Isotope data such as δ13C and 14C of dissolved inorganic carbon (DIC) have been widely used for dating groundwater with ages in the range up to 40,000 years. Various single-sample-based (SSB) correction models have been developed for 14C dating including those proposed by Eichinger (Radiocarbon, 25(2):347–356, 1983), Evans et al. (Isotope Hydrol II, 679–708, 1978), Fontes and Garnier (Water Resour Res 15(2):399–413, 1979), Han and Plummer (Chem Geol 351:105–114, 2013), Ingerson and Pearson (Estimation of age and rate of motion of groundwater by the 14C method, 263–283, 1964), Mook (On the reconstruction of the initial 14C content of groundwater from the chemical and isotopic composition, 1972; The dissolution-exchange model for dating groundwater with 14C, 213–225, 1976; Handbook of Environmental Isotopes Geochemistry, The Terrestrial Enviroment A, pp 49–74, 1980), the IAEA model by Salem et al. (Groundwater flow patterns in the western Libyan Arab Jamahiriya evaluated from isotopic data, 165–179, 1980), Tamers (Isotope Tech Hydrol Cycle 11:143–152, 1967; Geophys Surv 2(2):217–239, 1975) and Wigley (Nature 263(5574):219–221, 1976). However, the reliability of these models has never been practically tested, because groundwater data with known ages are not available. This study synthesized geochemistry data, including inorganic carbon isotopes, for different groundwater samples by simulating various geochemical processes that could occur in both the unsaturated and saturated zones. The theoretical groundwater ages were estimated using mass-balance approaches that account for these processes. The same data sets were also applied in SSB models, and their corrected ages were compared with the theoretical ages from the mass-balance simulations to evaluate model performances. Results showed that most SSB models provided acceptable ages in scenario A, where only calcite dissolution was considered, with age deviations less than 11 years. However, when groundwater experienced complicated chemical reactions, the differences became much more significant. In scenario B, which added carbon exchange between DIC and soil gas CO2 in the unsaturated zone into consideration, only two models developed by Mook and Han and Plummer provided reasonable ages with deviations of less than 100 years, while other models underestimated groundwater ages by about 1500–4489 years. In scenario C, which included silicate weathering and organic matter decomposition in the saturated zone, all SSB models overestimated groundwater ages, with typical differences ranging from over 1400 to nearly 2000 years, and some models showing deviations as significant as 3294 or 4996 years. We found that the results were most reliable when the geochemical reactions that affected the carbon isotopic compositions were corrected from geochemical data. In this sense, we suggest using mass balance methods by analyzing various kinds of geochemical data together with isotopes of dissolved inorganic carbon to get better dating results.
