Cloud seeding—sowing clouds with small particles to make them rain or snow—has a reputation as dodgy as the weather. That’s because even though scientists have been seeding clouds since the 1940s, there was precious little proof the technique worked. Now, researchers flying two small planes through a bank of clouds in Idaho have shown, for the first time outside the lab, that humans can artificially turbocharge snowfall.

“What they’ve done is identify the chain of events from seeding to precipitation on the ground, which has been sorely needed for the last 80 years,” says William Cotton, a former professor of atmospheric science at Colorado State University in Fort Collins who was not involved with the research.

In the 1940s, atmospheric scientist Bernard Vonnegut found that particles of silver iodide can cause supercool clouds of water vapor to freeze into snow in the lab. (He is the brother of writer Kurt Vonnegut, whose fictional ice-nine was partially inspired by the discovery.) Particles like silver iodide can provide a scaffold on which water molecules can align themselves into a crystalline structure or, in other words, freeze. The technique has been used by militaries and civilian governments on and off ever since.

But despite decades of cloud seeding operations, proof that the technique works outside miniaturized clouds created in the lab has been elusive. One reason: Instruments of decades past couldn’t measure water droplet size in clouds in real time. Without knowing how a cloud evolves after seeding, scientists were unsure whether the silver iodide was doing anything at all. Another: The chaotic nature of weather makes controlled, natural experiments almost impossible. “Once you seed, you’re contaminating the cloud. You can’t repeat the experiment because you’ll never have the same atmospheric conditions again,” says Katja Friedrich, an atmospheric scientist at the University of Colorado in Boulder.

But newer instruments convinced Friedrich and her colleagues that the time was ripe for another approach—and the National Science Foundation and Idaho Power provided the funding. The team took its experiment to the mountains of southwestern Idaho, where it waited until supercooled clouds appeared in the sky. At temperatures of 0°C to –15°C, they are cold enough to freeze, but are at low odds of doing so.

When the right clouds came along, the team sprang into action. It launched one aircraft that made laps between two ground-based radars, dropping canisters that spread silver iodide into the clouds. The same plane also flew through the cloud while streaming silver iodide from its wings. Another plane loaded with cloud measuring equipment paced a perpendicular path to take readings.

At first, there was nothing. “The radar can only see [water] particles that are big enough, and these clouds had tiny droplets not detectable by radar,” Friedrich says. “Suddenly, we saw lines appear. It was really astonishing.” The zig-zagged lines matched the flight path of the first plane. Within these lines, the cloud’s water particles were getting bigger as they hit the silver iodide and froze. After a couple of hours, the snowflakes had grown from a few microns in diameter to 8 millimeters in diameter—heavy enough to fall to the ground, Friedrich and her colleagues report today in the Proceedings of the National Academy of Sciences. “We were super, super excited. Nobody had seen that before,” she says.

The experiments have also been met with enthusiasm from cloud seeding companies. “Those of us working on cloud physics for a long time have felt that [cloud seeding] was working,” says Bruce Boe, a meteorologist at cloud-seeding company Weather Modification in Fargo, North Dakota. “This verification and incontrovertible evidence this is occurring is really, really nice for us.”

Still up for study is whether the approach is economical. “Does it make enough snow to make an impact on a water budget?” Friedrich wonders. “We still have to answer those fundamental questions.”