Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Integrating Delay-Absorption Capability into Flight Departure Delay Prediction

Jianyang Zhou

TL;DR

This work tackles flight departure delay prediction by explicitly modeling delay-absorption during turnaround. It introduces a two-stage framework: Stage I uses CatBoost to estimate the probability of absorbing upstream delays (AbsorbScore), and Stage II incorporates AbsorbScore into an XGBoost classifier to predict departures exceeding 15 minutes late. The approach yields substantial performance gains over baselines (AUC ~0.898) and provides interpretable insights into airport-specific recovery dynamics, informing proactive delay-management strategies. The methodology leverages BTS on-time performance data and NOAA weather information to produce a practically actionable framework for resilience in airline operations.

Abstract

Accurately forecasting flight departure delays is essential for improving operational efficiency and mitigating the cascading disruptions that propagate through tightly coupled aircraft rotations. Traditional machine learning approaches often treat upstream delays as static variables, overlooking the dynamic recovery processes that determine whether a delay is absorbed or transmitted to subsequent legs. This study introduces a two-stage machine learning framework that explicitly models delay-absorption behavior and incorporates it into downstream delay prediction. In Stage I, a CatBoost classifier estimates the probability that a flight successfully absorbs an upstream delay based on operational, temporal, and meteorological features. This probability, termed AbsorbScore, quantifies airport- and flight-specific resilience to delay propagation. In Stage II, an XGBoost classifier integrates AbsorbScore with schedule, weather, and congestion indicators to predict whether a flight will depart more than 15 minutes late. Using U.S. domestic flight and NOAA weather data from Summer 2023, the proposed framework achieves substantial improvements over baseline models, increasing ROC-AUC from 0.865 to 0.898 and enhancing precision to 89.2% in identifying delayed flights. The results demonstrate that modeling delay absorption as an intermediate mechanism significantly improves predictive performance and yields interpretable insights into airport recovery dynamics, offering a practical foundation for data-driven delay management and proactive operational planning.

Integrating Delay-Absorption Capability into Flight Departure Delay Prediction

TL;DR

This work tackles flight departure delay prediction by explicitly modeling delay-absorption during turnaround. It introduces a two-stage framework: Stage I uses CatBoost to estimate the probability of absorbing upstream delays (AbsorbScore), and Stage II incorporates AbsorbScore into an XGBoost classifier to predict departures exceeding 15 minutes late. The approach yields substantial performance gains over baselines (AUC ~0.898) and provides interpretable insights into airport-specific recovery dynamics, informing proactive delay-management strategies. The methodology leverages BTS on-time performance data and NOAA weather information to produce a practically actionable framework for resilience in airline operations.

Abstract

Accurately forecasting flight departure delays is essential for improving operational efficiency and mitigating the cascading disruptions that propagate through tightly coupled aircraft rotations. Traditional machine learning approaches often treat upstream delays as static variables, overlooking the dynamic recovery processes that determine whether a delay is absorbed or transmitted to subsequent legs. This study introduces a two-stage machine learning framework that explicitly models delay-absorption behavior and incorporates it into downstream delay prediction. In Stage I, a CatBoost classifier estimates the probability that a flight successfully absorbs an upstream delay based on operational, temporal, and meteorological features. This probability, termed AbsorbScore, quantifies airport- and flight-specific resilience to delay propagation. In Stage II, an XGBoost classifier integrates AbsorbScore with schedule, weather, and congestion indicators to predict whether a flight will depart more than 15 minutes late. Using U.S. domestic flight and NOAA weather data from Summer 2023, the proposed framework achieves substantial improvements over baseline models, increasing ROC-AUC from 0.865 to 0.898 and enhancing precision to 89.2% in identifying delayed flights. The results demonstrate that modeling delay absorption as an intermediate mechanism significantly improves predictive performance and yields interpretable insights into airport recovery dynamics, offering a practical foundation for data-driven delay management and proactive operational planning.

Paper Structure

This paper contains 27 sections, 8 equations, 9 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Machine Learning Approaches
  4. Delay Propagation in Air Transportation
  5. Analytical Foundations
  6. Analytical--Econometric Framework
  7. Network-Level Classification
  8. Research Gap and Contribution
  9. Dataset Review and Feature Engineering
  10. Dataset Overview
  11. Departure Delay Characteristics
  12. Airport's performance on delay absorption
  13. Methodology
  14. Overview
  15. Definition of DelayAbsorbed
  16. ...and 12 more sections

Figures (9)

  • Figure 1: Time-space graph of a single flight and ground buffer time illustration
  • Figure 2: Distribution of departure delay categories
  • Figure 3: Top 20 busiest airports: traffic vs delay absorption index
  • Figure 4: Airport absorbed delay across previous delay levels
  • Figure 5: ATL's delay absorption capability in different weather
  • ...and 4 more figures