Narco-trafficking clandestine airstrip mapping
Unpermitted landing strips used for drug transshipment leave a geometric and spectral signature that satellite imagery can resolve. Sentinel-2, WorldView-3 and Sentinel-1 SAR together cover the cloud-prone Andean and Central American corridors where most clandestine strips operate.
Sensors
- Sentinel-2 MSI: 10 m resolution in visible and near-infrared bands; 5-day revisit at the equator with both satellites. Detects strips longer than roughly 300 m through geometric linearity and the spectral contrast between compacted bare earth and surrounding canopy. Free and open; archive from 2015.
- Maxar WorldView-3: 30 cm panchromatic, 1.24 m multispectral. Resolves individual tyre tracks, fuel drums and aircraft wheel ruts. Tasked on demand; typical revisit 1–4 days depending on latitude and cloud. The primary tool for confirmation and court-quality documentation.
- Sentinel-1 SAR (C-band): 20 m resolution in Interferometric Wide Swath mode; 6-day repeat. Cloud and darkness are irrelevant. Detects the backscatter discontinuity where forest is cleared and soil compacted. Particularly useful in the persistently overcast wet-season months across the Colombian and Guatemalan lowlands.
- Planet SkySat: 50 cm resolution, tasked daily. Faster cadence than WorldView-3 for monitoring known strips for renewed activity, fuel deliveries or aircraft presence. Multispectral coverage supports vegetation-recovery assessment to estimate how recently a strip was last used.
Why a landing strip cannot hide in plain sight
A clandestine airstrip imposes a specific set of physical requirements on its operators. The strip must be long enough to land a loaded light aircraft, typically 400 to 800 metres for the Cessna 206 and similar types documented in UNODC field assessments. It must be flat, compacted and clear of vegetation taller than about half a metre. Those requirements produce a shape that is, from above, unmistakable: a narrow rectangle of low or absent vegetation punched through otherwise continuous forest or scrub.
That rectangle has a spectral identity too. Compacted bare earth reflects strongly in the red and shortwave-infrared bands relative to the dark, moisture-rich canopy surrounding it. The normalised difference vegetation index (NDVI) of a cleared strip drops sharply compared with adjacent forest, a contrast detectable at Sentinel-2's 10 m resolution if the strip exceeds roughly 300 m in length. Shorter strips require VHR tasking. Neither the shape nor the spectral signature can be disguised without defeating the operational purpose of the strip.
The aeronautical negative space
Every licensed airstrip appears in a national aeronautical information publication (AIP) and, in most jurisdictions, in ICAO regional databases. A strip that appears in satellite imagery but carries no AIP registration is, by definition, anomalous. That cross-reference is the first filter in any systematic survey: identify all linear clearings above a minimum length threshold, then subtract every registered facility. What remains is a candidate list for further investigation.
Published UNODC monitoring work in the Andean region has used exactly this approach, combining optical change detection with aeronautical registry checks to produce annual inventories of suspected clandestine infrastructure. The method is not perfect. Legitimate private strips in some countries are poorly registered, and agricultural airstrips used for crop-spraying can produce false positives. Disambiguation requires checking for associated infrastructure: fuel storage, access tracks, perimeter clearance and the absence of agricultural equipment or irrigation hardware.
Cloud cover, SAR and the wet-season problem
The Andean foothills and the Pacific lowlands of Central America are among the cloudiest environments on Earth during the wet season. Optical sensors can be blocked for weeks at a time. Sentinel-1 SAR operates through cloud and at night, imaging the same geometry in C-band backscatter rather than reflected light. A freshly cleared strip shows as a low-backscatter rectangle against the higher, rougher return of surrounding forest. The contrast is cleaner when the soil is moist, which is common in exactly the environments where cloud is worst.
SAR does carry its own ambiguities. Smooth water bodies, sand bars and some agricultural fields can produce similar low-backscatter signatures. Slope-induced geometric distortion in mountainous terrain is a real problem: strips cut into hillsides can appear compressed or stretched depending on their orientation relative to the satellite's look angle. These limits mean SAR is best used as a screening and alerting layer, with optical confirmation at the earliest cloud-free opportunity.
What VHR imagery adds, and what it cannot settle
WorldView-3 at 30 cm panchromatic resolution can resolve tyre tracks, individual fuel drums, aircraft wheel ruts and the scorching patterns left by engine exhaust on compacted earth. That level of detail supports activity assessment: is this strip recently used, or abandoned? Are there vehicles present? Is there a perimeter guard post? These are questions that 10 m Sentinel-2 imagery cannot answer.
What VHR imagery cannot settle on its own is attribution. Identifying a strip as clandestine is an analytical conclusion; identifying who operates it requires corroboration from signals intelligence, human sources or law-enforcement interdiction. Satellite analysis defines the where and the when, and can track the pattern of activity over time. It does not replace the investigative chain. Buyers who expect satellite data alone to produce prosecutable evidence will be disappointed; buyers who use it to direct scarce aviation and ground assets to the right locations will find the economics compelling.
