Abstract

Estimating the water transparency (SD) is one of the traditional applications of remote sensing for water environmental studies. Monitoring the change of SD in space and time help quantitative-ly assess the river water environmental state and better understand pollutant sources and transporta-tion by the river. This study aims to develop a model to estimate SD using Sentinel 2A (S2A) satel-lite image basing on in-situ data of water reflectance and SD measured concurrently at the Red River section in Lao Cai city on four days in both wet and dry seasons. The result showed that SD corre-lated highly to the spectral ratio corresponding to the ratio of the NIR band (band 5) versus the red band (band 4) of S2A image and can be accurately calculated by an exponential function of this ratio (R2 = 0.89, RMSE = 0.12). Resultant maps of SD estimated from S2A images acquired in 2018 show clearly two trends of SD in the river section water in space and time: a) SD of the river water in the wet season is lower than that in the dry season; b) SD of the Red River water is lower than SD of the river tributaries’ water, i.e.: Nam Thi River and Ngoi Dum stream. Result and methods presented in this study need be advanced by adding auxiliary data from the whole river system to provide critical information for the river management, particularly transboundary pollution control task.