Many people look at trees to relax, taking in the soothing fractal patterns to adjust their eyes from too many hours spent at a computer. Dominick “Dom” Ciruzzi, assistant professor of geology at William & Mary, takes tree watching to a whole new level. An ecohydrologist, Ciruzzi is fascinated by how ecosystems and water interact, and he’s taken a special interest in arboreal life. 

Along with a bevy of collaborators, he recently co-authored a paper in Agricultural and Forest Meteorology. The study validates video cameras as a tool to monitor changes in tree sway — a biomarker of overall health.

“Videos give us a path forward to develop a non-invasive, cost-effective and easy-to-use technology to measure important ecohydrological processes like tree hydration, phenology — seasonal changes such as leaf-out and leaf-drop — and overall health,” said Ciruzzi. 

Down the road, that kind of tool could have far-reaching applications. Frequent video monitoring could help detect drought stress before wildfires ignite. It could guide prescribed thinning and burning. And this approach could prompt cities spending millions of dollars a year on stormwater protection to incorporate more trees as green infrastructure.

“We hope this video-based approach will transform how we monitor tree function and deepen our understanding of water-tree interactions under changing environmental conditions,” said Ciruzzi.  

This research builds on the team’s previous work, highlighted by the Wall Street Journal and Wired, developing what they like to think of as Fitbits for trees, devices that strap to tree trunks to gather health metrics. Using videos gives the researchers the ability to scale this technology from the level of individual trees to forests.

A language unto their own

It turns out that just as animals and humans have their own body language, so do trees. But rather than communicating with raised eyebrows or slumped shoulders, trees speak through their sway — a graceful movement that follows the arc of a pendulum.

“All trees sway with a certain frequency when excited by the wind,” said Ciruzzi. “This frequency depends primarily on their mass and stiffness, along with some geometrical attributes like diameter and height.”

As a tree gains mass, through increased foliage in the spring or a cargo of snow in the winter, it sways more slowly. Conversely, after a heavy rain, a tree engorged with water will gain rigidity and sway faster. 

“By observing changes in the sway frequency and combining that information with measurements like temperature and soil moisture, scientists can study a host of important ecohydrological processes,” said Ciruzzi. “We can tell if a tree is water stressed, observe seasonal changes in foliage, measure how much water trees capture during rain and snowstorms as well as tell if a tree is rotting and sick.” 

Beyond academic curiosity, these insights have practical value: They could help forest and water managers make smarter, data-driven decisions. To unlock that potential, scientists are working to refine the tools that capture and interpret a tree’s subtle sway. 

A new way to measure tree sway

Over the course of earning his Ph.D. and conducting research at W&M, Ciruzzi has shown that accelerometers — devices that track shifts in motion — can accurately monitor tree sway frequency. Sling-shotting the chunky waterproof boxes into the branches, he and his students secure them with a pulley system, sparing themselves the precarious climb up a playfully swaying tree. 

Unfortunately, accelerometers have one major drawback — they can only measure one tree at a time. 

“We’ve gotten really good at monitoring individual trees with these devices, but now we’re at the stage where we need a scalable solution,” said Ciruzzi. 

To Ciruzzi and his collaborators, videos seemed like the next natural step. Pointing cameras at individual trees, they collected footage and then ran it through a special algorithm. Tracking subtle shifts in pixel brightness, the algorithm was able to detect the back-and-forth motion of trees and ascertain the sway frequency. Accelerometer data taken from the same trees confirmed that the video-produced values were accurate. 

Proving the method on a single tree gives the researchers confidence it can scale to many — opening the door to using sway frequency as a practical metric for forest health.

New frontiers

“These proof-of-concept results indicate that sway frequency can be leveraged as a powerful tool with real-world applications,” said Ciruzzi. 

He finds this approach particularly enticing as it could supplement — or potentially even replace — some of the more expensive and laborious techniques currently used to monitor forests.

“Take drought monitoring for example. This traditionally relies on satellite imagery or sometimes destructive measures, like drying a plant to see how much water it’s holding,” said Ciruzzi. “There are obvious downsides to both of those methods. It’s just not convenient or time effective to wait 25 days, as can happen, so you can capture the next cloud-free satellite image.”

Beyond the verdant hills of land undisturbed by human development, video cameras and algorithms could prove useful in more densely populated areas.

“Municipalities spend millions of dollars on infrastructure to manage flooding risks. But the role of trees in safeguarding cities is often poorly quantified,” said Ciruzzi. “If we understand how much water different types of trees capture and store, planners might incorporate more green infrastructure — trees — into their designs instead of relying solely on gray infrastructure like pipes and drains.” 

Ciruzzi and W&M student Gabriel Imholte ’28 began to investigate these questions this past summer, looking at trees on the university’s campus. 

Ultimately, Ciruzzi envisions a Fitbit for forests: accelerometers tracking the health of individual trees, with videos expanding that insight to the larger ecosystem.