Potential fishing zone advisory from SST and chlorophyll
Pelagic species aggregate where thermal fronts and chlorophyll-rich upwelling edges concentrate prey. MODIS, VIIRS and Sentinel-3 make those features mappable at operational frequency, giving fishing fleets and fisheries managers a daily probabilistic advisory.
Sensors
- MODIS-Aqua: 1 km SST (11 and 12 µm thermal bands) and 1 km chlorophyll-a via ocean colour algorithms (OC3M). Daily global coverage from a single overpass, though cloud contamination frequently forces composite periods of 8 days or longer to achieve gap-free imagery in cloudy tropical and mid-latitude regions.
- VIIRS (Suomi-NPP and NOAA-20): 750 m day/night band and 375–750 m thermal infrared for SST; 750 m ocean colour for chlorophyll-a. Two satellites together provide twice-daily coverage at mid-latitudes. VIIRS SST is operationally produced by NOAA STAR and is considered a successor to MODIS for operational oceanography.
- Sentinel-3 OLCI: 21-band ocean and land colour instrument at 300 m spatial resolution, the highest-resolution freely available ocean colour sensor in routine operation. Chlorophyll-a retrievals follow the OC4Me algorithm. Revisit is approximately 2 days at the equator with two satellites (Sentinel-3A and 3B), better at higher latitudes.
- Sentinel-3 SLSTR: Dual-view thermal infrared radiometer co-located with OLCI on the same platform, providing 1 km SST in the 3.7, 10.8 and 12 µm channels. Simultaneous acquisition with OLCI means SST and chlorophyll fronts can be co-registered without inter-sensor timing errors, which matters when fronts move 5–20 km per day.
Why fish pile up where physics says they should
Pelagic species do not distribute themselves randomly. They follow prey, and prey follows nutrients. Nutrients reach the sunlit surface layer through upwelling, where wind-driven or topographically forced divergence lifts cold, nutrient-rich water from depth. The result is a thermal gradient at the surface, a front, and within days to weeks, a phytoplankton bloom measurable as elevated chlorophyll-a concentration. Zooplankton graze the bloom; forage fish eat the zooplankton; tuna, billfish and large pelagics eat the forage fish. The thermal front is therefore both a direct habitat preference signal and a proxy for the trophic chain beneath it.
This is not a new observation. Fishers have read sea-surface colour and temperature by eye for generations. What satellites add is daily, synoptic coverage of features that span hundreds of kilometres and shift position faster than a vessel can survey them conventionally. A front that was productive yesterday may have advected 30 km overnight. An advisory derived from yesterday's imagery is already wrong in detail; one derived from this morning's overpass is actionable.
What the sensors actually see, and what they miss
MODIS-Aqua and VIIRS retrieve SST from the difference in emitted thermal radiance at two atmospheric window channels, typically around 11 and 12 µm. Under clear sky, absolute accuracy is better than 0.5 °C for well-validated products. Chlorophyll-a is inferred from the ratio of water-leaving radiance in blue and green bands: high chlorophyll shifts the ocean colour from blue toward green. The OC3M algorithm used for MODIS and equivalent formulations for VIIRS and OLCI are well-characterised in open ocean conditions, with retrieval uncertainty typically cited at 30–35% for Case 1 waters. In optically complex coastal and shelf waters where coloured dissolved organic matter and suspended sediment co-vary with chlorophyll, uncertainty rises substantially and the algorithm can fail entirely.
Cloud is the dominant operational constraint. In persistently cloudy regions such as the western tropical Pacific or the Bay of Bengal during monsoon, single-day clear-sky coverage may be below 20% of the ocean area. Operational potential fishing zone (PFZ) advisories typically rely on 3-day or 8-day composite SST and chlorophyll fields to fill gaps, which means the advisory reflects conditions averaged over that window rather than a snapshot. Fronts that form and dissipate within 48 hours will be blurred or missed entirely. This is an honest limit of the method, and any advisory product should state the composite period explicitly.
Deriving the advisory: fronts, gradients and probability
A PFZ advisory is not simply a map of where chlorophyll is high. The productive signal must coincide with a thermal gradient strong enough to indicate a front rather than a diffuse bloom. Front detection algorithms, of which the histogram-based method published by Cayula and Cornillon in the 1990s remains a widely cited baseline, identify pixel windows where the SST distribution is bimodal, indicating two distinct water masses in contact. Gradient-based methods using the Sobel operator or similar edge-detection filters are computationally simpler and adequate for operational use at 1 km resolution.
The two layers, SST front position and chlorophyll-a concentration, are then combined. A simple overlay flags zones where both criteria exceed thresholds. More sophisticated approaches use species-specific habitat models that weight SST range, chlorophyll level and sometimes sea-surface height anomaly (from altimetry) to generate a probability surface. The output is a gridded advisory, typically at 1–4 km resolution, showing zones of low, moderate and high aggregation probability for a named target species group. Fisheries agencies in India (INCOIS), Thailand and several Pacific Island states have run operational PFZ services based on exactly this architecture for over a decade, demonstrating that the method is mature and institutionally accepted.
