Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
Geodesy and Geodynamics 2012,3 ( 1) :1 -7
http://www. jgg09. com
Doi:10.3724/SP.J.l246.2012.00001
Co-seismic deformation and gravity changes of the 2011
India-Nepal and Myanmar earthquakes
Liu Chengli 1 ' 2 , Zheng Y ong1 , Shan Bin 1' 2 and Xiong Xiong1
1 State Key Laboratory of Geodesy and Eanh ' s Dynamic, Instituu of Geodesy and Geophysics, ChWse Academy of Sciences, Wuhan 430077, China
2 Graduate University of Chinese Academy of Sciences, Beijing 100049 , China
Abstract: Co-seismic deformation and gravity field changes caused by the 2011 Mw6. 8 Myanmar and Mw6. 9
India-Nepal earthquakes are calculated with a finite-element model and an average-slip model, respectively,
based on the multi-layered elastic half-space dislocation theory. The calculated maximum horizontal displace-
ment of the Myanmar earthquake is 36 em, which is larger than the value of 9. 5 em for the India-Nepal earth-
quake. This difference is attributed to their different focal depths and our use of different models. Except cer-
tain differences in the near field, both models give similar deformation and gravity results for the Myanmar
event.
Key words: India-Nepal earthquake; Myanmar earthquake; average-slip model; finite-element rupture mod-
el; gravity
1 Introduction
Two strong earthquakes occurred near the southwestern
border of China in 2011 , and caused serious damages
in the neighboring Yunnan and Tibet regions, respec-
tively. One is an Mw6. 8 earthquake occurring in Moog
Hpayak, Myanmar on March 24. The epicenter is loca-
ted at 20. 698 °N, 99. 889 °E, and the focal depth is a-
bout 10 km. The focal mechanism shows a left-lateral
strike slip. Myanmar and the southern part of Yunnan
province belong to Yunnan-Myanmar seismotectonic
block[I,z], which is located to the south of the eastern
structure knot of Tibetan plateau , which includes
southern Yunnan, Myanmar, Vietnam and other re-
gions. Its western boundary is the " arc" subduction
Received :2011-11-27; Accepted :2011-12-{)7
Corresponding authar:Tel: +86-13971417171; E-mail : lcl8669@ 126. com
This work is supported by grant 201008007 from China Earthquake
Administmtion ,National Natural Science Foundation of China ( 40974034,
41174086)
zone, where the Indian plate subducts beneath the My-
anmar plate, resulting in some right-lateral strike-slip
faults and large earthquakes ( Fig. 1 ( b) ) . Its eastern
and southern boundaries are, respectively, Jinshajiang-
Honghe fault, and the Sumatra-Andaman sea. It is di-
vided into western and southern tectonic blocks by Nu-
jiang-Lancangjiang fault.
The other earthquake has a magnitude of Mw6. 9 ,
occurring on September 18 in India-Nepal; it is located
at 27. 43°N, 88. 33°E, and has a focal depth of about
19. 7 km (USGS) and a focal mechanism showing left-
lateral strike slip with an oblique thrust component.
The location is near the boundary between the India
and Eurasia plates, in the mountainous region of north-
eastern India near the Nepalese boarder, where many
earthquakes had occurred ( Fig. 1 ( a) ) . According to
USGS , the earthquake source is located within either
the upper Eurasian plate or the underlying India plate,
rather than at the plate boundary.
In this paper , we present an analysis of the co-seis-
mic surface deformation and gravity changes caused
2 Geodesy and Geodynamics Vol. 3
30•N
22•N
28•N
20•N
18•N
90•E 92 OE 96 OE 98 OE 100 OE 102 OE 104 OE
(a) Location of the India-Nepal Mw6.9 earthquake (b) Location of the Myanmar Mw6.8 earthquake
Figure 1 Location of the India-Nepal Mw6. 9 earthquake and the Myanmar Mw6. 8 earthquake
( Yellow stars indicate epicenters ; white circles , historical earthquakes ( Mw > 4. 5) ; green circles , main cities ; yellow circles , aftershocks)
by these two events. The results may provide some use-
ful information about the boundary between the India
and Eurasia plates, the eastern structural belt of Tibet-
an plateau, and the long-term tectonic motion and in-
ter-seismic deformation field [5 -9l.
2 Theory and models
For the Myanmar earthquake, we used a finite-element
method to invert its co-seismic rupture parameters.
