Accepted
Manuscript
2
7
July 2015
Slip pulse and resonance of Kathmandu basin during the 2015
Gorkha earthquake,
1
Nepal
2
3
Authors:
John Galetzka
1
+
, Diego Melgar
2
, Joachim F. Genrich
1
, Jianghui Geng
3
, Susan Owen
4
,
4
Eric O. Lindsey
3
, Xiaohua Xu
3
, Yehuda Bock
3
, Jean
-‐
Ph
ilippe Avouac
5,1*
, Lok Bijaya Adhikari
6
,
5
Bishal Nath Upreti
7
, Beth Pratt
-‐
Sitaula
8
,
Tara Nidhi Bhattarai
9
, Bhairab P Sitaula
9,
, Angelyn
6
Moore
4
, Kenneth W. Hudnut
10
, Walter Szeliga
11
, Jim Normandeau
12
, Michael Fend
12
,
7
Mireille Flouzat
13
, Laurent Bollinger
1
3
,
Prithvilal S
hrestha
6
,
Bharat Koirala
6
, Umesh
8
Gautam
6
, Mukunda Bhatterai
6
, Ratnamani Gupta
6
, Thakur Kandel
6
, Chintan Timsina
6
, Soma
9
Nath Sapkota
6
, Sudhir Rajaure
6
, Naresh Maharjan
6
10
11
Affiliations:
12
1
California Institute of Technology,
Department
of Geol
ogy and Planetary Sciences,
13
Pasadena, CA, USA.
14
2
University of California Berkeley, Seismological Laboratory, Berkeley, CA, USA.
15
3
Cecil H. and Ida M. Green Institute of Geophysics and Planetary Physics, Scripps Institution
16
of Oceanography, University of Cal
ifornia San Diego, La Jolla, CA, USA.
17
4
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA.
18
5
University of Cambridge, Department of Earth Sciences, Cambridge, UK.
19
6
Department of Mines and Geology, Kathmandu, Nepal.
20
7
Nepal
Academy of Science and Technology (NAST) Khumaltar, Lalitpur, Nepal
21
8
Geological Sciences Department, Central Washington University
22
9
Tri
-‐
Chandra Campus, Tribhuvan University,
Ghantaghar, Kathmandu Nepal
23
10
US Geological Survey, Pasadena, CA, USA.
24
11
PAcifi
c Northwest Geodetic Array and Geological Sciences, Central Washington University
25
12
UNAVCO Inc., Boulder, CO, USA
26
13
CEA, DAM, DIF, Arpajon, France
, France.
27
+
Now at UNAVCO Inc., Boulder, CO, USA.
28
*Correspondence to: avouac@gps.caltech.edu
29
30
Detailed
geodet
ic
imaging of earthquake rupture
enhances
our understanding of
31
earthquake
physic
s
and induced ground shaking.
The April 25, 2015 Mw 7.8 Gorkha,
32
Nepal earthquake is the first example of a large continental megathrust r
upture
33
beneath a high
-‐
rate (5 Hz) GPS n
etwork.
W
e
use
GPS
and InSAR
data to
model
the
34
earthquake rupture
as a
slip puls
e of ~20 km width, ~
6
s duration
,
and
with
peak
35
sliding velocity of
1.1
m/s
that
propagated toward Kathmandu basin
at
~
3.
3
km/s
36
over ~1
4
0
km.
The
smooth
slip onset
,
indicating
a large
~5
m
slip
-‐
weakening
37
distance,
caused
moderate ground shaking at
high
>1Hz
frequencies (~16%
g)
and
38
limited damage to regular dwellings.
W
hole
b
asin resona
nce
a
t
4
-‐
5
s period caus
ed
39
collapse of tall structures
, including
cultural artifacts.
40
41
One sen
tence summary:
High
-‐
rate GPS record
s
reveal that
t
he
Gork
h
a earthquake
42
resulted f
r
o
m
eastward
propagation of
a ~6s long
slip pulse
, with smooth onset
which
43
generated mild ground shaking
but
exited resonance of
Kathmandu bas
in
at ~4
-‐
5 s
.
44
45
The shape of the slip
-‐
rate time function (STF) during
seismic rupture
provides
critical
46
insight into constitutive fault properties
. The abruptness of slip onset
determine
s
the high
47
frequency content
a
nd hence the intensity of the near
-‐
field ground motion
(
1
)
,
whereas
the
48
tail
, which discrimi
nates p
ulse
-‐
like and crack
-‐
like ruptures
(
2
)
, has a low frequency
49
signature.
