Search

It is clear that flash flood events are on the rise, and the detection of flash flood-prone areas has now become a top priority for governments throughout the world. Flash-flood potential mapping can provide very useful information that can be used in order to adopt the most appropriate measures to reduce the risk of these phenomena. A precise model is needed in order to predict flash flood potential, which can help to improve flood management strategies, as it can be used to develop more effective flood management methods. In this present research, Index of Entropy (IOE), Logistic Regression (LR), four machine learning models (, Support Vector Machine (SVM), Classification and Regression Tree (CART), Naive Bayes (NB), k-Nearest Neighbour (k-NN)) and their 5 stacking ensembles were used to construct a new GIS-based framework for flash flood-potential mapping in the Suha River Basin, Romania. For the purposes of running the models, 8 topographical and environmental factors (slope, land use, plan curvature, profile curvature, convergence index, aspect, hydrological soil groups and tpi), along with 111 torrential locations were used to run the five applied models. In order to facilitate training and validation purposes, the entire dataset was divided into 70:30 parts. ROC Curve and statistical measures were employed in order to verify the model reliability and also validate the results. It is noteworthy here that in the training procedure, the stacking ensemble constructed through the combination of all machine learning and statistical learning models (LCkNS) was better than all the other models with a maximum accuracy of 0.972. The models used, revealed that between 10% and 22% of Suha river basin is occupied by high and very high flash-flood potential. As a result, the data generated by this research provides reliable results that can assist policymakers at the local and national levels in developing a concrete strategic plan to reduce flood frequency in their area by utilizing a flood warning system. [ABSTRACT FROM AUTHOR]

It is difficult to predict and model with an accurate model the floods, that are one of the most destructive risks across the earth's surface. The main objective of this research is to show the prediction power of three ensemble algorithms with respect to flood susceptibility estimation. These algorithms are: Iterative Classifier Optimizer – Alternating Decision Tree – Frequency Ratio (ICO-ADT-FR), Iterative Classifier Optimizer – Deep Learning Neural Network – Frequency Ratio (ICO-DLNN-FR) and Iterative Classifier Optimizer – Multilayer Perceptron – Frequency Ratio (ICO-MLP-FR). The first stage of the manuscript consisted of the collection and processing of the geodatabase needed in the present study. The geodatabase comprises a number of 14 flood predictors and 132 known flood locations. The Correlation-based Feature Selection (CFS) method was used in order to assess the prediction capacity of the 14 predictors in terms of flood susceptibility estimation. The training and validation of the three ensemble models constitute the next stage of the scientific workflow. Several statistical metrics and ROC curve method were involved in the evaluation of the model's performance and accuracy. According to ROC curves all the models achieved high performances since their AUC had values above 0.89. ICO-DLNN-FR proved to be the most accurate model (AUC = 0.959). The outcomes of the study can be used to guide future flood risk management and sustainable land-use planning in the designated area. [Display omitted] • New 3 deep learning hybrid models were proposed for flood susceptibility modelling. • 14 flood predictors were used to represents the spatial extension of flood susceptibility. • All the 3-ensemble obtained very high performance in terms of flood susceptibility computation. • ICO-DLNN-FR was the most performant algorithm. [ABSTRACT FROM AUTHOR]

• Flash-flood susceptibility modeling was done using highly accurate models. • A number of 111 flash-flood locations were used for modelling. • A number of 8 flash-flood predictors were used to estimate the susceptibility. • DLNN-AHP was the most performant machine learning model. • Flash-flood susceptibility maps achieved a very good prediction accuracy. The present study was done in order to simulate the flash-flood susceptibility across the Suha river basin in Romania using a number of 3 hybrid models and fuzzy-AHP multicriteria decision-making analysis. It should be noted that flash-flood events are triggered by heavy rainfall in small river catchments. To achieve the proposed results, a total of 8 flash-flood predictors (slope angle, plan curvature, hydrological soil groups, land use, convergence index, profile curvature, topographic position index, aspect) along with a sample of 111 torrential phenomena points were used as input datasets in the next four algorithms: Fuzzy-Analytical Hierarchy Process (FAHP), Deep Learning Neural Network -Analytical Hierarchy Process (DLNN-AHP), Multilayer Perceptron - Analytical Hierarchy Process (MLP-AHP) and Naïve Bayes - Analytical Hierarchy Process (NB-AHP). The Analytical Hierarchy Process was used to calculate the coefficients for each class/category of flash-flood predictors. The torrential points sample was split into training (70%) and validating samples (30%). The modelling was done in Excel, SPSS and R software (H2O package), while the result mapping was performed in ArcGIS 10.5 software. The analysis revealed that the high and very high susceptibility degrees are spread over a maximum of 35.01% of the study area. The best performances, demonstrated by an AUC-ROC of 0.984, are associated with the Deep Learning Neural Network – Analytical Hierarchy Process model, followed by Naïve Bayes – Analytical Hierarchy Process model (AUC = 0.976), Multilayer Perceptron - Analytical Hierarchy Process model (AUC = 0.882) and Fuzzy-Analytical Hierarchy Process (AUC = 0.807). These results indicates that Deep Learning Neural Network is a promising machine learning model which can provide outcomes with very high precision. Also, according to the present research results the deep learning neural network, having many hidden layers, is able outperform the multilayer perceptron that contains a single hidden layer. The main novelty of the present research is the application of the three ensemble models (DLNN-AHP, MLP-AHP and NB-AHP) and also the use of H2O package for the first time in literature, to evaluate the flash-flood susceptibility in small river catchments. [ABSTRACT FROM AUTHOR]