2010:057
M A S T E R ' S T H E S I S
Moraine as a Source of Gold in Stream Sediments and the Relationship to a Shear Zone in the Rombak-Skjomen Area
of Norway
David Whitehead
Luleå University of Technology Master Thesis, Continuation Courses Exploration and Environmental Geosciences Department of Chemical Engineering and Geosciences
Division of Applied Geology
Keywords
Orogenic gold; Moraine sampling; Rombak basement window; Svecokarelian orogeny;
Fennoscandian shield.
Abstract
The Rombak-Skjomen area of northern Norway consists of a c.2.3-1.78 Ga Proterozoic basement window overlain by Phanerozoic metamorphosed sedimentary rocks from the Caledonian Orogeny. The window consists of volcanic arc and sedimentary rocks deformed during the Svecokarelian Orogeny when the area was situated on the margin of the Norrbotten craton. The area is geologically similar to other important mining areas of northern
Fennoscandia such as the Skellefte District and the Gold Line of northern Sweden.
Exploration for gold and base metal deposits during the 1980s produced a lot of geological, geophysical and geochemical data. Stream sediment survey data has been used together with new moraine sampling data to see if there is any relation between a recently identified shear zone and unexplained gold anomalies in nearby stream sediments (Larsen et al., 2010).
Different statistical methods have been used to identify anomalous metal concentrations in stream sediment surveys to produce anomaly maps that can be used alongside the new data to identify areas of mineralisation.
The area has the potential to host orogenic gold style mineralisation as has been found in other areas of northern Fennoscandia. Moraine samples have been taken in profiles across the shear zone and analysed to see if the moraine is a possible source of gold in these stream sediment anomalies and if the shear zone is gold bearing. Analysis of the moraine showed that the concentration of gold is very low and it is possible that the moraine is an indirect source of gold in stream sediment anomalies. Gold can enter moraine by weathering of gold bearing mineralisations in the surrounding bedrock, but because of its fine grained nature, it is quickly washed out of the moraine to be re-deposited in stream sediments several kilometres away from its parent rock. The area probably contains more than one type of gold mineralisation that are small, widely dispersed and have different mineralogy making them hard to locate using stream sediment sampling alone.
The sampling results show that the moraine contains very low gold and arsenic concentrations.
However, one sample collected close to a known mineralisation shows anomalous
concentrations in several base metals. This means that moraine sampling can be used to locate base metal mineralisations by identifying metal anomalies down ice of the mineralisation.
Higher concentrations of zinc where also found in samples collected close to the inferred shear zone and moraine sampling may be able to identify large scale structures such as shear zones if they contain higher concentrations of certain metals such as zinc or arsenic.
There is the potential to find new mineralisation in the Rombak basement window especially at depth which has only been drilled at a small number of locations. A gold bearing quartz-
ankerite boulder was found in the moraine may be similar to a new gold mineralisation found in the Mauken basement window further to the north. It is important to understand the
relationships between these two basement windows and how they fit in to the tectonics at the
margins of the Norrbotten craton during the Proterozoic.
Contents
1. Introduction ... 1
2. Geology of the Rombak Area ... 7
2.1 Lithology... 7
2.2 Structure ... 9
2.3 Metamorphism ...11
2.4 Mineralisation ...11
3. Tectonic Setting ...13
4. Quaternary Geology ...15
4.1 Quaternary Geology of the Norddalen Area ...15
4.2 Quaternary Geology of the Gautelis Area ...15
4.3 Description of the Moraine ...15
5. Previous Exploration Work ...19
5.1 Background ...19
5.1.1 Exploration Work Conducted by ARCO ...19
5.1.2 Exploration Work Conducted by Folldall Verk AS ...20
5.1.3 Exploration by Other Companies ...20
5.2 Gold in Stream Sediment Samples ...20
5.2.1 Sampling Methods...20
5.2.2 Analysis Methods ...21
5.2.3 Sources of Gold ...22
6. Methods ...23
6.1 Geochemistry ...23
6.1.1 Choice of Anomalous Values ...26
6.1.2 Moraine Sampling Method ...38
6.1.3 Analytical Method ...40
6.2 GIS ...40
7. Results and Analysis ...41
7.1 Norddalen ...41
7.1.1 Moraine Sampling ...41
7.1.2 Geological Profile. ...42
7.1.3 Geochemical Anomaly Maps ...42
7.2 Gautelis ...50
7.2.1 Moraine Sampling ...50
7.2.2 Geological Profile ...50
7.2.3 Geochemical Anomaly Maps ...51
7.3 Bjørnfjell ...58
8. Discussion ...59
9. Conclusions ...63
Acknowledgments ...65
References ...67
Appendix I: Moraine Sample Data...71
Appendix II: Analystical Results ...77
List of Figures
Figure 1: Location of the Rombak basement window ... 1
Figure 2: Geology of the Rombak basement window ... 4
Figure 3: Exposed basement windows in the northern part of the Fennoscandian Shield ... 7
Figure 4: Profile across the Rombak basement window... 8
Figure 5: Map showing the inferred location of the shear zone cutting through the Rombak basement window and locations of some known mineralisations ... 10
Figure 6: Field sketch showing shearing and folding of the greywacke-granite contact on Haugfjellet ... 10
Figure 7: Tectonic setting of northern Fennoscandia during the Jatulian rift event at c. 2.1Ga ... 13
Figure 8: Tectonic setting of the Rombak basement window during the Svecokarelian Orogeny at c.1.9 Ga ... 14
Figure 9: Photos of moraine from the sample area ... 16
Figure 10: Quaternary geology map of the Norddalen area ... 18
Figure 11: Quaternary geology map of the Gautelis area ... 18
Figure 12: Map showing stream sediment anomalies in the Norwegian part of Rombak basement window ... 21
Figure 13: A: Hypothetical frequency of samples from mineralised and background populations ... 23
Figure 14: Example of primary geochemical dispersion from the Happy Jack-Bulletin (HJB) shear zone at Bulletin, Norseman-Wiluna Belt ... 24
Figure 15: Box plots of trace element concentrations in rock samples collected by NGU and Folldall Verk AS .. 25
Figure 16: Box plots of trace element concentrations in stream sediment samples collected by NGU and ARCO 25 Figure 17: Frequency distribution of As in stream sediment samples collected by ARCO ... 28
Figure 18: Normal distribution of As in stream sediment samples collected by ARCO ... 28
Figure 19: Frequency distribution of Cu in stream sediment samples collected by NGU ... 29
Figure 20: Normal distribution of Cu in stream sediment samples collected by NGU ... 30
Figure 21: Frequency distribution of Cu in stream sediment samples collected by ARCO ... 31
Figure 22: Normal distribution of Cu in stream sediment samples collected by ARCO ... 31
Figure 23: Frequency distribution of Zn in stream sediment samples collected by NGU ... 32
Figure 24: Normal distribution of Zn in stream sediments collected by NGU ... 33
Figure 25: Frequency distribution of Zn in stream sediment samples collected by ARCO ... 34
Figure 26: Normal distribution of Zn in stream sediment samples collected by ARCO ... 34
Figure 27: Frequency distribution of Pb in stream sediment samples collected by NGU ... 35
Figure 28: Normal distribution of Pb in stream sediments collected by NGU... 36
Figure 29: Frequency distribution of Pb in stream sediment samples collected by ARCO ... 37
