Litcius/Paper detail

Flood susceptibility analysis to sustainable development using MCDA and support vector machine models by GIS in the selected area of the Teesta River floodplain, Bangladesh

Shapla Akhter, Md. Mostafizur Rahman, Md. Moniruzzaman Monir

2024HydroResearch18 citationsDOIOpen Access PDF

Abstract

Managing flood risks to ensure sustainable development, this research analyses the flood susceptibility in the selected area of the Teesta River floodplain in Bangladesh using the MCDA model with the SVM model. The different geological and climatic flood vulnerability factors used in this study were collected from USGS, BMD, BARC, BWDB, and BBS. This study reveals that the channel pattern changed, drainage density decreased by 42.28 %, and TWI became high during the study period (2000−2020). This study shows that the very high, high, and medium flood-susceptible zones increased by 5.66 %, 2.7 %, and 7.74 %, and the low-risk zone decreased by 10.79 % during the study period. The drainage system was found to be a significant flood conditioning factor, and the waterbody and river area decreased by 5.31 %. From the flood inventory prediction rate curve analysis 70 % of validation points agreed with field data. According to AUC, the average success rate is 91.51 %. • The MCDA model with RBF kernel from the SVM model was utilized. • The drainage density decreased by 42.28 % from 2000 to 2020. • The high flood-susceptible zones increased by 8.36 % and the low-risk zone decreased by 10.79 %. • The drainage system was found a significant flood-conditioning factor.

Topics & Concepts

FloodplainFlood mythSustainable developmentWater resource managementSupport vector machineMultiple-criteria decision analysisHydrology (agriculture)Decision support systemGeographic information systemEnvironmental scienceGeographyGeologyRemote sensingEngineeringComputer scienceOperations researchCartographyData miningArchaeologyArtificial intelligenceEcologyGeotechnical engineeringBiologyFlood Risk Assessment and ManagementHydrology and Watershed Management StudiesHydrological Forecasting Using AI