Your independent source for Harvard news since 1898 | SUBSCRIBE

Your independent source for Harvard news since 1898


Daniel Schrag and David Keith: Can Solar Geoengineering Help Fight Climate Change?


Daniel Schrag and David Keith headshots over an orange background

Professors David Keith and Daniel Schrag

Professors David Keith and Daniel Schrag



Climate changes now underway are occurring at a rate 100 times more rapid than at any time in Earth’s geological history. But in human terms, at this early stage of climate impacts such as rising seas, rampant wildfires, and intensifying storms, the pace of change has been perceived as slow. In this episode, Hooper professor of geology and professor of environmental science and engineering Daniel Schrag, and Gordon McKay professor of applied physics David Keith discuss the nature of the climate change crisis, the choices and tradeoffs in transitioning to a carbon-free economy, and the risks associated with engineered global approaches to controlling Earth’s thermostat.


Transcript from the interview (the following was prepared by a machine algorithm, and may not perfectly reflect the audio file of the interview):


Jonathan Shaw: Climate change may be the hardest problem the human race has ever confronted. In a single century, humans have set in motion events that will unfold on a geological timescale, ultimately redrawing coastlines around the globe as ice sheets melt and sea level rises. What to do about the warming is dominated by uncertainties. Can humanity agree to meet its energy needs with renewables such as wind and solar power to run the world’s economies on carbon-free energy? If so, how quickly could the transition be made? Is there a threshold beyond which the effects of greenhouse gases will become irreversible? Can solar geoengineering, the lofting of reflective particles into the stratosphere, help stop this runaway train? Welcome to the Harvard Magazine podcast, “Ask a Harvard Professor.” Joining us today are two guests expert in these questions. Daniel Schrag is director of the Harvard University Center for the Environment, Sturgis Hooper professor of geology and professor of environmental science and engineering and an authority on what ancient climates can tell us about the future of our earth. He is the recipient of a MacArthur Fellowship and of the James B. Macelwane Medal from the American Geophysical Union. From 2009 to 2017, he served on President Obama’s Council of Advisors on Science and Technology. At the Harvard Kennedy School, he is co-director of the Science, Technology and Public Policy Program. David Keith is Gordon McKay professor of applied physics in the School of Engineering and Applied Sciences, and professor of public policy at the Harvard Kennedy School. He is the faculty director of Harvard’s Solar Geoengineering Research Program, and the founder of Carbon Engineering, a company developing technology to capture carbon directly from ambient air. For more than two decades, he has been a leading thinker on ways to improve the effectiveness of solar geoengineering and to reduce its risks. Welcome to you both.

Daniel Schrag: Thanks, Jon.

David Keith: Thank you. Great to be here.

Jonathan Shaw: Dan, my first question is for you. What does paleoclimate data suggest lies in store for us at current CO2 concentrations, which reached a seasonal peak of 417 parts per million in May 2020?

Daniel Schrag: Well, I think there are many lessons from the paleoclimate record. But I think it’s important for everybody to understand the difference between what we see in the geologic past with warmer climates and higher CO2 levels and what we’re experiencing today. In the past, when we’ve had times with higher carbon dioxide levels and warmer temperatures, the changes have been very slow. What that means is that ecological communities have had time to migrate, to adapt. And the slow timescale means that what we’re seeing in the geologic past is kind of the equilibrium climate, when things have adjusted on slow timescales to reach a kind of steady state. What we’re doing is roughly 100 times faster than the natural pace of climate change over the last, say, few hundred thousand years. The Ice Age cycles that come and go on tens of thousands of years, we’re doing it on century timescales. And so what that means is, we have a sense of the direction of the change, but there’s a lot of uncertainty about how fast things are going to change. A great example would be the ice sheets in Greenland and Antarctica. We’re pretty sure at this point every summer now, the Greenland ice sheet melts much, much more than new snow forms in the wintertime. So it’s losing mass every summer. And the net over the course of the year raises sea level by something like about a millimeter of sea level rise per year.

It’s already very likely that it’s past the point of no return. That is, if we go to still higher CO2 levels, 450, 500, 550, essentially, we’re going to melt the entire Greenland ice sheet. That’s something like seven meters of sea level rise overall, that’s a huge amount. But what we don’t know is how fast it’s going to happen. When we look back at the geologic past, we can see times where CO2 was just a little bit higher and there was no ice on Greenland. And that tells us that the Greenland ice sheet is fundamentally unstable. But it doesn’t tell us how long it takes to melt it, given the experiment we’re doing to the climate today.

