ZONASI GEOMORFOLOGI DAN KOREKSI KOLOM AIR UNTUK PEMETAAN SUBSTRAT DASAR MENGGUNAKAN CITRA QUICKBIRD

Bias may occur on attenuation coefficient ratio estimated from water column correction method. This bias then contribute to thematic accuracy of bottom substrate images. This study used geomorphologic spatial zonation to improve thematic accuracy of bottom substrate maps that produced from water column correction method. Quickbird pixel values were converted to the top of atmosphere radiance and followed by water column correction to make bottom substrate map with three themes i.e. sand, seagrass and coral reef. Field data were grouped using Bray Curtis method and become basis of image reclassification. Geomorphological profile was extracted from green and red composite images, refer to a bathymetric survey. This combined method was significantly reach the thematic accuracy up to more than 80%.