ALGORITMA UNTUK ESTIMASI KEDALAMAN PERAIRAN DANGKAL MENGGUNAKAN DATA LANDSAT-7 ETM+

Abstract

The need for accurate shallow water mapping is highly desireable for various engineering work and navigational safety. The current research attempts to estimate shallow water depth using Landsat-7 ETM+ and to formulate the best combination of Van Hengel and Spitzer algorithm and the best original digital number model. The best digital number model is obtained by correlation analysis, Principal Component Analysis (PCA) and application of regression model to 1st, 2nd and 3rd bands of Landsat-7 ETM+ satellite images. The results show that: (1) Shallow water depth in Pari Island, as a study site, can be estimated from Landsat-7 ETM+; (2) Combination of 321 bands of Land Satellite-7 Enhanched Thematic Mapper Plus (Landsat-7 ETM+ )is the best combination of algorithm of Van Hengel and Spitzer; (3) The best model to estimate shallow water depth from original digital number is exponential equation: y = 21.07e-0.0591x; (4) The best original digital number model has smaller mean error at the depth of less than 9 meter while the Van Hengel and Spitzer algorithm has smaller mean error at the depth of more than 9 meter.