Comparative Analysis of Flood Conditioning Factors and Model Transferability Across Tetouan, Tata Regions, Morocco

This is the authors' abstract. We don't add key points for this paper.

Climate change has caused a severe natural disaster; flooding is one of the famous consequences. Therefore, the scientific community focuses on controlling proactively these changes. This work focuses mainly on flood prediction through a deep understanding of the factors leading to it. As a result, 11 factors were extracted from various data sources from two zones in Morocco: Tata and Tetouan. The primary aim of this comparison is to identify the principal clusters impacting flood occurrence, whether the dominance lies with hydrological factors or with topography and land cover. The application of Random Forest (RF) on Tetouan and Tata on datasets of 1000 and 708 samples respectively achieved good accuracy of 97% and 96.24%. The results of this study are relevant for helping planners and authorities reduce flood-related damages.