And so, without intervention, things look very bad. We’ve started in motion a series of changes in the system, and the challenge is those changes all have very long timescales, from centuries to thousands of years. And figuring out how to control those systems that are so massive and have such long timescales is an enormous challenge.

Jonathan Shaw: This is a question for both of you. First, David, how difficult would it be, from a purely technological point of view, to meet all the world’s power needs with renewable energy? And how long would that conversion take?

David Keith: First of all, I want to push back on something you said in the introduction, which is that this is the hardest problem humanity’s ever faced. I forget your precise words. As somebody who—I worked on climate my entire career now, 30 years, but I’ve paid attention to other public policy problems, and I just don’t think that’s true. I think the problem of managing nuclear weapons in a divided world, I think the problem of managing biological weapons, perhaps, there are other problems that are probably sharper when it comes to consequences for humanity and sort of the intense difficulty of getting to agreements. So I do see climate as the most important global-scale environmental problem. But not all problems have to be the absolute most important problem. Not all knobs need to be turned to 11 for things to be worth solving.

Daniel Schrag: Let me just comment on that, because again, this is part of the fun. What I say, David, sometimes, and you may disagree with this, but I say that many people claim that climate change is the most urgent problem we face. And to me, it’s backwards, actually. Climate change is one of the most difficult problems we face precisely because it’s never the most urgent problem. There’s always something more urgent, which makes climate change so difficult. Dealing with a problem that’s inherently long timescales is painful and very, very difficult. 

David Keith: Because of this central fact to know about climate, which is that climate change is in some way proportional to cumulative emissions, that has all these profound consequences. Among them, the fact that even among them, the fact that even once we bring emissions to zero, we haven’t in any way solved the problem, we’ve just stopped the problem getting worse. But the other part of that cumulative nature is it is, in fact, rationally true that in a particular year, it doesn’t really make any difference what our emissions are. It’s this cumulative emissions. And that means that as much as we feel like we want to make it a crisis in order to force political action, in some objective sense, no year will it be a crisis, even though overall it has this terrible environmental impact.

Daniel Schrag: That’s right.

David Keith: And human impact.

Daniel Schrag: That’s right. Which is, again, I think some people ... Certainly the environmental groups want to make this an urgent problem. And the frustrating thing about this problem is, it’s important. It’s always easy to worry about something that’s right in front of you that’s going to have a short timescale response.

David Keith: Actually, the question of how quickly we could cut emissions, I think you said, what is the technical possibility for how quickly we cut emissions? Is that basically how you phrased it?

Jonathan Shaw: Yes. From a technological point of view, how long would it take to meet all the world’s power needs with renewable energy?

David Keith: So I think that problem is simply ill-posed. I mean, the short, ugly answer is you could do it in 10 minutes with a big war where you just shut everything down. So presumably, what you mean is do it in a way that keeps something happening. But then you have to define what are the necessary services. Unless you do that, the question is ill-posed. If you mean do it with no impact at all on the economy, no slowing of the economy, then the answer might be infinity. It’s extremely hard to transition. So that answer does not have any technocratic closed form answer. These are inherently trade-offs.

So my view is that it’s possible for us to decarbonize in just a few decades, technically possible to do that and still keep the fundamentals of modern civilization going. Still keep transportation and communication networks happening, still keep people fed. But if you want to do that in a couple decades, you need a level of command and control that’s akin to wartime where you simply nationalize a bunch of industries, shut down a bunch of things you don’t want. Take over a bunch of siting decisions in a way that’s very different from what we’re familiar with in democracies.

And I think with decisions like that, I think you can do in just a few decades, but with enormous social and actually other environmental consequences. So, the bottom line is the answer to that question is entirely dependent on one’s political assumptions and on questions about what the side impacts are, both social and economically, you’re willing to tolerate.

Jonathan Shaw: And Dan, why has decarbonizing economies the world over proven so difficult to effect from a political point of view?

Daniel Schrag: Well, I actually think until recently, decarbonization was relatively expensive. It required either foregoing inexpensive fossil fuels or replacing them with technologies that were pretty much all much more expensive. And it’s really just in the last 5 to 10 years that we’ve seen, for example, solar and wind emerge as one of the least expensive forms of electricity generation, at least, at low penetration where you don’t have to worry so much about storage or the intermittency of the electricity. That’s extraordinary. That’s like a new world all of a sudden.

