Learning Spatio-Temporal Vessel Behavior using AIS Trajectory Data and Markovian Models in the Gulf of St. Lawrence

Abstract: Maritime Mobility is at the center of the global economy, and analyzing and understanding such data at scale is critical for ocean conservation and governance. Accordingly, this work introduces a spatio-temporal analytical framework based on discrete-time Markov chains to analyze vessel movement patterns in the Gulf of St. Lawrence, emphasizing changes induced during the COVID-19 pandemic. We discretize the ocean space into hexagonal cells and construct mobility signatures for individual vessel types using the frequency of cell transitions and the dwell time within each cell. These features are used to build origin-destination matrices and spatial transition probability models that characterize vessel dynamics at different temporal resolutions. Under multiple vessel types, we contribute with a temporal evolution analysis of mobility patterns during pandemic times, highlighting significant but transient changes to recurring transportation behaviors. Our findings indicate vessel-specific mobility signatures consistent across spatially disjoint regions, suggesting that those are latent behavioral invariants. Besides, we observe significant temporal deviations among passenger and fishing vessels during the pandemic, indicating a strong influence of social isolation policies and operational limitations imposed on non-essential maritime activity in this region.