When Did This Water Last See the Sky?
An interactive primer on dating young groundwater — the decades-old water that supplies most wells and carries most contaminants. Three clocks that start the moment water loses contact with the atmosphere (³H/³He, the CFCs, SF₆), the corrections that set each clock to zero, the ways each clock breaks, and the lumped-parameter models that turn one concentration into a distribution of ages. Builds on Week 3; no tracer background needed.
A well is dated at 27 years. Fertilizer arrived 15 years ago. Is the nitrate there yet?
A water-supply well is sampled in 2017. Its dissolved SF₆ is 0.93 fmol L⁻¹ — the concentration that water in equilibrium with the 1990 atmosphere would carry, so the textbook reading is "recharged in 1990" — 27 years before the sample was drawn. The farmland upgradient switched to heavy nitrogen fertilizer 15 years ago, and nitrate travels with the water. (The situation is the worked example in Chambers et al.'s Fig. 4.) You are asked whether the well is at risk.
Chambers et al. run exactly this comparison. Give the well a single piston-flow age and its outflow contains no nitrate at all. Give it instead a broad (log-normal) age distribution with the same SF₆ concentration and the same mean — and more than 30% of its water is younger than 15 years, nitrate included. Same measurement, same "age," opposite management decisions. Their conclusion is the week's thesis: "it is the shape of the age distribution that will determine the 'breakthrough' of a contaminant."
The three readings divide the labour: Chambers et al. (2019) review the gas tracers and how their clocks are set and broken; Solomon & Cook (2000) do the same for tritium and its daughter helium-3; Jurgens, Böhlke & Eberts (2012) document the workbook that turns tracer concentrations into age distributions. This page is a map to all three, not a substitute for reading them.
Why the under-100-year window gets its own week
Modern groundwater — recharged within the last ~50–100 years — is, by Chambers et al.'s reckoning, the most abundant and most accessible part of the active water cycle, roughly three times the volume of fresh surface water, and also the part most exposed to fertilizer, industry, land-use change and climate. It sits below the radiocarbon window (Week 3) and above the range where nothing happens fast enough to measure. Three clocks work here, and each starts the same way: the moment water stops exchanging with the atmosphere. Click each one.
Pick a clock to see how it keeps time, what it needs, and what it cannot see.
Five curves the whole method hangs on
CFC-12 and CFC-113 production began in the early 1940s, CFC-11 and SF₆ in the 1950s (refrigeration and air-conditioning, semiconductor manufacture, electrical insulation). They leaked, mixed globally, and rose in the air until the Montreal Protocol bent the CFC curves over in the 1990s; SF₆, a greenhouse gas with no protocol, is still climbing. Tritium's history is a different animal: a cosmogenic background of a few TU, then thermonuclear tests from 1952 that pushed Northern-Hemisphere rain past 1,000 TU — a monthly mean above 5,000 TU at several stations, an annual mean of 3,278 TU at Ottawa in 1963 — followed by a long decay back toward background. Hover the chart.
Input histories, Northern Hemisphere
CFCs and SF₆ in air
Left axis: CFCs · right axis: SF₆ (always ~100× less)
Tritium in precipitation, Ottawa
Log axis. Annual means (Fig. 13.1 plots monthly values); the pre-bomb level is the 8 TU TracerLPM assumes for Ottawa (Solomon & Cook: 3–6 TU in Europe and North America).
Reconstructed histories: smoothed annual curves anchored to the USGS Reston Groundwater Dating Laboratory air curves (the data behind Chambers et al.'s Fig. 1) and to the Ottawa record as summarized by Solomon & Cook (Fig. 13.1). The Southern Hemisphere lags by 1–2 years for the gases, about one year for the tritium peak, and its peak was ~100× smaller (37.7 TU at Kaitoke, New Zealand, in 1964). For real work use the stored histories in TracerLPM.
What the shapes give you
- A rising limb is a one-to-one calendar. Pre-1990 CFC concentrations, and SF₆ to this day, map a concentration to a single year.
- Different shapes are a mixing detector. CFC-12 flattened while SF₆ kept climbing; a mixture of old and young water lands off the piston-flow curve in a CFC-12 vs SF₆ plot (Section 11).
