A comparative assessment of flood susceptibility modelling of GIS-based TOPSIS, VIKOR, and EDAS techniques in the Sub-Himalayan foothills region of Eastern India

Abstract

In the Sub-Himalayan foothills region of eastern India, floods are considered the most powerful annually occurring natural disaster, which causes severe losses to the socio-economic life of the inhabitants. The present study tested three comprehensive and systematic MCDM techniques such as Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), Vise Kriterijumska Optimizacijaik Ompromisno Resenje (VIKOR), and Evaluation Based on Distance from Average Solution (EDAS) in Koch Bihar district for comparative assessment of the flood susceptible zones.The 21 indicators, i.e., elevation, slope, drainage density, geomorphology, rainfall deviation, aspect, curvature, lineament density, lithology, long-term annual rainfall, land use land cover, modified fournier index, modified normalized difference water index, normalized difference vegetation index, roughness, soil, sediment transport index, stream power index, topographic positioning index, terrain ruggedness index, and topographic wetness index were considered as the essential flood conditioning parameters. The multicollinearity statistics were employed to erase the issues regarding highly correlated parameters (i.e., MFI and long-term annual rainfall). The Spearman's rank (𝑟ₛ) tests among the three models revealed that TOPSIS-EDAS persisted in a high correlation (𝑟ₛ = 0.714) in contrast to the relationships between VIKOR-EDAS (𝑟ₛ = 0.651) and TOPSIS-VIKOR (𝑟ₛ = 0.639). The model's efficiency was statistically judged by applying the Receiver Operating Characteristic-Area Under the Curve (ROC-AUC), Mean Absolute Error (MAE), Mean Square Error (MSE), and Root Mean Square Error (RMSE) techniques to recognize the better-suited models for mapping the flood susceptibility. The performance of all techniques is found good enough (ROC-AUC = >0.700 and MAE, MSE and RMSE = <0.300). However, TOPSIS and VIKOR have manifested an excellent outcome and are highly recommended for identifying flood susceptibility in such active flood-prone areas.

GIS-based data-driven bivariate statistical models for landslide susceptibility prediction in Upper Tista Basin, India

“…Concave, flat, and convex are the three elements that are represented by curvature. Positive values are related to convexity, whereas negative values are associated with concavity, and zero is represented by the flat region [ 72 ]. Curvature values that are both positive and negative implications that the terrain is highly susceptible to landslides.…”

“…This study calculated the LCF roughness based on Eq. (1) (Evans, 1972; [ 72 ]. where, represent focal statistics mean, represent focal statistics minimum and represent the focal statistics maximum.…”

“…The SPI illustrates the surface water flow's erosional power, and it is a significant hydraulic factor for modelling LSMs [ 64 , 72 , 75 ]. SPI also describes how susceptible a terrain is to runoff-induced erosion and its indirect impact on slope stability.…”

Flood Susceptibility Modelling by Advanced Convolutional Neural Networks (CNN) in the foothills of Southern Western Ghats, Kerala, India

“…2020). MCDM method with fuzzy AHP approach (Tella, et al 2020), and other MCDM methods like Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), VIekriterijumsko KOmpromisno Rangiranje (VIKOR), a Serbian term for "multi-criteria optimization and compromise solution," Enhanced Data Authentication System (EDAS) are also used (Mitra, et al 2023) in ood susceptibility studies. A ood susceptibility model with a small number of parameters in the data-scarce area of northwest Bangladesh was carried out using the AHP method (Sarkar, S. K., et al 2022).…”

“…Tree-based algorithms are extensively employed for ood susceptibility modelling (Chen, Wei et al 2020). Random Forest (Lee, Sunmin, et al 2017) and Decision Tree (Khosravi, Khabat, et al 2018) hybrid models for ood susceptibility are also popularly used because hybrid approaches can harness the characteristics of different methods to generate more robust and accurate ood susceptibility models by integrating several Machine Learning techniques (Mitra, et al 2023). Recently ensemble models gained popularity because they integrate the predictions of numerous base models to provide a more accurate and robust prediction (Al-Abadi and Alaa M. 2018).…”