Understanding groundwater age distribution is crucial for assessing aquifer vulnerability to contamination and evaluating the sustainability of water extraction. Anthropocene water (post-1953 recharge) indicates susceptibility to anthropogenic contamination, while Holocene and Pleistocene water signifies potential groundwater mining. While single-tracer analyses offer insights, they don't provide the complete picture of age distribution. This study employs multiple tracers (tritium, tritiogenic helium-3, sulfur hexafluoride, chlorofluorocarbons, carbon-14, and radiogenic helium-4) to comprehensively characterize groundwater age distributions within 21 Principal Aquifers (PAs) in the continental US, representing roughly 80% of public water supply. The research aims to provide a detailed, spatially unbiased assessment of groundwater age across these significant aquifers and the continental US, ultimately informing sustainable groundwater management practices.

Previous research has highlighted the importance of groundwater age in assessing vulnerability to contamination (Ferguson et al., 2020) and the need for multiple tracers to accurately determine age distributions (Lindsey et al., 2019). Studies using single tracers have provided global-scale evaluations of groundwater vulnerability (Gleeson et al., 2016; Jasechko et al., 2017), but these lack the detail offered by analyzing multiple tracers and the complete age distribution. The authors also reference previous work using lumped parameter models (LPMs) to estimate groundwater age distributions, demonstrating their similarity to more computationally intensive numerical models (Eberts et al., 2012; Jurgens et al., 2016). The paper builds upon this existing literature by providing a high-resolution, spatially explicit analysis of groundwater age across a significant portion of the US public water supply.

The study analyzed groundwater samples from 1279 public-supply wells across 21 PAs, using multiple tracers with overlapping dating ranges. These tracers were selected to identify both Anthropocene and pre-Anthropocene water. Lumped parameter models (LPMs) were employed to fit the measured tracer concentrations and derive age distributions for each sample (Jurgens et al., 2012). The sampling design was spatially unbiased, ensuring representative data at both PA and continental scales. Each PA was divided into equal-area cells, with one well selected per cell (Scott, 1990; Belitz et al., 2010). An aquifer-scale cumulative distribution function (ACDF) was calculated for each PA, representing the spatial average of individual sample age distributions. The ACDF allowed the researchers to determine the proportion of Anthropocene, Holocene, and Pleistocene groundwater within each PA. Data processing involved correcting tracer concentrations for various factors, including geochemical dilution for carbon-14, solubility, excess air, and gas fractionation for other gas tracers (Jurgens et al., 2020). The TracerLPM software (Jurgens et al., 2012) was used to model groundwater age distributions, utilizing either a single dispersion model or a bimodal mixing model to fit the tracer data. The authors discuss the considerations and uncertainties associated with data processing and model fitting. ACDF analysis provided a comprehensive understanding of groundwater age distributions within and across different PAs.

The mean age of groundwater samples ranged from 1 to ~750,000 years, with a median of ~2500 years. Approximately half the samples contained pre-Anthropocene water, while about 20% were Anthropocene. Notably, 30% of the samples exhibited bimodal age distributions, reflecting mixtures of Anthropocene and pre-Anthropocene water. These bimodal distributions highlight the limitations of single-tracer methods. The study found that nationally, 38% of groundwater is Anthropocene, 34% Holocene, and 28% Pleistocene. However, this varies significantly across PAs. Western unconsolidated aquifers primarily contain Holocene water, reflecting low recharge rates in arid climates. Coastal clastic aquifers predominantly have Pleistocene water due to confinement and long lateral flow paths. Interior sandstone and carbonate aquifers also show a mix of Holocene and Pleistocene water. The Glacial aquifer system shows the highest proportion of Anthropocene water. Carbonate aquifers exhibit diverse age distributions, influenced by both climate and confinement, and the oldest groundwater was found in deeply buried parts of the Cambrian-Ordovician aquifer. The two fractured-rock aquifers showed a strong influence of climate on groundwater age. The ACDFs effectively distinguished hydrologic characteristics across different PAs. The study underscores that aquifers with high Anthropocene proportions are susceptible to anthropogenic contamination, while those with high Holocene or Pleistocene fractions may face geogenic contamination, although geology also plays a critical role. The sustainability of groundwater extraction varies depending on the PA, with some exhibiting signs of mining while others maintain a hydrologic balance through enhanced recharge.

The findings highlight the importance of considering the full groundwater age distribution, rather than relying on mean ages or single-tracer data, for assessing both contamination risk and the sustainability of groundwater use. The spatial variability in groundwater age across different aquifer types and climatic regions underscores the need for regionally specific management strategies. The high percentage of Anthropocene water in many aquifers emphasizes the potential impact of human activities on groundwater quality. The study's methods provide a robust framework for evaluating groundwater resources, and the results offer crucial insights for effective groundwater management. The detailed analysis of ACDFs contributes significantly to understanding the complex hydrogeology of different aquifer types and its implications for water resource management.

This study provides a comprehensive assessment of groundwater age distribution across major US aquifers, demonstrating the importance of considering both anthropogenic and geogenic contamination risks and the sustainability of groundwater extraction. The use of multiple tracers and ACDF analysis offer a powerful approach to characterize groundwater age distributions and their spatial variability. Future research could focus on improving the spatial resolution of groundwater age assessments and integrating this information into dynamic groundwater models for more refined prediction of future conditions. Further investigation into the relationship between groundwater age and specific contaminant occurrence in different hydrogeological settings would also be beneficial.

While the study covers a significant portion of US public water supply, it does not encompass all aquifers. The accuracy of age estimations is influenced by the precision of tracer measurements and the assumptions underlying the LPMs. The study's focus is on public-supply wells, which may not fully capture the age distribution across the entire aquifer. Uncertainties associated with tracer corrections could influence the results, particularly for pre-Anthropocene water.