Revealing Delay Propagation Patterns and Interactions in the Chinese Air Traffic System Using Bayesian Network Analysis

Abstract

Abstract Flight delays in the air traffic system exhibit strong correlations among different airports, making it crucial to comprehend the dynamics of delay propagation for optimal structural design and prevention of cascading delays. In this study, we adopt a Bayesian network (BN) approach, assuming that flight delays in individual airports are influenced by various potential airports. Through the application of complex network analysis, we enhance our understanding of delay propagation characteristics, while employing a BN-based analytical framework to uncover spatial and temporal patterns of delay propagation. Using Chinese air traffic data from 2015 to 2017 as a case study, our findings reveal that each airport's flight delays have an impact on approximately eight external airports, indicating a significant degree of interconnectedness within the air traffic system. Larger airports tend to be affected by delays originating from upstream airports, whereas smaller airports typically influence downstream airports. Additionally, economically developed regions experience broader delay propagation, impacting a larger number of airports compared to less developed regions. These insights provide valuable guidance for the development of delay propagation damping strategies and the implementation of efficient delay management techniques. By understanding the patterns of delay propagation and their spatial-temporal dynamics, stakeholders can take proactive measures to mitigate the impact of delays and optimize overall system performance.