Vessel Exhaust Plume Monitoring for Sulphur Compliance
Satellite SO2 and aerosol sensors can detect sulphur-rich exhaust plumes from shipping corridors, giving port-state authorities and charterers an independent check on IMO 2020 compliance. Detection is real but attribution to a single hull requires corroborating data.
Sensors
- Sentinel-5P TROPOMI: Measures SO2 vertical column density at 3.5 × 5.5 km ground pixel (improved from original 7 × 3.5 km after 2019 processor update). Daily global coverage. Sensitive to column enhancements above roughly 0.5 Dobson Units in clean marine air, which is sufficient to detect aggregated plumes from busy shipping lanes such as the English Channel or Malacca Strait but generally insufficient to isolate a single vessel.
- Planet SuperDove (PlanetScope): Eight-band multispectral imagery at approximately 3 m ground sample distance, with near-daily revisit over most latitudes. Can resolve individual vessel smoke trails as visible aerosol streaks under clear-sky conditions. No dedicated SO2 or gas-phase absorption band; detection relies on aerosol optical depth contrast in visible and near-infrared channels, not direct sulphur chemistry.
- Sentinel-2 MSI: 13-band multispectral at 10 m (visible/NIR) to 60 m (atmospheric bands). Five-day revisit at the equator with two satellites. Useful for confirming smoke plume geometry and correlating vessel position with trail origin. Like SuperDove, it lacks a dedicated SO2 absorption channel; its aerosol band (Band 1, 443 nm) gives a coarse proxy for aerosol loading.
- VIIRS EDR Aerosol: Suomi-NPP and NOAA-20 VIIRS produce daily global aerosol optical depth products at roughly 750 m resolution. Useful for tracking the spatial extent of shipping-corridor aerosol loading at regional scale. Temporal resolution is one or two overpasses per day, so transient plumes from individual vessels are frequently missed.
What the IMO 2020 cap actually requires, and why satellites care
Since January 2020, the International Maritime Organisation's MARPOL Annex VI regulation limits sulphur content in marine fuel to 0.5 percent by mass globally, and to 0.1 percent inside designated Emission Control Areas covering the North Sea, Baltic, North American coasts and the US Caribbean. The penalty for non-compliance falls on the shipowner and flag state, but enforcement historically depended on port-state inspectors sampling bunker fuel or reading engine logs. Both methods are retrospective, slow and geographically limited.
Satellites offer something different: an independent, continuous, geographically broad signal that does not depend on a vessel choosing to cooperate. Burning high-sulphur fuel produces SO2 and sulphate aerosol at concentrations measurably above background marine atmosphere. That contrast is the physical basis for remote detection. The question is not whether the signal exists; it does. The question is how precisely it can be resolved, and what it actually proves.
What TROPOMI can and cannot resolve
TROPOMI is the most capable operational SO2 sensor in orbit for this application. Its 3.5 km pixel width sounds fine-grained until you consider that a Panamax container ship is roughly 300 m long. A single vessel's plume, dispersed by even modest wind, typically occupies a fraction of one TROPOMI pixel within minutes of emission. What TROPOMI reliably detects is the cumulative SO2 enhancement over shipping lanes where dozens or hundreds of vessels transit daily. Studies published in journals including Remote Sensing have shown statistically significant SO2 column elevations over the English Channel, the Strait of Malacca and the Red Sea corridor, consistent with fleet-level sulphur loading.
That corridor-level signal is genuinely useful for regulators monitoring whether a shipping lane's aggregate emissions are trending down after a policy change. It is not, on its own, evidence against a named vessel. Wind dispersion, atmospheric mixing height and the geometry of the overpass all introduce ambiguity that grows quickly with time since emission. TROPOMI's daily revisit helps with trend analysis but rarely catches a plume within the narrow window before it disperses below detection threshold.
High-resolution optical imagery: seeing the smoke, not the chemistry
Planet SuperDove's 3 m imagery can, under clear-sky conditions, image a visible exhaust plume trailing from an individual vessel. This is not a niche capability: researchers have demonstrated plume detection from high-sulphur fuel combustion as a visible aerosol signature in multispectral imagery. The limitation is fundamental. Neither SuperDove nor Sentinel-2 MSI carries a band tuned to SO2 absorption wavelengths (around 300 to 320 nm, well into the ultraviolet). What the optical sensors see is particulate matter and water vapour in the exhaust, not the gas-phase sulphur compounds that define the regulatory violation.
A dark, dense smoke trail is consistent with high-sulphur fuel combustion but is not diagnostic of it. A vessel burning compliant fuel with a poorly maintained engine may produce a similar visual signature. Conversely, a vessel burning high-sulphur fuel at sea in humid conditions may produce a plume that blends into low cloud within seconds. Optical imagery is best used to confirm plume geometry and establish vessel position and heading at a known time, feeding a chain of evidence rather than standing alone.
