Seabed movement detection above subsea pipelines and cables
Satellite InSAR and photon-counting lidar can detect millimetre-scale onshore settlement and nearshore bathymetric change above subsea corridors between survey campaigns, but both methods have hard depth and turbidity limits that operators must understand before trusting the data.
Sensors
- Sentinel-1 C-band SAR (ESA): Repeat-pass InSAR at 6-day revisit (two-satellite constellation) delivers line-of-sight displacement maps at 5 x 20 m ground range resolution in IW mode. Coherence is maintained over stable ground; onshore tie-in zones and nearshore intertidal flats are routinely measurable. Open ocean and water-saturated beach faces decorrelate completely.
- ICESat-2 ATL03/ATL07 photon-counting lidar (NASA): 532 nm green laser penetrates clear water to roughly 10-15 m depth under low-turbidity conditions. Along-track bathymetric profiles at ~0.7 m horizontal spacing resolve sub-metre bed elevation change when differenced across repeat passes (91-day exact repeat, with sub-cycle crossovers). Turbid estuaries and harbours block the signal entirely.
- TanDEM-X bistatic SAR (DLR): Single-pass interferometry eliminates temporal decorrelation, producing DEMs at 12 m posting with vertical accuracy of ~1 m (0.2 m in high-coherence areas). Useful for establishing a high-quality baseline DEM of intertidal and supratidal terrain at pipeline landfall; not a repeat-monitoring tool by default.
- Planet SuperDove multispectral: Eight spectral bands at 3-4 m resolution with near-daily revisit. Used here as a turbidity proxy: coastal aerosol and blue-band reflectance ratios indicate suspended sediment load, which predicts whether ICESat-2 passes will achieve useful bathymetric penetration on a given date.
What the seabed is actually doing between inspections
Subsea pipelines and cables fail for a small number of reasons: corrosion, anchor strikes, fishing gear, and geotechnical movement of the seabed itself. That last category, scour, slope creep, free-spanning caused by sediment redistribution, is both the hardest to inspect and the most amenable to satellite monitoring in shallow water. A remotely operated vehicle survey costs time and money and produces a snapshot. Satellite data produces a time-series.
The physics that matter are straightforward. Scour deepens the bed around a pipeline, creating free spans that increase bending stress. Slope creep on a continental shelf edge can displace a pipe laterally over months. Onshore tie-in structures settle differentially as coastal sediments compact. Each of these processes leaves a measurable signal somewhere in the satellite record, if you know which sensor to interrogate and what its honest detection floor is.
What a floating roof gives away: InSAR on the onshore side
Sentinel-1 InSAR is most useful at the landfall and the onshore tie-in, where the ground is coherent enough to sustain phase measurements. Over stable, dry ground, the technique resolves vertical displacement to roughly 3-5 mm per epoch in favourable conditions, degrading quickly over vegetated or waterlogged surfaces. A pipeline landfall structure sitting on reclaimed coastal fill is a classic InSAR target: slow differential settlement of 10-20 mm per year is well within detection range and has been documented at coastal industrial sites using Sentinel-1 data.
The limit is the shoreline itself. Intertidal zones decorrelate because the surface changes state between acquisitions. Anything seaward of the mean low-water mark is invisible to C-band InSAR. TanDEM-X can produce a precise single-epoch DEM of the intertidal flat, giving a baseline against which future topographic surveys can be differenced, but TanDEM-X is not a free repeat-monitoring service. Operators need to plan and commission repeat acquisitions explicitly.
Green laser in clear water: what ICESat-2 can and cannot resolve
ICESat-2 carries the Advanced Topographic Laser Altimeter System, which fires 532 nm photons at 10 kHz and records individual photon returns. In water with a diffuse attenuation coefficient below roughly 0.1 m⁻¹ (typical of clear tropical or temperate coastal waters in calm conditions), the instrument resolves the seabed to depths of 10-15 m. Published studies using ATL03 data have demonstrated bathymetric retrieval with vertical uncertainties of 0.25-0.5 m in clear conditions, which is sufficient to detect significant scour events or sediment accumulation above a pipeline corridor.
The 91-day exact repeat cycle is the primary temporal constraint. Sub-cycle crossover passes exist but do not cover every corridor. For a specific pipeline route, the analyst must check whether ICESat-2 ground tracks actually intersect the corridor of interest; coverage is not global at useful density. And turbidity is a hard blocker. A single storm resuspending fine sediment can raise attenuation coefficients by an order of magnitude, rendering that pass useless for bathymetry. Planet SuperDove imagery acquired within a day or two of an ICESat-2 pass provides a practical turbidity screen: if the coastal aerosol band shows high reflectance, the lidar return is likely to be a water-column signal, not a seabed signal.
Where the methods stop working
Depth beyond 15 m is beyond satellite lidar reach under any realistic water clarity. The North Sea, the Gulf of Mexico deepwater, and most mid-ocean cable routes are simply inaccessible to current spaceborne optical systems. For those corridors, satellite data contributes indirectly: regional seabed mobility indicators from SAR-derived surface roughness, storm intensity and wave height from altimetry, or sediment plume tracking as a proxy for active transport. These are useful contextual inputs, not direct seabed measurements.
