Science and Technology @ UCSB
The global fishing fleet is so big it can be seen from space. Really.Fishing activity now covers at least 55 percent of the world’s oceans — four times the land area covered by agriculture — and can now be monitored, in near real time, to the level of individual vessels. In fact, 70,000 vessels of the global fishing fleet traveled 460 million kilometers in 2016, equivalent to traveling to the moon and back 600 times.
Using satellite tracking, machine learning and common ship-tracking technology, scientists from UC Santa Barbara teamed up with colleagues at Global Fishing Watch, National Geographic Society’s Pristine Sea project, Dalhousie University, SkyTruth, Google and Stanford University to illuminate the extent of global fishing — down to single vessel movements and hourly activity. Their findings appear in the journal Science.
“I think most people will be surprised that until now we didn’t really know where people were fishing in vast swaths of the ocean,” said co-author Christopher Costello, a professor at UCSB’s Bren School of Environmental Science & Management. “This new real-time dataset will be instrumental in designing improved management of the world’s oceans that is good for the fish, ecosystems and fishermen.”
While the dataset is hundreds of times higher in resolution than previous global surveys, the total area of the ocean fished is likely higher than the 55 percent estimated. That’s because some fishing efforts in regions of poor satellite coverage or in exclusive economic zones with a low percentage of vessels using the automatic identification system (AIS) were not included.
The team used machine learning technology to analyze 22 billion messages publicly broadcasted from vessels’ AIS positions from 2012 to 2016, to answer the question, “What drives commercial fishing behavior?” Based solely on vessel movement patterns, the Global Fishing Watch algorithm was able to identify more than 70,000 commercial fishing vessels, the sizes and engine powers of these vessels, what type …