ST. PETERSBURG, Fla. — According to scientists, some parts of our ocean ecosystems are on life support, and swift action is needed to save them. Satellite images could be AI's new "eyes" in the sky.
At the USF St. Petersburg campus, hundreds of leaders from across the globe met for the fifth International Ocean Colour Science (IOCS) meeting convened by the International Ocean Colour Coordinating Group (IOCCG) in partnership with the University of South Florida, NASA, and NOAA.
The conference's goal is to bring the brightest minds in their field together to find solutions and advance novel ideas to combat issues like red tide, algal blooms, oil spills, coral bleaching, marine biota collapse, Sargassum inundation, sea level rise, and extreme heat. This summer, the ocean temperature reached a possible record 101.1 degrees.
"This (conference) is extremely important. And you know, in the old days, people go to the field and take a couple of measurements to tell, this is what's happening. But they have no way to tell the large picture. And this conference is about using the most cutting-edge technology to look at a large picture, global picture," Dr. Chuanmin Hu, Professor of Oceanography at the USF College of Marine Science, told ABC Action News reporter Michael Paluska. "We need more people. And this is a part of the objectives of this conference, to let people know what we are doing, what we are lacking, and how we can mitigate or fill the gaps in the future."
"If the oceans collapse, can humanity survive?" Paluska asked.
"No way," Dr. Hu said. "Oceans provide half of the oxygen of the entire atmosphere. Perhaps, and most people don't realize that. If there's no ocean, we don't have half of the oxygen. How do we breathe?"
"How bad is it?" Paluska asked.
"Right now, we are okay. But it could be too late to take action," Dr. Hu responded.
"How much danger are Floridians in since we're surrounded by water?" Paluska said.
"Floridians should be more aware of those problems than others," Dr. Hu said. "They're under pressure, stress of hurricanes, and harmful algal blooms. And occasionally oil spills and heatwaves. This past summer, we had a heatwave, and we all felt it was so hot. The water feels even more severe, and the coral is dyed or bleached under heat. So we have all kinds of problems."
One potential solution is using AI and machine learning to highlight parts of the ocean that are healthy or in distress.
"In the open ocean, it's very much blue or very deep blue," Dr. Brian Barnes, Research Assistant Professor at USF College of Marine Science, said. "And then as you get into coastal systems, you'll get more greens, and depending on what kind of what's in the water, you'll get a white if it's sediment or kind of like a dark brown if it's, you know, like runoff from river tannins and things like that. So, depending on what place you're in, is whatever is natural for that color to make."
Ocean color science is precisely what it sounds like studying colors with millions of data points, including satellite images, to forecast red tide events, for example, or debris fields following hurricanes.
"What we're trying to do is really zoom in on the coastal regions, and couple what we can detect from satellites being, you know, individual patches of Sargassum and say, we couple that with some models saying which way the water is moving. And from that, we can extrapolate and say this particular beach may be impacted by Sargassum within the next couple of days," Dr. Barnes said.
"How long would it take for like a human to crunch data like that?" Paluska asked.
"It's unfathomable," Dr. Barnes responded. "It cannot physically be done to look at all of the individual images that are collected. These machine learning algorithms, the fast computers we use, allow us to project that out. Quick ingest, quick processing, and then distribute those data to people impacted by particular Sargassum inundation or other coastal hazards."