澳大利亚西北部皮尔巴拉地区Yandi河谷沉积型铁矿岩调研分析

冯 睿中国辽宁金鼎集团驻津巴布韦津鼎矿业有限责任公司 115100



澳大利亚西北部皮尔巴拉地区Yandi河谷沉积型铁矿岩调研分析

冯 睿
中国辽宁金鼎集团驻津巴布韦津鼎矿业有限责任公司 115100

摘要：Yandi Channel Iron deposit is located in the Hamersley Province, Pilbara region of northwestern Australia and contains one of the world’s largest iron resources. Yandi CID was deposited in pre-existing palaeochannel incised into Hamersley Group. Hamersley Group contains BIFs and provides source of ore. The deposition began between Late Oligocene and Mid Miocene. During Early-Mid Miocene, the local climate increased continuously and accompanied with heavily rainfall, resulting in weathering and rework processes to the ore body. The ore body is contained in Marillana Formation, which consists of Munjina Member, Barimunya Member (main CID zone) and the uppermost Iowa Eastern Member. Yandi CID is characterised by a simple mineralogy, dominantly goethite and hematite. Hematite granules consist of pelletoids and pisoids. The pelletoids formed upper part of ore body and pisoids formed lower part. Weathering of iron fragments leads to dissolution and precipitation of goethite and formed vitreous goethite surrounding the hematite granules. Some goethite formed matrix that contains large amount of porosities, quartz and hematite granules. Silica mobilization happened multiple times. Some dissolved and precipitated in hematite granules, and some dissolved in groundwater and precipitated in macro voids in matrix. The climate change also leads to the replacement of vegetation. Wood fragments were driven by river flow and deposited in upper CID. Lowering of the water table resulted in developing of extensive alteration zones which reduced the quality of ore. Overall, Yandi CID is unique production of iron-rich materials, erosion, global climates changes, plates drifting, rainforest vegetation and palaeochannel.

关键词：Hamersley Group, Marillana Formation, hematite, pelletoid, pisoids, matrix, vitreous goethite

Introduction

The Pilbara region of northwestern Australia has dominant global iron-ore resources, there are three major types of iron ores in this region which are bedded iron deposits, detrital iron deposits and channel iron deposits, with resources totalling over 32000 Mt[1]. Channel Iron Deposits currently supply over 40% of the iron ore exported from Hamersley Province in southern part Pilbara region[2]. Yandi CID is characterised by simple mineralogy, dominantly goethite and hematite, and forms a huge ribbon-like body. However, it is overprinted by silica and alumina alternation with clay-filled pods and pipes, resulting in zones of lower quality ore and waste within the orebody[3]. More importantly, there is limited understanding of how the iron-rich fragments were deposited in the palaeochannel, the timing of deposition or the source of the material and how these facts relate to the genesis. The concept of continuous weathering points out the hypothesis that CID are the product of intense weathering and erosion processes depositing iron-rich detritus in areas of low relief[4][5]. Channel iron deposits could therefore be considered the result of a climatic accident; developed from highly unusual local conditions in a geologically short period of time[6].

Geological setting of Yandi CID

This chapter presents the geologic and geomorphologic setting of the Yandi CID. Description of lithostratigraphic units combined with observations made by previous work

Regional geological setting

The Hamersley Province covers an area of about 80 000 km2 and is located in the southern part of the Archaean to Paleoproterozioc Pilbara Craton in northwest Australia[7]. The rocks of the Hamersley Province are composed of ～2,800-2,300 Ma sedimentary rocks of the Mount Bruce Supergroup, which overlies Archaean granitoid-greenstone terranes of the Yilgarn and Pilbara blocks[7].

The Mount Bruce Supergroup is composed of the conformable Fortescue, Hamersley and Turee Creek Groups, and is unconformably overlain by the Lower Wyloo Group[1]. Mt Bruce Supergroup is a 2.5 km thick succession of BIFs, cherts, dolostones, pyroclastic and hemipelagic claystone and felsic volcanics intruded by dolerite sills and dykes[1].

The majority of faults and fold in the Hamersley Province is broadly east-west[8]. Two structural environments are represented which include west to northwest trending open folds and E-W trending folds in the southeast[8].

Exhumation of the Wyloo Group led to the development of colluvium during the Tertiary, associated with calcareous sediments, silcrete and ferricrete[3]. The colluvial deposits are across the Hamersley Province and host CID in the Marillana and Robe Formations[3]. These formations are parts of ancient drainage systems that contain the world-class CID at Yandi.

