Predicting Arrival Delays: An Application of Spatial Analysis

Analysts have many tools available to forecast delays. However, spatial analysis does not readily come to mind when predicting delays. Based on the case study of Newark Liberty International airport (EWR), this study proposes to illustrate the potential application of prevalent geostatistical techniques to delay forecast. Arguably, there is a high degree of dependence among delays in a space or neighborhood defined by hour of operation and by day. Local spatial autocorrelation statistics can help determine how delays in a space are autocorrelated to other ones. Among the other spatial analytical techniques, kriging enables the interpolation of delay estimates at unobserved spaces based on the values at observed spaces. Error estimates can be mapped to define spatial patterns (spaces where delays are likely to be more intense). Finally, spatial error regression provides a method for analysts to test the reliability of their findings when spatial dependency in errors is not taken into account.


Issue Date:
2010-03
Publication Type:
Conference Paper/ Presentation
PURL Identifier:
http://purl.umn.edu/207260
Total Pages:
16




 Record created 2017-04-01, last modified 2017-08-28

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)