Por Manuel Ribeiro (CERENA, Instituto Superior Técnico, Universidade de Lisboa).
Outdoor air pollution is an important issue for public health and can contribute to the incidence of several diseases affecting the central nervous system, cardiovascular system or respiratory system, just to name a few. Still, apart from few exceptions, there is not a definite knowledge about the impacts of air pollution on health. So, a better understanding and modelling of the impacts is of great importance. In most studies correlating health outcomes with air pollution, personal exposure assignments are based on air pollution data collected at very specific locations, not coinciding with health data locations. So, spatial interpolators are needed to predict air pollution at unsampled locations and to assign exposures at health data locations. Methodological developments in geostatistical framework over the last two decades have been playing a positive role to interpolate air pollution exposures accurately and to provide measures of spatial uncertainty for each prediction. The latter is of key importance since predicted exposures can be misleading if they do not take into account that the extent of uncertainty varies throughout the spatial domain. In this seminar, we present how geostatistical models are used for exposure and exposure uncertainty assessment, and how to combine these with regression methods to predict and assess spatial uncertainty of air pollution impacts on health. We illustrate the methods with previous studies developed in the industrial region of Sines (Portugal) to assess the impacts of an air quality index on birth weight.