The regional landscape consists of erosional and depositional surfaces which include scarps and gorges, alluvial-colluvial fans and plains, mesas, plateaux and inselbergs[1]. Yandi CID is exposed in mesas along palaeodrainage in an intermontane-valley within the Weeli Wolli Formation[7]. Mesa formation occurred as the valley floor and walls were eroded leaving the CID in inverted relief[7]. Water flow through the present day Marillana Creek drainage system occurs after heavy rainfall generally during the winter[9]. The creek bed contains angular to rounded, sand- to boulder-size clasts of CID, iron-formation, shale, dolerite, alluvium/ colluvium and calcareous sedimentary rocks[9]. The latter cover the CID in the headwater region. Vegetation around the creeks includes paperbarks, river red gums, snappy gums, coolibahs, and wattles[1]. Spinifex occurs in alluvial outwash areas[1].

Local geological setting

The Yandi CID was deposited in Marillana Creek palaeochannel that incised into the BIF – dolerite - shale Weeli Wolli Formation along the axis of the Yandicoogina Syncline[6]. The palaeochannel is approximately 500-650 m wide and up to 120 m deep along over 90 km length[6]. The palaeochannel host the CID in Marillama Formation.

The Marillana Formation is composed of ironrich sediments that fill in the Tertiary palaeochannel of Marillana Creek[7]. The CID shows as hollow mesas around the present-day creeks. It is unconformably overlain by calcrete and silcrete of the Oakover Formation (upstream parts of the channel) and unconsolidated alluvium/colluvium[7].

The Marillana Formation can be divided into the basal Munjina, the central Barimunya and the uppermost Iowa Eastern Members[10].

The Barimunya Member is the main ore-containing unit of the Marillana Formation and consists of the Lower and Upper CID units with an ochreous clay zone in the middle.

The Lower CID is variable in thickness, ranging from 50 m in the Western deposits to less than 10 m in the Eastern deposits[7]. It is variably weathered by alteration of the original red-brown hematite-goethite granules and brown goethite-rich matrix, to yellow-brown ochreous and goethite-dominant weakly textured material, to near massive ochre and tertiary brown goethite[1] [5] [6] [10]. This progressive alteration of original textures via iron oxide solution-reprecipitation processes may be related to postdepositional alteration below earlier water table(s)[3] [10].

The top of the Lower CID is defined as semicontinuous ochreous clay that is interpreted to have been deposited from suspension during waning flow[7]. It occurs in the Western and Central deposits at Yandi, and absences in the Eastern deposits[7].

The Upper CID consists of largely unaltered redbrown hematite-goethite granules within a goethitecemented matrix.

Ferruginised wood fragments are common and maghemite is present in the upper parts[1] [5] [6] [10]. The Upper CID is low to moderate silicified as a result of postdepositional alteration from groundwater[7].

Deposition of the Barimunya Member has been interpreted to reflect increasingly warmer and dry conditions in northwestern Australia during the mid-late Miocene (15-11 Ma), consistent with the Miocene climatic optimum, characterised by global warming and high sea levels[6].

Petrology

In order to better identify lithology and mineralisation, one diamond drill core and five thin sections were provided by Applied Geology, Curtin University, Western Australia. This chapter mainly to describe the core logging and thin sections observation.

Geological logging

The core logging had been finished from August to September 2013. This diamond drill core starts from 24m and ends on 68m.

The upper 24m to 47m consist of approximately 30% largely fine grained red-brown hematite and goethite granules, the grain size ranges from 1mm to 6mm. The hematite and goethite granules are surrounded by brown vitreous goethite (25%) and yellow-brown ochreous goethite (25%). The vitreous goethite is silicified textured and translucent that indicates post-depositional alteration from groundwater. Some small white clay zones also present, generally 50cm thick, which may be a result of palaeo-water table fluctuations.

Ferruginised wood fragments are common, only celltissue is replaced by goethite, maintaining the original cellular porosity of the replaced organic material. The percentage of wood fragments is up to 20% but distributed variably.

Some root structures have been found cutting the entire core, which related to post-depositional vegetation.

The middle 47m to 47.85m consist of increased yellow-white ochreous goethite and clay (90%), which may be a result of pre-existing flood plain. It formed a boundary that separates the upper and lower parts of core, judging from the highly extent of weathering characteristic.