However, because of the uncertainties of the source in-
formation about the India-Nepal earthquake, it is diffi-
cult to use this approach to obtain a co-seismic rupture
model. Thus, we used an average-slip model based on
the empirical formula provided by Donald L, et al[ 10J.
We also used the PSGRN/PSCMP[llJ software and a
multi-layered crust model in our calculation. Wang[ 12l
proposed an orthogonal normalization method to calcu-
late the Green function of seismic stress field. Based
on this method and a viscoelastic multi-layered model,
he established a co- and post-seismic deformation mod-
el and produced a corresponding numerical meth-
od[13'14l. By using this method we calculated the theo-
retical deformation and gravity changes caused by these
two earthquakes.
2. 1 Average-slip model
We used some teleseismic data and the inversion proce-
dure developed by Ji , et al [ 15] to generate a rupture
model for the Myanmar earthquake, and an empirical
equation to generate an average-slip model for the Indi-
a-Nepal earthquake, with the following regressions be-
tween average displacement (AD ) , subsurface rupture
length ( RLD ) , rupture width ( RW) , and moment
magnitude (M) [1o]
Log (AD) = a 1 + b1M
Log (RLD) = a2 + b2M Log (RW) = a 3 + b3M
the empirical coefficients of a 1 , a2 , a 3 , b1 , b2 , b3 were
found by Donald L, et al [ JOJ to be -4. 80, - 2. 44,
- 1. 01 , 0. 69, 0. 59 and 0. 32, respectively, for a
strike-slip earthquakes. By applying this average-slip
model to the India-Nepal earthquake, we obtained a
length along the strike direction of 55 km and a width
along down-dip direction of 16 km. The distribution of
aftershocks shows a dip of 74 degrees and a strike of
217 degrees. Global CMT (http://www. globalcmt.
org/CMTsearch. html) gave a rake of -18 degrees;
USGS, a focal depth of 19. 7 km.
2. 2 Finite-element rupture model
Using broadband seismograms recorded by Global Seis-
mic Network ( GSN) (Fig. 2( b)), and based on the fi-
nite-element inversion method [ 15 '16] , we derived a rup-
ture model for the Myanmar earthquake ( Fig. 2 ( a) ) .
No.1
Liu Chengli,et al. Co-seismic deformation and gravity changes of the 2011
India-Nepal and Myanmar earthquakes 3
4
l 8 l 12 Cl
16
a- 4
"' 2 -18 -12 16 0 6
~~~~~~ .... ~~~em 0 50 100 150 200 250 300 350 400 450 500
(a) Co-seismic slip model (White arrows
represent amplitude and direction of the slips)
12
(B) Distribution of used seismic
Stations( triangles; The red star indicates the hypo/epicenter)
Figure 2 Co-seismic slip model of the Myanmar earthquake
Mter comparing with the distribution of aftershocks, we
determined a fault plan with the strike of 250° and dip
of 86 °. In order to get a high resolution image of the rupture process, we divided the fault plane into 171 el-
ements with a spatial dimension of 2. 0 km by 2. 0 km.
2. 3 Layered crustal model
According to the velocity model of eastern Nepal[!?]
and based on the model of Crust 2. 0 , we established
the multi-layered crustal models in India-Nepal and
Myanmar areas, respectively. In this paper, we only
discuss co-seismic effect caused by these two events ,
and hence assume the crust and upper mantle to be
purely elastic layers, and the model parameters are lis-
ted in table 1 and table 2.
Table 1 A multi-layered earth model for the
India-Nepal earthquake
Serial No.
2
3
4
Serial No.
2
3
4
5
Depth Vp Vs Density (km) (km·s- 1 ) (km·s- 1 ) (km·s- 3 )
0.0-3.00 5.5 3.2 2 100
3.00-23.0 5.7 3.2 2 400
23.0-55.0 6.3 3. 7 2 700
55.0 8.0 4.5 3 450
Table 2 The multi-layered model for the
Myanmar earthquake
Depth Vp Vs Density (km) (km·s- 1 ) (km·s- 1 ) (km·s- 3 )
0.0-1.00 2.5 1.2 2 100
1.00-19.0 6.1 3.5 2 750
19.0-36.0 6.3 3.6 2 800
36.0-45.0 7.1 4.0 3 100
45.0 8.0 4.6 3 350
3 Modeling co-seismic deformation and gravity changes
Here we present the result of our calculation for the
surface co-seismic deformation and gravity changes of
the Myanmar and India-Nepal earthquakes according to
the above-mentioned models.
