Summary and Conclusions


Chlorofluorocarbon (CFC), sulfur hexafluoride (SF₆), and tritium/helium-3 (³H/³He) tracer data collected from 1,399 NAWQA Program groundwater sites across the United States were compiled in this report. The data were from Land-Use Study (LUS), Major-Aquifer Study (MAS), and Reference (REF) networks, and from several OTHER networks composed of miscellaneous wells not represented in other NAWQA groundwater-age archives. The period of data collection focused primarily on Federal fiscal years 1992–2005, although two networks of samples from 2006 to 2007 were included (tracer data collected during 2006–07 from one group of wells was superior to tracer data collected in the 1990s from the same network, and tracer data collected during 2007 from another group of wells was included to fill a data gap). Tracer data from other NAWQA Program components (Flow System Studies and various topical studies) were not compiled herein. Tracer data collected after 2005 were not compiled in this report, except as noted.


Tracer data that previously had been interpreted and published were compiled (618 sites), and tracer data that previously had not been interpreted and published were evaluated (781 sites) using established methods. For the newly interpreted tracer data, the following are compiled: aqueous concentrations, equivalent atmospheric concentrations (for CFCs and SF₆), and estimates of tracer-based piston-flow ages; selected ancillary data (for example, well construction and redox-chemistry data) and brief summaries of each tracer dataset also are included. For the previously interpreted and published tracer data, published tracer-based piston-flow ages are listed, along with documentation of the original references and brief discussions of each dataset.


Tracer-based piston-flow ages have limitations, yet they are derived from point measurements that provide understanding that may be difficult to obtain from other methods. Tracer-based piston-flow ages commonly can be used to represent relative groundwater ages, and (less commonly) can closely represent actual groundwater ages. Tracer-based piston-flow ages can be used in establishing directions of hydrologic progression or geochemical evolution. Tracer-based piston-flow ages also can be used to establish timescales of hydrologic progression or kinetics of geochemical reactions. The degree to which simple tracer-based piston-flow ages can be used for purposes such as these depends on the degree of confidence associated with the interpretations. Multiple tracers or multiple lines of evidence can increase the understanding of tracers and any interpretations based upon those tracers. For the datasets in this report, multiple tracers were in most cases limited to multiple CFCs (which are limited in that the different CFCs have differing degrees of reactivity), or to ³H (only sporadically collected) in addition to CFCs, SF₆, or ³H/³He. The summaries of the individual tracer datasets provide information about these datasets and interpretations that can help the users of these tracer-based piston-flow ages identify strengths and weaknesses and thus identify appropriate uses for the different datasets.


LUS sites, generally composed of monitoring wells installed near the water table in recharge zones, often seem to yield groundwater that, at least in part, has been recharged in recent years. MAS sites, generally composed of domestic wells and other types of supply wells situated in the shallow regions of principal aquifers also yield young tracer-based piston-flow ages, but these ages tend to be slightly greater (older) than those of LUS sites. The median tracer-based piston-flow age of the dated LUS sites in this compilation was 11 years and for the dated MAS sites was 17 years. Of course, the median tracer-based piston-flow ages of the dated LUS and MAS sites are not likely to represent the median tracer-based piston-flow ages of the undated LUS and MAS sites precisely. Additional tracer analysis at these undated sites is needed.


In the process of analyzing tracer data, major dissolved-gas data also were analyzed where available. Because recharge temperature often is inferred using climate data (recharge temperature often being approximated by MAAT+1°C), the major dissolved-gas data provide an opportunity to compare N₂/Ar-inferred recharge temperatures to climate-based recharge temperatures for a diverse group of aquifers across the United States. For aquifers composed of sediments (the most common aquifer type in this report), differences between N₂/Ar-inferred recharge temperatures and recharge temperatures based on MAAT+1°C were, on average, within about 0.1°C. However, although the average differences were small, the standard deviation of these differences was 3.1°C, indicating that climate data may be useful for characterizing average recharge temperatures, but it also can indicate that recharge temperatures vary greatly from site to site.


Characterizing groundwater age is a complex undertaking. As evidenced in this compilation, collection of tracer data does not necessarily lead to determination of groundwater age. Collecting tracer data is only part of the investigation, but careful selection of tracers as well as ancillary geochemical data, and preparation of a well-thought-out study design can improve the outcome. Some of the more salient lessons from the effort documented in this report are discussed below. 


Where characterizing the distribution of groundwater ages in a sample is a goal, multiple tracers along with appropriate use of analytical or numerical models are needed. Even where only a tracer-based piston-flow age is to be reported, multiple tracers can provide support for tracer-based piston-flow age interpretations, or identify cases where the piston-flow model might be inappropriate. If the goal is simply to determine a tracer-based piston-flow age, the availability of multiple tracers can help increase the chances that at least one tracer will yield an interpretable tracer-based piston-flow age.


In the planning phase of a project, design of a network with a flow-system approach provides hydrologic context during tracer interpretation, and a design focusing on short-screened wells will reduce well-bore mixing. Selection of long-screened wells will mandate the use of multiple tracers and perhaps depth-dependent tracer sampling. The choice of tracers and approaches for age dating can be optimized to local environmental conditions. For example: (1) sites in urban areas (which often may contain elevated concentrations of CFCs) might be best served by tracers other than CFCs, (2) sites in silicic igneous rocks (a source of natural SF₆) might be best served by tracers other than SF₆, and (3) sites with thick unsaturated zones (where CFC and SF₆ transport may be impeded) may require that these tracers either not be used or be used in conjunction with unsaturated-zone sampling for CFCs and SF₆ to provide understanding of unsaturated zone transport. Prior to full project sampling, sampling a subset of wells for a wide range of tracers to assess tracer reliability for a given study area may provide information not available by other means. Sampling efforts can benefit where on-site information is used. For example, if degassing is observed during sampling, collection of He samples is likely to result in fractionated samples, whereas if suboxic conditions are observed, collection of CFCs may result in degraded tracers. Major dissolved gases have not been collected consistently with tracers. Major dissolved gases provide critical constraints on recharge temperature, excess air, and redox state. Similarly, ³H has been collected sporadically. Because ³H is a conservative tracer with a known historical record, it can provide valuable insights when used in conjunction with age-dating tracers. Other supporting data are collected infrequently. For example, noble gases can support the improved determination of recharge temperature, gas fractionation at the water table, and recharge altitude. These are but a few of the considerations, many of which have been discussed in published literature, which can contribute to improved age-dating results.


First posted January 27, 2011