In the process of groundwater environmental monitoring, while ensuring the representativeness of groundwater quality evaluation, the number of monitoring indicators should be optimized as much as possible, which is of great significance to groundwater environmental management. Based on monitoring data of shallow groundwater in Nanchang in 2014 and 2019, the chemical characteristics of water and the changes in water quality were analyzed via statistical analysis, a Piper three-line diagram, and the entropy-weighted water quality index (EWQI). Furthermore, a key indicator optimization method based on water quality evaluation was constructed by coupling stepwise multiple linear regression analysis. The feasibility of this method was also evaluated. The results showed that the water chemistry type of groundwater in 2014 and 2019 was mainly HCO3-Ca, and the five abnormal indicators pH value, NO3-, I-, Fe, and Mn were the main influencing factors of water quality change. The water quality in 2019 was generally higher than that in 2014, which was considered as overall "moderate," and the average EWQI values of the two years were 53.72 and 82.34, respectively. The optimal model EWQImin-4 constructed based on the key indicator optimization method could better represent the actual EWQI; the key indicators included Mn, NO3-, TH, Fe, pH value, and I-; and the determination coefficient (R2) and percentage error (PE) values were 0.865 and 10.61%, respectively. Therefore, the optimization method of groundwater monitoring indicators based on entropy-weighted water quality evaluation could be used as an important reference for optimizing monitoring indicators and provide a method for regional groundwater environmental management.
Keywords: entropy-water quality index(EWQI); groundwater quality; key indicators; optimization method; stepwise multiple linear regression.