TITLE:
Correspondence Analysis on a Space-Time Data Set for Multiple Environmental Variables
AUTHORS:
Palma Monica
KEYWORDS:
Space-Time Data, Indicator Transform, Correspondence Analysis
JOURNAL NAME:
International Journal of Geosciences,
Vol.6 No.10,
October
29,
2015
ABSTRACT: Applications of the
multivariate technique called correspondence analysis for environmental studies
are relatively new and are limited to spatial multivariate data set. In this
paper, a procedure of applying correspondence analysis to a large space-time
data set for multiple environmental variables is shown. In particular, nitrogen
dioxide and carbon monoxide hourly concentrations measured during January 1999
at several monitored stations in a district of Northern Italy are analyzed. The
procedure consists in transforming the continuous variables into categorical
ones by the means of appropriate indicator variables, generating special
contingency tables and applying correspondence analysis. The use of this
classical multivariate technique allows the identification of important
relationships among pollution levels and monitoring stations and/or
relationships among pollution levels and observation times.