Marine εNd Neodymium Isotopes: Water-Mass Provenance

Marine neodymium isotopes can help scientists investigate where ocean water has been and which continental sources influenced it. Yet a sediment-derived εNd value is not a direct measurement of ocean current speed, geographic distance or elapsed travel time. The laboratory determines a radiogenic isotope ratio involving neodymium-143 and neodymium-144 in material thought to preserve a seawater-related signature. Researchers translate that ratio into epsilon notation and compare it with potential water-mass sources. Every stage, from which mineral phase was dissolved to how ocean boundary exchange operated, affects what the number can mean.

Did you know that two water masses mixed in equal volumes need not produce a neodymium isotope value halfway between their εNd numbers? The waters may contain very different dissolved Nd concentrations, so they contribute unequal amounts of the actual element carrying the isotopic signature. A small volume of Nd-rich water can have disproportionate influence on the measured ratio. In addition, neodymium can be exchanged with particles and sediments along a flow path. This makes isotope provenance powerful but not perfectly conservative.

This guide begins with the isotope definition, then explains the marine sources and analytical phases, builds checked mixing examples, discusses measurement and preservation uncertainty, and tests common interpretation errors. The ratios and concentrations in worked exercises are explicitly hypothetical. Source papers and primary studies are linked at the end so readers can distinguish a teaching calculation from an observation of a particular ocean basin. Our central rule is to separate isotope measurement, carrier material, water-mass inference and physical circulation mechanism.

What is radiogenic neodymium?

Neodymium is a rare-earth element with several naturally occurring isotopes. In geological materials, the isotope neodymium-143 is partly produced through radioactive decay of samarium-147 over immensely long time scales. Different rock reservoirs have distinctive histories of parent-daughter fractionation and isotopic evolution. As continental rocks weather and interact with water, their neodymium can help imprint regional isotope signatures on marine reservoirs. The resulting dissolved seawater pattern reflects sources and exchanges as well as circulation.

The relevant laboratory ratio is 143Nd/144Nd: the abundance of neodymium-143 divided by the abundance of neodymium-144, determined by an appropriate mass-spectrometric method with standards and corrections. Because the numerical ratio varies only slightly between many environmental samples, researchers commonly report epsilon neodymium, written εNd, for convenient comparison. Epsilon is a scaled relative difference from a reference isotope ratio. It is dimensionless, even though scientists often discuss its value as “epsilon units.”

A frequently used reference is CHUR, the Chondritic Uniform Reservoir, with 143Nd/144Nd conventionally taken as approximately 0.512638 in this calculation. The epsilon expression is εNd = [(Rsample/RCHUR) − 1] × 10,000, where R is the isotope ratio. A negative result means the sample’s ratio is lower than the selected CHUR reference; a positive result means it is higher. The sign does not mean that the water contains negative neodymium or that the current flows in a negative direction.

The factor 10,000 matters. Epsilon units are parts per ten thousand of the fractional ratio difference, not per thousand. They should not be confused with δ13C or δ18O, which use different isotope systems, standards and conventional scaling. Even though all three are ratios, their environmental mechanisms differ. An εNd value of −10 does not say that seawater lacks ten per cent of an element; it says its 143Nd/144Nd isotope ratio is one-thousandth lower than the reference ratio.

Calculate epsilon from a raw ratio

Take a hypothetical sample with 143Nd/144Nd = 0.512125362 and use RCHUR = 0.512638. The sample ratio equals 0.999 times the reference. Therefore εNd = (0.999 − 1) × 10,000 = −10.0. The arithmetic is exact for these intentionally selected figures. An instrument would carry uncertainty, and standards may involve additional correction or normalization choices. The meaningful information is the relative isotopic composition, not a direct count of water-mass kilometres.

If a second hypothetical sample has εNd = −5.0, reverse the equation: Rsample = RCHUR × (1 + εNd/10,000). Substituting −5 gives 0.512638 × 0.9995 = 0.512381681. This inverse calculation is useful for checking spreadsheets and for converting published epsilon values back to the isotope-ratio scale. Always use the same reference convention when comparing samples. A change in reference choice can move reported epsilon values slightly without changing the physical sample.

An epsilon difference of three units is not the same as three per cent. It represents an isotope-ratio difference of roughly three parts in ten thousand relative to the reference, subject to the exact ratio values. Reporting a comparison as “the Nd concentration rose by three” is also wrong: εNd and dissolved Nd concentration are different variables. One describes isotopic composition, the other describes the amount of Nd in the water or sample. Both are often necessary for an interpretable mixing calculation.

