The benthic habitats of shallow waters of Sebaru Island and Lancang Island have different water characteristics from geographical location. Data and information about benthic habitat are needed to maintain and preserve ecosystems in the waters. This study aims is to know the effect of different satellite image resolution, different algorithms and water quality e.g chlorophyll-a (Chl-a) and total suspended solid (TSS) on the accuracy of shallow-water benthic habitats mapping on Sebaru Besar Island and Lancang Island. The accuracy (OA) of the application of different classification algorithms showed a good results. The highest OA results in shallow waters of Sebaru Island with Wordview 2 imagery were obtained from the SVM and DT algorithms with the same value of 76.24%, while the Sentinel 2B image with the DT algorithm obtained results (OA) of 68.08%. In Lancang Island the highest OA value of Wordview 2 imagery was obtained by DT algorithm with a value of 74.44%, while Sentinel 2B imagery was obtained from KNN algorithm with a value of 59.0%. High concentrations of Chl-a and TSS cannot yet be said to affect the low accuracy in mapping shallow water benthic habitats.
Copyrights © 2020