Author: Paul Tullis
Hundreds of thousands of fishing vessels. Millions of square miles of ocean. Billions of radio transmissions. The constant stream of data can overwhelm even the most dedicated fisheries managers trying to combat the $23 billion illegal fishing industry.
In economically underdeveloped countries, a small team of analysts must pore over the surveillance profiles of thousands of fishing enterprises; often the environmental cops can be as much as six months behind. By the time they see that a vessel in their jurisdiction is acting suspiciously, the ship has sailed.
That’s according to Chris Wilcox, principal research scientist in the oceans and atmosphere business unit for Commonwealth Scientific and Industrial Research Organization (CSIRO), the Australian government’s main research arm.
So he and colleagues are deploying algorithms to help authorities make better and faster decisions about which vessels to pursue or inspect. “We want to help port inspectors and fisheries managers make quick, informed decisions and target high-risk vessels,” Wilcox told Oceans Deeply.
International maritime law requires that large ships use an automated identification system (AIS), which transmits the vessel’s identity, direction, speed and sometimes other data as a means of avoiding collisions. Around 15 years ago, someone discovered that these signals are detectable by satellite. Wilcox’s team leverages this data, which is virtually free.
“We designed a bunch of statistical algorithms that look for suspicious patterns in how vessels move around,” he explained. The formulas look for three things: Whether the vessel moves in an unusual way given the type of vessel it is and its location, whether it seems to be shutting off its I.D. transponder, which there is really no good reason to be doing, and whether the vessel has a history of sailing in waters of countries with high levels of crime or corruption, or in areas with high levels of illegal fishing.
“A cargo vessel sitting just outside the maritime border of a country – that’s weird,” Wilcox said. Such ships are expensive to operate and ought to be moving from port to port as quickly as possible. Staying put might indicate it’s taking on cargo from other vessels trying to avoid catch limits.
According to the United Nations Food and Agriculture Organization, illegal fishing costs nations billions of dollars a year and contributes to the collapse of fish populations around the world. “We have to find a way to get the reins on that problem,” said Jackie Savitz, a senior vice president at the environmental nonprofit Oceana. “Otherwise we will not be able to provide a fish meal every day to the 1.5 billion people around the world” who rely on the ocean for protein. Successful management of depleted fisheries in the United States has led to their recovery, according to the National Oceanic and Atmospheric Administration. “These are tools that can help us,” Savitz said of CSIRO’s new system.
Based on the AIS data, vessels are assigned a risk score. Authorities can register an area of interest – where in their territorial waters they suspect illegal fishing may be occurring, for instance, or an area they don’t have the capacity to patrol. Whenever a vessel identified as suspicious enters that area, the authorities receive an alert. They can also query CSIRO’s database with a specific vessel and receive a risk report. Port inspectors can use the information to help decide which of the many vessels entering port they ought to board. The system, called Monitoring, Control and Surveillance Analytics for Illegal, Unreported and Unregulated Fishing, launches in October. It was developed in collaboration with Vulcan, the Seattle-based company owned by Microsoft cofounder Paul G. Allen.
“This valuable tool will enable enforcement agencies to identify and locate suspicious vessels all over the world,” Mark Powell, illegal fishing program officer for Vulcan, said in a statement released by CSIRO.
Wilcox’s algorithms build on Global Fishing Watch, a free online tool Oceana helped develop with Google and SkyTruth to enable anyone with an Internet connection to monitor fishing activity anywhere on Earth. “The value add [with CSIRO’s system] is the risk report, which a really good idea,” Savitz said. “Scoring based on suspicious behavior and allowing port authorities to prioritize their attention is a good strategy.”
Global Fishing Watch and CSIRO’s tool are just the first of several new weapons to come in the fight against high crimes on the high seas. The first international agreement against illegal fishing, the international Port State Measures Agreement, became law last year, encouraging the development of such tools. CSIRO is focused on repurposing low-cost, widely available data not designed for surveillance. AIS is one such data set; another comes from the Visible Infrared Imaging Radiometer Suite, a sensor on a NOAA satellite that looks for visible light. It was conceived as a means of tracking human development, but it turns out that VIIRS can also see large ships.
“If you’re a fisheries manager looking for noncompliant vessels over a large area, VIIRS is something you can use and it doesn’t cost anything,” Wilcox said. “But they need someone to set up the infrastructure to process the data.”
His team also hopes to deploy hydrophones to listen for coral reef-destroying dynamite fishing, one of the most ecologically destructive methods of harvesting the ocean. Scientists use hydrophones to track behavior of marine life and underwater volcanic activity, and the US military uses them to listen for submarines. CSIRO hopes to develop a system that listens for sounds that might indicate dynamite fishing and automatically alerts the local maritime authority.
Vulcan last month committed, as part of the U.N.’s Sustainable Development Goal 14, to develop a detection system to give law enforcement evidence of illegal fishing activity in near-real time. Together with its cofunding of CSIRO’s system, the company is working both the surveillance approach and the analytics approach to combat illegal fishing.
“There’s really no such thing as too much work in this space,” Savitz said. “It’s a planet-sized problem, and we need planet-sized solutions. If we lay off these depleted fish populations, they will start to rebuild.”