Therefore,
resolv
ing
the STF
with band
-‐
limited strong motion records
is difficult
.
50
The combination of
high
-‐
rate GPS waveforms
(
3, 4
)
, which capture both dynamic and
51
permanent deformation
,
overcomes this limitation.
52
The April 25
th
2015 M
w
7.8 Gorkha, Nepal earthquake resulted from unzipping of the
53
lower edge of the locked portion of the Main Himalayan Thrust (MHT) thrust faul
t,
along
54
which the Himalayan wedge is thrust over India
(
5
)
. The earthquake nucleated
~
8
0
km
55
northwest of Kathmandu
and
ruptured a 1
4
0
km long segment of the fault (F
ig
ure
1
A
)
with
56
a hypocentral depth of ~15 km and a dip angle of 7
-‐
12
°
(
5, 6
)
. The MHT
accommodates th
e
57
majority of the convergence between India and southern Tibet w
ith a rate
between 17 an
d
58
21 mm/yr
(
7
)
. For the 2015 event, which resulted in over 8,000 deaths, mostly in the
59
Kathmandu and adjacent districts, Mercali shaking intensities (MMI)
reported by the
60
National So
ciety for Earthqua
ke Technology
(
8
)
reach
ed
up to IX (violent) and exceeded VI
61
(strong)
over a 170x40
km
2
area
.
Kathmandu has been struck by repeated
earthquakes in
62
the past, with major destruction (MMI>X
, extreme
) i
n 1255, 1344, 1408, 1681, 1833 and
63
1934
(
9
-‐
11
)
. These earthquakes all
occurred close to Kathmandu
and
have been
assigned
64
magnitudes
between
Mw 7.5
and 8.
4
. Damages in the Kathmandu basin were probably
65
amplified by
site effects
during the Gorkha earthquake as happened with past
events
(
12,
66
13
)
.
The
basin is filled with 500
-‐
600 m of fluviolacustrine sediments resting on
67
metamorphic basement
(
14
)
.
68
The
damage to the
most vulnerable
vernacul
ar
dwellings in Kathmandu
,
which rarely
69
exceed 4 stories, was in fact
much
less than
expected in view of the 2015 earthquake’s
70
magnitude and its proximity to Kathmandu. By contrast, some taller structures were more
71
severely affected, such as the
60
m tall Dharahara t
ower
which collapse
d
,
but
had
partially
72
survived the
Mw 8.1
-‐
8.
4
1934 earthquake
.
. The 1934 event
induced
much
more extensive
73
destruction to
vernac
ular
dwel
lings in Kathmandu than in 2015
(20% of the buildings in
74
Kathmand
u were destroyed in 1934
, less than 1% in 2015
)
(15)
.
These observations reflect
75
the combine
d effects of the source characteristics and local geological conditions, in
76
addition to evolution of building practices.
77
The 2015
Gorkha earthquake ruptured a subhorizontal portion of the MHT lying
78
directly beneath a network
(16)
of continuous GPS (cGPS) s
tations recording at a high rate
79
of 5 samples per second, and one accelerometer station
(17)
(
Fig.
1A). I
n addition, surface
80
displacements were measured with
interferometric
synthetic aperture radar
,
InSAR,
(
18,
81
19
)
(
fig.
S1). While a number of recent earthquake
s were documented with similar
82
techniques
(20, 21)
, the Gorkha event is the first occurrence of a large continental thrust
83
earthquake to be observed by high
-‐
rate
c
GPS
stations
at very close distances to
and
84
completely encompassing
the rupture area
. The combination of these measurements
85
provide the opportunity to image the kinematics of the source process and the strong
86
ground motion that led to the particular pattern
of structural damage observed during this
87
earthquake.
88
The records of seismic displacements and accelerations (Figs
.
2 and S2) show
89
southward motion of up to 2 m, with a rise time on the order of 6 seconds. The pulse is
90
particularly clear at
cGPS
station
KKN4 located
on bedrock
just north of Kathmandu
and
91
only ~13
km above the fault
.
The displacement at this station started at about 25
s
after the
92
onset of rupture
,
corresponding to 15 seconds after P
-‐
waves arrival time (Fig.2),
and
93
reached its final static
value by about 32
s, using the USGS origin time of radiated direct P
94
waves at 06:11:26.270
UTC
(
6
)
.
