The Gold Standard of Stratigraphy

Chapter contents

A photograph may show red rock below pale rock. That is an observation about appearance and position in the image. Calling the red rock older requires additional information about the section’s orientation, its contacts and the processes that produced the arrangement.

Start by describing what is visible. Are the boundaries planar, irregular, sharp or gradational? Do they cut across another feature? Are fragments embedded within another material? Distinguish the observation from the history proposed to explain it.

In an invented notebook, “coarse layer truncates inclined fine laminae” records a relationship. “A flood happened after uplift” is an interpretation requiring more support. The interpretation may become reasonable, but keeping the two statements separate makes the reasoning auditable.

A practical first task is to draw a small section with no geological story attached. Label the boundaries, texture changes and any uncertainty. Then write the simplest sequence of events consistent with those observations.

Ask which observation would distinguish that sequence from an alternative. A boundary might represent erosion, an intrusive contact or a structural displacement. Similar geometry in a small photograph does not guarantee the same origin.

The field’s wider setting is developed in The Gold Standard of Geology. Stratigraphy contributes a particular discipline within it: preserve the relationships that constrain history before filling the page with an attractive narrative. A careful description is already useful work, even when the full explanation remains unresolved.

Contents · Next chapter

In an upright, undisturbed depositional succession, lower layers formed before overlying layers. The condition matters. Folding, overturning, faults and intrusions can produce arrangements that cannot be read as a simple upward time sequence.

Consider an invented succession A below B below C. If the three beds are established as an upright depositional sequence, the order is A, then B, then C. If a dyke cuts all three, the dyke is younger than the rocks it cuts. The interpretation rests on the demonstrated cutting relationship.

Now add a fault that displaces the dyke. The fault movement responsible for the displacement occurred after the affected dyke formed. That statement does not date every episode of movement on the fault. A structure may have a longer history than the visible displacement records.

These distinctions teach an important habit: state which event the evidence constrains. “The fault is younger” can be too broad if it conflates initiation, reactivation and the specific movement observed.

Relative reasoning is powerful precisely because it can work without a numerical date. It establishes a partial order from relationships while leaving unresolved intervals and events honestly unresolved.

Contents · Previous chapter · Next chapter

Lithostratigraphy groups rocks by observable lithological properties and relationships. Chronostratigraphy organises rock bodies in relation to intervals of geological time. A physical change in rock type is not automatically a boundary between the same times everywhere.

The International Stratigraphic Guide on lithostratigraphic units makes physical properties central to recognition. Its chronostratigraphic guide addresses time-related rock units. These are complementary ways of organising the record.

Imagine a coastline moving landward. Sand can accumulate near the shore while mud accumulates farther offshore. As environments migrate, the sand-to-mud transition at one location can occur earlier or later than a similar transition elsewhere.

Calling every sand-to-mud boundary the same moment would confuse a moving environmental boundary with a synchronous event. The rocks can belong to a useful mapped unit without sharing exactly the same age at every point.

The distinction also protects vocabulary. A formation is not defined by a universal fixed duration or thickness. A time interval is not the same object as the rocks deposited during that interval.

For teaching, place two labels beside a specimen: one describing the rock’s features, the other describing what is known about its age. Ask which evidence supports each label. Students should see that classification by material and correlation through time need different observations. Memorising names without that distinction makes later diagrams unnecessarily confusing.

Contents · Previous chapter · Next chapter

A measured section is a record of vertical change through a succession, tied to location, orientation and a measurement method. Its value comes from consistent observation, not from the neatness of the final column.

Record the reference point and which direction counts upward. Describe bed thickness, grain size, sedimentary structures, fossils, contacts and sampling positions where relevant. Distinguish a covered interval from an exposed interval that contains no visible change.

An invented section might contain 1.2 metres of sandstone, 0.3 metres of mudstone, a covered interval estimated at 0.8 metres and a further 0.6 metres of sandstone. Drawing the covered interval as continuous exposed mudstone would turn missing observation into fabricated geology.

Thickness also depends on geometry. A slope distance measured across inclined bedding is not automatically true stratigraphic thickness. The appropriate correction depends on the relationship between the measurement direction and the bedding orientation.

For a simple invented case where the measured line is perpendicular to strike on a horizontal surface, a 10-metre horizontal separation across beds dipping 30 degrees corresponds to a normal thickness of 10 × sin 30 degrees = 5 metres. This special geometry should not become a universal field formula.

A good notebook records the geometry so the calculation can be checked. If the assumptions are unavailable, report the measured distance and its limitations. That is more informative than attaching an unjustified true-thickness label to a precise-looking number.

