Robust identification of air traffic flow patterns in Metroplex terminal areas under demand uncertainty

• Published in: Transportation research. Part C, Emerging technologies Vol. 75; pp. 212 - 227
• Main Authors: Sidiropoulos, Stavros, Han, Ke, Majumdar, Arnab, Ochieng, Washington Y.
• Format: Journal Article
• Language: English
• Published: Elsevier India Pvt Ltd 01.02.2017
• Subjects: Air traffic demand; Air traffic management; Distributionally robust optimization; Multi-airport system; Terminal area operation; Air traffic demand; Multi-airport system; Terminal area operation; Air traffic management; Distributionally robust optimization
• ISSN: 0968-090X; 1879-2359; 1879-2359
• DOI: 10.1016/j.trc.2016.12.011

•The current static airspace structure in Metroplex is sub-optimal.•A new concept of operation based on dynamic route service policy is proposed.•We provide a robust framework for characterizing dynamic traffic demand in Metroplex.•Distributionally robust optimization is employed to account for demand uncertainties.•A case study in the New York Metroplex is conducted and validated by SMEs. Multi-Airport Systems (MAS), or Metroplexes, serve air traffic demand in cities with two or more airports. Due to the spatial proximity and operational interdependency of the airports, Metroplex airspaces are characterized by high complexity, and current system structures fail to provide satisfactory utilization of the available airspace resources. In order to support system-level design and management towards increased operational efficiency in such systems, an accurate depiction of major demand patterns is a prerequisite. This paper proposes a framework for the robust identification of significant air traffic flow patterns in Metroplex systems, which is aligned with the dynamic route service policy for the effective management of Metroplex operations. We first characterize deterministic demand through a spatio-temporal clustering algorithm that takes into account changes in the traffic flows over the planning horizon. Then, in order to handle uncertainties in the demand, a Distributionally Robust Optimization (DRO) approach is proposed, which takes into account demand variations and prediction errors in a robust way to ensure the reliability of the demand identification. The DRO-based approach is applied on pre-tactical (i.e. one-day planning) as well as operational levels (i.e. 2-h rolling horizon). The framework is applied to Time Based Flow Management (TBFM) data from the New York Metroplex. The framework and results are validated by Subject Matter Experts (SMEs).