HTML
-
Geophysical field has the function of penetrating through the covered rocks and detect the deep geological structures and the orebodies (Wang et al., 2000). Geophysical methods are particularly well suited for discovering concealed ore deposits (Shah et al., 2013). Geophysical data such as magnetic and gravity surveys may contain ore-finding information from greater depths and has not been fully explored due to the limitation of the traditional mathematical model like the Fourier transformation. The exploration of buried mineral deposits represents a frontier in the Mineral exploration (Zhao et al., 2008). In fact, regional geophysical anomalies are generally originated from the large and/or deeply buried geological bodies with low frequencies (e.g., sedimentary basin and big intrusions), while local geophysical anomalies are usually caused by the small and/or shallow geological bodies with high frequency (small intrusions, folds, and alteration mineralized bodies) (Wang et al., 2013). It is necessary for geophysical anomaly interpretation to decompose it into a series of constituents with different frequencies in accordance with the corresponding geological objects (Wang et al., 2000). Spatial and frequency analysis like the Fourier transformation has been widely used in decomposing geophysical data in past decades, yet the interpretation of geophysical anomalies face rigorous challenges because of the ambiguity of the interpretation derived from the heterogeneity of the geological bodies created during complicated geological processes (Wang et al., 2000). Even the same lithological units in different spatial locations can cause different geophysical anomalies, whereas different lithological units can cause similar geophysical fields. This non-unique correspondence can cause difficulties in inferring deep-seated geological structures and in delineating buried deep ore objects (Pan and Harries, 2000). The challenge causing by geophysical anomaly interpretation diversity is ultimately attributed to the nonlinear features of the geological bodies formed by a number of overlapped geologic processes (Cheng, 2008, 2006; Li and Cheng, 2004). The nonlinearity of geological bodies makes the traditional methods such as Fourier transform and geostatistics not suitable for describing such processes (Chen et al., 2006; Lovejoy et al., 2005; Huang et al., 1998). To tackle this challenge, Huang et al.(2014, 2008, 2005, 1998) developed an adaptive method for dealing with nonlinearity data and goes by the name of Hilbert-Huang transform (HHT) which consists of empirical mode decomposition (EMD) and Hilbert spectral analysis. Nunes et al.(2005, 2003) developed EMD for image texture analysis in 2D called the BEMD. However the BEMD is only suitable for handling continuous raster data in 2D but is not well suited for processing the scattered data in 2D like the geophysical and the geochemical exploration data. Thus some researchers (Chen et al., 2019, 2017; Tao et al., 2019; Zhao et al., 2016; Hou et al., 2012; Jian et al., 2012; Huang et al., 2010) developed a BEMD for handling scattered geophysical and geochemical data in 2D to extract their anomalies associated with the mineralization. It is possible to extract deeper geophysical anomalies related to the mineralization from their raw data by using nonlinear data processing techniques to construct more exact digital ore-finding pattern.
In previous studies, we applied successfully the bi-dimensional empirical mode decomposition (BEMD) to extract gravity and geochemical anomalies associated with mineralization from their original data (Chen et al., 2019, 2017; Jian et al., 2012; Huang et al., 2010). In this paper, BEMD is used for decomposing the aeromagnetic data surveyed with the scale of 1 : 200 000 at southeastern Yunnan, which covers three giant polymetallic ore regions including the Gejiu Sn-Cu, the Bozhushan Ag-Pb-Zn-W, and the Laojunshan W-Sn-Zn-In polymetallic deposits (Fig. 2). The extracted magnetic components can be used for inferring deeply buried Sn polymetallic mineralization and related granitic intrusions. A better result in mineral exploration has been made than the previous result from Zhao et al. (2016).
Figure 1. Sifting procedure of BEMD (revised after Nunes et al., 2005, 2003).
Figure 2. (a) Geological outline map of southeastern Asia, showing major tectonic units and location of southeastern Yunnan Sn-W polymetallic region is situated (Cheng et al., 2013); and (b) simplified geological and mineral map showing the spatial distribution of intrusive rocks and the various types of mineral deposits in the SE Yunnan (modified after 1 : 1 000 000 geological map from Yunnan Bureau of Geology and Mineral Resources, 1991). CB. Cathaysia Block; YB. Yangtze Block; SB. Sibumasu Block; TP. Tibet Plate; ICB. Indo-China Block; IP. India Plate.
