Content uploaded by Tero Niiranen
Author content
All content in this area was uploaded by Tero Niiranen on Nov 02, 2019
Content may be subject to copyright.
Central Lapland Greenstone Belt 3D modeling project
Final report
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
GEOLOGICAL SURVEY OF FINLAND
Report of Investigation 209
2014
GEOLOGIAN TUTKIMUSKESKUS GEOLOGICAL SURVEY OF FINLAND
Tutkimusraportti 209 Report of Investigation 209
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
CENTRAL LAPLAND GREENSTONE BELT
3D MODELING PROJECT FINAL REPORT
Front cover: 3D gravity model of the Kittilä terrane and adjacent lithological units.
ISBN 978-952-217-302-7 (PDF)
ISSN 0781-4240
Layout: Elvi Turtiainen Oy
Espoo 2014
Niiranen, T., Lahti, I., Nykänen, V. & Karinen, T. 2014. Central Lapland Greenstone Belt 3D modeling
project nal report. Geological Survey of Finland, Report of Investigation 209, 78 pages, 55 gures and
3 tables.
is report presents results of the Central Lapland Greenstone Belt 3D modeling project carried out in
2007–2012. e report documents the utilization of the multiscale edge detection, or “worming”, method
for regional gravity data and in prospectivity modeling, presents a new Ni-Cu prospectivity model for the
Central Lapland Greenstone Belt (CLGB) district, and documents ve 3D geomodels from the Central
Lapland area and their implications for the geological interpretation of the area.
e worming method was tested on gravity data from Central Lapland. e results indicate that worms,
i.e. gravity gradient maxima, are highly useful in interpreting the major shear, thrust, and fault zones, as
well as the contacts of lithological units, their depth extent and their dip orientations. Weights of evidence
analysis indicates that in the Central Lapland area, the known orogenic gold and iron oxide-Cu-Au deposits
display a spatial correlation with the gravity worms, and hence the gravity worms can be used as a tool for
locating the most prospective areas for epigenetic mineral deposits.
A simple Ni-Cu prospectivity model of the CLGB was constructed using airborne magnetic and ground
gravity data together with till geochemistry (Ni, Cu, Co) data. A high-pass ltering technique was used for
till geochemistry to lter out regional anomalies and the data were integrated using fuzzy logic. e result-
ing prospectivity map identies over 40 target areas for Ni-Cu deposits, including the Sakatti Ni-Cu and
Kevitsa Ni-Cu-Au-PGE deposits.
A regional 3D model of the Kittilä terrane and the key adjacent structures was constructed using a
multidisciplinary approach and a wide array of geophysical and geological data. e modeling results indi-
cate that the Kittilä terrane forms a keel-shaped unit that is ca. 9 km thick at the thickest part, thinning out
towards the margins. e shape and thickness suggest that the terrane was considerably thickened during
thrusting from the S and NE. A rough estimate of the orogenic gold potential was carried out by applying
a metamorphic source model for the gold deposits. Based on the modeling results, potentially up to 228
Moz of gold was mobilized from the Kittilä Group rocks alone during the metamorphic events related to
the Svecofennian orogeny. is gure is about 30 times greater than the currently reported gold resources in
the known deposits of the area, suggesting that signicant undiscovered gold resources remain in the Kittilä
terrane area.
A 3D model of the Kolari region was constructed using new seismic (HIRE) data from the Hannukain-
en-Rautuvaara area. e modeling results reveal that the Savukoski and Sodankylä group rocks form a gen-
tly SW-plunging open fold structure with an internal small-scale dome and basin structure. e SW–NE-
striking Äkäsjoki shear zone has been formed in the axial plane of this fold. e modeling results indicate
that the current stratigraphical interpretation is incorrect and needs revising. e results also indicate that
the known Fe ± Cu-Au deposits in the area cannot be hosted by the same stratigraphical unit horizon, and
are thus not strata-bound deposits, as they have previously been interpreted.
A 3D model of the Lapland Granulite Belt (LGB) was constructed using seismic FIRE and geophysical
data. According to the model, the LGB consists of at least 4 tectonic blocks. e Vuotso complex imme-
diately SW of the LGB consists of at least 2 tectonic blocks. ese units comprise a listric thrust package
limiting the Kittilä terrane at its NE contact.
Old Outokumpu Oyj drill core and ground geophysical data were used to constrain a deposit-scale
model of the Saattopora Au-Cu deposit. e model shows the main shear zones and lithological units, as
well as a block model of the two Au-Cu lodes. e modeling data together with the geological observations
indicate that the northern ‘A’ lode appears to be controlled by the albitized phyllite unit between a komatiite
unit in the south and a mac tu unit in the north. e southern ‘B’ lode is controlled by an ESE-striking
subvertical shear zone. e modeling data imply that the deposit is open to depth at ca. 160 m below the
surface. e data indicate that the ore-hosting veins cross-cut the regional F3 folding visible in outcrops and
also in the modeled geological units, suggesting that the mineralization took place in the late stages of the
regional deformation and metamorphism.
e Lauttaselkä 3D model is based on geophysics and the recent bedrock mapping and drilling cam-
paign in the area. e modeling shows a west-vergent thrust system, with thrust folding explaining the
repeating pattern of the Kautoselkä and Vesmajärvi formations in the western part of the area. In the eastern
part of the study area, the Salla group rocks on top of the Sodankylä and Savukoski groups is explained by
west-vergent thrusting of the Salla group on the latter two units.
Keywords (GeoRef esaurus, AGI): mineral exploration, gold ores, nickel ores, copper ores, iron ores,
gravity methods, edge detection, seismic methods, three-dimensional models, Central Lapland Greenstone
Belt, Lapland Granulite Belt, Kittilä, Kolari, Sodankylä, Lapland, Finland
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
Geological Survey of Finland, P.O. Box 77, FI-96101 Rovaniemi, Finland
E-mail: tero.niiranen@gtk.
Niiranen, T., Lahti, I., Nykänen, V. & Karinen, T. 2014. Central Lapland Greenstone Belt 3D modeling
project nal report. Geologian tutkimuskeskus, Tutkimusraportti 209, 78 sivua, 55 kuvaa ja 3 taulukkoa.
Tässä raportissa esitetään vuosina 2007−2012 toimineen Keski-Lapin vihreäkivivyöhykkeen 3D-mallin-
nushankkeen tulokset. Raportissa esitetään monimittakaavaisten gradienttimaksimien käyttö potentiaa-
likenttäaineistolle ja niiden käyttöä prospektiivisuusmallinnuksessa, uusi Ni-Cu-prospektiivisuusmalli
Keski-Lapin alueelle sekä dokumentoidaan viisi geologista 3D-mallia ja esitellään niiden pohjalta tehdyt
johtopäätökset.
Monimittakaavaisten gradienttimaksimien käyttöä testattiin Keski-Lapin painovoima-aineistolla. Pai-
novoimagradienttimaksimit ovat erittäin käyttökelpoisia suurten hierto- ja ylityöntörakenteiden sekä lito-
logisten kontaktien tulkinnassa. Gradienttimaksimien perusteella pystytään linjaamaan näiden rakentei-
den paikkaa, kaateen suuntaa ja syvyysulottuvuutta. Weight-of-Evidence-analyysin tulokset osoittavat, että
Keski-Lapin alueen painovoimagradienttimaksimit korreloivat tunnettujen orogeenisten kulta- ja rautaok-
sidi-kulta-kupariesiintymien kanssa. Näin ollen gradienttimaksimeja voidaan käyttää apuna epigeneettis-
ten malmiesiintymien paikantamisessa.
Keski-Lapin alueelta tehtiin uusi Ni-Cu-esiintymien prospektiivisuusennuste käyttäen hyväksi mag-
neettista lentomittausaineistoa, painovoima-aineistoa sekä moreenigeokemiaa (Ni, Cu, Co). Moreenigeo-
kemian aineisto käsiteltiin käyttämällä ylipäästösuodatusmenetelmää suodattamaan alueelliset anomaliat
pois ja aineistot yhdistettiin käyttäen hyväksi sumean logiikan menetelmää. Tuloksena saatu prospektiivi-
suuskartta osoittaa yli 40 Ni-Cu-potentiaalista kohdetta mukaan lukien tunnetut Sakatin Ni-Cu- ja Kevit-
san Ni-Cu-Au-PGE-esiintymät.
Kittilän terraanin geologinen 3D-malli ja sen välittömässä läheisyydessä olevat tärkeimmät rakenteet
mallinnettiin käyttäen hyväksi laajaa geofysiikan ja geologian aineistoa. Mallinnustulokset osoittavat Kitti-
län terraanin olevan kölin muotoinen, maksimissaan noin yhdeksän kilometriä paksu yksikkö, joka ohenee
reunojaan kohti. Yksikön muoto ja paksuus viittaavat siihen, että terraani paksuuntui merkittävästi eteläs-
tä ja koillisesta suuntautuneiden ylityöntöjen vaikutuksesta. Kittilän alueen orogeenisten kultaesiintymien
esiintymispotentiaalia tarkasteltiin siten, että tehtiin mallinnustulosten pohjalta kvantitatiivinen arvio kul-
lan mobilisuudesta käyttäen oletuksena metamorsta metallinlähdemallia orogeenisille kultaesiintymille.
Tulokset viittaavat siihen, että jopa 7 000 tonnia kultaa mobiloitui metamorfoosin vaikutuksesta pelkästään
Kittilän ryhmän kivistä. Luku on noin 30 kertaa suurempi kuin tällä hetkellä on raportoitu alueen tunne-
tuista esiintymistä. Tämä viittaa siihen, että alueen vielä löytymättömät kultavarannot ovat merkittävät.
Kolarin alueen 3D-malli tehtiin pääasiassa käyttäen hyväksi uutta seismistä (HIRE) aineistoa Han-
nukaisen-Rautuvaaran alueelta. Mallinnustulosten perusteella Savukosken ja Sodankylän ryhmien kivet
muodostavat loivasti lounaaseen kaatuvan poimurakenteen, jossa on pienimittakaavainen sisäinen doomi-
allasrakenne. Koillis-luodesuuntainen Äkäsjoki-hiertovyöhyke on muodostunut kyseisen poimurakenteen
akselitasoon. Mallinnustulokset osoittavat, että nykyinen stratigranen tulkinta alueelta on virheellinen ja
vaatii korjauksen. Tulokset myös osoittavat, että Kolarin alueen rauta ± kupari-kultaesiintymät eivät voi liit-
tyä yhteen stratigraseen yksikköön eivätkä siten ole kerrossidonnaista tyyppiä, kuten aiemmin on joissain
lähteissä esitetty.
Lapin granuliittivyöhyke mallinnettiin käyttäen hyväksi geofysiikan aineistoja sekä seismistä FIRE-
proilia. Mallinnuksen perusteella granuliittivyöhyke koostuu vähintään neljästä tektonisesta yksiköstä.
Granuliitin välittömässä luoteiskontaktissa olevasta Vuojärvikompleksista voidaan erottaa vastaavasti kaksi
tektonista yksikköä. Nämä yksiköt muodostavat listrisen ylityöntöpinkan, joka rajaa Kittilän terraanin koil-
lisosan.
Saattoporan Au-Cu-esiintymä mallinnettiin käyttäen hyväksi vanhaa Outokumpu Oyj:n kairaus- ja
maanpintageofysikaalista aineistoa. Esiintymästä mallinnettiin merkittävimmät hiertovyöhykkeet, pää-
kivilajiyksiköt sekä kaksi Au-Cu-malmiota. Mallinnuksen perusteella pohjoista A-malmiota kontrolloi
voimakkaasti albiittiutunut fylliittiyksikkö, joka rajaantuu etelässä komatiitti- ja pohjoisessa maseen tuf-
yksikköön. Eteläistä B-malmiota kontrolloi itäkaakko-länsiluode suuntainen, lähes pysty hiertovyöhyke.
Mallinnuksen perusteella esiintymä on avoin noin 160 metrin syvyydellä maanpinnasta. Malmipitoiset juo-
net leikkaavat alueen F3-poimutusta, joka on havaittavissa niin mallinnetuissa geologisissa yksiköissä kuin
paljastumillakin. Tämä viittaa siihen, että malmi on syntynyt myöhäisessä vaiheessa suhteessa alueelliseen
deformaatioon ja metamorfoosiin.
Lauttaselän 3D-malli tehtiin käyttäen hyväksi alueella tehtyjä uusia kairaus- ja kartoitustöitä sekä ole-
massa olevaa geofysiikkaa. Malli kuvaa länteen suuntautuneen ylityöntösysteemin, johon liittyvä ylityöntö-
poimutus selittää Vesmajärven ja Kautoselän muodostumien vuorottelun alueen länsiosassa. Sallan ryhmän
kivien esiintyminen Sodankylän ja Savukosken ryhmien päällä alueen itäosassa selittyy Sallan ryhmän yli-
työntymisellä näiden päälle.
Asiasanat (Geosanasto, GTK): malminetsintä, kultamalmit, nikkelimalmit, kuparimalmit, rautamalmit,
painovoimamenetelmät, reunantunnistus, seismiset menetelmät, kolmiulotteiset mallit, Keski-Lapin vih-
reäkivivyöhyke, Lapin granuliittivyöhyke, Kittilä, Kolari, Sodankylä, Lappi, Suomi
Tero Niiranen, Ilkka Lahti, Vesa Nykänen ja Tuomo Karinen
Geologian tutkimuskeskus, PL 77, 96101 Rovaniemi
Sähköposti: tero.niiranen@gtk.
4
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
CONTENTS
PREFACE AND SUMMARY ...................................................................................................................... 5
CHAPTER I.
Gravity worms in the exploration of epigenetic gold deposits:
New insight into the prospectivity of the central Lapland Greenstone Belt,
northern Finland ..........................................................................................................................................8
Ilkka Lahti, Vesa Nykänen and Tero Niiranen
CHAPTER II.
Ni prospectivity mapping using Fuzzy Logic and Receiver Operating Characteristics
(ROC) model validation in the Central Lapland greenstone belt, northern Finland ........................18
Nykänen V., Lahti I. and Niiranen T.
CHAPTER III.
3D model of the Kittilä terrane and adjacent structures ........................................................................ 27
Tero Niiranen, Ilkka Lahti and Vesa Nykänen
CHAPTER IV.
3D model of the Kolari region ..................................................................................................................42
Tero Niiranen, Vesa Nykänen and Ilkka Lahti
CHAPTER V.
e Lapland Granulite belt 3D model ...................................................................................................... 53
Tero Niiranen, Vesa Nykänen and Ilkka Lahti
CHAPTER VI.
e Saattopora Au-Cu deposit 3D model ................................................................................................63
Tero Niiranen, Vesa Nykänen and Ilkka Lahti
CHAPTER VII.
e Lauttaselkä 3D model ..........................................................................................................................73
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
ACKNOWLEDGEMENTS .......................................................................................................................76
REFERENCES .............................................................................................................................................76
4
5
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
PREFACE AND SUMMARY
e Paleoproterozoic Central Lapland Greenstone
Belt (CLGB) extends ca. 450 km from Norway
through Finnish Lapland to the western part of
Russian Karelia, covering an area of roughly 35
000 km2. It records a prolonged history of riing
stages, sedimentation, and magmatism between
2.44 Ga and 2.0 Ga, culminating in deformation
and metamorphism during the Svecofennian
orogeny in 1.91–1.79 Ga. A considerable amount
of geological research has been carried out relat-
ing to the stratigraphy, metamorphism, deforma-
tion, and age determinations, providing insights
into the geological evolution of the CLGB. Active
exploration has been ongoing in parallel with the
academic research, and as a result of this, several
signicant mineral deposits as well as numerous
interesting prospects have been discovered. e
most signicant deposits relate to gold and base
metals, but there is also a potential for other metals
in the district. e discoveries clearly indicate that
the CLGB is one of the most prospective Paleo-
proterozoic greenstone belts in Europe. Despite
the several-decades-long exploration history, most
parts of the CLGB remain underexplored, and the
region is very likely to host a number of undiscov-
ered mineral deposits.
Geological research and exploration in the
CLGB have resulted in extensive geological and
geophysical datasets on the area. Most of the in-
terpretations of this data have been carried out in
2D, focusing on the surface geology, with limited
interpretations of the geology at depth. e few
studies focusing on the geology below the rst few
hundreds of meters of the current bedrock surface
were also carried out in 2D as cross sections. A few
deposit-scale 3D models have been constructed
for the district. Almost all of them have focused on
resource modeling and resolving the form, shape,
size, and/or continuation of individual ore bodies.
Geological 3D modeling or geomodeling became
a standard tool for the gas and oil industry dur-
ing the early 1990s, when geomodeling soware,
the development of which started in the 1980s,
became mature. During the past 15 years, the use
of geomodeling in visualizing and solving geologi-
cal problems in mineral deposits and crystalline
bedrock areas has become increasingly popular.
In 2007, the Geological Survey of Finland initiat-
ed a 3D modeling project on the Central Lapland
Greenstone Belt. e aims of the project were to:
update the geological maps in selected areas, test
the suitability of 3D modeling methods for geo-
logical modeling in targets at local and regional
scales, develop methods for processing geophysi-
cal data sets to support the modeling, and develop
GIS-based prospectivity maps. e modeling was
carried out by combining available extensive geo-
logical and geophysical data sets, using a multi-
disciplinary approach in the interpretation. is
report presents the results of the Central Lapland
Greenstone Belt 3D modeling project in the fol-
lowing chapters. e locations of the study areas
are presented in Figure 1.
