BOTTOM TYPES IDENTIFICATION IN SHALLOW CORAL REEF ECOSYSTEMS USING IMAGERY SATELLITE DATA

Abstract

Satellite data provide information about spectral signatures of objects in detail, based on the wide range of spectral wavelengths. Bottom types in a coral reef Ecosystems are diverse and each object has a different spectral signature. The aim of this research is to define bottom types using Multispectral and Hyperspectral imagery satellite data. Six processes were applied to Hyperspectral Images to identified bottom types using modification of Analytical Imaging and Geophysics LLC (AIG) hyperspectral analysis. The multispectral analysis was focused on correcting water column noise by applying the radiative water column algorithm (Lyzenga, 1978, 1981) and the modified image correction algorithm (Lyzenga et al., 2006). The results showed that multispectral image analysis was able to identify a fine complexity of b bottom types classes with 68.57% overall accuracy. In contrast, Hyperion image identified a coarse complexity of bottom types classes with 61.57% overall accuracy. This low result was caused by low spatial resolution which created a mixing pixel around image of thin and narrow shallow coral reef ecosystem. Spatial resolution, atmosphere and water scattering played an important role in bottom types identification.