Remote sensing analysis for flood risk management in urban sprawl contexts

Remote sensing can play a key role in risk assessment and management, especially when several concurrent factors coexist, such as a predisposition to natural disasters and the urban sprawl, spreading over highly vulnerable areas. In this context, multitemporal analysis can provide decision-makers with tools and information to reduce the impacts of disasters (e.g. flooding) and to encourage a sustainable development. The present work focuses on the employment of multispectral satellite imagery to produce multitemporal land use/cover maps for the city of Dhaka, which is subject to frequent flooding events. In particular, the evaluation of the urban growth, the analysis of the annual dynamics of flooding and the study of the 2004 catastrophic event were performed. For the change-detection procedure, Landsat images were used. These images allow the quantification of the very rapid growth of the metropolis, with an increase in built-up areas from 75 to 111km(2). The image of 2009 showed that an ordinary flood affects about 115km(2) (on a studied area of 591km(2)). On the other hand, the analysis of the 2004 extreme flooding event, performed on a wider area, showed that the affected lands added up to 750km(2) (on about 3845km(2)).