SHALLOW WATER DEPTH MAPPING USING SINGLE BAND AND BAND RATIO ON HIGH-RESOLUTION IMAGERY

Abstract

One of the availabilities of remote sensing satellite imagery can be used as a provider of shallow sea depth information using the Satellite-Derived Bathymetry (SDB) technique. This technique's main problem is the variation in the bottom cover of waters such as coral reefs and seagrass, which distorts the spectral values. The use of band ratios can normalize variations in bottom water cover. This study compares the single band algorithm's accuracy with the band ratio depth data obtained by field survey around the port of Karimunjawa Islands, Central Java. The image used in this study is high-resolution imagery, Worldview 3. Preprocessing includes Sunglint correction to reduce the effect of sunglint in the waters and correction of depth data so that the data are free from tides' influence. The bands used are red, green, blue, and Near-Infrared, which results in 10 combinations. This study indicates that the band ratio method produces a smaller RMSE value than the single band. The blue/green ratio makes the best depth values with an RMSE of 1.669 meters at a depth of 0-5 meters. In comparison, single-band use shows that the best estimation result is with an RMSE of 2.373 meters in the green band. This study shows that the band ratio method produces better depth estimates than the single band method.