Forward air-defence battery emplacement and redeployment tracking
Very-high-resolution optical and SAR imagery reveals the revetment geometry, antenna shadows and vehicle parks that characterise surface-to-air missile batteries. Sequential scene comparison tracks redeployment, though decoy emplacements of similar geometry create a persistent false-positive problem.
Sensors
- Maxar WorldView-3: 30 cm panchromatic resolution (commercially available product at 30–50 cm). Eight SWIR bands allow surface-material discrimination between compacted soil berms and natural ground. Revisit approximately 1–4.5 days depending on latitude and tasking priority.
- Airbus Pléiades Neo: 30 cm native resolution, four-satellite constellation achieving sub-daily revisit over priority areas. Stereo and tri-stereo modes produce digital surface models useful for measuring berm height, a discriminator between operational revetments and shallow decoys.
- ICEYE SAR constellation: Stripmap mode at approximately 3 m resolution; spotlight mode at approximately 1 m. SAR penetrates cloud and operates at night, critical for monitoring sites in continental climates with persistent overcast. Coherence change detection between repeat passes flags ground disturbance within days of emplacement.
- Planet SkySat: 50 cm resolution, with the broader Planet Dove fleet providing daily 3–5 m context imagery. SkySat tasking confirms the presence or absence of vehicles and equipment at a battery site identified in a prior high-resolution scene.
What the geometry gives away
Surface-to-air missile batteries are not random in their layout. A typical S-300 or S-400 battery occupies a characteristic footprint: a central command-and-control vehicle, launcher transporter-erector-launchers arranged in an arc or star pattern around it, a search radar at a standoff distance of roughly 100–200 m, and a track-and-engagement radar closer in. Each element sits within or beside an earthwork revetment, a berm of compacted soil designed to reduce blast and radar cross-section. The berms themselves are the most durable signature. They survive long after the vehicles depart.
At WorldView-3 or Pléiades Neo resolution, individual berm walls 3–5 m high cast measurable shadows. The shadow length, combined with known solar elevation at acquisition time, gives an independent height estimate. Radar antenna shadows in SAR imagery are equally diagnostic: a large phased-array or dish antenna produces a characteristic double signature, a bright return from the structure itself and a dark shadow behind it. Open-source analysts applied exactly this logic to publicly available commercial imagery of S-400 sites in Syria, Turkey and Ukraine between 2018 and 2023, identifying battery positions and, in several cases, tracking lateral moves of a few kilometres within 48-hour windows.
SAR coherence and the problem of fresh earth
Optical imagery is weather-dependent. A single overcast day can break a monitoring sequence at exactly the wrong moment. SAR fills that gap, and it adds something optical cannot provide: coherence change detection.
When two SAR acquisitions over the same area are compared interferometrically, pixels that have not changed remain coherent. Ground disturbance, whether from vehicle movement, excavation or berm construction, decorrelates the signal. ICEYE spotlight-mode passes at approximately 1 m resolution can detect the decorrelation signature of heavy vehicle tracks and freshly turned earth within a day or two of activity, even under complete cloud cover. The practical limit is that coherence change is sensitive to almost any surface disturbance, including agriculture and civilian construction, so context from optical and pattern-of-life data is necessary to avoid false alerts.
Tracking redeployment: the scene-differencing method
Redeployment detection rests on a straightforward principle: compare a reference scene with a follow-on acquisition, flag pixels where the signature of a known battery element has disappeared and search the broader area of interest for a matching new signature. In practice this is harder than it sounds. Vehicles move between passes. Shadows shift with season. Vegetation changes. A battery that has dispersed its launchers into tree cover may leave only the radar and command post visible.
The published open-source record, particularly work from the Middlebury Institute, Bellingcat and the James Martin Center for Nonproliferation Studies, demonstrates that redeployment of S-300-class systems over distances of tens of kilometres can be detected within one to three days using commercial tasking, provided the destination area is already under periodic observation. If it is not, the battery can vanish into a large search area. This is the core operational limit of the method: it is a detection tool, not a tracking tool. Once a battery moves out of a monitored zone, finding it again requires either tip-off from other intelligence or a broad-area search that may take days.
Archive depth matters here. WorldView imagery archives extend back to 2007 for some areas. Comparing a current scene against a multi-year baseline distinguishes a newly prepared hide site from a long-standing civilian feature of similar geometry.
