Evaluation of Water Depth Extraction Using Empirical Method with Sentinel 2 Image on Google Earth Engine in South Sea Island of Bangka Belitung

The development of remote sensing technology based on cloud computing is an important concern, especially in providing convenience in processing and storing images for various importance. This can be utilized in hydrographic activities in terms of determining depth values using the Satellite Derived Bathymetry (SDB) method. This research focuses on the utilization of Google Earth Engine (GEE) technology in obtaining depth values using the SDB method with Stumpf and Lyzenga algorithms. The area used in this research is on the island of Bangka Belitung which has different characteristics of waters such water depth, water clarity, and water bottom cover. The results of the research conducted were evaluated with the Indonesian National Standard (SNI) 8202 in 2015 regarding the accuracy of the base map. Depth extraction using 2,662 depth sample data. The results of this study show that the Stumpf model provides a better accuracy value compared to the Lyzenga model at 0 - 10 meters. But at 10-25 meters the Lyzenga model produces a better accuracy value than the Stumpf model. The best correlation value was obtained with the Lyzenga model, which is equal to 0.841 while the correlation value of the Stumpf model is 0.753. The best RMSE was obtained at 1.7 m with a depth range of 0 - 5 m with the Stumpf method. The results of the accuracy evaluation of the two empirical bathymetry models pass the requirements of SNI 8202 of 2015 regarding the Accuracy of the Base Map for making Lingkungan Pantai Indonesia (LPI) and Lingkungan Laut Nasional (LLN) maps at a scale of 1: 50,000 with a 20 m contour interval.