Author ORCID Identifier
https://orcid.org/0009-0006-1139-4027
Biosketch
I am a GIS and Remote Sensing professional with expertise in flood mapping, hydrological modeling, climate change impact assessment, and spatial data analysis. My work focuses on integrating satellite remote sensing, geographic information systems (GIS), and numerical modeling techniques to analyze flood dynamics and support sustainable water resource management.
I have extensive experience in flood inundation mapping using Synthetic Aperture Radar (SAR) datasets such as Sentinel-1 and RISAT imagery, combined with machine learning approaches for accurate flood detection and monitoring. I am proficient in hydrodynamic modeling using HEC-RAS 2D for simulating historical and future flood scenarios under changing climatic conditions. My research experience includes rainfall trend analysis, flood frequency assessment, flood risk zonation, and the integration of climate projection datasets such as CORDEX Regional Climate Models to evaluate future flood risks and adaptation strategies.
My technical expertise includes ArcGIS, QGIS, Google Earth Engine, HEC-RAS, Bentley SewerGEMS (StormCAD), spatial statistics, geospatial modeling, remote sensing analysis, and data visualization. I have worked on projects involving flood hazard assessment, urban growth analysis, population exposure mapping, groundwater studies, and environmental monitoring. I also have experience applying machine learning techniques, multi-criteria decision-making methods, and spatial regression models to address environmental and engineering challenges.
Currently, I contribute to urban water infrastructure projects involving stormwater management, water supply systems, sewerage networks, and drainage planning. My responsibilities include GIS-based analysis, hydraulic and hydrological modeling, technical report preparation, and infrastructure planning support. I am committed to applying geospatial technologies and engineering solutions to improve flood resilience, water resource management, and sustainable urban development.
Date of Award
11-12-2025
Document Type
Thesis
School
School of Civil Engineering
Programme
Ph.D.-Doctoral of Philosophy
First Advisor
Dr.R.Selvakumar
Keywords
Flood Risk Hazard Susceptibility, Climate driven, Spatial Data, HEC-RAS Modelling, Rainfall induced flooding
Abstract
Background: Tamil Nadu’s coastal corridor (Cuddalore–Sirkazhi; ~2,740 km²) experiences recurring floods driven by climate change, rapid urban expansion, and inadequate drainage across low-lying deltaic terrain. Existing flood assessments remain largely reactive and rely on static methods that fail to incorporate future climate projections, demographic growth, and evolving urban patterns.
Methods: This research develops a dynamic, climate-resilient flood risk framework integrating (i) long-term hydroclimatic trend analysis (1984–2023), (ii) satellite-based flood hazard mapping using multi-temporal Sentinel-1A and RISAT-1A SAR, (iii) CORDEX REMO2015 (RCP 4.5)–driven HEC-RAS 2D hydrodynamic modelling, (iv) machine-learning-based susceptibility modelling (RF, DT, SVM), and (v) scenario-based evaluation of structural mitigation using DEM-altered configurations.
Results:
Rainfall Trends: Annual and NEM rainfall increased by ~11 mm/year, with wet-day frequency rising and coastal flood-trigger thresholds as low as 50–60 mm/day (inland: 70–100 mm/day).
Flood Hazard Mapping: Integrated Flood Analysis (frequency + duration) achieved ~96% accuracy, identifying persistent inundation around Perumal Lake, Chidambaram, and the Kollidam basin.
Hydrodynamic Modelling: HEC-RAS 2D achieved 96% inundation accuracy and 90% depth accuracy (validated with SAR & TN-SMART). Simulations showed non-linear behaviour, with rainfall >400 mm/day causing rapid lateral floodplain expansion due to drainage saturation.
Flood Risk Mapping: Random Forest performed best (79% accuracy, AUC = 0.85). SHAP identified elevation, flood depth, and drainage density as dominant predictors. Ensemble classification and Spatial Agreement Index located high-confidence hotspots in Parangipettai, Keerapalayam, and Melbhuvanagiri.
Mitigation Evaluation: Channel deepening and new drainage pathways reduced flooded area by 33% (448 → 300.5 km²), reduced maximum depths by ~0.6 m, and lowered Very-High-risk villages from 126 (~260,000 people) to 76 (~105,000 people).
Novelty & Contributions:
(1) First Indian study to integrate CORDEX REMO2015 projections with high-resolution HEC-RAS 2D for future flood simulation.
(2) Introduces a dynamic flood risk framework linking climate projections, hydrodynamics, and projected 2030 socio-demographics.
(3) Provides comparative benchmarking of ML classifiers for flood risk (RF, DT, SVM).
(4) Demonstrates pre-implementation evaluation of structural mitigation under future climate extremes.
(5) Offers a scalable methodology applicable to global deltas (Ganges–Brahmaputra, Mekong, Nile).
Limitations & Future Work: Constraints include the absence of LiDAR-grade DEMs, uncertainties in CORDEX projections, coarse socio-economic datasets, and computational limitations. Future efforts should incorporate LiDAR/UAV-based terrain data, AI-driven realtime decision support, multi-hazard modelling (cyclones, surges, sea-level rise), and high-resolution vulnerability metrics.
Conclusion: This thesis advances flood risk science by integrating climate projections, hydrodynamic modelling, socio-economic forecasting, and machine learning into a unified, adaptive framework. The results offer a robust decision-support tool for climate-resilient planning in coastal Tamil Nadu and contribute to global strategies for managing flood risk in deltaic systems under accelerating climate change.
Recommended Citation
DR, Sakthi Kiran Mr, "Climate-Driven Flood Risk Mapping and Adaptive Strategy Modelling for Coastal Tamil Nadu" (2025). Theses and Dissertations. 192.
https://knowledgeconnect.sastra.edu/theses/192