Resolution, latency and the practical limits of the advisory
At 1 km (MODIS, VIIRS SST) and 300 m (Sentinel-3 OLCI chlorophyll), the sensors resolve fronts that are meaningful at the scale of a fishing ground. They do not resolve sub-kilometre filaments or the fine-scale eddy structure that can concentrate fish over areas of a few hundred metres. A vessel's captain still has to search within the advisory zone; the satellite narrows the search area, it does not replace it.
Latency from satellite overpass to delivered advisory depends on processing chain design. Near-real-time VIIRS SST products from NOAA STAR are typically available within 3–6 hours of acquisition. Sentinel-3 Level-2 ocean colour products from the Copernicus Data Space are generally available within 3 hours of sensing. A well-engineered advisory pipeline can therefore deliver a daily product before the fishing fleet departs at dawn, which is the operationally relevant window. Archive depth for MODIS-Aqua extends to 2002, and for VIIRS to 2012, enabling multi-year climatological baselines against which anomalous frontal positions can be flagged.
Who uses this and for what decision
The primary user is a fisheries management agency or a fishing cooperative that wants to reduce fuel expenditure by directing vessels toward productive zones. Fuel is typically the largest variable cost in pelagic fishing operations, and a credible PFZ advisory that reduces search time by even a modest fraction has a clear economic case. Secondary users include maritime surveillance agencies that want to anticipate where fishing pressure will concentrate, and conservation bodies monitoring whether vessels are operating within or outside protected areas in relation to productive frontal zones.
Satellize runs PFZ analytics on open constellation data, including Sentinel-3 and VIIRS, and can configure species-specific probability layers for a client's target fishery and geographic domain. The same analytical infrastructure used in the Kingdom of Tonga crop-estimation programme, reading multi-sensor open data and delivering structured outputs, applies directly to ocean colour and SST compositing. Clients receive a daily GIS layer and a summary advisory report, with the composite period and cloud-cover fraction stated explicitly so the uncertainty is never hidden.
Typical figures
| SST spatial resolution | 1 km (MODIS-Aqua, VIIRS, Sentinel-3 SLSTR); 375 m (VIIRS day/night band thermal) |
| Chlorophyll-a spatial resolution | 1 km (MODIS-Aqua OC3M, VIIRS); 300 m (Sentinel-3 OLCI OC4Me) |
| Revisit frequency | Daily (MODIS-Aqua single overpass); twice-daily at mid-latitudes (Suomi-NPP + NOAA-20 VIIRS); ~2 days equatorial (Sentinel-3A + 3B OLCI) |
| SST retrieval accuracy (clear sky) | Better than 0.5 °C absolute for validated MODIS and VIIRS products in open ocean |
| Chlorophyll-a retrieval uncertainty | ~30–35% in open-ocean Case 1 waters; substantially higher in coastal or turbid Case 2 waters |
| Cloud impact | Single-day clear-sky fraction can fall below 20% in persistently cloudy regions; 3- or 8-day composites used operationally to fill gaps |
| Spectral bands used | Thermal infrared 10.8 and 12 µm (SST); visible blue (~443 nm) and green (~555 nm) ratio (chlorophyll-a) |
| Advisory product latency | 3–6 hours post-overpass for near-real-time VIIRS SST (NOAA STAR); ~3 hours for Sentinel-3 Level-2 ocean colour (Copernicus Data Space) |
| Archive depth | MODIS-Aqua from 2002; VIIRS from 2012; Sentinel-3A OLCI from 2016 |
| Minimum detectable front gradient | Typically ~0.5 °C over 5–10 km at 1 km resolution; weaker or narrower gradients may not be resolved |
Analytics Satellize can run
| Daily SST front map | Edge detection (Sobel gradient or Cayula-Cornillon histogram method) applied to composited VIIRS or MODIS SST fields | GeoTIFF layer showing front magnitude and position, delivered daily with cloud-cover fraction metadata |
| Daily chlorophyll-a concentration layer | OC4Me (Sentinel-3 OLCI) or OC3M (MODIS-Aqua) band-ratio algorithm applied to atmospherically corrected Level-2 data | GeoTIFF at 300 m or 1 km resolution, flagged for cloud, glint and turbid-water pixels |
| Potential fishing zone probability grid | Threshold overlay or weighted habitat model combining SST range, SST gradient magnitude and chlorophyll-a concentration for a named target species group | Classified GIS layer (low / moderate / high probability zones) in GeoTIFF and GeoJSON, with advisory summary PDF |
| Frontal anomaly alert | Comparison of current front position against multi-year climatological baseline (MODIS or VIIRS archive from 2002 or 2012 respectively) using percentile thresholds | Email or API alert when front position or chlorophyll level deviates beyond the 90th or 10th percentile for the calendar week |
| Seasonal productivity climatology | Monthly and 8-day composite statistics derived from full archive depth, producing mean, standard deviation and percentile fields for SST and chlorophyll-a | Static GIS atlas and tabular summary by fishing zone polygon, used for licence planning and stock assessment context |
| Composite quality report | Per-pixel valid-observation count and cloud-fraction statistics computed for each composite period | Companion layer delivered with every advisory product, allowing the client to assess spatial confidence before acting on the advisory |
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.