What the measurement can and cannot tell you
Groundwater CFC and SF₆ tracer ages are interpretations of dissolved gas concentrations, not birthdays stamped on individual water molecules. Chlorofluorocarbons such as CFC-11 and CFC-12 and sulfur hexafluoride, abbreviated SF₆, have changed in the atmosphere over decades. If water dissolved these gases during recharge and then became isolated from the atmosphere, its measured concentration can be compared with a historical atmospheric input curve. The date that matches the corrected concentration is called an apparent recharge date. Subtracting that date from the sampling date yields an apparent age, conditional on the chosen model.
Readers searching for how to interpret CFC groundwater age, SF6 groundwater dating, or excess-air corrections often encounter a deceptively precise output: for example, a report may show a single recharge year. The useful question is not whether that one number looks plausible. It is whether field sampling, gas solubility, atmospheric history, local contamination, chemical preservation and groundwater mixing all permit that conclusion. Different tracers should be compared as independent tests of the same proposed history rather than averaged automatically.
This guide begins with the language of the laboratory report, translates every term into a physical process, demonstrates checked hypothetical calculations, and then shows how to reject an interpretation that cannot survive evidence from another tracer. The numerical cases are deliberately illustrative, not measurements of an aquifer. A good final statement will identify a recharge-date window and its assumptions, describe alternative explanations, and say what additional measurement could discriminate among them.
A dissolved concentration is not an atmospheric concentration
A laboratory reports the dissolved mass or amount of a compound per unit mass of water, perhaps picograms per kilogram for some CFC analyses or femtomoles per litre for SF₆. Historical atmospheric records usually report a dry-air mole fraction in parts per trillion by volume, pptv. The units differ because one describes a substance in liquid water and the other describes a ratio in air. There is no direct one-to-one numeric conversion between them without gas solubility, temperature, pressure and sometimes salinity.
Henry-type equilibrium relationships connect a partial pressure in air with an equilibrium dissolved concentration in water. In a simplified notation, C_eq=K(T,S)×p, where K depends on temperature and salinity and p is the relevant partial pressure. Different references define Henry constants in reciprocal ways. The safe practice is to write the definition used by the chosen software or laboratory, together with its units, rather than borrowing a numerical coefficient from a source that uses the inverse convention.
Temperature matters because cooler water generally dissolves more gas at the same pressure. Recharge elevation matters because atmospheric pressure changes with elevation. Salinity changes solubility as well. A groundwater sample from a cool upland recharge area cannot be interpreted correctly by entering the warmer temperature recorded beside the well during sampling. The measured discharge temperature may reflect geothermal warming, long travel or pump conditions; it need not equal temperature at the moment of original gas equilibration.
The timeline: atmosphere, recharge, isolation, pumping
Imagine a clean rainwater parcel infiltrating an unsaturated zone. Before complete saturation and isolation, its dissolved volatile tracers can exchange with soil gas whose composition may approximate or depart from local atmosphere. The parcel reaches the water table, acquires dissolved air under particular temperature and pressure conditions, and then flows along a hydraulic pathway. Its tracer concentrations are carried along that pathway unless later mixing, reaction, diffusion, gas loss or exchange alters them.
The ‘clock’ used by CFC and SF₆ methods is primarily the historically changing atmospheric input, not radioactive decay of the molecules after recharge. This is why a CFC apparent age and a tritium–helium-3 age are not obtained by the same equation. For transient atmospheric gases, a measured dissolved concentration is converted to an equivalent atmospheric concentration at recharge, then mapped to an input curve. For a radioactive system, decay and daughter production have different governing balances.
A pump can destroy the parcel story by collecting water from multiple screens or fractures at once. When old and young waters combine, the concentration in the bottle is often a mixture of tracer concentrations, not the concentration corresponding to the arithmetic mean recharge year. If a short screened well intercepts several thin inflows, even a visibly stable field parameter does not prove that all entering water has one recharge date. The sampling system belongs in the interpretation.
Why historians of air are essential to groundwater geochemistry
The concentration of a tracer in the atmosphere is the input function. Published reconstructions and measurements can differ with hemisphere, sampling period, local industrial releases and changes in the compound’s emissions. CFC atmospheric curves rose strongly in much of the twentieth century but later slowed, flattened or declined after restrictions. This creates situations in which one dissolved concentration may correspond to more than one recharge year or cannot resolve neighboring recent years reliably.
