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

Abstract

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

Mapping Gully Erosion Variability and Susceptibility Using Remote Sensing, Multivariate Statistical Analysis, and Machine Learning in South Mato Grosso, Brazil

“…To ensure a proper evaluation of the modeling performance of the four machine learning models, we used four types of classification results provided by the confusion matrix, namely, accuracy (ACC; Equation (2)), sensitivity (SST; Equation (3)), specificity (SPF; Equation (4)), and precision (PRC; Equation ( 4)) [11,62]. In general, the higher the ACC, SST, SPF, and PRC values, the better the performance of the models.…”

“…where TP, TN, FP, and FN are true positive, true negative, false positive, and false negative, respectively. Model evaluation was also performed using the Receiver Operating Characteristic curve (ROC) statistic, which is a common criterion for evaluating spatial modeling performance [11]. The ROC curve value represents the probability that a test point is accurately differentiated from a random point in the predetermined context of the study area.…”

“…In the past decade, studies aiming at creating erosion susceptibility maps through statistical, machine learning, data mining, or multi-criteria decision analysis methods have multiplied. Machine learning methods such as multivariate discriminant analysis (MDA), random forest (RF), logistic regression (LR), support vector machine (SVM), and others have recently been implemented, for example, to understand the parameters controlling the development of drainage network types, urban flooding, landslides, soil subsidence, snow avalanches, and gully erosion [10][11][12][13][14][15][16][17][18][19][20]. These methods have demonstrated their superiority over traditional statistical-based techniques by having the ability to model highly dimensional and nonlinear data sets, allowing complex environmental interactions to be evaluated [9,21,22].…”

“…CART has been widely used in many studies of natural hazards such as landslides, subsidence, urban flooding, etc. [11,19,20,52]. Here, it is be applied to the prediction of gully development following a four-step procedure: (1) building the tree, (2) stopping the building of the tree, (3) pruning the tree, and (4) selecting the optimal tree for classifying gully or non-gully classes [19,54].…”

Flood risk management in New Zealand: A case study of the Northland urban community

“…A major cause of urban flooding is by poorly managed urbanisation [21], increasing areas of impermeable surfaces [22], poor flood management strategies [23], lack of flood early warning systems [24], and disposal of solid waste in drainage lines [25]. Whilst urban flood issues are prevalent in almost every megacity globally, it also happens in small towns built on floodplain like in many cases within New Zealand.…”