My prediction is that in the next decade or two, we will be building solar and wind like crazy. This year in the U.S., 75% of new electricity generating capacity was wind and solar. That’s incredibly good news. Now, the bad news is that the total amount we’re building is something on the order of 20 gigawatts. And 20 gigawatts is not nearly fast enough. Just to put it in sort of round numbers, the total US electricity consumption is something like 500 gigawatt years. And if we were to electrify transportation, and electrify home heating, and electrify industry, we’d probably double or even triple that electricity demand. But let’s assume that we’re very, very big on conservation and we only double our electricity demand. That means we would need something like 1,000 gigawatt years of electricity in the near future.

A thousand gigawatt years if we were to do it all with wind and solar, and forget the problem of intermittency. Let’s imagine that it somehow matched perfectly and we could just ignore the times when there’s no wind or no sun. That’s a minimum of roughly 4,000 gigawatts of wind and solar, because the capacity factor between wind and solar average is going to be something like 25%. So you just to do that 20 gigawatts a year, 4,000 gigawatts. And that’s the minimum required. You probably need many times that. You’re talking about at current rates something like 200 years of time.

Now, I’m not saying it’s going to require 200 years. I think David’s exactly right. It’s certainly possible to accelerate it. But what is not just technically feasible, but foreseeable, I think it’s likely to be much longer than a few decades, unfortunately. I’d love to be wrong about this. I just don’t see that people are going to be willing to make the kind of sacrifices that like World War II required. We’re talking about real sacrifices, financial sacrifices, giving up health care, giving up education, things that people care about. And not only do we have to do it here in the US, but it has to be done globally. And so when I look at that, I say, “Boy, I’d love to imagine that it could happen in 20 or 30 years. But it doesn’t seem very likely to me.”

David Keith: Yeah. Just to be clear, I think it’s extraordinarily unlikely that it happens in 20 years and I actually don’t think I would vote for it. I don’t think it would be ethical. Doing it that fast, this is a trade-off about the benefits to future generations of cutting CO2 emissions very sharply, which are huge against other environmental damage of doing this too quickly and social damage. You can’t do this in 20 or 30 years in a way that maintains things we care about, from health care to human rights. If you’re going to do it that fast, you’re going to tear up a bunch of rights and stop doing a bunch of other things we care about.

David Keith: So I personally don’t think that trade-off would be ethical, but I don’t think it’s technologically impossible. I think we ought to be very careful about saying that and defining carefully. I think the key point I’m trying to make is these are human choices, political choices, and we shouldn’t imagine there’s some technocratic, turn the crank value-free answer. There’s no value-free answer for how quickly we should cut emissions. There’s no value-free threshold or what is the maximum amount. 

Daniel Schrag: I think maybe one way to frame it, because I think something David said it’s really important to pay attention to just when you think about wind and solar and building something that fast. Forget the Green New Deal, it’s talking about 10 years, let’s talk about 20 or 30 years. What David said is exactly right, that doing something on that timescale means forget permitting, you’re just going to build stuff. And people don’t like that.

David Keith: Let’s talk about how this actually would work. Yeah. How this would actually work. So let’s say that you want to do it on wind and solar. So, in the near term in the U.S., you do a lot of solar, because the solar capacity factors are much higher, and you do that associated with a huge amount of long-distance transmission. Right now, the time to permit big, long distance transmission lines is basically infinity, we can’t do it. So you’d have to start this, if you imagine Congress really wanting to do this in 15, 20 years, your first law is the national clean power permitting law. And it says that local people get to complain, they get to argue, but they get like three weeks to do it. And then there’s a final binding decision and it’s backed up by troops. Good farmers don’t like it, they get pushed out of the way. It has to be that way if you’re going to go that fast. If you’re going to allow every individual local person to fight every installation all the way through the court system, you’ll never do it.

Daniel Schrag: Another nice example for people who live in the northeast is, why is it that we don’t have a fast train? Using 1970s French TGV technology, we should be able to go from New York to Boston in about 45 minutes. Why does it take three and a half hours? And that’s on the fast train. And the answer is actually very simple. The answer is called Connecticut. The tracks through Connecticut on the coast of New London, trains go 20 miles an hour. So, what would it take to actually have high speed rail to get rid of airplanes that are flying short distances from New York to Boston to Washington DC?

The answer is we could do it with existing technology. It just would require seizing people’s homes and building straight tracks. And the governor of Connecticut is the only person who has the authority to do that. Why would the governor of Connecticut seize people’s homes in Connecticut so that people could go from Boston to New York very quickly? Very hard to imagine in our current political system.

