Correlations between the radon concentrations in soil gas and the activity ofthe Anninghe and the Zemuhe faults in Sichuan, southwestern of China
Yao Yanga,b, Ying Lia,∗, Zhijun Guanb, Zhi Chena,∗∗, Lei Zhangc, Chao Jia Lva, Fengxia Suna
a CEA Key Laboratory of Earthquake Prediction, Institute of Earthquake Science, Beijing 100036, Chinab Sichuan Earthquake Agency, Chengdu 610041, Chinac Institute of Crustal Dynamics, China Earthquake Administration, Beijing 100085, China
The Anninghe fault (ANHF) and the Zemuhe fault (ZMHF) with left-lateral strike-slip, located along the easternboundary of the Sichuan-Yunnan block (southwestern of China), are some of the most active faults. These faultsmainly control the seismicity of southwestern area of China. Measurement of soil gas radon (Rn) emitted fromfault along the ANHF and the ZMHF has been carried out for the research of tectonic activity. We obtained theRn concentrations at 394 sampling points along 15 profiles across the ANHF and the ZMHF in 2016. Themeasurement results show that the values of Rn in the ANHF are significantly higher than that in the ZMHF. Therelative coefficient KQ of Rn activity attained in profiles of the ANHF ranges from 3.3 to 9.1, which are obviouslyhigher than that of 2.1–2.5 in profiles of the ZMHF. The radon flow brings up the deeper and radon-richer gasupward through the high-level cracked strata caused by the tectonic activity accounts for the anomalously highvalues attained. The spatial variation of Rn in the concentration profile and the relative coefficient KQ calculatedindicate that the tectonic activity of the south segment of the ANHF is significantly higher than that of the northsegment of the ZMHF.
1. Introduction
Survey of anomalously Rn concentration is an effective way to studyvarious manifestations of geodynamic activity in the upper crust (Kinget al., 1996; Toutain and Baubron, 1999; Immè et al., 2005; Seminskyand Demberel, 2013). The crust with strong inhomogeneity containsnumerous pores and fractures with fluids that have different chemicalcompositions at different places (King et al., 2006). These fluids can beactivated and migrate up to the surface under different conditions suchas pressure, temperature and stress/strain changes related to crust de-formation and earthquakes (King, 1986; Toutain and Baubron, 1999;Bernard, 2001). Gas discharge through seismically active faults is a longterm and permanent phenomena. Active faults can act as drains in thecrust which sustains the gas geochemical characteristics in a certainarea, due to their high permeability and porosity (Irwin and Barnes,1980; Ghosh et al., 2009; Walia et al., 2009, 2010; Li et al., 2013). Thepermeability through pathways increases with the enhanced tectonicactivity and results in more significant gas migrating to the surface ofthe earth (Lombardi and Voltattorni, 2010). Related researches of fluidgeochemistry have shown that the change of Rn concentration is one ofthe most effective tracers of structure (López et al., 2016) and seismic
activities (Ciotoli et al., 2007). Rn is a product of uranium decay seriesand is also ubiquitous in the earth crust because Uranium and Thoriumare present in almost all rock and soils. It is inert, colorless, odorless gasand is the heaviest rare gas in nature under ambient temperature andpressure conditions. Rn displays poor intrinsic mobility due to its shorthalf-life (3.8 d). In a diffusive system, therefore, it obviously comesfrom a short distance near the sampling site. The soil gas concentrationis sensitive to the changes of stress accumulation and tectonic activityin the crust (Fu et al., 2008; Zhou et al., 2010; Yuce et al., 2017).Changes of elastic strain cause rocks to dilate or compress and make anopen fractures rupture, which in turn allows the flux of gases. Since ithas a very poor intrinsic mobility in soil and very short period of half-life, Rn would not be expected to travel upwards long distances.However, the enhanced upward velocity of Rn movement is explainedby the existence of rising fluids that play as a carrier of Rn, such asunderground water, carbon dioxide, nitrogen and methane which areproduced and liberated by the enhanced tectonic activity (Baubronet al., 2002).
The spatio-temporal variations of soil gas Rn at an active fault zonemay reflect the regional crustal stress/strain changes related to seismo-tectonic activity (Yang et al., 2006; Fu et al., 2008; Zhou et al., 2010;
https://doi.org/10.1016/j.apgeochem.2017.11.006Received 12 June 2017; Received in revised form 20 October 2017; Accepted 14 November 2017
Koike et al., 2014). Utkin and Yurkov (2010) suggest a model of Rnemanation under compression or extension, and indicate that a relativestrain change of n × 10−7 leads to an increase of Rn activity in con-centration by 200%. This high sensitivity means that Rn data can beused as good tracers of tectonic activity. Lombardi and Voltattorni(2010) studied the geochemistry Rn, He and CO2 soil gas in two Italianareas characterized by different seismo-tectonic activities. Their resultindicates that due to seismic activity favoring gas migration from thedeep, the concentrations of soil gases are much higher in Colpasquale (aseismically active region) than that in Guasila-Suelli (a seismically in-active area). Ghosh et al. (2011) reported that Rn concentration canincrease abruptly under high stress condition before an earthquake, andthe average Rn concentration is much higher in an active area than thatin non-active region. Han et al. (2014) investigated the spatiotemporalvariations of soil gas Rn and CO2 across the Tangshan faults, the seis-mogenic fault of the Ms7.8 Tangshan earthquake in 1976, and con-cluded that, the annual increases of soil gas Rn and CO2 concentrationsin the northeastern segment in 2012 relative to 2011 were caused bythe enhancement of seismic activity in the area.
In this paper, the Rn concentrations in soil gas are measured acrossthe ANHF and the ZMHF. The geochemical characteristic of Rn in theseareas is determined and its correlation to the activity of fault is dis-cussed. In addition, the tectonic activity of the ANHF and the ZMHF isevaluated according to the geostatistics of the data attained and var-iation among the measuring points.
2. Study area
Our study areas are located in the intersection of the ANHF and theZMHF on the southeastern margins of the Tibetan Plateau (Fig. 1). TheANHF and the ZMHF systems show characteristics of a left-lateralstrike-slip movement since late Quaternary (Xu et al., 2003a,b). Thesefaults are located along the eastern boundary of the Sichuan-Yunnanblock (Zhang et al., 2003), which is part of the southeastward extrusiveactive tectonic system in the eastern and southeastern Tibetan Plateau(Tapponnier et al., 1982; Peltzer and Tapponnier, 1988) and theycontrol the main seismic activity in southwest China (Wen et al., 2008).The ANHF terminates at the ZMHF around Xichang city region, wherehighly populated and multiple historical earthquakes have been re-corded, including 814 AD M7, 1536 AD M71/2, 1850 AD M71/2 and1952 AD M63/4 (Fig. 2; Department of Earthquake Disaster Prevention,1995).
