Multivariate Analysis and Machine Learning Approach for Mapping the Variability and Vulnerability of Urban Flooding: The Case of Tangier City, Morocco

Urban flooding is a complex natural hazard, driven by the interaction between several parameters related to urban development in context of climate change, which makes it highly variable space and time challenging predict. In this study, we apply multivariate analysis method (PCA) four machine learning algorithms investigate map variability vulnerability floods city Tangier, northern Morocco. Thirteen that could potentially affect were selected divided into two categories: geo-environmental socio-economic parameters. PCA processing allowed identifying classifying six principal components (PCs), totaling 73% initial information. The scores on PCs spatial distribution allow highlight interconnection topographic properties characteristics (population density building density) as main source flooding, followed relationship drainage (drainage distance channels) properties. All show excellent performance predicting flood (ROC curve > 0.9). Classifications Regression Tree Support Vector Machine models best prediction (ACC = 91.6%). maps highlight, one hand, low lands with high recent buildings, other, higher, steep-sloping areas old buildings population density, very-high vulnerability.