Marine Sediment Grain-Size Analysis and Laser Diffraction

A grain-size curve is a measurement, not a current-speedometer

Marine sediment grain-size analysis converts a mixture of particles from a core or seabed sample into a distribution across diameter classes. The resulting laser-diffraction curve, sieve table or cumulative plot can tell a reader something about physical texture and particle populations. It cannot, by itself, measure ocean-current speed. The first translation task is to name the material, fraction, instrument and reporting basis before interpreting anything environmental. A median size of 25 micrometres can be perfectly measured while the geological story attached to it remains uncertain.

Did you know that two laboratories can analyze the same marine mud and report different percentages of “clay” without either misreading its equipment? One may use a 2-micrometre threshold and another a 4-micrometre threshold. One may calculate laser-derived optical equivalent volume, while another separates clay through a settling method. Some laboratories remove carbonate or organic matter before determining terrigenous grain size. If the conditions differ, the numbers describe different operational populations. Translate the methods before you translate the conclusions.

This guide explains marine sediment particle-size data from first principles: how a sample becomes a size curve, how sieve and laser techniques differ, what D10, D50 and D90 mean, how phi units reverse the apparent size direction, and why fine fractions are sensitive to dispersion and particle shape. We work through checked mass accounting, conditional volume-to-mass conversion, sample comparison and independent practice questions. The goal is a dependable explanation for readers who need to understand the result rather than memorize an instrument label.

A useful reading order begins with the core description, then sample interval and preparation protocol, then instrument data, then derived statistics, then geological interpretation. Reversing that order invites confirmation bias. If the paper opens by claiming vigorous currents, the reader may treat every coarse tail as proof. Starting with the raw texture helps identify storm layers, turbidity flows, shell-rich horizons, local sediment supply and sample preparation as alternatives. Good translation makes those pathways visible without burying the reader in equipment advertising.

Track every physical fraction

A marine core has position, orientation, section and depth metadata. A laboratory aliquot might come from 112–114 centimetres within one core section or from a narrow lamina visible on a split face. That thickness matters. If a sample straddles an erosional boundary, its grain-size distribution mixes two units rather than characterizing either independently. A publication may convert section depths into metres below seafloor or a composite depth scale across multiple holes. Keep the sample identifier and original interval in the record to prevent assigning one distribution to the wrong event.

The denominator of a reported fraction can be the original wet mass, dried bulk mass, carbonate-free dry residue, mass retained in a sieve class, or a laser-inferred particle-volume distribution. These are not interchangeable. A sample containing 30 grams of shell carbonate and 70 grams of silicate material has a different bulk denominator from the same sample after carbonate dissolution. If the treated residue has 20% sand, that corresponds to 14 grams of residue sand under the specified simplified mass accounting, not automatically 20 grams of bulk sediment.

Water and salt complicate a dry-mass measurement. Marine sediments include pore water that can leave salts when evaporated, and procedures vary in whether and how samples are rinsed. A wet mass of 120 grams cannot be compared directly with 80 grams of oven-dried sediment as though both represented solid material. The laboratory should describe temperature, drying, salt correction where appropriate, pretreatment and the recovery of solids. A reader need not perform the procedure but should insist that the basis of every percentage is clear.

Suppose a sample is first passed through a 63-micrometre sieve and the coarse retained portion is weighed. The finer material enters a laser analyzer. A laser report normalized to 100% of the fine fraction does not automatically give whole-sample percentages. To reconstruct the bulk distribution, one must know the fine fraction’s contribution to the original sample and any losses. Without those masses the correct narrative is “within the prepared fine fraction,” not “of all particles in the seafloor core.”

Representative splitting is another hidden challenge. A few large shell fragments can dominate the coarse mass of a small subsample, while very fine material may aggregate or segregate during handling. Repeat scans of one well-mixed suspension assess instrument consistency, but they do not test whether a tiny aliquot represents the core interval. Repeated independent subsamples offer a stronger check of heterogeneity. If a sediment layer is genuinely patchy, variability is scientific information and should not be disguised by averaging without explanation.

