A Review of Relationship between the Metallogenic System of Metallic Mineral Deposits and Lithospheric Electrical Structure: Insight from Magnetotelluric Imaging

Abstract

In development over 70 years, magnetotelluric (MT) sounding, a high-resolution technique for subsurface electrical resistivity imaging, has been widely applied in resource exploration in the Earth. The key factors of the metallogenic system of metallic mineral deposits can be closely correlated to the electrical anomalies of the lithosphere. In this paper, we review the relationship between the electrical resistivity model of the lithosphere and the metallogenic system. At the beginning, we indicate why the electrical parameters relate to the metallogenic system in all geophysical parameters. The advantage of MT sounding in sketching an electrical resistivity model of the lithosphere is subsequently discussed, and some methods of data processing, analysis and inversion are also introduced. Furthermore, we summarize how to bridge the relationship between the electrical resistivity model of the lithosphere and metallogenic system, and analyze the influence of the rheological variation estimated from conductivity in the lithosphere on mineralization. In the end, we list some typical cases of the application of MT sounding in mineral exploration, and also give some suggestions for future work. This study is aimed at providing guidance in discussing the metallogenic system using an electrical resistivity model.

1. Introduction

The material circulation between the Earth’s sub-systems maintains the dynamic equilibrium in the Earth’s system, which allows the abnormal enrichment of metals to form the mineral resources that human society depends on. The orogenic belts formed by oceanic subduction and continental collision are regions of the most active material circulation and the richest mineral resources in the Earth [1,2]. Statistical data show that 92.5 percent of the giant deposits use water as the driving force of mineralization, and 2/3 of giant deposits are related to magmatic–hydrothermal fluids, in which most metals are precipitated [3]. Therefore, the magmatic–hydrothermal deposits are the largest deposit family [4]. The upwelling, flow and thermal erosion of the asthenosphere with different types induce the partial melting of the underlying lithosphere, the rheological deformation of the crust and surface–crust–mantle interaction, which cause the activation and massive enrichment of metals and form the various ore deposits produced by the circulation of magmatic–hydrothermal fluids [5,6,7,8].

Previous studies stated that the components of a metallogenic system can be categorized into seven parts [9,10]; that is, (1) the metallogenic geodynamic setting, which facilitates the production of the metal-rich magmatic–hydrothermal fluids; (2) favorable lithospheric architecture; (3) sources of fluids; (4) drivers of fluid flow; (5) fluid-flow pathways; (6) physical and chemical processes of metal migration and deposition, and (7) preservation conditions of the ore deposit. The ore deposits can be formed on multiple scales, while the single metallogenic system can be extended over hundreds of thousands of square kilometers, from the entire crust to the upper mantle [11]. Although geophysical methods mainly represent the current subsurface material structure, the most recent geological event can modify the previous mineralization, and furthermore, the large-scale structural features revealed by geophysical methods (e.g., the magnetotelluric sounding and seismic imaging) could be preserved over a long historical period [12]. Therefore, it is feasible to discuss the metallogenic system based on the spatial variation of geophysical parameters.

At present, our discussion of the metallogenic system of ore deposits can concentrate on (1) an integrative model for revealing crust–mantle material structure; (2) the interaction of layers in the Earth; (3) tracing the circulation of metallogenic materials; (4) illustrating the enrichment mechanism of metals and (5) establishing the metallogenic model [9,12]. Therefore, besides the abundance of geological and geochemical research, geophysical models of the lithosphere are also needed to provide the geophysical evidence required to analyze the source–migration–deposit process of the magma–fluid–mineralization system.

2. Electrical Parameter and Electrical Prospecting

Electrical conductivity (EC; simplified as conductivity, $\sigma$) is a material’s ability to conduct electric current, and takes the SI units of a Siemens per meter (S/m), which is the reciprocal of electrical resistivity (simplified as resistivity; SI units of ohm-meters (Ω·m)). Conductivity is related to the transmission of the electric current by free charge carriers, and can be described as the special charge carrier (Equation (1)).

$$\sigma =cq\mu$$

(1)

where c, q and μ are the charge concentration, charge and mobility of the charge carrier, respectively.

