We Have No Way to Predict the Unintended Consequences of Geoengineering

Using technology to fix climate change requires careful research—but that's easier said than done.

To prove that reflecting sunlight with sulfates can safely cool Earth, experiments must be large enough in geography and long enough in time frame. Dado Ruvic/Reuters

This story originally appeared on Slate and is reproduced here as part of the Climate Desk collaboration.

The historic agreement forged in Paris among 195 countries in December holds the promise of triggering a global shift to combat climate change—and harbors a hidden warning.

Regardless of what happens next, the Paris accord is a triumph of diplomacy among nations that have starkly disagreed in years past about who is responsible for cutting carbon dioxide emissions—and who should bear the cost. But success in heading off the worst climate disruptions hinges on whether countries fulfill the pledges each made leading up to the Paris talks and make bolder ones this decade. The teeth come in the form of sunshine and shame: The accord requires transparency and monitoring of emissions from each country. And it relies on countries to be motivated by the ignominy they would face if they reneged.

There is a danger that shame will not be enough. In the United States, Republican presidential candidates have already vowed to undo President Obama's climate change policies, including the pivotal Clean Power Plan that regulates emissions from the electricity sector. Senate Majority Leader Mitch McConnell warned that the Paris accord stands on shaky ground; he thinks, contrary to most credible legal experts, that the Obama power plant rules are illegal. It's unlikely the prospect of infamy would deter a climate change–denying president or Congress from shattering the U.S. pledge. That's the kind of consideration that pertains to the reasonable and rational.

What would happen if the U.S. failed to keep up its end of the agreement? The United States' actions greatly affect other countries', because the world sees us as responsible for the mess everyone's in. Leaders of developing countries and island nations hold industrialized countries responsible for the reckless carbon binges of past decades that have pushed low-lying territories to the brink of disaster. The Obama administration summoned all its leverage to persuade other countries to develop without emitting as much carbon as we did on our path to economic dominance. In Paris, the U.S. had to play the parents who say, "Do as I say, not as I do," while also promising a generous allowance.

The United States faces strong internal pressure to keep burning fossil fuels, reflected in our divisive politics; other nations—especially island nations like Tuvalu and Kiribati—face strong pressure to keep the planet cooler at any cost. The seas are already rising. The mood is ripe for private-sector companies or individual nations to seek drastic ways to change the climate, either to avoid the cuts agreed to in Paris or to hedge their bets in case of political failure. Yet absent from the Paris agreement and absent from U.S. political discourse is any robust discussion of what could be a growing threat, especially after the November presidential election: that countries, people, or businesses will take it upon themselves to directly cool the planet.

Experiments in geoengineering have already been tried. In 2012, a rogue scientist dumped 120 metric tons of iron into the Pacific Ocean to grow plankton blooms to remove carbon dioxide from the atmosphere. He violated no law. In 2015, China announced plans to seed clouds with chemicals to boost rainfall, building on its standing artificial weather program. Based on the cost and potential cooling effect of various geoengineering technologies, the most likely scenario today is that someone would attempt to change the atmosphere by pumping sulfate aerosols in the stratosphere—to reflect sunlight away from Earth. (This technology, dubbed solar radiation management, is being promoted and researched by a small but vocal group of scientists.)

Now is the time, with the wind from Paris at our backs, to set international norms for how geoengineering technologies are tested and deployed and to consider how the U.S. would navigate a global landscape in which different nations want to engineer the climate to different ends. Would Russia want to warm Earth beyond 2 degrees Celsius to turn Siberia into a fertile growing region? Will Vanuatu find a sympathetic billionaire to shield the planet from the sun so that sea levels do not rise so high?

More research on geoengineering could help us anticipate the possible ways the technologies could be used. But we should be clear about what each stage of research can actually show us. In order to prove that the technique of reflecting sunlight with sulfates can cool the planet consistently without terrible consequences, experiments must ultimately be large enough in geography and long enough in time frame. And those characteristics raise the possibility for widespread, unintended consequences. Rutgers climate scientist Alan Robock has argued that trials in the atmosphere won't show a significant climate response, "unless an experiment is so large as to actually be geoengineering" and lasts at least a decade. (Experimenters would need to confirm that any changes in climate were not just coincidental.)

Scientists have studied the effects of volcanoes that temporarily cooled the planet, such as the 1991 Mount Pinatubo eruption in the Philippines. But incidents in the past cannot simulate what it would really be like to try to cool the entire planet today with such technologies, over a time span of not years, but decades. Launching small experiments of limited duration—or gradual deployment, as has been advocated by Harvard geoengineering scientist David Keith—may also help advance our knowledge. But the world should be aware that this method cannot tell us whether the technology is safe enough or too dangerous to deploy—it can't give us the kind of insights that we have, for instance, when we test a drug for its side effects in a randomized controlled trial. (Even with that gold standard for assessing risk, we still have cases like Vioxx, where the dangerous side effects for large populations can be far worse than what appears in a limited clinical trial. In geoengineering, we have only one planet, not many patients, and the benefits and risks are collective, not individual.)

This conundrum of conducting large-scale solar geoengineering experiments poses an ethical dilemma that cannot be resolved by scientists alone. We need a robust public debate to ask when and to what extent it is ethical to experiment with the planet. We need global norms that take into account the uncertainty and serious risks that solar radiation management could pose, such as manipulation of weather patterns and damage to the food supply, air pollution deaths, depletion of the ozone layer, and other impacts we may not yet anticipate in the dynamic, complex system that is Earth's climate. If early experiments epically fail, will they be counterproductive to the technology over the long term, like the nuclear meltdown in Three Mile Island?

A recent global summit on gene editing technologies hosted by national scientific councils from the United States, the United Kingdom, and China could provide a model for how policymakers, ethicists, scientists, and the public can set boundaries on the use of technologies with unknown and intergenerational consequences. After the summit, the three councils published principles to guide when the use of gene editing is sound and ethical, and when it is too risky—namely, when it poses unpredictable and irreversible impacts for future generations. While such norms won't stop rogue engineers, they at least keep the wise and the willing from unleashing unforeseen consequences. And they can help scientists navigate the fine line between expanding knowledge and deploying technologies in the absence of international support.

The chance of success in cutting climate change emissions has never seemed more palpable. But we should seize the goodwill generated in Paris to talk about the engineering elephant in the room. The United States can and should lead not just by making politically determinate commitments but by anticipating what could happen if we fail to keep up our end of the bargain.