A good specimen ledger states depth, section, lithologic unit, mass before and after treatment, fraction boundaries, instruments, replicate identifiers, dates and any visible disturbance. If an analyst removes only a shell fragment judged to be contamination, the ledger should say so. If the scientific aim is to measure sediment as deposited, removing natural shells may be inappropriate. The preparation follows the question. An accurate translation reports which population the laboratory intended to characterize, not merely which percentages emerged from software.

What a sieve measures and what a laser infers

Sieving is a physical separation by aperture. A stack of decreasing openings divides material into retained mass fractions under a defined agitation and sample-loading procedure. Grain shape affects passage: a long thin fragment may orient through an opening differently from a compact grain of similar maximum dimension. Cohesion and moisture can cause fine particles to adhere to coarse ones. Clogged sieves or overloading distort the distribution. The dry masses retained at each stage provide a check that the fractions plus documented losses reconcile with the starting sample.

Laser diffraction infers particle sizes from how a suspended mixture scatters light. The instrument and its software use a scattering model to estimate an equivalent diameter distribution. The result is commonly volume weighted and reflects optical assumptions. It is not a direct photograph of each grain measured with a ruler. The curve can be sensitive to refractive index, background light, particle concentration, multiple scattering, bubbles, circulation and dispersion. Those are reasons to report method settings, not evidence that laser diffraction is useless.

Consider platy clay minerals. A particle has length, width and thickness; an “equivalent sphere” cannot represent all three dimensions simultaneously. Light scattering and settling need not assign the same characteristic diameter. Diatom frustules, broken shells and elongated mineral shards introduce similar shape complications. Comparing a laser-derived size with a settling velocity therefore requires additional assumptions about shape, density and aggregation. The phrase “mean grain diameter” should always be read as a statistic under a specified measurement method.

Mixed sieve–laser workflows are often practical. USGS sediment laboratories describe using coarse sieves with laser diffraction for finer particles. The join is meaningful only when the original mass fractions are preserved. For example, if 20% of bulk dry mass remains above the sieve threshold, and 25% of the fine branch is smaller than a selected laser cutoff under an admissible mass-equivalent assumption, the bulk fraction below that cutoff is 0.80 times 0.25, or 20%. Stating 25% as the bulk value would count the wrong denominator.

A microscope supplies independent context that an instrument histogram cannot. It can reveal shell fragments, angular quartz, intact diatoms, clumps, organic debris or crystals produced after deposition. The same optical mode may represent different material populations in two samples. When a secondary peak appears, confirm its existence with repeat measurements and image inspection before treating it as an extra transport regime. The peak is an observation in a computed curve; its origin is an interpretation requiring separate evidence.

Instrumentation precision does not equal geological specificity. A machine might reproduce D50 within a small percentage across several scans while a core interval contains multiple sedimentary processes. Conversely, a variable instrument result could reflect genuine aggregating behavior rather than an environmental event. A careful report separates the reliability of repeated measurements from the validity of inferring river supply, ocean currents, resuspension or climate history. Each inference needs appropriate physical context.

Understand size classes, log scales and sorting

Common size terminology classifies sand, silt and clay, but the exact boundaries are operational. One scheme places the clay boundary near 4 micrometres; another uses 2 micrometres. Sand commonly begins around 63 micrometres and may extend to about 2 millimetres, while other schemes differ slightly. Do not rewrite percentages under a new scale without the original distribution data. If study A reports 30% below 2 micrometres and study B reports 30% below 4 micrometres, numerical equality does not establish equivalent clay composition.

Many grain-size plots use micrometres on a logarithmic horizontal axis. Equal widths on a log axis correspond to equal ratios of diameter, not equal additions. A jump from 2 to 4 micrometres occupies the same log distance as 20 to 40 micrometres. Some outputs display cumulative “percent finer than” while others plot “percent coarser than” or retained mass. Look at the axis label and accumulation direction before describing whether a curve shifts toward fine or coarse grains.

The phi scale is defined as φ = −log₂(d in millimetres). This means a larger phi number corresponds to a smaller diameter. Diameter 0.125 mm is 3 phi because 2⁻³ is 0.125; diameter 0.03125 mm is 5 phi because 2⁻⁵ is 0.03125. A sample shifting from 5 phi to 3 phi becomes physically coarser despite the plotted phi value decreasing. The numerical operation is simple but frequently reverses interpretations when a reader assumes “bigger number” means “bigger grain.”

