WATER POLITICS

Bill Kornell has spent most of his half-century-long career flying into bad weather. A former bull-riding champion, the sinewy 80-year-old has been a pilot since the 1960s, when he realised that travelling to far-flung rodeo towns across the American West was more efficient by plane than by car. After an injury in the late 1970s, Kornell left the bull-riding circuit and took a job as a bush pilot, ferrying supplies and commuters deep into the Alaskan interior. Flying in the Arctic presented a host of challenges: bad weather, freezing temperatures and poor visibility.

Those hazards pale in comparison with the rigours of his current job. For the past three years, Kornell has been trying to make it snow. When he’s called to action by his employer, North American Weather Consultants, Kornell hops into a two-engine Cessna 414. Minutes later the tiny aeroplane, its wings and fuselage covered in dozens of ungainly metal tubes, is soaring through bucking turbulence above Salt Lake City, Utah. Kornell does not avoid the roiling storms that mass along the high peaks of the Wasatch mountains, but flies directly into them, releasing a chemical solution into the clouds.

Proponents of this technique, known as cloud-seeding, believe that it can increase precipitation by around 10%. In the drought-stricken American West, cloud-seeding is used in the hope of making snow, which, when it melts, feeds rivers and reservoirs. The aim is to trigger a stormfront to release snow over an entire mountain range.

Kornell lives in a small trailer on the outskirts of Logan, a town in the foothills of the Wasatch mountains, north of Salt Lake City. During cloud-seeding season – late October to early April – he checks in daily with Todd Flanagan, a local meteorologist, who tracks storms rolling into the region. “It’s a waiting game,” Kornell told me, in a nasal western drawl. “I go to the gym, otherwise I’d go crazy with all the sitting around.”

An array of meteorological conditions must be met – the right temperature, humidity, wind direction and speed – before Flanagan gives Kornell the go-ahead to fly. But the critical component he’s looking for is “supercooled water”. For water to freeze it needs a “seed” – something solid around which ice can form. In clean air where seeds are rare, water droplets can remain liquid well below 0°C. The goal of cloud-seeding is to turn that supercooled water into ice, which will then fall as snow. Flanagan pores over forecasts and sifts through pilot reports of icing on aircraft in the region, an indication that supercooled water might be lurking.

Once the meteorological tumblers align, Flanagan calls Kornell – it could be 3am or 3pm. Kornell then drives to the airport, fires up the propellers of his Cessna and takes to the sky. “When you get into a cloud, you have no depth perception, no lateral perception,” he said. Powerful winds sometimes accompany these storms, tossing his little plane to and fro in a high-altitude bull ride. “Your normal senses that you use to survive in the daytime and night-time – you don’t have that,” he said. “You have to transition to the aeroplane’s instruments. Then you have to believe that instrument is correct. And you stay with that until you get back out of the clouds.”

When Kornell reaches the right altitude, his plane delivers a burst of silver iodide. The compound’s molecular structure closely resembles that of ice and therefore makes near-perfect seeds for ice crystals to form on, even at temperatures well above those needed to create the conditions for snowfall.

The chemical is delivered in cylindrical flares, which emit particles of silver iodide as they burn. Twenty-four of these incendiaries – which resemble oversize Roman-candle fireworks – are secured to each wing of his plane. Each burns for four minutes. All Kornell has to do is push a button.

The automation allows Kornell to focus on the main task at hand – keeping his Cessna in the air. Among the many hazards Kornell faces in the clouds, one stands out. “Ice is your biggest enemy,” he said. Large aeroplanes have mechanisms, such as heating elements in their wings, to prevent ice from forming. But Kornell’s plane is in constant danger of icing up – and this could mean stalling and plummeting from the sky. “There is a slim margin between living and dying,” Kornell said. “When the emergency happens, you have no room for error, otherwise, you’re gonna die. It has to be done perfectly.”

To Kornell, the risks are worth it – not only for the pay cheque, but because he feels like he is doing his part to allay the region’s mounting water crisis. He recalled with fondness some flights he took outside Saint George a few years earlier that preceded a series of major snowstorms. “We produced enough extra water that they actually asked us to quit flying,” he said.

Water-strapped regions across the world are using cloud-seeding to try to reduce the worst effects of climate change. According to the World Meteorological Organisation, 52 countries had cloud-seeding programmes in 2014. In Australia Snowy Hydro, a utility company, uses cloud-seeding generators – machines that release puffs of silver iodide into the air – to increase snowfall that will, in turn, feed rivers that supply its hydroelectric dams. The United Arab Emirates (UAE) has used the technology for more than 30 years in an attempt to bolster its water supply for a rapidly growing population. In 2016 Chinese officials announced the Sky River Project: an enterprise with the goal of diverting precipitation from the Yangzi river basin to the parched tributaries of the Yellow river, which supplies drinking water for 420m people.

