Article

Geophysics June 2011 Vol.56 No.16: 16911699

doi: 10.1007/s11434-011-4487-y

SPECIAL TOPICS:

Uppermost mantle structure of the North China Craton: Constraints

from interstation Pn travel time difference tomography
LI ZhiWei1,2, HAO TianYao1* & XU Ya1
1
Key Laboratory of Petroleum Resources Research, Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100029, China;
2
Key Laboratory of Dynamical Geodesy, Institute of Geodesy and Geophysics, Chinese Academy of Sciences, Wuhan 430077, China

Received October 21, 2010; accepted March 16, 2011

The uppermost mantle is the key area for exchange of heat flux and material convection between the crust and lithospheric mantle.
Spatial variations of lithospheric thinning and dynamic processes in the North China Craton could inevitably induce the velocity
heterogeneity in the uppermost mantle. In this study, we used Pn arrivals from permanent seismic stations in North China and
surrounding regions to construct a tomographic image of the North China Craton. The tomographic method with Pn travel time
difference data were used to study the velocity variations in the uppermost mantle. Pn velocities in the uppermost mantle varied
significantly in the Eastern, Central and Western blocks of the North China Craton. This suggests that the lithosphere beneath
different blocks of the North China Craton have experienced distinct tectonic evolutions and dynamic processes since the Paleo-
zoic. The current uppermost mantle has been imprinted by these tectonic and dynamic processes. Fast Pn velocities are prominent
beneath the Bohai Bay Basin in the Eastern Block of the North China Craton, suggesting residuals of the Archean lithospheric
mantle. Beneath the Tanlu Fault Zone and Bohai Sea, slow Pn velocities are present in the uppermost mantle, which can be at-
tributed to significant lithospheric thinning and asthenospheric upwelling. The newly formed lithospheric mantle beneath Yanshan
Mountain may be the dominant reason for the existence of slow Pn velocities in this region. Conversely, the ancient lower crust
and lithospheric mantle already have been delaminated. In the Central Block, significant slow Pn velocities are present in Tai-
hangshan Mountain, which also extends northward to the Yinchuan-Hetao Rift on the northern margin of the Ordos Block and
Yinshan Orogen. This characteristic probably is a result of hot asthenospheric upwelling along the active tectonic boundary on the
margin of the Western Block. The protracted thermal erosion and underplating of hot asthenospheric upwelling may induce litho-
spheric thinning and significant slow velocities in the uppermost mantle. Fast velocities beneath the Western Block suggest that
the thick, cold and refractory Archean lithospheric keel of craton still is retained without apparent destruction.

North China Craton, Lithospheric thinning, Uppermost mantle, Pn travel time difference, Cratonic destruction

Citation: Li Z W, Hao T Y, Xu Y. Uppermost mantle structure of the North China Craton: Constraints from interstation Pn travel time difference tomography.
Chinese Sci Bull, 2011, 56: 16911699, doi: 10.1007/s11434-011-4487-y

A craton consists of ancient continental crust and subconti- (NCC) was formed in the Early Precambrian, and has lost a
nental lithospheric mantle (SCLM). Due to its relatively significant proportion of its lithospheric keel (more than 100
small density, high melting point and large thickness (~200 km thick) since the Paleozoic. The thick lithospheric mantle
km), the cratonic lithosphere can be preserved on the sur- has been replaced by depleted asthenosphere or oceanic
face of the Earth and become the most stable tectonic unit mantle, and the properties of the lithospheric mantle also
[1]. Although most Precambrian cratons have maintained have been essentially modified. This indicates lithospheric
their stability since their formation in the Paleoproterozoic, modification and even destruction of the Eastern Block of
some of them have experienced significant destruction. For the NCC (Figure 1) [1–6]. Although lithospheric thinning
example, the Eastern Block of the North China Craton and modification of lithospheric properties and thermal state
have been characterized by many geochemical and geo-
*Corresponding author (email: tyhao@mail.iggcas.ac.cn) physical studies, the dynamic mechanism of cratonic

© The Author(s) 2011. This article is published with open access at Springerlink.com csb.scichina.com www.springer.com/scp
1692 Li Z W, et al. Chinese Sci Bull June (2011) Vol.56 No.16

