments describe three main atmospheric sunlight reflection approaches: stratospheric aerosol injection, marine cloud brightening (which goes by MCB), and cirrus cloud or mixed-phase cloud thinning (which goes by CCT or MCT), alongside more limited surface albedo modification methods and more speculative space-based concepts. Different sunlight reflection approaches fit different goals. Some of those interventions are best understood as targeted risk reduction tools. Surface albedo modification, for example, could be valuable for reducing urban heat. There is interesting work happening in Australia about marine cloud brightening to protect coral reefs.
David Roberts
Just so all the listeners are keeping up, albedo is the reflectiveness of the surface. The idea is if you replace black or dark surfaces with white or more reflective surfaces, you’ll bounce more radiation out into space.
Dakota Gruener
Exactly.
David Roberts
You can get some local, as you say, local urban cooling with this.
Dakota Gruener
Others are candidates, at least in principle, for meaningful cooling at the global scale. If your objective is global cooling, SAI — stratospheric aerosols — is the leading candidate.
David Roberts
Interesting. I want to get into the research and some of the details and what we do and don’t know. I was going to ask this question later, but I might as well ask it first since I threw this out onto social media, that I would be doing this. You will not be surprised to hear that one question, above all — the top question I get back, I’m sure you have heard it a million times and we should just discuss it right off the bat — which is the moral hazard question. This is what strikes everybody when they hear about this: we have a way to mask the effects of climate change temporarily.
Immediately people’s minds go to, “It seems if you did that it would take the pressure off people decarbonizing.” That it would give people an excuse not to decarbonize or an excuse to delay decarbonization. I’ve had these discussions before. Everybody I’ve talked to who’s involved in SAI research or SRM research generally, or really geoengineering research generally, says the same thing to this, which is, “No, no, no, it’s not a substitute for decarbonization. It’s just to buy time. We still need to decarbonize.” Everybody involved in the research says the right thing. Your website says it very straightforwardly with no caveats.
Dakota Gruener
Yep.
David Roberts
But people still worry, and what the worry is — if I’m trying to articulate concretely what the worry is — is that good-hearted researchers will get it going. But if you good-hearted researchers show that it works, then it will be taken over by national leaders that might not be so good-hearted and that might use it to substitute for or escape the need for decarbonization.
That’s what I worry about — not that you have ill motives or that anybody involved has ill motives now, but what will happen if it proves out. I’m curious how you think about the moral hazard question generally.
Dakota Gruener
Yeah, I think three points on this. The first, and maybe this is me emphasizing what you just said about good-hearted researchers in this for very genuine reasons. I care and I think we all care about decarbonization. The framing is that this is a potential complement, not a substitute. There is a genuine commitment to that across the entire field.
David Roberts
Just to make that clear, I’m sorry, this might be obvious. The reason people are worried is that this will block some heat, but it will still allow greenhouse gases to accumulate. It will still be driving ocean acidification. If the greenhouse gases continue to accumulate, the minute you stop putting these sulfur particles up, all that accumulated warming will just come back. This is literally physically not a substitute for decarbonization. All the warming is still happening. You’re just disguising it.
The analogy that came to my mind as I was thinking about it is a rent freeze — like in New York. They’re doing a rent freeze, which is fine in that it solves the immediate problem. It stops rent from getting worse. But all the structural forces that are causing rents to rise are still operating behind the scenes. The minute you take the cap off, it’s all going to come back. You’re not solving the structural problem. I know this. You know this. I think everybody involved in SAI knows this.
Dakota Gruener
Yeah. There is interesting and somewhat counterintuitive evidence of how SAI lands at an individual level. Public perception studies — and this has now been repeated multiple times — generally find that people, after learning about sunlight reflection, do not on average show a change in their support for policies to cut emissions. Rather than people saying, “Great, we can stop worrying, we’re off the hook,” they say, “If we’re talking about that, things must be worse than I realized.”
David Roberts
That makes sense. If you are not familiar with this discussion, the whole thing sounds crazy.
Dakota Gruener
Yep. That is one piece. The second piece — I found this to be a really compelling case and it is difficult because we will have no counterfactual to evaluate this against — but every climate catastrophe, floods, hurricanes, fires, requires politicians to mobilize enormous funding and political capital to respond. That is money that is diverted from the slow, steady work of decarbonizing. There is an interesting thing to explore around whether measures like sunlight reflection, by averting some of these near-term disasters, could protect the political space and resources needed for the energy transition.
David Roberts
Yes, because there is certainly an argument to be made that the stresses of climate change itself are going to make the political coordination to solve climate change more difficult as you get more disasters, etc. If you can hold some of those disasters off, maybe you create more space. It is really hard to predict these things. Very hard to say what would happen if this —
Dakota Gruener
It’s impossible, and there is no mechanism that can guarantee that this doesn’t allow fossil fuel companies to continue emitting, just as there is no mechanism that can guarantee that carbon removal won’t be used as an excuse to delay cuts. What we think we can do, though, is make the research transparent and the evidence public and ensure that there is good information that is being generated rather than just a flurry of speculation.
