Fishing fleet activity monitoring for fisheries-bond and quota compliance
SAR satellites detect fishing vessels regardless of whether their transponders are switched on. Correlating radar detections with AIS and VMS records identifies dark fishing activity, giving bond issuers and ESG auditors an independent, physics-based view of quota compliance.
Sensors
- Sentinel-1 (ESA): C-band SAR at 5.405 GHz. Interferometric Wide Swath mode delivers 250 km swath at 5 x 20 m resolution, with a 6-day repeat at the equator (3-day with both satellites when both are operational). Free and open; suitable for wide-area EEZ surveillance on a routine basis.
- RADARSAT-2 (MDA): C-band SAR with selectable modes from 3 m (Spotlight) to 100 m (ScanSAR). The 3 m Spotlight mode resolves vessel superstructure detail useful for vessel-type classification. Commercially tasked; revisit depends on latitude and purchased capacity.
- ICEYE SAR constellation: X-band SAR (9.65 GHz) with Spotlight resolution down to approximately 1 m and Strip mode at roughly 3 m. The constellation offers same-day revisit to any point on Earth under commercial tasking. X-band is more sensitive to smaller metallic structures than C-band, aiding detection of mid-sized vessels.
- AIS / VMS records: Automatic Identification System (AIS) broadcasts vessel identity, position, speed and heading. Vessel Monitoring Systems (VMS) are fisheries-authority-mandated transponders with position reports typically every 1-2 hours. Neither is tamper-proof; AIS can be spoofed or simply switched off. Both are essential as the reference layer against which SAR detections are matched.
Why a fishing vessel cannot hide from a radar satellite
Steel and aluminium hulls, masts and winch frames are efficient radar reflectors. At C-band and X-band frequencies, Bragg-resonance backscatter from metallic superstructures produces a bright return against the darker ocean surface, making vessels visible in SAR imagery even at night and through cloud cover. A typical ocean-going trawler of 30 to 60 metres length presents a radar cross-section large enough to be detected reliably in Sentinel-1 IW mode at 20 m ground resolution, provided sea state is moderate.
The detection limit is not a secret: small wooden or fibreglass vessels below roughly 10 to 15 metres length have a radar cross-section that can fall below the noise floor in standard wide-swath modes, particularly when sea clutter is elevated by winds above Beaufort 5. This is an honest constraint. High-resolution spotlight modes on RADARSAT-2 or ICEYE push the detectable size threshold down, but at the cost of narrower swath and higher tasking expense. For a bond or quota-compliance programme, the practical approach is to accept that artisanal small-scale fleets operating wooden vessels are underdetected, and to scope the financial instrument accordingly.
The dark vessel problem in fisheries finance
Fisheries-backed bonds and quota-linked instruments depend on catch volumes that are self-reported by operators and aggregated by flag-state authorities. The incentive to under-report or over-fish is obvious. IUU fishing, which the FAO estimates accounts for a substantial share of global catch though precise figures are contested, directly erodes the stock value that underlies the bond collateral. An investor holding a fisheries bond in a Pacific tuna zone has no independent visibility into whether licensed vessels are staying within their quota areas, or whether unlicensed vessels are fishing the same stock.
SAR-AIS correlation changes that. The method is straightforward: run a vessel-detection algorithm over a SAR scene covering the licensed zone, extract centroid positions and estimated vessel lengths from the radar returns, then cross-reference each detection against AIS and VMS position logs for the same acquisition window. Detections with no matching AIS or VMS record within a plausible positional tolerance are flagged as dark contacts. The analyst then assesses whether the dark contact is a licensed vessel that has switched off its transponder, an unlicensed intruder, or a false alarm caused by sea clutter or fixed infrastructure such as a fish aggregating device buoy.
What the physics gives you, and what it does not
Sentinel-1's 6-day revisit (or 3-day when the constellation is complete) is adequate for trend monitoring across a season but insufficient for real-time interdiction. ICEYE's commercial constellation can revisit a given ocean area multiple times per day under a tasking contract, which narrows the window in which a vessel can fish undetected. Even so, a vessel that fishes for four hours between SAR passes leaves no direct evidence in the imagery; only catch-landing records or VMS gaps can fill that window.
Vessel-length estimation from SAR is approximate. Published studies using Sentinel-1 and RADARSAT data report length estimation errors of roughly 10 to 20 percent for vessels above 20 metres, which is sufficient to distinguish a large industrial trawler from a small artisanal boat but not to make precise vessel-class assignments. Heading can be inferred from wake geometry when the vessel is moving at speed. Vessel type, flag state and operator identity cannot be read from SAR alone; they require AIS correlation or, where AIS is absent, cross-referencing against optical imagery or port-call records.