The records clearly indicate a pulse
-‐
like rupture
(
22
)
with
95
slip on any given portion of the fault occurring over a short fraction of the total ~70
s
96
duration of the earthquake source
(5)
. Given the ~78 km distance of KKN4 to the epicenter,
97
the pulse must have propagated at ~3 km/s, a value consistent with wav
eform modeling
98
and back projection of high frequency seismic waves recorded at teleseismic distances
(5)
.
99
Surface velocities reached values of
~0.7
m/s. The cGPS station
NAST
within Kathmandu
100
basin shows, in addition to the pulse seen at KKN4, strong oscillations of period of about 3
-‐
101
4 seconds lasting for ~20 s (Figs
.
2 and 3A). Th
e Gorkha earthquake must have excited a
102
resonance of the Kathmandu basin as a whole. The resonance is clearly shown in the
103
response spectra from these stations as well as from the accelerometer station
K
A
TNP
(
F
ig
104
3G
-‐
I).
105
To retrieve the kinematics of the se
ismic rupture, we carried out a
formal
inversion of
106
time
-‐
dependent slip on the fault
(23, 24)
and compared the recorded waveforms with
107
forward predictions assuming a propagating slip pulse
with varied characteristics
. We
108
assumed a planar fault geometry with a strike of 295
°
and a dip of 11
°
in accordance with
109
the teleseismic W
-‐
phase moment tensor solution from the USGS
(
6
)
. We tested shallower
110
dips up to 7
°
but found that 11
°
provided a better fit to the data. The fault was discretized
111
into 10x10 km subfault segments. We jointly inverted the three
-‐
component, 5 Hz GPS
112
derived velocity
waveforms,
the GPS static offsets,
and the InSAR line of sight (LOS) static
113
displacements measured between February 22 and May 3 (
fig.
S1). The GPS
displacement
114
time series show
s
large
postseismic
motion
at only one station (CHLM) with less than 2
cm
115
magnitude
on both the
horizontal and vertical over the week following the earthquake.
116
Therefore, for our purposes,
we neglect
t
he contribution of postseismic deformation to the
117
LOS displacements
.
.
The model fit
s both data sets
closely
(Figs. 1A
)
, with 8
6
% variance
118
reduction
for
the InSAR
and GPS
coseismic displacements and
7
4% variance reduction
for
119
the GPS velocity waveforms (
Figs.
S2, S
4
). The model indicates predominantly unilateral
120
rupt
ure to the southeast with peak slip of ~
6.5
m on a large asperity to the north of
121
Kathmandu. The event duration is 65 s (
fig.
S
4
) with peak moment release at 23 s when the
122
slip pulse is less than 10 km north of Kathmandu (movie S1)
,
and peak slip
-‐
ra
te is
1.1
m/s.
123
Most of the slip is concentrated within a narrow
region
between
the
10 and 20 km
fault
124
depth contours.
We find a
large
asperity with 3.
0
m of slip due east of the main asperity and
125
between 20 and 23 km depth. The ruptu
re velocity
of the propagating slip pulse
indicated
126
by the onset of slip in our best
-‐
fitting model
is ~3.2
km/s and has a maximum allowed
127
velocity of
3.3 km/s (
fig.
S4)
.
This velocity
corresponds to
~
9
5
% of the shear wave speed at
128
the depth of the m
ajority of slip (15 km) according to the local velocity model used to
129
calculate the Green
’s
functions (Table S2)
,
indicat
ing
a very fast rupture propagation. Slip
130
tapers at 17
-‐
20 km depth along the edge of the locked zone of the MHT.
The inversion has a
131
la
rge number of parameters
,
which
allows for
a relatively complex rupture history
.
132
However, t
he resulting model is remarkably simple with essentially a single propagating
133
slip pulse. The spatio
-‐
temporal evolution of the slip pulse matches well the
location of the
134
sources of high frequency (0.5
-‐
2Hz) seismic waves derived from the back
projection of
the
135
teleseismic waveforms
(
5
)
(Movie S1).
136
We calculated the static stress change on the fault plane due to the earthquake (
Fig.
1B).
137
It shows loading of the fault around the main asperity where most of the aftershocks
138
occur
red,
including the Mw 7.
3
aftershock of May 12,
as
expected from triggering by
139
coseismic stress transfer
(25)
. The model predicts a pattern of uplift of the Kathmandu
140
basin and subsidence at the front of the high range (
fig.
S
4
), approximately opposite to the
141
pattern observed in the
interseismic period as expected from simple models of the seismic
142
cycle on the MHT
(
26, 27
)
.
143
The record at
station
KKN4
should be a close representation of the slip
-‐
rate time
144
function as it
lies only ab
out
13
km above the propagating slip pulse and is not affected by
145
the site effects seen at the stations in Kathmandu basin
.