Chapter I describes the use of the multiscale
edge detection method or “worming” for the re-
gional gravity data set on the CLGB. e work
demonstrated that using worms enables us to bet-
ter outline the major thrust and shear zones and
that the worming data may be utilized in prospec-
tivity analysis for epigenetic mineral deposits. e
worming data presented in Chapter I were also
used in 3D modeling presented in following chap-
ters.
Chapter II presents a new Ni-Cu prospectivity
model of the CLGB area, which was constructed
using combined till geochemistry, airborne mag-
netics and regional gravity data. e results have
identied more than 40 target areas favorable for
Ni-Cu deposits. ese include the Kevitsa Ni-Cu-
Au-PGE deposit and the Sakatti Ni-Cu discovery,
the former of which is currently under mining and
the latter u nder active exploration and develop-
ment. e high-pass ltering technique was used
in ltering out the regional anomalies and enhanc-
ing the local anomalies in the data. Model valida-
tion indicates a clear spatial association with the
current Ni exploration areas dened by explora-
tion licenses.
6
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
3330000 3360000 3390000 3420000 3450000 3480000 3510000
7500000 7530000 7560000 7590000 7620000
Paleoproterozoic hypabyssal suites
Paleoproterozoic intrusive suites
Nattanen granite suite
Haaparanta suite
Paleoproterozoic supracrustal suites
Nuttio suite
Olostunturi suite
Eastern Lapland layered intrusion suite
Archean complexes
Paleoproterozoic complexes
Vuotso complex
Hetta complex
Central Lapland granitoid complex
Lapland granulite complex
Pomokaira and Muonio complexes
Diverse lithodemes
Haaskalehto gabbro- wehrlite suite
Nilipää granite suite
Major Structures
Shear/Fault zone
Thrust zone
Keivitsa layered intrusion suite
Paleoproterozoic Groups
Kumpu Group
Kittilä Group
Sodankylä Group
Kuusamo Group
Salla Group
Vuojärvi Group
Savukoski Group
35°0'0"E25°0'0"E15°0'0"E
65°0'0"N60°0'0"N
Kittilä
Kolari
Nuttio
Granulite arc
Saattopora
Fig. 1. Location of the study areas. Geology modied aer the Bedrock of Finland − DigiKP. Contains data from the National
Land Survey of Finland Topographic Database 03/2013 © NLSand HALTIK.
7
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
Chapter III describes a regional-scale 3D model
of the Kittilä terrane and adjacent structures, in-
cluding the northward-directed Sirkka and Vene-
joki thrust systems, the gold-critical NE–SW-
striking Muusa and Kiistala shear zones, and the
crustal-scale Enontekiö shear zone. e results
demonstrate that the Kiistala terrane is up to 9 km
thick in its thickest part and shows considerable
thickness variation, which is most likely a result
of thickening via stacking and folding during the
compressional tectonic events. Based on the mod-
el, the gold potential of the Kittilä terrane has been
estimated. is suggests that the undiscovered
gold resources within the terrane may total up to
several tens of millions of ounces.
Chapter IV presents a visual 3D model of the
Kolari region based on new seismic data. e
modeling results reveal issues in current 2D geo-
logical maps of the area. e results indicate that
the current division of the Savukoski Group into
the Rautuvaara and Kolari formations is not justi-
ed, but the two units should be merged into one.
is also has direct implications for the explora-
tion model of the Fe ± Cu-Au deposits in the area,
which have been considered by some researchers
to be strata-bound deposits. e results imply that
the structural control rather than certain strata
should be emphasized in exploration.
Chapter V presents a visual 3D model of the
Lapland Granulite Belt (LGB) and adjacent Vuot-
so complex units. e model supports the work of
several authors, suggesting that the Lapland Gran-
ulite Belt represents a listric, SE-vergent thrust
system. Combining the modeling results and pub-
lished metamorphic data on the LGB and adjacent
Vuotso complex rocks, it appears that there was
considerable variation in the transport distances of
the dierent tectonic blocks during the thrusting
event at 1.91–1.89 Ga.
A visual 3D model of the Saattopora Au-Cu
deposit, an Au-Cu resource model, and new eld
observations are presented in Chapter VI. e
modeling results reveal a folding pattern of the
lithological units, the orientation of which cor-
responds to the F3 folding from the outcrops. e
southern ‘B’ ore body is controlled by a sub-ver-
tical ESE-striking shear zone and the northern ‘A’
ore body is for the most part hosted by an intensely
albitized phyllite unit located between komatiite
and a mac tu unit. e mineralization appears
to have taken place during the late phase of the re-
gional D3, or post-dates it. Based on the available
drill core data, the mineralization is open to depth
at ca. 120 below the surface.
Chapter VI presents a visual 3D model of the
Lauttaselkä area located in the eastern margin of
the Kittilä terrane. e model shows a thrust zone
with west-vergent thrust folding, and reveals that
the older Salla Group rocks have been thrusted on
top of the younger Kittilä Group rocks, and the
Nuttio suite ophiolite fragments have been ob-
ducted into the rock package.
8
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
CHAPTER I.
GRAVITY WORMS IN THE EXPLORATION OF EPIGENETIC GOLD
DEPOSITS: NEW INSIGHTS INTO THE PROSPECTIVITY OF THE
CENTRAL LAPLAND GREENSTONE BELT, NORTHERN FINLAND
Ilkka Lahti, Vesa Nykänen and Tero Niiranen
BACKGROUND
e spatial relationship between hydrothermal
mineral deposits and faults and crustal disconti-
nuities has been recognized and discussed by sev-
eral authors (e.g., Groves et al. 1998, Goldfarb et al.
2001, Bierlein et al. 2006). Large deposits appear
to have formed from large hydrothermal systems
that are commonly located around major crustal
structures (Jaques et al. 2002). However, second-
ary structures adjacent to larger primary struc-
tures also have an important role in ore-forming
processes, as they act as pathways for mineralized
uids to the uppermost crust, and orogenic gold
deposits, for example, are typically hosted by these
second or lower order structures in relation to the
larger structures (e.g., Sibson et al. 1988, Groves et
al. 1998, McCuaig & Kerrich 1998, Chernico et
al. 2002).
Various fault types and other geological discon-
tinuities are commonly associated with petrophys-
ical variations, causing potential eld gradients.
ese gradients can be detected by geophysical
measurements and subsequent interpretation of
geophysical maps. e conventional interpreta-
tion of a geophysical map image essentially in-
volves tracing contacts or edges between bodies of
contrasting density or magnetic susceptibility by
separating local extremes. Sun angles, vertical de-
rivatives and upward continuations are commonly
applied to enhance and simplify the image. Specif-
ically, the horizontal derivatives of potential eld
data are oen used to map the edges of bodies that
generate gravity or magnetic anomalies. In both
cases, the maxima of the horizontal derivative will
be above a vertical contact. Multiscale edges, or in
other words, worms, a term introduced by Hornby
et al. (1999), are representations of potential eld
gradient maxima at various upward continuation
levels. e technique has also been discussed and
developed by Archibald et al. (1999) and Holden et
al. (2000). Figure 2 shows the concept and process-
ing stages of multiscale edge detection for gravity
data.
e technique automatically detects the posi-
tions of gradient maxima at various upward con-
tinuation levels. e upward continuation level
models the response of measurements collected at
dierent heights above ground level. Worms ob-
tained from low upward continuation levels are
short wavelength gradient maxima that usually
relate to shallow sources. High upward continua-
tion worms, in turn, are long wavelength gradient
maxima, which typically result from deeper crus-
tal sources.
Worms are commonly associated with geologi-
cal discontinuities such as faults, thrusts, and al-
teration zones that might be prospective for ore
deposits (e.g. Bierlein et al. 2006, Austin and Blen-
kinsop 2008). For example, Archibald et al. (2001)
noted a strong correlation between gravity worms
of the Australian continental gravity dataset and
the locations of large Zn–Pb and Cu deposits. A
spatial relationship between worms and orogenic
gold deposits has also recently been recognized
(e.g. Bierlein et al. 2006).
In general, a worm image that contains results
from several upward continuation levels can pro-
vide information on the apparent dip direction of
geological contacts and boundaries. It is notewor-
9
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
Fig, 2. e concept and processing stages of gravity worms.
thy that worms can identify geological features
provided there is signicant petrophysical varia-
tion to generate a contrast in the magnetic or grav-
ity data set. Although worms are usually blind to
geological structures that lack petrophysical con-
trast, some such boundaries (e.g. cross faults) may
be inferred through linear truncations and osets
of gravity worms.
Recent advances in the processing and integra-
tion of large and diverse data sets have enabled ge-
oscientists to increasingly apply computer-based
GIS and conceptual strategies that can make the
exploration process more eective. In this chap-
ter, we describe the use of the worming technique
on the gravity dataset from the Central Lapland
Greenstone Belt. By including more than 19 000
ground gravity observations with an average site
distance of ca. 0.5–2 km, the gravity Bouguer
dataset can be considered relatively extensive and
dense. e aim of the present study was to evalu-
ate the spatial correlation of gravity worms and
known gold deposits using the weights of evidence
calculation procedure. We also aimed to reveal
new prospective structures and thereby improve
the condence in area selection during gold explo-
ration in the area.
10
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
REGIONAL GEOLOGICAL SETTING
Figure 3 presents a geological map of the study
area. e Central Lapland Greenstone Belt (CLGB)
is one of the largest Proterozoic greenstone ter-
rains in the world. e CLGB consists of Paleopro-
terozoic volcanic and sedimentary cover (2.5–1.97
Ga) on the Archean granite gneiss basement (3.1–
2.6 Ga) (Lehtonen et al. 1998, Hanski & Huhma
2005). e Paleoproterozoic bedrock of the CLGB
consist of a 2.44–1.98 Ga supracrustal sequence of
mac to ultramac metavolcanic rocks, quartzites,
phyllites, and graphitic schists that are intruded
by 2.2–2.05 Ga mac dykes and sills, 1.91–1.86
Ga mac to felsic intrusions, and 1.80–1.77 Ga
felsic intrusions (e.g. Hanski et al. 2001, Hanski &
Fig. 3. Geological map of the study area. e red dashed line indicates the area of gravity measurements. Contains data from
the National Land Survey of Finland Topographic Database 03/2013 © NLSand HALTIK.
11
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
Huhma 2005). Lehtonen et al. (1992) and Hanski
(1997) have linked the volcanites of the CLGB to
an allochthon that is overthrusted. e core of the
CLGB consists of a thick sequence of mac volcan-
ic rocks of the ca. 2.0 Ga Kittilä group, which has
been interpreted to represent an allochthonous to
para-autochthonous unit bound by tectonic con-
tacts with the surrounding units (Lehtonen et al.
1992, Hanski 1997, Hanski & Huhma 2005) Inter-
pretations from reection seismic studies (Patison
et al. 2006, Niiranen et al. 2009) and potential eld
modeling (Elo et al. 1989, Lehtonen et al. 1998)
suggest that the volcanic rock-dominated green-
stone belt has a maximum thickness of roughly
6–10 km.
Two operating gold mines (Pahtavaara and Su-
urikuusikko) together with more than 30 drilling-
indicated gold occurrences are located within the
CLGB (Eilu 1999, 2007). e majority of the gold
occurrences fall into the orogenic gold category.
ese greenstone-hosted gold mineralizations are
similar in many ways to those in more established
mineral districts such as the Yilgarn region of
Western Australia and Superior Province of Can-
ada (Patison 2007). Currently, the largest known
orogenic gold deposit in the area is the Suuri-
kuusikko deposit, with current resources exceed-
ing 6 million ounces of Au. Iron oxide-copper-gold
(IOCG) deposits with extensive albite alteration
haloes around them are also known within the re-
gion (e.g. Eilu et al. 2007). e gold-bearing oc-
currences in the western part of the study area are
mainly of the IOCG-type. e largest known gold
resources in an IOCG-type deposit are in the Han-
nukainen deposit, with ca. 200 000 ounces of gold.
ere are also a few Paleoplacer gold deposits in
the study area, but these are not considered in this
work.
DESCRIPTION OF THE DATA AND PROCESSING
Gravity data
Finland is exceptionally well covered by gravity
measurements. e Finnish Geodetic Institute has
established a national gravity net with an aver-
age station separation of 5 km (e.g. Kääriäinen &
Mäkinen 1997). In addition, the Geological Survey
of Finland (GTK) has performed both target and
regional scale gravity surveys for decades. In 1972,
GTK initiated regional gravity surveys in order
to obtain more dense regional gravity data than
that of the Finnish Geodetic Institute. e meas-
urements have primarily focused on regions with
high mineral potential and areas having key roles
in crustal studies (Elo 1998). e average station
interval is about 1–6 sites per km2. In 2008, the re-
gional gravity register of GTK contained 264500
gravity observations covering an area of 75 150
km2, which is more than 20% of the Finnish ter-
ritory. Data acquisition has primarily been carried
out using Worden, Scintrex CG-3, and CG-5 grav-
ity meters.
e gravity measurements of the CLGB area
started in 1972, while the latest measurements
used in this analysis were carried out in the sum-
mer of 2009. e total number of gravity observa-
tions used in this study was 19273. e density of
gravity observations is mainly 1–4 sites per square
kilometer. e most dense observations have been
made in the northern and northwestern parts of
the study area, representing the latest measure-
ments using novel data acquisition procedures.
Figure 4 displays the Bouguer gravity grid that
was obtained using the minimum curvature grid-
ding algorithm and the cell size of 250 x 250 m. e
gold mines are indicated by yellow stars. Bouguer
gravity values range from -58 to 16 mGal, with a
mean value of -24.7 mGal in the study area. e
extensive Bouguer maximum in the central part of
the study area mainly results from mac volcanic
rocks of the CLGB. e Bouguer minimum in the
northwest is caused by granitic rocks of the Hetta
complex. e Bouguer minimum to the north of
Pahtavaara mine is due to tonalitic and granodior-
itic gneisses of the Archean basement (Pomokaira
Complex).
12
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
Fig. 4. Bouguer gravity map of the study area and major thrust and shear zones. Abbreviations: PC = Pomokaira Complex, SKB
= Sattasvaara Komatiite Belt, CLGB = Central Lapland Greenstone Belt, HC = Hetta Complex, HS = Haaparanta Suite, SG =
Sodankylä Group, LGB = Lapland Granulite Belt.
Worming
Worms were processed using the WormE wiz-
ard of the Intrepid Soware v4.2. Processing was
performed using 12 upward continuation levels
with a multiplying factor of 1.4 times the cell size
(250 m), resulting in upward continuation levels
from 350 m to 14172 m. Figure 5 shows calcu-
lated gravity worms on a grey-shaded Bouguer
anomaly map. Red and blue worms represent the
lowest and the highest upward continuation levels,
respectively. e gure also indicates the location
of the known gold deposits, prospects and active
or closed mines of the study area. Mineral deposit
data were acquired from the FINGOLD database
(Eilu 1999, 2007) maintained by the Geological
Survey of Finland. All of the known orogenic gold
and IOCG deposits in the CLGB show intimate
spatial correlations with shear zones of varying
scale. Figure 5 illustrates that gravity worms have a
clear spatial correlation with the known orogenic
gold and IOCG deposits
13
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
Fig. 5. Gravity worms, gold deposits and prospects of the study area. e background image is a grey-shaded Bouger gravity
data.
MAIN GEOLOGICAL IMPLICATIONS
Figure 6 presents the gravity worms on a geologi-
cal map of the study area. e worms highlight
major geological boundaries that are also evident
in the source Bouguer data and well known from
previous bedrock mapping and exploration stud-
ies. One of the main advantages of worming is the
capability to detect weak gravity gradients that are
not clearly seen in the source data. e small-scale
structures adjacent to larger ones have an impor-
tant role in ore-forming processes, as they can act
not only as pathways for mineralized uids to shal-
low depths, but also as suitable traps for gold dep-
osition in the study area. ese worms are mainly
located within the main lithological units. Several
approximately N–S-trending worms are seen in
the CLGB. For example, the large Suurikuusikko
orogenic gold deposit and associated shear zone
correlates well with the N–S-elongated worm. e
worm breaks into two parts in the mining camp
area (Fig. 5). To the east of the Suurikuusikko ore
deposit, a number of N–S-trending worms indi-
cate other possible prospective structures.
e major gold hosting structure in the study
area is the Sirkka shear zone (Fig. 6), which is lo-
cated at the southern edge of the central Lapland
Greenstone Belt (Figs. 5 and 6). is shear zone
contains two closed gold mines (Saattopora and
Sirkka) and a number of gold deposits and pros-
pects. Besides the shear zone itself, the gravity
worm partly results from the density dierence
between the mac volcanic rocks of the CLGB and
the volcano-sedimentary rocks and quartzites to
the south of the CLGB. e worm could be also
interpreted as a lithological contact zone or tec-
tonic boundary.