Running a systematic survey: method and honest limits
A practical survey workflow begins with a change-detection pass over the target corridor using Sentinel-2 time series. Automated line-detection algorithms flag rectangular clearings above a length threshold, typically 250 m, against a baseline image stack. The candidate list is then cross-referenced against aeronautical registries and filtered by associated-infrastructure signatures. High-priority candidates are escalated to tasked VHR imagery for confirmation. Sentinel-1 provides continuity during cloud-blocked periods.
Detection probability drops significantly for strips shorter than 250 m, for strips covered by camouflage netting or planted with low crops, and for strips that are used infrequently enough that vegetation recovery masks the clearing between revisits. Planet SkySat's daily cadence helps with the last problem: a strip used once a month may look green on a monthly Sentinel-2 composite but show bare earth on a daily SkySat pass taken the morning after a flight. Archive depth matters too. Sentinel-2 data runs back to 2015, enabling retrospective analysis of when a strip first appeared and how its activity has evolved.
Satellize can run this full detection-and-monitoring pipeline as a standing analytical service, with outputs delivered as GIS layers and periodic intelligence summaries. The methodology is the same class of spectral and geometric analysis applied in the Kingdom of Tonga crop-estimation programme, adapted to a very different target signature.
Prioritising the corridor: where the signal concentrates
Not every clandestine strip is equally significant. A strip used once and abandoned is a historical artefact. A strip showing repeated activity signatures, fresh tyre tracks on successive VHR passes, fuel-drum deliveries, vehicle movements at unusual hours, is operationally active infrastructure. Prioritisation should focus enforcement resources on the latter.
Published DEA and UNODC casework identifies the Colombian Vichada and Guainía departments, the Peruvian VRAEM valley, and the Guatemalan Petén as areas of historically high clandestine strip density. These are not the only areas; the corridor shifts as enforcement pressure moves. A standing monitoring service that flags new clearings within weeks of their appearance is more useful than a periodic survey that documents what was happening six months ago.
Typical figures
| Minimum detectable strip length (Sentinel-2) | ~300 m under good contrast conditions; shorter strips require VHR |
| Spatial resolution (screening layer) | 10 m (Sentinel-2 MSI visible/NIR bands) |
| Spatial resolution (confirmation layer) | 30 cm panchromatic / 1.24 m multispectral (WorldView-3) |
| SAR resolution (cloud-penetrating layer) | 20 m (Sentinel-1 IW mode) |
| Revisit cadence (screening) | 5 days (Sentinel-2 dual-satellite); 6 days (Sentinel-1) |
| Revisit cadence (confirmation) | 1–4 days tasked (WorldView-3); daily tasked (Planet SkySat) |
| Spectral bands used | Visible (RGB), NIR, SWIR for NDVI and bare-soil indices; C-band SAR backscatter |
| Archive depth | Sentinel-2 from 2015; WorldView archive from 2014 (commercial licence); Sentinel-1 from 2014 |
| Latency (alert to delivery) | 24–72 hours from satellite pass to analyst-reviewed output, depending on cloud and tasking queue |
| Delivery formats | GeoTIFF, GeoJSON, KML, PDF intelligence summary, GIS-ready vector layers |
Analytics Satellize can run
| New-strip detection alert | Automated change detection on Sentinel-2 NDVI and bare-soil index time series; geometric line-detection filter applied to candidate pixels | Georeferenced alert with coordinates, detection date, estimated strip length and cloud-cover flag; delivered as GeoJSON or email notification within 72 hours of a clear-sky pass |
| Aeronautical registry cross-reference | Spatial join of detected clearings against national AIP and ICAO regional database footprints; unmatched candidates flagged as anomalous | Filtered candidate list with registration status, nearest registered facility and distance; exported as attributed GIS layer |
| Activity recency assessment | Multi-date VHR (WorldView-3 or SkySat) comparison of tyre-track freshness, vegetation encroachment rate and infrastructure presence using published vegetation-recovery rate models for tropical soils | Per-strip activity rating (active/recently active/dormant/abandoned) with supporting imagery chips and analyst commentary; PDF or structured JSON |
| Wet-season SAR continuity monitoring | Sentinel-1 C-band backscatter change detection; low-backscatter rectangular anomalies flagged against forest baseline; slope-correction applied using SRTM DEM | Monthly SAR-derived suspect-strip layer covering defined area of interest; GeoTIFF and summary statistics |
| Historical timeline reconstruction | Retrospective Sentinel-2 and available commercial archive analysis to establish first-appearance date, periods of use and any abandonment or reactivation events | Strip-level activity timeline chart and supporting imagery stack; suitable for intelligence briefings or legal proceedings |
| Corridor-level density and trend report | Aggregated strip inventory across a defined geographic corridor; year-on-year change in strip count, spatial clustering analysis and correlation with known enforcement events | Quarterly or annual strategic intelligence report with maps, trend charts and methodology annex; PDF and GIS package |
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.