The vast majority of the world's industrial fishing vessels are not publicly tracked. Three-quarters of the world's large fishing vessels and one-quarter of transport and energy vessels are “dark vessels” that do not share their location publicly. The discovery comes from analyzing satellite imagery using artificial intelligence, an approach that could help better track human activities that impact the ocean.
“We had this idea that we were missing a large part of the activity that was happening in the ocean, but we didn't know how much we were missing,” he says. Fernando Paolo At Global Fishing Watch, a nonprofit organization based in Washington, DC. “And it turned out to be so much more than we had imagined.”
Paolo et al. used satellite images taken between 2017 and 2021 covering coastal areas where large-scale fishing and other industrial activities occur most often, revealing objects regardless of clouds or darkness. (including radar images that can be used). The researchers trained several AIs to detect and classify boats and marine structures in this dataset.
Researchers compared the ship's global map with a database of ships that publicly broadcast their locations and found that the vast majority did not have automatic identification systems turned on. Such identification is not always required, but if it is not used, it may indicate illegal fishing or other activity.
An AI learned to distinguish fishing boats from other types of boats based on their movement patterns and location. It found that 42 to 49 percent of approximately 63,000 ships fell into this category.
Other AI has identified 28,000 offshore structures related to wind power and oil production, with rapidly growing swarms of offshore wind turbines outnumbering oil infrastructure such as oil rigs. Although such activities other than offshore development and fishing boats are expanding, fishing activity is almost “at its limit,” he said. david kurzma At Global Fishing Watch.
“We have to plan for all non-fishing activities because we are encroaching on fishing grounds,” Krusma said. “The ocean is becoming increasingly crowded, so we need to consider how everything fits together.”
Publicly available satellite images do not have the resolution to detect small fishing boats less than 20 meters in length. Konstantin Kremer with microsoft Esther Rolfe at Harvard University Nature Articles commenting on research. But they said such efforts could improve monitoring of human activity near protected areas and unregulated parts of the ocean.
topic:
Source: www.newscientist.com
