Corrected prediction intervals for change detection in paired watershed studies

Hydrological data may be temporally autocorrelated requiring autoregressive process parameters to be estimated. Current statistical methods for hydrological change detection in paired watershed studies rely on prediction intervals, but the current form of prediction intervals does not include all ap...

Full description

Saved in:
Bibliographic Details
Published in:Hydrological sciences journal 2012, Vol.57 (1), p.134-143
Main Authors: Som, Nicholas A, Zégre, Nicolas P, Ganio, Lisa M, Skaugset, Arne E
Format: Article
Language:English
Subjects:
covariance
débit
Earth sciences
Earth, ocean, space
estimateur d'erreur quadratique moyenne de Prasad-Rao
Exact sciences and technology
generalized least squares
hydrologic data
Hydrology. Hydrogeology
moindres carrés généralisés
paired watershed study
Parameter estimation
Prasad-Rao mean-squared error estimator
prediction
Regression analysis
stream discharge
Studies
séries chronologiques
Time series
uncertainty
watersheds
étude de bassins versants appariés
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Hydrological data may be temporally autocorrelated requiring autoregressive process parameters to be estimated. Current statistical methods for hydrological change detection in paired watershed studies rely on prediction intervals, but the current form of prediction intervals does not include all appropriate sources of variation. Corrected prediction intervals for the analysis of paired watershed study data that include variation associated with covariance and linear model parameter estimation are presented. We provide an example of their application to data from the Hinkle Creek Paired Watershed Study located in the western Cascade foothills of Southern Oregon, USA. Research implications of using the correct prediction limits and incorporating the estimation uncertainty of autoregressive process parameters are discussed.Editor D. KoutsoyiannisCitation Som, N.A., Zégre, N.P., Ganio, L.M. and Skaugset, A.E., 2012. Corrected prediction intervals for change detection in paired watershed studies. Hydrological Sciences Journal, 57 (1), 134–143.
ISSN:2150-3435
0262-6667
2150-3435
DOI:10.1080/02626667.2011.637494