A modelação geoestatística e os Sistemas de Informação Geográfica : uma abordagem exploratória aplicada em Saúde Pública

• Autor(a) principal: Ribeiro, Manuel Luís Castro
• Data de Publicação: 2006
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/19980
• topic: Environment; Geostatistcs; Health; GIS

In this study a data driven approach based on exploratory spatial data analysis is used with geographic information technologies in order to formulate hypothesis about effects of particulate matter concentration (PM10) on child respiratory diseases and allergies. In order to integrate the spatial data in a common environment and to correlate the studied variables, a geographical referenced database was created for every variable with a Geographic Information Systems (GIS) tool with capabilities to store, manage and perform spatial modelling considering one or more variables. An exploratory spatial data analysis based on graphical methods and user interaction for analysis of spatial structures was conducted for both environmental and health variables and geostatistical modelling was performed to build child respiratory diseases and allergies and air pollution concentration, uncertainty and local correlation maps. Given the scarcity of data and the spatial location of both variables, a co-regionalization model was not possible to be driven from the previous analysis. Hence a stochastic spatial simulation was conducted independently for each variable. PM10 concentrations were related to child respiratory diseases and allergies incidence by local correlation coefficients obtained with simulated realizations of air pollution and health data. In the last step, local correlation maps associated to critical values were obtained.