Understanding the age and movement of groundwater is important for predicting the vulnerability of wells to contamination, constraining flow models that inform sustainable groundwater management, and interpreting geochemical signals that reflect past climate. Due to both the ubiquity of groundwater with order ten-thousand-year residence times and its importance for climate reconstruction of the last glacial period, there is a strong need for improving geochemical dating tools on this timescale. Whereas ¹⁴C of dissolved inorganic carbon and dissolved ⁴He are common age tracers for Late Pleistocene groundwater, each is limited by systematic uncertainties related to aquifer composition and lithology, and the extent of water-rock interaction. In principle, radiogenic ⁴⁰Ar in groundwater acquired from decay of ⁴⁰K in aquifer minerals should be insensitive to some processes that impact ¹⁴C and ⁴He and thus represent a useful, complementary age tracer. In practice, however, detection of significant radiogenic ⁴⁰Ar signals in groundwater has been limited to a small number of studies of extremely old groundwater (>100 ka). Here we present the first high-precision (<1‰) measurements of triple Ar isotopes (⁴⁰Ar, ³⁸Ar, ³⁶Ar) in groundwater. We introduce a model that distinguishes radiogenic ⁴⁰Ar from atmospheric ⁴⁰Ar by using the non-radiogenic Ar isotopes (³⁶Ar, ³⁸Ar) to correct for mass-dependent fractionation. Using this model, we investigate variability in radiogenic ⁴⁰Ar excess (Δ⁴⁰Ar) across 58 groundwater samples collected from 36 wells throughout California (USA). We find that Δ⁴⁰Ar ranges from ~0‰ (the expected minimum value) to +4.2‰ across three study areas near Fresno, San Diego, and the western Mojave Desert. Based on measurements from a network of 23 scientific monitoring wells in San Diego, we find evidence for a strong dependence of Δ⁴⁰Ar on aquifer lithology. We suggest that Δ⁴⁰Ar is fundamentally controlled by the weathering of old K-bearing minerals and thus reflects both the degree of groundwater-rock interaction, which is related to groundwater age, and the integrated flow through different geological formations. Future studies of Late Pleistocene groundwater may benefit from high-precision triple Ar isotope measurements as a new tool to better interpret ¹⁴C- and ⁴He-based constraints on groundwater age and flow.

了解地下水的年龄和运动对于预测水井对污染的脆弱性、约束为可持续地下水管理提供信息的流动模型以及解释反映过去气候的地球化学信号非常重要。由于地下水的普遍存在，其停留时间为一万年，而且它对末次冰期气候重建的重要性，因此迫切需要在这个时间尺度上改进地球化学定年工具。而溶解无机碳的¹⁴ C 和溶解的⁴ He 是晚更新世地下水的常见年龄示踪剂，每个都受到与含水层组成和岩性以及水-岩相互作用程度相关的系统不确定性的限制。原则上，放射性⁴⁰从含水层矿物中⁴⁰ K衰变获得的地下水中的 Ar应该对影响¹⁴ C 和⁴ He 的某些过程不敏感，因此是一种有用的补充年龄示踪剂。然而，在实践中，地下水中显着的放射性⁴⁰ Ar 信号的检测仅限于对极古老地下水（> 100 ka）的少数研究。在这里，我们首次对地下水中的三 Ar 同位素（⁴⁰ Ar、³⁸ Ar、³⁶ Ar）进行了高精度 (<1‰) 测量。我们引入了一个模型，区分放射⁴⁰从大气压的Ar ⁴⁰通过使用非放射性的Ar氩同位素（³⁶Ar, ³⁸ Ar) 以校正质量相关的分馏。使用该模型，我们研究了从整个加利福尼亚州（美国）的 36 口井中收集的 58 个地下水样本中放射性⁴⁰ Ar 过量 (Δ ⁴⁰ Ar) 的变异性。我们发现，在弗雷斯诺、圣地亚哥和莫哈韦沙漠西部附近的三个研究区域中，Δ ⁴⁰ Ar 的范围从 ~0‰（预期最小值）到 +4.2‰。根据圣地亚哥 23 口科学监测井网络的测量结果，我们发现证据表明 Δ ⁴⁰ Ar 对含水层岩性有很强的依赖性。我们建议 Δ ⁴⁰Ar从根本上受古老含钾矿物的风化作用控制，因此既反映了与地下水年龄有关的地下水-岩石相互作用的程度，又反映了通过不同地质地层的综合流动。对晚更新世地下水的未来研究可能会受益于高精度三重 Ar 同位素测量，作为一种新工具，可以更好地解释¹⁴ C 和⁴ He 对地下水年龄和流量的限制。