The ANHF trends N-S for ∼200 km along the Anninghe valley, in-tersecting with the Xianshuihe fault to the northwest (Allen et al.,1991), the Longmenshan Thrust Belt to the northeast and the ZMHF tothe south (see Fig. 1; Ren, 2014). The ANHF is divided into two seg-ments by Mianning County with different geometrical distributions (Peiet al., 1988). The ZMHF trends NNW-SSE for ∼120 km along thetrough-like Zemuhe valley. It extends from the north of the XichangCity to the town of Qiaojia and terminates at the Xiaojiang fault in thesouth (Ren et al., 2010). The strata from bottom to up in the study areainclude Precambrian, Paleozoic (Ordovician), Mesozoic (Triassic, Jur-assic, Cretaceous), and Cenozoic (Late Tertiary and Early Quaternary).Cenozoic strata are mainly consisted of interbedded mudstone andsandstone with interlayered coals (Geological Bureau of SichuanProvince, 1966). Quaternary fluvial, alluvial, and glacial deposits aredistributed mainly in the lowlands along the flat Anninghe valley and inintermontane basins. Granites of the Precambrian, Cambrian and Me-sozoic are widely distributed in the western parts of the ANHF-ZMHFzone (see Fig. 2). The soil type in the 3 survey sites is same, mainlyconsists of yellow silt and clay (Seismological Bureau of SichuanProvince, 2013a,b).
2.1. Site 1
Yangfushan (site 1) is located at the eastern Shagou Village, Yuehua
Township. The Mesozoic strata are distributed along the eastern side ofthe fault and Quaternary sediments are located along the west side.Three soil gas profiles for Rn concentration with the strike of NEE-SWW, which cross the southern segment of the ANHF, were measured(see Fig. 2). The sampling points are located in the hillside farmlandand wasteland covered by dry silty soil. Around site 1, the fault trace ofthe ANHF is distributed as intermittent and echelon lineation, whichseparates the terraces from mountain areas. The displaced landformsare extremely developed and the flowing channels have systematicallydisplaced by 400–600 m (Ren, 2014).
2.2. Site 2
The Xiaomiao Township (site 2) is located at the suburb of theXichang City, about 1000 m north of the Lijingbao Village. The segmentof the ANHF in site 2 is strike near N-S direction, and the fault scarpsare clearly observed along the fault. However, the level of the dis-placement cannot be accurately obtained due to the horizontal dis-location of the surface is difficult to be confirmed. The Jurassic sand-stones are distributed on the western side of the ANHF, while Neogeneand Quaternary sediments or slope deposits which is developed intoyellow loess layers on the other side. The unconformable interface be-tween the Jurassic bedrock and the Quaternary loess has been observedin the gullies. In this site, nine soil gas profiles which strike near E-Wdirection in layout were measured for Rn concentration. The measuringpoints are located in the hillside farmland and wasteland, where drysilty soil is developed (see Fig. 2). A scarplet caused by the M71/2earthquake in 1850 AD is well preserved.
2.3. Site 3
Dajingliangzi (site 3) is located at the Dajing Village, southeasternof the Qionghai Lake. The transverse profiles are across the northernsegment of the ZMHF zone. The three soil gas profiles strike NEE-SWWdirection in layout (see Fig. 2). The measured points on the west side ofthe fault trace are distributed on the fault cliff and the river terrace, andthe points on opposite side of fault trace are all distributed in thefarmland. The soil is similar at each profile and consists mainly ofyellow silt and clay. It can be observed in the site that the tectoniclandforms such as offset stream channels, gullies, alluvial fans, and faultscarps have been shifted by the left-lateral movements of the ZMHF.Displacements on the scale of meters even up to decameters occur in thelowest terrace and the youngest alluvial fans along Zemuhe valley. Alarge numbers of stream channels have been abandoned and offset left-laterally as the result of the displacement of the ZMHF. In addition,fault scarps are widely developed along the fault in this region, inter-secting mountain ranges and slopes, as well as river channels, alluvialfans, and terrace risers (Ren et al., 2010).
3. Material and methods
For the first time, we carried out a cross-sectional soil gas Rnmeasurement along 15 profiles at Yangfushan in the southern segmentof the ANHF (site 1, 3 profiles), Dajingliangzi in the northern segmentof the ZMHF (site 3, 3 profiles) and Xiaomiao Township, the intersec-tion part of them (site 2, 9 profiles), respectively. Totally, 394 samplingpoints were measured for soil gas Rn concentration. Interval betweenneighbor sampling points is similar and varies from 10 to 15 m in thefield (see Fig. 2). In each profile, 2 or 3 parallel survey lines with in-terval distance of 5 m were laid out. Due to that the emission of soil gascan be affected by soil moisture, air temperature and barometric pres-sure (Hinkle, 1994; Iakovleva and Ryzhakova, 2003; Fu et al., 2017), atime duration with relatively stable meteorological conditions (Clearweather, temperature 5-20 °C) in the region was selected from 10th to20th January 2016, for soil gas survey in the field. Furthermore, inorder to minimize the diurnal temperature variation on the measuring
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
24
results, all the sampling and measuring work were completed in thefield from 9:30 a.m. to 17:30 p.m.
Soil gas Rn was sampled by inserting a hollow stainless steel sam-pler into the ground at a depth of 80 cm. The sampler was connected toa detector through rubber tubes. Rn survey was carried out in the fieldusing a RAD 7 Rn detector (Durridge COMPANY Inc. Bedford, USA)which counts the α particles emitted during the decay of 222Rn to 218Pofor the counting of Rn concentration. The sensitivity and measurementerror of Rn detector are 14.8 Bq/m3 and± 5%, respectively (Zhouet al., 2010). The detection system consists of a solid-state ion-im-planted planar silicon alpha detector placed in a 0.7 L accumulationchamber (Fig. 3). The soil gas is drawn through two filters: the firstremoves moisture and the second, only allows Rn to enter the detectionsystem chamber by removing atmospheric particulate as well as Rndaughters (Giammanco et al., 2009).
