There are several geological structural units between the STDS and the YD. The typical structure is the CR, which is oriented north to south. The CR is situated at 92°E and stretches from Sangri County in the north to Cuona County in the south along a total length of about 220 km and an average width of 5–15 km (Wu et al., 2008b; Zeng et al., 2009; Xie et al., 2017). From the south to the north, it crosses the Cuona Graben and the Qiongduojiang Graben (Figure 1C). The Qiongduojiang graben cuts through the eastern part of the YD (Wu et al., 2007a; 2008a; Zeng et al., 2011). A canyon, with a deep basin base of 200–300 m, and a wide valley bottom in the south, occurs north of the Qiongduojiang Graben at an altitude of about 4,400–4,500 m. Snowy mountain ridges border its eastern and western sides. These mountains extend in the northern and southern directions. The Cuona graben is a wide valley graben with gently undulating hills and valleys at an elevation of about 4,600–4,800 m. Its width varies greatly, ranging from about 10 km at its widest point to approximately 1 km at its narrowest. It is located in the Tethys Himalayan block, but passes through the STDS in the south and extends into the high Himalayan block. The Cuona Graben and the Qiongduojiang Graben cut through the east-west oriented structural belt and thrust fault zone, where sediments of varying thicknesses are deposited (Wu et al., 2007a; 2007b; 2008a; Ha et al., 2018).

Several thrust faults, including the Qiongduojiang fault (QDJF), the Rongbu-Gudui fault (RGF), the Zhegucuo-longzi fault (ZLF), and the Lhozhag fault (LZF) are developed in the study area (Figure 1C). Two essential and geologically complex dome structures, the CD and YD, also occur in the study area. The CD is a newly discovered dome structure located close to the CR and the STDS. Granites of different ages are developed in their core, while metamorphic rocks occur in the eastern part of the dome (Luo et al., 2020). Several rare metal mines occur in this area (Li GM et al., 2021; Li HL et al., 2021; Liang et al., 2021; Fan et al., 2021; Zhang et al., 2022; Xia et al., 2022). The YD is situated in the central extension area of the CR (Zhang et al., 2007) and forms the easternmost part of the northern Himalayan dome belt. Similar to the CD, the YD is also composed of gneiss and leucogranite, with the CR cutting through the YD (Zeng et al., 2009). The STDS is located in the southernmost part of the study area, in southern Tibet in the northern Himalayas. The STDS is composed of weakly metamorphic or non-metamorphic sedimentary rocks, with strong deformation in some areas (Burchfiel et al., 1992; Yin et al., 2006).

The Institute of Geology of the Chinese Academy of Geological Sciences laid a nearly south-north oriented deep reflection seismic profile between the STDS and the YD from 2018 to 2019. This profile passed through the STDS, CR, CD, ZOCA, and YD and had a total length of about 112 km. A total of 478 single seismic shots with a shot interval of 250 m and a trace interval of 50 m were collected. Every single shot has 720 traces, and the sampling rate is 2 ms. Table 1 shows the specific seismic acquisition parameters. From the original single shot record (Figure 2), it can be seen that the surface wave signal is evident in both the far-offset (green circle) and the near-offset (red circle). Therefore, the surface wave signal can be fully extracted and used to obtain the shallow shear wave velocity structure through inversion. After removing the poor signal-to-noise ratio data, it retained 453 dispersion curves.

This study uses the multichannel surface wave data processing method, extracts the surface wave signal from the deep reflection data, and obtains the S-wave velocity structure by inversion (Figure 3). Currently, multi-surface wave imaging technology is primarily used for urban geological exploration, with a particular emphasis on shallow layer imaging (Park et al., 1999; Xia et al., 1999, 2015; Yin et al., 2018; Andajani et al., 2019). There are relatively few studies on the application of surface wave extraction from deep reflection data. Compared to the reflection data during petroleum seismic exploration, the deep reflection data collected by explosive sources has a more intense surface wave signal and deeper detection depth. This is beneficial for inversion imaging. Before data processing, it is necessary to intercept the original data. Single-shot data can only obtain the shear wave velocity information of a specific underground position in multichannel wave imaging. Therefore, the direct use of 720 channels of a deep reflection single shot data does not accurately reflect the information of a specific underground location. When there are a limited number of receiving channels, the mapping resolution of dispersion energy will be poorer, complicating the extraction of the dispersion curve. This study selected the most appropriate 60 channels bilateral for single-shot dispersion extraction. The phase shift method was applied to transform the single shot record into a dispersion energy diagram. Finally, based on the dispersion curve, the one-dimensional S-wave velocity structure was obtained by inversion, and the two-dimensional profile was derived by interpolation.

