Soil Degassing From the Xianshuihe–Xiaojiang Fault System at the Eastern Boundary of the Chuan–Dian Rhombic Block, Southwest China

The Chuan–Dian region, situated in the middle part of the north-south seismic zone of mainland China in a highly deformed area of the eastern margin of the Tibetan Plateau, is one of the principal areas for monitoring earthquake activities in China. In this study, the geochemical characteristics of soil degassing (of CH4, H2, CO2, Rn, and Hg) and, the relationship between degassing and fault activity, were investigated in the Xianshuihe–Xiaojiang fault system (XXFS) at the eastern boundary of the Chuan–Dian rhombic block. The mean soil-gas concentrations of CH4, H2, CO2, Rn, and Hg in the XXFS were 8.1 ppm, 9.9 ppm, 0.5%, 15.1 kBq/m3 and 12.9 ng/m3, respectively. The δ13CCO2 and δ13CCH4 values of the hot-spring gases varied from −11.9‰ to −3.7‰ and −62.5‰ to 17‰, respectively. The He-C isotopic ratios indicate that the carbon in the northern and middle parts of the XXFS may have originated from deep fluids, whereas the carbon in the southern part of the XXFS is of organic origin. The high concentrations of soil gas were distributed near the faults, indicating that the faults could act as channels for gas migration. The distributions of the high soil-gas concentrations in the XXFS coincide with the highest stress and maximum strain rates, indicating that the fault activity enhanced permeability and increased the emission rates of the gases. The results of this study will be helpful for degassing in active fault zones and earthquake monitoring.

Northward subduction of the Indian plate beneath the Tibetan Plateau makes south-eastern Tibet an important domain involved in the convergence and interaction between the Eurasian and Indian plates ( Figure 1A, Tapponnier et al., 2001). The Chuan-Dian region, in the southern part of the north-south seismic zone of mainland China, is tectonically located in the highly deformed area of the eastern Tibetan Plateau. Enclosed by the Xianshuihe-Xiaojiang fault system (XXFS), the Red River fault, and the Jinshajiang fault ( Figure 1B), the Chuan-Dian rhombic block is one of the principal areas of earthquake monitoring in China.
The XXFS is located at the eastern boundary of the Chuan-Dian rhombic block ( Figure 1B) and has a curved left-lateral strike-slip structure extending at least 1,400 km into the eastern margin of the Tibetan Plateau (Allen et al., 1991). Fieldwork confirms that the XXFS is one of the largest and most seismically active fault systems in China. The XXFS has experienced at least 35 earthquakes of M > 6 in the last 300 years, and almost all segments of the system have been the locus of major earthquakes within the historical record (BGMRSP, 1993; Figure 1B).
Previous studies primarily focused on the soil gas and hotspring gas at the western boundary of the Chuan-Dian rhombic block (e.g. Li et al., 2018;Zhou et al., 2020). However, at the eastern boundary of the rhombic block, only the geochemical characteristics of several gas species (CO 2 , Rn, and Hg) in the middle part of the XXFS have been reported (Zhang et al., 2019a;Zhange, 2019b). So far, systematic investigation of multi-gas components in soil gases at the eastern boundary of the rhombic block has rarely been reported. In this study, the geochemical characteristics of soil degassing of CH 4 , H 2 , CO 2 , Rn, and Hg and the relationship between degassing and fault activity in the XXFS were reported.
The NW-striking XSH fault is a highly active strike-slip fault in the eastern margin of the Tibetan Plateau that begins in Luding, passes through Kangding and Luhuo, and extends into Qinghai Province ( Figure 1B). The XSH fault has a long active history. The regional uplift tectonic motion of the Tibetan Plateau in the Neogene produced the XSH NW-NNW striking left-lateral slip fault that cuts through the sialsphere, producing the belts of intensive faults and cleavage. The XSH fault is very active and is accompanied by many destructive earthquakes, such as the M 7.9 earthquake that occurred in Luhuo in 1973(Figure 1B, BGMRSP, 1993. The ANH fault starts from the town of Jintang, passes through Mianning, Dechang, and Huili, and extends into Yunnan Province. The ANH fault is multi-segmented. The northern, middle, and southern segments are separated at Shimian and Dechang. In this study, the northern and middle segments of the ANH fault were selected for targeting. The northern segment of the ANH fault is characterized by thrust and ductile shear, whereas the middle segment is characterized by deep faulting. Seismic activities in these segments are severe, with many M ≥ 6 earthquakes occurring (BGMRSP, 1993).
