Flood Susceptibility Assessment Using Geographic Information System and Statistical Index in the Mae Wang Watershed, Chiang Mai Province

Abstract

The objectives of this study are: 1) to assess flood susceptible areas in Mae Wang Watershed, and 2) to evaluate the performance of the model using the area under the curve method. The flood data were used to analyze the relationship between flooding and ten conditioning factors, namely: elevation, slope, aspect, normalized difference vegetation index (NDVI), topographic wetness index (TWI), rainfall, soil drainage, stream density, road density, and land use. The flood data were randomly categorized as training data (70%) and testing data (30%). The training data were analyzed using the ten conditioning factors mentioned above to calculate statistical index values. A flood susceptibility score was then assigned to each factor. The results revealed that the conditioning factors that had the highest susceptibility index pertained to poor soil drainage followed by the road density of 5.31 to 7.34 kilometers per square kilometer. According to the results from the flood susceptibility map, 65.11 square kilometers of Mae Wang Watershed can be classified as highly susceptible, accounting for 11.42% of the study area. This susceptible area is primarily located in lowlands to the East of the watershed, especially in Yang Khram Subdistrict, Santi Suk Subdistrict, and Ban Kat Subdistrict. The validation results using the area under the curve showed that the success and prediction rates were 94.84% and 94.85%, respectively.