in the Central Block. This indicates significant lithospheric

thinning and destruction. Other than the thickness variation
of the lithosphere, distinct discrepancies of velocity struc-
ture also are present by the seismic travel time tomography
in the crust and upper mantle [15,16]. Meanwhile, signifi-
cant spatial variations of lithospheric thinning and destruc-
tion of the NCC have been revealed by the abovementioned
results, and even the Central-Western blocks shows some
signs of lithospheric thinning.
The uppermost mantle is the key area for exchange of
heat flow and material convection between the crust and
lithospheric mantle. Spatial variations of lithospheric thin-
ning and dynamic processes in the NCC could inevitably
induce velocity heterogeneity in the uppermost mantle.
Thus, it is possible to investigate spatial variation and dy-
Figure 1 Simplified tectonic map of the NCC (after [4]). ENCC, Eastern
namic mechanisms of cratonic destruction using the veloci-
Block of the NCC; CNCC, Central Block of the NCC; WNCC, Western
Block of the NCC; CAOB, Central Asian Orogen Belt; LU, Luxi Uplift; ty structure of the uppermost mantle. Pn waves dive into the
YinM, Yinshan Mountain; YM, Yanshan Mountain; Y-H Rift, Yinchuan- mantle from the base of the crust and travel along the up-
Hetao Rift; S-S Rift, Shaanxi-Shanxi Rift; BBB, Bohai Bay Basin. permost mantle. Temperature, pressure, composition and
anisotropy vary the velocity of the uppermost mantle and
destruction still is widely disputed. Thermal-chemical ero- affect the Pn travel time. Thus, Pn travel times provide an
sion and delamination are two contrasting hypotheses for important avenue to study dynamic processes of continental
NCC destruction [1–3,6,7]. The delamination hypothesis deformation and differentiation [17–20]. Given the uneven
states that the lower crust and upper mantle were laterally distribution of earthquakes and stations in the NCC, rela-
foundered into the asthenosphere due to gravity instability. tively large earthquake mislocations may contaminate the
The hot asthenospheric upwelling replaced the former lower velocity imaging by absolute travel time data from sparse
crust and upper mantle [8–11]. In contrast, the thermal- seismic arrays. In this study, we used relative travel time
chemical erosion hypothesis proposes that small amounts of data (e.g., interstation Pn travel time difference data) to
SCLM were gradually digested by asthenospheric upwelling, construct a tomographic image of the uppermost mantle in
and induced final destruction of the lithosphere [3,12]. the NCC. Our tomographic results provide further con-
While the delamination and thermal-chemical erosion hy- straints on deep tectonics, spatial variations and dynamic
potheses differ, both include alterations of thickness, physi- mechanisms with respect to destruction of the NCC.
cal and chemical properties, and thermal states. According
to the delamination hypothesis, the ancient lower crust and
lithospheric mantle sank into the asthenosphere and were
1 Methods and data
replaced by the newly formed lithospheric mantle. Con-
1.1 Tomographic methods with interstation Pn travel
versely, the thermal-chemical erosion hypothesis states that
time difference data
mineral composition and rheology were modified only at
the lower part of lithospheric mantle, and the current pre- Standard Pn travel time tomography utilizes the absolute Pn
served lithospheric mantle still remains ancient [1]. travel time data to inverse the velocity structure of the up-
Spatial variation of lithospheric thickness shows signifi- permost mantle. Event and station delays accommodate
cant discrepancy in different blocks of the NCC. Significant variations of crustal velocity and thickness, and earthquake
lithospheric thinning took place in the Eastern Block of the mislocation [17,18,20,21]. Assuming an isotropic velocity
NCC, and the current lithospheric thickness varies from model in the uppermost mantle, Pn travel time residuals can
60–100 km in this region. The Central Block of the NCC be approximately described by the following equation [18]:
also shows lithospheric thinning to some extent. Although N

the entire area of the Western Block of the NCC still re- tij  tijo  tijc  tist  t eqj   dijk sk , (1)
k 1
mains stable, thinned lithosphere also is found in some
marginal areas. Seismological results from receiver function where tij is the Pn travel time residual. tijo and tijc are
and surface wave tomography indicate that lithospheric the observed and calculated Pn travel times from the jth
thickness in the Ordos Block can reach 160–200 km [5,13, earthquake to the ith station. tist is the difference between
14]. This suggests that the block is stable. However, a thin-
ner lithosphere (~100 km) has been revealed in the Yin- observed and calculated travel times traveling from the
chuan-Hetao Rift on the northern margin of Ordos Block, Moho discontinuity to the ith station, which relates to crus-
and in the Shaanxi-Shanxi Rift and Taihangshan Mountain tal velocity and thickness. t eqj is the difference between
 Li Z W, et al. Chinese Sci Bull June (2011) Vol.56 No.16 1693

