Acta Geophys.

(2017) 65:1133–1144
https://doi.org/10.1007/s11600-017-0101-2

RESEARCH ARTICLE – APPLIED GEOPHYSICS

A conceptual model of the copper–porphyry ore formation based

on joint analysis of deep 3D geophysical models: Sorskoe complex
(Russia) case study
Viacheslav V. Spichak1 • Alexandra G. Goidina1

Received: 6 February 2017 / Accepted: 14 November 2017 / Published online: 18 November 2017
Ó Institute of Geophysics, Polish Academy of Sciences & Polish Academy of Sciences 2017

Abstract Joint analysis of deep three-dimensional models Introduction

of the electrical resistivity, seismic velocity, and density of
the complex hosting the Sorskoe Cu–Mo deposit (Russia) It is a well-known fact that most of polymetallic, gold-ore,
is carried out aimed at finding geophysical markers char- uranium, and crystal-bearing deposits have endogenous
acterizing the areas of ore generation, transportation and (hydrothermal) origin, which is driven by the internal
deposition. The three-dimensional lithology model of the energy of the Earth. Endogenic ore mineral deposits are
study area is built based on the empirical relationship most often extracted by open-pit mining technologies,
between the silica content of the rocks and seismic veloc- which currently provide more than 2/3 of the total volume
ities. It is in agreement with geological and geochemical of ore. Underground mining provides about 20% iron, up to
studies provided in this area earlier and could be used as a 45% copper, up to 70% zinc, etc. Due to the depletion of
basis for forecasting locations of the copper–molybdenum easily accessible shallow metallic ore deposits, the depth of
ore deposits at depth. A conceptual model of the copper– mining increases nowadays. For example, endogenic ore is
porphyry complex explaining the mechanisms of ore gen- mined from a depth of up to 2.6 km in Canada, up to 3 km
eration, transportation from the lower to the upper crust in the USA, up to 3.5 km in India, from more than 5 km in
and deposition in the upper crust is suggested. In particular, the Republic of South Africa, and this trend becomes
it is supposed that post-magmatic supercritical gas–water consistently more intense.
ore-bearing fluids are upwelling through the plastic crust The prospecting for shallow mineral ore deposits, which
due to the sliding of the fluid films along the cleavage are located at a depth of a few hundreds of meters, typically
planes of the foliated rocks while at the depths of the brittle includes geological surveying together with geochemical
upper crust this mechanism could be changed by volu- and shallow-depth geophysical methods [e.g., gravity and
metric fluid transportation along the network of large pores magnetic prospecting, natural and induced potential log-
and cracks. ging and transient electromagnetic techniques (Heinrich
et al. 1999)]. These methods provide the basis for typifying
Keywords Conceptual model  Copper–porphyry  ore-bearing intrusions, which play the key role in ore for-
Geophysical markers  Lithology model  Ore mation, and for delineating the potentially ore-bearing
transportation mechanism  Sorskoe deposit (Russia) deposits. At the same time, prospecting and exploration of
rather deep deposits (particularly, large ore clusters)
require the methods of large-depth geophysics including
seismic and magnetotelluric (MT) soundings [see, for
instance, (Tavakoli et al. 2016)]. In the search for
hydrothermal solid mineral deposits, magnetotelluric
& Viacheslav V. Spichak soundings are most sensitive to the ore and ore-hosting
[email protected]
formations since, on one hand, ore components are
1
Geoelectromagnetic Research Centre IPE RAS, GEMRC, deposited in most permeable ore-bearing rocks, which are
Troitsk Restrict, POB 30, Moscow 142190, Russia capable of adsorption, and, on the other hand, metasomatic