4. Data processing
Using the SPSS 16.0 software package, descriptive and statistic re-sults of the soil gas data of the 15 traverses profile are obtained(Table 1). Geochemical background represents the normal abundanceand reflects the accumulation level of a given element or component ina certain area. The Kolmogorov-Smirnov test of the data shows that allprofiles of the data are subject to normal distribution (sig.> 0.05).
Consequently, in this work area, the background value is expressed bythe mean value of the remained data after iteratively removing valueshigher than three times of the standard deviation (Zhang et al., 2006).The calculated background values of Rn concentration in the Yang-fushan, Xiaomiao Township and Dajingliangzi areas are 21036, 7575and 12265 Bq/m3, respectively.
The statistical threshold for an anomaly value of soil gas could bedetermined by several methods, such as a cumulative probability plot(or Q-Q) (Sinclair, 1991; Ciotoli et al., 2007; Zhou et al., 2010; Li et al.,2013), empirical distribution (Seminsky and Demberel, 2013) and themean plus n time the standard deviation (SD) (Ciotoli et al., 1998;Guerra and Lombardi, 2001; Fu et al., 2005; Walia et al., 2008; Al-Hilaland Al-Ali, 2010). In present study, the number of sampling point ineach profile is insufficient for using the method of cumulative prob-ability plots. In this paper, the threshold for an anomaly value of Rn isusually determined using the method of mean + n SD (n = 1/2 or 1) forsoil gas Rn data that obeys normal distribution. If taking n = 1, theanomaly range is narrow and the anomaly values is scattered and iso-lated in the traverses profile curves, which is not suitable for judgingthe anomaly variation and fault activity. In order to observe the con-tinuity and ensure the changing trend of anomalies, consequently, thethreshold for anomaly values is fixed at “mean + 1/2SD". The part ofthe value that exceeds the threshold for anomaly is represented by darkcolor in Figs. 4–6. Width of Rn anomalous domain (H) in the sampling
Fig. 1. Tectonic map of the eastern Tibetan Plateau and the study area (modified after Wang et al., 2013).ANHF = Anninghe fault, EKLF = eastern Kunlun fault, GZ-YSF = Ganzi-Yushu fault, HHF = Honghe fault, JSJF = Jinshajiang fault, LMSF = Longmenshan fault, XJF = Xiaojiang fault,XSHF = Xianshuihe fault and ZMHF = Zemuhe fault.
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
25
profile curve is determined by the length of the profile with fracturesand cracks, throughout of which Rn concentrations are anomalouslyassociated with the fault. However, in the profiles of the study areas,the most of common cases are the alternatively appearance of high andlow values of Rn, and the core zone is often located within the domaincharacterized with values lower than the threshold for anomaly (seeFig. 4, profile A-A′, Fig. 5, profiles D-D′, L-L′ and Fig. 6). Domains with
low permeability appear in the fault zones due to the presence of finelydispersed filler or small blocks, almost not broken by open fractures(Torabi and Berg, 2011). In such cases, the shape of Rn concentrationcurve along a profile is often multi-peaks features. In our measuringprofiles, the small segments with Rn concentration lower than thespecific threshold value are regarded and included in the anomaly partsassociated with the fault, pending their location between two
Fig. 2. Geological map of study area (left) (modified after Ren et al., 2010 and Ren, 2014) and plot map of the points of the soil gas Rn sampled along the ANHF and the ZMHF (right). 814AD (7), 1536 AD (71/2), 1850 AD (71/2), 1952 AD (63/4): historical earthquakes in the study area (number outside the brackets represents the year of the earthquake occurred and numberinside the brackets represents the magnitude of the earthquake, AD = Anno Domini).
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
26
anomalous domains if they are smaller than at least one of such do-mains in length (Seminsky and Demberel, 2013) (see Figs. 4 and 5,profiles D-D′, E-E′, H-H′, I-I′, J-J′, K-K′, and Fig. 6, profile N-N′, in thegrey part).
The Rn concentrations (Q) in profiles across different faults arecommonly characterized by high variability, which could be due todifferent causes related either to the fault itself (tectonic and structure)or to local features (Rn source, geology, weather conditions, etc.) andmakes it difficult to compare faults in different areas (Seminsky andBobrov, 2009). To overcome this issues, Seminsky and Bobrov (2009)and Seminsky and Demberel (2013) used relative Rn activity of faults(KQ), which is the ratio of maximum concentration (Qmax), to minimumconcentration (Qmin) in any fault wall for a fault next to the fault-relatedanomaly. The KQ reflects the contrast of the anomaly detected in theprofile. It has been identified that the KQ is clearly related to the scaleand seismic potential of the fault and reflects the degree of geodynamicactivity of the fracture. In addition, it is less dependent on such factorsas weather conditions, radioactivity of rocks, thickness of sediments,etc., which complicate the estimation of activity of the faults in seismicregions (Seminsky and Bobrov, 2009; Seminsky and Demberel, 2013).The comparison of the Rn activity of the faults is more efficient if re-lative parameters are considered (Guerra and Lombardi, 2001;Ioannides et al., 2003; Moussa and Arabi, 2003). To calculate thisindex, we use either an average of two values of Qmin, determined foreach wall of the fault, or one values of Qmin, if the damage zone is notcompletely cut by the profile (e.g., Fig. 4, profile A-A′). In the curves,
the value of each point is determined by the mean value of the Rnconcentrations attained in the 2 or 3 parallel survey lines in the specificposition of the profile.
5. Profile Rn concentration survey results
5.1. Site 1
In order to analyze the Rn concentration changes in site 1, the 77measurement points totally were sampled along 3 transverse profiles(see Fig. 2) across the ANHF in the Yangfushan. The measured radonconcentrations show the prominent spatial variation along the fault.The parameters of Rn anomalies in the Yangfushan are list in Table 1. Inthe profiles A-A′, B-B′ and C-C′, the Rn concentrations in soil gas varyfrom 6842 to 42200, 6114 to 31202 and 4595–77650 Bq/m3, and themean arithmetic values are 18866, 18283 and 27550 Bq/m3, respec-tively. The variations of Rn concentration along the three profiles areoscillatory, where anomalous domains alternate with the profile's seg-ments, where Rn concentrations are lower than the threshold value. Inthe profile A-A′, the analysis of ratios of adjacent sections values withQ < Qmean+1/2SD and Q > Qmean+1/2SD reveal that the anomalyalong the fault is about 52 m long (see Fig. 4 and Table 1) and appearsin the entire western half of the profile. The maximum value of 42200Bq/m3 appears at a site about 7 m away from the fault trace. In theprofile B-B′, the width of Rn anomalous associated with the fault is upto 70 m and anomalous Rn concentration distributes on both sides of
Fig. 3. Sampling scheme and instrument employed for soil gasRn concentration measurement.
