Analysis of tailing pond contamination in Galicia using generalized linear spatial models

  1. Javier Taboada 1
  2. Ángeles Saavedra 2
  3. María Paz 1
  4. Fernando G. Bastante 1
  5. Leandro R. Alejano 1
  1. 1 Department of Natural Resources, University of Vigo, Vigo, Spain
  2. 2 Department of Statistics, University of Vigo, Vigo, Spain
Revista:
DYNA: revista de la Facultad de Minas. Universidad Nacional de Colombia. Sede Medellín

ISSN: 0012-7353

Ano de publicación: 2015

Volume: 82

Número: 189

Páxinas: 76-83

Tipo: Artigo

DOI: 10.15446/DYNA.V82N189.42069 DIALNET GOOGLE SCHOLAR lock_openAcceso aberto editor

Outras publicacións en: DYNA: revista de la Facultad de Minas. Universidad Nacional de Colombia. Sede Medellín

Resumo

We statistically analysed the chemical components present in waste water from mines in Galicia (NW Spain). These elements pose a risk to public health and the environment, most particularly in the event of a failure in the containment structure of a pond or dam. The statistical processing of the data, which started with an analysis of the typical contaminants present in mining ponds and dams, pointed to the potential limitations of using non-spatial models for spatially structured data. Our results indicate the greater potential of the generalized linear spatial model over the generalized linear model for analysis of spatially structured data. We also show how a misspecification of the model for analysing spatial data can lead to misleading conclusions, which might lead, in turn, to poorly designed protective or corrective measures.