- A spike is a marker. The 1963 tritium peak is a dated horizon you can find in a profile (Section 8).
What the shapes take away
- The Montreal bend. After ~1990 a single CFC concentration matches two years (Section 5).
- A fading input. Tritium in Ottawa rain fell from 3,278 TU (1963) to 18.5 TU (1993); the Southern Hemisphere never had much. A lone ³H value rarely dates anything now.
- Tiny numbers. Atmospheric equilibrium means picomoles per litre (CFCs) and femtomoles per litre (SF₆): the detection limits Chambers et al. quote are 0.01 pmol L⁻¹ and 0.1 fmol L⁻¹, and the same smallness makes contamination trivially easy (Section 6).
Henry's law, and the four numbers you must know about the recharge zone
A gas tracer in water tells you nothing until you convert it back to the air it came from. That conversion is Henry's law — dissolved concentration = solubility × partial pressure — and the solubility depends on temperature (colder water holds more gas), the partial pressure on elevation (thinner air holds less), a little on salinity, and the whole budget on excess air: bubbles trapped and forced into solution as the water table rises, adding gas in atmospheric proportions on top of equilibrium. Chambers et al. call this setting the groundwater-dating "clock" to zero, and every age you read afterward inherits whatever you assumed here.
The clock-setter
Choose a tracer and the recharge conditions. The chart shows the concentration water would carry if it recharged in each year — the curve you match a sample against. Then set a measured concentration and read the recharge year(s) it implies. Move the temperature 2 °C and watch the year move.
Dissolved CFC-12 expected in water recharged each year
Temperature and elevation
Recharge temperature is best fixed from other dissolved gases — N₂/Ar or the noble gases (Week 7) — and in a pinch from shallow groundwater temperature. Chambers et al.: an uncertainty of ±2 °C typically shifts the age of a pre-1990 sample by less than 3 years. Elevation lowers pressure exponentially, but the lapse rate (6.5 °C km⁻¹) cools the recharge and raises solubility, so the two corrections tend to cancel rather than compound. Salinity at infiltration is usually near zero; even 1,000 µS cm⁻¹ water (<1‰) changes the answer by under 1%.
Excess air and its opposite
A few cm³ L⁻¹ of excess air is typical of sedimentary aquifers, but it varies, and it matters most for the least soluble gases — SF₆ and ³He — where a small bubble is a large fraction of a tiny equilibrium amount (compare the CFC-12 and SF₆ curves of Fig. 3 in the clock-setter above). The reverse problem, degassing, is increasingly common: methane from anoxic organic matter or N₂ from denitrification forms a gas phase below the water table that strips dissolved gases — SF₆ can be lost entirely. Both are diagnosed with the noble gases, which is why Chambers et al. recommend them as the standing companion measurement.
The Montreal bend, and why you measure three gases
Chambers et al.'s Fig. 1 makes the point with a single red line: a CFC-11 concentration a little below the peak matches the rising limb in the late 1980s and the falling limb in 2005. One measurement, two recharge years. Their fix is not a better measurement but a second tracer with a different shape — SF₆, still rising — whose value picks out the right branch (their dashed line, drawn for a recharge year of 2005). (All of this still assumes piston flow; mixing is Section 9's problem.)
Which branch?
CFC-11 (left) and SF₆ (right), dissolved at 10 °C, sampled 2017
Non-correctable, but diagnosable
Temperature, pressure and salinity are corrections; Chambers et al. reserve a harsher word — non-correctable — for the processes that add or remove tracer after recharge. You cannot fix them with arithmetic. You can only detect them, usually because three gases with different chemistries stop agreeing. Learn the signatures, then try the diagnosis board.
addsContamination
Equilibrium concentrations are so small (pmol L⁻¹) that "only minor contamination is required to render the CFCs somewhat meaningless in terms of tracers of groundwater age." Chambers et al.'s scale-setter: one-tenth of the CFC-12 in a refrigerator of older design could push a moderately sized aquifer above ten times modern atmospheric equilibrium. Sources are industrial activity and leaking landfills, from the air or from below — worst in urban and peri-urban aquifers where fractured horizons carry them fast. The sample reads "younger than today," which is impossible, or the CFCs disagree wildly with each other. SF₆ is far less prone (high-voltage switchgear, Mg and Al smelting, landfills), but has its own problem, next.