3.1 Co-seismic horizontal displacement
Figure 3 shows the calculated co-seismic horizontal
surface displacements caused by the India-Nepal and
Myanmar earthquakes. It may be seen that their maxi-
mum displacements are quite different, about 9. 6 em
and 36 em, respectively, although their magnitudes are
comparable. The difference may be due to either differ-
ent models used or different focal depths ( Tab. 3 ) .
The depth of the India-Nepal earthquake we used is
19. 7 km (from USGS) , which is almost twice as large
as the depth of Myanmar earthquake ( 10 km) ; hence,
the smaller displacement.
3. 2 Co-seismic vertical displacement and gravity
changes
Figures 4 and 5 show , respectively, the calculated co-
seismic vertical deformation and gravity changes
caused by the India-Nepal and the Myanmar earth-
quakes. The vertical displacements of the India-Nepal
earthquake show an irregular pattern in four quadrants
( Fig. 4 ( a) ) , which is consistent with its focal mech-
anism, which shows a left-lateral strike slip with an
oblique thrust component. This pattern indicates up-
ward displacements in northern and southern near-
field, and western and eastern far-field (due to com-
4 Geodesy and Geodynamics Vol. 3
30"N
I r
Mw6.9e
5cm
CHINA
~ asa
(a) Horizontal co-seimic surface displacements of the India-Nepal earthquake
22"N
~---Chiang Khong
. ·L~2o·N
oe.D~ ;: 100 "E 101 "E 102 "E
(b) Horizontal co-seimic surface displacements of the Myanmar earthquake
Figure 3 Horizontal co-seimic surface displacements ( Stars indicate epicenters ; circles , main cities)
Table 3 Estimated depth of the India-Nepal
earthquake by different institutes
Institutes Depth (km)
IGP-CEA I 20
USGS 11 19.7
Harvardm 47.4
I http :I /www. csi. ac. en; ll http :I I earthquake. usgs. gov; m http :I /www. globalcmt. org
pression) , with a maximal uplift of 4. 65 em. It also
indicates downward displacements in the northern and
southern far-field, and western and eastern near-field
( due to tension ) , with a maximal subsidence of
- 1. 26 em. The gravity change shows a similar pattern
( Fig. 4 ( b ) ) . The gravity changes are large in the
near-field , with extreme values of 7. 68 f.LGal and
- 1. 49 f.LGal, respectively, but relatively small in the
far field.
On the other hand , the co-seismic vertical displace-
ment and gravity changes caused by the Myanmar
earthquake show a clear symmetrical pattern in four
quadrants ( Fig. 5 ) : compressive uplift of as much as
3. 89 em in the two N-W and E-S uplifting lobes and
tensile subsidence of as much as - 1. 28 em in the two
N-E and S-W lobes. The gravity changes are consist-
ently similar, with extreme changes of 10. 99 f.LGal and
- 8. 02 f.LGal, respectively ( Fig. 5) .
4 Sensitivity of parameters
4.1 Focal-depth uncertainty
In the inversion of the India-Nepal earthquake the focal
depth is quite uncertain for lack of effective con-
straints , and this uncertainty affects our simulation re-
sults. In order to asses this effect, we select four near-
field points around the hypocenter to examine the sur-
face deformation and gravity changes with different fo-
cal depths ( Fig. 6) . The results show that the effect is
relatively small on vertical displacement and gravity
changes ( Fig. 6 ( b) and ( c) ) , but relatively large
on horizontal displacement (Fig. 6 (d) ) , at near-field
locations.
