Citation

Diamoutene A, Kamsu-Foguem B, Noureddine F, Barro D. Transp. Res. A Policy Pract. 2018; 109: 65-75.

Copyright

(Copyright © 2018, Elsevier Publishing)

DOI

10.1016/j.tra.2018.01.022

PMID

unavailable

Abstract

General Aviation is the main component of the United States civil aviation and the most aviation accidents concern this aviation category. Between early 2015 and May 17, 2016, a total of 1546 general aviation accidents in the United States has left 466 fatalities and 384 injured. Hence, in this study, we investigate the risk of U.S. General Aviation accidents by examining historical U.S. General Aviation accidents. Using the Peak Over Threshold approach and Generalized Pareto Distribution, we predict the number of fatalities resulting in extreme GA accidents in the future operations. We use a graphical method and intensive parameters estimates to obtain the optimal range of the threshold. In order to assess the uncertainty in the inference and the accuracy of the results, we use the nonparametric bootstrap approach.

Language: en

Keywords

Extreme value statistics; General Aviation accident; Graphical diagnosis; Nonparametric bootstrap; Return level; Uncertainty