Figure 30: Normal distribution of Pb in stream sediments collected by ARCO ... 37
Figure 31: Quaternary geology map showing the moraine sampling profiles in the Norddalen area ... 38
Figure 32: Quaternary geology map showing the moraine sampling profiles in the Gautelis area ... 39
Figure 33: Location of boulder 42 in streams cutting moraine on the south side of the Norddalen valley ... 40
Figure 34: Geology profile and moraine samples in the Norddalen area ... 44
Figure 35: Gold concentrations in moraine in the Norddalen area ... 45
Figure 36: Arsenic anomalies in the Norddalen area ... 46
Figure 37: Copper anomalies in the Norddalen area ... 47
Figure 38: Zinc anomalies in the Norddalen area ... 48
Figure 39: Lead anomalies in the Norddalen area... 49
Figure 40: Location of Moraine Samples in the Gautelis area ... 51
Figure 41: Au concentrations in moraine in the Gautelis area ... 53
Figure 42: Arsenic anomalies in the Gautelis area ... 54
Figure 43: Copper anomalies in the Gautelis area ... 55
Figure 44: Zinc anomalies in the Gautelis area... 56
Figure 45: Lead anomalies in the Gautelis area ... 57
Figure 46: Geochemical anomalies in the Bjørnfjell area ... 58
Figure 47: Photograph of Moraine Boulder 41 ... 61
List of Tables
Table 1: Summary statistics of Au in stream sediment samples collected by ARCO ... 26
Table 2: Coordinates and results of heavy fraction samples collected by ARCO in the Norddalen area. ... 27
Table 3: Summary statistics of As in stream sediment samples collected by ARCO ... 27
Table 4: Summary statistics of Cu in stream sediment samples collected by NGU ... 29
Table 5: Summary statistics of Cu in stream sediment samples collected by ARCO... 30
Table 6: Summary statistics of Zn in stream sediment samples collected by NGU. ... 32
Table 7: Summary statistics of Zn in stream sediment samples collected by ARCO ... 33
Table 8: Summary statistics of Pb in stream sediment samples collected by NGU ... 35
Table 9: Summary statistics of Pb in stream sediment samples collected by ARCO ... 36
Table 10: Summary of anomalous values of elements from NGU and ARCO stream sediment surveys. ... 37
Table 11: Analysis results of moraine samples for selected elements in the Norddalen area ... 42
Table 12: Analysis results of moraine samples for selected elements in the Gautelis area. ... 50
Table 13: Selected sample results from rock analyses from the Gautelis area ... 60
1. Introduction
The Rombak-Skjomen area is located in northern Norway between Narvik and the Swedish border (Figure 1). The area is approximately 1,900 km
2and consists of a c.2.3-1.78 Ga Proterozoic basement window overlain by Phanerozoic metamorphosed sedimentary rocks from the Caledonian Orogeny. The window extends into Sweden where it is known as Sjangeli but in this report the window is called the Rombak basement window as it is known in
Norway.
The window can be roughly divided into two areas based on the geology. The eastern part of the window from the Norwegian-Swedish border and further east contains the oldest rocks in the area and is dominated by mafic to ultramafic volcanic rocks. In the central and western part of the window the rocks are dominated by granites with inliers of meta-volcanic and
metasediment rocks. The Proterozoic rocks of the central and western parts of the window are c. 1.9 to 1.8 Ga old and of a similar age to other mineralised belts in northern Fennoscandia, e.g. the Skellefte District and the Gold Line in northern Sweden.
Figure 1: Location of the Rombak basement window. From NGU map service.
Mineral exploration in the area started at the end of the 19
thCentury and was mainly focused around lead-bearing quartz veins that are common in the area. These were soon realised to be of no economic interest but there has been some mining in the neighbouring area of Ballangen in periods from the early 1900s until the 1990s. In the 1920s, a local occurrence of arsenic with some gold was discovered at Gautelis and was test mined at the time. The area was seen as of little interest until the 1980s when the area was explored by various companies searching for gold and base metal deposits. Some gold was discovered in the stream sediments but follow up work in the rocks surrounding the gold anomalies contained only background levels of gold.
Focus then shifted to the zinc-lead mineralisations found on in the northern part of the window such as Haugfjellet. Work in the area was abandoned after diamond drilling found these mineralisations too small to be of economic interest.
Work recommenced in 2007 by the Geological Survey of Norway (NGU) when it was realised that the geology of the area had many similarities to the Fäboliden area in Sweden. The area is being investigated for potential orogenic gold style mineralisation as has been found
throughout northern Sweden and Finland since the mid-1980s (Korneliussen & Nilsson, 2008).
This work has included identifying mineralised outcrops using a handheld XRF and taking samples for analysis. Outcrops with anomalous concentrations of arsenic have then been sampled with a hammer or circle saw and sent for analysis and analysed for gold. Work in the area is continuing with focus around a recently identified shear zone that runs approximately north-south through the area (Larsen et al., 2010).
Orogenic gold deposits are structurally controlled, epigenetic deposits hosted in
metamorphosed terranes. They form at depths of 5-20 km by deposition of gold and arsenic in quartz-carbonate veins by low-salinity, volatile-rich fluids formed during metamorphism. Gold is normally confined to these vein networks, but can also be found in iron-rich sulphidised wall rocks or silicified and aresenopyrite rich replacement zones (Dubé & Gosselin, 2007).
Understanding the precipitation mechanisms can help in identifying the most suitable
lithologies for gold deposition as fluid-rock interaction and wallrock sulphidation are the most important gold precipitation mechanisms (Mikucki, 1998).
The main focus of this study is to see if moraine is a possible source of gold in stream sediment samples taken in the 1980s in the Norddalen area. The source of the gold in the stream
sediments has never been established despite some follow up work at the time. To test this idea, a profile of moraine samples has been taken from the area with gold anomalies back to the supposed source of the gold in the shear zone. The location of the areas can be seen in Figure 2.
A second area with known gold mineralisation at Gautelis, approximately 15 km southwest from Norddalen has also been tested to add support to the theory. Two short moraine profiles have been sampled from the shear zone and outwards. This would help to confirm the theory if it is shown that the moraine is enriched in gold relative to background levels.
The moraine cover in the area is generally sparse but there are some areas where it is possible
to sample long profiles. Information collected from the moraine sampling as well as existing
geochemical data has been analysed using GIS software to see if there is an increase in gold
concentration towards the shear zone and if there is any spatial correlation with gold in the
moraine and other elements such as arsenic associated with a dispersion halo. Geochemical
dispersion haloes around gold mineralisation are possible to detect and studies have shown that
chemical dispersion of elements such as arsenic, antimony and tungsten can define exploration
targets, e.g. Kishida & Kerrich, 1987.
