Flood risk assessment in arid and semi-arid regions using Multi-criteria approaches and remote sensing in a data-scarce region
Mohamed Adou Sidi Almouctar, Yiping Wu, Shantao An, Xiaowei Yin, Caiqing Qin, Fubo Zhao, Linjing Qiu
Abstract
Flooding is a natural disaster that poses a threat to both people and the environment, necessitating proactive assessment and mitigation strategies to protect vulnerable communities and ecosystems. These measures are necessary to reduce the risk of flooding and moderate the impact of rainfall. In this study, an Analytical Hierarchy Process (AHP) was used to evaluate flood risk in a data-limited region by integrating Remote Sensing (RS) and Geographic Information System (GIS) methods. The study identified several key flood risk indicators, including topographic wetness index, elevation, slope, land cover, precipitation, distance to river, distance to road, and NDVI. The flood risk map had a score range of 8.71–30.99 %, with higher scores indicating a greater susceptibility to flooding. These scores were then used to classify the flood risk into five categories: very low, low, moderate, high, and very high. The percentages of regions falling into each category were 8.71 %, 23.52 %, 30.99 %, 22.68 %, and 14.09 % respectively. The area under the Curve (AUC) approach was used to validate the flood risk map, which showed a high degree of accuracy (0.86). The results of this study provide valuable insights for monitoring, and forecasting the probability of floods in the Dosso Region. • AHP and RS enhance the assessment of flood risks in vulnerable areas. • The integration of AHP, RS, and GIST in Dosso results in the production of flood risk maps. • The availability of data and model precision have an impact on accuracy in arid regions. • The AUC measure evaluates the performance of the model in classifying high-risk areas.