Development of novel hybridized models for urban flood susceptibility mapping

Omid Rahmati,Stefanos Stefanidis,Dieu Tien Bui,Hamid Darabi,Aiding Kornejady,Zahra Kalantari,Seyed Amir Naghibi,Farnoush Mohammadi,Zahra Karimidastenaei,Carla Sofia Santos Ferreira,Mahdi Panahi,Ali Torabi Haghighi

doi:10.1038/s41598-020-69703-7

Abstract

Floods in urban environments often result in loss of life and destruction of property, with many negative socio-economic effects. However, the application of most flood prediction models still remains challenging due to data scarcity. This creates a need to develop novel hybridized models based on historical urban flood events, using, e.g., metaheuristic optimization algorithms and wavelet analysis. The hybridized models examined in this study (Wavelet-SVR-Bat and Wavelet-SVR-GWO), designed as intelligent systems, consist of a support vector regression (SVR), integrated with a combination of wavelet transform and metaheuristic optimization algorithms, including the grey wolf optimizer (GWO), and the bat optimizer (Bat). The efficiency of the novel hybridized and standalone SVR models for spatial modeling of urban flood inundation was evaluated using different cutoff-dependent and cutoff-independent evaluation criteria, including area under the receiver operating characteristic curve (AUC), Accuracy (A), Matthews Correlation Coefficient (MCC), Misclassification Rate (MR), and F-score. The results demonstrated that both hybridized models had very high performance (Wavelet-SVR-GWO: AUC = 0.981, A = 0.92, MCC = 0.86, MR = 0.07; Wavelet-SVR-Bat: AUC = 0.972, A = 0.88, MCC = 0.76, MR = 0.11) compared with the standalone SVR (AUC = 0.917, A = 0.85, MCC = 0.7, MR = 0.15). Therefore, these hybridized models are a promising, cost-effective method for spatial modeling of urban flood susceptibility and for providing in-depth insights to guide flood preparedness and emergency response services.

Highlights

Based on the area under the ROC curve (AUC), Wavelet-Support Vector Regression (SVR)-Bat had the highest accuracy in training (AUC = 0.986), followed by Wavelet-SVRGWO (AUC = 0.984) and the standalone SVR (AUC = 0.949)
Since hydrometric stations are not available in urban areas, data scarcity is the main problem in spatial modeling of flood susceptibility
This study presents two new hybridized models, wavelet-SVR-Grey Wolf Optimizer (GWO) and wavelet-SVR-Bat, based on historical urban flood inundation events and using metaheuristic algorithms and wavelet transformation analysis

Summary

Introduction

The accuracy of these maps is affected by (i) nonlinear dynamic characteristics of floods, as a result of distinct factors such as precipitation and human a ctivities[19]; (ii) limitations in data availability, including detailed hydrological and hydraulic data, in developing countries, despite advances provided by new remote sensing techniques, such as satellites, multisensory systems, and/or radar[11]; and (iii) limited applicability of methods at different s cales[20] These drawbacks have encouraged the use of advanced data-driven models, e.g., machine learning (ML), a field of artificial intelligence that uses computer algorithms to analyze and predict information through learning from training d ata[21]. SVMs are suitable for both linear and nonlinear classification and are efficient and reliable tools for producing flood susceptibility maps in a GIS e nvironment[19]

Objectives

Methods

Results

Discussion

Conclusion