The moraine sampling data has been combined with stream sediment data collected in the 1980s to try to identify possible mineralised areas that could be the source of the gold found in stream sediments. These areas should then be followed up with more detailed investigation and further bedrock sampling.
The data has been plotted using ArcGIS and the statistical analysis has been done using Excel.
The existing data can also be combined with new data, such as structural measurements to identify areas that could be followed up by further bedrock sampling or diamond drilling. This was done in the Varden Ridge area in the northern part of the window in 2004 by a Canadian company Golden Chalice Resources (Coller, 2004).
The tectonic setting of the area would make the area prospective for different kinds of
mineralisation but deep erosion would have removed many of these. This study should help in
evaluation the recently identified shear zone as a potential source for orogenic gold type
mineralisation in the surrounding rocks. Work is currently ongoing with workers from both the
NGU in Trondheim and the University of Tromsø focusing on the structure of the area and
geochronology.
Figure 2: Geology of the Rombak basement window. Profile AB in Figure 4.
Geological data provided by NGU.
A B
Narvik
A B
Geology Legend
ROCKS
Overthrust rocks Intrusive rocks
Pegmatite Ophiolite
Granite (medium grained) Quartz-norite
Norite Peridotite
Quartz-feldspar gneiss (metamorphosed granite) Metamorphosed gabbro
Volcanic rocks Amphibolite
Mixed volcanic and sedimentary rock
Mixed intrusive and metamorphosed sedimentary rocks Sedimentary rocks
Garnet-mica schist with iron formation Garnet-mica schist with calcareous mica schist Calcite marble
Calcite and dolomite marble Quartzite
Phyllite
Calcareous mica schist with sulphide/graphite Magnetite-bearing schist
Sulphide- graphite- bearing schist Garnet-mica gneiss
Quartz-feldspar gneiss Mica schist
Parautochthonous rocks (Thrust over a limited distance) Sedimentary rocks
Biotite schist
Basal conglomerate sandstone
Autochthonous rocks Intrusive rocks
Granite (fine grained)
Granite (coarse grained, light grey) Granite (coarse grained, dark grey) Gabbro
Peridotite
Supracrustal rocks, variably metamorphosed Tuff
Andesite/Dacite Basalt
Greywacke
Quartz-sandstone
Siltstone
Limestone
2. Geology of the Rombak Area
The area comprises a Precambrian basement window that is exposed within younger
allochthonous Phanerozoic rocks of the Caledonian Orogeny. The window is situated on the western margin of the Fennoscandian Shield and is one of a number of tectonic windows within the Caledonides (Figure 3). The window is dominated by younger plutonic intrusive rocks with north-south orientated inliers of older volcanic and sedimentary rocks (Figure 4).
Figure 3: Exposed basement windows in the northern part of the Fennoscandian Shield.
Modified from Bax, 1989
The Proterozoic rocks of the window are overlain by a thin cover of autochthonous sediments of the Late Proterozoic to Cambrian Dividal Group. These are covered by allochthonous nappe complexes of the Caledonian Orogeny (Gustavson 1974 a & b). Deep erosion during the Quaternary has removed the overlying nappe rocks and exposed the underlying Proterozoic rocks of the Fennoscandian Shield.
2.1 Lithology
The Rombak Window comprises a number of N-S trending belts of supracrustal rocks of
varying compositions (Figure 2) separated by extensive areas of younger plutonic rocks
(Korneliussen & Sawyer, 1989 and references therein). The belts differ in composition and the
geology across the area is variable in terms metamorphic grade and extent of deformation. A
detailed description of the lithologies within the area can be found in Dalen & Martinsen
(1983).
Figure 4: Profile across the Rombak basement window. Based on information from Birkeland, 1976.
The oldest rocks in the area are a suite of high-Mg, low-K
2O basalts in the Ruvssot-Sjangeli area on the eastern side of the Muohtaguobla Shear Zone that have been dated at a maximum of 2.3 Ga using Rb-Sr (Romer, 1989a). However this age may be too high because of
contamination from a different source of Sr. These rocks are different in composition to the rocks on the western side of the shear zone which are younger in age. They are perhaps related to the nearby Kiruna greenstone group more than 100 km to the east and may be the equivalent of the Ädnamvare Formation that is also komatiitic in composition (Martinsson, 1997).
Tonalite basement that is at least 1.94 Ga old (Romer, 1994) is exposed in Gautelis area in the southern end of the window and may be associated with the older rock units. This is overlain by a thin autochthonous sequence of basal conglomerate containing large pebble-sized, sub- rounded clasts derived from the basement. This unit which is several metres thick at
Gautelisvattnet is overlain by a dolomitic carbonate and a turbidite sequence.
Poorly preserved and highly deformed volcanic and sedimentary units consisting of alkali- calcic volcanic rocks ranging from basalt to rhyolite with abundant andesite, turbidites,
greywackes and conglomerates have probable ages of 1.91 – 1.88 Ga (Korneliussen & Sawyer, 1989). These volcanic and greywacke rocks are similar in character to the meta-volcanic- greywacke sequences of the Lycksele-Storuman area of Sweden around the Fäboliden gold deposit (Korneliussen & Nilsson, 2008, Bark & Weihed, 2007). The turbidites from
Rombaksbotn and Gautelis formed in an active marginal basin adjacent to a mature volcanic arc possibly of Andean-type (Sawyer & Korneliussen, 1989). This is supported by the nature of the volcanic rocks with a mature continental arc on tonalitic continental crust. The remains of black shales can also be found throughout the area as small, highly deformed graphitic schists.
These can be seen along the main shear zone in some areas as thin outcrops which have been almost entirely destroyed by deformation.
The coarse-grained granites and syenites that intrude older volcanic and sedimentary rocks have been dated at 1.78 Ga by Rb-Sr (Gunner, 1981). Similarities in the geochemistry of these intrusives and the older volcanic rocks suggest a related source and possibly represents a later and deeper stage of mantle activity as the arc thickened (Korneliussen & Sawyer, 1989).
The area is cut by swarms of mafic dykes that are less than 1 metre to several metres wide that run parallel to the shear zones. These dykes occur both in and outside of the shear zones. The precise age of these dykes is unknown, but field relationships would suggest that they post-date the supracrustal rocks and pre-date the granitic intrusions. At Gautelis, both the tonalite and mafic dykes are cut by veins of granite. At some locations such as on Haugfjellet, the mafic dykes are also sheared and were possibly emplaced at the time of the main period of shearing during the Svecokarelian Orogeny.
Following a period of deep erosion during the Precambrian, the area was covered by clastic sequences of conglomerates and arkosic sandstones during the Late Proterozoic to Cambrian (Bergström & Gee, 1985). These rocks are known as the Dividal Group and are overlain by the lowermost Caledonian nappe complex which consists of metasediments and locally derived crystalline rocks (Bax, 1989).
