Empirical and Mechanistic Models for Flight Delay Risk Distributions

Abstract

In the field of robust flight optimization, disruption and delay propagation play a central role. Recently, huge amounts of data became available for analysis of these delays. For example, empirical distribution functions can be estimated and used as input for scheduling systems. But for a better understanding of the underlying mechanisms, distribution models with explicit dependencies and interpretable parameters are more desirable. In this paper we present new patterns in flight delay data and an initial model of the underlying physical processes.