Table 1Main statistic parameters of soil gas Rn data in study areas.
Maximum/Qmax - maximum value of the Rn concentrations in soil gas; Minimum - minimum value of the Rn concentrations in soil gas; Qmin - minimum value of the Rn concentrations insoil gas just outside the fault anomaly (counted as average value on estimations in both fault wings); H - width of Rn anomalous domain; KQ - index of soil gas Rn activity.
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
27
the fault. The maximum value (31202 Bq/m3) is recorded at a siteabout 21 m away from the fault trace. In the profile C-C′, the spatialscope of Rn anomaly is 75 m wide and distributes on both the sides ofthe fault. The values in the eastern side of the fault are significantlyhigher than that in the west side. The Rn activity indices, KQ, de-termined for the selected anomaly within profiles A-A′, B-B′ and C-C′are 6.2, 3.3 and 9.1, respectively (see Table 1).
5.2. Site 2
Totally, 222 sampling points were measured along 9 transverseprofiles (see Fig. 2) in site 2, the intersection part of the ANHF and theZMHF. Statistic parameters of the Rn anomalies in site 2 are listed inTable 1. The spatial distribution of the Rn concentration along theprofile F-F′ shows an increasing trend that the maximum value appearsat the distance 0 m of the profile. Variation of Rn concentration alongthe profile L-L′ shows a U-shape pattern. The curve shape of otherprofiles is oscillatory (see Fig. 5). In the ∼150-m-long profile D-D′, anabout 55-m-wide linear Rn anomaly is distributed on the east part of thefault trace, where a Rn concentration peak of 13986 Bq/m3 appearsabout 21 m east of the fault trace. However, it is an obvious fact that thefault is associated with about 40 m long domain of the minimum values(Q < Qmean) of Rn concentration in this site. The fault trace is locatedat the central part of this domain and is marked by the minimum of Rnvalue (2373 Bq/m3). The similar pattern appears in the profile L-L'. Inthe profile F-F' (135 m long), an about 28 m wide anomalous Rn as-sociated with the fault presented in the east side of the fault. Themaximum value (10400 Bq/m3) has been observed in the east side ofthis profile about 70 m away from the fault trace. On the east side offault the Rn value is significantly higher than those on the west side.This may be explained by the increase in the depth of bedrock from theeast side of the fault to the west. In the profile L-L' (180 m long), thewidth of anomalous Rn concentration is about 30 m, which is located atthe farthest side of fault. The position of maximum value (15922 Bq/m3) is about 80 m away from the west side of the fault trace. In theprofile G-G' (165 m long), the maximum value (17728 Bq/m3) appearsat a site about 10 m away from the west side of the fault trace. The Rnconcentration is significantly higher in the vicinity of the fault trace anda sharp peak is identified on the fault trace. In the profiles I-I′ and J-J′,the anomaly width of Rn concentration varies from 86 to 95 m re-spectively, and the anomaly presents on both sides of the fault trace. Inthe eastern side of the fault, anomalous amplitude is significantlygreater than that on the west side. In the profiles E-E′, H-H′ and K-K′,anomaly of Rn concentration along the fault appears on both sides ofthe fault and high values mostly present near the fault and the max-imum value measured in these three profiles are all located on the eastside of their fault traces. The KQ, determined within 9 profiles rangesfrom 3.3 (profile G-G′) to 8.7 (profile J-J′). According to Table 1, themaximum value appears on profile I-I'.
5.3. Site 3
In site 3, 70 sample points were attained along three transverseprofiles (see Fig. 2) across the ZMHF in the Dajingliangzi area. Therecorded Rn concentrations show prominent spatial variation along theprofiles. The key parameters of the Rn concentration anomalies in theDajingliangzi area are listed in Table 1. In the profiles M-M′, N-N′ andO-O′, the Rn concentrations of the soil gas vary from 7495 to 15883,8273 to 20119 and 6219–22433 Bq/m3, respectively and the arithmeticmeans are 11417, 14366 and 14666 Bq/m3, respectively. Variation ofthe Rn concentrations along the 3 profiles shows an oscillatory. In theprofile M-M' (135 m long), the Rn concentration variation shows theanomaly associated with the fault is about 34 m wide that is distributedon the west side of the fault trace. The maximum value (15883 Bq/m3)is measured at a position about 39 m away from the fault trace. In theprofiles N-N' (135 m long) and O-O' (135 m long), the Rn concentration
Fig. 4. Variation of the Rn concentrations in soil gas along the profiles in Yangfushan area(Site 1). 1 – Profile number; 2 – curve of variation of the Rn concentrations (Q) along theprofiles; 3 – position of fault in curves; 4 – levels and values of arithmetic mean (Qmean) ofsoil gas Rn activity for profiles; 5 – levels of Qmean± 1/2SD (SD = standard deviation); 6 –value of soil gas Rn concentration, representing major maximums in the fault-relatedanomaly and minimum(s) next to the fault-related anomaly; 7 – profile domains withanomalous Rn concentrations in soil gas (Q > Qmean+1/2SD); 8 – soil gas Rn fault-related anomaly associated with fault. In each profile, 2 or 3 parallel survey lines withinterval distance of 5 m were laid out, and 10-13 sampling points were laid out in eachsurvey line. Interval between neighbor sampling points is similar and varies from 10 to15 m according to the ups and downs of the terrain and the presence of obstacles such asconstructions on the ground. In the curves, each point represents the Qmean attained in the2 or 3 parallel survey lines in the specific position of the profile.
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
28
anomalous width associated with the fault is 53 and 45 m respectively,both of them are distributed on the east side of the fault trace. The Rnconcentration maximum (22433 Bq/m3) measured at a site in the eastin the profile O-O′ and is about 36 m away from the fault trace. The Rnactivity indices KQ for the profile M-M′, N-N′ and O-O′ are 2.1, 2.4, and2.5, respectively (see Table 1).
6. Discussion
6.1. Spatial characteristics of soil gas Rn
The results of the profile surveys conducted for the three sites in-dicate that the Rn concentrations are unevenly variable at faults.However, the sampling points near the fault can be easily distinguishedby high Rn concentration of soil gas. The anomalies of Rn concentrationrevealed by the profile survey data are variable in curve shape. In orderto reveal the shape of soil gas Rn anomalies with the aim to determinethe factors that control the Rn concentrations, in earlier study(Seminsky and Demberel, 2013) the fault-related anomalies have beendivided into anomalies of continuous and discontinuous types.