123
1134 Acta Geophys. (2017) 65:1133–1144

ores are typically represented by highly conductive uplifting and relative subsidence. The current stage of
sulfides. neotectonic activation manifests itself by amplification of
High sensitivity of the electromagnetic methods to the vertical motions in the uplifted and subsided areas and in
ore and water-saturated rocks opens the possibility to map the further differentiation of the structures, especially in
these rocks by audiomagnetotelluric sounding (Bauer et al. near-fault zones of junction of the ridges and basins. The
2003). Seismic methods using the converted waves from presence of the northeast- and northwest-striking faults of
the earthquakes (CWE) and deep seismic sounding (DSS) different orders is a characteristic feature of the region
techniques are suitable for identifying structural and for- (Kadurin et al. 2008).
mational positions of the ore-bearing regions within poly- The neotectonic plan of the region is dominated by
metallic deposits (Aleinikov et al. 1986). At the same time, alternation of subsided graben-like blocks and uplifted
contrasts in the electrical resistivity between the ore-bear- horst-like blocks with asymmetric wings and different
ing and host rocks in the porphyry media are rather low. amplitudes of the displacement. Velocities of vertical
Therefore, magnetotelluric methods, which have diffusion neotectonic motion of these units attain 0.3–0.4 mm/year
nature, are rarely used for prospecting, since their sensi- in the basins and 0.5–0.6 mm/year in the ridges, which is
tivity is in many cases insufficient for distinguishing such supported by the 500–600 m thickness of Neogene–Qua-
targets. Seismic methods, in turn, are not quite efficient in ternary sediments in the basins.
prospecting for ore bodies in the media having a complex The deep Sardino–Sorskii fault borders the region of
structure and inclined boundaries between layers (Heinrich study in the northeast and the Sayano–Batanevskii fault in
et al. 1999). The situation is aggravated by the fact that the southeast. Just as the lower-order faults of the Pre-
geophysical reconstructions in ore prospecting typically cambrian origin, these faults have an inherited character of
retrieve two-dimensional models, which do not enable evolution. They supposedly play the key role in the ore
reliable interpretation in terms of the locations and physical formation, since throughout their entire depth these faults
properties of the ore-bearing rocks. Therefore, the problem are framed by the fractures of varying orientation and by
of geophysical prospecting for ore deposits should be the vast zones of crushed rocks, which provides favorable
solved in the context of joint interpretation of the geo- conditions for the development of hydrothermal processes.
physical data, which can enhance reliability of the con- Besides mantle processes, the duration of the faults’ evo-
clusions. Besides, instead of two-dimensional models, lution was also affected by horizontal compression from
reconstructions should yield three-dimensional models, the south, which resulted in the significant displacements of
which are suitable for identifying the target geological crustal blocks along the faults by up to 70 km. Water
objects. mineralization is quite high within the basins, e.g., the
This study is aimed at finding geophysical characteris- potassium salt concentration in the brine attains 255 g/kg.
3
tics of the probable areas of generation, transportation and He concentration within the basins is 0.3–1.6%, which
deposition of the copper–porphyry ore deposits by joint indicates that the faults reach large depths up to the upper
analysis of deep 3-D geophysical models of the Sorskoe mantle (presumably more than 40 km).
copper–molybdenum complex located in Kuznetskii Alatau Regional MT soundings conducted in the Altai-Sayan
(Russia). With this purpose, we provide a joint analysis of folded belt by Kadurin et al. (2008) revealed a series of
three-dimensional models of electrical resistivity, seismic conductive structures with resistivity of up to 100 Xm,
velocities, density and lithology followed by building of which are confined to the large faults and basins. Most of
the conceptual model of the copper–porphyry ore these anomalies are correlated to the sources of the earth-
formation. quakes, maximal vertical displacements of the crustal
blocks, and the areas marked by the attenuation of seismic
waves. According to the DSS data, the Moho rises to a depth
Geological and geophysical setting of 45 km within the considered segment of the Altai-Sayan
region while the thickness of the upper, middle, and lower
The region of the study (Fig. 1) is located in the north- crust is approximately 15, 10, and 18–20 km, respectively.
western segment of the Altai-Sayan folded area, which is a Basins inherited from the earlier tectonic development
part of the orogenic belt of Inner Asia that separates the of the territory are marked by the minima in the distribu-
Siberian and Indian Platforms. Activation of the belt started tion of gravity field, which are probably due to the sig-
in the Mesozoic and intermittently continues up to the nificant thickness of sediments or to the reduced density
present. (deconsolidation) of the mantle and crust. The regional
The Early Mesozoic activation has enhanced the frag- fault zones are controlled by the maximal gradients of
mentation of the heterogeneous basement by large faults anomalies while granite intrusions are traced by low neg-
and predetermined the areas involved into the block-dome ative anomalies.

123
Acta Geophys. (2017) 65:1133–1144 1135

Fig. 1 Map of the geological formations and mineral deposits in the 10-coal-bearing paralic, 11-granite–syenite–granite, 12-trachybasalt–
Kuznetskii Alatau (Russia): 1-granite, 2-tonalite–plagiogranite–gran- trahyaterit–trachyliparite. The square bounds the Sorskoe copper–
odiorite, 3-trachybasalt, 4-crust of weathering formations (undiffer- molybdenum complex studied in this research; dots mark magne-
entiated), 5-lime, 6-carbonate–terrigenous–ash (tuffite), 7-silt– totelluric sounding sites; thin lines indicate faults
sandstone, 8-lime–sand–siltstone, 9-carbonate–terrigenous–ash,

