Characteristics of Tsunami-Affected Areas in Moderate-Resolution Satellite Images

Abstract

The massive 2004 Indian Ocean tsunami caused vast devastation along the coastal areas in countries around the Indian Ocean rim. Satellite images of various spatial resolutions could quickly capture the affected areas and were used for emergency response after the catastrophe occurred. To figure out the extent of affected areas, moderate-resolution satellites (e.g., Terra-ASTER) images are more suitable than high-resolution satellites (e.g., Ikonos) images. Basically, tsunami-affected areas can be observed and detected through land cover changes. Based on the nature of the tsunami attack, we chose the normalized difference vegetation index, soil index, and water index as indicators to help detect changes. This paper first investigates the fluctuations of these indexes and their differences using ASTER images of southern Thailand. The investigation is carried out in two cases: one using only the data acquired after the tsunami, and the other using both data acquired before and after the tsunami. Consequently, the thresholds of index differences are set up for the detection of tsunami-affected areas. In addition, since landform is a significant factor to determine the extent of tsunami runup, Shuttle Radar Topography Mission data are employed to perform geomorphological classification and to assess its relationship with the tsunami-affected areas