A Plan to Answer the Sunlight Reflection Question
Geoengineering News
A Plan to Answer the Sunlight Reflection QuestionA field-wide plan for the technical research needed to determine whether SAI could ever responsibly be used to temporarily reduce global temperatures.
The first papers on sunlight reflection were published in the 1960s. Today, Stratospheric Aerosol Injection (SAI), releasing reflective sulfate particles into the stratosphere that reflect sunlight back to space, is the most researched form of sunlight reflection. Six decades later, nobody can answer the basic question: is SAI a tool the world could ever use responsibly, or one it should rule out? We do not have a shared plan for answering that question, either.Another decade of papers, funded one at a time and pursued in no particular order, will not change that. The problem is not that the research is bad. It is that the field is not organized to converge on a decision. Individual teams pursue individual questions. Funders make individual bets. Governments may see the importance of the issue without having a concrete view of what research to support, in what order, or at what scale. SAI may be one of very few interventions capable of reducing global temperatures fast enough to affect climate risks over the next few decades. It cannot be dismissed casually. Its risks and uncertainties mean it cannot be embraced casually either. It is not a substitute for cutting emissions or adapting to warming, and research on it should add to—not displace—the much larger investments those efforts require. Today, we are publishing the SAI Research Roadmap: a field-wide plan for the technical research needed to determine whether SAI could ever responsibly be used to temporarily reduce global temperatures. The Roadmap works backward from that decision. It contains 45 research activities, each with a scientific justification, scope of work, estimated cost and timeline, the work it depends on, and what it unlocks. An activity belongs only if it materially reduces an uncertainty that matters to the decision, or is necessary for research that does. If the research shows SAI is a bad idea, the Roadmap will have done its job. The headline numbers: even when every activity that can run in parallel does, reaching a well-informed decision about whether to begin monitored cooling will take at least a decade and at least $370 million. Those are lower bounds, not a forecast. They assume substantial, coordinated investment starting soon and no avoidable gaps along the critical path. Borrowing the logic of clinical trialsTo build the Roadmap, we borrowed the logic of clinical trials: learn everything you can at a given scale before graduating to a larger, potentially riskier phase. Phase 0 is where the field is now—using models, laboratory studies, engineering work, and existing observations to estimate the potential benefits and risks of SAI. Phase 1 adds small-scale outdoor research to measure how sulfate particles actually form and grow in the stratosphere. Phase 2, if reached, involves a larger, sustained release designed to measure how stratospheric winds transport those particles around the globe. Phase 3, if it is ever reached, is monitored cooling: a slow ramp-up, measured continuously against predictions, that could be paused, adjusted, or reversed. Outdoor experiments anchor Phases 1 and 2, but neither phase is just an experiment. Each phase also includes the modeling, impact research, engineering, and observing systems needed to interpret the results and determine whether any subsequent step would be justified. Between Phases 1 and 2, and between Phases 2 and 3, there are gates: checkpoints where the full body of evidence is assessed against criteria set in advance. One phase doesn’t automatically lead to the next. We haven’t tried to outline the specific criteria – that’s not our role – but the essential questions are: Are projections meaningfully better constrained? Do they point to any scenario that would reduce climate risks? Would the next phase resolve uncertainties that still matter, and are the systems needed to conduct it responsibly in place? All three have to be true. Otherwise, the work is repeated or the program stops. Reflective does not decide whether a gate has been met. The Roadmap proposes evidence tests. The final criteria, and the process for judging them, should be agreed by independent stakeholders before the next phase begins. Technical evidence is necessary for a legitimate decision. It is not sufficient, and this Roadmap does not replace the governance process that would have to make that decision.
Reducing critical uncertaintiesBefore we could scope the research program, we first had to identify which uncertainties actually mattered to the decision. So we built the Uncertainty Database first. Over the past year, we worked with researchers around the world to translate those uncertainties into concrete research activities. An activity earns a place on the Roadmap only if it materially reduces an uncertainty that matters for the decision, or is necessary for work that does. Interesting is not an inclusion criterion. Cooling efficacy—how much the planet cools per tonne of sulfur injected—is one of the most consequential uncertainties. Today’s models disagree by roughly a factor of two on how much cooling a given injection would produce, and about 40 percent of that disagreement traces to how aerosol particles form and grow in the stratosphere. Phase 1 is designed to address that gap. Its outdoor research would involve releases totaling roughly 10 tonnes of sulfur dioxide, at a scale far too small to affect surface climate. The purpose is to observe directly how sulfate aerosols form and grow in the stratosphere. We estimate that Phase 1 could reduce uncertainty in cooling efficacy by roughly 25 percent. By the end of Phase 2, that reduction could reach roughly two-thirds. Narrowing that gap does more than sharpen one number. Everything that depends on how much sulfate is in the stratosphere, and in what form, becomes clearer with it—including potential effects on the ozone layer, air quality, and the climate more broadly. The Roadmap’s estimates become less precise the further out an activity sits. That is unavoidable. The design, cost, and value of later work depend on what earlier research finds. The Roadmap gives ranges rather than pretending to have precision we do not have, and it is designed to change as the evidence does. Why coordination changes the timelineOne of the Roadmap’s most important outputs is an estimated timeline of (at least) ten years to a well-informed decision. Ten years is not a target or a shortcut. It is a critical-path calculation: the shortest route through a long chain of activities that depend on one another. A Phase 1 experiment cannot happen simply because someone funds an airplane. The plume-tracking instruments, measurement strategies, multi-scale modeling tools, and other prerequisites all have to be ready first. If one is late, the experiment shifts. If the experiment shifts, the model improvements that depend on it happen later. The whole program slips. The Roadmap maps those dependencies across the field: what has to happen first, what can run in parallel, and where a single stalled activity delays everything after it. More funding can start more of the program at once. It cannot make the underlying science instantaneous, move an activity ahead of the findings it depends on, or shortcut a gate. While Reflective is actively working on some of these activities, this is not a work plan for Reflective. Many are already under way at institutions around the world. Others still need funding or teams. By putting them in one place—with their dependencies, costs, timelines, and expected contribution—we want researchers and funders to see not only which work matters, but how the pieces fit together. Best case, funders use the Roadmap to identify which activities would actually move the program forward. Governments use it to see what they can fund domestically, where international coordination would help, and where pooling resources might make sense. Researchers use it to understand where their work fits and what it enables. And over time, the field can return to the Roadmap and ask a simple question: are we actually getting closer to an answer? A plan concrete enough to argue withFor each activity, we consulted researchers with relevant expertise, then cross-checked the proposed scope, cost, and timeline against in-house expertise and public information. We have almost certainly gotten some things wrong. This is a public proposal, not a consensus document. It covers technical research in the physical sciences and engineering. It does not map the equally necessary work on governance, politics, legitimacy, or public consent. That is why we are making it public now, with a way to provide feedback on every page. We expect the activities, estimates, dependencies, and even the structure of the plan to change as the field responds and the science advances. We would rather publish a plan concrete enough to argue with than another document everyone can endorse because it commits to nothing. Sunlight reflection may never be used. Research may show that its risks are too great, its benefits too uneven, or its practical requirements too difficult to meet. That would be valuable knowledge—and far better than leaving the question unresolved until climate pressure is greater, time for careful deliberation is shorter, and the outcome is shaped by whoever moves first. Reflective is agnostic about whether SAI should ever be deployed. We are not agnostic about decisions being made in the dark. A credible “no” should be reachable early. A responsible “yes” should be hard. This Roadmap is our first public proposal for how to make both possible. Thanks for reading Upon Further Reflection! Subscribe to receive notifications of new posts.
© 2026 Reflective
|
