Offshore wind site screening using multi-parameter satellite data

Combining SAR wind fields, radar-altimeter wave climatology, sea-ice extent, and AIS vessel density lets developers shortlist offshore lease areas before committing to costly met-ocean campaigns. Satellite data accelerates the funnel; it does not replace in-situ measurement for bankable energy yield.

Sensors

Sentinel-1 SAR-C (ESA): C-band SAR at 5.405 GHz. IW mode delivers 10 m spatial resolution across a 250 km swath with 6-day repeat at the equator (3-day with both satellites). Backscatter inversion via CMOD5.n or CMOD7 retrieves 10 m wind speed and direction fields; documented retrieval uncertainty is roughly ±1.5 m/s against buoy reference in open-ocean conditions.
CryoSat-2 radar altimeter (ESA): Ku-band SAR/SARIn altimeter. Provides significant wave height (SWH) and sea-surface height along a ground track with ~350 m along-track resolution in SAR mode. Repeat cycle is 369 days (sub-cycle ~30 days), so climatological SWH products require multi-year stacking; single-pass data is not useful for operational wave monitoring.
Sentinel-6 Michael Freilich altimeter (ESA/EUMETSAT/NASA): High-resolution mode Poseidon-4 altimeter on a 10-day exact-repeat orbit. Delivers SWH with a noise floor around 0.25 m and sea-surface height accuracy better than 2 cm (RMS). The 10-day repeat makes it practical for seasonal wave climatology when combined with the 25-year TOPEX/Jason heritage record.
AIS vessel-tracking feeds (multiple aggregators): Automatic Identification System messages decoded from spaceborne receivers (e.g. Spire Global, Orbcomm). Positional accuracy is typically 10 m. Coverage gaps exist at high traffic density due to VHF signal collision; terrestrial AIS fills coastal gaps. Vessel-density rasters are computed by binning message positions into 0.01-degree grid cells over multi-year windows.
NSIDC / EUMETSAT passive microwave sea-ice products: AMSR2 and SSMIS passive microwave radiometers produce daily sea-ice concentration maps at 3.125–12.5 km resolution depending on channel. Ice-extent climatologies derived from the 1979-present SSM/I–SSMIS record define seasonal operability windows for high-latitude sites. Ice edges carry a positional uncertainty of roughly 10–20 km in marginal ice zones.

Why screening from orbit makes financial sense

A met-ocean buoy deployment costs hundreds of thousands of dollars per year and must run for at least twelve months before a developer can characterise seasonal variability with any confidence. Satellite data costs a fraction of that and, for open-ocean areas, already spans decades. The strategic question is not whether to use satellite data but which satellite parameters are genuinely decision-relevant at the pre-feasibility stage.

The answer is four. Wind-speed climatology determines whether the energy resource justifies further investigation. Significant wave height informs preliminary foundation-load envelopes. Sea-ice extent defines the operational calendar at high latitudes. And vessel-density maps identify where shipping conflicts will trigger regulatory objection before a single turbine is proposed. Each of these can be estimated from existing satellite archives without commissioning new data collection.

Wind resource: what SAR backscatter actually tells you

Sentinel-1 IW mode imagery, processed through the CMOD5.n geophysical model function, converts C-band backscatter to 10 m wind speed at 10 m above the sea surface. At 10 m spatial resolution and 250 km swath width, a single pass captures enough spatial structure to identify wind-speed gradients driven by coastal jets, island wakes, and atmospheric boundary-layer rolls. These features are invisible to coarse reanalysis products such as ERA5, which resolves roughly 31 km.

The honest limits matter here. SAR wind retrieval degrades in rain (rain cells attenuate and scatter the signal simultaneously), at very low wind speeds below about 2 m/s where the sea surface is glassy and backscatter loses sensitivity, and at speeds above roughly 25 m/s where the relationship between roughness and wind speed becomes ambiguous. Retrieval also requires an independent wind-direction input, typically from a numerical weather model, because SAR alone cannot unambiguously separate speed from direction. For pre-feasibility screening these limitations are acceptable. For bankable P50/P90 yield estimates, they are not.

What a floating roof gives away: altimeter-derived wave climatology

Significant wave height from Sentinel-6 and CryoSat-2 feeds directly into preliminary foundation design. Monopile fatigue life is sensitive to the long-term distribution of wave loads, not just the extreme event. A 20-year SWH climatology built from the Jason series (TOPEX through Jason-3) and extended with Sentinel-6 gives developers a credible picture of the P99 wave environment before a single wave-rider buoy is deployed.

The caveat is geometry. Altimeters measure along a narrow ground track. In areas with complex bathymetry or strong wave-refraction patterns, interpolating between tracks introduces uncertainty that grows with distance from the nearest pass. Significant wave height from altimetry also integrates over a footprint of several kilometres, so localised swell shadowing behind a headland or shoal will not appear in the record. Foundation engineers will still require in-situ wave spectra for detailed load calculations. Satellite climatology narrows the design envelope; it does not close it.

Bathymetric depth, a critical input for foundation-type selection (monopile, jacket, or floating), is not reliably retrievable from optical or SAR satellite data in open ocean. Satellite-derived bathymetry from ICESat-2 photon-counting lidar works in clear, shallow coastal water to roughly 40 m depth. Beyond that, developers must rely on ship-borne multibeam surveys or existing hydrographic charts.

