Journal of Asian Earth Sciences 90 (2014) 1–14
Contents lists available at ScienceDirect
Journal of Asian Earth Sciences
journal homepage: www.elsevier.com/locate/jseaes
Crustal shear-wave velocity structure beneath northeast India
from teleseismic receiver function analysis
Dipok K. Bora a,⇑, Devajit Hazarika b, Kajaljyoti Borah c, S.S. Rai c, Saurabh Baruah d
a
Department of Physics, Diphu Government College, Diphu, Karbi Anglong 782 462, Assam, India
Wadia Institute of Himalayan Geology, Dehradun 248 001, India
c
Seismic Tomography Division, CSIR-National Geophysical Research Institute, Uppal Road, Hyderabad 500 606, India
d
Geoscience Division, CSIR-North-East Institute of Science & Technology, Jorhat 785 006, Assam, India
b
a r t i c l e
i n f o
Article history:
Received 20 December 2013
Received in revised form 1 April 2014
Accepted 8 April 2014
Available online 20 April 2014
Keywords:
Shillong–Mikir plateau
Crustal shear-wave velocity structure
Receiver function
Low velocity zone
a b s t r a c t
We investigated the seismic shear-wave velocity structure of the crust beneath nine broadband seismological stations of the Shillong–Mikir plateau and its adjoining region using teleseismic P-wave receiver
function analysis. The inverted shear wave velocity models show 34–38 km thick crust beneath the
Shillong Plateau which increases to 37–38 km beneath the Brahmaputra valley and 46–48 km beneath
the Himalayan foredeep region. The gradual increase of crustal thickness from the Shillong Plateau to
Himalayan foredeep region is consistent with the underthrusting of Indian Plate beyond the surface collision boundary. A strong azimuthal variation is observed beneath SHL station. The modeling of receiver
functions of teleseismic earthquakes arriving the SHL station from NE backazimuth (BAZ) shows a high
velocity zone within depth range 2–8 km along with a low velocity zone within 8–13 km. In contrast,
inversion of receiver functions from SE BAZ shows high velocity zone in the upper crust within depth
range 10–18 km and low velocity zone within 18–36 km. The critical examination of ray piercing
points at the depth of Moho shows that the rays from SE BAZ pierce mostly the southeast part of the plateau near Dauki fault zone. This observation suggests the effect of underthrusting Bengal sediments and
the underlying oceanic crust in the south of the plateau facilitated by the EW-NE striking Dauki fault dipping 300 toward northwest.
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
The northeastern region (NER) of India is one of the most complex tectonic domain in the world which is manifested by the
ongoing India-Asia collision to the north and Indo-Burmese subduction to the east (Bilham and England, 2001; Angelier and
Baruah, 2009; Kayal et al., 2010, 2012). The region has experienced
two great earthquakes (Ms 8.7), the 12 June 1897 Shillong earthquake (Oldham, 1899) and the 15 August 1950 Assam earthquake
and more than 20 large (M > 7.0) earthquakes (Nandy, 2001; Kayal,
2008).
Several hundred earthquakes (M > 4.0) are recorded during the
past few decades in the region. In addition to the national network
run by the India Meteorological Department (IMD), the North-East
Institute of Science and Technology-Jorhat (NEIST-J), the National
Geophysical Research Institute-Hyderabad (NGRI-H), Geological
Survey of India (GSI) and several universities established
⇑ Corresponding author. Tel./fax: +91 03671272317.
E-mail address: [email protected] (D.K. Bora).
http://dx.doi.org/10.1016/j.jseaes.2014.04.005
1367-9120/Ó 2014 Elsevier Ltd. All rights reserved.
temporary and permanent analog networks since 1982. Since
2001 the network is upgraded to broadband digital stations with
global positioning systems (GPS) timing. The tectonic complexity
of the region is evidenced by the sustained seismicity as well as
surface geological features. The Shillong–Mikir Plateau, Brahmaputra valley, Indo-Burmese Subduction zone and Eastern Himalayan
Syntaxis are the distinctive geological units of the region.
A number of studies have been made on seismicity and seismotectonics of the region (Angelier and Baruah, 2009; Kayal et al.,
2006; Baruah et al., 2013; Bora et al., 2013). However, the subsurface structure of the distinctive geological units of the region is less
studied, which is crucial to understand the geodynamic of the
region. Prior to 2001, most of the studies (De and Kayal, 1990;
Kayal and Zhao, 1998; Rai et al., 1999; Sitaram et al., 2001;
Bhattacharya et al., 2005, 2008, 2010) were focused on the crustal
structure of the NER India in general and Shillong–Mikir Plateau
(SMP), in particular, using the analog data. Now, most of the seismic stations are upgraded to broadband digital stations. The recent
broadband waveform data accrued from these stations allowed
few researchers to adopt receiver function (RF) analysis technique
2
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
(Kumar et al., 2004; Mitra et al., 2005) to estimate the depth of
Moho. They observed a dipping Moho, reaching a depth of 48 km
to the north of SMP. Based on gravity model constrained by the
seismological data, Nayak et al. (2008) suggested that the Moho
beneath the Shillong Plateau exists at a shallow depth of about
35 km. Baruah et al. (2011) presented the Conrad (18–
20 ± 0.5 km) and Moho depth (30–35 ± 1.0 km) beneath the Shillong–Mikir Hills Plateau, estimated through travel time analysis.
Recently, Bora and Baruah (2012a), by inverting the travel-time
residuals of direct (P) and reflected Moho (PmP) phases, estimated
thinner Moho (35–38 km) beneath the Shillong Plateau and the
deepest Moho is found beneath the Brahmaputra valley to the
north at about 39–41 km depth, which is deeper by 4–5 km compared to the Shillong Plateau with simultaneous observation of
thinnest crust (33 km) in the western part of the Shillong Plateau
in the Garo Hills region.