The observed positive magnetic anomaly is caused Characteristics of the Sorskoe copper–
either by basic rocks that compose the crust or by the low molybdenum deposit
position of the Curie point before the Oligocene, since this
area is currently marked by the high heat flow and the Eastern offsets of the Kuznetskii Alatau hold a series of
Moho there is maximally close to the surface. The regional polymetallic ore fields and copper–molybdenum porphyry
faults are marked by negative magnetic anomalies up to – deposits confined to the Saralinskii and Uibatskii plutons of
20 nT, which are associated with acidic (granite) intrusions Cambrian age. The largest deposit (Sorskoe) is located in
in their bodies. the intersection of the northwestern and northeastern tec-
tonic zones (bounded by square in the Fig. 1).
The formation of the Sorskoe deposit, which is related to
the development of the ore-bearing porphyry complex, was
preceded by a long evolution of multi-pulse granitoid
magmatism originating from the common ore formation
area located at large depths (Sotnikov 2006). By a number
of its characteristics, the Sorskoe deposit corresponds to the

123
1136 Acta Geophys. (2017) 65:1133–1144

molybdenum–porphyry type described by the ‘‘granite locations in the bounded area of the Fig. 1) at 45 fre-
model’’, in which the ore mineralization is associated with quencies. In the context of this approach, both observations
the subalcalic magma series. Such deposits are marked by and model parameters (resistivities) were considered as
the predominance of the mantle–crustal interaction, random variables. Bayesian analysis determines the pos-
increased oxidation potential and high salinity of the ore- terior probability density function of the resistivity, i.e., the
forming fluids (Sokolova and Duchkov 2008). conditional probabilities of the resistivities given the data,
The ore field of the Sorskoe stockwork copper–molyb- prior information in terms of a resistivity palette and the
denum deposit is composed of the Lower Paleozoic gran- noise level.
itoids of the Uibatskii batholite which compositionally The solution of the inverse problem was reduced to the
range from gabbro–diorites to leucocratic granites (Berzina search for the posterior resistivity distribution by means of
et al. 2010). The formation of the ore body occurred in a successive solution of the forward problem for the prior
few stages of ore mineralization: the pyrite-potassium values of the resistivities in all domains of search. The
feldspar-(albite)-quartz stage was followed by the sulfide– inversion process was stopped when the marginal mean
quartz stage—the key stage for molybdenum; then, posterior resistivities in all grid cells are stabilized, i.e., the
molybdenum–quartz stage took place and, eventually, the normalized resistivity difference between two successive
process was finalized by the formation of the quartz–car- iterations is less than preset threshold value e (= 0.01):
bonate veins with rare impregnations of pyrite, galena, 0 ! 11:2
X qðnþ1Þ ðnÞ 2
sphalerite and chalkopyrite. These stages were separated in @ 1 0k  q k A  e;
ðnÞ
ð1Þ
time by intrusions of dikes and stocks of subalcalic por- K k¼1 q k
phyry granites. More than 80% of ores are vein-impreg-
ðnÞ
nated sulfides; the increase in the molybdenum where qk is an electrical resistivity in the k-th cell at
concentration is revealed in the veins and veinlets with a iteration n; K is a total number of cells in the grid (in our
width of up to 20 cm and even 1 m. Molybdenite, pyrite case K = 40 9 40 9 60 = 96,000).
and chalcopyrite are the main ore minerals, and quartz is
the main vein mineral. Seismic survey
To provide general geophysical characteristics of the
crust domains related to ore generation, transportation and Seismic survey was carried out in the study area by
deposition 3D models of electrical resistivity, seismic Kadurin et al. (2008) by means of a joint registration of
velocities, density and lithology of the study area were arrival times of compressional, shear and converted waves
built using magnetotelluric and seismic data collected by generated by local earthquakes. Seismic signals in the
Kadurin et al. (2008). Below we consider the methods used frequency range 0.5–10 Hz were recorded along a number
to this end. of profiles by a Russian made 4-channel digital recorder
‘‘Delta-Geon’’. The sampling of events was carried out
from 4 recorders in a 3-min window. After appropriate pre-
Methods processing of the P-, S- and PS-wave data, the travel-time
cross-sections were built.
Magnetotelluric sounding The reconstruction of seismic velocities was conducted
by Kadurin et al. (2008) in the context of a block-layered
Array magnetotelluric data were collected in the study area isotropic model of the Earth crust with spacing 1 km in
(bounded square in the Fig. 1) in the nodes of a uniform depth and 1 km in horizontal direction, which corresponds
rectangular network with 1 km spacing. MT measurements to the declared resolution at depth and horizontal site
were carried out using Phoenix instruments in the fre- spacing. The accuracy of the reconstruction of seismic
quency range 300–0.0001 Hz. Impedance amplitudes and velocities is estimated as 0.1 km/s for both compressional
phases were estimated with accuracy about 5% and 2–3°, (VP) and shear (VS) waves.
respectively. Relative discrepancies between the ampli-
tude–frequency responses of the magnitude and phases of Retrieving density from seismic velocities
the principal components of the tensor impedance in the
control MT soundings did not exceed 5% (Kadurin et al. Generally speaking, rock density depends on its mineral
2008). composition, structural and textural features, porosity, type
We have built the resistivity model of this area of the material filling its pores and voids and the conditions
(20km 9 20 km 9 50 km) using 3-D Bayesian statistical of rock formation and bedding. Since seismic velocities
inversion (Spichak et al. 1999) of the magnetotelluric depend on density, the knowledge of the elastic coefficients
impedances estimated at 441 (21 9 21) sites (see their