All the main structures associated with the
IOCG deposits (Hannukainen and Rautuvaara)
and prospects are clearly indicated by gravity
worms (Fig. 5). ese deposits are epigenetic in
origin owing to a similar genetic link to the struc-
tures as the orogenic gold deposits. In Hannukain-
en and Rautuvaara, the observed worms are prob-
14
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
Fig. 6. Gravity worms at ve continuation levels on a geological map.
ably caused by lithology, as deposits are located in
the contact zone of large units of granitoids and
volcanic-sedimentary rocks. However, the Rautu-
vaara worm extends northeastwards through vol-
canic-sedimentary rocks, suggesting a primarily
structural rather than lithological source. Drillings
performed in the target area have shown that the
contact zone between the two units is tectonized in
the NE–SW direction. It is therefore likely that the
prolongation of the Rautuvaara worm is due to the
structure that extends northeastwards. is could
be a link to a minor IOCG prospect (Lauttaselkä)
that is located approximately 20 km from Rau-
tuvaara at the northeastern prolongation of the
worm (Fig. 5). e observed structure might be
prospective and hopefully increases condence in
area selection of IOCG exploration in the target
area.
e gravity worms also indirectly reect thrust
zones, shear zones, lithological contacts, and al-
teration structures. For example, the presence of
the major Hanhimaa shear zone (HaZ) is indicat-
ed by the truncation of worms at its location. e
NE–SW-striking shear zone is marked as a black
line in Figure 6. e NW–SE cross-cutting faults
of the Pomokaira formation (PoF in Fig. 6) are an
excellent example of the truncation of worms due
to fault structures. To the east, the southern con-
tact of the Nattanen Granite intrusion (NaG) can
be traced to greater depths. Besides these, it is nec-
essary to emphasize other sources that generate
worms: the unexposed 3D geometry of geological
units, density dierences between adjacent lith-
ologies (schists, granitoids, mac volcanites, and
banded iron rocks in the CLGB), and their relative
displacements cause gravity worms. As the pro-
cessing of worms is a straightforward technique,
possible errors in worming results are mainly
caused by errors in the source data and gridding
of this data.
15
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
SPATIAL CORRELATION OF GRAVITY WORMS AND GOLD OCCURRENCES
For this study, the multivariate empirical approach
called weights of evidence (WofE) was applied to
gravity worm data. e WofE method is a Bayesian
approach for combining data to predict the occur-
rence of events. Although the technique was ini-
tially developed as a diagnostic tool in medicine, it
has been extensively used in mineral prospectivity
studies (e.g. Bonham-Carter et al. 1988, Bonham-
Carter 1994, Agterberg et al. 1990, Raines 1999,
Nykänen & Salmirinne 2007). e method is based
on the presence or absence of a characteristic fea-
ture or pattern and the occurrence of an event.
Two types of probabilities, i.e. W+ and W-, can
be computed for each class in the themes of the
model. For example, a class could be dened either
by a high magnetic intensity or a selected proxim-
ity class from the structural lineament. For each
class, a W+ probability value is computed from the
presence of a feature (or training point) in the class
area. In our study, gravity worms were described
as a proximity map classied into 16 classes using
the quantile method in GIS. e W+ probability
for each class was calculated using the equation:
(1)
where
N(Bi ∩ D) = number of gold occurrences within a
map class i
N(D) = total number of gold occurrences within
the study area
A(Bi) = area of map class i
A(T) = total study area
Similarly, a W- probability value is calculated from
its absence from the class area. e commonly used
value is the contrast C, which is calculated from
the dierence between W+ and W-. It can be used
as measure of the correlation power between the
tested theme and the modeled occurrence of the
feature. One of the main advantages of the WofE
technique is the possibility to identify those data
and data combinations that contribute most to the
results.
Figure 7 presents a proximity map of the select-
ed worms. e 119 yellow points representing the
Fig. 7. Training points and proximities to the worms.
N(Bi ∩ D) ⎜N(D)
Wi+ = 1n A(Bi) ⎜A(T)
16
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
gold occurrences were used for the calculation of
weights and the 983 grey random points were used
for comparison. e results of the weight calcula-
tions are provided in Figure 8.
e prospective area based on the calculation of
weights is shown in Figure 9. e calculation re-
veals that, while the calculated worms cover only
30% of the total area, 70% of the known gold oc-
currences lie within this area, i.e. within 675 me-
ters from the gravity worms. is demonstrates
that gravity worms are spatially correlated with
known gold deposits and prospects in the CLGB.
Fig. 8. Results of the calculation of weights.
Fig. 9. Prospective area for gold based on the worms.
17
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
CONCLUSIONS
e relatively large gravity Bouguer dataset ena-
bled us to quantitatively test the spatial correlation
of gravity worms and structurally controlled gold
deposits within the CLGB, northern Finland. All
of the known orogenic and IOCG deposits show
intimate spatial correlation with shear zones of
varying scale. Gravity worms spatially correlate
with known gold deposits and prospects within
the CLGB. For example, the Suurikuusikko oro-
genic gold deposit, which is the largest known
gold deposit in Europe, is associated with an N–S-
trending worm in the study area. e worm breaks
into two separate parts in the mining camp area.
Numerous orogenic gold prospects and two closed
gold mines at the Sirkka shear zone are indicat-
ed by the gravity worms. All IOCG deposits and
prospects have a positive correlation with gravity
worms in the study area.
WofE analysis revealed that while the calculated
worms cover only 30% of the total area, 70% of the
known gold occurrences lie within this area, i.e.
within 675 meters from the gravity worms.
18
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
CHAPTER II.
NI PROSPECTIVITY MAPPING USING FUZZY LOGIC AND RECEIVER
OPERATING CHARACTERISTICS ROC MODEL VALIDATION IN THE
CENTRAL LAPLAND GREENSTONE BELT, NORTHERN FINLAND
Vesa Nykänen, Ilkka Lahti and Tero Niiranen
INTRODUCTION
The Central Lapland Greenstone Belt (CLGB)
is located in the Northern Fennoscandian Shield,
approximately 100 km north of the Arctic Circle
(Fig. 10). The CLGB consists of Paleoproterozoic
volcanic and sedimentary cover (2.5–1.97 Ga) on
the Archean granite gneiss basement (3.1–2.6 Ga)
(Lehtonen et al. 1998, Hanski & Huhma 2005).
Rifting events of the Archean continent from 2.5
Ga to 1.97 Ga resulted in mostly tholeiitic mac
intrusions, dykes, and lavas. There is one operat-
ing nickel mine within the CLGB, and active nick-
el exploration is ongoing in the surrounding areas.
Traditionally, mineral potential assessments are
based on expert opinions on potential areas for a
particular deposit type. However, modern digital
geological maps allow quantitative analysis of data,
Fig. 10. Location of the study area in the Northern Fennoscandian Shield.
19
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
and thus numerical modeling for prospectivity
mapping or exploration targeting. e advantage
of these techniques is easy documentation of the
model parameters and the possibility to rene the
numeric models in an iterative manner. erefore,
it is straightforward to test several model scenarios
(i.e. exploration models) with subtle or more dras-
tic changes. ere are two main approaches to per-
forming a prospectivity analysis (Bonham-Carter
1994):
(1) e conceptual approach uses expert opinions
and knowledge to re-formulate a theoretical or
practical exploration model into a set of cri-
teria that can be described by a mathematical
formula. is approach is suitable for ‘green-
eld’ mineral exploration terrains with only a
limited number of known deposits available
for statistical assessment. ese techniques
include fuzzy logic or expert weights of evi-
dence. Unsupervised classication can also be
considered to belong to this category.
(2) e empirical approach uses the known min-
eral occurrences within the study area as ‘train-
ing points’ for examining spatial relationships
or correlations between the known occurrenc-
es and spatial data. ese techniques are suit-
able for mature ‘browneld’ exploration ter-
rains with abundant data available. Supervised
classication, e.g. neural networks, weights of
evidence, or logistic regression, belong to this
categor y.
e conceptual approach uses the expertise of the
exploration geologists, geochemists, and geophysi-
cists to dene the threshold values for the eviden-
tial datasets. In classical set theory, the membership
of a set is dened as true or false (1 or 0), whereas
membership of a fuzzy set is expressed on a con-
tinuous scale from 1 to 0 (e.g. somewhere between
‘anomalous’ vs. ‘not anomalous’). e values of
fuzzy membership can be chosen based on the sub-
jective judgment of an expert. Membership reects
the degree of truth of some proposition or hypoth-
esis, which is oen a linguistic statement such as
high magnetic values are anomalous for gold de-
posits. To dene the membership function, one
needs to dene the thresholds for ‘not anomalous’
and ‘anomalous’ values, and then a function de-
scribing the ‘maybe – probably’ values in between
these two thresholds. e fuzzy membership values
reect the relative importance of each class of the
maps used. e closer the fuzzy membership value
is to one, the more signicant is the map pattern.
Aer dening the fuzzy membership functions for
each evidential map, a variety of operators can be
used to combine the membership values.
Fig. 11. Inference network describing the exploration model.
20
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
For the study reported in the present chapter, we
dened a simple exploration model that integrates
aeromagnetic data with data from regional gravity
and till surveys. Data integration was performed
using the fuzzy logic technique as described, for
example, by Bonham-Carter (1994) and Nykänen
et al. (2008b). is conceptual knowledge-driven
method was used because there are not enough
known nickel deposits within the study area to be
able to carry out empirical data-driven prospectiv-
ity analysis. e selection of these data was based
on the assumption that Ni deposits are related to
local magnetic and gravimetric anomalies caused
by Ni-critical lithologies. Since Ni is associated
with suldes, it is also assumed that these deposits
would show elevated concentrations of Ni, Cu and
Co in regional till assays compared to background
values. erefore, an inference network described
in Figure 11 was draed.
All the data preparation and spatial data anal-
ysis for this study was performed in ArcGIS 9.3
enhanced with the Spatial Data Modeller (SDM)
toolbox (Sawatzky et al. 2009). The procedure for
using these tools in mineral prospectivity mapping
is extensively described by Nykänen (2008) and
references therein.
DATA USED FOR MODELING
Magmatic sulde deposits can be divided into
two main groups (Naldrett 2011): 1) sulde-rich
Ni and Cu deposits, and 2) sulde-poor PGE
deposits. These deposits form from the segrega-
tion and concentration of sulde liquid droplets
from ultramac or mac magmas, and the parti-
tioning of chalcophile elements into these drop-
lets from the silicate magma forming the ultra-
mac or mac rocks. These rock types typically
have a higher density and magnetic susceptibility
than surrounding felsic rock types, and they can
therefore be seen as magnetic anomalies or grav-
ity anomalies. Due to glacial dispersion, the out-
crops of these deposits can cause elevated values
of chalcophile elements in till deposits. Therefore,
the exploration model we use for predicting mag-
matic nickel deposits denes that the host rocks
of Ni deposits are characterized by local magnetic
and gravimetric anomalies. A high-pass lter was
applied to these data to exclude the long wave-
length regional anomalies and enhance the locally
derived shallow anomalies caused by possible
Fig. 12. Filtering of the gravity data using a high-pass lter.
21
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
mac-ultramac intrusions. The principle of this
ltering technique is dened in Figure 12. In GIS,
this ltering can be performed by using a mov-
ing window of a xed radius across the study area
to calculate the median value, which is subtracted
from the original data values (e.g. Nykänen et al.
2008a). The resulting residual grid is then used as
the input in the prospectivity model. This ltering
was applied to the total intensity of the magnetic
eld (Fig. 13), regional gravity (Fig. 14), and to
km
km
Mining concession
Claim (Ni-Cu)
30°E
30°E
25°E20°E
20°E
69°N
69°N
66°N
66°N
63°N
63°N60°N
High pass AM Magnetic field total intensity
B
A
High
Low
High
Low
Fig. 13. A) Magnetic eld total intensity. B) Residual of the magnetic eld total intensity using a high-pass lter with the
median value over a 2000-m-radius neighborhood. Contains data from the National Land Survey of Finland Topographic
Database 03/2013 © NLSand HALTIK.
22
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
km
km
30°E
30°E
25°E20°E
20°E
69°N
69°N
66°N
66°N
63°N
63°N60°N
Mining concession
Claim (Ni-Cu)
B
A
High
Low
Fig. 14. A) Regional gravity Bouguer anomaly. B) Residual of the Bouguer anomaly using a high-pass lter with the median
value over a 2000-m-radius neighborhood. Contains data from the National Land Survey of Finland Topographic Database
03/2013 © NLSand HALTIK.
23
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
km
km
30°E
30°E
25°E20°E
20°E
69°N
69°N
66°N
66°N
63°N
63°N60°N
B
A
Ni concentrations in regional till geochemistry
(Fig. 15). For the geophysical data, we used a
moving window with a 2000-m radius, whereas
for the regional till data we used a neighborhood
with a 6000-m radius. This was due to the spa-
tial resolution, which was much coarser for the till
data, with 1 sample per 4 square kilometers, when
compared with the geophysical data. The airborne
magnetic data were recorded using a line spacing
of 200 m, and the sampling density for regional
Fig. 15. A) Ni in regional till interpolated using inverse distance method in GIS. B) Residual of Ni anomaly map using a high-
pass lter with the median value over a 6000-m-radius neighborhood. Contains data from the National Land Survey of Finland
Topographic Database 03/2013 © NLSand HALTIK.
24
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
gravity was 1 measurement per square kilometer.
The ltering enabled us to recognize the local
anomalies, which would otherwise be masked by
the long wavelength regional anomalies derived
from deeper sources. The selection of the radius
of the regional eld was, however, somewhat
problematic and denitely a source of uncertain-
ty. Nickel, copper and cobalt in regional till were
combined using a fuzzy AND operator (Fig. 16).
This map was used as one of the inputs in the nal
prospectivity map described below.
km
30°E
30°E
25°E20°E
20°E
69°N
69°N
66°N
66°N
63°N
63°N60°N
Mining concession
Claim (Ni-Cu)
And till (Ni, Cu, Co)
High
Low
Fig. 16. Combined till geochemistry (Ni, Cu and Co). Contains data from the National Land Survey of Finland Topographic
Database 03/2013 © NLSand HALTIK.
FUZZY LOGIC MODEL
Prior to data integration, the grid cell values of
each map were rescaled into a common scale from
0 to 1. This was achieved by applying the fuzzy
membership function Fuzzy Large (after Bonham-
Carter 1994). The closer the values are to 1, the
more favorable is the corresponding location for a
magmatic nickel deposit. After rescaling, the input
data sets were combined using the Fuzzy Gamma
operator (Bonham-Carter 1994). Figure 11 de-
nes the inference network used to combine the
input maps into a single prospectivity map (Fig.
17). This map denes the most favorable areas for
nickel deposits as a red color in the map. In these
areas, all the exploration criteria are met. We also
plotted the exploration licenses, reservations, and
mining concessions related to nickel on the map.
It is clear from a visual inspection that the red ar-
eas are mostly concentrated in the areas with cur-
rent Ni exploration activity. The known deposits,
Kevitsa and Sakatti, are also in areas with very
high favorability.
The residual maps (Figs. 13–15) and the -
nal prospectivity map (Fig. 17) also include the
lithological boundaries from the 1:200 000 scale
25
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
bedrock map. The locations of the exploration li-
cense areas were used to validate the model results,
because there are only two known Ni occurrences
within the study area. Kevitsa is an operating Ni
mine of First Quantum Minerals plc. It is clearly
seen as an anomaly on all of the input maps and
on the prospectivity map. The Sakatti occurrence
is a prospect of Anglo American Exploration. In
addition to the visual validation, we used so-called
receiver operating characteristics (ROC) valida-
tion, which can be used to measure the accuracy of
diagnostic systems (Obuchowski 2003) or the per-
formance of a spatial prediction model (Nykänen
2008). An ROC curve is a plot of sensitivity (true
positive rate) on the y-axis vs. 1 - specicity (false
positive rate) on the x-axis. The area under ROC
curve (AUC) can be used as a measure of the ac-
curacy of a diagnostic test, and can also be used as
a measure of the classication of a spatial predic-
tive model. AUC values vary from 0 to 1. A test
that results in an AUC value of 1 is perfectly accu-
rate, having a sensitivity value of 1 and a 1-speci-
km
30°E
30°E
25°E20°E
20°E
69°N
69°N
66°N
66°N
63°N
63°N60°N
Mining concession
Claim (Ni-Cu)
Ni prospectivity
Very low
Low
Moderate
High
Very high
Fig. 17. Prospectivity map dening areas favorable for magmatic nickel deposits. Contains data from the National Land Survey
of Finland Topographic Database 03/2013 © NLSand HALTIK.
Fig. 18. ROC validation diagram.
city value of 0 (Fig. 18). A totally random model
would result in an AUC value of 0.5, and the curve
would follow the chance diagonal.
26
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
For ROC validation we need two sets of test
locations within the study area: true negative and
true positive sites. The selection of true negative
sites is problematic in spatial prospectivity mod-
els of this kind unless we have tested sites where
we can be sure that there is no mineral deposit.
We decided to use random points for this purpose.
True positive sites were represented by the loca-
tion of the exploration licenses for nickel explora-
tion. This validation presented here includes a sig-
nicant level of uncertainty, because not all of the
claims are really Ni deposits and the true-negative
sites are random points. However, this validation
indicates a high level of correlation between the
prospectivity model presented here and the reason-
ing for claiming areas for Ni exploration among
the companies executing Ni exploration within the
study area. Whether these high potential areas in-
clude new deposits remains to be seen in the near
future, when the exploration campaigns have been
executed.