Several western states in America have tried it out. In the beleaguered Colorado river basin, where desert cities such as Las Vegas and Phoenix rely on dwindling supplies from the ever-shrinking river, water agencies devote millions of dollars a year to enhancing snowfall. Idaho now spends $3.9m annually on cloud-seeding, up from $500,000 in 2017. In Utah the state legislature approved $18m for cloud-seeding operations in 2025, up from $350,000 only two years ago.

Supporters of cloud-seeding insist it’s the most cost-effective way to increase water supply. Established methods – dams, canals, reservoirs – are expensive. The Delta Conveyance Project in California, which would involve the construction of a 45-mile-long tunnel to syphon water from rivers in the north of the state and send it to the drier and more populous reaches in the south, is projected to cost $20bn.

Yet questions remain over the technique’s effectiveness and its potential for accidental climate disruption. Some reckon the benefits of the technology have been oversold. “Science never developed the ability to back a lot of these claims,” said Jeffrey French, a professor of atmospheric science at the University of Wyoming. James Rodger Fleming, a historian of the field with a sceptical stance, said that with ground-based methods such as generators, “It wasn’t even clear that the silver iodide was getting up into the clouds.” Funding for research dried up in the 1980s and 1990s only to re-emerge more recently as worries about climate change increased. The Sky River Project was criticised by several prominent Chinese scientists, including Lu Hancheng, a climate professor at the National University of Defence Technology. “It’s unbelievable”, he told the South China Morning Post in 2018, “that a project that has neither scientific evidence nor technical feasibility was approved.” Eight years after the plan was unveiled, it’s unclear how much headway China has made.

Some scientists have suggested that seeding clouds in one place could result in the loss of precipitation in other places. That this is even a possibility could be enough to cause a slew of lawsuits in places such as the American West, where water resources are divided based on interstate agreements. In less stable parts of the world, it could even lead to wars.

In the 1940s Bernard Vonnegut, a scientist working for General Electric (GE) in Schenectady, New York, was investigating the problem of supercooled water damaging aircraft when it turned to ice. The goal of the research, designated “Project Cirrus”, was to discover a compound that could transform that substance to snow and ice.

The scientists worked out that by breathing into a freezer filled with “dry ice” – frozen carbon dioxide – they could trigger the formation of tiny flecks of water ice. The dry-ice particles had transformed the water vapour in their breath into individual ice crystals. To see if this might have practical applications the Cirrus team took to the air and dropped 2.7kg of dry-ice pellets over Mount Greylock in Massachusetts. Irving Langmuir, a Nobel laureate and the lab’s associate director, who watched the flight, described “a radical transformation of the cloud” from which “streamers of snow began to pour out.”

Yet dry ice proved to be an unwieldy cloud-seeding agent. It is not only heavy, but sublimates to carbon-dioxide gas, which, in the closed confines of an aeroplane fuselage, might asphyxiate the crew. So Vonnegut plunged into chemistry textbooks, looking for other seeding materials and came across a substance with a crystalline structure uncannily similar to that of ice: silver iodide.

American military planners hoped to use silver iodide to control the weather for strategic ends. In his memoirs, Edward Teller, father of the hydrogen bomb, recounts a meeting with Langmuir at Los Alamos National Laboratory in 1947. According to Teller, Langmuir “talked so much about the amount of damage done by a storm his seeding had caused that I began to wonder whether he saw the technique as competition to the atom bomb.”

In Vietnam the army put Langmuir’s ideas to the test, trying to prolong the monsoon season to bog down the North Vietnamese army. The secrecy of the operation (code name, “Popeye”), made it hard for independent researchers to evaluate its effectiveness in real time. In 1967 a government memo claimed that the programme had been “outstandingly successful”, with 82% of seeded clouds producing rain shortly after, “a percentage appreciably higher than normal expectation in the absence of seeding”.

Bernard’s brother Kurt, a budding novelist who worked for GE as a publicist, had a front-row seat to the research. The experiments would become the basis for his novel, “Cat’s Cradle”, which was published in 1963. The book tells the story of a Nobel-prizewinning nuclear researcher named Felix Hoenikker who develops a substance, “ice-nine”, that is capable of transforming water to ice at room temperature. By way of a convoluted plot twist, ice-nine finds its way into the environment and brings on global cataclysm as it freezes all the world’s water. It’s a wry warning about the unforeseen consequences of meddling with nature.

Though some cloud-seeding in the American West is conducted by daring pilots like Bill Kornell, most is carried out at ground level by small, mobile, remotely operated machines. One of these is on the 8,600-foot summit of Ward Peak, the highest point of the Alpine Meadows ski resort in the Sierra Nevada mountains in northern California.

