Two-Stage Approach to Identify Flight Delay Patterns

As the demand for air travel grows, more efforts have been made to maintain high schedule fidelity. To develop an effective strategy for reducing delay, it is important to observe the trend of flight delay and identify the factors associated with it. Given that flight schedule is affected by cyclic variation in passenger demand and weather, it is expected that flight delay follows certain periodic pattern. In this regard, this study proposes two-stage approach to identify flight delay patterns. Two-stage approach comprises 1) a frequency analysis technique for detecting the periods of any regularly repeating delay patterns and 2) statistical analysis techniques (ANOVA and logistic regression) for identifying the factors correlated with detected frequencies of delay. This study uses the on-time performance data for non-stop domestic flights arriving at Orlando International Airport during 2002-2003. Using the proposed approach, the study detected daily, weekly and seasonal patterns of arrival delay and identified important factors associated with delay such as time of day, day of week, season, flight distance, precipitation and time intervals between successive flights.