David Keith: Maybe this is a useful back and forth. I mean, I don’t think it’s going to happen in 20, 30 years. But I also think that in order to get any substantive, fast-enough-to-satisfy-me-or-Dan progress, we will need some new political reality. And the answer is, in the end, the governor of Connecticut can be overruled federally, just to take that example. I think we will get to some level of federal consensus and power around decarbonization that will be slower than a kind of 20-year pace that I said, but quicker than the 200-year pace Dan said, but it will require some political discontinuity between now and then. I don’t mean a revolution. But I mean, something that really moves this towards the center of the political agenda. And I think that is actually quite possible, but it’s very hard to guess when it will happen.

Daniel Schrag: I think a useful way to think about this, and I agree, David, and I wasn’t trying to say that we were going to do it in 200 years, I was just saying that the current pace if you just sort of have to scale it out, it’s on a multi-hundred-year timescale and that’s an important number to know. I would sort of see two end member pathways and the reality of the world is likely to be somewhere in between. One would be a kind of free market timescale where wind and solar get cheap and so they get built. And gradually, they replace old coal and natural gas eventually. Batteries keep getting cheaper as we install them.

That’s a kind of technological progress and transition, maybe with a little bit of guidance from government, but mostly letting the market and letting technology do the work. And people buy electric cars because they’re better choices, not because the government tells you have to, or because the government subsidizes them heavily. The other end member is more what David was talking about, that sort of command control. A little bit what like London has done with electric taxis. There will be no more diesel taxis in London in the next few years. They’re all getting phased out. That’s one way to do it. And you could imagine the problem with that being ultimately in a democracy, people revolt.

Imagine if you said in New York City that in the next 10 years we’re going to phase out all gasoline powered taxi cabs, and the price of taxis are going to rise by a factor of three but you’re going to all have electric vehicles. I suspect that a lot of New Yorkers would erupt. And whoever made that law, whether it was the governor of New York or the mayor of New York City, would get thrown out of office. That’s the challenge. But I think somewhere in that range is where we’re likely to see. And my hope is that technology can be wind in our sails. That is, right now some of the new technology is looking like it’s going to make this problem a lot easier. But there are still big obstacles ahead.

Jonathan Shaw: Great. Thank you both. David, what is solar geoengineering? And how might that be used to mitigate global warming?

David Keith: Solar geoengineering is the idea that humans might deliberately alter the amount of heat the earth absorbs from the sun. That might be done by putting reflective aerosols—these are just tiny little particles of dust—into the upper atmosphere, the stratosphere, maybe 20 kilometers above our heads. But it could in principle, also be done by some shields in outer space or by making some cirrus clouds thinner or by making some low-level clouds a little lighter, or maybe even by painting roofs or changing the way crops are planted. There’s a variety of ways humans could deliberately alter what we call the radiative forcing, the amount which the Earth’s energy is out of balance, is caused to be out of balance by the CO2 that we put in the atmosphere. And the net effect of this would be to somewhat reduce the climate changes—changes in rainfall, changes in temperature—that come from the accumulated amount of long-lived greenhouse gases like carbon dioxide. So the goal would be to reduce the climate risk for a given amount of carbon dioxide in the atmosphere.

Jonathan Shaw: And David, do you imagine solar geoengineering being used as part of a larger decarbonization project?

David Keith: I have no idea what will happen. And I think people’s track record of being able to predict what will happen is just terrible. But personally, I only think it would make sense and I would only support using solar geoengineering in combination with deep emissions cuts and the ability to remove CO2 from the atmosphere in some mode where it’s used to reduce the peak, the climate damage during the peak of CO2 concentrations.

Jonathan Shaw: Dan, what kind of global governance structures would that kind of a geoengineering project require?

Daniel Schrag: Well, I think it’s a broader question than that, Jon. I think that climate change itself is going to ultimately push us into new forms of global governance. The top-down Kyoto style treaty wasn’t terribly effective. And so we moved to a very loose Paris agreement where countries offer their own commitments and set their own targets with maybe a little bit of peer pressure but certainly nobody’s telling anybody what to do. There aren’t negotiations, per se, the way there were part of the Kyoto process.

