A Flood Hazard Risk Assessment Map in Growing Urban Areas by Integrating Remote Sensing and DEM Data

This article presents a new approach based on an integration of multi-attributes semantic partitions (MASP) coming from the three following attributes: an urban growth layer computed from a couple of ERS-SAR images -exploited as a vulnerability map-, a dispersion flow (DF) layer estimated from a linear DF model -used as a flood hazard map-, and a NDVI layer used as an urban/non-urban interpretation measure, in order to produce a flood risk (FR) map. Thus, input mono-attribute semantic partitions (imASP) are first defined from attributes, and their membership degree functions are built based on the fuzzy subset theory. Then imASP are selected to form MASP according to the degree of confidence given to each one to perform the flood risk (FR) and membership degrees of MASP are calculated to provide individual degrees of FR worsening (FRW). Lastly, the global degree of FRWis computed by aggregating previous individual degrees of MASP by using a fuzzy integral (FI) to achieve the resulting FR map.