Developing a Decision Tree Data Mining Method for detecting the Effective Parameters for Determining the Power of Flood Destruction

Floods, as one of the natural disasters, cause irreparable damages to urban infrastructures, agricultural lands, and natural resources. Therefore, access to comprehensive information on effecting factors the extent of flood damage can be useful in estimating the extent of damage. Therefore, the aim of this study was to create a database of Effective parameters on flood destruction power using case study of Landsat-7 satellite images with ETM+ sensor and ASTER DEM data using decision tree data mining method. In this study, environmental parameters such as canopy, natural slope, and slope direction were considered in order to evaluate flood degradation power in the study area and the decision tree model was created using these criteria. Finally, based on these parameters, the number of changed pixels (after the flood) in the study area is 692361 which indicates 62312.49 hectares of degraded land in the study area. According to the findings of the present study, lands with low canopy characteristics, namely normalized differential vegetation index (NDVI) between 0.2 to 0.4, low slope 0 to 45 degrees and Southern slope direction caused the most damage caused by floods. Also, areas with dense NDVI, high slope, and northern slope orientation have preventative influence on floods-caused-damages. Finally, it can be concluded that the decision tree, as data mining method is capable of yeilding better accuracy and quality in determining the effective parameters in estimating flood destruction power by increasing the input variables.