SF₆ is often useful for relatively young groundwater because its atmospheric mole fraction has generally increased over the modern record. That directional signal helps distinguish recharge years when CFC curves are flat. Nevertheless, sulfur hexafluoride can be introduced by equipment, anthropogenic contamination or some rocks and geothermal systems. A steadily increasing global curve does not make every SF₆-bearing sample a valid atmospheric clock.
Be precise about chronology. A laboratory reference curve prepared years ago should not be called the current atmospheric concentration simply because its website is accessible today. Likewise, a paper’s dating window reflects the years and analytical conditions described there. For a 2026 sample, consult the actual updated input function available to the laboratory rather than extend a plotted curve by eye. This guide makes no invented claim about a 2026 mixing ratio.
What an apparent date means
The word ‘apparent’ signals model dependence, not careless science. Under a piston-flow assumption, a sampled water parcel has experienced a single recharge time and no alteration of tracer concentration along the pathway. With a corrected sample concentration and a historical curve, one may choose an equivalent recharge year. In a real aquifer, each of those assumptions is a hypothesis to be tested against site observations.
Suppose a validated tracer analysis under an explicit recharge temperature, elevation and excess-air correction corresponds to atmospheric conditions of 2008, with a plausible matching interval from 2005 to 2011. If the water was sampled in 2026, the nominal apparent age is 2026−2008=18 years, and the interval under those limited assumptions is 15 to 21 years. No single gas concentration needs to be invented to demonstrate this calendar subtraction. The interval would widen if recharge conditions or input-curve ambiguities were poorly constrained.
The same result should not be rewritten as ‘the aquifer is 18 years old.’ Aquifers and sedimentary units can be much older than flowing pore water. Nor does the result necessarily identify a travel time between a named catchment and the well. A hydraulic source attribution requires potentiometric surfaces, geology, water chemistry and sometimes flowpath modeling. The apparent date is an observation-based clue about groundwater renewal, not a complete flow model.
Field sampling: the easiest stage at which to lose a good result
Volatile gas samples are unusually vulnerable to contact with air. A bottle with headspace, an aerating faucet, leaky tubing, turbulence through a pump or prolonged transfer can alter the composition before the laboratory sees it. Degassing can remove dissolved constituents; contact with modern air can add them. Neither bias has to be obvious from colour, smell or routine electrical conductivity. Therefore the field log should identify sampling connection, tubing, container, fill method, flow conditions and whether bubbles appeared.
The USGS groundwater dating laboratory publishes method-specific sampling directions. A practitioner should follow the laboratory actually performing the analysis, because container specifications and analytical methods vary. Avoid treating an attractive generic ‘water sample’ checklist as interchangeable with CFC or SF₆ requirements. The central goal is to preserve the dissolved-gas signal and ensure that the laboratory interprets the collection procedure as intended.
A field duplicate that disagrees with the primary sample can indicate variability or handling error; it does not independently diagnose which bottle is correct. Equipment blanks, paired well samples and checks for headspace are powerful when selected for a clear purpose. If a well crosses several zones, collect depth-discrete samples when feasible, rather than later assigning one integrated composite concentration to a single stratigraphic unit. A qualified hydrogeologist should decide how to sample safely.
Why recharge temperature has to be inferred rather than guessed
Recharge temperature is a model input that often controls calculated solubility strongly. An inexpensive approximation might use mean annual ground conditions, but climates with strong seasonality, snowmelt recharge, mountain recharge or irrigation returns can violate that assumption. A sample that reached the water table during a cold event may retain gas ratios associated with that event despite being measured decades later in a warm well casing.
Dissolved noble gases and major gases may help estimate temperature at recharge. Neon, argon, krypton and xenon each respond to temperature and other processes differently, so a measured suite can fit temperature and excess-air parameters more securely than a single arbitrary assumption. In simpler settings, dissolved nitrogen and argon can be informative if other nitrogen-producing processes are understood. The practical interpretation is a fitted parameter set with uncertainties, not a thermometer reading from a depth probe.
