The time that water takes to travel through the terrestrial hydrological cycle and the critical zone is of great interest in Earth system sciences with broad implications for water quality and quantity. Most water age studies to date have focused on individual compartments (or subdisciplines) of the hydrological cycle such as the unsaturated or saturated zone, vegetation, atmosphere, or rivers. However, recent studies have shown that processes at the interfaces between the hydrological compartments (e.g., soil‐atmosphere or soil‐groundwater) govern the age distribution of the water fluxes between these compartments and thus can greatly affect water travel times. The broad variation from complete to nearly absent mixing of water at these interfaces affects the water ages in the compartments. This is especially the case for the highly heterogeneous critical zone between the top of the vegetation and the bottom of the groundwater storage. Here, we review a wide variety of studies about water ages in the critical zone and provide (1) an overview of new prospects and challenges in the use of hydrological tracers to study water ages, (2) a discussion of the limiting assumptions linked to our lack of process understanding and methodological transfer of water age estimations to individual disciplines or compartments, and (3) a vision for how to improve future interdisciplinary efforts to better understand the feedbacks between the atmosphere, vegetation, soil, groundwater, and surface water that control water ages in the critical zone.

Noble-gas concentrations in ground water have been used as a proxy for past air temperatures^{1,2,3,4,5,6,7}, but the accuracy of this approach has been limited by the existence of a temperature-independent component of the noble gases in ground water, termed ‘excess air’, whose origin and composition is poorly understood^7,8,9. In particular, the evidence from noble gases in a Brazilian aquifer for a cooling of more than 5 °C in tropical America during the Last Glacial Maximum⁴ has been called into question⁹. Here we propose a model for dissolved gases in ground water, which describes the formation of excess air by equilibration of ground water with entrapped air in quasi-saturated soils^10,11,12. Our model predicts previously unexplained noble-gas data sets, including the concentration of atmospheric helium, and yields consistent results for the non-atmospheric helium isotopes that are used for dating ground water. Using this model of excess air, we re-evaluate the use of noble gases from ground water for reconstructing past temperatures. Our results corroborate the inferred cooling in Brazil during the Last Glacial Maximum⁴, and indicate that even larger cooling took place at mid-latitudes.

Concentrations of atmospheric noble gases (neon, argon, krypton, and xenon) dissolved in groundwaters from northern Oman indicate that the average ground temperature during the Late Pleistocene (15,000 to 24,000 years before present) was 6.5° ± 0.6°C lower than that of today. Stable oxygen and hydrogen isotopic groundwater data show that the origin of atmospheric water vapor changed from a primarily southern, Indian Ocean source during the Late Pleistocene to a dominantly northern, Mediterranean source today. The reduced northern water vapor source is consistent with a drier Last Glacial Maximum through much of northern Africa and Arabia.