Abstract
Maritime Domain Awareness increasingly depends on publicly available vessel tracking data to support commercial fisheries oversight, but it remains unclear how closely publicly available AIS-derived products approximate the restricted Vessel Monitoring System data that have long served as the regulatory standard. Here we evaluate whether two no-cost public AIS-derived platforms can substitute for restricted VMS data in the eastern Bering Sea, one of the world's largest commercial fishing regions. We compared monthly vessel presence hours, coverage area, and hot spot distributions across three datasets, namely Global Fishing Watch (GFW), the Global Maritime Traffic Density Service (GMTDS), and federally regulated VMS data, for the period 1 January 2021 through 31 December 2022 through statistical analyses of vessel presence, temporal trends, and hot spot detection. AIS-derived data mirror VMS in capturing broad-scale seasonal dynamics and peak activity, with high temporal and spatial coverage correlations. Differences emerge at fine scales, particularly in vessel classification, nearshore activity, and spatial hot spot identification. VMS-only hot spots clustered in waters deeper than 100 m near the U.S.-Russia Maritime Boundary Line, while AIS-only hot spots clustered nearshore in Bristol Bay and along the Alaska Peninsula at depths under 50 m. The annual footprint of AIS-only hot spots expanded from 0.4 percent in 2021 to 2.4 percent in 2022. Between 11.0 and 12.6 percent of VMS-tracked vessels operated entirely without detectable AIS, and an additional 13 to 15 percent operated intermittently, with non-detection peaking in summer and late autumn months. Publicly available AIS-derived datasets can reliably support broad-scale Maritime Domain Awareness and fisheries management, particularly in jurisdictions where VMS access is restricted. They are less reliable for fine-scale applications: deep-water federal fisheries near international boundaries are systematically under-represented in AIS, while nearshore state-managed fleets are largely absent from VMS. We recommend integrated VMS and AIS-derived monitoring for high-resolution applications, with attention to inter-annual processing changes in public AIS pipelines that can mimic real changes in fishing activity.