Spectrum analyses were performed on the single-shot records of five geological units: STDS, CR, CD, ZOCA, and YD (Figure 4). Based on the frequency spectrum, single-shot energy is primarily concentrated between 0 and 50 Hz, with a main frequency of 10 Hz. The amplitude values of the single-shot records near the STDS (No. 9275), the CR (No. 9863), and the CD (No. 10194) are small. This is due to the proximity of these points to faults and rifts, which contain many Quaternary sediments. These sediments cause the rapid attenuation of seismic wave energy. The amplitude values of the single-shot near the ZOCA (No.10599) and the YD (No.11260) are large because the rock mass integrity is better. The lithology of the ZOCA and the YD consists of leucogranite and metamorphic rocks with a high density; therefore, the seismic energy waves weaken slowly. The amplitude spectrum of the ZOCA varies significantly with frequency and contains many peaks. This can be due to the complex geological structure and diverse lithological types in the mining area (Guo et al., 2019; Jiao et al., 2019).

This paper selects the fundamental dispersion curve according to the maximum value in the dispersion energy diagram. The single-shot dispersion curves indicate that the frequency is maintained between 1 and 10 Hz, and the phase velocity is between 500 and 2,500 m/s. The ratio of the energy of the measuring point to the maximum energy at this frequency is 1. It considers the signal-to-noise ratio of the picked dispersion curve is the best. This ensures that the chosen fundamental dispersion value is the maximum. As shown in Figure 5, the single shot dispersion in different research areas demonstrates that the corresponding signal-to-noise ratios of various signal frequencies are close to 1, resulting in more reliable inversion findings.

The dispersion curves of specific geological units, including the STDS, CR, ZOCA, and CD, were analyzed (Figure 6). The overall phase velocity near STDS and CR is less than 1,500 m/s, which can be attributed to the presence of Quaternary sediments and weathered and broken rocks between the fault and rift. These rocks cause low S-wave velocities. The dispersion near the CD and the ZOCA has a high phase velocity of 2000 m/s and 2,500 m/s, respectively, which can be due to granite and metamorphic rocks in these two areas (Jiao et al., 2019). The dispersion curve of ZOCA shows a significant curved bulge near 3 Hz, demonstrating certain high-low velocity anomalies underground. These anomalies can be attributable to the change of ore veins or the mining goaf (Jiao et al., 2019; Fan et al., 2021; Wu et al., 2021; Zhang et al., 2022). Analysis of the dispersion curves at different positions can reveal the formation-related details.

A reasonable initial model is necessary to provide proper constraints for inversion. The regional velocity changes greatly as the deep reflection profile passes through several geological units. Therefore, it is necessary to establish different initial models for the positions. The phase velocity of the fundamental Rayleigh surface wave is the most sensitive to the S-wave velocity at a depth of 1/3 of its wavelength. The relationship between the phase velocity and S-wave velocity satisfies formula C=0.92 Vs. in a uniform half-space Poisson medium. If the measured Rayleigh wave phase velocity is divided by 0.92 to obtain the S-wave velocity, the depth is 1/3 of its wavelength, which is a suitable initial model (Xia et al., 1999; Luo et al., 2008). Therefore, this paper used the dispersion curve measured by different single-shot records and calculated the S-wave velocity corresponding to different frequencies using the above formula. Then, based on various frequencies, the wavelength corresponding to different frequencies was calculated using the relationship between wave velocity and wavelength, and 1/3 of its wavelength was taken as the inversion depth. S-wave velocities at various depths can be obtained through the calculation presented above. Finally, the above S-wave velocity variation with depth was utilized as the initial model for inversion (Figure 7, red dotted line). The initial model was divided into ten layers, from shallow to deep, based on the principle of “shallow subdivision and deep coarsening”.