The AUC values of the input data and the -
nal prospectivity models were all above 0.5, but
did not exceed 0.76 (Table 1). The ‘fuzzy gam-
ma’ model performed slightly better than the
‘fuzzy AND’ model. The ROC curve in Figure
19 indicates that the model is not perfect, but it is
denitely better than random sampling would give
us. Therefore, we can conclude from these re-
sults that this prospectivity model can be used for
selecting new exploration target areas for more
detailed assessment.
Fig. 19. ROC curve for the fuzzy gamma model. AUC is 0.68.
SUMMARY
is prospectivity model combined till geochem-
istry (Ni, Cu and Co) with airborne magnetics and
regional gravity data. e data were ltered us-
ing a high-pass ltering technique, in which the
long wavelength signal is removed from the data
revealing local anomalies. e resulting prospec-
tivity map identies more than 40 targets areas fa-
vorable for Ni-Cu deposits within the study area
in the northern Fennoscandian Shield. Most of
these areas are under active exploration, includ-
ing the Sakatti Ni-Cu deposit and Kevitsa Ni-Cu-
Au-PGE mine. A receiver operating characteristics
(ROC) curve was used for model validation. ROC
validation indicated that the input datasets and the
resulting prospectivity maps have a clear spatial
association with the current Ni exploration areas
dened by exploration licenses.
Table 1. AUC values of ROC tests.
Raster AUC
Co 0.76
Cu 0.74
Combined till geochemistry 0.74
Gamma 2 0.68
AND 2 0.62
Ni 0.60
Gravity 0.55
AM 0.54
27
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
CHAPTER III.
3D MODEL OF THE KITTILÄ TERRANE AND ADJACENT STRUCTURES
Tero Niiranen, Ilkka Lahti and Vesa Nykänen
INTRODUCTION
e Central Lapland Greenstone Belt (CLGB)
is one of the largest Paleoproterozoic greenstone
belts in the world, covering an area of ca 30 000
km2 (Fig. 20). e bulk of the CLGB consists of
a volcano-sedimentary sequence deposited on the
Archean basement during multiple episodes of
riing between 2.44–2.0 Ga. However, the mac
volcanic rock-dominated Kittilä terrane comprises
the core (Fig. 20).
Economically, the CLGB is very interesting, as it
not only hosts the largest gold-only deposit in Eu-
rope, namely the 7.67 Moz Suurikuusikko deposit,
but also a number of other gold-bearing deposits
and occurrences. e economic signicance of the
area was recognized in the early 1980s, and as a
result of exploration and bedrock mapping, a con-
siderable amount of geophysical and geological
data has been gathered from the district. e data
CALEDONIAN OROGENIC BELT
NORRBOTTEN
CRATON
KOLA CRATON
LAPLAND GRANULITE
BELT
SKELLEFTEÅ
BELT
KARELIAN CRATON
FINLAND
RUSSIA
SWEDEN
NORWAY
27°E
21°E
66°N
70°N
Central Lapland Greenstone belt
Kittilä Terrane
Study area
km
0 50 100
Fig. 20. Location of the study area, ex-
tent of the Central Lapland Greenstone
Belt, and location of the main geologi-
cal units. Contains data from the Na-
tional Land Survey of Finland Topo-
graphic Database 03/2013 © NLSand
HALTIK.
28
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
include high-quality airborne geophysics, ground
gravity data and several seismic proles, as well
as extensive drill hole and outcrop observation
datasets. A number of studies have been carried
out based on the existing data. As a result, several
geological interpretations have been presented as
maps, cross sections, and so on. So far, the 3D in-
terpretations have been limited to deposit-scale
studies focusing on mineral resource estimates.
In this chapter, we present the rst regional-scale
3D interpretation of the Kittilä terrane and key
structures in and around it. We also present a gold
potential estimation for the area based on the 3D
modeling results.
GEOLOGY OF THE STUDY AREA
e geology of the central part of the CLGB and
its immediate surroundings is illustrated in Figure
21. e CLGB records a complex geological his-
tory, including a series of riing events between
2.44 and 2.0 Ga followed by compressive tectonic
events and metamorphism related to Svecofennian
orogenic events at 1.91–1.79 Ga. e supracrustal
rocks of the CLGB can be divided into a Karelian
volcano-sedimentary sequence deposited on the
Archean basement during riing events at 2.44 Ga,
2.2 Ga, and 2.0 Ga, and a Svecofennian sequence
deposited on the Karelian units during 1.89–1.77
Ga (Lehtonen et al. 1998, Hanski et al. 2001, Hans-
ki & Huhma 2005). Hanski (1997) interpreted the
Nuttio suite serpentinites near to the eastern mar-
gin of the Kittilä Group as ophiolite fragments,
and suggested that the Kittilä Group represents an
allochthonous unit that is at least partly oceanic in
origin. is hypothesis is supported by the distinct
geochemical anity of the Kittilä Group mac
volcanic rocks, and the tectonic and/or tectonized
contacts between the unit and surrounding Kareli-
an rocks (e.g. Lehtonen et al. 1998, Hanski & Huh-
ma 2005). e geological evolution of the CLGB
culminated during the multi-stage deformation,
intrusive magmatism, and metamorphism related
to Svecofennian orogenic events at 1.91–1.79 Ga.
Stratigraphy
e supracrustal rocks of the CLGB have been
divided into seven lithostratigraphical groups,
which from oldest to youngest are the Vuojärvi,
Salla, Kuusamo, Sodankylä, Savukoski, Kittilä,
and Kumpu Groups (Fig. 22). In the northeastern
part of the study area, the Salla Group represents
the oldest unit deposited on the Archean base-
ment. Recently, a new lithostratigraphical unit, the
Vuojärvi Group, was distinguished in the south-
ern part of the CLGB (GTK digital bedrock data-
base 2013). e current interpretation is that the
quartzites and sericite-quartzites of the Vuojärvi
group overlie the Archean basement, or the Group
is of Archean age.
e Salla and Kuusamo Groups predominantly
consist of felsic to mac volcanic rocks represent-
ing the earliest volcanism related to the riing
of the Archean basement. e Sodankylä Group
chiey consists of clastic sedimentary rocks with
minor volcanic and carbonaceous intercalations.
e Sodankylä rocks were deposited on the Vuo-
järvi, Salla, and Kuusamo Groups, or in some cases
on the Archean basement. Widespread quartzites
with locally preserved cross-bedding, herringbone
and mud crack textures suggest that the deposi-
tional environment was at least locally tidal and
that the ri basin was considerably widened aer
the cessation of volcanism related to the Salla and
Kuusamo Groups (Lehtonen et al. 1998, Hanski &
Huhma 2005). e Savukoski Group is lithologi-
cally complex, consisting of mixed sequences of
komatiitic to tholeiitic volcanic rocks and phyl-
lites, graphite- and sulde-bearing schists, as well
as carbonaceous rock intercalations (Lehtonen et
al. 1998). In a number of locations, the volcanic
rocks preserve primary pillow lava textures, indi-
cating a sub-aqueous depositional environment.
e Kittilä Group is dominated by Fe- and Mg-
tholeiitic massive lavas, pillow lavas, and pyroclas-
tic rocks. Sedimentary interbeds and more wide-
spread sedimentary units have also been reported
within the volcanic rock units (Lehtonen et al.
1998). ese consist of metagraywackes, phyllites,
graphite- and sulde-bearing schists and tutes,
29
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
3360000 3390000 3420000 3450000 3480000
7500000 7530000 7560000
Gold deposits
Paleoproterozoic hypabyssal suites
Paleoproterozoic intrusive suites
Nattanen granite suite
Haaparanta suite
Paleoproterozoic supracrustal suites
Nuttio suite
Olostunturi suite
Eastern Lapland layered intrusion suite
Archean complexes
Paleoproterozoic Groups
Quartzite, siltstone, conglomerate,
intermediate to felsic volcanic rocks
Kumpu Group
Tholeiitic volcanic rocks, graphite- and sulphide-bearing tuffite,
BIF, phyllite, mica schist, graywacke
Kittilä Group
Quartzite, mica schist, mica gneiss, mafic volcanic rock
Sodankylä Group
Tholeiitic and komatiitic volcanic rocks
Kuusamo Group
Intermediate to felsic volcanic rocks
Salla Group
Vuojärvi Group
Quartzite, mica gneiss, possibly volcanic in origin
Tholeiitic and komatiitic volcanic rocks, phyllite, graphite and
sulphide-bearing schist, tuffite, dolomite
Savukoski Group
Paleoproterozoic complexes
Vuotso complex
Hetta complex
Central Lapland granitoid complex
Lapland granulite complex
Pomokaira and Muonio complexes
Diverse lithodemes
Haaskalehto gabbro-wehrlite suite
Nilipää granite suite
Major Structures
Shear/Fault zone
Thrust zone
Orogenic gold
Paleoplacer
Iron oxide-Cu-Au
Keivitsa layered intrusion suite
Fig. 21.Geology of the Kittilä terrane and adjacent area. Location of the known gold deposits and occurrences. Modied GTK
digital bedrock map database and FINGOLD.
30
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
carbonate rocks, and banded iron formations.
Lehtonen et al. (1998) divided the mac volcanic
rocks of the Kittilä Group into two formations:
the Vesmajärvi and Kautoselkä formations. e
geochemical data indicate that the former volcanic
rocks have an anity for oceanic basalts, and the
latter bear characteristics of passive margin tholei-
ites (Lehtonen et al. 1998, Hanski & Huhma 2005).
e depositional basement of the Kittilä Group
rocks is unknown.
e Kumpu Group represents the youngest
supracrustal unit in the CLGB. It predominantly
consists of clastic, molasse-like sedimentary rocks
deposited on a major stratigraphic break. Minor
amounts of intermediate to felsic volcanic rocks
occur in the unit in the western part of the CLGB
(Lehtonen et al. 1998).
Intrusive magmatism
e early intrusive magmatism consists of 2.44 Ga
and 2.05 Ga mac layered intrusions divided into
the Eastern Lapland and Kevitsa suite intrusions,
respectively (Fig. 21). Abundant mac dyke mag-
matism indicating repeated riing took place at
2.22 Ga, 2.05 Ga, and 2.0 Ga (Hanski et al. 2001).
e 2.2 Ga mac dykes and dierentiated sills of
the Haaskalehto gabbro-wehrlite suite are the most
abundant, predominantly occurring within the
Sodankylä Group rocks along an E–W-trending
zone south of the southern margin of the Kittilä
terrane (Fig. 21). e 2.05 and 2.0 Ga magmatism
predominantly consists of diabase dykes and small
gabbroic intrusions, although in places coeval
felsic quartz-feldspar dykes have been detected
(e.g. Lehtonen et al. 1998, Hanski et al. 2001). e
1.91–1.86 Ga Haparanda Suite mac to felsic in-
trusions comprise the syn-orogenic, and 1.82–1.79
Ga granitoids the late-orogenic intrusions in the
area. Nilipää suite granites occur in the southern
margin of the CLGB and have an enigmatic age of
2.1 Ga, suggesting that they were intruded during
or between the extensional stages (e.g. Hanski et
al. 2001).
Archean Basement
Vuojärvi Gp
Salla Gp
Kuusamo Gp
Sodankylä Gp
Savukoski Gp
Kittilä Gp
Kumpu Gp
1.80
1.89-
1.86
1.91
2.05
2.2-
2.14
2.44
2.1
>1.88 Ga terrestrial sedimentary rocks,
intermediate to felsic volcanic
rocks
ca. 2.0 Ga tholeiitic volcanic rocks,
shallow to deep marine sedimentary rocks
>2.05 Ga komatiitic to mafic volcanic
rocks, shallow marine sedimentary rocks
>2.2 Ga terrestrial to shallow marine
sedimentary rocks, mafic to intermediate
volcanic rocks
tholeiitic and komatiitic
volcanic rocks
2.44 Ga felsic to intermediate
volcanic rocks
undefined quartzite, mica gneiss
and felsic to intermediate volcanic
rocks, possibly Archean in age
felsic intrusions
mafic to felsic intrusions
mafic layered intrusions,
mafic sills and dykes
tectonic contact with
ophiolite fragments
2.1 igneous age in Ga
Central Lapland Greenstone Belt
Fig. 22. Stratigraphy and igneous ages of the Central Lapland Greenstone Belt. Modied aer Lehtonen et al. (1998), Hanski et
al. (2001) and GTK’s FINSTRATI database.
31
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
Deformation and metamorphism
e CLGB records a complex multi-stage defor-
mation history that, despite numerous investi-
gations, is still not completely understood. e
following sequence of deformation events is mod-
ied aer the interpretations of Ward et al. (1989),
Lehtonen et al. (1998), Väisänen et al. (2002), Tu-
isku and Huhma (2006), Hölttä et al. (2007), and
Patison (2007). e deformation history of the
CLGB area is divided into three ductile stages
followed by a completely brittle stage. e main
deformation stages, D1 and D2, lack clear over-
printing relationships and geochronology, and are
therefore grouped here as the D1-2 stage. e D1-2
stage relates to thrust tectonics and resulted in the
main ductile deformation features of the CLGB.
Features of the stage indicate S–SW-vergent trans-
port and N–NE-vergent transport in the north-
ern and southern parts of the CLGB, respectively.
e S–SW-vergent thrusting in the north relates
to collision of the Karelian and Kola cratons and
thrusting of the Lapland Granulite belt to the SW.
e northward-directed thrusting in the south
involved the initial generation of the south-dip-
ping Sirkka and Venejoki thrust zones (Fig. 23). It
is currently unknown whether the two thrusting
events were contemporaneous or successive.
e D3 stage resulted in deformation features
with highly variable vergence, depending on the
location, and it may have consisted of several
stages that do not necessary relate to each other.
e D3 deformation resulted in the generation of
N- to NE-striking shear zones and local refolding
3360000 3390000 3420000 3450000 3480000
7500000 7530000 7560000
Major Structures
Unspesified shear/fault zone
Thrust zone
DGRF-65 anomaly (nT)
High : 17859
Low : -8950
1. 1.
2. 1.
2.
2.
3.
3.
3.
4.
5.
6.
7.
Fig. 23. Aeromagnetic map of the study area with the location of the key structures. 1. Sirkka thrust, 2. Venejoki thrust,
3. Kiistala shear zone, 4. Muusa shear zone, 5. Äkäsjoki shear zone, 6. Jerisjärvi thrust, 7. Enontekiö shear zone. GTK data.
32
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
of the earlier deformation features. In the western
part of the Sirkka thrust zone, clear indications ex-
ist of reactivation of the structure during this stage
(e.g. Patison 2007, Saalmann & Niiranen 2010).
e external events resulting in the D3 deforma-
tion and the robust timing of it are unknown. e
minimum age and at the same time the maximum
age for the following brittle D4 stage is 1.77 Ga
(Väisänen et al. 2002). e probable age frame for
the D3 is either 1.89–1.86 Ga or 1.82–1.77 Ga, and
it may vary between dierent parts of the CLGB
(e.g. Väisänen et al. 2002, Patison 2007).
e peak metamorphic conditions in the CLGB
were reached during the D1-2 stage. A characteristic
feature of the metamorphism in the CLGB is that
excluding the northern and western margin, the
metamorphic conditions within the Kittilä terrane
are of mid-greenschist facies (Hölttä et al. 2007).
e metamorphic grade increases away from the
Kittilä terrane, being upper amphibolite facies in
the south close to the rocks of the Central Lapland
Granitoid Complex and mid-amphibolite facies
in the western margin. e highest metamorphic
conditions are recorded next to the Vuotso and
Lapland granulite complex rocks, in which a peak
metamorphic pressure and temperatures of up to
12 kbar and 800–850 oC, respectively, were reached
(Tuisku & Huhma 2006). e peak metamorphic
temperature estimates for the Kittilä terrane are ca.
350 oC, although no reliable pressure estimate ex-
ists for the lowest grade metamorphic area (Hölttä
et al. 2007). e western margin of the Kittilä ter-
rane has metamorphosed at a higher temperature
due to heat ow from adjacent granitoids. A garnet
thermobarometry from this area yields a pT esti-
mate of 3.2 kbars and 550oC (Hölttä et al. 2007).
DATA AND METHODS
e datasets used in this study comprised the Geo-
logical Survey of Finland’s geophysical data, includ-
ing airborne magnetic, EM, and radiometric data,
ground gravimetric data, and various ground geo-
physical data sets. In addition GTK’s digital bed-
rock database, bedrock observation data and drill
core data were utilized. Much of the interpretation
of deep structures relied on seismic data, for which
data were utilized from the Finnish Reection Ex-
periment FIRE 2001–2005 and High Resolution
Reection Seismics for Ore Exploration 2007–2010
(HIRE) programs (Kukkonen & Lahtinen 2006,
Kukkonen et al. 2011). e locations of the seismic
proles are indicated in Figure 24.
e data were processed in several ways. For
example, potential eld datasets were processed
using multiscale wavelet edge detection (“worm-
ing”), described in detail in Chapter I of this vol-
ume. In addition, various derivative maps were
utilized for the magnetic data to establish the con-
tacts and dips of units and structures.
Modeling of the Kittilä terrane was based on
3D modeling of regional Bouguer gravity data
and seismic data. A multidisciplinary approach
was used in all modeling, and much of the mod-
eling process was iterative. e visualization,
interpretation, and modeling were carried out
using GocadTM soware with Mira Mining utilit-
iesTM and GRGPack research plugins provided by
the Gocad Research Group of the Nancy School
of Geology.
