Regolith geology of the Yilgarn Craton, Western Australia: Implications for exploration

This paper presents and reviews the processes responsible for the distribution and formation of regolith and associated landscapes of the Yilgarn Craton and highlights their implications for mineral exploration. It provides an analysis of regolith geology investigations that were conducted in many districts of contrasting contemporary geomorphic and climatic conditions.The Yilgarn Craton is composed of Archaean rocks, predominantly granitoids, that are crossed by north‐northwest‐trending belts of greenstones. It has an arid to humid climate at present. The gently undulating landsurface forms a partial etchplain and the topography is largely related to bedrock lithologies and a complex history of valley development and aggradation. Deep weathering has affected most lithologies and geological provinces across the craton. The depth of the weathered mantle may be as much as 150 m, but it varies considerably and rock outcrop may occur in any part of the landscape. The main factors influencing extent of weathering are rock type, mineralisation and deformation. Palaeomagnetic dating of deeply weathered regolith profiles suggest that they formed throughout the Phanerozoic.An idealised profile commonly comprises fresh bedrock, grading upwards into saprock and saprolite, commonly bleached towards the top, especially on felsic or sheared mafic rocks. This is overlain by a clay‐rich and/or quartz‐rich zone, a mottled zone and a ferruginous, bauxitic or siliceous upper zone. These horizons are formed by a combination of weathering and landscape processes. Landscape processes would have been continuous throughout the weathering period, with major environmental changes triggering particular erosional and depositional events. Thus, upper horizons, mottled zone, ferruginous duricrust and silcrete have developed in residuum, colluvium and alluvium of various ages.Weathering is the result of interaction between the hydrosphere, biosphere and lithosphere. During weathering some of the components of primary minerals are leached and secondary minerals are formed as residua. The pathways by which these minerals form are varied and complex. Biota were present in the regolith and it is likely that they and their associated chemical reactions played a significant part in the weathering process, as well as inorganic chemical processes. The final product of weathering of all rocks is a mineral assemblage of least soluble minerals (kaolinite, hematite, goethite, maghemite, gibbsite, anatase and boehmite) and the most resistant primary minerals (quartz, zircon, chromite, muscovite and talc), although neoformation of several generations of hematite, goethite, kaolinite and gibbsite may occur. Poorly crystalline minerals are an important constituent in surface or near‐surface materials. In addition, the more soluble minerals, including carbonates, sulfates and halides, occur in arid environments.The principal effects of weathering on element distributions relates leaching and retention of a range of elements to mineral transformations in the principal regolith profile horizons. However, the chemical patterns may not be consistent from profile to profile even for similar lithologies due to differences in groundwater regimes and topography. Minor and trace elements may be displaced from their primary host mineral. Thereafter, they may occur camouflaged in the newly formed secondary minerals, as major components of accessory new minerals, or in resistant minerals.In the present landscape, deeply weathered profiles may be preserved or partly eroded and buried beneath a variety of sediments. Three major regolith‐landform terrains were identified and the extents of these regolith‐landform terrains vary across the Yilgarn Craton. The first is dominated, in its upper part, by sands, ferruginous gravels and duricrust that commonly overlie mottled zones and saprolites. Ferruginous duricrust is developed in weathered Archaean bedrock and younger sediments that have been extremely weathered and/or indurated by Fe oxides. A second regolith‐landform terrain comprises saprolite with fresh rock in places. These may have been exposed by erosion of a pre‐existing weathered material, or may represent the most weathered form of the parent rock, which had never been capped with a ferruginous duricrust. A third terrain is of sediment‐dominated areas with fluvial, aeolian and/or lacustrine deposits, commonly several metres thick, that may be underlain by ferruginous duricrust, saprolite or bedrock. Sediments are highly variable in genesis, provenance, composition and thickness and were either derived from erosion of fresh and weathered Archaean bedrocks or the reworking of older pre‐existing sediments. They vary according to region, topography and age.Numerous buried palaeochannels occupy the lower parts of the landscape and are up to several hundreds of metres wide and many kilometres long. Drainage incision along palaeovalleys on the weathered landsurface resulted in development of channels that were subsequently filled with sand, lignite and kaolinite‐ and smectite‐rich sediments with lenses of ferruginous gravel. Palaeochannels are younger than the palaeodrainage system of broad, shallow valleys in which they occur and were probably incised during the final stages of rifting between Australia and Antarctica during the Early‐Middle Tertiary. Sediments were deposited under fluvial, lacustrine, estuarine and marine environments during the Middle‐Late Eocene. Further deep weathering occurred in both sediments and bedrock. Mixing occurred between the accumulating sediments and the underlying saprolite, possibly as a result of the formation of palaeosols. The collapse features and associated nodular and pisolitic materials in the underlying saprolite were probably formed during this period. Hematitic megamottles have developed in sediments