observed and calculated travel time from the jth event to the ence. Thus, this approach can be implemented in areas with
Moho discontinuity, which also relates to crustal velocity sparse seismic arrays and large earthquake mislocations,
and thickness and focal depth. dijk is the ray path length in such as the Western Block of the NCC and the Bohai Sea.
the kth cell, and sk is the slowness perturbation of this cell.
Due to the difficulty of relocating earthquakes with only 1.2 Data
Pn travel time data, event and station delays are needed to
minimize the bias of earthquake mislocation and heteroge- The data used in this study were obtained from the Chinese
neity of crustal velocity. As shown in the appendix, when Seismic Network and the International Seismological Cen-
one earthquake approximately lies on the great circle pass- tre (ISC) (1980–2008). The focal depths of selected events
ing through two stations, but not between them, Pn wave were shallower than 30 km, and the α angle between the
travel times along the overlapped ray paths are very close (< back azimuths from the event to the two stations was less
~0.1 s). Thus, interstation Pn travel time differences cannot than 6° (Figure S2). The distance between two stations for
easily be contaminated by earthquake mislocations and the interstation Pn travel time difference was greater than 20
complex crustal velocity structure near the source. The ve- km to ensure enough sampling of the uppermost mantle. For
locity structure also can be largely improved after the inver- data selection, only travel time difference data with residu-
sion [22–24]. Moreover, the undulated Moho discontinuity als less than 4.0 s and 10% of total travel time difference
model also can be introduced into the inversion to simulta- were selected. Finally, a set of 9815 interstation Pn travel
neously improve the tomographic image. time differences were selected for the inversion. The events,
For one event, the interstation Pn travel time difference stations and ray paths are shown in Figure 2. The data cov-
recorded by station i and j can be presented as erage is exceptionally good in most areas of the NCC, but is
poor in the marginal areas of the NCC. Due to the ad-
 Ni

tij  ti  t j    dik sik  tieq  tista  vantage of the interstation Pn travel time difference tomo-
k 1  graphic method, events located in the Bohai Sea and Yellow
Sea also could be involved in the inversion, which is usually
 Nj

   d jk s jk  t eq
j  t j 
sta not possible for standard Pn tomography due to its large
 k 1  earthquake mislocation problem.
 
 Ni Nj
 The averaged Pn velocity (~8.0 km/s) of the uppermost

 k 1 
  dik sik  d jk s jk 

mantle was determined by linear fitting with selected inter-
 k 1  station Pn travel time difference data, and it was consistent
 (tieq  t eqj )  (tista  t sta with previous studies [18,19,21]. Figure 3(a) shows the re-
j ), (2)
duced travel time versus epicenter distance for Pn arrivals.
where ti and tj are the Pn travel times at station i and j. tij is Figure 3(b) shows the interstation Pn wave travel time dif-
the interstation travel time difference. dik and djk are the kth ference versus epicenter distance difference from one
ray path lengths traveling through the uppermost mantle, Ni earthquake to two stations. The linear trend was clear and
and Nj segments in total. Sik and Sjk are the corresponding Pn the slope equaled the averaged Pn velocity of the uppermost
wave slowness at the kth ray path segment. tieq and
t eq
j are the travel times from the event to the Moho dis-
continuity. tista and t sta
j are the travel times from the
Moho discontinuity to the stations. In equation (2), the
origin time of event in tieq and t eq
j has been elimi-
nated, and only the effect of event mislocation (especially
the error of focal depth) has been retained. Based on the
analyses in Appendix, tieq and t eqj in equation (2) are
almost the same when the earthquake approximately lies on
the great circle passing through the two stations. Hence, the
interstation Pn travel time difference can be expressed as:

 Ni Nj

j )    d ik sik   d jk s jk  , (3)
tij  ti  t j  (tista  t sta
 k 1  Figure 2 Map of interstation Pn travel time difference data used in this
 k 1  study. Black triangles indicate the seismic stations, black crosses are the
earthquakes, and gray lines show the ray paths approximately sampling the
where the effect of the event has been eliminated, and the uppermost mantle between two stations, which are implemented into the
event mislocation has no effect on the Pn travel time differ- tomographic inversion.
1694 Li Z W, et al. Chinese Sci Bull June (2011) Vol.56 No.16

