Earthquake Research in China  2019, Vol. 33 Issue (3): 461-477     DOI: 10.19743/j.cnki.0891-4176.201903006
Deformation Evolution Characteristics Revealed by GPS and Cross-fault Leveling Data before the MS8.0 Wenchuan Earthquake
ZHAO Jing1,2, LIU Jie2, REN Jinwei3, YUE Chong1,2, LI Jiaojiao4     
1. State Key Laboratory of Earthquake Dynamics, Institute of Geology, CEA, Beijing 100029, China;
2. China Earthquake Networks Center, Beijing 100045, China;
3. Institute of Earthquake Forecasting, CEA, Beijing 100036, China;
4. China Aero Geophysical Survey and Remote Sensing Center for Natural Resources, Beijing 100083, China
Abstract: Based on GPS velocity during 1999-2007, GPS baseline time series on large scale during 1999-2008 and cross-fault leveling data during 1985-2008, the paper makes some analysis and discussion to study and summarize the movement, tectonic deformation and strain accumulation evolution characteristics of the Longmenshan fault and the surrounding area before the MS8.0 Wenchuan earthquake, as well as the possible physical mechanism late in the seismic cycle of the Wenchuan earthquake. Multiple results indicate that:GPS velocity profiles show that obvious continuous deformation across the eastern Qinghai-Tibetan Plateau before the earthquake was distributed across a zone at least 500km wide, while there was little deformation in Sichuan Basin and Longmenshan fault zone, which means that the eastern Qinghai-Tibetan Plateau provides energy accumulation for locked Longmenshan fault zone continuously. GPS strain rates show that the east-west compression deformation was larger in the northwest of the mid-northern segment of the Longmenshan fault zone, and deformation amplitude decreased gradually from far field to near fault zone, and there was little deformation in fault zone. The east-west compression deformation was significant surrounding the southwestern segment of the Longmenshan fault zone, and strain accumulation rate was larger than that of mid-northern segment. Fault locking indicates nearly whole Longmenshan fault was locked before the earthquake except the source of the earthquake which was weakly locked, and a 20km width patch in southwestern segment between 12km to 22.5km depth was in creeping state. GPS baseline time series in northeast direction on large scale became compressive generally from 2005 in the North-South Seismic Belt, which reflects that relative compression deformation enhances. The cross-fault leveling data show that annual vertical change rate and deformation trend accumulation rate in the Longmenshan fault zone were little, which indicates that vertical activity near the fault was very weak and the fault was tightly locked. According to analyses of GPS and cross-fault leveling data before the Wenchuan earthquake, we consider that the Longmenshan fault is tightly locked from the surface to the deep, and the horizontal and vertical deformation are weak surrounding the fault in relatively small-scale crustal deformation. The process of weak deformation may be slow, and weak deformation area may be larger when large earthquake is coming. Continuous and slow compression deformation across eastern Qinghai-Tibetan Plateau before the earthquake provides dynamic support for strain accumulation in the Longmenshan fault zone in relative large-scale crustal deformation.
Key words: MS8.0 Wenchuan earthquake     GPS data     Cross-fault leveling data     Fault locking     Block deformation    

INTRODUCTION

The MS8.0 Wenchuan earthquake on May 12, 2008 occurred on the NE-trending Longmenshan fault zone at the junction of the Bayan Har block and the South China block. The earthquake basically ruptured unilaterally along the NE-trending fault zone. The Wenchuan earthquake is the first strong earthquake with magnitude 8.0 of high angle thrusting type that occurred on the low slip rate faults in the interior of the continent since the records of history. Its genesis mechanism and occurrence mechanism may be different from other large thrusting earthquakes (Zhang Peizhen et al., 2009). To understand the gestation process of the Wenchuan earthquake, it is not enough to study only its seismogenic fault—the Longmenshan fault zone, but also a larger space-time scale. Therefore, it is necessary to observe and study the movement and deformation of adjacent geological tectonic units in different time periods and scales before earthquakes, which are related to seismogenic faults, in order to obtain complete and reliable information, including cross-fault GPS velocity profile, large-scale GPS strain rate field of the surface, distribution of locking degree on fault plane, time series of large-scale GPS baseline, short leveling data across faults of near-field, etc. Till May 12, 2018, the Wenchuan earthquake had occurred for ten years. During this period, many scientists used pre-earthquake GPS data and cross-fault data to study the locking and deformation characteristics of Longmenshan fault zone, the movement and deformation evolution process of Bayan Har block and South China block, and so on. Chen Z. et al. (2000), Zhang Peizhen et al.(2008, 2009), Hubbard J. et al. (2009), Jiang Zaisen et al. (2009), Du Fang et al. (2009), Zhang Peizhen et al. (2010), Zhao Jing et al. (2012) and Wu Yanqiang et al. (2015) used GPS velocity profile method to analyze the deformation characteristics of Longmenshan fault zone in the far and near field; Jiang Zaisen et al. (2009), Wu Yanqiang et al. (2011a), Zhao Jing et al. (2012), Wei Wenxin et al. (2015) and Wu Yanqiang et al. (2015) carried out relevant analysis of strain rate field of Longmenshan fault zone and its surrounding areas; Shen Zhengkang et al. (2005), Meade B. J. et al. (2007) and Zhao Jing et al. (2012) established block and fault models to study the movement characteristics, locking degree and spatial deficit distribution of Longmenshan fault zone; Jiang Zaisen et al. (2009) used large-scale GPS baseline to analyze large-scale crustal deformation of the North-South seismic zone; Jiao Qing et al. (2008), Su Qin et al. (2009), Cheng Wanzheng et al. (2011) and Bo Wanju et al. (2011) analyzed the vertical movement of Longmenshan fault zone using short leveling data across faults; Yang Guohua et al. (2006), Chen Zu'an et al. (2011), Chen Yuntai et al. (2013) and Xu Jing et al. (2017) analyzed the impact of large earthquakes and surrounding strong earthquakes on Longmenshan fault zone and adjacent areas. In addition, the predecessors have done a lot of researches in other areas. The above research helps us further understand the deformation characteristics of the Longmenshan fault zone in the near and far fields before the Wenchuan earthquake and to study the gestation process of the Wenchuan earthquake. However, most of these studies only involve one or several methods, without using many kinds of data and methods to comprehensively analyze the movement and deformation of the surface and deep fault, horizontal and vertical movement and locking of the fault, and so on. Therefore, there is no comprehensive analysis of far-near field, deep-shallow and three-dimensional of Longmenshan fault zone. The purpose of this paper is to summarize the ten-year research results in this field, so that readers can understand the evaluation characteristics and current understanding of GPS and cross-fault data before the Wenchuan earthquake.

