Fishing vessel activity monitoring during quota seasons
Combining spaceborne AIS with SAR-detected vessel positions gives fisheries managers an independent record of fleet effort during open quota seasons, filling the gaps that self-reporting and coastal radar leave behind.
Sensors
- Spire Global spaceborne AIS: A constellation of more than 110 low-Earth-orbit satellites receiving Class-A and Class-B AIS transponder messages. Revisit at high latitudes can reach 20-plus passes per day per location, dramatically outperforming coastal VHF receivers whose range caps at roughly 40 nautical miles. Message latency to delivery is typically under 30 minutes on the commercial feed.
- exactEarth HD AIS: A complementary spaceborne AIS service using a dedicated small-satellite constellation plus hosted payloads. Adds redundancy in mid-ocean regions and provides a second independent record useful for cross-validation when a vessel's transponder behaviour is disputed.
- Sentinel-1 C-band SAR (ESA): Dual-polarisation synthetic aperture radar at 5.405 GHz. In Interferometric Wide Swath mode, ground resolution is 5 x 20 m with a 250 km swath. Revisit at the equator is 6 days with a single satellite; at 60 degrees north or south it shortens to roughly 2-3 days due to orbit geometry. Detects vessel-sized radar cross-sections regardless of cloud or darkness.
- ICEYE X-band SAR: Commercial X-band constellation operating at approximately 9.65 GHz. Stripmap mode delivers roughly 3 m resolution; Spot mode reaches approximately 1 m. Tasking can be directed to a specific ocean area within hours, making it practical for surge coverage when a quota season opens or closes. Smaller wavelength than C-band gives higher sensitivity to small wooden or fibreglass hulls.
Why self-reporting alone cannot close the quota books
Quota management rests on a simple assumption: that the fishing effort applied to a stock is accurately reported. In practice, logbook entries and VMS pings are produced by the same operators whose income depends on the quota outcome. That is not an accusation; it is a structural problem that independent monitoring exists to resolve.
Vessel Monitoring Systems transmit position at intervals set by the flag state, commonly every hour or every two hours. A vessel can change fishing behaviour substantially within that window. AIS, which was designed for collision avoidance rather than compliance, transmits more frequently but can be switched off or manipulated. Neither system provides an independent record. Satellite observation does.
What AIS tracks actually tell you, and what they do not
A spaceborne AIS track gives you a time-stamped position sequence for every vessel broadcasting a valid MMSI. From that sequence you can derive speed over ground, heading changes and time spent in a given grid cell. The standard heuristic for distinguishing active trawling from steaming is speed: trawlers typically fish at 2-4 knots, while transit speeds run 8-12 knots. Purse-seiners show a different signature, with a rapid encircling manoeuvre followed by a near-stationary haul period.
The limits are real. Speed-based classification produces false positives when vessels slow to avoid weather or manoeuvre around gear. It produces false negatives when a vessel tows fast or drifts with current while hauling. Published studies using Global Fishing Watch's publicly documented methods report overall classification accuracy in the range of 85-90 percent for well-represented vessel classes, but accuracy drops for smaller vessels and less common gear types. AIS dark periods, where a vessel stops transmitting, are a separate problem addressed in the dark-vessel detection page in this library.
SAR as the independent witness
SAR does not depend on a transponder. It illuminates the ocean surface with microwave pulses and records the backscatter. A steel or aluminium hull of any size produces a strong point-like return well above the surrounding sea clutter, provided sea state is not extreme. At Beaufort 6 and above, wave clutter can mask smaller wooden vessels, so SAR detections have a practical lower size limit that varies with conditions and incidence angle.
The operational workflow is a cross-match: take every SAR-detected vessel position within a quota zone during the season window, compare it against the AIS record for the same time and location, and flag positions where no AIS transmission was received. That flag does not automatically mean a violation. The vessel may have experienced a transponder fault, or the satellite pass may have missed a short transmission. What it does produce is a list of events requiring explanation, which is precisely what a quota enforcement authority needs.
Sentinel-1's 6-day equatorial revisit is adequate for fleet-level statistics but insufficient for real-time interception. ICEYE tasking, directed to a specific area within hours of a request, fills that gap when the situation warrants it. The cost of commercial SAR tasking is a real constraint; it is most efficiently spent on areas of known high effort or suspected non-compliance rather than blanket ocean surveillance.
Building a fishing effort map for stock assessment
Post-season stock assessment models need spatially resolved estimates of fishing mortality. The standard input is catch per unit effort (CPUE) by grid cell, but the effort denominator is only as good as the reporting. Satellite-derived effort maps offer an alternative denominator.
The method is straightforward in principle. Assign each AIS position classified as active fishing a unit of effort proportional to the time interval since the last position. Aggregate those units into a grid, typically 0.1 or 0.25 degree cells for open-ocean fleets. Apply a correction factor derived from the SAR cross-match to account for vessels that were present but not transmitting. The result is a gridded effort surface that can be ingested directly into surplus production or age-structured assessment models.
