Embed Size (px)
Citation preview
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
400 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
25 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
300 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
194 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
159 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
1 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
125 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
169 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
250 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
350 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
5 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
451 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
96 97 98 99 100 101 102 103 104 105
−6
−5
−4
−3
−2
−1
0
1
1−bsat
2−btet
3−lais
4−lnng5−mkmk
6−mlkn
7−mnna
8−msai9−ngng
10−pbjo
11−ppnj
12−prkb
13−pski
14−slbu
15−tiku
5 days, γ = 0.2, rake = 73°
0
200
400
600
800
1000
1200
mm
1 day 5 days
25 days 125 days
159 days 169 days
200 days 250 days
300 days 350 days
400 days 451 days
A. P. Kositsky*, J.-P. Avouac*, J. Genrich* , J. Galetzka, K. Sieh*, D. H. Natawidjaja, B. [email protected], [email protected]*1200 E. California Blvd. Mail Code 100-23 Pasadena CA 91125
Padang
Bengkulu
1833(Mw~8.9-9.1)
1797(Mw~8.7-8.9)
1935(Mw~7.7)
Wha
rton
Ridge
.2 .4 .6 .8 1
Coupling
0
96 98 100 102
-
-
-
6
4
2
2
4
5.7 cm/yr
Enggano
Inve
stig
ator
Frac
ture
Zone
2005(Mw8.7)
2000(Mw7.9)
2004(Mw9.1)
Pagai Is.
Sipora
Siberut
Mentaw
ai Islands
Batu
2007(Mw 8.4)
2007(Mw7.9)
1B
2B
1A
1C
2A
Afterslip on the Sumatra megathrust following the 2007 Mentawai earthquake sequence
AbstractA magnitude 8.4 and twelve hours later, a magnitude 7.9 earthquake occurred on September 12, 2007 occurred on the Sumatra megathrust in the Mentawai islands area, rupturing partially a patch of the plate interface that had remained locked in the decades preceding the earthquake (Figure 1) [Konca, et al., in press]. Here we analyze the postseismic time series collected at the per-manent GPS stations from the Sumatra Geodetic Array (SuGAr) (http://www.tectonics.caltech.edu). Displacements were determined first relative to ITRF200 and then expressed relative to the Australia plate using the pole of [Bock, et al., 2003]. The time series show trenchward displacements amounting to as much as 80 cm over the first 125 days after the earthquakes, and as much as 120 cm after 451 days. The stations located on the islands show postseismic subsidence (Figure 2), while these same stations had experienced uplift during the earthquake sequence. This pattern suggests afterslip on the shallow portion of the plate interface updip of the ruptured area as was observed following the Mw 8.6 Nias earthquake of 2005 [Hsu, et al., 2007]. The geodetic times series were inverted for slip on the megathrust using the PCAIM algorithm [Kositsky and Avouac, submitted] (Figure 4). The best fitting model obtained from the inversion 3 first principal components does yield a good fit to the time series (Figure 3), showing that the data are consistent with the hypothesis that postseismic deformation is dominated by afterslip on the megath-rust. Moreover the distribution of afterslip is mostly complementary of the co-seismic slip distribution in the area where the data distribution yields some reasonable spatial resolution (Figure 5). Finally, the temporal evolution pattern is consistent with a logarithmic increase of slip as expected from velocity strengthening friction laws (Figure 3). These results suggest that postseis-mic deformation following the 2007 Mentawai earthquakes is probably reflecting frictional slip on velocity strengthening patches on the megathrust, in particular updip of the rupture area. Af-terslip released a geodetic moment of about 1.5 x 10^21N.m over the 451 days following the mainshock, equivalent to as much as 25% of the moment released by seismic slip during the earth-quake sequence.
Figure 1: Interseismic coupling on the Sunda megathrust, offshore Sumatra, and rupture area of major recent earthquakes. The pat-tern of coupling, defined as the ratio of interseismic slip rate to plate convergence rate, is derived from the modeling of geodetic and paleogeodetic data [Chlieh, et al., 2008]. Slip distribution of the 2005 Mw 8.6 earthquake of 2005 is shown with 5 meter con-tour lines in green Gray and black polygons show estimated rup-ture areas of the 1797 and 1833 earthquakes. Dark and pale blue lines show the 1 m and 5 m slip contour lines of the Mw 8.4 and 7.9 seismic ruptures of 2007, stars show the epicenters.
Figure 6: Distribution of afterslip over different time periods obtained with PCAIM. A series of earthquakes occured near slbu near day 159, and the contribtuion of the largest shocks has been removed from slbu using a parametric model. Note the slip shift of slip toward the south during the last ~200 days.
