Abstract
Transnational cocaine trafficking, or ‘narco-trafficking’, networks often move large shipments of harmful and illegal drugs by sea through the use of unregulated boats in remote maritime spaces. This study presents a framework for identifying narco-trafficking drop-off zones by detecting and analyzing Unreported and Unregulated Boats (UUBs) potentially linked to narco-trafficking in Costa Rica’s Guanacaste and Puntarenas regions, utilizing a combination of high-resolution satellite imagery, machine learning, and spatial analysis. By training a YoloV5 convolutional neural network model, we detected boat wakes in PlanetScope satellite images, which were then cross-referenced with legal vessel traffic data (Automatic Identification System and Vessel Monitoring System) to isolate UUBs. A spatial autocorrelation analysis revealed a positive association between UUB locations and narco-trafficking activity indicators, such as drug-related media reports, court arrest records, and cocaine seizure data. High-high clusters of UUBs and trafficking indicators suggested that particular coastal districts may serve as primary landing zones for illicit shipments, a finding consistent with secondary data on cocaine trafficking in the region. By integrating geospatial intelligence with contextual data sources, this study advances methodology for identifying narco-trafficking drop-off zones and contributes a spatially explicit perspective to the broader understanding of cocaine and other illicit supply chains. Despite the limitations of cloud cover and restricted nighttime visibility, this framework offers a proof-of-concept approach for identifying UUB concentrations and cocaine drop-off shipment zones. Future work should consider expanding temporal coverage and multiple imagery to further enhance the identification of narco-trafficking zones.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call