The ionic conduction, electronic conduction and semi-conduction (proton conduction and small polaron conduction) are the primary conduction mechanisms in the minerals and rocks of the lithosphere, and the conditions of temperature and pressure have different influences on different conduction mechanisms. The conductivity of minerals or rocks in the lithosphere may be related to one or several conduction mechanisms. Note that it is common for a single mechanism of a material to dominate in a given thermal regime. The conductive phases in the lithosphere mainly include saline fluids, partial melts, graphite, sulfides and water (hydrogen) in nominally anhydrous minerals (NAMs). Previous studies have summarized the conducive characteristics of these phases [13,14] (Figure 1), and in this paper, these are not further described. Conductive mediums, such as graphite/graphite film, metallic oxide/sulfide, aqueous fluid and molten silicate/carbonate melt, have a conductivity that is several orders of magnitude higher than that of country rocks [15]. When conductive minerals form a connected net, the bulk conductivity can be controlled by a small number of highly conductive minerals.

Furthermore, the vapor phase (mainly volatile) also influences the bulk conductivity of rocks. The deposit types of different volatiles in the lithosphere depend on the solubility of different volatiles in the different fluids or melts. In comparison with the conductivity of the vapor phase, the volatiles in the fluids or melts have a positive effect on the bulk conductivity of rocks. Taking carbon dioxide (CO₂), which is the second most abundant volatile in the lithosphere, as an example, (1) CO₂ has much higher solubility in alkali-rich melts [16,17,18,19]; (2) the conductivity of the CO₂ presented as the vapor phase is approximately 0.01 S/m (resistivity of 100 Ω·m), and 5–8 vol% of CO₂ can reduce the conductivity from 0.04 S/m (resistivity of 250 Ω·m) to 0.01 S/m (resistivity of 100 Ω·m) [20]; (3) when CO₂ is in the melt, the carbonate ion (CO₂⁻³) can act as the effective charge carrier, and contribute significantly to electrical conduction [16,21].

Figure 1. A cartoon showing key features of the crust and upper mantle and their typical ranges of resistivity, with numbers denoting resistivity (Ω·m) [14,22].

Figure 1. A cartoon showing key features of the crust and upper mantle and their typical ranges of resistivity, with numbers denoting resistivity (Ω·m) [14,22].

Because conductivity is extremely sensitive to the high-conductivity, high-temperature and high-rheology materials (e.g., metallic minerals, melts and aqueous fluids) [23], the electrical structure (the spatial variation of conductivity or resistivity) may be preferred when studying the metallogenic system (especially metallic minerals). Electric resistivity (direct current, DC) and electromagnetic (EM) methods (such as the time-domain electromagnetic (TEM) method and magnetotelluric method (MT)) can be used to sketch the variation of resistivity/conductivity in the underground medium [24,25]. Because the probing depth of electric resistivity methods is shallow, the EM methods are the main methods used to establish the lithospheric electrical structure and study the metallogenic system. Furthermore, the composition of the crust and mantle is heterogeneous, and therefore, the EM method is used in measuring volume current, which represents the average volume of the compositive material assembly. In comparison with the time-domain electromagnetic method (TEM) and controlled source electromagnetic methods (such as controlled sources audio magnetotelluric sounding (CSAMT) and wide-field electromagnetic method (WFEM)), magnetotelluric (MT) sounding is the primary geophysical method that can image the electrical properties of the lithosphere–asthenosphere system. At present, the geophysical electromagnetic methods play an important role in the exploration of copper, molybdenum, lead, zinc, bauxite, uranium and other metal mineral resources all over the world (see more details in Section 6).

3. Construction of the Lithospheric Electrical Structure

3.1. Magnetotelluric Sounding and Data Processing

Magnetotelluric (MT) sounding is a passive electromagnetic geophysical method, which can measure signals in a broad frequency of the electromagnetic field varying from 10⁵ Hz to 10⁻⁵ Hz, and includes audio magnetotelluric (AMT) sounding, broad-band magnetotelluric (BMT) sounding and long-period magnetotelluric (LMT) sounding [23]. High-frequency signals (>1 Hz) originate from worldwide lightning storms, and low-frequency signals (<1 Hz) are derived from the interactions of solar wind with the Earth’s magnetosphere. In a uniformly half-space medium, the penetration depth of EM fields is described by skin depth $\delta =503\sqrt{\rho /f}$ (m), where $\rho$ denotes the subsurface resistivity and f is the frequency (Figure 2).

MT sounding can be used to image the subsurface electrical structure by measuring the time variation of natural electric (E) and magnetic (H) fields in the orthogonal horizontal direction (Ex, Ey, Hx, Hy; x and y are south–north and east–west directions) and the vertical direction (Hz) at the Earth’s surface (see Chave and Jones, 2012 [23]; Figure 3).

