Characteristics and implications of the stress state in the Longmen Shan fault zone, eastern margin of the Tibetan Plateau

Tectonophysics 656 (2015) 1–19

Contents lists available at ScienceDirect

Tectonophysics journal homepage: www.elsevier.com/locate/tecto

Characteristics and implications of the stress state in the Longmen Shan fault zone, eastern margin of the Tibetan Plateau Wen Meng a,b, Qunce Chen a,b,⁎, Zhen Zhao c, Manlu Wu a,b, Xianghui Qin a,b, Chongyuan Zhang a,b a b c

Institute of Geomechanics, Chinese Academy of Geological Sciences, Beijing 100081, China Key Laboratory of Neotectonic Movement and Geohazard, Ministry of Land and Resources, Beijing 100081, China Chinese Academy of Geological Sciences, Beijing 100037, China

a r t i c l e

i n f o

Article history: Received 29 August 2014 Received in revised form 15 April 2015 Accepted 22 April 2015 Available online 1 May 2015 Keywords: Longmen Shan thrust belt Stress state Hydraulic fracturing Coulomb frictional-failure criterion Fault activity

a b s t r a c t Using stress data measured in 16 boreholes along the strike of the Longmen Shan fault zone by hydraulic fracturing from 2008 to 2012 after the Wenchuan earthquake and before the Lushan earthquake, we characterize the contemporary stress state in the Longmen Shan thrust belt along the eastern margin of the Tibetan Plateau to understand the implications of in-situ stress for fault activity. The stress regimes are generally conducive to reverse faulting and partly to strike-slip faulting characterized by σH N σh N σv and σH N σv N σh, indicating that the regional stress field is definitely dominated by the maximum horizontal stress. The fracture impression results reveal that the maximum horizontal principal stresses are predominantly NE in the northern segment of the Longmen Shan fault zone and NW in the southern segment, postulating a preliminary understanding of the coupling between the shallow crustal stress field and lower crustal flow. According to Coulomb frictional failure criteria, horizontal principal stresses can be predicted as functions of rock density, ρ, frictional coefficient, μ, depth, H, and water level, HW, in frictional equilibrium. The influence of HW on critical stresses is discussed, and the decrease in the stress values corresponds to an increase in the water level. The depth profiles of the stress magnitudes in different segments are illustrated, indicating that the stress values are relatively higher in the southern and northern segments and lower in the middle segment. The stress state in the southern segment, specifically, near the epicenter of the Lushan earthquake, favors the occurrence of earthquakes. Under the stress state in the northern segment, the Longmen Shan fault might be the optimally oriented failure plane, assuming that the plane is critically stressed. This finding may imply that the northern segment of the Longmen Shan fault is likely to be active when the stress builds up sufficiently to destroy the frictional equilibrium. © 2015 Published by Elsevier B.V.

1. Introduction Stress measurements play an essential role in geodynamic process analyses, such as those used to determine plate tectonic driving forces and crustal stability. Measurements of the stress field within the crust can provide the most useful information concerning the forces responsible for various tectonic processes, such as earthquakes (McGarr and Gay, 1978). Earthquakes are a major hazard and commonly result in great disasters worldwide. Fault instability and earthquakes have a close relationship with crustal stress. Exploring the crustal stress state and its characteristics is critical to understand the physical processes inside the crust and fault activity (Li, 1973). Previous studies have focused on fault stability, crustal activity, and seismogenics by using measurement stress data. Liao et al. (2003) and Guo et al. (2009) revealed an obvious stress change before and after the 2001 MS 8.1 West Kunlun Pass and 2008 Wenchuan earthquakes. Lin et al. (2011, 2013) analyzed the ⁎ Corresponding author at: No.11 Minzu University South Road, Haidian District, Beijing, China. Tel.: 86 010 88815081; fax: 86 010 68422326. E-mail address: [email protected] (Q. Chen).

http://dx.doi.org/10.1016/j.tecto.2015.04.010 0040-1951/© 2015 Published by Elsevier B.V.

stress state before and after the 2011 Tohoku-Oki earthquake, determining that the aftershock faulting stress regime clearly differed from that prior to the earthquake and fault slip could result in a completely reduced stress state. Chang et al. (2010) analyzed the interaction between the regional stress state and faults by examining borehole insitu stress data and the earthquake focal mechanism in southeastern Korea. Wu et al. (2013) analyzed stress changes before and after large earthquakes using both hydraulic fracturing and piezomagnetic methods in the Tibetan Plateau. We characterize the present-day stress state in the Longmen Shan fault zone along the eastern margin of the Tibetan Plateau, which is one of the most developed areas of seismic activity and has complex geological structures and active faults caused by the continued northward subduction of the Indian continental plate (Molnar and Tapponnier, 1975; Tapponnier and Molnar, 1977). The Longmen Shan fault zone and surrounding blocks, as part of the eastern boundary of the Tibetan Plateau and located in the middle section of the famous north-south seismic belt in China, consists of a series of parallel imbricate thrust belts (Fig. 1) that trend 40°N–50°E, with a total length of 500 km and width of 25–40 km. This zone contains the Maoxian-Wenchuan,

2

W. Meng et al. / Tectonophysics 656 (2015) 1–19

Fig. 1. Geological sketch map of the Longmen Shan region along the eastern margin of the Tibetan Plateau. TIBET, Tibetan Plateau; YZ, Yangtze block; LMSF, Longmen Shan fault zone; XSHF, Xianshuihe fault zone.

Pingwu-Qingchuan, Yingxiu-Beichuan, and Anxian-Guanxian faults (Xu et al., 2008; Dirks et al., 1994; Wu et al., 2009). The Longmen Shan fault zone has experienced the most devastating earthquake of the century, the 2008 Ms 8.0 Wenchuan earthquake, which collapsed buildings and killed thousands in major cities in the western Sichuan Basin in China (Parsons et al., 2008). This earthquake produced a seismic rupture zone of ~240 km along the causative fault, the Yingxiu–Beichuan fault, and a surface rupture zone of ~70 km along the Anxian-Guanxian fault (Lin et al., 2009; Xu et al., 2009). Moreover, thousands of large-scale landslides and debris flows were triggered in the Longmen Shan Mountains. Five years later, the Ms 7.0 Lushan earthquake occurred on April 20, 2013, in the southwestern Longmen Shan fault zone, inducing geological hazards and resulting in the deaths or disappearances of more than 200 people. Because the stress state is responsible for present-day earthquake activity, its characterization is essential. In the present study, the results of hydraulic fracturing in-situ stress measurements conducted in 16 boreholes from 2008 to 2012 after the Wenchuan earthquake and before the Lushan earthquake are used to characterize the present-day stress state in the Longmen Shan fault zone, and the

implications of the in-situ stress in terms of the stress magnitude and orientation are discussed for fault activity. 2. In-situ stress measurements 2.1. General conditions of the test boreholes As shown in Fig. 1, HZ-1 and HZ-2 are located in the transition zone between the northeastern Longmen Shan thrust belt and western margin of the Hanzhong Basin at a distance of ~33 m apart. Hydraulic fracturing stress measurements were performed in Proterozoic granite. GY-1 is located in the footwall of the Longmen Shan fault zone and contains exposed feldspar quartz sandstone and purple siltstone from the Middle Jurassic Shaximiao Formation (J2s). GY-2 is also located in the footwall of the Longmen Shan central fault, ~ 2 km from the YingxiuBeichuan fault, and contains exposed dolomite, quartz sandstone, and siltstone from the Middle Devonian Guanwushan Formation (D2g). PW-1 and PW-2 are located in the hanging wall of the YingxiuBeichuan fault, ~30 m apart, and contain exposed biotite leptynite and

W. Meng et al. / Tectonophysics 656 (2015) 1–19

3

Fig. 2. Schematic illustration of the hydraulic fracturing system.

pebbly biotite leptynite from the Lower Sinian Muzuo Formation (Z1m). Although parts of the test intervals developed fractures, the measurement results show high breakdown pressure and rock tensile strength. The rocks in PW-1 and PW-2 are hard and dense, resulting in stress accumulation in complete sections. JY-1 is ~ 19 km from the YingxiuBeichuan fault. The strata are mainly purple fine sandstone from the Upper Jurassic Lianhuakou Formation (J3l) and include mudstone and

Fig. 3. Typical impression packer test record at 96.00 m depth in GY-1. (a) Packer picture with imprints of induced fractures; (b) schematic diagram of an impression packer test record showing hydraulic fracturing traces on the borehole wall.

shale. BC-1 is ~3 km from the Yingxiu-Beichuan fault and ~1 km from BC-2. The strata exposed at the two sites are both sandstone and siltstone from the Middle Devonian Ganxi Formation (D2g). LG-1 is situated in the hanging wall of the Yingxiu-Beichuan fault, ~ 2.5 km away. Hydraulic fracturing stress measurements were conducted in Silurian limestone. YA-1 is located in the footwall of the Yingxiu-Beichuan fault and the hanging wall of the Anxian-Guanxian fault, which is less than 1 km away, and contains exposed limestone and dolomitic limestone, with underlying bedrock from purple Permian sandstone. YX-1 is located ~13 km northeast from the epicenter of the Ms 8.0 Wenchuan earthquake and was drilled into the granitic rocks in the Peng-Guan

Fig. 4. Acoustic televiewer images showing the pre-existing natural fractures (a) and intact rock section (b).

4 W. Meng et al. / Tectonophysics 656 (2015) 1–19

Fig. 5. Typical records of pressure versus time for hydraulic fracturing in boreholes GY-1, GY-2, PW-1, LG-1, YA-1 and HZ-1.

