Enhancing Flood Area Mapping Accuracy Using Advanced SAR Data Processing
S. Pazhanivelan *
Centre for Water and Geospatial Studies, Tamil Nadu Agricultural University, Coimbatore, India.
K. P. Ragunath
Centre for Water and Geospatial Studies, Tamil Nadu Agricultural University, Coimbatore, India.
N.S. Sudarmanian
Centre for Water and Geospatial Studies, Tamil Nadu Agricultural University, Coimbatore, India.
S. Satheesh
Department of Remote Sensing and GIS, Tamil Nadu Agricultural University, Coimbatore, India.
K. Sneka
Centre for Water and Geospatial Studies, Tamil Nadu Agricultural University, Coimbatore, India.
*Author to whom correspondence should be addressed.
Abstract
Aim: To assess the spatial distribution of floods in 2024 using remote sensing data, specifically Synthetic Aperture Radar (SAR), a powerful tool in flood monitoring and mapping due to its ability to capture data under all weather conditions, including rain and cloud cover provides high-resolution imagery suitable for identifying and analyzing flood extents.
Study Area and Duration: North-Eastern districts of Tamil Nadu viz., Tiruvannamalai, Ranipet, Chengalpat, Kancheepuram, Tiruvallur, Viluppuram, Cuddalore, Nagapattinam, Thanjavur, Tiruvarur, Kallakurichi and Mayiladuthurai.
Methodology: Flood mapping uses imagery collected from the European Space Agency's (ESA) Sentinel-1A satellite to identify and map flooded regions. Flood mapping receives assistance from this satellite's C-band SAR sensor, which can capture images in any weather condition without affecting the data.
Results: The flood vulnerability assessment using Sentinel-1A satellite data has provided critical insights into the extent and impact of flooding across Tamil Nadu in 2024. With a total of 90,369 hectares of agricultural land affected, the study highlights the urgency of implementing targeted flood management strategies. 350 ground truth points were collected, out of which 309 points coincided with the flood-affected areas. Among these 309 points, 214 were flood points and 95 were non-flood points. The overall accuracy of the results was 90.00 per cent. The producer and user accuracy for flood-affected areas was 92.10 per cent and 93.40 per cent, respectively. The producer and user accuracy for non-flood areas. was 85.30 per cent and 82.70 per cent with Kappa index of 0.80.
Conclusion: These findings underscore the importance of integrating advanced remote sensing technologies with ground-level data to better understand flood dynamics and provides a foundation for sustainable disaster risk management and resource allocation, ensuring long-term agricultural and environmental security in Tamil Nadu.
Keywords: Flood affected areas, rainfall, sentinel 1A satellite data and SAR data