A B
2.2 Structure
The supracrustal rocks are steeply dipping within the enclosing plutonic rocks and may be the result of the collapse of a mature volcanic arc at the same time as the plutons were emplaced at lower crustal levels (Figure 4). The Paleoproterozoic supracrustal rocks in the area were then intensely deformed during the Svecokarelian Orogeny at c.1.8 Ga.
Compressional deformation included the formation of N-S trending upright folds, fold-thrust belts and large, steeply dipping shear zones (Larsen et al., 2010). During the orogeny, the area underwent possible oblique northeast to southwest compression forming roughly northwest to southeast thrust and fold belts. This was later followed by the formation of a large roughly north-south striking shear zone to accommodate the continued compression (Iain Henderson, NGU, pers. comm.).
The large N-S trending shear zone is exposed at many locations within the basement window, e.g. Muohtaguobla and in Norddalen (Figure 5). The shear zone varies in thickness from a few metres to over 300 m in Norddalen with parallel zones of sheared and unsheared greywackes and granite on either side of the main shear zone. Shear sense indicators such as sigma clasts are common in the area and show a complex history with both dextral and sinistral shear and it is likely that the shear zone has been reactivated several times in the past (Bax, 1989).
Folds in the area vary from low angle open folds to tightly closed folds and they have been cut by low angle thrusts in places. Some of the folds are cut by shears and some folds are
themselves sheared.
The rocks within the window are variably deformed and show a general N-S almost vertical foliation. The contacts between the supracrustal belts and the enclosing granites are often sheared and steeply dipping. At Haugfjellet, the greywackes have been folded prior to the intrusion of the granite and the contact between the two is sheared (Figure 6).
Some of the mineralisations in the area are structurally controlled, such as the zinc
mineralisations at Čunojávri and Haugfjellet. Small pockets of mineralisation can be found in duplex structures and fold hinges and may represent possible reworking of small syngenetic base metal mineralisations during the Svecokarelian Orogeny. These small mineralisations are well preserved in low strain zones along the shear zone. The mineralisations are generally not preserved in the higher strain parts of the shear zone.
Caledonian deformation in the area has reactivated or modified some of these older structures
as well as forming entirely new structures (Fossen & Rykkelid, 1992). Extensional ductile
deformation has also been documented along the northern edge of the basement window and
indicates some deformation of the basement as the overlying thrust sheets were emplaced
(Cashman, 1989).
Figure 5: Map showing the inferred location of the shear zone cutting through the Rombak basement window and locations of some known mineralisations.
(Shear zone location from Larsen, 2010)
Figure 6: Redrawn field sketch showing shearing and folding of the greywacke-granite contact on Haugfjellet.
Greywackes
Sheared contact (10 cm)
Granite
Sheared contact (10 cm)
Folded Greywackes 1m
N
2.3 Metamorphism
The area has undergone amphibolite-grade metamorphism during the Proterozoic at P-T conditions of 6kb and 575C (Sawyer, 1986). This metamorphism has been dated at 1825173 Ma by the Sm-Nd method (Romer, 1989a) and was followed by retrogression to greenschist to lower-amphibolite facies during the Caledonian Orogeny at around 400 Ma (Korneliussen &
Sawyer, 1989; Romer, 1994). At Gautelis, U-Pb zircon data indicate a younger metamorphic event between 1796 to 1780 Ma (Romer, 1994), which is rather later than the main
Svecokarelian event. The dates of the metamorphism in the area vary depending on the location and the method used, Sm-Nd or U-Pb, but the dates are generally consistent with the timing of the Svecokarelian Orogeny (Weihed et al,. 2002; Eilu & Weihed, 2005).
2.4 Mineralisation
The area hosts many different types of mineralisation. Some of these are:
galena hosted in quartz veins within granite;
zinc mineralisation hosted in the Proterozoic metasediments;
different types of copper mineralisations in the Sjangeli area (Romer, 1989b);
chalcopyrite hosted in quartz veins cutting the main shear zones;
arsenic-gold mineralisation in carbonates at Gautelis;
disseminated aresenopyrite-gold mineralisations in the greywackes which have been the main reason for recent studies.
The galena in quartz-calcite veins is similar in style to that seen in other areas such as at Flåsjön in Jämtland in Sweden and is likely contemporaneous with the Caledonian Orogeny (Romer, 1994). The zinc and copper mineralisations within the Proterozoic metasediments may represent small, low-grade SEDEX-type deposits that formed on the sea floor during a period of extension before or during the Svecokarelian Orogeny. At Čunojávri and at Haugfjellet, the zinc mineralisation in the metasediments has been remobilised into shear zones and hence must be older than the shearing event.
The other mineralisation styles are either related to the Svecokarelian Orogeny at 1.9 to 1.8 Ga or possible remobilisation during the Caledonian Orogeny (Romer, 1994). It is at this older period when any orogenic gold mineralisation in the area would have formed. It is then important to understand the timing of mineralisation relative to the deformation of the area.
Any orogenic gold mineralisation should have formed at or just after peak metamorphism in the region at 1825173 Ma. These dates fit with the present ages of orogenic gold mineralising events in the Fennoscandian Shield at 1.90 to 1.86 and 1.85 to 1.79 Ga (Eilu & Weihed, 2005).
The arsenic-gold mineralisation at Gautelis may have been added or remobilised during the Caledonian metamorphic event (Skyseth & Reitan, 1990) but the source of the metals may still originate from mineralisations formed during the older orogenic event.
Fluids would have been focused along pre-existing faults and fractures during remobilisation
and metals would be re-deposited or added at shallower crustal levels. This remobilisation
could explain why the area is anomalous in arsenic and lead which can be transported by low-
temperature fluids as sulphur complexes and may represent leaching of the basement at deeper
levels (Romer, 1992).
The mineralisations across the basement window tend to be small and intensely deformed making them difficult to date precisely. It is possible that small syngenetic mineralisations formed in the basin sediments during the Proterozoic and were later deformed and remobilised during the Svecokarelian Orogeny and the younger Caledonian Orogeny. This makes it
unlikely that any sizeable deposits remain in the area but there is still some potential to find
mineralisation formed by the remobilisation and trapping of metal bearing fluids under
favourable conditions.
3. Tectonic Setting
The oldest rocks in the area, dated at a maximum of 2.3 Ga, formed during a period of crustal extension possibly related to the Jatulian rift event at 2.3-2.0 Ga. This event formed thick units of tholeiites, locally associated with komatiites and picrites in northern Sweden possibly because of the existence of a mantle plume in the area at the time. The eastern edge of the Rombak basement window was situated on the edge of a failed rift associated with a triple junction located to the south of Kiruna (Martinsson, 1997) (Figure 7) and the tectonic boundary at Muohtaguobla may represent the deformed boundary of the greenstone domain.
Figure 7: Tectonic setting of northern Fennoscandia during the Jatulian rift event at c. 2.1Ga.
Modified from Martinsson, (1997).
