B4 · Publication Volume 9

Geomorphology, Regolith and Surficial Geology

Explains landscape and regolith processes that preserve, transport or conceal exploration signals.

Purpose of this book

Geomorphology explains how landforms are produced and reorganised by uplift, weathering, erosion, transport and deposition. Regolith geology explains the loose or altered material between fresh rock and the land surface, including material formed in place and material carried from elsewhere. Surficial geology connects these processes to deposits, landforms and near-surface boundaries that can be mapped. Together they answer a practical question: what does a surface observation represent, and how far can it be traced back toward its source?

A soil anomaly, ferruginous gravel, drainage line or smooth digital-elevation surface is not self-interpreting. The sampled material may be residual, colluvial, alluvial, aeolian, lacustrine or anthropogenically disturbed. Its chemistry may reflect bedrock, weathering, hydromorphic transfer, mechanical concentration, adsorption, contamination or a mixture. A present valley may reuse an older valley; a resistant cap may preserve an abandoned surface; a sharp geochemical boundary may be a transported contact rather than a bedrock contact. This book therefore treats landscape, regolith and sampling medium as one coupled evidence system.

The sequence begins with landscape mass balance, rivers, slopes and sediment pathways. It then develops physical and chemical weathering, regolith-profile architecture, ferruginous materials, gossans and supergene enrichment. The final lessons turn these ideas into exploration-under-cover decisions, responsible interpretation of remotely sensed terrain and a transferable arid-zone case. The completion task requires a sampling-medium design and an explicit test of residual versus transported signal origins.

General and institution-neutral scope

This is a general, institution-neutral tutorial. It has no affiliation with, sponsorship by, endorsement from or curriculum relationship to any company or individual. It is not written for a named project, mine, consultancy, university, software product or private dataset. All numerical examples use synthetic instructional data, and all field scenarios are invented unless a published regional study is explicitly identified in a source note.

People, standards bodies, public agencies and journals named in source notes identify technical sources only. Their citation does not imply authorship, approval, partnership or participation in this tutorial. A website may host the material without becoming its publisher, scientific authority or subject. The Australian case in the final lesson is a regional process example, not a commercial case study or an endorsement of any organisation.

Diagrams are simplified process models rather than surveyed maps, assay results or predictions for a real place. Colours do not encode universal lithology, grade, risk or confidence. Site-specific exploration, engineering, environmental, heritage or land-use decisions require authoritative local data, appropriate permissions, competent professional review and the laws and standards of the relevant jurisdiction.

Learning objectives

By the end of the book, you should be able to:

  • construct a landscape mass balance that distinguishes uplift, weathering, erosion, transport and storage;
  • trace sediment through hillslope, channel, floodplain, lake, coast and wind-driven pathways;
  • explain how physical disintegration and chemical reactions create mobile solutes, secondary minerals and residual mass;
  • log a regolith profile without assuming that every vertical change is a soil horizon or that every lower unit is residual;
  • distinguish lateritic residuum, ferricrete and gossan by process, geometry and evidence rather than colour alone;
  • build a mass-conserving supergene model with leached, oxidised and enriched domains;
  • select a sampling medium by testing how a target signal could be preserved, transported, masked or contaminated;
  • derive and audit basic terrain metrics from a digital elevation model while respecting scale and resolution;
  • use an arid-zone regolith-landscape model as a transferable hypothesis rather than a regional template; and
  • design an investigation that exposes the boundary between defensible interpretation and unresolved alternatives.

What you should already know

You should be able to distinguish observation from interpretation, read geological maps and cross-sections, work with units, ratios, logarithms, basic probability and uncertainty, and describe common minerals, rocks, sediments and stratigraphic relations. Familiarity with sedimentology, mineralogy, geochemistry, remote sensing and map projections is helpful. Equations are explained at an introductory level, but this volume is not a substitute for full courses in soil science, hydrology, geochemistry, geotechnical engineering or quantitative geomorphology.

The essential prerequisite is support-aware evidence discipline. Every observation needs a location, date, scale, depth or elevation reference, medium, particle-size fraction, preparation and analytical method where relevant, uncertainty and a stable identifier. A concentration measured in a fine soil fraction cannot be compared uncritically with a whole-rock value. A pixel value cannot be treated as a point measurement. A mapped regolith unit needs criteria, boundary confidence and the observations used to recognise it.

Core process model

Use a linked five-reservoir model:

  1. Fresh and altered substrate: bedrock, saprock and saprolite supply minerals, solutes and particles.
  2. Residual regolith: material remains near its parent while volume, chemistry and mineralogy change.
  3. Transported regolith: gravity, water, wind, ice or human activity moves and mixes material.
  4. Temporary stores: soils, fans, terraces, floodplains, dunes, lake beds and duricrusts preserve selected parts of the history.
  5. Export: sediment and dissolved load leave the local system or enter a deeper sink.