-
Assume that Ori(m, n) is the original magnetic data set and can be decomposed into a finite number of bi-dimensional intrinsic mode functions (BIMFs).
From high frequency to low frequency, aero-magnetic data set, Ori(m, n) can be
where Bi(m, n) is the ith BIMF component (from high to low frequency, B1(m, n), B2(m, n), …, Bt(m, n)), and Res(m, n) is the residue. The BIMF1's frequency is higher than the other BIMFs' overall and everywhere. The different IMF components represent different oscillatory mode, while the residual component describe the overall trend of the data.
In the process of filtering, similar to the one-dimensional case, different filters can be designed and used for high pass, band pass and low pass filters respectively. Certain two-dimensional IMF components reflecting the characteristics of specific frequency (scale) structures can also be selected as filtering results.
Filters such as high-pass (SHP), band-pass (SBP) and low-pass (SLP) defined by Freire and Ulrych (1988) can be established in BEMD as following
Besides, we need to set the stop condition of the sifting process, which can be accomplished by limiting the size of the standard deviation, SD. SD can be calculated from the two consecutive sifting results as follow
The number of BIMFs will increase with the smaller SD. The 2D sifting process algorithm first proposed by Nunes et al.(2005, 2003) is used, and the envelope interpolation is improved in this study, and is shown in Fig. 1.
The orthogonality of the one dimensional IMFs was defined by Huang et al. (1998) using any two of the IMFs, Cf and Cg, by calculating the index of orthogonality (IO) according to the formula 6
In this study, two vector IMFs are considered. If they are orthogonal, the inner product should be zero (IMFi(t)·IMFj(t)=0). For BEMD, if the index of orthogonality (IO) value is less than 0.05, orthogonality is satisfied. Here, if an m×n matrix is treated as an m×n vector, the pseudo orthogonality for bi-dimensional data is defined as follow,
where BIMFi(p, q) and BIMFj(p, q) are BIMF components. This new definition of the IO will decrease the error caused by the shortage of data in a given direction.
-
The southeastern Yunnan is located at the western margin of the South China Block, which is well known in the world for its abundant Sn-W polymetallic mineralization. It is bounded by the NE-trending Mile-Shizong fault on the northwestern side where it is adjacent to the Yangtze Block. On the southwestern side, it is bounded by the NW-trending ASRR fault and adjacent to the Indo-China Block, while on the eastern side, it is bounded by the Malipo-Wenbshan fault. Besides, the southeastern Yunnan extends eastward to the Dachang-Damingshan W-Sn polymetallic mineralization region in Guangxi Province (Chen et al., 2017; Xu et al., 2015) (Fig. 2).
The exposed strata in this study area are mainly Proterozoic, Paleozoic and Mesozoic. The Proterozoic can be subdivided into the Paleo-Proterozoic and Neo-Proterozoic. The former covers the Dahongshan, the Ailaoshan and the Yaoshan group in the west part of the study area and so does the Mengdong Group in the east part of the study area. They are a set of migmatitic metamorphic rock series which consist of biotite granulitite, gneiss, marble intercalated amphibolite with strong to medium intensity of magnetism Their protolith is inferred to be mafic volcanic rocks and a group of clastic sedimentary formed in the active tectonic setting. The Neo-Proterozoic includes the Xinzhai Formation which is composed of schists, phyllites, and slates. Its protolith is inferred to be a set of clastic rocks containing littlemafic volcanic sediments (Wang et al., 2013). Paleozoic includes Cambrian–Middle Ordovician carbonate rocks and Devonian carbonate rocks, as well as Lower Permian clastic sedimentary rock and volcanic sedimentary rock with intercalated basalt. Mesozoic mainly includes the Meso–Triassic Farlang group clastic sedimentary formation and the Gejiu carbonate formation with intercalated basalt. Basalts display strong magnetism. In the eastern part of the study area, the main strata containing Sn-W and Pb-Zn-Cu polymetallic deposits are the Nanyangtian Formation of the Paleo-Proterozoic Mengdong Group and the Middle Cambrian Tianpeng Formation respectively (Wang et al., 2013). The main strata containing Sn-Cu polymetallic deposits in the western part of the study area is the Middle Triassic Gejiu Formation.