The time-domain electromagnetic signal is transformed into frequency-domain data by fast Fourier Transform (FFT). Subsequently, the frequency-domain MT data are processed with the least square method or statistically robust algorithm [27], and noise is reduced at the measurement locations by employing the remote reference method, which can be effective in improving the coherence and signal–noise ratio of the MT data [23]. The electric and magnetic field components are related by the complex impedance tensor Z (Equation (2)).

$$\left[\begin{array}{l}{E}_x\\ E_y\end{array}\right]=\left[\begin{array}{l}{Z}_{xx}\hspace{1em}{Z}_{xy}\\ Z_{yx}\hspace{1em}{Z}_{yy}\end{array}\right]\left[\begin{array}{l}{H}_x\\ H_y\end{array}\right]=\left[Z\right]\left[\begin{array}{l}{H}_x\\ H_y\end{array}\right]$$

(2)

The apparent resistivity (${\rho }_{ij}$) and phase (${\varphi }_{ij}$) can be calculated from four components of the impedance tensor matrix $\left[Z\right]$,

$${\rho }_{ij}=\frac{1}{\omega \mu }{\left|Z_{ij}\right|}^2$$

(3)

$${\varphi }_{ij}=\arctan (\frac{\mathrm{Im}{Z}_{ij}}{\mathrm{Re}{Z}_{ij}})$$

(4)

where ω is the angular frequency and μ is the permeability; Im Z_ij and Re Z_ij are the imaginary and real parts of Z_ij, respectively; i and j are directions. The phase angle is commonly expressed in the first quadrant (0° ≤ ϕ ≤ 90°) such that it indicates whether apparent resistivity increases (ϕ_ij < 45°) or decreases (ϕ_ij > 45°) as frequency decreases.

Furthermore, under the condition of horizontal heterogeneity, the vertical component of the magnetic field is not equal to zero ($H_z\ne 0$), and therefore, the ratio of horizontal and vertical components of magnetic field is defined as the tipper data (T),

$$H_z=\left[T_x,T_y\right]\left[\begin{array}{l}{H}_x\\ H_y\end{array}\right]$$

(5)

The tipper data include information about the spatial variation of resistivity; however, the absolute resistivity value needs to be constrained by the data related to the electric field. The tipper data can offer supplementary information to impedance data.

3.2. Magnetotelluric Data Analysis and Inversion

Following the processing of the MT data, some basic analyses of the MT data are required [27]. (1) Analysis of the phase tensor. This is mainly used to analyze the dimensionality of the MT data and the regional tectonic strike, which results from the advantage of not being affected by a three-dimensional distortion effect [28,29]. (2) Analysis of the impedance tensor. This is used to analyze and estimate the tectonic strike. Before the two-dimensional inversion of the MT data, the tectonic strike needs to be estimated in the research region, and then the MT data must be rotated according to the tectonic strike [30], as part of which the MT data can be divided into two perpendicular, independent polarization modes (transverse electric (TE) mode; transverse magnetic (TM) mode). (3) Analysis of the induction vector. This is sensitive to the horizontal variation in conductivity of the underground medium. Under Parkinon’s regulation, the arrows of induction vector point to the converging areas of current [31], which are the conductive zones, while under Wiese’s regulation [32], the arrows point to the high-resistance zones. The induction vector is mainly used to judge the positions of conductive zones, and can also evaluate the dimensionality. Under the ideal two-dimensional condition, the distribution of conductive zones must extend infinitely along the tectonic strike. (4) Analysis of the galvanic distortion. This results from the conductive inhomogeneity in the shallow area, which can give rise to frequency-independent field amplifications and rotations, and is also a key issue of the electromagnetic method [33]. Previous studies have proposed many methods to tackle this problem [33,34,35,36].

Subsequently, according to the different purposes of different research, the one-dimensional, two-dimensional and three-dimensional inversions of the MT data can be carried out to obtain the electrical resistivity model of the lithosphere. The nonlinear conjugate gradient algorithm (NLCG), limited-memory Broyden–Fletcher–Goldfarb–Shanno (L-BFGS) algorithm, OCCAM algorithm, base reduction OCCAM algorithm (REBOCC) and fast relaxation algorithm (RRI) are the main inversion algorithms [37,38,39,40]. Furthermore, although the natural electromagnetic (EM) signal can penetrate the resistive zones and detect the conductive zones, the EM signal cannot be easily used to determine the bottom boundary of the conductive zones because of the diffusion of the EM signal in the Earth’s interior [41,42]. Therefore, after the inversion, some sensitivity tests must be carried out to verify whether the MT data have good constraints on the electrical resistivity model.

