Cloud seeding consists of dispersing particles into a cloud — most often silver iodide — that serve as a support for the formation of ice crystals, in order to trigger or increase precipitation. The technique is real, documented since 1946, and practised openly in many countries. But its measured effectiveness remains modest: it adds a few percent of precipitation under good conditions, far from the "climate control" the popular imagination ascribes to it.
Definition. Cloud seeding is a weather-modification technique (not climate modification) that injects glaciogenic nuclei — silver iodide, dry ice, or salt — into clouds that are already present and suitable, to facilitate the condensation and the fall of water. It is a local, occasional intervention, not a remote command of the weather.
The story begins in a laboratory, not in a conspiracy. In November 1946, the chemist Vincent Schaefer, working at General Electric under the direction of Irving Langmuir (Nobel Prize in Chemistry, 1932), demonstrated that by dispersing dry ice — solid CO2 — into a supercooled cloud, one triggers the formation of ice crystals and then precipitation. Shortly afterwards, his colleague Bernard Vonnegut established that silver iodide produces the same effect: its crystalline structure mimics that of ice so closely that supercooled water droplets aggregate on it and freeze. Cloud seeding was thus born as a published and patented discovery, in the journal Science.
This point of origin matters for understanding the nature of the process: it was never secret. It was announced, described, patented, and reproduced. The whole operation rests on a known principle of cloud physics — ice nucleation — and not on a hidden chemistry.
The mechanism presupposes an already existing and suitable cloud: one does not make a cloud out of a clear sky. In a cold cloud, part of the water remains liquid below 0 °C (the so-called "supercooled" state) for lack of a support on which to crystallise. Seeding provides that support.
The right conditions are therefore required: a cloud present, favourable humidity and temperatures, the right moment. Without an appropriate cloud, seeding does nothing. It is an assistance to a natural process, not its creation.
This is the point most often glossed over, in both directions. For a long time, the effectiveness of seeding was hard to prove, precisely because one cannot compare the same cloud with and without treatment: one cannot know what it would have done on its own. Operations multiplied without a direct and rigorous demonstration of their effect. The 2024 report of the Government Accountability Office (GAO) puts it plainly: in several studies, the estimated effect was not distinguishable from zero with a high degree of statistical confidence.
The SNOWIE project, conducted in Idaho from 2017 and published in PNAS in 2020 (Friedrich et al.), marked a milestone: thanks to modern radars and instruments, researchers were able to directly observe the chain "seeding → crystal formation → snowfall" and quantify the additional snow attributable to the treatment. The honest conclusion: the effect exists and is now measured, but it remains modest — a few percent of additional precipitation under good conditions — and hard to generalise to all weather situations.
In other words, cloud seeding genuinely works, but marginally. It can scrape together a little water where exploitable clouds already exist. It can neither break a drought at will, nor manufacture a storm, nor "set" the weather of an entire region.
Far from secret, cloud seeding is the object of declared, contracted, and sometimes subsidised operations, for specific uses. In the United States, the Weather Modification Reporting Act of 1972 even requires that any operation be reported to NOAA at least ten days in advance; NOAA records these activities but does not conduct them itself.
These programmes are public and regulated. The World Meteorological Organization (WMO) notes, moreover, that weather-modification operations take place in more than fifty countries. Their existence proves that it is true that "we modify the weather" — but in a limited, local, and declared manner, which has nothing to do with a hidden poisoning of populations.
Because it is true that humans modify the weather, the mind sometimes concludes too quickly that "the climate is being controlled in secret," or even that aircraft trails are chemical spraying — the famous "chemtrails." The leap is enormous and illegitimate. Cloud seeding is a local, declared intervention, of modest effectiveness, on already formed clouds. The white trails behind aircraft are condensation trails (contrails): water vapour that freezes at altitude, a known physical phenomenon. NOAA, which is required by law to track weather-modification activities, has publicly refuted the thesis of a secret spraying programme.
