cloud seeding – A study published June 24 in PLOS Water argues that small, carefully timed cloud-seeding could shift major weather systems—citing examples like Superstorm Sandy and California’s atmospheric river flooding. But scientists and critics warn there is currently no
When Atlantic hurricane season starts to feel less like a forecast and more like a countdown, a new study in the journal PLOS Water is pushing a provocative idea into the open: use technology to nudge storms off course.
The proposal is tied to cloud seeding, a decades-old practice that uses supporting technologies for research and operations. The study’s authors argue that if cloud seeding could be “supercharge[d]” with high-tech data and analysis. humans might disrupt huge weather systems and protect densely populated areas.
The idea arrives with a warning label attached—especially for hurricanes. Critics say it’s far-fetched and that proving any effect on large-scale severe weather remains exceptionally difficult.
The study’s timeline, and what it claims could change
The research appeared in PLOS Water on June 24. It suggests that small, carefully-timed cloud seeding operations applied days before the peak of an extreme weather event could have changed the outcomes of multiple disasters.
The study authors point to three examples: a proposed track shift of about 300 miles that they say could have kept Superstorm Sandy in 2012 from missing New York City; raising the low temperature of the 2021 Texas freeze by about 18 degrees Fahrenheit; and reducing the amount of precipitation carried by a 2022 atmospheric river that caused flooding in California by about 5%.
The authors argue the growing impact of weather extremes on society means traditional approaches—specifically dams, levees, and insurance—may not be sufficient on their own to address widespread consequences.
Even the study’s supporters acknowledge the leap from concept to control
To explain how small interventions could produce meaningful changes, one of the study’s co-authors drew a comparison to “jiu-jitsu.”
Upmanu Lall of Arizona State University said in an email that the basic principle of jiu-jitsu is achieving “maximum efficiency with minimal effort,” relying on leverage, proper weight distribution, and momentum rather than brute strength.
He and the other authors argue that humans could potentially leverage a few tools already at their disposal to change the course of huge weather events.
But outside experts say the science is not there yet.
A sharp response from researchers: “NO scientific evidence”
Katja Friedrich. an assistant professor in the department of atmospheric and oceanic science at the University of Colorado. said via email that there is currently “NO scientific evidence” supporting the idea that existing cloud-seeding technologies can modify large-scale weather systems—especially severe weather systems like hurricanes and thunderstorms—driven by large dynamic forcing.
Her concern cuts to the core of the study’s promise: not whether cloud seeding can affect rain or hail at small scales, but whether it can meaningfully steer storms once they are already operating under the physics of large-scale atmospheric forcing.
What cloud seeding can do—and what it can’t do alone
Cloud seeding is not new. The Government Accountability Office (GAO) describes it as a decades-old approach to modifying weather that uses a range of supporting technologies for research and operations.
In the U.S., nine states are currently using cloud seeding. Another ten states have banned it or have considered banning cloud seeding or weather modification in general. Florida is included among the bans, having banned it in 2025.
The most common uses are to increase precipitation or suppress hail, typically by adding tiny particles of silver iodide, according to the National Oceanic and Atmospheric Administration.
The study’s own logic, however, depends on much more than dropping particles and hoping for the best.
Lall said the theory would require a high density of observations available in near real time. AI models able to rapidly process that information. and a delivery mechanism to “identify locations that are suitable for nudging.” He also cautioned that researchers are not even sure cloud seeding is the right technology to physically nudge weather systems.
“We are still working on the best technology for delivering the nudge,” he said. “A very large-scale cloud seeding effort at the right time and place, is a possible mechanism, but not necessarily the definitive one.”
And even before the technology question, there is the measurement problem.
Previous research has proven tricky for the same reason critics are now wary. Kara Lamb, an associate research scientist at Columbia University, said in an email that it is “very difficult to separate the effects of human intervention from the natural variability of hurricanes.”
Lamb’s warning points to what could become the central challenge if any future effort claims success: hurricanes are not lab experiments. Their behavior varies widely, and nature can blur cause and effect.
One storm might be too ethical to steer
The debate also runs into ethics and control.
In a best-case scenario, the concept of steering a natural disaster carries an ethical dilemma the authors say they are not solving in the paper itself.
Saving New York City from a direct hit could, theoretically, put people in New England in harm’s way. The study authors frame that as a question for others in the future.
Lall said the effort is currently focused on developing “the science and engineering capacity,” and he acknowledged ethical and control issues, but said the team is not focused on them “at this point.”
Critics say even if the physical hurdle could be overcome, the fallout might still stop the idea.
Robert Rauber, a director emeritus in the Department of Climate, Meteorology and Atmospheric Sciences at the University of Illinois, said via email that “even if humans could move storms, the political, international and legal fallout would be so great that it would never happen.”
A pattern of blame—and conspiracy narratives—already travels faster than evidence
Cloud seeding has become fertile ground for conspiracy narratives after real disasters.
After the tragic Texas floods of 2025 that killed dozens of people. many of them children. viral claims alleged a nearby cloud-seeding operation caused the disaster. according to CBS News. Meteorologists said there was “zero evidence” of that claim. and they emphasized that seeding can only modestly enhance rainfall—not create catastrophic floods.
Internationally, after record rainfall and flooding in the United Arab Emirates in April 2024, social media widely blamed cloud seeding, according to AFP. Scientists and officials said no seeding occurred during the storm, and they pointed instead to natural weather systems and climate change.
Those episodes show how quickly public narratives can harden around a plausible-sounding technology—regardless of what the data can prove.
The through line in the debate right now is not whether cloud seeding exists. It does. The disagreement is whether it can do what this study suggests—steering hurricanes and other large-scale severe events—with enough confidence to justify action.
For now. the paper pushes toward a future where high-density near real-time observations. AI modeling. and new delivery mechanisms could make “nudging” storms technically feasible. But until evidence clears the gap critics cite—especially the lack of proof Friedrich described—cloud seeding remains. for many scientists. a tool with limits far narrower than the study’s most dramatic examples.
cloud seeding hurricane control PLOS Water Michael Brennan weather modification silver iodide AI weather modeling Superstorm Sandy Texas freeze atmospheric river flooding ethics