When reading a technical table, distinguish a laboratory measured gas concentration from an estimated recharge temperature computed by fitting those measurements. A useful audit records the gas dataset, pressure or altitude assumptions, model of gas equilibration, statistical fit and plausible parameter ranges. Reporting ‘recharge temperature 14 °C’ without how it was inferred encourages users to propagate that value into other dates as though it were independent evidence.
Excess air: an ordinary hydrologic process with an extraordinary dating effect
Excess air is atmospheric gas dissolved into groundwater beyond ideal equilibrium solubility, often when trapped bubbles are compressed or dissolved as the water table rises. Its label does not mean air remained visibly as bubbles in the collected sample. It can represent a dissolved component that increases measured gas amounts even after sampling appears perfectly bubble-free. Models may assume unfractionated excess air or describe partial bubble dissolution and gas fractionation.
For SF₆, even a relatively small excess-air contribution can matter because atmospheric concentrations are tiny and solubility is low. If that additional gas is mistaken for a larger past atmospheric mixing ratio, a tracer whose atmospheric abundance was rising can yield an apparently too-recent recharge date. A correction is a quantitative subtraction or model adjustment made before comparing the input curve, not a verbal footnote appended after the age has been selected.
Do not assume every excess measurement is due to compressed air. A geological source of SF₆ can produce another kind of excess; contamination from a handling system can do likewise. Noble-gas patterns, multiple tracers, geological context and field quality controls distinguish possibilities. The operator should test whether the fitted excess-air amount is compatible with noble gases that are not subject to the same environmental releases.
A transparent hypothetical excess-air calculation
Consider an explicitly synthetic gas budget in which total measured SF₆ is 1.40 femtomoles per litre. A defensible independent excess-air model estimates that 0.20 femtomoles per litre of that result came from excess atmospheric air. Assume no other source and no in-transit loss. The equilibrium component to map to an atmospheric input curve is 1.40−0.20=1.20 femtomoles per litre. The correction is 0.20/1.40=0.142857, or approximately 14.3% of the measured amount.
If the atmospheric-equivalent conversion were mistakenly run on 1.40 instead of 1.20 femtomoles per litre, the inferred atmospheric value would be greater, for the same Henry constant. On a rising SF₆ input curve, that greater value normally maps to a later recharge date and thus a younger apparent age. Without the actual solubility model and year-specific input curve we cannot truthfully calculate a numerical age shift. The correct worked answer is the corrected concentration and the direction of bias.
A sensitivity test makes this more useful. Suppose the estimated excess-air contribution is 0.20±0.05 femtomoles per litre while the total is treated as exact for teaching. The corrected component ranges from 1.15 to 1.25 femtomoles per litre. One should map each endpoint through the actual atmospheric and solubility model, because the relationship between concentration uncertainty and apparent-year uncertainty need not be linear.
CFCs: why three compounds are not three copies of one clock
CFC-11, CFC-12 and CFC-113 have different atmospheric input functions, properties, contamination pathways and susceptibility to transformation. A concordant set can support an interpretation. Discordance is a diagnostic signal requiring more work. For example, in reducing aquifers the degradation of certain CFC compounds can produce an erroneously old apparent recharge date if the analyst attributes the reduced amount entirely to an earlier atmosphere.
The common novice move is to average three apparent ages because they came from one laboratory form. That hides exactly the information needed to test preservation. A better table gives compound, measured concentration, reporting uncertainty, corrected equilibrium concentration, apparent recharge year, valid atmospheric branch, and a note about stability or contamination. Mark a result as unresolved rather than inventing an average.
USGS guidance treats agreement among CFC tracers as useful, while presenting patterned mismatches as clues to mixing or degradation. This does not license automated identification from order alone. Consider whether local landfill emissions, industrial spills, older refrigeration equipment or field fittings could alter one analyte. Also check whether a declining atmospheric branch makes a concentration correspond to more than one year.
CFC source contamination versus atmospheric recharge
An unusually high CFC concentration can indicate a source unrelated to historical air at recharge: industrial chemicals, equipment, contaminated groundwater or laboratory cross-contamination. If the inferred equivalent air mixing ratio exceeds physically plausible historical values for the selected location and date range, assigning a negative age or a date in the future is not a scientifically useful repair. The result should instead be flagged as incompatible with an uncomplicated atmosphere-equilibrium model.