This study used open-source software to inverse the dispersion curve with the initial model (Herrmann, 2013). In order to ensure the reliability of the results, the inversion result error is greater than 5%, which will be discarded. Moreover, it carried out 30 iterative inversions (Figures 7B, D, F). Based on the inversion results, the S-wave velocity of the CR (No. 9863) changes greatly in the depth range of 0–200 m and slightly in the depth range of 200–600 m (Figure 7A), which is less than 1,500–1700 m/s, indicating the low-velocity characteristics between the rift and the fault. The S-wave velocity increases in a “trapezoidal” trend within 300 m of the CD (No.10194), reflecting the sediment’s characteristics around the CD (Figure 7C). At 300 m, the velocity suddenly increases and stays at about 2000 m/s, indicating the intrusive leucogranite body (Jiao et al., 2019). The velocity near the ZOCA (No.10599) is generally high and alternates greatly between ‘high’ and ‘low’ velocities, which can be related to mineralization (Figure 7E). The extension of the vein and the mined-out area will affect the S-wave velocity, resulting in alternating high and low velocities. Therefore, the velocity changes significantly from shallow to deep in the one-dimensional S-wave profile. The actual change in strata characteristics can be obtained from the inversion results based on the various initial models established at different points.

Numerous peaks with a continuous elevation of over 5,500 m between STDS and YD are accompanied by many rivers. The fluvio-lacustrine and glacial sediments provide provenance filling for the CR and its surroundings, forming a certain thickness of sediments. Previous studies indicate that the sedimentary thickness of the Qiongduojiang graben ranges from 400 to 600 m, while the Cuona graben is at least 300 m thick (Wu et al., 2007b; 2008b). The S-wave velocity profile demonstrates that the average thickness of the sediment layer between the STDS and the YD is 400–500 m (Figure 8B, black dotted line). The STDS, LZF, ZLF, and RGF, can reach sedimentary thicknesses of up to 800 km or deeper (Wu et al., 2007b; 2008b). The sediment thickness is the thinnest near the ZOCA. The average sediment thickness of the survey line is about 600 m since it passes through the western edge of the CD close to the CR. The sediment thickness of YD in the north is less than that near CD, at about 400 m thick.

The east-west extensions of the Tibetan Plateau have formed a series of north-south rifts in southern Tibet since the Cenozoic (Armijo et al., 1986; Tapponnier et al., 2001; Hou et al., 2006; Zhang et al., 2007; Xu et al., 2006; Yin, 2006; Zhang., 2007a, 2007b; Zhang et al., 2012; Xue et al., 2021). A high-conductivity and low-velocity layer, which is interpreted as a partial melting layer in the middle and lower crust, is widely distributed under these north-south trending rifts (Unsworth et al., 2004, 2005; Nabelek et al., 2009; Jin et al., 2010; Xie et al., 2017; Liang et al., 2018; Pang et al., 2018; Xue et al., 2021). Over time, some of these melts migrated upwards in the CR through the rock boundaries and weak zones (Hill et al., 2015). These melts condensed and crystallized to form leucogranite when they reached a position close to the surface.

