Scientists Focused on Geoengineering Challenge the Inevitability of Multi-Millennial Global Warming

UPDATED Sept. 5 | I encourage anyone interested in climate change science and policy to explore the rich discussion below about geoengineering, in this case mainly focused on managing incoming solar radiation to counter CO2-driven global warming — particularly in the context of the long (and building) commitment to warming already baked into the climate system. I have long rejected the utility of a focus on this response to global warming because I still can’t imagine a scenario in which a single actor would initiate it or, in contrast, a global consensus could be reached on its deployment. (In many past posts, I’ve asked, “Who gets to set the global thermostat?”)

But I’ve been shifting my thinking based on recent conversations with some of the authors below about the feasibility of incremental management of sun-blocking aerosols, even as warming aerosols contributing to conventional air pollution are reduced. See the postscript. I still see scant prospects for action, but the conversation is vital. (A good separate starting point is this National Research Council report: “Climate Intervention: Reflecting Sunlight to Cool Earth.”)

The context of this chat was a paper published early this year on the long commitment to climate change, as explored on Dot Earth. This exchange was prompted by this tweet by Gernot Wagner of Harvard:

Oliver Morton, author of “The Planet Remade: How Geoengineering Could Change the World

You were kind enough to ask on Twitter what I thought, in the context of geoengineering, of the Clark et al. Nature Climate Change perspective on deep-time consequences of 21st century climate policy and the Dot Earth post to which it led. I think the paper is fascinating but also somewhat frustrating, so I thought I’d take the opportunity your inquiry provided to think through some of the issues in rather more detail than twitter allows. (I’m cc’ing various people who have a stake in this, including some authors on the NCC paper who I have come across in the past….)

I agree with the authors that taking a view measured in millennia is an appropriate, and underappreciated, part of assessing the impacts of anthropogenic climate change and the need for climate action; I also agree with their conclusion (and starting point) that a zero-net-emissions human world is a desirable target. And it is inspiring to see such progress being made in the detail with which models of ice sheet dynamics and other forms of change can be applied to the moderately far future.

But I have a problem with the way in which, while discussing some plausible — if far from inevitable — forms of technological change, the authors choose to ignore others. They repeatedly describe change on the timescales they are looking at as “irreversible”. There is a potent sense in which this is true: once a change has happened it cannot be made not to have happened; once they have risen, neither carbon-dioxide concentrations nor sea levels can be made not to have risen. But that’s not what most people mean by “irreversible”: the term normally means pretty straightforwardly “can’t be reversed”. And at a couple of points in their paper the authors make it clear that that is indeed the sense in which they mean it.

Specifically, they say: “The implication is that, in the absence of efficient, large-scale capture and storage of airborne carbon (emphasis mine), carbon emissions that have already occurred or will occur in the near future result in a commitment to climate change that will be irreversible on timescales of centuries to millennia and longer.” Which is to say: in the absence of technologies to reverse this, it is irreversible.

Well yes. And in the absence of a reverse gear, a car is irreversible, too. That’s why cars have reverse gears.

That may read as a flippant trivialization. But I think the authors of the NCC paper trivialize the issue, too; they just do it more subtly, through neglect. The sentence I just quoted implies pretty strongly that, in the presence of efficient (or for that matter inefficient) large-scale capture and storage of airborne carbon, carbon emissions that have already occurred or will occur in the near future might not result in a commitment to climate change that is irreversible on timescales of centuries to millennia and longer. That’s quite an important statement. But it receives almost no follow-up at all in the subsequent six pages. A paper that makes use of the concept of the Anthropocene — a concept predicated on the idea that human activity is a dominant factor in the state of the earth system — does not spend any time at all looking at what humans might try to do, or be able to do, about the problems it discusses over the periods it imagines.

It is easy to understand why not. This is a paper by natural scientists, and the natural sciences give you no way of understanding what people will attempt to do in the time-scales under discussion. (Nor does anything else, which is one reason Rob Socolow’s notion of “Destiny studies” is, as you say on Dot Earth, a welcome one.) But rather than discuss the impossibility of predicting or modeling human decisions or capabilities, the authors choose not to address the issue at all, and this undercuts the seriousness of their undertaking. When there is a huge source of uncertainty in your analysis you should be explicit about it. You should try to assess the limits it puts on the salience of your results and what might be done about those limits. Instead the authors choose to proceed as though humans will do nothing other than what they are doing today, and make no effort to justify their decision to privilege that singular scenario. That is what I mean when I say I think they trivialise the issue.