4. 2 Differences between the average-slip and
rmite-element rupture models
As shown in figures 7 and 3 (b) , for the Myanmar
earthquake the maximum co-seismic near-field surface
horizontal displacements calculated on the basis of the
average-slip model are smaller than those of the finite-
element model. As to the corresponding vertical dis-
placements ( Fig. 8 ) , both models give symmetric dis-
tributions of comparable size in the far field with
change of sign in near field , where the maximum uplift
and decline are 1. 87 em and -4. 39 em, respective-
ly, according to the average-slip model; these are
slightly larger than the values of the finite-element
No.1
Liu Chengli,et al. Co-seismic deformation and gravity changes of the 2011
India-Nepal and Myanmar earthquakes
em 6.00 3.00 2.00 1.00 0.10 0.03 0.01
-0.10 -0.03 -0.10 -0.30 -0.50 -1.00
24'NL-----~----~~---L----~--~ -3.00
84'E 86'E 88'E 90'E 92'E 84'E 86'E 88'E 90'E (a) Co-seimic surface vertical displacement (b) gravity changes
92'E
Jl Gal 12.00 6.00 1.50 1.10 0.03 O.Ql
-0.01 -0.10 -0.30 -0.50 -1.00 -3.00
Figure 4 Co-seimic surface vertical displacement and gravity changes caused by the India-Nepal earthquake
( Positive is downward ; yellow stars indicate the epicenter; green circles , main cities)
96'E 100 'E 102 'E
(a) Co-seimic surface vertical displacement
em 4.00 2.00 0.50 0.30 0.10 0.03
-0.01 -0.03 0.10 0.30 0.50 1.50 3.00
96'E 100 'E 102 'E
(b) Gravity changes
J1Gal 12.00 6.00 3.00 1.00 0.10 0.00 0.01
-0.01 -0.03 -1.00 -2.50
-10.00
Figure 5 Co-seimic surface vertical displacement and gravity changes caused by the Myanmar earthquake
( Positive is downward ; yellow stars indicate the epicenter; green circles , main cities)
Figure 6 Variation of vertical displacement ( positive downwards in b) , gravity ( c) ,
and horizontal displacement (d) with focal depth for the India-Nepal earthquake at
four locations shown as stars in corresponding colors. Dash line
indicates the focal depth shown in table 3.
5
6 Geodesy and Geodynamics Vol. 3
98 'E 99 'E 100 'E 101 'E 102 'E
Figure 7 Horizontal co-seismic surface displacements of the Myanmar earthquake obtained
by using a finite-element rupture model ( red arrows) and an average-slip model.
(green arrows) . Star indicates the epicenter; circles, main cities
20
96'E 100 'E
(a)
102'E 96'E 100'E
(b)
102'E
4.00 2.00 0.50 0.30 0.10 0.03
-0.01 -0.03 -0.10 -0.30 -0.50 -1.50 -3.00
Figure 8 Vertical co-seismic surface displacement of the Myanmar earthquake (positive is downward)
obtained by using ( a) a finite-element rupture model, and ( b) an average-slip model
rupture model. shown in figure 8 ( a) . This discrepancy
is due to the simplification of the slip parameter in
the average-slip model. However, without a reliable
co-seismic rupture model, the simulation result of the
average-slip model is still acceptable.
5 Conclusions and discussion
1 ) Although both earthquakes are strike-slip events
with comparable magnitudes, they have different calcu-
lated maximum co-seismic horizontal displacements
(India-Nepal, 9. 5 em; Myanmar, 36 em). This dis-
crepancy may be due to their different focal depths and
the use of different slip models.
2) The vertical displacement and gravity changes of
both earthquakes exhibit a pattern of four quadrants.
Different from Myanmar earthquake, the signs of the
gravity changes in the near-field for the India-Nepal
earthquake are almost opposite to those in the far field,
and the near-field displacement changes show asym-
metric distribution. The gravity changes are consistent
with the vertical deformation, but in the near-field, it
is only at the epicenter that gravity has increased; with
increasing epicentral distance, the gravity shows a de-
crease and then an increase.
3 ) The results based on the two different models are
only slightly different in the near-field due mainly to
the simplified parameters of average-slip model. Thus,
No.I
Liu Chengli,et al. Co-seismic deformation and gravity changes of the 2011
lodia-Nepal and Myanmar earthquakes 7
without a reliable finite-element rupture model , the use
of the average-slip model is acceptable.
Our analysis of co-seismic surface displacement and
gravity field changes for the India-Nepal earthquake
and Myanmar earthquake may provide useful informa-
tion for the research of seismic activity, long-term tec-
tonic activity , and deformation field in the houndary
zone between the India and Eurasia plates, the east-
em structural belt of Tibetan plateau , and surrounding
areas.
It should be noted that the complexity nf rupture
have a significant effect on the model computation , es-
pecially in the near field. In this study, we used the
average-slip and finite-fault models. H more data
( such as near-field GPS , InSAR and teleseismic ) are
available in the combined inversion to obtain a more
accurate co-seismic rupture model, we should be able
to obtain a better result.
Acknowledgement:
We would like to thank Prof. Wang Rongjiang for
providing calculation software. Seismic data used in
this paper is provided by the United States IRIS seismic
data management center. The figures were made using
free GMT (Generic Mapping Tools) software.
