Band ratio model for remote estimating the water quality parameters in small inland water bodies based on Landsat ETM+ data

Abstract

This work tries to explore how the remote sensing could be used in monitoring of selected water quality (WQ) parameters in small inland water bodies. The respective models to estimate the water quality parameters were proposed based on the Landsat 7 ETM+ images taken in 6 samplings from May 2012 to September 2014. The images used were scenes of WRS-2, path and row 191/25, as well as 190/25, respectively. Samples were taken from 13 water bodies, from which 9 water bodies (20–90 Ha) were used in modelling (and some removed due to clouds and imagery gaps). The WQ parameters were chlorophyll-a (Chl-a), Total Carbon (TC), Total Organic Carbon (TOC), Total Nitrogen (TN), Temperature (T), and Secchi Disk Depth (SDD). The 3 × 3 moving average-window technique with a water-only-mask approach was used in order to limit the process toward the water areas only. The best models based on the surface reflectance (T based on brightness temperature [K]) showed a correlation r2 between 0.78 and 0.90 and NRMSE between 16.6 % and 8.0 %, respectively, for all the water quality indicators. This has proved that all the parameters can be remotely estimated. The models and workflow scheme created are intended to help to institutions mandated in the monitoring of water bodies.