David Roberts
If there is any topic in the world that is ripe for misinformation and conspiracy theories, it is a handful of scientists trying to tweak the global atmosphere. You need aggressive, good communication. Along those lines, let’s talk about the database. The idea behind this is injecting these particles, doing this — there are a lot of things we don’t understand about it. Let’s get a map of what we do and don’t understand. Just explain quickly the database and how it’s constructed.
Dakota Gruener
First, on why we built this database. At present, research into sunlight reflection — research into stratospheric aerosols — is fragmented and underfunded. Different groups are chasing their own questions, and those are often really interesting questions. But there isn’t a shared view of what uncertainties really matter for real-world decisions and what it would take to reduce them.
Which means that if you’re a funder thinking about funding work in this area, you have no idea where to place your resources. If you’re a researcher saying, “How do I spend my time?” you’re maybe second-guessing how you choose to spend time. Policymakers are stuck feeling skeptical of whether this can ever deliver actionable guidance.
David Roberts
It’s worth saying up front that your organization says at the top of all this, “We are not currently advocating doing this. We don’t know yet whether it will work or not.” I think this is important for the audience to hear. You’re not out there banging the drum saying, “Do this, do this, do this.” It says on your website, “We just don’t know yet. What we want to do is find out.”
Dakota Gruener
Absolutely. We are genuinely agnostic about whether SAI should ever be deployed. If the science says it’s a bad idea, that clarity would be a good outcome, and if we have to close up shop because the science says this shouldn’t be done, we’d celebrate that. What we are not agnostic about is decisions being made in the dark. The debate often flattens into this pro versus con conversation. The challenge is that we don’t have the evidence for that pro versus con yet. What matters is how fast we can get to a place where decisions on if, when, and how are evidence-based.
David Roberts
You break up the uncertainties — these gaps — into four basic categories. There are dozens, and we can’t possibly go through all of them. Maybe you could just talk about some of the big categories to give people a sense of the things researchers are looking into.
Dakota Gruener
Absolutely. We have broken it up into a few different categories of technical uncertainties. I should stress here that we have only focused on technical uncertainties.
David Roberts
That’s important.
Dakota Gruener
This first version does not try to tackle all of the sociopolitical, big open questions surrounding all of this. For those technical uncertainties, we focused on four categories: aerosol evolution, climate response, Earth system response, and engineering. For each one of those, we have gone through the scientific literature to describe each uncertainty and to place it into a pretty simple matrix which tries to essentially rank things based on how likely we are to be wrong about this uncertainty, and how much decision making on SAI would be impacted if we were wrong about that uncertainty?
David Roberts
The significance of it.
Dakota Gruener
Exactly. There may be some things that we understand pretty well, have reasonably high confidence on, but if we’re wrong even by a little bit, that would make a big difference. There are other things where we might not understand it all that well, but it doesn’t seem like that would change any decisions.
David Roberts
Aerosol evolution, I’m guessing, is just discussion over what particles. Is that an active — is that a big discussion or...
Dakota Gruener
Yeah. One of the examples that I’ll give here, an aerosol evolution uncertainty, is around aerosol size distribution. I apologize, this is going to get wonky quickly. Aerosols — if somebody were to do SAI — would be released from a plane, and a gaseous precursor would be released from a plane, but those aerosols would oxidize and then coagulate. The resulting size of the aerosols themselves really matters because it affects how reflective it is and the lifetime, how quickly it falls out. This is something that we don’t understand all that well because while we have empirical data on aerosols and how they form after a volcanic eruption, nobody’s ever intentionally gone to the stratosphere, released material, and then said, “What happens in that setting and what does that mean for the resulting size distribution?”
And the implications of that, if you use the UK climate model, UKESM, and you use CESM, the Community Earth Systems Model out of NCAR, and in each of them you say, “We want to produce a given amount of cooling. Let’s say we want to hold temperatures flat at 1.5 degrees Celsius,” they produce pretty similar plots of what that looks like. If you’re saying we want to hold temperatures at 1.5 degrees Celsius, that is a global average temperature. You’re going to see some distribution — some places will be a little cooler than that, a little warmer than that — and these two models produce roughly similar maps of temperature. They agree on precipitation to a large extent.
Where they disagree is that they disagree by a factor of two on how much stuff you have to put up there to produce that amount of cooling. This is that aerosol size distribution. They have very different assumptions around the aerosol sizes. That has real implications.
David Roberts
That’s pretty basic. A factor of two is...
Dakota Gruener
If you put up two times as much stuff, two times as much stuff is going to come out. That may have implications for ecosystems, for human health. It certainly has implications for what would be required — the physical infrastructure that would be required to do it.
David Roberts
Or what if you inadvertently get twice the cooling you think you’re going to get?
Dakota Gruener
Exactly. The aerosol size distribution uncertainty — if you look at this database that we’ve released, you’ll see that it is far up and to the right because it is something that we think is very —
David Roberts
Very important and very uncertain.
