Flood Susceptibility Map Data of Jeju Island Using Probabilistic Methods

Floods are among the most widespread and destructive natural disasters globally, and Jeju Island in South Korea is particularly vulnerable due to its unique geological and hydrological characteristics. This study aims to produce a flood susceptibility map data for Jeju Island using a probabilistic method known as the frequency ratio (FR) model. The research integrates diverse spatial datasets including topography, geology, soil properties, land use, vegetation, and rainfall to analyze their correlation with historical flood occurrences. A flood inventory was constructed from multiple sources such as satellite images and disaster records. These data were used to train and validate the FR model. The model’s predictive performance was evaluated using the receiver operating characteristic-area under curve (ROC-AUC), achieving an accuracy 82.33%, indicating high accuracy. The resulting map data offers valuable insights for disaster mitigation, urban planning, and climate adaptation, especially in geologically complex regions like Jeju Island.