Ultimately, as climate change gets worse and worse, and I’m talking about 20, 30, 40 years from now, I think that managing the impacts and managing the decarbonization process is going to require some innovations in the form of interactions between countries in ways that are hard to imagine right now, but I think they’re inevitable. And I think solar geoengineering is an important part of that. I wouldn’t see it in isolation, I would see it as part of a broader framework for dealing with a global scale problem. In the same way that the world invented, in some ways, with the rise of terrorism, post 9/11, the world reinvented a variety of structures that have helped manage terrorism. I think the same is going to be true with the threat of climate change.

I think some kind of innovation in how nations deal with this is going to be essential. I can’t imagine just one or two superpowers implementing a solar geoengineering scheme that affects every living thing on the planet without any kind of consultation or buy-in. I just don’t think that’s very stable and that just wouldn’t be very smart. I guess I can imagine a couple of countries still doing it. But I think that wouldn’t be a very good outcome, because in the long run, I suspect that’s not very stable. And we’re talking about, even if you were to do it in concert with carbon removal that would ultimately reverse the problem, you still need to do it for a very long time on any kind of political timescale. David, what do you think about that?

David Keith: I agree. I mean, new technologies and new interconnections require different kinds of global governance. And it’s happening. I mean, we have global governance imperfectly that manages imperfectly, infectious diseases and nuclear weapons and the internet and trade. And these things actually work, have enforcement powers. We have an Outer Space Treaty, to take one really interesting example, which applies in principle to solar geoengineering because it’s very general, which had the power to make Russia pay Canada back when they dropped a nuclear reactor literally, on Canada in the early 80s.

So I think there are lots of examples like that where global governance has grown. It’s not a single top-down thing, it certainly isn’t just what happens to be in some overarching treaty like the Kyoto Protocol. There are a lot of underlying governance mechanisms, and we need them and we need to grow them faster in order to manage all sorts of problems of which climate is just one. For solar geoengineering, I see kind of two tracks. So one track is the, you might call it the UN system process, the IPCC, all these meetings, all this conversation about something that points to a global consensus.

I think that conversation is actually very important, because it’s a way for people to share information, and for nations to begin to see what their interests and divergent interests are. I think there’s virtually no chance that such a thing will produce a clear answer to implement or not implement, although, I think it might produce some interesting enabling or framework treaties that might reduce the chance of really ill-thought out, ill-considered unilateral action.

But I think that if solar geoengineering does get on the international agenda for implementation, it’ll be because a small set of countries likely not one, force it on the agenda by basically pushing towards deployment. By essentially saying, it’s in our fundamental right for preserving us against environmental risks, that we will move forward towards this technology, and that will precipitate the discussion about what actually does happen. I’m not saying that that’s good or bad, but I think it’s likely that is how it plays out.

Daniel Schrag: I think it’s a really interesting. There’s a quote from H. G. Wells that I like from The Time Machine where he wrote, “We are kept keen on the grindstone of pain and necessity.” And that sounds a little harsh, but I honestly think we’re in the early stages of experiencing the impacts of climate change. And I think the real question that we don’t understand is how the political will to do something about climate change will change as people become more and more frightened about the impacts. I think the fires in California and Australia and Greece are just a little bit of a wake-up call. I mean, that’s just to use a bad phrase, the tip of the iceberg.

I think it’s really difficult for us today to imagine the political context of the international discussion of climate change 30 years from now. When, literally, many weeks in the summer, it may be too hot to go outside in India to do any kind of physical labor or just walk down the street. It’s hard to imagine the kind of enabling capacity that gives to political discussions. And I think our perception of the global politics will change entirely. But again, maybe that’ll be wrong, but I think that’s possible.

David Keith: No, I think that’s very likely. I think there’s ways in which we’re still in a bit of a phony war on climate, where there’s lots of talk about how it’s a top priority. And many national leaders find it important to keep part of their political constituency happy to articulate how important climate is. But when you measure it by actual actions, it’s clear that it’s really not at the top of the political agenda, essentially anywhere. But it will get to the top of the political agenda, and it will get to it as these impacts get sharper. And it will be later than it should be, but I think that’s the way it will play out.

Daniel Schrag: Right. That’s exactly right. I mean, imagine now when the G7 get together, and they come out with some statement on climate, which I guess President Trump didn’t want to join, but imagine they actually did all come out with some statement on climate. It’s a posturing right now. But when they actually get together to talk about a terrorist attack in Paris, or in New York, or whatever, they actually talk with their top military advisors. And they are serious about, what are they going to implement? How are they going to coordinate to actually solve this problem? And you could imagine at some point when the discussions on climate change between world leaders have that sense of urgency, the same as you could imagine after 9/11, that’s when we’re going to see real change.

