Academic researchers and computer giant IBM are aiming to make Lake George, a 52-kilometer-long body of water in New York state, one of the smartest lakes in the world. Late last month, scientists formally began to capture data from the first of 40 sensing platforms that will give researchers a detailed glimpse into lake behaviors such as water circulation and temperature. The information will be fed into computer models that the researchers say could help managers protect Lake George from threats such as invasive species, excessive nutrients, road salt, and pollution.
The effort, known as the Jefferson Project, involves more than 60 scientists from the Rensselaer Polytechnic Institute (RPI) in Troy, New York; the FUND for Lake George, a regional conservation group; and IBM research labs in Brazil, Ireland, Texas, and New York. The researchers are using Lake George as a test bed for an array of sophisticated “smart” sensors that will monitor 25 different variables, including biological characteristics and water chemistry and quality. The sensors will not only report data back to laboratories, often in real time, but be able to respond to changes in the lake environment. “Our sensors can look at other sensors around [them] and say, ‘I’m seeing something a little unusual, are you seeing it too?’” says RPI’s Rick Relyea, director of the Jefferson Project. “If so, the sensor can make the decision to sample more frequently or sample in a particular depth of water more. They have a great deal of intelligence.”
The data the sensors collect will be fed to an IBM supercomputer that will help researchers develop five different computer models that will enable one of the Jefferson Project’s main goals: visualizing Lake George’s behavior. For example, using high-resolution weather forecasting technology developed by IBM, researchers will be able to see how runoff from big storms moves through the 600-square-kilometer Lake George watershed. Other models will allow researchers to examine the impact of the use of road salt on water quality, see how water circulates throughout the lake, and visualize lake food webs.
A map of the lake bed, shoreline, and watershed of Lake George, from the team’s 2014 bathymetric and topographic survey. The bright colors show the depth of the lake bed below the waterline.
The Jefferson Project isn’t the only effort to harness new technologies to wire up and study lakes. The U.S. National Science Foundation’s National Ecological Observatory Network is using similar approaches to study the impact of climate change, land-use change, and invasive species on aquatic ecosystems. Internationally, the Global Lake Ecological Observatory Network (GLEON), a grassroots network of ecologists, IT experts, and engineers, also uses new technologies to study how lakes respond to environmental change.
This Jefferson Project isn’t the first time IBM has experimented with instrumenting a body of water, says Harry Kolar, an IBM researcher and an adjunct professor of physics at Arizona State University, Tempe. The company has helped develop many of the technologies being used at Lake George by participating in other projects, including the River and Estuary Observatory Network, an observatory system tracking the Hudson River at Denning’s Point in Beacon, New York. In 2009, IBM also launched a joint project with Ireland’s Marine Institute to monitor water quality and marine life in Ireland’s Galway Bay.
What makes the Jefferson Project different, Kolar says, is not only the smart sensors and the high frequency with which they will collect data, but how the data will be used to help inform the models. And Paul Hanson, a limnologist at the University of Wisconsin (UW), Madison, says that although the Jefferson Project is similar to other lake-monitoring projects, “they’re doing it on steroids. More variables, more frequency, and with better integration [with] models.”
Overall, researchers plan to equip the lake with 40 sensor-carrying platforms, some on land and some in the water; they have deployed 14 thus far. The platforms come in four “flavors”: vertical profilers that send instruments into the lake’s depths to monitor things such as water temperature, chlorophyll, and dissolved organic matter; weather stations that measure humidity, barometric pressure, and wind velocity; tributary stations that study water entering the lake; and acoustic Doppler profilers, underwater sensors that measure lake currents.
Kevin Rose, a postdoctoral associate at UW Madison, who is active in GLEON, says IBM’s involvement makes the Jefferson Project stand out. “Private-public partnerships are going to be a hallmark of how more research is done in the future and this is a great model to see that in action,” he says.
The ultimate test of the Jefferson Project’s value, Hanson says, will be whether local and regional officials are able to use the information to better manage and protect the body of water known as “the Queen of American Lakes.”And project director Relyea says they are aiming high. “Ultimately,” he adds, “our goal is to make this project a blueprint for understanding lakes” that can be replicated elsewhere.
The project, which is expected to run for at least 3 years, is jointly funded by the three groups; leaders say it has a total budget “in the millions,” including direct spending and in-kind contributions. Researchers expect the Jefferson Project to have all of its systems fully integrated by the end of 2016.