Beyond these conceivable dangers, though, a more fundamental problem lurks: Solar geoengineering simply might not work. It has been a subject of intense debate among climate scientists for roughly a decade. But most of what we know about its potential effects derives from either computer simulations or studies on volcanic eruptions like that of Mount Pinatubo in 1991, which generated millions of tons of sunlight-scattering particulates and might have cooled the planet by as much as 0.5 degrees Celsius, or nearly 1 degree Fahrenheit. The lack of support for solar geoengineering’s efficacy informs Keith’s thinking about what we should do next. Actively tinkering with our environment — fueling up the Gulfstream jets and trying to cool things down — is not something he intends to try anytime soon, if ever. But conducting research is another matter.
A decade ago, when Keith was among the few American scientists to advocate starting a geoengineering research program, he was often treated at science conferences as an outlier. “People would sort of inch away or, really, tell me I shouldn’t be doing this,” he said. Geoengineering was seen as a scientific taboo and Keith its dark visionary. “The preconception was that I was some kind of Dr. Strangelove figure,” he told me — “which I didn’t like.”
Attitudes appear to have changed over the past few years, at least in part because of the continuing academic debates and computer-modeling studies. The National Academy of Sciences endorsed the pursuit of solar geoengineering research in 2015, a stance also taken in a later report by the Obama administration. A few influential environmental groups, like the Natural Resources Defense Council and the Environmental Defense Fund, now favor research.
In the meantime, Keith’s own work at Harvard has progressed. This month, he is helping to start Harvard’s Solar Geoengineering Research Program, a broad endeavor that begins with $7 million in funding and intends to reach $20 million over seven years. One backer is the Hewlett Foundation; another is Bill Gates, whom Keith regularly advises on climate change. Keith is planning to conduct a field experiment early next year by putting particles into the stratosphere over Tucson.
The new Harvard program is not merely intent on getting its concepts out of the lab and into the field, though; a large share of its money will also be directed to physical and social scientists at the university, who will evaluate solar geoengineering’s environmental dangers — and be willing to challenge its ethics and practicality. Keith told me, “It’s really important that we have a big chunk of the research go to groups whose job will be to find all the ways that it won’t work.” In other words, the technology that Keith has long believed could help us ease our predicament — “the nuclear option” for climate, as one opponent described it to me, to be considered only when all else has failed — will finally be investigated to see whether it is a reasonable idea. At the same time, it will be examined under the premise that it may in fact be a very, very bad one.
Climate change already presents a demoralizing array of challenges — melting ice sheets and species extinctions — but the ultimate severity of its impacts depends greatly on how drastically technology and societies can change over the next few decades. The growth of solar and wind power in recent years, along with an apparent decrease in coal use, suggest that the global community will succeed in curtailing CO₂ emissions. Still, that may not happen nearly fast enough to avert some dangerous consequences. As Keith likes to point out, simply reducing emissions doesn’t reverse global warming. In fact, even if annual global CO₂ emissions decrease somewhat, the total atmospheric CO₂ may continue to increase, because the gas is so slow to dissipate. We may still be living with damaging amounts of atmospheric carbon dioxide a half-century from now, with calamitous repercussions. The last time atmospheric CO₂ levels were as elevated as they are today, three million years ago, sea levels were most likely 45 feet higher, and giant camels roamed above the Arctic Circle.
Recently, I met with Daniel Schrag, who is the head of the Harvard University Center for the Environment, an interdisciplinary teaching and research department. Schrag, who helped recruit Keith to Harvard, painted a bleak picture of our odds of keeping global temperatures from rising beyond levels considered safe by many climate scientists. When you evaluate the time scales involved in actually switching our energy systems to cleaner fuels, Schrag told me, “the really depressing thing is you start to understand why any of these kinds of projections — for 2030 or 2050 — are absurd.” He went on: “Are they impossible? No. I want to give people hope, too. I’d love to make this happen. And we have made a lot of progress on some things, on solar, on wind. But the reality is we haven’t even started doing the hard stuff.”
Schrag described any kind of geoengineering as “at best an imperfect solution that is operationally extremely challenging.” Yet to Schrag and Keith, the political and technical difficulties associated with a rapid transition to a zero-carbon-emissions world make it sensible to look into geoengineering research. There happens to be a number of different plans for how to actually do it, however — including the fantastical (pumping seawater onto Antarctica to combat sea-level rise) and the impractical (fertilizing oceans with iron to foster the growth of algae, which would absorb more CO₂). Some proposals involve taking carbon out of the air, using either immense plant farms or absorption machines. (Keith is involved with such sequestration technology, which faces significant hurdles in terms of cost and feasibility.) Another possible approach would inject salt crystals into clouds over the ocean to brighten them and cool targeted areas, like the dying Great Barrier Reef. Still, the feeling among Keith and his colleagues is that aerosols sprayed into the atmosphere might be the most economically and technologically viable approach of all — and might yield the most powerful global effect.