We conducted synthetic tests with
146
the
same Earth structure model used in the inversion (
T
able S1) to assess the distortion
147
and smoothing introduced by the elastic half space response (
fig.
S
5
).
We found
a vertical
148
velocity amplitude of about 70% of the peak slip rate on the fault direct
ly beneath it along
149
with a well
-‐
preserved temporal shape.
Furthermore,
the
tests
demonstrate that
the smooth
150
onset of slip is not an artifact
resulting from
the transfer through the elastic medium
151
represented by the elastodynamic Green’s functions. The shape of the slip pulse can also be
152
retrieved from the G
PS records at NAST and strong motion
vertical
records at K
A
TNP
153
which are less affected by site effects than the horizontal records
(
Fig.
1
). All three records
154
indicate a ~6 s duration
pulse.
The
shape of the pu
lse fits
the regularized Yoffe function
155
(28)
yield
ing
a rather smooth rise
,
with an accel
eration time to peak slip rate
of
τ
s
=
1.7
s
,
a
156
rise time of
τ
R
=3.3 s
and a total effective duration of
τ
eff
=
6.7
s. The slip
-‐
rate pulse derived
157
from the inversion is also well fit using the same values of
τ
s
and
τ
R
s and peak slip
-‐
rate of
158
~0.9 m/s (
F
ig.
4).
We compared the recorded waveforms with predictions from a suite of
159
forward models to test the robustness of our results.
W
e use
d
the static slip model
in these
160
tests
deduced from the inversion of the GPS static and InSAR measuremen
ts
(
Fig.
S7)
.
We
161
assumed a propagating slip pulse with varying characteristics using the regularized Yoffe
162
STF. We varied the rupture velocity between 2.8 and 3.6
km/s,
and
the rise time between
2
163
and 10s (
fig.
S8).
We also
tested
the re
solution power of the inversion and the limited bias
164
introduced by the regularization
applied to the
inversion
s
by inverting synthetics
165
calculated from forward modeling
(24, fig. S10, fig. S11)
.
Together,
t
hese test
s
demonstrate
166
the duration of the slip pulse is probably
less
than 10
s
and
the time to the peak
-‐
slip rate
167
can
not
be shorter than 1
s (
we would otherwise observe a
much larger amplitude at hig
h
168
frequencies
) and the average propagation rate of the slip pulse is
not less than
~3.0
km/s
169
over the first 30
s (until
KKN4, NAST and KATNP
records a pulse signal
)
.
170
Tinti et
al
(28)
analyzed
how
the
shape of the
STF
relate
to the characteristi
cs of the
171
friction law governing the dynamics of the rupture.
Based on th
is
rationale (
their
equations
172
6 and 11),
we estimate the slip
-‐
weakening distance to
be
~
5
m (for a peak
-‐
slip of
6.5
m).
173
The distance is a
large value compared to
those estima
ted from
kinematic
and dynamic
174
modeling of seismic ruptures
(
29, 30
)
,
which
tend to be overestimated
(1)
and
are typically
175
on the order of 0.5 to 1 m
.
The large value
we
obtained
is
possibly
related to the earthquake
176
occurring close to the brittle
-‐
ductile transition at the lower edge of the locked portion of
177
the MHT. Th
e modeled
smooth onset of the STF and the related large slip
-‐
weakening
178
distance
provide an explanation of the relatively low amplitude of shaking at frequencies
179
above 1 Hz.
T
he observed slip
-‐
weakening behavior does not require the friction law to be
180
actually slip
-‐
weakening. A fault obeying rate and state friction can show a
n effective slip
-‐
181
weakening behavior with an effective critical distance
several orders of magnitude larger
182
that the critical distance entering the
friction law
(31)
.
Aspects of the rupture kinematics
183
and ground strong motion observed during the Gorkha event may also be due to hanging
184
wall effects, the importance of which
could be
assess
ed
through dynamic modeling of the
185
rupture
(
32, 33
)
.
186
Our study provides insight into the main factors that determined damage sustained
187
during the Gorkha earthquake. Whil
e the hypocenter was ~80 km away from the city, the
188
main asperity that radiated most of the energy
was
much closer, just north of the basin
and
189
at relatively shallow depth
. Comparison of the waveforms recorded within the sedimentary
190
basin at NAST and KATNP
(
fig.