W. Meng et al. / Tectonophysics 656 (2015) 1–19

5

Fig. 6. Graph showing the determination of the reopening pressure at 240.00 m depth in borehole GY-1.

complex in the hanging wall of the central Longmen Shan fault. Stress measurements were conducted just 4 months after the strong Wenchuan earthquake. The hydraulic fracturing stress measurements at BX-1 were carried out in the granite rock in the Bao-Xing complex, which is a major tectonic feature of the southwestern segment of the Longmen Shan fault zone. The site is ~20 km from the epicenter of the Ms 7.0 Lushan earthquake. KD-1 and KD-2 are located at the conjunction zone between the NE-trending Longmen Shan thrust belt and the NWstriking Xianshuihe fault zone. KD-1 is in granite rock, and the lithology exposed at KD-2 is metamorphic quartz sandstone. SM-1 is situated at the southwestern end of the Longmen Shan fault zone. The lithology of the test intervals is mainly quartzite with minor interbeds of mud shale. 2.2. Methods of in-situ stress measurement Hydraulic fracturing is a borehole field-test method suggested by the International Society for Rock Mechanics (ISRM) and is designed to assess the state of in-situ stress in the Earth's crust (Haimson and Cornet, 2003). Successful hydraulic fracturing measurements generally result in an estimate of the in-situ principal stresses, including both their magnitudes and directions, in the plane perpendicular to the borehole axis, assuming that the rock mass is linearly elastic, homogeneous, and isotropic. The vertical stress is taken as being principal and equal to the overburden weight. The equipment used for the measurements were developed and improved by the authors' institute; see Fig. 2. The measurements are based on the principle of sealing off a section (i.e., test interval) in a borehole using two inflatable rubber packers that are sufficiently pressurized so that they can adhere to the borehole wall. These packers have a diameter of 72 mm for boreholes with a diameter of 76 mm. Water was pumped into the test interval using an electrical pump on the surface with a maximum working pressure of 50 MPa, gradually increasing the pressure on the borehole wall until a fracture is initiated in the rock. A mechanically controlled transfer valve on the top of the packer was able to switch from packer pressurization to injection into the test interval. The flow rate was controlled by a pressure control unit. A pressure meter was employed to give real-time information of the hydraulic fluid pressure, and a pressure sensor was used to monitor and transmit the pressure data to a recording device, including a 12 bite A/D (analog/digital) board and a computer. Thus, pressure–time curves were recorded at the same time. A follow-up re-fracturing test was carried out by injecting water into the test interval again until the previously closed fractures were reopened. The pressurizations in each isolated test interval were repeated for 5 cycles to induce hydrofractures, which were reopen and shut-in in our measurements. The key pressure values used in the computation

of the in-situ stresses are picked from the pressure–time records, including the breakdown pressure (Pb), instantaneous shut-in pressure (Ps), and reopening pressure (Pr). Then, the maximum and minimum horizontal principal stresses (σH and σh, respectively), and vertical principal stress (σv) can be obtained according to the correlation formulas given by Eqs. (1)–(3) (Haimson, 1978; Klee et al., 1999; Haimson and Cornet, 2003). The orientation of the maximum horizontal principal stress was determined using an impression packer with an automatic orientation device. The impression packer was covered by a sulfonated rubber layer, and the distinct geometric copy of the induced fracture was recorded on the packer surface due to the strong squeeze from the rubber wall into the fracture, as illustrated in Fig. 3. The strike of the fracture traced on the packer was determined by the automatic orientation device. The direction of the maximum horizontal principal stress is perpendicular to that of the minimum horizontal principal stress, shown in Eq. (4).

σ H ¼ 3P s −P r −P P

ð1Þ

σ h ¼ Ps

ð2Þ

σ v ¼ γH

ð3Þ

σ H direction ¼ direction of vertical hydraulic fracture strike

ð4Þ

where PP is the pore pressure, γ is the unit weight of the overlying rock, and H is the depth of the test interval. PP is an essential parameter to determine the maximum horizontal principal stress. The present research and actual measurements show that the pore pressure is approximately equal to the hydrostatic pressure in low permeability rocks in the shallow crust (Zoback and Haimson, 1982; Haimson and Doe, 1983; Moos and Zoback, 1990; Barton et al, 1995). We adopted this basic assumption in the present study. 3. Stress measurement results 3.1. Data process of the characteristic hydraulic pressure parameters To ensure the accuracy of the hydraulic fracturing measurements, leakage checking of the drilling rods and associated pipes in the process of pumping was performed before each measurement at a test pressure of 15–20 MPa, and the test boreholes were routinely image-logged using an acoustic televiewer before testing to ensure that the test intervals were devoid of pre-existing fractures. Based on the information in the borehole logs and acoustic televiewer images (Fig. 4), the test intervals were selected in intact rock mass sections.

6

W. Meng et al. / Tectonophysics 656 (2015) 1–19

specific as the interval pressure would decrease significantly once the induced fracture opened and grew further. The reopening pressure (Pr) was marked as the pressure in a subsequent cycle where the test interval pressure versus time curve departed from its tangent slope. However, there are uncertainty issues in determining the maximum principal stress by hydraulic fracturing using the reopening pressure (Ito et al., 1999; Rutqvist et al., 2000). The apparent reopening pressure detected in the conventional manner is larger than the true reopening pressure. As we used stiff drill rods to inject water, the discrepancy between the true and apparent reopening pressures is small, especially where the measurement boreholes are small (with a diameter of 76 mm) and the test intervals are relatively shallow (Ito et al., 1999; Zoback, 2007). Furthermore, five pressurization cycles (one fracturing test cycle and four reopening test cycles) were adopted at each test interval in our study. The last three injection cycles (2., 3., and 4. Refracs in Fig. 5) were used to determine the reopening pressure value so that the fractures could be adequately reopened, and we took the average value as the final reopening pressure (see Fig. 6) to decrease the error that was detected from a single cycle. The shut-in pressure (Ps) was the pressure reached when the hydraulically induced fracture closed (Haimson and Cornet, 2003). Because the shut-in pressure is considered equal to the minimum horizontal stress for vertical boreholes (Eq. (2)), it is strongly recommended that more than one method be used to obtain the crucial Ps parameter to ensure the reliability of the shut-in pressure value to ensure good hydraulic fracturing measurement results. The pressure decay rate versus pressure (dP/dt–P) method (Zoback and Haimson, 1982; Lee and Haimson, 1989), the inverted pressure decay rate versus pressure (dt/dP–P) method (Hayashi and Sakurai, 1989; Cheung and Haimson, 1989; Hayashi and Haimson, 1991), and the inflection point methods (Gronseth and Kry, 1983), all recommended by the ISRM (Haimson and Cornet, 2003), were adopted and the average values from these three calculations were used to determine the final Ps. Fig. 7 shows the graphical interpretations of Ps at a measurement depth of 240.00 m in borehole GY-1 by the three approaches using software developed by the authors' institute.

3.2. Stress magnitudes Based on the relationships between the pressure and time, we successfully obtained the characteristic hydraulic pressure parameters for all the test intervals. Because hydraulic fracturing is a 2D stress measurement approach that is only applicable to the determination of the maximum and minimum stresses in the horizontal plane, using Eqs. (1)–(2), the vertical stress is estimated from the overburden weight. As shown in Eq. (3), the unit weight of the overlying rock (γ)

Fig. 7. Graph showing three different methods for determining shut-in pressure (Ps) proposed by the ISRM. The graphs are an example from the GY-1 borehole at 240.00 m depth showing (a) the dt/dP vs. P method, (b) the dP/dt vs. P method, and (c) the inflection point method. The average value from these three calculations was used to calculate the final Ps.

Individual pressure–time curves were generated for each hydraulic fracturing measurement. Typical hydrofracture records in partial boreholes are shown in Fig. 5. Due to space limitations, the pressure–time curves for other test intervals cannot be presented, but their patterns are very consistent. The key pressures used to calculate the in-situ stress magnitudes should all be derived from the field measurements curves; thus, the accuracy of the hydrofracture stress measurements strongly depends on the correct interpretation of the pressure–time records obtained during the tests. The breakdown pressure (Pb) was taken as the peak pressure attained in the first pressure cycle (Frac in Fig. 5). Its value is usually

Fig. 8. Vertical stress estimated with the actual densities obtained by rock mechanics tests.