From radar detections to a compliance report a bond trustee can use
The analytic chain for a fisheries-bond application typically runs in three stages. First, a baseline fleet census is established for the licensed zone: all vessels with valid licences are registered with their AIS MMSI numbers and expected VMS reporting schedules. Second, SAR acquisitions are scheduled across the zone at a cadence agreed with the bond trustee, typically weekly or fortnightly for routine monitoring with surge tasking triggered by anomaly alerts. Third, each acquisition is processed through a vessel-detection pipeline, dark contacts are flagged, and a compliance summary is issued showing the ratio of detected vessels to licensed vessels, the number and estimated size of dark contacts, and any spatial clustering near zone boundaries that might indicate quota-area violations.
The output is not a prosecution file. It is a statistically grounded signal that a bond trustee or ESG auditor can use to ask harder questions of the operator or flag-state authority. If dark-contact rates in a given zone rise over consecutive observation periods, that is material information for a bond covenant review. Satellize structures this kind of periodic monitoring programme for sovereign and institutional clients; the methodology is the same one underpinning fisheries-surveillance work published by organisations including Global Fishing Watch.
ESG supply-chain applications: where the audit trail starts at sea
Seafood processors and retailers face growing pressure to demonstrate that their supply chains are free of IUU-caught fish. The challenge is that IUU catch enters the legal supply chain at landing, where documentation can be falsified. Satellite monitoring of fishing grounds provides an upstream signal: if a supplier's named vessels show consistent AIS gaps in a protected zone, or if SAR detections in that zone exceed the number of licensed vessels on record, the supply-chain auditor has a basis for further inquiry before the fish reaches port.
This application is distinct from the bond-compliance use case in one important respect: the relevant geographic scope is often much larger, spanning multiple EEZs and high-seas areas, and the monitoring is continuous rather than tied to a specific financial event. Archive depth matters here. Sentinel-1 data extends back to 2014, and RADARSAT-2 to 2007, which allows retrospective analysis of fishing pressure in a zone before a supplier relationship was established. That historical baseline is something no port inspection can reconstruct.
Typical figures
| Spatial resolution (wide-area mode) | 5 x 20 m (Sentinel-1 IW); 8 x 8 m (RADARSAT-2 ScanSAR Narrow) |
| Spatial resolution (spotlight mode) | ~1 m (ICEYE Spotlight); ~3 m (RADARSAT-2 Fine Quad) |
| Swath width | 250 km (Sentinel-1 IW); 50-500 km depending on RADARSAT-2 mode; 5-30 km (ICEYE Spotlight/Strip) |
| Revisit interval | 3-6 days (Sentinel-1, equatorial); sub-daily possible (ICEYE commercial tasking) |
| Radar frequency | C-band 5.405 GHz (Sentinel-1, RADARSAT-2); X-band 9.65 GHz (ICEYE) |
| Minimum detectable vessel length (typical sea state) | ~15 m in Sentinel-1 IW; ~8-10 m in ICEYE Spotlight; smaller vessels unreliable in high sea clutter |
| AIS/VMS positional tolerance for matching | Typically ±500 m accounting for AIS broadcast latency and SAR acquisition geometry |
| Archive depth | Sentinel-1 from 2014; RADARSAT-2 from 2007; ICEYE from 2018 |
| Processing latency (routine) | 6-24 hours post-acquisition for automated detection; 24-72 hours for analyst-reviewed compliance report |
| Delivery formats | GeoJSON vessel-detection layer, PDF compliance report, CSV dark-contact log, GIS-ready shapefile |
Analytics Satellize can run
| Dark vessel detection layer | Constant False Alarm Rate (CFAR) detector applied to calibrated SAR sigma-naught backscatter, followed by AIS/VMS positional cross-match within acquisition window | GeoJSON point layer of unmatched SAR detections with estimated vessel length, heading (where wake visible) and confidence score; updated per acquisition |
| Licensed-fleet compliance ratio | Comparison of SAR-detected vessel count within defined zone polygon against licence registry; dark-contact rate expressed as percentage of total detections | Periodic compliance summary table (weekly or fortnightly) formatted for bond trustee review, with trend chart across monitoring period |
| Zone-boundary incursion alert | Spatial intersection of dark-contact centroids with licensed-zone boundary buffer; clustering analysis to distinguish transiting vessels from fishing behaviour (low speed, loitering pattern) | Alert notification with georeferenced contact positions, estimated vessel size and timestamp; delivered within 24 hours of SAR acquisition |
| Historical fishing-pressure baseline | Multi-year SAR archive analysis (Sentinel-1 2014-present) to compute vessel-density heatmaps by season; compared against licence records for the same periods | Retrospective PDF report with seasonal density maps; suitable for ESG due-diligence and supply-chain audit packages |
| Vessel-type classification | SAR-derived length and backscatter profile compared against AIS vessel-type codes and optical imagery cross-reference where available; rule-based classifier distinguishing trawlers, longliners and carrier vessels by size range | Annotated vessel-detection layer with probable vessel-type field; confidence intervals stated explicitly |
| AIS gap analysis for named vessels | Time-series analysis of AIS broadcast continuity for vessels on the licence registry; gaps correlated with SAR detections in the zone to distinguish transponder failure from deliberate deactivation | Per-vessel AIS-gap log with SAR corroboration flag; formatted as CSV and narrative summary for flag-state or bond-covenant reporting |
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.