Knowledge extraction from open-access maritime positioning data

Introduction: Monitoring how vessels move and concentrate is essential for maritime safety, environmental protection, and global trade analysis. Open-access data from the vessel Automatic Identification System (AIS) provide a massive repository for Earth science research, but they are often fragmented by gaps, signal interference, and inconsistent quality that can obscure the true state of maritime traffic. This contribution presents how suitable data-processing techniques can overcome these hurdles to produce reliable operational and economic insights.
Methods: We developed a series of workflows to turn raw AIS messages into structured observations, tailored to the fluctuating quality and coverage of public data streams. Our approach includes (i) cleaning and reconstruction steps and (ii) a journey model that converts individual messages into metrics such as total vessel counts, traffic flow, and gridded spatial maps. The flexibility of these methods is demonstrated through two case examples, one focusing on broad regional patterns and another on high-resolution port activity.
Results: By analyzing 10⁸ messages from a three-month period in 2024, we accurately mapped the density of moving and stationary vessels, calculated traffic rates and long-term trends, and identified individual berths with up to 30-meter precision in the Baltic Sea and Tokyo Bay (Japan). Additionally, we found that signal shadows caused by urban infrastructure can be used to pinpoint the exact locations of terrestrial receiver antennas.
Conclusions: Our work shows that open-access AIS data, when handled with the right processing tools, provide a level of detail that matches official statistics and expensive private datasets. These methods scale effectively from individual ports to entire sea regions, offering a low-cost way to study shipping trends, port utility, and the human footprint on the marine environment.