GPS and cross-fault observation data can effectively monitor the three-dimensional dynamic deformation characteristics of the crust at different scales in the fault zone and its surrounding areas, and the deep and shallow movement and locking of the fault greatly affect the crustal deformation. Therefore, this paper will mainly analyze the movement and deformation characteristics of the far and near fields of the fault zone based on 1999-2007 GPS velocity field and the results of previous cross-fault velocity profiles, and analyze the strain rate field in the surrounding area based on the results of the least squares collocation method, and estimate the locking degree and slip deficit distribution of the Longmenshan fault zone before the Wenchuan earthquake using DEFNODE negative dislocation inversion program, analyze the crustal movement characteristics of the North-South Seismic Belt based on the results of the previous large-scale GPS baseline time series, and analyze annual average vertical change rate of different sites and spatial distribution of fault deformation accumulation rate of the Longmenshan fault zone using the short leveling data across faults from 1985 to 2008. Based on the above research results, the evolution characteristics of horizontal and vertical deformation, surface and deep fault movement and deformation in the Longmenshan fault zone before the Wenchuan earthquake are comprehensively analyzed, and the development process and earthquake mechanism of the Wenchuan earthquake are discussed in order to have a more comprehensive and in-depth understanding of the high-angle thrust MS8.0 earthquake on the low-slip-rate faults in the Chinese mainland.

1 CROSS-FAULT GPS VELOCITY PROFILE

GPS velocity profiles across fault zones can be used to analyze the long-field and near-field deformation characteristics of fault zones (Savage J. C. et al., 1973; Meade B. J. et al., 2005), and many scholars (Zhang Peizhen et al., 2008, 2009; Jiang Zaishen et al., 2009; Du Fang et al., 2009; Zhang Peizhen et al., 2010; Zhao Jing et al., 2012; Wu Yanqiang et al., 2015) made analysis on the Longmenshan fault zone after the Wenchuan earthquake. Fig. 1 is a GPS velocity profile perpendicular to the Longmenshan fault zone. The data on the northwestern side of the profile show that both the horizontal dextral strike-slip component parallel to the Longmenshan fault zone and the horizontal shortening component perpendicular to the Longmenshan fault zone reveal that the western Sichuan plateau with a width of 500km was subjected to obvious continuous deformation before the earthquake. Its dextral strike-slip rate is about 9-11mm/a and the horizontal shortening rate is about 4-5mm/a. The linear gradient of velocity profile in the western plateau indicates that its deformation and strain are quasi-uniform (Zhang Peizhen et al., 2009). Data from the southeastern side of the profile show that the velocity of GPS stations on the side of the Sichuan Basin is near zero, indicating that there is no obvious trend of deformation in the Sichuan Basin, and the velocity of GPS stations across the whole Longmenshan fault zone is consistent with that in the Sichuan Basin, where the deformation trend can hardly be seen (Jiang Zaisen et al., 2009). The above results show that the western Sichuan plateau experienced strong deformation before the earthquake, and its spatial deformation range is much larger than that of the Longmenshan fault zone and Sichuan Basin (Zhang Peizhen et al., 2009).

Fig. 1 Active tectonic map of the Longmenshan area and GPS velocity profile before the Wenchuan earthquake (adapted after Zhang Peizhen et al., 2010) (a)Active tectonic map and GPS velocity field in the Longmenshan area (relative to the South China block). (b) GPS velocity profile across the Longmenshan fault zone before the Wenchuan MS8.0 earthquake
2 GPS STRAIN RATE FIELD

The strain rate field calculated by the GPS velocity field can reflect the deformation characteristics such as tension and shear in the upper part of the crust. Therefore, the strain rate field results of the Longmenshan fault zone and its vicinity before the Wenchuan earthquake can reflect some of the characteristics of the earthquake. The GPS horizontal velocity field data in the range of 99°—107°E and 25°—34°N is selected and calculated by the least squares collocation method (Wu Yanqiang et al., 2009, 2011b; Jiang Zaisen et al., 2010) to get the distribution of the principal strain rate, the east-west strain rate, and the maximal shear strain rate of the region (Fig. 2). The principal strain rate results (Fig. 2(a)) show that the strain accumulation rate in the southwestern part of the Longmenshan fault zone is much larger than that in the middle and northern sections, and has obvious zoning characteristics. The whole fault zone is characterized by extrusion and right-lateral shear deformation. The deformation of the northwestern side of the mid-northern part of the fault zone decreases from the far fault zone to the near fault zone, indicating that the long-term deformation of the western Sichuan Plateau continuously provides strain accumulation in the Longmenshan fault zone and was longer before the earthquake. During the period, high strain energy has been accumulated, and the Longmenshan fault zone has only weak inter-seismic deformation due to its strong locked state and strong blockage of the Sichuan Basin, making it impossible to observe the fault zone deformation before the Wenchuan earthquake with GPS observations (Zhao Jing et al., 2012). The deformation of the southeastern part of the mid-northern part of the fault zone is very weak, indicating that the Sichuan Basin as a supporting unit has a stable blocking effect on the compression of the western Sichuan Plateau before the Wenchuan earthquake, which is also a necessary condition for locking and high strain accumulation of the Longmenshan fault zone (Zhang Peizhen et al., 2009). The east-west strain rate results (Fig. 2(b)) indicate that a significant east-west compressive strain concentration zone is formed in the southwestern section of the Longmenshan fault zone, because of the eastward movement of the Bayan Har block. The northwestern side of the mid-northern section is obviously deformed, while the southeastern side of the fault zone is an east-west tensile deformation zone. This result indicates that the Longmenshan fault zone acts as a boundary zone of east-west deformation and has a background of long-term accumulation of compression strain (Zhao Jing et al., 2012), and the Wenchuan earthquake occurred in the edge region of the east-west strong extrusion deformation. The maximal shear strain rate result (Fig. 2(c)) shows that the strain rate is higher in the Xianshuihe-Anninghe-Zemuhe-Xiaojiang fault zone (the fault slip rate is not deducted during calculation, so fault creeping may affect the strain rate), but the strain rate is not obvious in the mid-northern section of the Longmenshan fault zone(Wu Yanqiang et al., 2015), indicating that the right-lateral shearing motion of the mid-northern section of the Longmenshan fault zone was also weak before the Wenchuan earthquake, and the fault was in locked state. In addition, from the east-west deformation characteristics of the western China (Wei Wenxin et al., 2015; Wu Yanqiang et al., 2015), the Qinghai-Tibetan block is mainly manifested as east-west tension west of 92.5°E reflecting the eastward flow of crustal material on the Qinghai-Tibetan Plateau, and the main manifestation of the deformation is east-west extrusion east of 92.5°E, which continues until the Longmenshan fault zone, reflecting that the eastward flow of crustal material in the Qinghai-Tibetan Plateau has been blocked by the North China and South China blocks. The extrusion zone is more than 900km wide, and the Longmenshan fault zone, as the boundary of the extrusion zone, has been receiving the accumulation of large-scale extrusion strain energy from the Qinghai-Tibetan Plateau.