A useful way to check a reported isotope conversion is to estimate whether the sign makes sense. If Rsample is below the CHUR ratio, εNd must be negative. If it is above, εNd must be positive. If a spreadsheet returns a large positive number for a lower raw ratio, inspect the subtraction direction and scale factor. If a sample near the reference produces thousands of epsilon units, check whether a decimal point or percent conversion was misplaced. Unit discipline prevents spectacular but preventable errors.

Where seawater receives its neodymium

Continental weathering, rivers, dust and interactions with continental margins can supply neodymium with different isotopic signatures to the sea. Regional geology matters because young volcanic terranes, old continental crust and mixed sediment sources need not have the same 143Nd/144Nd ratios. The resulting seawater imprint can help distinguish influences from different source regions. But geology does not map to water in a single direct step: transport, scavenging, dissolution and local exchange affect the final dissolved Nd mixture.

Ocean circulation then moves water containing those signatures. A particular water mass may carry a distinctive isotope composition from its source region while also exchanging Nd with surrounding particles and sediments as it travels. The degree of conservation varies with ocean setting and processes. Therefore an apparent match between seawater εNd and a continental source does not prove that a water parcel flowed directly from that source without intermediate exchange. A provenance claim needs supporting hydrography and an understanding of chemical cycling.

A water-mass tracer is not automatically a current meter. It may distinguish contributions from different reservoirs without measuring transport velocity. Suppose a site shifts from εNd = −12 toward −7 over time. One possible explanation is increased influence of a more radiogenic end member. Another is intensified exchange with local sediment or a changed riverine input. A third is altered particle cycling. The observation constrains the interpretation but does not select the cause by itself.

Water-mass end members are not permanent constants. The isotope character of a formation region can evolve as weathering regimes, dust inputs or marine boundary exchange change. When comparing ancient and modern samples, researchers should avoid assuming every source water had an identical past εNd signature. Where necessary, a study can use coeval regional records, multi-proxy constraints or model ensembles rather than inserting present-day values as unquestioned geological boundary conditions.

Why neodymium concentration matters

Isotopic mixing is tied to the amount of neodymium contributed by each reservoir. Imagine one kilogram of water A with dissolved Nd concentration of 10 picomoles per kilogram and εNd = −12, and one kilogram of water B with 30 picomoles per kilogram and εNd = −4. The masses of water are equal, but B contributes three times as much Nd. Under a small-difference approximation, the mixture εNd is [10 × (−12) + 30 × (−4)]/(10 + 30) = −6.0. A naive equal-water average would give −8.0.

The same principle applies with unequal water fractions. Suppose water A comprises 75 per cent of a mixture at Nd = 10 pmol/kg and εNd = −12, while water B comprises 25 per cent at Nd = 30 pmol/kg and εNd = −4. Their relative neodymium contributions are 0.75 × 10 = 7.5 and 0.25 × 30 = 7.5. The approximate mixed epsilon is therefore −8.0, although the water itself is three quarters A. Equal isotope carrier contributions can emerge from unequal water volumes.

For rigorous work, isotope amounts and ratios should be conserved explicitly rather than assuming epsilon values mix exactly linearly. Because the isotope-ratio differences are small, concentration-weighted epsilon mixing is commonly a useful first approximation for educational calculations. Yet the approximation has a domain: it presumes that the measured concentrations represent the relevant dissolved pools and that no additional sources or sinks operate during mixing. If scavenging or particle exchange is important, a simple closed mixture may be inappropriate.

Now suppose the Nd-rich end member is also chemically altered after mixing. Particle scavenging might remove some Nd, or sediment dissolution might add Nd with a new isotope signature. The measured concentration and epsilon could then reflect both hydrodynamic mixture and chemical exchange. One isotope observation plus one concentration measurement may still be insufficient to solve for all unknown processes. A physically plausible inversion identifies the sources, sinks and independent constraints rather than forcing a two-end-member answer.

What is an authigenic phase?

In sediment chemistry, “authigenic” describes material formed or modified in place within the depositional environment, as distinct from detrital mineral grains eroded from preexisting rock. Seawater-derived iron–manganese oxyhydroxides and some biogenic particles can retain information about seawater-related neodymium. Researchers may extract Nd from fish teeth, foraminiferal coatings or selectively leached sediment phases, depending on the scientific setting and method. The critical question is which reservoir the analysed material actually sampled.