Contents · Previous chapter · Next chapter

An unconformity records a significant break associated with erosion or non-deposition. A thin boundary can therefore separate rocks whose histories are far apart in time. Physical thickness does not measure the duration missing at the contact.

The International Stratigraphic Guide on unconformities provides the terminology. In practice, identifying a surface and explaining its history are connected but distinct tasks. A regionally important boundary requires evidence beyond one suggestive photograph.

In an invented example, tilted beds are truncated and overlain by less inclined strata. One possible sequence includes deposition, tilting, erosion and renewed deposition. The contact constrains that sequence without telling us how long each step lasted.

A parallel contact can also hide missing time. The absence of an obvious angular discordance does not establish continuous deposition. Conversely, not every small interruption deserves the same regional significance.

Ask what would reveal the break: truncation, an erosional relief surface, a weathered horizon, missing independently recognised intervals or another appropriate observation. The answer depends on the setting and quality of exposure.

A useful classroom analogy is a diary with several pages removed. The two remaining adjacent pages can be close together physically while far apart in date. The analogy should stop there: geological gaps are not always simple removals from a once-complete sequence. Some missing record was never deposited locally. That distinction matters when reconstructing the past.

Contents · Previous chapter · Next chapter

Suppose an invented upright succession contains 24 metres of preserved sediment between two independently constrained depositional horizons. The older horizon is 12 million years old and the younger is 8 million years old. For this exercise, treat the ages as exact and ignore compaction.

The elapsed interval is four million years. Dividing preserved thickness by elapsed time gives six metres per million years. This is a net preserved accumulation rate across the whole interval, including any missing time.

Now suppose independent evidence establishes a two-million-year hiatus within the interval. The preserved sediment accumulated during the remaining two million years. Its average accumulation over active depositional time becomes twelve metres per million years under these simplified assumptions.

Both numbers can be correct because their denominators answer different questions. Calling either “the sedimentation rate” without qualification hides the distinction.

The example still does not recover short-term deposition rates. A storm bed might form rapidly within a much longer interval. Compaction and erosion could alter the preserved thickness. Numerical age uncertainties would also affect the calculation.

The useful transfer check is to ask what each number describes before applying it elsewhere. A long-term average should not be used to estimate the duration of a single bed unless the additional assumptions are justified.

Distinguish a local episode from a long average

Extend the invented 24-metre succession with a two-metre bed that independent sedimentological evidence suggests formed during one short episode. The average across four million years remains six metres per million years. That number does not tell us how long the two-metre bed took to form.

A student might divide two by six and report one third of a million years. The division is valid arithmetic, but applying the whole-interval average locally is an added model assumption. The evidence for a brief episode gives a reason to question it.

A more careful answer separates three statements: the thickness was measured as two metres; the larger interval spans four million years under the stipulated age constraints; and the local duration remains unresolved by that average alone. Additional process evidence can inform a duration estimate, but it should be identified explicitly.

This distinction matters whenever a section is converted into a time series. Evenly spaced samples in depth do not automatically represent evenly spaced samples in time. A chart can look regular because the sampling positions were regular, while the history represented by those positions remains uneven.

An original event-order problem with an unresolved branch

Suppose bed A is overlain by B. An intrusion cuts A but ends below the exposed base of B. An erosion surface separates A from B. What can be ordered confidently?

A predates the intrusion where the cutting relationship is established. B postdates the surface on which it was deposited. But the supplied view does not say whether the intrusion was eroded before B or simply did not reach B. The intrusion’s termination geometry and relationship to the surface need closer examination.

Two narratives can fit the incomplete description. In one, intrusion follows A, erosion truncates it and B covers the result. In another, erosion occurs, the intrusion enters A beneath the surface, and B is later deposited without the intrusion reaching its base.

The purpose is not to make relative dating feel unreliable. It is to show how strongly it depends on the relationship actually observed. A textbook drawing can supply a complete truncation contact; a field exposure may not.

Ask the learner to mark the unresolved relation rather than force every event into one line. A partial order can be a complete answer to incomplete evidence. This habit prepares students to distinguish a well-constrained historical sequence from a story that becomes definite only because an ambiguous connection was silently filled in.

Contents · Previous chapter · Next chapter

Correlation asks how records at different places relate. Similar lithology is useful evidence, but many environments repeatedly produce similar materials. One pale layer is not automatically the same pale layer across a valley.