In addition to these studies, an effort has been made for details
investigation of shear-wave velocity structure beneath the SMP
and adjoining region by applying RF analysis to the new data set
recorded at nine broadband seismographs (Fig. 1). The additional
dataset presented in this paper provides new information constraining shear-wave velocity structure beneath the study region.
2. Geology and tectonics
The NER India is jawed between the two arcs, the Himalayan
frontal arc to the north and the Indo-Burmese arc to the east and
is seismically very active. The Shillong Plateau (SP) in the northeast
India is a part of the Indian shield (Evans, 1964). Rajasekhar and
Mishra (2008) suggested that predominant E–W, N–S and NW–
SE oriented faults in the SP might be the results of complex tectonic forces from both the Himalayan collision zone and the
Indo-Burma subduction zone (Kayal, 2001). The long E–W trending
Dauki fault separates the Plateau to the north and the Bengal basin
to the south (Fig. 1). The Dauki fault demarcates the thick (18 km)
alluvium of the Bengal Basin from the Archaean gneiss of the Shillong Plateau, and the Dapsi-Thrust, a western segment of the Dauki
fault, on the other hand, demarcates the Tertiary meta-sediments
from the Archaean gneisses within the Plateau, and it is seismically
active (Kayal et al., 2012). Several studies exist on the seismicity
associated with the Dauki fault. The fault plane solutions of earthquakes in the close vicinity of the Dauki fault indicate a thrust fault
with an oblique strike (NE–SW) (Chen and Molnar, 1990). Kayal
et al. (2006) report a strike slip mechanism for two earthquakes
at depths of 48 and 55 km near the Dauki fault. The Brahmaputra
River to the north, separates the Shillong Plateau from the Himalaya, and is named as Brahmaputra river fault (Nandy, 2001). The
Mikir Plateau, a fragmented part of the Shillong Plateau, moved
to the northeast; separated by the 400 km long NW-SE trending
Kopili fault (Nandy, 2001). Based on the observation of intense
seismicity, fractal dimension, and b-value studies, the long Kopili
fault is identified as the seismically most active fault in the region
(Bhattacharya et al., 2002). The Assam Valley is an ENE-WSW
trending to the narrow valley, which lies between the Shillong Plateau and the eastern Himalayan tectonic domains. Further, the Plateau is delimated to the west by the NS Dhubri fault which
generated a large earthquake (M 7.1) in 1930. In addition to these
geologically as well as seismically mapped known tectonic faults,
Bilham and England (2001), based on geodetic and GPS data, identified a hidden fault at the northern boundary of the Plateau; they
named it Oldham fault (Fig. 1). They further proposed that the
great 1897 earthquake occurred by pop-up tectonics of the Plateau
between the south dipping Oldham fault and north dipping Dauki
fault by reversing faulting. Owing to regional tectonic stresses from
Fig. 1. Major tectonic features of the study region. The digital broadband seismic stations are shown by red triangles. The major tectonic features in the region are indicated:
Main Central Thrust, Main Boundary Thrust, Kopili Fault, Dauki Fault, DHF: Dudhnoi Fault; DT: Dapsi Thrust; OF: Oldham Fault; CF: Chedrang Fault; BS: Borapani Shear Zone.
Inset Map of India indicating the study region (box). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this
article.)
3
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
the Himalayan collision zone to the north and from the IndoBurma subduction zone to the east, the Shillong Plateau has a history of vertical uplift since Cretaceous and is characterized by a
large positive EW trending isostatic anomalies, with an average
Plateau elevation of about 1 km (Verma and Mukhopadhyay,
1977). Details of the seismicity and the seismotectonics of the
region are reviewed by several researchers (Chen and Molnar,
1990; Bilham and England, 2001; Rajendran et al., 2004; Kayal
et al., 2006, 2012; Kayal, 2008; Baruah and Hazarika, 2008;
Angelier and Baruah, 2009; Bora and Baruah, 2012a, 2012b; Bora
et al., 2013; Baruah et al., 2013).
3. Data
In this work, we use seismological data recorded in the period
from 2001 to 2009 by a network consisting of nine broadband stations in the SMP and its adjoining region (Fig. 1). These stations
are operated by different agencies/university viz. National Geophysical Research Institute (NGRI)-Hyderabad, North-East Institute of
Science and Technology (NEIST)-Jorhat, Indian Institute of Geomagnetism (IIG)-Mumbai and Gauhati University (GU). The stations are
equipped with CMG-3T/3ESP/Trillium-240 seismometers with REFTEK/Taurus data logger. Data were continuously recorded at 100
samples/s and Global Positioning System (GPS) receivers are used
for time synchronization. The seismograph at SHL, NGL, MND, and
TUR lie on the Precambrian crystalline basement exposures on the
Shillong Plateau; GAU, TZR, and JPA are on the low-lying basement
ridges in the Brahmaputra Valley, and BKD, RUP is located on bedrock in the Lesser Himalaya and a few kilometers south of the MCT
(Fig. 1). The TZR, JPA, NGL, BKD, and RUP seismic stations are operated by NGRI, Hyderabad; MND and TUR are operated by NEIST, Jorhat; SHL by IIG, Mumbai and GAU by Gauhati University. Details of
the stations and their locations are presented in Table 1.
We used over 1200 three-component seismograms from earthquakes of magnitude P 5.5 and epicenter distance within the
range 30–95°, with clear P-wave arrivals. The computation of RF
includes; decimation of the waveform data to 20 samples per second, filtering the data using Butterworth high pass filter with a corner frequency of 0.02 Hz, windowing the data to 150 s of the
waveform (30 s prior to P-wave arrival and 120 s post length),
removal of mean (D.C. effect) and trend (tilt in the base line) from
the seismograms etc. The selected earthquakes used in the study
are shown in Fig. 2. The earthquakes are mostly from NE and SE
directions.
