SCIENCE CHINA Earth Sciences

© Science China Press and Springer-Verlag Berlin Heidelberg 2012 earth.scichina.com www.springerlink.com

*Corresponding author (email: wangqi@cug.edu.cn)

• NEWS & FOCUS • May 2012 Vol.55 No.5: 695–698

doi: 10.1007/s11430-012-4412-5

A precise velocity field of tectonic deformation in China as inferred from intensive GPS observations

LI Qiang1, YOU XinZhao1, YANG ShaoMin2, DU RuiLin2, QIAO XueJun2, ZOU Rong3 & WANG Qi3*

1 National Infrastructure of Earthquake Centre, Beijing 100036, China; 2 Institute of Seismology, China Earthquake Administration, Wuhan 430071, China;

3 Planetary Science Institute, China University of Geosciences, Wuhan 430074, China

Received February 3, 2012; accepted March 5, 2012

Citation: Li Q, You X Z, Yang S M, et al. A precise velocity field of tectonic deformation in China as inferred from intensive GPS observations. Sci China Earth Sci, 2012, 55: 695–698, doi: 10.1007/s11430-012-4412-5

Active tectonics, e.g., faulting, folding, and rifting are clearly manifested over the vast territory of China. The in-tensive crustal deformation with long-lived structures has given rise to numerous highly-elevated mountains in the western China with the Yangtze and Yellow rivers down- streaming to the east, as well as resulting in catastrophic earthquakes with huge fatalities in the historical times. Quantifying crustal deformation into its amount, distribu-tion, and timing is prerequisite for an understanding of the nature of geological evolution, climate-tectonics interaction, and ultimately the geodynamic aspects that control these processes. Moreover, monitoring active deformation with great precision is of fundamental importance in trying to forecast earthquake or assess seismic hazard for an earth-quake-prone country like China.

In the late 1980s, GPS geodesy was first used to measure crustal deformation [1]. About 300 GPS survey-mode sites were installed in a decade-long effort to monitor tectonic activities in Yunnan, Sichuan, Fujian, Xinjiang, Tibet, Qinghai, Gansu, Ningxia, and several provinces in North China [2–4]. In 1997–2001, a big scientific infrastructure was constructed for a nationwide geodetic network CMONOC , which refined a relatively poor description of ongoing deformation of China with a total of 1056 survey- mode sites and 27 continuously recording stations (cGPS). However, this network was still heterogeneous in site den-sity [5, 6]. A limited spatial coverage provided insufficient

information about the tectonic processes in critical tectonic zones. In 2007–2011, a continued construction of CMONOC was implemented with addition of 1000 sur-vey-mode sites and 233 cGPS stations. In this paper, we analyzed all observations acquired from the enhanced CMONOC to present the most comprehensive velocity field of tectonic deformation in China.

Our velocity field is based primarily on all GPS data ob-tained at 1056 sites from 1998 to 2007 and two campaigns for a total of 2056 sites respectively in 2009 and 2011. The fieldwork was arranged from March to August each year, with each site occupied for one session of four consecutive days using dual-frequency receivers and geodetic antennas. Before 2011 only 27 cGPS stations equipped with choke- ring antennas were used and 233 new cGPS stations were gradually put in operation by the time of the 2011 cam-paign.

GIPSY software was used to analyze GPS raw data col-lected before 2008 and four different software packages, i.e., GIPSY, PANDA, BERNSE, and GAMIT, were used for data analysis of the 2009 and 2011 campaigns in order to access the data-quality of 1000 new sites. With GIPSY and PANDA, a precise-point-positioning algorithm is adopted in which one 24-hour session of undifferenced carrier phase and pseudorange measurements from a site was processed to produce one daily coordinate estimate of that site. With BERNSE and GAMIT, carrier phases and pseudoranges from a group of sites surveyed concurrently on one day are combined by a conventional double-difference algorithm to

696 Li Q, et al. Sci China Earth Sci May (2012) Vol.55 No.5

yield one daily network solution of site positions. Daily estimates of site positions are all transformed onto the In-ternational Terrestrial Reference Frame (ITRF) 2005 [7]. The daily solutions produced by the four different software packages are averaged into single daily solution each day. Assuming simply a linear function of time for surface dis-placement at each site, the velocities are estimated from combining all daily solutions with a weighted-least-squares adjustment algorithm.

