Subsea cable route seabed disturbance detection

Multispectral water-quality indices and SAR surface roughness anomalies can detect turbidity plumes and seabed disturbance from cable-laying operations in shallow water, with honest limits in deep routes where no surface signature exists.

Sensors

Sentinel-2 MSI: 10 m resolution in visible bands, 20 m in red-edge (B5, B6, B7) and shortwave infrared (B11, B12). Five-day revisit at the equator with both satellites. Red-edge and SWIR bands are sensitive to suspended particulate matter and allow turbidity indices such as NDTI and SPM retrieval. Useless through cloud or in water deeper than roughly 20–25 m where bottom reflectance is lost.
Sentinel-1 SAR C-band: 6–25 m resolution depending on mode (IW mode standard at ~10 m range, ~14 m azimuth). Six-day revisit per satellite. C-band backscatter detects surface roughness anomalies caused by upwelling sediment and bubble columns in shallow water. Penetrates cloud and works at night, but the signal is ambiguous with wind-roughened sea states above roughly Beaufort 4–5.
Planet SuperDove: 3 m resolution, up to daily revisit in mid-latitudes. Eight spectral bands including red-edge at 705 nm, useful for suspended sediment mapping at finer spatial scale than Sentinel-2. Commercial tasking allows same-day acquisition over active cable-lay corridors. Archive depth shorter than Sentinel and costs apply.
MODIS (Terra/Aqua): 250–500 m resolution in ocean colour bands, one to two passes per day per sensor. Useful for tracking large, persistent turbidity plumes over tens of kilometres in very shallow shelf seas. Too coarse for vessel-scale attribution; best used as a wide-area screening layer to flag anomalous water colour before tasking finer sensors.

What suspended sediment looks like from orbit

When a cable plough cuts into the seabed, it ejects a column of disturbed sediment into the water column. In water shallower than roughly 20–25 m, enough of that sediment reaches or stays near the surface to alter the spectral reflectance of the water body in ways a multispectral sensor can measure. Suspended particulate matter increases reflectance across the visible and near-infrared, with the red and red-edge channels responding most strongly relative to clear water. Sentinel-2 bands B4 (665 nm), B5 (705 nm) and B6 (740 nm) are the primary diagnostic channels. Indices derived from these bands, including the Normalised Difference Turbidity Index and empirical suspended particulate matter retrieval algorithms published in the ocean-colour literature, can separate plume water from background with reasonable confidence when the plume is wide enough to fill at least a few 10 m pixels.

The physics sets a hard floor. In water deeper than about 25 m the seabed contribution to top-of-atmosphere reflectance becomes negligible, and a plume that does not reach the surface is invisible to any passive optical sensor. Most transoceanic cable routes cross water hundreds to thousands of metres deep. For those routes, satellite optical detection is simply not applicable, and any vendor claiming otherwise is selling something other than physics.

SAR adds a complementary, cloud-independent view

Sentinel-1 C-band SAR measures microwave backscatter from the sea surface. A rising column of fine sediment and entrained gas bubbles can alter surface roughness at scales detectable by SAR, particularly in calm sea states. Published studies of dredging plumes in shallow estuaries and coastal construction zones have shown backscatter anomalies of 2–4 dB above background in IW mode imagery. That is a detectable but modest signal, easily confused with wind streaks, ship wakes or rain cells. Reliable SAR-based plume attribution therefore requires concurrent wind and wave data from numerical weather models to screen out false positives.

SAR's real advantage here is availability. Sentinel-1 acquires regardless of cloud cover and at any hour, which matters in tropical cable corridors where persistent convective cloud routinely blocks optical sensors for days at a time. Used together, Sentinel-1 and Sentinel-2 form a complementary pair: SAR flags potential disturbance events, optical imagery confirms and quantifies them when skies clear.

Fusing AIS with imagery to name the responsible vessel

Detecting a turbidity plume is only half the analytical task. Attributing it to a specific cable-lay operation requires knowing which vessel was where and when. Automatic Identification System data, broadcast by vessels under IMO regulations, provides position, speed and heading at intervals of a few seconds to a few minutes for vessels underway. Cable-lay ships operating at plough speed, typically 0.3–1.5 knots, have a distinctive AIS signature that differs clearly from transiting or anchored vessels.

Spatial and temporal co-registration of AIS tracks with satellite imagery acquisition timestamps allows an analyst to test whether a detected plume is geometrically consistent with the reported track of a named vessel. If the plume centroid trails the vessel position by a distance consistent with the current speed and direction, attribution is defensible. If no AIS-reporting vessel is present, that itself is informative: some operators disable or spoof AIS, and a plume without a matching track warrants further investigation. The fusion is probabilistic, not conclusive, and should be documented as such in any regulatory submission.

Honest limits: where this method stops working

The depth constraint is absolute. Deep-water cable routes, which account for the majority of international fibre capacity by length, generate no detectable surface signature from ploughing or burial operations. Satellite remote sensing cannot help there.