4. Methodology
4.1. Receiver function analysis
We analyzed teleseismic P-wave RFs to obtain shear-wave
velocity structure of the crust. The teleseismic P-wave coda
contains S-waves generated by P-to-S (Ps) conversions at Moho discontinuity below the seismograph site. The RF is a time series computed by deconvolving the vertical component waveform from the
radial and tangential component waveforms to isolate the receiver
site effect from other information contained in the teleseismic Pwave (Langston, 1977, 1979). The detail crustal structure beneath
a recording site can be obtained by modeling the amplitude and
timing of these converted phases (Ps) and their reverberations
(Owens et al., 1984; Ammon et al., 1990). We have computed
RFs of teleseismic earthquakes recorded by nine broadband seismological stations using the iterative time domain deconvolution
technique of Ligorria and Ammon (1999). This technique is based
on least square minimization of the difference between the
observed horizontal and a predicted signal generated by convolution of an iteratively updated spike sequences with vertical component seismogram. The Gaussian filter of width a = 1.6 s is used for
RF computation. The RFs with good signal to noise ratio (SNR) and
more than 80% waveform fit are used for further analysis. The RFs
are grouped by observations with similar backazimuth (BAZ) and
epicentral distances and then stacked to increase SNRs.
Fig. 3 shows the RFs plots for the nine stations with respect to
the BAZ. The individual RFs at each station show clear and consistent Moho converted Ps phase at all BAZ, however strength of Ps
phase and intra-crustal phases vary with BAZ in few stations. The
seismic stations NGL, SHL, TUR, and MND located on Shillong Plateau shows clear Ps phase at 4–4.5 s, whereas the three stations
located over the Brahmaputra valley (JPA, GAU and TZR) shows
Ps phase at 5 s. In JPA station, clear Ps is observed at 5 s, however their amplitudes are less corresponds to the other stations.
The amplitudes of the Moho Ps phase in RF depend on the incidence angle (and hence epicenter distance) at Moho boundary
and the size of the velocity contrasts generating the conversions
(Ps) (Ammon, 1991). The RFs of BKD and RUP shows Ps phase at
7 and 6 s, respectively. The RFs of BKD stations show 1 s delay
of the first P arrival followed by positive amplitude at 3.5 s. BKD
station sits in the foothills of the Himalayan convergence zone.
RFs for this station has large conversions between 0.5 and 1 s
which correspond to the conversions from the thick sediments
beneath BKD. Due to this sedimentary thickness variations along
with the complex crustal pattern beneath BKD station make Ps
more complicated. The crustal multiples of Ps phase e.g. PpPms is
observed at TZR, SHL, TUR, and MND around 14–17 s while it is
not very clear at JPA, BKD, and GAU stations. Most of RFs at individual stations show consistent delay time of Ps phase with respect to
direct P-wave arrival. The NGL station, however, shows azimuthal
variation of delay time of Ps phase which is 4.5 s and 4.2 s for the
earthquakes arriving from the NE and SE BAZ, respectively (Fig. 3).
This station does not show any intra-crustal layer. On the other
hand, the RFs of SHL station show a positive arrival at 2 s for
the earthquakes arriving from NE BAZ and a negative arrival at
3 s for the earthquakes from SE BAZ. The variation of crustal
structure beneath SHL station is discussed in later section. The
Table 1
Locations of seismographs operated in our study region and number of receiver functions (RFs) used in our study.
Tectonic blocks
Station
Station code
Latitude (°N)
Longitude (°E)
Elevation (m)
No. of RFs used
Shillong Plateau
Tura
Nangalbibra
Manikganj
Shillong
TUR
NGL
MND
SHL
25.546
25.472
25.924
25.566
90.243
90.702
90.676
91.859
305
330
40
1590
21
69
14
83
Assam Valley
Jogighopa
Gauhati University
Tezpur
JPA
GAU
TZR
26.239
26.152
26.617
90.575
91.667
92.783
42
69
140
54
11
85
Lesser Himalaya
Bharabkunda
Rupa
BKD
RUP
26.890
27.203
92.115
92.401
210
1470
15
74
4
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
Fig. 2. Epicenter locations of the earthquakes (red circles), of magnitude more than 5.5, recorded by our stations (black star). (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of this article.)
Table 2
Comparison of Moho depth from different studies.
Tectonic blocks
Moho depth (km)
Methodology
References
Shillong Plateau
35 ± 1
35
35 ± 2
37–38
30 ± 1
35–38) ± 0.5
(34–38) ± 2
Receiver function analysisa
Receiver function analysisa,b
Gravity modeling
Gravity modeling
Travel-time analysis
Inversion of travel time residuals
Receiver function analysisb,c
Kumar et al. (2004)
Mitra et al. (2005)
Nayak et al. (2008)
Rajasekhar and Mishra (2008)
Baruah et al. (2011)
Bora and Baruah (2012a)
Present study
Assam Valley
35 ± 3
41
42
(36–37) ± 0.5
(39–41) ± 0.5
(37–38) ± 2
Receiver function analysisa
Receiver function analysisa,b
Gravity modeling
Inversion of travel time residuals
Kumar et al. (2004)
Mitra et al. (2005)
Nayak et al. (2008)
Bora and Baruah (2012a)
Receiver function alalysisb,c
Present study
42–47
(46–48) ± 2
Receiver function alalysisa
Receiver function alalysisb,3
Kumar et al., 2004
Present study
Lesser Himalaya
a
b
c
H-k stacking approach (Zhu and Kanamori, 2000).
Inversion Algorithm (Herrmann and Ammon, 2004).
Neighborhood Algorithm (Sambridge, 1999).
intra crustal positive arrival at 2.5 s is also visible in the RFs of
TZR and RUP.