The Himalayan leucogranites mainly occur along two belts. The one leucogranite occurs near the STDS is called the high Himalayan leucogranite. The second leucogranite occurs in the Tethys Himalayan belt and is called the Tethys Himalayan leucogranite (Wu et al., 2015; Zeng and Gao., 2017; Zhang et al., 2018; Gao et al., 2019; Cao et al., 2022). Jiao (2019) used gravity, electrical, and magnetic methods to image the three-dimensional structure of the ZOCA and its adjacent area. Based on this imaging, he postulated an intrusive body of leucogranite under the CD and the ZOCA. It also found discontinuous high-velocity anomalies below the study area based on the S-wave velocity profile (Figure 8B). These anomalies can represent the intrusive body of leucogranite for different geological periods (Guo et al., 2019; Jiao et al., 2019). The high-velocity anomaly is divided into five areas: A, B, C, D, and E. The high-velocity anomalies in areas A and B are close to the STDS and belong to the high Himalayan terrane and the Tethys Himalayan terrane, respectively. These anomalies were caused by the upwelling of deep thermal materials along the STDS (Jin et al., 2010; Hill et al., 2015; Xie et al., 2017), which formed the leucogranite located in the high Himalayan metamorphic rock series and the Tethys Himalayan sedimentary rock series. Area C is about 1 km underground and located along the western boundary of the CD, consistent with the gravity inversion results obtained by Jiao (2019). The high-velocity anomaly in Area C is connected with Area D, indicating that the ZOCA and the CD can have the same source of deep material (Jiao et al., 2019). Due to the existence of several secondary faults around ZOCA. The ZOCA has higher contact between the high-velocity body and the shallow surface than the CD, which can cause reactions between the magma and its surroundings as it rises to the surface. These interactions can develop different metal deposit types (Liang et al., 2014; Guo et al., 2019; Jiao et al., 2019; Zhang et al., 2022). The high-velocity anomaly in Area E is located under the south-dipping YD. The deep main body of the anomaly extends more widely to the south and in the same direction as the Qiongduojiang graben.

The deep melting magma source area of the lower crust and the local low-velocity melting of the middle crust were formed during the collision and compression of Eurasian plates. Magma pockets and local melting that formed at different depths of the upper crust during the tectonic evolution of the area display a unique crust-mantle thermal structure in the Tibetan Plateau and its surrounding areas (Zhang et al., 2014; Liu et al., 2014; Wang et al., 2017; Wang et al., 2022; 2022b). Several faults developed between the STDS and the YD, with high- and medium-temperature geothermal systems existing in the intersections of these faults. The deeper CR and STDS have a larger fracture zone, which can conduct heat and water and serve as a high-quality water storage structure. In the shallow surface, Jurassic and Triassic metamorphic sandstone and quaternary sandstone (Figure 8A) serve as natural heat storage layers (Wang et al., 2022). These conditions between STDS and YD have good geothermal characteristics and promote mineralization.

The thermal conductivity was determined at different depths below the survey line using the relation of S-wave velocity and the thermal conductivity in experimental petrology (Tian et al., 2020). The average thermal conductivity along the survey line direction was then calculated using the thermal conductivities of different depths. The thermal conductivity reveals that the CGF, the GGF, and the ZOCA have high values and a more intense thermal radiation capacity (Figure 9). The S-wave velocity profile (Figure 8B) reveals a continuous high-velocity body under the CD and ZOCA and a low-velocity channel above the high-velocity body (Figure 8B, red dotted line). While the high geothermal gradient enhances the circulation of fluid, the low-speed channel between ZOCA and CD provides favorable conditions for the migration and reservoir of ore-forming fluid.

The partially molten body in the deep part of the Tethys Himalaya (Brown et al., 1996; Nelson et al., 1996; Wei et al., 2001; Guo et al., 2019) increases the geothermal gradient of the region when it invades upward. More heat is radiated to the shallow geological layers under high thermal conductivity conditions, which drives geothermal circulation and causes the circulation of the underground ore-forming fluid in the CD, ZOCA, and the surrounding fracture system (Zhou et al., 2017). Useful ore-forming elements (Au, Sb, Pb, and Zn) are gradually extracted from granitic rock, surrounding rock strata, and basic dikes to form ore-rich fluid. The circulating fluid mixes with the infiltrated atmospheric water and then precipitates, causing elements such as lead, antimony, and silver in the fluid to form sulfide mineral deposits, such as the Zhaxikang antimony, lead, zinc, and silver polymetallic deposits (Cox, 2007).

In geothermal fields, the deep molten magma migrates upward to heat the upper and middle crust. Atmospheric precipitation and meltwater that permeate downward along the fault zone move upward along the fissure after being heated by the local underground molten material and hot fluids (Wang et al., 2017; Wang et al., 2020a; Wang et al., 2020b; Wang et al., 2022). The S-wave velocity profile contains three low-velocity channels. These low-velocity channels are located in the STDS, CR, and RGF (Figure 8B) and are connected with deep heat sources. The heat radiation is enhanced by the high thermal conductivity (Figure 9), increasing the temperature of the superficial fluid. The CGF and the GGF formed when the high-temperature fluid was exposed.