I know that there are not yet any negative emissions technologies up to the task. But the possibility of such things is under active discussion. Indeed negative emissions are already being incorporated into the sort of integrated assessment models that inform discussion like those of COP 21 in Paris. Those scenarios typically demonstrate a lack of specificity about the costs and potentially enormous impacts of such technologies, which makes their ready and convenient acceptance of a currently hypothetical capability disturbing. But to ignore the potential capability completely does not redress that problem. The NCC paper makes great play of taking a long view, putting itself forward as a necessary and clear-sighted corrective to analyses that content themselves with looking out only 85 years. To do so without making any attempt to examine the role negative emissions might play in coming centuries demands a pretty explicit justification.

Maybe they cannot be developed in time. In his contribution to Dot Earth on the subject, Ray Pierrehumbert (one of the authors of the NCC paper, and someone I know a bit and admire a great deal) says “it’s true that given a thousand years or so — if technological civilization survives — it becomes likely that we could develop ways to remove carbon dioxide from the atmosphere.” If that is indeed the timescale appropriate to this discussion then maybe it really does not matter all that much. But why on Earth should we accept that such technologies are a thousand years away? After all, Ray’s analysis says that for technological civilization to be likely to survive there has to be a decisive global shift to new energy technologies within the current century. If the mix of energy technologies cheap, powerful and acceptable enough to bring this shift about includes one or more of solar, nuclear fusion or nuclear fission (and who, seriously, thinks it won’t?) then energy scarcity in subsequent centuries seems unlikely. And with ample energy on a global scale is it really likely that airborne-carbon reduction technologies will not be available for another millennium or so?

(Incidentally, I think it’s worth noting an apparent asymmetry in Ray’s approach to geoengineering technologies. He has said in various places that he finds it hard to imagine albedo-enhancement being carried out consistently for a millennium or more; but he finds it easy to imagine airborne-carbon reduction consistently not being tried for similar periods.)

Ray is at least specific about a time scale for thinking about such things. His NCC co-authors are not. They say towards the end of their paper that it is necessary to research technologies for removing carbon from the atmosphere, but they do not say what effect, over what timeframes, the success of such research might be expected to have on their analysis. This is not because of an overall lack of confidence in technological prognostication. For example they say that a technological change which they equate to a “fourth industrial revolution” will “inevitably lead to the decarbonization of current energy systems”.

I have no idea where that “inevitably” comes from, but it certainly demonstrates confidence. They make other broad assertions, too, such as seeing this fourth industrial revolution as an occasion for optimism about its “opportunities for… positive change” on the basis that the first three “created new jobs, new wealth and shifted power structures”. They leave undiscussed the fact that the “shifted power structures” of the first industrial revolution led to, among other things, the subjugation and impoverishment of many non-European nations and peoples, or that the “new jobs” created by the second included tank commander, torpedo officer and air force bombardier, all parts of a transformation which introduced mechanized killing on a wholly unprecedented scale. (And though it is a less important point, I doubt all the authors of the NCC paper are entirely happy about the distribution of the third revolution’s copious “new wealth”.)

I don’t want to be taken as being against industrial revolutions (though in fairness I’m not quite sure what it would be to take such a stance) and I am all for optimism, properly tempered. But I am against double standards. I think the authors’ keenness to wrap up the “inevitability” of one set of profound technological changes in Davos-friendly boosterism while not giving any serious discussion to other relevant and complementary technologies makes the paper less than it could have been.

An alternative would have been additional runs of the same models with carbon-dioxide reduction assumptions built in, looking at varying rates, varying start times, and different levels of intermittency. This at least would expose the limits of what such technology might or might not be able to achieve. More runs with more assumptions means more computer time and more effort, and I know it’s all too easy to call for such things when the time and effort are other people’s. I can also see that a result which could be interpreted as “Oh, if we start doing 2ppm a year in carbon-dioxide reduction in 2100 everything except RCP8.5 looks more or less OK, ice-cap-wise” would carry risks, just as using unspecified negative-emissions technologies in integrated assessment models does.

