An Improved Flood Susceptibility Model for Assessing the Correlation of Flood Hazard and Property Prices using Geospatial Technology and Fuzzy-ANP

This study proposes an integrated Geographic Information System (GIS) Fuzzy Multi-Criteria Decision Making (F-MCDM) model to assess the impacts of flood on residential property prices. Triangular Fuzzy numbers was implemented to address limitations such as uncertainty, bias and ambiguity inherent in the conventional Analytic Network Process (ANP) flood model criteria ranking thereby improving the accuracy and reliability of the susceptibility model. The developed Fuzzy-ANP's (F-ANP) pair-wise comparison technique was used to rank the relative importance of nine flood conditioning criteria based on experts' input. Utilizing GIS and re-mote sensing data and techniques on Kelantan, a perennially flooded state in Malaysia, FANP based criterion maps were generated and aggregated to produce flood susceptibility maps of the area, showing the flood vulnerability levels of different locations. A 10-year inventory of real estate prices from the National Property Information Centre (NAPIC), Malaysia was analyzed to investigate the trend in market prices of residential properties situated in the high flood probable zones highlighted by the spatial F-ANP model. Model validation results showed that 59.42% and 36.23% of past flood events fall within the very high and high susceptible locations of the susceptibility map respectively, confirming its high accuracy. A weak positive correlation also exists between the highly susceptible flood class and housing locations vs market prices. We conclude that the ensemble GIS-FANP flood susceptibility model can produce maps capable of conveying accurate risk information to a broad range of stakeholders thereby facilitating decision making. However, other factors such as supply and demand, construction cost, macro-economy and micro-economy tend to also exert some influence on real estate prices, together with location in hazard-prone areas.