Engineering Environment News releases Research Science Video
June 13, 2022
In a citizen research project created by UW researchers, participants looked at time-lapse photos from Colorado and Washington and noted photos taken when trees had snow in their branches. Here is a close-up of a camera image of the AmeriFlux tower in Level Ridge, Colorado. This image is archived in the PhenoCam network and is one of the images analyzed by scientists in this project.AmeriFlux Tower
The snow that falls in the mountains is good not only for skiing, snowshoeing and breathtaking views. The snow cover it creates will eventually melt and this water can be used for hydropower, irrigation and drinking water.
Researchers want to predict how much water we will get later in the year based on the snow cover. But in forested areas, trees affect the calculations. When falling snow is spotted by trees, it sometimes never makes its way to the ground, and current models struggle to predict what will happen.
To improve the models and study what is happening with this detected snow, researchers from the University of Washington have created a science project for citizens called Snow Spotter. Participants viewed photos from Colorado and Washington and noted photos taken when the trees had snow in their branches. This information gave a first idea of how the interactions of snow and trees can vary between climates and how this can affect forecasts for summer water supplies.
The team published these findings on May 18 in AGU Water Resources Research.
“We, as skiers or snow enthusiasts, know that the snow in Colorado is really different than in Washington. But so far, there has not been an easy way to observe how these differences manifest themselves in the canopy of trees, “said lead author Cassie Lumbrazo, a doctoral student at UW who studies civil and environmental engineering. “This project uses volunteers to get some solid data on these differences. Another advantage is that it introduces our volunteers to how the research works and what snow hydrology is.
There are three possible scenarios for snow that has been caught by trees. It can fall to the ground like snow, adding to the current snow cover. It can blow out and turn into water vapor, so add nothing to the snow cover. Or the snow may melt and drip on the ground, which, depending on the conditions, may or may not add to the total amount of water in the snow cover.
One current problem with the mathematical models that describe these processes is that researchers do not know the weather – in a year, how often there is snow in the trees and what happens to it? – and how this time varies in different climatic conditions.
But acceleration cameras can record what’s happening in remote locations by taking pictures every hour, every day for years, creating a huge set of image data.
This is where civilian scientists intervene. Snow Spotter shows the volunteers a photo with the question: “Is there snow in the branches of the trees?” The volunteers then choose “yes”, “no”, “uncertain” or “it’s dark” before moving on to the next photo.
Using Snow Spotter, a total of 6,700 civilian scientists scanned 13,600 images from a number of sites in the western United States. The team focused on four sites for this study: Mount Hopper, Washington; Level Ridge, Colorado; and two different sites in Grand Mesa, Colorado.
“When the project started, I don’t think anyone really knew how successful it would be,” said Lumbrazo, who is currently doing research in Norway as part of Valle’s scholarship and Scandinavian exchange program. “But civilian scientists were processing it so fast that we were constantly left with images for people to classify. We received feedback that this task is really relaxing. Civic scientists can download these photos in the Zooniverse app and can just sit on the couch and click very quickly.
Civic scientists often dealt with the photographs they classified, for example, by calling out animals that appeared in the frame. Here is a screenshot of a participant pointing to a bird in the lower right corner of the image.Screenshot: University of Washington; image from the AmeriFlux tower
Each photo had between nine and 15 different volunteers who classified it, and the volunteers agreed between 95% and 98% of the time. From there, researchers could gather what the snow in the trees looks like during the year for each site.
“Our data physically show the difference in snow,” Lumbrazo said. “You can see the snow in Washington just cemented in the shed and never go away, which is how it feels when you drive that snow. Unlike the snow in Colorado, where it often snows, but it blows away. It is dry and dusty. ”
Researchers have used this data set to estimate current snow patterns. One limitation, however, is that currently the team only knows when there is snow in the trees. This method does not say how much snow there is in the trees, another component needed to make the models even better.
“But a limitation that does not exist is the number of scholars who want to process these images,” Lumbrazo said. “We’ve signed countless volunteer classes for students, and in the end they even have great discussions about certain images, and it’s becoming more of a scientific conversation.”
In addition, the data set generated by these volunteers could be used to train a machine learning algorithm to classify images in the future, the team said.
Researchers are working to expand their image data set to include photos from around the world so they can continue to study how different climatic conditions and precipitation patterns affect snow cover, which will also help make models more accurate.
Additional co-authors are Andrew Bennett and William Ryan Carrier, both of whom completed their research as PhD students in civil and environmental engineering at UW; and Bart Naisen and Jessica Lundqvist, both professors of civil and environmental engineering. Snow Spotter was created by Max Moser, who started this project as a student at UW studying civil and environmental engineering. This study was funded by the National Science Foundation and a presidential scholarship from Steve and Sylvia Burgess.
For more information, contact Lumbrazo, which is currently in Central European Time, at [email protected]
Grant number: CBET-1703663
Tag (s): Cassie Lumbrazo • College of Engineering • Department of Civil and Environmental Engineering • Jessica Lundquist