123
Acta Geophys. (2017) 65:1133–1144 1137

of the medium could probably provide the basis for direct Electrical resistivity model
estimation of density. On the other hand, the presence of
the known functional dependence between these parame- Figure 2 indicates horizontal cross-sections of the loga-
ters can be used for deriving the empirical formulas by rithmic electrical resistivity in the depth interval from 1 to
analyzing the experimental data from the same region. 50 km. The analysis of these slices suggests the following
To build the spatial model of rock density (1), the inferences. First, the background electrical resistivity of the
regression equation derived from the multivariate correla- medium is about 104 Xm, i.e., the area is overall highly
tion between the VP, VS, and VP/VS parameters (Aleinikov resistive. Second, an SE–NW striking listric fault is iden-
et al. 1986) was used (the standard deviation d in the tified at a depth of 10–40 km. It flattens out at the depths of
density estimation was 0.032 g/cm3 and the coefficient of Moho boundary in the northeast of the territory. Southwest
multivariate correlation R was 0.975): of this fault, the model contains a relatively conductive
structure with a resistivity less than 100 X.m, which
1 ¼ 0:107  VP  0:0537  VS þ 0:0026  VP VS
stretches from the surface to below 40 km (Fig. 3a), while
þ 0:0463ðVP2  4=3 VS2 Þ þ 2:66: ð2Þ
northwest of the fault a highly resistive batholite with a
resistivity of 105–106 X.m is located (Fig. 3b).
Deriving the lithology 3-D model of logarithmic resistivity for the study area is
presented in Fig. 4. It clearly indicates a conductive zone
There are different approaches to derive the lithology from with a diameter of about 5 km located at a depth of
the geophysical data. In particular, Bauer et al. (2003) 32–37 km. In the central part of this domain, the electrical
assessed the rock lithotypes in the subsurface part of the resistivity ranges from 10 to 40 Xm. Two areas with a
crust (up to a depth of 1–2 km) by constructing a cluster resistivity below 150 Xm are identified in the upper crust.
model in the two-dimensional space of seismic P-wave The channels with a resistivity of 500–1000 X.m connect
velocities and Poisson ratio. Alternative approach to the these areas with the lower-crustal anomaly. Table 1 sum-
typification of rocks could be based on the general classi- marizes relative anomalies of the electrical resistivity (to-
fication of all magmatic rocks of the Earth’s crust by their gether with other geophysical markers) in different parts of
chemical composition and, primarily, by the contents of the earth crust.
silica and alkali and their ratio in the rocks. Deep anomaly in the electrical resistivity could probably
To estimate the silica percentage in the crust of the study be caused by partial melting of hosting rocks (see, for
area, we used an empirical formula derived by Aleinikov et al. instance, Jones (1992) and references therein) heated by
(1986) from the multivariate correlation between the VP, VS, high-temperature basalt magma that has been previously
and VP/VS parameters (the standard deviation was 2.2% and transported to the reservoir from the mantle depths through
the coefficient of multivariate correlation was 0.97): the channel which has a resistivity of 150 to 500 Xm (see
  the front plane in Fig. 4) or by predominance of acidic
KSiO2 ð%Þ ¼ 152:420:82 VP VS  2:58 V2P 4=3 V2S :
(including granite) rocks which have relatively low elec-
ð3Þ trical resistivity (Parkhomenko 1967). However, taking
Based on the silica content K, it is possible to roughly into account a large time gap between the melting pro-
estimate the prevailing rock types in the crust, disregarding cesses indicated above and present geophysical observa-
the percentage of the sodium and potassium salts. tions we are inclined to suppose that the latter hypothesis is
According to the classification scheme presented by Mid- more preferable.
dlemost (1994), different bins of K can be associated with The observed low resistivity of the anomalous domains
the following main lithotypes: peridotite (K \ 40%), gab- in the upper crust could have been related to the aqueous
bro (40% \ K \ 52%), gabbro–diorite (52% \ K \57%), fluids present in fractured zones. On the other hand, it
diorite (57% \ K \ 63%), granodiorite could also be generated by conductive stockwork segments
(63% \ K \ 70%), and granite (K [ 70%). represented by thin veins and impregnated sulfides as well
as by thick metallic veins. And, finally, the relatively low
resistivity in a thick channel connecting two anomalous
Results areas in the lower and upper crust can be associated with
the presence of ore-bearing fluids.
In this section, we consider the results of successive Considering the fact that resistivity is sensitive to the
application of the methods presented above for building 3D fluid saturation, mineralization, porosity, permeability,
geophysical and lithology models characterizing the study lithological composition, temperature and pressure, the
area. nature of the mentioned anomalies can barely be unam-
biguously determined without additional information about