The ZMH NW-striking fault conjoins with the ANH fault at Xichang and is currently active (BGMRSP, 1993). The basement of the ZMH fault primarily consists of metamorphic rocks derived from granite, rhyolite, and basalt (Ren et al., 2010). Quaternary alluvial sediments, which are primarily composed of sandy gravel, fine sand, silt, and clay, are widely distributed in the study area (Ren et al., 2010).
The XJ fault is a zone of intense seismic activity that starts from Zhaojue in Sichuan Province and terminates in Yunnan Province. The fault divides into two branches south of the city of Dongchuan ( Figure 1B). A wide crushed zone formed along the fault. The XJ fault is distinctly currently active and is characterized by frequent seismic activities and many hot springs (BGMRSP, 1993).

Measurement Method of Soil Gas
The soil-gas measurements in the XXFS were conducted on 1-13 March 2019 to minimize the effect of weather factors. Ten measurement sites (M1 to M10 in Figure 1B) along the XXFS were selected for measurement. The measurement sites of M1-M3, M4-M5, M6-M7, and M8-M10 were in the XSH, ANH, ZMH, and XJ faults, respectively ( Figure 1B). M8 and M10 were in the western branch of the XJ fault, whereas M9 was in the eastern branch of the XJ fault ( Figure 1B).
Four parallel survey lines with an interval of approximately 10 m were laid across the fault at each measurement site. Gas concentrations were measured at 16 points on each survey line; 72 measurement points at M3 were laid out across the XSH fault. The distance between each measurement point was approximately 10 m, whereas those far from the fault were separated by approximately 20-100 m at places that were confined by local conditions. A total of 648 soil-gas measurement points for each gas species were measured in the XXFS.
The measurement method of the soil-gas concentration was the same as that reported by Zhou et al. (2016). Before measurement, two samplers (S1 and S2), 1-m in length and 2.3-cm in diameter, were hammered into the soil at a depth of ∼80 cm ( Figure 2). The concentrations of soil gas at a depth of ∼80 cm were measured. The distance between the two samplers was 50 cm. S1 was used to measure CO 2 , CH 4 , H 2 , and Rn concentrations. The CO 2 analyser was connected to S1 to first measure the CO 2 concentration. The CH 4 analyser was connected to the outlet of the CO 2 analyser, and the H 2 analyser was connected to the outlet of the CH 4 analyser ( Figure 2A).
When the measurements of CO 2 , CH 4 , and H 2 concentrations were finished, the Rn analyser was connected to S1 for the measurement ( Figure 2A). S2 was only used to measure Hg concentration ( Figure 2B).
The CO 2 concentration was measured using the nondispersive infrared gas analyser GXH-3010 with a measuring resolution and accuracy of 0.001% and 4%, respectively . The concentration of CH 4 was measured using the Huberg LASER ONE with a measurement resolution and accuracy of 1 ppm and 5%, respectively. The concentration of H 2 in the soil gas was measured using the ATG-300H analyser, which has been widely used for the measurement of fault degassing. The measurement range and resolution of the H 2 analyser were 0-5,000 ppm and 0.1 ppm, respectively. An RAD7 Radon Detector and an RA-915+ Zeeman Mercury Analyser were used to measure Rn and Hg concentrations in soil gas, respectively.

Sampling and Analyzing Methods of Hot-Spring Gas
Ten bubbling hot-spring gas samples were collected from the hot springs of ①-⑩ in the XXFS ( Figure 1B) in 2017 and 2018 using the water displacing method. The gases were collected in sodalime glass bottles with a volume of 500 ml. An inverted funnel was submerged and placed on top of the bubbles to collect the hotspring gas. The collected gas in the funnel flowed into a bottle   filled with hot-spring water and replaced the water in the bottle (Zhou et al., 2015;Zhang et al., 2017). Details of the sampling method can be found in Zhou et al. (2015) and Zhang et al. (2017).
The δ 13 C CO2 and δ 13 C CH4 values of the gas samples were analysed at the Key Laboratory of Petroleum Resources Research, Institute of Geology and Geophysics, Chinese Academy of Science. A GC-IRMS analytical system gas chromatograph (Agilent 6890)-stable isotope ratio mass spectrometer (Thermo-Fisher Scientific Delta Plus XP) was used for carbon isotopic ratio analysis. The δ 13 C CO2 and δ 13 C CH4 values were reported relative to PDB.