A detrital fraction might carry Nd from continental rocks deposited as particles, while an authigenic coating may record Nd exchanged with bottom or pore waters. Both can be geologically interesting, but they answer different questions. If a study intends to reconstruct deep-water provenance and accidentally dissolves detrital silicate grains, the resulting isotope ratio can be driven toward the sediment source rather than the ancient water. The laboratory has not merely introduced noise; it may have changed the identity of the archive.

The problem is particularly important where sediments include volcanic ash or reactive lithogenic material. A 2025 investigation of Adriatic samples examined extraction methods and demonstrated how tephra-associated material can shift some leachate signatures. Such studies remind us that the chemical selectivity of a leach solution is empirical and site-dependent. An acid or reducing reagent does not carry an inherent label saying “seawater only.” The phase dissolved must be evaluated through chemistry, mineralogy, extraction tests and comparisons with other carriers.

Why the leaching protocol matters

A selective leach is intended to dissolve target coatings or other seawater-related phases while limiting attack on unwanted detrital material. Reagent strength, pH, duration, temperature, sample grain size and mineralogy influence what dissolves. A procedure validated in one sediment type may behave differently in a volcanic or clay-rich core. Authors should document the exact extraction sequence and its validation instead of presenting “leached Nd” as though it were a self-explanatory environmental category.

Suppose a weak extraction yields εNd = −9 and a stronger extraction yields −5 from the same sediment layer. The difference could mean that the stronger reagent mobilized an additional radiogenic detrital component, but it might also reflect another authigenic phase or a genuinely mixed pore-water archive. One cannot decide solely from the direction of the shift. Measure associated element concentrations and phase indicators, examine mineralogy and compare independently prepared samples. Different leach results are evidence about the material’s chemical compartments.

A leachate with very low total Nd can be particularly vulnerable to laboratory blanks or small contaminant contributions. If a sample carries ten picomoles of recovered Nd and a blank adds one picomole with a different isotope signature, the blank is ten per cent of the sample amount. Its isotopic leverage is not necessarily trivial. Blank correction and uncertainty should be documented alongside the extraction yield. The ratio can be precisely measured yet reflect a mixture that includes unwanted laboratory-derived Nd.

Pore waters, bottom waters and boundary exchange

Authigenic phases do not always register bottom-water εNd exactly. Sediment pore waters can exchange neodymium with particles, and their composition can differ from the overlying ocean. Reactive boundaries can modify dissolved signals over time. A study in the Gulf of Alaska, for example, investigated differences between authigenic phases and bottom waters and considered exchange processes. The general lesson is that proximity to the seafloor is not proof of complete isotopic equilibration with the water mass being reconstructed.

Imagine a sediment surface with bottom-water εNd = −8, but pore-water exchange with local particles tends toward −5. A newly formed coating might reflect some mixture or evolving exchange history rather than exactly −8. If a researcher assigns −5 to the bottom water without further evidence, they may infer a different water-mass source. The discrepancy could instead be a sediment-water interface effect. Comparing modern bottom waters, pore waters, coatings and detrital phases helps estimate which interpretation is plausible.

Boundary exchange can also change a water mass’s isotope signature along continental margins without a large change in water volume proportions. A shift in margin interaction or sediment reactivity might move εNd while the bulk circulation pathway remains similar. An interpretation should identify the expected spatial pattern under boundary exchange compared with end-member mixing. If a signal is strongest near margins and weakens offshore, that geography provides a testable clue, although it is not conclusive by itself.

Instrumental precision and normalization

Neodymium isotope measurements commonly require chemical separation and high-precision mass spectrometry. Standards, interference corrections, mass-bias normalization, blanks and repeatability govern the quality of the reported ratio. Some interfering species or matrix residues can affect signals unless separated or corrected appropriately. The details are method-specific and should be found in the original analytical description. A ratio reported with six digits is not automatically accurate to six digits in terms of palaeoceanographic inference.

A laboratory can exhibit excellent reproducibility for a dissolved standard and still have uncertainty related to sample dissolution or sediment leaching. This is analogous to measuring the right number on the wrong material: the mass spectrometer may work perfectly while the extracted Nd fails to represent the intended seawater reservoir. Distinguish instrument precision, chemical blank uncertainty, extraction selectivity and environmental representativeness. They should not be compressed into a single small error bar that describes only the final analytical step.

When comparing studies, ensure they use compatible normalization conventions, reference values and standards. A paper may report the raw 143Nd/144Nd ratio, a normalized ratio, εNd relative to CHUR, or values adjusted for particular methodological issues. The numerical records can be reconciled only after the conventions are understood. If different reference ratios are used, convert transparently using the same base rather than assuming a discrepancy between two laboratories must represent an actual ocean difference.