THREEDIMENSIONAL GRAVITY MODELING OF THE KITTILÄ TERRANE
ree dimensional gravity modeling was carried
out to obtain information on the deep geometry of
the Kittilä terrane and surrounding areas (Fig. 25).
e Kittilä terrane is favorable for gravity mode-
ling, as a large number of ground gravity measure-
ments have been carried out in the area, and the
belt itself produces a signicant positive Bouguer
anomaly of about ~45 mGal (Fig. 24). e anom-
aly is caused by thickness variations in the high-
density mac volcanites of the terrane in contrast
to lower density surrounding felsic rocks such as
granitoids, quartzites, and schists. e data and
measurements used have been described in more
detail in Chapter II of this volume.
e regional Bouguer gravity data (~20 000
measurements, 1–4 measurements/km2) were
33
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
3360000 3390000 3420000 3450000 3480000
7500000 7530000 7560000
Bouguer
[mgal]
High : 15.6853
Low : -57.9864
Seismic profile
Kolari HIRE
Petäjäselkä HIRE
Suurikuusikko HIRE
FIRE 4B
FIRE 4A
FIRE 4A
Fig. 24. Bouguer anomaly map of the study area with locations of the reection seismic proles used in the modeling.
gridded, and modeling lines were created by sam-
pling the grid along 20 N–S-oriented 80-km-long
lines, spaced 5 km apart. e modeling area was
therefore 95 km in the E–W direction and 80 km
in the N–S direction (Fig. 26). e model was con-
structed from a number of layer-type three-di-
mensional bodies shown in Figure 26. Due to the
thousands of bedrock observations and drillings
performed in the modeling area, the surface geol-
ogy is well known. erefore, the topmost part of
the model is primarily based on the known lithol-
ogy. Model densities, geometries, and the regional
trend were manually modied to improve the data
t between the measured and calculated data along
the modeling lines. Although such a “trial and er-
ror” modeling procedure is very time consuming,
it allows the results to be checked against the geol-
ogy and other geophysical results during the pro-
cess. e model was revised several times in 3D
using the Gocad soware together with reection
seismics to obtain good agreement between the
model and the seismic data. In order to gain a bet-
ter nal modeling result, the most signicant geo-
logical units around the terrane causing positive
and negative anomalies were also incorporated in
3D modeling, since the use of a single background
density would yield inaccurate modeling results.
e area is characterized by granitoids in the north
and east, such as the Hetta, Tepasto, and Postojoki-
Sovasjoki granitoids, which cause negative Bougu-
er anomalies. Densities of 2600–2650 kg/m3 and
thicknesses of 1–6 km were used for granitic in-
trusions to t the negative Bouguer anomalies en-
countered in the north and east. To the south of
the Kittilä terrane, more dense rock types such as
schist, mica schist, and mac–ultramac intrusions
34
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
Measured vs. calculated
Regional
Measured Calculated
Depth [m] Bouguer
[mGal]
Distance [m]
Fig. 25. 3D gravity modeling of the Kittilä terrane. Upper images show measured and modeled gravity data. e lower panel
presents an example of one of the modeling proles.
a) b)
Fig. 26. a) 3D model of the Kittilä terrane and surrounding geological blocks used in the modeling. View from the northwest
and b) Kittilä terrane gravity model from the northeast. Modeling proles are presented in both images.
35
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
create short-wavelength positive anomalies. High-
er density blocks (2650–3050 kg/m3) were added
in the southern part of the model. e Kittilä ter-
rane was modeled using the density of 2950 kg/m3,
which is a representative estimate for mac volcan-
ites in area. e background density of 2780 kg/m3
was assigned for the rest of the 3D model, as this
value represents an average between the mac and
felsic rock types in the area. Several smaller-scale
high- or low-density bodies were necessary to add
in the terrane model. ese anomalies are caused
by banded iron formations (Porkonen), quartzites
(Kumpu), and granites (Ruoppapalo).
Figure 26 presents three-dimensional perspec-
tive views of the nal model. e terrane can be
modeled using 8 layers, each having a thickness of
1 km. e terrane is relatively shallow (1–2 km)
in the eastern and western parts, but becomes
considerably thicker around the Sirkka area. e
abrupt thickening is in good agreement with the
reection seismic results of the FIRE seismic pro-
les, which show steeply dipping reectors in the
same area (Patison et al. 2007). Although it is chal-
lenging to accurately estimate the thickness of the
terrane, this modeling suggests that the thickest
part occurs around the Kiistala-Lintula area.
MODELED STRUCTURES AND UNITS
Kittilä terrane
e bottom of the Kittilä terrane has been con-
structed on the basis of the 3D gravity model and
seismic data. e surface expression is as outlined
by GTK’s digital bedrock map. In many locations
of the FIRE proles, there is a sub-horizontal,
highly reective unit below the Kittilä terrane that
corresponds well to the base outlined by the grav-
ity forward model (Patison et al. 2006, Niiranen
et al. 2009). e contact can be also outlined rela-
tively clearly in a number of HIRE proles from
the central and eastern part of the Kittilä terrane.
e base of the Kittilä terrane was modeled as a
surface, and the terrane was also modeled as a 3D
block and voxet model to calculate the volume and
for further numerical estimations. e modeled
base of the Kittilä terrane is presented in Figure 27.
e base reaches the depth of 9500 meters in the
central parts, where the unit appears to have been
bulged down. is is slightly thicker than the 8 km
thickness of the gravity model. However, the nal
Fig. 27. e base of the Kittilä terrane with depth contours in meters below the current bedrock surface.
36
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
model is based on the combination of the gravity
model and the seismic data smoothed using DSI
interpolation. e lowermost 1.5 km volume cov-
ers only a small area and is in accordance with the
gravity anomaly. e unit thins toward its bounda-
ries from this location. e unit is thinnest in the
western part, where the base reaches a depth of ca.
1000 m, and in SE corner, where the base surfaces.
Venejoki and Sirkka thrusts
e Venejoki and Sirkka thrusts, which form
northward-directed thrust systems, were modeled
as surfaces (Fig. 28). Both structures were inter-
preted using geophysical and bedrock maps. e
FIRE 4A prole intersects both structures, and
they can be seen as low reectivity zones cross-
cutting or truncating sub-horizontal reectors
(e.g. Patison et al. 2006, Niiranen et al. 2009). e
apparent dips of the structures are 45 and 30 de-
grees for the Sirkka and Venejoki thrusts, respec-
tively, being very close to the true dips, as the FIRE
4 prole crosses the structures almost perpendicu-
larly (Fig. 29). e Sirkka thrust reaches a depth of
at least 9 km, and the Venejoki thrust can be traced
to reach the mantle at a depth of ca. 42 km (Patison
et al. 2006).
Kiistala and Muusa shear zones
Kiistala and Muusa shear zones are the two strike
slip systems cross-cutting the Kittilä terrane for
which there was sucient data to model them
(Figs. 21, 23, and 28). ey both are NE- to N-
striking strike-slip structures, the Kiistala being
interesting for gold, as the Suurikuusikko orogenic
gold deposit is hosted by it. e Muusa shear zone
appears as a low magnetic, cross-cutting linear
feature on aeromagnetic map and as a sub-verti-
cal NW-dipping transparent zone cross-cutting
sub-horizontal seismic reectors on the FIRE 4B
prole (Fig. 30, Niiranen et al. 2009). ere is no
direct drilling or outcrop evidence of the struc-
ture, and the geological features are consequently
1.
1.
1.
2.
2.
2.
3.
3.
3.
4.
5.
6.
7.
7.
Fig. 28. Modeled key structures and the base of the Kittilä terrane (green) as surfaces. 1. Sirkka thrust, 2. Venejoki thrust,
3. Kiistala shear zone, 4. Muusa shear zone, 5. Äkäsjoki shear zone, 6. Jerisjärvi thrust, 7. Enontekiö shear zone.
37
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
Fig. 29. Sirkka thrust (le) and Venejoki thrust (right) with aeromagnetic data and the seismic FIRE 4 prole. Oblique view
from the west.
Fig. 30. Modeled Muusa shear zone (bluish gray with red borders) on the aeromagnetic map and seismic FIRE 4B prole. e
structure cross-cuts the highly reective gently dipping units on the seismic data. Also note the apparent sinistral characteristic
of the shear zone on the aeromagnetic map.
38
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
unknown. Based on the aeromagnetic data, it has
an apparent sinistral component.
e outcrop and drill core data of the Kiistala
shear zone are mainly from the Suurikuusikko de-
posit and adjacent area. Patison (2007) and Patison
et al. (2007) described the structure as sub-vertical
strike-slip shear zone with early sinistral and late
dextral movement. Based on the drilling data from
the Suurikuusikko deposit, the structure extends to
a depth of at least 1500 meters. On aerogeophysi-
cal maps, the structure can be outlined to extend
across the Kittilä terrane, potentially reaching the
Lapland granulite arc in the NE and cross-cutting
the Sirkka thrust in the SW, splaying to at least two
segments around the Suurikuusikko deposit (Fig.
21). e structure can be perceived on the seismic
HIRE proles from the Suurikuusikko area. How-
ever, being vertical to sub-vertical, it is not a very
prominent feature. e structure is modeled as
two planes, of which the southern segment dips ca.
80 degrees to the west and the northern segment
85 degrees to the east (Fig. 28).
Structures in western part of the area
ree key structures were modeled from the
western part of the study area: the Äkäsjoki and
Enontekiö shear zones, and the Jerisjärvi thrust
zone limiting the western margin of the Kittilä ter-
rane (Figs. 21, 23, and 28). All of these probably
form part of the crustal-scale Balthic-Bothnian
megashear (BBSZ), as described by Berthelsen &
Marker (1986). e Äkäsjoki shear zone was mod-
eled based on the Kolari HIRE seismic data and
geophysical data, and the Enontekiö shear zone
and Jerisjärvi thrust were modeled based on the
seismic interpretation on FIRE 4B prole and
geophysical interpretation (Patison et al. 2006, Ni-
iranen et al. 2009).
e Äkäsjoki shear zone appears in the airborne
magnetic map as a linear low magnetic zone that in
places truncates magnetic low and magnetic high
zones, clearly being a local-scale tectonic block
boundary. It appears in the Kolari HIRE seismic
data as a transparent zone that cross-cuts the gen-
tly dipping reector packages (see Chapter IV, this
volume). e modeled shear zone dips 85 degrees
to the NW. ere are no direct observations on the
structure, although it may be a southwestern ex-
tension to the Muusa shear zone.
e Jerisjärvi thrust zone limits the western
margin of the Kittilä terrane. It is most likely a
northern extension to the thrust zone, extending
from Hannukainen in Kolari, being truncated by
the Enontekiö shear zone. In the seismic FIRE 4B
prole, it occurs in a highly tectonized zone with
numerous vertical to west- and east-dipping trans-
parent zones, indicating the presence of a set of
shear and thrust zones relating to the BBSZ (e.g.
Patison et al. 2006, Niiranen et al. 2009). e Jer-
isjärvi thrust is modeled as a half bowl-shaped
thrust limited at its western margin by the Enon-
tekiö shear zone (Fig. 28). It reaches a depth of 8.5
km.
e Enontekiö shear zone appears on the aero-
magnetic map as a distinct, >100-km-long, NNW-
striking zone that clearly divides the area into two
blocks with dierent magnetic patterns (Fig. 23).
e structure has an apparent sinistral strike-slip
component and appears to be linked to the Äkäs-
joki shear zone at its southern end. Based on the
seismic data, the structure is sub-vertical, dipping
80 degrees to the WSW, and reaches a depth of
at least 12 km (Patison et al. 2006, Niiranen et al.
2009). e structure braids into at least two seg-
ments at its southern end and has NNE–SSW-
striking conjugates next to the Jerisjärvi thrust
zone (Fig. 28).
DISCUSSION
e 3D model of the Kittilä block indicates that
it is considerably thicker at its thickest part than
previous 2D interpretations suggested. According
to Lehtonen et al. (1998), the base of the Kittilä ter-
rane is at ca. 6 km in the Tepasto-Pomokaira area.
is 2D interpretation prole is about 8 km north
of the thickest part of the Kittilä terrane in our
model. e base of the deepest part in our model
at the location of the Tepasto-Pomokaira prole
of Lehtonen et al. (1998) is 1.5 km deeper than
39
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
in their 2D interpretation. In the Kumputunturi
area, the previous 2D investigation prole suggest-
ed that the Kittilä terrane is up to 5 km thick. At
the corresponding location in our 3D model, the
thickness is about 5.5 km, being relatively consist-
ent with previous investigations.
e Kittilä terrane has been suggested to rep-
resent a fragment of oceanic crust obducted and
emplaced to its current position during the early
stages of the Svecofennian orogenic events (e.g.
Hanski 1997, Hanski & Huhma 2005). e 9.5 km
thickness indicated by the model is about twice the
thickness of an oceanic crust. is, together with
dramatic thickness variation and the 3D shape of
the modeled Kittilä terrane, strongly suggests that
the unit has been thickened via stacking and fold-
ing during the Svecofennian orogenic events.
A number of studies on the structural geology
of the CLGB have suggested bivergent thrusting
towards the Kittilä terrane: thrusting of the Lap-
land Granulite Belt to the SW during the collision
of the Karelian and Kola cratons around 1.91 Ga,
and northward-directed thrusting that was either
synchronous or a subsequent event (e.g. Ward et al.
1989, Lehtonen et al. 1998, Tuisku & Huhma 2006,
and Patison 2007). e 3D model of the Lapland
Granulite Belt and its foreland units is described
in Chapter V of this volume. e 3D model of the
Venejoki and Sirkka thrusts illustrates the north-
ward-directed thrusting systems. e exact tim-
ing and ultimate external driving force is unclear,
but the thrusting most likely took place between
either 1.91–1.89 Ga or 1.86–1.77 Ga (e.g. Patison
2007 and references therein). e characteristics
of the Venejoki thrust are poorly known and all
interpretations of this structure are based on seis-
mic and geophysical data. In contrast to this, the
Sirkka thrust has been explored for gold since the
1980s and it has been described in a number of
publications (e.g. Ward et al. 1989, Lehtonen et al.
1998, Patison 2007). A cluster of 2.2 Ga dolerite
dykes and sills has been detected along the Vene-
joki thrust, and Sattasvaara formation komatiites
occur along the Sirkka thrust (e.g. Lehtonen et
al. 1998). ese features suggest that the Vene-
joki and Sirkka thrust systems may already have
been initially formed during the riing stage(s) as
normal ri faults and reactivated as thrust zones
during the basin inversion. In addition, there are
number of indications that the Sirkka thrust was
reactivated aer the northward-directed thrusting
as a strike-slip shear (e.g. Patison 2007, Saalmann
& Niiranen 2010). If these scenarios are true, the
mineral potential of these structures is even higher
than has previously been considered, as many hy-
drothermal deposit types with structural control
are most abundant in structures with a prolonged,
multi-stage deformation history.
e modeled Muusa and Kiistala shear zones
are just two of the several NE–SW-striking strike-
slip shear zones cross-cutting the Kittilä terrane.
e data on this set of shear zones suggest that
they have been active during the regional D3 stage.
However, they may also have had a considerably
longer history. As already suggested by Patison
(2007), among others, these structures may have
originally been initiated during the D1-2 thrusting
events as transfer faults relating to the thrust zones
prior to the D3, and were multiply reactivated dur-
ing the later stages. Visual examination of the 3D
models presented in this volume suggest that this
is a viable hypothesis, but robust timing constrains
are needed to verify this.
e BBSZ, which in some publications is re-
ferred to as the Pajala Shear Zone, or Kolari Shear
Zone, is clearly a signicant crustal-scale feature.
However, the geological evolution, timing, and
external driving mechanisms of the structure are
poorly known. It has been proposed to represent a
collisional boundary between the Norrbotten cra-
ton in the west and the Karelian craton in the east
(Lahtinen et al. 2005), although very little concrete
evidence for this has been presented. Only a few
indications of the timing of the tectonic evolu-
tion have been reported. From the Kolari region,
Hiltunen and Tontti (1976) and Hiltunen (1982)
suggested that the thrusting and shearing was co-
eval with the 1.89–1.86 Ga Haparanda suite mag-
matism, and Niiranen et al. (2007) suggested that
the structures related to the BBSZ at Kolari were
(re-)activated during 1.82–1.79 Ga. e cross-
cutting evidence for the modeled Jerisjärvi thrust,
the Äkäsjoki and Enontekiö shear zones with the
Sirkka thrust and Kittilä terrane, suggests that
the initial formation of the BBSZ structures post-
dated events in relation to the northward-directed
thrusting at the southern margin of the CLGB.
Clearly, more research, including robust dating,
needs to be carried out before the geological evo-
lution of the BBSZ can be unveiled.
40
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
Implication for the gold potential of the Kittilä terrane
e Kittilä terrane is an economically prospective
area, especially for gold deposits, not least due to
the discovery of the world-class Suurikuusikko de-
posit within the central part of the area. Numer-
ous other showings and deposits are known within
and in the southern margin of the terrane. All but a
few of these fall into the category of orogenic gold
deposits. Groves et al. (1998) presented a genetic
model for orogenic gold deposits in which the met-
als and uids involved are released from the coun-
try rocks during metamorphism (with or without
input from intrusions) and are focused into suit-
able structures that act as pathways, and metals are
subsequently deposited into suitable traps. Pitcairn
et al. (2006) presented data from the Otago and Al-
pine belts in New Zealand showing a clear pattern
of depletion in gold and gold-related metals from
mac volcanic and sedimentary rocks undergoing
progressive metamorphism. eir work indicates
that gold concentrations in sedimentary and mac
volcanic rocks metamorphosed in amphibolite fa-
cies conditions show 50–80% depletion compared
with their unmetamorphosed varieties, which is in
agreement with the model presented by Groves et
al. (1998).
