ANÁLISIS POR COMPONENTES PRINCIPALES DE DATOS PLUVIOMETRICOS
A) APLICACIÓN A LA DETECCIÓN DE DATOS ANÓMALOS
Carlos López - Jorge Goyret
Centro de Cálculo
Elizabeth González
Instituto de mecánica de los fluidos e Ingeniería ambiental
Engineering Faculty - University of Uruguay,
Montevideo, URUGUAY
ABSTRACT
The techniques employed in the treatment of a pluviometric data base used during the development and calibration phases of a Flow-Rain, Flow hydrological model are presented.It's well known that such a model is strongly affected by errors (outliers) in the data, both random and systematic. So it is needed to remove them prior to use the data bank.For this case, some different methodologies have been applied. From them, the most successful was the one based in Principal Component Analysis (PCA).The methodology is liable to be used in real time, involving minimum computer resources. For the stages described here, only errors coming from manually digitizing are considered. However, it is suggested that PCA may help in detecting random errors from the observer himself, and also some kind of systematic errors, all of which is still in an investigation phase.
Published in:
ESTADISTICA(1994), 46, 146,147, pp. 25-54.
© Instituto Interamericano de Estadística.
If you are still interested in it; here you have:
(224KB)
AN ENGLISH TRANSLATION (.PDF)(698KB)
carlos.lopez@ieee.org
elizabet@fing.edu.uy