W. Meng et al. / Tectonophysics 656 (2015) 1–19

7

Fig. 9. Simplified geologic map of boreholes LG-1 and YA-1 (a) and KD-1 and KD-2 (b). Photograph of disked cores in KD-2 (c); pressure–time curve at 174.00 m in HZ-2 (d).

is critical to vertical stress estimation. Brown and Hoke (1978) summarized some measurement vertical stress values at various sites around the world, and the measured vertical stresses do cluster around the line of the lithostatic stress model, which corresponds to a mean rock density of 2700 kg/m3 and can be used to estimate the vertical stress when in-situ stress measurements are not available. This generic unit weight provides a good predictive estimate of the average vertical stress from global stress data (Zoback, 2007) but is not appropriate to estimate specific test boreholes. In our measurements, drill cores at different depths along some boreholes were selected to conduct rock mechanics tests, and we took the average unit weight of the overlaying rocks, 26.8 kN/m3 (see Fig. 8), as the generic value to evaluate the vertical stress in analyses such as Fig. 11. For individual test boreholes, the actual density corresponding to the lithology based on density tests on extracted core samples was used to calculate the vertical stress, as listed in Table 1. Normally, in-situ stresses at very shallow depths tend to be subjected to topography effects. We rejected data that were not appropriate to be used for regional stress characterization based on some criteria. First, measurement curves that were not standard were not considered because the key pressures used to calculate the in-situ stress magnitudes

should all be derived from the field measurement curves. Second, shallow stress orientations that deviated from the deeper stress orientations were ruled out because they were likely to be perturbed by topography. Third, we left out data measured at very shallow depths, which would decrease the correlation coefficient (R) of the best-fit line for each individual borehole. Stress magnitude plots for individual boreholes are illustrated in Fig. 10. As the number of data is small for a given borehole, the value of R is not very high. Boreholes HZ-2, GY-2, PW-1, LG-1, KD-2, and SM-1 show the most scattered datasets. Note that the rock type in boreholes PW-1, PW-2, KD-2, and SM-1 is metamorphic rock, in which the stress is considered to be more scattered with depth than other rocks such as sedimentary and magmatic rocks (Zhu and Tao, 1994). Whether the scattered data in these boreholes are suitable to such interpretation is indeterminate, as the previous study was based on large scale data from around the world. For LG-1, we consider that the YingxiuBeichuan fault should be the main influencing factor, as illustrated in Fig. 9a. Due to the nearby fault (maybe the Wenchuan earthquake), the stress data in LG-1 are much lower and scattered than those in YA-1, which was in a relatively simple tectonic environment. The stress

8

W. Meng et al. / Tectonophysics 656 (2015) 1–19

Fig. 10. Stress magnitude plots for individual boreholes.

magnitudes in KD-2 are much higher than those in KD-1, which is not far away (see Fig. 9b). With high RQD values greater than 94 % (i.e., excellent rock quality), the cores in this borehole are hard and

intact. In addition, disked cores appeared at depths of 171.02– 173.00 m, which is a phenomenon when diamond drill cores are retrieved from brittle hard rock masses in high stress environments

W. Meng et al. / Tectonophysics 656 (2015) 1–19

Fig. 11. Depth profile of the principal stresses in the Longmen Shan thrust belt (a), and the in-situ stress ratios determined from the stress data (b).

(Lim and Martin, 2010), as shown in Fig. 9c. The stress magnitude at a depth of 174.00 m in HZ-2 disturbs the regular stress trends with depth. As seen in Fig. 9d, the pressure–time curve of this test interval is good for deriving stress parameters; therefore we take it as a good data. Fig. 11 shows the magnitudes of the horizontal principal stresses measured in the shallow crust in the Longmen Shan fault zone and as well as the vertical stress calculated from the overburden (an average rock density of 2.68 g/cm3 is assumed). Note that the horizontal principal stresses are notably scattered in the test depths, varying from approximately 50 to 600 m as the dataset was collected from different locations with different lithologies, local structures, etc. Nonetheless, the stress magnitudes obtained by hydraulic fracturing are the direct responses of the in-situ stress state in the region, which should give insight into the stress information for the region, at least for the depths they cover (Chang et al., 2010). Despite this scattering, the horizontal principal stress values are definitely higher than those of the vertical stress; thus, the prevailing stress regime in the region favors reverse faulting (σH N σh N σv) and, partly, strike-slip faulting (σH N σv N σh) (Anderson, 1951), which is a consequence of the ongoing squeezing pushover process from the Tibetan Plateau to the Sichuan Basin. Yang et al. (2012) analyzed the characteristics of the measured stress in the Chinese mainland and its active blocks and the North– South seismic belt based on the “Database of Crustal Stress in China and Adjacent Area” and obtained regression equations between the stresses and depths of different study areas. Jing et al. (2007) collected stress data from approximately 400 drillings in the Chinese mainland, mainly derived from hydraulic fractures and borehole breakouts, and established the trends of the principal stresses with depth. Zhu and Tao (1994) collected in-situ stress data from various rocks worldwide and revealed the distribution characteristics of the principal stresses with depth. The horizontal principal stress dataset is poorly constrained because the measurements that were conducted vary spatially (Fig. 11), so the best-fit linear line was fitted for each given test borehole. Although this may be meaningless as the number of data for individual locations is small (Chang et al., 2010), we should obtain general information for this region, namely, that the measured stress variations

9

with depth in the Longmen Shan fault zone differ from those in larger study areas (see Table 2). It is widely considered that horizontal principal stress is stronger than vertical stress because horizontal tectonic activities such as plate movement and block collision control the formation of shallow crustal stress. However, the dominant role of horizontal stress weakens with depth because the increasing rate of vertical stress, normally derived from the overlying rock density, is usually higher than that of horizontal stress. This may result in dominant stress transitions (Yang et al., 2012) or a hydrostatic state in the deep crust (Zoback, 2007). Gao et al. (1987) used a depth of 1500 m as a variation key point for σH/σv. Yang et al. (2012) calculated the transition depth of the in-situ stress state in each study region of China to be 170–309 m and 714–4667 m for when the vertical stress became the intermediate and maximum principal stresses, respectively. However, in the Longmen Shan thrust belt, the horizontal principal stress is always higher than the vertical stress in the measurement depth range. We conclude that the most definite feature of stress state around Longmen Shan fault zone which locates at the eastern boundary of the Tibetan Plateau is represented by the maximum horizontal stress being the maximum principal stress, caused by collision of Tibetan Plateau and the Sichuan Basin. The above discussion indicates that the region characterizes the stress state. The horizontal principal stress magnitudes as functions of depth appear to be distinct in different regions (Amadei and Stephansson, 1997). The stress states differ among the study areas because of factors such as geological structure, topography, and lithology. Because of the complex tectonic environment, the Longmen Shan thrust belt has a special stress state. Furthermore, the regression equations between the stresses and depth are restricted by the number and depth ranges of the measurement data. In addition to the principal stresses, the magnitudes of the in-situ stress ratios determined from the stress data are illustrated in Fig. 11b. The ratio of σH to σh (KHh) is mainly limited between 1.0 and 2.0 and only varies slightly with depth. The difference between the maximum and minimum horizontal principal stresses is small, with an average value of 1.48. The ratios of σH to σv (KHv) and σh to σv (Khv) and the ratio of the average of the horizontal stresses to the vertical stress, defined as Ka = (σH + σh)/2σv, mostly exceed values of 1.0. The coefficients KHv, Khv, and Ka appear to be scattered in the superficial crust and more concentrated in the deeper crust (under 400 m), converging to 1.47, 0.97 and 1.22. Brown and Hoek (1978) used Ka to study the changes in in-situ stress with depth. Jing et al. (2007) took KHv as a more effective parameter to reflect the horizontal stress intensity considering the significant difference between the maximum and minimum horizontal principal stresses, particularly in shallow positions. We did not give fitting expressions for the stress ratios because of the limited deep stresses, and the shallow stress data would significantly affect the fitting. Thus, a general trend for the stress state in the Longmen Shan fault zone based on hydraulic fracturing should be obtained. 3.3. Stress orientation The maximum horizontal principal stress directions were estimated from the induced fractures in the hydraulic fracturing tests using Eq. (4). Fig. 12a shows the directions of the maximum horizontal stresses in a depth profile. Maximum horizontal stress orientations are lacking for test intervals that were not measured or for which the measurements failed. Note that the direction data are widely scattered. However, for a given location, the stress directions are well constrained, as indicated by both Fig. 12a and Table 3, because shallow stress data that systematically deviated from the deeper stress orientations have been ruled out by taking topography perturbations into account. Considering the relatively large spatial distribution of the measured boreholes, a map view and a strike profile are shown in Figs. 13 and 12b, clearly showing that the orientations of σH along the strike of the Longmen Shan fault zone present evident segmentation. The northern and southern ends of the

10

W. Meng et al. / Tectonophysics 656 (2015) 1–19

Table 1 Results of hydraulic fracturing stress measurements in the Longmen Shan thrust belt. Borehole

HZ-1

HZ-2

GY-1

GY-2

PW-1

PW-2

JY-1

BC-1

BC-2

No.

1 2 3 4 5 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 12 13 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 9 1

Test interval depth ( m)

105.30 140.40 155.32 229.43 256.00 83.00 112.80 174.00 215.00 221.00 230.50 235.50 125 147.83 170.5 240 270 292 311.6 323 165.5 174.5 194 218 239 258.5 280.24 296.93 315.02 325.5 348 120.8 171.1 178.2 202 230 253.96 284.65 292.3 331.7 348.36 366 386.2 439 87.5 122 144 149 173 194 208 252 85.5 94.94 104.5 123.5 134.5 152.5 163.47 178.5 58 80 91.85 117 124 133 148 185 195 77

Time/Water level/Rock type/Rock density

2010.08 / 12.80 m / Granite / 2.66 g/cm3 2010.10 / 2.70 m / Granite / 2.66 g/cm3

2011.11 / 3.00 m / Feldspar quartz sandstone, Siltstone / 2.67 g/cm3

2011.12 / 149.40 m / Dolomite, Quartz sandstone, Siltstone / 2.67 g/cm3

2011.10 / 19.00 m / Biotite leptynite, Pebbly biotite leptynite / 2.73 g/cm3

2011.12 / 25.00 m / Biotite leptynite, Pebbly biotite leptynite / 2.73 g/cm3

2009.10 / 0.00 m / Sandstone / 2.65 g/cm3

2009.11 / 5.00 m / Sandstone, Siltstone / 2.67 g/cm3

Principal stresses (MPa) σH

σh

σv

6.79 5.68 7.20 9.02 9.54 6.13 8.02 19.52 13.47 11.38 12.08 12.02 3.81 5.27 5.71 11.28 12.89 12.05 16.74 33.12 11.43 10.07 6.91 14.63 9.17 18.59 17.48 25.24 16.39 14.98 12.81 8.05 10.21 17.29 19.76 26.77 45.17 20.46 30.51 22.06 18.59 46.82 34.87 37.55 3.81 5.93 2.93 4.26 4.28 7.37 15.40 29.71 6.23 3.86 6.11 6.64 5.98 8.82 8.19 10.66 2.46 3.06 4.64 4.93 6.21 5.57 4.40 7.07 6.55 6.00

