Using Artificial Intelligence and Machine Learning to predict Flood Susceptibility in the Kikou Watershed in the Beni Mellal region (Morocco)

¹ Natural Resources and Durable Development Laboratory, Science and Technical Faculty, Ibn Tofail University, Kenitra, Morocco
² Laboratory of Applied Geology and Geo-Environment, Faculty of Sciences, Ibn Zohr University, Agadir, Morocco
³ Hydraulic system Analysis Team, Mohammadia Engineering School, Mohammed V University, Rabat, Morocco

^* Hamza LEGSABI: This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

Flood susceptibility prediction is a complex subject due to the interactions of multiple factors related to hydrology, meteorology, urbanization and finally climate change. This paper addresses these complexities by investigating flood susceptibility in watershed situated in the urbanistic area of Beni Mellal city. Prone to extreme weather events, the flood susceptibility is evaluated using three popular Machine Learning (ML) techniques: Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Networks (ANN). The performance of these models is evaluated by the ML performance evaluation metrics (accuracy, recall, and f1 score). The SVM is the best fitting algorithm with a performance of 100% in all evaluation metrics (accuracy, recall, and f1 score). The RF algorithm is also well suited to the study area. It has a performance of 87% in accuracy, 100% in recall and 90% in f1 score. On the other hand, ANN algorithm performed poorly with only 25% in accuracy and 35% in recall.