Following the period of rifting, the area was situated at the passive margin on the border of Norrbotten craton until about 1.9 Ga. Erosion of the tonalite basement to form conglomerates and deposition of carbonate and turbiditic rocks occurred during this period. The black shales would have formed at this time in deep water basins away from the input of eroded material being washed off the nearby craton. This was followed by a transition into an active margin basin during the development of a volcanic arc in an Andean-type setting (Sawyer &
Korneliussen, 1989). The volcanic successions and turbiditic greywackes were deposited during this episode prior to the Svecokarelian Orogeny.
The Svecokarelian Orogeny was a period of complex accretionary processes and oblique subduction at the margins of an Archean craton from about 1.9 Ga to 1.8 Ga (Figure 8)
(Weihed et al., 2002). During this period, the area was situated at the margin of the Norrbotten craton to the east and the rock types suggest an Andean-type setting with subduction. This period fits with one of the main periods of global orogenic gold mineralisation during the Precambrian (Goldfarb et al., 2001). Here the Svecokarelian refers to the deformation and metamorphism that took place at 1.9 to 1.8 Ga and the Svecofennian to the supracrustal rocks suites 1.95 to 1.85 Ga ia age.
Rift margin and failed rift 1 Kiruna 6 Ostrobothnia MORB-type pillow lava 2 Kautokeino 7 Kuopio Paleoproterozoic Greenstones 3 Repparfjord-Alta 8 Thornajärvi Dike Swarms c.2.1 Ga 4 Kittilä Rombak Basement Inferred Archean boundary 5 Peräpohja Window
Figure 8: Tectonic setting of the Rombak basement window during the Svecokarelian Orogeny at c.1.9 Ga.
Modified from:
http://en.gtk.fi/ExplorationFinland/Commodities/gold_in_finland.htmlThe oldest rocks in the area predate the Svecokarelian Orogeny and the supracrustal rocks are contemporaneous with the orogenic period. The metamorphism and regional deformation are coeval with the formation of other orogenic gold deposits in the northern part of the
Fennoscandian Shield, e.g. Fäboliden in Sweden and Laivakangas in Finland.
Most studies of the Svenofennian Orogeny have been focused on events in Sweden and Finland, e.g. Nironen, (1997), Weihed et al., (2005) and little is known of the eastern margin which is largely covered by younger Phanerozoic rocks. Much is still unknown or poorly understood and the models are still very much theoretical. Much more work still needs to be carried out, especially on the western margin of the Precambrian terranes of Norway and how they fit into the overall sequence of events.
The area has endured a long period of large scale tectonic activity which plays an important role in different ore forming processes. These tectonic processes have the potential to form layered intrusions, syngenetic volcanic Cu-Zn deposits, SEDEX, iron formations and
epigenetic orogenic gold type deposits. Although not all of these deposit types can be found in the area, the large number of different styles of mineralisation indicates that the area was rich in a variety metals that have been transported in various metal bearing solutions over time.
Some of these metals could have been transported along reactivated structures during orogenic episodes and lead to the formation of structurally controlled epigenetic deposits which may be present in the area.
Rombak Basement Window
4. Quaternary Geology
The area has been affected by several periods of ice sheet cover and glaciation throughout the Quaternary period. During the last Ice Age the area was situated on the margin of the
Fennoscandian Ice Sheet. This mass of ice began to retreat from the area about 9000 years ago when thick outwash sheets formed on the floors of some valleys (Andersen, 1975). The
mountain glaciers have generally been in retreat since, but several episodes of glacial advance have been recorded on the Swedish parts of the Rombak Window around the Kebnekaise mountains (Karlén, 1973). Cirque and valley glaciers are still present in the highest parts of the area although they appear to have been retreating quite rapidly since the first aerial
photographs were taken of the area.
The glaciers have carved out valleys and lakes in the direction of the ice movement. The
general direction of ice movement was east to west in the direction of the sea. The valley floors and surrounding slopes have been partly covered by various types of glacial debris that was deposited as the ice melted and retreated (Figure 10 & Figure 11).
The areas are situated in the mountains above the tree line and are only covered by sparse vegetation of grasses and mosses with lichens on some rock surfaces.
4.1 Quaternary Geology of the Norddalen Area
Norddalen is a glacier carved valley that runs approximately east-west and is surrounded by higher mountains on three sides. Various types of moraine of different thickness cover parts of the valley floor and sides (Figure 9a). The sides of the valley are covered by thin lateral
moraines that vary from a few cm to several metres thick. The centre of the valley is covered by the remnants of a near 3 km long medial moraine that is flanked on both sides by
depressions that contain rivers and fluvial material. Striations on the granitic rocks in the side of the valley show that the direction of ice flow was along the valley and that the moraines are parallel to the direction of ice flow.
4.2 Quaternary Geology of the Gautelis Area
The Gautelis area is on the northwest side of an elongated glacier-carved lake called
Gautelisvattnet. There are many striations on the surrounding rocks indicating a southeast to northwest direction of ice movement.
The area is also a glacier carved valley but does not have the distinctive U-shape. The valley is flanked by peaks to the southwest and northeast and is likely the place where several mountain glaciers converged. There is an almost continuous 3 km long moraine on the western side of the valley with some small and sporadic areas on the eastern side. The area contains some perched boulders and erratics that are absent in Norddalen.
4.3 Description of the Moraine
The term “moraine” covers a wide range of depositional features associated with glaciers and ice sheets and are classified based on their position and how they were formed (Embleton &
King, 1975). Moraine consists of rock fragments that have been eroded from the surrounding mountains and have fallen onto the ice surface and material that has been eroded and
transported at the base of the glacier. These fragments are often poorly sorted in terms of grain
size and fragments of different rock types often become mixed together. It is possible to determine the most likely source of the moraine, but this has not been done in this study. In addition, the distance that the moraine has been transported and the degree of reworking and resorting are unknown. It can be possible to get a rough idea of the minimum transport distance by seeing how far bedrock from a specific locality has been transported. The angular shape of the moraine would suggest that the material has not been transported far within the glacier but it is likely that the moraine has been transported several kilometres on or buried within the upper layers of the glaciers.
a b
c d
Figure 9: Photos of moraine from the sample area.
a: Moraine on the southern side of the Norddalen valley. b: Rock fragments in the moraine in Norddalen valley.
c: View of the moraine north of Gautelisvattnet looking south. d: Moraine at a road cutting near Bjørnfjell.
The moraine is very variable in composition across the area from a scale of a few meters to several kilometres. It is composed of material derived from the local area and is generally composed of blocky boulders to very fine sand and silt particles. The larger boulders and stone blocks are angular to sub-angular and are 20 cm to several meters in size. Some of the small pebbles that are 3-10 mm in size can be rounded but they are generally sub-angular to angular.
The moraine is poorly sorted and has well developed soil horizons that contain a lot of clastic material (Figure 9b).