The lower 47.85m to 68m (EOH) consist of fine grained red-brown hematite-goethite granules (1-6mm) and brown goethite-rich matrix, to yellow-brown ochreous and goethite-dominant weakly textured materials, to near massive ochre and tertiary brown goethite. The ochreous goethite increased to nearly 40%. It also shows more clay zones compared with upper 24-47m. These characteristics indicate the weathering of original textures via iron oxide solution and precipitation processes that may be attributed to post-depositional alteration below earlier water table. Rare fossil wood fragments can be found in this part and secondary silicification is common throughout this part.

Stratigraphy

The Lower CID ranges from 47.85m to EOH. It is variably weathered as a result of alteration of the original red-brown hematite-goethite granules and brown goethite-rich matrix, to yellow-brown ochreous and goethite-dominant weakly textured materials, to near massive ochre and tertiary brown goethite.

The middle ochreous clay zone overlies the top of the Lower CID, approximately 0.75m thick, which is defined by continuous ochreous clay or kaolinite marker horizon. It is interpreted that the clay zone has been deposited from pre-existing flood plain. It forms a natural boundary of Lower CID and Upper CID, and this boundary is interpreted mainly based on the extent of alteration.

The Upper CID forms from 24m to 47m. It consists of largely unaltered red-brown hematite-goethite granules within goethitic matrix. Fossil wood fragments are common and root structures present. The Upper CID is weakly to moderately silicified as a result of postdepositional alteration from groundwater. The Upper CID becomes more weathered towards the base that indicates the potential water table.

Thin Sections Observation

Five thin sections had been provided. The depths of these five thin sections are 30.4m, 33.6m, 35.2m, 43.9m and 54m. The first four sections are from Upper CID and the last one (54m) is from Lower CID.

The first thin section (30.4m) represents the large red irregular hematite pelletoids a.k.a nodules are leached by yellow-yellowish brown goethitic cortex in the goethite cemented matrix. The average grain size of hematite pelletoids is 10mm. The goethitic cortex is generally very similar to the matrix, but judging from their textures and position, they could be formed in different processes. The matrix formed interlayered texture that indicates the solution and precipitation of goethite happened in an open space environment. The porosities are common within the matrix that may be the result of intense weathering process under the earlier water table.

The next thin section taken from 33.6m shows the large amount of secondary quartz formed in macro voids in matrix, the average grain size is 3mm. This indicates the later silica mobilization – silica dissolved and precipitated in the voids of iron-rich materials with the water flow. The source of silica might be local surrounding siliciclastic rocks and migrated by river flow. There are large amount of fossil wood fragments leached by goethite in matrix, but they still present the general characteristics of organic materials. The wood fragments are irregular coarse grain, yellowish coloured, the average grain size is 10mm and some even over 1cm. The fragments are without goethitic cortex, these are likely to have been replaced within the river sediment, since they lack the features of soil, and leached by goethite rather than hematite, indicating a protected environment[6].

Thin section taken from 35.2m represents the most common feature of Yandi CID which is large amount of hematite pisoids with goethitic cortex within the goethiterich matrix (Fig. 1). The hematite pisoids are rounded cream colour, fine grain, the grain size is generally 8mm, some larger than 1cm. The hematite pisoids are much different compared with pelletoids, which indicates the two different types of hematite, each with different depositional process. The outer goethitic cortex is grey colour, generally 1mm thick. The cortex probably formed together with hematite in the earlier process. The goethite-rich porous matrix is grey colour, filled by quartz and wood fragments, which indicates the later weathering process below the water table.

Thin section taken from 43.9m represents the general features of goethitic alteration. It contains the large wellpreserved fossil wood fragment impregnated by goethite, the cell tissue still retains the organic features. The fragment is over 1cm and without cortex, which indicates the alteration happened in a protected environment. Besides, it also shows the well-preserved cream coloured hematite pisoids surrounded by grey coloured goethiticcortex within the porous goethite-rich matrix. Goethitic cortex shows varied levels of dehydration to hematite due to periodic surface exposure. Large amount of porosities within the goethite-rich matrix indicates the highly intense weathering process below the water table.

Figure 1 Photomicrographs of Yandi CID thin section – 35.2m. Large amount of hematite pisoids surrounded by goethitic cortex within the goethite-rich porous matrix. Photo taken in transmitted light, FOV=5mm.

