Quinn Thomas has a bold vision: He wants us to predict the natural environment like we currently predict the weather.

“Imagine waking up to a 10-day forecast of the risk of interacting with ticks that have Lyme disease,” said Thomas, an associate professor of forest resources and environmental conservation in the College of Natural Resources and Environment and an associate professor of biology in the College of Science. “Imagine being able to forecast the carbon uptake that a forest is going to collect for a number of years or knowing which reservoirs a resource manager should pull water from and when. And then imagine doing all of that with shared tools, shared frameworks, and a shared computational infrastructure so that learning in one domain can easily transfer to another.”

To help make these possibilities a reality — utilizing data to predict outcomes as diverse as infectious disease spread to the changing colors of leaves — the Virginia Tech Board of Visitors has approved the awarding of a Data Science Faculty Fellowship to Thomas by the College of Science.

The fellowship, established in 2021 with funds from an anonymous donor, aims to enhance the study and teaching of data science at Virginia Tech by recognizing faculty dedicated to extraordinary research in data science within and across disciplines. The College of Science also awarded a second Data Science Faculty Fellowship to Xinwei Deng in the Department of Statistics.

For Thomas, who teaches environmental informatics and climate science in the Department of Forest Resources and Environmental Conservation, data is the key driver for new possibilities in environmental forecasting.

“In many areas of environmental science, we are starting to have observational networks that rival those in meteorology,” he noted. “Harvesting that supply of real-time data and using it to help galvanize and advance the field of ecological and environmental forecasting is my passion.”

“Advancements in science often occur at the boundaries or intersections of disciplines,” said Paul Winistorfer, dean of the College of Natural Resources and Environment. “This partnership of faculty intellectual talent between the colleges is one example, and it is fostering significant advancement in ecological and environmental forecasting with a foundation on data science.”

Thomas has already started to apply data to real-time challenges. As the lead investigator for a National Science Foundation grant aimed at utilizing data collected by the National Ecological Observatory Network, Thomas is helping to train a new generation of researchers who are adept at turning information into ecological forecasts.

"Through a grassroots, community-driven, decision-making process, our Ecological Forecasting Initiative settled on five themes to examine: tick abundance, beetle community richness, carbon storage, water quality, and the timing of leaf colors in the spring and fall," explained Thomas, an affiliate of the Global Change Center housed in Virginia Tech’s Fralin Life Sciences Institute. "We’re asking teams to contribute forecasts from different sites across the U.S. to help grow the field. So far, we’ve had 15 different teams submit 1,500 unique forecasts."

Quinn has also worked collaboratively with Associate Professor Cayelan Carey of the Department of Biological Sciences to provide forecast data to better safeguard water quality standards. Using an open-sourced software called Forecasting Lake and Reservoir Ecosystems (FLARE), Thomas and Carey are working with the Western Virginia Water Authority to provide long-term environmental forecasts to ensure that clean water is protected against the ecological challenges of the future.

“Data Science is now a key tool in the advancement of science and it is why we established the Academy of Data Science and these Data Science Fellowships in the College of Science,” said Ron Fricker, interim dean. “The goal is to support our incredibly talented faculty, like Quinn, who work at the intersection of their discipline and data science.”

For Thomas, working at that intersection helps strengthen his understanding of how to tackle the daunting task of predicting the environment.

“The process of tackling ecological forecasting across different areas has helped hone my vision of how ecological forecasting systems might operate,” said Thomas. “By trying to interpret ecological data across different domains of environmental and ecological science, I’ve gained a kind of synthetic understanding and a generalized way of viewing forecasting problems.”

Thomas, who admits that he learns by doing, has incorporated a problem-first approach to learning in his courses.

“My classes start with a question,” he noted. “Are lakes losing ice earlier in the year? Which forests are taking up more carbon? If you start there and then help students see which data science techniques can be used, which data sources are most critical, and how they can best communicate their findings, they can try again with a new question and adapt their strategies to that new challenge.”

Thomas says that the challenges of the environment are subjects that transcend the boundaries of any one academic discipline, and notes that data is a tool that allows collaboration to advance our understanding of those challenges.

“We have strong statistical science and data analytics and a technology-based focus at Virginia Tech, and we also have a very strong natural resources, environmental science, and agricultural focus,” he said. “In the next phase of my career, I’d like to think about how all of these strengths can be brought together to emphasize what we can do to meet both the technology and land grant missions of the university, so we can provide actionable information for the commonwealth and the globe.”

Written by David Fleming

Share this story