On a sunny afternoon in June 2023, I boarded a chairlift at Alpine Meadows. Even in early summer, the slopes were still blanketed in white powder, the result of an unusually snowy winter. I skied a few hundred feet down the mountain, then stumbled in my boots over rocky ground towards a green metal box the size of a shed where I met Frank McDonough, the chief cloud researcher at the University of Nevada’s Desert Research Institute (DRI).

McDonough, compactly built and sporting a tidy moustache, explained that this was one of 30 machines operated by the DRI across California’s Sierra Nevada mountains and nearby ranges in Nevada. Its purpose, he said, was to enhance snowfall in the Truckee river basin, some 20 miles away. This wasn’t about making more powder for skiers. Half a million people in and around the city of Reno, Nevada, depend on the river for drinking water, and climate change threatens the snowpack that feeds into the river.

The generator’s 55-gallon (210-litre) tank, McDonough said, contains sufficient silver iodide to coax more than 5,000 acre-feet (6.2m cubic metres) of water from the clouds – enough to supply roughly 10,000 homes with water for a year. “This place gets absolutely crushed with water,” he added, pointing to the billowing snow clouds above our heads.

We should take McDonough’s estimates with a pinch of salt. Though scientists have been able to show that silver iodide can, under certain circumstances, stimulate snowfall, cloud-seeding’s effects remain stubbornly difficult to measure. “Once you seed a cloud you don’t know what would have happened if you didn’t seed it,” said Sarah Tessendorf, a project scientist at the National Centre for Atmospheric Research (NCAR) in Boulder, Colorado.

The unpredictability of local weather patterns, notably in mountainous regions, also challenges researchers’ ability to devise controlled experiments, Tessendorf said. And collecting high-quality data could take decades. Each winter “You might be able to seed 30 [storms] per year,” Tessendorf told me. But researchers have to analyse thousands of storms to tease out the effect.

Over the past decade, supercomputer-powered models such as NCAR’s WRF (pronounced “wharf,” an acronym for Weather Research and Forecasting) have advanced greatly and helped scientists run simulations that mimic controlled experiments. But even the best models can predict only the “average” behaviour of a cloud, according to French. “It might be the outlying behaviour that occurs 1% of the time, or 0.01% of the time, that actually drives the outcome.”

Though we were outside the official season for seeding on Ward Peak, McDonough wanted to show me how the generator worked. Some of the machines can be controlled via a smartphone app, but this one required a phone call to the DRI’s headquarters in Reno, where a technician fired up the generator remotely.

A flame ignited in a small smokestack projecting from the top of the generator. It emitted only a faint hiss, like the whisper of a backyard grill on a summer evening. Once the cauldron reached a pre-set temperature, a nozzle sprayed fine jets of a solution of silver iodide suspended in acetone into the flame. The reaction didn’t produce any smoke or burning smell – just invisible particles of silver iodide that drifted into the clouds.

McDonough and I peered northwards, into the Donner Lake basin, where, in the winter of 1846, 42 pioneers perished. An early and unintentional form of cloud-seeding, McDonough has speculated, could have sealed the Donner party’s fate. During a cold snap, great billowing clouds of chimney smoke poured from nearby cities and towns. That smoke, suggests McDonough, may have served as a nucleating agent, triggering massive blizzards and record snowfalls across the mountains.

In the winter of 2022 the amount of water detected in the snow was higher than average, which meant that cloud-seeding wasn’t allowed. Despite the safeguards, a series of blizzards and atmospheric river storms pounded California in the spring of 2023, causing extensive flooding in parts of the Central Valley. It is just this sort of year – with copious snowfall taking place without seeding – that makes it difficult to say whether the technology is worth it.

Art Rangno, a retired atmospheric scientist who once worked at the University of Washington, believes the positive effects of cloud-seeding are too marginal to justify the money spent on the technique. Along with his colleague Peter Hobbs, Rangno analysed several studies that had concluded cloud-seeding was effective. When they looked at the data, they were underwhelmed. “Not one was what the experimenters reported them to be,” Rangno told me. He pointed to a series of Israeli experiments carried out from 2013-20, which showed that a combination of ground-based and aerial seeding (using aeroplanes) increased average rainfall by just 1.8%. Those lacklustre findings prompted the Israel Water Authority to abandon its cloud-seeding programme.

Rangno echoed Tessendorf’s concerns about the difficulties of measuring cloud-seeding’s effectiveness, especially with ground-based methods. “You can’t control what concentrations of the seeding material are going to enter the clouds you’re targeting, whether the clouds have natural ice that might compete with the seeding agent, and if the seeding agent gets to them at the right place to have precipitation fall out where you want it.”