In this work, we applied the genetic model by
Groves et al. (1998) and developed a scenario in
which the mac volcanic rocks of the Kittilä ter-
rane are the dominant source for the uids and
gold, and these are released from the rocks in met-
amorphic reactions involving devolatization via
the breakdown of hydrous silicates. In mac rocks,
the devolatization is generally strongest around
the greenschist-amphibolite facies boundary at a
temperature of around 525oC.
Based on the work of Hölttä et al. (2007), the
peak metamorphic temperature of the Kittilä ter-
rane at the current erosional level is ca. 350oC and
the pressure 3.2 kbar. Using these gures and a
density of 2.95 kg/dm3 for the Kittilä terrane, the
temperature gradient during the peak of the meta-
morphism was 37oC/km. e metamorphic peak
a.
b.
c.
V = 9500 km3
V = 1500 km3
Fig. 31. a. Voxet model of the Kittilä terrane. Northern, northwestern, and westernmost high-T metamorphic zones are ex-
cluded. V is the total volume of the model. b. Calculated peak metamorphic temperature on the voxet model. e blue area is
in greenschist facies, the red area is in amphibolite facies, and the white zone illustrates the transition zone between greenschist
and amphibolite facies at 500–550oC. c. Amphibolite facies part of the modeled Kittilä terrane with the calculated volume (V).
Oblique view from the SE in all images. e scale is the same in all images.
41
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
temperature was calculated in a voxet model of the
Kittilä terrane as a function of depth using these
gures (Fig. 31). e volume of the lower part of
the Kittilä terrane in amphibolite facies is 1500
km3, and using the average density of 2.95 kg/dm3,
the total mass of this part is 4425 Bt.
Data on the background levels of Au in mac
volcanic rocks of the Kittlä terrane are lacking.
However, the data from the literature suggest that
the average background gold concentrations for
mac volcanic rocks vary between 0.75 ppb and
4.7 ppb, depending on the type of mac volcanic
rocks (Pitcairn, 2012 and references therein). We
performed our calculations using a background
concentration of 2 ppb Au. Based on this, the to-
tal amount of gold in mac metavolcanic rocks of
the Kittilä terrane subjected to amphibolite facies
metamorphism was 285 Moz (8850 t). If 50–80%
of this was mobilized, as indicated by the work by
Pitcairn (2006), a total of 143–228 Moz gold was
mobilized from the mac volcanic rocks of the Kit-
tilä terrane during the metamorphism. ese g-
ures are considerable compared with the currently
known gold resources within the Kittilä block (Su-
urikuusikko and Kuotko ~7.8 Moz, Agnico Eagle
data). Obviously, the gold and uids released at
depth by metamorphism need to be focused into a
suitable structure such as the KiSZ, and gold needs
to be precipitated in suitable trap, both events oc-
curring with an unknown eciency. Also, some of
the gold may have precipitated above the current
erosional level. Despite the numerous assump-
tions and hypothetical nature of the calculations
presented above, it seems very likely that the un-
discovered gold resources in the Kittilä terrane
are potentially several tens of millions of ounces.
Obviously, until the true background gold concen-
trations are available for rening the calculations,
the above-presented gures should be considered
as “blue sky” estimates.
CONCLUSIONS
e work presented above suggests that the Kit-
tilä terrane is thicker than previous investigations
have suggested. e considerable thickness varia-
tions of the terrane are due to the thrusting-relat-
ed tectonics, resulting in the folding and stacking
of the mac volcanic rocks and associated sedi-
ments during the main deformation stages of the
Svecofennian orogeny. e Venejoki and Sirkka
thrust zones form a northward-directed thrust
system. ese structures may already have formed
as normal fault zones during the riing stages and
may have subsequently been reactivated as thrust
zones during the inversion of the basin, and yet re-
activated during the later orogenic events, making
them long-lived crustal-scale structures and thus
highly prospective for the discovery of epigenetic
metal deposits.
e Äkäsjoki and Enontekiö shear zones and the
Jerisjärvi thrust zone form part of the Baltic-Both-
nian megashear, which, based on the seismic data,
is clearly a crustal-scale feature. Until further work
on these structures is carried out, the exact nature
and timing of this system remains unknown.
Based on this work, potentially up to 228 Moz
gold was mobilized from the Kittilä Group rocks
alone. is gure is almost 30 times greater than
the total reported gold resources in known depos-
its in the Kittilä terrane. Although rough, our esti-
mate shows that metamorphic processes alone can
easily have liberated enough gold for the known
deposits, and a considerable amount of undiscov-
ered gold most likely remains in the terrane.
42
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
CHAPTER IV.
3D MODEL OF THE KOLARI REGION
Tero Niiranen, Vesa Nykänen and Ilkka Lahti
INTRODUCTION
e Kolari district is located in the western part of
Finnish Lapland. e area has been under active
exploration for decades, and some 15 iron ± Cu-
Au deposits have been discovered in the area. Vari-
ous genetic interpretations have been suggested
for the known deposits. ey have been proposed
to represent metamorphosed syngenetic iron
formations (e.g. Hiltunen & Tontti 1976, Mäkelä
& Tammenmaa 1978, Väänänen 1998), strata-
bound, intrusion-related epigenetic skarn deposits
(Hiltunen 1982), or epigenetic iron oxide-Cu-Au
deposits (Niiranen et al. 2007). In 2010, GTK and
Northland Resources S.A., a company that is cur-
rently carrying out exploration and development
of the Kolari deposits, carried out in co-operation
a deep seismic prole program (HIRE) over key
targets areas in the Kolari district. e data became
available for research aer a one-year quarantine
period and were used in the Central Lapland 3D
modeling program of GTK. e results of the 3D
modeling and its implication for the stratigraphy
of the area and the genetic interpretation of the
known deposits are presented in this chapter.
GEOLOGICAL FRAMEWORK
e Kolari district forms the westernmost ex-
tension of the Central Lapland Greenstone Belt
(CLGB). e bedrock of the area consist of a 2.44–
2.05 Ga Karelian supracrustal sequence of ri
related mac volcanic rocks and associated sedi-
mentary rocks, a >1.89 Ga Svecofennian sequence
of clastic sedimentary rocks, 2.2 Ga dolerite dykes,
1.89–1.86 Ga Haparanda suite intrusions, and
1.82–1.77 Ga granitoids (Fig. 32, Hiltunen 1982,
Väänänen 1998, Niiranen et al. 2007). e domi-
nant Karelian rocks in the Kolari region belong to
the Kuusamo, Sodankylä, and Savukoski Groups,
the latter two dominating the study area (Figs. 32
and 33). ese are overlain by clastic sedimentary
rocks of the Svecofennian Kumpu Group.
e CLGB was subjected to multi-phase defor-
mation and metamorphism during the Svecofen-
nian orogenic events between 1.91–1.77 Ga. e
deformation has been divided into three ductile
stages, D1-3, followed by a completely brittle D4
stage (e.g. Väisänen 2002, Hölttä et al. 2007, Pa-
tison 2007). e earliest stage, D1, resulted in
bedding parallel foliation rarely visible in macro-
scopic observations (Lehtonen et al. 1998, Hölttä
et al. 2007). e D2 stage in the CLGB relates to
thrust tectonics resulting in the main foliation in
the CLGB, S2, and recumbent or reclined F2 fold-
ing (Väisänen 2002, Hölttä et al. 2007). e ver-
gence of the F2 folding varies within the area, being
to the SW to W in the northern and northeastern
part of the CLGB and to the N to E in the south
and west (Väisänen 2002, Hölttä et al. 2007). e
S2 is in general gently dipping or at-lying and
sub-parallel to bedding. e D2 structures were
overprinted by sets of F3 folds and late shear zones
with varying orientations. e F3 folds are upright
to steeply inclined, being northward vergent in the
central parts of the CLGB and eastward vergent in
SW and W parts of the area (Väisänen 2002, Hölttä
et al. 2007). In the eastern and northern part of the
43
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
((
((
((
((
((
((
((
((
((
((
((
((
((
((
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
3365000 3370000 3375000 3380000
7486000 7492000 7498000 7504000 7510000
Legend
Supracrustal units
!
(
Iron(-Cu-Au) deposit
HIRE profiles
Unspecified shear zone
((
Thrust zone
Intrusives
1.82-1.79 Ga granite
1.89-1.86 Ga Haparanda Suite diorite
2.0 Ga dolerite dykes
1.89-1.86 Ga Haparanda Suite monzonite
Niesakero Fm
Kumpu Group
Savukoski Group
Vuojärvi Group
Sodankylä Group
Juurakkojärvi Fm
Tapojoki Fm
Tuulijoki Fm
Ylläs Fm
Kolari Fm
Rautuvaara Fm
Kuervitikko
Äkäsjoki Shear Zone
0 4 82
km
E1
E2
V1
V1
V2
V2
V3
V3
V4
V5
V6
Haisujupukka Fm
3D model extent
Tuohilehto
Hannukainen
NE-Rautuvaara
SW-Rautuvaara
Cu-Rautuvaara
Sivakkalehto
Rautuoja
Rautuhelukka
Rytijänkä
Taporova
Suuoja
Fig. 32. Geological map of the study area with the location of the known Fe ± Cu, Au deposits, seismic proles, and the extent
of the 3D model. Geological map modied aer Bedrock of Finland − DigiKP.
CLGB the vergence is to the west. D2 took place be-
tween 1.91 Ga and 1.86 Ga, and D4 is younger than
1.77 Ga (Väisänen 2002, Hölttä et al. 2007, Pati-
son 2007). D3, which most likely consists of several
sub-stages varying in age in dierent parts of the
CLGB, took place between 1.89 Ga and 1.77 Ga
(Patison 2007). e peak metamorphic conditions
vary from upper amphibolite to mid-greenschists
facies and were reached prior to the development
of D3 shear zones (Hölttä et al. 2007). In the Ko-
lari district, the peak metamorphic conditions are
of mid-amphibolite facies conditions (Hölttä et al.
2007).
Stratigraphy of the study area
e initial detailed stratigraphy of the Kolari re-
gion was laid out by Hiltunen and Tontti (1976)
and Hiltunen (1982). According to these works,
the lowermost unit, the Niesakero-Kuertunturi
quartzite complex, consists of quartzite, arkosite,
conglomerate and mica gneiss, with thin amphi-
bolite intercalations (Fig. 33). ese rocks are cor-
related with the >2.2 Ga Sodankylä Group rocks.
e Niesakero-Kuertunturi quartzite complex is
overlain by the Rautuvaara and Kolari greenstone
Formations, representing the Savukoski Group in
the area (Fig. 33). e Kolari greenstone and Rau-
tuvaara Formations have similar lithology con-
sisting of mac volcanic rocks, graphite-bearing
44
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
Archean Basement
Vuojärvi Gp
Salla Gp
Kuusamo Gp
Sodankylä Gp
Savukoski Gp
Kittilä Gp
Kumpu Gp
1.80
1.89-
1.86
1.91
2.05
2.2
2.14
2.44
2.1
>1.88 Ga terrestrial sedimentary rocks,
intermediate to felsic volcanic
rocks
ca. 2.0 Ga tholeiitic volcanic rocks,
shallow to deep marine sedimentary rocks
>2.05 Ga komatiitic to mafic volcanic
rocks, shallow marine sedimentary rocks
>2.2 Ga terrestrial to shallow marine
sedimentary rocks, mafic to intermediate
volcanic rocks
tholeiitic and komatiitic
volcanic rocks
2.44 Ga felsic to intermediate
volcanic rocks
undefined quartzite, mica gneiss
and felsic to intermediate volcanic
rocks, possibly Archean in age
felsic intrusions
mafic to felsic intrusions
mafic layered intrusions,
mafic sills and dykes
tectonic contact with
ophiolite fragments
2.1 igneous age in Ga
Lithostratigraphical
units
Yllästunturi
quartzite
Luosujoki
conglomerate
Kolari
greenstone
formation
Rautuvaara
formation
Niesakero-
Kuertunturi
quartzite
complex
thickness
(m)
?-1000
-800?
50-150
-500
50-150
-500
Quartzite, arkosite
Conglomerate
Mica gneiss,
quartz-feldspar gneiss
Amphibolite
Graphite-bearing
schist
Skarn and magnetite
Carbonate rocks
Albite schist
Central Lapland Greenstone Belt
Kolari
Fig. 33. Stratigraphy of the Central Lapland Greenstone Belt (modied aer Lehtonen et al. 1998, Hanski et al. 2001, and GTK’s
FINSTRATI database) and the Kolari region (redrawn aer Hiltunen 1982).
schists and carbonate rocks (Fig. 33). e Rautu-
vaara Formation is separated into its own forma-
tion based on the occurrence of skarn rocks and
magnetite ore in it (Hiltunen 1982). In geophysi-
cal maps, it can be followed as magnetic highs and
locations where iron ± Cu-Au deposits are known
(Fig. 34). e uppermost units in the Kolari area
consist of Luosujoki conglomerate and Yllästun-
turi quartzite units, which correlate with the <1.89
Ga Kumpu Group (Fig. 33). e stratigraphical
scheme of Hiltunen (1982) has changed very little
during the past 30 years, with only minor changes
in the nomenclature and separation of the Kumpu
Group into four separate formations (Fig. 32).
Recent unpublished GTK age data from western
Lapland suggest the presence of a previously unre-
corded unit of clastic sedimentary rocks that have a
maximum depositional age of ca. 1.91 Ga, which is
slightly older than the Kumpu Group age of <1.89
Ga. One of the dated samples of the 1.91 Ga unit
is from the sillimanite quartzite taken from ca. 6
km E of the Rautuvaara Formation, representing
the Niesakero Formation (Niesakero-Kuertunturi
quartzite complex in Fig. 33). ese results indi-
cate that the correlation of the Niesakero Forma-
tion with the Sodankylä Group is most likely cor-
rect, and that the quartzites in the eastern part of
the study area overlie the Rautuvaara and Kolari
Formation rocks and are overlain by the Kumpu
Group rocks (Fig. 32).
Structural features
Hiltunen (1982) described the structural pattern
in the Kolari region as a series of synclines and
anticlines with axial trace striking NE to N, with
discontinuity at the Äkäsjoki shear zone (ÄSZ)
dividing the area into two tectonic domains (Fig.
32). In the northern domain, the layering dips
gently to the west, whereas in the southern domain
the dips are steeper (e.g. Hiltunen & Tontti 1976,
Hiltunen 1982). A characteristic feature of the Ko-
lari area is SW-plunging folding with the fold axis
plunging 20–55 degrees to the SW (Hiltunen 1982,
Väisänen 2002). A pronounced stretching line-
ation has been recorded in the district, having a
similar SW-plunging orientation to the F2 fold axis.
45
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
Northeast-, east-, and southwest-vergent folding
has been reported in the broader Kolari area, the
rst of which is probably the youngest generation
and relates to the regional D3 stage (Hiltunen 1982,
Väisänen 2002, Hölttä et al. 2007).
Several shear and thrust zones have been out-
lined in the Kolari region (Fig. 32). ese comprise
the crustal-scale Baltic-Bothnian megashear zone
(BBMZ) in the Kolari area, sometimes referred
to as the Pajala Shear Zone, or Kolari Shear Zone
(Berthelsen & Marker 1986). is N–S-striking
shear zone system reaches from Kalix in Sweden
in the south through western Finnish Lapland up
to the Norwegian Caledonides in the north, being
about 400 km in length and up to 75 km in width.
Lahtinen et al. (2005) proposed that the BBMZ
represents a cratonic boundary between the Norr-
botten craton in the west and the Karelian craton
in the east.
Fe ± Cu-Au deposits
About 15 Fe ± Cu-Au deposits are known from the
Kolari district. e deposits are located near to the
contact zone between the 1.86 Ga Haparanda Suite
monzonite and diorite intrusions and the Savuko-
ski and Kumpu Group supracrustal rocks, being
structurally controlled by thrust and shear zones
(Hiltunen 1982, Niiranen et al. 2007). e depos-
its consist of lenticular to tabular disseminated to
semi-massive magnetite bodies that host the Cu-
Au mineralization (Hiltunen 1982, Niiranen et al.
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
!(
3365000 3370000 3375000 3380000
7486000 7492000 7498000 7504000 7510000
Legend
!
(
Iron(-Cu-Au) deposit
HIRE profiles
Unspecified shear zone
((
Thrust zone
Rautuvaara
Hannukainen
Kuervitikko
Äkäsjoki Shear Zone
0 3 61.5
km
E1
E2
V1
V1
V2
V2
V3
V3
V4
V5
V6
3D model extent
Rautuvaara Formation
DGRF-65 anomaly [nT]
High : 17859
Low : -8950
Fig. 34. Aeromagnetic map of the study area with shear zones, seismic proles, and known deposits. e Rautuvaara Formation
is as in Figure 1.
