Quantitative remote sensing methods for real-time monitoring of inland waters quality

TLDR

Spaceborne remote sensing of inland water quality assumes a priori knowledge of the relationship between reflectance and concentrations of key constituents. The study aims to develop and validate empirical relationships linking spectral reflectance to chlorophyll‑a, suspended matter, and dissolved organic matter concentrations. Measurements of upwelling/downwelling irradiance, chlorophyll‑a, suspended matter, and dissolved organic matter were collected at over 20 water bodies across diverse trophic states and climates using a single high‑resolution radiometer, and statistical modelling was applied to derive and test reflectance‑based estimation functions. The derived estimation functions achieve maximum errors of 3 mg m⁻³ for chlorophyll‑a, 4 mg L⁻¹ for suspended matter, and 0.065 mg C m⁻³ for dissolved organic matter, demonstrating a viable approach for space‑based eutrophication monitoring.

Abstract

Abstract Spaceborne remote sensing of inland water quality is based on the assumption that the relationship between the reflectance and the concentration of relevant water quality constituents is known a priori. Simultaneous measurements of the upwelling and downwelling irradiances, along with phytoplankton chlorophyll-a, suspended matter and dissolved organic matter concentration at over 20 water bodies throughout former USSR, Hungary, Germany and Bulgaria, are reported in this article. The measurements cover different trophic states of water bodies, from oligotrophic to hypertrophic, and different climatic conditions. The range of chlorophyll-a is 0·1 to 350 mgm-3, suspended matter is 0·1 to 66mg1-1, and dissolved organic matter absorption at the wavelength 380nm is 0·1 to 12m-1. All radiometric measurements are performed with a single radiometer in the 400-750 nm range with spectral resolution better than 1 nm. Factor and signature analysis, as well as multi-spectral statistical modelling of water quality parameters versus simulated spectral band ratios, makes possible the determination of appropriate functions of reflectance for estimating phytoplankton chlorophyll-a, suspended matter, and dissolved organic matter concentrations. One set of measurements is used to develop empirical relationships among the spectral reflectance and the above-mentioned water quality parameters. The other sets of measurements are used to test these relationships. The maximum error of estimation are: 3mgm-3 for chlorophyll-a, 4mgl-1 for suspended matter and 0·065mgCm-3 (mg of carbon per m3) for dissolved organic matter, The results are used to develop an appropriate methodology for monitoring of eutrophication process in inland waters and to test concepts of inland water quality monitoring from space