It is not a new idea. In 2000, Keith published a long academic paper on the history of weather and climate modification, noting that an Institute of Rainmaking was established in Leningrad in 1932 and that American engineers began a cloud-seeding campaign in Vietnam a few decades later. A report issued in 1965 by President Lyndon B. Johnson’s administration called attention to the dangers of increasing concentrations of CO₂ and, anticipating Keith’s research, speculated that a logical response might be to change the albedo, or reflectivity, of the earth. To Keith’s knowledge, though, there have been only two actual field experiments so far. One, by a Russian scientist in 2009, released aerosols into the lower atmosphere via helicopter and appears to have generated no useful data. “It was a stunt,” Keith says. Another was a modest attempt at cloud brightening a few years ago by a team at the Scripps Institution of Oceanography at the University of California, San Diego.
Downstairs from Keith’s Harvard office, there is a lab cluttered with students fiddling with pipettes and arcane scientific instruments. When I visited in early March, Zhen Dai, a graduate student who works with Keith, was engaged with a tabletop apparatus, a maze of tubes and pumps and sensors, meant to study how chemical compounds interact with the stratosphere. For the moment, Keith’s group is leaning toward beginning its field experiments with ice crystals and calcium carbonate — limestone — that has been milled to particles a half-micron in diameter, or less than 1/100th the width of a human hair. They may eventually try a sulfur compound too. The experiment is called Scopex, which stands for Stratospheric Controlled Perturbation Experiment. An instrument that can disperse an aerosol of particles — say, several ounces of limestone dust — will be housed in a gondola that hangs beneath a balloon that ascends to 70,000 feet. The whole custom-built contraption, whose two small propellers will be steered from the ground, will also include a variety of sensors to collect data on any aerosol plume. Keith’s group will measure the sunlight-scattering properties of the plume and evaluate how its particles interact with atmospheric gases, especially ozone. The resulting data will be used by computer models to try to predict larger-scale effects.
But whether a scientist should be deliberately putting foreign substances into the atmosphere, even for a small experiment like this, is a delicate question. There is also the difficulty of deciding on how big the atmospheric plumes should get. When does an experiment become an actual trial run? Ultimately, how will the scientists know if geoengineering really works without scaling it up all the way?
Keith cites precedents for his thinking: a company that scatters cremation ashes from a high-altitude balloon, and jet engines, whoseexhaust contains sulfates. But the crux of the problem that Harvard’s Solar Geoengineering Research Program wrestles with is intentionality. Frank Keutsch, a professor of atmospheric sciences at Harvard who is designing and running the Scopex experiments with Keith, told me: “This effort with David is very different from all my other work, because for those other field experiments, we’ve tried to measure the atmosphere and look at processes that are already there. You’re not actually changing nature.” But in this case, Keutsch agrees, they will be.
During one of our conversations, Keith suggested that I try to flip my thinking for a moment. “What if humanity had never gotten into fossil fuels,” he posed, “and the world had gone directly to generating energy from solar or wind power?” But then, he added, what if in this imaginary cleaner world there was a big natural seep of a heat-trapping gas from within the earth? Such events have happened before. “It would have all the same consequences that we’re worried about now, except that it’s not us doing the CO₂ emissions,” Keith said. In that case, the reaction to using geoengineering to cool the planet might be one of relief and enthusiasm.
In other words, decoupling mankind’s actions — the “sin,” as Keith put it, of burning fossil fuels — from our present dilemma can demonstrate the value of climate intervention. “No matter what, if we emit CO₂, we are hurting future generations,” Keith said. “And it may or may not be true that doing some solar geo would over all be a wise thing to do, but we don’t know yet. That’s the reason to do research.”
There are risks, undeniably — some small, others potentially large and terrifying. David Santillo, a senior scientist at Greenpeace, told me that some modeling studies suggest that putting aerosols in the atmosphere, which might alter local climates and rain patterns and would certainly affect the amount of sunlight hitting the earth, could have a significant impact on biodiversity. “There’s a lot more we can do in theoretical terms and in modeling terms,” Santillo said of the Harvard experiments, “before anyone should go out and do this kind of proof-of-concept work.” Alan Robock, a professor of atmospheric sciences at Rutgers, has compiled an exhaustive list of possible dangers. He thinks that small-scale projects like the Scopex experiment could be useful, but that we don’t know the impacts of large-scale geoengineering on agriculture or whether it might deplete the ozone layer (as volcanic eruptions do). Robock’s list goes on from there: Solar geoengineering would probably reduce solar-electricity generation. It would do nothing to reduce the increasing acidification of the oceans, caused by seawater absorbing carbon dioxide. A real prospect exists, too, that if solar geoengineering efforts were to stop abruptly for any reason, the world could face a rapid warming even more dangerous than what’s happening now — perhaps too fast for any ecological adaptation.
