Evgeny Podolskiy; (inset, top right) Osamu Abe

Tail of ice. Feathery "shrimp tails" form on slanted surfaces (main image and left inset). Scientists have also spotted other unusual ice formations (right inset).

Of Ice and Men

Emily is a staff writer at Science.

In 2009, glaciologist Evgeny Podolskiy was on vacation, climbing Japan's Mount Zao volcano, when he stumbled on something he'd never seen: soft, feathery fronds of frost called shrimp tails. He left the mountain with photographs and a question: What weather conditions produced the strange formations? Now, with the help of a powerful climate model, he and a team of international scientists have reconstructed how the tails formed. The discovery is a potential boon for wind energy companies building high-altitude transmission lines and windmills.

It was late April when Podolskiy began his hike up Mount Zao, joined by his family and cryospheric scientist Osamu Abe of the National Research Institute for Earth Science and Disaster Prevention in Shinjo, Japan. By then, the famous rime accretions called "snow monsters," which form on Mount Zao's trees and draw thousands of tourists every winter had melted, and cherry trees were blooming at the bottom of the mountain. "We had no expectation to see anything icy," says Podolskiy. But as they neared the top, they saw strange feathers of ice attached to boulders and the walls of a shrine near the emerald-green lake in the volcano's crater.

Abe explained to the group that the feathers were called Ebi no shippo, or "shrimp tails," and that they form when tiny water droplets in the clouds over the volcano's dome collide with obstacles such as rocks or buildings. Although the water droplets are below freezing temperature, they do not form crystals until the instant they hit an object. When millions of droplets accumulate on surfaces that have a roughly 25° slant, the shrimp tails grow in discrete feathers. Similar "lobster tails" have been studied on the wings of aircraft, says Podolskiy, but to his knowledge no one had determined the weather conditions that cause the tails to form on the ground.

After taking a few photographs, Podolskiy went back to his home institution, Nagoya University in Japan, where he was finishing his Ph.D. "I forgot about the shrimp tails for a while," he says. A year later, however, he attended a talk by Bjørn Egil Nygaard, an atmospheric scientist at the Norwegian Meteorological Institute in Oslo, who had used a meteorological model called the Weather Research and Forecasting (WRF) model to study the icing of structures such as telecommunication towers and ski lifts in Europe. It occurred to Podolskiy that this model might solve the mystery of the shrimp tails' origin, but there was a problem: He wasn't a modeler. Undaunted, he began teaching himself the necessary physics and recruited ice modeling experts to help him solve the puzzle.

To figure out what the weather was like on Mount Zao when the shrimp tails formed, the team made use of the WRF model's ability to deduce past storm patterns by analyzing hour-to-hour historical meteorological measurements taken all over the world. According to the model, the shrimp tails formed as a result of two cold, windy periods, each lasting several days. Temperatures dipped to -6.3°C, and winds gusted to nearly 26 meters per second. The model suggested that the amount of liquid water in the clouds over the volcano was several times higher than used in lab studies of similar tails, which could help scientists predict future icing events.

Next, the team took the atmospheric conditions simulated by the WRF model and entered them into an ice-accretion model commonly used to predict how much frost will accumulate on manmade structures. When they ran the model to include a low wind angle, it predicted shrimp tails that matched the length of those on Mount Zao within centimeters. They reported their results last month in the Journal of Geophysical Research.

Greg Thompson, the atmospheric scientist at the National Center for Atmospheric Research (NCAR) in Boulder, Colorado, who helped develop the WRF model, says the study is convincing. "The results are solid—they took care of any arguments against them." Hugh Morrison, a NCAR climate modeler who also helped develop WRF, says he was surprised by how closely the wind and temperature measurements produced by the model aligned with those taken at the ski station. "That's a pretty robust result."

To Thompson, the take-home point of the study is that models are getting better all the time. Twenty years ago, he says, it would have been "a joke" to try to reproduce storm conditions in a certain place on a specific day with a regional weather model. But today, state-of-the-art climate models such as WRF can recreate certain atmospheric phenomena with impressive accuracy. In climate science, he says, "modeling has been a bright spot."

Alain Heimo, a physicist who studies icing on energy infrastructure, describes the results of the study as "really impressive." Expanding renewable energy sources such as wind power, he says, will require building more windmills and transmission lines in mountainous regions where icing is a major concern. Studies like Podolskiy's, he says, show that the "skill of modeling icing is increasing constantly."

*Correction, 3 August: It was previously reported that the Weather Research and Forecasting Model is a global climate model. In fact, WRF is a regional weather model.