Submarine base pier activity and sortie-rate inference
Optical and SAR imagery of submarine berths reveals vessel presence, pier-side equipment states and sortie patterns. Submerged boats are invisible to every open-source sensor; only surface and pier-side signatures are observable.
Sensors
- ICEYE X-band SAR: Spotlight mode delivers approximately 0.5 m ground resolution, enabling individual hull detection and double-bounce signature extraction at the waterline. Revisit to a named location can be tasked within hours on the commercial constellation. Operates through cloud and at night.
- Capella Space X-band SAR: Spotlight products at 0.5 m resolution with sub-metre geolocation accuracy. Particularly useful for detecting the metallic hull-to-pier double-bounce return that distinguishes a berthed submarine from an empty berth. Rapid tasking turnaround supports time-sensitive collection.
- Maxar WorldView-2: Panchromatic resolution of 0.46 m and eight-band multispectral imagery. Resolves hull colour, deck fittings, crane configuration and tender positioning in good visibility. Cloud cover is a hard constraint; the sensor provides no night capability.
- Planet SkySat: 0.5 m panchromatic resolution with flexible tasking and a revisit cadence of several times daily over priority sites. Useful for building a temporal baseline of pier occupation states, though image quality is somewhat below WorldView-2 in low-contrast conditions.
- ESA Sentinel-1 C-band SAR: Free-to-access, 5 x 20 m IW mode resolution is too coarse to resolve individual submarine hulls reliably, but useful for detecting large tender vessels and monitoring pier-area change over time. Six-day repeat at mid-latitudes provides a consistent historical archive back to 2014.
What a floating roof gives away: the physics of SAR double-bounce
A submarine hull sitting at the waterline creates a near-perfect dihedral reflector. The SAR pulse bounces off the flat water surface, then off the curved metal hull, and returns to the sensor with a characteristic bright double-bounce signature. In X-band imagery at 0.5 m resolution, this return is visually distinct from the pier structure, the water background and any moored surface vessel. It is the single most reliable open-source indicator that a submarine is present at a berth.
The physics also define the method's hard ceiling. Once a submarine submerges, the dihedral vanishes. No passive optical sensor and no SAR system operating at open-source frequencies can detect a submerged submarine. This is not a gap that better resolution will close; it is a physical constraint. The entire analytical value of this method rests on what can be inferred from surface and pier-side states, not from direct observation of the boat after it leaves.
Reading the pier: cranes, tenders and the logistics of departure
Pier-side activity is often more informative than hull presence alone. Crane arm position and extension indicate loading or maintenance operations. Fuel and stores barges moored alongside suggest pre-departure provisioning. Tender vessels, which support crew transfer and some replenishment functions, shift position predictably in the hours before and after a sortie. High-resolution optical imagery, when cloud-free, resolves deck equipment, open hatches and personnel movement at a level that SAR cannot match.
Change detection between successive passes is the core analytical operation. An analyst compares berth occupation state, crane configuration and tender positioning across an image time series. A berth that was occupied in the previous pass and is now empty, combined with a tender that has moved to the outer anchorage, is a stronger sortie indicator than either observation alone. Conversely, a crane extended over an open hatch with a barge alongside suggests the boat is undergoing maintenance and is unlikely to depart imminently.
The limit here is temporal resolution. Commercial tasking can achieve revisit intervals of a few hours over a named site, but gaps remain. A submarine can sortie and return within a window between passes without appearing absent in the archive. Sortie-rate estimates derived from imagery are therefore minimum bounds, not complete counts.
Building a sortie-rate baseline: what the time series actually shows
A meaningful sortie-rate estimate requires a sustained collection programme, not a single image. Over weeks or months, a time series of berth-occupation states allows an analyst to calculate the fraction of observation windows in which each berth was empty, the typical duration of absences, and whether patterns cluster around particular days or seasons. These statistics describe operational tempo in a way that a single snapshot cannot.
Published open-source analysis of facilities such as the Russian Pacific Fleet base at Vilyuchinsk and the Chinese PLAN base at Yulin has demonstrated that this method is viable at unclassified resolution. Analysts working from commercial imagery have tracked submarine movements and inferred deployment patterns at both sites. The method is not novel; what changes is the availability of sub-metre SAR with rapid tasking, which reduces the gap between collection and delivery from days to hours.
Baseline disruption is itself an analytical product. A facility that historically shows low sortie rates and then shifts to sustained high pier-vacancy rates is signalling a change in operational posture. That change is detectable even without knowing where the submarines have gone.