The last thin section taken from 54m represents a range of accretionary pisoids in a porous goethitic matrix. The hematite pisoids are rounded, fine grain, with cream colour, the average grain size is 5mm. Pisoids with accretionary ferruginised soil layers on complex accretionary nuclei to demonstrate the genesis of weathering process. The goethite layers show varied levels of dehydration to hematite due to periodic surface exposure. The grain size of Lower CID is smaller than Upper CID and the quartz filling in the macro voids is much small than 33.6m, which indicates the two different silica mobilization processes. The quartz only exists in hematite pisoids and goethite-rich matrix, not in the goethitic cortex. It suggests that the mobilization of silica happened in multiple processes, with different period and environment, and the source of silica may be also different. No wood fragments can be observed in this thin section, which indicates the vegetation only participated in later weathering process and deposited in the Upper CID.

Geochemistry

23 elements has been analysed using whole-rock method (XRF) from Applied Geology, Curtin University, Western Australia. Plotting the geochemistry for selected elements (Fe, Al, Si, P, Mn etc.) against main lithological units can interpret the different weathering processes.

The average grade of iron is approximately 59%. The Lower CID is generally higher than Upper CID. There is an apparent drop around 52m that indicates the middle ochreous clay zone. The lower parts of Upper CID indicate the potential clay zones that exist in the ore body. Generally, the Upper CID has more clay zones than Lower CID which may be result of later weathering process after the deposition of iron-rich materials.

The compositions of SiO2between 15m to 32m are much higher than others in Upper CID, this is probably because the later silica dissolution and precipitation process after the deposition of CID. The Yandi CID experienced multiple weathering processes after earlier deposition. The Upper CID has more silica compositions than Lower CID, the silica mobilization happened together with clay zones and the source of silica may be local surrounding siliciclastic rocks and migrated by river flow. The Lower CID has less silica compositions and this probably can explain the reason why the silica filling in macro voids in goethite-rich matrix in Upper CID is much larger than Lower CID, and distribution of silica appears to be different, only in hematite granules and goethite-rich marix.

Al2O3is one of chemical components of kaolinite, which is the major mineral of clay zone. The Al2O3- XRFScatter Plot shows the trends of Al2O3match the clay zones perfectly. The Upper CID has more Al2O3than Lower CID that indicates the later weathering process on the Upper parts. The middle 52m increased dramatically and points out the depth of ochreous clay zone.

Generally, the geochemical analysis shows that the Lower CID has higher iron grade than Upper CID. The Upper CID has more clay zones that may be result of later weathering process after deposition. Due to the highly intense river flow and flood activities, the Upper CID was deeply altered, resulting more waste than Lower CID.

Genesis of Yandi CID

Time of Formation

Tentative timing of the deposition of the Cenozoic Detritals of the Hamersleys is based largely on the detrital-filled strike valleys formed, possibly in the Early Palaeogene[1][3], preferentially along the lower carbonaterich section of the Wittenoom Formation, where the potentially complete sequence is preserved[6].

Palaeomagnetic weathering ages around Australia [6] show an interesting clustering at 60 ± 10 Ma and others at 10 ± 5 Ma and 180 ± 10 Ma, with the 10 Ma group broadly consistent with the timing. Heim et al.[11]have determined 14 - 5 Ma uncorrected uranium- helium ages for post-depositional supergene infill goethite from CID.

Deposition Process

The Yandi CID was deposited in pre-existing palaeochannels that incised into the Weeli Wolli Formation of Hamersley Group within Mt Bruce Supergroup. The Hamersley Group consists of BIF, rhyolite, basalt, shale, chert and carbonate. The exhumation of Weeli Wolli Formation resulted in development of colluvium, siliciclastic rocks, calcareous sediments, silcrete and ferricrete. The primary mineral, which is mainly hematite, is deeply detrital iron fragments from surround deeply weathered BIFs (Fig.2). During late Oligocene to mid Miocene, these detrital iron fragments transported by river flow and deposited in channels. Some fragments are rounded and some are irregular.

During Late Oligocene and Early Miocene, the climate of northwestern Australia increased continuously, the widespread distribution of brown coals across Australia indicates that an absolute increase in precipitation over the Pilbara region also occurred[3]. Increased climate resulted in heavily rainfall, the river flow carried iron fragments, gravels and sands to the depositional position. The pisoids formed lower parts of ore body and pelletoids a.k.a nodules formed upper parts. These materials were weathered, resulting the dissolution and precipitation of goethite, goethite formed vitreous goethite and surrounded hematite granules to formed cortex.