4.2. Modeling of receiver functions
Considering the fact that radial RF is sensitive to subsurface Swave velocity, the crustal velocity structure beneath recording stations is determined by inverting the radial RF waveforms using a
non-linear and derivative free direct search technique, neighborhood algorithm (NA) of Sambridge (1999). The non-linearity and
non-uniqueness are the two main problems in RF inversion
(Ammon et al., 1990), which can be overcame using NA. In NA
modeling of RF, the Earth is divided into six horizontal layers; Sediment, Basement, upper, middle, lower crust and upper mantle.
The model comprises of four parameters in each layer; the thickness of the layer (km), S-wave velocity at the topmost and the bottommost point of the layer (km/s), and Vp/Vs ratio in that layer,
making a 24 (=6 4) dimensional parameter space. The NA inversion is applied to the receiver function stacked in narrow epicentral distance and BAZ ranges. The objective function for the
measure of waveform misfit is the L2 norm of the difference
between observed receiver function and synthetic receiver function obtained from a model. Each inversion run involved 2000 iterations, generating 40,020 velocity models. Stability of the inversion
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
5
Fig. 3. Receiver functions of the nine stations in increasing backazimuth plotted with equal spacing. Black dashed lines are drawn at every 5 s interval. Moho P-to-S
conversion (Ps) is marked by red dashed line. Station name is shown at the top right corner of each plot. For each station, right panel shows the backazimuth (°) (red circle)
and epicentral distance (°) (blue circle) values for individual earthquake plotted in respective left panel. (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
solutions was tested using a large range of initial random seeds,
incidence angles, and velocity model parameterizations. Vertical
depth resolution from the RF depends on the Gaussian width at
which RF was created. For a Gaussian Width (GW) of 1.6 s corresponding to a seismogram low-pass filtered to 0.8 Hz and average
Vs of 3.7 km/s, the minimum wavelength is 4.63 km leading to
minimum resolvable layer thickness of 1.15 km.
To illustrate the modeling procedure, we briefly discussed the
inverse and forward modeling of RFs of JPA station (Figs. 4 and
5). Fig. 4a shows example of RFs with small bin of epicentral dis-
tance (DEL: 43–45°) and BAZ (60–67°) along with the stacked RF
with ±1r (standard deviation) bounds. These bounds are determined from the variance of the stacked data and are used to check
the fitting of RF in inversion modeling. About 40,020 velocity models (Fig. 4b) have been generated using the NA algorithm
(Sambridge, 1999) and synthetic RFs are generated corresponding
to each model and compared with the original stacked RF. Based
on waveform matching between the synthetic and observed RFs,
1000 best models (colored section) are selected and shown in
Fig. 4b. The thick red and black lines in Fig. 4b correspond to the
6
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
Fig. 4. Inversion results using Neighborhood algorithm (NA) for selected station JPA. (a) Receiver functions in a narrow epicenter (DEL) and backazimuth (BAZ) range as
mentioned in the top right corner with their stacked receiver function with ±1r bounds. (b) The gray regions are the range of models (40020 models) searched to find the best
one. The yellow and green regions indicate the best 1000 models with least error between observed and computed RF. The best fitting and average model is indicated by the
red and black lines, respectively, that overlies the model density plot. The red and black lines (in the left side) are the best fitting and average Vp/Vs ratio model. Station code is
mentioned in the bottom right. (c) Synthetic receiver function (red line) obtained from NA inversion with ±1r bounds (black dashed line). (For interpretation of the references
to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 5. Forward modeling results for the JPA station receiver function. The match of the observed (thin black line) and synthetic receiver function from inversion (red line) and
forward modeling (blue line) with ±1r bounds (black dashed line) shown on the left (a, b, c) calculated for the velocity model shown on the right (d, e, f). (a, b) Inversion (red
line) and forward (blue line) synthetic RF for the thin inverted (red line) and forward model (blue line) plotted in (d, e). (c) Test to determine bounds on the Moho depth
(±2 km) plotted in models (f). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
best fitting and the average Vs and Vp/Vs, respectively. Fig. 4c
shows the fit of the observed (stack) RF with the one computed
from the best predicted model with error bounds.
To map the Moho as well as other intra-crustal layers, we used
the velocity–depth section for individual station. Christensen and
Mooney (1995) and Christensen (1996) show that the shear-wave
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
7
velocity in the lower crust usually do not exceed 4.3 km/s and
wave velocities exceeding this value indicate the presence of lithology of mantle composition. Therefore, the crustal thickness at most
of the stations was determined by placing Moho at the depth
where the shear-wave velocity exceeds 4.3 km/s. However, at
few stations (e.g. JPA and GAU) shear wave velocity at lower crust
is less than 4.3 km/s (Figs. 5 and 12). In such case, we apply forward modeling technique of Herrmann and Ammon, 2004 to find
simpler crustal structure and mark Moho at the depth where there
is a step jump in shear-wave velocity. Forward modeling is also
applied to those stations showing any complexities like azimuthal
variation in shear-wave velocity structure (e.g. SHL station) or gradation jump of shear wave velocity at Moho (e.g. RUP station). The
forward modeling scheme is illustrated for RF data of JPA station
(Fig. 5). Fig. 5d shows a step jump in shear-wave velocity (Vs
4.1 km/s) at a depth of 37 km, which fit the RF very well
(Fig. 5a). This step jump in velocity at 37 km may indicate Moho
beneath the JPA station. The forward modeling is repeated making
Vs > 4.3 km/s at Moho (Fig. 5e) that fit the RF but their amplitude is
outside the error bounds (Fig. 5b) suggesting the Moho Vs should
not be greater than 4.1 km/s. We also vary the Moho depth by
±2 km (Fig. 5f) and compare the synthetic RF with the observed
RF to determine the bounds of the Moho depth. The test shows that
the error in Moho depth is ±2 km (Fig. 5c). All other station data
were analyzed in the same manner as JPA and their results are discussed below.
