Using INLA to Estimate a Highly Dimensional Spatial Model for Forest Fires in Portugal

Isabel Natário, Maria Manuela Oliveira, Susete Marques

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

Abstract

Within the context of accessing the risk of forest fires, Amaral-Turkman et al. (Environ. Ecol. Stat. 18:601–617, 2011) have proposed a spatio-temporal hierarchical approach which jointly models the fire ignition probability and the fire’s size, in a Bayesian framework. This is recovered and applied to Portuguese forest fires data, with some necessary modifications in what concerns the format of the data (not available in a regular lattice over the territory) and also because of the estimation complications that arise due to the high dimensionality of the neighbouring structure involved. To address the latter, as it compromises the estimation via Markov Chain Monte Carlo (MCMC) methods, and having the model be recognized as a latent Gaussian model, it was chosen to do the Bayesian estimation also using an Integrated Nested Laplace Approximation approach, with real computational advantages. Corresponding methodologies and results are described and compared.
Original languageEnglish
Title of host publicationNew Advances in Statistical Modeling and Applications
EditorsAntónio Pacheco, Rui Santos, do Rosário Maria Oliveira, Daniel Carlos Paulino
Place of PublicationCham
PublisherSpringer International Publishing
Pages239-247
Number of pages9
VolumePart IV
ISBN (Electronic)978-3-319-05323-3
ISBN (Print)978-3-319-05322-6
DOIs
Publication statusPublished - 3 Apr 2014

Publication series

NameStudies in Theoretical and Applied Statistics
PublisherSpringer International Publishing

Keywords

  • Forest Fire
  • Markov Chain Monte Carlo
  • Markov Chain Monte Carlo Method
  • Deviance Information Criterion

Fingerprint

Dive into the research topics of 'Using INLA to Estimate a Highly Dimensional Spatial Model for Forest Fires in Portugal'. Together they form a unique fingerprint.

Cite this