But I would rather that such a result (if that would indeed be the result — I obviously don’t know) prompted a renewed emphasis in subsequent discussion on just how hard large-scale removal of airborne carbon is, and on how foolish efforts to that end would be if a fossil-fuel-free system is not put in place first, than not see such results at all. Similarly, if an a priori more plausible 0.2 ppm a year does more or less nothing to the long-term picture I want to know that, too. And in either scenario, I want to know what the models can say about the hysteresis in the system: how much ice lost at higher carbon-dioxide levels does not come back at lower ones. I might even want to know if albedo modification could be used to overcome those problems of hysteresis, and if so how much of it would be needed for how long.

I cannot bring such analysis about. But it seems to me that an impressive body of scientists such as the authors of the NCC perspective could go some way towards it, and I would urge them to do so. Giving people a sense of the scale and consequence of current human interference in the earth system, as this paper seeks to do, is an important and profound task, and I welcome anyone’s willingness to take it on. But if you want to illuminate the earth system’s possible Anthropocene futures, you have to deal with the profound uncertainty that unknowable human capacities and intentions on an earth system scale impose on the picture. They are its very essence.

The earth system is not just a human artefact. The Earth is not a car to which a reverse gear can simply be added. But the earth system is under the sway of human influence, intentional and otherwise. Using the tools of natural science to analyze the earth system’s future while not addressing the question of what human technologies might be available and what effect they might have seems to me so incomplete an approach as to end up misleading.

Oliver Morton
Senior Editor, Essays and Briefings
The Economist


Raymond Pierrehumbert, Halley Professor of Physics, Oxford

The respect Oliver expresses for me is certainly reciprocated, but I think Oliver has more faith in the predictability of technological progress than I do. The reason I agree with my co-authors on the notion of irreversibility is that carbon dioxide removal technology does not exist in a form that is deployable at acceptable cost today, and the current technology might not be deployable at scale at any cost. Technological progress can be amazing, but it rarely proceeds along predictable paths. Fifty years ago, few would have envisioned the great advances in computing power we’ve seen, but many thought controlled fusion power was practically around the corner — a state it is still stuck in, except that many are now less optimistic the problems will ever be solved. Energy technology is one area where progress has been frustratingly slow, though (as Andy has often noted) that may mainly reflect under-investment.

If somebody demonstrates CDR technology that is deployable at scale within the next twenty years, that would be a game changer, but that is not the state we are in now. Decisions about energy systems need to be made now, and we do not know if CDR will proceed along the path of controlled fusion or the path of the Internet, so it would be highly imprudent to proceed on the basis that CDR will eventually come along in time to save us, and that it will bring down CO2 levels swiftly enough to avoid severe climate damage.

One of the appropriate reactions to the perspective in our article should indeed be to invest massively (and wisely) in CDR technology. Oliver and I have very different views on the dangers of albedo modification (at least in the next millennium or so); I consider it more a form of geo-vandalism than “geo-poetry” as some of Oliver’s fans like to call it. But if we are ever driven into the desperate circumstances of needing to rely on albedo hacking, surely it would be better to have some form of CDR deployable first, so as to provide an exit strategy that avoids the dangers of a millennial commitment.

— Ray


Ken Caldeira, Carnegie Institution

Working on solar geoengineering and carbon dioxide removal today is overall a good thing except insofar as actions taken today to reduce emissions reflect an expectation that these technologies should and will be deployed at large scale at some point in the future.

Risk aversiveness suggests that we should be trying to purchase these insurance policies through R&D at the same time that we are trying to reduce harm by deploying and improving near zero emission energy systems.

Broadening our options can only be a good thing, unless the broader palette prevents us from taking decisive action today.

Ken Caldeira

Carnegie Institution for Science, Dept. of Global Ecology


Andy Revkin:

I agree with Ray’s point on surprises — the biggest being how old on-shelf science in entrepreneurs’ hands opened the vast shale energy resource that was thought untappable. Perhaps the biggest blow to ANY clean-energy path?


David Keith, Gordon McKay Professor of Applied Physics, Harvard

Ray asserts that carbon dioxide removal (CDR) technologies do not exist at scale now and that we cannot have any confidence that they will exist during this century. He compares the uncertainty about their development with the uncertainty in predicting when we will have fusion power.

