Bathymetry mapping using PlanetScope imagery on Kemujan Island, Karimunjawa, Indonesia

Abstract

Bathymetrical information plays an important role in the management of coastal resources such as fisheries and shipping lanes, and is basic information of coastal resources besides benthic habitat. The extraction of satellite data for bathymetry mapping or SDB (Satellite-Derived Bathymetry) has emerged as an effective method to obtain information in terms of cost and time. Kemujan Island is an island with unique geomorphology and thus selected as a study area for bathymetry mapping. The method applied single band and band ratio methods on PlanetScope imagery. PlanetScope was chosen because it has high spatial and temporal resolution. The image used was acquired on June 30, 2017. The results show that linear regression is more representative than SVR-based modeling, with the best single band modeling from the (ln) Green band and the best band ratio modeling from the ratio transformation of Blue (ln)/(ln) Green. Single band had higher correlation and accuracy than band ratio (0.89 and 0.66, respectively) and RMSE value of 0.78 and 0.96, respectively. However, based on bathymetry profiling, the results of band ratio modeling are more appropriate as the single band method cannot afford to minimize the effect of various bottom types.