The topmost layer of the soil (A horizon) is dark and is composed of sandy material and
organic matter. This layer varies considerably in thickness from less than 5 cm to 30 cm in
some places. Where this layer was thick it is often saturated with water and sampling sites were
chosen that had a thin A horizon and where the underlying material was dry. Underneath the dark material is a well developed leached zone that is light grey in colour and is often 10-30 cm thick. Beneath this layer are layers of brown sandy to gravelly material with abundant small to large rock fragments which are sometimes underlain by a layer of pale green clayey material that is often saturated and contained few rock fragments. This clay layer was absent where the samples were taken directly above the underlying bedrock.
The moraine is often water saturated especially where the top soil horizon was very thick.
Some samples of muddy moraine were taken and dried before sieving. Once they were dried it could be seen that they consisted of sandy-clayey material with abundant angular rock
fragments from less than 5 mm to more than 8 cm in size.
The moraine cover in Norddalen is generally thin and varies from a few cm to several metres.
It is composed of poorly sorted clastic material covered by a well developed soil layer. The fine grained material is often sandy and brown in colour with green to grey coloured clay layers. The size of the fragments varies from less than 5 mm to more than 10 cm and large granitic boulders can be found on the slope on the northern side of the valley. These boulders are not covered by soil and have fallen down the slope in more recent times. The fragments are generally angular, but some are more rounded. They are composed of different lithologies including granite fragments similar to the bedrock.
The medial moraine that runs down the centre of the valley is also composed of very variable moraine. Some sample sites contain mostly smaller fragments whereas other sites contained larger fragments. Large granitic boulders are found where the moraine is terminated by an outflow for the hydroelectric plant. One sample was collected on what may have been the top of one of these large boulders or it could have been granitic bedrock.
The moraine in the Gautelis area varies from several centimetres to several metres thick in the depressions. The moraine is covered with angular blocks that range in size from about 10 cm to more than one metre (Figure 9c). Some of these blocks were of metamorphic rocks that are not present in the immediate area. They are dark green with a schist-like texture. Most of the samples collected in the area had granitic bedrock nearby and some samples, e.g. 23 and 30 were collected above the bedrock. The moraine at some sample sites is gravelly and sandy but clayey in others. The amount of fragments vary between the sample sites as did the size and shape. The moraine in the shorter profile appears to contain more rounded fragments but this is more a general observation.
The moraine at Bjørnfjell is very thin and sporadic compared to the other areas. The moraine is water saturated almost everywhere and is composed of large blocks to fine sand grains with angular fragments several cm in length and is very poorly sorted. In places where there are deep road cuttings the moraine can be more than three metres thick and composed of sandy to gravelly material with abundant angular blocks up to several decimetres in size (Figure 9d).
The sample was taken above granitic bedrock where the moraine cover was generally thin. The
three composite samples were taken within depressions in the bedrock where the moraine was
much thicker. As the moraine is water saturated, samples of roughly 10 kg were taken for each
of the composite samples and air dried before being sieved later to make one 10 kg sample.
Figure 10: Quaternary Geology Map of the Norddalen Area. Modified from NGU Map Service (Karttjenst).
Figure 11: Quaternary Geology Map of the Gautelis Area. Modified from NGU Map Service (Karttjenst).
5. Previous Exploration Work
A considerable amount of exploration work has been conducted in the area in the past and a lot of data has been collected and is available in various exploration reports and publications.
5.1 Background
The earliest exploration in the area was in the late 19
thand early 20
thcenturies and focused on the mineralisations in the Sjangeli area along the Swedish-Norwegian border. Investigations included trenching and pitting and evidence of these workings can still be seen such as at the zinc mineralisation north of the Čunojávri tourist hut. The Gautelis As-Au deposit was test mined in the period from 1916 to 1920 (Bugge, et al., 1922) and about 500 metric tons of ore was mined at an average grade of 30 % As and 11 ppm Au (Priesmann, 1984a).
Much of the recent exploration work in the area was carried out in the 1980s. Since the
beginning of the 1980s, over 6000 rock, soil and stream sediment samples have been collected and analysed by various groups working in the area. Most of the exploration and sampling was done by two main companies ARCO and Folldall Verk AS. ARCO Norway Minerals Inc.
carried out exploration from 1983 to 1989 looking for zinc-lead and gold deposits in the central and northern parts of the window and from 1983 to 1986 Folldall Verk AS, initially in a joint venture with the Amoco Norwegian Oil Company, was exploring for gold in the southern part of the window.
NGU carried out smaller investigations of zinc-lead and uranium mineralisations in the area in the 1970s and early 1980s (Lindahl, 1976, 1978, 1980a, 1980b and Lindahl & Furuhaug, 1977). Falconbridge Nikkelverk AS investigated the mineralisations around the
Rosokkatoppen Sjangeli area in the early 1970s looking for copper. Some stream sediment samples were collected and some small geophysical surveys were conducted (Band, 1971).
Most of the data collected is available in exploration reports stored at Bergvesenet (the Norwegian Mining Inspector) in Trondheim. The majority of the available data has not been digitised and many of the sample points lack co-ordinates. Some of the data mentioned in the reports is missing altogether. A database of all the available samples should be made from the available exploration reports in the future to aid in exploration and further research in the area.
5.1.1 Exploration Work Conducted by ARCO
Regional exploration was conducted in 1983 and 1984. About 800 stream sediment samples were collected across the area and a total of 168 follow up samples were analysed by heavy mineral washing. This survey identified several gold anomalies including visible gold grains in the Norddalen area (Figure 5). Follow up heavy fraction sampling (panning) in the streams often confirmed the anomalies which ranged from 0.1 to 23 ppm Au (Korneliussen et al., 1986). Focus then switched to the northern part of the window after several large base metal anomalies were identified.
In the Haugfjellet area in the northern part of the window, the highest zinc and lead
concentrations in the stream samples occurred close to visual mineralisation in the bedrock.
The stream sediments also showed that the areas with meta-sedimentary units contained very
high arsenic concentrations as well as some gold anomalies. The area was explored quite
extensively with geophysics, rock and soil sampling and the most promising anomalies in the
Haugfjellet area were tested with diamond drilling.
5.1.2 Exploration Work Conducted by Folldall Verk AS
Exploration activities were concentrated around the Gautelis As-Au deposit in the southern part of the window. Sampling was also carried out in the Kjørisvann, Vavrat, Čunojávri and
Cainhavarre areas. Over 1300 rock samples were collected along with over 500 soil samples from the Gautelis area. Diamond drilling totalling 1347 m was carried out around the Gautelis deposit and a number of sections were sent for assay. Some geophysical surveys were also conducted in the area around the Gautelis deposit. These included CEM horizontal shootback and IP, protonmagnetometer and VLF measurements (Priesmann, 1984a & b).
Systematic sampling of soil and bedrock was carried out along profiles around the Gautelis deposit and were analysed for Au, Cu, Zn, Pb, As, and Ag. The results of the sampling showed spotty occurrences of high gold concentrations (100-200 ppb in soils and 100-300 ppb in rocks). A small geochemical halo with high Au was identified around an outcrop assaying 0.6 g/t Au. High zinc concentrations in soils (up to 1500 ppm) occur close to faults. Some samples also contained anomalous copper and lead values (Priesmann, 1984b).