I don’t agree. While I suspect our disagreement is actually about political strategy not technology, I will tackle the technological claim first.

What does it mean to say that a technology exists?

I would say that there are three functional categories:

A. Market. Something that multiple vendors can supply in quantity in a market today.

B. Feasible. Something that multiple vendors could readily supply using current capabilities if there was a significant market.

C. Future. And finally, something that is physically possible but where there’s no clear route to supplying it at a reasonable price.

Fusion power is future tech. The world could spend 50 $bn over the next 20 years and while we would certainly get some burning plasmas, there is no basis for confidence that after, say, two decades of development you could make electricity at 70 $/MWhr or some other reasonable price. Nor, would final costs be much better understood after spending the first 10%.

How about a liquid hydrogen powered 100 passenger transport aircraft? This is feasible. Gas turbine hydrogen combustors have been developed and tested , but there is not a model number ready-to-go from GE. Liquid hydrogen tankage is straightforward and airframe development is a known quantity. Airbus did a major study. The total development cost of such an aircraft might be 20 $bn — three times the 787 — but there is no doubt that it could be built. Most the design would rest on Commercial Off-The-Shelf-Technology (COTS), the risk would be integration and execution. Furthermore, after spending 10% of the money on an initial design study, one could know fairly accurately the overall cost performance of the final product. The reason that we don’t have such aircraft is that there is no strong market incentive to build one.

What about CDR? Is it COTS-feasible or is it future tech? I take the claim that we can’t be confident that it will exist late this century as a claim that CDR is future not feasible.

Biomass with capture is clearly in the COTS-feasible category. Multiple independent vendors (e.g., MHI and Fluor) could build biomass systems with carbon capture, compression, and cleanup. And they could do it at well understood prices with performance guarantees. Companies like Schlumberger could manage the injection of CO2. Markets for biomass are relatively well understood. If the US wanted to have 100 Mt-CO2/year of BECCS by 2030, it could do it with confidence if it set a clear price and regulatory structure now. This is not a technology question. We can argue about what the price would be but it is very likely well under 200 $/tCO2 at that scale.

Then there is industrial scale burial of wood wastes.

Of course, the problem with biomass with capture is scalability and environmental impact. But current biomass supply suggests that humanity could scale BECCS to 1 GtC/year without substantially increasing land footprint. Here’s my 2001 opinion piece on the tradeoffs, and here is Jamie Rhode’s PhD, the first on BECCS, which still provides a nice overview.

I am not arguing for a grand commitment to BECCS — my vote is for higher intensity energy systems that spare more land for nature — but, it could be done.

Now consider adding alkalinity to the ocean using the old-fashioned idea of calcining limestone with CO2 capture and then adding the lime to the ocean. Limestone calcination with post-combustion amine capture is COTS-feasible. Multiple vendors could do it today with performance guarantees and well understood costs. Though environmentally acceptable engineering methods to distribute that lime in the ocean do not exist, there are a host of technologies from which they could be developed. And my guess is we could get some useful answers in less than a decade for a cost of less than 50 $m.

Unlike biofuels — there is no obvious scale limit for CDR by alkalinity addition.

Work is not being done because there no policy incentive to do it — -not because it’s all that hard.

Obviously, I am conflicted on the topic of direct air capture since I helped to found Carbon Engineering, a startup that works on the topic. But the technologies that we are using are mostly the old blue-collar ideas oxy-fuel limestone calcination, an aqueous pellet reactor used in water treatment, forced draft cooling tower technologies. We have industrial partners who can provide us with performance guarantees for all of these things. We have a pilot plant which is demonstrating integration using industrial hardware the cost of about 10 $m. There are lots of scale up and execution risks, we are a small company, but we and our investors and industrial partners are confident that a large-scale version of our current system could do air capture with CO2 compression at a cost well under 200 $/tCO2. Whether or not we’re right, we developed all the engineering and hardware for our pilot at a total cost that of about 20 $m. So that says that one could make a lot of progress in this field if there were five different efforts similar to ours. Of course the others don’t agree with our assessment. Disagreement is healthy, but with our technology and that of Climeworks now being piloted at scales near 1 kt-CO2/year, there is sufficient engineering for third parties do a far more realistic assessment than the APS was able to do.

