Most people living in Utah probably think they have a good idea of what snowflakes look like even if they do not ski or snowboard. But two researchers at the U are changing the way people look at snowflakes, and their research could help improve the accuracy of the weather forecast.
Back in 2008, Cale Fallgatter, then a master’s student in mechanical engineering, suggested to Tim Garrett, professor of atmospheric sciences, that they could build an instrument that would time snowflakes in free fall while simultaneously photographing the flakes from multiple angles.
The camera can easily capture 10,000 images a day without any supervision and detects falling snowflakes using infrared sensors. Garrett said the lack of human intervention is what makes these snowflake images different from their predecessors.
“We’re seeing forms of snowflakes that have never been seen before,” Garrett said. “What we’re seeing is these snowflakes without actually having to touch them. We see them as they actually exist in nature.”
But the Multi-Angle Snowflake Camera is turning out more than just pretty pictures.
“A snowflake falls through our camera and it triggers a bank of [infrared] sensors,” Fallgatter said. “The IR sensors record the time stamp when the trigger happens. It continues falling and triggers a second bank of sensors. It sends another time stamp. If you subtract the time stamps … you get fall speed.”
The instrument then captures images of the flake from several angles, and each fall speed is reported to a computer with images of the corresponding snowflake.
Knowing the size and fall speed of snowflakes could provide weather forecasters with the information they need to predict accurate snowfall.
“You’ve probably experienced times when the forecast for how much snow there is going to be has been quite hopelessly off,” Garrett said.
The major reason meteorologists get snow count numbers wrong is they have not been able to accurately predict the size of the snowflake or how quickly it will fall.
“Graupels,” or ice pellets formed when water droplets collide with a snowflake, are much more aerodynamic and thus fall faster, whereas fluffy “aggregates,” made up of clusters of flakes, float to the ground more slowly, Garrett said.
Meteorologists often get their snow predictions wrong because they are making their calculations based on measurements made in the Cascade mountains back in the early 1970s, Garrett said.
“This is quite primitive data,” he said. “We will have immensely better statistics.”
Garrett said the pictures could also help with avalanche prediction by providing a better description of the snow pack.
The camera is hyper-fast and equipped with a bright flash mechanism.
“One of the biggest challenges to capturing snowflakes in free fall is lighting,” Fallgatter said. “We are running the camera at 1/250,000 of a second, which is extremely fast, and at that speed you have to flood the picture with light.”
If the camera were to shoot without a flash, the picture would come out black because the fast shutter speed. Fallgatter chose to use LED lighting because it is much brighter than traditional lighting and doesn’t produce heat, which means the snowflakes will not melt when passing through the camera.
“We’re using three 40-watt LEDs, which is a lot of light,” Fallgatter said.
Two of Fallgatter’s instruments are currently snapping photos at the Alta Ski Area. The equipment does all the shooting and data collection without human help, but Fallgatter said it has to be monitored for snow buildup. He plans to start working to solve that problem once this year’s snow season is finished.