GIS and Machine Learning Approaches in Flood Hazard Mapping: A Case Study of Lower Niger River Basin
Publication Date : 31/10/2024
Author(s) :
Volume/Issue :
Abstract :
Flooding is a recurrent and destructive natural disaster intensified by elements such as extreme rainfall, urbanization, climate change, topography, and human activities. This study primarily aims to integrate Geographic Information System (GIS) and Machine Learning (ML)techniques in flood hazard mapping in the lower Niger River basin in Nigeria .Twenty flood influencing factors including elevation, slope, aspect, flow direction, flow accumulation, drainage density, distance from river, plan curvature, profile curvature, roughness, topographic wetness index (TWI), stream power index (SPI), sediment transport index (STI), normalized difference 15 vegetation index(NDVI), normalized difference moisture index (NDMI), land use/land cover LULC), soil, geology, temperature, and rainfall, were considered and analyzed within the GIS framework. The Extreme Gradient Boosting(XGBoost) model was applied to generate the flood hazard zones within the study area. Based on historical flood events within the study area, 1164 flooded and non-flooded points were identified and utilized to train and test the model. The ML model achieved high accuracy of 0.905(90.5%), and an ROC-AUC score of 0.88. The generated flood susceptibility map indicated that 4.67%, 4.98%, 10.31%, 11.13%, and 68.91% of the basin are respectively at very high, high, moderate, low, and very low risk of flooding. The successful integration of GIS with machine learning validates the potential to improve flood hazard prediction and mitigation efforts in the Niger River basin and other similar flooding environments in Nigeria. Keywords: Flood Hazard Mapping, Geographic Information System, Machine Learning, XGBoost, Niger River Basin
No. of Downloads :
0