3) with the bedrock records at KKN4 shows
prominent
191
differences even though the stations are less than 13 km apart. The waveforms at the
192
bedrock station KKN4 are simple, mostly dominated by the single pulse, while within the
193
basin
peak horizontal ground velocities of 0.5 to 0.8 m/s (considered severe to violent,
194
(34)
) are sustained for
20
s at KATNP and
40
s at NAST. The ratio of the amplitude spectra
195
of the basin waveforms to those at the hill station (
Fig.
2D
-‐
F) shows
amplification of long
196
perio
d energy between 1 and 9
s with the basin amplitudes being 6
-‐
7 times larger in the
197
horizontal direction than at the
bedrock
station. The response spectra (
Fig.
2G
-‐
I) show that,
198
within this amplified period band, it was the 4
s period shaking that wa
s the strongest at
199
the basin stations.
200
The 4
s peak in the response spectra coincides with the observation that the source
201
time function beneath Kathmandu likely had a duration of ~6
-‐
7
s. The net effect of this long
202
source duration with slow onset time is
to produce radiation that is depleted of high
203
frequency energy
(
fig.
S11)
. This explains why
vernac
ular dwellings with only a few stories
204
were not severely affected despite
the anticipated
short period site effects
from
205
microz
oning
(13)
. Furthermor
e, high frequency intensity measurements such as peak
206
ground accelerations were modest (
Fig
2, ~1.6 m/
s
2
, MMI VI), while longer period intensity
207
measures such as peak ground velocity (
Fig
3) were very large (80 cm/s, MMI IX).
208
Kathmandu was faced with
a combination of source and site effects.
Rupture
directivity
209
focused radiated seismic energy towards the city; the smooth onset and 6
-‐
7 second
210
duration of the pulse excited a resonance of the Kathmandu basin, producing protracted
211
duration of violent shak
ing at a period around 4s.
212
213
Acknowledgments
214
The GPS data are available from the UNAVCO website.
The INSAR data are available at
215
http://topex.ucsd.edu/nepal/
.
The Nepal Geodetic Array was funded
by
internal fun
ding to
216
JPA
from
Caltech and DASE and by
the Gordon and Betty Moore Foundation, through Grant
217
GBMF 423.
01 to the
Caltech
Tectonics Observatory and was main
tained thanks to NSF
218
Grant EAR
13
-‐
45136.
Andrew Miner and t
he
PAcific Northwest Geodetic Array
(
PANG
A
)
at
219
Central Washington University are thanked for technical assistance with the construction
220
and operation of the Trib
h
uvan
University
-‐
CWU
network.
Additional funding for the TU
-‐
221
CWU network came from United Nations Development Programme and Nepal Academy
for
222
Science and Technology.
The high rate data were recovered thanks to a rapid intervention
223
funded by NASA (US) and the Department of Foreign International Development (UK).
We
224
thank T
rimble Navigation Ltd and the
Vaidya family for supporting the rapid r
esponse as
225
well.
The accelerometer record at KATN
P
was provided by USGS.
Research at UC Berkeley
226
was funded by the
Gordon and Betty Moore
Fo
undation through grant GBMF 3024. A
227
portion of this
work was carried out at the Jet Propulsion Laboratory, California Institute of
228
Technology, under a contract with the National Aeronautics and Space Administration.
The
229
GPS data were processed by ARIA
(JPL) and the Scripps Orbit and Permanent Array Center.
230
The effort at the Scripps Institution of Oceanography was funded by NASA grants
231
NNX14AQ53G and NNX14AT33G.
ALOS
-‐
2 data were provided under JAXA (Japan) PI
232
Investigations 1148 and 1413.
JPA thanks the
Royal Society for support. We thank Susan
233
Hough, Doug Given,
Irving Flores
and J
im Luetgert for contribution to the installation of
234
this station.
We thank Doug Dreger for discussion and Walter Mooney for comments.
235
236
Authors contributions: Jean
-
Philippe Avou
ac led the study and wrote the article. Diego Melgar
237
did the kinematic modeling and wrote the article. Yehuda Bo
ck supervise
d
the high
-
rate data
238
processing and
wrote the article. John Gale
t
zka led the field operations. Ji
a
nghui Geng
239
conducted the high rat
e data processing. Sue Owen
, Angelyn Moore,
Walter Szeliga
and Jeff
240
Genrich
conducted the low rate data analysis to estimate co
-
seismic offsets.
Eric Lindsey and
241
Xiaohua Xu conducted the InSAR data processing.
Lok Bijaya helped
organizing
the field
242
operati
ons. All other authors contributed to building and servicing the GPS stations and to the
243
post
-
earthquake d
a
ta recovery. All authors edited the article.