The lower anomaly thresholds (LAT) of CH 4 , H 2 , CO 2 , Rn, and Hg in the XXFS were 1.6 ppm, 2.0 ppm, 0.2%, 2.3 kBq/m 3 , and 5.0 ng/m 3 , respectively ( Figure 3 and Table 2). The upper anomaly thresholds (UAT) of CH 4 , H 2 , CO 2 , Rn, and Hg were 8.1 ppm, 47.7 ppm, 0.7%, 24.9 kBq/m 3 , and 29.0 ng/m 3 , respectively ( Figure 3 and Table 2). Accordingly, the data of each soil-gas species in the XXFS could be divided into the background (A), mixing (B), and anomaly (C) populations. The proportion and mean concentrations of each population of CH 4 , H 2 , CO 2 , Rn, and Hg in the XXFS are listed in Table 2. The mean concentration of each gas species was calculated using Eq. 1 (Sinclair, 1974): where P M is the mean concentration of each gas species; f A , f B and f C are the proportion of each population, and P A , P B , and P C are  Frontiers in Earth Science | www.frontiersin.org March 2021 | Volume 9 | Article 635178 the mean gas concentrations of each population. Therefore, the mean concentrations of CH 4 , H 2 , CO 2 , Rn, and Hg in the XXFS were calculated to be 8.1 ppm, 9.9 ppm, 0.5%, 15.1 kBq/m 3 , and 12.9 ng/m 3 , respectively ( Table 2), which were close to the soilgas concentrations measured in the surface rupture zone produced by the Wenchuan Ms 8.0 earthquake in 2008 and 2010 ( Figure 4, Zhou et al., 2010;Zhou, et al., 2017). The spatial distributions of soil-gas concentrations at the 10 measurement sites (obtained by using Kriging interpolation) are shown in Figure 5. From the spatial distributions of each gas species, high concentrations of soil gas were spatially coincident with the location of the fault and appeared close to the fault ( Figure 5).

Origin of Carbon in the CO 2 and CH 4
Gas emissions from the fault zones are from multiple sources (e.g. Williams-Jones et al., 2000;Hong et al., 2010). Mantle, metamorphism of carbonate-bearing rocks, decomposition of organic material and surface biological activity are the principal sources of CO 2 in active fault zones (Irwin and Barnes, 1980;Toutain and Baubron, 1999), whereas microbiological and thermogenic genesis are the major processes of CH 4 genesis (e.g. Toutain and Baubron, 1999;Milkov and Etiope, 2018).
In this study, the δ 13 C CO2 and δ 13 C CH4 values of the hot-spring gas samples varied from -11.9‰ to -3.7‰ and -62.5‰ to 17‰, respectively (Table 3), and are consistent with the δ 13 C CO2 and δ 13 C CH4 values reported in previous studies ( Figure 6A; Zhou et al., 2015;Shen et al., 2007). Combined with the carbon isotopic values reported in previous studies, the δ 13 C CO2 and δ 13 C CH4 values in the XXFS varied from -21.37‰ to -2.9‰ and -62.5‰ to 19‰, respectively ( Figure 6A and Supplementary Table S1; Zhou et al., 2015;Shen et al., 2007).  Previous studies indicated that upper mantle-derived CO 2 has δ 13 C CO2 values between −3‰ and −9‰ with an average value of −6.5‰ (Javoy et al., 1986;Marty and Jambon, 1987). Soil organic matter derived from C3 and C4 plants have δ 13 C CO2 values varying from −12‰ to −14‰ and −24‰ to −27‰, respectively (Cheng, 1996). The R c /R a -δ 13 C CO2 plot reveals that the gas samples in the XXFS show characteristics of mixing among the end-members of the mantle, crust, and sediment ( Figure 6B). However, the He-C isotopic ratios in the XSH and ANH faults showed obvious characteristics of mixing between the end-members of the mantle and crust, whereas those in the ZMH and XJ faults showed greater addition of organic sources ( Figure 6B). The δ 13 C CO2 values near M9 and M10 in the XJ fault are −21.37‰ to −6.9‰ (Table 3  and Supplementary Table S1; Shen et al., 2007), with majority of the δ 13 C CO2 values within the domain of organic carbon source.
The spatial distributions of the CO 2 concentration in the XXFS show that the highest concentrations of CO 2 were found at M9 and M10 in the XJ fault ( Figure 5C). Therefore it could deduce that the carbon in the XJ fault in southern part of the XXFS has an organic carbon source.