Bock, Y., L. Prawirodirdjo, J. F. Genrich, C. W. Stevens, R. McCaffrey, C. Subarya, S. S. O. Puntodewo, and E. Calais, Crustal motion in Indonesia from Global Positioning System measurements, Journal of Geophysical Research-Solid Earth, 108, doi: 10.1029/2001KB000324, 2003.
Chlieh, M., J.-P. Avouac, K. Sieh, D. H. Natawidjaja, and J. Galetzka, Het-erogeneous coupling on the Sumatra megathrust constrained from geodetic and paleogeodetic measurements, J. Geophys. Res.,, 113, doi:10.1029/2007JB004981, 2008.
Hsu, Y. J., P. Segall, S. B. Yu, L. C. Kuo, and C. A. Williams, Temporal and spatial variations of post-seismic deformation following the 1999 Chi-Chi, Taiwan earthquake, Geophysical Journal International, 169, 367-379, 2007.
Konca, A. O., et al., Partial rupture of a locked patch of the Sumatra megathrust during the 2007 earthquake sequence, Nature.
Kositsky, A. P., and J. P. Avouac, Inverting geodetic time series with a Principal Component Analysis based Inversion Method, J. Geophys. Res. B., in press.
This study was partly funded by the Gordon and Betty Moore Fundation and NSF grant EAR 0838495. Andrew Kositsky thanks the Caltech Summer Under-graduate Research Fellowship (SURF) program, George W. Housner Student Dis-covery Fund, and Burns Prize in Geology.
References and Acknowledgements
Evolution of AfterslipGeological Context
Figure 4: Time variation of slip at depth at the the patches of largest slip within the updip, downdip, and seismic crisis regions. The slip at the seis-mic crisis region is consistent with the timing of the later earthquakes in that region.
2007.8 2008 2008.2 2008.4 2008.6 2008.8
−400
−200
0
200
400
600
800
1000
1200
Slip on Select Patches at Depth
Slip
in m
m
Updip PatchDowndip PatchSeismic Crisis Patch
Figure 5: Original GPS imtseries plus model predictions. Most features in most of the timeseries are faithfully represented by the model.
0
100
200
−100
0
100
200
−50
0
50
−100
−20
−10
0
10
20
0
50
2007.8 2008 2008.2 2008.4 2008.6 2008.8 2009−40
−20
0
20
40−50
0
50
−50
0
50
100
150
2007.8 2008 2008.2 2008.4 2008.6 2008.8 2009−200
−100
0
100
−50
−50
0
50
100
150
−100
0
100
200
−50
0
50
−40
−20
0
20
40
−40
−20
0
20
40
−40
−20
0
20
40
−20
0
20
40
60
−20
−10
0
10
20
−100
−50
0
50
−20
−10
0
10
20
−20
−10
0
10
20
−40
−20
0
20
40
−20
−10
0
10
20
−20
0
20
40
−40
−20
0
20
40
−50
0
50
100
−50
0
50
100
2007.8 2008 2008.2 2008.4 2008.6 2008.8 2009−40
−20
0
20
40
−20
−10
0
10
20
−40
−20
0
20
40
−40
−20
0
20
40
0
50
100
−100
0
100
200
2007.8 2008 2008.2 2008.4 2008.6 2008.8 2009−200
−100
0
100
-50
−20
0
20
40
60
−40
−20
0
20
40
−40
−20
0
20
40
−50
0
50
100
150
−100
0
100
200
2007.8 2008 2008.2 2008.4 2008.6 2008.8 2009−40
−20
0
20
40
−50
0
50
100
150
−100
0
100
200
−40
−20
0
20
40
−100
0
100
200
−100
0
100
200
−40
−20
0
20
40
bsat btet lais lnng mkmk
mlkn mnna msai ngng pbjo
ppnj prkb pski slbu tiku
Raw DataModeled Signal
E
N
U
E
N
U
E
N
U
E
N
U
E
N
U
E
N
U
GPS Timeseries and Model Predictions
0 1 2 3 4 5 6 70
20
100
120
Number of Components
χ2red vs. Number of Components
105.8131
3.1132 1.8009 1.4755 1.3012 1.2313 1.2073 1.1902
χ2 re
d
1 2 3 4 5 6 7 8 910−300
10−250
10−50
100
p1→2
p2→3
p3→4
p4→5
p5→6
p6→7
=4.8e−297
=8.9e−48=1.5e−22
=8.9e−08
=0.0015
=0.0041
Added Component
p−value
p-Value For Each Component
100 102 104-6
-4
-2
0
98
Seismicity: 451 Days Post-Mainshock Model Statistics
Figure 2: Location of seismicity following the September 12, 2007 mainshocks. Primary shock shown in cyan. Note the three clusters of seismicity, updip, downdip, and to the Northwest (“seismic crisis region”).
Figure 3: Two measures of the significance of the seven components used.p-values are the chance that the additional ith component explains primarily noise, and χ should be about one.2
red