Regionally, in addition to the Red River fault, Mile-Shizong fault and Wenshan-Malipo fault act as a boundary between different blocks within the study area. The major ore-controlling faults in the study area include also the Xiaojiang-Gejiu fault and the Pingbian-Mengzhi fault.
Strong tectonic magmatic activity took place during the Late Cretaceous forming the Gejiu complex (77–87 Ma, Chen et al., 2020), the Bozhushan granitic intrusion (86 to 91 Ma, Zhang et al., 2016) and the Laojunshan granitic intrusions (85 to 95 Ma, Yang et al., 2020; Zhao et al., 2018) and is related to the Gejiu Sn-Cu (Chen et al., 2020; Cheng et al., 2013), the Bozhushan Ag-Pb-Zn-W (Zhang et al., 2016; Liu et al., 2007), and the Laojunshan W-Sn-Zn-In giant polymetallic deposits (Wang et al., 2017, 2013; Xu et al., 2015).
The giant Gejiu Sn-Cu polymetallic deposits is located at the western margin of the South China Block (Fig. 2). The ore deposits in the Gejiu district exist mostly within the Mid-Triassic Gejiu Formation carbonates. Skarn alteration mineralization were intensively developed in the contact zone between the granites and the carbonates. Skarn-sulfide ores are characterized by disseminated, veinlet and massive structures. The major ore minerals include arsenopyrite, pyrrhotite, cassiterite, chalcopyrite, marmatite, pyrite, scheelite, molybdenite and magnetite (Chen et al., 2020, 2017; Cheng et al., 2013).
The Bozhushan ore field is located at the western margin of the South China Block, about 150 km west to the giant Gejiu giant tin polymetallic ore field (Fig. 2). There are two main types of ore deposits in the Bozhushan district, and the Ag-Pb-Zn-W-Fe polymetallic mineralizations in the ore field is associated with Late Cretaceous granites (Chen et al., 2015): one is the Guanfang skarn W-Fe polymetallic deposit, and other is the Bainiuchang Ag-Pb-Zn polymetallic deposit. The former is situated at the southeastern side of the Bozhushan granitic intrusion, and formed in the contact zone between the granitic intrusion and the Paleozoic carbonate rocks (Fig. 2). The main ore minerals include scheelite, magnetite, pyrrhotite, pyrite, chalcopyrite, galena, sphalerite, arsenopyrite and a small amount of ilmenite, siderite and molybdenite (Zhang et al., 2016). Bainiuchang Ag-Pb-Zn polymetallic deposit related to the deep concealed granite is located at northwestern side of the Bozhushan granitic intrusion. The orebodies in lamellar, lenticular and vein mainly occur in the carbonate rocks of Tianpeng Formation of the Middle Cambrian. Ore minerals include mainly pyrite, pyrrhotite, sphalerite, cassiterite, arsenopyrite, galena, and siderite. Silver minerals cover mainly freibergit, antimonite, pyrargyrite, argentite and polyargyrite (Liu et al., 2007).
The Laojunshan W-Sn-Zn(In) polymetallic ore deposits are located at the southeastern part of the study area (Fig. 2). There are two types of significant ore deposits which are spatial-temporally associated with the Laojunshan granitic intrusion: (a) Dulong skarn Sn-dominant polymetallic deposit and (b) Nanyangtian stratabound W-Sn polymetallic deposit. The former situated at the south side of the Laojunshan intrusion is a typical skarn deposit related to the Late Cretaceous granitic intrusion (He et al., 2014; Zhang et al., 2006). Most of the ore bodies occur in the middle and lower carbonate rocks of the Middle Cambrian Tianpeng Formation. The ore minerals mainly include cassiterite, tetrahedrite, marmatite, chalcopyrite, pyrrhotite, pyrite, arsenopyrite and magnetite. The Nanwenhe W-Sn polymetallic deposit is located on the east side of Laojunshan granitic intrusion, and the orebodies are hosted within the stratiform-like skarn and marble in the Paleoproterozoic Nanyangtian Formation (Wang et al., 2020; Wang et al., 2017; Xu et al., 2015). Ore minerals include scheelite, pyrrhotite, pyrite, chalcopyrite, sphalerite, galena, arsenopyrite, molybdenum, etc. Although these ore deposits were produced by a set of complicated superimposed multi-stage of mineralization events, the Late Cretaceous granitic intrusions played a key role in the polymetallic mineralization (Yang et al., 2020; Wang et al., 2017; Zhang et al., 2006; Yang and Yan, 1994).