igppweb.ucsd.edu/~gabi/crust2.html) was taken into account

to calculate the Pn travel time. Trials suggest that the veloc-
ity images remain closely similar even though the crustal
velocity and Moho depth varied from 6.0–6.4 km/s and
32–36 km, respectively. This result verified that the inter-
station Pn travel time difference was not sensitive to the
initial crustal model. In standard Pn tomographic method,
station delay could reflect the influence of crustal velocity
and thickness beneath the station. Because the undulated
Moho discontinuity model of CRUST2.0 was used for Pn
travel time calculations, most of station delays were less
than 0.2 s, and only some of them were relatively large
(~0.6 s) at Taihangshan Mountain, Shaanxi-Shanxi Rift and
the marginal regions with strong crustal velocity and thick-
ness variations (Figure S6). Thus, it is difficult to analyze
the trend of crustal thickness with station delays [18]. After
the inversion, residuals of interstation Pn wave travel time
differences clearly were minimized with all back azimuths
Figure 3 (a) Plots of reduced travel times versus epicenter distance for (Figure 5) and lengths of Pn ray paths traveling in the up-
Pn arrivals involved in the inversion (reduced velocity is 8.0 km/s). (b)
Plots of interstation Pn wave travel time differences versus epicenter dis-
permost mantle (Figure 6). A posteriori residual standard
tance difference from one earthquake to two stations (dashed line repre- deviation was 0.54 s, a 57% reduction from that of the
sents the synthetic travel time with an averaged velocity of 8.0 km/s in the starting model, which suggests that the preferred model can
uppermost mantle). fit the observed data very well.
We performed the checkerboard test to estimate resolu-
mantle (~8.0 km/s). Most of the back azimuth differences of tion of our tomographic result. Figure 7 shows the recov-
interstation Pn travel time difference data were less than 3° ered images with 1.5°×1.2° and 2°×1.6° grid spacing mod-
(Figure 4), which satisfied the hypothesis of interstation Pn els with a perturbation of 3%. Results indicate that for much
travel time difference tomographic method (e.g. the Pn of the NCC the velocity anomalies were well resolved. In
wave travel times on the overlapped ray paths are approxi- the northern part of the NCC and sea areas, the resolution
mately the same). was relatively low because of sparse ray path coverage. Our
tomographic results were largely consistent with previous
tomographic results in much of the NCC [15,18,19,21,
2 Results

The cell size used to solve the travel time equations was
0.5°×0.4°. Average crustal P wave velocity of 6.2 km/s and
crustal thickness of 34 km were adapted in the inversion
after several trials. The average Pn wave velocity of 8.0
km/s from linear fitting also was used in the inversion. The
undulated Moho discontinuity model of CRUST2.0 (http://

Figure 5 Plots of residuals of interstation Pn wave travel time differences

Figure 4 Distribution of back azimuth differences of interstation Pn versus back azimuths of ray path before (upper) and after (below) the
travel time difference data. tomographic inversion.
 Li Z W, et al. Chinese Sci Bull June (2011) Vol.56 No.16 1695

feature in the Shaanxi-Shanxi Rift and Taihangshan Moun-

tain. The slow velocities also extended northward to the
Yinchuan-Hetao Rift and Yinshan Mountain to the north of
the Western Block. In contrast, fast velocities reached 8.1
km/s in the southern part of the Central Block, which dif-
fered greatly from slow velocities in the Qinling-Dabie
Orogen. In the Western Block of the NCC, significant fast
velocities were present in much of the Ordos Block.

3 Interpretation and discussion

Previous results of petrology and lithospheric thickness

provide evidences for spatial variations of lithospheric thin-
ning, which imply different mechanisms of cratonic de-
struction in different areas of the NCC. Our tomographic
images of the uppermost mantle also clearly indicate that
significant discrepancies exist in the NCC. The velocity
structure near the crust-mantle boundary may provide an-
Figure 6 Plots of residuals of interstation Pn wave travel time difference other important clue to investigate cratonic destruction and
versus the lengths of Pn wave ray paths traveling in the uppermost mantle its dynamic processes in the NCC.
before (upper) and after (lower) the tomographic inversion.

3.1 Eastern Block of the NCC

25–27], but more details of the uppermost mantle are shown
in our results, which could provide further information to Significant fast velocities were found beneath the Bohai
analyze spatial variations of destruction of the NCC. Bay Basin. Similar characteristics also have been found in
Figure 8 shows a Pn velocity image of the uppermost previous tomographic results of the lithospheric mantle [15,
mantle in the NCC and surrounding regions. In the Eastern 18,19,21,27–29]. Local tomographic images from P and
Block, fast velocities of the uppermost mantle (~8.2 km/s) PmP travel time data also indicate obvious fast velocities at
in the Bohai Bay Basin were pronounced. Conversely, the 42 km depth [30]. Since 42 km is already deep in the upper-
Luxi Uplift and its southern areas (the Tanlu Fault Zone, the most mantle in North China, the tomographic images at this
Jiaodong peninsula and the Yanshan Mountain) all were depth represent the velocity structure of the uppermost
characterized by slow velocities (7.8–7.9 km/s). In the Cen- mantle, which is consistent with our results from Pn travel
tral Block, slow velocities (~7.8 km/s) were a dominant time difference data. Moreover, fast S wave velocity struc-

Figure 7 Checkerboard resolution tests with different anomaly sizes. (a) 1.5°×1.2°; (b) 2°×1.6°.
1696 Li Z W, et al. Chinese Sci Bull June (2011) Vol.56 No.16

and inferred that the dominant mechanism of cratonic de-

struction in the Luxi Uplift is lithospheric delamination. If
this inference is correct, the significantly slow velocities in
the uppermost mantle could be attributed to effects of high
temperatures due to hot asthenospheric upwelling after lith-
ospheric thinning.
Yanshan Mountain is also characterized by apparent slow
velocities in the uppermost mantle. Delamination was sup-
ported as the dominant mechanism of cratonic destruction in
this area [9,34], and may have induced the SCLM sinking
into the asthenosphere. Thus, the newly formed and thinned
lithosphere after delamination of ancient lithosphere could
be the primary reason for slow velocities in the uppermost
mantle at Yanshan Mountain.