4. Connecting the Electrical Resistivity Model of Lithosphere and Metallogenic System

Metallic minerals are mainly hosted in the rocks as metal sulfides (e.g., bornite, chalcocite, pyrite, sphalerite, galena, molybdenite) and metal oxides (e.g., cuprite, magnetite). It is generally known that the conductivity of metal-bearing rocks is much higher than that of the surrounding rocks. The structure of minerals in the rocks also has an influence on conductivity; when the metallic minerals have a good interconnection in the rocks (e.g., lumpy structure, vein-like structure, spotted structure), the conductivity increases sharply; when the metallic minerals have a bad interconnection in the rocks (e.g., dot structure, disseminated structure), conductivity increases a little, or the variation of conductivity may not be recognized in the electrical resistivity model. Furthermore, because the metallic minerals are sparsely distributed in the ore bodies, which results in the limited distribution of conductive anomalies, the metallic minerals can represent a possible factor affecting conductive zones in the shallow zones of crust, and struggle to produce large-scale conductive anomalies in the deep Earth [41,42]. Therefore, the conductive zones in the subsurface depicted by MT sounding in the high-frequency band can be interpreted as large-scale ore deposits with interconnected minerals [43,44].

Based on different migration mediums in the mineralization, the ore deposits can be divided into three types, that is, magmatic deposit, hydrothermal deposit and superficial deposit. The enrichment of metallic minerals in the magmatic deposits and hydrothermal deposits are formed by the filling process and metamorphism, while the superficial deposits are formed by either the reverse process of the selective removal of surface water from rocks of the Earth’s surface, or the enrichment of valuable elements in the solid residues through chemical dissolution and mechanical separation. Therefore, the fractures and fault zones in the rocks not only act as important migration channels of the medium, but are also filled with minerals as a vessel during mineralization. Generally, fractures and fault zones are indicated as zones of conduction from the subsurface to the deep-seated zone, or electrical gradient belts between the conductive zones and resistive zones, or electrical interval belts between the resistive zones [45,46].

Because magmatic and hydrothermal deposits mainly form in the background of magmatic events (e.g., the geological environment related to the intrusive bodies), the upwelling and evolution of the magmatic–hydrothermal materials play an essential role in the mineralization of magmatic–hydrothermal deposits. The multi-stage upwelling of the dense mafic magmas from the mantle wedge underplated crust forms the large-volume periodical magma chambers, and meanwhile, the MASH process (melting, assimilation, storage and homogenization) occurs at the crust–mantle boundary [47] (Figure 4). The magmatic–hydrothermal materials subsequently go through the processes of liquation and multi-stage penetration, and migrate into the shallow area to form metal-rich small-volume magma chambers, which act as a key point in the evolution of ore deposits [48]. Previous studies have reported that deep-seated magmatic–hydrothermal materials migrate upwards along the existing faults and weakened zones in the lithosphere [49], in which the gathering and evolution zones can also be identified as a high-rheological feature. With the decrease in temperature following mineralization, melts in the magma chambers and migration channels gradually cool and crystallize. In this process, fluids are released and migrate to the low-pressure zone above. Following current mineralization, the late tectonic–magmatic events may reactivate the existing weakened zones and fault zones in the lithosphere, which can facilitate the migration and gathering of magmatic–hydrothermal materials [46].

During the mineralization of magmatic–hydrothermal deposits, the volatiles (e.g., chlorine, sulfur, carbon dioxide, methane, water) in the magmatic–hydrothermal materials also play an essential role in the mineralization of different deposits [5,50,51,52,53]. For example, the concentration of Cl ions can increase the distribution ratio of metal elements in the fluid [50,54]. Furthermore, the alkali-rich, volatile-rich and/or water-rich melts (e.g., anatexis in the crust) also contribute to most types of ore deposits [6,53,54,55,56,57,58,59]. During or after mineralization, these volatile residues may form a chlorate, sulfide and/or graphite film under specific temperature and pressure conditions, which can increase the conductivity of the rock system [13,14,60]. The ore deposits form in a specific geological time and the metallogenic events are transient; however, the most important features of metallogenic systems are shaped before the mineralization or a long period after the metallogenic event, and the evolution of metallogenic systems may be preserved for hundreds of millions to billions of years [2,12]. Because the large-scale tectonic feature depicted by MT sounding may not change a lot over a long historical period, the conductive zones in the lithosphere can be interpreted as electrical traces of residues related to mineralization, even though they may overlap with the latest magmatic activity. Therefore, an electrical resistivity model of the lithosphere can provide constraints on the migration of magmatic–hydrothermal materials from the source area to the subsurface shallow area.

