This study quantitatively analyzed the Closest Point of Approach (CPA) and distance from the channel centerline between vessels using AIS trajectory data for three representative channel environments: the Incheon Bridge-adjacent channel, the Incheon B...
This study quantitatively analyzed the Closest Point of Approach (CPA) and distance from the channel centerline between vessels using AIS trajectory data for three representative channel environments: the Incheon Bridge-adjacent channel, the Incheon Bridge open channel, and the Mokpo Port shallow-water channel.
The objective was to identify how structural constraints and environmental conditions influence the spatial behavior of vessels in maintaining safe distances and their overall traffic patterns within the channels. AIS-based CPA data were classified by vessel gross tonnage range and vessel type, and the lateral deviation from the channel centerline was calculated to examine the distribution characteristics of vessel trajectories. Furthermore, one-way ANOVA and Welch’s post-hoc tests were conducted to verify statistical differences among channels and vessel categories. The analysis results showed that as vessel tonnage increased, both CPA
and the distance from the channel centerline decreased, indicating that larger vessels tend to navigate more centrally and maintain stable trajectories. The average CPA followed the order bridge channel < shallow-water channel < open channel, demonstrating a consistent trend in which greater physical constraints lead to shorter inter-vessel distances. In particular, within 200 meters of the Incheon Bridge piers, CPA decreased by more than 10%, suggesting a localized influence of bridge structures on navigator behavior. A similar trend was observed in the analysis of centerline deviation. In restricted waterways such as the pier-adjacent and shallow channels, vessels exhibited centralized navigation patterns, while in the open channel, small vessels showed a more dispersed, peripheral navigation pattern. On the other hand, vessel type did not show a statistically significant correlation with CPA or distance from the centerline, indicating that differences in CPA are determined more by the degree of structural constraint of the channel than by vessel type. Overall, the results empirically demonstrate that vessel proximity and navigational distribution are governed more by the structural and environmental openness of the channel than by physical ship characteristics.
This study provides quantitative insights into vessel traffic behavior and offers a scientific basis for channel width design, bridge-area traffic management, and risk-based maritime traffic safety systems.