D10, D50 and D90 usually refer to percentiles of a cumulative finer-than distribution: 10%, 50% and 90% of the reported size basis are finer than the corresponding diameters. D50 is a median, not a representative grain that necessarily exists in the sample. A strongly bimodal distribution can have a median near a size where few actual particles occur. Percentiles also inherit the weighting convention: laser D50 is commonly a volume-equivalent median; a sieve-derived mass median is a different operational statistic.

Sorting refers to dispersion across sizes and can be expressed with several statistics. A simple D90/D10 ratio conveys the width of the central 80% of positive diameters: 80/4 equals 20, while 40/10 equals 4. The first is broader by this measure. Formal Folk-and-Ward graphic sorting uses selected phi percentiles and other coefficients. Never substitute a convenient ratio while retaining the literature’s formal sorting classification, because different formulas do not encode precisely the same feature.

The mean and median are not synonymous. A coarse tail can raise an arithmetic mean diameter substantially while the median barely changes. A distribution with equal material at very small and very large sizes is especially poorly summarized by one average. Report modes, tail proportions and interval-specific context when an interpretation depends on unusual fractions. A neat median line may conceal the important event sediment or dissolved biogenic component. Choose statistics for the scientific question rather than because the instrument prints them by default.

Preparation changes what counts as a particle

To disperse sediment, a laboratory may use chemical dispersants, controlled ultrasound, mixing and wet sieving. The goal might be to separate mineral grains that arrived as aggregates, but a strong treatment can damage fragile biological material. Natural marine flocs are physically meaningful entities for settling behavior, while dispersed primary particles may be more relevant for identifying mineral grain-size supply. These are different measurement targets. If the report does not explain which one was wanted, interpretation of the laser curve is incomplete.

Carbonate removal is sometimes used before characterizing terrigenous particles. Acid dissolves shell carbonate and some other reactive phases, changing the solid inventory. If bulk sediment contains abundant foram tests, a size curve of carbonate-free residue tells us about a selected fraction, not the deposited bulk texture. USGS procedures discuss chemical pretreatment and warn that its use depends on the presence of detrital carbonate. Always track starting and residual mass, chemistry, washing and whether silicate material was affected.

Organic matter can bind mineral particles into clumps or contribute its own fragments to the measured size population. Oxidative treatment may be needed for certain sediment questions, but it changes composition and can affect fragile constituents. Comparing an oxidized sample with an untreated sample requires more than aligning the x-axis. If a fineward trend appears only after changing the oxidation protocol, it may be methodological. Parallel untreated and treated aliquots can help quantify the processing effect rather than assuming it is negligible.

Salt concentration and water chemistry influence clay aggregation. Dispersing a salty marine mud in purified water can change its ionic environment; particles that existed as flocs may separate. Floc size, effective density and settling velocity are not the same as the diameter of each dispersed mineral plate. For a study of bottom nepheloid layers or floc transport, a dispersed laser histogram may be an incomplete proxy for behavior in situ. Keep the natural sediment setting separate from laboratory equivalent spheres.

Mechanical grinding is not ordinary dispersion. Breaking grains into smaller fragments manufactures new size material, while gentle disaggregation aims to release existing grains. Some soft sediment can be damaged by aggressive sonication or processing. The appropriate settings depend on lithology, desired fraction and instrument method. A trustworthy report records enough detail to replicate the procedure, including the stage at which a sample was divided, cleaned, washed or stored. “Standard preparation” is too vague when treatment changes the result.

A well-designed comparison uses paired aliquots and replicates to isolate effects. Measure a bulk mixture, a treated mineral residue or several independent subsamples when the question warrants it. If repeats are made from a single suspension, call them measurement repeats; if several separately collected slices are analyzed, call them independent or spatial replicates. Neither procedure alone diagnoses all bias. A figure should label the resulting curves distinctly so a reader does not mistake preparation differences for depositional changes.