addsTerrigenic SF₆
Rocks make SF₆. Anomalously high concentrations occur in some sedimentary aquifers and in aquifers with fluorite or metallic-sulphide mineralization — "a far greater concern" for SF₆ than industrial leaks. The signature is SF₆ above any possible atmospheric equilibrium while the CFCs look sane, or an SF₆ "age" much younger than the CFC ages. (Old, tracer-free groundwater with measurable SF₆ is the giveaway.)
removesMicrobial degradation
Under anoxic conditions the CFCs are consumed, CFC-11 typically faster than CFC-12. Under strongly reducing conditions they vanish below detection. So the field parameters — dissolved oxygen, redox potential — belong on the sampling sheet, and a CFC-11 age older than the CFC-12 age of the same sample is the classic tell. A detectable DO does not clear you: redox can change along the path, and a mixed sample can carry oxic water from one flow line and stripped water from another. SF₆ does not degrade.
removesSorption, degassing, the vadose zone
Sorption is minor except in organic-rich matrices, and increases in the order CFC-12 < CFC-11 < CFC-113. Degassing (methanogenesis, denitrification) strips the least soluble gas first — SF₆ can go to zero — and is caught by the noble gases. A thick unsaturated zone delays the arrival of the atmospheric signal at the water table, by an amount set by each gas's diffusion coefficient, solubility and the soil moisture; gas-phase diffusion is fast enough that in most practical situations the lag is negligible, but "most" is doing work in that sentence. Thermal effects (CFC-12 preferentially lost) have been seen at two springs, in Belgium and the UK, and nowhere else.
The diagnosis board
Each row is a pattern in the three-gas data from one well. Pick the most likely explanation. (One process per row is the best fit; the feedback explains the runner-up.)
³H/³He: a clock that carries its own zero
Tritium is hydrogen — it is part of the water molecule, so it goes exactly where the water goes and reacts with nothing. It decays by beta emission to ³He with a half-life of 12.32 years (Lucas & Unterweger, 2000; Solomon & Cook's chapter uses the older 12.43). The trouble with tritium alone is the input: to turn a concentration into an age you must know what the rain contained, and the rain's record is a decaying spike. Tolstikhin & Kamensky's 1969 idea sidesteps the whole problem. Measure the daughter too. If the ³He produced by decay stays in the water, the ratio of daughter to surviving parent depends only on elapsed time:
The catch is in the asterisk. Groundwater holds ³He from three other sources — atmospheric solubility, excess air, and nuclear reactions in the rock (plus mantle helium in volcanic terrain) — and each must be subtracted before the ratio means anything. One tritium unit is one ³H¹HO molecule per 10¹⁸ molecules of water; when a TU of tritium has fully decayed it has produced 2.487 pcm³ (STP) of ³He per kilogram of water — against an atmospheric-equilibrium background of about 63.7 pcm³ kg⁻¹ at 10 °C, and up to ~100 pcm³ kg⁻¹ more from excess air. Small signal, large background: the accounting is the method.
The ³H/³He clock
Set the tritium the water carried when it crossed the water table and let it age. The bars show what is in the water; the readouts show the apparent age the ratio returns — and how much an unrecognized cm³ of excess air would corrupt it. Solomon & Cook's worked example (20 TU, 2 years vs 30 years) is the first preset.
What the water holds (TU-equivalents)
Parent, daughter and their sum through time
The helium ledger (eqs. 13.3–13.6)
³Hetotal = ³Heatm + ³He* + ³Henuc + ³Heman, and ³Heatm = ³Hesol + ³Hee. The solubility term comes from temperature (helium solubility changes by only about 0.5% per °C near 10 °C, and the dissolved ³He/⁴He ratio, 1.36 × 10⁻⁶, sits just below air's 1.384 × 10⁻⁶). The excess-air term is fixed from neon, which has no source but the atmosphere. The nucleogenic term rides on radiogenic ⁴He: crustal helium has ³He/⁴He of order 10⁻⁸, a hundred times below air, so it matters only when ⁴Herad is large — easy to see, hard to convert. With no radiogenic helium and no excess air the whole ledger collapses to ³He* = ⁴Hem(R₀ − Rsol): the measured ³He/⁴He ratio minus the equilibrium one, times the helium you measured.
