A method for analyzing land-ocean intermodal transportation capability of harbors based on traffic big data

Harbors are critical nodes linking maritime transport systems with inland logistics networks, yet globally comparable measures that jointly describe land-side travel conditions, marine movement, and surrounding transport infrastructure remain limited. This study proposes a land-ocean integrated and reproducible big-data framework for characterizing the proxy accessibility and transportation capability of container harbors. Google Maps, OpenStreetMap (OSM), Automatic Identification System (AIS) data, remote-sensing images, and publicly available port information were integrated to transform heterogeneous traffic observations into six comparable indicators, including travel complexity, average speed, and travel coverage area for travel conditions, as well as harbor capacity, transport networks, and traffic facilities for infrastructure conditions. The framework was applied to 30 major container harbors worldwide using 2016 container throughput as a reference outcome, with a supplementary analysis of 18 mainland Chinese harbors examining year-specific associations between fixed indicator values and annual throughput from 2017 to 2025. Results show that indicator–throughput relationships vary across indicators and regions. Transport networks exhibit the strongest statistically significant positive association with throughput, while average speed and harbor capacity also show significant positive associations. The proposed framework differs from conventional port assessments that mainly rely on throughput, infrastructure capacity, or single-mode accessibility by simultaneously capturing land-side travel efficiency, sea-side vessel movement characteristics, and surrounding transport conditions. These results demonstrate the potential of multi-source traffic big data for comparative analysis of harbor transportation capability.