Jonathan Shaw: So you may have answered my next question already, but I’ll ask it anyway and see what you say. What is the risk that the fruits of solar geoengineering research might be used to lower Earth’s temperature without addressing the underlying cause of the warming? In other words, what is the risk that the fossil fuel industry or even private citizens who are vested in combustion engines, in everything from cars to home furnaces to lawnmowers will use the potential technology as cover to keep pumping CO2 into the atmosphere?

David Keith: You asked two completely different questions. One is what is the risk that some people will attempt to exploit this as cover? I think the answer is that risk is 100%. I think it’s certain that, I mean, it is a profound truth that humans act in their individual self-interest most of the time. So it is a no-brainer that some countries, oil companies or whatever, will over-claim at some point about how solar geoengineering might work as a way to avoid emissions cuts. I am certain of that.

But another question you asked is, what is the chance that it will mean that there’s sort of no action to cut emissions? I think the chance of that is actually also zero since we are, after all, already taking action to cut emissions. We’re spending something like $300 billion a year globally on clean energy now, which is really fundamentally motivated by CO2, most of all, and so we are not as quickly as I want or Dan wants, we are cutting emissions. So to me, reality is in between those two rails.

The question is, to what extent will solar-geo result in a little bit slower emission cuts than would have happened in a world without solar-geo? And the even sharper question is, to what extent it will result in slower emission cuts than it should have been in a world without solar-geo that was doing the right thing according to some ethical trade-offs? Because of course, it’s actually rational and ethical to do some risk trade-off.

So, if solar geoengineering really reduces some of the risks, long run risks of CO2, and if your choice about how quick to cut emissions is a trade-off between current costs to emissions cuts and long run risk of CO2, if solar geoengineering reduces that risk a little bit, then it’s actually rational and ethical to cut a little bit more slowly, although, you still have to cut in the end to zero that emissions.

Daniel Schrag: I step back from it. I agree with what David says. I think in the real world, again, as the impacts of climate change become more and more apparent to people, as their suffering gets greater, I expect political will to grow. And it doesn’t usually grow linearly, it grows in a kind of tipping point way. But as, suddenly, political leaders around the world are called upon to act because people are scared. And when people are scared, the political demand for action is intense. But here’s the problem, again, the timescale of the climate system is such that it’s as [Heinz professor of environmental policy] John Holdren said many years ago, “We’re driving a supertanker, we’re not driving a sports car.” And so, you just can’t turn around, you just can’t stop very quickly. And so the short timescale of solar geoengineering is one aspect of it that I suspect it makes it incredibly attractive to some political leader in the future. When the political will that demands action is so high, it’s actually something that can happen on short timescales. And I think that’s going to make it very attractive. That’s my suspicion. I wonder what David thinks about that.

David Keith: I think it’s likely it’ll be very attractive at some point. But I think it’s really hard to see. I’ve had lots of interviewers ask me if I think solar geoengineering is inevitable, and the answer is certainly not. I can easily imagine situations where it’s avoided. I think it’s just extraordinarily hard to guess how this will play out.

Jonathan Shaw: David, what are some of the scientific risks of solar geoengineering that you’ve been thinking about over the last couple of decades? And how can those be minimized?

David Keith: So first of all the risks are ... This isn’t like talking about the risks of some thing that’s a given, either the risk of a volcano or the risks of some thing that humans are doing, because solar geoengineering is fundamentally an engineering choice. There are a bunch of different ways to do it. And there are certainly ways to do it that would have very high risks that would be just crazy, wantonly destructive. So there’s no way to separate out discussion about what the risks are from discussion about what engineering choices are made in implementing solar geoengineering?

But I think that if you’re asking about, I think a good question is to say, what would be the risks of solar geoengineering implemented to try and provide global benefits and even global modification of what we call radiative forcing. And I think the risks of that, that seem like a given would be some risk of the aerosols themselves. Aerosol air pollution now kills millions of people a year, globally, and we’re talking about adding more aerosols to the atmosphere. And while there are lots of reasons why those risks might be small, they’re not going to be zero. And you’re definitely going to be perturbing the atmospheric chemistry, perturbing the distribution of aerosols in a way that will have consequences.

Another risk is the fact that solar geoengineering is not anti-CO2. It’s altering the climate in ways that may reduce in most places most climate changes we care about, but it will not reduce them in every place. In some places it will certainly increase some climate changes compared to pre-industrial with consequences. That’s very general. I mean, there’s a whole host of individual things, air pollution, ozone, stratospheric ozone loss, acidic rain, changes in the stratospheric circulation that change storm tracks. You can go on and on. Changes to direct versus diffuse radiation, solar radiation, which have consequences for plant growth. There’s a big list of physical risks like that.

