According to a new study from Columbia University in the United States, stratospheric aerosol injection technology may cause large-scale side effects depending on the location, time and materials used. From disrupted monsoon systems to supply chain constraints to uncertain chemical reactions, the technology faces formidable obstacles.
An idea once considered fanciful—cooling the Earth by spreading sunlight-reflecting particles in the upper atmosphere—has become a serious topic in climate science. The method, known as stratospheric aerosol injection (SAI), aims to combat global warming by simulating the natural cooling effects following volcanic eruptions. Hundreds of studies have modeled how such a system might function in theory. But researchers at Columbia University warn that proponents of the concept are ignoring the huge uncertainties, technical challenges and risks it could pose in practice.
"Even though simulations of SAI in climate models are very sophisticated, they are necessarily idealized," said V. Faye McNeil, an atmospheric chemist and aerosol scientist at the Columbia Climatological Institute and Columbia School of Engineering. "Researchers are simulating perfect particles of perfect sizes. In the simulations, they put exactly the amount they want where they want it. But when you start thinking about where we actually are, compared to that idealized situation, there's a lot of uncertainty in those predictions."
"If you try to implement this technology, there's a range of things that could happen - we think the range of possible outcomes is much wider than anyone has previously realized."
In the study published in Scientific Reports, McNeil and his collaborators explored the physical, political and economic barriers that could complicate SAI deployment. They compiled findings from previous studies to better understand how different design choices—such as timing, altitude, and injection location—affect Earth's climate response. Even small differences in how and where aerosols are released can drastically change the results.
Among many variables, latitude is one of the most important. For example, injecting particles over the poles could disrupt tropical monsoon systems, while focusing near the equator could disrupt the jet stream and alter interhemispheric heat circulation.
"It's not just a matter of pumping 5 teragrams of sulfur into the atmosphere. Where and when it's implemented matters," McNeil said. The changes suggest that if SAI is done, it should be done in a centralized, coordinated way. However, given geopolitical realities, researchers say this is unlikely to happen.
Model studies to date have focused almost exclusively on SAI methods using sulfate-rich gases, similar to those formed when volcanic plumes oxidize and condense in the stratosphere. Volcanic eruptions have cooled the Earth in the past: When Mount Pinatubo erupted in 1991, for example, the planet's temperature dropped by nearly 1 degree Celsius in the following years. This incident is often cited as proof of principle of how SAI works.
In addition to ground cooling, SAI can have adverse consequences, both expected and unintended. For example, the eruption of Mount Pinatubo also disrupted the Indian monsoon system, causing less rainfall in South Asia and causing warming and ozone depletion in the stratosphere. Using sulfates for SAI may pose similar risks or additional environmental concerns, including acid rain and soil contamination. These concerns prompted a search for other aerosol components of SAI.
Proposed mineral alternatives include calcium carbonate, alpha alumina, rutile and anatase titanium dioxide, cubic zirconia and diamond. Consideration of alternatives has focused primarily on their optical properties, but other factors have been neglected.
"When scientists discuss candidate materials for use in aerosols, they rarely consider how practical constraints might limit your ability to actually inject large amounts of these materials every year," said Miranda Harker, an aerosol scientist at Columbia University and the paper's lead author. "Many of the proposed materials are not particularly abundant."
Diamonds are optically good for the task, but there simply aren't enough of them. As for cubic zirconia and rutile titanium dioxide, supply may be able to meet demand, but the Columbia team's economic modeling suggests that increased demand will strain the supply chain and make them more expensive. There is sufficient supply of alpha alumina and calcium carbonate to absorb demand without pushing prices to prohibitive levels—but as with other candidate materials, dispersing them involves serious technical challenges.
At the tiny submicron particle sizes required for SAI, all mineral substitutes tend to aggregate into larger aggregates. According to the researchers' calculations, these aggregates are less effective at reducing sunlight than individual particles, and their climate impact is even less known. "Instead of these perfect optical properties, you get something much worse," Haack said. "Compared to sulfate, I don't think we're going to see the kind of climate benefits that have been discussed."
According to Columbia University researchers, each real-world challenge—from how to implement SAI to the types of particles to use—adds new layers of uncertainty to an already unpredictable idea. They argue that these complexities must be recognized before the deployment of stratospheric aerosol injection is seriously considered.
"When you look at solar geoengineering, it's all about risk trade-offs," said Gernot Wagner, a climate economist at Columbia Business School who works closely with the Climate Institute. Given the complex realities of SAI, he said, "it's not going to happen the way 99 percent of these papers model it."
Co-authors on the study also include Daniel Steingart, co-director of the Columbia Electrochemical Energy Center.
The study, published in the October 21, 2025, issue of the journal Scientific Reports, is titled "Engineering and logistical considerations add practical constraints to stratospheric aerosol injection strategies."
Compiled from /ScitechDaily