244
245
246
References
and Notes
247
References
248
249
1.
M. Guatteri, P. Spudich, What can strong
-‐
mot
ion data tell us about slip
-‐
weakening
250
fault
-‐
friction laws?
Bulletin of the Seismological Society of America
90
, 98
-‐
116
251
(2000).
252
2.
X. Lu, N. Lapusta, A. J. Rosakis, Pulse
-‐
like and crack
-‐
like ruptures in experiments
253
mimicking crustal earthquakes.
Proceeding
s of the National Academy of Sciences of
254
the United States of America
104
, 18931
-‐
18936 (2007).
255
3.
R. M. Nikolaidis, Y. Bock, P. J. de Jonge, P. Shearer, D. C. Agnew, M. Van Domselaar,
256
Seismic wave observations with the Global Positioning System.
Journal of
257
Geophysical Research
-‐
Solid Earth
106
, 21897
-‐
21916 (2001);
258
(10.1029/2001jb000329).
259
4.
G. L. Emore, J. S. Haase, K. Choi, K. A. Larson, A. Yamagiwa, Recovering seismic
260
displacements through combined use of 1
-‐
Hz GPS and strong
-‐
motion
261
accelerometers.
Bulletin
of the Seismological Society of America
97
, 357
-‐
378 (2007).
262
5.
J.
-‐
P. Avouac, L. Meng, S. Wei, W. Wang, J.
-‐
P. Ampuero, Lower edge of locked Main
263
Himalayan Thrust unzipped by the 2015 Gorkha earthquake.
Nature Geoscience
,
264
(submitted); (10.1038/ngeo2518).
265
6
.
USGS,
http://earthquake.usgs.gov/earthquakes/eventpage/us20002ejl
-‐
266
general_summary
.
267
7.
T. Ader, J. P. Avouac, J. Liu
-‐
Zeng, H. Lyon
-‐
Caen, L. Bollinger, J. Galetz
ka, J. Genrich, M.
268
Thomas, K. Chanard, S. N. Sapkota, S. Rajaure, P. Shrestha, L. Ding, M. Flouzat,
269
Convergence rate across the Nepal Himalaya and interseismic coupling on the Main
270
Himalayan Thrust: Implications for seismic hazard.
Journal of Geophysical R
esearch
-‐
271
Solid Earth
117
, (2012); (10.1029/2011jb009071).
272
8.
NSET, Intensity maps of the Mw7.8 Gorkha earthquake, National Society for
273
Earthquake Technology
-‐
Nepal,
274
http://www.nset.org.np/eq20
15/intensity_maps.php
. (2015).
275
9.
J. L. Mugnier, P. Huyghe, A. P. Gajurel, B. N. Upreti, F. Jouanne, Seismites in the
276
Kathmandu basin and seismic hazard in central Himalaya.
Tectonophysics
509
, 33
-‐
277
49 (2011); (10.1016/j.tecto.2011.05.012).
278
10.
R. Bilham,
Location and Magnitude of the 1833 Nepal Earthquake and Its Relation to
279
the Rupture Zones of Contiguous Great Himalayan Earthquakes.
Current Science
69
,
280
101
-‐
128 (1995).
281
11.
M. R. Pant, in
Adarsa
. ( Odisha, India, 2002), vol. 2, pp. 29
-‐
60.
282
12.
A. Dixit, Dwe
lly
-‐
Samant, L., Nakarmi,M., Tucker, B., Pradhanang, S., The Kathmandu
283
valley Earthquake management plan. Published by National Society for Earthquake
284
Technology
-‐
Nepal37 pp.
285
htt
p://www.preventionweb.net/english/professional/publications/
. (1998).
286
13.
Y. R. Paudyal, R. Yatabe, N. P. Bhandary, R. K. Dahal, A study of local amplification
287
effect of soil layers on ground motion in the Kathmandu Valley using microtremor
288
analysis.
Ear
thquake Engineering and Engineering Vibration
11
, 257
-‐
268 (2012);
289
(10.1007/s11803
-‐
012
-‐
0115
-‐
3).
290
14.
S. Moribayashi, Y. Maruo, Basement topography of the Kathmandu valley Nepal: an
291
application of gravitational method to the survey of a tectonic basin in Hima
layas.
J.
292
Jap. Soc. Eng. Geol.
21
, 30
-‐
37 (1980).
293
15.
B. S. J. B. Rana,
Nepalako mahabhukampa (1990 sala) Nepal's great earthquake
. B. S. J.