Geographical knowledge matters. A well near a historic disposal site may have an entirely different contamination hypothesis from an agricultural monitoring well in an undeveloped recharge basin. A nearby blank or uncontaminated reference well can help assess local sources, but only if the water histories and sampling systems are understood. A contamination explanation should name an independent test: repeat sampling with controlled materials, analyse other compounds, examine major ions or assess land-use records.
Even when CFC concentrations fall within an historical atmospheric range, contamination may still be present. A modest local addition can move a result along a rising or falling segment without triggering an obvious impossibility. Reconciliation across CFCs, SF₆, tritium, noble gases and chemistry is stronger than checking each value in isolation.
SF₆ sources in rocks and engineered systems
SF₆ is useful because it can remain stable in environments where some CFCs degrade, but the gas is not exclusively atmospheric. USGS work documents terrigenic SF₆ in some rock and geothermal settings. Specific volcanic or crystalline geology therefore belongs in the conceptual model. Electrical equipment and certain tracer tests can also introduce anthropogenic SF₆ that is not evidence of recent natural recharge.
If a sample’s SF₆ apparent age is much younger than CFC, tritium–helium-3 and hydrochemical evidence, test for non-atmospheric SF₆. Do not announce a geological source solely because the local rock is volcanic; show supporting data and consider the alternatives. Conversely, a low SF₆ reading could arise from gas loss, analytical limitations or a genuinely older component. Each hypothesis predicts something observable.
The purpose of a tracer comparison is to assign a defensible meaning to discordance. A reader might say: ‘SF₆ indicates a younger atmospheric-equivalent recharge date than CFC-12. Because noble-gas correction and geological sources remain unconstrained, the SF₆ age is provisional.’ That is more useful than a single falsely confident date.
Mixing: a perfectly sensible concentration can imply a misleading date
Consider an aquifer that contributes 40% of pumped water from an old interval with negligible CFC and 60% from a recent interval with dissolved concentration 3.0 arbitrary concentration units. The composite concentration is 0.40×0+0.60×3.0=1.8 units. A piston-flow lookup may match 1.8 to an atmospheric history midway between two calendar years, but no actual parcel need have recharged in that inferred year.
The weights in this example are water-volume fractions, not fractions of dissolved-solute mass. The numerator is the sum of concentration multiplied by volume; the denominator is the total volume. If the flow proportions change with pumping duration, a sequence of bottles from the same well can yield changing gas ages even though the aquifer itself has not suddenly become younger or older.
Tracer mixing is especially important for long-screen municipal wells and fractured rock systems with unequal permeabilities. A valid model might treat an age distribution, a binary mixture or dispersion rather than a single piston-flow age. One cannot uniquely infer the whole distribution from one scalar tracer concentration. Constrain proportions and ages using independent tracers and site hydraulics.
A contrasting hypothetical mixture and its calculation
Suppose sample A has 2.5 femtomoles per litre of a tracer and sample B has 0.5 femtomoles per litre. A mixing volume contains 300 litres of A and 700 litres of B, with no gas loss or gain. Tracer amount, expressed consistently in femtomoles, is 300×2.5+700×0.5=750+350=1,100 femtomoles. Divide by the 1,000-litre total to obtain 1.10 femtomoles per litre.
Some readers average 2.5 and 0.5 to obtain 1.5, which implicitly assumes a 50–50 mixture. The actual fraction of A is 300/1,000=0.30. A second common error is to average apparent ages rather than concentrations; because the input curve is time-varying and nonlinear, that operation has no general physical justification. Date interpretation comes after constructing a mass-balanced mixture of measured substances.
If a conservative co-tracer predicts a different mixing fraction, investigate whether the two endmembers, their assumed compositions or the no-reaction assumption are wrong. Agreement of one gas-derived mixture with one major-ion calculation does not guarantee uniqueness. Multiple combinations of source waters can occupy the same composite chemical point.
Sampling year, recharge year and hydraulic travel time
The calendar subtraction is simple but easy to misuse. An apparent recharge date is a date, not a duration. A sample dated 2026 with a model recharge date 1999 gives an apparent age of 27 years. If the lab instead supplies a corrected recharge date interval 1996–2004, the corresponding age interval is 22–30 years, expressed in ascending age order. Never label the year 1999 as 1999 years of residence.