Fig. 2 Results of GPS strain rate field from 1999 to 2007 (adapted after Zhao Jing et al., 2012; Wu Yanqiang et al., 2015) (a)Distribution of principal strain rates. (b)Distribution of east-west strain rates. (c)Distribution of maximal shear strain rates

The third-phase results of the east-west strain rate field in the Sichuan-Yunnan region before the Wenchuan earthquake (Fig. 3) reflect the evolution characteristics of the east-west extrusion deformation around the Longmenshan fault zone. From 1999 to 2001 (Fig. 3(a)), the high value area of the east-west extrusion deformation is small, and the highest value area is located in the Xianshuihe fault zone; the southwestern section of the Longmenshan fault zone shows strong extrusion deformation, while the extrusion deformation of mid-northern segment is weak. From 2001 to 2004 (Fig. 3(b)), the magnitude and the high value area of the east-west extrusion deformation increased, and the maximum extrusion deformation was still located in the Xianshuihe fault zone and moved eastward. The extrusion deformation range in the Longmenshan fault zone has increased, and the extrusion deformation in the southwestern section has been further enhanced, and that near the mid-northern sections is stronger. From 2004 to 2007 (Fig. 3(c)), the magnitude of the east-west extrusion deformation increased significantly, and the maximum extrusion deformation continued to migrate eastward to the junction of the Longmenshan fault zone and the Xianshuihe fault zone. During the gradual eastward migration of the east-west strain rate field, the east-west extrusion deformation of the Longmenshan fault zone gradually increased and reached its highest level from 2004 to 2007 (Wu Yanqiang et al., 2011a).

Fig. 3 East-west strain rate field in Sichuan-Yunnan region (a)1999-2001 results. (b) 2001-2004 results. (c) 2004-2007 results
3 FAULT LOCKING AND SLIP DEFICIT

Based on the analysis of the cross-fault GPS velocity profile and the GPS strain rate field, we have established a block and fault physical model to invert the three-dimensional distribution of locking degrees and slip deficit of the Longmenshan fault zone before the Wenchuan earthquake. The calculation attempts to gain a deeper understanding of the Wenchuan earthquake from the perspective of deep motion and deformation. The basic principles and methods of the DEFNODE negative dislocation inversion procedure are described in the references (McCaffrey R., 2002, 2005; Savage J. C. et al, , 2001; Zhao Jing et al., 2012, 2015), which assumes that the motion of the interior points of the block is a sum of the block rotation, the uniform strain inside the block and the surface elastic deformation are caused by fault locking, so the program can use the GPS data, grid search and simulated annealing to calculate the Euler pole of block rotation, the uniform strain inside the block, the three-dimensional distribution of fault locking and slip deficit on block boundary, etc. The locking degree is the ratio of the slip deficit rate to the long-term slip rate of the fault. The ratio of 0 indicates the complete creep of the fault, and the ratio of 1 indicates the complete lock of the fault. The ratio is between 0 and 1, indicating that the fault is partially locked and there is a partial creep motion. Firstly, the inversion reliability and stability of the program are verified. The results show that the inversion results are highly reliable under different fault dip conditions. The weight f has a weak influence on the inversion results, and the inversion results also have good stability. GPS velocity field data with 1° interval distribution along the longitude and latitude can basically constrain the inter-seismic deformation pattern of the seismogenic fault (Zhao Jing et al., 2013).

The data used in the GPS velocity field from 1999 to 2007 before the Wenchuan earthquake (provided by Research Professor Wang Min). A total of 186 stations participated in the inversion. Therefore, the spatial resolution of GPS should be able to constrain the inter-seismic deformation of the Wenchuan earthquake. Considering that the Longmenshan fault zone, the Xianshuihe fault zone and the Anninghe-Zemuhe-Xiaojiang fault zone form an intersection at the eastern edge of the Qinghai-Tibetan Plateau, the movement and deformation between the three faults zones interact, so when the physical model of the fault and block is established, the whole research area is divided into the Bayan Har block, the Sichuan-Yunnan block and the South China block by the above three faults. The inversion results of spatial distribution of locking degree and slip deficit in the Longmenshan fault zone are shown in Fig. 4. The results of Fig. 4(a) show that the entire Longmenshan fault is basically in strong locked state, and only the source of the Wenchuan earthquake is relatively weak, and the locking degree is about 0.88. About 20km width of the southwestern section is creeping in the depth of 12-22.5km, which is the only unlocked position in the whole fault zone, while the MS7.0 Lushan earthquake occurs in the transition from strong-locked to unlocked (Zhao Jing et al., 2018). The results of Fig. 4(b) shows that the parallel slip deficit rate of the Longmenshan fault is right-lateral deficit, and the deficit rate in the northeast is the largest, about 5.8-6.0mm/a; the deficit rate of the completely locked section in the mid-southern section is about 5.7-5.8mm/a, slightly weaker than that of northeastern section. The results of Fig. 4(c) shows that the vertical slip deficit rate of the Longmenshan fault is compressive deficit, and the slip deficit rate increases from the northeast to the southwest. The deficit rate in the northeast is the smallest, about 0.6-1.1mm/a; the deficit rate of the completely locked section in the mid-southern section is about 2.0-3.6mm/a.

Fig. 4 Spatial distribution of locking degree and slip deficit rate of Longmenshan fault zone before the Wenchuan earthquake (reconstructed by Zhao Jing et al., 2018) (a)Fault locking. (b) Slip deficit rate of parallel fault. (c) Slip deficit rate of vertical fault
4 LARGE-SCALE GPS BASELINES