3.2 Central Block of the NCC

As part of the Daxinganling-Taihangshan-Wuyishan gravity
Figure 8 Pn wave velocities of the uppermost mantle in the NCC. lineament, crustal and lithospheric thicknesses intensively
vary beneath the eastern and western sides of the Central
Block of the NCC. The gravity lineament also becomes the
tures also have been found in the surface wave tomography suture zone, which separates Eastern and Western blocks [1,
beneath Bohai Bay Basin [14]. Receiver function migrations 35]. The uppermost mantle beneath the Central Block is
depict the lithospheric thickness of the Eastern Block, and dominated by significantly slow velocities. The slow veloc-
indicate that the thinnest lithosphere (60–80 km) was located ities were closer to eastern Taihangshan Mountain and ex-
in the Bohai Sea, the Luxi Uplift and the Tanlu Fault Zone tended to the northern margin of the Ordos Block, which is
[5]. In contrast, in the Bohai Bay Basin, the lithosphere was largely consistent with the distribution of late Mesozoic and
to 80–110 km in thickness. Huang et al. [14] inferred that the Cenozoic basalts and the Shaanxi-Shanxi and Yinchuan-
lower crust and lithospheric mantle to as deep as 130 km in Hetao rifts [1,36]. We infer that the slow velocities here
the North China Basin already have been modified. However, may be attributed to magmatic activity in the upper mantle.
from surface wave tomography, some residuals of fast veloci- In previous travel time tomography studies, slow velocity
ties still remain in the uppermost mantle beneath Bohai Bay zones to 500 km depth, and extending in a N-S trend, also
Basin, which agrees with our results. Hence, we infer that the have been found at the base of lithospheric mantle in the
fast velocity anomaly in the uppermost mantle suggests re- Central Block [15,27]. The characteristics of slow velocity
siduals of Archean lithospheric mantle. were clearer in the S waves rather than P waves [27]. Since
As the region with the thinnest lithospheric mantle S wave velocities are more sensitive to variation of temper-
(60–70 km), the Tanlu Fault Zone has slow velocity layers ature compared with P wave velocities [37], the slow velocity
in the crust [5,7,31–33]. Previous studies show that the anomaly could be primarily induced by high temperature.
Tanlu Fault Zone has cut through the crust and reaches to Zhao et al. [27] indicated that the Central Block may have
the lithospheric mantle. Hence, it may have become the behaved as a sublithospheric corridor for hot mantle materi-
al, which may have induced reactivation of the Archean
corridor of asthenospheric upwelling during a period of
NCC. If this inference is correct, the significantly slow ve-
continental extension of the eastern Block and lithospheric
locities beneath Taihangshan Mountain and the Yinchuan-
thinning in the Mesozoic and Cenozoic [31]. This also may
Hetao Rift in our tomographic image suggest that the influ-
be the reason for the thickened crust-mantle transition zone
ence of the asthenosphere upwelling has already reached the
and significant low velocity anomaly in the uppermost man- uppermost mantle and crust-mantle boundary. Our results
tle beneath the Tanlu Fault Zone. Cratonic destruction in also provide possible evidence that the dominant cause of
this area may be largely attributed to protracted thermal lithospheric destruction may be protracted thermal erosion
erosion and underplating [7,12,32]. This also suggests that and underplating. Zheng et al. [34] found thickened crust
the lithospheric weak zone plays an important role in cra- and a thick crust-mantle transition zone at Taihangshan
tonic destruction of the NCC. Mountain by receiver functions, and inferred that the lead-
The Luxi Uplift has significantly slow velocities in the ing cause of lithospheric destruction also could be protract-
uppermost mantle. According to previous studies, the crus- ed thermal erosion and underplating, which is similar with
tal and lithospheric thicknesses vary from 32–36 km and the Tanlu Fault Zone. In addition, results from receiver
60–70 km, respectively [32]. Gao et al. [10] found evidence function and surface wave tomography studies both indicate
for lithospheric delamination in Feixian of the Luxi Uplift, that the lithosphere has been thinned to 100 km in the
 Li Z W, et al. Chinese Sci Bull June (2011) Vol.56 No.16 1697