46
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
2007). e typical metal association in the Kolari
deposits is Fe-Cu-Au, with locally elevated Co and
LREE. e size of the deposits ranges from 110 Mt
to about 1 Mt (Hiltunen 1982, Risto et al. 2010).
Multi-stage and -style alteration is associated with
the deposits. Early, pervasive albitization occurs
around all deposits, and this is overprinted by
calc-silicate (skarn) alteration that envelopes most
of magnetite bodies (Hiltunen 1982, Niiranen et
al. 2007).
THE DATA AND METHODS
e modeling utilized GTK’s aerogeophysical,
digital bedrock (DigiKP), bedrock observation,
and drill core databases. Old Rautaruukki maps
and cross sections were used, in addition to the
previously mentioned data sets. Much of the
modeling relied on the reection seismic proles
generated in the High Resolution Reection Seis-
mics for Ore Exploration 2007–2010 (HIRE) pro-
gram (e.g. Kukkonen et al. 2011). e HIRE data
on Kolari consist of a total of 80 km of lines in
8 proles (Fig. 32) (see Kukkonen et al. 2011, for
further details of the seismic data).
e visualization, interpretation, and modeling
were carried out using GocadTM soware with
Mira Mining utilitiesTM and GRGPack research
plugins provided by the Gocad Research Group
of the Nancy School of Geology. e key horizons
were interpreted on the geological and geophysical
maps, cross sections, and seismic proles in Gocad
as lines and point sets. e surfaces were interpo-
lated from interpreted lines and point sets using
the GRGPack StructuralLab plugin using methods
described by Caumon et al. (2007). A 3D block
model was generated from the surfaces using
Gocad inbuilt tools.
INTERPRETATION OF THE SELECTED HORIZONS AND 3D MODELING
e seismic section E1 at Hannukainen (Fig.
35) shows three highly reective, gently dipping
units stacked on top of each other. e thickness
of these units varies between 200 and 250 m. e
uppermost of the highly reective units can be
correlated with the mac volcanic rock that hosts
part of the ore at Hannukainen and correlates with
the Rautuvaara Formation. Comparison with the
E1 prole and cross section of the Hannukainen
deposit demonstrates that the Laurinoja ore body
also shows up as a reector on top of the Rautu-
vaara Formation reectors in the seismic data (Fig.
36). e second, lower highly reective unit (B in
Fig. 35) correlates at its upper part with the mica
schist and quartzite unit below the Rautuvaara
Formation at Hannukainen. Drilling has not, how-
ever, extended through this unit, and the composi-
tion of the lower part is therefore unknown. ere
are no drilling or outcrop data that would indicate
the lithology of the lowermost C unit.
e stacked highly reective units can be traced
from section to section, thus making them ideal
key horizons for 3D modeling. Of the key struc-
tures, only the ÄSZ has been modeled. It appears
as a set of transparent zones that cross-cut the re-
ectors, a feature that is clearest in seismic sections
V1 and V5, which cross-cut the shear zone at steep
angles (Fig. 37). e gently dipping to horizontal
A, B, and C units appear to steepen on both sides of
the shear zone, although no dramatic vertical shi
in these units can be seen. e ÄSZ, albeit clearly a
zone of fractures with an approximate thickness of
a few hundred meters, is modeled as a plane.
e modeled three highly reective units (A,
B, C in Fig. 35) form a broad open fold plung-
ing approximately 30 degrees to the SW, the ÄSZ
being located in the axial plane of the fold (Fig.
38). e internal form of the fold structure shows
small-scale dome and basin structures, suggest-
ing buckling and bulging of the A, B, and C units
during the folding events. e projected surface
geology of the horizons A and B correlates well
with the geological maps for the north of the
Äkäsjoki shear zone. e A unit correlates with
the Rautuvaara Formation mac volcanic rock
unit and B with the Sodankylä Group quartzites.
e quartzite-mica schist sequence south of the
ÄSZ, interpreted in geological maps to represent
47
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
WE
1000 m
2000 m
0 m
A
B
C
D
Fig. 35. Seismic E1 prole from Hannukainen with interpreted units (dashed green lines). See Figures 32 and 34 for the loca-
tion of the prole.
Fig. 36. Comparison of the geological cross section (Hiltunen 1982) and the E1 seismic prole spatially coinciding with the
geological prole. e western part of the Laurinoja semi-massive magnetite ore shows up as reectors, as well as the highly re-
ective mac volcanic rock footwall unit. Legend for the geological prole: 1. monzonite, 2. diorite, 3. skarn rock, 4. magnetite
ore, 5. quartz-feldspar schist, 6. mac volcanic rock, 7. quartzite, 8. mica gneiss, 9. ground surface, 10. drill hole.
Sodankylä Group rocks, correlates with a 300- to
500-m-thick, weakly reective topmost sequence
on the eastern segment of seismic prole V1. It
appears to be discordant with the modeled A
and B units, suggesting that contrary to the cur-
rent geological interpretation, the quartzite-mica
schist unit is younger than the Sodankylä Group
sedimentary rocks. is interpretation is further
supported by new age data for the sedimentary
rocks (see above). e aeromagnetic map of the
area (Fig. 34) shows that there appear to be two
dierent domains within the Sodankylä quartzite
unit in the geological map. e domain north of
the ÄSZ appears relatively monotonous and mod-
erately magnetic, whereas the domain south of the
ÄSZ is considerably more complex, consisting of
alternating magnetic lows and highs. Based on
seismic data and the geophysical characteristics
of the two domains, the quartzite-mica schist unit
south of the ÄSZ is modeled as a Svecofennian
unit discordantly overlying the modeled A, B, and
C units, and the northern domain is considered to
represent the Sodankylä Group rocks.
e seismic reectors become discontinuous,
weak, and dicult to interpret in the southeastern
corner of the modeled area and in the Rautuvaara
48
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
a.
b.
Fig. 37. A. Modeled three highly reective units with seismic prole V2 and the aeromagnetic map. e uppermost green unit
corresponds to the Savukoski Group mac volcanic rock at footwall at Hannukainen, the yellow unit to the quartzite below it.
Oblique view from west. B. Modeled uppermost highly reective unit and Äkäsjoki shear zone as the plane (yellow). Note the
folded geometry and the small-scale bulging and buckling of the mac volcanic rock unit. e Äkäsjoki shear zone appears to
be developed in the axial plane of the fold structure.
49
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
area. is may be due to steepening of the struc-
tures in this part, or the abundant monzonite intru-
sions that cut the units and cause the discontinuity,
or both. erefore, we modeled the intrusions and
the small mac volcanic rock units in this area as
a single block referred to as the Rautuvaara com-
plex. Other units modeled include the ‘basement’
on which the A, B, and C units overlie, and the low
reectivity unit north of the ÄSZ correlating with
the monzonite, diorite, and granite intrusions west
of Hannukainen. e complete model is presented
layer by layer in Figure 39.
E1
V1
V1
ÄSZ
A bottom
Unconf
Fig. 38. e Äkäsjoki shear zone, which shows up as a set of subvertical transparent zones (blue lines) cross-cutting the sub-
horizontal reectors on seismic proles. e transparent green plane is the modeled bottom of the A horizon of Figure 37.
e brown line indicates the unconformity on the Karelian units (see text). e gray-scale aeromagnetic map is presented as
a reference on the surface.
50
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
a. b.
c. d.
e. f.
g.
Fig. 39. e modeled units: a. “basement”; b. lowermost of
the three stacked highly reective units, probably sedimen-
tary rocks of the Sodankylä Group; c. middle of the highly re-
ective units, probably quartzite and mica gneiss/schist of the
Sodankylä Group; d. uppermost highly reective units cor-
responding to the Savukoski Group mac metavolcanic rocks
at Hannukainen; e. the <1.91 Ga quartzite-mica schist unit
overlying the older rocks unconformably; f. Haparanda suite
and granitoid intrusives with low reectivity in the seismic
data; g. Rautuvaara complex, consisting of Haparanda suite
intrusives and Savukoski Group supracrustal rocks. Oblique
view from the SE in all images. See the text for further details
of the units.
51
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
DISCUSSION AND CONCLUSIONS
e uppermost highly reective unit that corre-
lates with the Savukoski Group mac volcanic rock
at Hannukainen forms a broad open fold structure
plunging ca. 25 degrees to the SW, which is con-
sistent with the observed folding data from the
outcrops. In the Rytijänkä and Rautuoja areas, the
top of this unit is respectively at 1500 m and 3000
m below the current bedrock surface. Although
the model does not extend to the Rautuvaara area,
it is clear that the same unit that partly hosts the
deposits north of the ÄSZ cannot be exposed at
the surface in the Rautuvaara area, and that the
present interpretation of the Rautuvaara Forma-
tion in current maps is not correct.
e proposed strata-bound nature of the known
deposits has been somewhat dubious, as some of
the deposits are partially hosted by the Haparan-
da-type diorite intrusions and younger Kumpu
Group rocks (e.g. Hiltunen 1982, Niiranen et al.
2007). e seismic data show that even the depos-
its that in current 2D geological maps are hosted
by the so-called Rautuvaara Formation cannot be
within the same stratigraphic level. us, the stra-
ta-bound model for the deposits is very unlikely,
as already suggested by Niiranen et al. (2007). e
division of the Savukoski Group at Kolari into the
Rautuvaara and Kolari formations in previous in-
terpretations was mainly based on the presence of
skarns and iron deposits in the former, the lithol-
ogy being otherwise the same. e seismic data
do not support the presence of a uniform strati-
graphic layer that could be connected through the
area. erefore, there remains no justication for
separating the Savukoski Group into two separate
formations in the Kolari area.
e highly reective unit below the modeled
Savukoski Group mac volcanic unit is somewhat
problematic. Based on the drilling data, the upper
part of this consist of quartzites and mica gneiss
and the unit outcrops at the location where the bed-
rock has been interpreted to represent Sodankylä
Group rocks in current bedrock maps. However,
the new age data combined with the seismic data
indicate that the Niesakero-Kuertunturi quartzite
complex south of the ÄSZ is an early Svecofen-
nian unit overlying the older supracrustal rocks.
e possible scenarios for nature of the quartzite
north of the ÄSZ are: a) it belongs to the Sodanky-
lä Group as it is mapped and shown in our model,
b) it belongs to the same Svecofennian unit as the
rocks south of the ÄSZ, and the Savukoski Group
rocks in the Hannukainen area and parts north of
it have been thrusted on top of the quartzite, or
c) the mac volcanic rocks that host the deposits
at Hannukainen are also Svecofennian in age. We
consider option c to be very unlikely. e dier-
ence between the two quarzite-mica-schist do-
mains in aeromagnetic maps suggests that the rst
option would be the most viable, although fur-
ther age dating would be required to verify this.
e lowermost highly reective unit C cannot be
correlated with any known exposed rocks. If the
B unit belongs to the Sodankylä Group, it is very
likely that the C unit represents the lower part of
the same group.
e results of this work provide some interesting
insights into the tectonic evolution of the area. e
broad open fold shape of the modeled Savukoski
Group and orientation of the fold axis coincides
with the previously measured fold axis. Hiltunen
and Tontti (1976) suggested that this was the earli-
est folding event in the area, and that it predates or
is synchronous with the intrusion of the 1.89–1.86
Ga Haparanda suite rocks. e fold axis orienta-
tion suggests that the dominant compression was
NW–SE-oriented at this stage. Väisänen (2002)
presented evidence of NE-vergent thrusting from
the SW part of the Kolari area, and the very prom-
inent SW-plunging stretching lineation supports
this. e small-scale dome and basin structures
observed in the modeled Savukoski and Sodanky-
lä Group units may result from refolding during
the D3 stage thrusting. However, this is somewhat
uncertain, as the doming may already have taken
place during the D2 folding event.
e modeled Äkäsjokisuu shear zone appears
to have been developed into the axial plane of the
broad fold structure, supporting the earlier inter-
pretations of Hiltunen and Tontti (1976) for the
generation of the NE-striking shear zones in the
area. is suggests that there was a shearing com-
ponent during the early folding event or that the
NE-striking shear zones were developed during
the following NE vergent thrusting.
e unconformity between the modeled <1.91
Ga quartzite-mica schist unit and the lower su-
pracrustal units suggest that there was a hiatus
between the two supracrustal sequences. e out-
crop data and the geophysical map of the <1.91 Ga
quartzite-mica schist unit indicate that it is folded,
52
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
and that the folding appears to have taken place in
the same stage as when the Savukoski Group rocks
were folded. Nevertheless, the Karelian rocks ap-
pear to have been tilted prior to the deposition of
the younger sedimentary rocks on them.
53
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
CHAPTER V.
THE LAPLAND GRANULITE BELT 3D MODEL
Tero Niiranen, Vesa Nykänen and Ilkka Lahti
INTRODUCTION
e Central Lapland greenstone belt (CLGB) is
bound at its NE end to a tectonized zone consist-
ing of the Hetta and Vuotso complexes intermin-
gled with Karelian Salla, Sodankylä, Savukoski,
and Kittilä group rocks (Fig. 40). e Lapland
Granulite Belt (LGB) comprises the northeast-
ernmost unit of the tectonized zone. e LGB has
been interpreted by several authors as a SW-ver-
gent thrust zone related to the continent–conti-
nent collision of the Kola craton in the NE and the
Karelian craton in the SW (e.g. Marker 1988, Gaál
et al. 1989, Korja et al. 1996, Tuisku et al. 2006).
Many of the main tectonic features of the CLGB
have been interpreted to be related to compression
from the NE and linked to LGB thrusting, making
the area one of the key locations in understanding
the structural evolution of the CLGB (e.g. Ward et
al. 1989, Väisänen 2002, Hölttä et al. 2007, Pati-
son 2007). Several 2D interpretations of the area
have been presented based on the magnetotel-
luric, gravity and seismic POLAR and FIRE data
(e.g. Elo et al. 1989, Gaál et al. 1989, Korja et al.
1989, Patison et al. 2006). is chapter presents a
3D visualization of the part of the LGB closest to
the CLGB.
GENERAL GEOLOGY
e LGB has been divided into the Kuttura and
Saariselkä suites (Fig. 40). e Saariselkä suite
consists of garnet-sillimanite gneisses, originat-
ing from pelitic to psammitic sediments. e
Kuttura suite consists of concordant sheets of an-
ortosites and arc-type noritic to enderbitic intru-
sions. e detrital zircons of the Saariselkä suite
sedimentary rocks fall within the range of 1.94–
2.9 Ga, clustering around 1.97 and 2.2 Ga, and
the age data for intrusions of the Kuttura suite
norite-enderbite series indicate that they were
intruded into the Saariselkä suite sediments at
ca. 1920–1905 Ma (Tuisku & Huhma 2006). e
bedrock adjacent to the southwest of the granu-
lite consists of Vuotso complex arcositic gneisses
and amphibolites and mac to ultramac intru-
sions intermingled with ri-related 2.44–2.0 Ga
Salla, Sodankylä, and Savukoski group suprac-
rustal rocks. Further southwest to these are the
Hetta complex granitoids and mac metavolcan-
ic rocks of the 2.0 Ga Kittilä Group, the latter of
which are referred here as the Kittilä Terrane (e.g.
Lehtonen et al. 1998). e ca. 1.77 Ga Nattanen
granite intrusions and Salla and Laanila dykes
cross-cut these rocks and comprise the youngest
rocks in the area.
e LGB and Vuotso complex rocks comprise
the highest metamorphic grade rocks of northern
Finland. e data for the LGB sedimentary rocks
indicate a clockwise metamorphic path, with peak
pT conditions of 5.5–8.0 kbars and 800–850 oC.
e peak metamorphic conditions for the Vuotso
complex rocks, at the SW margin, reach ca. 12
kbars and 650oC (Tuisku et al. 2006). e meta-
morphic conditions decrease to the SW, being
7.4–9.9 kbars and 600–690oC around the Hetta
complex and the NE corner of the Kittilä Terrane,
and decreasing to 3.2 kbars and 350–400 oC for
54
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
3450000 3475000 3500000 3525000
7575000 7600000 7625000
Laanila and Salla dykes
Nattanen granite suite
Nuttio suite
Diverse lithodemes
Kittilä group, Pyhäjärvi formation
Kittilä group, Uurrekarkia formation
Kittilä group, Porkonen formation
Kittilä group, Kautoselkä formation
Savukoski group, Matarakoski formation
Savukoski group, Sattasvaara formation
Savukoski group, Peurasuvanto formation
Savukoski group, Peuralampi formation
Sodankylä group, Postojoki formation
Salla group, Kuortisoja formation
Lapland granulite complex, Kuttura suite
Lapland granulite complex, Saariselkä suite
Kaamanen complex
Vuotso complex
Vuotso complex, Kussuolinkivaara suite
Hetta complex
Pomokaira complex, Diverse lithodemes
Unspecified major fault/shear zone
Major thrust fault
Unspecified minor fault/shear zone
Minor thrust fault
Fracture
FIRE4A profile
3D model boundary
Fig. 40. Bedrock map of the study area with the location of the 3D model boundary and the FIRE4A seismic prole. Aer
Bedrock of Finland − DigiKP.
55
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
the central and eastern parts of the Kittilä terrane
(Hölttä et al. 2007).