Dakota Gruener
Exactly.
David Roberts
Got it. Climate response is pretty obvious. What are the other climate systems going to do in response to this? Earth system, I guess, is hydrology and plants and what they do. What is engineering? What is that about?
Dakota Gruener
Yeah, engineering is how would this be done? There’s an interesting thing where, despite people using the term geoengineering to describe this whole field, there has been very little engineering work done. There’s a huge number of assumptions about what is feasible. One oft-quoted and pretty obvious example: most of the scenarios that have been modeled assume that you are flying an aircraft up to 22 km and releasing material at 22 km.
David Roberts
Is that where the troposphere becomes the stratosphere? Around there?
Dakota Gruener
That is a couple kilometers above what is called the tropopause, the zone between the troposphere and the stratosphere. The tropopause is much higher at the tropics or at mid-latitudes than it is at high latitudes. These scenarios have assumed that you are releasing near the tropics at 22 kilometers, which is just over 72,000 ft. For those here who think about aviation, you will know that that is much higher than commercial aircraft fly. It is much higher than really any large payload planes that we have currently. We do not have the aircraft that would be required for the scenario that is most commonly modeled.
David Roberts
We’re talking about building new kinds of planes.
Dakota Gruener
If the goal is to deploy at the tropics, yes, you will have to do new aircraft development. What’s interesting, I just mentioned that the tropopause is much lower at the poles. It is hypothetically possible to loft material in either the northern hemisphere at high latitudes using existing aircraft.
David Roberts
Aha. But then you have to know, are they going to spread all the way down to the tropics? This gets back to Earth system response, but this is an interesting area of research to me, which is that I think people who think about climate change are used to thinking about global effects. Carbon dioxide, famously global. It affects everyone equally. But this is not necessarily global, or rather, we don’t necessarily know how global it is. In other words, if you dump a bunch of these things up by the poles, will they spread down and shelter the tropics?
How regional is this?
Dakota Gruener
How SAI is deployed matters enormously. There are stratospheric circulations that will move material around the stratosphere. What research has shown is that material injected in the tropics essentially is circulated poleward. If you put material into the northern hemisphere, that will get circulated towards the North Pole. If you put material into the southern hemisphere, that gets circulated towards the South Pole. These are pretty well understood large-scale dynamics.
There’s an interesting thing in that, first, as you note, because it moves poleward, we don’t see much crossing of material over the equator. If you put stuff into the northern hemisphere, you can cool the northern hemisphere, and if you put stuff into the southern hemisphere, you can cool the southern hemisphere. But you would need to be putting material into both hemispheres to cool the planet.
David Roberts
Got it. There’s no prospect of a single source of this affecting the entire globe. It’s probably physically impossible.
Dakota Gruener
We think it would affect the entire globe. It wouldn’t cool the entire globe.
David Roberts
I see.
Dakota Gruener
One thing that feels pretty evident looking at modeling — this is not a new topic, this is something that LBJ and his climate advisors were talking about in 1965 — there have been decades of modeling work done and pretty consistently what is shown is if you cool significantly in only one hemisphere, you end up with cooling in that one hemisphere, no cooling in the other. That can lead to really material and I would argue obviously deleterious impacts for precipitation.
David Roberts
Oh, interesting. Can you generalize, like what exactly?
Dakota Gruener
Yeah. If you cool evenly, what you see is you are not resetting the clock to zero. You’re not just turning down the temperature. Relatively even cooling doesn’t seem to lead to any particularly drastic changes in precipitation patterns. I should note that that is looking at it at the 10,000-foot level.
David Roberts
These are the big models.
Dakota Gruener
These are bigger systems models. If you cool in only one hemisphere, there’s this thing called the intertropical convergence zone. You can move the ITCZ up or down based on relative temperatures between the two hemispheres. You could imagine if rainfall shifted significantly north or significantly south, that would threaten food production for hundreds of millions of people. One of the pretty clear things is that even cooling is going to be important if this is ever pursued.
David Roberts
One thing to keep in mind is the things you’re talking about are capable of producing relatively rapid, relatively large changes in climate systems. We just know that human beings are not — we cannot adapt to those things rapidly. I think a lot of people are thinking, “How do you even test this without risking horrors?” People have trouble imagining what are the incremental steps. One of the things you’ve written lays them out really interestingly in an analogy to what they call the three-phase trial framework in pharmaceuticals.
Maybe you could just take us through that analogy and what it would look like to do phased research. What would the phases look like?
Dakota Gruener
We’d argue that SRM research needs a framework. It is unclear at this point what work can proceed under what conditions, who must be consulted. That creates this dangerous catch-22 where responsible researchers are trying to satisfy this amorphous, ever-expanding list of governance demands. Careful, small experiments that might answer really critical uncertainties aren’t moving forward. Meanwhile, the risk grows that less scrupulous actors will bypass these debates altogether.
David Roberts
We’re going to discuss some of those later.
