PROPERTIES OF TROPICAL COASTAL WATERS FROM MERIS DATA AROUND THE BERAU ESTUARY, INDONESIA

The MERIS sensor on ENVISAT satellite has given a great prospect to understand the dynamic of water quality with high temporal and spatial resolutions. The objectives of this study were to derived coastal water quality from MERIS Level 1 imageries using Artificial Neural Networks (ANN) and then validated with in situ observation. The study was done in Berau Estuary, East Kalimantan, Indonesia, where its situated in a specific interest for studying the complex system and dynamic of tropical coastal environment. The MERIS image processed to water quality product by applying the ANN algorithm (Case-2 Regional, C2R, and Freie Universität Berlin, FUB) for retrieving remote sensing reflectance and waters constituents. Next, inter-comparison among the results of MERIS L1 and MERIS L2 with in-situ observation was done. The result of the study indicate that remote sensing reflectance from C2R algorithm has low RMSE with the in-situ data compared to the remote sensing reflectance resulted from FUB and MERIS L2. The channel of 490 nm, 510 and 560 nm is the most correlated channel between MERIS data set and in-situ data set. The research shows also that the irregular structure of water constituents along the Berau Estuary can be accessed significantly by using MERIS full resolution data.