Illegal, unreported and unregulated fishing vessel detection

SAR imagery detects vessel hulls regardless of AIS status, exposing fishing activity inside EEZs and marine protected areas that transponder data alone would miss. Matching SAR detections against AIS and VMS gaps identifies the dark vessels most likely to be operating without authorisation.

Sensors

Sentinel-1 SAR (ESA): C-band SAR at 5.405 GHz. Interferometric Wide Swath mode covers 250 km at 5 × 20 m resolution; Extra Wide Swath covers 400 km at 20 × 40 m. Revisit is 6 days at the equator for a single satellite, 3 days with both operational. Free and open; the backbone of Global Fishing Watch's published detection pipeline.
ICEYE SAR constellation: X-band commercial SAR. Stripmap mode delivers approximately 3 m resolution; Spot mode reaches sub-1 m. Revisit to any ocean point can be less than 24 hours depending on tasking. Useful for confirming vessel class and gear type where Sentinel-1 resolution is insufficient.
Capella Space SAR: X-band commercial SAR with Spotlight mode at approximately 0.5 m resolution. At that resolution, vessel superstructure and deck gear are distinguishable, supporting gear-type inference. Tasking latency of hours makes it practical for rapid-response enforcement.
Spire AIS receiver network: Space-based AIS from a constellation of over 100 LEO satellites. Provides near-global vessel tracking with latency of minutes to a few hours. The gap between a Spire AIS track and a SAR-detected position is the primary signal for dark-vessel identification.
VIIRS Day/Night Band (NOAA/NASA): Detects vessel lighting at night. Squid and saury fleets using high-intensity lights are visible at 375 m resolution. Not vessel-specific, but useful for identifying fishing grounds where declared effort is absent, and for cross-checking SAR detections in known squid-fishing regions.

What a radar return tells you about a vessel you have never met

SAR backscatter from a vessel hull is governed by its radar cross-section (RCS). RCS scales with vessel size and the geometry of metal surfaces oriented toward the sensor. A 50-metre trawler presents an RCS orders of magnitude larger than a 10-metre artisanal boat, which means length class can be inferred from backscatter intensity and the physical extent of the bright return, even without any registry data.

Global Fishing Watch's peer-reviewed methodology, published in Nature (2022) and using Sentinel-1 as its primary SAR input, demonstrated detection of vessels down to roughly 15 metres in calm sea states. In high sea states, wave clutter raises the detection floor considerably. That is not a software problem; it is physics. Analysts working enforcement cases need to understand that a Beaufort 6 sea will suppress small-vessel detections in C-band imagery, and X-band commercial sensors offer only partial mitigation.

The gap between AIS and reality is where the crime lives

AIS is mandatory for vessels over 300 gross tonnes on international voyages and for all passenger ships, under SOLAS. Fishing vessels below those thresholds may be required to carry AIS under flag-state or coastal-state rules, but enforcement is uneven. VMS (Vessel Monitoring System) is separately mandated by many fisheries authorities for licensed vessels, but VMS data is rarely public.

The analytic workflow is a join, not a search. You take SAR-detected vessel positions across a defined area, match them against space-based AIS positions within a time window, and flag the residual: detections with no corresponding AIS broadcast. Those are dark vessels. Not all dark vessels are fishing illegally; AIS equipment fails, and some vessels are legitimately exempt. But a dark vessel inside a closed MPA, or inside a foreign EEZ with no licence on record, is a materially different risk category from one in open international waters.

Spire's space-based AIS is important here because terrestrial AIS receivers have a coastal range of roughly 40 nautical miles. Beyond that, only satellite AIS provides coverage. A vessel that silences its transponder 50 miles offshore and steams into a productive fishing ground will be invisible to coastal authorities relying on terrestrial AIS alone.

Vessel length as a proxy for fleet class, and why that matters for enforcement

Fisheries enforcement agencies need to distinguish an industrial trawler from an artisanal canoe. The legal regimes, the likely catch volumes, and the diplomatic consequences of interdiction differ enormously. SAR-derived length estimates, cross-referenced against vessel registries such as the FAO Global Record or Lloyd's, allow analysts to assign probable vessel class before a patrol asset is committed.

The method is not precise. SAR-derived length carries an uncertainty of several metres depending on incidence angle and vessel orientation relative to the flight path. A vessel measured at 45 metres in one pass might be 38 or 52 metres in reality. That range is usually sufficient to distinguish a 200-metre supertrawler from a 20-metre artisanal boat, but it will not reliably separate two trawlers of similar size. Honest use of this method means presenting length estimates with their uncertainty bounds, not as definitive identifications.

Revisit, latency and the practical limits of enforcement support

Sentinel-1's 3-to-6-day revisit is adequate for building statistical pictures of fishing pressure over weeks or months. It is not adequate for real-time interdiction. A vessel detected in a Sentinel-1 pass acquired at 10:00 UTC and processed by 14:00 UTC is already 100 nautical miles from its detected position if it is making 12 knots. Patrol vessels need a heading, not a historical position.

