The possibility on estimation of concentration of heavy metals in coastal waters from remote sensing data

Abstract

The heavy metals in waters cannot be decomposed but can be transferred and accumulated with food chains. Many heavy metals are toxic to human beings. It is very important to measure concentration of heavy metals in coastal waters for water quality research, monitoring, and environmental management. On consideration of geochemistry behavior of heavy metals, their distribution is related with water components which determined waters' optical properties. The in-situ remote sensing reflectance data and heavy metal concentration data at 48 sampling points collected from three cruises in the Pearl River estuary were analysed. For single band among all the 57 bands ranging from 365 to 935 nm, the band centered at 711 nm (B <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">711</inf> ) has highest correlation coefficient (R=0.51) with concentration of both Cu and Zn. The band ratio, B <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">711</inf> / B <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">406</inf> has the highest correlation coefficient with Cu (R=0.749), and band ration, B <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">711</inf> / B <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">416</inf> has the highest correlation coeficient with Zn (R=0.804). The band and band ratio were employed for algorithm development using the symbolic regression method, and the results showed the possiblity to retrieve concentration of heavy metal from remotely-sensed data.