Jonathan Shaw: What about the risk of termination shock? Maybe you would consider that to be a political risk. But it could be a scientific risk, if something were discovered about the sulfate aerosols that had been lofted into the atmosphere, that there was some deleterious effect and you wanted to stop the engineering project.

David Keith: I personally do not see that as a risk in the same categories as others. So first of all, it’s certainly true that we will discover new surprises and new bad outcomes. And that may cause people to change how much they’re doing or to transition from one kind of solar geoengineering to another. But I think the risk of very sudden turn off of large-scale solar geoengineering is pretty low because of individual country level self-interest. Even countries that initially opposed deployment of solar geoengineering have a very strong self interest in maintaining the ability to start it once deployed because of the risks of sudden termination. And sudden termination effectively requires unanimity, global unanimity among countries of significant scale, in shutting it off. And I think that’s a very unlikely outcome.

Daniel Schrag: I think David’s right about this, that the termination shock, first of all, I think I’m not so convinced that we would ever put sulfate aerosols up there, at this point, given what we know. But let’s imagine we put something else. Suppose we put calcium carbonate up there. And then we discovered that there was some effects of calcium carbonate that we didn’t like. The timescale, and this is why I actually think the way you do it is really important, some people have advocated low level cloud modification. And I have always said that’s a terrible idea because the timescale is too short.

Daniel Schrag: What we want to do is be able to, if we’re going to do anything, it should be on a timescale that gives us a year, two years, three years timescale that allows us to adjust without too much fine control. Sort of like a steady hand on the tiller on a very large ship. Let’s imagine that we put calcium carbonate up in the air, up in the stratosphere, and then found that it was a problem. If we had to scramble and find out something that we needed as an alternative in the next year, we could probably do that.

I think we would certainly do that rather than have the termination of geoengineering altogether if it was something that really there was real urgency to do because the impacts of climate change were intolerable. So, I think David’s right. I think that termination shock is not, for me, one of the biggest fears ahead.

David Keith: It’s not clear to me if it’s a feature or a bug, you wouldn’t want the solar geoengineering technology you couldn’t turn off. You want it to be adjustable with a timescale of probably a few years. So, it’s a very odd one. I mean, there are people who hypothesize that what if we just somehow forget how to do this technology? Or what if there’s a global war? So it’s important to say that the timescale we need to do this for if you’re doing it as part of a strategy with emissions cuts and carbon removal is a century or two. We’ve now run transatlantic telecommunications for well over a century, over two world wars. We’ve run electricity systems in the midst of war for decades or in some cases through centuries. I think that given how cheap this is, I think if people want to maintain it, it’ll be maintained in anything short of an absolutely apocalyptic global war. And under such a circumstance, the temperature shock from termination will be tiny compared to the effects of that war anyway, so it just doesn’t matter. So, I actually think there are a lot of terrible things about solar geoengineering, but I don’t see this one is really all that important.

Jonathan Shaw: My last question, for both of you. Since natural mechanisms of removing CO2 from the atmosphere operate slowly, is there a way to remove carbon from the atmosphere at scale? Or do you imagine that new innovation might be needed to meet that need?

David Keith: Well, there’s no question you can remove CO2 from the atmosphere at scale. There’s also no question that we don’t have mechanisms to do it now that are really cost effective and have low environmental impacts and can be scaled to what we need. But because this is a large-scale kind of industrial project, and would happen over a century, there will of course be innovation, there’s no way you do it always with the same technology. I should say I have a self-interest in being involved in helping to start one company working on one technology.

My personal view is that if you really are talking about large scale removal, one of the technologies that never is popular, but I would put a lot of effort into is adding alkalinity to the ocean. Essentially, the CO2 is a weak acid. And if you add alkaline to the ocean, you tend to push the ocean pH back towards pre-industrial and you tend to permanently remove the CO2, so it’s dissolved in the ocean water in a way that’s safe. I think that that technology is one that requires real work and may not turn out to be useful. But I think there’s huge room for innovative effort funded by governments in the public interest.