A preliminary adjustment of the daily solutions indicates that with respect to neighboring sites, velocities at 104 sites appear to be anomalous, of which anomalous velocities of 59 sites are attributed to possible monument instabilities, instrument malfunctions or observational blunders and those at the remaining 45 sites obviously to recent large earth-quakes. These sites are all excluded from the final velocity field calculation. In fact, seismic perturbations on long-term velocities of CMONOC sites are not restricted to those mentioned above. For instance, the 2011 Tokohu-Oki, Japan earthquake (Mw9.0) caused measureable far-field effects in northeastern and northern China, illustrated by GPS- inferred surface displacements and hundreds of CMONOC sites may have transient displacements of up to 35 mm re-sulted from this event [8].

In order to correct the effects, we isolated coseismic off-sets of the cGPS stations in this region and around by dif-ferencing their coordinates averaged over 3-days immedi-ately before and after the Tohoku-Oki earthquake, then with bi-cubic spline functions, and interpolated these observed offsets onto other survey-mode sites to within an uncertain-ty of 3 mm. The coseismic offsets were then removed from the post-earthquake daily solutions in the adjustment. The 2001 Kokoxili Mw7.8, 2004 Sumatra Mw9.2, 2008 Wen-chuan earthquake Mw7.9 and 2010 Yushu Mw6.9 earth-quakes also resulted in significant coseismic transients on a large spectrum of spatial scales [9–12]. Given a much sparser cGPS array by the times of these earthquakes, we adopted a different approach based on site coordinate time series to estimating coseismic offsets in the following ways. For near-field sites, their post-earthquake coordinate time series were removed because postseismic deformation would continue for decades long and might bias velocity estimation. We simply made use of their pre-earthquake coordinate time series to estimate the velocities. As a result, numerous sites with no or one-epoch observation prior to the earthquakes were excluded from the velocity estimation. For the far-field sites with coseismic offsets anticipated to be lager than 20 mm, we fitted a line plus offset respectively to the pre- and post-earthquake coordinate time series of an affected site to infer its offset with an uncertainty of 4–5 mm.

The final adjustment provides a data set of 1979 velocity vectors defined on the ITRF2005. After subtracting rigid rotations of the Eurasia Plate from these velocity vectors [7], the residual components represent a steady tectonic defor-

mation with amplitudes in range of 0–40 mm/yr relative to stable Eurasia Plate (Figure 1). In addition, we incorporated published GPS velocities obtained from the surrounding regions [13–17] to fully understand the relations between different geologic units. The new result with most of uncer-tainties less than 2 mm/yr has a much higher resolution than the previous velocity fields [4–6], therefore providing a more accurate and comprehensive description about the complex pattern of continental tectonics.

Much of the deformation in China seems to be attributa-ble to a single cause—the collision between India and Eura-sia. Active tectonics in China resemble its topography and geomorphology, which all appear to contrasting features between eastern and western partstwo equally-sized areas divided roughly by longitude 105–106E. In the east, the pattern of deformation is quite simple, characterized by a coherent movement directed southeasterly or east-south- easterly with decreasing magnitude from north to south. The overall motion in the northeastern China amounts to 2–7 mm/yr in rate, and roughly 4–9 mm/yr for most part of North China as well as 6–11 mm/yr for South China. Dif-ferential motion less than 2 mm/yr across two adjacent re-gions is evident from the densified velocity field, though no boundaries are defined unambiguously between them. Fur-thermore, the denser networks for the northeastern China and South China do not observe any abrupt velocity gradi-ent within their interiors either, consistent with a previous interpretation of little deformation throughout themselves [4, 5]. In North China, differential motions at level of a few mm/yr across several grabens along the southern and east-ern margins of the Ordos block are observed and the slip rate across the east-west trending Zhangjiakou-Bohai fault is constrained as low as 1–3 mm/yr [18].

In the west, there appears most of deformation associated with an abrupt rise of the Tibetan Plateau. Diverse types of ongoing tectonics have attested to this spectacular process with reverse faulting and crustal shortening normal to the margins as well as normal and strike-slip faulting responsi-ble for crustal extension within the high mountains [19]. The continental deformation in response to the India’s northward motion is not confined to the plateau, but extends farther north into the Tianshan, Altay, Mongolia and even Baikal, ~3000 km northeast of the Himalayas [13–15].