Checked worked example: mass and percentages

Imagine an 80.0-gram dry whole sample. After coarse sieving, 12.0 grams remain above 63 micrometres, 67.2 grams enter the fine-fraction workflow, and 0.8 gram is documented as processing loss. Dividing by 80.0 grams gives 15.0% coarse, 84.0% fine and 1.0% lost. These sum to 100%. It would be wrong to call the fine fraction 85% simply because the sieve did not retain that last gram. The unassigned loss should remain visible until the lab knows what happened to it.

Suppose an analysis of that 67.2-gram fine fraction assigns 55% below 16 micrometres on a justified compatible mass basis. Then 0.55 × 67.2 = 36.96 grams would be below 16 micrometres under the stated ideal assumption, corresponding to 36.96 / 80.0 = 46.2% of the original sample. If the actual laser instrument reports volume percentage and densities differ, the conversion is conditional rather than exact. A scientifically precise article states that proviso and does not conceal the measurement basis.

A separate example starts with 100.0 grams of whole sediment from which 45.0 grams of carbonate material are dissolved, leaving 55.0 grams of noncarbonate residue under a simplified complete-recovery assumption. The residue contains 20% sand by dry mass, or 11.0 grams. This represents 11% of the initial whole dry sample as preserved residue sand. It does not establish that original bulk sand was 11%, because some dissolved carbonate may itself have been sand-sized. The mass accounting is correct, while the geological total remains unknown.

Now consider two median diameters of 18 and 24 micrometres from samples processed comparably. The difference is 6 micrometres; the relative increase is 6/18 = 33.3%. Converting to millimetres gives 0.018 and 0.024 mm. Phi values are approximately 5.80 and 5.38 respectively. The 24-micrometre median is coarser despite its smaller phi value. This numerical comparison does not imply the current was 33.3% faster; the relation between sediment size and transport energy is not a direct proportionality.

Compare two distributions. Sample A has D10 = 3 micrometres, D50 = 20 micrometres and D90 = 100 micrometres. Sample B has D10 = 8, D50 = 20 and D90 = 40 micrometres. A’s D90/D10 is about 33.3, whereas B’s is 5. Both have the same median, but A spans a much broader central range. A bimodal source mixture is one possibility; it is not proved by percentiles alone. The full histogram and mineral observations are needed before attributing a mechanism.

Independent subsamples yield sand percentages of 14%, 16% and 17%. Their arithmetic mean is 15.67%, and the sample standard deviation is approximately 1.53 percentage points. The mean can reasonably be rounded to 15.7% for this demonstration. Three values do not by themselves establish the uncertainty of every section in a core or a regional environmental effect. If a second nearby interval averages 16%, the difference may be smaller than the sampling variation. Replicate design and scientific question determine whether the difference is meaningful.

A denominator trap can occur without any extra coarse grains. At time A an interval receives 70 grams of fine material and 30 grams of coarse material, giving 30% coarse. At time B it receives 50 grams of fine material and the same 30 grams coarse, giving 37.5% coarse. The coarse share rises, but the absolute coarse amount remains unchanged. A core’s relative percentages cannot distinguish increased coarse supply from reduced fine supply without chronology, dry bulk density and appropriately modeled mass accumulation.

For a cumulative curve plotted in classes rather than individual values, do not claim a precision that its bins cannot support. If the 50% crossing lies somewhere inside the 16–32 micrometre class, the defensible direct reading is that D50 lies in that interval. Interpolation can estimate a value only after specifying how the distribution varies within the bin, often with a logarithmic size axis. A figure copied into a presentation should preserve that uncertainty instead of converting a rough visual guess into a six-decimal statistic.

Reading geological signals without overclaiming

A sudden coarse spike may be related to an event bed, river pulse, storm transport or sediment gravity flow. Look for an erosional base, normal grading, ripples, shell breakage, provenance changes and other sedimentary structures. A turbidite can deliver coarse particles into deep water without requiring a persistent increase in background bottom-current speed. If a spike coincides with a core-section boundary, check for drilling disturbance or sample mixing before telling an environmental story. Texture is one line of evidence, not a complete cause.

A gradual fineward shift can reflect quieter deposition but can also result from greater fine-particle supply, altered biology, changed sorting, carbonate dissolution or sample processing differences. Compare multiple proxies and adjacent cores where available. If the preparation protocol changed halfway through a long analytical series, that transition is a plausible alternative explanation. A strong interpretation says which competing causes were tested. Merely repeating that a coarser median indicates energy does not remove the ambiguity.