Commercial SAR constellations change this calculus. ICEYE and Capella can be tasked to revisit a specific ocean area within hours, not days. Used together with a Sentinel-1 detection that flags a suspicious cluster, a follow-up commercial task can narrow the position uncertainty enough to be operationally useful. The cost is real; commercial SAR tasking is priced per acquisition, and covering large ocean areas at high revisit is expensive. Clients need to define the geographic priority zone carefully rather than attempting to monitor an entire EEZ at commercial-SAR revisit rates.

Cloud cover is not a constraint for SAR, which is one of its principal advantages over optical sensors for maritime surveillance. Rain at very high intensity can introduce artefacts in C-band, but routine cloud and darkness have no effect.

What imagery cannot tell you, and what fills the gap

No satellite sensor currently in operation can confirm what species a vessel is catching, or how much. Catch identification requires biological sampling. Gear type can sometimes be inferred at very high resolution from the arrangement of deck equipment, but this is interpretive and not suitable as standalone legal evidence.

The strongest enforcement cases combine satellite detection with other data streams. AIS history from Spire or similar providers shows where a vessel has been over weeks or months. Port-state records show where it declared its catch. Cross-referencing a vessel's SAR-detected positions inside a closed area with its declared port landings can reveal discrepancies that constitute evidence of unreported catch, even without direct imagery of the hold.

Satellize structures this kind of multi-source correlation as a standard analytic product, drawing on the same open-constellation and commercial-tasking pipeline used in its Tonga crop-estimation programme. The output is a vessel-level activity report with position history, AIS gap events, estimated vessel class, and flag-state information, formatted for handoff to a fisheries authority or legal team.

Building a persistent picture rather than a single snapshot

Single-pass detection is useful for incident response. Persistent monitoring is what changes fishing behaviour. When fishing operators know that an EEZ is covered by a systematic SAR programme with near-daily commercial tasking and automatic AIS-gap flagging, the risk calculus for entering without a licence shifts.

The practical architecture for a small island developing state, which typically has a large EEZ relative to its enforcement budget, is a tiered system. Sentinel-1 provides free, systematic wide-area coverage. Automated detection algorithms flag anomalous vessel clusters or AIS gaps. Commercial SAR is tasked only against those flagged areas, keeping costs proportionate. Alerts go to a patrol coordination centre with enough lead time to intercept, or at minimum to build a documented record for port-state and flag-state enforcement actions. That record, accumulated over months, is often more effective than a single at-sea interception.

Typical figures

Spatial resolution (Sentinel-1 IW mode)	5 × 20 m (range × azimuth), 250 km swath
Spatial resolution (commercial SAR, Spotlight)	0.5–3 m depending on sensor and mode (ICEYE, Capella)
Revisit (Sentinel-1, equatorial)	3–6 days (2 satellites operational); single-satellite 6 days
Revisit (commercial SAR tasking)	Sub-24 hours to any ocean point, subject to constellation capacity and cost
Minimum detectable vessel length (C-band, calm sea)	Approximately 15 m (published Global Fishing Watch threshold)
AIS latency (space-based, Spire)	Minutes to a few hours depending on orbital geometry
SAR archive depth (Sentinel-1)	From 2014 (Sentinel-1A launch); accessible via Copernicus Data Space
Sensor frequency (Sentinel-1)	C-band, 5.405 GHz; all-weather, day/night
Delivery formats	GeoJSON vessel-detection layers, CSV position logs, PDF enforcement summary reports, GIS-compatible alert feeds
Coverage	Global ocean; Sentinel-1 prioritises coastal and EEZ zones; open-ocean commercial tasking by request

Analytics Satellize can run

Dark-vessel detection layer	SAR vessel detection (CFAR algorithm on backscatter) cross-referenced against space-based AIS within a configurable time window; residual detections flagged as AIS-absent	GeoJSON layer of dark-vessel positions with estimated length, RCS class and time of detection, updated per Sentinel-1 pass
EEZ intrusion alert	Spatial join of dark-vessel positions against client-supplied EEZ or MPA boundary polygons; alerts generated for detections inside closed zones	Near-real-time alert feed (email or API) with vessel position, estimated class, and distance from nearest legal boundary
Vessel activity history report	AIS track reconstruction from Spire historical data combined with SAR detection timestamps; gap events and position anomalies identified across a defined period	PDF or structured JSON report per vessel MMSI or SAR-assigned track ID, covering up to 12 months of activity
Fleet-pressure heatmap	Density aggregation of SAR detections and AIS fishing-activity hours (Global Fishing Watch methodology) over a defined area and time window	Raster heatmap (GeoTIFF) and summary statistics showing fishing effort by grid cell, suitable for stock-assessment or diplomatic briefing
Vessel class estimation	SAR-derived length measurement combined with RCS intensity; cross-referenced against FAO Global Record and public registry databases for probable vessel type	Tabular output with estimated length range, probable class (artisanal, coastal, industrial), and registry match confidence score
Night-light fishing-ground identification	VIIRS Day/Night Band detection of vessel lighting in target EEZ; compared against declared licensed-fleet positions to identify unlicensed light-fishing activity	Monthly GIS layer of night-light fishing detections with AIS correlation status, formatted for fisheries authority review

Sources

Who does the work

We can get this done for you. Satellize runs its own analyst desk and a strong science team. You do not buy a data feed and work out what it means; our people source the imagery, run the analysis described on this page, and hand you the answer with its confidence limits stated. Discuss this requirement.