5. Results
5.1. Geometry of Moho across NS profile
In order to compare the features of the crust, we plot individual
RFs for the stations along the N–S profile as shown in Fig. 6. These
RFs were corrected for the distance moveout of the Moho Ps phase
referenced to 67° epicentral distances. The stacked RF is presented
in the left panel. The individual RFs are ordered, first according to
the station location and second according to the BAZ and plotted
equispaced in the right panel (Fig. 6). The dominant feature in
the RFs observed in the transect is the gradual increase in the delay
time of Moho Ps phase with respect to direct P-wave arrival. The
delay time of Ps phase varies from 3.5 to 4.0 s at the stations of
the Shillong Plateau (NGL, TUR, MND, and SHL) to 4.5–5.0 s
observed at the stations of Brahmaputra valley (GAU, JPA, and
TZR). This delay time further increases to 6.0–7.0 s beneath the
Lesser Himalayan stations (RUP, BKD). The increase in delay time
along the NS profile is clearly observed in the RFs.
We discuss the velocity models with reference to continental
crust of average thickness 40 km (Christensen, 1996; Rudnick
and Gao, 2003) divided into upper, middle and lower crust corresponding to depth interval of 0–11, 11–23, and 23–40 km with
the corresponding Vs of 2.8–3.5, 3.5–3.8, and 3.8–4.1 km/s, respectively. Usually Vs 4.0 km/s and above is considered as the representative for the basal layer and the Moho is defined by a jump in
seismic wave velocity to values greater than 7.6–8.0 km/s for Pwave and 4.3–4.6 km/s for S-wave. The results obtained from RF
modeling at different tectonic units are discussed below.
5.2. Shillong Plateau
The data from MND, TUR and NGL stations samples the western
part of the Shillong Plateau, whereas, data from SHL station samples the eastern part of the plateau. The inversion of a few selected
RFs of teleseismic earthquakes (DEL: 60–67°, BAZ: 106–110°)
recorded at MND station shows a simple crust with average Vs of
3.4 km/s and Moho at 34 km (Fig. 7, upper panel). Inversion mod-
Fig. 6. Plots of receiver functions along profiles N–S profile. In each figure, left panel
shows the stacked receiver functions plotted as a function of relative distance along
profile, and right panel shows individual traces used to generate the stacked
receiver function. Individual traces are ordered by backazimuth from south (lower
trace) to north (upper trace) positioned along the profile. P and Ps phases are
marked by thick lines.
eling of TUR (Fig. 7, middle panel) shows Moho at a depth of
37 km with average Vs of 3.6 km/s. Inversion results for NGL station, using six RFs show simple crustal model with Moho at 38 km
and average Vs of 3.7 km/s (Fig. 7, lower panel). For this station
middle crust is thicker (28 km) than the MND and TUR station.
The SHL station shows strong local shear-wave velocity variation in the crust. The RFs of NE BAZ (45–50°) shows a positive arrival at 2 s (Fig. 8, top left panel). Inversion modeling of stacked RFs
shows a high velocity layer of thickness 6 km within 2–8 km depth
range followed by low velocity layer. Moho is observed at a depth
of 36 km with average Vs of 3.8 km/s. The model shows a 15 km
thick basal layer (Vs P 4.0 km/s) within the depth range 21–36 km.
Forward model for that inversion also shows 6 km thick high
velocity followed by 4 km thick low velocity layer (Fig. 9c). The
upper most mantle velocity is 4.5 km/s within the depth range
36–46 km. A decrease in mantle velocity is observed beyond
46 km depth (Vs 4.24 km/s).
Stacked RFs from NW BAZ (BAZ: 298–299°, DEL: 59°) have two
positive peaks at 1.5 s and 4 s followed by Moho Ps phase at
5 s (Fig. 8, top right panel). The inversion shows Moho at 34 km
depth with average Vs of 3.65 km/s. A 15 km thick middle-crust
and a 17 km thick lower-crust are observed with no corresponding
basal layer in the inverted model. RFs from SE BAZ (110–114°)
show clean Ps and PpPms phase with no intracrustal layer (Fig. 8,
lower left panel). Inversion modeling shows average Vs of
3.85 km/s, 29 km lower crust with a 24 km thick basal layer. Moho
is observed to be at 35 km depth. We observed typical RFs from
158° to 162° BAZ (Fig. 8, bottom right panel). The stacked RF show
a negative phase at 2.5 s and a strong Ps amplitude at 4 s. To
8
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
check whether the observed negative amplitude is actual converted phase or a reverberation of some initial phase before 2.5 s,
we use higher Gaussian width (Gw 5.0) and repeated the inversion
as shown in Fig. 10. Both the inversion results (Fig. 8, bottom right
panel and Fig. 10) and forward modeling (Fig. 9d) show high
velocity of Vs 4.1 km/s at a depth of 10 km followed by low velocity of thickness 16 km at a depth of 20 km. Average Vs of
3.7 km/s and thick mid crust with no lower crust are observed
in the inversion modeling. Forward modeling result shows Moho
at 36 km (Fig. 9d).
5.3. Brahmaputra Valley
Brahmaputra valley is sampled by three stations, JPA, GAU, and
TZR. The RFs for TZR station show 2.5 s positive peak, followed by
5 s Moho Ps. The inversion modeling shows Moho at 38 km (Fig. 11,
left panel) and average Vs of 3.4 km/s. The upper crust beneath this
station is thicker (26 km) than middle (5 km) and lower crust
(7 km). For JPA station, the inversion modeling shows, a Moho
depth of 38 km with an average Vs of 3.3 km/s and a 30 km thick
upper crust. For GAU station, the data is limited. The right panel
of Fig. 11 shows the stack of the two RFs from the 32–33° BAZ
which has Ps at 4.5 s. The inversion modeling shows similar crustal
structure as that in JPA. Forward modeling shows Moho at 38 km
depth (Fig. 12). The RF modeling results for all the stations of Brahmaputra Valley do not show any basal layer.