5.02 4.00 5.24 6.95 6.71 3.81 4.80 10.90 8.35 6.86 7.26 7.30 3.52 4.05 3.96 7.44 8.75 8.52 11.60 18.98 6.69 7.41 5.73 9.08 5.61 10.74 9.71 13.76 11.33 8.67 7.69 5.52 5.95 9.96 12.89 16.06 25.39 10.41 14.77 12.12 11.42 23.77 19.43 19.14 2.93 3.71 2.63 3.71 4.00 5.83 9.14 15.26 4.20 3.20 4.06 4.78 4.88 5.29 6.19 7.39 1.82 2.75 4.18 3.70 4.58 4.61 3.86 6.48 5.35 4.25

2.80 3.73 4.13 6.10 6.81 2.21 3.00 4.63 5.72 5.88 6.13 6.26 3.34 3.95 4.55 6.41 7.21 7.80 8.32 8.62 4.42 4.66 5.18 5.82 6.38 6.90 7.48 7.93 8.41 8.69 9.29 3.30 4.67 4.86 5.51 6.28 6.93 7.77 7.98 9.06 9.51 9.99 10.54 11.98 2.39 3.33 3.93 4.07 4.72 5.30 5.68 6.88 2.27 2.52 2.77 3.27 3.56 4.04 4.33 4.73 1.55 2.14 2.45 3.12 3.31 3.55 3.95 4.94 5.21 2.06

Borehole

BC-2

LG-1

YA-1

YX-1

BX-1

KD-1

KD-2

SM-1

No.

2 3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 1 2 3 4 5 6 7 1 2 3 4 5

Test interval depth (m)

86 95.5 105.5 123 133 144 152.2 166 176.78 193 65.5 87.6 93.3 134 165.5 181 197 221 293.5 335 364.5 468 486 516 553 574 126.6 139.9 150.94 167.26 210.87 220.46 244.2 288.95 306.1 329.45 455.65 495.75 550.72 561.47 90 128 142 171 178 185 189 219 258 283 317 354 55 81 117 136 172 186 108.86 114.5 121.5 131 162.5 171.9 185.17 97.35 117.30 124.25 131.00 201.70

Time/Water level/Rock type/Rock density

2009.12 / 3.50 m / Sandstone, Siltstone / 2.67 g/cm3

2012.11 / 28.5 m / Limestone / 2.68 g/cm3

2012.12 / 5.6 m / Limestone, Dolomitic limestone / 2.68 g/cm3

2008.09 / 8.00 m / Granite / 2.66 g/cm3

2008.09 / 0.00 m / Granite / 2.66 g/cm3

2008.09 / 24.00 m / Granite / 2.66 g/cm3

2010.12 / 21.00 m / Metamorphic quartz sandstone / 2.72 g/cm3

2011.08 / 0.00 m / Quartzite / 2.68 g/cm3

Principal stresses (MPa) σH

σh

σv

4.14 4.31 9.09 8.02 8.45 8.76 6.74 11.05 11.26 16.19 4.81 3.48 6.22 13.73 6.69 9.44 10.05 12.06 18.89 9.93 8.72 14.79 19.70 18.87 16.10 12.02 7.24 4.40 9.93 6.00 9.05 21.28 8.41 6.85 10.39 22.00 16.47 18.41 33.95 34.54 3.48 8.34 7.81 15.62 16.36 13.11 9.87 8.83 14.60 18.22 17.28 25.67 1.74 2.84 2.46 3.67 6.41 6.31 4.89 5.47 18.19 9.52 23.95 30.69 16.61 7.59 6.91 13.13 21.42 19.30

2.78 3.10 5.54 5.48 5.41 5.77 5.11 7.59 6.85 9.19 4.37 2.39 3.75 7.61 5.78 6.84 6.71 7.53 11.18 6.81 6.79 11.23 12.54 12.50 11.27 9.36 5.74 3.88 7.60 5.01 6.07 11.95 6.26 5.98 7.04 13.12 11.69 11.84 19.47 20.94 2.60 7.44 5.95 11.81 9.68 8.35 7.36 6.45 9.81 12.27 11.09 15.75 1.70 2.44 2.42 3.55 6.12 4.94 3.26 3.40 12.83 7.14 15.70 14.54 9.80 5.79 5.31 7.39 11.45 10.76

2.30 2.55 2.82 3.28 3.55 3.84 4.06 4.43 4.72 5.15 1.76 2.35 2.50 3.59 4.44 4.85 5.28 5.92 7.87 8.98 9.77 12.54 13.02 13.83 14.82 15.38 3.39 3.75 4.05 4.48 5.65 5.91 6.54 7.74 8.20 8.83 12.21 13.29 14.76 15.05 2.39 3.40 3.78 4.55 4.73 4.92 5.01 5.80 6.84 7.50 8.40 9.38 1.46 2.15 3.11 3.62 4.58 4.95 2.96 3.11 3.30 3.56 4.42 4.68 5.04 2.61 3.14 3.33 3.51 5.41

Note: The boreholes in which rock mechanics tests have been conducted include HZ-2, PW-2, JY-1, BC-1, BC-2, and KD-2, and their actual densities were recorded. Considering the same lithology, the rock densities of HZ-1 and PW-1 are referred to those of HZ-2 and PW-2, and the actual densities of BC-1, BC-2, and HZ-2 were used to estimate the densities of GY-1, GY-2, YX-1, BX-1 and KD-1. The rock densities of the remaining boreholes are referred to the average value.

W. Meng et al. / Tectonophysics 656 (2015) 1–19

11

Table 2 Regression equations between the horizontal stresses and depth (m). Study area

Worldwide Chinese mainland Middle segment of the North-South seismic belt in China Longmen Shan thrust belt

σH (MPa) = AH + B

σh (MPa) = AH + B

A

B

A

B

0.0210 0.0229 0.0216 0.0209

12.290 4.738 6.781 6.134

0.0150 0.0171 0.0182 0.0153

6.290 1.829 2.233 3.589

0.0205–0.2356

–

0.0148–0.1118

–

fault are dominated by NW–NWW stress, which reflects the background stress state controlled by the eastward extrusion of the Tibetan Plateau. However, the orientations of σH determined in the test boreholes located in the northern part are NE–NEE, suggesting a local stress feature. The stress orientations change from NW–NWW to NE–NEE from south to north, which agrees with the orientation characteristics analyzed by the focal mechanism solutions and fault plane solutions (Cui et al., 2011; Liu et al., 2012). The evolution of the Tibetan plateau involves the subduction of the Indian lithosphere, thickening of the Tibetan crust, and eastward extrusion of the Tibetan lithosphere (Royden et al., 2008). Because the lack of significant upper crustal shortening across much of the eastern plateau's margin implies that crustal thickening occurs mainly in the deep crust (Clark and Royden, 2000), a lower crustal flow hypothesis was proposed that describes how the lower crust escapes from beneath thickened, elevated central plateau through regions where the crust is weak (Clark and Royden, 2000; Kirby et al., 2000; Harris, 2007; Royden et al., 2008), which has received considerable but not unanimous support. Clark and Royden (2000) modeled the topography of the eastern margin of the Tibetan Plateau

Number of statistic samples

Maximum depth (m)

Reference

322 3365 450 362

5100 3984 5500 749

Zhu et al., 1994 Yang et al., 2012 Jing et al., 2007 Yang et al., 2012

140

574

This study

in terms of a Newtonian fluid through a lower crustal channel of uniform thickness. They demonstrated a viscosity of 1018 Pa s for the lower crust beneath the low-gradient margins and 1021 Pa s beneath the steep, abrupt margins. Strong crust such as the Sichuan Basin, which is an old, intact craton, should inhibit the flow of lower crustal material from central Tibet and therefore build a steep topographic margin, which is exemplified along the eastern plateau's margin. For example, as illustrated in Fig. 14, profile A′-A (profile location shown in Fig. 13) rises westward from the Sichuan Basin at ~500 m elevation to peak elevations of ~5000 m over horizontal distances of ~50 km. Meanwhile, the weak lower crust steers clear of and flows around regions of cold, strong, continental material, and low topographic gradients were presented. The elevations differences in profile B′-B and profile C′-C are less than 2500 m over horizontal distances of ~300 km. It is interesting to find that the directional rotation of the maximum horizontal principal stress in the shallow crust is consistent with the lower crustal flow hypothesis. The maximum horizontal principal stresses are dominantly oriented to the NE in the northern segment of the Longmen Shan fault zone and to the NW in the southern segment.

Fig. 12. Variations in the orientations of the maximum horizontal stresses, σH, with depth (a). Along the Longmen Shan fault strike (b), the mean orientations of σH in each test borehole are shown as open triangles, and the horizontal bars represent the angular standard deviation. The gray rectangles represent the orientations of σH at each test interval.

