Global Fishing Watch has released updates to the map. Anyone can now:
- Analyze and visualize activity of vessels at sea that may not appear in cooperative tracking systems.
- Global Fishing Watch has added an open global radar detections layer, derived from synthetic aperture radar or SAR, to display at-sea vessels, in any weather conditions.
- SAR can show vessel detections that can not be picked up by other cooperative tracking systems (e.g. AIS, automatic identification systems, or VMS, vessel monitoring systems).
- The addition of an experimental AIS matching filter allows anyone to interrogate the SAR vessel detections layer and determine which detections are likely matched to identifiable vessels and which detections may not be transmitting identification information in cooperative tracking systems.
- For vessels that don’t have an AIS match, Global Fishing Watch has introduced an experimental machine learning filter to determine whether the vessel is likely fishing, likely non-fishing or unknown, based on various factors including inferred vessel length, bathymetry, distance from port and environmental values.
- Understand where fixed infrastructure – like oil platforms or wind farms – are located and how vessels may interact with them.
- A fixed infrastructure reference layer can be added to any workspace, and filtered to show points of infrastructure associated with oil or wind activities.
- Using the SAR data, Global Fishing Watch use an algorithm to detect “anomalies” or bright pixels scene in satellite scenes, and then use a machine learning model to exclude vessels and determine the probability that it’s a known class of infrastructure: wind infrastructure, oil infrastructure, other infrastructure, and noise.
- Download vessel tracks for further analysis or research.
- Vessel tracks can be downloaded from the map for ongoing analysis and research activities.
Find our user guide, supporting information and tutorials on the Global Fishing Watch website. Share feedback through the built-in feedback form in the left sidebar on the map or add feature requests to our public Canny board. Email [email protected] for support.
Release – Thursday 4 January
