Berkeley researchers use wireless sensors to study redwoods
The network — a ‘new kind of microscope’ — will offer environmental scientists a ‘web of data,’ promote efficiencies of scale

By Sarah Yang, Public Affairs | 13 August 2003

This wireless mote, which fits into a container the size of a film canister, is so small and light that entire networks of them can be used to monitor redwood physiology, yielding a wealth of data about the giant trees’ fog-belt ecosystem. Noah Berger photo

Fear of heights is not an option for graduate students working with professor of integrative biology Todd Dawson. For years, Dawson’s research on the moisture that giant redwoods absorb from fog has involved the installation of 30 pounds of gear — including data loggers, sensors, and wires — onto trees that stand 300 feet tall in the redwood groves of Santa Cruz and Sonoma counties.

“We use climbing equipment to attach our data loggers, which are the size of a small printer, to the trees,” said Dawson. “Imagine doing that 200 feet or more above the ground.”

But now this daunting task is being made orders of magnitude easier as Dawson teams up with David Culler, professor of computer science, in an innovative project that employs a network of miniature wireless sensors designed to measure critical variables, such as light, temperature and humidity, within a redwood grove.

In the project’s first phase, Dawson and Culler, who is also director of the Intel Research Berkeley laboratory, have begun installing 50 sensor nodes on five redwood trees in the UC Botanical Garden’s Mather Redwood Grove.

“Wireless sensors are perfect for environmental monitoring because they are low impact and low visibility,” says Dawson. “Instead of a printer-size data logger, we’ll have something the size of a film canister.”

Each wireless sensor, or micromote, measures less than three cubic inches in size and is capable of transmitting radio signals at 50 kilobytes per second. Culler and a team of graduate students developed the sensor boards and networking software for the motes.

“These devices need to run for months on a size-C battery, streaming a variety of environmental data out of the trees for data processing,” says Culler. “We worked with Todd’s team to design a system that would generate trustworthy data and withstand the harsh environment in the forest, while making it easy to install many sensors in each tree. The network of sensors will provide a web of data for environmental scientists.”

The researchers plan to expand the wireless network later this year to include redwood groves in Big Basin Redwoods State Park in Santa Cruz County and at a site in Sonoma County.

The fog-belt boundary
In the project’s later phases, Dawson says, additional sensors will likely be used to measure the flow of water within trees as part of his lab’s ongoing efforts to quantify the amount of water redwoods obtain from the soil and absorb from fog. He says that quantifying the redwood’s dependence on fog and the unique environments found in the fog belt could provide important insights into the factors influencing the geographic range of the coast redwood, which covers the fog belt of coastal northern California to southern Oregon.

“We’d like to better understand why the natural range of the coast redwoods is largely restricted to the fog belt,” says Dawson. “One reason could be that fog provides an essential source of water for the trees. The summertime fog-drip accounts for 25 to 40 percent of the water the redwoods obtain each year. We’re now interested in studying whether a fungus that lives on redwood leaves helps the trees absorb water directly from the fog. We have a lot of questions, and these new micromotes are the solution to getting them answered.”

The wireless sensor array project, part of the Center for Information Technology Research in the Interest of Society (CITRIS), further expands the environmental applications that are possible with wireless-sensor networks developed by Berkeley engineers. In 2002, for example, researchers here and at Intel began using sensors to monitor the habitat of Petrel seabirds on Great Duck Island, off the coast of Maine.

“Wireless-sensor networks are like a new kind of microscope,” says Culler, who in 2002 approached Ellen Sims, director of the UC Botanical Garden, with the idea of using wireless sensors in environmental research. “They give you the ability to perceive what you haven’t been able to see before. In this case, we are visualizing in detail what’s happening over a large space.”

Culler points out that because wireless sensors are significantly cheaper and easier to install, biologists can deploy multitudes in a given area to create a higher-fidelity picture of the ecosystem in a redwood forest than would be possible with the current sensor system.

“The use of wireless sensors drives scale,” says Culler. “We could potentially use hundreds of sensors per tree to create a three-dimensional representation of the forest.”

Not only can the sensors paint a picture of life in a redwood grove, they can also provide details on the dynamic behavior of individual redwoods by measuring a tree as it uses water.

“The sensors can measure how a tree swells and shrinks in response to water uptake or loss,” says Dawson. “As any large tree takes up water, it’s stored in the trunk, which causes it to expand. As a tree uses the water, the trunk will shrink. It could be a change of a few millimeters within an hour, which is significant.”

The wireless network will also make it easier to recover information. Instead of climbing up to the data logger to retrieve the measurements, researchers will be able to download the data onto a laptop from the safety of the base of the tree.

“This will revolutionize our ability to collect high-precision environmental data,” says Dawson. “We’ll also be able to get rid of about 10 kilometers of wire now strung along redwood trees. The wireless sensor network will change the way people do ecological research.”