New Study Uses Satellite Images and AI to Illuminate Hidden Ocean Industrial Activities
A groundbreaking study published in the prestigious journal Nature has revealed the previously hidden extent of human industrial activities across the world’s oceans. The study, led by Global Fishing Watch in collaboration with researchers from Duke University, the University of California, Santa Barbara, and SkyTruth, employed a combination of satellite images, vessel GPS data, and artificial intelligence (AI) to track and analyze human activities in the ocean over a five-year period.
The findings of the study highlight the urgent need for transparency in monitoring ocean activities. The researchers emphasize that by shedding light on this blue acceleration of industrial growth at sea, critical environmental degradation, overexploitation of fisheries and marine resources, and lawless behavior such as illegal fishing and human trafficking can be prevented.
Until now, the sheer size of the ocean has made it impractical to track industrial activities on a large scale. However, the newly developed methodology used in this study provides the most comprehensive and accurate picture of industrial uses of the ocean to date.
One of the challenges faced in tracking industrial activities at sea is the reliance on the Automatic Identification System (AIS), a device carried by many vessels that broadcasts identity, position, course, and speed. However, AIS systems have limitations, including blind spots, vessels not being required to use them, and the potential for vessels engaged in illegal activities to disable or tamper with them. Additionally, location data for offshore structures such as oil platforms and wind turbines is often incomplete, outdated, or kept confidential.
To address these limitations, the researchers utilized AI models to analyze 2 million gigabytes of satellite-based radar and optical images taken between 2017 and 2021. They then matched these results with 53 billion AIS vessel position reports to identify vessels that were publicly trackable at the time of the image. Through this innovative approach, the researchers discovered that a significant amount of industrial activity occurs outside of public monitoring systems.
One remarkable finding was that approximately 75% of the fishing vessels detected were missing from public AIS monitoring systems, with a high concentration of this activity around Africa and South Asia. This previously invisible fishing activity significantly altered the understanding of the scale, scope, and location of global fishing.
The study also revealed that around 25% of non-fishing vessels, primarily transport- and energy-related, were missing from public AIS monitoring systems. Many of these missing vessels were in areas with poor AIS reception, suggesting that they may have broadcast their locations, but satellite detection was limited.
Moreover, the researchers identified more than 28,000 offshore structures, including oil platforms, wind turbines, piers, and aquaculture farms. The number of wind turbines in the ocean surpassed the number of oil structures in 2020, indicating the growing importance of offshore wind energy.
The availability of this comprehensive data through the Global Fishing Watch data portal holds significant potential for monitoring and managing ocean resources effectively. Fishery managers in developing countries can use this information to monitor local stocks in data-poor regions, while enforcement agencies can identify and target illegal fishing activities. The data can also aid in tracking maritime activities that may breach international economic sanctions and contribute to climate change mitigation efforts. In addition, the mapping of offshore energy impacts will enable a better understanding of environmental footprints and the identification of sources of marine pollution.
The research team has made the data freely available, allowing for ongoing updates and expansion. By providing this unprecedented level of information about human activities in the ocean, decision-making processes related to ocean management can be evidence-based, fair, effective, and sustainable.
The study’s findings underscore the vital importance of transparency and comprehensive monitoring of ocean industrial activities. The combination of satellite images, vessel GPS data, and AI technologies has opened new avenues for understanding and managing the world’s oceans. With this newfound knowledge, stakeholders can work towards sustainable practices that ensure healthy oceans for generations to come.
