Glossary

Dark Vessel Detection

Dark Vessel Detection identifies ships whose AIS transmitter is off or unresolvable, inferring presence from satellite radar (SAR) detections and AIS-gap events where the broadcast record goes silent. Global Fishing Watch publishes SAR vessel detections from 2017 to about 5 days ago alongside AIS-gap events.

What is Dark Vessel Detection?

A dark vessel is a ship whose Automatic Identification System transmitter is switched off or cannot be resolved, which makes it invisible to anyone relying on the broadcast record alone. Detection therefore comes from a second sensor: synthetic aperture radar, which images metal hulls regardless of what the ship broadcasts. In the cataloged Global Fishing Watch records, the tags include "dark-vessels" and the offering includes "AIS-gap events"; SAR vessel detections run "2017 to about 5 days ago", complementing the AIS-derived record. The same API family covers "encounters, loitering, port visits and AIS-gap events", so a gap is never interpreted in isolation - it sits next to meeting events, station-keeping behavior and port calls built from the same behavioral layer that produces apparent fishing effort.

Why does Dark Vessel Detection matter when choosing a dataset?

Any monitoring workflow premised on AIS completeness fails exactly where enforcement cares most. A vessel that switches off its transceiver disappears from broadcast-only products while continuing to fish, transfer cargo or loiter near protected zones - which is why counter-illegal-fishing programs treat gap detection as core rather than optional. Recency is the second trap: SAR detections running "2017 to about 5 days ago" support historical investigation and near-real-time screening alike, but the roughly five-day lag means yesterday's incident may not yet appear, so operational decisions need a latency check rather than an assumption. Cost asymmetry matters too - AIS data is cheap and plentiful, while radar detection capacity is scarce, so vendors differ sharply in how much of the dark fleet their products can plausibly see.

How do you evaluate Dark Vessel Detection in a data source?

  1. Ask whether SAR detections complement the AIS record Global Fishing Watch pairs its AIS products with "SAR vessel detections: 2017 to about 5 days ago" - a source without radar sees only broadcasting ships.
  1. Check detection history depth The documented window runs from 2017 to about 5 days ago, which bounds both historical studies and how current operational alerts can be.
  1. Look for AIS-gap event products The cataloged APIs cover "encounters, loitering, port visits and AIS-gap events" - gap events are the explicit signal that a transmitter went quiet.
  1. Test the behavioral context layers Encounter and loitering layers let you judge whether a silence followed a rendezvous pattern typical of transshipment.

Neighboring entries cover the signals and behaviors this detection technique builds on:

apparent-fishing-effort estimates fishing activity from AIS track behavior, the record dark vessels drop out of. automatic-identification-system is the broadcast signal whose absence defines a vessel as dark.

Browse curated sources in marine-transportation data, where these detection products are cataloged.

Frequently asked questions

What is a dark vessel?

A ship whose AIS transmitter is off or unresolvable. Because it stops appearing in the broadcast record, detection depends on satellite radar imagery and AIS-gap events rather than the vessel's own transmissions.

How recent is dark vessel detection data?

Global Fishing Watch's cataloged SAR vessel detections run from 2017 to about 5 days ago, so the archive supports decade-scale investigation while leaving a lag of roughly five days before the latest activity appears.

Datasets containing this field

Datasets containing Dark Vessel Detection

6 datasets carry dark vessel detection in the catalog. Open one, count the fields, judge for yourself.

Every listing shows the field dictionary, sample rows, and coverage before you commit. API, files, or your warehouse. Daily, weekly, or hourly.

Get sample rows