The honest caveat is that the correction factor itself carries uncertainty. SAR coverage of the quota area during the season is rarely complete, so the dark-vessel correction is an extrapolation. Sensitivity analysis on that factor should be standard practice, and the assessment report should document it explicitly.
High-latitude and remote ocean coverage: where the gap is largest
Coastal AIS receivers cover the areas where most fisheries management infrastructure already exists. The coverage gap is in the Southern Ocean, the central Pacific, the Bering Sea and Arctic waters opening under retreating ice. These are also the areas where some of the world's highest-value quota fisheries operate and where flag-state oversight is thinnest.
Spire's constellation geometry means revisit frequency actually improves at higher latitudes, because polar orbits converge near the poles. A vessel at 60 degrees south may receive 20 or more satellite passes per day, compared with 6-8 at the equator. That is a counterintuitive advantage: the regions hardest to monitor by any other means are precisely where spaceborne AIS performs best. Sentinel-1's revisit similarly shortens at high latitudes for the same geometric reason, giving better SAR coverage of sub-Antarctic quota zones than of tropical ones.
Satellize runs quota-season monitoring analytics on open constellations and adds commercial SAR tasking on client licence; the architecture is the same one underlying the Tonga crop-estimation programme, adapted from land to ocean observation.
What a quota-season monitoring package delivers, and what it cannot guarantee
A well-constructed monitoring package produces three things: a fleet-level effort map by grid cell and time window, a vessel-level activity log with classified behaviour states, and a cross-match report listing SAR-detected positions without corresponding AIS records. These are inputs to enforcement and assessment decisions, not substitutes for them.
What the package cannot do is prove catch quantity from space. Vessel presence and behaviour are observable; the weight of fish in the hold is not. Connecting effort estimates to catch requires either the vessel's own logbook or an observer programme. Satellite monitoring strengthens the credibility of those self-reported figures by providing an independent check on where and when effort was applied. That is a meaningful contribution to quota integrity, but it is a partial one. Managers who treat satellite effort maps as a complete audit trail will be disappointed; managers who treat them as a structured cross-check on self-reporting will find them genuinely useful.
Typical figures
| AIS positional accuracy | GPS-derived, typically better than 10 m; accuracy is that of the vessel's own GPS receiver |
| Spaceborne AIS revisit (high latitude, >55°) | Up to 20+ passes per day per location; equatorial revisit 6-8 passes per day |
| Sentinel-1 SAR resolution (IW mode) | 5 x 20 m ground range; 250 km swath |
| Sentinel-1 revisit | 6 days at equator; 2-3 days at 60° latitude with single satellite |
| ICEYE SAR resolution | ~3 m (Stripmap); ~1 m (Spot mode) |
| Minimum detectable vessel (SAR) | Approximately 10-20 m length in moderate sea state (Beaufort 3-4); degrades at Beaufort 6+ |
| AIS data latency (commercial feed) | Typically under 30 minutes from message reception to delivery |
| Archive depth (Sentinel-1) | From October 2014 (Sentinel-1A launch); continuous operational record |
| Effort grid resolution (standard output) | 0.1° or 0.25° cells; finer grids available for inshore quota zones |
| Delivery formats | GeoTIFF effort rasters, GeoJSON vessel tracks, CSV event logs, PDF season summary report |
Analytics Satellize can run
| Gridded fishing effort map | Speed-and-heading classification of AIS tracks (published Global Fishing Watch methodology) aggregated to spatial grid cells by time window | GeoTIFF raster and GeoJSON polygon layer, one file per quota period |
| Vessel-level activity log | Per-MMSI state classification (steaming, fishing, at anchor, dark) derived from AIS position sequence and speed thresholds | CSV table with vessel identifier, flag state, gear type, classified state, position, timestamp |
| SAR cross-match report | Automated spatial join of SAR-detected vessel centroids against AIS record within ±15-minute time window; unmatched detections flagged for review | PDF report listing unmatched detections with image chip, position, and nearest AIS vessel if within 2 km |
| Quota-zone effort summary statistics | Aggregation of classified fishing hours by vessel flag, gear type and sub-zone polygon; comparison against declared effort from logbooks where available | Excel workbook with pivot tables; optional dashboard feed via API |
| Dark-period correction factor for stock assessment | SAR detection rate within AIS-covered periods used to estimate fraction of unobserved effort; bootstrapped uncertainty range reported | Coefficient table with confidence intervals, formatted for ingestion into standard stock assessment software |
| Season-on-season effort trend comparison | Year-over-year comparison of gridded effort surfaces using the Sentinel-1 archive; change detection on fleet spatial distribution | Comparative GeoTIFF pair and written interpretation in season summary report |
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.