References
[ 1 ] Zhang GM, Ma H S, Wang H, et al.. Bmmd.aries between active-
tectonic blocks and strong earthquakes in the China mainland.
Chineoe l . Geophy•. , 2005 , 48 ( 3 ) , 602 - 610. ( in Chineoe)
[ 2 ] Zhang G M, Ma H S, Wang H, et al. The relation between ac-
tive blocks and strong earthquakes in China. Science in China
(Seri. D). 2004 ,34 (7) , 591-599. (in Chine•e)
[ 3 ] Qin S Q and Xue L. A summary of prediction for the Y:ingjiang
Ms5. 8 earthquake in Yunnan and the BUI1Ila Ms7, 2 earthquake
as well as the analysis on the earthquake situation after the earth-
quake. Progre.s in Geoph)", 2011,26(2), 462-468. (in Chi-
nese)
[ 4 ] Chang Zufeng, An Xiaowen, ChenGang, et al. Brief analysis on
seismotectonics of Mong Hpayak Ml. 2 earthquake in Myanmar.
Recent Developments in World Seismology. April, 2011, 4(Ser-
ial No.388) :52-56. (in Chinese)
[ 5 ] Gan W J, Zhang P Z, Shen Z K,et al. Present-day crustal mo-
tion within the Tibetan plateau inferred from GPS measurements.
Journal of Geoph)"ical Researoh, 2007, 112, 808416.
[ 6 ] Deng J and Sykes L R. Evolution of the stress field in southern
California and triggering of moderate-size earthquakes: a 200-year
perspective. J Geophys Res. , 1997, 102: 9859-9886.
[ 7 ] Deng J and Sykes L R. Stress evolution in southern California and
triggering of mOOerate-, small-, and micro-size earthquakes. J
Geophy• Res.' 1997' 102, 24411-24435.
[ 8 ] Feod A M. Emthquake triggering hy •tatic, dynamic, and pom-
eismic stress transfer. Annu Rev Earth Planet Sci. , 2005, 33:
335-367.
[ 9] 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. Journal of Geophysi-
cal Re""""h, 2003, 108(810)' 2501.
[10] Donald L, Wells and Kevin J and Coppersmith. New empirical
relationships among magnitude, rupture length, rupture width,
rupture area, and surlace displacement. Bulletin of the Seismo-
logical Society of America. 1994, 84 ( 4) : 974 -1002.
[11] WangR, LorenroF-M,.,andRothF. PSGRN/PSCMP-anew
code for calculating co- and post-seismic deformation, geoid and
gravity changes based on the viscoelastic-gravitational dislocation
theory. Computers and Geosciences, 2006, 32(4): 527-541.
[ 12 ] Wang R. A simple orthonormalization method for the stable and
efficient computation of Green' s functions. Bull Seism Soc Am. ,
1999' 89' 733 -741.
[ 13] Wang Rand Kuempel H J, Poroelasticity: e:Eiicient modeling of
strongly coupled, slow deformation processes in a multi-layered
half-•pace. Geophyoic•, 2003 , 68 ( 2) , 705 -717.
[ 14] Wang R, Lorenzo F and Roth F. Computation of deformation in-
duced by earthquakes in a multi-layered elastic crust-FOR-
TRAN programs EDGRN/EDCMP. Computers and Geosciences,
21XJ3, 29, 195-200.
[ 15] Ji C, D J Wald and Heimberger D V. Source description of the
1999 Hector Mine, California earthquake. Part I. Wavelet do-
main inversion theory and resolution analysis. Bull. Seismol.
Soc. Am. ' 2002' 92: 1192 - 1207.
[ 16] Bassin C, Laske G and Masters G. The current limits of resolu-
tion for surface wave tomography in North America. EOS Trans
AGU ,2000, 81 : 897.
[17] Monsalve G, Sheehan A, Schulte-Pel.kum, et al. Seismicity and
one-dimensional velocity structure of the Himalayan collision
zone: Earthquake in the crust and upper mantle. J. Geophys.
a.,. -Solid Emth, 2006, lll, 810301.
Co-seismic deformation and gravity changes of the 2011 India-Nepal and Myanmar earthquakes1 Introduction2 Theory and models3 Modeling co-seismic deformationand gravity changes4 Sensitivity of parameters5 Conclusions and discussion