Trails of the killer tsunami: A preliminary assessment using satellite remote sensing technique

• Published in: Current science (Bangalore) Vol. 88; no. 5; pp. 709 - 711
• Main Authors: Ramakrishnan, D., Ghosh, S. K., Raja, V. K. M., Chandran, R. Vinu, Jeyram, A.
• Format: Journal Article
• Language: English
• Published: Current Science Association 10.03.2005
• Subjects: Coastal water; Coasts; Oceans; Radiance; Remote sensing; SCIENTIFIC CORRESPONDENCE; Sea water; Sediments; Tsunamis; Wavelengths; Waves
• ISSN: 0011-3891

The recent tsunami that paralysed the entire southeast and southern parts of India, owes it origin to a devastative earthquake off the west coast of Sumatra (5.51 deg N 92.92 deg E) with a magnitude of 8.9 on the Richter scale and focus5 at a depth of 33.9 km. Here an attempt has been made to understand the trails of the recent tsunami during its propagation and aftermath using satellite remote sensing and GIS techniques. The adopted methodology includes a three-phased approach, viz. 1. Acquisition of MODIS - Terra and Aqua calibrated radiance data sets (level 1B) of 25 December 2004 (07 : 30 GMT) and 26 December 2004 (5: 10, 5 : 20, 8: 15GMT). MODIS band 1 (0.62-0.67 mum) and band 2 (0.841-0.876 mum) (250 m spatial resolution) are used to map coastal flooding. Band 3 (0.459-0.479 mum) and band 4 (0.545-0.565 mum) have spatial resolutions of 500 m. For sediment load study, band 2 was re-sampled to 500 m. As MODIS-Terra data was acquired less than 3 h after the earthquake, special emphasis is made to understand the trails during the propagation and after its impact on the coast. 2. The raster data preparation and image analyses were carried out using ERDASimagine (ver. 8.5) following the standard image processing procedures6. Precise image-to-image registration was achieved between pre- and post-event datasets by way of selecting about 60 ground control points (GCP) in a second order polynomial equation. By this way, the root mean square error was restricted to 0.05. The post-event data is normalized to pre-event data by fitting linear regression equation. Principal component analysis and band ratios are derived to document the moisture changes (on land) and suspended sediment concentration (in ocean). 3. Mapping the tsunami-affected areas using NDVI ratios. Vector layer of the inundated area was generated using ARC/Info ver. 7.1.1. The inundated area statistics was computed from this layer.