5.4. Himalayan foredeep region
The two seismological stations, BKD and RUP sample the crust
beneath the Lesser Himalaya of north east India. The individual
stacked RF at BKD station shows 1 s delay of first P-wave arrival
followed by positive peaks around 3.5 s and 7.0 s (Fig. 13a, left
panel). The inverted model shows 4 km thick sedimentary layer
(Vs < 2.8 km/s) beneath the station followed by a high velocity
(3.7 km/s) layer within the depth range 4–10 km. The Moho is
observed at a depth of 48 km. Average velocity of the crust
beneath BKD is 3.2 km/s. Crustal structure beneath the RUP station
is sampled by the RFs from 299° to 301° BAZ. Stacked RF shows two
positive peaks at 2 and 3.5 s and Ps phase at 6 s. Inversion (Fig. 13,
right panel) and forward modeling (Fig. 14) shows 6 km thick sedimentary layer (Vs < 2.8 km/s), 18 km thick upper and middle crust,
and average Vs of 3.4 km/s. Lower crust observed to be thin (4 km)
and hence the basal layer (4 km). Forward modeling shows Moho
depth at 46 km (Fig. 14) beneath RUP station.
6. Discussions
6.1. Crustal thickness
Fig. 7. Inversion results using Neighborhood algorithm (NA) for the stations from
Shillong Plateau. Description of the figure is same as Fig. 4.
The crustal thickness obtained at nine broadband seismic stations of the Shillong–Mikir plateau and its adjoining region provides important information related to the characteristics of the
crust beneath the region. The inverted shear wave velocity models
show 34–38 km thick crust beneath Shillong Plateau which
increases to 37–38 km beneath the Brahmaputra valley and
46–48 km beneath the Himalayan foredeep region. The gradual
increase of crustal thickness from the SP to the Himalayan foredeep region is consistent with the underthrusting of the Indian
plate beyond the surface collision boundary as observed by several
studies (Nayak et al., 2008; Rajasekhar and Mishra, 2008; Mitra
et al., 2005; Kumar et al., 2004; Devi et al., 2011). The estimation
of crustal thickness beneath the seismic stations of the present
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
9
Fig. 8. Inversion results using Neighborhood algorithm (NA) for station SHL for different backazimuth ranges. Description of the figure is same as Fig. 4.
study is also in agreement with results obtained from gravity modeling (Nayak et al., 2008; Rajasekhar and Mishra, 2008), previous
receiver function analysis (Kumar et al., 2004; Mitra et al., 2005)
and inversion of travel time residuals (Bora and Baruah, 2012a).
However, travel time study by early analog data suggested comparatively higher crustal thickness (45–49 km) beneath the SP
(Rai et al., 1999). On the other hand, Moho depth beneath the Brahmaputra valley obtained through travel time analysis of Pg, P* and
Pn waves recorded on analog seismograph (Sitaram et al., 2001)
shows comparable values of crustal thickness with the present
study (Table 2). Based on gravity data and broadband seismological
data, Nayak et al. (2008) suggest that the Moho discontinuity is
inferred to be 40 km below the Bengal basin, to the south of Dauki–Dapsi Thrust, 35 ± 2 km below the Shillong Plateau and
42 ± 2 km below the foredeep/Assam shelf, south of the MBT.
Therefore the Moho beneath the Plateau is shallower by 5 km
in contrast to its immediate south and north. Thinner crust and
higher elevation of the Shillong Plateau, where the Bouguer gravity
anomaly has a small positive value of 20–40 mGal (Verma and
Mukhopadhyay, 1977; Gaur and Bhattacharya, 1983; Das Gupta
and Biswas, 2000), supported the model by Bilham and England
(2001) that the Plateau is formed by slip on two reverse faults,
the south-bounding Dauki and a north-bounding Oldham Fault.
The high gravity anomaly over this plateau may not be entirely
attributed to the absence of sedimentary overburden and hence
propose a high density layer emplaced in the crust beneath the
Shillong Plateau (Gokarn et al., 2008). The RF analysis (Mitra
et al., 2005) delineate about 10 km thick high velocity crust
beneath the Shillong Plateau (S-wave velocity 4 km/s as against
3.6 km/s in the surrounding region), which seems to support this
conjuncture. Thinning of the crust beneath the Shillong Plateau
seems to have evolved by a cyclic process of erosion and uplifting
(Molnar and England, 1990). A deeper depth of Moho discontinuity
as well as lower crust is observed below the foredeep of the Himalaya. This structural deformation may be associated with a flexure
of the down going Indian plate below the Himalayas.
10
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
Fig. 9. Inversion and forward modeling results for station SHL. (a, b) Inversion (red line) and forward (blue line) synthetic RF for the thin inverted (red line) and forward
model (blue line) plotted in (c, d) for two backazimuth ranges. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of
this article.)
Fig. 10. Inversion results using Neighborhood algorithm (NA) for station SHL from south-east backazimuth. Description of the figure is same as Fig. 4.
6.2. Lateral variation of the crust
The broadband seismological stations recorded sufficient number of teleseismic earthquakes arriving from wide range of BAZ
that facilitates study of the azimuthal variation of crustal structure
beneath the study region. Critical examination of converted phases
of teleseismic waves arriving from different BAZ at individual station have been carried out which shows identical crustal structure
at most of the stations observed from different BAZ. However, the
SHL station shows significant variation in intra-crustal shear-wave
velocity structure as observed in inverted velocity models obtained
from RFs of NE and SE BAZ. The shear wave velocity model
obtained from RF of teleseismic wave arriving the SHL station from
NE shows a high velocity zone within depth range 2–8 km along
with a low velocity zone within 8–13 km (Fig. 9). This observation agrees with the inversion results (unsimplified model) at
BPN station (located northeast of SHL station) carried out by
Mitra et al. (2005). In contrast, inverted and forward models
obtained from the RFs of teleseismic waves arriving from SE shows
high velocity zone in the upper crust within 10–18 km and low
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
Fig. 11. Inversion results using Neighborhood algorithm (NA) for the stations from Assam Valley. Description of the figure is same as Fig. 4.
