Integrated Index for Flood Risk Assessment Development Using Bayesian Network

Abstract

The purpose of this study is to develop a reliable flood vulnerability index that can be used to verify scientific data. Existing flood vulnerability assessment uses methods commonly used in calculating the weight between indicators. However, because they contain similarities in their estimation measurement, such methods cannot be differentiated, and their verification is not reliable. Therefore, it is necessary to improve flood vulnerability assessment by differentiating its method from that used in existing research and to use the new method to reliably predict and evaluate the flood damage risk. In order to realize this, optimal weighting is derived by using a Bayesian network, which can make reasonable and integrated decisions after considering the existing weighting method. As a result, the integrated index for Flood Risk Assessment (InFRA) was developed through presentation of the formula. Qualitative and quantitative comparison with existing evaluation methods revealed that InFRA showed better results than previous methods and thus proved its applicability.