Modelagem espaço temporal do uso do solo e potencial de erosão no Vale do Ribeira / Spatio temporal modeling the soil use and potential of erosions in the Ribeira Valley

AUTOR(ES)
DATA DE PUBLICAÇÃO

2008

RESUMO

This paper comprises an experiment to map hidden patterns in multi-source data that could be associated with Pb geochemical soil anomalies. The study area is confined between the Apiaí Fold-Belt, hosts numerous Pb and Zn deposits; it was under the influence of regional Pb mining and Pb refinery factory activities until 1995; and exhibits soils enriched in Pb, As and other heavy metals yielded from the supergenic alteration of metavolcanic and metasedimentary rocks that host the metallic mineralizations. Using Neural Networks available in the ArcSDM (Arc Spatial Data Modeler) package, it was demonstrated that there is a signature indicated by the data that agrees with anomalous concentrations of Pb in soil over control areas. This signature was also detected in sectors for which no soil geochemical surveys were available, but a regional stream sediment survey. This survey also revealed a coincidence among tracts with local multi-source signatures and high concentration in Pb, indicating the consistency of the pattern against both soil and stream sediment Pb anomalies. This results have several implications to the use of scarcely spaced sampling geochemistry to detect patterns that can be linked to environmental risk areas

ASSUNTO(S)

celuar automata arsenic lead runoff redes neurais (computação) arsênio neural networks (computer science) runoff erosion erosão chumbo automato celular - modelos matemáticos

ACESSO AO ARTIGO

Documentos Relacionados