Vessel traffic conflict mapping around offshore wind development zones
Integrating satellite AIS with Sentinel-1 SAR reveals the full picture of vessel traffic inside proposed offshore wind lease areas, including dark vessels that never appear in transponder records. The analysis directly supports maritime spatial planning and cable-route risk assessment.
Sensors
- Sentinel-1 SAR (C-band, ESA): Synthetic aperture radar at 5.405 GHz. In Interferometric Wide Swath mode: 250 km swath, 5 × 20 m ground resolution. Detects vessels as bright radar returns regardless of cloud or darkness. Minimum detectable vessel length is roughly 20–30 m under typical sea states; smaller craft are below the noise floor. Revisit over European waters is typically 1–3 days with both satellites combined.
- Satellite AIS (multiple commercial providers, e.g. Spire Global): VHF transponder signals received by low-Earth-orbit satellites. Provides vessel identity, position, speed, heading, and vessel type for AIS-equipped ships. Latency from observation to delivery can be under 30 minutes for near-real-time feeds. Coverage is global, but signal collision in congested ports and busy sea lanes degrades position accuracy and completeness; reception rates can fall below 60 % in the busiest corridors.
- Sentinel-2 MSI (ESA): Multispectral optical imager at 10 m resolution (visible and near-infrared bands), 290 km swath, 5-day revisit at the equator and faster at higher latitudes. Useful for mapping the visible footprint of installed turbine arrays, identifying construction staging areas, and detecting vessel wakes in clear-water conditions. Cloud cover is the primary constraint; a single overpass over an overcast North Sea site may yield no usable data.
- Copernicus Marine Service (CMEMS) ocean products: Provides modelled and satellite-derived sea-state, surface current, and significant wave-height fields at spatial resolutions of 1/12° to 1/32°. Used to contextualise vessel routing decisions and to flag periods when sea state would suppress small-vessel activity, avoiding false conclusions about traffic absence.
Why AIS alone gives an incomplete picture of who is in the lease area
Automatic Identification System transponders are mandatory on vessels over 300 gross tonnes on international voyages, and on fishing vessels over 15 m in many jurisdictions. That leaves a large population of smaller fishing craft, recreational vessels, and, occasionally, vessels deliberately operating dark, entirely outside the record. In a typical offshore wind lease area in the southern North Sea or Celtic Sea, the economically significant conflict is often with vessels that fall just below the mandatory threshold.
Even among vessels that do carry AIS, satellite reception is imperfect. In congested waters, multiple transmissions arrive at the satellite simultaneously and corrupt each other, a phenomenon called signal collision. Published analyses of AIS reception in areas like the English Channel suggest that position-report loss rates can exceed 30–40 % during peak traffic periods. A conflict analysis built solely on AIS will therefore systematically undercount traffic density in the zones that matter most: those already busy enough to create planning problems.
What Sentinel-1 SAR adds, and what it cannot see
SAR illuminates the sea surface with microwave pulses and records the backscattered energy. A vessel's metal hull produces a bright return, detectable regardless of cloud, rain, or night conditions. Sentinel-1's Interferometric Wide Swath mode covers 250 km in a single pass at 5 × 20 m resolution, which is sufficient to detect most commercial fishing vessels, service vessels, and cargo ships operating in a lease area.
The honest limit is size. Vessels shorter than roughly 20–30 m produce returns that can be indistinguishable from wave clutter, particularly in sea states above Beaufort 4–5. Small open fishing boats, jet skis, and rigid inflatables are effectively invisible. SAR also captures a single snapshot per overpass, not a track. Fusing SAR detections with AIS records, by matching position, heading, and speed, allows analysts to identify which SAR targets have no corresponding AIS signal. Those are the dark vessels. The residual after matching is an imperfect count, not a census, because some legitimate vessels will have had a temporary transponder outage, and some small craft will simply be below the detection threshold.
Revisit cadence matters here. Over UK and northern European waters, Sentinel-1A and 1B together typically achieve a 1–3 day repeat. That is adequate for building a statistical picture of traffic density over weeks or months, but it will miss vessels that transit quickly through a lease area between passes.
Building the conflict map: from detections to spatial planning evidence
The core analytic product is a density grid. AIS position reports and SAR vessel detections are binned into a regular spatial grid, typically 500 m or 1 km cells, across the lease area and a buffer zone around it. Counts are normalised by observation time and satellite coverage to produce vessel-hours per cell per month. Separate layers are produced for vessel type where AIS data permits: fishing, cargo, tanker, service vessel.
Cable-route risk assessment uses a corridor version of the same approach. A candidate export cable corridor is intersected with the density grid, and the resulting traffic intensity profile identifies the highest-conflict crossings. Anchor-dragging risk is a function of both traffic density and the proportion of vessels that slow or stop within the corridor, which can be extracted from AIS speed histories.