Particles contain history beyond their size. Quartz grains may arrive through rivers or aeolian transport, feldspars may record different mineral sources, and porous biological tests may respond differently to resuspension. The same size bin can therefore represent distinct origins and hydrodynamic behavior. Mineralogy, shape, microfossils and geochemistry can help, but they also have their own uncertainties. Use multiple independent observations rather than pretending a laser analyzer directly identifies provenance or sediment transport pathways.

Bioturbation mixes layers after deposition. In an interval with intense burrowing, a once-sharp sand pulse can become a gradual grain-size shoulder. Dissolution removes some carbonate grains; authigenic minerals add particles after burial; physical coring may smear a soft contact. Read the measured histogram beside the lithologic log and core images. If the timing of change matters, account for how mixing and preservation affect apparent signal sharpness. A smoothed trend is not automatically a slowly changing ocean condition.

The chronology must also support the story. A one-centimetre sample in a slowly accumulating core might cover centuries, while a rapidly accumulating event bed may capture minutes or hours of deposition. Particle-size resolution and age resolution are independent. A finely binned laser distribution cannot identify one named storm if the dated depth is uncertain by many decades. Translate the time attribution with the age model’s uncertainties, the sample thickness and evidence of disturbance. Instrument precision cannot make an undated event historically precise.

A regional comparison requires consistency across sites. A shelf core influenced by local rivers differs from a slope core with gravity-flow deposits, even when their medians match. Density, organic content, shell abundance and mineral supply affect settling and sorting. When an author says “grain size tracks current strength,” ask whether a calibration or independent flow measure supports that relation in this setting. A proxy can be useful without being universal. State the domain where its assumptions are credible and flag circumstances in which they fail.

Laboratory diagnostics and uncertainty

A laser instrument should be checked with appropriate calibration or reference material, background measurements, particle concentration controls, repeat runs and recorded optical settings. These checks assess instrument performance under defined conditions. They do not guarantee representative subsampling of a heterogeneous core or prove the optical assumptions for every clay and shell shape. A coherent audit distinguishes systematic uncertainty, short-term repeatability and sample heterogeneity. A small scan-to-scan variation can coexist with large uncertainty in geological interpretation.

Sieve quality checks include dry mass recovery, appropriate sieve cleaning, limited sample overload and consistent shaking. An 80-gram sample yielding 78 grams of retained fractions has 2 grams unaccounted for, equivalent to 2.5%. The analyst should locate or report the missing mass before normalizing retained fractions. If the loss preferentially involves very fine particles, the reported distribution can be biased coarse. Renormalizing to 100% changes the denominator but does not restore lost material. This is a mathematical repair only in appearance.

Method transitions can create artificial time trends in long datasets. A core analyzed on one laser model and a second core measured with another optical setting may display systematic differences. Before attributing them to climate, measure shared standards and paired sediment aliquots across the platforms where possible. Document software version, optical parameters, dispersion and class definitions. The comparison may still be scientifically useful, but a reader needs a bridge between the operational measurement systems.

If a histogram has a strange secondary peak, consider coarse contamination, bubbles, aggregated fines, incomplete dispersion or a genuine two-source mixture. Repeat the analysis after checking the suspension, inspect a microscope slide and compare independent aliquots. A peak that disappears with proper bubble removal is not evidence for a lost ocean current. A peak that persists and coincides with mineralogical contrast and a sedimentary boundary has stronger support. Show what observations changed the diagnosis rather than deleting inconvenient data without explanation.

Numerical uncertainty should not be confused with mechanism uncertainty. D50 might be reproducible to one micrometre while the choice among storm, river and gravity-flow explanations remains unresolved. Conversely, a clear event bed may be unmistakable even if its fine-grained tail varies analytically. Report precision for the actual measured quantity and confidence in the hypothesis separately. Avoid constructing a fictitious flow-speed percentage by treating a percentage change in grain diameter as a proportional change in current velocity.

