Water isotopes help scientists investigate where water came from and what happened along its route. The measurements compare forms of hydrogen and oxygen within water molecules. They can constrain mixing and evaporation, but they are not a unique address label for every river, cloud or aquifer.
The gold standard is an interpretation that keeps the reference scale, sampling context and hydrological assumptions visible. Begin with a ratio, calculate a simple mixture and then ask which alternative histories could produce the same result.
eduKateSG · Water Isotopes
Find your next learning step
Read a delta value correctly, test a two-source model and recognise when another measurement is needed.
Full chapter index · Practice and transfer · Education root · How X Works
Full chapter index
Read the tracer · Chapters 1–2
Calculate the signal · Chapters 3–4
Test the interpretation · Chapters 5–7
Connect and practise · Chapters 8–10
Choose the next explanation · Chapters 11–12
Isotopes of an element have the same proton number but different neutron numbers. Stable hydrogen and oxygen isotopes occur in water molecules; δ²H and δ¹⁸O are common ways to describe their relative abundances.
A tracer question should be specific: is stream water consistent with a mixture of two sampled sources? Has a stored water body experienced evaporation? Those questions require different comparisons. Write the candidate sources and processes before interpreting the laboratory number.
This guide concerns stable water isotopes. It does not treat every isotope dissolved in water as an equivalent tracer, and it is not a field sampling or laboratory operating protocol.
Let R be a heavy-to-light isotope ratio. Delta notation is δ = (Rsample/Rreference − 1) × 1,000, reported in per mille, ‰. A negative value means a lower ratio than the reference, not a negative number of heavy atoms.
The reference and scale normalisation belong with the result. Two values should not be pooled merely because both carry a δ symbol. Check the isotope pair, reference scale and laboratory reporting convention first.
Do not read −10‰ as ‘ten per cent heavy water’. It means a ratio difference of −1% relative to the specified reference ratio. Ratio, relative difference and abundance are different quantities.
CHAPTER 3 OF 12 · Calculate the signal
3. Worked example: move between delta and ratio
Take a hypothetical normalised ratio Rsample/Rreference = 0.992. Then δ = (0.992 − 1) × 1,000 = −8‰. Conversely, δ = −12‰ gives Rsample/Rreference = 1 − 12/1,000 = 0.988.
The second water has a lower heavy-to-light ratio on the same scale. Its δ value alone does not show that it is older, cleaner or from a particular mountain. Those are additional claims requiring a process model and contextual evidence.
A useful self-check is to set Rsample = Rreference. The formula must return zero. Then try a ratio slightly above one: the delta value must be positive.
For a dilute stable-isotope teaching model, approximate δmix = fδA + (1 − f)δB, where f is the fraction of water from source A. Assume only two sources, comparable water composition and no fractionation during mixing; exact isotope mass balance uses isotope abundances.
Let A be −10‰, B be −4‰ and the mixture be −7‰. Then f = (−7 + 4)/(−10 + 4) = 0.50. The model assigns half the mixture to each source.
Substitution checks the result: 0.5(−10) + 0.5(−4) = −7‰. This is a conditional estimate, not proof that only those two sources exist. A third source could change the interpretation even if the arithmetic stays correct.
CHAPTER 5 OF 12 · Test the interpretation
5. Source separation controls the fraction estimate
Keep the previous endpoints fixed. If the mixture were −7.3‰, f = 0.55; if it were −6.7‰, f = 0.45. A ±0.3‰ change in the mixture produces a ±0.05 change in the inferred fraction in this model.
Now reduce the endpoint separation to 1‰. The same ±0.3‰ mixture change corresponds to ±0.30 in the fraction. Nearly overlapping sources are difficult to distinguish, even with a tidy calculation.
These are sensitivity calculations, not a complete uncertainty budget. Variable source signatures, sampling error and the two-source assumption also matter. If a mixture lies outside the endpoint interval, investigate the model rather than silently forcing the fraction between zero and one.
Evaporation and condensation can partition isotopes differently. A water body’s signature can therefore change without mixing with a new source. This is why isotope hydrology considers both source differences and processes acting along the route.