Dakota Gruener
They’ll conduct large experiments or deployment without oversight. Medicine faced a similar dilemma. The introduction of phase clinical trials was a response to that: how do we ensure that there is a way to test incrementally that a drug is safe and effective before we roll it out?
David Roberts
Without potentially sacrificing the lives of the testees.
Dakota Gruener
Exactly. The idea is a disciplined pathway that can minimize undue risk by sequencing experiments and requiring review at each step. This helps to prevent the slippery slope concerns because there are explicit stop or continue criteria, while also producing the evidence that policymakers need to evaluate whether this is a good idea.
David Roberts
What would that first phase be like? How do you do a small experiment on the atmosphere? It’s quite large. What does that look like?
Dakota Gruener
I’d argue phase zero, preclinical, is where we are now. That’s advanced modeling, ensuring that multiple research groups around the world are asking the questions that matter to them about potential impacts, flagging things that might be a concern, and identifying what uncertainties move policy decisions. What we’ve been trying to do with the uncertainty database.
Phase one would be a small experiment designed to address the uncertainties that we were just talking about around aerosol size distribution. The idea would be to observe whether aerosols, when released from a plane, behave as models predict. How do they form and evolve in those first few days after injection?
This would require — best estimates are somewhere in the order of 10 tons of SO2. This is much smaller than something that would affect climate. It is less than 1.5% of what planes, just as part of daily aviation, emit each day. But it would be large enough that you would be able to observe all of this.
David Roberts
That first phase, you’re not even really studying impact, you’re just studying what do the aerosols do when we let them go. You can do that with a relatively small amount of aerosols?
Dakota Gruener
Yeah. Phase two would also not be at the scale of studying impacts. This would be at a scale that is too small to produce any meaningful cooling. It would be smaller than volcanic eruptions. The idea would be to track the distribution of those aerosols over months to see if the aerosols go where models predict they will.
David Roberts
Based on these stratospheric circulations.
Dakota Gruener
Yeah.
David Roberts
How do you track an aerosol particle? It’s quite small and invisible. How are these being tracked?
Dakota Gruener
Nobody is doing this yet, but there are instruments that we can use. There are satellite-based instruments that allow us to detect that there is SO2 and increased concentrations in the atmosphere. You could imagine using in situ observation — a plane with a whole instrument suite that could then fly through the plume from the lofting aircraft in order to sample and collect all of that data.
David Roberts
Phase one is just what do aerosols do immediately upon release — how do they behave? Phase two is tracking the circulation of aerosols, both of which you can do without any particular risk of having any particular effect on the climate. What is phase three?
Dakota Gruener
This is where the analogy breaks down a little bit. This is an imperfect analogy. All analogies are. We don’t at the moment see an obvious phase 3 analog in SRM. Instead, at that point the remaining open questions would be things you could only observe once you had decided you actually wanted to produce a little bit of cooling. In medicine there’s phase four, post-licensure monitoring. They’ve decided the FDA has approved a drug, it’s out there, people can be prescribed it, but there are things you might not detect until it is in a much larger population.
We think there’s a phase one, there’s a phase two, and then there might be that decision to —
David Roberts
Go or no go.
Dakota Gruener
Go or no go. At that point that’s not a “We’re going to start cooling by half a degree tomorrow.” It would be a very gradual increase in injection rate over a decade or longer and during that whole period, continued monitoring to say, “Is there anything that we didn’t expect?”
David Roberts
Because you’re talking about monitoring, which — and I think that’s so vitally important. What are the pieces of a good monitoring system? Among other questions I have is just, who’s doing this?
Dakota Gruener
Who monitors?
David Roberts
Who do all the countries of the world trust to engage in this in a transparent way? Just technologically, what do we have to have in place before we are doing good monitoring?
Dakota Gruener
On the monitoring side, there are existing satellite instruments. We have a handful of valuable instruments for in situ monitoring, but there is a significant gap in our monitoring capabilities and that is only going to get worse. One of the satellites that is used to monitor this — when I say monitor this, I mean to monitor SO2 and study atmospheric composition — is the AURA satellite. This is a microwave limb-sounding instrument. It has been a central piece of all of our SO2 monitoring now for decades.
It was launched in 2004 and was meant to have a five-year lifespan. We are well outside of that window and it will go out of commission soon. There is a very vital piece here around how do we ensure that the monitoring capabilities — and these are not just important for a scenario where SAI is deployed. These are important so that we better understand our atmosphere and the implications of warming.
David Roberts
I just wanted to interject here with a line from one of the papers you wrote that really made my eyes bug out. To give people a little context to all this stuff we have been talking about. The world currently emits 80 million tons of sulfur dioxide and that is all going into the troposphere. If a quarter of that — 20 million tons — went into the stratosphere instead of the troposphere, both the warming since the Industrial Revolution and all the future warming we expect would be balanced out. That is quite striking. If a quarter of existing sulfur emissions were redirected to the stratosphere instead of the troposphere, that wipes out all historical warming.
That is striking. I don’t even have a question. I’m just struck by it.