Continuous anomalies are represented by segments where Rn con-centrations are variable and are always higher than the Qmean anddiscontinuous anomalies refer to the domains where Rn concentrationsare lower than the Qmean, but the length of each of such domain is lessthan at least one of the two segments with Q > Qmean which are lo-cated side by side according to Seminsky and Demberel (2013). How-ever, in this paper, in order to make the anomalies highlighted com-pared with the background value, and due to the lack of earlierresearches in the study area, the threshold for anomaly value is fixed at"Qmean + 1/2SD". Therefore, we use Qmean + 1/2SD instead of Qmean
which is suitable for the evaluation of the anomaly scale. Continuousanomalies are defined as the segments where Rn concentrations arevariable and are always higher than the Qmean + 1/2SD, discontinuousanomalies refer to the domains where Rn concentrations are lower thanthe Qmean + 1/2SD in present study. The profiles F-F′, G-G′ and L-L′ inFig. 5 and the profiles M-M′, O-O′ in Fig. 6 are classified to continuousanomaly type. The rest of the profiles, such as all profiles in Fig. 4,profiles D-D′, E-E′, H-H′, I-I′, J-J′, K-K′ in Fig. 5 and profile N-N′ inFig. 6, are classified to discontinuous anomaly type. The geometry ofanomalous patterns of soil gas Rn concentration is correlated highly
Fig. 5. Variation of the Rn concentrations in soil gas along the profiles in Xiaomiao Township (Site 2). The legend refers to the caption of Fig. 4.
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
29
with the tectonic line. Fault trace that in some place manifest fault scarpin the topography appears in different positions within the boundariesof anomaly domain. In most cases, such faults are located inside theanomaly domain (see Fig. 4, profiles B-B′, C-C'; Fig. 5, profiles E-E′, G-G′, H-H′, I-I′, J-J′, K-K′). Anomaly area mostly have been identified inthe bilateral sides of activity tectonic line (see Fig. 5, profiles D-D′, L-L';Fig. 6), except profile A-A′ where anomaly values are mostly located on
the fault and profile F-F′ where anomaly values are located far from thefault trace.
The Rn concentrations in soil gas are influenced by many factors,such as the permeability of the substrate (Seminsky and Demberel,2013), sediment thickness, soil type (Lehmann et al., 2000), soil par-ticle size (Choubey et al., 1999), average radium content of the sur-rounding geological formation (Papp et al., 2008) and the meteor-ological conditions (Winkler et al., 2001; Moreno et al., 2016). In thestudy area, granite is widely distributed on the west side of the ANHFand the ZMHF, where the distance from these granite to the measuringsites is about 6–35 km respectively. Because of granite weathering anddenudation, the content of uranium in the soil in granite areas and itssurrounding areas is higher than other areas (Cho et al., 2004). Thismay affect the determination of the background value of soil gas Rnconcentration. However, in order to explain the sharply inhomogeneousfield of Rn across the faults, it is assumed that the rock type changesevery 10-20 m, which cannot be considered as an ordinary case in thegeological setting. Consequently, the radioactivity of rocks under thesoil may affect the background value rather than the sharply in-homogeneous field. Besides, taking into account the soil type and thesoil stratum thickness, which remains similar along the profiles in the 3sites, revealed discontinuous anomalies are not related to the soil.
Logically, it is reasonable to conclude that the spatial shapes of theRn concentration anomalies are dominated by the permeability of theunderlying rocks and the structure of the faults which are characterizedby different properties and geometries (Fig. 7). In site 1, the fault showsa left-lateral strike-slip movement with thrust component, N-S trending,70-75° dipping and thrust toward the east. On the east side of the mainfault, there are three branch faults which thrust toward in the oppositedirection. The rocks in this region have been intensely deformed andare highly cracked due to the movement of the fault under compression.In site 2, the faults is left-lateral strike-slip accompanied by normal-dip-slip, N-S trending and dipping towards east at an angle of 75–80°. Twofaults with similar properties of the target fault on the east side diptowards west at a high dip angle. The above set of faults together formsa graben system. In site 3, the fault is a left-lateral strike-slip fault witha normal dip-slip, and has a NNW-SSE trending and NE dipping with ahigh dip angle. The rock formation has not experienced serious de-formation both in site 2 and 3 compared with site 1. The coincidence ofthe fault with the narrow minimum of Rn concentration, as revealed bysome profiles (see Fig. 5, profiles D-D′, L-L'; Fig. 6), can be caused by thegouge and ultracataclasite with low permeability that prevent the gasmigration from the deep up to the surface (King et al., 1993). In brief,the discontinuity and specific characteristics of the soil gas Rn con-centration anomalies are attributed to the complex structure of the faultzones.
6.2. Fault activities
The Rn concentrations of soil gas revealed by the profile survey dataare obviously variable in shape across the faults. According to themeasurement results, the average of Rn concentration in site 1 is thehighest compared with the other two areas. Inasmuch as the absoluteRn values may bear local effects (intensity of the primary Rn source,geology, weather dynamics, etc.), we use the parameter KQ to analyzethe fault activity. The parameter KQ activity is higher in normal than instrike-slip faults, and increases with the size and tectonic activity offaults, with the latter factor being the basic geodynamic control of theRn emanation (Seminsky and Bobrov, 2009). This parameter has beenapplied by Seminsky and Bobrov (2009) and Seminsky and Demberel(2013) to estimate the activity of faults in Central Mongolia, westernBaikal and southern Angara areas.
The relative index of the Rn activity KQ, reflecting the contrast ofthe anomaly within the profiles the study area, varies in from 2.1 to 9.1(see Table 1). The arrangement of KQ in the decreasing series (Fig. 8)allows us to identify the characteristic values of the index (3 and 6) that
Fig. 6. Variation of the Rn concentrations in soil gas along the profiles in Dajingliangziarea (Site 3). The legend refers to the caption of Fig. 4.
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
30
separate groups of the profiles with similar Rn activity. The faults canbe classified by contrast of the associated soil gas Rn concentrationanomalies into three groups with high (KQ > 6), medium(6 > KQ > 3) and low (KQ < 3) activity. As can be seen from Fig. 8,the KQ values of profiles A-A′ and C-C′ in site 1 (Yangfushan area) are in
the group of high activity, and the profile B-B′ falls in the group ofmedium activity. In site 2 (Xiaomiao Township), the KQ values aremostly in the group of medium activity, except the values of profiles J-J′and F-F′ that fall into the group of high activity and K-K′ in the group oflow activity. The KQ values attained from site 3 (Dajingliangzi area) arein the group of low activity.