Consider two invented sections. Section West contains mudstone, a thin distinctive ash and sandstone. Section East contains mudstone, a similar ash and limestone. If the ashes are demonstrated to represent the same event, they can connect the sections even though the rocks above differ.

The phrase “demonstrated to represent the same event” carries the work. Appearance alone may be insufficient. Composition, stratigraphic position, independent age information and the regional setting can strengthen or weaken the match.

Now add a second ash to one section. A correlation based only on colour becomes ambiguous. A better interpretation compares the sequence of markers and asks which proposed match preserves all independently supported relationships.

Draw alternative correlation lines before choosing one. Which line requires an unobserved fault? Which implies a missing interval? Which fits the fossils but conflicts with another secure marker? The alternatives make assumptions visible.

This is an especially useful teaching task because the first plausible answer may not be the best-supported answer. Give learners enough information to revise a match when a new observation arrives. Revision then becomes evidence of improved reasoning, rather than a sign that the original work was pointless.

Contents · Previous chapter · Next chapter

Biostratigraphy organises strata through their fossil content. Fossils can support correlation, but their usefulness depends on identification, distribution, preservation and the relationship between the fossil and the enclosing material.

The International Stratigraphic Guide on biostratigraphic units distinguishes fossils believed to belong to the depositional assemblage from reworked or introduced material. A fossil carries both a biological history and a history of movement and preservation.

Imagine an invented shell from an older bed eroded and deposited in younger sediment. Dating the younger bed solely by the shell would ignore that transport. Abrasion, a mismatch with the surrounding assemblage or other contextual evidence may prompt investigation.

The reverse problem can occur when younger material enters an older layer through roots, burrows or later disturbance. A position inside a rock body does not automatically make the introduced object the age of deposition.

Biostratigraphic interpretation also depends on taxonomy. If two previously combined forms are distinguished, a recorded range may need reassessment. Changes in classification can alter the biological signal without changing the physical section.

Continue with The Gold Standard of Palaeontology for the wider fossil record and The Gold Standard of Systematics and Taxonomy for the evidence behind biological identification. The teaching habit is to ask two questions together: what organism is represented, and how did this evidence enter the layer in which it was found?

Contents · Previous chapter · Next chapter

The lowest observed occurrence of a fossil in a section is an observation about that section and its sampling. It is not automatically the moment the lineage originated. The highest observed occurrence is not automatically the moment of global extinction.

Imagine an invented fossil present in six sampled beds. In the next three beds, no specimen is found. Possible explanations include local absence, unsuitable preservation, insufficient sampling or actual disappearance. The missing observations require interpretation.

Increase the sampling effort and suppose one specimen is found in the second previously empty bed. The observed last occurrence moves upward. The organism did not return from extinction because the notebook changed; the evidence about its range improved.

This example reveals why an absence statement needs a search account. How much material was examined? Were comparable environments sampled? Could the organism have lived elsewhere while conditions changed at this site?

A useful correlation uses assemblages and multiple indicators where appropriate. It also preserves differences between local records and wider inferred events. A convenient line on a diagram should not erase the uncertainties in how the line was placed.

For a learner, ask what evidence would be needed to move from “not found here” to a regional or global claim. The answer encourages a scale change in reasoning: observations at one site form part of a larger argument. They do not become universal simply because the diagram uses a familiar geological period name.

Contents · Previous chapter · Next chapter

Geochronology supplies numerical age information from particular materials and processes. Stratigraphy supplies relationships that help interpret where that information belongs in a sequence. The two are strongest when their different jobs remain clear.

A date from a mineral does not automatically date deposition of the rock around it. The mineral may have crystallised earlier, been transported, or recorded a later thermal event. The interpretation depends on the material, method and geological context.

In an invented sequence, a primary ash bed below a fossil-bearing interval is dated at 10.2 million years, and a primary ash above it at 9.8 million years. Under the stated interpretation and upright ordering, deposition of the fossil interval lies between those ages.

That bracket does not establish an age of exactly 10.0 million years. A midpoint is a calculation, not an additional observation. Assigning it as a precise age would require a model for accumulation and the relevant uncertainty.

Now suppose the upper dated material is recognised as reworked crystals rather than a primary ash. The original upper bracket may no longer be valid. The numerical laboratory result could remain correct while its use as a depositional boundary fails.

The existing Gold Standard of Geochronology develops the clocks and events in greater depth. The transferable question is always “the age of what?” Attach that answer to every date before using it to narrate Earth history.

Contents · Previous chapter · Next chapter

An age model links position in a record with estimated time. Between dated horizons, interpolation supplies information through assumptions. Different assumptions can produce different histories even when the dated points are identical.