Yichuan-Hetao and Shaanxi-Shanxi rifts [13,14], and sug- and thermal state of the lithosphere [1]. The spatial varia-
gest that lithospheric thinning occurred in these areas. tions of lithospheric structure also were clearly indicated by
our uppermost Pn velocity image, and may have been asso-
3.3 Western Block of the NCC ciated with different destruction mechanisms and dynamic
Due to sparse seismic stations and earthquakes in the West- processes in different areas. Clear discrepancies of litho-
ern Block of the NCC, sparse ray path coverage reduces the spheric destruction exist in the Eastern, Central and Western
resolution in this area. Thus, the block boundaries were not blocks of the NCC. The lithosphere of the Eastern Block is
clearly delineated in our tomographic image. However, tec- in destruction mode and is thinned overall, while the Western
tonic features at a regional scale generally were reflected in Block basically remains stable. As for the Central Block,
the velocity image of the uppermost mantle. Obvious fast the lithospheric thickness increases from east to west, sug-
velocities are found in most areas of the Western Block, gesting cratonic destruction is exacerbated in this direction
which is consistent with previous Pn tomography at larger [7]. This also is consistent with subduction and dehydration
scales [18,19]. Seismic travel time tomography also indi- of the stagnant slab of the Pacific Plate, which could result
cated that the fast velocity anomaly beneath the Western in the increase of melting and fluid in the mantle, and in-
Block extends to 200 km depth or deeper [15,16,27,28]. duce upwelling of hot mantle materials [7,39–41]. As the
Given the thick lithosphere (160–200 km) [13,14], relative- stagnant slab of the Pacific Plate only reached Taihangshan
ly low heat flow (~40 mW/m2) [38], absence of volcanoes Mountain to the west [28], it is reasonable that the entire
or strong earthquake activity, we infer that the lithosphere lithospheric destruction and thinning only took place to the
of the Ordos Block remains a stable Archean craton. Con- east of Taihangshan Mountain. In addition, subduction of
versely, marginal areas of the Ordos Block become weak the Pacific Plate promoted the mantle convection and
zones for lithospheric thinning. Based on results from re- upwelling of hot mantle in the Central and Western blocks
ceiver function and surface wave tomography studies [13, [7], which thinned the weak lithosphere on the margin of
14], the lithosphere has been thinned to 100 km in the Cen- the Ordos Block, and also decreased the velocity in the up-
tral Block to the east of the Ordos Block, and strong thin- permost mantle. Benefiting from the obstruction of the thick
ning to 80 km has occurred in the Yinchuan-Hetao Rift and Archean lithosphere in the Ordos Block, the Western Block
Yinshan Mountain to the north of the Ordos Block. As the has not been destroyed in any obvious manner. However, if
suture of the Yangtze and the North China blocks, the lith- the destruction had happened to a certain extent in the
osphere in the Qinling-Dabie Orogen also was thinned to Western Block, we infer that the protracted thermal erosion
100–130 km to the south of the Ordos Block. In our tomo- and underplating should be the dominant mechanism.
graphic results, the above mentioned areas for lithospheric
thinning all showed significant slow velocities, suggesting
considerable influence from lithospheric thinning to the 4 Conclusions
uppermost mantle. Based on the locations of slow velocities,
we infer that the asthenospheric upwelling may have been The uppermost mantle structure of the NCC was construct-
obstructed by the ancient and stable lithosphere of the Or- ed by using interstation Pn travel time difference data. Pn
dos Block. As a result, upwelling acted on a pre-existing velocities in the uppermost mantle vary significantly in dif-
weak zone on the margin of the Ordos Block and formed ferent blocks, suggesting the lithosphere of the NCC has
the upwelling corridor of hot sublithospheric materials. The experienced distinct tectonic evolution and dynamic pro-
upwelling corridor also exacerbated lithospheric destruction cesses since the Paleozoic. The current uppermost mantle
and thinning. The Western Block may not have been obvi- has been imprinted by these tectonic and dynamic processes.
ously affected by Phanerozoic lithosphere destruction en- The fast velocity anomaly in the uppermost mantle of the
tirely, and the cratonic keel of the Western Block apparently Bohai Bay Basin suggests residuals of the Archean litho-
remains stable. However, if the lithospheric thickness of spheric mantle. The slow velocities beneath the Tanlu Fault
160 km in the Western Block indeed has been thinned to Zone and Bohai Sea area can be largely attributed to inten-
some extent, as shown in the surface wave tomography [14], sive lithospheric thinning and hot asthenospheric upwelling.
then the mechanism for lithospheric destruction here should The newly formed and thinned lithosphere after delamina-
be protracted thermal erosion and underplating, which was tion could be the reason for the slow velocity in the upper-
also the case in the Central Block. most mantle at Yanshan Mountain. The protracted thermal
erosion and underplating are likely responsible for litho-
3.4 Spatial variation of cratonic destruction and the spheric thinning and slow velocities in the uppermost man-
discrepancy of dynamic processes tle in the Central Block. The lithosphere of the Western
Block remains relatively stable and shows characteristics of
Cratonic destruction and lithospheric thinning not only al- the Archean craton, which is consistent with the fast veloci-
tered lithosphere thickness, but also affected the petrology ties in the uppermost mantle.
1698 Li Z W, et al. Chinese Sci Bull June (2011) Vol.56 No.16

