This spring a camera will begin taking pictures of the Ohio River at California, Kentucky to identify rare but toxic algal blooms as much as a day before they become a danger to drinking water.
A partnership between Thomas More College, Northern Kentucky University, the Environmental Protection Agency and the Ohio River Valley Water Sanitation Commission (ORSANCO) is developing a network of cameras that will take pictures of the Ohio River and analyze the information in a computer algorithm.
Using the color of the river, the goal is to detect an algal bloom earlier than traditional methods, which use a microscope. Scientists would also like to find out what causes a bloom to form.
In 2015, an Ohio River algal bloom stretched hundreds of miles from West Virginia to Louisville forcing authorities to issue health advisories. This was the same year blooms in Lake Erie forced Toledo to shut down its drinking water supply.
Algae naturally occur in waterways, but harmful blooms are the product of excessive runoff of nutrients like phosphorus and nitrogen, combined with sunny, warm conditions and low-flow water levels. If people drink algae-contaminated water it can make them sick with gastrointestinal symptoms. Some people have reported breathing problems, along with skin, eye, and throat irritation, after contact with affected waters.
The first camera in this Ohio River project is in California, Kentucky at the Thomas More College Biology Field Station. Professor Chris Lorentz has partnered with Mike Waters, an NKU math professor. Lorentz says Waters has revolutionized the method used to detect algal blooms.
"So, traditionally, the detection method would involve gathering a water sample, looking at the water under a microscope, counting the cells, trying to detect the toxins, and that is still in vogue today, but what Dr. Waters has brought to the table is the ability to detect harmful algal blooms using a camera and taking photographs."
Waters' algorithm uses an artificial neural network. It models the way the brain solves problems.
“The images from the camera are fed once an hour to a server near my office," he says. "Those images are analyzed by a computer and over time should get better and better at recognizing a harmful algal bloom’s color.”
Waters has also developed a citizen scientist app he's testing in Lake Erie. It uses the same algorithm.
"The idea is that the application will be released to the public in the next couple of years. They will be able to use that information to help determine whether the water is safe or could have a harmful algal bloom and they should stay out," he says.
It will likely be four to five years before there are any concrete results with the river cameras and computer algorithm.