Sea ice: defining the construction and maintenance calendar

For sites above roughly 60°N or in enclosed seas such as the Baltic or Bohai, sea-ice extent is a direct constraint on vessel access and turbine operability. Passive microwave products from AMSR2 provide daily ice-concentration maps at 3.125 km resolution (89 GHz channel), and the SSM/I record extends back to 1979, giving a 45-year climatology of ice-edge position and seasonal duration.

Ice concentration maps are not ice-thickness maps. A 30% concentration field could mean thin, navigable frazil ice or a mix of open water and thick floes. For construction-season planning, developers need ice-thickness data, which currently comes from CryoSat-2 freeboard retrievals (uncertainty roughly ±0.1 m in first-year ice) or from in-situ drilling. Satellite ice-extent data is sufficient to flag whether a site has a seasonal ice problem at all. Quantifying that problem precisely requires additional data.

Shipping conflict: AIS density as a pre-screening filter

Vessel-density rasters derived from multi-year AIS archives identify traffic separation schemes, ferry routes, fishing grounds, and naval exercise areas that will create regulatory friction or physical exclusion requirements. A grid cell showing more than, say, 500 vessel transits per year in a proposed lease area is a material planning risk, not an insurmountable one, but one that needs to be quantified early.

AIS has a structural weakness: ships that switch off their transponders do not appear in the record. Fishing vessels in some regions are notorious for this. SAR imagery can detect vessel targets independently of AIS, and the gap between SAR-detected vessels and AIS-reporting vessels is itself informative, though interpreting it requires care. Satellize runs this kind of multi-layer conflict analysis as part of its data-analytics work; the Tonga crop-estimation programme is a different domain, but the methodological pattern of fusing passive registries with active sensor detections is the same.

The output of an AIS conflict screen is a density raster and a ranked list of sub-areas by traffic intensity, delivered as a GIS layer. It is not a legal exclusion determination. Maritime authorities in each jurisdiction hold the definitive view on what constitutes a prohibited zone.

What satellite data cannot do, and when to stop relying on it

The hard limit is bankability. Project finance for an offshore wind farm requires a P50 annual energy production estimate with a defined uncertainty band, typically derived from at least one year of on-site anemometry at hub height, validated against a long-term reference dataset. No satellite product currently meets that standard. SAR wind fields at 10 m height must be extrapolated to hub heights of 100–150 m using a wind-shear model, and the uncertainty in that extrapolation alone can exceed the uncertainty in the SAR retrieval itself.

Satellite screening is best understood as a funnel-narrowing tool. It can eliminate areas with demonstrably poor wind resource, unacceptable wave environments, year-round ice cover, or irresolvable shipping conflicts. The areas that survive that filter are the ones worth spending money on met-ocean buoys, LiDAR profilers, and detailed geophysical surveys. That is the honest value proposition: satellite data reduces the number of expensive in-situ campaigns a developer needs to run, not the need for them entirely.

Typical figures

SAR wind-field spatial resolution	10 m (Sentinel-1 IW mode); 100–500 m typical after wind-retrieval smoothing
SAR revisit (dual-satellite)	~3 days at mid-latitudes; 1–2 days above 60°N
Significant wave height noise floor	~0.25 m (Sentinel-6 Poseidon-4 HR mode)
Altimeter ground-track repeat	10 days (Sentinel-6); ~30-day sub-cycle (CryoSat-2)
Sea-ice concentration resolution	3.125 km (AMSR2 89 GHz); 25 km (SSM/I standard product)
AIS vessel-density grid resolution	Typically 0.01° (~1 km); positional accuracy of individual messages ~10 m
Satellite-derived bathymetry depth limit	~40 m in clear, shallow coastal water (ICESat-2 photon-counting lidar)
Archive depth (wind climatology)	Sentinel-1: 2014–present; SAR wind heritage (ERS, Envisat): 1991–2012
Archive depth (wave climatology)	TOPEX/Jason/Sentinel-6 altimeter series: 1992–present (~33 years)
Delivery formats	GeoTIFF, NetCDF, GeoPackage/Shapefile; API tile feeds on request

Analytics Satellize can run

10 m wind-speed climatology raster	CMOD5.n or CMOD7 backscatter inversion applied to Sentinel-1 IW archive; seasonal and annual percentile stacking	GeoTIFF stack of mean, P10, P50, P90 wind speed per season; accompanying uncertainty layer
Significant wave height climatology	Along-track SWH from Sentinel-6 and CryoSat-2 gridded by kriging interpolation; validated against ECMWF ERA5 wave reanalysis	NetCDF climatology with seasonal SWH percentile fields and extreme-value (Gumbel) fit for P99 design wave
Sea-ice operability calendar	AMSR2 and SSM/I passive microwave ice-concentration time series; threshold at 15% concentration for ice-edge definition per NSIDC convention	Monthly ice-free probability raster and tabulated open-water season statistics per candidate site polygon
Vessel-density conflict map	Multi-year AIS message binning to 0.01° grid; traffic separation scheme overlay from official ENCs; SAR vessel detection cross-check for dark-target identification	GeoPackage with density raster, ranked conflict-score polygons, and dark-vessel anomaly points
Multi-parameter site-score ranking	Weighted composite of wind resource, SWH environment, ice-free days, and vessel-conflict score; weights configurable per client criteria	Ranked shortlist table and GIS polygon layer with per-parameter breakdown for each candidate area
SAR-vs-AIS vessel gap analysis	Co-registration of Sentinel-1 CFAR ship detections with contemporaneous AIS positions; unmatched detections flagged as non-reporting targets	Per-scene matched/unmatched vessel report; monthly dark-vessel frequency map as GeoTIFF

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.