5. Influence of the Rheological Structure on the Metallogenic System

• Relationship of conductivity and melt/fluid fraction

As mentioned above, the migration of lithospheric materials is controlled by faults and weakened zones, which show up as high-conductive anomalies. Furthermore, the weakened zones are closely related to partial melting in the lithosphere. A petrophysical experiment reported that when the volume fractions of melts are more than 5%, the effective viscosity of rocks can be reduced by an order of magnitude; when the volume fractions are more than 20%, the effective viscosity may be reduced by at least two orders of magnitude [61,62,63,64].

Silicate melts and carbonate melts, both of which are formed by the partial melting of different rocks, are the main melts in the lithosphere [65,66]. The partial melting of carbonate is generated at depths of more than 300 km in the mantle, and carbonated melts are stable under the condition of more than 2.5 GPa, while the silicate melts are more likely to exist at depths of less than 180 km [67,68]. When the upwelling carbonate melts react with normally anhydrous minerals, melts change from carbonate melts to carbonate-bearing (CO₂-bearing) silicate melts [69]. Furthermore, free fluids (alkali-bearing/saline fluids) are generally present in the porous sedimentary rocks, brittle fractures and shear zones in the mid-upper crust, and dehydrated areas [70], and the chlorate and bicarbonate are the main alkali-bearing fluids in the lithosphere. Because the conductive feature of bicarbonate is controlled by the content of carbon dioxide in the fluid, the saline (alkali-bearing) fluids can be simply described as the chlorate of potassic and sodium [70,71,72].

Although MT sounding struggles in obtaining the specific distribution of conductive zones, it can estimate the total volume of melt, that is, the volume fraction of melt [73]. Note that a small number of fluids can also greatly increase the bulk conductivity of rocks, and therefore, both the thin layer, with high volume fractions of melts, and the thick layer, with low volume fractions of melts, have the same bulk conductivity, and produce the same observation regarding MT sounding [23].

2.

Mixing Models in the Lithosphere

Rocks in the crust and upper mantle generally consist of some minerals and mixtures, the bulk conductivity of which is determined by the conductivity, percentage by volume (vol%), percentage by weight (wt. %) and spatial distribution of every component. Therefore, the volume fractions of melts or fluids can be estimated by the specific mixed model when the information above is known. Because of the occurrence of some uncertainties when calculating bulk conductivity in the lithosphere, a variety of numerical models have been proposed to calculate the bulk conductivity of rocks in different geometric distributions. The mixing models include the parallel model, the perpendicular model, the Hashin–Shtrikman bound model, Archie’s law model, the random model, the Waff model, the modified brick model, the Lichtenecker–Rother equation model and the Bussian model [74,75,76,77,78,79,80,81], of which the Hashin–Shtrikman bound model and Archie’s law model the two most common models.

The Hashin–Shtrikman bound model (Equation (6)) [75] can be divided into the Hashin–Shtrikman upper (HS+) and lower bounds (HS−). The HS+ bound model is a model wherein the melt phase is completely connected and contains low-conductive spherical grains. In contrast, the HS− bound model is the resistive aggregate containing discontinuous conductive minerals.

$${\sigma }_{H{S}_{\pm }}\equiv {[{\sum }_{i=1}^N\frac{v_i}{{\sigma }_i+2{\sigma }_{\max ,\min }}]}^{-1}-2{\sigma }_{\max ,\min }$$

(6)

where ${\sigma }_{H{S}_{\pm }}$ are the maximum and minimum conductivity in the multi-phase system, N is the total number of minerals, $v_i$ is the volume percentage of every mineral, ${\sigma }_i$ is the conductivity of every mineral, and σ_max,min are the maximum and minimum minerals.