Seasonal stratification is important. Fishing effort in many North Sea and Atlantic lease areas peaks in spring and autumn. A conflict analysis based on a single season's data will misrepresent the annual risk profile. A minimum of 12 months of fused AIS and SAR data, with monthly breakdowns, gives planners defensible evidence for an Environmental Impact Assessment or a Marine Licence application.
Sentinel-2 and ocean context: the supporting layers
Sentinel-2 optical imagery contributes two things. First, it provides a precise, independently verified footprint of any turbines already installed in an adjacent operational farm, which matters when a proposed lease area sits next to an existing one and the planning question is about cumulative spatial pressure on fishing grounds. Second, vessel wakes are sometimes visible in high-contrast optical imagery over calm, clear water, providing a qualitative cross-check on traffic patterns in areas where SAR passes are sparse.
Sea-state data from CMEMS is used as an analytical covariate rather than a primary input. Days with significant wave height above roughly 2.5 m tend to suppress small-vessel activity. Including sea-state as a filter prevents the analysis from attributing reduced traffic to the presence of the wind farm when the real cause was weather. This distinction matters in post-construction monitoring, where developers must demonstrate whether their installation has displaced fishing effort or merely coincided with a rough season.
Honest limits planners should state in their submissions
Any conflict map produced by this method carries three irreducible uncertainties that should be stated explicitly in planning submissions. First, vessels below roughly 20–30 m are undercounted in SAR, and their share of total traffic varies strongly by fishery type and region. Second, AIS signal collision means that congested areas are systematically undercounted in the transponder record; the degree of undercounting is not directly measurable from the data alone, though published correction factors exist in the academic literature. Third, SAR revisit of 1–3 days means that short-duration traffic events, a single night's gill-netting fleet, for example, may fall between passes entirely.
These limits do not make the analysis unusable. They mean the output is a lower-bound estimate of traffic intensity, which is the conservative assumption most appropriate for conflict assessment. A developer who presents this analysis alongside its stated limits is in a stronger evidential position than one who presents uncaveated AIS statistics alone.
From analysis to submission: what a delivered product looks like
Satellize runs this class of analysis on open-access Sentinel-1 and Sentinel-2 archives combined with licensed satellite AIS feeds. The output is a GIS-ready set of density layers, a conflict-intensity report with seasonal breakdowns, and a cable-corridor risk profile, formatted for submission to a maritime spatial planning authority or inclusion in an EIA appendix. The analytical approach follows published methods for SAR vessel detection and AIS-SAR fusion that are well-established in the remote-sensing literature.
Typical figures
| SAR spatial resolution (Sentinel-1 IW mode) | 5 × 20 m (range × azimuth) |
| SAR revisit (Sentinel-1A + 1B, northern Europe) | 1–3 days typical |
| Minimum detectable vessel length (SAR) | Approximately 20–30 m; smaller vessels below noise floor in moderate sea states |
| Optical resolution (Sentinel-2 MSI, visible/NIR) | 10 m |
| Optical revisit (Sentinel-2, mid-latitudes) | 5 days; cloud cover may extend effective revisit to weeks in winter |
| AIS latency (satellite receiver, near-real-time feed) | Under 30 minutes for near-real-time; historical archives available |
| AIS completeness in congested corridors | Position-report loss rates of 30–40 % or more reported in high-density areas |
| Density grid cell size (typical output) | 500 m or 1 km, configurable to lease-area geometry |
| Archive depth (Sentinel-1) | From April 2014 (Sentinel-1A launch); Sentinel-1B data available 2016–2021 |
| Delivery format | GeoTIFF density grids, GeoPackage or Shapefile vessel tracks, PDF conflict report |
Analytics Satellize can run
| Vessel traffic density grid | Spatial binning of AIS position reports and SAR vessel detections, normalised by observation time; separate layers by vessel type | GeoTIFF raster layers (monthly and annual), GIS-ready, with metadata on data completeness |
| Dark-vessel identification | SAR vessel detection via constant false-alarm rate (CFAR) processing followed by AIS-SAR positional matching; unmatched SAR targets flagged as dark | Point shapefile of dark-vessel detections per SAR overpass, with confidence scores |
| Cable-corridor conflict profile | Intersection of candidate cable corridor polygons with vessel density grid; speed-profile analysis from AIS to identify anchoring and slow-steaming behaviour | Corridor risk report with traffic intensity transect, anchor-risk scoring, and recommended crossing angles |
| Seasonal traffic stratification | Monthly aggregation of fused AIS and SAR data over a minimum 12-month baseline; sea-state filtering using CMEMS significant wave-height fields | Seasonal density maps and summary table for EIA appendix |
| Wind-farm footprint delineation | Sentinel-2 MSI object-based image analysis to map installed turbine positions and array extent; cross-checked against SAR bright-point targets | Polygon shapefile of confirmed turbine positions and exclusion-zone boundaries |
| Post-construction displacement monitoring | Comparison of pre- and post-construction traffic density grids within lease area and buffer; sea-state-normalised to isolate displacement from weather effects | Before-and-after density comparison report, suitable for regulatory compliance submission |
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.