The Spatiotemporal variations of the Rn concentrations in soil gas inthe study region with strong tectonic activity may reflect the change ofregional crustal stress/strain. The slip rate of fault also can reflect theactivity of fault related to stress/strain changes. Thus, whether there isa clear correlation between soil gas Rn activity and the slip rate of thefault, it is the place worth exploring. In this paper, we attempt to dis-cuss the relationship among the slip rate, tectonic activity of the faultsand the relative index of Rn activity (KQ) based on the soil gas datameasured in the field. The slip rates of the ANHF and the ZMHF havebeen studied in detail mainly in the Yangfushan (site 1), southernsegment of the ANHF and in the Dajingliangzi (site 3) of northernsegment of the ZMHF. No detailed study of the fault slip rate has beencarried out in the Xiaomiao Township (site 2), because the horizontaldislocation of the surface is difficult to be confirmed. Hence, we willfocus our discussion on site 1 and 3.
On the southern segment of the ANHF, the left-lateral displacementrate of this fault in late Holocene calculated by Ran et al. (2008) is6.2 mm/yr, using the methods of detailed geomorphic and geologicalsurvey. Xu et al. (2003b) estimated a slip rate of about 6.5 mm/yr basedon measurement of a gully offset on a terrace and thermoluminescence(TL) dating. On the northern segment of the ZMHF, Wang et al. (2011,2014) estimated its left-lateral displacement rate to be 2.4–3.6 mm/yrbased on the field investigations, trench excavations and radiocarbonage data. The results of previous studies show that the strike-slip
Fig. 7. Geological profiles of the survey sites(modified after Seismological Bureau of SichuanProvince, 2013a,b), the trace of geological pro-files are shown in Fig. 2.
Fig. 8. Curve for comparison of soil gas Rn anomalies revealed in study area in terms ofcontrast, KQ values of the profiles (Table 1) are shown in the curve in the order of de-creasing. The several profiles in site 1 are marked with red, in site 2 are purple and in site3 are sky-blue. (For interpretation of the references to colour in this figure legend, thereader is referred to the web version of this article.)
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
31
movement velocity of the southern segment of the ANHF is higher thanthat of the northern segment of the ZMHF. On the fault activity, theindex KQ of the Yangfushan profiles on the southern segment of theANHF ranged from 3.3 to 9.1. The southern segment of this fault isbelonging to groups with high and medium activity. However, the KQ
value of the Dajingliangzi profiles on the northern segment of the ZMHFranged from 2.1 to 2.5 and falls in the group with low activity (seeFig. 8). It is obvious that the KQ value of southern segment of the ANHFis higher than that of the northern segment of the ZMHF. The researchresults indicate that the strike-slip fault with high slip rate has higherKQ value. The reliability and rationality of the parameter KQ proposedby Seminsky and Bobrov (2009) and Seminsky and Demberel (2013) isverified again and it can indeed reflect the level of the geodynamicactivity of the fracture. In addition, the observation results show thatthe background value and maximum value of Rn concentrations in thesouthern segment of the ANHF are significantly higher than that of thenorthern segment of the ZMHF, indicating that the fault with higher sliprate is more conducive to discharge gas from the deep to the surface.Based on the profile survey data, for the first time in the western Si-chuan, it is concluded that the tectonic activity of the south segment ofthe ANHF is stronger than that of the north segment of the ZMHF. Thetectonic activity of the Xiaomiao Township, the intersection part of theANHF and the ZMHF, is between them.
7. Conclusion
The Rn concentrations in soil gas have been measured across theANHF and the ZMHF for the first time. The results of the profile surveysconducted for the three sites in the Western Sichuan show that the Rnconcentrations are very unevenly changed along the fault. The highvariability of the Rn concentrations near faults is deduced to be asso-ciated with the complex structure of the fault zones and the perme-ability differences of the underlying bedrock according to the geologicalsections across the fault. The background and maximum value of the Rnconcentrations in the southern segment of the ANHF are significantlyhigher than that of the northern segment of the ZMHF. It indicates thatthe faults with high slip rate are more favorable for discharging gasfrom the deep.
The relative index KQ of the Rn activity of the southern segment ofthe ANHF is obviously higher than that of the northern segment of theZMHF. It indicates that tectonic activity of the south segment of theANHF is stronger than that of the north segment of the ZMHF. Thetectonic activity level of the intersection part of the ANHF and theZMHF is in the median of them.
Acknowledgments
This work is financially supported by the Natural ScienceFoundation of China (Grant Nos 41573121 and 41402298), ChinaEarthquake Science Experiment Field Project (Grant Nos 20150113 and20150105) and Sichuan Earthquake Science and Technology Project(Grant Nos LY1701). The authors thank the two anonymous reviewerswhose criticisms and suggestions helped to greatly improve the article.The authors are grateful to Prof. Fang Du, Jiang Wu and XiaochengZhou for helpful comments on earlier manuscript. The first author isalso grateful for the help as a visitor during 3 month study at theDepartment of Seismic Fluids, Institute of Earthquake Science in 2016.
References
Al-Hilal, M., Al-Ali, A., 2010. The role of soil gas radon survey in exploring unknownsubsurface faults at Afamia B dam, Syria. Radiat. Meas 45, 219–224.
Allen, C.R., Luo, Z., Qian, H., Wen, X., Zhou, H., Huang, W., 1991. FIeld study of a highlyactive fault zone: the Xianshuihe fault of southwestern China. Geol. Soc. Am. Bull.103, 1178–1199.
Baubron, J.-C., Rigo, A., Toutain, J.-P., 2002. Soil gas profiles as a tool to characteriseactive tectonic areas: the jaut pass example (Pyrenees, France). Earth Planet Sci. Lett.
196, 69–81.Bernard, P., 2001. From the search of 'precursors' to the research on 'crustal transients'.
Tectonophysics 338, 225–232.Cho, J.S., Ahn, J.K., Kim, H.C., Ramola, R.C., 2004. Radon concentration in groundwater
in Busan measured with a liquid scintillation counter method. J. Environ. Radioact.75, 105–112.
Choubey, V.M., Bist, K.S., Saini, N.K., Ramola, R.C., 1999. Relation between soil gasradon variation and different lithotectonic units, Garhwal Himalaya, India. Appl.Radiat. Isot. 51, 587–592.