Archie’s law (Equation (7)) [74,80,81] is a method that can be used to connect the conductivity of a clean reservoir rock with its porosity and the conductivity of its pore fluid. The generalized form can be expressed as:

$${\sigma }_{eff}={\displaystyle \sum _{i=1}^n{\sigma }_i{\Phi }_i^{m_i}}$$

(7)

where ${\sigma }_{eff}$ is the bulk conductivity, ${\sigma }_i$ is the melt conductivity of every phase, ${\Phi }_i$ is the percentage, and m_i is the cementation exponent.

Therefore, based on the electrical resistivity model of the lithosphere, the volume fractions of melts or fluids are estimated according to different mixing models.

3.

Rheological constraints from electrical conductivity

Besides the rheological estimation of the lithosphere based on the relationship of conductivity and volume fractions of melts or fluids, the conductivity–viscosity relationship is also simplified as a ratio relationship between these two factors. Liu and Hasterok (2016) [82] put forward an empirical formula for the conductivity–viscosity relationship (Equation (8)):

$$\frac{{\eta }_{eff}}{{\eta }_0}=C_0{\left(\frac{\rho }{{\rho }_0}\right)}^{C_1}$$

(8)

The parameters are defined as follows: ${\eta }_{eff}$ is the effective viscosity, ${\eta }_0$ is the reference viscosity, $\rho$ is the measured resistivity, ${\rho }_0$ is the reference resistivity, and both C₀ and C₁ are controlling coefficients [82]. Here the reference viscosity is taken to be 10²⁰ Pa·s and the reference resistivity is taken to be 100 Ω·m. The controlling coefficients are estimated over a narrow range and act in the following way: C₀ is the controlling coefficient of the reference resistivity (values possibly in the range of 0.5–8.0), and C₁ controls the maximum contrast of the viscosity (values possibly in the range of 0.5–1.5) [82]. With the systematic variation of C₀ and C₁, the observed conductivity can match the viscosity of the lithosphere. In addition, in the upper mantle, C₀ and C₁ can be set as 1 and 0.6667, respectively [83].

Certainly, the presence of partial melting can cause a correlating response in both resistivity and viscosity, with an increasing melt fraction leading to lower values of both. However, some materials with a conductive phase may be present in the crust that lower the resistivity but have a small (or no) impact on the viscosity. Examples include graphite films and sulfide minerals (important specifically in the shallow crust), both of which are important specifically in the upper crust.

The strength of rocks, which is related to a variation in viscosity, determines whether the “channel flow” occurs under the given boundary conditions, and at what rate the deformation occurs [84,85]. The “channel flow” in the partial melting zones can be represented as the Couette flow and/or Poiseuille flow. Couette flow is the laminar flow of a viscous fluid in the space between two parallel plates that are moving relative to each other. Poiseuille flow is pressure-induced flow (“channel flow”) in a layer, usually a pipe, and is distinct from drag-induced flow such as Couette flow.

In brief, the weakened zones characterized by low viscosity in the lithosphere can provide the channels of migration and the emplacement of the magmatic–hydrothermal materials, and can also influence the stress condition of the overlying rigid stratum. Therefore, rocks in the overlying stratum may locally deform, which has an impact on the exposure state and distribution feature of deposits following mineralization.

6. Typical Cases

Based on the temporal and spatial coupling relationships of other geophysical, geological and geochemical data, the electrical resistivity model obtained by MT sounding has been successfully applied in studying the metallogenic dynamics of ore deposits, and the metallogenic regularity and models of some typical ore deposits (e.g., multi-store metallogenic model [12]) have also been summarized. In the following section, three typical cases are shown that depict the relationships between the electrical resistivity models of the lithosphere and the metallogenic system.

In Australia [86,87], a cross-sectional electrical resistivity model to a depth of 60 km in the Olympic Dam is shown in Figure 5. Based on the depth and shape of the conductive zone C3 (~1 S/m), the brittle–ductile transition zone is approximately 15 km, and three narrow low-resistance zones above this zone branch to the surface. Heinson et al. (2018) [87] reported that lithospheric imaging helps portray a deep metallogenic system and map some of the pathways of metalliferous fluids moving from crust–mantle sources to discrete locations in the surface.

