The response patterns of riverbank to the components carried by different pollution sources in the river: Experiments and models

Abstract

Intensive agriculture/industrial activities as well as natural processes result in pollution in river water and groundwater. However, how riverbank groundwater responds to the components carried by different pollution sources in the river remains unclear. In this study, the surface water and riparian groundwater quality at the Laonanguan (LNG) and Hongqierzha (HQEZ) stations of the Gusong River, China, were jointly investigated via hydrochemical monitoring and water quality analyses, and the main pollution sources, i.e., anthropogenic activities and mineral weathering, were identified by principal component analysis (PCA). The absolute principal component score-multiple linear regression (APCS-MLR) revealed that anthropogenic inputs only accounted for 11.4 % of the total pollution sources for riparian groundwater, which was much lower than that for river water (24.8 %). The riverbank hence acted as the primary sink for the components (e.g., NH4+-N and TP) from non-point pollution and could degrade 13.7 % of the pollutants from the river water caused by anthropogenic input. Conversely, riverbank played a source role for the species (e.g., SO42- and Cl-) from mineral weathering and contributed additional 19.3 % for the groundwater based on river water input. Eventually, the combination of physical-based numerical model (i.e., using an advection–dispersion-filtration transport equation) and data-driven model was proposed to directly simulate and predict groundwater quality in response to the variation of river water quality. The results showed that both models could predict the status of riparian groundwater quality, and the physical-based model could better interpret the results compared to the data-driven black-box model. This study promotes our understanding of the importance of riverbank in improving groundwater quality in the areas with intensive agriculture practices, and riverbank restoration strategies should be carefully designed to maximize their potential to degrade non-point pollutions.