123
1138 Acta Geophys. (2017) 65:1133–1144

Fig. 2 Horizontal slices of the logarithmic electrical resistivity derived by 3D inversion of MT data

123
Acta Geophys. (2017) 65:1133–1144 1139

Fig. 3 Locations of the relatively conductive domain (a 75 Xm isosurface) and highly resistive batholite (b 105 Xm isosurface)

Seismic velocity models

Figure 5 indicates 3-D models of VP (a), VS (b) and VP/VS

(c) derived from seismic survey data by Kadurin et al.
(2008). VP ranges between 5.99 and 7.15 km/s (median
value is equal to 6.57 km/s) and is generally increasing
with depth. It is worth mentioning three peculiarities in its
behavior (Fig. 5a): first, VP values in the western part of the
upper crust are generally less than in the eastern part,
separated by the SE–NW striking listric fault determined
above from the electrical resistivity model (Fig. 2); second,
at depths 10–15 km in the former part VP jumps from 6.1 to
6.2 km/s marking the Conrad boundary; finally, at depths
below approximately 32 km an anomalous domain is seen
with decreased VP ranging between 5.96 and 6.19 km/s. In
the channel connecting the anomalous domains in the
lower and upper crust VP is slightly less than its median
value (6.34/6.57 km/s).
VS (Fig. 5b) ranges between 3.22 and 4.21 km/s (median
Fig. 4 Model of the logarithmic electrical resistivity value is equal to 3.71 km/s). Its spatial behavior is more
complicated; however, one can detect two anomalous
other rock properties. In the following section, we will domains with enhanced values of VS ranging between 3.81
return to this question after analyzing seismic velocity and 3.91 km/s: one is located at depths 25–45 km while
models. another one is located in the upper crust. VS values in the
connecting channel in the middle crust range between 3.61
and 3.71 km/s, which is slightly less than the median value.
The VP/VS ratio (Fig. 5c) varies from 1.66 to 1.88 [with
a median value of 1.77, which corresponds to the thickness

123
1140 Acta Geophys. (2017) 65:1133–1144

Table 1 Relative anomalies (in per cents from the median values) of the logarithmic electrical resistivity (Log q), compressional (VP) and shear
(VS) seismic velocities, ratio VP/VS and density (1)
Location Log q VP VS VP/VS 1

Upper crust (- 58.6)/(- 44.0) (- 9.3)/(- 5.8) (2.7)/(5.4) (- 5.1)/(- 3.9) (– 2.7)/0
Middle crust (- 29.3)/(- 14.6) (- 3.5)/0 (- 2.7)/0 (- 2.8)/1.1 (- 3.1)/0
Lower crust (- 73.3)/(- 58.6) (- 9.3)/(- 5.8) (2.7)/(5.4) (- 5.1)/(- 3.9) (- 8.6)/(- 5.8)

Fig. 5 Models of seismic velocities VP (a), VS (b), VP/VS (c) (after Kadurin et al. 2008) and density (d) estimated from seismic velocities’ data

123
Acta Geophys. (2017) 65:1133–1144 1141

of the consolidated crust of 43–45 km (Aleinikov et al. metasomatic processes (e.g., amphibolization), which are
1986)]. In the anomalous domains located in the lower and accompanied by the reduction in density. Generally, the
upper crust VP/VS is less than the median value and ranges reduced density at shallow depths of the upper crust could
between 1.68 and 1.70 while in the channel connecting have been due to increased concentration of less dense
these areas it ranges between 1.72 and 1.79. minerals, deconsolidation of rocks or to the fluids that fill
Relative anomalies of seismic velocities detected above the fractures branching from the listric fault and open
in different parts of the crust are summarized in the pores.
Table 1. Reduced VP and VP/VS and slightly enhanced VS in To further reduce the uncertainty regarding the origin of
the lower crust may characterize granite rather than basalt the revealed geophysical anomalies, we have built the
rocks (Schon 2015) and support the hypothesis of the lithological model of the study area.
predominance of acidic rocks formulated above in the
‘‘Electrical resistivity model’’ based on the electrical Lithology model
resistivity behavior.
The reduction in VP/VS in the subsurface is characteristic Figure 6 shows the horizontal slices of the 3-D lithology
of many shallow ore deposits (see, for instance, Vjunov model built in the way described above (see ‘‘Deriving of
et al. 2007). The anomalies in electrical resistivity and the lithology’’). This model indicates that the upper crust in
seismic velocities, which are revealed at shallow depths in the study area is dominated by gabbro–diorites and diorites.
the upper crust, can probably be due to the higher (relative Against this background, there are areas in the western part
to the assemblages of the hosting rocks) concentration of of the model that are mainly composed of predominantly
low-elastic minerals including silica with impregnated acidic (granite) magmas, which are traced from a depth of
metallic ore (Castagna et al. 1985). Note in this relation
that the latter factor should be considered as a sufficient but
not a necessary condition of the mentioned anomalies in
VP/VS: according to Kuzin (2012) such anomalies could be
caused also by ore mineralization in silica-poor rocks.
The channel of fluid transportation in the middle crust,
which connects the anomalous resistivity zones in the
lower and upper crust (Fig. 4) and has slightly lower
resistivity than surrounding medium, is characterized by
decrease of VP, VS and VP/VS (Fig. 5a–c). This could be
explained if we suppose that these anomalies are caused by
a supercritical gas–water fluid (Ito et al. 1979), which has a
decreased density. To ascertain these hypotheses, a three-
dimensional density model for the study area was built.