12

W. Meng et al. / Tectonophysics 656 (2015) 1–19

Table 3 Directions of the in-situ stress in the Longmen Shan fault zone. HZ-1 and HZ-2, PW-1 and PW-2, BC-1 and BC-2, and KD-1 and KD-2 are each treated as one for this analysis because of their close proximity. Borehole no.

Predominant direction of σH

Standard deviation

Borehole no.

Predominant direction of σH

Standard deviation

HZ-1, HZ-2 GY-2 GY-1 PW-1, PW-2 BC-1, BC-2 JY-1

N65°W N76°E N77°E N85°E N51°E N68°E

10.5° 11.5° 8.3° 11.9° 20.2° 14.4°

LG-1 YA-1 YX-1 BX-1 KD-1, KD-2 SM-1

N29°W N73°W N56°W N59°W N39°E N44°W

16.8° 4.2° – 31.3° 28.5° 7.8°

Furthermore, the stress directions present a clockwise and anticlockwise pattern in the northern and southern ends of the Longmen Shan fault zone, as if the maximum principal stress directions are driven by weak crustal material flow. It can be inferred that the stress field in this region is dominated by deep movement. We postulate a preliminary understanding of the coupling between the shallow crustal stress field and the lower crustal flow. However, further research is necessary to determine the deep stress field in this region.

shear and normal stresses on this fault are given, respectively, by Eqs. (5) and (6), considering the effect of pore pressure, PP (Zoback and Townend, 2001).

τf ¼

S1 −S3 sin2β 2

ð5Þ

σn ¼

S1 þ S3 −2P P S1 −S3 þ cos2β 2 2

ð6Þ

4. Implications of in-situ stress for fault activity 4.1. Theoretical basis inferred from crustal frictional strength constraints The frictional strength of brittle crust can be computed by considering a pre-existing fault whose normal, NF, makes an angle β with respect to the direction of the maximum principal stress, S1, seen in Fig. 15a. The

where S3 is the minimum principal stress. Hence, the shear and normal stresses on the fault depend on the magnitudes of the principal stresses, the pore pressure, and the orientation of the fault with respect to the maximum principal stresses. Moreover, assuming that the angle and coefficient of internal friction are Φ and μ, respectively, and the cohesive strength of the fault is zero,

Fig. 13. Regional topography of the Longmen Shan region. The short red lines represent the orientations of the maximum horizontal stresses measured by hydraulic fracturing in this study. The inset shows the approximate weak crustal regions (white areas) with topographic contours. The black arrows are the inferred directions of lower crustal flow (modified after Clark and Royden, 2000). SB: Sichuan Basin; QB: Qaidam Basin.

W. Meng et al. / Tectonophysics 656 (2015) 1–19

13

Fig. 14. Topographic profiles for the eastern margins of the Tibetan Plateau. The locations of profiles are shown in Fig. 13.

the shear stress can be defined as Eq. (7) if the plane is critically stressed in terms of the Coulomb frictional-failure criterion, as shown in Fig. 15b. τ ¼ μσ n ¼ σ n tanϕ

ð7Þ

μm ¼

and ϕ ¼ 2β−π=2:

ð8Þ

Combining Eqs. (5) and (7), we obtain S1 −S3 ¼

stress, S1 − PP, to the minimum effective principal stress, S3 − PP, that corresponds to the case in which a critically oriented fault is at the frictional limit, are given by (Jaeger and Cook, 1979; Townend and Zoback, 2000)

μ ðS1 þ S3 −2P P Þ sin2β−μ cos2β

ð9Þ

such that, for Eqs. (7) and (8), qffiffiffiffiffiffiffiffiffiffiffiffiffiffi sin2β ¼ cosϕ ¼ 1= μ 2 þ 1 qffiffiffiffiffiffiffiffiffiffiffiffiffiffi cos2β ¼ − sinϕ ¼ −μ= μ 2 þ 1:

ð10Þ ð11Þ

Hence, the values of μm, which is the ratio of the maximum shear stress, (S1 − S3)/2, to the effective mean principal stress, (S1 + S3)/ 2 − PP, and K, which is the ratio of the maximum effective principal

K¼

qffiffiffiffiffiffiffiffiffiffiffiffiffiffi S1 −S3 ¼ μ= μ 2 þ 1 S1 þ S3 −2P P

S1 −P P ¼ S3 −P P

 qffiffiffiffiffiffiffiffiffiffiffiffiffiffi2 μ þ μ2 þ 1 :

ð12Þ

ð13Þ

It has been clearly shown that the intraplate continental crust is in a state of failure equilibrium by in-situ stress measurements in six deep boreholes with a laboratory-derived frictional coefficient, μ, of 0.6–1.0 (Byerlee, 1978; Zoback, 2007; Zoback and Townend, 2001). The fault will experience frictional sliding along the optimally orientated plane, which is given by   β ¼ π=2 þ tan−1 μ =2:

ð14Þ

From this perspective, Eqs. (12) and (13) can be used to evaluate intense stress and fault activity. The upper bounds of μm and K are 0.5–0.7 and 3.12–5.83, respectively, when the frictional coefficient is assumed to be in the range of 0.6–1.0. As illustrated in Table 4, the Anderson classification scheme defines three stress states—normal, strike-slip, and reverse—in terms of the corresponding fault nature (Anderson, 1951), which can be used to determine which principal stress, i.e., σH, σh, or σv, corresponds to S1, S2, and S3. For example, the vertical stress, σv, is the maximum principal stress (S1) in the normal faulting scheme, the minimum principal stress (S3) in reverse faulting regimes, and the intermediate principal stress (S2) in strike-slip regimes. 4.2. Fault activity

Fig. 15. Schematic diagram of the stress analysis on a pre-existing plane (a) and Mohr circle (b).

Fault activity is an important concept for crustal stability evaluation. Wang et al. (2014) suggested adopting μm to reflect the capability and

14

W. Meng et al. / Tectonophysics 656 (2015) 1–19

Table 4 Faulting regimes and relative stress magnitudes. Schematic illustrations of fault regimes

Relative stress magnitudes

Steady state Derived from Eq. (12) pffiffiffiffiffiffiffiffiffiffiffiffiffiffi σ v −σ h 2 σ v þσ h −2P P ≤μ= μ þ 1

Normal

σv N σH N σh

Reverse

σH N σh N σv

σ H −σ v σ H þσ v −2P P

pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ≤μ= μ 2 þ 1

Strike-slip

σH N σv N σh

σ H −σ h σ H þσ h −2P P

pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ≤μ= μ 2 þ 1

stress level of the upper crust and analyzed stress build-up in the shallow crust prior to the Lushan earthquake. Chen et al. (2012) considered that K values can reveal the stress intensity acting on a fault. Li et al. (2012) analyzed the fault activity in the city of Urumqi, China, by discussing values of μm. In this paper, we discuss the implications of

Derived from Eq. (13) σ v −P P σ h −P P

 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi2 ≤ μ þ μ2 þ 1

σ H −P P σ v −P P

 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi2 ≤ μ þ μ2 þ 1

σ H −P P σ h −P P

 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi2 ≤ μ þ μ2 þ 1

in-situ stress for fault activity in the Longmen Shan thrust belt based on in-situ stress measurements. Fig. 16 illustrates the dependence of the maximum effective principal stress on the minimum effective principal stress (Fig. 16a) and the maximum shear stress on the effective mean principal stress (Fig. 16b). The profiles in Fig. 16a and b are

Fig. 16. Dependence of (a) the maximum effective principal stress on the minimum effective principal stress and (b) the maximum shear stress on the effective mean principal stress in the Longmen Shan thrust belt based on in-situ stress measurements.

W. Meng et al. / Tectonophysics 656 (2015) 1–19

15

Fig. 17. Depth profiles of the critical stress magnitudes of different fault regimes with μ assumed to be 0.6.

nearly in accordance because they are derived from the same theories. It is evident that the stress levels are almost beneath the limit defined by μ = 1.0 at relatively shallow depths in the Longmen Shan thrust belt, and the lower limit can be defined by μ = 0.2. With increasing depth, the stress magnitudes are consistent with μ = 0.6 and 1.0, and even partial data exceed the upper limit; that is, any well-oriented faults, if they exist, are generally at frictional equilibrium with the friction coefficient given by Byerlee's experimental value (Zoback and Townend, 2001). We consider that point values exceeding the limit indicate high stress. The vertical stresses in the crust are relatively stable and approximately equal to the overburden gravity (Hoek and Brown, 1980) and are assumed in this paper to correspond to a mean rock density of 2.68 g/cm3. Therefore, when the pore pressure, Pp, is approximately equal to the hydrostatic pressure and the frictional coefficient, μ, is given a certain value, the depth profiles of the critical stress magnitudes can be displayed as shown in Fig. 17. Under the assumption that Pp is equivalent to the hydrostatic pressure, the stress increases at a rate of 10 MPa/km, in which the most important factor is the water level (Hw) of the borehole. When a high water level is close to the borehole's orifice, HW can be simply assumed to be zero, and Pp can be defined as Pp = 0.01 H (in units of MPa). However, this parameter cannot be ignored when the water level is low, in which case Pp = 0.01 (H − HW). Eqs. (12) and (13) can then be expressed as linear relationships of the critical principal stresses and depth for the fault regimes of normal and reverse faults, as shown in Table 5. Of the 16 test boreholes along the Longmen Shan thrust belt, the water levels of three are close to the orifice; those of the others are 12.8 m, 2.7 m, 3.0 m, 149.4 m, 19.0 m, 25.0 m, 5.0 m, 3.5 m, 28.5 m, 5.6 m, 8.0 m, 24.0 m, and 21.0 m. Fig. 18 illustrates the estimated fault activity in the study area when considering two extremes for the borehole water levels: HW = 0 and 150 m. A considerable part of the maximum principal stress falls in the critical zone for reverse faulting, and almost no minimum principal stress falls in the critical zone for normal faulting. These results indicate that the stress state is conducive to reverse faulting and normal faulting is weak. Frictional sliding is more likely to occur when decreases in the critical value correspond to increases in the water level, which is the reason for fault instability during water injection in oil and gas extraction. As previously discussed, the stress orientations indicate the segmentation of the Longmen Shan thrust belt. Moreover, the Wenchuan earthquake formed a seismic rupture zone toward the northeast from its epicenter, Yingxiu, with a length of ~ 240 km along the YingxiuBeichuan fault. In contrast, the southwestern epicenter had weaker energy radiation (Lin et al., 2009; Xu et al., 2009). Considering the tectonic stress features in addition to previous research, the Longmen Shan thrust belt can be generally divided into three segments along its strike:

the northern, middle and southern segments (Meng et al., 2013). The results from the fault stability evaluation of the three segments based on measurement data are given in Fig. 19, based on the above analysis that the stress regime in this region favors reverse faulting. It has been confirmed that the frictional coefficient of most rocks ranges from 0.6 to 1.0 (Byerlee, 1978; Brace and Kohlstedt, 1980; Zoback, 2007; Townend and Zoback, 2000). However, it has been reported that the coefficient of friction is lower in an actual fault than expected (Boatwright and Cocco, 1996; Carpenter et al, 2009, 2011). Verberne et al. (2010) and Zhang and He (2013) determined the steady-state friction coefficients of a natural fault gouge collected from the Yingxiu-Beichuan fault to be 0.4 and 0.2–0.7, respectively, by conducting frictional sliding experiments. Wang et al. (2014) considered that the ratio of the halfmaximum differential stress to the mean effective stress, μm, has a similar physical meaning with the frictional coefficient, μ, and we can replace μ with μm to perform rock friction analysis. Based on the value of μm calculated by the measured stress data in the Longmen Shan fault zone, the critical value of μ = 0.4 was added into the fault activity analysis in different segments. Note that the stress values along the Longmen Shan thrust belt are relatively higher in the southern and northern segments and lower in the middle segment, which may be the result of stress release due to the Wenchuan earthquake. Chang et al. (2010) conducted a complementary analysis of the borehole in-situ stress and earthquake focal mechanism in southeastern Korea. They suggested that the magnitude of the stress field appears to be inversely correlated with the density of regional-scale faults because of the stress released as a result of faulting. Meng et al. (2015) analyzed the interaction between the stress state and active faults by considering the Longmen Shan fault zone as an example. Their results indicate that stress build-up changes relative to changes in fault properties such as frictional strength, and areas with higher stress accumulation are more evident near faults with higher friction coefficients. The tectonic stress field of the Longmen Shan thrust belt is still adjusting after the Wenchuan earthquake. The stress released because of the earthquake and its aftershocks migrated to and accumulated in the northern and southern ends of the Longmen Shan thrust belt, resulting in high-stress Table 5 Critical value expressions of the horizontal principal stress of fault activity. Fault regime Reverse Normal

Expressions pffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi  WÞ μ 2 þ 1 þ μ  μ ðHH μ2 þ 1 þ μ 50 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi  ρH WÞ σ h ¼ 100 μ 2 þ 1  μ þ μ ðHH μ2 þ 1  μ 50 ρH σ H ¼ 100

H is the depth, in units of m; HW is the static water level, in units of m; ρ is the rock density in units of g/cm3; μ is the frictional coefficient.

16

W. Meng et al. / Tectonophysics 656 (2015) 1–19

Fig. 18. Schematic illustration of the fault activity in the study area of the Longmen Shan fault zone. The red lines represent the critical values when HW = 0 m; the blue lines represent the critical values when HW = 150 m.

environments in the northern and southern segments. As the contemporary crustal stress is in critical frictional equilibrium or exceeds the equilibrium state and is conducive to reverse faulting, strong earthquakes are likely to occur because of the enhanced fictional strength in special tectonics. On April 20, 2013, Beijing time, the Ms 7.0 Lushan earthquake occurred along the eastern marginal thrust of the southwestern Longmen Shan fault zone. Because the earthquake did not cause significant surface rupture, the seismogenic structure of the Lushan earthquake remains controversial. Hong et al. (2013) proposed that the seismogenic fault is either the Shuangshi-Dachuan fault or Dayi fault, or both (see Fig. 20a). Xu et al. (2013) concluded a blind reverse fault that

strikes 212° and dips toward the NW with a dip angle of 38° ± 2° was the seismogenic fault, as demonstrated by the focal mechanism solutions and surface structural geology (Fig. 20a, red dashed line). Fang et al. (2013) also suggested that a blind thrust fault on the eastern side of the Shuangshi-Dachuan fault might be the seismogenic fault by relocating the Lushan earthquake and aftershocks. In addition, the fault is listric, with approximately 63° in the shallow crust, 41° near the mainshock, and 17° at the bottom. The focal mechanism inversion results show that the Lushan earthquake is a thrust earthquake with a strike of 220°and dip of 35° (Fang et al., 2013), in accordance with analysis results based on in-situ stress data, which reveal a reverse faulting environment in the Longmen Shan fault zone. The normals of the optimally oriented failure planes defined by Eq. (14) were at angles of 56° and 68° to the maximum horizontal principal stress when the lower and upper limits of the frictional coefficient, μ, are 0.4 and 1.0, respectively, in light of the above analysis. The maximum principal stress direction of borehole BX-1, which is near the Lushan earthquake's epicenter, was taken to calculate the dip angle of the thrust fault. For different boundary values of μ, 34° and 22° were obtained. Frictional coefficient (μ) values of 0.4 or lower approach an actual fault in the southern part of the Longmen Shan thrust belt, combined with previous research. The inferred thrust fault striking 31° (211°) only considered the stress environment (see Fig. 20b, c), which is consistent with that studied by Xu et al. (2013). The tectonic stresses resulting from the convergence of crustal material moving from the high Tibetan Plateau to the west against strong crust underlying the Sichuan Basin and reconstruction resulting from the Wenchuan earthquake may have directly induced the Lushan earthquake. Under the ongoing power source of continental collision from the India and Eurasia plates, which are converging at a relative rate of 40–50 mm/year (USGS, 2013), the stresses in the Tibetan Plateau accumulate and contribute to fault activity. At the intersection area of the Min Jiang, Xue Shan, Huya, and Longmen Shan faults and Min Shan, the complicated tectonic framework of the northern segment of the Longmen Shan fault zone is conducive to stress build-up. The orientations of the maximum principal stresses, which favor frictional failure, were calculated based on the boundary values of μ, namely, 0.4 and 1.0. The Longmen Shan fault was taken as the failure fault. The calculated directions, N 67°–79° E, were close to measured directions in the

Fig. 19. Depth profiles of the principal stresses compared to the critical stress magnitudes in the Longmen Shan thrust belt when μ is set to 0.4 and 1.0 in the northern segment (N), middle segment (M), and southern segment (S). The red shapes represent the maximum horizontal principal stresses, and the blue shapes represent the minimum horizontal principal stresses.

W. Meng et al. / Tectonophysics 656 (2015) 1–19

17

Fig. 20. (a) Distributions of the mainshock and aftershock sequences for the Lushan earthquake, modified after Fang et al. (2013). The red star represents the mainshock and the circles represent aftershocks, with color changes from blue to red (unit of color scale: day). (b) Seismogenic fault model inferred from the direction of the maximum horizontal principal stress. (c) Planar graph showing the direction of the maximum horizontal principal stress and strike of the surface intersection of a blind thrust fault.

northern part, N 51°–85° E (see Fig. 21). It can be concluded that the northern segment of the Longmen Shan fault is likely to be active when the stress builds-up enough to destroy the frictional equilibrium. Although a reverse stress regime dominates at shallow depths according to the in-situ stress measurements, the stress directions and fault strike are conducive to strike-slip faulting. Therefore, the actual faulting is likely to be characterized by reverse and strike-slip motion.