5.1.3 Exploration by Other Companies
Further investigations were carried out in 1994 by Geologiske Tjenester in partnerships with Hendricks Mineral Canada Ltd around Kjørisvatnet and Haugfjellet. The investigations included 186 soil samples from Kjørisvatnet and 295 soil samples from Haugfjellet. This program highlighted some geochemical anomalies that coincided with some geophysical anomalies (Flood, 1994). In the summer 2004, a Canadian junior company, Golden Chalice Resources conducted a twelve day field program on Varden Ridge in the northern part of the window. Structural measurements were collected at more than 300 locations and plotted on aerial photographic maps together with some of the previous geochemical samples from the area. The results were plotted using ArcGIS to try to identify any relationship between geochemical anomalies and the structural geology. A drill program was planned but was not carried out (Coller, 2004). At present, the company GEXCO has some small claims across the area.
5.2 Gold in Stream Sediment Samples
Several gold anomalies were identified in the Norddalen area from Losi in the west to the Muohtaguobla area in the east. Follow up heavy fraction sampling confirmed anomalies at Losi and Cunojokka, but did not confirm the anomaly at Muohtaguobla. It is likely that these
different anomalies are derived from different mineralisations within the area. The fact that only some of the initial anomalies were confirmed by the follow up sampling could be explained by the fine grained nature of the gold.
5.2.1 Sampling Methods
The stream sediment samples collected by ARCO were sieved in the field with an aluminium sieve with a nylon cloth to obtain the -80M (0.18 mm) or +80M (0.5 mm) fraction sizes depending on what fraction size was available in the stream. Both the fractions were collected for some samples and the ratio of smaller to larger fractions is about 1:1.6 (Flood, 1983).
Heavy mineral sampling of the stream sediments were conducted in the streams that showed
anomalous Au concentrations (>5 ppb). The samples consist of sieved material (<1 mm)
bulked with panned material to make a sample of between 210 to 1750 g with an average of 880 g (Skaldebø, 1984).
5.2.2 Analysis Methods
The stream sediment samples collected by ARCO were semi-dried and sent to the Tucson laboratory for analysis of Au, Ag, As, Cu, Pb, Zn, Co, Fe and Mn. The analytical method is not mentioned in the reports.
The heavy mineral samples were analysed on the Goldhound 18” concentrating wheels at LKABs laboratory in Grängesberg in Sweden. The weight of the samples varied from 0.54 to 11.72 g with an average of 4.7 g. Two runs were made for each sample with each run
consisting of 3 dl of material that was kept in the machine for 10 minutes. The water flow was adjusted to obtain enough heavy minerals from the sample for analysis and the dip of the wheel was fixed at 45° for the whole program (Skaldebø et al., 1983,). The samples were then
visually examined under a binocular microscope to look for free gold and to estimate the percentage of other heavy minerals such as garnet and pyrite. The samples were then sent to the laboratory in Tucson for assaying of Au and Ag (Skaldebø, 1984).
The sampling results were combined with some of the other available data to produce a multi- element stream sediment anomaly map (Figure 12) that identified a number of geochemical anomalies (Korneliussen, et al., 1986). Some of the sources for these anomalies remain unknown. After the regional sampling programs ARCO decided to concentrate on the large anomalies in the northern part of the window with some limited exploration work carried out in other areas.
Figure 12: Map showing stream sediment anomalies in the Norwegian part of Rombak Basement Window.
Modified from Korneliussen et al., (1986).
5.2.3 Sources of Gold
There are several possible sources of gold in the stream sediments including:
weathered material from the surrounding bedrock;
transported weathered material from upstream;
transported as a suspended fraction in the numerous valley streams;
moraine.
Analysis of the bedrock in the area around the gold anomalies did not show any anomalously high gold concentrations. These small gold anomalies were not followed up further at the time because the source of the gold was probably invisible and efforts were focused on other areas with known mineralisation. These anomalies are false anomalies as the gold is unlikely to be coming from the underlying bedrock and has probably been transported by some unknown mechanism and deposited in the stream sediments. Gold can be transported over long distances in the surface environment and the source of the gold can be very difficult to locate. Gold could also be transported in a suspended fraction in the streams that run down the valley sides. This would mean that small, dispersed areas with locally high gold concentrations could be the source in a number of anomalies.
If the source of the gold is from weathered material from the surrounding rock then that too would have high concentrations of gold. So far the largest source of gold in the bedrock is the As-Au deposit at Gautelis and no other larger sources of gold have been found in the area.
Rock and soil sampling around other gold anomalies in the area have not detected any larger
sources of gold in bedrock. Some of the soil surveys conducted in the area report spotty gold
anomalies e.g. in the Kjorisvattnet area (Flood, 1994).
6. Methods
6.1 Geochemistry
Geochemistry can be used on local or regional scales to identify anomalous patterns in the bedrock or covering overburden that can be used to identify mineralised bodies. This can be of great benefit in mineral exploration programs especially when the bedrock is at least partly covered by glacial overburden or if the area to be investigated is very large. There are various methods that can be used to find a population group within large collections of geochemical data that are associated with mineralisation. The use of computer software for univariate and multivariate analysis means that fewer samples are needed to identify subtle anomalies within a broad area (Govett & Kouda, 1975).
The basic concept of exploration geochemistry is that mineralising processes are different from background processes and this difference can be reflected in frequency distributions of one or more elements and that these elements may be concentrated spatially in a way that becomes apparent at the sampling scale used (Figure 13).
Figure 13: A: Hypothetical frequency of samples from mineralised and background populations.
B: Hypothetical clustering of mineralised samples on a map. From Singer & Kouda 2001.
Anomaly maps can be produced from careful analysis of all the available geochemical data.
Mineralised zones associated with shear zones should be enriched in Au, As, Ag, W, B, Sb and Mo with very low concentrations of base metals (Dubé & Gosselin, 2007; Groves, et al., 1998).
These deposits can produce geochemical dispersion haloes that surround the mineralisation on a scale of less than 100 m. As any mineralised zone is probably vertical to almost vertical in nature, a profile across the shear zone starting from known geochemical anomalies should indicate any mineralisation hosted within the shear zone (Figure 14). These alteration zones are possible to detect with large data sets (Harris; 2000, 2001). However, the techniques are
difficult to apply here with the limited amount of data that has been collected so far. To produce geochemical anomaly maps the anomalous values have to be separated from
background values. This is difficult to achieve with a small data set but the gold values can be said to be anomalous if they are higher than the background levels in the surrounding bedrock.
Some bedrock samples from the area have been collected and analysed and these are included
in the geochemical map. The dispersion patterns need to be interpreted carefully as not all of
the dispersion is likely to be caused by any mineralisation event. Some elements such as
arsenic can be mobile under a range of conditions and the surface of the area will have been
reworked during the latest period of glaciation.
Figure 14: Example of primary geochemical dispersion from the Happy Jack-Bulletin (HJB) Shear Zone at Bulletin, Norseman-Wiluna Belt. From Eilu & Groves, (2001).
