An Alternative Model to Estimate Total Suspended Solids Concentrations using Landsat 8 Imagery in Indonesia

Abstract

A regular monitoring program of water quality is generally performed using a direct measurement method, which requires substantial efforts and resources. These issues can be minimised using several options, one of which is Landsat 8 (L8). This imagery has been broadly used to measure several water quality parameters, especially Total Suspended Solids (TSS) concentrations, even in Indonesia. This paper will compare several models from previous studies and a modified model generated using data from various sites. The comparison is based on their competencies to estimate TSS concentrations. The competencies are determined by the coefficient of correlation (r), correlation of determination (R2), and residual standard error (RSE) parameters as these three parameters are strongly correlated, generally applied, and provide distinctive determinations. The best model should have the highest r and R2 values, while the RSE value should be the lowest. The results imply that TSS model 4 generated in this study provides comparable results with TSS model 1, which has been generally used in Indonesia and provided favourable results. Thus, it can be an alternative model to estimate TSS concentrations in Indonesia.