Traffic analysis of Yangtze river delta multi-port system (china) using hierarchical clustering

Abstract

This paper focuses on the container throughput of 15 ports in the Yangtze River Delta (YRD) multi-port system from 1992 to 2019 to explore the evolution of container traffic share using concentration indexes and hierarchical clustering. Compositional Data (CoDa) analysis techniques are used to investigate the similarity and linkage of data. Then hierarchical clustering is used to cluster the traffic data of the YRD multi-port system from 1992 to 2019 into five periods based on the similarity among the yearly data. The consistency of the analysis is verified by combining the concentration index with the ternary diagram. The periods obtained from hierarchical clustering exhibit different patterns in the composition of traffic shares and they are consistent with traffic fluctuations. The combination of these methods provides a good interpretation of the spatial and temporal evolution of the YRD multi-port system.