Even in shallow water, tidal flushing can disperse a plume within hours, and a five-day Sentinel-2 revisit may simply miss the event. Commercial tasking of Planet SuperDove narrows that gap considerably, but same-day cloud-free acquisitions are not guaranteed. Turbid background conditions in naturally sediment-laden estuaries or near river mouths also raise the detection threshold significantly: a plume must exceed background turbidity by a margin detectable above sensor noise and atmospheric correction error, which in practice means the method is most reliable in clear-water coastal environments such as coral-reef margins or oligotrophic shelf seas.

SAR plume detection in sea states above Beaufort 4 becomes unreliable because wind-driven surface roughness swamps the sediment signal. Cable-lay operations in exposed offshore areas during normal working weather frequently exceed that threshold.

Practical applications: environmental compliance and route verification

The primary commercial use case is environmental compliance monitoring. Cable-lay permits in territorial waters often specify sediment disturbance corridors, seasonal exclusion zones near sensitive habitats, and post-lay reinstatement requirements. Independent satellite monitoring provides a permit authority or a project lender with a contemporaneous, third-party record of where disturbance occurred and whether it remained within the consented corridor. That record has value precisely because it is generated independently of the operator's own survey logs.

A secondary application is route verification for project finance. Lenders funding subsea cable infrastructure need confidence that the physical asset has been installed along the consented route. Combining AIS track archives with imagery-derived plume mapping provides a spatial audit trail covering the shallow-water portions of the route. Satellize structures this kind of monitoring as a time-series delivery covering the active lay period, with flagged events reported within 48 hours of a cloud-free acquisition. For context on how we apply similar time-series analytics to agricultural monitoring, the Kingdom of Tonga crop-estimation programme is the one public example we can point to.

Post-lay, repeat Sentinel-2 acquisitions over the first one to two seasons can track whether turbidity has returned to background levels, providing evidence for permit closure. That archive is free and goes back to 2015 for Sentinel-2A.

What the data cannot replace

Satellite monitoring of seabed disturbance is a screening and audit tool, not a substitute for in-situ water-quality sensors or acoustic seabed surveys. It cannot measure sediment concentration at depth, characterise grain size, or detect plumes below the photic zone. It also cannot verify burial depth or confirm that the cable has been laid rather than merely that a vessel passed. Those questions require ROV surveys, sub-bottom profilers and physical sampling.

Used within its limits, however, satellite monitoring offers something those methods cannot: continuous, spatially comprehensive coverage of an entire route corridor throughout the lay campaign, at a cost per kilometre that makes systematic monitoring of long shallow-water sections economically practical for the first time.

Typical figures

Optical spatial resolution	10 m (Sentinel-2 visible), 20 m (red-edge/SWIR), 3 m (Planet SuperDove)
SAR spatial resolution	~10 m range × ~14 m azimuth (Sentinel-1 IW mode)
Revisit interval	5 days (Sentinel-2, both satellites); 6 days per satellite (Sentinel-1); up to daily (Planet SuperDove, commercial)
Maximum detectable depth	~20–25 m water depth for seabed-origin plumes in optical bands; surface expression only for SAR
Spectral bands used	Sentinel-2 B4 (665 nm), B5 (705 nm), B6 (740 nm), B11 (1610 nm), B12 (2190 nm); Sentinel-1 C-band (5.4 GHz)
Minimum detectable plume width	Approximately 30–50 m (optical, requiring 3–5 contiguous pixels above background); coarser for SAR
Latency after acquisition	Sentinel data available within ~3 hours of overpass via Copernicus Data Space; analysis delivery typically 24–48 hours
Archive depth	Sentinel-2: from 2015; Sentinel-1: from 2014; MODIS: from 1999; Planet: from 2016 (variable coverage)
AIS data fusion	Historical and near-real-time AIS from commercial providers; position accuracy typically ±10–50 m
Delivery formats	GeoTIFF turbidity rasters, vector plume polygons, AIS-attributed event reports, PDF compliance summaries

Analytics Satellize can run

Turbidity plume mapping	Normalised Difference Turbidity Index and empirical suspended particulate matter retrieval from Sentinel-2 red-edge and SWIR bands	GeoTIFF raster and vector polygon per acquisition, showing plume extent and relative turbidity class
SAR surface roughness anomaly detection	Change detection on Sentinel-1 IW backscatter time series, wind-corrected using ERA5 reanalysis fields	Flagged anomaly polygons with confidence score, delivered as GIS layer within 48 hours of acquisition
Vessel-to-plume attribution	Spatial and temporal co-registration of AIS track archives with imagery acquisition timestamps; geometric consistency test	Event attribution report naming vessel, position, speed and plume offset distance at time of imagery
Consented corridor compliance check	Overlay of plume polygons against permit boundary shapefile; area and centroid distance calculations	Per-event compliance flag with exceedance area in hectares; tabular summary for regulatory submission
Post-lay turbidity recovery monitoring	Time-series analysis of Sentinel-2 SPM retrievals at fixed transects across the cable corridor	Monthly trend chart and GIS layer showing return to background turbidity levels; supports permit closure evidence
Wide-area screening with MODIS	Ocean colour anomaly detection in MODIS 250 m bands over the full route corridor	Daily alert layer flagging large-scale turbidity events for follow-up tasking of higher-resolution sensors

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.