Based on the metamorphic and age data for the
LGB, Tuisku et al. (2006) and Tuisku and Huhma
(2006) divided the evolution of the LGB into sev-
eral stages: burial and heating of the metasedi-
ments aer ca. 1.94 Ga, high-T metamorphism
and migmatization at ca 1.91 Ga, accompanied
by emplacement of norite-enderbitic intrusions at
ca. 1.92 Ga, thrusting, erosion, and cooling of the
whole LGB sequence from ca. 1.91–1.88 Ga, and
nal crystallization of leucosomes and exhuma-
tion to upper crust in 1.89–1.88 Ga. Tuisku et al.
(2006) considered high-strain rocks of the Vuot-
so complex as a foreland slab related to the LGB
thrust.
MATERIAL AND METHODS
e study was based on the airborne geophysical,
bedrock observation and digital bedrock databas-
es of GTK. In addition, seismic data from the Fin-
ish Reection Experiment 2001–2005 (FIRE, see
Kukkonen & Lahtinen 2006) were used in mode-
ling. e modeling was carried out by interpreting
the lithological units on geological and geophysi-
cal maps as well as on the seismic FIRE 4A prole.
e visualization, interpretation and modeling
were performed using GocadTM soware with
Mira Mining utilitiesTM and GRGPack research
plugins provided by the Gocad Research Group of
the Nancy School of Geology. e surfaces were
interpolated from these using the GRGPack Struc-
turalLab plugin following methods described by
Caumon et al. (2007). A 3D block model was gen-
erated from the surfaces using Gocad inbuilt tools.
INTERPRETATION OF THE KEY HORIZONS
In airborne magnetic maps, the LGB shows up
as variable thick bands of magnetic low and high
zones with a general strike of NW–SE (Fig. 41).
e bands are cut and locally displaced by NE–
SW-striking magnetic low lines interpreted as fault
or shear zones in geological maps (e.g. Fig. 40).
e area SW of the LGB representing the Vuotso
complex rocks has similar characteristics in air-
borne magnetic maps, although the banding is less
clear. e LGB and Vuotso complex rocks appear
as highly reective units in the seismic section,
with the reectors having a gently NE-dipping lis-
tric pattern (Fig. 42a). e reectors correlating
with the LGB and Vuotso complex rocks atten
out at depth, and based on the seismic data, the
interpreted base of the southwesternmost Vuotso
complex rocks reach a depth of ca. 12.5 km. e
seismic pattern of these units correlates with the
interpreted SW-directed thrust. However, the lis-
tric thrust pattern extends further SW on the seis-
mic prole, the southwesternmost listrict reec-
tors reaching the surface at the northeasternmost
margin of the Hetta complex (Figs. 40 and 42a).
Based on seismic, aeromagnetic, and bedrock
data, a total of 9 tectonic blocks were distinguished
(Figs. 42 and 43). e LGB in the modeled area
is divided into four tectonic blocks separated by
discontinuities in the seismic data, which appear
as tectonic features (thrust planes). For the Vuotso
complex, three tectonic blocks were distinguished
on a similar basis. In addition, one small tectonic
slice was outlined, correlating with Vuotso com-
plex rocks on the surface, and the last unit in the
seismic prole consists of Kittilä terrane material
(Figs. 42 and 43).
e apparent dips of the seismic reectors of
the thrust zones closest to the surface level vary
between 18 and 32 degrees to the NE along the
thrust zone. As the seismic prole is almost per-
pendicular to the lithological banding, the dips
on the seismic data are relatively close to the true
dips. is is further supported by the bedrock ob-
servation data. GTK’s databases include a total of
712 foliation measurements from the LGB and
Vuotso complex rocks within the modeled area
and its close proximity. We divided these meas-
urements into three groups based on their spatial
distribution: measurements in the NW, central,
and E part of the model (Fig. 44). e average dip
orientations of these groups are in agreement with
56
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
3450000 3475000 3500000 3525000
7575000 7600000 7625000
FIRE4A profile
3D model boundary
High : 17859
Low : -8950
DGRF-65 anomaly (nT)
Fig. 41. Airborne magnetic map of the study area with the location of the 3D model boundary and the FIRE4A seismic prole.
GTK data.
apparent dip data measured from the seismic pro-
le. is also shows that the dips appear to steepen
when moving from the NE to the E along the LGR
and Vuotso complex units, a feature that has also
previously been reported (e.g. Tuisku et al. 2006).
is at least partly explains the apparent thicken-
ing of the lithological units moving to the NE.
57
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
a.
b.
2. 3. 4. 5. 6. 7.
8. 9.
Fig. 42. a. Airborne magnetic map and the FIRE4A seismic prole. e bottom of the seismic prole is at 20 km depth. Oblique
view from the NW. b. e tectonic blocks, interpreted based on the seismic data. e same view as in a. See Figure 43 and the
text for an explanation of the numbering.
58
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
3450000 3475000 3500000 3525000
7575000 7600000 7625000
High : 17859
Low : -8950
DGRF-65 anomaly (nT) 3D model boundary
FIRE4A profile
Modeled faults/shear zones
Modeled unit contacts
1.2.3.4.
5.
6.
7.
8.
9.
Fig. 43. Interpreted tectonic block boundaries (blue) and modeled faults (green) on the airborne magnetic map. Units 1–4
comprise the LGB blocks, 5–7 the Vuotso complex blocks, 8 Hetta complex rocks, and 9 Kittilä terrane rocks (cf. Figure 1).
59
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
e 3D model
Based on the interpreted boundaries presented in
Figures 42 and 43, and the foliation measurement
data, the thrust planes limiting units 1–6 were
constructed as surfaces, and from the surfaces
the units representing the four LGB tectonic and
2 Vuotso complex blocks were built as 3D blocks
(Fig. 45). e two most signicant NE–SW-strik-
ing faults were constructed in the model as planes.
However, as there are no direct dip data, the faults
were assumed to be vertical. e modeled units are
presented in Figure 46. e vertical extent of the
model is 10 km.
Based on the modeled units, all tectonic blocks
are relatively uniformly thick along their horizon-
tal axis, the apparent thickness variation observ-
able on their surface expressions relating to the
variation in dips. e true thickness of units 2–4
varies between 2.5 and 3 km. ey thicken towards
n = 442
Aver. 033/30
n = 163
Aver. 027/33
n = 107
Aver. 039/37
Fig. 44. Location of the foliation measurement data and the average foliation orientations in dierent parts of the study area.
Dip orientation vectors as points on the lower hemisphere, equal area projection.
60
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
a. b.
c. d.
e. f.
g.
Fig. 45. 3D model of the study area. a.–d. are the LGB tectonic blocks (1–4 in Fig. 43), e.–f. the Vuotso complex tectonic blocks
(5–6 in Fig. 43). e two planes are modeled NE–SW-striking faults. Oblique view from the N. g. All modeled tectonic blocks
in an “exploded” view. Oblique view from the SW.
61
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
a.
b.
Fig. 46. e variations in the dip of the base of the modeled tectonic blocks. a. Block 6; b. block 1. Histograms of the dips and
strikes for each node in the corresponding planes.
62
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
the top at the locations where their dips steepen.
A similar pattern exists with units 5 and 6, repre-
senting Vuotso complex rocks. However, there is
also thickness variation along the horizontal axis
in these units. e spatial variation in dips of the
base of units 1 and 6 is presented in Figure 46. is
illustrates the attening of the modeled units at
depth, the steepest parts generally being close to
the surface.
DISCUSSION
e constructed 3D model shows classical features
of a thrust front with listric thrust planes, being
in good agreement with the previous studies on
the LGB. e NE–SW-striking shear or fault zones
probably represent tear or transfer faults relating to
the thrust planes. Clearly, the thrust front extends
a considerable distance to the SW from the LGB
rocks, including all of the Vuotso complex rocks,
and the eects of the SW-directed thrusting events
extend further SW to the Kittilä terrane rocks.
Based on the metamorphic data, Tuisku et al.
(2006) and Tuisku & Huhma (2006) estimated that
the high-pressure assemblages of the LGB were
formed at depths of ca. 28 km, and the melting and
crystallization of the leucosomes took place at ca. 7
km depth some 30 Ma later. e highest peak met-
amorphic pressure estimates come from the Vaulo
area, which corresponds to the modeled block 5.
For these, the metamorphic assemblages indicate
pressures of ca. 11 kbar, corresponding to a depth
of ca. 40 km (Tuisku et al. 2006). Tuisku et al.
(2006) presented a model for LGB metamorphism
in which the sedimentary rocks were deposited
into a trench environment and were subsequently
obducted by a subducting slab into the mantle,
where they were subjected to high-pressure high-
temperature metamorphism. e subsequent low-
pressure assemblages were formed via post-peak
metamorphic thrusting and exhumation of the
LGB rocks.
Based on the seismic data and the 3D model
presented here, the thrusting of ca. 40 km of the
LGB rocks in a horizontal direction would have re-
sulted in a ca. 10 km change in depth if the shape
of the LGB had been the same as at present. is
suggests that an approximately 18-km-thick se-
quence of rocks had to be exhumed from the top
of the LGB rocks, and ca. 30 km from the Vuotso
complex rocks. ese gures are clearly too high
to be explained by simple erosion and upli, and
the tectonic transport during the later stages of
evolution has clearly accounted for much more
than the 10 km upli. e dramatic peak pres-
sure dierences between the Vuotso complex and
the LGB data indicate that the depth dierence of
these units was more than 10 km during the for-
mation of the peak pressure assemblages. Hence,
there were probably considerable dierences in
transport distances between the dierent tectonic
blocks within the thrust belt.
63
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
CHAPTER VI.
THE SAATTOPORA AuCu DEPOSIT 3D MODEL
Tero Niiranen, Vesa Nykänen and Ilkka Lahti
INTRODUCTION
e Saattopora Au(-Cu) deposit was discovered
in 1985 by Outokumpu Oyj and was subsequently
mined during 1988–1995 from two open pits and
underground workings. e total amount of gold
produced was 6279 kg, and 5177 tonnes of cop-
per was produced as a by-product. e deposit is
located in the municipality of Kittilä, in Northern
Finland. e Saattopora is one of several gold de-
posits discovered along the E–W-trending Sirkka
thrust, although it is the only one that has been
under full-scale mining. In 2008, Outokumpu Oyj
delivered much of the existing data from the Saat-
topora deposit to GTK. is was the early stage of
the 3D modeling project on the Central Lapland
Greenstone Belt, and it was soon decided to use
the data to construct a deposit-scale 3D model of
the Saattopora. e aims were to test the suitability
of Gocad soware in deposit modeling, to evaluate
the possible continuation of the mineralization to
depth, and to assess whether the modeling would
provide new insights into the geological setting of
the deposit, including its relation to the regional
deformation.
GEOLOGICAL BACKGROUND
e Saattopora deposit is located in the western
part Central Lapland Greenstone Belt (CLGB),
next to the major, crustal-scale Sirkka thrust zone
(Fig. 47). e deposit consists of two roughly E–
W-trending lodes that are hosted by variably al-
tered mica schists, phyllites, komatiites, mac tus
and mac lavas of the >2.2 Ga Savukoski Group
(Fig. 48). e intrusives in the area consist of 2.2–
2.0 Ga diabase dykes and 2.0 Ga felsic porphyry
dykes, the former of which occur in the ore host-
ing sequence. e mineralization consists of N–S-
striking, sub-vertical to vertical quartz-carbonate-
sulde-gold veins, vein arrays, and hydrothermal
breccias (Fig. 49). e total size of the deposit is
2.163 Mt and the average grades are 3.29 g/t Au
and 0.28% Cu (Lahtinen et al. 2005).
e host rock sequence was subjected to poly-
phase deformation and metamorphism during the
Svecofennian orogenic events in 1.91–1.79 Ga. e
earliest ductile deformation stages, D1-2, relate to
thrusting from the south and northeast, and the
last D3 stage in the western part of the CLGB re-
lates to thrusting from the west or southwest (e.g.
Ward et al. 1989, Hölttä et al. 2007, Patison 2007).
e regional metamorphic grade at Saattopora is
of mid-greenschist facies and the peak conditions
were reached during the D1-2 stage (Hölttä et al.
2007). e Saattopora deposit is structurally con-
trolled by second or lower order structures relat-
ing to the Sirkka thrust, which was initially formed
during the northward-directed thrusting in the
D1-2 stage and subsequently re-activated during D3
as a strike-slip shear (Patison 2007, Saalmann &
Niiranen 2010).
Multi-stage and -style alteration has been de-
tected at Saattopora. e earliest alteration stage is
a regional-scale albitization that was folded in D3
and possibly already during D1-2 (e.g. Patison 2007,
Saalmann & Niiranen 2010, Niiranen et al. 2012).
At Saattopora, the albitized rocks are overprinted
64
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
((
((
((
(( ((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
(( ((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
(( ((
((
((
((
((
((
((
((
((
((
((
((
((
((
((
3380000 3385000 3390000 3395000 3400000
7520000 7525000 7530000
Ü
(( ((
((
(( ((
(( (( (( ((
((
(( ((
((
(( ((
(( ((
(( (( ((
(( ((
(( (( ((
(( ((
(( ((
(( ((
(( ((
((
(( ((
((
(( ((
((
(( ((
(( (( (( ((
((
(( ((
((
(( ((
(( ((
(( (( (( (( (( (( ((
(( (( (( ((
3390000 3391000 3392000 3393000
7524000 7525000 7526000
Tholeiitic basalt
Tholeiitic basalt
Mafic tuff, mafic lava
Graphitic tuffite
Komatiite
Conglomerate
Quartzite
Phyllite, mica schist, locally graphite-bearing
Kumpu Gp
Kittilä Gp
Savukoski Gp
(( ((
Major thrust fault
Ca. 1.80 Granite
1.89-1.86 Ga Monzonite
2.2-2.0 Ga Diabase
Unspecified minor shear zone
150 km
Saattopora
km
0 2.5 5
Gold deposit
3D model boundary
STZ
STZ
Fig. 47. New geological map of the Saattopora-Pahtavuoma region with the location of known orogenic gold deposits. Loca-
tion of the mapping area in the inset (upper part). Detailed view of the Saattopora area geology and location of the 3D model
extents. STZ = Sirkka thrust zone. Lithostratigraphical units aer Lehtonen et al. (1998). Contains data from the National Land
Survey of Finland Topographic Database 03/2013 © NLSand HALTIK.
65
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Central Lapland Greenstone Belt 3D modeling project nal report
7525300
3390500
7524600
3391900
A
A’
200 m
AA’
230 m
30 m
Mafic tuff, phyllite & mafic lava intercalations
Phyllite, mica schist intercalations
Komatiite
Tholeiitic lava
Shear zone
Au lodes (1 g/t cut off)
Fig. 48. Geology of the Saattopora deposit with 1 g/t Au envelopes. Redrawn aer the generated 3D model.
66
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014
Tero Niiranen, Ilkka Lahti, Vesa Nykänen and Tuomo Karinen
A
B
C
10 cm
A B
C
10 cm
A
B
C
10 cm
Fig. 49. A. Typical Au-bearing quartz-carbonate-sulde
veins at Saattopora. Completely brittle sub-vertical to verti-
cal N–S-striking veins cross-cutting albitized phyllite. ‘A’ pit.
B. Dierent alteration stages at Saattopora. Intense albitiza-
tion in phyllite, overprinted by early carbonate alteration
(brown material), which is cross-cut by Au-bearing veins.
Note that albitized phyllite is folded in a ductile manner,
and the early carbonation is either folded or follows the pre-
existing folded foliation. e mineralized veins cross-cut all
the early ductile deformation and early alteration in a com-
pletely brittle manner. e small-scale folding visible in the
photo relates to the D3 stage. C. Polymictic hydrothermal
breccia in the ‘B’ pit. Quartz, carbonates, suldes, and gold
brecciate the host rocks consisting of albitized phyllite, mica
schist, and mac volcanic rocks.
by carbonation, which was either folded in the D3
stage or follows the foliation folded in the D3 stage
(Fig. 49b). e mineralized veins crosscut these
early alteration stages, as well as all ductile defor-
mation features in the area. Other alteration styles
reported are chlorite and talc-chlorite alteration
detected in mac and ultramac rocks. However,
they may be a regional feature related to the de-
velopment of thrust and shear zones, as they have
been reported from well outside the mineralized
zone along the Sirkka thrust zone (e.g. Saalmann
& Niiranen 2010). Sericite alteration has been also
reported from the felsic host rocks, although it is
unclear how it is related to the mineralization.
Fluid inclusion data from the mineralized
quartz-carbonate veins indicate that the aqueous-
carbonic mineralizing uid was moderately saline
(ca. 9% NaCleq.) and the mineralization took place
at 300–350oC (Niiranen et al. 2012). Fluid inclu-
sion data also suggest that the mineralization was
most likely due to phase separation between the
aqueous and carbonic phases of the gold-carrying
uid, a feature that is further supported by the
presence of hydrothermal breccias (Fig. 49c).
e general features of the Saattopora deposit t
the categorization of an orogenic gold deposit with
atypical metal association outlined by Groves et al.
(1998).
DATA AND PROCESSING
In the modeling, all available regional geological,
geophysical, and digital elevation data of GTK
were utilized, including the bedrock observation
database and airborne geophysics. Ground geo-
physical data provided by Outokumpu, including
magnetic, VLF-R, and gravity data, were addition-
ally used (further details in Lahtinen et al. 2005).
Based on these data and new eld observations
67
Geologian tutkimuskeskus, Tutkimusraportti 209 – Geological Survey of Finland, Report of Investigation 209, 2014