Density model

According to the density model built in the way described

above (see ‘‘Retrieving density from seismic velocities’’)
density across the entire area (Fig. 5d) varies within
2.50–3.33 g/cm3 with a median value of 2.91 g/cm3. This
span of density covers almost the entire range of the rock
types from acidic to ultrabasic. The overall distribution of
density is quite uniform except for the areas that hold the
mentioned anomalies in electrical resistivity and seismic
velocities, where the density varies from 2.67 to 2.75 g/
cm3, i.e., is by 5.7–8.5% lower than average.
This negative density anomaly in the depth interval of
the lower crust can be associated with the predominance of
acidic rocks (including granitic magma) in comparison
with surrounding basalt rocks characterized by higher
density (Parkhomenko 1967). In the middle crust, the Fig. 6 Lithology model based on the silica percentage in rocks:
density deficiency can be indicative of hydrothermal 1-gabbro, 2-gabbro–diorite, 3-diorite, 4-granite–diorite, 5-granite

123
1142 Acta Geophys. (2017) 65:1133–1144

about 32–37 km, where it is possible to presume the from assuming that the acidic melt was formed during the
presence of the ore-generating source up to the surface. crystallization differentiation of basalt magma that has
Remarkably, the SE–NW strike of this type of rocks on the risen from the mantle depths to the Moho. Thus, the lower-
surface coincides with the strike of the main deep fault crustal rocks enriched in amphibolites could be treated as a
revealed by MT soundings (Fig. 2). At the same time, the potential source for ore-bearing fluids, which are involved
model shows that similar ore-bearing segments can also be in hydrothermal processes in the lower–middle crust at the
located at a depth of up to 5 km, from which mining is in stage of development of leucogranites and ore-bearing
principle possible. porphyry granites.
The behavior of geophysical markers characterizing the
middle crust channel indicates that the displacement of
Discussion granite magma from the lower to the middle crust is
accompanied by its decompression degassing, metamor-
Joint analysis of revealed geophysical markers together phism and differentiation, which can give rise to the for-
with the lithology model enables to suggest a conceptual mation of post-magmatic supercritical gas–water fluids
model of the copper–porphyry ore formation. Unlike geo- (Williams-Jones and Heinrich 2005) resulting in reduction
logical models (Holliday and Cooke 2007; Berger et al. of VP, VS and VP/VS values. They are capable of extracting
2007; Sillitoe 2010) it indicates processes and mechanisms significant amounts of metallic ore from granite magma
of ore generation, transportation and deposition (Table 2). and becoming enriched in ore on their way as a result of
The anomalous area in the lower crust is characterized progressive and regressive metamorphosis of the host rocks
by reduced electrical resistivity, VP, VP/VS and density, and (Sinclair 2007). The latter process leads to some decrease
enhanced VS values. According to the combination of these of the electrical resistivity in comparison with surrounding
geophysical markers and the constructed lithotype model, medium.
this area can be interpreted as a reservoir of acidic granite Note in this relation that the hypothesis of ore minerals
magma with high concentration of salts. It could be formed transportation from the magmatic melt by purely gaseous
in this area at a lower-crustal level under the pressure of fluids (Henley and McNabb 1978; Heinrich et al. 1999;
above 10–12 Kbar (which corresponds to a depth of about Ulrich et al. 1999) is not supported by corresponding
40 km) due to partial melting of the previously crystallized anomalies in the rocks’ physical properties revealed above.
basites against the background reactivation of the mantle On the other hand, the leakage of supercritical gas–water
processes during the transition of the region to the stage of ore-bearing fluids through the plastic medium at depths of
Devonian rifting (Berzina et al. 2010). Note that a large the middle crust could occur due to sliding of fluid films
time gap between leucogranites and gabbro prevents one along the cleavage planes of the foliated rocks, e.g., in the

Table 2 Processes, mechanisms of fluid and magma transportation and geophysical markers characterizing copper–porphyry ore generation,
transportation and deposition
Location Processes Mechanisms of fluid and magma Geophysical markers
transportation
q VP VS V P/ 1
VS