The time series of a single GPS station will be affected by the reference frame and seasonal weather, and these effects are related to a certain extent. Therefore, it is necessary to find some parameters which have physical significance and are less affected by common errors to describe crustal deformation information. Extracting the time series of baselines between two GPS stations can have little correlation with the reference frame and weaken the influence of some homologous errors. The information of crustal deformation can be highlighted, and the use of the arc length of the geodetic line as the baseline time series can avoid the influence of GPS normal observation error (its non-structural effect is more prominent). Filtering can weaken the influence of non-structural factors in many annual cycles and shorter frequency bands, retain relatively low-frequency trend changes with stability, and the crustal movement information can be more reliable (Jiang Zaisen et al., 2009). Therefore, the method of extracting the low-frequency information of the geodetic arc between two GPS stations is commonly used in crustal deformation analysis. After the Wenchuan earthquake, Jiang Zaishen et al. (2009) processed the continuous observation data of GPS base stations on the Chinese mainland. The results show that the NE-trending GPS baseline time series across the North-South Seismic Belt region has experienced a general turning since 2005 (Fig. 5(a)), reflecting the enhanced relative movement of NE-trending crustal shortening in the North-South Seismic Belt and the NE-trending of the Qinghai-Tibetan block relative to the South China block. The NE or NNE baselines associated with KMIN (Kunming), XIAG (Xiaguan) and LUZH (Luzhou) stations in the southwestern region show that after the Sumatra MW9.0 earthquake of December 26, 2004, they began to elongate relatively in 2004-2005 and then changed to shortening compression in 2005-2006, such as the KMIN (Kunming) -YANC (Yanchi) baseline (Fig. 5(b)), indicating the southwestern region is obviously affected by the Sumatra earthquake (Yang Guohua et al., 2006). The northeast-oriented baseline inside the Qinghai-Tibetan block and the South China block did not show a trend of shortening and strengthening. At the same time, triangular deformation unit strain parameter time series results containing LUZH (Luzhou)-LHAS (Lhasa)-YANC (Yanchi) GPS base stations (Jiang Zaisen et al., 2009) show that the large-scale crustal deformation changes in the North-South Seismic Belt have enhanced the east-west crustal shortening and northeast right-lateral shear activity, which are consistent with the tectonic dynamics required for seismicity in Longmenshan fault zone, so the Sumatra earthquake may have contributed to the Wenchuan earthquake.

Fig. 5 The GPS geodesic curve of northeast direction with the linear trend removed in the vicinity of North-South Seismic Belt (Jiang Zaisen et al., 2009) (a) LHAS (Lhasa)-YANC (Yanchi). (b) KMIN (Kunming)-YANC (Yanchi) The ordinate is the change in the length of the baseline, the thin line is the result of the least squares collocation processing, the thick line is the result of 9-order trend using DB6 wavelet processing, and the long-term linear rate of geodesic measurements is indicated at the lower right of the figure
5 CROSS-FAULT LEVELING DATA RESULTS

Fig. 6 shows the distribution and elevation variation of short-leveling sites across faults in Longmenshan fault zone before the Wenchuan earthquake. The short-leveling sites include Guanxian, Qipangou, Gengda, Shuanghe and Pujiang. The time span of leveling data is 1985-2008. Comprehensive analysis shows that there are no obvious short-term anomalies in the cross-fault deformation data of the five sites before the Wenchuan earthquake (Su Qin et al., 2009; Bo Wanju et al., 2011).

Fig. 6 Distribution of cross-fault leveling sites in the Longmenshan fault zone (a) and (b) variation of level difference before the Wenchuan earthquake The Guanxian, Qipangou and Gengda sites in the north are represented by dots, the Shuanghe sites in the middle are represented by blocks, and Pujiang in the south is represented by diamonds

It can be seen from Fig. 6 that there was a large extension anomaly of A-B line in Gengda before the earthquake, but it was verified that the change was caused by environmental factors at point A, and the period of the anomaly coincided well with the period of local building construction, so it was not regarded as an earthquake precursor anomaly (Su Qin et al., 2009); the line 3-4 in Guanxian had a significant extension change from 1992 to 1994, with a maximum of 2.47mm. After 1994, the fault is in a weak active state, and there is a weak change of compression and extension from 2005 to 2008. The line 1-2 in Pujiang has been in a continuous extension state since 1987. From September 1993 to September 1995, an abnormal process occurred in which the extension rate increased and then weakened (Jiao Qing et al., 2008). The curve recovery trend from 1995 to 2002 indicates that the extension movement of the fault segment recovered. The curve flattened from 2002 to 2004, the leveling change weakened obviously, and the fault activity disappeared. Since the end of 2004, the leveling extension change has increased again, and the extension turned to compression in the second half of 2006, then the Wenchuan earthquake occurred (Cheng Wanzheng et al., 2011). The annual variation period of the leveling of Qipangou and Shuanghe sites are regular, and at the same time, it presents certain compressional change.

In order to further study the short-leveling changes across faults before the Wenchuan earthquake, we calculated the average annual vertical change rate (Fig. 7). From the results of the annual variation rate of several sites, it can be seen that the average annual variation rate along the Longmenshan fault zone fluctuates around 0, indicating that the vertical movement of the whole fault is weak and is in a state ofstrong locking, which is in line with the fault locking inverted by 1999-2007 GPS results. The average annual vertical change rate of Pujiang site and Guanxian site changed slightly in 2006 and 2007 respectively, which may indicate that the activity of faults increased slightly before the Wenchuan earthquake. However, it is not difficult to see from the deformation data of all faults in Longmenshan fault zone that there was no obvious short-term jump or rate change anomaly before the Wenchuan earthquake.

Fig. 7 Average annual vertical change rate of cross-fault leveling in the Longmenshan fault zone before the Wenchuan earthquake

In order to study the spatial distribution of short-leveling vertical variation across faults in the vicinity of Longmenshan fault zone before the Wenchuan earthquake, we calculated the accumulative rate of vertical deformation of faults (Fig. 8). Dc is the ratio of the average annual variation rate of long-term fault deformation to the average variation amplitude (dimensionless). Its absolute value ranges from 0 to 1, and if Dc tends to 1, it indicates that the linear trend of the curve is better and the fluctuation change is smaller. The more Dc tends to 0, the more stable the curve changes and the less the deformation and strain accumulation occur (Yue Chong et al., 2017). The results of Fig. 8 shows that the absolute value of Dc of Zhuwo, Goupu and Anshunchang sites along the Xianshuihe fault zone is tending to 1, indicating that they are in the process of deformation and strain accumulation for a long time, while the rate of deformation and strain accumulation of other sites is low. Based on the results of 8 sites along the Xianshuihe fault zone, it can be seen that the locking of the northwestern segment of the fault zone is weak, while the locking of the southeastern segment is strong. The Gengda, Qipangou and Shuanghe sites in Longmenshan fault zone have very small accumulative rate of deformation. Combining with the results of Fig. 7, we can see that the fluctuation of the average annual vertical change rate curves of these three sites are not strong, which indicates that the activities of these 3 sites are very weak. The accumulative rate of deformation of Pujiang and Guanxian sites are slightly larger. Combining with the results of Fig. 6 and Fig. 7, we can see that the larger accumulative rate of deformation of Guanxian sites is mainly caused by the observation data from 1992 to 1996. Based on the results of 5 sites in Longmenshan fault zone, it can be seen that the cumulative rate of deformation in the whole fault zone is very low, indicating that there is no obvious change in the fault movement rate and the fault is in strong locked state.