A worked two-end-member marine investigation

Take hypothetical end member A at εNd = −13 with concentration 12 pmol/kg and end member B at εNd = −5 with concentration 24 pmol/kg. Suppose a sample contains 40 per cent A and 60 per cent B by water mass, and imagine no chemical sources or sinks after mixing. A contributes 0.40 × 12 = 4.8 concentration-weighted units; B contributes 0.60 × 24 = 14.4. The approximate mixed epsilon is [4.8 × (−13) + 14.4 × (−5)]/19.2 = −7.0.

The first tempting error is to ignore concentration and calculate 0.40 × (−13) + 0.60 × (−5) = −8.2. That gives a different answer because it treats isotope composition as though equal masses of water transported equal Nd amounts. The second temptation is to interpret −7 as proof of precisely 60 per cent B in the real ocean. Without end-member concentration histories, exchange corrections and an independent mixing constraint, the inverse solution may not be unique. The forward toy model is not a uniquely determined field reconstruction.

Suppose an observed sample is −7, but a sediment leach also displays anomalous aluminium and other detrital indicators. That observation raises the possibility that part of the isotope signal came from mineral dissolution rather than water-mass mixing. Repeating the mixing arithmetic more precisely will not resolve the carrier problem. The next useful work is chemical and mineralogical: compare extraction strengths, carrier phases, blanks and relevant sediment inputs. Correct equations cannot rescue an invalid source assumption.

A worked geological trend

Imagine an ancient core recording εNd values of −11, −10, −8 and −7 from old to young. The observations show a trend toward higher, more radiogenic isotope ratios on the epsilon scale. They do not by themselves show that a named current accelerated. A possible interpretation is greater contribution from a radiogenic water source, but alternatives include changing continental weathering, boundary exchange, pore-water effects or detrital contamination. A trustworthy narrative states which alternatives have been examined and how.

A nearby core might show the same trend at one depth but little change at another. That spatial difference could support an evolving water-mass boundary, particularly when independent hydrographic tracers agree. Yet preservation and leaching biases may also vary with sediment type or location. Align the age models, compare extraction protocols, document isotopic end-member assumptions and examine whether the signal follows a coherent regional structure. Proxy gradients become powerful when they constrain a model that makes further predictions.

If the isotope trend spans millions of years, source-rock evolution and changing erosion patterns deserve particular attention. The isotopic composition of continental material reflects deep geological history, while regional delivery pathways and climate-controlled weathering can vary on much shorter intervals. Assuming modern εNd end members were fixed over an entire geological epoch can obscure those changes. A careful palaeoceanographic reconstruction uses the best available coeval evidence for both water sources and the material supplying neodymium.

Water age is not an epsilon number

A water parcel bearing εNd = −9 is not nine years, nine centuries or nine thousand years old. The epsilon notation expresses relative isotopic composition, and the radiogenic 143Nd system reflects source geology over far longer timescales. It does not function as a direct clock for seawater ventilation. Ocean mixing and chemical exchange alter its distribution. Time since atmospheric contact requires other tracers and models, with their own assumptions. This distinction is especially important when a public summary casually calls all isotope ratios “ages.”

A related misconception treats more radiogenic values as always indicating younger water. “Radiogenic” refers to an isotope ratio influenced by the long-term decay history of geological source material, not the contemporary age of the seawater parcel. New water can carry an unradiogenic source signature and old water can carry a radiogenic one. Source identity, transport and exchange must be separated conceptually from ocean residence time. The vocabulary is scientifically precise only when the object being dated is named.

Independent exercises with answers

Exercise one: use CHUR = 0.512638 and a hypothetical sample ratio 0.512381681. Find εNd. Divide the sample ratio by CHUR to obtain 0.9995, subtract one and multiply by 10,000. The answer is −5.0. The sign confirms a lower ratio than the reference. Explain why this is not “minus five per cent neodymium”: it is a relative isotopic difference in ten-thousandth units.

Exercise two: reverse an εNd value of −12 under the same reference. Rsample = 0.512638 × (1 − 12/10,000) = 0.512638 × 0.9988 ≈ 0.512022834. A calculator can verify the last digit under the stated precision. The useful skill is to carry the reference ratio and the factor 10,000 faithfully through the transformation. If a spreadsheet instead uses 1,000, the answer will be much too far from the reference.