The GPS velocity field based on either the new or the old dataset in this region has shown [20–22] ongoing tectonics in Tibet that can be dominated by three kinds of defor-mation regimes. That is, the continuously-distributed con-traction across the plateau with gradually-decreasing veloc-ity components along the direction of the plate convergence; the approximately east-west extension manifested by veloc-ity vectors changes in orientation from north-northeast in the south to due east in the north; and the clockwise rotation of crustal material around the eastern Himalayan Syntaxis as illustrated by opposite orientation for the velocity vectors in Lhasa and Yunnan, all of which show a clockwise rota-

Li Q, et al. Sci China Earth Sci May (2012) Vol.55 No.5 697

Figure 1 The red arrows show the horizontal displacements of 1979 GPS sites with 67% confident ellipses based on the CMONOC campaigns in 1998–2011. The black triangles mark the new sites. A total of 788 pinks arrows represent the velocities of GPS sites outside China, publicly available from refs. [13–17].

tion of almost 180 from north-northeast to southwest. Furthermore, the new results have revealed new aspects

of the pattern of deformation. For example, one velocity profile running roughly along the Qinghai-Tibet highway shows GPS sites moving to northeast with a nearly-linear decrease in rate along the profile direction. A total of 36–38 mm/yr of the plate convergences is distributed uniformly along a section of 2000 km-long across the Himalaya, Tibet, and Mongolia. In contrast, another profile running roughly along longitude 80°E demonstrates a heterogeneous con-traction. GPS sites along the profile move to north or north-northwest at the rates that decline northward in a non-linear fashion. Approximately a half of the India’s 33–36 mm/yr northward motion is accommodated by the contraction across the northwestern Himalaya and southern margin of the Tibetan Plateau.

The densified GPS measurements in the Tarim and Dzungar Basins attest to no, if any, deformation. The basins separated by Tianshan rotate as rigid blocks clockwise to-wards northeast or north-northeast. The western Tianshan (west of longitude 80°E) takes up a convergent deformation of 16–18 mm/yr, almost another half of the India’s north-

ward motion so that a vast region in the Kazakh platform exhibits little deformation [15]. By contrast, the eastern Tianshan (east of 88°E) deforms at a convergent rate of 2–6 mm/yr. As a result, the region north of Tianshan including the Dzungar Basin and Gobi Altay moves northeasterly at an average rate of 7–9 mm/yr, which is absorbed in part by 4–6 mm/yr of contraction deformation in Altay and the western Mongolia. The remaining convergence deformation between India and Eurasia is partitioned finally by a modest eastward extrusion of Mongolia and its adjacent area to the north.

One of the goals of pursuing a high-precision and inten-sive mapping of present-day deformation is to accurately describe the kinematics of continental tectonics. Although, diverse sorts of data for tectonics studies of the Tibetan Plateau were growing steadily in the past two decades, an unambiguous kinematical model has not yet been completed. The debate persists on which of the two end-member mod-elsrigid block rotations as opposed to diffusive defor-mation of continuum materials is a better kinematics for the Tibetan Plateau, or if a combination of the two models is most appropriate [23]. The previous GPS velocity field

698 Li Q, et al. Sci China Earth Sci May (2012) Vol.55 No.5

failed to differentiate the two end-members in that the data with limited site density and velocity accuracy actually fit-ted both of them equally well [20, 24, 25]. It is anticipated that a refined velocity field with the help of new data com-ing from subsequent CMONOC campaigns would provide much better constraints on the kinematic models, thereby improving our understanding of the dynamics of continental tectonics.

We thank all participants involved in the CMONOC project. It is apprecia- ted that GNSS analysis center at Institute of Seismology provided GAMIT, GIPSY, and BERNESE solutions and Satellite Navigation and Positioning Center at Wuhan University provided PANDA solution for our study. We are grateful to Jeffrey Freymueller at Geophysical Institute, University of Alaska, Fairbanks for thoughtful reviews that improve this manuscript. This work was supported by the National Natural Science Foundation of China (Grant Nos. 40674054, 41074016) and the CMONOC Project.

