We know trees get stressed. What if they could tell us when they’re stressed?
A new project from NAU aims to build a system using satellites and instruments on the International Space Station that will track in near-real time when trees are stressed either because of lack of water or a bark beetle infestation and make that information available not only to researchers and forest managers but also to the public who have trees in their yards that they want to keep healthy.
Alexander Shenkin, an assistant research professor in the School of Informatics, Computing, and Cyber Systems, is the principal investigator, with Michael Gowanlock, an assistant professor in SICCS, as co-PI. The Arizona Tree Stress Explorer and Alert System is funded by a three-year, $1.26 million grant from the Arizona Board of Regents’ Technology and Research Initiative Fund (TRIF).
The system will rely on the changing optical and thermal properties of leaves when a tree is stressed and will be able both to detect stress and alert stakeholders of significant changes in tree stress. When it’s completed, every 9-square-meter patch of land in Arizona with at least one tree will be monitored daily by satellite and ISS-borne platforms for signs of stress, with the goal of getting current data into the hands of the people who can help protect trees, which play an outsized role in removing carbon from the atmosphere and make places like Phoenix a more enjoyable place to be in the summer.
“I’m excited to build systems that actually interact in the world, not just write a paper,” Shenkin said. “This system will make a difference on the ground by getting information to policymakers, forest and fire managers and the public, who can check on trees in areas that they care about.
How trees feel (and show) stress
Two of the biggest causes of tree stress in Arizona are the never-ending drought and the increasing frequency of bark beetle infestations. Researchers know this, and they’ve been measuring the adverse effects for a while. And most of us, scientists or no, can pinpoint a severely stressed or dying tree.
But tree stress, like in humans, is a changing factor, and trees are not just healthy or unhealthy. But every green leaf does not indicate the same level of stress; a number of small factors like where the color falls on the electromagnetic spectrum, the shape of that spectrum and how reflective the leaf is go into pinpointing how stressed a tree is. Additionally, scientists can’t individually diagnose every tree from the ground. This is where big data comes in.
“The big-data approach combined with parallel computing will enable near real-time updates of tree-stress maps across all of Arizona,” Gowanlock, the project’s big-data expert, said. “Without these large datasets and significant computational resources, we would not be able to make accurate predictions of tree health in Arizona.”
How this system works
There are two pieces to building this system—collecting the images that will provide the data, and calibrating the system so it can sort out signs of stress in trees. This is all the big data.
Shenkin plans to use satellite remote sensing data from Planet Labs that covers the world at 3-meter pixel resolution daily. It does this by taking countless pictures and breaking those images down into data. That data will go into Monsoon, NAU’s supercomputing cluster, to be analyzed for signs of tree stress.
The second part is teaching the system what it’s looking for. That part will happen on the ground; Shenkin said they will install stem psychrometers on trees, to measure the amount of water in a branch. Knowing when a tree is actually stressed will allow researchers to pull satellite images for those times when trees are stretched and determine the specific factors that indicate stress.
It’s a similar concept for bark beetles, though instead of psychrometers, they will use data that USGS is already collecting.
It’s yet another use for big data, which is transforming how problems are solved on Earth and beyond. Gowanlock does a lot of his work in astronomy; there’s a significant need for real-time monitoring, as conditions in space can change disconcertingly quickly as well.
“We want to be able to detect or predict interesting events as fast as possible so that we can either study these events in more detail and/or provide actionable information to stakeholders, such as reporting the risks of forest fires to the state,” he said.
This is one of many interdisciplinary projects at SICCS that aims to solve problems with big data. Researchers use informatics on issues related to ecology and the environment, health, electrical engineering, astronomy, business and even arts to respond to rapidly evolving, multifaceted problems and come up with creative, adaptive solutions.
Why it matters
Trees are important for biodiversity; they provide habitats for insects (which are more critical than many of us would like to admit), animals and birds. They also make humans’ habitats more pleasant. In Phoenix, for example, trees have a major impact on the urban environment; the process of evapotranspiration cools the air, in addition to offering shade and just making cities, parks and yards more attractive.
They also play a quiet but critical role in climate change; trees are a significant carbon sink. When trees die, that carbon goes back into the atmosphere. So, planting trees is good, but keeping the trees we have healthy and alive for longer is even better.
“While reforestation isn’t going to solve all our climate change problems, it will and needs to be a bridge to a lower carbon future,” Shenkin said. “If we can keep trees on our landscapes, the tail of climate change, once we get past peak warming, if we’ve done our work, that temperature tail can be much shorter.”