New study provides first global dataset examining intentional disabling of automatic identification system devices across commercial fisheries
The ocean is vast and ship crews rely on several tools to navigate it safely. One of these tools is the automatic identification system, commonly referred to as AIS, which uses GPS transponders to regularly broadcast critical information, such as a vessel’s location, identity and speed to nearby ships, satellites and receivers on land. Designed as an anti-collision mechanism, AIS was first required by the International Maritime Organization for particular ships with the intent of boosting safety on the sea. But it soon became a valuable device for monitoring compliance and tracking fishing behavior globally.
AIS can generally be turned off without penalty, unlike a government-mandated vessel monitoring system, or VMS, which is trackable by a ship’s flag State—the country in which it is registered. Access to VMS data is often tightly restricted by national governments, so AIS disabling may prevent other governments, coastal States and nearby ships from detecting such vessels. Therefore AIS disabling can limit the usefulness of the device as a fisheries monitoring tool and ultimately hinder transparency of fishing activity at sea.
In a recently published paper, “Hotspots of Unseen Fishing Vessels,” Global Fishing Watch, along with NOAA Fisheries and the University of California, Santa Cruz, provides the first-ever global dataset examining intentional AIS disabling events across commercial fisheries, identifying hotspots of where these events occur, as well as what may have caused them.
What are the challenges of AIS?
AIS is a really powerful tool for monitoring vessel activity, but it wasn’t originally designed that way. It is a maritime navigation safety system that was created to help vessels track each other and avoid collisions. AIS devices can be intentionally disabled for a variety of reasons, including to hide productive fishing spots from competitors and to potentially mask illegal activity.
Not all AIS messages that are broadcast by vessels are recorded by receivers for technical reasons, such as poor satellite reception and limited range of receivers on land. So, as a result, it’s not uncommon to see gaps in AIS data, for hours or perhaps even days. This limitation of AIS presents a key challenge: how to differentiate suspected AIS disabling events from gaps in the data caused by technical issues.
How do current AIS regulations make it challenging to identify illegal disabling events?
AIS devices are not universally mandated. They are required by the International Maritime Organization for large vessels and those on international voyages, but most fishing vessels are exempt. Policies on AIS usage can also vary by region, regional fisheries management organization, flag State and vessel size. Some regulations also don’t require AIS devices to remain on at all times and allow them to be switched off for security concerns, such as transiting through heavily pirated areas. Therefore, while switching off an AIS device is generally not illegal on its own, the activities undertaken during the time the device is off could very well be outside the law.
How did you determine the drivers behind AIS disabling?
We used a popular type of machine learning model, which allows computers to “learn from experience,” to understand how the locations of AIS disabling and fishing activity differ. Our model included numerous environmental and behavioral factors that represented the quality of fishing in the area and the proximity to activities and boundaries important for proper oversight, including transshipment and exclusive economic zones (EEZ). Our results showed that motivations behind disabling a ship’s AIS went beyond just a desire to hide good fishing spots. Sometimes they occurred when a vessel was nearby the sovereign waters of another country or within close proximity of a refrigerated cargo vessel, which are used to transship catch.
How were you able to demonstrate strong links between suspected disabling events and behaviors associated with illegal, unreported and unregulated fishing?
We drew strong links between suspected disabling events and illegal, unreported and unregulated (IUU) fishing in two ways. First, we identified numerous hotspots of AIS disabling in regions known to have problems with illegal fishing, such as the North Pacific and West Africa. AIS disabling may limit the ability of enforcement agencies to detect IUU fishing in these regions if government-mandated VMS data is not accessible. Second, the results of our model revealed that AIS disabling was highest near areas often associated with IUU fishing. These areas can include transshipment activity—where fishing vessels transfer catch at sea to refrigerated cargo vessels—and jurisdictional boundaries like EEZs. Vessels were more likely to disable their AIS near the borders of foreign EEZs, particularly contested ones, suggesting that vessels may disable their AIS prior to operating in areas where they are not authorized.
The analysis showed that there is evidence that disabling may occur for legal activities unrelated to IUU – why are these events also important for providing insights into ecological impacts?
Our analysis identified chlorophyll—one indicator of fishing ground quality—as an important driver of AIS disabling, especially for trawlers. Chlorophyll is found in phytoplankton–the microscopic plants at the base of marine food webs–and high chlorophyll levels may signal areas with healthy marine food webs and an abundance of fish. This suggests, not surprisingly, that fishing vessels disable their AIS to hide good fishing grounds. While this practice may not be illegal, doing so may obscure fishing impacts on sensitive habitats from the views of fisheries managers if the vessel is not tracked by a vessel monitoring system.
Can this data be used to look at activity around marine protected areas and help identify potentially problematic behavior?
Yes! However, because AIS coverage nearshore can be inconsistent when terrestrial receivers are lacking and satellite reception is poor, our analysis only identifies potential AIS disabling events that take place more than 50 nautical miles from shore.This restriction limits the utility of this data for monitoring activity around most MPAs, which are predominantly located closer to shore, but it can provide insights for many of the largest protected areas around the world.
How can this dataset be used to improve fisheries management?
AIS data can tell us a lot, but so can the lack of it. Understanding when and where a fishing vessel’s location is intentionally hidden may signal potential blindspots for fisheries managers, including areas of potential IUU fishing concern. It can also help authorities prioritize where to deploy surveillance and enforcement resources to address those blindspots. Additionally, this information can help focus in-port IUU fishing inspections required by the Port State Measures Agreement. Ironically, it is the absence of AIS data itself that can contain a wealth of information and serve as a valuable tool in the data-deficient fight against IUU fishing activity.
Tyler Clavelle is a senior data scientist at Global Fishing Watch.