Geophysical studies (e.g. Liu and Wang, 2020) indicate that anomalies of low velocity and high Poisson's ratio are present at the XSH, ANH, and ZMH faults in the XXFS (26.5°N, 28°N, 29.5°N , 31.5°N, Figure 1B and Liu and Wang, 2020) and are interpreted as fluid intrusions (partial melt of deep material) from the lower crust and/or uppermost mantle. The δ 13 C CO2 values in the XSH, ANH, and ZMH faults are higher than those in the XJ fault ( Figure 8A; Table 3 and Supplementary Table S1). According to previous studies, the 3 He/ 4 He ratios in the XXFS are 0.05-3.8 R a (where R a is the air-corrected 3 He/ 4 He ratios, Supplementary Table S1; Shen et al., 2007;Zhou et al., 2015), which lie within the typical crust and mantle domains (e.g. Sano and Marty, 1995;Zhang et al., 2017). The 3 He/ 4 He ratios in the XSH, ANH, and ZMH faults range from 0.19 to 3.8 R a , 0.1-1.3 R a , and 0.07 R a , respectively, which are higher than those in the XJ fault (0.05-0.35 R a ) ( Figure 8B and Supplementary Table S1). Analysis of He-C isotopic ratios in the XXFS was used to deduce that the carbon in the XSH, ANH, and ZMH faults might have originated from deep fluids, whereas the carbon in the XJ fault is of organic origin.

Anomaly and Degassing of Rn in the XXFS
Rn is inert and has low mobility in the geological environment (Tanner, 1964;Dubois et al., 1995). The uranium decay and the upward migration of Rn along the fault could represent the two sources of Rn in the fault zone (Toutain and Baubron, 1999;Sciarra et al., 2015;Sciarra et al., 2017;Sciarra et al., 2018;Sciarra et al., 2020). Previous studies indicate that the emanating power of the rock and soil, the permeability of the host rock, and the flow of the carrying gas are the primary factors that influence Rn activity in a diffusive system (e.g. Ball et al., 1991;Morawska and Phillips, 1993). Therefore, we used the formula of Åkerblom (1993) (Eq. 2), which is based on the 222 Rn activity at equilibrium with parent radionuclides (radium) in the soil, by verifying the anomaly threshold to determine the origin of Rn in the XXFS (Åkerblom, 1993;Sciarra et al., 2018): where C Rn and C Ra are the concentrations of Rn and radium in the soil gas (Bq/m 3 ) and soil (Bq/kg), respectively. ε is the emanation power coefficient (non-dimensional), ρ is the soil density (kg/m 3 ), and n is the effective porosity coefficient (non-dimensional). The colour symbols are data from bubbling hot-spring gas in this study. The grey symbols are hot-spring gas data reported in Shen et al. (2007) and Zhou et al. (2015). The He-C isotopic data in June 2009 were selected in this study in consideration of the significant influence of earthquake on the variations of He-C isotopic ratios (Zhou et al., 2015).
Frontiers in Earth Science | www.frontiersin.org March 2021 | Volume 9 | Article 635178 Previous studies show that the average content of soil radium-226 in China is 36.5 Bq/kg (e.g. Sun et al., 2004). The ε values are variable in different rock and soil types (e.g. Righi and Bruzzi, 2006;Jobbágy et al., 2009;Qin, 2012;Shen et al., 2012). The surrounding rocks and strata outcrop in the XXFS can be divided into four types: 1) volcanic rocks and acidic intrusions; 2) dolomite, limestone and carbonate; 3) sandstone; and 4) polyterrigenous materials ( Table 4, BGMRSP, 1993). Therefore, the ε values of the four types of surrounding rocks and strata in the XXFS have values of 0.48 (Qin, 2012), 0.074 (Righi and Bruzzi, 2006;Jobbágy et al., 2009), 0.4 (Qin, 2012, and 0.12 (Shen et al., 2012), respectively ( Table 5). The values of the soil porosity (n) were calculated using Eq. (3): n 1 − subsoil bulk density 2.65 ( 3) where 2.65 g/cm 3 is the mean density of soil porosity. The values of subsoil bulk density (kg/m 3 ) at a depth of 100 cm in the XXFS were selected from the Harmonized World Soil database (HWSD), which is elaborated by the Food and Agriculture Organization (FAO) (FAO/IIASA/ISRIC/ISSCAS/JRC, 2012). Therefore, the calculated Rn concentrations from M1-M10 in the XXFS ranged from 8.5 kBq/m 3 to 61.1 kBq/m 3 (Table 5), with mean value of 40.2 kBq/m 3 . The calculated Rn concentration in the XXFS was higher than the statistically