Travel time from a mapped infiltration area to the well is a more demanding claim. It assumes the gas-equilibration and water-table isolation event coincides with the source-area boundary of the chosen flowpath, which may not hold. Unsaturated-zone transit, recirculation and sampling integration can separate the events. When groundwater movement is slow, mixing and diffusion may blur an apparent age beyond the resolution of a simple subtraction.
Reports should present sampling date and recharge year separately. Time-zone and month detail matter when comparing very young waters, but annual input curves rarely justify daily recharge precision. If a software package returns many decimal places, round only after uncertainty and model identifiability are assessed; precise arithmetic is not the same as precise groundwater history.
What can a tracer date say about a contaminant?
A tracer apparent age can help frame when a water component might have been at the surface, but it does not directly date nitrate, chloride, pesticide molecules or a contaminant release. Different chemicals can move at different rates, degrade, sorb or be introduced along the pathway. An age distribution and source history are needed to connect a present concentration to an earlier land-use event.
Suppose a monitoring sample has an apparent CFC date close to a known agricultural transition. It would be tempting to declare that the nitrate present must come from fields operating in that year. Yet the water may mix recharge dates and nitrate inputs, while nitrogen undergoes nitrification or denitrification. A careful report calls the timing a hypothesis and tests it with nitrate isotopes, excess nitrogen, other age tracers and spatial observations.
The distinction matters in policy and public communication: one result may support the statement that young groundwater is contributing to the well, but not prove the identity of an individual polluter or forecast how quickly a basin will recover. Uncertainty about sources is separate from analytical uncertainty in gas concentration.
Error budget: organize uncertainty rather than attaching one vague percentage
An uncertainty budget should separate laboratory repeatability, field handling, recharge temperature, recharge elevation or pressure, salinity, excess-air model, historical atmospheric curve, and hydrogeological mixing. These do not necessarily combine as independent Gaussian errors. For instance, a wrong recharge temperature can bias all three CFC dates together; it is not reduced by averaging three instrument measurements.
A useful sensitivity table reruns the interpretation at plausible low, central and high recharge temperatures; at alternative excess-air models; and with competing mixing structures. Report which input causes the largest change in inferred recharge interval. If varying one assumption makes a recent date impossible, say so. Quantify uncertainty only within models that remain physically credible.
A numerical uncertainty expression without a model definition is misleading. ‘+/− two years’ might represent laboratory precision alone, or include recharge conditions, or simply reflect the spacing of plotted years. The report should specify. When there is more than one atmospheric matching branch, presenting two plausible windows is better than disguising a multimodal answer inside one symmetric error bar.
A field-to-report checking workflow
Begin with the question: is the intended reader determining a likely recharge decade, comparing sources to an abstraction well, checking recent vulnerability, or estimating an age distribution? Next, document aquifer material, flow directions, screened intervals, pumping regime and potential sources of contamination. These features determine whether a piston-flow age could be representative.
Confirm that sample containers, tubing and preservation matched the laboratory method. Examine blanks, duplicates, detection limits and uncertainty flags. Record the measured CFC/SF₆ values and units before reading any derived dates. Validate the atmospheric dataset and recharge location assumptions, then estimate recharge temperature and excess air independently where possible. Map corrected concentrations to the matching historical input curve and inspect every possible date branch.
Finally, compare tracer ages and indicators of gas or chemical transformation. Document how each competing explanation was tested. The final narrative should lead with what survives the cross-checks, not the software’s most visually prominent number. Include the source and date of input curves, so a later analyst can reproduce the interpretation.
Reading an apparently contradictory field report
Imagine a fictional monitoring well where CFC-12 suggests an apparent recharge interval around the early 2000s, while SF₆ suggests the late 2010s. The most responsible first sentence is that the tracers disagree under the stated piston-flow assumptions. A report should not choose whichever result fits a preconceived picture of land use.
Check four distinct mechanisms. First, does SF₆ contain a terrigenic or anthropogenic excess? Second, was excess atmospheric air included using an appropriate gas model? Third, could CFC-12 have been lost or degraded under the well’s redox conditions? Fourth, does a mixture of old and young water explain both measured concentrations together? Each mechanism implies different additional checks, including noble-gas fitting, field duplicates, redox markers, other CFCs and tritium evidence.