For southeastern China [88], an electrical resistivity model produced from a 3D MT data array in the Nanling–Xuancheng area is shown in Figure 6. The conductive zones C7 and C8 in the mid–lower crust can be attributed to hydrous fluids that are possibly related to partial melting. The upwelling fluids create a weak zone at the ductile–brittle transition (conductive zones C1–C5), and provide the conditions for crustal decoupling caused by the NW-directional compression of the Paleo-Pacific plate. The fluids migrate through NE–SW-directional channels and interact with the upper crustal rocks. Therefore, according to the electrical structure and rheological parameters as indicated by electrical conductivity, a new mechanism is offered, referred to as “crustal decoupling”, which explains the tectonic–magmatic history of the magmatic metallogenic system in the Nanling–Xuancheng area.

In the context of North China, a multi-scale 3D electrical structure of the lithosphere of the Caosiyao porphyry Mo ore district is obtained by combining the joint inversion of the AMT and MT data [44] with the electrical structure in the lithospheric mantle previously derived by Zhang et al. (2016) [89]. The multi-scale electrical structure indicates a giant upper-mantle basaltic magma reservoir, indicated by the conductor DTHCZ beneath the Datong volcanic area, which moves through a lithosphere-scale shear zone, the mid–lower crustal magma chamber CZ3, the crust-scale weak zone, and a conduit (CC1 and CC2) to transport ore-forming magmas and fluids (Figure 7).

Other typical results include the Ning (Nanjing)-Wu (Wuhu) basin in eastern China [90], the Shihu–Xishimen gold–iron ore district in North China [91], the Middle–lower Yangtze Metallogenic Belt (including Ningwu, Luzong, Xuancheng, Anqing, Guichi, and Tongling areas) in East China [12,92,93,94,95,96], the Jiaodong gold deposits in northern China [97], the Liaodong Qingchengzi orefield [98], the Yixingzhai gold deposits (Boqiang Cu–Mo–Au deposits and Nanling W-Sn ore district in China) [45,99,100], the northeastern Jiangxi metallogenic province in southeastern China [101], the Caosiyao porphyry Mo ore district in North China [43], the Beiya Cu ore district in southwestern China [46], the large-scale Hatu epithermal gold deposit in western Junggar, NW China [102], Zhaxikang in the Tethys–Himalaya area in southwestern China [103,104], the Baogutu porphyry copper–gold deposit, the Hongqiling Cu–Ni sulfide intrusions in the Central Asian Orogenic Belt [105], the Xiangshan volcanogenic uranium deposit [106,107], the Narusongduo Pb–Zn–Fe–Cu ore district, and the Qulong–Jima porphyry ore Cu deposit in southwestern China [41,42]. In the Olympic Dam mine and Tennant ore district in Australia [108], the Tsagaan Tsahir Uul Au deposit in southern Mongolia [109], the Norrbotten district in northern Sweden [110,111], and the orogenic gold district in the Red Lake greenstone belt, western Superior craton, Canada [51], the electrical resistivity models obtained by MT sounding provide better geophysical evidence of the metallogenic dynamics of metallic mineral deposits.

7. Summary

In recent years, the relationship between the metallogenic system and the multi-scale electrical structure obtained by MT sounding in the lithosphere has become a hot topic, and a frontier in the academic field. Based on many geological and geochemical studies, geophysical imaging, especially regarding geoelectrical structures, can be used to sketch the migration channels of materials and ancient magma chambers from a macro perspective. Meanwhile, the temporal and spatial coupling relationship of the geological, geophysical and geochemical data contribute to enriching our understanding of the metallogenic background and metallogenic mechanisms. Aside from metallic mineral deposits, MT sounding can also be used in the prospecting of nonmetallic deposits, such as gas hydrate and natural gas [112].

Furthermore, some suggestions regarding the relationship between the electrical structure and metallogenic system can also be given, as follows: (a) how to obtain a high-resolution, multi-scale electrical resistivity model; (2) how to synthetically analyze the different parameters obtained by different electrical methods; (3) how to improve the temporal and spatial coupling of geological, geophysical and geochemical data; (4) how to choose a mixed model of the lithosphere and the values of different parameters used for the estimation of conductivity and melt/fluid volume fraction; (5) how to improve the interpretation of coupling mechanisms in the deep and shallow lithosphere based on an electrical resistivity model; (6) how to perform a petrophysical experiment on specific rocks and minerals.

Finally, as Engels said in the «Dialectics of Nature»: “We can understand the issue under the conditions of The Times, and the extents we understand depend on what these conditions reach”. This paper seeks to throw out a minnow so as to catch a whale.