Using a non-homogeneous Poisson model with spatial anisotropy and change-points to study air pollution data
Nenhuma Miniatura disponível
Data
2019-06-01
Autores
Rodrigues, Eliane R.
Nicholls, Geoff
Tarumoto, Mario H. [UNESP]
Tzintzun, Guadalupe
Título da Revista
ISSN da Revista
Título de Volume
Editor
Springer
Resumo
A non-homogeneous Poisson process is used to study the rate at which a pollutant's concentration exceeds a given threshold of interest. An anisotropic spatial model is imposed on the parameters of the Poisson intensity function. The main contribution here is to allow the presence of change-points in time since the data may behave differently for different time frames in a given observational period. Additionally, spatial anisotropy is also imposed on the vector of change-points in order to account for the possible correlation between different sites. Estimation of the parameters of the model is performed using Bayesian inference via Markov chain Monte Carlo algorithms, in particular, Gibbs sampling and Metropolis-Hastings. The different versions of the model are applied to ozone data from the monitoring network of Mexico City, Mexico. An analysis of the results obtained is also given.
Descrição
Palavras-chave
Anisotropic spatial model, Bayesian inference, Change-points, Markov chain Monte Carlo algorithms, Non-homogeneous Poisson process
Como citar
Environmental And Ecological Statistics. Dordrecht: Springer, v. 26, n. 2, p. 153-184, 2019.