Upper Accumulation of silica rich minerals with impregnations of Upward volumetric fluid transportation ; ; : ; ;
crust metallic ore along the network of large pores and
Ore deposition in the sedimentary cover cracks
Middle Decompression degassing, metamorphosis and differentiation of Upward sliding of the fluid films along ; ; ; ; ;
crust granite magma the cleavage planes of the foliated
Formation of post-magmatic supercritical gas-water fluids capable rocks
of extracting metallic ore from granite magma and metamorphic
hosting rocks
Transportation of ore-bearing supercritical gas-water fluids
Lower Slow cooling of the water bearing basalt melt and partial melting Upwelling of the basalt magma from ; ; : ; ;
crust of the hosting rocks the mantle
Formation of acidic granite magma reservoir with high
concentration of salts
Up and down arrows indicate signs of relative geophysical anomalies

123
Acta Geophys. (2017) 65:1133–1144 1143

zones of high-temperature blastomylonites characteristic of depths of the lower and middle crust due to the sliding of
the deep parts of the faults and folded dislocations (Let- the films of the fluid along the cleavage planes of the
nikov 2000). foliated rocks. At the depths of the upper crust, this
In the upper crust, where plastic deformations are mechanism could be changed by volumetric fluid trans-
changed by brittle ones, the film-type mechanism of fluid portation along the network of large pores and cracks.
transport could be changed by the mechanism of volu- The proposed methodology of successive constraining
metric transport (along the network of large pores and the areas of ore generation, transportation and deposition is
cracks). It is accompanied by decrease in density, VP, VP/ not restricted by only copper–porphyry ores. On the other
Vs, and electrical resistivity in the fluid pathway channels hand, prior knowledge or guess about the type of ore could
(to 300 Xm on average). At the depths of sedimentary affect the process of interpretation and lead to conclusions
cover, the ore-bearing fluids reach the ore deposition specific for only this type of ores.
location. It is marked by a more significant reduction in the
electrical resistivity (to 80 Xm), seismic P-wave velocity, Acknowledgements We acknowledge Mr Kadurin who offered us
the MT data collected in the study area and the results of 3D inversion
VP/VS ratio, and density. Considering this combination of of seismic data. We also thank Ass. Editor Acta Geophysica Dr Mario
the anomalies in the physical properties of rocks and taking Zarroca and two anonymous reviewers for constructive remarks and
into account the constructed lithological model, it is pos- suggestions, which helped us to significantly improve the manuscript.
sible to assume that they are caused by the increased
concentration of less dense minerals (e.g., silica) with
impregnations of metallic ore, which can reduce the bulk References
resistivity of rocks [e.g., see (Chen et al. 2012)].
Aleinikov AL, Nemzorov NI, Khalevin NI (1986) Multiwave seismic
Finally, the proposed methodology of determining pos- exploration for studying of the ore deposits. Nedra, Moscow,
sible locations of the ore generation, transportation and p 86 (in Russian)
deposition is essentially based on joint analysis of 3D rock Bauer K, Schulze A, Ryberg T, Sobolev S, Weber M (2003)
properties’ models revealed from geophysical data and is Classification of lithology from seismic tomography: a case
study from Messum igneous complex, Namibia. J Geophys Res
not restricted by only copper–porphyry ores. On the other 108(B3):2152–2167. https://doi.org/10.1029/2001JB001073
hand, prior knowledge or guess about the ore type could be Berger BR, Ayuso RA, Wynn JC, Seal RR (2008) Preliminary model
taken into account at the interpretation stage and lead to of porphyry copper deposits, USGS Open-file Report
conclusions specific for only this type. 2008–1321. 55 pp
Berzina AP, Berzina AN, Serov PA, Gimon VO (2010) The
petrogenic relationship between mafic and felsic rocks from
the Sora porphyry Cu-Mo center (Kuznetsk Alatau): A geo-
Conclusions chemical and Sm-Nd isotope study. Dokl Earth Sci
430(1):28–33. https://doi.org/10.1134/S1028334X1001006X
Castagna JP, Batzle ML, Eastwood RL (1985) Relationships between
Based on the joint analysis of the rock physical properties compressional-wave and shear-wave velocities in elastic silicate
derived from the geophysical observations, among which rocks. Geophysics 50(4):571–581. https://doi.org/10.1190/1.
the magnetotelluric and seismic soundings play the key 1441933
role, geophysical markers characterizing the areas of ore Chen Xiang-Bin, Lü Qing-Tian, Jia-Yong Yan (2012) 3D electrical
structure of porphyry copper deposit: a case study of Shaxi
generation, transportation and deposition are determined. copper deposit. Appl Geophys 9(3):270–278. https://doi.org/10.
Lithology model of the study area is built based on 1007/s11770-012-0337-1
estimating the percentage of silica in the rocks from the Heinrich CA, Gunther D, Audetat A, Ulrich T, Frischknecht R (1999)
velocities of seismic waves. It is in a good agreement with Metal fractionation between magmatic brine and vapor, deter-
mined by microanalysis of fluid inclusions. Geology 27:
previous geological and geochemical studies of this area 755–758. https://doi.org/10.1130/0091-7613(1999)027\0755:
and could be used for forecasting the location of the MFBMBA[2.3.CO;2
potential ore-bearing zones at depth. Henley RW, McNabb A (1978) Magmatic vapor plumes and
Comparative analysis of revealed geophysical anomalies groundwater interaction in porphyry copper emplacement. Econ
Geol 73:1–20
together with a lithology model enables to suggest a con- Holliday JR, Cooke DR (2007) Advances in geological models and
ceptual model of the copper–porphyry deposit formation exploration methods for copper and gold porphyry deposits. In:
explaining possible mechanisms of ore generation, trans- Milkereit B (ed) Proceedings of fifth decennial international
portation from the lower to the upper crust and deposition conference on mineral exploration, 791–809
Ito H, De Vilbiss J, Nur A (1979) Compressional and shear waves in
in the upper crust. saturated rock during water-steam transition. J Geophys Res
The mechanisms of ore transportation from the lower to 93:12081–12106
middle and further to the upper crust are hypothesized. It is Jones A (1992) Electrical conductivity of the continental lower crust.
supposed that post-magmatic supercritical gas–water ore- In: Fountain DM, Arculus RJ, Kay RW (eds) continental lower
crust. Elsevier, Amsterdam, pp 81–143
bearing fluids are moving through the plastic crust at the