Consider an invented ten-metre interval between 100,000 and 80,000 years before present. Constant net accumulation would place its midpoint at 90,000 years. That is a simple linear model.

Now introduce evidence for a 10,000-year hiatus at the midpoint. The preserved lower and upper parts must share the remaining depositional time. A single continuous straight line through the interval would spread the gap across sediment that did not record it.

Alternatively, suppose compaction differs strongly between the upper and lower sediment. Equal present-day thickness may represent unequal original accumulation. The relationship between depth and time again requires more than two endpoints.

The practical response is to show the age constraints, the model and the unresolved intervals separately. Label interpolated ages as estimates. Report uncertainty that reflects the model where possible, rather than displaying only analytical laboratory precision.

The Palaeoclimatology root and its guide to shared age anchors and lead–lag uncertainty extend this issue into environmental comparisons. A small apparent time difference between records can depend on how both chronologies were built.

A good student answer identifies which conclusion remains stable across reasonable models. If the event order changes when interpolation changes, the evidence may support a weaker timing claim than the most attractive graph suggests.

Contents · Previous chapter · Next chapter

Stratigraphic architecture reflects interacting processes: sediment supply, space available for accumulation, relative water level, tectonics, erosion and the movement of depositional environments. One vertical pattern rarely identifies one cause uniquely.

A coarsening-upward succession can suggest a changing relationship between sediment delivery and the depositional setting. It does not, on its own, prove a particular global sea-level history. A local process may produce a similar pattern.

In an invented basin, increased sediment supply advances a delta while water level remains unchanged. In another, changing relative water level shifts environments while supply is steady. A small vertical section could look similar even though the histories differ.

To distinguish them, ask for lateral geometry, regional relationships, independent constraints and the distribution of surfaces. The useful observations are those that make competing explanations predict different records.

The International Stratigraphic Guide on relationships among unit types supports keeping different classification criteria distinct. A descriptive unit should not silently acquire a causal interpretation merely because the explanation is popular.

For teaching, give two histories that could generate the same simplified column. Ask learners to design the next observation rather than choose by confidence. The subject becomes more interesting when uncertainty is treated as a route to investigation. Knowing which evidence would discriminate is a more substantial achievement than attaching a dramatic process label to every boundary.

Contents · Previous chapter · Next chapter

An outcrop exposes a surface. A core samples a narrow subsurface path. A geophysical log records a measured response along a borehole. These sources can complement one another, but they do not show identical information.

A core can preserve material and small structures that are difficult to infer from a log. An outcrop may reveal lateral geometry absent from one borehole. A log can provide continuity through intervals that were poorly recovered, while still requiring calibration and interpretation.

Imagine an invented core with a missing 40-centimetre interval. A corresponding log shows a sharp response change. The signal suggests a change worth examining, but it does not reconstruct the missing rock texture by itself.

A useful interpretation connects the physical sample, depth reference and instrument response. Depth offsets, core expansion, recovery gaps and measurement resolution can affect alignment. A neat overlay can hide a registration problem.

Ask what each source directly measures and what has been inferred. “The instrument recorded this response” is a different claim from “this bed has this composition”. The inference may be justified, but its supporting calibration should be available.

The professional skill is integration with traceability. Preserve raw observations and explain transformations. For learners, a simple exercise can use a fictional core description and a fictional log, then introduce a small depth shift. The task reveals how apparent mismatches can arise from alignment rather than an entirely new geological process.

Contents · Previous chapter · Next chapter

Archaeological stratigraphy also reasons through deposits, cuts, fills and relationships. Its materials and questions differ from many geological sections, but the discipline of contextual order remains useful.

Consider an invented pit cut into two older deposits and later filled. The cut is younger than the deposits it truncates; the fill follows the creation of the cut. An object inside the fill can have been made long before the pit was dug.

That sequence distinguishes manufacture, use, discard, deposition and later disturbance. Treating all those moments as one date can produce a persuasive but misleading historical story.

The Teaching Archaeology guide develops context and material evidence. Teaching Geoarchaeology connects sediments with site formation, while Teaching Soil Micromorphology examines small-scale structures that can clarify processes invisible in a broad drawing.

An original classroom task can use paper layers with a drawn cut and a filled feature. Ask students to order the events before supplying any calendar labels. Then add an old object to a young fill and ask which inference changes.

The answer should preserve the physical sequence. The object constrains some possibilities, but it does not override the documented cutting relationship. This is a useful transfer from geological reasoning into human history.