Dakota Gruener
If we take a step back, the Earth’s warming is dictated by an energy budget and an energy balance. There is bad news on all fronts on this. Earth is becoming less reflective.
David Roberts
Yes, it’s getting worse. I don’t know if people know this.
Dakota Gruener
It is getting worse. We usually think about climate in terms of greenhouse gases trapping more outgoing heat. That is why most of our policy has focused on what we should do on that side of the equation.
There’s an interesting thing where, counterintuitively, the Earth is also radiating more and more energy into space. Earth is becoming less reflective. The main suspected driver of that is declining cloud cover. We would expect to have less heat escape in a world with increasing greenhouse gas concentrations. But the Earth is actually radiating more energy into space.
Outgoing longwave radiation has gone up significantly since 2000. Multiple factors could contribute. Here too, declining cloud cover is potentially a really important contributor. We’re in this position where we’re messing with some of these systems in ways that we don’t understand.
David Roberts
We are geoengineering.
Dakota Gruener
We are certainly geoengineering.
David Roberts
Everybody should understand that.
Dakota Gruener
The challenge is we’ve now seen the significant decline in reflectivity that has real implications for warming. These acronyms are stacking up. I apologize, but we use the term Earth’s energy imbalance (EEI) to talk about that gap between energy coming in and energy going out. Since 2001, there’s been a net increase in energy imbalance of about 1 watt per meter squared. That’s nearly twice the increase predicted by climate models.
David Roberts
Yes.
Dakota Gruener
It’s equivalent to increasing atmospheric CO2 by 86 parts per million, which is a huge amount.
David Roberts
Yes. This is what’s going on right now — the Earth system is responding more dramatically to greenhouse gases than the models say. This is the ongoing mystery in climate science: why are we seeing effects that seem more severe than what we would expect from the models? What you’re saying is that reflectivity is part of this? Cloud cover might be part of this.
Dakota Gruener
Yeah. The net of all of it is what you just led with, which is somewhere on the order of a quarter of today’s annual SO2 emissions, if delivered into the stratosphere, would be in the range needed to counterbalance this.
David Roberts
I know people get this already, but just as a footnote, when we say balance out, we don’t mean wipe out the effects of — you still have ocean acidification, etc. It’s just the heat of it. Although the heat of it is not a small thing.
Dakota Gruener
Yeah, it would not cancel out greenhouse gases, it would not restore a pre-industrial climate. This is admittedly — and I don’t want to say proudly — but this is clearly a band-aid.
David Roberts
Right. I think a lot of people are just intuitively thinking, when we put these particles up here and call them pollution, they are doing all sorts of bad things. Pollution has bad effects. Throwing sulfur up into the air has bad effects. It causes ozone holes and respiratory problems. People think of pollution negatively. If you are talking about deliberately throwing a bunch of pollution up there, people’s minds automatically go to, “Will this not make all that worse? Will this not have all the ill effects that pollution has?”
Will this affect our respiratory health, and will it affect plants? If you are measurably reducing the solar radiation reaching the surface, are you affecting the performance of photosynthesis? Are you affecting the performance of solar panels? What are the possible negative effects of doing this?
Dakota Gruener
Vis-a-vis health, I’ll start there. The tradeoff that people are thinking about — and this is a reductive initial way to try and think about some of the health implications — is if you add a small amount of sulfate to the stratosphere, what will be the effects on respiratory health when that precipitates out versus how will that same amount of sulfur reduce the mortality associated with extreme heat events?
David Roberts
What a macabre cost-benefit analysis.
Dakota Gruener
Absolutely. The hard thing in thinking about the risks and impacts of SAI — and this is true for any sunlight reflection method, and probably true for many of the tradeoffs that we’re considering around climate policy — is that you have to be thinking about this kind of risk versus risk analysis.
David Roberts
There’s no non-risk.
Dakota Gruener
There’s no non-risk. There’s the risks of doing something. There’s the risks of not doing something and having continued unabated warming. On that particular measure, there’s been research showing that for every one additional death that you would see due to those respiratory effects from the sulfate, you would save 100 lives due to heat-related mortality.
David Roberts
Interesting. That’s pretty — maybe not precise, but pretty clear, at least directionally.
Dakota Gruener
Directionally. There’s much more work required to dig into that further. Another thing that people talk about is ozone and potential interactions with ozone-depleting chemicals. Sulfate aerosols do interact with stratospheric ozone chemistry. That’s a known concern. One of the things that people have worried about is we’ve put a lot of work into trying to course-correct from the ozone hole. If this in any way caused a backslide, that would obviously not be good. There are a few interesting pieces here. One is that we’re on track for a super recovery of the ozone hole.
David Roberts
What does that mean?
Dakota Gruener
That we’re going to have more ozone than we did. With that, the sulfate particles would interact with these ozone-depleting chemicals because we have done a really good job of arresting the use of ozone-depleting chemicals. If somebody ever decides to deploy, when that happens, there will be different concentrations of those chemicals in the stratosphere. There are important implications to when deployment might start — and hence what those potential interactions would be. In some of the research, it suggests that because we are on track for a super recovery, all it would do is delay that by a couple years.