These magnetic minerals like magnetite and pyrrhotine developed in the above ore deposits lay a foundation for the magnetic exploration.
-
In this study area, the magnetic susceptibility of the Paleo-Proterozoic metamorphic basement rocks containing mafic lava and Permian and Triassic basalts reaches up to 0.005 03–0.289 03SI, which cause medium to strong magnetic anomaly. The susceptibility of the sedimentary rocks dominated by carbonates closes to 0.00 SI, which can not cause magnetic anomalies. Granitic rocks with low of susceptibility of 0.001 32 to 0.039 6SI may cause low-gradient magnetic anomaly. The magnetite-bearing skarn rocks with susceptibility of 0.062 8–11.523 4SI, which can cause distinctive magnetic anomalies (Xiong and Shi, 1994).
Aeromagnetic data are available for this study with a 2 km spatial resolution (ΔT) in this study, and cover an area including the Gejiu Sn-Cu polymetallic deposits, the Bozhushan Ag-Pb-Zn-W-Fe polymetallic deposits, and the Laojunshan W-Sn-Pb-Zn-In polymetallic deposits (Fig. 3) and is provided by the Yunnan Geological Survey. The aero-magnetic anomaly image (Fig. 3) is obtained by the reduction-to-the-pole of the ΔT magnetic anomalous data (magnetic dip: -1.65°, magnetic inclination: 35.23°).
Figure 3. The image of reduction-to-the-pole of the ΔT magnetic anomaly in the SE Yunnan. Pt. Proterozoic; Pz1. Lower Paleozoic; Pz2. Upper Paleozoic; Mz. Mesozoic; Kz. Cenozoic.
To effectively use the geophysical field, like the magnetic field, for the mineral exploration, it is necessary to establish a set of signatures that characterize shape, size and depths, as well as composition of various geological objects and their relationship to mineralization (Pan and Harries, 2000). As far as magnetic fields are concerned, a high magnetic measured value indicates the presence of the geological objects with higher average magnetic field intensity than the materials surrounding them. Conversely, a low magnetic measured value indicates the presence of the geological bodies with relatively lower average magnetic intensity than the materials surrounding them. The scale of magnetic anomalies is related not only to the size, but also to the shape as well as the depth of geological bodies. Decomposition of magnetic fields is crucial for investigation of specific geological structures and geological bodies such as faults, concealed granites and the basement. According to mentioned above magnetism features of the matamorphic basements, sedimentary rocks, granitic rocks and skarn mineralization a typical magnetic criterion for recognition of the granitic intrusions related to mineralizayion is the low gradient magnetic anomalies. The pattern with high magnetic anomalies around the round shape areas with low magnetic anomalies may be caused by skarn mineralization within contact zones between the granitic intrusions and their host rocks (Xiong et al., 1994).
2.1. Mineralization
2.2. Aeromagnetic Datasets
-
The ΔT magnetic anomaly data was decomposed into four BIMFs and one residue Res(m, n) by the BEMD with the stepwise criterion of stoppage (SD=0.02). According to the Eq. 8 using the sifting process of the BEMD described above
where Ori(m, n) are the original 2D aeromagnetic data; Bi(m, n) are the 2D IMFs and Res(m, n) are the 2D residual component. The mixed aeromagnetic data in two dimensions in the study area can be decomposed into four intrinsic mode functions BIMF1, BIMF2, BIMF3, and BIMF4, which decrease in frequency and a residue with the lowest frequency.
The orthogonality of the BIMFs was checked (Table 1). The results show that except for IO34 and IO4R being obviously bigger than zero, other values are close to zero and thus orthogonality is approximately satisfied.