and the upper part of some palaeochannel sediments contain ferruginous nodules and pisoliths. All this indicates post‐depositional weathering within the sedimentary sequence. The similarities in the nature and characteristics of palaeochannel sediments in the southern and northern regions suggest that similar conditions prevailed not only during the deposition of sediments, but also during their subsequent weathering. Parts of the palaeochannel sediments were eroded prior to deposition of Quaternary colluvium and alluvium. Colluvial, alluvial and aeolian sediments unconformably overlie palaeochannel sediments, ferruginous duricrust or saprolite. These sediments have been derived from increased erosion following the change to a more arid climate during the Late Miocene‐Pliocene, in part a result of instability caused by a reduction in the vegetative cover.A variety of soils occur in the three major regolith‐landform terrains, the nature of which is controlled largely by the parent material and erosional and depositional processes. There is a significant aeolian component in many soils, commonly in the finer fractions. Where these soils occur over basic or ultramafic rocks, the Ti/Zr ratio and quartz contents can be used to identify any aeolian contribution.Duricrusts occur in many different associations. In places, duricrusts of Fe, Si and Ca occur together in a single profile. Inland, ferruginous duricrust did not form on a simple, extensive, peneplained surface, but as a discontinuous cover on a broadly undulating plateau. It is developed on all rock types, but is particularly well developed on mafic and ultramafic bedrocks. By contrast, in the humid region of the Darling Range of the western Yilgarn, ferruginous duricrusts of various morphologies form an almost continuous sheet on all rock types. Transported hematite‐maghemite‐rich gravels at the base of the palaeochannel sequence are evidence that ferruginous duricrust existed prior to deposition of palaeochannel sediments in the Eocene. Ferruginous duricrusts have formed in residuum, in colluvium and in alluvium of various ages and are not necessarily associated with deep weathering. Thus, it is unwise to assign a single age to all ferruginous duricrusts or imply a single extensive surface of planation of continental extent. Residual duricrust (lateritic residuum) has developed from bedrock by essentially residual processes. In contrast, transported duricrusts (ferricretes) develop by impregnation and cementation of sediments by Fe oxides precipitated from groundwater, so that ferricretes have little direct relationship with the underlying rocks. In places they now form low hills because of relief inversion. Evidence of ferruginisation from the pre‐Eocene through to at least Pliocene can be seen in the arid parts of the Yilgarn Craton. However, this process is currently operating in humid regions of the Darling Range, where goethite‐rich ferricrete is forming on the edges of valley floors. Recent overprinting of ferruginous duricrusts has continued both inland and in the Darling Range. In the Darling Range, degradation of bauxitic ferruginous duricrust continues by dissolution of gibbsite and goethite. In the Kalgoorlie region, ferruginous duricrust is extensively modified by precipitation of carbonates to form calcretes.Inland with increased aridity, soil and groundwater‐related processes became dominated by gypsification, silicification and calcification. Silicification and calcification have affected a wide variety of regolith materials; calcification post‐dates silicification. There are at least two widespread episodes of silicification. One episode is related to silcrete formation and the other with red‐brown hardpan formation. It is possible that red‐brown hardpans were once more extensive, but were subsequently replaced by carbonates to form calcretes. The red‐brown hardpans are formed by partial replacement and cementation of the matrix and clasts by amorphous Si and aluminosilicates with minor goethite and hematite, accompanied by clay illuviation. Silicification of hardpan is associated with weathering within and outside soil profiles and appears to be still active. Pedogenic and groundwater calcretes are end members of a continuum that varies according to landscape setting and origin. Pedogenic calcretes occur in soils in association with a variety of rock types, but are more abundant in soils developed on greenstones. Their thickness and forms are largely controlled by topographic setting and the nature of the host material. Airborne accession is an important source of calcium for the formation of pedogenic calcretes, but weathering of bedrock, lateral transport by soil creep and soil solutions, and redistribution by biological processes are equally important. Groundwater calcretes are linear, tabular bodies occurring at or close to the surface and forming gentle mounds. They are Ca– and Mg–cemented valley deposits that are up to tens of metres thick.Many of the regolith types resulting from a complex array of processes of weathering, erosion and deposition have a distinctive pattern, and could be potential sampling media. However, a combination of a long weathering history and a variable degree of erosion have resulted in a landscape of highly variable and complex regolith. Thus, assessment of the nature and origin of the regolith, weathering history, geomorphological processes and regolith‐landform relationships are essential in determining the optimum geochemical sampling medium applicable in a particular terrain. A regolith‐landform framework and models of regolith evolution of the Yilgarn Craton provide a basis for exploration models and exploration strategy that, with appropriate modification, may be extended into similar terrain elsewhere.

Keywords:

• duricrusts
• exploration
• landscape evolution
• mineralogy
• petrology
• regolith
• regolith‐landform mapping
• sediments
• weathering
• Yilgarn Craton