Treffer: Application of ensemble fuzzy weights of evidence-support vector machine (Fuzzy WofE-SVM) for urban flood modeling and coupled risk (CR) index for ward prioritization in NCT Delhi, India.

NCT Delhi, the heart of India, remains vulnerable to urban flooding from July to October, when southwest monsoon is active over its area of 1483 km2. To address the paucity of a comprehensive susceptibility map, this study employs urban flood modeling to quantify the spatial sensitivity of NCT Delhi to water logging. Fifteen flood related variables, including elevation, slope, slope aspect, profile curvature, plan curvature, Topographic Wetness Index (TWI), Stream Power Index (SPI), Topographic Roughness Index (TRI), geology, soil type, land use/land cover (LU/LC), Modified Fournier Index (MFI), water level depth, distance from storm drain, and distance from the Yamuna were analyzed. The study applies the data-driven form of ensemble fuzzy weights of evidence-support vector machine (Fuzzy WofE-SVM). The first-level flood susceptibility map was generated using Fuzzy WofE technique. Using this map and secondary data on water logging locations identified by the Delhi Traffic Police, flood inventory datasets were created. To execute the ensemble methodology, SVM was sequentially integrated with the Fuzzy WofE technique. Urban flood susceptibility zonation maps of NCT Delhi were created using four SVM kernel functions: linear (LN), polynomial (PL), radial basis function (RBF), and sigmoid (SIG). The kernel parameters i.e., regularization parameter (C), kernel width (γ) and degree (d) were optimized using python-mediated k-fold cross validation method. Area under the receiver operating characteristic curve (AUROC) analysis validated that all the SVM kernels yield excellent success and prediction rates. However, in terms of success rate, RBF (AUROC = 0.976) outperforms PL (AUROC = 0.968), LN (AUROC = 0.966), and SIG (AUROC = 0.956). With an AUROC of 0.966, RBF again outperforms SIG (AUROC = 0.964), LN (AUROC = 0.955), and PL (AUROC = 0.952) in terms of predictive performance. The novel Coupled Risk (CR) Index has been developed and presented in this study to increase the applicability of flood modeling by translating macro-hazard perspectives into a polished vulnerability scenario of administrative convenience. Spatial analysis of hazard intensity and endangered population using this novel tool would benefit take specific actions for disaster management by prioritizing wards for mitigation strategy planning and implementation. [ABSTRACT FROM AUTHOR]

Copyright of Environment, Development & Sustainability is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)