Over the last decade, flood disasters have affected millions of people and caused massive economic losses. Social vulnerability assessment uses a combination of several factors to represent a population's differential access to resources and its ability to cope with and respond to hazards. In this paper, social vulnerability assessment to flood risk was applied to the third most populous Portuguese municipality. The study was developed at the neighbourhood level, allowing for social vulnerability analysis at inter civil parish, intra civil parish, and municipality scales. A geographic information system-based multicriteria decision analysis (GIS-MCDA) was applied to social vulnerability and allows for an increased understanding and improved monitoring of social vulnerability over space, identifying 'hot spots' that require adaptation policies. Mafamude, Oliveira do Douro, Vila Nova de Gaia, and Avintes civil parishes display the greatest vulnerability to flooding. According to the most pessimistic scenario 57%À68% of the area of these civil parishes is classed at a high or very high level of social vulnerability. The GIS-MCDA helps to assess what and who is at risk, and where targeted impact-reduction strategies should be implemented. The results demonstrate the importance of an urban-scale approach instead of a river basin scale to urban flood risk management plans.

Southern Portugal is characterized by an irregular distribution of precipitation which is highly variable from year to year, requiring a clear ascertaining if the pattern of the precipitation is changing due to climate variability or climate change.In this paper, the authors applied several statistical methods to annual and seasonal time series of precipitation during the period 1931-2006, in order to detect trends and evaluate the variability in the frequencies of occurrence of dry and wet periods and also to assess spatial distribution patterns.The statistical methods include homogeneity tests, Mann-Kendall test, simple moving averages and cluster analysis. In spring, the authors detected a lack of homogeneity in the precipitation time series and a significant decreasing trend of precipitation was identified, by using simple moving averages and the Mann-Kendall test.Cluster analyses also confirmed the results and identified significant trends in the interannual and spatial distributions of dry, normal and wet years. It is concluded that the annual precipitation regime in the North interior sector is becoming drier, as the winter and spring also become drier. The only season in which a precipitation increase is observed is at fall and near the sea, but even for that region springs are also becoming drier.

No abstract