Real excavation involves permissions, preservation and specialist methods. The educational task here is reconstruction from supplied records. It encourages careful reasoning without treating a classroom analogy as authority to disturb a site.

Contents · Previous chapter · Next chapter

A reproducible interpretation preserves enough of the observation trail for another competent reader to understand and test it. That means more than supplying a final colourful cross-section.

Keep locations, orientation, measurement methods, sample identifiers, photographs with scale and the distinction between observed and inferred boundaries. Record where exposure was incomplete. Preserve revisions so the reader can see why a correlation changed.

Imagine two invented teams drawing the same contact differently. One placed it at the first colour change; the other used a texture change ten centimetres higher. Their disagreement may concern the recognition rule rather than the existence of the boundary.

The useful repair is to state the criterion and revisit the interval. A shared label without a shared recognition rule creates false agreement. Conversely, a visible disagreement can improve the eventual definition.

Uncertainty can be shown directly: an approximate contact, an interval of possible positions, an alternative correlation or a statement that exposure is insufficient. Each is more informative than drawing every boundary with equal confidence.

A professional report also separates the evidence from its decision use. A stratigraphic model can inform groundwater, construction, resource assessment or conservation, but those applications require additional expertise and site-specific information.

Students can practise by exchanging notebooks and reconstructing a section without verbal assistance. Wherever the second reader must guess, the first record needs clarification. The exercise makes scientific communication part of the evidence rather than an afterthought.

Contents · Previous chapter · Next chapter

These checks expose different mistaken moves: confusing order with duration, resemblance with identity, and an average with a local rate. The correction should address the particular error. Once repaired, change the drawing or the numbers and ask for an explanation again. Understanding becomes visible when the relationship survives the change.

Contents · Previous chapter · Next chapter

Is stratigraphy just naming rock layers? Naming and mapping units are important, but the subject also establishes relationships, correlations and the record’s connection with time. A useful name comes with criteria for recognising the unit.

Is the oldest rock always at the bottom? That rule applies to an upright depositional succession under its stated conditions. Overturning, faults, intrusions and other relationships can change what a photograph appears to show.

Does a thicker bed mean a longer duration? Not necessarily. Deposition rates, compaction, erosion and interruptions vary. Thickness is an observation; duration requires a defensible temporal interpretation.

Can similar fossils date two beds exactly? They can support correlation, but identification, local occurrence, reworking and sampling matter. A shared fossil signal does not automatically provide an exact numerical age.

Why do stratigraphers use several classification systems? Rock properties, fossil content and time relationships answer different questions. Keeping those criteria separate makes the combined interpretation clearer.

Can a sequence tell a complete history? It can constrain a history, sometimes strongly. Missing deposition, erosion and incomplete exposure mean that preserved record and elapsed history are not identical.

What is a good beginner’s next step? Describe one supplied section carefully, establish its relative order, identify one uncertainty and propose an observation that would help resolve it. A modest answer with a clear evidence trail is a strong foundation.

Contents · Previous chapter · Next chapter

Build the physical and temporal framework How Science Works: Stratigraphy provides the wider technical subject map. The Gold Standard of Geology sets the Earth-science context. The Gold Standard of Sedimentology explains transport, deposition and the evidence behind sedimentary structures. The Gold Standard of Geochronology develops numerical ages and the events recorded by clocks.

Follow fossils, environments and chronology The Gold Standard of Palaeontology connects fossils with biological history. The Gold Standard of Systematics and Taxonomy explains the naming and identification decisions behind fossil taxa. The Gold Standard of Palaeoclimatology follows environmental reconstruction, and Shared Age Anchors and Lead–Lag Uncertainty examines comparisons between chronologies. Teaching Palynology develops pollen evidence and its contextual limits. Teaching Dendrochronology explains crossdating in tree-ring records.

Read human deposits with their context intact Teaching Archaeology, Teaching Geoarchaeology and Teaching Soil Micromorphology develop distinct scales of contextual evidence. Each belongs to a specialist practice rather than serving as an interchangeable method for every sediment.

Sources and further learning The International Commission on Stratigraphy’s Guide, lithostratigraphy chapter, unconformity chapter, biostratigraphy chapter, chronostratigraphy chapter and relationships chapter support the formal distinctions. Checked on 21 October 2026. The invented sections, calculations and teaching tasks illustrate reasoning rather than establish formal geological units or dates. Continue through How X Works or revisit The Gold Standard of Education for the learning framework.

Contents · Previous chapter · Continue through How X Works