David Roberts
Interesting. If I can crudely summarize, maybe we can spare a little ozone.
Dakota Gruener
I want to be careful about saying that, but it is a concern that many researchers feel is overblown.
David Roberts
Interesting.
Dakota Gruener
It is also a reason that alternative particles get discussed, like calcite.
David Roberts
We skipped past that, but I would just ask that now. Is there even the prospect of us being able to custom design particles to do the things we want them to do and not to do the things you don’t want them to do? Is that even a thing?
Dakota Gruener
We focus on sulfates because it has a natural analog. Volcanic eruptions are not a perfect proxy for what an SAI deployment would be, but they do provide real-world constraints on what would happen if you were to release sulfate aerosols. What would those climate effects be? There is work ongoing to explore other natural aerosols. There has been research around alumina and calcite, etc.
There is a group, Stardust, which is a private company. They have not released any science, so it is hard to dig into this. They have been trying to do essentially what you were describing — can we engineer a particle to meet a certain set of desired criteria?
David Roberts
I’m guessing you, as just a nonprofit research organization, are agnostic to this. If a better particle comes along, you’re fine.
Dakota Gruener
Yes, our sense is sulfates make sense to study as a Gen1. There’s finite time and finite resources available until we think decisions about this could be on the table. Given that, prioritization and focus is a necessity. We’d rather understand something well than poorly understand lots of things. Because sulfate has this natural analog, we understand it relatively well. It is the thing that is most commonly discussed. It’s likely to be on that agenda and there’s a pretty clear critical path to producing the decision-relevant evidence. More novel particles may eventually matter, but they do add new unknowns in terms of chemistry or feasibility and monitoring.
David Roberts
Then you have to do your whole phased research all over again.
Dakota Gruener
All over again, yeah.
David Roberts
Just to return to the question before we move past it, would this affect photosynthesis? Are we talking about blocking enough solar radiation to measurably affect photosynthesis or to measurably affect the performance of solar panels?
Dakota Gruener
There’s really interesting evidence on plants and photosynthesis. SAI diffuses some sunlight, it reflects some and it diffuses it. In large-scale crop models, that leads to better crop yields, not less — plants like that more diffuse light.
David Roberts
Interesting.
Dakota Gruener
Counterintuitive, but generally the sense is, you’re in an environment where there is more CO2 and diffuse sunlight.
David Roberts
We understand that dynamic well enough to say with some confidence that this might boost rather than restrain photosynthesis.
Dakota Gruener
That is a gross oversimplification — all of these are. Different plants respond to different things. You’ll see regional differences in the amount of cooling. You’ll see regional differences in precipitation. Those in aggregate need to be studied. We’re at a point in the research where we know big, broad-strokes things, but those specific questions really matter.
David Roberts
Solar panels — has anybody done the research on that or tried to estimate?
Dakota Gruener
I would need to go pull up more research on this, but the Earth naturally reflects about 30% of incoming sunlight. Even in the proposals where people are saying “let’s cool pretty significantly,” they are talking about an increase to reflecting 31% of incoming sunlight. The net effect on solar panels will be modest.
David Roberts
As you say, the officialdom is moving slowly on this stuff. There’s not a structured research program that’s been given a legitimate thumbs up from all the legitimate organizations. Consequently, there’s some cowboy activity happening around the edges.
There’s this private company that’s popped up called Make Sunsets, where you can now literally send them — I think it’s $10 or whatever — and they’ll send a little balloon of particles up for you. That’s ridiculous, but also it’s sinister, I think.
What do you think our collective response should be to those types of things? My instincts, my priors are that we should go overboard to shut that down and to try to channel this into some common, central, transparent, shared thing. Am I overreacting? How do you think about that?
Dakota Gruener
I don’t think you’re overreacting. There are two companies active in the space right now. Make Sunsets, as you just referred to, is one of them. Stardust is another. I would treat them as quite distinct. Make Sunsets is, as you say, you can go and give them $10 and they will release a small weather balloon with a nominal amount of SO2. It is not an amount that will cool the planet — far, far, far from it. They’re also not doing any science. You’re not learning anything in that process.
It is neither improving the quality of our information nor producing cooling. I don’t see a pathway for it to evolve to support either of those goals.
David Roberts
Maybe they would say they’re just doing a stunt to push everybody else to take this more seriously.
Dakota Gruener
I think that’s true. They would say, “The time is nigh and everyone is being too precious about all of this,” and that they are trying to...
David Roberts
Nonetheless, I have the urge to tell them to go to their room.
Dakota Gruener
As do I, as do I.
David Roberts
Go spend some time in your room and think about what you are doing.
Dakota Gruener
There is significant pushback against that very rogue kind of cowboy actor.
David Roberts
I’m curious what you think about Stardust. They seem a little more — at least they’re scientific.
Dakota Gruener
Yeah. They have a different scale of funding. They have a different level of seriousness, and they might be doing good science. Given the pedigrees of the folks on their team, that feels plausible.