IO12 IO13 IO14 IO1R IO23 IO24 IO2R IO34 IO3R IO4R 0.13 0.044 0.009 0.035 0.076 0.098 0.018 0.425 0.008 0.182 Table 1. Orthogonality assessment
-
Because decomposed four BIMFs (BIMF1, BIMF2, BIMF3, and BIMF4) and one residue Res(m, n) from the original ΔT magnetic anomaly data by BEMD decrease in frequency within the same range, the BIMF1 represents the magnetic image with the highest frequency, and Res(m, n) represents the magnetic image with the lowest frequency. The frequency of the BIMF2, the BIMF3 and the BIMF4 images ranges between that of the BIMF1 and the Res(m, n).
The above-mentioned BIMFs and Res(m, n) have distinct geological implications. The low-pass filtered magnetic component image (Res(m, n)) (Fig. 4) depicts the deep geological architecture within the study area, as suggested by Hou et al. (2012). Taking 0 nT as a boundary in the residual component map (Fig. 4) the magnetic anomaly can be divided into a positive and a negative magnetic anomaly zone (Ⅰ and Ⅱ). It is inferred that the zone Ⅰ with the positive magnetic anomaly represents the spatial distribution of the Yangtze Block with high grade metamorphic basement, which consists of biotite plagioclase gneiss, hornblende gneiss, plagioclase amphibole, mica quartz schist and calc-silicate granulite and marble (Zhu et al., 2001).
Figure 4. Image of the magnetic anomaly reconstructed from its residues component in the SE Yunnan district. Pt. Proterozoic; Pz1. Lower Paleozoic; Pz2. Upper Paleozoic; Mz. Mesozoic; Kz. Cenozoic.
The protolith of the plagioclase amphibole is tholeiite basalt, while the protolith of the intermediate acid gneiss is calc-alkali igneous rocks (Zhai et al., 1990). The magnetic anomaly intensity decreases from the west to the east, which implies that the ferruginous mass contained by the high grade metamorphic basement is gradually reduced from the west to the east. The zone Ⅱ with the negative magnetic anomaly reflects the spatial distribution of the Cathaysia Block basement with less or no magnetism, which consists of a suit of low grade metamorphic flysch and carbonate formation with intercalated felsic rocks with less or no ferruginous rock mass (Zhao, 1999; Zhao et al., 1995).
Because the orthogonal coefficient between BIMF3 and BIMF4 (0.425) is obviously bigger than 0, it indicates that there is a significant correlation between them. Thus, we define the BIMF3+BIMF4 component as a band-pass filtering image that is shown in Fig. 5, which depicts the middle-lower geological architecture. There are three positive magnetic anomalies (Ⅰ+1, Ⅰ+2, Ⅰ+3) around a negative anomaly (Ⅰ-1) in the Yangtze Block. The Ⅰ-1 negative magnetic anomaly coincides with the Mesozoic-Cenozoic Jianshui-Shiping sedimentary basin (Fig. 5). The positive magnetic anomalies (Ⅰ+1) and (Ⅰ+2) correspond with the Ailaoshan Group, while the (Ⅰ+3) does with the Dahongshan Group, which consist of high the grade metamorphosed basic volcanic rocks and volcanic-sedimentary rocks (Yang et al., 2014; Zhu et al., 2001). In the Cathaysia Block there are one negative magnetic anomaly (Ⅱ-1) and six positive magnetic anomalies (Ⅱ+1, Ⅱ+2, Ⅱ+3, Ⅱ+4, Ⅱ+5, Ⅱ+6). The negative magnetic anomaly (Ⅱ-1) spatially may coincide with the deep buried granitic intrusion where the Gejiu giant Sn-Cu polymetallic deposits were produced. The positive magnetic anomalies (Ⅱ+1, Ⅱ+2, Ⅱ+3, Ⅱ+4) coincide with the Permian and/or the Triassic basalts, the positive magnetic anomaly (Ⅱ+5) coincide Cenozoic basalts from Maguan and Pingbian area (Huang et al., 2013), while the positive magnetic anomaly (Ⅱ+6) may coincide with Indosinian basic valcanic ricks and intrusive ricks which were developed in the Youjiang rift during Early Triassic to Meso–Triassic (Liu et al., 1986).