The world now spends about 300 $bn a year clean energy. If there was an effort to develop CDR technologies at a scale that was even a 1% of this amount, it is hard to believe that scalable technologies could not be developed given that many of them don’t require some deeply uncertain technological leap as it the case for fusion or space solar power. These technologies can be developed by applying today’s large industrial technologies in novel configurations.

CDR is much more like the hydrogen airplanes than it is fusion power.

The world spent over $ 1T on solar and wind before average carbon avoided costs of good projects got significantly under 200 $/tCO2. Thanks to a myriad small innovations but no big breakthrough cheap solar PV is now a reality. CDR could proceed in much the same way.

It is of course entirely possible that there will be no large-scale CDR technologies late this century. But if that’s true, it’s a choice not a forced outcome.

While my disagreements with Ray are framed as if they were about technology and science I suspect root disagreement is political.

Arguments that these technologies are very unlikely to exist late this century are, I suspect, motivated by a well-intentioned belief that the only way to get rapid investment in mitigation is to convince the public that it’s the only way to avoid the climate crisis.

Emissions mitigation is necessity. Nothing I know about carbon removal or solar geoengineering suggest that we can avoid the need to cut emissions. On a political level it’s possible that Ray is correct and I am wrong. That the best strategy is to downplay CDR and demonize solar geoengineering because bringing them into the mainstream discussion will sap political will to cut emissions. But the claim that we should stick to only one tool — mitigation — is a political claim, not a technical or scientific one.




Gernot Wagner, research associate, Harvard School of Engineering and Applied Science, co-author, with Harvard’s Martin Weitzman, of “Climate Shock

A couple of thoughts, building on the references to politics as the main driver of some — perhaps most — of the differences in opinions here:

I’d say the perception/communications question goes to the very core of this debate, whether it’s carbon dioxide removal or albedo modification aka ‘solar geoengineering’. The question is whether talking about either is a complement to or a substitute for mitigation.

There are two competing theories: What you might call the “Copenhagen Theory of Change” describes the complementarity viewpoint. One thing leads to another, and soon enough you have half of Copenhageners biking to work in the bitter cold.

No, biking alone won’t stop climate change, but that’s not the point here. The point is how lots of individual policies added up to much more over time.

The opposite is inherently more comfortable to economists. There are trade-offs after all. Or as philosopher Jagger once put it, you can’t always get what you want. It’s one or the other.

That works on an individual level. It might go for policy, too. Obama has only so much time left to spend on thinking about climate policy. If he spends it on one aspect of the problem, he won’t think as much about another. Call it the “Crowding-Out Bias.”

Both of these theories are at play here, though it’s entirely unclear which dominates.

Medical analogies abound. Taking statins lowers your risk of heart disease. Now, when the doc tells you about Lipitor, does that mean you should stop dieting and exercising? Of course not. If anything, the hope is that when the doc mentions Lipitor, it may also draw attention to the fact that you may want to eat your greens and take the stairs more often.

Something similar goes for carbon dioxide removal and, perhaps most obviously, for research into albedo modification. Does talk of it make you want to stop mitigating in the first place, or is there an element of complementarity here, too: Wait, serious climate scientists are talking about doing what? Maybe we should take another look at that carbon tax after all.

Seeing what it’s like to go through chemotherapy first-hand may well motivate you to stop smoking. In any case, just because chemo works in some lab rat somewhere doesn’t mean you should pick up a cigarette.

The real challenge then is to design conversations around geoengineering and mitigation that reinforce rather than hamper each other.

But let’s also make sure to keep that task separate from the technology question itself.

Your doc would clearly be wrong to suggest to you to try a quick dose of chemo, even if you were healthy. But he would be similarly remiss not to have done his homework and know enough about chemo to know to suggest it when necessary.

In other words, don’t let the perception or possibility of ‘moral hazard’ stop you from doing the research on albedo modification in the first place.

Best, Gernot


INSERT, Sept. 5, 2016 | Ken Caldeira added these thoughts, first a close-focus point, then a broader reflection:

At some point, a strong motivation to engage in solar geoengineering may be to allow additional time for adaptation.