David Roberts
I think it’s supposed to be coming out with stuff next year. There’s supposed to be a big release of their early...
Dakota Gruener
Yes. It is impossible to know because they are not releasing any science publicly. In a field where public trust is fragile, that lack of transparency poses a risk not only to them, but to the entire field and to efforts to responsibly learn about sunlight reflection.
David Roberts
It really does seem, especially in this area, you just want to overcorrect in the direction of transparency in these early stages. If you want this to go anywhere, you really need to scrupulously be open about it. That’s my instinct.
Dakota Gruener
I could not agree more. For us, we publish all of our funders on our website. We are publishing all of the research that’s done. We put out the uncertainty database. We’ve released the simulator. All of them are designed to lay bare what we know, what we don’t know, what we think it will take to get to the point of informed decisions. Who’s guiding those decisions, who’s guiding our thinking. The transparency is absolutely critical. There’s another piece, which is with Stardust, the profit motive itself is structurally problematic, not because companies are inherently bad.
David Roberts
Well…
Dakota Gruener
We can have that debate later. Even in a scenario where research exists entirely in the public sector, at some point, if there is a decision to deploy, there will be companies involved. Somebody will have to develop those aircraft, for example.
David Roberts
Yeah.
Dakota Gruener
But in the absence of any independent authority with the capacity to evaluate claims, there is a real risk that Stardust or any other private actors overstate their particle safety or efficacy to cut a deal. That could lead to very real consequences. Going back to the medical analogy, there is no FDA equivalent.
David Roberts
Yeah.
Dakota Gruener
There’s a good reason the FDA exists. Without that, the incentive structure is dangerous.
David Roberts
I want to address what I feel is one of the knottiest and most difficult aspects of all this, which is the who decides, who’s in charge kind of thing. There’s recently a call by governments for a solar geoengineering non-use agreement, and ministers from 54 African nations signed this. There is already a lot of sentiment in the global south that, “Here they go again, the global north presuming to take charge of the temperature of the globe,” without consultation, doing things that affect our countries without asking us.
This is obviously a whole nest of difficult issues about colonialism, etc. This is where my brain breaks on this stuff. How do you even think about trying to get the degree of international cooperation you would need to really have this be fully transparent, fully common, something we’re all involved in? We can’t agree on anything, Dakota. Look at the US right now. Even assuming we get a better regime in the US, how do you get everybody on the same page like that? How do you think about that challenge?
Dakota Gruener
I want to be precise about this because the framing matters. The non-use calls are primarily driven by European NGOs. They are not coming from African scientists, African research institutions, or necessarily originating with African governments. In fact, the opposite is happening. After an article was run suggesting that this was really being driven by researchers and policymakers on the continent, over 30 African climate scientists signed a public statement saying, “Excluding African scientists from global research conversations would be far more colonial.”
David Roberts
We have European NGOs using African countries as tokens or shields for their own ideological preferences. That certainly does not strike me as outlandish.
Dakota Gruener
The point is research needs to be inclusive and decision making needs to be inclusive. There is really important work ongoing to make that real.
David Roberts
What does that look like though? Is it a framework? Is it an organization? Is it a list of principles?
Dakota Gruener
Some of the things that are happening at the moment: there’s a recently launched African Climate Intervention Research Hub that’s stewarding this within the continent. There’s an organization called the Degrees Initiative that we work closely with and love, that funds researchers across the global south to study SRM. We ourselves have funded researchers in six countries outside of the US and soon to be many more. There is a lot of work ongoing to ensure that the evidence base and the people producing the evidence base aren’t exclusively northern.
That doesn’t get you to the point of who is making the decision, but it does improve the quality and the inclusivity of the evidence base itself. In terms of where these decisions should be taken, we see a future in which intergovernmental or multilateral processes take up responsibility for SAI research and deployment decisions as the scenario likeliest to lead to good outcomes.
David Roberts
Do you mean by that the UN or are you deliberately not specifying the UN?
Dakota Gruener
The existing institutions that people look to, whether it is the UN or the Montreal Protocol, I don’t think any of these are credible or desirable answers. They are not designed to sponsor experimentation, nor do they have the mandate or capacity for ongoing management of something that is ethically and technically complex. The question becomes what are models that might be appropriate? We can draw from other sectors because something new will need to be built.
David Roberts
One of the things that climate heads are probably thinking right now is whatever you do, don’t put it in a context where you need unanimity among all the world’s hundreds of countries to do anything, because that’s what —
Dakota Gruener
We’re not in a world where unanimity happens on anything.
David Roberts
That’s what we did for climate change. Consequently, the UN did nothing for decades on end because you can’t get unanimity. There’s no flat answer to this, but you’re on this sort of tradeoff: you’re never going to get total unanimity to do this. What counts as enough unanimity to pull a trigger on something that affects the entire globe?