Daniel Schrag: I agree that governments should be funding research. Because if someone did develop a way of doing this at a much, much lower cost, it would be incredibly consequential. And so it’s worth funding a lot of different ideas at this stage because we just don’t know at this point. That being said, barring a radical breakthrough in cost, I think there’s a growing consensus, at least on the range of what this is likely to cost. Maybe it’ll be wrong, but I think there is a convergence of views with some range, but I think we have some sense of what that’s likely to be.

I’ll tell you what my worry is about this. It’s that humans are very good at adaptation. We often in the climate community underestimate human ingenuity when it comes to adapting to difficult situations. For example, the literature’s full of papers about how climate change is going to reduce agricultural output, because they take a climate model, look at a relationship between crops and temperature, and run the climate model and project out 100 years. And what they never do is think about what farmers are going to do in response. They do the same thing with malaria, with all sorts of problems.

It sometimes feels a little bit like fear mongering. And again, these are well-intentioned scientists who are doing this. But to me, I actually have a lot of trust in humans’ ability to deal with very difficult situations that happen to them locally. We’re really bad at collective action problems. We’re tribal, we’re nationalistic, and we’re really bad at long time scale problems. But if your home is under attack, you’re going to figure out. Hopefully, that doesn’t mean people won’t suffer. But it does mean that people are very good at adapting to difficult circumstances.

My worry is that, let’s imagine we stabilize CO2 in the atmosphere and reduce emissions to zero, and solar geoengineering is part of the mix. The question is, at that point, will people’s willingness to pay for carbon removal persist, even as they become used to the current climate circumstances? And that’s a question that I just don’t know. But I worry that humans’ ability to kind of get used to their surroundings means their willingness to sacrifice to return to something that has much lower carbon dioxide is going to be a challenge. I’d love to hear what David thinks about that.

David Keith: One thing is, I think this problem ... I think that threats to humans are the most important problem, but at least as I personally see it, they’re not the only problem. I think there’s some value in leaving the natural world that we evolved out of, that we inherited, leaving as much of it as we can for future generations to love, to serve as an anchor for their civilizations. And I think that this is not just for my mind about managing human risks. I think one of the things that actually excites me about solar geoengineering is the possibility of really reducing climate risks in places where adaptation isn’t relevant, in natural systems, in ways that we could not otherwise do.

I think that Dan’s question about willingness to pay once the kind of acute human risk is reduced, I think is very real. But it’s an interesting trade-off because it’s also true that societies get more rich, they’re often more willing to pay for things beyond just the most necessary and pay to protect the natural world in more idealistic ways. And we see this in all sorts of measurable ways. And the technology cost gets cheaper. So when you play this out in these simple integrated assessment models that just allow you to kind of have these knobs, the dollars on carbon looks very high in 2,100 or 2,150 in these models. But the economy is very big, and richer people tend to be willing to pay more to protect things they love, as you see in natural parks and other natural protection. So, I’m actually kind of cautiously optimistic that people will pay to do this, because at that point will be a pretty tiny fraction of an economy and an economy where people care more about the natural world as sort of a fraction of their total utility.

Daniel Schrag: Unlike the solar geoengineering situation, the carbon removal problem has a huge collective action component of it and a huge free rider component of it. So how willing are people to pay for something if there are a few rogue countries that are continuing to put CO2 into the atmosphere? It’s going to be hard. I think it’s going to be really difficult.

David Keith: Yep. It will be.

Jonathan Shaw: Thank you both for joining us today. Really appreciate it.

Daniel Schrag: Thanks, Jon.

David Keith: Thanks a lot, Jon.

This episode of Ask a Harvard Professor was hosted by Jonathan Shaw and the season is  produced by Jacob Sweet and Niko Yaitanes. Our theme music was created by Louis Weeks. This third season is sponsored by the Harvard University Employees Credit Union and supported by voluntary donations from listeners like you. To support the podcast, visit If you enjoyed this episode, please consider rating and reviewing us on Apple Podcasts. Contact us with questions at


You Might Also Like:

Jeannie Suk Gersen headshot over an orange background.

Jeannie Suk Gersen

Photograph by Nina Subin

Ask a Harvard Professor with Jeannie Suk Gersen

Danielle Allen headshot over an orange background.

Danielle Allen

Photograph by Laura Rose

Ask a Harvard Professor with Danielle Allen

You Might Also Like:

Jeannie Suk Gersen headshot over an orange background.

Jeannie Suk Gersen

Photograph by Nina Subin

Ask a Harvard Professor with Jeannie Suk Gersen

Danielle Allen headshot over an orange background.

Danielle Allen

Photograph by Laura Rose

Ask a Harvard Professor with Danielle Allen