Exercise three: two equal masses of waters have εNd values −10 and −2, with dissolved Nd concentrations 10 and 30 pmol/kg. Using a concentration-weighted small-difference approximation, calculate [10 × (−10) + 30 × (−2)]/40 = −4.0. The naive average −6.0 is wrong for the modeled dissolved element inventory. Name the assumption that no Nd sources or sinks act during mixing, and explain why boundary exchange could invalidate a simple closed model.

Exercise four: 80 per cent of the water mass is source A with Nd = 5 pmol/kg and εNd = −15; 20 per cent is source B with Nd = 20 pmol/kg and εNd = −5. The weighted Nd contributions are 0.8 × 5 = 4 and 0.2 × 20 = 4. The approximate mixture epsilon is (4 × −15 + 4 × −5)/8 = −10. Equal isotope contributions have arisen despite an 80–20 split by water mass. That is the central reason to keep concentrations alongside isotope ratios.

Exercise five: a bottom-water sample has εNd = −8 and an adjacent sediment leach yields −4. May the leach be interpreted as a newly formed, more radiogenic deep current? Not without testing. Pore-water exchange, detrital dissolution, volcanic material or analytical issues could shift the extracted ratio. Compare different carrier phases, leach conditions, elemental indicators and regional evidence. A mismatch invites a phase-specific investigation, not an immediate circulation story.

Exercise six: two analytical sessions measure the same standard with nearly identical values, but sediment replicates differ by one epsilon unit. What does the standard agreement establish? It supports instrumental comparability for that standard; it does not guarantee identical leaching, sample heterogeneity or phase selectivity. The appropriate next checks are preparation blanks, extraction yields, sediment mineralogy, independent leaches and sample replicates. The key is to locate the uncertainty where it enters the chain.

Exercise seven: a published figure labels its values “Nd, ppt” but the caption discusses εNd. Can the two be read as the same variable? No. Concentration units such as parts per trillion describe how much material is present, while εNd describes isotope composition relative to a reference. Both can be reported for the same sample and jointly support mixing analysis. If the axis and caption conflict, consult the original data and methods before quoting the graph.

Exercise eight: a core record shifts from εNd = −12 to −7 while detrital mineralogy changes markedly. Identify at least three tests. Examine whether the authigenic extraction increasingly dissolves detrital sources; compare alternative carriers such as suitable foraminiferal phases or fish debris; check leachate elemental ratios and mineral-specific signatures; and compare regional seawater-related proxies. The study needs to establish that its archive represents the intended water reservoir before attributing the shift to current reorganisation.

Exercise nine: a mixing inversion returns a source contribution greater than 100 per cent. Does that reveal a new ocean pathway? More likely, at least one assumption, input value or model is wrong or incomplete. Recheck end-member values, concentrations, units, reference ratios and possible exchange terms. An observed ratio outside the two-end-member range cannot be explained by simple closed mixing of those end members. Do not force the fraction into zero to one without explaining the model mismatch.

Exercise ten: an author writes, “The epsilon neodymium value proves that a current moved five times faster.” Identify the unsupported leap. εNd may trace source contributions or exchange processes but has no universal direct conversion to velocity. Transport rates require independent dynamics or transport constraints. The correct translation might say the isotope pattern is consistent with changing water-mass provenance under a specified model. State the observational support and test alternatives rather than attaching an invented speed multiplier to a dimensionless ratio.

How to audit a real marine εNd paper

First identify the material measured: dissolved seawater, fish teeth, foraminifera, authigenic Fe-Mn coatings, pore water or bulk detritus. Read the sampling and chemical extraction methods before interpreting the epsilon value. Check the laboratory standards, reference ratio, blanks and uncertainty reporting. Recalculate at least one raw-ratio-to-epsilon transformation to catch scaling errors. Record whether the study measures present-day seawater or reconstructs it indirectly from buried sediment.

Second, assess whether the chosen carrier reliably records seawater or has been modified by detrital contamination, pore-water exchange or recrystallisation. Look for extraction tests and chemical indicators rather than trusting the word “authigenic” alone. Identify regional geological sources and their isotope signatures. Consider whether water-mass end members and neodymium concentrations may have changed through the study period. A successful reconstruction aligns the chemistry of the archive with the physical ocean question.

Third, use an explicit mixing or transport framework with clearly stated assumptions. If the analysis estimates water fractions from εNd, inspect how dissolved concentrations and nonconservative exchanges are handled. Compare with independent hydrographic, nutrient, carbon-isotope or sedimentological evidence. Build uncertainty into the result and distinguish a source shift from a rate change. Where multiple explanations fit, report the alternatives and the observation that would discriminate between them.

