�պ�����

Last year, in the middle of a warm, humid summer, in the Jericho Research Forest, associate professor in �պ�����’s College of Engineering and Mathematical and Sciences (CEMS) Arne Bomblies and his team of researchers were waiting for snow. Starting in April, Bomblies and his team began installing a series of snow sensors and meteorological instruments throughout the state, from the top of Mount Mansfield to the shores of Lake Champlain.

“What we’re after is a better predictive model of snow in Vermont and in the northeast in general,” Bomblies said. “The goal is to ultimately understand how things like trees, slope aspect, elevation, rainfall, and cloudiness impact snow and be able to model that.”

“Snow is critically important to the state and the region,” Beverley Wemple said, a professor of Geography & Geosciences and member of the research team. “Our winter recreation economy depends on our snowpack. Snowmelt events deliver carbon and nutrients to our waterways, and the increasing frequency of mid-winter rainfall and icing events creates hazards for travel and alters fragile ecosystems. Our winters are shifting rapidly and we need more information about these dynamic changes.”

Long-term observations of snow in Vermont come from various sources, including a network of volunteer observers and notably a measurement station near the summit of Mt. Mansfield, Vermont’s highest peak.

“The Mount Mansfield snow stake is a critical source of high altitude snow information but it records only snow depth,” Bomblies said. “Compare that to places in the western United States where they have snow measuring stations monitoring the full range of winter weather dynamics, including the important snow-water equivalent.”

The snow-water equivalent is the depth of water that would cover the ground if the snow cover was liquid. Snow-water equivalents can vary depending on how dense snow is. Lighter, powdery snow, ideal for skiing and other snow sports, has a lower snow-water equivalent.

“If it snows 10 inches in Vermont, all we know is that we’ve got 10 inches of snow,” Bomblies said. “We have no idea much water that corresponds to and what that means for water runoff or how sensitive the snow is to the sun.”

The sensors Bomblies and his team installed include anemometers to measure wind speed, sensors to measure humidity and temperature, soil moisture sensors, and ground platform sensors to measure snow density and water equivalent.

“These platforms will sense directly the amount of liquid water in the snow, the amount of ice, and the amount of air space,” Bomblies said. “We’ll be able to directly sense all of the components that make up snow and see how that changes as, say, rain starts to fall or how a particularly sunny stretch affects the snow.”

Beyond those important meteorological measurements, however, more researchers at �պ����� are capitalizing on the project to study other aspects of snow, including capturing images of snowflakes as they fall.

“The condition of snow as it falls determines the properties of the top layer of the snow pack. This is important for a wide range of activities, including recreational sports such as skiing, road conditions, snow drifting, and resuspension of wind-blown snow,” Jeffrey Marshall, associate dean for research in CEMS and member of the research team, said. “Some properties of falling snow, such as temperature, humidity, and velocity of snow flakes, are easier to measure than other properties, such as wetness and size of flakes.”

Taking a range of measurements of properties such as snowflake size, wetness and fall velocity over the next two winter seasons, Marshall and CEMS professor Safwan Wshah are using a combination of physical modeling and machine learning to study the more difficult-to-measure properties of snowflakes from measurements of the properties more easily obtainable, which will help make monitoring of snow easier and more efficient.

The implications of this project are deeply important, not just for snow monitoring, but snow tourism in Vermont and climate change.

“There’s a growing concern in the northeast that the warming climate is going to make winter recreation – skiing, snowmobiling, ice fishing – much less available,” Bomblies said. “Vail Resorts, which owns Stowe Mountain resorts, has invested money into snowmaking equipment, but sustaining artificial snowmaking in a warming climate will be challenging. New data on the drivers of snow distribution and melt, along with climate model predictions for our future, will help us better predict the susceptibility of our winter recreation industry to warming.”

Bomblies explained that while overall warming is a worry for snow research, increased weather variability during a winter season is much more drastic and a larger cause for concern.

 “It used to be that once it got cold, it stayed cold with maybe one or two ‘January thaw’ events, commonplace surges in temperature often accompanied with rain,” Bomblies said. “Those have become much more frequent recently and it’s one of the features of the changing climate.”

According to the Gund Institute for Environment’s recent , the state’s traditional winter season will be shortened by as much as a month in some parts of the state.

“Downhill skiing, with the help of snowmaking, will likely remain largely viable in Vermont up until approximately 2050,” the assessment stated. “By 2080, the Vermont ski season will be shortened by two weeks (under a low emissions scenario) or by a whole month (under a high emissions scenario), and some ski areas will remain viable.”

Bomblies hopes the research will provide Vermont and the entire northeast region a better understanding of how meteorological elements and climate change affect snowpack.

“It's certainly a concern around here, what climate change will look like in Vermont, where winter is such a integral component of our identity and livelihood.” Bomblies said. “With data collection starting now, we can improve modeling and follow research projects and help significantly here at �պ�����.”