Keith is well aware of Robock’s concerns. He also makes the distinction that advocating research is not the same as advocating geoengineering. But the line can blur. Keith struck me as having a fair measure of optimism that his research can yield insights into materials and processes that can reduce the impacts of global warming while averting huge risks. For instance, he is already encouraged by computer models that suggest the Arctic ice cap, which has shrunk this year to the smallest size observed during the satellite era, could regrow under cooler conditions brought on by light-scattering aerosols. He also believes that the most common accusation directed against geoengineering — that it might disrupt precipitation patterns and lead to widespread droughts — will prove largely unfounded.
But Keith is not trained as an atmospheric scientist; he’s a hands-on physicist-engineer who likes to take machinery apart. There are deep unknowns here. Keutsch, for one, seems uncertain about what he will discover when the group actually tries spraying particulates high above the earth. The reduction of sunlight could adversely affect the earth’s water cycle, for example. “It really is unclear to me if this approach is feasible,” he says, “and at this point we know far too little about the risks. But if we want to know whether it works, we have to find out.”
Finally, what if something goes wrong either in research or in deployment? David Battisti, an atmospheric scientist at the University of Washington, told me, “It’s not obvious to me that we can reduce the uncertainty to anywhere near a tolerable level — that is, to the level that there won’t be unintended consequences that are really serious.” While Battisti thought Keith’s small Scopex experiment posed little danger — “The atmosphere will restore itself,” he said — he noted that the whole point of the Harvard researchers’ work is to determine whether solar geoengineering could be done “forever,” on a large-scale, round-the-clock basis. When I asked Battisti if he had issues with going deeper into geoengineering research, as opposed to geoengineering itself, he said: “Name a technology humans have developed that they haven’t used. I can’t think of any. So we can work on this for sure. But we are in this dilemma: Once we do develop this technology, it will be tempting to use it.”
Suppose Keith’s research shows that solar geoengineering works. What then? The world would need to agree where to set the global thermostat. If there is no consensus, could developed nations impose a geoengineering regimen on poorer nations? On the second point, if this technology works, it would arguably be unethical not to use it, because the world’s poorest populations, facing drought and rising seas, may suffer the worst effects of a changing climate.
In recent months, a group under the auspices of the Carnegie Council in New York, led by Janos Pasztor, a former United Nations climate official, has begun to work through the thorny international issues of governance and ethics. Pasztor told me that this effort will most likely take four years. And it is not lost on him — or anyone I spoke with in Keith’s Harvard group — that the idea of engineering our environment is taking hold as we are contemplating the engineering of ourselves through novel gene-editing technologies. “They both have an effect on shaping the pathway where human beings are now and where will they be,” says Sheila Jasanoff, a professor of science and technology studies at Harvard who sometimes collaborates with Keith. Jasanoff also points out that each technology potentially enables rogue agents to act without societal consent.
This is a widespread concern. We might reach a point at which some countries pursue geoengineering, and nothing — neither costs nor treaties nor current technologies — can stop them. Pasztor sketched out another possibility to me: “You could even have a nightmare scenario, where a country decides to do geoengineering and another country decides to do counter-geoengineering.” Such a countermeasure could take the form of an intentional release of a heat-trapping gas far more potent than CO₂, like a hydrochlorofluorocarbon. One of Schrag’s main concerns, in fact, is that geoengineering a lower global temperature might preserve ecosystems and limit sea-level rise while producing irreconcilable geopolitical frictions. “One thing I can’t figure out,” he told me, “is how do you protect the Greenland ice sheet and still have Russia have access to its northern ports, which they really like?” Either Greenland and Siberia will melt, or perhaps both can stay frozen. You probably can’t split the difference.
For the moment, and perhaps for 10 or 20 years more, these are mere hypotheticals. But the impacts of climate change were once hypotheticals, too. Now they’ve become possibilities and probabilities. And yet, as Tom Ackerman, an atmospheric scientist at the University of Washington, said at a recent discussion among policy makers that I attended in Washington: “We are doing an experiment now that we don’t understand.” He was not talking about geoengineering; he was observing that the uncertainty about the potential risks of geoengineering can obscure the fact that there is uncertainty, too, about the escalating disasters that may soon result from climate change.
His comment reminded me of a claim made more than a half-century ago, long before the buildup of CO₂ in the atmosphere had become the central environmental and economic problem of our time. Two scientists, Roger Revelle and Hans Suess, wrote in a scientific paper, “Human beings are now carrying out a large-scale geophysical experiment of a kind that could not have happened in the past nor be reproduced in the future.”
If anything could sway a fence-sitter to consider whether geoengineering research makes sense, perhaps it is this. The fact is, we are living through a test already.