Locally accurate prediction standard errors with spatially varying regression coefficient models A comparison of techniques

Paul Harris, Stewart Fotheringham, Chris Brunsdon

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This study assesses the prediction and prediction uncertainty performance of models that cater for both: (i) a nonstationary relationship between the response and a contextual variable and (ii) a nonstationary residual variance (or variogram), at point locations for a single realisation spatial process. Here the crucial aspect of the model specification is allowing the residual variance (or variogram) to vary across space. Without this, the estimated prediction standard errors are only likely to be accurate in a global (or overall) sense and not the desired, local sense. Locally-accurate prediction standard errors, allow locally-relevant prediction confidence intervals and/or locally-relevant estimates of risk (e.g. the risk of exceeding some critical threshold) which is not only valuable to researchers who attempt to model spatial processes, but also to policy makers who need to plan and manage the outcomes of spatial processes at different spatial scales.

Original languageEnglish (US)
Title of host publicationAccuracy 2010 - Proceedings of the 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences
PublisherInternational Spatial Accuracy Research Association (ISARA)
Pages137-140
Number of pages4
StatePublished - 2010
Externally publishedYes
Event9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences, Accuracy 2010 - Leicester, United Kingdom
Duration: Jul 20 2010Jul 23 2010

Other

Other9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences, Accuracy 2010
CountryUnited Kingdom
CityLeicester
Period7/20/107/23/10

Fingerprint

prediction
variogram
confidence interval
comparison
plan
policy

Keywords

  • Bayesian prediction models
  • Georaphically weighted regression
  • Heteroskedastic
  • Moving window kriging

ASJC Scopus subject areas

  • Environmental Science(all)

Cite this

Harris, P., Fotheringham, S., & Brunsdon, C. (2010). Locally accurate prediction standard errors with spatially varying regression coefficient models A comparison of techniques. In Accuracy 2010 - Proceedings of the 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences (pp. 137-140). International Spatial Accuracy Research Association (ISARA).

Locally accurate prediction standard errors with spatially varying regression coefficient models A comparison of techniques. / Harris, Paul; Fotheringham, Stewart; Brunsdon, Chris.

Accuracy 2010 - Proceedings of the 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences. International Spatial Accuracy Research Association (ISARA), 2010. p. 137-140.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Harris, P, Fotheringham, S & Brunsdon, C 2010, Locally accurate prediction standard errors with spatially varying regression coefficient models A comparison of techniques. in Accuracy 2010 - Proceedings of the 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences. International Spatial Accuracy Research Association (ISARA), pp. 137-140, 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences, Accuracy 2010, Leicester, United Kingdom, 7/20/10.
Harris P, Fotheringham S, Brunsdon C. Locally accurate prediction standard errors with spatially varying regression coefficient models A comparison of techniques. In Accuracy 2010 - Proceedings of the 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences. International Spatial Accuracy Research Association (ISARA). 2010. p. 137-140
Harris, Paul ; Fotheringham, Stewart ; Brunsdon, Chris. / Locally accurate prediction standard errors with spatially varying regression coefficient models A comparison of techniques. Accuracy 2010 - Proceedings of the 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences. International Spatial Accuracy Research Association (ISARA), 2010. pp. 137-140
@inproceedings{22c90335ad604ee9a5eba23371ed4464,
title = "Locally accurate prediction standard errors with spatially varying regression coefficient models A comparison of techniques",
abstract = "This study assesses the prediction and prediction uncertainty performance of models that cater for both: (i) a nonstationary relationship between the response and a contextual variable and (ii) a nonstationary residual variance (or variogram), at point locations for a single realisation spatial process. Here the crucial aspect of the model specification is allowing the residual variance (or variogram) to vary across space. Without this, the estimated prediction standard errors are only likely to be accurate in a global (or overall) sense and not the desired, local sense. Locally-accurate prediction standard errors, allow locally-relevant prediction confidence intervals and/or locally-relevant estimates of risk (e.g. the risk of exceeding some critical threshold) which is not only valuable to researchers who attempt to model spatial processes, but also to policy makers who need to plan and manage the outcomes of spatial processes at different spatial scales.",
keywords = "Bayesian prediction models, Georaphically weighted regression, Heteroskedastic, Moving window kriging",
author = "Paul Harris and Stewart Fotheringham and Chris Brunsdon",
year = "2010",
language = "English (US)",
pages = "137--140",
booktitle = "Accuracy 2010 - Proceedings of the 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences",
publisher = "International Spatial Accuracy Research Association (ISARA)",

}

TY - GEN

T1 - Locally accurate prediction standard errors with spatially varying regression coefficient models A comparison of techniques

AU - Harris, Paul

AU - Fotheringham, Stewart

AU - Brunsdon, Chris

PY - 2010

Y1 - 2010

N2 - This study assesses the prediction and prediction uncertainty performance of models that cater for both: (i) a nonstationary relationship between the response and a contextual variable and (ii) a nonstationary residual variance (or variogram), at point locations for a single realisation spatial process. Here the crucial aspect of the model specification is allowing the residual variance (or variogram) to vary across space. Without this, the estimated prediction standard errors are only likely to be accurate in a global (or overall) sense and not the desired, local sense. Locally-accurate prediction standard errors, allow locally-relevant prediction confidence intervals and/or locally-relevant estimates of risk (e.g. the risk of exceeding some critical threshold) which is not only valuable to researchers who attempt to model spatial processes, but also to policy makers who need to plan and manage the outcomes of spatial processes at different spatial scales.

AB - This study assesses the prediction and prediction uncertainty performance of models that cater for both: (i) a nonstationary relationship between the response and a contextual variable and (ii) a nonstationary residual variance (or variogram), at point locations for a single realisation spatial process. Here the crucial aspect of the model specification is allowing the residual variance (or variogram) to vary across space. Without this, the estimated prediction standard errors are only likely to be accurate in a global (or overall) sense and not the desired, local sense. Locally-accurate prediction standard errors, allow locally-relevant prediction confidence intervals and/or locally-relevant estimates of risk (e.g. the risk of exceeding some critical threshold) which is not only valuable to researchers who attempt to model spatial processes, but also to policy makers who need to plan and manage the outcomes of spatial processes at different spatial scales.

KW - Bayesian prediction models

KW - Georaphically weighted regression

KW - Heteroskedastic

KW - Moving window kriging

UR - http://www.scopus.com/inward/record.url?scp=84975726994&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84975726994&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84975726994

SP - 137

EP - 140

BT - Accuracy 2010 - Proceedings of the 9th International Symposium on Spatial Accuracy Assessment in Natural Resources and Environmental Sciences

PB - International Spatial Accuracy Research Association (ISARA)

ER -