A conceptual comparison helps: mixing joins contributions, whereas fractionation changes isotope proportions through a process. If both occur, a single delta measurement may not separate their contributions.
Plot paired δ²H and δ¹⁸O data alongside appropriate local precipitation observations and the sampling history. Do not treat a familiar line on a diagram as a universal certificate of origin. The alternative explanations must still be examined.
CHAPTER 7 OF 12 · Test the interpretation
7. Keep the sample and the reported quantity connected
A useful evidence record identifies place, date, water body, sample handling, analytical method and scale. The interpretation also needs to say whether a value represents one sample, a time average or a weighted combination.
For a teaching comparison, suppose event A supplies 10 units of water at −2‰ and event B supplies 30 units at −6‰. The volume-weighted approximation is [10(−2) + 30(−6)]/40 = −5‰, not the unweighted mean −4‰.
The weights describe the question. An average over dates is not automatically an average over water volume. Choose the weights before seeing which answer looks more convenient.
| Question | Needed information | Common overclaim |
|---|---|---|
| Compare isotope ratios | Same isotope pair and scale | A delta value is an absolute abundance |
| Estimate source fractions | Distinct endpoints and a mixing model | Two matching values prove a unique origin |
| Describe a period | Defined sampling and weights | One bottle represents every flow condition |
Stable isotope values can help investigate water sources and histories. They do not decay with a half-life and do not, by themselves, turn a δ¹⁸O reading into a groundwater age.
Radioactive tracers can provide age-related information under their own assumptions. A reported water age may involve mixing and a model of recharge or residence time. It is a different inference from comparing two stable-isotope signatures.
When reading a result, ask which observation supports origin, which supports elapsed time and which supports replenishment. Keep the three claims separate even when they appear in one report.
For a catchment study, draw inputs, storage and outputs before plotting isotope values. A changing river signature could reflect changing source proportions, changing source signatures or processes affecting the water.
For climate evidence, preserve the distinction between the water-derived proxy and the climate quantity inferred from it. Ice-core interpretation also needs chronology and other evidence; a delta value is not a thermometer with a universal conversion.
For a learner, the next move is to identify the missing link: notation, mass balance or process explanation. Repair that step with a small example before adding more sources or a longer time series.
First, calculate δ for a normalised ratio of 1.003. Second, use endpoints −12‰ and −4‰ with a mixture at −6‰. Third, explain why matching a river sample to one rainfall sample does not establish a unique source.
Checks: the first answer is +3‰. In the second model f = (−6 + 4)/(−12 + 4) = 0.25 for the −12‰ source. The final question has no unique numerical answer: temporal variation, mixing and process changes remain possible.
Transfer the reasoning to a mixture measured outside the endpoints. State which assumption you would check first and what new sample or independent tracer could help. An explanation that names the failed assumption is stronger than a clipped percentage.
CHAPTER 11 OF 12 · Choose the next explanation
11. Connected water and climate reading
Build the hydrological model
- Hydrology — connect isotope evidence with water pathways, storage and budgets.
- Isotope Geochemistry — compare water tracing with wider isotope systems and dating questions.
Follow climate evidence
- Ice cores and past-climate evidence — separate water-derived signals, trapped air and chronology.
- Paleoclimatology — place isotope proxies alongside other climate records.
Use the hydrology route when the difficulty is source accounting. Use the climate route when the question concerns what an archived signal can reveal about an earlier environment.
CHAPTER 12 OF 12 · Choose the next explanation
12. Sources, scope and the next learning step
- IAEA: isotope hydrology for water security — stable and radioactive tracers serve different hydrological questions.
- IAEA TECDOC 1723 — delta notation and tracer use in recharge investigations.
- USGS: Lake Tahoe isotope observations — an applied example of paired isotope ratios and mixing evidence.
All numerical values in the worked examples are invented for teaching, not measurements of a Singapore water supply. They illustrate algebra and assumptions rather than certify environmental conditions.
A successful next step is to explain one isotope inference in four sentences: what was measured, what comparison was used, what model linked the two and what remains unresolved. Continue through the Education root for the wider principle of learning that transfers.