Honest limits: what this method cannot tell you
The method has four irreducible constraints that any honest assessment must state plainly. First, submerged submarines are unobservable. Second, cloud cover degrades optical collection; at high-latitude bases, this can mean weeks of optical blackout. SAR fills part of that gap but cannot replicate optical detail. Third, the inference from pier-side indicators to operational intent involves judgement, not measurement. A crane extended over a hatch might mean loading a torpedo or replacing a pump. Fourth, the archive depth of sub-metre commercial SAR is limited; ICEYE and Capella have operated at full capability for only a few years, constraining long-term baseline construction.
Deception is also possible. A facility aware of satellite collection schedules can manage pier activity to present a misleading picture. Pass timing is not secret; orbital mechanics are published. Analysts should treat imagery-derived sortie rates as indicators, not ground truth, and weight them alongside other open-source signals.
Integrating SAR and optical into a working collection plan
No single sensor is sufficient. A practical collection architecture combines Sentinel-1 as a persistent, free-to-access change-detection layer with commercial SAR tasking triggered when Sentinel-1 detects a state change at a monitored berth. High-resolution optical tasking is then added opportunistically when cloud cover permits, to resolve detail that SAR cannot provide. This layered approach balances cost against coverage.
Satellize builds collection plans of this type for government clients, integrating open constellations with commercial tasking on client licence. The analytical workflow is similar in structure to the crop-area estimation work the company runs for the Kingdom of Tonga, where the challenge is also one of combining sensor types with different revisit and resolution characteristics to produce a reliable time series from imperfect inputs.
For a submarine base monitoring programme, the practical output is a structured berth-state log updated on each collection pass, with automated alerts on state changes and a monthly sortie-rate summary report. The collection plan should specify which berths are primary targets, what the minimum acceptable revisit interval is for each, and how cloud-gap periods will be handled by SAR substitution.
Typical figures
| Best available spatial resolution (SAR) | 0.5 m (ICEYE and Capella Space spotlight mode) |
| Best available spatial resolution (optical) | 0.46 m panchromatic (Maxar WorldView-2) |
| Minimum detectable target | Submarine hull at berth detectable at 0.5 m SAR resolution; individual deck fittings and crane configuration require 0.5 m optical in good light |
| Tasked revisit (commercial SAR) | Sub-daily to a named site with coordinated ICEYE or Capella tasking; exact interval depends on constellation geometry and competing task load |
| Free-tier revisit (Sentinel-1) | 6-day repeat at mid-latitudes; 12-day at some high-latitude geometries; 5 x 20 m IW resolution |
| SAR frequency | X-band (9.6 GHz, ICEYE and Capella); C-band (5.4 GHz, Sentinel-1) |
| Cloud and darkness limitation | SAR is unaffected; optical collection (WorldView-2, SkySat) is blocked by cloud and requires daylight |
| Archive depth | Sentinel-1 from 2014; Maxar archive from early 2000s at lower resolution; ICEYE and Capella sub-metre SAR from approximately 2019 onwards |
| Delivery latency (commercial tasking) | Typically 2 to 6 hours from collection to analyst-ready product for priority tasking; varies by provider and processing pipeline |
| Delivery formats | GeoTIFF imagery, GIS-compatible berth-state vector layers, structured CSV time series, PDF monthly summary reports |
Analytics Satellize can run
| Berth occupation state log | Automated SAR double-bounce detection combined with optical change detection; binary occupied/vacant classification per berth per pass | Structured GIS layer updated on each collection pass, with timestamp and sensor provenance per record |
| Sortie-rate estimate | Time-series analysis of berth vacancy frequency and duration over a rolling window; minimum-bound calculation accounting for collection gaps | Monthly summary report with per-berth vacancy statistics and trend comparison against baseline period |
| Pier-side activity indicator | Object detection on high-resolution optical imagery for crane arm position, barge presence, tender vessel location and personnel density | Annotated image report with activity classification (maintenance, provisioning, departure preparation, idle) per collection pass |
| Operational posture change alert | Statistical threshold detection on berth-vacancy rate deviation from rolling baseline; triggered when vacancy rate exceeds defined sigma threshold | Near-real-time alert (email or API push) with supporting imagery thumbnail and context note |
| Tender and support vessel tracking | Vessel detection and position logging from SAR and optical; cross-referenced against AIS where available to identify vessels that have disabled transponders | GIS point layer of tender positions per pass; flagged events where tender moves to outer anchorage coincide with berth vacancy |
| Long-term baseline construction | Retrospective processing of Sentinel-1 archive from 2014 and available commercial optical archive to establish multi-year berth-occupation history | Historical berth-state database with seasonal and annual sortie-rate statistics; delivered as CSV and GIS layer with methodology note |
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.