“Our software has predicted the high snowpacks that occurred in the Rockies this year and the low snowpacks of previous years—useful for forecasting floods and droughts,” said USask post-doctoral fellow Chris Marsh who developed the model as part of his PhD project supervised by hydrologists John Pomeroy and Howard Wheater.
“By determining how much snow accumulates in the winter and melts in the spring, our software enables better planning for crop irrigation and municipal water usage to help produce food and support communities.”
Still in development, Marsh’s computer model can aid the prediction of floods or droughts by integrating information with other tools, and can predict snow in high mountains, a vital source for freshwater and rivers that flow to the Pacific, Arctic and Atlantic oceans. The snow that feeds these water sources is significantly affected by climate change, causing fluctuations in water availability for agriculture, fish, hydroelectricity, and mining, Marsh said.
A recent paper by USask and Environment and Climate Change Canada researchers shows that scientists would have needed a model like Marsh’s that simulates precisely high mountain snowpacks to better predict the 2013 Calgary flood—no Canadian or U.S. models were able to do so accurately at the time.
One pilot application of Marsh’s software is SnowCast, a tool that helps scientists estimate the depth of snowdrifts and avalanches by simulating detailed patterns of snowpacks and their melt in the Bow River Basin above Calgary, including Banff National Park.
“Due to the COVID-19 pandemic, we could not manually monitor snowpacks in the Canadian Rockies this spring, so SnowCast has become essential to quantifying how much snow was on the ground and its location to inform possible flood risks downstream,” said Pomeroy, Canada Research Chair in Water Resources and Climate Change and director of the USask Centre for Hydrology. “This provides timely and important information for both scientists and authorities.”
Freely available on GitHub, SnowCast is the first in Canada to combine weather forecast information, precipitation, and wind flow changes, as well as the terrain and vegetation features of the mountains.
The software was developed as part of the USask-led Global Water Futures (GWF), a national research program led by the university’s Global Institute for Water Security, one of the world’s top water research institutes.
“These complex water simulations usually require lots of computer power, which means it takes a long time to run them, but Marsh’s novel computer techniques have made the USask model faster and more efficient than those currently available,” said collaborator and computer science professor Raymond Spiteri.
Working with USask computer scientist Kevin Green and other GWF researchers, the team will add river flows, vegetation, groundwater, streams, dams, and other water management controls to the model. The team will also expand predictions to other areas of Canada, and eventually to key mountain-sourced rivers worldwide.
Pomeroy noted that last year the World Meteorological Organization urged that more work be done to predict snowpacks in high mountain regions that supply water for over half of humanity.
“Integrating SnowCast with data from the 35 USask mountain research stations near Calgary to track snow conditions will be a great example of doing exactly that,” said Pomeroy.
The research was funded by the Canada First Research Excellence Fund, the federal agency NSERC and Alberta Innovates.
This article first ran as part of the 2020 Young Innovators series, an initiative of the USask Research Profile and Impact office in partnership with the Saskatoon StarPhoenix.
Federica Giannelli is a graduate student intern in the University of Saskatchewan research profile and impact unit. This content runs through a partnership with The StarPhoenix.