Fig. 8 The spatial distribution contour of the cumulative rate of fault deformation before the Wenchuan earthquake (results of 1990-2008)
6 DISCUSSION AND CONCLUSION

(1) Dividing the coseismic displacement of Wenchuan earthquake by the interseismic fault slip rate obtained by the GPS method, the recurrence interval of the Wenchuan earthquake is 2500-5000 years (Zhao Jing et al., 2012), 3190-5952 (Zhang Peizhen et al., 2008), and 2000-10000 years (Burchfiel B. C. et al., 2008), but whether the fault slip rate obtained by the 10-year GPS results can represent the slip rate of faults in the past few thousand years or even longer is the key to calculating the interval of earthquake recurrence (Yin An, 2010). Large faults such as the Himalayan fault can accumulate elastic strain away from the fault for about 500km (Feldl N. et al., 2006), while the near-field GPS observations do not show significant deformation, but when the accumulated energy is released, it can promote the occurrence of an earthquake in the fault. The Longmenshan fault zone has a certain similarity with the Himalayan fault, and both are intracontinental thrust fault zones, except that the Longmenshan fault zone is the obduction fault zone (the active block is the hanging wall block) and the Himalayan fault zone is the subduction fault zone (the active block is the footwall block), so in addition to the Longmenshan fault zone on the eastern edge of the Qinghai-Tibetan Plateau, there should be a large-scale strain accumulation area in the eastern part of the Bayan Har block, and the cross-fault GPS velocity profile results show that the entire 500km width of the western Sichuan Plateau experienced a strong continuous deformation before the earthquake. The seismic geological results show that the slip rate of the Longmenshan fault zone is higher than that given by the GPS results before the Wenchuan earthquake. Therefore, dividing the Wenchuan earthquake coseismic displacement by the fault slip rate obtained by seismic geological methods, the recurrence of the Wenchuan earthquake is 2233-4167 years, which is shorter than the recurrence interval obtained by the GPS results (Zhang Peizhen et al., 2008).

(2) The elastic strain released by the large earthquake is derived from the long-term strain accumulation caused by the crust movement before the earthquake. The Wenchuan earthquake may have experienced the elastic strain accumulation process for thousands of years, and the observation time of the cross-fault data is less than 30 years. The observation time of GPS data is only 10 years, and it is in the final stage of the Wenchuan earthquake period. Therefore, the elastic strain accumulation of the crust near the Longmenshan fault zone at this stage has reached the limit, and it is in a state of very slow accumulation or even no accumulation. This may be the reason why the precursor anomalies, especially the near-fault anomalies, are not easy to highlight, and it is one of the important reasons that we have underestimated the seismic hazard on the medium and long-term scale. After the Wenchuan earthquake, we realized that faults at low motion and deformation rates do not necessarily represent low potential seismic hazards. In particular, thrust faults may reflect the fact that they are in a state of strong locking and high accumulation of strain energy. Since there is no GPS continuous observation stations in the northwestern block of the Longmenshan fault zone before the Wenchuan earthquake, we have not obtained good observations on the continuous deformation of the crust near the fault before earthquake, and we cannot analyze the evolution feature of continuous deformation near the fault with continuous GPS data. After the Wenchuan earthquake, Jiang Zaisen et al. in Institute of Earthquake Forecasting, China Earthquake Administration recognized that there is still a risk of strong earthquakes in the southern section of the Longmenshan fault zone where no rupture occurred. Therefore, they deployed 10 GPS continuous observation stations around the southern section, and successfully captured the continuous deformation evolution characteristics of the southern section of the Longmenshan fault zone before the Lushan earthquake (Liu Xiaoxia et al., 2015), this further improves our understanding of the pre-earthquake deformation and seismogenic mechanism of the intracontinental thrust fault.

(3) In seismological research, inter-seismic deformation is one of the challenging problems due to the influence of long time windows and various interference factors. In the roughly half a century before the Wenchuan earthquake, there were many strong earthquakes of magnitude 6.0-7.0 in the surrounding area of the Longmenshan fault zone, including the Songpan M7.2 earthquake on August 16, M6.7 earthquake on August 22, and M7.2 earthquake on August 23, 1976, the M6.0 earthquake in the Luding-Tianquan area on June 1, 1941, and the Dayi M6.2 earthquake on February 24, 1970. The results of the Coulomb stress change caused by the coseismic and post-earthquake effects of large earthquakes show that the Songpan earthquake sequence has a weak impact on the 2008 Wenchuan earthquake, the 2013 Lushan earthquake and the entire Longmenshan fault zone (Xu Jing et al., 2017). The seismic energy released in the 1941 Luding-Tianquan M6.0 earthquake and the 1970 Dayi M6.2 earthquake is too small, and the release of seismic wave energy is only about 6% to 13% of MW6.8 earthquake (the magnitude of the unruptured section between the Wenchuan earthquake and the Lushan earthquake), which is not enough to change the seismic estimation of dangerous faults in the southwestern section of the Longmenshan fault zone (Chen Yuntai et al., 2013). The impact of the earthquake has both coseismic and post-earthquake creep adjustments. Although the above studies indicate that the impact of these strong earthquakes on the Wenchuan earthquake may be weak, the main factor is still the energy accumulation of the Wenchuan earthquake itself, but the inter-seismic deformation of the Longmenshan fault zone and surrounding areas given in this paper is still affected by these earthquakes, which is also a reason causing the difference in deformation between different earthquakes.

(4) The GPS velocity profile across the fault zone shows that both the horizontal shortening component perpendicular to the Longmenshan fault and the horizontal right-lateral strike-slip component parallel to the Longmenshan fault reveal that the western Sichuan Plateau of 500km wide has been suffering from obvious and continuous deformation before the earthquake; the velocity of the GPS stations on one side of the Sichuan Basin and the velocity of the GPS stations across the Longmenshan fault zone are all near zero, and there is basically no deformation trend, indicating that the western Sichuan Plateau is continuously providing energy accumulation for the Longmenshan fault zone that was already in locked state before the Wenchuan earthquake.

(5) The results of GPS principal strain rate show that the strain accumulation rate in the southwestern section of the Longmenshan fault zone is significantly larger than that in the mid-northern sections, and the whole fault zone is characterized by extrusion and right-lateral shear deformation. The deformation of the northwestern side of the mid-northern section of the Longmenshan fault zone is gradually reduced from far field to near fault zone, and the deformation in the fault zone does not occur or only weak inter-seismic deformation occurs due to its strong locked state. The results of GPS area strain rate show that the southwestern section of the Longmenshan fault zone has formed a significant east-west compression strain concentration zone; taking the mid-northern section of the Longmenshan fault zone as the boundary, the northwestern side of the fault zone is obviously deformed, and the southeastern side of the fault zone belongs to the east-west tensile deformation zone. The results of GPS shear strain rate show that the maximal shear strain rate in the Xianshuihe-Anninghe-Zemuhe-Xiaojiang fault zone is very high, while the shear strain in the mid-northern section of the Longmenshan fault is not obvious. A variety of results of GPS strain rate indicate that the Longmenshan fault zone is the boundary of movement and deformation between the Bayan Har block and the South China block, and has a background of long-term accumulation of extrusion and right-lateral shear strain.