We are grateful to the Academician Liu G D for the direction. We also phys Res, 2007, 112: B05312
thank Dr. Li J and Dr. Zhao L F for their constructive comments and help- 20 Li Z W, Roecker S, Li Z H, et al. Tomographic image of the crust and
ful modifications of the manuscript. We thank two anonymous reviewers upper mantle beneath the western Tien Shan from the MANAS
for their insightful reviews and thoughtful comments, which improved the broadband deployment: Possible evidence for lithospheric delamina-
manuscript. This work was supported by the National Natural Science tion. Tectonophysics, 2009, 477: 49–57
Foundation of China (90814011), Open Fund of Key Laboratory of Geo- 21 Wang S Y, Xu Z H, Pei S P. Velocity structure of uppermost mantle
detection (China University of Geosciences, Beijing), Ministry of Educa- beneath North China from Pn tomography and geological structure (in
tion (GDL0905), Fund of China Geological Survey (GZH200900504) and Chinese). Sci China Ser D-Earth Sci, 2003, 33(Suppl): 91–98
the China Postdoctoral Science Foundation (20090460511). 22 Li J. Research of seismic tomography in Capital Circle Area from Pn
and PmP wave travel time inversions (in Chinese). Dissertation for the
Doctoral Degree. Beijing: Institute of Geophysics, Chinese Admin-
1 Wu F Y, Xu Y G, Gao S, et al. Lithospheric thinning and destruction istration of Earthquake, 2003. 1–110
of the North China Craton (in Chinese). Acta Petrol Sin, 2008, 24: 23 Phillips W S, Rowe C A, Steck L K. The use of interstation P wave ar-
1145–1174 rival time differences to account fore regional path variability. Ge-
2 Fan W M, Menzies M A. Destruction of aged lower lithosphere and ophys Res Lett, 2005, 32: L11301
accretion of asthenosphere mantle beneath eastern China. Geotecton 24 Seward A M, Henderson C M, Smith E G C. Models of the upper
Metallogen, 1992, 16: 171–180 mantle beneath the central North Island, New Zealand, from speeds
3 Menzies M A, Fan W, Zhang M. Palaeozoic and Cenozoic lithoprobes and anisotropy of subhorizontal P waves (Pn). J Geophys Res, 2009,
and the loss of > 120 km of Archaean lithosphere, Sino- Korean 114: B01301
Craton, China. In: Prichard H M, Alabaster T, Harris N B W, et al., 25 Liang C, Song X, Huang J. Tomographic inversion of Pn travel times
eds. Magmatic Processes and Plate Tectonics. Geol Soc Spec Pub, in China. J Geophys Res, 2004, 109: B11304
1993, 76: 71–78 26 Li C, van der Hilst R, Nafi T M. Constraining P-wave velocity varia-
4 Zhao G, Wilde A, Cawood A, et al. Archean blocks and their bounda- tions in the upper mantle beneath Southeast Asia. Phys Earth Planet
ries in the North China Craton: Lithological, geochemical, structural Inter, 2006, 154: 180–195
and P-T path constraints and tectonic evolution. Precambrian Res, 27 Zhao L, Allen R M, Zheng T, et al. Reactivation of an Archean craton:
2001, 107: 45–73 Constraints from P- and S-wave tomography in North China. Geophys
5 Chen L, Wang T, Zhao L, et al. Distinct lateral variation of lithospher- Res Lett, 2009, 36: L17306
ic thickness in the Northeastern North China Craton. Earth Planet Sci 28 Huang J, Zhao D. High-resolution mantle tomography of China and
Lett, 2008, 267: 56–68 surrounding regions. J Geophys Res, 2006, 111: B09305
6 Zheng Y F, Wu F Y. Growth and reworking of cratonic lithosphere. 29 Li Z W, Xu Y, Hao T Y, et al. Seismic tomography and velocity
Chinese Sci Bull, 2009, 54: 3347–3353 structure in the crust and upper mantle around Bohai Sea area (in Chi-
7 Zhu R X, Zheng T Y. Destruction geodynamics of the North China nese). Chinese J Geophys, 2006, 49: 797–804
Craton and its Paleoproterozoic plate tectonics. Chinese Sci Bull, 2009, 30 Lei J, Xie F, Lan C, et al. Seismic images under the Beijing region inferred
54: 3354–3366 from P and PmP data. Phys Earth Planet Inter, 2008, 168: 134–146