Turbid coastal environments, river deltas, estuaries, and harbours present the same problem at shallow depth. The Ganges-Brahmaputra delta, the Yangtze estuary, and similar high-sediment coasts have attenuation coefficients that block green-wavelength lidar even in calm conditions. InSAR coherence in these areas is also poor because of vegetation, tidal inundation, and rapid surface change. Satellite methods provide limited direct value for pipelines in these settings; they are better used to characterise the sediment transport regime that drives risk, rather than to observe the seabed directly.
A final ambiguity: bathymetric differencing detects elevation change but not its cause. A 0.4 m deepening of the bed above a pipeline could be scour, could be a data artefact from differing sea-state conditions between passes, or could be a navigation offset in the ICESat-2 ground track. Validation against at least one in-situ survey point per corridor is not optional; it is the step that converts a satellite observation into an actionable finding.
Building a monitoring cadence that matches the risk
The practical workflow combines sensors by zone. Onshore tie-in: Sentinel-1 InSAR time-series, processed at every available acquisition, producing a displacement velocity map updated monthly. Intertidal and nearshore to 15 m depth: ICESat-2 bathymetric differencing at each repeat cycle, quality-gated by same-week Planet turbidity imagery. Beyond 15 m: storm and current climatology from altimetry and reanalysis, used to flag periods of elevated seabed mobility for prioritising ROV inspection.
Satellize runs this multi-sensor pipeline on open constellations and adds commercial tasking where clients hold the relevant licences. The analytic architecture is the same one used in the Tonga crop-estimation programme: open-data ingestion, sensor-specific pre-processing, and change-detection outputs delivered as GIS layers with uncertainty bounds attached. The difference here is that the consequence of a missed signal is a pipeline failure rather than a harvest estimate, which raises the bar for validation considerably.
The honest summary: satellite methods extend the temporal resolution of seabed monitoring from annual ROV surveys to near-continuous observation in the zones where they work. They do not replace in-situ inspection. They tell you where to look and when to look harder.
Typical figures
| Sentinel-1 IW mode ground range resolution | 5 x 20 m (range x azimuth) |
| Sentinel-1 InSAR minimum detectable displacement (stable ground) | 3-5 mm per epoch under favourable coherence; degrades rapidly over wet or vegetated surfaces |
| Sentinel-1 revisit (two-satellite constellation) | 6 days at mid-latitudes; 12 days with one satellite |
| ICESat-2 bathymetric depth limit (clear water) | 10-15 m; dependent on diffuse attenuation coefficient below ~0.1 m⁻¹ |
| ICESat-2 vertical uncertainty (clear water, seabed returns) | 0.25-0.5 m (published range from ATL03 validation studies) |
| ICESat-2 exact repeat cycle | 91 days; sub-cycle crossovers available but track-dependent |
| TanDEM-X DEM vertical accuracy (high coherence) | ~0.2 m relative; ~1 m absolute at 12 m posting |
| Planet SuperDove revisit (turbidity screening) | Near-daily at 3-4 m resolution; 8 spectral bands including coastal aerosol |
| Sentinel-1 archive depth | From April 2014 (Sentinel-1A launch); globally available via Copernicus Data Space |
| ICESat-2 archive depth | From October 2018; ATL03 and ATL07 products via NASA Earthdata |
Analytics Satellize can run
| Onshore tie-in settlement map | Sentinel-1 persistent scatterer or small-baseline InSAR time-series (SBAS or PS-InSAR) | GeoTIFF velocity map (mm/year) with uncertainty layer; monthly update; flagged anomaly report when displacement exceeds operator-defined threshold |
| Nearshore bathymetric change raster | ICESat-2 ATL03 photon-cloud seabed extraction, differenced between repeat passes; quality-gated by concurrent turbidity index | Along-corridor elevation-change profile (CSV and GeoJSON) per 91-day cycle, with per-point confidence flag based on photon return density |
| Turbidity suitability forecast for ICESat-2 passes | Planet SuperDove coastal aerosol and blue-band ratio mapped to Kd(490) proxy; compared against ICESat-2 pass schedule | Pass-by-pass suitability score delivered 48 hours before acquisition; retrospective quality flag appended to bathymetric product |
| Intertidal baseline DEM | TanDEM-X bistatic SAR single-pass interferometry; phase-to-height conversion with tidal correction | One-time GeoTIFF DEM at 12 m posting, with vertical accuracy metadata, for use as change-detection reference |
| Scour anomaly alert | Statistical change detection on bathymetric time-series; threshold set against operator-supplied free-span tolerance | Email or API alert with location, magnitude estimate, and ICESat-2 pass ID; linked to GIS layer showing affected corridor segment |
| Sediment transport regime characterisation | Multi-temporal Planet imagery time-series for suspended sediment plume extent; storm climatology from satellite altimetry (Copernicus Marine Service reanalysis) | Annual report: dominant transport directions, storm recurrence statistics, high-mobility season identification; used to prioritise ROV inspection scheduling |
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.