Data-driven modeling of systemic delay propagation under severe meteorological conditions

TL;DR

This paper presents a data-driven agent-based model that simulates delay propagation in the US air transportation network during severe weather, aiding managers in evaluating intervention strategies.

Contribution

It introduces a novel agent-based modeling approach using real flight data to accurately reproduce and analyze delay propagation under extreme meteorological conditions.

Findings

Model accurately reproduces delay dynamics during the 2010 Superstorm

Intervention measures improve prediction accuracy

Model can assist in planning to reduce delays during severe weather

Abstract

The upsetting consequences of weather conditions are well known to any person involved in air transportation. Still the quantification of how these disturbances affect delay propagation and the effectiveness of managers and pilots interventions to prevent possible large-scale system failures needs further attention. In this work, we employ an agent-based data-driven model developed using real flight performance registers for the entire US airport network and focus on the events occurring on October 27 2010 in the United States. A major storm complex that was later called the 2010 Superstorm took place that day. Our model correctly reproduces the evolution of the delay-spreading dynamics. By considering different intervention measures, we can even improve the model predictions getting closer to the real delay data. Our model can thus be of help to managers as a tool to assess different intervention measures in order to diminish the impact of disruptive conditions in the air transport system.