8 Gao S, Rudnick R, Carlson R W, et al. Re-Os evidence for replace- 31 Chen L, Zheng T, Xu W. A thinned lithospheric image of the Tanlu Fault
ment of ancient mantle lithosphere beneath the North China Craton. Zone, northeastern China, constructed from wave equation based receiv-
Earth Planet Sci Lett, 2002, 198: 307–322 er function migration. J Geophys Res, 2006, 111: B09312
9 Gao S, Rudnick R, Yuan H, et al. Recycling lower continental crust in 32 Zheng T, Zhao L, Zhu R. Insight into the geodynamics of cratonic re-
the North China Craton. Nature, 2004, 432: 892–897 activation from seismic analysis of the crust-mantle boundary. Geo-
10 Gao S, Rundnick R, Xu W, et al. Recycling deep cratonic lithosphere phys Res Lett, 2008, 35: L08303
and generation of intraplate magmatism in the North China Craton. 33 Xu Y G, Li H Y, Pang C J, et al. On the timing and duration of the de-
Earth Planet Sci Lett, 2008, 270: 41–53 struction of the North China Craton. Chinese Sci Bull, 2009, 54:
11 Gao S, Zhang J F, Xu W L, et al. Delamination and destruction of the 3379–3396
North China Craton. Chinese Sci Bull, 2009, 54: 3367–3378 34 Zheng T, Zhao L, Xu W W, et al. Insight into modification of North
12 Xu Y G. Thermo-tectonic destruction of the Archean lithospheric keel China Craton from seismological study in the Shandong Province.
beneath the Sino-Korean Craton in China: Evidence, timing and Geophys Res Lett, 2008, 35: L22305
mechanism. Phys Chem Earth, 2001, 26: 747–757 35 Hao T Y, Liu Y K, Duang C. The characteristic of geophysical field in
13 Chen L, Cheng C, Wei Z. Seismic evidence for significant lateral vari- the east China and adjacent regions and its tectonics significance (in
ations in lithospheric thickness beneath the central and western North Chinese). Chin J Geophys, 1997, 40: 677–690
China Craton. Earth Planet Sci Lett, 2009, 286: 171–183 36 Tang Y, Zhang H, Ying J. Asthenosphere-lithospheric mantle interac-
14 Huang Z X, Li H Y, Zheng Y J, et al. The lithosphere of North China tion in an extensional regime: implication from the geochemistry of
Craton from surface wave tomography. Earth Planet Sci Lett, 2009, Cenozoic basalts from Taihang mountains, North China Craton. Chem
288: 164–173 Geol, 2006, 233: 309–327
15 Tian Y, Zhao D, Sun R, et al. Seismic imaging of the crust and upper 37 Cammarano F, Goes S, Vacher P, et al. Inferring upper-mantle tem-
mantle beneath the North China Craton. Phys Earth Planet Inter, 2009, peratures from seismic velocities. Phys Earth Planet Inter, 2003, 138:
172: 169–182 197–222
16 Xu P F, Zhao D. upper-mantle velocity structure beneath the North 38 Hu S B, He L J, Wang J Y. Heat flow in the continental area of China:
China Craton: Implications for lithospheric thinning. Geophys J Int, A new data set. Earth Planet Sci Lett, 2000, 179: 407–419
2009, 170: 1095–1099 39 Zhao D. Global tomographic images of mantle plumes and subducting
17 Hearn T M. Anisotropic Pn tomography in the western United States. J slabs: Insight into deep Earth dynamics. Phys Earth Planet Inter, 2004,
Geophys Res, 1996, 101: 8403–8414 146: 3–34
18 Hearn T M, Wang S, Ni J F, et al. Uppermost mantle velocities beneath 40 Lei J, Zhao D. P-wave tomography and origin of the Changbai intraplate
China and surrounding regions. J Geophys Res, 2004, 109: B11301 volcano in northeast Asia. Tectonophysics, 2005, 397: 281–295
19 Pei S, Zhao J, Sun Y, et al. Upper mantle seismic velocities and ani- 41 Lei J, Zhao D. Global P-wave tomography: On the effect of various
sotropy in China determined through Pn and Sn tomography. J Geo- mantle and core phases. Phys Earth Planet Inter, 2006, 154: 44–69

Open Access This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction
in any medium, provided the original author(s) and source are credited.