Where it is sensitive, and where it is not
- Young water is fragile. Two-year-old 20-TU water holds 5.25 × 10⁻¹² cm³ kg⁻¹ of ³He* (5.29 with the 12.32-year half-life the demo uses) — about 10% of the 6.4 × 10⁻¹¹ dissolved from the air. One cm³ kg⁻¹ of excess air adds 7.2 × 10⁻¹², as much as the signal.
- Bomb-era water is robust — because of what it started with. At 30 years the same water's ³He* has grown to 4.0 × 10⁻¹¹, comparable to the solubility background, so "the sensitivity of ³H/³He ages to determinations of atmospheric ³He decreases the older the water becomes." The error from an unrecognized bubble, though, scales with 1/³H₀: Cook & Solomon's figures — roughly −5 years per cm³ kg⁻¹ of excess air for very young water, falling to −0.25 by 25 years — fall because water recharged in the early 1970s carried ten times the tritium of today's rain.
- The clock does not start until the water table. In the vadose zone, helium exchanges freely with soil gas, so ³He* is lost; Solomon & Cook's Fig. 13.11 shows ³H/³He ages offset from total travel time by exactly the unsaturated-zone transit.
- The clock can leak. Below the water table, ³He can diffuse back up and out. With an effective ³He diffusion coefficient of 1.3 × 10⁻⁴ m² day⁻¹, loss is under 20% if the vertical velocity exceeds ~0.1 m yr⁻¹ (≈30 mm yr⁻¹ recharge at 30% porosity), under 1% above 0.5 m yr⁻¹; below ~0.01 m yr⁻¹ diffusion dominates and little travel-time information survives (Fig. 13.9). Slow recharge, no clock.
- Mantle helium lies. Its ³He/⁴He is ~10× atmospheric, indistinguishable from decay. Rare in young groundwater; a real problem near young volcanic rock and in geothermal water.
Watching the bomb peak sink
The 1963–64 rain is a dated layer. Smith et al. found it 4 m down in the English Chalk in October 1968 — a 600-TU spike, four to five years after the peak, hence 0.9 m yr⁻¹ downward. Andersen & Sevel tracked it through 22 m of Danish outwash at 4.5 m yr⁻¹. At Sturgeon Falls, Ontario, Solomon et al. (1993) found the ³H and ³He peaks together at 11 m in 1991 — with the ³H peak shrinking between the 1986 and 1991 samplings while the ³He peak grew. The elevator below animates the idea: profile shape from the input history, decay in the parent, ingrowth in the daughter, and a ³H/³He age line whose slope is the velocity.
The bomb-peak elevator
One-dimensional downward flow at velocity v below a water table, driven by the Ottawa input; the unsaturated-zone lag shifts the ³H/³He clock. (The Smith et al. Chalk profile was in the unsaturated zone; here its 0.9 m yr⁻¹ is run below a water table.) Presets transcribe the reported velocities; dispersivities and lags are illustrative. Drag the sampling year forward and watch the parent fade, the daughter grow, and their sum — which "effectively corrects for radioactive decay" (Schlosser et al., 1988) — keep the peak readable for decades longer than tritium alone.
Concentration vs depth below the water table
Age vs depth
Three ways tritium alone has been read
- Peak displacement (Fig. 13.2–13.3): depth of the marker ÷ time = velocity; × water content = recharge. Fails where vapour transport or root uptake moves the tracer without moving the water, which is why Tyler & Walker insist on velocities measured below the root zone, and why arid profiles often carry their maximum tritium right at the surface.
- Mass balance (eq. 13.1): total tritium stored in the profile ÷ decay-corrected tritium that fell = the fraction of rain that recharged. Akrotiri, Cyprus: 75.2 TU m in the ground against 849 TU m fallen → 36 mm yr⁻¹, versus 48–63 mm yr⁻¹ from peak displacement — "reasonably in agreement," which is also a statement about the method's precision. No flow assumption needed, but the local input must be known accurately — Atakan et al. put the error in estimating local fallout at a German site near 20%.