An excellent data table includes original core reference, depth interval, pretreatment sequence, dry mass, fraction recovery, instruments, reference standards, D10/D50/D90, class boundaries, distribution weighting and replicate details. For a chart with interpreted hydrodynamic changes, add independent sedimentological and chronological evidence. Such metadata are not an obstacle to clear writing. They are what allow a concise explanation to preserve truth as findings are translated between an instrument report, a classroom lesson and a broader synthesis.

Additional cross-checks that change the interpretation

An additional test concerns density and the difference between volume and mass fractions. Imagine a sediment mixture with equal solid volumes of quartz-like material and a much denser heavy-mineral fraction. Its optical volume percentages could be approximately equal while its mass percentages are not, because mass is density times volume. If the heavier material has density twice that of the lighter material under the illustrative assumption, equal volumes correspond to a two-to-one mass ratio. A report that turns volume fractions directly into sediment mass accumulation without using component densities can create a serious systematic error. The repair is to identify composition, measure or justify effective densities, and show the conversion rather than implying it was part of the instrument output.

Another diagnostic uses mass closure across laboratory branches. Suppose 50 grams of dry sediment enter a preparation, 7 grams are isolated as a coarse fraction, 40 grams reach the analyzed fine suspension, and 3 grams are lost during washing. The 3-gram loss equals 6% of the original sample. That is not a harmless zero merely because the recovered 47 grams can be renormalized to 100%. If fine material was preferentially lost in decanting, the surviving sample would appear artificially coarse. A responsible laboratory description preserves absolute masses and investigates where losses occurred. When the loss composition cannot be recovered, a scientific interpretation should carry that uncertainty rather than invent a corrected histogram.

Core-scanning instruments and laboratory size distributions operate at different spatial scales. A scanner may acquire near-continuous physical-property measurements along an archive half while a grain-size laboratory takes discrete plugs several centimetres apart. A narrow sand layer visible in the scanner may be missed entirely by the laboratory samples or overrepresented if a plug targets the layer. Correlating the two datasets requires explicit depth scales and sampling support, including whether a specimen includes neighboring material. If a proxy peak and a size maximum are separated by a few centimetres, check the measurement footprints and core disturbance before proposing that a current change led or lagged a chemical event.

The surrounding stratigraphy is often the best independent check. A coarse tail aligned with sharp basal erosion, graded bedding and abundant fragmented shells suggests a different process family from a comparable tail in continuously bioturbated fine mud. Both might produce the same D90, yet one can be an episodic gravity-flow deposit while the other reflects gradual mixing and resuspension. A size histogram does not encode bedding geometry. The repair is to write two sentences: one describes the measured distribution; the second describes the independent core features and the process hypotheses they support. That distinction improves both scientific rigor and readability.

Sortable-silt studies may try to infer past bottom-flow strength from grain size in a carefully selected fine fraction. Such proxy interpretations are not identical to a bulk grain-size report. The selected fraction, exclusion of cohesive clay, possible effects of provenance and sediment supply, and site-specific calibration matter. Readers should not apply a published sortable-silt relationship to unsorted whole sediment simply because both graphs contain micrometres. Ask for the operational selection, statistics and assumptions, then compare with independent evidence. Grain size can be a valuable hydrodynamic proxy within a defensible domain; it is not a universal speed sensor across all marine sediment environments.

Finally, consider whether a change in percentages reflects a change in amount. With only normalized grain-size data, increasing the coarse proportion can mean added coarse particles, decreased fine particles, or both. Chronology and dry bulk density allow researchers to estimate mass accumulation, but those quantities introduce their own uncertainty and may be disturbed by event beds or hiatuses. An article should not turn “coarse share doubled” into “coarse delivery doubled” without a defensible absolute mass calculation. In worked examples, express mass accumulation as dry mass per unit area per time and keep the measured distribution basis aligned with that physical denominator.

Transfer practice with answers

Question one: a mixed workflow separates 25% of bulk dry mass as coarse sediment, and a valid fine-fraction measurement assigns 40% of the remaining 75% to particles below 20 micrometres. What share of bulk dry mass does that represent under equivalent-basis assumptions? Answer: 0.40 × 0.75 = 0.30, or 30%. The common wrong answer is 40%, which mistakenly uses the within-fraction denominator as the whole-sample denominator. If the laser output is volume weighted, verify density assumptions before calling the calculated result a mass share.