Despite the shortcomings of the evidence, he said, cloud-seeding is still a “win-win” for both private contractors and the governments that fund them, “Because they look like they’re doing something about a drought.”

Whether cloud-seeding is worth it or not, new technology is still being rolled out in pursuit of ever more efficient methods of delivering silver iodide to the skies. Jesse Juchtzer, a research technician at the DRI, is among the vanguard. His workspace, in a former warehouse at the airport in Reno, looks like a car mechanic’s workshop, filled with winches, welders, air compressors and drill presses. Scattered across the floor were several frames of square steel tubing. These, he explained, were chassis for a new prototype of a smaller, lighter, more mobile seeding generator. The plan is to put these units in rugged, remote areas that to date have been inaccessible to larger generators.

Juchtzer and his colleagues are also looking into expanding beyond snowfall enhancement into seeding summer rainstorms. This is common in the Midwest, but rare in the western states where these kinds of storms are less predictable, shorter and more violent. As these storms can be dangerous for pilots, the DRI and others are exploring the use of drones to deliver pinpoint plumes of aerosols inside thunderclouds. Juchtzer showed me a quadcopter with silver-iodide emitters on long flanges.

Elsewhere, some of the most cutting-edge experimentation involves “clean” drone technology. Rather than chemicals, one method sends zaps of electricity into a cloud, mimicking lightning. The goal is to give an electrical charge to the water molecules in the cloud, causing them to stick to one another and form droplets large enough to fall from the sky.

Despite the increase in innovation, cloud-seeding is largely unregulated. In 1976 Congress passed the National Weather Modification Policy Act, requiring “the development of a comprehensive and co-ordinated national weather modification policy”. No such policy has ever been developed, however, and rules vary widely between states. In California, for example, the practice cannot be carried out during large public gatherings or during weekend holidays. But unless neighbouring states adopt similar policies, these limits could prove futile.

The lack of a national policy, say critics, is cause for concern. Kathryn Sorensen, a researcher at Arizona State University, calls cloud-seeding “water wrangling” because of its potential to “steal” atmospheric moisture from elsewhere. As the climate is complex, small perturbations in one place can possibly lead to chaotic consequences elsewhere.

“Once I would have said, hell, you could probably take some moisture from the Mississippi river basin and no one would really care that much,” said Sorensen. “But then New Orleans ran into really serious problems [in 2023] because of drought and saline water pushed farther up into the river system.” There’s no evidence to suggest the drought was due to cloud-seeding, but Sorensen is worried we don’t yet know enough about its effects. “I’m just not sure where this mystical place is, where you can steal moisture from, and no one will care.”

In 2018 Iran’s government claimed that Israel and the UAE were using weather modification to steal its rainfall. Iran, which has its own cloud-seeding programme, levelled similar charges against Turkey last year, after photos circulated online showing Turkish mountains covered in snow while adjacent Iranian peaks were snowless.

Although proponents of cloud-seeding insist that it does not appreciably reduce the amount of water vapour in the atmosphere, some believe the mere perception that the technique is causing a reduction could be enough to cause legal problems. In 2011 Michael J. Brown, an industrial engineer, presented a hypothetical scenario in which cloud-seeding in California coincided with a diminishment of precipitation in bordering states. “If states that are farther east of California perceive they are receiving less water from the Colorado river than may occur naturally, they may request an amendment to the [Colorado river apportionment, which decides how water is allocated between seven western states] to the detriment of current California water users.”

Critics and advocates alike say expanding cloud-seeding programmes may open the door for more ambitious geoengineering schemes, such as the spraying of sulphate particles high into the atmosphere to cool the planet. Recent surveys have shown that the public is becoming more amenable to the use of weather modification and geoengineering to combat global warming. But as seductive as large-scale weather modification is, policymakers must be on guard for scientific hubris, says Fleming. “[Scientists] become attached to one-horse mechanisms, like we need to make it colder, so we need to put some more chemicals out,” Fleming said. “It’s like Jupiter or Zeus. They get this idea that they can make it rain on demand, that they can control the weather. But, of course, they can’t.”

He believes the solution lies not in trying to make it rain or snow more, but in aggressive conservation. He points to Israel, which after abandoning its cloud-seeding programme, shifted its focus to the demand side, for example by developing irrigation techniques in agriculture that needed less water. “They found that the brute force atmospheric approach just does not work.”

Meanwhile, proponents of cloud-seeding press ahead with efforts to wring more water from the sky. Back on Ward Peak, Frank McDonough gestured to the green box busily sending puffs of silver iodide into the air. “There is more moisture in one of these clouds than we could ever use,” he said, looking out at the snow-covered Sierra extending to the horizon. “We’re just taking a little extra.”