For example, if high temperatures lead to widespread crop failures in the tropics, people may be motivated to delay warming or slow the rate of warming in the hopes of providing additional time to develop heat resistant crop varieties.

Imaginably, people might be motivated to prevent further warming for a while to give people time to prepare for warmer temperatures to come. Or people might decide to slow rates of warming to allow for a more orderly transition to a warmer world.

Doug MacMartin, David Keith and I wrote a paper in 2014 discussing how people might be motivated to use solar geoengineering to limit rates of warming, but not eventual amounts of warming, and also discussing how a solar geoengineering deployment might be phased out if society concluded that the solar geoengineering was not improving environmental outcomes:

Solar geoengineering to limit the rate of temperature change

17 November 2014.DOI: 10.1098/rsta.2014.0134

[In a theme issue of Philosophical Transactions of the Royal Society A: ‘Climate engineering: exploring nuances and consequences of deliberately altering the Earth’s energy budget’ compiled and edited by John Latham, Philip J. Rasch and Brian Launder]

Further, there is no such thing as millennial commitment to solar geoengineering. The aerosols remain in the atmosphere for only a year or two. If people maintain a solar geoengineering for 1000 years or more, it will be because they feel the solar geoengineering is making the world better than it otherwise would be. If people felt the solar geoengineering was making the world worse, they would ramp it out. There is no millennial scale commitment to solar geoengineering, but if solar geoengineering is making things better, they might have strong millennial-scale incentives to continue using solar geoengineering to maintain a better environment.

Absent carbon dioxide removal, carbon dioxide emissions commit us to millennia of climate change. Solar geoengineering does not commit us to millennia of anything.

By the way, we also did a paper on millennial-scale solar geoengineering (Cao et al, 2016) showing that, in at least one climate model, solar geoengineering behaves quite well on the 1000-year time scale with no substantial long term growth in climate change as ocean circulation and such adjusts to the new conditions.

Lastly, and this has just been submitted and so is not something to write about now, we submitted a paper to a Crutzen+10 special section of Earth’s Future. I attach it. You might find the second and third parts amusing (i.e., starting on line 112 to line 146 and from line 184). While this is not formally under embargo, I would like to save if for tweeting about etc when it comes out. This commentary has just a submitted draft and has not been reviewed yet, so will be revised prior to publication.

The reflection:

My PhD work, 25 years ago now, was on what happened to the carbon cycle when the dinosaurs went extinct 65 million years ago. So, I have long been sensitive to the importance of considering long time scales.

One of my first papers concluded that CO2 emissions would be removed from the Earth’s surficial environment on a time scale of about 300,000 years (Caldeira and Rampino, 1991). The piece on ocean acidification that I had in Nature in 2003 explicitly put our current activities in the context of geologic time. I’ve worked with Dave Archer on a paper in 2009 characterizing the multi-millennium time-scales associated with CO2 removal from Earth’s surficial environment. In 2015, I co-authored a paper in Science Advances on the response of Antarctica to our CO2 emissions over the next 10,000 years.

I co-authored a paper with Long Cao on the 1000-year response to solar geoengineering. Also in 2015, Xiaochun Zhang and I published a paper pointing out that, over the several hundred thousands of years that today’s CO2 emissions from fossil-fuel burning will perturb atmospheric content, the radiative forcing from that CO2 will warm the Earth more than 100,000 times more than the direct thermal emissions from the combustion of fossil fuel.

So, when Clarke et al emphasize the importance of considering time scales that are considered to be long by politicians, but nearly instantaneous by geologists, that message falls on very receptive ears. To me, one of the most important changes in thinking over the 25 years since I got my PhD is that 25 years ago, most scientists thought of the “lifetime” of CO2 in the atmosphere as being several decades. Now, it is clear to everyone that today’s CO2 emissions will continue perturbing atmospheric CO2 concentrations for many tens of thousands, if not hundreds of thousands of years.

Where I might differ with Clarke et al is the emphasis on a particular time frame, in their case 20 millennia. Every time scale is important. What happens in the here and now is important and what happens over hundreds of thousands of years is also important. One of the reasons the climate problem is so tough is that it asks people to quantitatively assess the relative importance of the influence of the actions I take today on what happens today, next year, next decade, next century, next millennium, etc. This balance depends both on our values and what we think the influence of our actions will be. Available empirical facts are insufficient to induce consensus with respect to either values or expectations. We will have to muddle through understanding that consensus will never be reached on these issues. Nevertheless, it is clear that society has been undervaluing the long-term view.

