Establishment and application of a new method identifying nitrogen pollution sources using CHEMTAX model combined with hydrochemical fingerprint indicators

Source identification of total nitrogen (TN) in rivers is important for water environment protection. This study aimed to establish a TN source identification method using CHEMTAX combined with fingerprint indicators and to apply this method in the basin of the upper Hanjiang River. Forty-eight end-member samples, including cropland, forest, grassland, wastewater, litter and manure samples, were collected and mixed to obtain 95 receptor samples with preset proportions. Four cations (K+, Ca2+, Na+, and Mg2+), two anions (Cl- and SO42-), five trace elements (B, Sr, Si, La, and Se) and four dissolved organic matter (DOM) spectral indices (FI, SR, SUVA254, and HIX) of both the end-member and receptor samples were measured as fingerprint indicators. Finally, a TN source identification method based on CHEMTAX using concentrations of four cations (K+, Ca2+, Na+, and Mg2+) and two elements Sr and Si was established. In addition, methods of hydrochemical fingerprint indicator combined with the CHEMTAX model and Bayesian mixing model based on nitrogen and oxygen stable isotopes were employed for the nitrate source identification in the upper Hanjiang River Basin. The principal coordinates analysis results, variance partitioning analysis and random forest models analysis, and the source contribution correlations between the MixSIAR and CHEMTAX method indicated that nitrogen contribution of three sources including soil, wastewater and litter & manure calculated by this CHEMTAX based method was credible. The established method not only reduced test cost, but also showed potential for online nitrogen source identification monitoring. This study opened a new window for TN source identification.