B4 · Publication Volume 9

Australian Arid-Zone Regolith

deep weathering, transported cover and exploration design

Transferable arid-zone domains, old weathering, transported cover and sampling decisions
Transferable arid-zone domains, old weathering, transported cover and sampling decisions

Learning objectives

After this lesson, you should be able to explain how old deep weathering and later arid reworking can coexist; distinguish residual uplands, erosional plains, alluvial corridors, aeolian sheets, calcareous or saline zones and palaeodrainage; translate a regional case into process hypotheses rather than fixed rules; design domain-specific sampling media; and complete a residual-versus-transported interpretation with explicit transfer limits.

Start with a field problem

A low-relief arid landscape contains isolated ferruginous rises, broad sand sheets, ephemeral channels, calcareous soil and buried valleys. Bedrock exposure is sparse. A surface anomaly occurs on one rise and another occurs in fine material over a palaeodrainage. Should both be interpreted vertically? The regional setting says no: long weathering histories can be preserved on remnants, while younger wind and water pathways redistribute material across them.

This lesson uses public Australian research because deeply weathered, tectonically old and later arid landscapes clearly expose the problem. It does not provide a universal Australian profile or a template for a particular property. The transferable task is to recognise inherited weathering, truncation, burial and remobilisation in any covered dryland.

Core process model

An arid-zone regolith-landscape system may preserve several overlapping histories:

  1. Deep weathering: older warm or humid conditions produce saprolite, mottling and ferruginous or bauxitic differentiation.
  2. Erosion and relief inversion: resistant duricrust remnants survive while surrounding softer material is stripped; former low positions may become capped rises.
  3. Valley incision and fill: palaeodrainage cuts older profiles and is filled by alluvial, lacustrine or other sediment.
  4. Arid reworking: wind sheets, dunes, sheetwash, ephemeral channels, deflation lags and dust redistribute particles.
  5. Salt and carbonate redistribution: evaporation and groundwater movement precipitate carbonate, gypsum, silica or salts in position-specific horizons.
  6. Modern disturbance: access, grazing, excavation and infrastructure alter surface material and drainage.

The model predicts juxtaposition rather than one complete vertical profile. Residual saprolite may occur on a low rise beside transported clay in a buried channel. Ferruginous gravel may be detrital, while younger iron cement records groundwater. Calcrete may contain components acquired from atmosphere, dust, substrate and groundwater.

Landscape domains should be defined from process and material: residual upland, erosional plain, colluvial footslope, active alluvium, palaeovalley fill, aeolian cover, saline depression and duricrust remnant. Each domain has a different likely sample support and transport direction. Boundaries are hypotheses to be tested by profile and provenance evidence.

Evidence and measurement

Integrate landform mapping, regolith logging, elevation, imagery, drainage, shallow geophysics, mineralogy and geochemistry. On ferruginous rises, test whether fabric and immobile components connect to underlying substrate. Across sand sheets, estimate thickness, grain provenance and degree of pedogenic overprint. Along palaeodrainage, distinguish channel fill, groundwater cement and underlying saprolite.

For carbonate or salt-rich material, record morphology, host fabric, depth, mineralogy and relation to present and past water movement. Carbonate can dilute whole-sample metal concentration, bind selected elements or create hardpan that redirects flow. Selective extractions need operational interpretation and mineralogical verification.

Sampling media may include residual soil, lag, ferruginous gravel, specific fine fractions, calcrete, vegetation or shallow subsurface material. No medium is universally best. A pilot orientation study compares background variability, target contrast, source displacement, collection reproducibility and contamination sensitivity within each domain.

Worked example

Consider a synthetic 12 km transect divided before chemical analysis:

| Domain | Cover model | Chosen first-pass medium | Vertical interpretation allowed? | |---|---|---|---| | residual rise | 0.5–2 m residual soil over saprolite | soil B horizon plus saprolite check | conditionally, after parent test | | sand sheet | 3–8 m aeolian sand | fine fraction and basal interface | no for surface sample alone | | palaeovalley | 15–40 m alluvial/lacustrine fill | targeted basal or hydromorphic medium | no; flow geometry required | | erosional plain | thin mixed lag over variable substrate | lag separated by clast type | only for proven residual clasts |

Synthetic results show 60 mg/kg of an element on the residual rise, 45 mg/kg in ferruginous coatings above the palaeovalley and 12 mg/kg in bulk sand. The values cannot be ranked as one population. Model R treats the rise as local residual expression and the palaeovalley value as displaced hydromorphic signal. Model T treats both iron-rich media as transported products sourced upslope. Provenance, substrate continuity, coating age and cross-valley asymmetry discriminate them.

The completion decision boundary is explicit: vertical source inference is permitted only within mapped residual domains where profile continuity and parent equivalence are demonstrated. Elsewhere, results guide pathway tests rather than vertical targeting. A map should show that boundary as prominently as the anomalies.

Misinterpretations and uncertainty

Do not equate aridity with shallow weathering. Thick profiles may be inherited. Do not assume a hardpan is the top of bedrock or a palaeodrainage is visible in present drainage. Deflation can concentrate resistant grains, dunes can bury earlier surfaces, and exceptional floods can move material across normally disconnected domains.

Regional analogues are not site models. Climate history, lithology, relief, groundwater, cover age and disturbance vary. A sampling medium validated in one domain cannot be transferred without an orientation test. Sparse outcrop makes circular reasoning especially dangerous: regolith interpretation should not be accepted merely because it improves the apparent anomaly pattern.

Practical investigation

Create a domain map for a hypothetical arid covered terrain using elevation, imagery, regolith observations and five shallow profiles. Assign each polygon an origin hypothesis, confidence, likely transport direction, valid sampling media and a no-inference condition. Keep residual and transported models visible simultaneously.

Design the completion programme: choose media and fractions, orientation transects, quality controls and at least one profile test in every domain. Define the interpretation boundary before viewing chemistry. After synthetic results are supplied, test residual and transported origins and nominate one observation that could reverse the preferred model.

Mastery check

  1. How can deep weathering coexist with a modern arid climate?
  2. Why can duricrust remnants produce relief inversion?
  3. Which evidence permits vertical interpretation on a residual rise?
  4. Why must palaeovalley signals be interpreted with flow geometry?
  5. What makes a regional analogue transferable without becoming a template?

Sources and further reading