Flood Susceptibility Mapping in District Jhang Using Remote Sensing and Advanced Machine Learning Techniques

Floods are often considered as a significant natural disaster, primarily due to their devastating effects that result in considerable socio-economic losses. The mapping and assessment of Flood Susceptibility Mapping (FSM) plays a key role in flood prevention and mitigation strategies. Flooding is a natural disaster which is critical issue for many regions, including District Jhang of Pakistan, which faces significant monsoon-related flood risks. In this study, 13 influencing factors of flood occurrence have been assessed by integrating the power of cloud-based computing through Google Earth Engine (GEE) with Geographic Information System (GIS) and programming languages R and Python. For this purpose, Sentinel-2 data were utilized, along with different Machine Learning (ML) and Deep Learning (DL) models. Based on the results from different models, 27.20% to 39% of the study area were projected to be very highly susceptible to flooding. The developed tools and methods in the present research, particularly the FSM visualization app, can be utilized to help decision-makers prepare and act upon the onslaught of a flood event.