Joe Romm extracted something from an email I once wrote to him:

Economists estimate that it might cost something like 2% of our GDP to convert our energy system into one that does not use the atmosphere as a waste dump. When we burn fossil fuels and release the CO2 into the atmosphere, we are saying “I am willing to impose tremendous climate risk on future generations living throughout the world, so that I personally can be 2% richer today.”

I believe this to be fundamentally immoral. We are saying we want to selfishly reap benefits today while imposing costs on strangers tomorrow.

Would we like it today if the Romans had developed a modern technological society like ours, and their scientists told them that using the atmosphere as a waste dump for greenhouse gases would melt the ice caps, acidify the oceans, overheat the tropics, cause species extinctions, etc, and then they decided to go ahead and do it anyway, just because they were selfish and didn’t care about other people? Perhaps their economists too would do a net present value calculation that would tell them that selfishness is the way to go. Would we be happy to have all of this environmental damage comforted by the knowledge that they knowingly imposed these costs on us in order to be 2% richer?

All I am asking is that we follow the golden rule: “Do unto others as you would have them do unto you.” This is fundamentally a moral issue, not an economic issue. Given what we know now, it is simply unethical to impose risk of grave damage on future generations just so that we can have a few more consumer products today.



In July, The Times published an Op-Ed article by Adam Sobel of Columbia University on a Science paper he co-authored finding that air pollution may be temporarily limiting an expected boost in hurricane power in a warming climate. Keith and Wagner sent me the following note, which they’d submitted, as well, as a letter to the editor (it didn’t run):

In “Where Are the Hurricanes?” Adam Sobel points to an important tradeoff between reducing greenhouse gases that warm and reducing aerosols that cool the planet (op-ed, July 15). Aerosols in the lower atmosphere have reduced global warming and the frequency of intense hurricanes by reflecting a small fraction of sunlight back to space.
These aerosols are air pollution that kills 3–6 million people a year worldwide. We must reduce pollution, but we don’t want more warming.
What if we reduce aerosol pollution in the lower atmosphere (saving lives) while deliberately injecting sufficient aerosol into the upper atmosphere to keep global temperatures (and perhaps hurricanes) in check? This is called ‘solar geoengineering’, it appears possible, and it might require only one fiftieth of the amount of aerosols now polluting the lower atmosphere to achieve the same cooling. The health and climate benefits could be significant, but much more research is needed — both on the efficacy and on the potential risks. Policy makers should no longer ignore that research.

By email, I asked Wagner and Keith if they or others had analyzed a “tailored” approach to solar radiation management, essentially synched to counter the added warming pulse as the world progressively cleans up the cooling pollutants Sobel points to? I mused that this might be an easier sell as a test of the concept because it keeps things in the range of today’s atmospheric composition and dynamics.

Keith replied:

We have talked about lots, but never wrote a paper saying that because I think it’s too trivial to be worth publishing.
We did write a general paper about moderate solar geo and the fact that it can be ramped up and down to match some goal: 174. David W. Keith and Douglas G. MacMartin. (2015) A temporary, moderate and responsive scenario for solar geoengineering. Nature Climate Change, DOI: 10.1038/NCLIMATE2493.

I replied:

Trivial climatologically, yes, but not strategically… (to my mind).

Demonstrates this can be done in a “do no [new] harm” way.

Sequencing matters — as Steve Schneider said so well in context of climate policy.

See below but substitute solar management for CO2 emissions>

I’ve cited the late Stephen H. Schneider several times on his notion of getting the sequence of initiatives right to build public support for the energy quest that’d be needed to foster progress while limiting risks of disruptive climate change. I’ll do so again here. He made this point most clearly while a “thinker in residence” in Adelaide, Australia, in 2006: “It is important that the sequencing of policy steps for achieving the emissions target build from obvious win-wins to more difficult steps such as establishing a shadow price for carbon.”


ACR: There’s more to the original conversation, which I will append when I have time. (Apologies to participants for the long delay in posting this!)