123
1144 Acta Geophys. (2017) 65:1133–1144

Kadurin IN, Belyavsky VV, Egorkin AV (2008) Geophysical study of Sotnikov VI (2006) Copper-molibdenum ore formation: origin,
the deep structure of Altai-Sayan folded belt: seismic and problems of volume and bounds. Russ Geol Geophys
resistivity prospecting along the 3300-km long profile network. 47(4):355–363. https://doi.org/10.1134/S107570150606002X
Tech. Rep. Geolfond, Moscow. 244 pp. (in Russian) Spichak VV, Menvielle M, Roussignol M (1999) Three-dimensional
Kuzin AM (2012) On possible nature of relatively low values of Vp/ inversion of MT fields using bayesian statistics. In: Oristaglio M,
Vs in the ore deposits of fluid genesis. Geofizika 2:7–15 (in Spies B (eds) 3D Electromagnetics. SEG Publ., GD7, Tulsa,
Russian) pp 406–417
Letnikov FA (2000) Fluid regime of the endogenic processes in the Tavakoli S, Dehghannejad M, Garcı́a Juanatey MA, Bauer TE,
continental lithosphere and the problems of metallogeny. In: Weihed P, Elming S-A (2016) Potential Field. Geoelectrical and
Adushkin A (ed) Problemi globalnoi geodinamiki. GEOS, reflection seismic investigations for massive sulphide explo-
Moscow, pp 204–224 (in Russian) ration in the skellefte mining District, Northern Sweden, Acta
Middlemost EAK (1994) Naming materials in the magma/igneous geophysica 64(6):2171–2199. https://doi.org/10.1515/acgeo-
system. Earth Sci Rev 37:215–224 2016-0088
Parkhomenko EI (1967) Electrical properties of rocks. Plenum Press, Ulrich T, Gunther D, Heinrich CA (1999) Gold concentrations of
New York, p 314 magmatic brines and the metal budget of porphyry copper
Schon JH (2015) Physical properties of rocks: fundamentals and deposits. Nature 399:676–679
principles of petrophysics. Elsevier, Amsterdam, p 512 Vjunov DL, Nosirev MYu, Stepanov VA (2007) Complex interpre-
Sillitoe RH (2010) Porphyry copper systems. Econ Geol 105(1):3–41. tation of geophysical and geochemical data for the prognosis of
https://doi.org/10.2113/gsecongeo.105.1.3 mineralization: a case study of the upper Amur region. Dokl
Sinclair WD (2007) Porphyry deposits. In: Goodfellow WD (ed) Earth Sci 413(2):384–387. https://doi.org/10.1134/
Mineral deposits of Canada. vol 5. Geological Association of S1028334X07030142
Canada Special Publication. Ottawa, pp 223–243 Williams-Jones AE, Heinrich CA (2005) Vapor transport of metals
Sokolova LS, Duchkov AD (2008) Heat flow of the Altai-Sayan and the formation of magmatic-hydrothermal ore deposits. Econ
region: new data. Russ Geol Geophys 49:940–950. https://doi. Geol 100:1287–1312. https://doi.org/10.2113/gsecongeo.100.7.
org/10.1016/j.rgg.2008.03.007 1287

123