294
B. Rana, Ed., (1935).
295
16.
The Nepal Geodetic Array
296
(
http://www.tectonics.caltech.edu/resources/kmlnepal.html
) was deployed as a
297
result of a collaboration between the Caltech Tectonics Observatory (US), the
298
Department of Mines and Geology of (Nepal) and the Department Analyse et
299
Surveillance de l
’Environnement (CEA, France).
300
17.
USGS, Earthquake Hazards Program
-‐
NetQuakes: Station KATNP_NQ_01,
301
http://earthquake.usgs.gov/monitoring/netquakes/station/
KATNP_NQ_01/201504
302
25061138/
.
303
18.
E. Lindsey, R. Natsuaki, X. Xu, M. Shimada, H. Hashimoto, and D. Sandwell,, Gorkha
304
Mw 7.8 Earthquake: Line of Sight Deformation from ALOS
-‐
2 Interferometry,
305
http://topex.ucsd.edu
/nepal
.
Seismological Research Letters
, (in prep).
306
19.
ESA, Nepal earthquake: Sentinel
-‐
1 InSAR analyis, April 2015,
http://insarap.org/
.
307
20.
C. Ji, K. M. Larson, Y. Tan, K. W. Hudnut, K. H. Choi, Slip history of the 2
003 San
308
Simeon earthquake constrained by combining 1
-‐
Hz GPS, strong motion, and
309
teleseismic data.
Geophysical Research Letters
31
, (2004); (10.1029/2004gl020448).
310
21.
H. Yue, T. Lay, S. Y. Schwartz, L. Rivera, M. Protti, T. H. Dixon, S. Owen, A. V. Newman
,
311
The 5 September 2012 Nicoya, Costa Rica M
-‐
w 7.6 earthquake rupture process from
312
joint inversion of high
-‐
rate GPS, strong
-‐
motion, and teleseismic P wave data and its
313
relationship to adjacent plate boundary interface properties.
Journal of Geophysical
314
Rese
arch
-‐
Solid Earth
118
, 5453
-‐
5466 (2013); (10.1002/jgrb.50379).
315
22.
T. H. Heaton, Evidence for and implications of self
-‐
healing pulses of slip in
316
earthquake rupture.
Physics of the Earth and Planetary Interiors
64
, 1
-‐
20 (1990).
317
23.
D. Melgar, Y. Bock., Kinem
atic Earthquake Source Inversion and Tsunami Runup
318
Prediction with Regional Geophysical Data.
Journal of Geophysical Research: Solid
319
Earth
120
, (2015); (10.1002/2014JB011832).
320
24.
Materials and methods are available as supplementary material on Science On
line.
321
25.
R. S. Stein, The role of stress transfer in earthquake occurrence.
Nature
402
, 605
-‐
322
609 (1999).
323
26.
R. Bilham, K. Larson, J. Freymueller, F. Jouanne, P. LeFort, P. Leturmy, J. L. Mugnier, J.
324
F. Gamond, J. P. Glot, J. Martinod, N. L. Chaudury, G. R
. Chitrakar, U. P. Gautam, B. P.
325
Koirala, M. R. Pandey, R. Ranabhat, S. N. Sapkota, P. L. Shrestha, M. C. Thakuri, U. R.
326
Timilsina, D. R. Tiwari, G. Vidal, C. Vigny, A. Galy, B. deVoogd, GPS measurements of
327
present
-‐
day convergence across the Nepal Himalaya
.
Nature
386
, 61
-‐
64 (1997).
328
27.
R. Cattin, J. P. Avouac, Modeling mountain building and the seismic cycle in the
329
Himalaya of Nepal.
Journal of Geophysical Research
-‐
Solid Earth
105
, 13389
-‐
13407
330
(2000).
331
28.
E. Tinti, E. Fukuyama, A. Piatanesi, M. Cocco, A ki
nematic source
-‐
time function
332
compatible with earthquake dynamics.
Bulletin of the Seismological Society of
333
America
95
, 1211
-‐
1223 (2005); (10.1785/0120040177).
334
29.
S. Ide, M. Takeo, Determination of constitutive relations of fault slip based on
335
seismic wave
analysis.
Journal of Geophysical Research
-‐
Solid Earth
102
, 27379
-‐
336
27391 (1997).
337
30.
T. Mikumo, K. B. Olsen, E. Fukuyama, Y. Yagi, Stress
-‐
breakdown time and slip
-‐
338
weakening distance inferred from slip
-‐
velocity functions on earthquake faults.