(6) The three-dimensional distribution inversion results of the locking degree and slip deficit show that the Longmenshan fault is basically in strong locked state before the Wenchuan earthquake, and only the source area of the Wenchuan earthquake is relatively weakly locked (locking degree is about 0.88), and in southwestern section, there are faults of about 20km wide in creeping state at 12-22.5km depth, while the Lushan earthquake occurred in the transition from strong locked to unlocked. The slip deficit rate of parallel Longmenshan fault is the right-lateral deficit rate. The deficit rate in the northeast is about 5.8-6.0mm/a, and the deficit rate is about 5.7-5.8mm/a in the completely locked section of the mid-southern section; The slip deficit rate of vertical Longmenshan fault is the extrusion deficit rate. The deficit rate in the northeastern segment is about 0.6-1.1mm/a, and the deficit rate is about 2.0-3.6mm/a in the completely locked section of the mid-southern segment. In addition, the randomly selected two groups, which only contain 80% GPS data inversion calculations, are basically consistent with the original results, only having slightly difference at the creeping position of the southwestern segment. The weight f is adjusted from the optimal value of 1.74 to 1.64 and 1.84, and the calculated results are basically consistent with the original results, indicating that the inversion results have good reliability and stability.

(7) Large-scale GPS baseline results show that the long-term trend of the NNE-trending GPS baseline time series across the North-South Seismic Belt has generally changed since 2005, reflecting the relative motion enhancement of the NNE-trending crustal shortening. The NE or NNE baselines associated with the stations in the southwestern region show that the NNE baselines relatively extended after the 2004 Sumatra MW9.0 earthquake in 2004-2005, and then shortened in 2005-2006. The time series results of strain parameters of triangular deformation units formed by GPS datum stations show that the large-scale crustal deformation in the North-South Seismic Belt is shortened in the near east-west direction and strengthened in the NE-direction right-lateral shear activity, which is consistent with the tectonic dynamics required for the Longmenshan fault zone to generate earthquakes.

(8) The calculation results of the short-leveling sites along the Longmenshan fault zone (Guanxian, Qipangou, Gengda, Shuanghe, Pujiang sites) show that there are basically not significant short-term sudden jump or rate change anomaly in the original curve and average annual vertical change rate before the Wenchuan earthquake. The deformation rate of the entire Longmenshan fault zone is very low, indicating that the activity near fault is weak and is in state of strong vertical locked. The horizontal locked state of the GPS velocity field is consistent with the vertical strong locked state reflected by the short-leveling data of the fault, indicating that before the occurrence of the Wenchuan earthquake, the movement is in strong locked state and cannot move freely regardless of the deep-shallow section or the horizontal-vertical direction of the fault.

(9) Through the correlation analysis of GPS and cross-fault data before the Wenchuan earthquake, we can see that in relatively small-scale crustal deformation, the deep and shallow parts of Longmenshan fault zone are in strong locked state before the Wenchuan earthquake, and the horizontal and vertical deformation of the fault zone are very weak. Regarding the recurrence interval of more than two thousand years, this phenomenon may have undergone a long process. And the nearer the earthquake occurs, the wider the range of weak deformation may be. At the same time, in the relatively large-scale crustal deformation, the eastern part of the Bayan Har block on the western side of the Longmenshan fault zone experienced a slow and sustained crustal shortening and compression deformation before the Wenchuan earthquake, which provided dynamic support for the continuous strain accumulation of the Longmenshan fault zone. The left-lateral strike-slip dislocation occurred in the seismogenic fault of the Kunlun M8.1 earthquake in 2001, which resulted in the further eastward expansion and deformation of the Bayan Har block on the southern side of the East Kunlun fault zone, and was blocked by the Sichuan Basin on the eastern side, which increased the Coulomb failure stress of the Longmenshan fault zone and enhanced the accumulation of compressive strain energy (Jiang Zaisen et al., 2009; Chen Zu'an et al., 2011). The Sumatra MW9.0 earthquake in 2004 promotes the northward movement of the Indian plate, while the southwestward movement of the Sichuan-Yunnan block and South China block accelerated (Yang Guohua et al., 2006), resulting in a significant increase in the relative movement of large-scale NE-trending crustal shortening in the North-South Seismic Belt. This large-scale effect is consistent with the dynamic action required for generating right-lateral shear rupture in the Longmenshan fault zone. The Sumatra earthquake may promote the accumulation and enhancement of dextral shear strain energy in the Longmenshan fault zone (Jiang Zaisen et al., 2009).