If no hypothesis can be distinguished, the professional conclusion is ‘unresolved apparent-age discordance’. This preserves the observed concentrations and prevents future readers from treating one convenient number as ground truth. Good explanations state the limits of inference with the same clarity as the calculation.
Where translation and technical language matter
The word ‘age’ in everyday English suggests a person’s duration since birth. In isotope hydrology, ‘apparent age’ may mean a model inversion of environmental tracer concentration under a simplified transport assumption. That semantic difference matters when translating an environmental report for students, public agencies or a second-language reader. Preserve the qualifier ‘apparent’ or ‘model’, and explain what was assumed.
Likewise, ‘excess air’ does not necessarily mean a sample was badly collected with an air bubble. It can describe air dissolved during recharge and retained in solution. ‘Recharge’ is water entering the saturated system, not a battery being charged. ‘Concentration’ is an amount per specified volume or mass, whereas atmospheric ‘mixing ratio’ is a fraction. Equating these phrases erases the mechanism that controls the calculation.
Technical vocabulary works best when a term is anchored to an observable, a unit and a decision. This approach connects to eduKate’s tritium–helium-3 groundwater ages guide, which interprets a distinct age mechanism. Readers can contrast the two rather than treating all dating methods as interchangeable.
Independent transfer checks and complete answers
Problem 1: A 2026 sample’s validated atmospheric-equivalent SF₆ input corresponds to a 2012 recharge date, with plausible date endpoints 2009 and 2016. What ages should a report show? Answer: nominal age 14 years; interval 10–17 years because 2026−2016=10 and 2026−2009=17. The 2012 result is an apparent recharge date, not a measured residence-time distribution.
Problem 2: A bottle contains 2.10 femtomoles per litre of tracer. An independent gas fit assigns 0.35 femtomoles per litre to excess air. Under the simple additive model, which amount is used for historical-air matching, and what fraction was corrected? Answer: 1.75 femtomoles per litre; 0.35/2.10=0.1667, or 16.7%. A full age cannot be obtained without temperature, pressure, solubility and atmospheric history.
Problem 3: A well collects 80% water at 0.30 concentration units and 20% at 3.30 concentration units. What is the mixed concentration? Answer: 0.80×0.30+0.20×3.30=0.24+0.66=0.90 units. A simple average, 1.80 units, would misrepresent the 80–20 water balance. Neither result is a directly averaged groundwater age.
Problem 4: CFC-derived dates cluster around 1995 while SF₆ suggests 2018. Name two non-equivalent interpretations and a discriminating observation. Answer: geological excess SF₆ is one interpretation, testable with lithology and independent gas tracers; CFC degradation is another, testable with redox conditions and other CFC patterns. Binary mixing is a third, for which a multi-tracer mass balance can be tested.
Frequently asked questions
Can groundwater dating tell the exact day it rained? No. Most CFC and SF₆ apparent-age procedures cannot resolve a single storm or exact recharge day; input histories, solubility uncertainty and mixing generally dominate that degree of precision. The quoted date should be no more precise than the measurement and model support.
Is a young apparent age always proof of recent recharge? No. An excess non-atmospheric source, excess air not removed, or mixing with a small young high-tracer fraction can create a misleadingly young piston-flow interpretation. Conversely, loss or decay-like degradation of a gas can make some apparent ages look older. Evaluate preservation and hydrologic context.
Are all chlorofluorocarbons reliable in low-oxygen groundwater? No. Some can undergo degradation in reducing settings, and vulnerability differs among compounds. Anoxic conditions warrant explicit diagnostic attention rather than automatically rejecting every compound or accepting all concentrations.
Can I estimate a 2026 age using the 1998 atmospheric chart in an old USGS page? Not safely for young groundwater beyond that input curve. Historical explanations remain useful, but current sampling requires an appropriate atmospheric time series and laboratory model. Never extrapolate a curve with assumed growth rates and then present the result as an institutional data product.
Why compare CFC, SF₆ and tritium–helium rather than buy one test? They answer overlapping but not identical questions and fail under different mechanisms. Agreement can improve confidence; disagreement identifies assumptions needing investigation. The scientific gain lies in a constrained interpretation, not simply in the number of analyses performed.