5. Discussion and conclusions The Longmen Shan thrust belt is undergoing active convergence, leading to the development of seismicity. As shown in Fig. 22, we plotted contemporary earthquakes with a magnitude higher than 2 that occurred after the in-situ stress measurements. Based on the hydraulic fracturing in-situ stress measurements conducted along the Longmen Shan thrust belt between 2008 and 2012, the characteristics of the stress state and implications for fault activity are discussed in this study. However, the absolute magnitudes of the in-situ stress may not accurately represent the tectonic stress states in the region because of the locality of the measurement boreholes. The test boreholes discussed in this paper are distributed throughout the Longmen Shan thrust belt, and measurement intervals in each borehole were conducted as much as possible to provide reliable information. The in-situ stresses may at least indicate the overall characteristics and relative spatial variations in the stress states. Based on the shallow crustal stress measurement data, the stress regimes are generally favorable to reverse faulting and partly to strike-slip faulting characterized by σH N σh N σv and σH N σv N σh, indicating that the regional stress field is definitely dominated by the maximum horizontal stress. The maximum horizontal principal stresses are dominantly oriented to the NE in the northern segment of the Longmen Shan fault zone and to the NW in the southern segment. Furthermore, the stress directions present a clockwise and anti-clockwise pattern in the northern and southern ends of the Longmen Shan fault zone, as if the maximum principal stress directions are driven by weak crustal material flow. It can be inferred that the stress field in this region is dominated by deep movement. We postulate a preliminary understanding of the coupling between the shallow crustal stress field and the lower crustal flow. However, the argument for this hypothesis is inadequate because the stress state in the deeper crust is not clear due to the limited

measurement depth. Further research is necessary to distinguish the deep stress field in this region. The parameters μm, which is the ratio of the maximum shear stress, (S1 − S3)/2, to the effective mean principal stress, (S1 + S3)/2 − PP, and K, which is the ratio of the maximum effective principal stress, S1 − PP, to the minimum effective principal stress, S3 − PP, can be used to calculate the stress state comparable to that predicted using Coulomb frictional failure criteria. Furthermore, the horizontal principal stresses can be predicted as functions of the rock density, ρ, frictional coefficient, μ, depth, H, and water level, HW, in frictional equilibrium. The influence of HW on the critical stresses is discussed, and the decrease in stress values corresponds to an increase in the water level. The depth profiles of the stress magnitudes in different segments are illustrated, showing that the stress values are relatively higher in the southern and northern segments and lower in the middle segment. The tectonic stress field of the Longmen Shan fault zone may still be adjusting as a result of the Wenchuan earthquake. The stress released because of the Wenchuan earthquake and its aftershocks migrated to and accumulated in the northern and southern ends of the Longmen Shan fault zone, resulting in high-stress environments in the northern and southern segments. Because the contemporary crustal stress is in critical frictional equilibrium or exceeds the equilibrium state and is conducive to reverse faulting, strong earthquakes are likely to occur. The Lushan earthquake may be a

Fig. 21. Calculated stress orientations that are favorable for frictional-failure, taking the Longmen Shan fault as the failure fault.

18

W. Meng et al. / Tectonophysics 656 (2015) 1–19

Fig. 22. Distribution of earthquake epicenters (2012.12–2015.2) after the in-situ stress measurements.

verification of this inference. Discussion based on the measured results reveals that the stress state in the southern segment, specifically, near the epicenter of Lushan earthquake, favors the occurrence of earthquakes. As the intersection area of the Min Jiang, Xue Shan, Huya, and Longmen Shan faults and Min Shan, the complicated tectonic framework of the northern segment of the Longmen Shan fault zone is conducive to stress build-up. Favorable orientations for the maximum principal stresses, which favor frictional failure, were calculated based on the boundary values of μ, namely, 0.4 and 1.0. The Longmen Shan fault was taken as the failure fault, and the calculated directions, N 67°–79° E, were close to the measured directions in the northern part, N 51°–85° E. It can be concluded that the northern segment of the Longmen Shan fault is likely to be active when the stress builds up enough to destroy the frictional equilibrium. Acknowledgments The excellent collaboration of all participants is highly appreciated. The authors also thank Qimei An for the beneficial discussions. This research was supported by the research funds of the Institute of Geomechanics, Chinese Academy of Geological Science (No. DZLXJK201404) and the Chinese Government's Executive Program SinoProbe (grant 06-03). Critical and constructive comments made by Weiren Lin and an anonymous reviewer greatly improved this manuscript. References Amadei, B., Stephansson, O., 1997. Rock Stress and Its Measurement. Chapman & Hall, London. Anderson, E.M., 1951. The Dynamics of Faulting and Dyke Formation with Application to Britain. Oliver & Boyd, Edinburgh, U.K. Barton, C.A., Zoback, M.D., Moos, D., 1995. Fluid flow along potentially active faults in crystalline rock. Geology 23, 683–686.

Boatwright, J., Cocco, M., 1996. Frictional constraints on crustal faulting. J. Geophys. Res. 101 (B6), 13895–13909. Brace, W.F., Kohlstedt, D.L., 1980. Limits on lithospheric stress imposed by laboratory experiments. J. Geophys. Res. 85 (B11), 6248–6252. Brown, E.T., Hoek, E., 1978. Trends in relationships between measured in-situ stresses and depth. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 15, 211–215. Byerlee, J.D., 1978. Friction of rocks. Geophys 116 (4/5), 615–626. Carpenter, B.M., Marone, C., Saffer, D.M., 2009. Frictional behavior of materials in the 3D SAFOD volume. Geophys. Res. Lett. 36, L05320. http://dx.doi.org/10.1029/2008g l036660. Carpenter, B.M., Marone, C., Saffer, D.M., 2011. Weakness of the San Andreas fault revealed by samples from the active fault zone. Nat. Geosci. 4, 251–254. http://dx.doi. org/10.1038/ngeo1089. Chang, C., Lee, J.B., Kang, T.S., 2010. Interaction between regional stress state and faults, complementary analysis of borehole in situ stress and earthquake focal mechanism in southeastern Korea. Tectonophysics 485, 164–177. http://dx.doi.org/10.1016/j. tecto.2009.12.012. Chen, Q.C., Feng, C.J., Meng, W., Qin, X.H., An, Q.M., 2012. Analysis of in situ stress measurements at the northeastern section of the Longmenshan fault zone after the 5.12 Wenchuan earthquake. Chinese J. Geophys. 55 (12), 3923–3932. http://dx.doi.org/10.6038/j.issn.0001-5733.2012.12.005 (in Chinese with English abstract). Cheung, L.S., Haimson, B.C., 1989. Laboratory study of hydraulic fracturing pressure data—how valid is their conventional interpretation? Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 26 (6), 595–604. Clark, M.K., Royden, L.H., 2000. Topographic ooze: building the eastern margin of Tibet by lower crustal flow. Geology 28 (8), 703–706. Cui, X.F., Hu, X.P., Yu, C.Q., Tao, K., Wang, Y.H., Ning, J.Y., 2011. Research on focal mechanism solutions of Wenchuan earthquake sequence. Acta Sci. Nat. Univ. Pekin. 47 (6), 1063–1072 (in Chinese with English abstract). Dirks, P.H.G.M., Wilson, C.J.L., Chen, S., Luo, Z.L., Liu, S., 1994. Tectonic evolution of the NE margin of the Tibetan Plateau, evidence from the central Longmen Mountains Sichuan Province, China. J. SE Asian Earth Sci. 9 (1/2), 181–192. Fang, L.H., Wu, J.P., Wang, W.L., Lu, Z.Y., Wang, C.Z., Yang, T., Cai, Y., 2013. Relocation of the mainshock and aftershock sequences of Ms 7.0 Sichuan Lushan earthquake. Chin. Sci. Bull. 58 (28–29), 3451–3459. http://dx.doi.org/10.1007/s11434-013-6000-2 (in Chinese with English abstract). Gao, J.L., Ding, J.M., Liang, G.P., Guo, Q.L., 1987. Variation of crustal stress with depth in the basin of North China. Earthquake Research In China 3 (4), 82–89 (in Chinese with English abstract). Gronseth, J.M., Kry, P.R., 1983. Instantaneous shut-in pressure and its relationship to the minimum in-situ stress. In: Zoback, M.D., Haimson, B.C. (Eds.), Hydraulic Fracturing Stress Measurements. National Academy Press, Washington, D. C.