Cloud cover is the operational constraint that neither sensor family can overcome. The North Sea and Baltic ECA zones, where enforcement interest is highest, carry significant cloud cover for much of the year. A clear-sky detection rate of even 30 to 40 percent on any given day is optimistic for those regions.
Building a corridor-to-hull attribution chain
The practical workflow for actionable compliance intelligence combines layers. TROPOMI provides the corridor-level flag: SO2 column density elevated above a statistically defined baseline for that lane and season. VIIRS aerosol products can confirm regional aerosol loading. A high-resolution optical pass, if timed and cloud-free, can resolve individual plumes and associate them with vessel positions. AIS records, cross-referenced against the optical geometry, narrow the candidate list to vessels whose reported position and heading are consistent with the observed plume origin and drift direction.
Even this chain leaves a gap. AIS can be manipulated or simply absent for non-cooperative vessels. Wind-field uncertainty compounds with time: a plume observed 40 minutes after emission may have drifted several kilometres from its source, and the error in back-trajectory modelling grows accordingly. What satellite evidence can deliver is a probabilistic shortlist, not a conviction. Port-state authorities in several jurisdictions have begun treating satellite-derived plume flags as grounds for elevated inspection priority, which is the appropriate evidentiary use of the data.
Honest limits and the cases where the method fails
Scrubber-equipped vessels present a specific ambiguity. Open-loop exhaust gas cleaning systems remove SO2 from stack emissions but discharge sulphate-rich washwater into the sea. The stack plume of a scrubber-equipped vessel burning high-sulphur fuel may appear compliant to both TROPOMI and optical sensors, even as the vessel discharges acidic washwater. Satellite sulphur monitoring cannot distinguish this case from genuine fuel compliance without additional data.
Slow steaming reduces fuel consumption and emission rate, which can push plume SO2 concentrations below TROPOMI's detection floor even for non-compliant vessels. Conversely, vessels accelerating through a waypoint may produce transient plumes that look alarming but reflect engine load rather than fuel quality. The method works best as a screening tool applied across many transits over time, where statistical patterns of excess SO2 in a corridor reveal systematic non-compliance rather than isolated events.
Satellize applies this multi-sensor workflow on open constellations, adding commercial optical tasking when a client needs targeted coverage of a specific route or flag-state fleet.
Typical figures
| SO2 spatial resolution (TROPOMI) | 3.5 × 5.5 km per pixel (post-2019 processor) |
| SO2 detection floor (TROPOMI, clean marine air) | Approximately 0.5 Dobson Units column enhancement; single-vessel plumes typically sub-threshold |
| TROPOMI revisit | Daily global coverage; one overpass per day at most latitudes |
| Optical plume resolution (SuperDove) | ~3 m GSD; individual smoke trails visible under clear sky |
| Optical revisit (SuperDove) | Near-daily at most latitudes; cloud-free acquisition not guaranteed |
| Sentinel-2 MSI resolution | 10 m (visible/NIR), 60 m (aerosol band 443 nm); 5-day revisit with two satellites |
| VIIRS aerosol product resolution | ~750 m; one to two overpasses daily |
| Effective attribution latency | 12 to 48 hours for corridor-level flag; same-day if cloud-free optical pass is available |
| Archive depth | TROPOMI from October 2017; Sentinel-2 from 2015; VIIRS from 2012 |
| Key spectral channels | UV 300–320 nm (SO2 absorption, TROPOMI); 443 nm aerosol proxy (Sentinel-2 Band 1); visible/NIR (plume geometry, SuperDove) |
Analytics Satellize can run
| Shipping-corridor SO2 anomaly map | TROPOMI SO2 column retrieval differenced against seasonal baseline; statistical threshold flagging | Weekly GIS layer showing corridor segments with statistically elevated SO2, delivered as GeoTIFF or GeoJSON |
| Individual vessel plume detection report | High-resolution optical change detection; aerosol optical depth contrast in visible/NIR bands correlated with AIS position | Per-event PDF report with annotated imagery, vessel candidate list and confidence rating |
| Back-trajectory plume attribution | Wind-field-driven Lagrangian back-trajectory modelling from observed plume centroid to candidate emission source | Probabilistic source polygon with time-stamped AIS vessel candidates ranked by positional consistency |
| Fleet-level compliance trend analysis | Multi-month TROPOMI time series aggregated by lane segment; trend decomposition controlling for traffic volume and meteorology | Quarterly trend report with lane-by-lane SO2 index, suitable for regulatory submission or charter due diligence |
| Targeted optical tasking alert | TROPOMI anomaly triggers commercial satellite tasking request for next available clear-sky pass over flagged corridor segment | Tasked imagery delivered within 24–72 hours of anomaly detection, with plume annotation overlay |
| Scrubber-washwater risk flag | Cross-reference of stack-plume SO2 absence with vessel equipment registry data and sea-surface turbidity proxies from optical bands | Vessel-level risk flag appended to compliance dossier; noted as indicative, not conclusive |
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.