Physical Vulnerability Modeling Based On Flood Inundation Model and Image Mining

Abstract

Flash flood disaster occurred within the City of Garut, West Java, Indonesia, on 20 th September 2016, which caused many casualties and damages. Flood model could be performed to model the already-occurring disaster, as well as to depict future events that may occur to overcome any potential disasters, where the inundation flood model depicted the element at risk. In order to assist the analysis for the damages occurred, image mining could be used as part of the approach, where online media was utilized as well. The image mining resulted information about building damages caused by the flood. Afterwards, the physical vulnerability (buildings/residents) model could be further performed. Finally, the relationship between vulnerability and the flood inundation were portrayed. The resulted physical vulnerability model showed that larger height of the flood water caused higher degree of loss of the building, in which portrayed the need for total rebuild of houses as well. Considering available open source data and fast data acquisition, the approach showed such efficient approaches, where the results could be used in order to establish recommendation for building reinforcement, spatial planning, or protection wall in flood prone areas within the future time.