At landscape scale, elevation change can be written schematically as


\frac{\partial z}{\partial t}=U-E-\nabla\cdot\mathbf{q}_s,

where U represents surface production by uplift or another vertical boundary motion, E represents local removal or lowering, and \nabla\cdot\mathbf{q}_s represents divergence of sediment flux. The symbols are bookkeeping terms, not directly interchangeable measurements. Chemical mass loss can lower a surface without exporting an equal volume of visible sediment; deposition may raise a surface while burying an older weathering profile.

For a component j, a regolith mass balance is


\Delta M_j=M_{j,\mathrm{in}}-M_{j,\mathrm{out}}+M_{j,\mathrm{reaction}}.

The reaction term may be positive for a precipitated secondary phase and negative for a dissolved primary phase. Apparent enrichment in concentration can occur even when total mass is lost, because other components are removed more strongly. This distinction between concentration and inventory is central to laterite, gossan and supergene interpretation.

Evidence and measurement

Build every interpretation from four linked evidence layers. First, map landform geometry: relief, slope, curvature, drainage, surface texture, breaks of slope, terraces, fans and closed depressions. Second, log material: grain size, sorting, clast shape, fabric, matrix, cement, colour using a stated method, mineralogy, structures and boundary character. Third, measure chemistry and physics with declared support: fraction, depth interval, density, moisture state, preparation, digestion or sensing method, detection limits and quality controls. Fourth, reconstruct sequence using superposition, truncation, soil development, cement generations, palaeosurfaces and independent age constraints where available.

Keep three boundaries separate:

  • the mapped boundary, drawn from observable changes at a chosen scale;
  • the process boundary, where the dominant origin or transport mechanism changes; and
  • the decision boundary, beyond which a sampling or interpretation rule is no longer considered valid.

These boundaries may not coincide. A visually uniform sand sheet may cross several source domains. A ferruginous horizon may cut both residual and transported units. A terrain-class boundary derived from elevation may not be the correct boundary for soil chemistry.

Worked example

Consider a synthetic 1 km² catchment used only for instruction. Bedrock weathering supplies 2,400 tonnes of solid material per year and atmospheric input supplies 100 tonnes. Dissolved export removes 700 tonnes, channel sediment export removes 1,100 tonnes and measured floodplain storage increases by 500 tonnes. The residual is


2{,}400+100-700-1{,}100-500=200\ \text{t yr}^{-1}.

The 200 tonnes is not automatically an error. It may represent hillslope storage, an unmeasured grain-size fraction, density uncertainty or inconsistent time windows. If the combined uncertainty of the terms is larger than 200 tonnes per year, the budget is compatible with closure. If it is much smaller, the conceptual model is missing a pathway.

Now suppose a surface sample contains 80 mg/kg of an element while nearby parent rock contains 40 mg/kg. Calling this “twofold enrichment” describes concentration only. If weathering reduced bulk mass to 40% of the parent inventory and no element entered or left, the concentration would rise to 100 mg/kg. The observed 80 mg/kg could therefore represent net loss of the element despite its higher concentration. Inventory, density, thickness and an immobile reference are needed before inferring addition.

Misinterpretations and uncertainty

Avoid equating present climate with the climate that produced the profile. Deep weathering can be inherited and later dissected, buried or re-cemented. Avoid treating soil, regolith and transported cover as synonyms. A soil is organised by pedogenic processes; regolith is broader; cover may include unweathered sediment. Avoid assuming vertical sequences are continuous time series: unconformities, stone lines, lag deposits, channel cuts and imported dust can interrupt them.

Colour is useful but non-unique. Red or yellow material may reflect iron oxidation, coatings, source lithology, fire or imported sediment. A hard iron-rich layer may be residual laterite, groundwater ferricrete, reworked gravel or gossan. A geochemical anomaly can be displaced laterally or vertically, diluted, sharpened by adsorption, or introduced by access tracks and previous disturbance.

Uncertainty has at least five sources: observation error, spatial variability, process non-uniqueness, hidden thickness or boundaries, and temporal change. Report all five rather than compressing them into one confidence adjective.

Practical investigation

Choose a small catchment or a fully documented teaching dataset. Divide it into source, transport, storage and export domains. Map at least two competing versions of the residual-versus-transported boundary. For each observation, record what it measures, its support and which boundary interpretation it favours. Construct a sediment and component budget with uncertainty ranges.

Then design a transect that crosses a hillslope, drainage margin and channel store. Specify the medium, depth, particle-size fraction, replicate spacing and quality controls. Before seeing any analytical results, write predictions for a residual-source model and a transported-source model. The investigation succeeds when the observations can reject or narrow at least one model, not when a visually attractive map is produced.

Mastery check

  1. Why can a surface sample be chemically enriched while having lost total element mass?
  2. How do a mapped boundary, process boundary and decision boundary differ?
  3. Which observations distinguish residual material from a transported unit that resembles its source?
  4. Why is present topography not a complete record of landscape history?
  5. What would make a sediment budget meaningfully auditable?

Sources and further reading