REFERENCES
Bo Wanju, Yang Guohua, Zhan Wei, Zhang Fengshuang, Wan Wenni, Zhang LichengBo Wanju, Yang Guohua, Zhan Wei, Zhang Fengshuang, Wan Wenni, Zhang Licheng. Preparatory mechanism of MS8.0 Wenchuan earthquake evidenced by crust-deformation data[J]. Geodesy and Geodynamics, 2011, 2(2): 23-28. DOI:10.3724/SP.J.1246.2011.00023
Burchfiel B.C., Royden L.H., Van Der Hilst R.D., Hager B.H., Chen Z., King R.W., Li C., Lü J., Yao H., Kirby E.Burchfiel B.C., Royden L.H., Van Der Hilst R.D., Hager B.H., Chen Z., King R.W., Li C., Lü J., Yao H., Kirby E. A geological and geophysical context for the Wenchuan earthquake of 12 May 2008, Sichuan, the People's Republic of China[J]. GSA Today, 2008, 18(7): 4-11. DOI:10.1130/GSATG18A.1
Chen Yuntai, Yang Zhixian, Zhang Yong, Liu ChaoChen Yuntai, Yang Zhixian, Zhang Yong, Liu Chao. From 2008 Wenchuan earthquake to 2013 Lushan earthquake[J]. Scientia Sinica Terrae, 2013, 43(6): 1064-1072 (in Chinese). DOI:10.1360/zd-2013-43-6-1064
Chen Z., Burchfiel B.C., Liu Y., King R.W., Royden L.H., Tang W., Wang E., Zhao J., Zhang X.Chen Z., Burchfiel B.C., Liu Y., King R.W., Royden L.H., Tang W., Wang E., Zhao J., Zhang X. Global positioning system measurements from eastern Tibet and their implications for India/Eurasia intercontinental deformation[J]. Journal of Geophysical Research: Solid Earth, 2000, 105(B7): 16215-16227. DOI:10.1029/2000JB900092
Chen Zu'an, Lin Banghui, Bai Wuming, Cheng Xu, Wang YunshengChen Zu'an, Lin Banghui, Bai Wuming, Cheng Xu, Wang Yunsheng. A study on the rupture process of the 2001 MS8.1 Kunlunshan earthquake and its influence on pregnant process and occurrence of the MS8.0 Wenchuan earthquake in 2008[J]. Chinese Journal of Geophysics, 2011, 54(1): 108-120 (in Chinese with English abstract).
Cheng Wanzheng, Guan Zhijun, Su Qin, Ruan Xiang, Zhang ZhiweiCheng Wanzheng, Guan Zhijun, Su Qin, Ruan Xiang, Zhang Zhiwei. Precursory anomalies in Sichuan region before 2008 Wenchuan MS8.0 earthquake and their statistical analysis[J]. Acta Seismologica Sinica, 2011, 33(3): 304-318 (in Chinese with English abstract).
Du Fang, Wen Xueze, Zhang Peizhen, Wang QingliangDu Fang, Wen Xueze, Zhang Peizhen, Wang Qingliang. Interseismic deformation across the Longmenshan fault zone before the 2008 M8.0 Wenchuan earthquake[J]. Chinese Journal of Geophysics, 2009, 52(11): 2729-2738 (in Chinese with English abstract).
Feldl N., Bilham R.Feldl N., Bilham R. Great Himalayan earthquakes and the Tibetan plateau[J]. Nature, 2006, 444(7116): 165-170. DOI:10.1038/nature05199
Hubbard J., Shaw J.H.Hubbard J., Shaw J.H. Uplift of the Longmen Shan and Tibetan plateau, and the 2008 Wenchuan (M=7.9) earthquake[J]. Nature, 2009, 458(7235): 194-197. DOI:10.1038/nature07837
Jiang Zaisen, Fang Ying, Wu Yanqiang, Wang Min, Du Fang, Ping JianjunJiang Zaisen, Fang Ying, Wu Yanqiang, Wang Min, Du Fang, Ping Jianjun. The dynamic process of regional crustal movement and deformation before Wenchuan MS8.0 earthquake[J]. Chinese Journal of Geophysics, 2009, 52(2): 505-518 (in Chinese with English abstract).
Jiang Zaisen, Liu JingnanJiang Zaisen, Liu Jingnan. The method in establishing strain field and velocity field of crustal movement using least squares collocation[J]. Chinese Journal of Geophysics, 2010, 53(5): 1109-1117 (in Chinese with English abstract).
Jiao Qing, Yang Xuanhui, Xu Liqing, Wang BoJiao Qing, Yang Xuanhui, Xu Liqing, Wang Bo. Preliminary study on motion characteristics of Longmenshan fault before and after MS8.0 Wenchuan earthquake[J]. Journal of Geodesy and Geodynamics, 2008, 28(4): 7-11, 37 (in Chinese with English abstract).
Liu Xiaoxia, Wu Yanqiang, Jiang Zaisen, Zhan Wei, Li Qiang, Wei Wenxin, Zou ZhenyuLiu Xiaoxia, Wu Yanqiang, Jiang Zaisen, Zhan Wei, Li Qiang, Wei Wenxin, Zou Zhenyu. Preseismic deformation in the seismogenic zone of the Lushan MS7.0 earthquake detected by GPS observations[J]. Science China Earth Sciences, 2015, 58(9): 1592-1601 (in Chinese). DOI:10.1007/s11430-015-5128-0
McCaffrey R. Crustal block rotations and plate coupling. In: Stein S., Freymueller J.T. (Editors). Plate Boundary Zones[M]. Washington: American Geophysical Union, 2002: 101-122.
McCaffrey R.McCaffrey R. Block kinematics of the Pacific-North America plate boundary in the southwestern United States from inversion of GPS, seismological, and geologic data[J]. Journal of Geophysical Research: Solid Earth, 2005, 110(B7): B07401.
Meade B.J., Hager B.H.Meade B.J., Hager B.H. Block models of crustal motion in southern California constrained by GPS measurements[J]. Journal of Geophysical Research: Solid Earth, 2005, 110(B3): B03403.
Meade B.J.Meade B.J. Present-day kinematics at the India-Asia collision zone[J]. Geology, 2007, 35(1): 81-84.
Savage J.C., Burford R.O.Savage J.C., Burford R.O. Geodetic determination of relative plate motion in central California[J]. Journal of Geophysical Research, 1973, 78(5): 832-845. DOI:10.1029/JB078i005p00832
Savage J.C., Gan Weijun, Svarc J.L.Savage J.C., Gan Weijun, Svarc J.L. Strain accumulation and rotation in the eastern California Shear Zone[J]. Journal of Geophysical Research: Solid Earth, 2001, 106(B10): 21995-22007. DOI:10.1029/2000JB000127
Shen Zhengkang, Lü Jiangning, Wang Min, Bürgmann R.Shen Zhengkang, Lü Jiangning, Wang Min, Bürgmann R. Contemporary crustal deformation around the southeast borderland of the Tibetan Plateau[J]. Journal of Geophysical Research: Solid Earth, 2005, 110(B11): B11409.
Su Qin, Zhu Hang, Yang YonglinSu Qin, Zhu Hang, Yang Yonglin. Short-line leveling anomaly at Gengda and Wenchuan MS8.0 earthquake[J]. Journal of Geodesy and Geodynamics, 2009, 29(S1): 103-105 (in Chinese with English abstract).