Question two: which is physically larger, 4 phi or 6 phi? Four phi is 2⁻⁴ mm = 0.0625 mm = 62.5 micrometres. Six phi is 2⁻⁶ mm = 0.015625 mm = 15.625 micrometres. The 4-phi particle is coarser. This is a sign-and-unit question as much as a sediment question. When a plotted phi trend decreases, the physical diameter generally increases under the defined formula.

Question three: two samples have identical D50 of 25 micrometres, but the first has D10 of 2 and D90 of 125 while the second has D10 of 10 and D90 of 50. Which is more broadly distributed under D90/D10? First ratio: 125/2 = 62.5. Second: 50/10 = 5. The first is much broader. No conclusion about a specific sediment source follows without examining the shape, composition and stratigraphy of the samples. Sharing a median does not imply shared environmental history.

Question four: an acid-treated residue is 60% of initial dry mass and contains 20% sand on a residue mass basis. What part of initial bulk mass is preserved residue sand? Answer: 0.60 × 0.20 = 0.12, or 12%. Can we conclude bulk sand was 12%? No; dissolved carbonate might include sand-sized tests. The correct answer distinguishes the known preserved fraction from the unknown material removed in pretreatment. That distinction is the difference between sound accounting and an invented measurement.

Question five: three laser scans of one suspension agree within 1%, but independently sampled core slices differ by 15 percentage points in coarse content. Is the instrument unreliable? Not necessarily. The sample may be heterogeneous or the preparation may vary between aliquots. Inspect sampling scale, mixing, lamination and preparation. The scans test short-term instrument repeatability; independent samples test something broader. It would be misleading to report all scans as independent evidence for one homogeneous sediment population.

Question six: the coarsest layer has a sharp erosional base and finer material upward. Does the median-size spike prove a stronger persistent background current? No. The observed sedimentary features are consistent with event deposition, including a gravity-flow mechanism. Test additional evidence of source, grading, location and dating. A sedimentary event can briefly change grain size without describing long-term bottom-current conditions. The careful answer identifies the plausible process and the limits of the claim.

Question seven: study A calls particles smaller than 2 micrometres “clay,” while study B calls particles smaller than 4 micrometres “clay.” Both report 18%. Are the results directly equivalent? No. A portion between 2 and 4 micrometres is counted differently, and the methods may also differ. Obtain the full cumulative distributions and standardize boundaries only when the data permit a defensible conversion. Never silently relabel both as the same percentage.

Question eight: a one-centimetre sample is analyzed at hundredths of a micrometre resolution, but the age model uncertainty at that depth spans several hundred years. Can the grain-size shift be attributed to a particular one-day storm? Not from those measurements alone. The limitation is chronology, not laser resolution. A detailed particle curve says what sizes were measured, while the age model sets the plausible timing. The scientifically useful result may be a broader interval of altered deposition.

Frequently asked questions and interpretation discipline

Is laser diffraction more accurate than sieving? They measure related operational properties in different ways; the appropriate choice depends on particle range, material and question. Can one read current speed directly from D50? No universal conversion exists without appropriate calibration and sedimentological context. Can chemistry change the size curve? Yes, when carbonate, organic matter or flocs are altered by treatment. What does a larger phi number mean? Smaller physical particle diameter. These simple answers should always be paired with the measurement assumptions of the actual study.

Why preserve the entire curve rather than only a median? Because tails, modes and sorting can reveal mixtures invisible to a single number. Why might a sand percentage rise without more sand arriving? The fine fraction may fall, raising the coarse share. Why record processing loss? It can alter denominators and bias the remaining distribution. Why verify links to original methods? They allow readers to distinguish source-defined measurements from a later summary. The goal is a readable explanation that can be audited, not a polished sentence that erases the experiment.

A well-translated grain-size result has a limited but important claim: a documented specimen, processed in a stated way, yielded a distribution with measured percentiles and uncertainty. Additional evidence may connect that distribution to transport, sediment supply, biology, mixing or environmental history. The connection is strongest when independent observations constrain alternatives. Make the data intelligible, calculate what the denominator allows, and stop at the edge of the evidence. That approach helps readers avoid treating a clear curve as an automatic cause.