Dakota Gruener
It is valuable to think about an incremental approach. One of the challenges that you see is there’s this chicken and egg around solving the governance challenges and trying to understand what we are actually governing. There have been attempts to preordain what the governance structure should look like to govern scale deployment. The problem is that might be designed for something that, quickly, the science rules out.
We think about it more as an incremental pathway where you’re developing governance and doing the research in parallel and in lockstep. With that, I could imagine essentially a coalition of the willing — a group of countries. It’s important that it’s countries who pool resources in order to fund research. Part of pooling those resources is then saying, “We’re going to openly debate the priorities and we’re going to work together to define those endpoint criteria for experiments.” In that you get a group of countries who are cooperating in a very open and transparent way before consensus or without consensus.
David Roberts
It would be nice to have that kind of rational organization addressing a variety of problems globally.
Dakota Gruener
I hear you and I don’t think this is unrealistic. The world that I come from — I mentioned earlier, I started my career in vaccine policy and worked for an organization called GAVI. GAVI was launched in 2000 to address fragmented funding and legitimacy gaps in terms of vaccine delivery. It was designed in a way that I think is pretty instructive. No one thinks of it as a governance organization per se, and yet that’s precisely what it is. Their board transparently wrestles with amazingly complex and very consequential questions about “Do we support Japanese encephalitis or do we support malaria?” That has real implications for whether kids are being protected and where kids are being protected.
Their management of their programs is adaptive. This isn’t the set-it-and-forget-it model of a treaty. There’s a secretariat with real technical depth, ensures that the decisions are scientifically grounded. It’s a focal point that discourages duplication and makes cooperative governance easier. That’s important in SRM, because without some type of pooled research fund, the field is likely to splinter into national research programs that may have competing priorities or methodologies.
David Roberts
Beyond that, nations who decide to go for it.
Dakota Gruener
That increases the odds of unilateral or poorly coordinated action. If you have this pooled research alliance, that creates a forum for those tradeoffs to be surfaced and debated in something that is inherently multilateral from the get-go. Probably the last thing to know about GAVI is it at least apocryphally emerged over a dinner. It was not the product of long UN deliberations. It was, “This is a thing that we think should exist. Let’s go create it.”
David Roberts
The last question is about the termination shock, since that seems like a good closing place. One of the sci-fi-esque fears about this is once you start doing it, you’re stuck doing it. You have to keep doing it or else the warming comes back all of a sudden. In addition to thinking about this multilateral incremental way of edging up to it, somebody’s also eventually going to have to think about multilateral incremental ways of undoing it at the end. Presumably this will buy us time. We’ll decarbonize and we’ll need less cooling, and then we’ll ramp down this sulfur injection. That’s the vision, right?. We ramp it up and then we ramp it back down. But then you get into who decides what the right temperature of the world is. It’s way too big a question to throw here, at the very end. But once your collective hand is on the thermometer...
Dakota Gruener
Yeah.
David Roberts
Then you own it. You break it, you own it. Then you own the temperature of the Earth. Some regions would benefit more than others from more or less heat. Once you are already doing it, have you thought through how to deal with those governance challenges or how to decide what the right target is?
Dakota Gruener
Termination shock is a legitimate concern. It’s one of the strongest arguments for doing this research carefully. Maybe not so much in the scenario that you’re talking about — we’ve managed to see the progress we want on decarbonization, we’ve seen CDR scale to the point where it’s meaningful, and we are now carefully considering how and when to start ramping down. The concerns around termination shock are often in the vein of, you start, you deploy at scale, and then you learn something you didn’t want to learn and there’s a very abrupt stop, or it introduces a geopolitical risk that causes an abrupt stop.
On the first of those — you learn something and then there is good reason to stop, and that leads to rapid warming — that’s part of why this clinical trials model, and particularly that phase four, is so important. You’d want to ramp up very slowly and monitor so that if things diverge from predictions at any point in that ramp up, you’d still be at a scale where you could ramp back down pretty easily and without termination effects. In terms of the geopolitics, that is an argument for why multilateralism matters so much. I will just keep banging on about that.
In terms of ramping down, careful, intentional action is needed. I’d like to think that if you have a group who is weighing the really complicated tradeoffs and decisions about how much cooling to do, it is just a continuation of that same set of decision-making criteria.
David Roberts
But that’s a lot of trust you’re asking for there. A lot. “Let us set the Earth’s temperature.” That’s putting a lot of trust in that group.
Dakota Gruener
That’s why it needs to be well-designed.
David Roberts
All right. Dakota, this has been super fascinating and I look forward to seeing this. I hope it moves forward in as multilateral and rational a way as you propose. I’ll just say that.
Dakota Gruener
Thank you so much for having me.
David Roberts
Thank you for listening to Volts. It takes a village to make this podcast work. Shout out, especially, to my super producer, Kyle McDonald, who makes me and my guests sound smart every week. And it is all supported entirely by listeners like you. So, if you value conversations like this, please consider joining our community of paid subscribers at volts.wtf. Or, leaving a nice review, or telling a friend about Volts. Or all three. Thanks so much, and I’ll see you next time.
