Urban Flood Susceptibility in Khardah Municipality Using Machine Learning Technique

Abstract

Waterlogging is an important in the urban areas. When there is heavy rainfall, some areas of the urban set up get blocked with water which brings about several environmental and social problems for the residents of the area. Khardah Municipality is not an exception in this regard. Certain wards of the study area often get waterlogged in and a situation of urban flood arises. The present research aims to prepare an urban flood susceptibility map of Khardah Municipality area through the application of geospatial technique and machine learning process. Analytical Hierarchy process has been used to assign criteria weights to the several conditioning factors. The research reveals that the southern, south eastern and northern parts of the study area are more prone to urban flood than the other areas. The model was found to be excellent with AUC value of 0.814 and damage to roads was turned out to be the most critical problem of the study area. Hence, it can be suggested that the urban flood map will aid in bringing about solution of problems of waterlogging in the study area.