Ciotoli, G., Guerra, M., Lombardi, S., Vittori, E., 1998. Soil gas survey for tracing seis-mogenic faults: a case study in the Fucino basin, central Italy. J. Geophys. Res. 103,23781–23794.
Ciotoli, G., Lombardi, S., Annunziatellis, A., 2007. Geostatistical analysis of soil gas datain a high seismic intermontane basin: fucino Plain, central Italy. J. Geophys. Res. 112,2637–2655.
Department of Earthquake Disaster Prevention, China Earthquake Administration (DEDP,CEA), 1995. The Catalogue of Chinese Historical Strong Earthquakes. SeismologicalPress, Beijing, China 514 In Chinese.
Fu, C.C., Yang, T.F., Du, J., Walia, V., Chen, Y.G., Liu, T.K., Chen, C.H., 2008. Variationsof helium and radon concentrations in soil gases from an active fault zone in southernTaiwan. Radiat. Meas. 43, S348–S352.
Fu, C.C., Yang, T.F., Tsai, M.C., Lee, L.C., Liu, T.K., Walia, V., Chen, C.H., Chang, W.Y.,Kumar, A., Lai, T.H., 2017. Exploring the relationship between soil degassing andseismic activity by continuous radon monitoring in the Longitudinal Valley of easternTaiwan. Chem. Geol. http://dx.doi.org/10.1016/j.chemgeo.2016.12.042.
Fu, C.C., Yang, T.F., Walia, V., Chen, C.H., 2005. Reconnaissance of soil gas compositionover the buried fault and fracture zone in southern Taiwan. Geochem. J. 39, 427–439.
Geological Bureau of Sichuan Province, 1966. 1: 200,000 Geological Map (Xichang).Geological Publishing House, Beijing (In Chinese).
Ghosh, D., Deb, A., Sengupta, R., 2009. Anomalous radon emission as precursor ofearthquake. J. Appl. Geophys. 69 (2), 67–81.
Ghosh, D., Deb, A., Sengupta, R., Bera, S., Sahoo, S.R., Haldar, S., Patra, K.K., 2011.Comparative study of seismic surveillance on radon in active and non-active tectoniczone of West Bengal, India. Radiat. Meas. 46, 365–370.
Giammanco, S., Immè, G., Mangano, G., Morelli, D., Neri, M., 2009. Comparison betweendifferent methodologies for detecting radon in soil along an active fault: the case ofthe Pernicana fault system, Mt. Etna (Italy). Appl. Radiat. Isot. 67, 178–185.
Guerra, M., Lombardi, S., 2001. Soil-gas method for tracing neotectonic faults in claybasins: the Pisticci field (Southern Italy). Tectonophysics 339, 511–522.
Han, X., Li, Y., Zhou, X.C., Zhang, W., 2014. Rn and CO2 geochemistry of soil gas acrossthe active fault zones in the capital area of China. Nat. Hazards Earth Syst. Sci. 14,2803–2815.
Hinkle, M.E., 1994. Environmental conditions affecting concentrations of He, CO2, O2 andN2 in soil gases. Appl. Geochem 9, 53–63.
Iakovleva, V.S., Ryzhakova, N.K., 2003. Spatial and temporal variations of radon con-centration in soil air. Radiat. Meas. 36, 385–388.
Immè, G., Delfa, S.L., Nigro, S.L., Morelli, D., Patanè, G., 2005. Gas radon emission relatedto geodynamic activity on Mt. Etna. Ann. Geofis. 48 (1), 65–71.
Ioannides, K., Papachristodoulou, C., Stamoulis, K., Karamani, D., Pavlides, S.,Chatzipetros, A., Karakala, E., 2003. Soil gas radon: a tool for exploring active faultzones. Appl. Radiat. Isot. 59, 205–213.
Irwin, W.P., Barnes, I., 1980. Tectonic relations of carbon dioxide discharges and earth-quakes. J. Geophys. Res. 85, 3115–3121.
King, C.Y., 1986. Gas geochemistry applied to earthquake prediction: an overview. J.Geophys. Res. 91, 12269–12281.
King, C.Y., King, B.S., Evans, W.C., Zhang, W., 1996. Spatial radon anomalies on activefaults in California. Appl. Geochem 11, 497–511.
King, C.Y., Zhang, W., King, B.S., 1993. Radon anomalies on three kinds of faults inCalifornia. Pure Appl. Geophys. 141 (1), 111–124.
King, C.Y., Zhang, W., Zhang, Z., 2006. Earthquake-induced groundwater and gaschanges. Pure Appl. Geophys. 163, 633–645.
Koike, K., Yoshinaga, T., Ueyama, T., Asaue, H., 2014. Increased radon-222 in soil gasbecause of cumulative seismicity at active faults. Earth Planets Space 66, 57.
Lehmann, B.E., Lehmann, M., Neftel, A., Tarakanov, S.V., 2000. Radon-222 monitoring ofsoil diffusivity. Geophys. Res. Lett. 27 (23), 3917–3920.
Li, Y., Du, J.G., Wang, X., Zhou, X.C., Xie, C., Cui, Y.J., 2013. Spatial variations of soil gasgeochemistry in the Tangshan area, Northern China. Terr. Atmos. Ocean. Sci. 24 (3),323–332.
Lombardi, S., Voltattorni, N., 2010. Rn, He and CO2 soil gas geochemistry for the study ofactive and inactive faults. Appl. Geochem 25, 1206–1220.
López, J., Dena, O., Sajobohus, L., Rodríguez, G., Chavarría, I., 2016. Correlation betweenunderground radon gas and dormant geological faults. J. Nucl. Phys. Mat. Sci. 4 (1),265–275.
Moreno, V., Bach, J., Font, L., Baixeras, C., Zarroca, M., Linares, R., Roqué, C., 2016. Soilradon dynamics in the Amer fault zone: an example of very high seasonal variations.J. Environ. Radioact. 151, 293–303.
Moussa, M.M., El Arabi, A.-G.M., 2003. Soil radon survey for tracing active fault: a casestudy along Qena-Safaga road, Eastern Desert. Egypt. Radiat. Meas. 37 (3), 211–216.
Papp, B., Deak, F., Horvath, A., Kiss, A., Rajnai, G., Szabo, C., 2008. A new method for thedetermination of geophysical parameters by radon concentration measurements inbore-hole. J. Environ. Radioact. 99, 1731–1735.
Pei, X.Y., Wang, X.M., Zhang, C.G., 1988. Basic character of segmentation of theQuaternary movement on the Anninghe fault. Earthq. Res. Sichuan 4, 52–61 (inChinese with English Abstract).
