This study tends to apply an appropriate range of flood models among diverse classes of flood modeling methods namely bivariate models including FR and machine learning algorithms including RF and LMT to flood suscep- tibility modeling of Ningdu catchment in China. For vali- dation and comparison, several quantitative methods, such as receiver operating characteristic (ROC), kappa, root mean square error (RMSE), and mean absolute error (MAE) were selected. Computations were made using WEKA 2.7.11 and ArcGIS 10.2 software.Result of this study can be useful for decision makers who are responsible for future flood mitigation of the Ningdu catchment in China.