- Discharge models (Fig. 13.6): for a spring or open borehole, a well-mixed-reservoir model relates tritium to recharge/volume — Polda Basin, 1.0–4.9 TU → R/V of 0.005–0.02 yr⁻¹ → ~30 mm yr⁻¹. Time series beat single values: Siegenthaler's spring falling from 290 to 230 TU over 3 years fit an exponential model with a 13-year mean residence time. A single value is non-unique (the Wairau River's decay line cuts the input curve twice, Fig. 13.5).
What ³H/³He added
- Sturgeon Falls. The 1963 peak at 11 m plus a 2-D flow model gave recharge of 0.15 m yr⁻¹; the ³H/³He age gradient alone gave a vertical velocity of ~0.45 m yr⁻¹, which at porosity 0.35 is 0.16 m yr⁻¹. Same answer — but the second needed only a few samples near the water table, not a profile deep enough to find the peak.
- Szigetköz, Hungary. Bank infiltration from the Danube makes flow one-dimensional and horizontal; apparent ³H/³He age vs distance gives ~530 m yr⁻¹ (Fig. 13.12), from ordinary long-screened wells.
- Cape Cod. A contaminant plume marks the flow path; ages along it give 90 ± 20 m yr⁻¹, date the spill (1975 ± 3), and locate its source to ±250 m — neither the time nor the place of the release was known beforehand (Fig. 13.13). Portniaguine & Solomon then inverted ages and heads together: heads constrain recharge/conductivity, ages constrain recharge/porosity; jointly they pin conductivity to ~50%, far better than pumping tests.
- Dispersion is the caveat. ³H and ³He diffuse at rates differing by ~4×, so hydrodynamic dispersion separates apparent age from travel time — mostly at and below the bomb peak (Fig. 13.10): minimal on Long Island sand, ~35% near the peak in Ontario silty sand.
The lumped-parameter zoo
Every age in Sections 4–8 assumed piston flow: one parcel, one path, one date. Real samples come from screens metres long, from springs where flow lines converge, from aquifers that disperse. Lumped-parameter models (LPMs) keep the arithmetic simple and drop the fiction: the aquifer is a black box whose only property is its exit-age distribution g(t) — the fraction of the sample that is each age. Maloszewski & Zuber gave the field its classic shapes in 1982; TracerLPM ships five, and lets any two be blended. Click a model to see the aquifer it stands for; then push its parameters.
Age-distribution explorer
The distribution g(t) for the selected model, on a fixed 0–150-year axis so that every slider visibly reshapes it. The readouts answer the Section 1 question directly: how much of this water is younger than 15 years?
Exit-age distribution
How a distribution becomes a concentration — and a concentration becomes a forecast
The convolution integral is the hinge of the week. Read it right to left: for every parcel in the sample, look up what the atmosphere (or the rain) held the year that parcel recharged, decay it for the years since, weight it by how much of the sample is that old, and add. Chambers et al.'s Fig. 4 draws it as a box with an input curve going in and a number coming out. The machine below lets you turn the crank.
Turn the crank
Pick a tracer, a model and a sampling year. The upper panel shows the input history; the shaded curve is the age distribution laid on the calendar (the sample's parcels, by recharge year); the lower panel is their product — the pieces that add up to the measured concentration.
Input history and the sample's parcels
Contribution of each recharge year to the sample
Mixtures and apparent ages
Blend water of one age with tracer-free old water, or with water of a second age, and ask each clock what it reads. The ³H/³He clock is a ratio, so diluting with tritium-free water does not move it; the calendar clocks read the diluted concentration as older. Neither is wrong about the water it can see.
Reading mixing without a model of the aquifer
Because the input histories have different shapes, every model traces a different path through a plot of one tracer against another, with age ticking along each path. Chambers et al.'s Fig. 5 (CFC-12 vs SF₆) shows the piston-flow "bow," the binary-mixing straight line from tracer-free water to modern recharge, and the exponential family between them. A sample lands somewhere on this map, and where it lands is a statement about its age distribution — obtained, in Chambers et al.'s words (citing Gooddy et al., 2006), "without any prior knowledge of the physical properties of the aquifer system." This is the tracer-tracer method that TracerLPM's first workgroup is built around.