Maps of the Norddalen and Gautelis areas have been produced using some of the existing geochemical data alongside the new data. These maps show samples that are anomalous in the following elements:
Au (from stream sediment samples);
As;
Cu;
Zn;
Pb.
These elements have been chosen because they can be used as pathfinder elements for various mineralisation types within the area.
Different techniques have been used to collect and analyse geochemical data from the area over the last 25 years. Before the data can be used in a meaningful way, it must be examined so that differences in collection or analytical techniques do not produce spurious results. The data from the different stream sediment and rock chip surveys can be shown in box plots that enable us to see differences in the distribution of the data (Figure 15 & Figure 16).
The box plots of the rock analyses show that the results for Pb and Zn are similar between the
two surveys. However, the results of the Cu analyses are very different between the two
surveys. This could be due to the sampling techniques or differences in the material that was
sampled. The results of the Cu analyses should be treated separately between the two surveys.
Figure 15: Box plots of trace element concentrations in rock samples collected by NGU and Folldall Verk AS.
(Data from Korneliussen & Sawyer, 1989; Priesmann, 1983).
Figure 16: Box plots of trace element concentrations in stream sediment samples collected by NGU and ARCO.
(Data from Næss, 1983 & Flood, 1983).
0 50 100 150 200 250 300 350
FV Pb Rock 1983 NGU Pb Rock FV Zn Rock 1983 NGU Zn Rock NGU Cu Rock FV Cu Rock 1983
ppm
Rock Data Collected by NGU and Folldall Verk
0 50 100 150 200 250 300 350
NGU Pb ARCO Pb NGU Zn ARCO Zn NGU Cu ARCO Cu
ppm
Stream Data collected by NGU and ARCO
(modifed data)
The box plots of the stream sediment data show that there are differences between the different surveys for each element. This could be explained by differences in sampling techniques, choice of sample material and analytical methods. Differences in the geology of the areas that have been sampled will also produce differences in the survey results. This means that the different surveys should be treated separately for the purposes of statistical analysis.
6.1.1 Choice of Anomalous Values
Some of the previous stream sediment survey data from the area have been digitalised and different statistical methods have been used to define the anomalous values. These include normal distribution, frequency distribution, 90
thand 95
thpercentile values and mean/median plus twice the standard deviation. Using different statistical methods alongside visualising the data produces a robust anomalous value for that data set and is better than just using one method to analyse all the different sets of data.
The stream sediment data are derived from two surveys conducted in the 1980s. A total of 777 stream sediment samples have been collected and analysed by NGU, but some of these are from the surrounding Caledonides to the north and east of the Rombak basement window. Only the samples collected from within the window have been included in the statistical analysis.
Some of the results have been reported by Næss, (1983). In 1983, ARCO collected 801 samples from across the Rombak basement window but most of these were collected from the northern parts (Flood, 1983). A summary of the anomalous values calculated from the different surveys is summarised in Table 10.
Au:
Some of the stream sediment samples taken in 1983 by ARCO were analysed for Au but the analytical method is not mentioned in the report (Table 1). The stream sediment samples collected by NGU were not analysed for Au. In the modified summary statistics, the three values greater than 1ppm have been removed and this affects the mean by a factor of 10. The statistics show that across the Rombak area, gold is only found in very low concentrations in stream sediments (~60% of the samples had 1 ppb). This is probably the detection limit for Au.
Summary Statistics Au Unmodified Modified
Minimum (ppm) 0,001 0,001
Maximum (ppm) 49,715 0,542
Mean 0,080 0,008
Median 0,001 0,001
Standard Deviation 1,789 0,027
Mean + 2x Standard Deviation 3,658 0,062 Median + 2x Standard Deviation 3,579 0,056
90th Percentile 0,015 0,015
95th Percentile 0,029 0,026
Number of Samples 781 778
Table 1: Summary statistics of Au in stream sediment samples collected by ARCO.
(Data from Flood, 1983).
Heavy fraction stream sampling was conducted by ARCO in 1983-1984 as a follow up to earlier surveys (Skaldebø, 1984). This survey detected and confirmed the presence of gold in streams in the Norddalen area. The coordinates and concentration of Au can be seen in Table 2.
These samples are very anomalous compared to the background level across the window and are worth following up.
Sample Year Map Sheet N Coord. E Coord. Weight (g) Au ppm
526 1983 1431 7568200 0615050 2,7381 0,204
532 1983 1431 7563650 0622100 0,3368 0,478
533 1983 1431 7563950 0621350 4,1571 0,154
535 1983 1431 7566950 0623300 0,908 0,5161
3552 1984 1431 7566700 0619350 0,082 22,67
3553 1984 1431 7566650 0619300 0,8 7,509
3559 1984 1431 7566850 0624350 0,61
3575 1984 1431 7567250 0624200 0,62 1,632
Table 2: Coordinates and results of heavy fraction samples collected by ARCO in the Norddalen area.
(Data from Skaldebø, 1984).
As:
Only samples from the survey conducted by ARCO were analysed for As. Summary statistics of the data is presented in Table 3 and shows that at large number of samples have been
omitted from the modified statistics. A large number of samples (514) have a concentration of 2 ppm and may represent concentrations below the detection limit of the analytical method.
This large number of low value results has a significant effect on the mean, median and standard deviation that in turn strongly affects the anomaly value. These values have been removed and the lowest concentration has been set to 10 ppm so that the frequency distribution of the concentration of As becomes lognormal (Figure 17). The results of the samples above the 95
thpercentile in the unmodified data have been omitted when calculating the summary statistics in the modified data. This reduces the mean and the standard deviation so that the mean plus twice the standard deviation value is close to the 95
thpercentile value after these samples have been removed.
Summary Statistics As Unmodified Modified
Minimum (ppm) 0 10
Maximum (ppm) 924 107
Mean 16,98 27,65
Median 2,00 21,50
Standard Deviation 55,69 18,89
Mean + 2x Standard Deviation 128,36 65,43 Median + 2x Standard Deviation 113,38 59,28
90th Percentile 35,6 53,1
95th Percentile 59,8 67
Number of Samples 785 250
Table 3: Summary statistics of As in stream sediment samples collected by ARCO
(Data from Flood, 1983).
Figure 17: Frequency distribution of As in stream sediment samples collected by ARCO.
The modified data has been used to plot a normal distribution curve to identify the anomalous samples (Figure 18). The curve shows that samples above 56ppm can be considered
anomalous. This value is relatively high and is perhaps caused by the sample area being enriched in As.
Figure 18: Normal distribution of As in stream sediment samples collected by ARCO.
0 10 20 30 40 50 60 70 80
10 11-15 16-20 21-25 26-30 31-35 36-40 41-45 46-50 51-55 ≥56
F r e q u e n c y
Concentration As ppm
Frequency Distribution As (modified data)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1 10 100
C u m u lati v e F r e q u e n c y
log Concentration As ppm
Normal Distribution As (modified data)
Anomalous samples marked in red.