defined anomaly threshold (24.9 kBq/m 3 ) ( Table 2). Therefore, the Rn concentrations that exceeded the calculated value were likely not supported by in-situ production. The measurement sites with high Rn concentrations were spatially consistent with the distributions of the volcanic and acidic intrusions in the XXFS ( Figure 5D and Table 4). The abovementioned analysis concluded that the concentrations of Rn lower than the UAT in the XXFS might originate from insitu production. Rn concentrations higher than the UAT in the XXFS might be related to the upward migration of Rn along the FIGURE 7 | Plot of δ 13 C CO2 vs. δ 13 C CH4 . The colour symbols are data from bubbling hot-spring gas in this study. The grey symbols are hot-spring gas data reported in Shen et al. (2007) and Zhou et al. (2015). Modified after Milkov and Etiope (2018). faults. Rn degassing from the fault was also verified by the spatial distributions of Rn concentrations in the XXFS ( Figure 5D). Additionally, the high concentrations of Rn in M5 ( Figure 5D) demonstrate the degassing of Rn from the fault according to the field survey.

Relationship Between Soil-Gas Concentration and Fault Activity
In our study, high concentrations of Rn were primarily found at the XSH fault ( Figure 5D and Table 1), which has the highest stress, fault slip rate, and maximum strain rate in the XXFS ( Figure 8C and Chen., 2006;He and Lu, 2007;Shi et al., 2012;Zhao et al., 2015;Yue et al., 2017;Li et al., 2019). Fault activity can facilitate the transport of fluids due to episodic fracturing events that can release quantities of deep fluids (Evans et al., 1997;Kulongoski et al., 2005;Tanikawa et al., 2010;Zhang et al., 2016). In our study, granites are well-developed in the XSH fault (BGMRSP, 1993; Table 4). In a previous study, the permeability of granites increased by three orders of magnitude as the slip rate increased, indicating that increasing the slip rate could create and maintain elevated fault permeability (Zhang et al., 2017).
From the spatial distributions of soil gases in the XXFS (Figure 5), the anomalously high concentrations of CH 4 , H 2 , CO 2 , Rn, and Hg appeared near the fault, which might indicate that the fault slip-rate influenced degassing from the active fault zone. The elevated permeability of the fault facilitates the migration of gases in the active fault zone. The different patterns of soil-gas concentrations across the fault might be influenced by the permeability of the fault zone, such as the porosity of the fault and soil characteristics of the measurement site (e.g. Fu et al., 2005).

CONCLUSION
In this study, the geochemical characteristics of soil gas and the relationship between degassing and fault activity in the XXFS were examined. Summarising: (1) The mean concentrations of CH 4 , H 2 , CO 2 , Rn, and Hg in the XXFS are 8.1 ppm, 9.9 ppm, 0.5%, 15.1 kBq/m 3 , and 12.9 ng/ m 3 , respectively. Anomalously high concentrations of soil gases primarily appeared near the fault, indicating the degassing of faults in the XXFS.
(2) In the XXFS, the δ 13 C CO2 and δ 13 C CH4 values varied from −11.9‰ to -3.7‰ and −62.5‰ to 17‰, respectively. He-C (1) sandstone, shale and carbonateous shale intercalated with carbonate rock (2) dolomite and limestone intercalated with siltstone and shale M9 (1) basic volcanic rock (2) dolomite and limestone M10 Basic volcanic rock a Data sources: BGMRSP, 1993. isotopic analysis indicates that the soil gas in the XSH, ANH, and ZMH faults might have originated from deep fluids, whereas the carbon in the XJ fault is of organic origin. The CH 4 in the XXFS might have thermogenic and abiotic origins. (3) The concentrations of Rn lower than the UAT in the XXFS might be due to the in-situ production. The anomaly of Rn concentrations (higher than the UAT) in the XXFS might be related to the upward migration of Rn along the faults. (4) The distributions of the high concentrations of soil gases were spatially coincident with the highest stress and the maximum strain rate in the XXFS, which might indicate that the fault activity enhanced permeability and, thus, increased gas emission rates.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.