339
Bulletin of the
Seismological Society of America
93
, 264
-‐
282 (2003).
340
31.
A. Bizzarri, M. Cocco, Slip
-‐
weakening behavior during the propagation of dynamic
341
ruptures obeying rate
-‐
and state
-‐
dependent friction laws.
Journal of Geophysical
342
Research
-‐
Solid Earth
108
, (2003); (1
0.1029/2002jb002198).
343
32.
J. E. Kozdon, E. M. Dunham, Rupture to the Trench: Dynamic Rupture Simulations of
344
the 11 March 2011 Tohoku Earthquake.
Bulletin of the Seismological Society of
345
America
103
, 1275
-‐
1289 (2013); (10.1785/0120120136).
346
33.
D. D. Oglesby
, R. J. Archuleta, S. B. Nielsen, Earthquakes on dipping faults: The effects
347
of broken symmetry.
Science
280
, 1055
-‐
1059 (1998).
348
34.
C. B. Worden, M. C. Gerstenberger, D. A. Rhoades, D. J. Wald, Probabilistic
349
Relationships between Ground
-‐
Motion Parameters a
nd Modified Mercalli Intensity
350
in California.
Bulletin of the Seismological Society of America
102
, 204
-‐
221 (2012);
351
(10.1785/0120110156).
352
353
354
355
356
Figure 1:
Cumulati
ve
slip distribution and
static stress drop
due to the Gorkha
357
earthquake.
(A) Slip inversion
results for the Mw7.8 Gorkha event. The red star is the
358
hypocenter. Dashed contours are depths to the fault. Orange diamonds are 5 Hz
c
GPS
359
stations
and
white
diamonds
are low rate (1/30 Hz) stations.
The green triangle is the
360
strong motion station.
Kathma
ndu is represented by the blue square. The black arrows
361
indicate the coseismic offsets measured at the sites (
the values and uncertainties are given
362
in
Table S1)
. Vectors with less than 10cm displacement are not shown
(B) Static stress
363
drop predicted by th
e model of figure 1A. Green circles are aftershocks with local
364
magnitude >4 recorded and located by the
Nepal
National Seismic Center. Focal
365
mechanisms represent the GCMT moment tensors for aftershocks with magnitude larger
366
than 6
.
367
368
369
370
Figure 2:
Records o
f ground displacement
s
and acceleration
s
during the Gorkha
371
earthquake.
Displacement waveforms at
cGPS
stations KKN4 and NAST (5 samples per
372
second) and acceleration waveforms at strong motion station KATNP (figure 1).
373
374
375
376
Figure
3
:
Evidence for resonance
of Kathmandu basin.
(A)
-‐
(C) three components of
377
ground velocity observed at two high
-‐
rate GPS stations (KKN4 and NAST) and one strong
378
motion station (KATNP) in the Kathmandu region. KKN4 is located on hard rock northwest
379
of
Kathmandu
while the other 2 sta
tions are on soft sediment in the basin. The GPS is
380
differentiated to velocity and the strong motion integrated after high
-‐
pass filtering at 0.02
381
Hz. (D)
-‐
(F) Ground motion amplification observed at the two basin stations. Plotted is the
382
ratio of the amplit
ude spectra of the basin stations to the amplitude spectra of the
383
reference bedrock station KKN4. (G)
-‐
(I) 5% damped velocity response spectra for all 3
384
stations. (J) Close up map showing the location of the basin and
bedrock
stations.
385
386
387
388
Figure 4
:
Slip p
ulse kinematics during the Gorkha earthquake
(A) Snapshot of slip rate
389
on
Main Himalayan Thrust
at 27 s after origin time during propagation of the seismic
390
rupture from the model in figure 1. The red star is the hypocenter and dashed lines
391
represent the de
pth to the fault. The white circles are the centers of 5 subfaults used to
392
compare against theoretical regularized Yoffe source time functions
(
28
)
. (B) STFs at the 5
393
locations from (A). Plotted are the inverted slip rates and the regularized Yoffe function
s
394
measured from the vertical velocity at KKN4 scaled to the maximum observed slip rate at
395
each point which is indicated numerically. Time is relative to the hypocentral origin
396
(
28.147°N 84.708°E
;
2015
-‐
04
-‐
25 06:11:26.270 UTC
)
.
397
398
Supplementary Materials:
399
Materials and Methods
400
Figs. S1 to S
1
1
401
Tab. S1 to S2
402
Movie S1
403
References (3
5
–
45)
404
405