A major period of emergence in the Early-Mid Miocene is interpreted to have resulted in the suspension settling of the Ochreous Clay and goethite alteration of hematite granules. The extensive silica in-filling of pore spaces in the upper part of the Lower CID and the transition to peloidal rocks in this zone probably reflects alteration along an ancient water table. This alteration is interpreted to have occurred post-emergence because the pisoids are overprinted by silica.

Figure 2 Genesis Model of Yandi CID, pic draw by Illustrator.

Pelletoid generation continued, or may havediminished, and pelletoids were stripped from the landscape[3]and a warming period followed. Erosion of the Hamersley Range provided the detritus for the granulepebble conglomerate and clay units in the base of the Iowa Eastern Member[6]. Clasts of Proterozoic Joffre Formation BIF accumulated with fire-generated maghemite granules, rare pelletoids and ferruginised wood to form the polymictic and maghemite conglomerates of the granulepebble conglomerate facies[1][3]. The coarse clast size of this facies suggests formation under similar flood conditions as the iron conglomerate facies[1][3]. Nontronitic clay settled from suspension to cap the unit[12]. These conglomerate and clay facies first occur near the junction of the palaeo-Herberts and Marillana Creeks and thin downstream, suggesting derivation of the clastic material specifically from the scarps of the Hamersley Range to the north[3].

Latter, pelletoid formation had probably ceased as a result of the increasingly arid conditions across much of the continent during the Late Neogene[6]. The Iowa Eastern CID therefore probably formed by headwater erosion of the Barimunya Member, and redeposition downstream on the Iowa Eastern Clay as indicated by the scoop-shaped upper contact of the Barimunya Member in the western extent of the CID[3]. This Late Miocene climatic change has been associated with the drift of the Pilbara region into relatively low palaeolatitudes at that time[3]. This interpretation is supported by burial of the Marillana Formation under calcareous sedimentary strata of the Oakover Formation in the Late Miocene[3]and Quaternary alluvium.

Both during and after deposition, the CID was subjected to groundwater movement with consequent modification in the more susceptible zones. Thus, silica mobilization happened multiple times. Silica dissolved and precipitated in macro voids in goethite-rich matrix. Some silica migrated by river flow and altered the Upper CID, resulting in lower grade zones. The distribution of quartz within the hematite granules and goethite-rich matrix, together with Al2O3support this hypothesis.

Zones of Lower Quality Ore and Waste

According to geochemical analysis, the upper part of Barimunya Member are likely to have higher silica and alumina contents than the lower part, and the higher silica and alumina are accompanied with ochreous clay zones, this relationship is a result of alteration during exposure and surficial weathering process.

Groundwater flow along ancient and pre-mining water tables resulted in alteration and changes in porosity producing zones of lower quality ore and waste. Large clay pods are common and formed preferentially along zones of higher porosity, such as bed contacts, and contribute to the increased alumina and silica[3]. Another major zone of alteration is coincident with the pre-mining water table. This zone consists of silicified peloidal rocks enriched in silica and alumina.

Lower quality ore is common along the margins of the deposit[1][3]. These marginal areas contain higher proportions of clay pods[1]. Goethite alteration resulting in increased rock friability is also common along these zones.

Discussion

(1) Heim et al.[11]using He ages measured a downward younging trend in upper 30m of profile, they argued that the iron cementation occurred at groundwater and atmosphere interface and was driven by water table drawdown. This is obviously opposite with my genesis model, but I cannot say which one is right, because data of mine do not show this trend and due to the limited number of samples studied, I can only suggest that the drawdownprocessing cementation model driven by water table cannot apply to my investigated in this study.

(2) The first silica mobilization happened together with iron-rich fragments been deposited, and after CID deposition, the silica was driven by groundwater consequently and altered the upper part of Barimunya Member. This hypothesis can explain the thin sections observation and geochemical data, and support my genesis model. However, the source of silica and how it migrated still remain uncertain. I personally assume the silica is from local siliciclast rocks and driven by river flow and ground water after deposition.

(3) The large amount of porosities within goethiterich matrix indicates that the weathering process happened in a protected environment, probably under the earlier water table. However, the interlayered texture of goethite-rich matrix indicates it happened in an openspace environment. How to explain these two different formations still need further work.

Conclusions

Although there are many unsolved mysteries about Yandi CID, the genesis of Yandi CID is generally due to the highly continuous weathering processes of local surrounding deeply weathered iron-rich materials. And there is no doubt that:

(1) The Yandi CID in Pilbara region is one of the world’s largest iron-ore resources.

