Multivariate Flood Damage Model: A Case Study of 2013 Kuantan Flood

Flood damage assessment is a crucial element to be considered in the implementation of efficient flood risk management. Hence, with the aim to examine the significant correlation between the impact and resistance parameters with the degree of flood damages, this study utilized the multiple linear regression (MLR) technique to develop a multivariate flood damage model of 2013 Kuantan flood. This study focuses on estimating flood damage in urban areas for residential and commercial buildings. The damage type assessed in this study is direct tangible damage which includes damage to buildings structure and its contents. The study’s findings revealed that socio-economy and property characteristics have also shown a significant correlation to the degree of flood damage. Flood depth was found to have a significant effect on all damage categories as in previous studies. However, in addition to flood depth, flood damage is also influenced by the value of properties, income, building material, building ownership status and type of properties. The MLR models validations are showing reasonable results compared with other studies with mean absolute error (MAE), root mean square error (RMSE), mean bias error (MBE), coefficient of variation (CV) which are closer to 0 and HR is close to 0.9.