Understanding the optical properties of snow could lead to many advances, from improved snow removal techniques to improved avalanche safety. You might think snow is easy to identify: It’s cold, white, and covers the landscape like a blanket. What else can be said about this?
Too many, according to Mathieu Nguyen. He recently defended his doctoral thesis on the optical properties of snow at the Norwegian University of Science and Technology (NTNU) in Hjovik.
“Snow reflects all wavelengths of light and can have very different colors depending on the conditions and the angle at which the light hits it. The age and density of snow and air pollution also affect its appearance. “The appearance of the abdomen is a very complex subject,” says Nguyen.
He analyzed more than a thousand snow images. “This type of method can be used in many sensor technologies, including everything from giving us a better basis for making decisions about when to open roads to more closely monitoring avalanche risk in the mountains.”
A landscape of mirrors
Nguyen studied, among other things, how snow absorbs and reflects light and how the sun makes snow crystals sparkle. He believes the beautiful appearance of winter landscapes may be the key to answering a series of questions that have puzzled researchers for years.
But first, what exactly makes snow sparkle on bright sunny days?
“Snow is an accumulation of ice crystals. When conditions are perfect, they act like tiny mirrors. If they’re at the right angle, they reflect sunlight directly at you and sparkle in the landscape,” Nguyen says.
great potential
There has been a lot of research into how different metals shine in this way, but the glow of snow is still not fully understood.
“If we want to have fully autonomous cars in Norway, this technology will also contribute to safer road travel in winter,” says Nguyen.
So Nguyen tried to find out how these sparkles vary in contrast and intensity in images of snow under different conditions. He hopes this will provide a method of analysis that will allow us to classify different types of snow from images. Now this is impossible.
“This type of method can be used in many sensor technologies, including everything from giving us a better basis for making decisions about when roads need to be cleared, to more closely monitoring avalanche risk in the mountains. If we want to have fully autonomous vehicles here in Norway, this The technology will also contribute to safer driving on winter roads,” says Nguyen.
Images are needed from around the world
So far, researchers have only collected data from different locations in eastern Norway. The results are promising and show that sparkles can be used to classify the size of snowflakes. But more accurate classification of snow type requires much larger amounts of data than has been studied so far. Preferably with images from around the world.
“It will be important to capture images from other locations where the environment is different. It will be critical to understand how different levels of pollution play a role in the appearance and characteristics of the snow,” Nguyen says.
Experience for the future
In addition to being difficult to interpret from images, snow has long been known to be surprisingly difficult to reproduce digitally.
“The images of artificial snow we have in computer games and simulations today are not much better than white surfaces,” Nguyen says.
Their discoveries have yielded promising results in this field as well. He believes his work will give people who don’t have access to snow a good winter experience; He believes that there may be much less snow in the future.
According to the Norwegian Meteorological Institute, by 2050 more than one million Norwegians will live in places with less than a month of winter. Additionally, a recent study in the journal Nature confirmed that the entire Northern Hemisphere faces a future with less snow as a result of anthropogenic climate change.
“If we want to teach someone who has never seen snow before what it is, we have to be able to recreate it in all its complexity,” Nguyen says.
Source: Port Altele