Acknowledging the modest fact — yes, we have been seeding clouds since 1946 — is precisely what allows the paranoid extrapolation to be refused. Reality does not need to be exaggerated to be interesting.
DOCUMENTED FACT. On 19 December 2024, the Government Accountability Office (GAO), the U.S. Congress's audit arm, published a technology assessment titled "Cloud Seeding Technology: Assessing Effectiveness and Other Challenges" (GAO-25-107328). Its central finding is sober: the benefits of cloud seeding are "unproven," and federal involvement remains minimal.
DOCUMENTED FACT. The report quantifies the picture: nine U.S. states currently use cloud seeding, while ten have banned or considered banning it. In the studies GAO reviewed, estimated additional precipitation ranged from 0 to 20 percent — a spread that itself shows how hard the real effect is to evaluate, for lack of reliable, standardized data.
DOCUMENTED FACT. On silver iodide safety, GAO is cautious: the few recent studies reviewed suggest no environmental or health concern at current use levels, but the effect of much wider use is unknown. The report identifies five policy options (status quo, targeted research, evidence-based operations, better monitoring, education) without mandating any.
WHAT THE REPORT ESTABLISHES — AND DEBUNKS. GAO explicitly flags a key point: the public often confuses cloud seeding with geoengineering, which affects the climate on much longer time scales. Cloud seeding locally enhances precipitation only when the right clouds are already present; it is neither climate control nor "chemtrails." The 0–20% range confirms a real but modest and uncertain effect — the opposite of weather steered at will.
Is silver iodide dangerous? At the quantities used in operational seeding, the available studies have not revealed any significant health or environmental effect; the WMO underlines this in its review, while recommending monitoring should the quantities used rise sharply. The silver is dispersed in very low concentrations over vast areas, well below the usual health thresholds. The subject is monitored, but there is, to date, no proof of toxicity at the doses used.
Can seeding cause or stop a drought? No. It can only modestly increase precipitation where exploitable clouds already exist. Without suitable clouds, it is inoperative; it does not create rain from a dry sky and does not "command" the regional weather.
Is cloud seeding geoengineering? In the broad sense, yes: it is a deliberate modification of an atmospheric process. But it is local weather modification, to be clearly distinguished from planetary climate geoengineering (solar radiation management, CO2 removal), whose stakes and risks are of an entirely different order. The WMO itself explicitly distinguishes weather modification, local to regional, from climate intervention at the global scale.
To place cloud seeding in the overall picture — and to learn to separate the real science from the "chemtrails" myth — see our pillar article on geoengineering: the science versus chemtrails. The full investigation, marked out level of certainty by level of certainty and supported by primary sources (from Schaefer 1946 to the SNOWIE project in 2020), is detailed in the ebook Geoengineering Without Consent.
At the quantities used in operational seeding, the available studies have not revealed any significant health or environmental effect; the WMO underlines this in its review, while recommending monitoring should the quantities used rise sharply. The silver is dispersed in very low concentrations over vast areas, well below the usual health thresholds. The subject is monitored, but there is, to date, no proof of toxicity at the doses used.
No. It can only modestly increase precipitation where exploitable clouds already exist. Without suitable clouds, it is inoperative; it does not create rain from a dry sky and does not "command" the regional weather.
In the broad sense, yes: it is a deliberate modification of an atmospheric process. But it is local weather modification, to be clearly distinguished from planetary climate geoengineering (solar radiation management, CO2 removal), whose stakes and risks are of an entirely different order. The WMO itself explicitly distinguishes weather modification, local to regional, from climate intervention at the global scale.
Published on 19 December 2024, the GAO report describes the benefits of seeding as "unproven": the studies reviewed estimate the precipitation gain at between 0 and 20%, with data deemed insufficient to settle the question. Nine US states use it, ten have banned it or considered banning it, and federal involvement remains minimal.
Dossier : Géo-ingénierie & modification du climat
MINI12: La Géo-Ingénierie Sans Consentement
Accueil · Collection · Journal · Dossiers · Sources