Wei Wenxin, Jiang Zaisen, Wu YanqiangWei Wenxin, Jiang Zaisen, Wu Yanqiang. Cognitions and questions regarding crustal deformation and location forecasts of strong earthquakes[J]. Geodesy and Geodynamics, 2015, 6(3): 210-219. DOI:10.1016/j.geog.2015.03.006
Wu Yanqiang, Jiang Zaisen, Yang Guohua, Fang Ying, Wang WuxingWu Yanqiang, Jiang Zaisen, Yang Guohua, Fang Ying, Wang Wuxing. The application and method of GPS strain calculation in whole mode using least square collocation in sphere surface[J]. Chinese Journal of Geophysics, 2009, 52(7): 1707-1714 (in Chinese with English abstract).
Wu Yanqiang, Jiang Zaisen, Yang Guohua, Liu Xiaoxia, Zhao JingWu Yanqiang, Jiang Zaisen, Yang Guohua, Liu Xiaoxia, Zhao Jing. Evolution characteristics of strain rate field reflected by GPS data before Wenchaun earthquake[J]. Journal of Geodesy and Geodynamics, 2011a, 31(5): 20-25, 29 (in Chinese with English abstract).
Wu Yanqiang, Jiang Zaisen, Yang Guohua, Wei Wenxin, Liu XiaoxiaWu Yanqiang, Jiang Zaisen, Yang Guohua, Wei Wenxin, Liu Xiaoxia. Comparison of GPS strain rate computing methods and their reliability[J]. Geophysical Journal International, 2011b, 185(2): 703-717. DOI:10.1111/j.1365-246X.2011.04976.x
Wu Yanqiang, Jiang Zaisen, Zhao Jing, Liu Xiaoxia, Wei Wenxin, Li Qiang, Zou Zhenyu, Zhang LongWu Yanqiang, Jiang Zaisen, Zhao Jing, Liu Xiaoxia, Wei Wenxin, Li Qiang, Zou Zhenyu, Zhang Long. Crustal deformation before the 2008 Wenchuan MS8.0 earthquake studied using GPS data[J]. Journal of Geodynamics, 2015, 85: 11-23. DOI:10.1016/j.jog.2014.12.002
Xu Jing, Shao Zhigang, Liu Jing, Ji LingyunXu Jing, Shao Zhigang, Liu Jing, Ji Lingyun. Analysis of interaction between great earthquakes in the eastern Bayan Har block based on changes of Coulomb stress[J]. Chinese Journal of Geophysics, 2017, 60(10): 4056-4068 (in Chinese with English abstract).
Yang Guohua, Jiang Zaisen, Wang Min, Zhang Zusheng, Liu Guangyu, Han Yueping, Ding Ping, Gong PingYang Guohua, Jiang Zaisen, Wang Min, Zhang Zusheng, Liu Guangyu, Han Yueping, Ding Ping, Gong Ping. Effect of Indonesia earthquake on horizontal crustal movement in Sichuan-Yunnan Region[J]. Journal of Geodesy and Geodynamics, 2006, 26(1): 9-14, 20 (in Chinese with English abstract).
Yin AnYin An. A special issue on the great 12 May 2008 Wenchuan earthquake (MW7.9): observations and unanswered questions[J]. Tectonophysics, 2010, 491(1/4): 1-9.
Yue Chong, Yan Wei, Li Xiaofan, Niu Anfu, Zhao Jing, Yuan ZhengyiYue Chong, Yan Wei, Li Xiaofan, Niu Anfu, Zhao Jing, Yuan Zhengyi. Sichuan fault activity analysis and correlation study of Wenchuan & Lushan seismic activity[J]. Journal of Geodesy and Geodynamics, 2017, 37(9): 888-892 (in Chinese with English abstract).
Zhang Peizhen, Xu Xiwei, Wen Xueze, Ran YongkangZhang Peizhen, Xu Xiwei, Wen Xueze, Ran Yongkang. Slip rates and recurrence intervals of the Longmen Shan active fault zone and tectonic implications for the mechanism of the May 12 Wenchuan earthquake, 2008, Sichuan, China[J]. Chinese Journal of Geophysics, 2008, 51(4): 1066-1073 (in Chinese with English abstract).
Zhang Peizhen, Wen Xueze, Xu Xiwei, Gan Weijun, Wang Min, Shen Zhengkang, Wang Qingliang, Huang Yuan, Zheng Yong, Li Xiaojun, Zhang Zhuqi, Ma Shengli, Ran Yongkang, Liu Qiyuan, Ding Zhifeng, Wu JianpingZhang Peizhen, Wen Xueze, Xu Xiwei, Gan Weijun, Wang Min, Shen Zhengkang, Wang Qingliang, Huang Yuan, Zheng Yong, Li Xiaojun, Zhang Zhuqi, Ma Shengli, Ran Yongkang, Liu Qiyuan, Ding Zhifeng, Wu Jianping. Tectonic model of the great Wenchuan earthquake of May 12, 2008, Sichuan[J]. Chinese Science Bulletin, 2009, 54(7): 944-953 (in Chinese). DOI:10.1360/csb2009-54-7-944
Zhang Peizhen, Wen Xueze, Shen Zhengkang, Chen JiuhuiZhang Peizhen, Wen Xueze, Shen Zhengkang, Chen Jiuhui. Oblique, high-angle, listric-reverse faulting and associated development of strain: the Wenchuan earthquake of May 12, 2008, Sichuan, China[J]. Annual Review of Earth and Planetary Sciences, 2010, 38: 353-382. DOI:10.1146/annurev-earth-040809-152602
Zhao Jing, Jiang Zaisen, Wu Yanqiang, Liu Xiaoxia, Wei Wenxin, Li QiangZhao Jing, Jiang Zaisen, Wu Yanqiang, Liu Xiaoxia, Wei Wenxin, Li Qiang. Study on fault locking and fault slip deficit of the Longmenshan fault zone before the Wenchuan earthquake[J]. Chinese Journal of Geophysics, 2012, 55(9): 2963-2972 (in Chinese with English abstract).
Zhao Jing, Jiang Zaisen, Wu Yanqiang, Liu Xiaoxia, Wei Wenxin, Wang Yuebing, Li Qiang, Xu JingZhao Jing, Jiang Zaisen, Wu Yanqiang, Liu Xiaoxia, Wei Wenxin, Wang Yuebing, Li Qiang, Xu Jing. Analysis of reliability and stability of inversion result with negative dislocation model of Defnode[J]. Journal of Geodesy and Geodynamics, 2013, 33(1): 21-24 (in Chinese with English abstract).
Zhao Jing, Jiang Zaisen, Niu Anfu, Liu Jie, Wu Yanqiang, Wei Wenxin, Liu Xiaoxia, Yan WeiZhao Jing, Jiang Zaisen, Niu Anfu, Liu Jie, Wu Yanqiang, Wei Wenxin, Liu Xiaoxia, Yan Wei. Study on dynamic characteristics of fault locking and fault slip deficit in the eastern boundary of the Sichuan-Yunnan rhombic block[J]. Chinese Journal of Geophysics, 2015, 58(3): 872-885 (in Chinese with English abstract).
Zhao Jing, Ren Jinwei, Jiang Zaisen, Liu Xiaoxia, Niu Anfu, Yan Wei, Yue Chong, Yuan ZhengyiZhao Jing, Ren Jinwei, Jiang Zaisen, Liu Xiaoxia, Niu Anfu, Yan Wei, Yue Chong, Yuan Zhengyi. Fault locking and deformation characteristics in southwestern segment of the Longmenshan fault[J]. Journal of Seismological Research, 2018, 41(2): 216-225 (in Chinese with English abstract).