Exploratory spatial analysis of pilot fatality rates in general aviation crashes using geographic information systems

Jurek G. Grabowski, Frank C. Curriero, Susan P. Baker, Guohua Li

Research output: Contribution to journalArticle

Abstract

Geographic information systems and exploratory spatial analysis were used to describe the geographic characteristics of pilot fatality rates in 1983-1998 general aviation crashes within the continental United States. The authors plotted crash sites on a digital map; rates were computed at regular grid intersections and then interpolated by using geographic information systems. A test for significance was performed by using Monte Carlo simulations. Further analysis compared low-, medium-, and high-rate areas in relation to pilot characteristics, aircraft type, and crash circumstance. Of the 14,051 general aviation crashes studied, 31% were fatal. Seventy-four geographic areas were categorized as having low fatality rates and 53 as having high fatality rates. High-fatality-rate areas tended to be mountainous, such as the Rocky Mountains and the Appalachian region, whereas low-rate areas were relatively flat, such as the Great Plains. Further analysis comparing low-, medium-, and high-fatality-rate areas revealed that crashes in high-fatality-rate areas were more likely than crashes in other areas to have occurred under instrument meteorologic conditions and to involve aircraft fire. This study demonstrates that geographic information systems are a valuable tool for injury prevention and aviation safety research.

Original languageEnglish (US)
Pages (from-to)398-405
Number of pages8
JournalAmerican journal of epidemiology
Volume155
Issue number5
DOIs
StatePublished - Mar 1 2002

Keywords

  • Accidents
  • Aviation
  • Geography
  • Information systems
  • Monte Carlo method

ASJC Scopus subject areas

  • Epidemiology

Fingerprint Dive into the research topics of 'Exploratory spatial analysis of pilot fatality rates in general aviation crashes using geographic information systems'. Together they form a unique fingerprint.

  • Cite this