W. Meng et al. / Tectonophysics 656 (2015) 1–19 Guo, Q.L., Wang, C.H., Ma, H.S., Wang, C.G., 2009. In-situ hydro-fracture stress measurement before and after the Wenchuan MS 8.0 earthquake of China. Chin. J. Geophys. 52 (5), 1395–1401. http://dx.doi.org/10.3969/j.issn.0001-5733.2009.05.029 (in Chinese with English abstract). Haimson, B.C., 1978. The hydrofracturing stress measuring method and recent field results. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 15, 167–178. Haimson, B.C., Cornet, F.H., 2003. ISRM suggested methods for rock stress estimation—part 3, hydraulic fracturing (HF) and/or hydraulic testing of pre-existing fractures (HTPF). Int. J. Rock Mech. Min. Sci. 40, 1011–1020. http://dx.doi.org/10.1016/j.ijrmms.2003. 08.002. Haimson, B.C., Doe, T.W., 1983. State of stress, permeability, and fractures in the Precambrian Granite of Northern Illinois. J. Geophys. Res. 88 (B9), 7355–7371. Harris, N., 2007. Channel flow and the Himalayan–Tibetan orogen: a critical review. J. Geol. Soc. 164, 511–523. http://dx.doi.org/10.1144/0016-76492006-133. Hayashi, K., Haimson, B.C., 1991. Characteristics of shut-in curves in hydraulic fracturing stress measurements and determination of in situ minimum compressive stress. J. Geophys. Res. 96 (B11), 18311–18321. Hayashi, K., Sakurai, I., 1989. Interpretation of hydraulic fracturing shut-in curves for tectonic stress measurement. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 26 (6), 477–482. Hoek, E., Brown, E.T., 1980. Underground Excavation in Rock. Institution of Mining & Metallurgy, London, pp. 382–395. Hong, H.C., Xu, H.G., Song, F., Zhang, X.H., Li, G.H., 2013. Discussion on seismo-geological hazards induced by 2013 Lushan Ms 7.0 earthquake and its seismogenic fault. Acta Seismol. Sin. 35 (5), 738–748 (in Chinese with English abstract). Ito, T., Evans, K., Kawai, K., Hayashi, K., 1999. Hydraulic fracture reopening pressure and the estimation of maximum horizontal stress. Int. J. Rock Mech. Min. Sci. 36, 811–826. Jaeger, J.C., Cook, N.G.W., 1979. Fundamentals of Rock Mechanics. Chapman & Hall, New York. Jing, F., Sheng, Q., Zhang, Y.H., Luo, C.W., Liu, Y.K., 2007. Research on distribution rule of shallow crustal geostress in China mainland. Chin. J. Rock Mech. Eng. 26 (10), 2056–2062 (in Chinese with English abstract). Kirby, E., Whipple, K.X., Burchfiel, B.C., Tang, W.Q., Berger, G., Sun, Z.M., Chen, Z.L., 2000. Neotectonics of the Min Shan, China: implications for mechanisms driving Quaternary deformation along the eastern margin of the Tibetan Plateau. Geol. Soc. Am. Bull. 112 (3), 375–393. http://dx.doi.org/10.1130/0016-7606(2000)112b375,NOTMSCN2. 0.CO;2. Klee, G., Rummel, F., Williams, A., 1999. Hydraulic fracturing stress measurements in Hong Kong. Int. J. Rock Mech. Min. Sci. 36, 731–741. Lee, M.Y., Haimson, B.C., 1989. Statistical evaluation of hydraulic fracturing stress measurement parameters. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 26 (6), 447–456. Li, S.G., 1973. Seismology and Geology. Science Press, Beijing (in Chinese). Li, H., Xie, F.R., Wang, H.Z., Dong, Y.K., Yu, J.J., 2012. Characteristics of in-situ stress measurements near the fault and fault activity in Urumqi City. Chin. J. Geophys. 55 (11), 3690–3698. http://dx.doi.org/10.6038/j.issn.0001-5733.2012.11.016 (in Chinese with English abstract). Liao, C.T., Zhang, C.S., Wu, M.L., Ma, Y.S., Ou, M.Y., 2003. Stress change near the Kunlun fault before and after the Ms 8.1 Kunlun earthquake. Geophys. Res. Lett. 30, 2027–2030. http://dx.doi.org/10.1029/2003GL018106. Lim, S.S., Martin, C.D., 2010. Core disking and its relationship with stress magnitude for Lac du Bonnet granite. Int. J. Rock Mech. Min. Sci. 47, 254–264. http://dx.doi.org/10. 1016/j.ijrmms.2009.11.007. Lin, A., Ren, Z.K., Jia, D., Wu, X.J., 2009. Co-seismic thrusting rupture and slip distribution produced by the 2008 Mw 7.9 Wenchuan earthquake, China. Tectonophysics 471 (3–4), 203–215. http://dx.doi.org/10.1016/j.tecto.2009.02.014. Lin, W., Saito, S., Sanada, Y., Yamamoto, Y., Hashimoto, Y., Kanamatsu, T., 2011. Principal horizontal stress orientations prior to the 2011 MW 9.0 Tohoku-Oki, Japan, earthquake in its source area. Geophys. Res. Lett. 38, L00G10. http://dx.doi.org/10.1029/ 2011GL049097. Lin, W., Conin, M., Moore, J.C., Chester, F.M., Nakamura, Y., Mori, J.J., Anderson, L., Brodsky, E.E., Eguchi, N., 2013. Expedition 343 Scientists, 2013. Stress state in the largest displacement area of the 2011 Tohoku-Oki earthquake. Science 339 (687). http://dx. doi.org/10.1126/science.1229379. Liu, J., Xiong, T.Y., Zhao, Y., Zhang, Y.S., Chen, Q.C., 2012. Kinematic characteristics of Longmenshan active fault zone and its tectonic implication. J. Jilin Univ. 42 (sup.2), 320–330 (in Chinese with English abstract). McGarr, A., Gay, N.C., 1978. State of stress in the earth's crust. Annu. Rev. Earth Planet. Sci. 6, 405–436.

19

Meng, W., Chen, Q.C., Wu, M.L., Li, G.Q., Qin, X.H., Feng, C.J., 2013. Research on segmentation and characteristic of tectonic stress field of Longmenshan Fault zone. Prog. Geophys. 28 (3), 1150–1160. http://dx.doi.org/10.6038/pg20130306 (in Chinese with English abstract). Meng, W., Chen, Q.C., Wu, M.L., Feng, C.J., Qin, X.H., 2015. Tectonic stress state changes before and after the Wenchuan Ms 8.0 earthquake in the eastern margin of the Tibetan Plateau. Acta Geol. Sin. (English Edition) 89 (1), 77–89. Molnar, P., Tapponnier, P., 1975. Cenozoic tectonics of Asia: effects of a continental collision. Science 189 (4201), 419–426. Moos, D., Zoback, M.D., 1990. Utilization of observations of well bore failure to constrain the orientation and magnitude of crustal stresses: application to continental, deep sea drilling project, and ocean drilling program boreholes. J. Geophys. Res. 95 (B6), 9305–9325. Parsons, T., Chen, J., Kirby, E., 2008. Stress changes from the 2008 Wenchuan earthquake and increased hazard in the Sichuan Basin. Nature 454, 509–510. http://dx.doi.org/ 10.1038/nature07177. Royden, L.H., Burchfiel, B.C., Van der Hilst, R.D., 2008. The geological evolution of the Tibetan Plateau. Science 321, 1054–1058. http://dx.doi.org/10.1126/science.1155371. Rutqvist, J., Tsang, C.F., Stephansson, O., 2000. Uncertainty in the maximum principal stress estimated from hydraulic fracturing measurements due to the presence of the induced fracture. Int. J. Rock Mech. Min. Sci. 37, 107–120. Tapponnier, P., Molnar, P., 1977. Active faulting and tectonics in China. J. Geophys. Res. 82 (20), 2905–2930. http://dx.doi.org/10.1029/JB082i020p02905. Townend, J., Zoback, M.D., 2000. How faulting keeps the crust strong. Geology 28, 399–402. http://dx.doi.org/10.1130/0091-7613(2000)28b399:HFKTCSN2.0.CO;2. USGS, 2013. M6.6–56 km WSW of Linqiong, China (BETA). http://comcat.cr.usgs.gov/ earthquakes/eventpage/usb000gcdd#summary. Verberne, B.A., He, C.R., Spiers, C.J., 2010. Frictional properties of sedimentary rocks and natural fault gouge from the Longmen Shan fault zone, Sichuan, China. Bull. Seismol. Soc. Am. 100, 2767–2790. http://dx.doi.org/10.1785/012009287. Wang, C.H., Song, C.K., Guo, Q.L., Zhang, Y.S., Ding, J.M., 2014. Stress build-up in the shallow crust before the Lushan earthquake based on the in-situ stress measurements. Chinese J. Geophys. 57 (1), 102–114. http://dx.doi.org/10.6038/cjg20140110 (in Chinese with English abstract). Wu, M.L., Zhang, Y.Q., Liao, C.T., Chen, Q.C., Ma, Y.S., Wu, J.S., Yan, J.F., Ou, M.Y., 2009. Preliminary results of in-situ stress measurements along the Longmenshan fault zone after the Wenchuan Ms 8.0 earthquake. Acta Geol. Sin. (English Edition) 83 (4), 746–753. Wu, Z.H., Chen, Q.C., Barosh, P.J., Peng, H., Hu, D.G., 2013. Stress rise precursor to earthquake in the Tibetan Plateau. Nat. Sci. 5 (8A1), 46–55. http://dx.doi.org/10.4236/ns. 2013.58A1006. Xu, X.W., Wen, X.Z., Chen, G.H., Yu, G.H., 2008. Discovery of the Longriba Fault Zone in Eastern Bayan Har Block, China and its tectonic implication. Sci. China Ser. D Earth Sci. 51 (9), 1209–1223. http://dx.doi.org/10.1007/s11430-008-0097-1. Xu, X.W., Wen, X.Z., Yu, G.H., Chen, G.H., Klinger, Y., Hubbard, J., Shaw, J., 2009. Coseismic reverse-and oblique-slip surface faulting generated by the 2008 MW 7.9 Wenchuan earthquake, China. Geology 37, 515–518. http://dx.doi.org/10.1130/G25462A.1. Xu, X.W., Chen, G.H., Yu, G.H., Cheng, J., Tan, X.B., Zhu, A., Wen, X.Z., 2013. Seismogenic structure of Lushan earthquake and its relationship with Wenchuan earthquake. Earth Sci. Front. (China University of Geosciences (Beijing); Peking University) 20 (3), 011–020 (in Chinese with English abstract). Yang, S.X., Yao, R., Cui, X.F., Chen, Q.C., Huang, L.Y., 2012. Analysis of the characteristics of measured stress in Chinese mainland and its active blocks and North–South seismic belt. Chin. J. Geophys. 55 (12), 4207–4217. http://dx.doi.org/10.6038/j.issn.00015733.2012. 12.032. Zhang, L., He, C.R., 2013. Frictional properties of natural gouges from Longmenshan fault zone ruptured during the Wenchuan MW 7.9 earthquake. Tectonophysics 594, 149–164. Zhu, H.C., Tao, Z.Y., 1994. In-situ stress distribution in different rocks. Acta Seismol. Sin. 16 (1), 49–63 (in Chinese). Zoback, M.D., 2007. Reservoir Geomechanics. Cambridge University Press. Zoback, M.D., Haimson, B.C., 1982. Status of the hydraulic fracturing method for in-situ stress measurements. Proc. U. S. Symp. Rock Mech. 23rd 143–156. Zoback, M.D., Townend, J., 2001. Implications of hydrostatic pore pressures and high crustal strength for the deformation of intraplate lithosphere. Tectonophysics 336, 19–30.