This June UVM’s Spatial Analysis Laboratory (SAL) received $108,000 in funding from the state of Vermont to update the state’s high-resolution land cover map. The first iteration of the map was created by SAL several years ago.
According to SAL director Jarlath O’Neil-Dunne, the change represents more than just an incremental improvement—the new digital map will track change on Vermont’s landscape at sub-meter resolution, making it thousands of times more detailed than existing maps.
“When the state makes management decisions, they want to have the best info possible,” explains O’Neil-Dunne. “Those decisions are based on things we can measure. For instance, to come up with paving budgets it’s important to know how many miles of roads we have.”
As a pioneer of GIS technology in Vermont, SAL operates as a non-profit in partnership with industry, government and non-profits to apply techniques in GIS, remote sensing, and spatial statistics aimed at solving unique problems.
The new land cover map his team will create (there are nine professionals and a small army of student workers and interns that make up the SAL lab) will include layers showing tree canopy, built infrastructure, grass or bare soil and water.
“I like to say it's about the green, the gray, the brown and the blue,” O’Neil-Dunne says, referring to color of each layer on the map. “Our role is turning this big data into useful information.”
Uses for the maps are almost limitless. Researchers depend on the visual imagery for analyzing forest growth patterns and impervious surfaces, or creating elevation and storm surge models.
Making Meaning from Maps
Rebecca Diehl is one of the many UVM researchers who relies heavily on the mapping data.
A research assistant professor in geography and a Gund Institute scholar, she’s interested in how floodplains retain sediment and nutrients that improve water quality and increase flood resiliency in the Lake Champlain Basin.
Her current work contributes to the Functioning Floodplain Initiative, a state effort to identify high priority projects to restore and protect stream, wetland and floodplain functions.
She overlays the map with LIDAR (Light Detection and Ranging) data, gathered from airplanes that fire off laser pulses to create very precise three-dimensional maps of the terrain. The flights are performed by private companies and funded by the state of Vermont.
“This helps us get a better understanding of inundation patterns on the landscape,” she explains. “We learn where the landscape is wet and where it's dry during different flood events and helps us anticipate where the worst flooding might occur.”
Diehl notes there can be gaps in the data. A bridge appearing over a river in an aerial shot can be interpreted as a barrier to water flow, when actually the water flows almost unimpeded underneath it. She uses Arc software to “hydro-enforce” the LiDAR images.
“Essentially we take out that bridge so as to make the image hydrologically correct,” she says.
Skill Building for Students
Both O’Neil-Dunne and Diehl are especially proud of how making maps—and using them as data tools for research—involves interns and work-study students who do real-world projects under the guidance of professionals in the lab.
Many recent UVM graduates have leveraged their SAL lab experience into jobs at global tech companies like Google, senseFly, and Uber; non-profits including the Chesapeake Conservancy; and research institutes like Duke University Marine Lab. Other grads have established careers in the commercial drone industry.
This year, Dayna Ullathorne '21 was hired by the lab to help with a project mapping the tree canopy across thousands of square miles surrounding Chesapeake Bay. The goal of the project is to help landowners and decision makers understand how the forests in the bay area are changing, with the ultimate goal of restoring healthy water systems for the entire watershed.
“We’re figuring out where the trees are, where new ones are growing, where they’ve been cut down,” she said in a recent interview. “Sometimes it will label shadows of trees as canopy, or there will be a building under the trees that needs to be identified,” she explains.
The project Ullathorne worked on was one of several out-of-state contracts SAL has secured over the years—the lab has completed more than 75 tree-canopy assessments in the United States and Canada.
“By working all over the country we build up a wonderful set of skills and workflows that allows us to reduce the cost of working in Vermont,” O’Neil-Dunne says.