Figure 5. Image of Southeast Yunnan magnetic anomaly reconstructed from the BIMF3+BIMF4 component. Pt. Proterozoic; Pz1. Lower Paleozoic; Pz2. Upper Paleozoic; Mz. Mesozoic; Kz. Cenozoic.
The high-pass filtered IMF1 image (Fig. 6) depicts the shallow geological architecture in the study area. In the Yangtze Block, the three negative magnetic anomalies correspond in turn to the Shiping Mesozoic–Cenozoic sedimentary basin (Ⅰ-1), the outcropped granites (Ⅰ-2) in the western Gejiu ore field and the buried granites (Ⅰ-3) in the eastern Gejiu ore field. Among these, two negative magnetic anomalies (Ⅰ-2 and Ⅰ-3) are bracketed by the approximate ring positive magnetic anomalies (Ⅰ+1, Ⅰ+2) (Fig. 6). The positive magnetic anomalies are conjectured to be created from the skarnization alteration with Sn-Cu mineralization, which is similar to the one inferred by the gravity anomaly component (Chen et al., 2017). In the Cathaysian Block, there are three negative magnetic anomalies (Ⅱ-1, Ⅱ-2, Ⅱ-3) bracketed by the approximate ring positive magnetic anomalies (Ⅰ+1, Ⅰ+2) (Fig. 6) respectively. The Ⅱ-2 negative magnetic anomaly with the ring positive magnetic anomaly corresponds to the Bozhushan Ag-Pb-Zn-W polymetallic mineralization. The Ⅱ-1 negative magnetic anomaly sharing the ring positive magnetic anomaly with the Ⅱ-2 magnetic anomaly is conjectured to be created by the buried granites, which could be a new area for exploring for Ag-Pb-Zn polymetallic deposits in the Bozhushan district. Western half part of the ring positive magnetic anomaly (Ⅱ+2) around the Ⅱ-3 negative magnetic anomaly is inferred to be the second new ore-forming target area of Cu-Pb-Zn polymetallic deposits where a series of lead-zinc occurrences occur. The Ⅱ-3 negative magnetic anomaly is conjectured to be created by the Paleozoic sendimentary basin dominated by carbonates.
-
It has been shown that BEMD can be effectively used for extracting magnetic components with different wavelengths reflecting the geological architectures at different depths associated with the Sn-W-Pb-Zn polymetallic mineralization and the related intrusions in the study area. The anomalous magnetic components are extracted by decomposing the original magnetic data into different bi-dimensional intrinsic mode factions (BIMFs). These anomalous magnetic components decomposed by BEMD have the following distinct geological implications.
(a) The high-pass filtered magnetic component image (BIMF1) depicts the shallow geological architecture within the SE Yunnan district. The image (BIMF1) indicates that the skarnization alteration with Sn-W mineralization displays positive magnetic anomalies around the granites with negative magnetic anomalies, which can be defined as a pattern for identifying new ore-finding targets of Sn-W polymetallic deposits. The Ⅱ-1 negative magnetic anomaly sharing the ring positive magnetic anomaly (Ⅱ+1) with the Ⅱ-2 negative magnetic anomaly is inferred to be one new ore-forming target for exploring Ag-Pb-Zn polymetallic deposits in the Bozhushan district. The the western half part of the Ⅱ+2 ring positive magnetic anomaly distributed around the Ⅱ-3 negative magnetic anomaly is conjectured to be another new ore-finding target for exploring Pb-Zn-Ag polymetallic deposits in the Laojunshan district.
(b) The band-pass filtered magnetic component image (BIMF2+BIMF3) depicts the middle-shallow geological architecture, which indicates that the two big negative magnetic anomalies spatially correspond to the Mesozoic–Cenozoic sedimentary basin and the Gejiu granitic complex respectively. The six small positive magnetic anomalies coincide with the Permian, Early Triassic to Meso–Triassic and the Cenozoic basalts in turn.
(c) The low-pass filtered magnetic component image (the Res(m, n)) depicts the lowest geological architecture within the study area, which indicates that the Yangtze Block has a strong magnetic basement characterized by high grade of metamorphic basic igneous rocks, while the Cathaysia Block does a weak magnetic basement characterized by different grade of metamorphic sendimentary rocks containing Intermediate-acid igneous rocks.