Limits, decisions and a usable reporting sentence
These tracers generally work best when the atmospheric input curve, gas preservation and sampling protocol support a resolvable date and when the chosen age-distribution model fits the aquifer. Long-screen mixing, chemical degradation, an uncertain gas excess and non-atmospheric sources can prevent unique interpretation. In such cases, describe the observed gas concentrations and model alternatives without publishing a false single-year verdict.
A defensible short conclusion might read: ‘Corrected SF₆ and CFC measurements support a substantial component of post-2000 recharge, but distinct tracer dates and well-screen mixing prevent assignment of one unique groundwater residence time. Dates remain model-dependent until dissolved noble gases and the source contributions are evaluated.’ The exact dates and caveats must be tailored to actual evidence; this is a language example, not a finding.
Readers seeking additional contexts can compare stable water-isotope evidence for evaporation and mixing and chloride mass-balance estimates of recharge. Each addresses a different measurement and inferential question. A mature groundwater explanation keeps those distinctions visible.
Additional audit: detection limits, calibration and model selection
The distinction between a detection limit and a defensible dating limit is often overlooked. A method can detect a tiny positive concentration yet be unable to identify a unique recharge decade because several historical atmospheric years, several gas-loss mechanisms, or several water mixtures would generate a similar amount. Report both the analytical limit of quantification and the interpretive resolution of the selected age model. A ‘detectable but not dateable’ finding can be valuable: it may rule out strictly tracer-free water or signal contamination, even though a calibrated apparent date cannot be supplied.
A positive result should also be evaluated against the blank and the stated handling blank, if available. Consider a synthetic analysis of 0.080 femtomoles per litre, with a laboratory blank of 0.025 femtomoles per litre and a quantification threshold of 0.060 femtomoles per litre. Blindly subtracting the blank gives 0.055 femtomoles per litre, but this corrected value may lie below the threshold and require qualification. Laboratories differ in whether reported measurements are already blank-corrected, so subtracting 0.025 a second time would be a double correction. Ask which quantity the report actually lists.
Another practical problem is calibration mismatch. Historic atmospheric curves may use dry-air mole fractions, while a field team may record ambient humidity, temperature and air pressure without distinguishing dry from wet gas conventions. Since part of total gas pressure is water vapour, the dry-air component must follow the model’s definition. A reader should not build an apparent recharge date from one solubility convention and a different atmospheric reference convention. Record the measured units, atmospheric unit basis, solubility relation, recharge pressure and every conversion. Only then compare the reconstructed mole fraction with a chronology.
Where groundwater is sampled repeatedly, design monitoring around stable interpretation rules rather than interpreting each small date shift as hydrologic change. A 2025 and 2026 sample from a mixed municipal well can have similar apparent dates even while young-water contributions and pumping proportions change; conversely, modest shifts in gas fitting can move the apparent year. For trend work, hold the analytical method and atmospheric model consistent when appropriate, or explicitly reprocess earlier measurements using updated models. Preserve original measured concentrations so historical inferences can be reconstructed.
An age distribution is not automatically better merely because its mathematics is more elaborate. Piston-flow, exponential, dispersion and binary-mixing models answer different questions and require different constraints. With a single concentration one can generally generate multiple age distributions that predict the same observation. Select a complex model only when measured tracers spanning complementary age ranges and independent hydraulic information provide identifiable parameters. A transparent simple model with an explicit warning can be more useful than an unconstrained complex fit with a narrow numerical confidence interval.
Before issuing advice about drinking-water vulnerability or aquifer management, inspect the decision scale. If an apparent CFC date identifies modern contribution to a production well, the decision may be to investigate source pathways or protection measures. It does not, on its own, establish compliance with a contaminant standard, the location of a pathogen source, or a legally sufficient attribution of pollution. Avoid turning a groundwater time indicator into a safety guarantee. For such decisions, combine analytical water-quality evidence, local regulatory requirements, repeated monitoring and the mapped flow system.
Primary methods and reading
USGS: Guidelines for assignment of apparent CFC age and tracer discordance.
USGS: SF₆ sampling, recharge temperature and excess-air effects.
USGS original research: terrigenic SF₆ as a groundwater dating confounder.
Google Search Central: people-first technical writing guidance.