1 Lai X, Wu X, Seeber G, et al. The first epoch western Yunnan GPS survey: Data reduction and analysis with Geonap software. Earthq Res China, 1992, 6: 68–90

2 Chen Z, Burchfiel B, Liu Y, et al. Global Positioning System meas-urements from eastern Tibet and their implications for India/Eurasia intercontinental deformation. J Geophys Res, 2000, 105: 16215– 16227

3 Shen Z K, Wang M, Li Y, et al. Crustal deformation along the Altyn Tagh fault system, western China, from GPS. J Geophys Res, 2001, 106: 30607–30622

4 Wang Q, Zhang P Z, Freymueller J, et al. Present-day crustal defor-mation in China constrained by Global Positioning System measure-ments. Science, 2001, 294: 574–577

5 Wang M, Shen Z K, Niu Z, et al. Contemporary crustal deformation of the Chinese continent and tectonic block model (in Chinese). Sci China Ser D-Earth Sci, 2003, 33(Suppl): 21–32

6 Niu Z, Wang M, Sun H, et al. Contemporary velocity field of crustal movement of Chinese mainland from Global Positioning System measurements. Chin Sci Bull, 2005, 50: 939–941

7 Altamimi Z, Collilieux X, Legrand J. ITRF2005: A new release of the International Terrestrial Reference Frame based on time series of stations and Earth Orientation Parameters. J Geophys Res, 2007, 112: B09401, doi: 10.1029/2007JB004949

8 Wang M, Li Q, Wang F, et al. Far-field coseismic displacements as-sociated with the 2011 Tohoku-oki earthquake in Japan observed by Global Positioning System. Chin Sci Bull, 2011, 56: 2419–2424

9 Lasserre C, Peltzer F, Crampé F, et al. Coseismic deformation of the

2001 Mw=7.8 Kokoxili earthquake in Tibet, measured by synthetic aperture radar interferometry. J Geophys Res, 2005, 110(B12), doi: 10.1029/2004JB003500

10 Qiao X, Yang S, Du R, et al. GPS-derived crustal deformation in southwestern China. Austrian J Earth Sci, 2006, 54: 521–529

11 Working Group of the Crustal Motion Observation Network of China Project. Coseismic displacement field of the 2008 Ms8.0 Wenchuan earthquake determined by GPS (in Chinese). Sci China Ser D-Earth Sci, 2008, 38: 1195–1206

12 Zha X, Dai Z, Ge L, et al. Fault geometry and slip distribution of the 2010 Yushu earthquakes inferred from InSAR measurement. Bull Seismo Soc Am, 2011, 101: 1951–1958

13 Calais E, Vergnolle M, Sankov V, et al. GPS measurements of crustal deformation in the Baikal-Mongolia area (1994–2002): Implications for current kinematics of Asia. J Geophys Res, 2003, 108(B10): 2501, doi: 10.1029/2002JB002373

14 Yang S, Li J, Wang Q. The deformation pattern and fault rate in the Tianshan Mountains inferred from GPS observations. Sci China Ser D-Earth Sci, 2008, 51: 1064–1080

15 Zubovich A, Wang X, Scherba Y, et al. GPS velocity field for the Tien Shan and surrounding regions. Tectonics, 2010, 29: TC6014, doi: 10.1029/2010TC002772

16 Banerjee P, Bürgmann R, Nagarajan B, et al. Intraplate deformation of the Indian subcontinent. Geophys Res Lett, 2008, 35: L18301, doi: 10.1029/2008GL035468

17 Maurin T, Masson F, Rangin C, et al. First global positioning system results in northern Myanmar: Constant and localized slip rate along the Sagaing fault. Geology, 2010, 38: 591–594

18 Jin S, Park P, Zhu W. Micro-plate tectonics and kinematics in North-east Asia inferred from a dense set of GPS observations. Earth Plant Sci Lett, 2007, 257: 486–496

19 Tapponnier P, Xu Z, Roger F, et al. Oblique stepwise rise and growth of the Tibet Plateau. Science, 2001, 294: 1671–1677

20 Chen Q, Freymueller J, Wang Q, et al. A deforming block model for the present-day tectonics of Tibet. J Geophys Res, 2004, 109: B01403, doi: 10.1029/2002JB002151

21 Zhang P Z, Shen Z K, Wang M, et al. Continuous deformation of the Tibetan Plateau from GPS data. Geology, 2004, 32: 809–812

22 Gan W, Zhang P, Shen Z K, et al. Present-day crustal motion within the Tibetan Plateau inferred from GPS measurements. J Geophys Res, 2007, 112: B08416, doi: 10.1029/2005JB004120

23 Thatcher W. Microplate versus continuum descriptions of active tec-tonic deformation. J Geophys Res, 1995, 100: 3885–3894

24 England P, Molnar P. The field of crustal velocity in Asia calculated from Quaternary rates of slip on faults. Geophys J Int, 1997, 130: 551–582

25 Meade B. Present-day kinematics at the India-Asia collision zone. Geology, 2007, 35: 81–84