CAP01: Radar at sea: Using land-based high-frequency radar systems to monitor risks of IUU fishing

With 1,535,538 km² of marine space, protecting South Africa from IUU (Illegal, Unreported and Unregulated) fishing is challenging for fisheries control authorities. Due to other countries running out of fish and starting to move south to catch more fish, South Africa has seen IUU Fishing on the rise, costing the economy around R6 billion each year.

Land-based high-frequency (HF) radar systems are at the forefront of surface oceanographic measurements with a spatial and temporal range and resolution that cannot be matched with traditional mooring or satellite approaches. The near-real time data feed can provide nowcasting opportunities and operational information on the surface currents, wave parameters, wind speed and direction. Additional features include Automated Identification System (AIS) coupling for ship detection and prevent IUU fishing. With the current configuration, we are able to supply wave and surface current data at 30-minute intervals enabling us to track the movement and speed of the Agulhas Current and associated eddies. With the multiple possibilities of this data, the challenge is to develop a Machine Learning application that uses high-frequency (HF) radar systems data to detect ships when they enter South Africa’s economic zone as well as display near real-time warnings to alert authorities when ships are detected through the ML algorithm, yet not visible on Global Fishing Watch.

A prototype of Machine Learning application that uses high-frequency (HF) radar systems data to detect ships at sea and hence risks of IUU fishing should the detected ship not be sharing its AIS signal.

Data analyst (ML, AI)
IT developer (full stack)
Oceanographer
Expert in maritime law