Tracer–tracer explorer
Choose the axes and the sampling year. Curves are the model outputs for mean ages 0–200 years (dots every 10 years, labels every 20). The dashed line is binary mixing between tracer-free water and water recharged in the sampling year, in 20% steps. Set a sample with the sliders and read the nearest model's mean age — TracerLPM's "lookup mean ages," by eye.
CFC-12 vs SF₆, dissolved at 10 °C
TracerLPM: the argument, operationalized
Jurgens, Böhlke & Eberts wrote TracerLPM for the USGS's "Transport of Anthropogenic and Natural Contaminants to supply wells" studies, after finding that piston-flow ages of public-supply wells "are often misleading," that contaminant forecasts built on distributions differ substantially from those built on single ages, and that LPMs calibrated to tracers can reproduce particle-tracking distributions from full flow models at a fraction of the cost. It is an Excel workbook with a compiled add-in that evaluates the convolution monthly for each model. Its worksheets encode a method, and the method is the point.
The workflow, worksheet by worksheet
- Corrections come first. The Samples sheet expects concentrations already corrected for recharge temperature, elevation, excess air, terrigenic helium and degradation — Sections 4 and 7 are prerequisites, not options. Gases are entered as the atmospheric mixing ratio (pptv) they were in equilibrium with, which strips local temperature and elevation out of the comparison.
- Conceptualize, then look, then fit. Start with PFM and EMM — one parameter each — and see where the sample falls. Off both curves, try EPM, PEM or DM; off all of them, suspect a binary mixture or a broken tracer. Only then run the fit, with bounds no wider than ~20 years on age and ~0.2 on a shape parameter, because "multiple local minima in the residual errors … can exist."
- Count your unknowns. At least p + 1 tracers for p parameters: PFM/EMM need 2, EPM/PEM/DM need 3, a BMM-DM-DM needs 6 — usually impossible at one well, hence the habit of borrowing constraints from nearby wells.
- Tritium is four tracers. Enter ³H once and the workbook derives ³Hetrit (eq. 16), initial tritium ³H₀ = ³H + ³Hetrit, and the ratio ³H/³H₀ — each with a different sensitivity to the distribution.
- Total mean age = saturated-zone mean + UZ travel time (eq. 22), with the UZ lag settable per tracer, because soluble ³H rides the water down while the CFCs can outrun it through soil air.
Forecasting: why the shape decides the future
A nitrate history of the Modesto type — rising loads from the 1950s, held constant from 2000, then cut to zero — pushed through two calibrated distributions. Jurgens et al.'s two examples had opposite tempers: the Modesto supply well (a PEM with mean age ~65 years and no water younger than the screen's depth allows) responds late and keeps rising after inputs stabilize; the upper Missouri River (a BMM-EMM-PFM: 84% groundwater with a 4.3-year mean, 16% prompt runoff) responds at once and then trails. Drag the cessation year and compare.
Nitrate at the water table (dashed) and in the discharge (solid)
Three readings, three blocks, one method
Chambers et al. supply the gas clocks and the discipline of setting them; Solomon & Cook supply the tritium clock and the field cases that made it credible; Jurgens et al. supply the instrument that turns concentrations into distributions. The guides below are maps, not substitutes: what each asks, where its load-bearing figures live, which equations to recognize, which numbers to be able to quote, and where a careful reader might press. The notes and the critique are yours to write.
Chambers, L. A., Gooddy, D. C., & Binley, A. M. (2019). Use and application of CFC-11, CFC-12, CFC-113 and SF₆ as environmental tracers of groundwater residence time: A review. Geoscience Frontiers, 10(5), 1643–1652.
Ten open-access pages (CC BY-NC-ND) that walk from sampling to the meaning of "age," recommending complementary techniques at every turn. doi:10.1016/j.gsf.2018.02.017
Solomon, D. K., & Cook, P. G. (2000). ³H and ³He. Ch. 13 of Cook & Herczeg (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer, pp. 397–424.
