[Groundwater Pollution Source Identification by Combination of PMF Model and Stable Isotope Technology]
Han Zhang,Xin-Yu Du,Fei Gao,Zhuo Zeng,Si-Qian Cheng,Yi Xu
DOI: https://doi.org/10.13227/j.hjkx.202110174
2022-08-08
Abstract:The pollution source identification methods based on traditional water quality monitoring and pollutant discharge loading typically require a high frequency of monitoring and generate a level of uncertainty in the identification results, owing to their limitations on the accurate and quantitative assessment of pollution source identification, migration, and transformation. This study combined multivariate statistical analysis and stable isotope technology to identify groundwater pollution sources in a typical multiple land-use area of the Chengdu Plain. A positive matrix factorization (PMF) model was adopted to reduce the interference of mass environmental factors on source identification and to determine the main factors influencing groundwater quality. Subsequently, a Bayesian stable isotope mixing model was developed to quantify the apportionment of each pollution source to groundwater nitrate (NO3-) with the consideration of hydro-chemical and land-use information. The results showed that the concentrations of NO3-, NO2-, NH4+, Mn, Fe, SO42-, and Cl- in groundwater of the study area exceeded the standard to different extents, presenting spatial variation. The main form of inorganic nitrogen in groundwater was NO3-. In general, concentrations of groundwater NO3- were the highest in vegetable fields (9.29 mg·L-1 on average), followed by livestock and poultry breeding farms (7.66 mg·L-1) and arable land (7.09 mg·L-1), whereas concentrations of groundwater NO3- in industrial areas were the lowest (2.20 mg·L-1). Groundwater quality in the study area was affected by geological processes, agricultural activities, hydrogeochemical evolution, and domestic and industrial discharges. Agricultural activities were the main contributor to the increase in groundwater NO3- in the study area. Chemical fertilizer (32%) and soil nitrogen (25%) contributed greatly to groundwater NO3- in agricultural areas, whereas sewage (28%) and atmospheric precipitation (27%) contributed most groundwater NO3- in industrial areas. Thus, the combination of multivariate statistical analysis and stable isotope technology could identify groundwater pollution sources and their apportionment effectively, providing scientific support for the prevention and control of groundwater pollution.