Spatial distribution and risk assessment of heavy metal in coastal waters of China.
Journal:
Marine environmental research
Published Date:
Apr 28, 2025
Abstract
In order to better understand the status of heavy metal pollution in surface seawater of China's coastal waters, this paper compiled research results on seven heavy metals (i.e. Hg, As, Cu, Cd, Cr, Pb, Zn) in surface seawater of 35 coastal areas of China's four major sea areas (i.e. Bohai Sea, Yellow Sea, East China Sea, and South China Sea) during 2011-2018, and evaluated the levels of heavy metal pollution and their potential ecological risk by using the methods of classical comprehensive pollution index (CPI) and potential ecological risk coefficient (PERC). The heavy metal pollution levels in the surface seawater were further compared across the different coastal areas. To enhance these evaluations, machine learning approaches, including support vector machines (SVM) and convolutional neural networks (CNN), were also employed to analyze and predict the complex nonlinear relationships between heavy metal concentrations and factors associated with human activities. The results showed that the highest concentrations(μg/L) of Hg, As, and Cd are found in the coastal waters of northern Liaodong Bay; the highest concentrations of Cu, Pb, and Cr are observed in the coastal waters of Tianjin, while the highest concentration of Zn is found in Huanghua Port. Most studied areas were classified as Class I or Class II. The maximum values among the individual pollution indexes were more frequently observed in the Yellow Sea. The areas with higher comprehensive pollution indices were generally due to higher individual heavy metal indices. Potential ecological risk in the coastal areas was mainly influenced by Hg, Cd, Pb, and Cu. Moderate or higher ecological risk was presented in 7 sea areas. The sea areas where the average values of CPI of heavy metals from high to low were Yellow Sea > South China Sea > East China Sea > Bohai Sea, and the average values of comprehensive potential risk factor are South China Sea > Yellow Sea > East China Sea > Bohai Sea. The SVM model exhibited effective performance on analyzing the individual pollution index of heavy metals, while the CNN model was effectively used in analyzing the PERC. The study suggested that machine learning is helpful to assess and predict the status of heavy metal pollution.
