Publicado: 15 setembro 2026 12:03 -03
Benjamin Redmond Roche, Visiting Research Fellow, Solar Radiation Management Research Governance & Independent Consultant, UCL
No, President Joe Biden did not engineer Hurricane Milton. No, the ice-crystal contrails left behind high-altitude aircraft are not “chemtrails” sprayed by governments to control the weather or manipulate populations. No, state-run cloud seeding did not cause the record floods in Dubai.
Yet claims like these are widespread online and increasingly find their way into political rhetoric. This blurs the distinction between conspiracy theories about weather manipulation and the very real scientific debate over solar geoengineering.
Solar geoengineering, also known as solar radiation modification (SRM), would try to temporarily cool the climate by altering the amount of energy entering or leaving the Earth system. Outdoor field experiments are particularly contentious, with researchers and other experts divided over whether they should proceed at all and, if so, under what conditions.
The best-known proposal, stratospheric aerosol injection, involves dispersing sulphur dioxide high in the atmosphere, where it can form reflective particles known as aerosols. These can remain in the stratosphere for long periods, broadly mimicking the cooling effect of major volcanic eruptions.
Stratospheric aerosol injection explained
The other main proposals are marine cloud brightening, which would add sea-salt particles to low clouds over the sea to make them more reflective, and cirrus cloud thinning, which would aim to reduce the net warming effect of high-altitude clouds.
What could outdoor experiments actually tell us?
In our recent study, colleagues and I examined what outdoor SRM experiments might look like in practice. We identified plausible experiments across the three approaches and examined how their scientific purpose and regulatory requirements change with scale.
Rather than a simple “small” or “large” distinction, we found discrete phases with increasingly stringent regulatory scrutiny at each stage – think of it like a steep staircase as scale increases. I focus here on stratospheric aerosol injection, the most extensively studied proposal, as it best illustrates the gulf between experimentation and deployment.
Recent simulations suggest that around 8 to 16 million tonnes of sulphur dioxide would need to be injected each year to produce 1ºC of cooling. An interactive simulator lets you experiment with these numbers yourself.
Proposed outdoor research could start with vastly smaller releases. Our study identified a range of increasingly large plausible trials, and the questions each could answer. The smallest trials could examine aerosol formation or compare how different particles behave in the stratosphere. Larger releases, involving tonnes of sulphur dioxide, could track how an aerosol plume evolves under different atmospheric conditions, providing evidence to test models of how particles spread and affect sunlight.
At the largest scale we considered, around 1,000 tonnes of sulphur dioxide, the resulting aerosol plume begins to approach the size of the individual 3D grid cells that climate models use to represent the atmosphere. This could allow computer predictions to be compared more directly with observations from an experiment.
To put these numbers into context, a Boeing 777 undertaking a transatlantic flight releases in the region of 80–100kg of sulphur dioxide, while 1,000 tonnes is comparable to the annual sulphur dioxide emissions of a mid-sized US coal-fired power station. These are comparisons of quantity, not environmental effect: a coal plant releases its sulphur dioxide much closer to the ground, alongside greenhouse gases such as CO₂. They nevertheless help make the physical scale of proposed experiments tangible.
The UN Environment Programme notes that small field experiments may reduce uncertainty around particular atmospheric processes but cannot resolve much larger uncertainties about the effects of a full-scale deployment. SRM researchers broadly accept this limitation: experiments are instead intended to answer specific questions that can improve computer models.
The debate goes beyond the science
Critics object more strongly to the research pathway itself. They argue that successive stages of research could generate technical and institutional momentum towards progressively larger trials, and eventually, deployment. On this view, each successive experiment could normalise the next, building infrastructure and scientific communities that make it progressively harder to step back. Others go further, calling for a ban on public funding, outdoor experiments, patents, dedicated research programmes and deployment.
A recent paper argues that some opposition to SRM cannot be settled through scientific evidence alone. For parts of the environmental movement, the concern extends to the kind of social and political order it might sustain: technological management of the climate, continued reliance on existing economic systems and less pressure for structural change. In this regard, evidence that deliberately managing solar radiation could work may in fact strengthen opposition, because it makes that future more plausible. More field research cannot resolve disagreements rooted in competing values.
Assessing SRM therefore involves weighing the risks of intervention against the risks of continued warming, including whether governments and international institutions could reliably oversee an intervention that might need to be sustained over time. One useful analogy is that it could function “like a tourniquet”: it would not cure the underlying issue, but it could reduce acute harm while treatment takes effect. In climate terms, rapid emissions reductions and carbon removal remain fundamental to tackling climate change, while SRM research examines whether temporary cooling could reduce acute climate risk as those measures scale up.
Where does research end and deployment begin?
That separation between research and deployment is already reflected in policy. European scientific advice recommends a Europe-wide moratorium on using SRM while simultaneously calling for research to continue under rigorous and ethical conditions. A decision by the Convention on Biological Diversity similarly leaves room for small-scale scientific research conducted under specific conditions. The problem is that neither “small-scale” research nor deployment has a clear, agreed boundary. Clearer definitions would make case-by-case scrutiny of field experiments more straightforward.
Ultimately, SRM may never be deployed, and further research may show that it should not be. But decisions about whether particular experiments are worth conducting should be based on what those experiments actually involve and what they can establish. Dismissing these discussions on behalf of the public is not a substitute for public debate. Public debate does not require agreement, but it does people to be arguing about the same issue.