Fig. 12. Inversion and forward modeling results for station GAU. Description of the figure is same as Fig. 9.
Fig. 13. Inversion results using Neighborhood algorithm (NA) for the stations from Lesser Himalaya. Description of the figure is same as Fig. 4.
11
12
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
Fig. 14. Inversion and forward modeling results for station RUP. Description of the figure is same as Fig. 9.
Fig. 15. (a): Locations of the Moho Ps piercing points (filled circles) for a few selected RFs recorded at SHL station (blue triangles) from different BAZ. The blue, orange and
pink filled circles represent Moho piercing points of earthquakes arriving the station from NW, NE, and SE respectively. Similarly, the piercing points of the same earthquakes
are estimated at 18 km depth. (b) All the piercing points are projected over the geoelectric section in the region obtained from previous study (Gokarn et al., 2008).
velocity zone within 18–36 km. In order to understand the origin
of the azimuthal variations beneath SHL station, the ray piercing
points have been estimated at the depth of Moho (38 km) by
ray tracing the incident Ps phase through the crust, using the ray
parameter calculated from the focal depth and epicentral distance
of the earthquake (Fig. 15a). The ray piercing points of the same
earthquakes are also estimated at 15 km depth. It is observed
that when the teleseismic waves arrives the station from SE BAZ,
piercing points at Moho mostly shuttle southern part of the plateau
near the Dauki fault zone. The low velocities observed in modeling
of RFs arriving from SE BAZ may be the effect of Bengal sediments
and the underlying oceanic crust beneath SE Shillong Plateau
(Gokarn et al., 2008). Magnetotelluric (MT) studies over the Shillong Plateau and lower Brahmaputra sediments have delineated
the Dauki fault as a NE-SW striking thrust zone with a dip angle
of about 30° along which the low resistivity layer of Bengal sediments and the underlying oceanic crust subduct toward the northwest (Gokarn et al., 2008). The piercing points for earthquakes of
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
SHL station estimated at 38 km and at 15 km are projected over
the geoelectric section obtained by Gokarn et al. (2008) (Fig. 15b).
A low resistivity is observed beneath region shuttled by rays piercing from SE BAZ. However, inverted model obtained from earthquakes arriving from NE BAZ pierces the low resistive zone (zone
‘‘G’’ as shown by Gokarn et al., 2008), but could not detect the
low velocity zone. Projection of piercing points corresponding to
depth of 18 km over the geoelectric section (Fig. 15b) clearly indicates that the waves arriving from SE pierce the extremely high
resistive zone that also shows high shear-wave velocity. This high
resistive and high velocity zone may indicate effect crystalline
rocks of the SP. Other geophysical study also detected mid-crustal
discontinuity at 18 ± 0.5 km beneath the Shillong Plateau (Bora and
Baruah, 2012b).
Low resistivity layer in the crust is usually explained by the
presence of partial melts, aqueous fluids or other conducting materials like graphite (Gokarn et al., 2002; Harinarayana et al., 2004;
Arora et al., 2007). The MT studies cannot distinguish between
these causing factors that also cause low shear-wave velocity. In
contrast, seismic observations detect fluid reach zones as low
shear-wave velocity zone. The low shear-wave velocity for waves
arriving from SE BAZ may be due to presence of underthrusting
sedimentary layer of the Bengal Basin along the Dauki fault.
The piercing points are also estimated at SHL station at shallow
depth (at 15 km depth) to examine the upper-crustal features.
The projection of these piercing points at 15 km depth over geoelectric section of Gokarn et al. (2008) shows that the piercing
points of earthquakes arriving from NE direction falls over the
low resistive zone (Zone ‘‘E’’ as shown in Fig. 8 of Gokarn et al.,
2008). Modeling of these earthquakes shows low velocity zone in
the depth range 8–13 km (Fig. 9c). The low resistivity and low
velocity zones as observed in the geophysical studies suggest the
presence of fluid rich zone. On the other hand, piercing points estimated at a depth of 15 km for teleseismic waves arriving from SE
BAZ fall over the extremely high resistive zone (Zone ‘‘F’’ with
resistivity 1000 ohm m, as shown in Fig. 8 of Gokarn et al.,
2008). The modeling of these RFs shows high velocity zone within
10–18 km depth. This high velocity zone may also be correlated
to the observation of high density with the help of gravity data
beneath the Shillong Plateau, based on the isostatic considerations
(Verma and Mukhopadhyay, 1977). In contrast to the observation
in SHL station, no such variation is observed beneath the stations
TUR and NGL located at the southernmost part of the plateau. This
indicates significant underthrusting of Bengal sediments at the
segment of the Dauki fault zone, where there is a significant
change in strike of the fault from EW to NE. Surface expression
in this part of the plateau also suggests that Bengal sediments
and underlying crust from south of the Shillong Plateau subducts
along NW dipping thrust. The high velocity in the upper crust
observed in the southern part of the plateau and the high density
observed by gravity survey (Verma and Mukhopadhyay, 1977) in
conjunction with basaltic nature of the crust beneath the Bangal
sediments (Baksi, 1965; Sengupta, 1966) clearly suggest oceanic
nature of the crust beneath the thrust zone (Gokarn et al., 2008).