(2) The Yandi CID was deposited in pre-existing palaeochannel incised into iron-rich rocks of Hamersley Group.

(3) Hamersley Group contains iron-rich materials such as BIFs and provides source of ore to Yandi CID.

(4) The deposition began between Late Oligocene and Mid Miocene. During Early-Mid Miocene, the local climate increased continuously and accompanied with heavily rainfall, resulting in weathering and rework processes to the ore body.

(5) The ore body is contained in Marillana Formation, which consists of basal Munjina Member, central Barimunya Member (contains the main CID) and the uppermost Iowa Eastern Member.

(6) The Yandi CID is characterised by a simple mineralogy, dominantly goethite and hematite. Hematite granules consist of pelletoids and pisoids. The pelletoids formed upper part of ore body and pisoids formed lower part. Weathering of iron fragments leads to dissolution and precipitation of goethite and formed vitreous goethite surrounding the hematite granules. Some goethite formed matrix that contains large amount of porosities and hematite granules.

(7) Silica mobilization happened multiple times. Some dissolved and precipitated in hematite granules, and some dissolved in groundwater and precipitated in macro voids in upper Barimunya Member.

(8) The goethite-rich matrix formed interlayered texture that indicates an open-space environment.

(9) The climate change also leads to the replacement of vegetation. Wood fragments were driven by river flow and deposited in upper CID. Large pieces of fossil wood fragments and some root structures are one of the classic features of Yandi CID.

(10)Lowering of the water table resulted in developing of extensive alteration zones which reduced the quality of ore.

(11)The Yandi CID is a unique production of ironrich materials, erosion, global climates changes, plates drifting, rainforest vegetation and palaeochannel.

References

[1] Macphail, M. K., and M. S. Stone. 2004. Age and palaeoenvironmental constraints on the genesis of the Yandi channel iron deposits, Marillana Formation, Pilbara, northwestern Australia. Australian Journal of Earth Sciences 51, 497-520. [J]

[2] Kepert, D. A. 2001. The mapped stratigraphy and structure of the MINING AREA C Region: The Block Monolith, unpublished report, BHP Billiton Iron Ore, 320 pp. [C]

[3] Michelle S. S. 2004. Depositional history and mineralisation of tertiary channel iron deposits at Yandi, Eastern Pilbara, Australia. PhD Thesis, University of Western Australia (unpublished), 322 pp. [D]

[4] Morris R. C., Ramanaidou E. R. & Horwitz R. C. 1993. AMIRA Project P75G - Channel iron deposits of the Hamersley Province. CSIRO Exploration and Mining Restricted Report 399R. [J]

[5] Ramanaidou, E. R., R. C. Morris, and R. C. Horwitz. 2003. Channel iron deposits of the Hamersley Province, Western Australia. Australian Journal of Earth Sciences 50, 669-690. [J]

[6] Morris, R. C., and E. R. Ramanaidou. 2007. Genesis of the channel iron deposits (CID) of the Pilbara region, Western Australia. Australian Journal of Earth Sciences 54, 733-756. [J]

[7] Storkey, A., A. Doecke, and A. Whaanga. 2010. Stratigraphy of the Western Channel Iron Deposits of the Marillana Creek Operations, Western Australia. Applied Earth Science 119, 2-11. [J]

[8] Thorne A. M. & Tyler I. M. 1997. 1:250000 geological series explanatory notes Roy Hill, Western Australia 2ndedition. Western Australia Geological Survey, 22 pp. [C]

[9] Edwards K. 2000. Historical review of tree stress management activities undertaken at Marillana Creek. Unpublished BHPBIO report, 11 pp. [C]

[10] Kneeshaw M., Kepert D. A., Tehnas I. J. & Pudovskis M. A. 2002. From Mt. Goldsworthy to Area C - Forty years of iron ore exploration in the Pilbara. In: Proceedings Iron Ore 2002, pp. 41 - 56. Australasian Institute for Mining and Metallurgy, Melbourne. [J]

[11] Heim, Jonathan A., Paulo M. Vasconcelos, David L. Shuster, Kenneth A. Farley, and G. Broadbent. 2006. Dating paleochannel iron ore by (U-Th)/He analysis of supergene goethite, Hamersley province, Australia. Geology 34, 173-176. [J]

[12] Morris, R. C. 1994. AMIRA Project P75G -Detrital iron deposits of the Hamersley Province. CSIRO Exploration and Mining Restricted Report 76R. [J]

（责任编辑：孙 嘉）