Causal analysis of flight en route inefficiency. (September 2021)

• Record Type: Journal Article
• Authors: Liu, Yulin; Hansen, Mark; Ball, Michael O.; Lovell, David J.
• Is Part Of: Transportation research. Volume 151(2021)
• Page Start: 91
• Page End: 115
• Publication Date: 2021-09
• Subjects: Air traffic flow management -- Air route structure -- ATM performance -- Trajectory clustering -- Discrete choice modeling -- En route inefficiency; Transportation -- Research -- Periodicals; Transportation -- Mathematical models -- Periodicals
• DOI: 10.1016/j.trb.2021.07.003
• Languages: English
• ISSNs: 0191-2615

Abstract

Abstract: En route inefficiency is measured in terms of extra distance flown by an aircraft, above a benchmark distance that relates to the theoretical shortest distance route (great circle route). In this paper, we have explored causal relations among en route inefficiency with multiple identified sources: convective weather, wind, miles-in-trail (MIT) restrictions, airspace flow programs (AFPs) and special activity airspace (SAA). We propose two mechanisms – strategic route choice and tactical reroute – to ascribe flight en route inefficiency to these factors. In our framework, we first propose an efficient trajectory clustering algorithm to identify nominal routes that reveal both air traffic flow patterns and air route structures in the national airspace system. Second, we develop a tree-based searching algorithm that matches different causal factors to the identified nominal routes in a 4-dimensional manner. Third, we employ a discrete choice modeling framework to quantitatively understand the flight route choice process, and establish a linear model to understand the tactical reroute process. Finally, we estimate the contributions of the identified causal factors to flight inefficiency through counterfactual analysis. Numerical experiments based on 8 representative airport pairs suggest strong negative impact of convection and headwind on route choice process, while MIT, AFP and SAA only demonstrate negative effects on some pairs. We have also shown that the contributions of convective weather, wind, traffic management initiatives, and SAA are up to 7.1%, 14.0%, 4.4%, 21.9%, respectively. Highlights: Propose two exact mechanisms interpreting sources of flight en route inefficiency. Design statistical framework to quantify causal factor en route inefficiency impact. Design a trajectory clustering algorithm to identify nominal air traffic flows. Design scalable trajectory matching algorithms to derive flight specific attributes. Discrete choice modeling the strategic route choice process of a flight.