Decoys: the persistent false-positive problem
Military forces have used decoy air-defence emplacements since at least the Second World War. Modern decoys are sophisticated. Inflatable radar replicas produce convincing SAR returns. Earthwork revetments can be constructed to match the footprint of a real battery without any operational equipment present. Russia, in particular, has documented use of decoy S-300 and S-400 positions in exercises and, according to open reporting, in operational deployments.
No single spectral or geometric indicator definitively separates a live battery from a well-made decoy. The most reliable discriminators are thermal emission from active electronics and engine heat, vehicle presence and movement over time, and cable or power-supply infrastructure. Thermal infrared imagery at the resolution needed to detect radar cooling systems requires either airborne collection or, at satellite altitude, sensors such as the Airbus Pléiades Neo thermal variant or future commercial thermal missions. Current commercial thermal satellites generally resolve to 3–5 m at best, which is marginal for this application. Pattern-of-life analysis over weeks, noting whether vehicles arrive and depart on a plausible operational schedule, is the most practical counter-decoy method available from open commercial imagery.
What this analysis can and cannot deliver
A well-designed monitoring programme can reliably detect the construction of new revetment complexes consistent with air-defence battery layout, flag the disappearance of known battery signatures within the revisit cadence of the tasked constellation, and identify candidate new positions in a defined search area. At WorldView-3 or Pléiades Neo resolution, individual launcher positions within a battery are distinguishable. SAR coherence monitoring can shorten detection latency to under 48 hours for sites with regular pass coverage.
The honest limits: cloud cover over continental Europe and parts of Asia can interrupt optical coverage for five to ten consecutive days in winter. SAR coherence is non-specific. Decoys of high quality cannot be reliably excluded from optical or SAR evidence alone. And commercial tasking over denied or contested territory carries legal and contractual constraints that vary by jurisdiction and operator. Satellize structures its analytics to be explicit about confidence levels, distinguishing a confirmed battery signature from a probable one and flagging cases where decoy probability is elevated.
For clients building a national air-picture or supporting treaty-verification work, the output of this analysis feeds directly into a GIS layer updated on each new acquisition, with a structured report noting geometry, shadow measurements, vehicle-count estimates and change flags.
Typical figures
| Optical spatial resolution (commercial) | 30–50 cm (WorldView-3, Pléiades Neo); 50 cm (SkySat) |
| SAR spatial resolution | ~1 m spotlight (ICEYE); ~3 m stripmap (ICEYE) |
| Revisit cadence (tasked optical) | Sub-daily to 4.5 days depending on constellation and latitude |
| SAR coherence change detection latency | 24–72 hours from acquisition to processed alert, subject to pass schedule |
| Spectral bands (optical) | Panchromatic, RGB, NIR, SWIR (WorldView-3 eight-band SWIR) |
| SAR frequency | X-band (ICEYE, ~9.6 GHz); C-band (Sentinel-1, ~5.4 GHz) for context |
| Minimum detectable revetment feature | Berm walls ≥3 m wide detectable at 30 cm optical; antenna shadows ≥2 m at 1 m SAR |
| Archive depth | WorldView series from 2007; Sentinel-1 from 2014; Pléiades from 2011 |
| Thermal resolution (current commercial) | 3–5 m (marginal for electronics heat detection; honest limit acknowledged) |
| Delivery formats | GeoTIFF change layers, GeoJSON feature annotations, structured PDF assessment report |
Analytics Satellize can run
| Battery-site geometry catalogue | Template matching and shadow-length photogrammetry on VHR optical imagery | GIS layer of confirmed and probable battery positions with berm dimensions and shadow-derived height estimates |
| Redeployment alert | Sequential scene differencing: pixel-level change detection between registered VHR optical acquisitions | Alert report flagging site vacated, with timestamp and last-known vehicle count |
| SAR coherence change flag | Interferometric coherence comparison between ICEYE or Sentinel-1 repeat passes | Raster coherence-loss map overlaid on site boundary, updated per pass |
| New-position candidate search | Broad-area template matching for revetment geometry within defined search radius following redeployment alert | Ranked list of candidate sites with confidence scores and supporting imagery chips |
| Decoy-probability assessment | Pattern-of-life vehicle-presence analysis over 30-day window; thermal anomaly check where data available | Structured assessment report with confidence band: confirmed active, probable active, indeterminate, probable decoy |
| Baseline and trend report | Multi-year archive comparison using available WorldView, Pléiades and Sentinel-1 scenes | Timeline chart of site activity with annotated imagery for each state change |
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.