Ending Hide & Seek at Sea: Global Fishing Watch in Science

The Phoenix Islands Protected Area (PIPA), located in the central Pacific between Hawaii and Australia, is the world’s largest UNESCO World Heritage Site. Spanning a swath of ocean roughly the size of California, its hosts a series of isolated seamounts and almost entirely uninhabited islands, all supporting rich, largely unspoiled ecosystems.

On January 1st of 2015, this unique park was officially closed to industrial fishing. But given that the protected area is so huge and isolated, how can we possibly monitor such a park and be sure that that fishing vessels are staying out?

In previous posts, we showed how Global Fishing Watch has done exactly that—verifying that most fishing vessels are following the new regulations in the Phoenix Islands. Today, in the prestigious journal Science, members of the Global Fishing Watch team outline in more detail how this monitoring was possible, and we make policy recommendations that will improve our ability to monitor the rest of the world’s oceans.

The key technology behind Global Fishing Watch is AIS – the “Automatic Identification System” that almost all large ocean-going vessels are required to carry. AIS transponders broadcast a vessel’s location and identity every few seconds to every few minutes. The system was originally designed as a ship-to-ship collision-avoidance system, but now we can use it to track, via satellite, the movements of large fishing vessels across the globe. In the Science paper we outline the strengths and weaknesses of using AIS to track fishing effort.

To get a sense of what we can see using AIS, we created the map below, which shows the movements of all vessels in the world that carry AIS in 2015. To produce this map we processed four and a half billion data points from over 200,000 vessels.

Out of these roughly 200,000 vessels, more than 35,000 are commercial fishing vessels whose movements we can track and analyze. Global Fishing Watch has developed algorithms—and is working to refine these algorithms—that use the movements of fishing vessels to estimate when and where the vessels are placing their hooks and nets. In the paper we show that in the central Pacific, the fishing estimate of our algorithms correlates very well with official, direct observations of fishing.


We then used these algorithms to estimate the fishing activity in the Phoenix Islands Protected Area before and after the closure. Below is a video that shows how the fishing vessels cleared out of the marine reserve. The yellow and white heat map shows the intensity of fishing effort as measured by Global Fishing Watch. In 2014, the park was heavily fished; in 2015, there was almost no fishing.

Today, also using PIPA as a case study, our partners at Oceana published their own report on the role of Global Fishing Watch in fisheries management and monitoring. 

As we make clear in the paper, AIS is not perfect. It is used primarily by large vessels, and is thus better for monitoring fishing on the high seas than in small artisanal fisheries. Also, mixed regulations means that in some regions of the world almost all fishing vessels carry AIS, while in others only a few vessels do. Nonetheless, this technology is an enormous leap forward, and in recent years, more countries are requiring more vessels to carry AIS. As the paper describes, “hide and seek” on the world’s oceans may be coming to an end.

[Updated] Fishing Fleet at MH370 Search Site May Have Moved North

Vessels in the Fu Yuan Yu fleet previously seen near the MH370 search site suddenly disappeared from our monitoring system last week as the entire fleet and associated fishing gear stopped broadcasting AIS. One of the vessels has now reappeared, the Fu Yuan Yu 076 is giving a current location in the Bay of Bengal near India. However looking at the vessel’s track over the past three months there is good reason to be skeptical. While the vessel showed a likely fishing track they appeared in a very unlikely location, the mountains of Tibet.

What we see is in fact a latitude reversal where a vessel’s position in the Southern Hemisphere is flipped over the Equator to appear in the Northern Hemisphere. This is one of a number of patterns of false AIS locations we have documented. We’re not sure if these simply result from errors in transmission or if the vessel operator intends to make his vessel difficult to track. Fortunately we have developed some tools to pick these false locations out of our data and correct them. The reappearance today of what is likely a beacon attached to the vessel’s fishing gear (identifying as Fu Yuan Yu 076 08) with a mirror track in the Southern Hemisphere is very strong evidence that this is also the true location of the parent fishing vessel.

The vessel’s latest positions show them moving on an irregular course at slow speed so it’s possible that they have resumed fishing at this new location. Though we can’t see any of the other vessels in the Fu Yuan Yu fleet there is a good chance they are nearby. We’ll see if any of them reappear in the next few days.

Fu Yuan Yu 76 and gear 3-2-16
Here are seen two AIS tracks over the past 3 months. The fishing vessel Fu Yuan Yu 076 broadcasting an impossible location on land in the Northern Hemisphere and the associated Fu Yuan Yu 076 08 (likely fishing gear) mirroring the track but in the Southern Hemisphere. The track of the Fu Yuan Yu 076 08 confirms our suspicion that the true location of the parent vessel can be found by reversing the latitude of their broadcast positions. While none of the other vessels in the fleet can be located it may be that the group has moved north to this new location.