Peltzer, G., Tapponnier, P., 1988. Formation and evolution of strike-slip faults, rifts, andbasins during the India-Asia collision: an experiment approach. J. Geophys. Res. 93
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
formation and displacement rates of the Anninghe fault around Zimakua. Seismol.Geol. 30 (1), 86–98 (in Chinese with English Abstract).
Ren, Z.J., 2014. Late quaternary deformation features along the Anninghe fault on theeastern margin of the tibetan plateau. J. Asian Earth Sci. 85, 53–65.
Ren, Z.J., Lin, A.M., Rao, G., 2010. Late pleistocene-holocene activity of the Zemuhe faulton the southeastern margin of the tibetan plateau. Tectonophysics 495, 324–336.
Seismological Bureau of Sichuan Province, 2013a. 1: 50,000 Geological Map of theAnninghe Active Fault. Earthquake Publishing House, Beijing (In Chinese).
Seismological Bureau of Sichuan Province, 2013b. 1: 50,000 Geological Map of theZemuhe Active Fault. Earthquake Publishing House, Beijing (In Chinese).
Seminsky, K.Zh, Bobrov, A.A., 2009. Radon activity of faults (western Baikal and southernAngara areas). Russ. Geo. Geophys. 50, 682–692.
Seminsky, K.Zh, Demberel, S., 2013. The first estimations of soil-radon activity near faultsin Central Mongolia. Radiat. Meas. 49, 19–34.
Sinclair, A.J., 1991. A fundamental approach to threshold estimation in explorationgeochemistry: probability plots revisited. J. Geochem. Explor 41, 1–22.
Tapponnier, P., Peltzer, G., Le Dian, A.Y., Armijo, R., Cobbold, P., 1982. Propagatingextrusion tectonics in Asia: new insights from simple experiments with plasticine.Geology 10 (12), 611–616.
Torabi, A., Berg, S.S., 2011. Scaling of fault attributes: a review. Mar. Pet. Geol. 28,1444–1460.
Toutain, J.P., Baubron, J.C., 1999. Gas geochemistry and seismotectonics: a review.Tectonophysics 304, 1–27.
Utkin, V.I., Yurkov, A.K., 2010. Radon as a tracer of tectonic movements. Russ. Geo.Geophys. 51, 220–227.
Walia, V., Mahajan, S., Kumar, A., Singh, S., Bajwa, D.S., Dhar, S., Yang, T.F., 2008. Faultdelineation study using soil-gas method in the Dharamsala area, NW Himalayas,India. Radiat. Meas. 43, S337–S342.
Walia, V., Yang, T.F., Hong, W.L., Lin, S.J., Fu, C.C., Wen, K.L., Chen, C.H., 2009.Geochemical variation of soil-gas composition for fault trace earthquake precursorystudies along the Hsincheng Fault in NW Taiwan. Appl. Radiat. Isot. 67 1855–1683.
Wang, H., Ran, Y.K., Li, Y.B., 2011. Growth of a small pull-apart basin and slip rate of
strike-slip fault: with the example of Zemuhe Fault on the southeastern margin of theTibetan plateau. Seismol. Geol. 33 (4), 818–827 (in Chinese with English abstract).
Wang, H., Ran, Y.K., Li, Y.B., Chen, L.C., 2013. Paleoseismic ruptures and evolution of asmall triangular pull-apart basin on the Zemuhe fault. Sci. China Earth Sci. 56,504–512.
Wang, H., Ran, Y.K., Li, Y.B., Chen, L.C., 2014. Paleoseismic behavior of the Anninghefault and its comparison with the Zemuhe fault in western sichuan. Seismol. Geol. 36(3), 706–717 (in Chinese with English abstract).
Wen, X.Z., Ma, S.L., Xu, X.W., He, Y.N., 2008. Historical pattern and behavior of earth-quake ruptures along the eastern boundary of the Sichuan-Yunan faulted-block,southwestern China. Phys. Earth. Planet 168, 16–36.
Winkler, R., Ruckerbauer, F., Bunzl, K., 2001. Radon concentration in soil gas: a com-parison of the variability resulting from different methods, spatial heterogeneity andseasonal fluctuations. Sci. Total Environ. 272, 273–282.
Xu, X.W., Wen, X.Z., Zheng, R.Z., Ma, W.T., Song, F.M., Yu, G.H., 2003a. Pattern of latesttectonic motion and its dynamics for active blocks in Sichuan-Yunnan region, China.Sci. China Earth Sci. 46, 210–226.
Xu, X.W., Wen, X.Z., Zheng, R.Z., Ma, W.T., Song, F.M., Yu, G.H., 2003b. Newly tectonicdeformation and its dynamic source of the Sichuan-Yunnan Block. Sci. China EarthSci. 33 (S1), 151–162 (in Chinese).
Yang, T.F., Fu, C.C., Walia, C.H., Chen, C.H., Chyi, L.L., Liu, T.K., Song, S.R., Lee, M., Lin,C.W., Lin, C.C., 2006. Seismo-geochemical variations in SW Taiwan: multi-parameterautomatic gas monitoring results. Pure Appl. Geophys. 163 793–709.
Yuce, G., Fu, C.C., D'Alessandro, W., Gulbay, A.H., Lai, C.W., Bellomo, S., Yang, T.F.,Italiana, F., Walia, V., 2017. Geochemical characteristics of soil radon and carbondioxide within the dead sea fault and karasu fault in the amik basin (hatay). Turk.Chem. Geol. http://dx.doi.org/10.1016/j.chemgeo.2017.01.003.
Zhang, P.Z., Deng, Q.D., Zhang, G.M., Ma, J., Gan, W.J., Min, W., Mao, F.Y., Wang, Q.,2003. Active tectonic blocks and strong earthquakes in the continent of China. Sci.China Earth Sci. 46, 13–24.
Zhang, X.C., Yang, Z.H., Ma, Z.S., Tang, J.H., 2006. Geochemical background and geo-chemical baseline. Geo. Bull. China 25 (5), 626–629 (in Chinese with English ab-stract).
Zhou, X.C., Du, J.G., Chen, Z., Cheng, J.W., Tang, Y., Yang, L.M., Xie, C., Cui, Y.J., Liu, L.,Yi, L., Yang, P.X., Li, Y., 2010. Geochemistry of soil gas in the seismic fault zoneproduced by the Wenchuan Ms=8.0 earthquake, southwestern China. Geochem.Trans. 11, 5.
Y. Yang et al. Applied Geochemistry 89 (2018) 23–33
