After World War II, it seemed that we could do anything. The world would do our bidding. We were going to build highways as had never been done, and we did. We were going to put our newfound understanding of splitting atoms to create useful energy, and we did — well yes, there were some problems, too.
We were also going to modify weather to suit us. The genesis for this ambition was at the General Electric Laboratory in Schenectady, N.Y. There, during the war, researchers studying the conditions under which aircraft icing occurred noticed that clouds at temperatures below freezing did not produce snow.
This wasn't new, exactly, says Chunglin Kwa in a book called Changing the Atmosphere. But this insight did lead directly to experiments in 1946 in which dry ice was used as seed material in first a household freezer and then in clouds to create snow.
Soon, clouds were being seeded willy-nilly, most generally by burning silver iodide in acetone, to produce smoke, enabling "seeding" from ground-based generators. Early on — and perhaps still — there were elaborate claims of seeded clouds gushing rain and snow.
The evidence for such claims? That was harder to establish. One of the best experiments was conducted from 1960 to 1970 on a mountain ridge in Colorado. There, in a randomized experiment, atmospheric scientist Lewis Grant began testing the efficacy of cloud-seeding. At last, he concluded the seeding had augmented snowpack by around 10 per cent.
In 1972, the Vail ski area launched a cloud-seeding program and, after a few hiccups, has been doing so continuously since 1977. Other ski areas have been involved in cloud-seeding. Telluride helps pay for cloud-seeding, and although they don't pay, generators are located at Winter Park and Crested Butte. Other generators in Utah are designed to benefit ski areas along the Wasatch Range.
Still, the evidence has not been totally satisfying. A National Academy of Sciences report in 2003 was almost scathing. It cited "extravagant claims, unrealistic expectations, and failure to provide scientifically demonstrable success."
What was needed, said the report, was a randomized statistical experiment. Wyoming in 2004 set out to do just that. With a budget that has grown to more than $14 million, the state authorized an experiment designed by the National Center for Atmospheric Research (NCAR). Two parallel mountain ranges, the Sierra Madre and Medicine Bow, were chosen.
Not just any passing clouds were seeded. Mountaintop temperatures needed to be -8C and cooler, and the clouds had to have what meteorologists call supercooled liquid. And finally, the wind had to be just right, to push the clouds across both ranges.
Then clouds were seeded randomly, sometimes one and then the other, but not both at the same time. The unseeded mountain range would be the control group. In this way, seeding began in 2008 and wrapped up last winter.
In a report submitted to the Wyoming Water Development Commission on Dec. 10, NCAR researchers said that the results "imply a 3 percent increase in precipitation with a 28 percent probability that the results occurred by chance." What that means, the report went on to say, is that the "primary statistical analysis indicated no significant seeding effect."
But hold on, there was more. By using modelling pegged to surface results, the researchers found suggestions of increased snowfall of five to 15 per cent.
As with art, this science was in the eyes of beholders. Those who liked cloud-seeding before thought they heard that their pre-existing beliefs had been confirmed. Those who thought cloud-seeding was one step removed from snake-oil retailing similarly heard reasons to scoff.
"It's good evidence that it works," said David Cole, of the Utah Division of Water Resources.
"There is always that question can you prove it beyond a shadow of a doubt."
From Los Angeles comes a similar appraisal. "These results are consistent with historic studies," said Tom Ryan, of Metropolitan Water District of Southern California, a consortium of agencies that collectively serve 18 million people.
Even if cloud-seeding produces only one per cent augmentation of snowpack, not the 10 per cent claimed by weather-modification companies, "that's a very cheap source of additional water for our basin," says Frank Krugel, the general manager of the Upper Gunnison Water Conservancy District, with headquarters 48 kilometres from the Crested Butte Ski Area.
In California and Idaho, companies pay for extensive cloud-seeding programs, believing that increased water yields will allow them to generate more electricity.
At the headwaters of the Colorado River in both Colorado and Wyoming, there are a variety of interests: the ski areas, Denver and other cities, and water districts serving farmers and ranchers.
But seeding programs are also paid for by water agencies in Los Angeles, Las Vegas and Phoenix, all of which draw water from the severely taxed Colorado River. This winter, for example, $930,000 is being spent on a cloud-seeding program in the Wind River Range. More than $1 million more is being spent in Colorado, at the headwaters of the Colorado River.
The two rivers join near Moab, Utah, their commingled waters slashing through the Grand Canyon and toward the Pacific Ocean. But except for last year, when there were special releases, the river has not delivered water into the ocean since the late 1990s and, not with any regularity since the 1960s. The giant reservoirs of Powell and Mead that were brimming 20 years ago both have bathtub rings, with broad speculation about the potential for both going dry altogether, or at least with too little water to produce electricity.
Nobody claims that seeding clouds will fill those giant reservoirs or get LA through this drought. "It's just one tool in the portfolio," says Metropolitan Water's Ryan. "You should do it whenever you can, but there are dozens and dozens of tools."