Further interpretation challenges

A margin-to-open-ocean transect provides a way to test whether local sediment exchange rather than a wholesale water-mass replacement explains an isotope gradient. Suppose bottom waters are hydrographically similar along the transect while εNd changes most strongly next to exposed, reactive margin sediment. The geography is compatible with boundary exchange, especially if dissolved Nd concentration also changes. However, the inference needs observed water structure and particle chemistry; closeness to land is not enough. Compare predicted isotope and concentration patterns under margin exchange versus mixing of independently identified water masses, and report what remains ambiguous.

A vertical profile can offer another diagnostic. If εNd varies sharply across an ocean water-mass interface and follows temperature, salinity or nutrient boundaries, a circulation-related explanation may gain support. But if the ratio changes without corresponding hydrographic structure, there may be an additional local supply or removal process. A single depth sample cannot distinguish these possibilities. Repeating observations along the same isopycnal surfaces can help separate movement along density layers from local vertical mixing, while acknowledging that isotope reactivity still complicates simple conservation.

Uncertainty in end members is often more important than the number of decimals in the sample result. If one source is estimated at εNd = −12 ± 1 and the other at −6 ± 1, a measured mixture of −9 does not uniquely imply equal Nd contributions once the uncertain endpoints are considered. A sensitivity analysis should vary both source values and dissolved Nd concentrations within justified ranges, recomputing possible mixing fractions rather than inserting one exact solution. Report the distribution of feasible fractions and explain whether chemical exchange broadens the range.

Think also about temporal resolution. A fish tooth, sediment coating and water sample integrate different formation or exchange histories. Even when they yield similar epsilon values, they may not represent exactly the same time interval. Sediment disturbance, bioturbation and age-model uncertainty can blur short-lived events. A high-resolution circulation claim requires evidence that the carrier material records the desired time scale. When a layer aggregates centuries, its isotope ratio is not an instantaneous photograph of a single eddy or overturning pulse.

Data translation should preserve raw analytical values, reference conventions, extraction details and the final interpreted epsilon number. A reader who only receives a temperature-style colour map of “water-mass age” cannot reconstruct how the authors reached the conclusion. Provide paired 143Nd/144Nd ratios and εNd where practical, state Nd concentration units, and show a simple cross-check against standards. Transparent metadata make the dataset reusable even if a later study revises a source end member or improves the understanding of boundary exchange.

Frequently asked questions

What does a negative εNd value mean? It means the measured 143Nd/144Nd isotope ratio is below the chosen reference ratio. It does not imply an absence of neodymium, a negative current or an automatically old water mass. The value is a compact way to describe isotopic composition and compare it with possible sources.

Does εNd uniquely identify an ocean current? No. Different sources and mixtures can produce similar values, and neodymium can be exchanged with particles and margins. Spatial structure and independent tracers can make provenance inferences stronger. Isotope matching should be treated as a constraint on possible origins rather than a unique geographic fingerprint in every environment.

Can a bulk sediment measurement replace seawater εNd? Not generally. Bulk sediment may be dominated by detrital particles carrying geological source signatures, while a water-mass reconstruction seeks a seawater-derived or otherwise defensible archive. Selective extractions and biogenic carriers can help, but they require validation. The sample’s carrier phase determines what reservoir the isotope measurement represents.

Why are concentration measurements needed alongside epsilon values? Mixing proportions by water amount and mixing proportions by neodymium amount differ when source waters have unequal Nd concentrations. εNd alone may leave a water-fraction inversion underdetermined. Concentration and other chemical constraints help, although boundary exchange can still complicate the result.

What is the most important caution for older sediments? The archive must be shown to retain a seawater-related signal, and assumed water-source isotope compositions must be appropriate for the ancient interval. Geological sources, boundary chemistry, pore-water exchange and analytical phase selectivity can all shift through time. An impressive isotope curve is not automatically a direct history of circulation.

The scientific meaning of translation

Neodymium isotopes become useful when readers understand both sides of the bridge: an extremely small radiogenic isotope-ratio difference and the large-scale geological and marine processes that can produce it. The laboratory result is precise chemical evidence. The water-mass story is a model constrained by that evidence, by where the neodymium came from, and by how it moved and exchanged between reservoirs. Learning the distinction makes εNd a powerful reasoning tool rather than an unexplained symbol on a palaeoceanographic chart.