The canonical method chapter: sources and units, sampling, four ways to read tritium alone, then the ³H/³He method with its helium ledger, its limitations, and five field applications. doi:10.1007/978-1-4615-4557-6_13
Jurgens, B. C., Böhlke, J. K., & Eberts, S. M. (2012). TracerLPM (Version 1): An Excel® workbook for interpreting groundwater age distributions from environmental tracer data. U.S. Geological Survey Techniques and Methods 4-F3, 60 p.
Public domain. Model theory (pp. 3–9), tracer notes (pp. 10–14), a worksheet-by-worksheet manual (pp. 14–35), and three worked examples (pp. 35–48) whose numbers are the ones to know. doi:10.3133/tm4F3
Four situations to reason through
Each scenario gives the kind of evidence a Week 5 reader should now be able to interrogate. Choose the reading you would defend first, then compare with the debrief.
Ten questions before the meeting
Immediate feedback, no grade, no record. If you miss one, the linked section is the fix.
Glossary
The Week 5 working vocabulary, with a few terms you will meet again in Weeks 6, 7 and 11. Search or browse.
The Week 5 readings — and two optional companions
Papers are not posted for this course: retrieving them from the citation and DOI — via the UGA Libraries, GALILEO, or the publisher — is part of the training. Two of this week's three are open access, and the workbook itself is a free download from the USGS.
Chambers, L. A., Gooddy, D. C., & Binley, A. M. (2019). Use and application of CFC-11, CFC-12, CFC-113 and SF₆ as environmental tracers of groundwater residence time: A review. Geoscience Frontiers, 10(5), 1643–1652.
Sampling and analysis; setting the clock (temperature, excess air, pressure, salinity); the non-correctables (contamination, terrigenic SF₆, degradation); three concepts of age; LPMs and tracer–tracer plots; links to flow models and hydrochemistry. Open access. doi:10.1016/j.gsf.2018.02.017
Solomon, D. K., & Cook, P. G. (2000). ³H and ³He. In Cook & Herczeg (eds.), Environmental Tracers in Subsurface Hydrology (pp. 397–424). Kluwer.
Tritium sources and units; profiles, mass balance and discharge models; the ³H/³He age equation and helium ledger; excess air, dispersion and ³He confinement; Sturgeon Falls, the Danube, Cape Cod. doi:10.1007/978-1-4615-4557-6_13
Jurgens, B. C., Böhlke, J. K., & Eberts, S. M. (2012). TracerLPM (Version 1): An Excel® workbook for interpreting groundwater age distributions from environmental tracer data. USGS Techniques and Methods 4-F3.
Five LPMs plus binary mixtures, tracer notes, the worksheet manual, and the Modesto, Albuquerque and Missouri River examples. Public domain; workbook and example files download with the report. doi:10.3133/tm4F3
Gilmore, T. E., et al. (2021). The ³H/³He groundwater age-dating method and applications. · Busenberg, E., & Plummer, L. N. (2000). Dating young groundwater with sulfur hexafluoride: natural and anthropogenic sources of sulfur hexafluoride. Water Resour. Res., 36(10), 3011–3030.
Gilmore and colleagues update the ³H/³He method two decades on, with applications; Busenberg & Plummer is the foundational SF₆ paper — solubility, the atmospheric record, and the terrigenic sources that Section 6 warns about. Not discussed in a block.
Next week — contaminant source identification I: nitrate and stable isotopes
The nitrate that this week's forecasts moved around gets its own fingerprint: δ¹⁵N and δ¹⁸O of nitrate, the dual-isotope cross-plot that apportions fertilizer, manure and soil sources, and the predictable slope along which denitrification drags a sample. Everything this week said about age distributions now applies to when a nitrate signal arrived; next week is about where it came from.
Carry-forward question, for your proposal one-pager (due Week 6): if your question needs the age of water younger than ~60 years, which of this week's clocks can see it — and what would you measure alongside it (noble gases? dissolved oxygen? a second gas? a time series?) so that a skeptic could not dismiss the reading in one sentence?