Abstract
Water in coastal and estuary areas needs to be investigated as human activities is allegedly decrease their quality. This has led to optical remote sensing for mapping optical water quality using empirical and semi-analytical approaches. Therefore, this study aims to estimate the Total Suspended Matter (TSM) concentration from Medium Resolution Imaging Spectrometer (MERIS) data using a spectral unmixing approach combined with a radiative transfer model. This approach was implemented in the equatorial tropical coastal water, the Berau estuary, Indonesia, by generating a look-up table (LUT) of top-of-atmosphere radiance spectra through the coupled forward models, and the endmembers were selected from the LUT. The spectral unmixing algorithm was employed to the MERIS data for decomposing the image into haze and sediment components. The data were then transformed into images of a constant haze level corresponding to 50 km visibility, and the atmospheric correction was applied. Furthermore, the TMS concentration was retrieved using the inverse semi-analytical Kubelka-Munk model. The result gave overestimated TSM concentration values on clear waters. However, in turbid waters, a lower RMSE was obtained, and the coefficient of determination was higher than in clear waters.
Highlights
Remote sensing (RS) is an instrument for mapping and monitoring the dynamics of coastal waters at various levels of scale, namely local, regional and global [1]
The LTOA spectrum for all 36 haze and sediment combinations is shown in Figure 3, which indicates a significant increase in the Total Suspended Matter (TSM) concentrations due to a change in the spectral shape of LTOA
The spectral unmixing technique with the semi-analytical radiative transfer model and MODerate spectral resolution atmospheric TRANsmittance (MODTRAN) atmospheric correction provided a robust algorithm for estimating TSM concentrations from satellite imagery in tropical coastal waters
Summary
Remote sensing (RS) is an instrument for mapping and monitoring the dynamics of coastal waters at various levels of scale, namely local, regional and global [1]. The RS has been used to map optical water quality such as concentration of total suspended sediment (TSM), chlorophyll-a, and gelbstoff absorption by using empirical/statistical and semi-analytical approaches [2]. The empirical approach is a simple method that correlates between TSM concentration measured and RS reflectance based on band ratio [3, 4]. This approach is not widely appropriate in coastal water due to gelbstoff and minerals from the freshwater plume, which significantly impacting the optical properties [5]. Spectral unmixing techniques are used to solve the mixed pixel problem [1]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
