Flood Susceptibility Mapping Using a Frequency Ratio Model: A Case Study of Chai Nat Province, Thailand

Abstract

AbstractFlooding has become more prevalent in many regions of Southeast Asian countries in recent decades. Intense precipitation, settlement in low-lying areas, population growth, and rapid urbanization can enhance vulnerability to floods and lead to serious hazards. This study developed flood susceptibility mapping for the Chai Nat province of Thailand using flood-conditioning factors and the frequency ratio (FR) method. The flood inventory (2005–2017) was randomly separated into a training dataset for FR analysis and a testing dataset for model validation. Eleven flood-conditioning parameters, i.e., altitude, slope, curvature, the topographic wetness index, rainfall, distance to drainage, drainage density, soil drainage, land use, the normalized difference vegetation index, and road density, were considered for this study. While constructing the flood susceptibility index, the relative frequency and predictor rate were used to create the flooding probability for each factor class and the weight of each factor in the model. The values for the flood susceptibility index were classified into five categories and used to make a flood susceptibility map. The area under the curve (AUC) was used to validate the model prediction. The results indicate that the AUC values for the success and prediction rates are 74.2% and 75.1%, respectively.KeywordsFlood susceptibilityFrequency ratioPredictor rateGISThailand