A three-pronged approach for identifying source and extent of nitrate contamination in groundwater

Abstract

A method to delineate the nitrate (NO₃⁻) contamination of a study area by sequential and statistical use of dual isotope data, principal component analysis (PCA), and land use/land cover (LULC) data was demonstrated using data from Eumseong, Korea. First, a dual isotope approach was applied to identify the possible NO₃⁻ sources and quantify their contribution to NO₃⁻ contamination using Bayesian statistics. Second, a PCA was performed to discriminate and evaluate the impact of NO₃⁻ contamination on chemical evolution in the aquifer. Lastly, we incorporated the LULC data into a regression analysis to identify the contribution of various land uses to NO₃⁻ recharge. Some samples had NO₃⁻ and iron (Fe) concentrations above the local drinking water quality standard, and the distributions of potassium (K⁺), sulfate (SO₄²⁻), Fe, and manganese (Mn) were skewed significantly. Trends from the dual isotope analysis suggested three major sources of NO₃⁻ contamination. Among the three sources, Bayesian statistics indicated that the NO₃⁻ contamination was largely attributable to influx from manure/sewage and soil nitrogen (N). Based on the PCA, following screening for skewed data, a contamination indicator was extracted. The indicator exhibited positive correlations with NO₃⁻, chlorine (Cl⁻), strontium (Sr²⁺), calcium (Ca²⁺), sodium (Na⁺), magnesium (Mg²⁺), K⁺, and SO₄²⁻, suggesting a single strong source. The regression analysis, using the LULC information, determined that agriculture activities in nonpaddy areas were responsible for the NO₃⁻ recharge. This study identified the benefits of combining dual isotope analysis, PCA, and LULC data for discriminating and evaluating sources of NO₃⁻ contamination when diverse contaminants are involved in geochemical evolution.