7. Conclusions
The investigation of shear-wave velocity structure beneath the
Shillong–Mikir plateau and its adjoining region with the help of
modeling of receiver functions provides important characteristics
of the underlying crust. The study reveals variations of shear-wave
velocity and Moho depth beneath the region. The crustal thickness
is 34–38 km thick beneath the Shillong Plateau which increases
to 37–38 km beneath Brahmaputra valley and 46–48 km
beneath the Himalayan foredeep region suggesting underthrusting
13
of the Indian Plate beneath the Eurasian plate to the north. Strong
local shear-wave velocity variation beneath SHL station is
reported. The shear-wave velocity model obtained from RFs of
earthquakes arriving to the seismic stations from SE shows a low
velocity zone in the lower crust which is not observed in case of
the earthquakes arriving from NE. The corresponding piercing
point analysis shows that these waves pierce the southeastern part
of the plateau. The most possible explanation of the low velocity
zone is the effect of Bengal sediments and the underlying oceanic
crust in the south of the plateau subducting along the NW dipping
thrust with the surface expressions near the Dauki fault.
Acknowledgements
Data obtained from NEIST, Jorhat; NGRI, Hyderabad; IIG, Mumbai and Gauhati University are acknowledged. The first author
(D.B.) thanks Indian Academy of Sciences, Bangalore, for the Summer Research Fellowship, 2012 to carry out the research and expedite the work at Seismic Tomography Division, NGRI, Hyderabad.
The authors (D.B. and D.H.) are thankful to the Director, Wadia
Institute of Himalayan Geology (WIHG), Dehradun, India, for giving
permission to carry out this collaborative research work. Two
anonymous reviewers and Associate Editor Prof. Dapeng Zhao are
gratefully acknowledged for their valuable comments and suggestions which upgrade the manuscript significantly.
References
Ammon, J.C., 1991. The isolation of receiver effects from teleseismic P waveforms.
Bull. Seismol. Soc. Am. 81 (6), 2504–2510.
Ammon, J.C., Randal, G.E., Zandt, G., 1990. On the nonuniqueness of receiver
function inversions. J. Geophys. Res. 95, 15,303–15,318.
Angelier, J., Baruah, S., 2009. Seismotectonics in Northeast India: a stress analysis of
focal mechanism solutions of earthquakes and its kinematic implications.
Geophys. J. Int. 178, 303–326.
Arora, B.R., Unsworth, M.J., Rawat, G., 2007. Deep resistivity structure of the
northwest Indian Himalaya and its tectonic implications. Geophys. Res. Lett. 34.
http://dx.doi.org/10.1029/2006GL029165.
Baksi, S.K., 1965. Stratigraphy of the Barail series in the southern part of the Shillong
Plateau, Assam, India. Am. Assoc. Pet. Geol. Bull. 49, 2282–2294.
Baruah, S., Hazarika, D., 2008. A GIS based Tectonic Map of India. Curr. Sci. 95, 176–
177.
Baruah, S., Bora, D.K., Biswas, R., 2011. Estimation of crustal discontinuities from
reflected seismic waves recorded at Shillong and Mikir Hills Plateau, Northeast
India. Int. J. Earth Sci. 100, 1283–1292.
Baruah, S., Baruah, S., Kayal, J.R., 2013. State of tectonic stress in Northeast India and
adjoining South Asia region: an appraisal. Bull. Seismol. Soc. Am. 103 (2A), 894–
910.
Bhattacharya, P.M., Majumdar, R.K., Kayal, J.R., 2002. Fractal dimension and b-value
mapping in northeast India. Curr. Sci. 82 (12), 1486–1491.
Bhattacharya, P.M., Pujol, J., Mazumdar, R.K., Kayal, J.R., 2005. Relocation of
earthquakes in the Northeast India region using joint Hypocenter
determination method. Curr. Sci. 89 (8), 1404–1413.
Bhattacharya, P.M., Mukhopadhyay, S., Mazumdar, R.K., Kayal, J.R., 2008. 3-D
seismic structure of the northeast India region and its implication for local and
regional tectonics. J. Asian Earth Sci. 33, 25–41.
Bhattacharya, P.M., Kayal, J.R., Baruah, S., Arefiev, S.S., 2010. Earthquake source
zones in northeast India: seismic tomography, fractal dimension and b value
mapping. Pure Appl. Geophys.. http://dx.doi.org/10.1007/s00024-010-0084-2.
Bilham, R., England, P., 2001. Plateau ‘‘pop-up’’ in the Great 1897 Assam earthquake.
Nature 410, 806–809.
Bora, D.K., Baruah, S., 2012a. Mapping the crustal thickness in Shillong–Mikir Hills
Plateau and its adjoining region of northeastern India using Moho reflected
waves. J. Asian Earth Sci. 48, 83–92.
Bora, D.K., Baruah, S., 2012b. Depth of mid-crustal discontinuity from reflected
seismic waves on local earthquake seismograms recorded at Shillong Plateau,
Northeast India. Geomatics, Nat. Hazards Risk 3 (4), 355–364.
Bora, D.K., Baruah, S., Biswas, R., Gogoi, N.K., 2013. Estimation of source parameters
of local earthquakes originated in Shillong–Mikir plateau and its adjoining
region of Northeastern India. Bull. Seismol. Soc. Am. 103 (1), 437–446.
Chen, W.P., Molnar, P., 1990. Source parameters of the earthquakes and Interplate
deformation beneath the Shillong Plateau and Northern Indo-Burman ranges. J.
Geophys. Res. 95 (B8), 12527–12552.
Christensen, N.I., 1996. Poisson’s ratio and crustal seismology. J. Geophys. Res. 101,
3139–3156.
Christensen, N.I., Mooney, W.D., 1995. Seismic velocity structure and composition
of the continental crust: a global view. J. Geophys. Res. 100, 9761–9788.
14
D.K. Bora et al. / Journal of Asian Earth Sciences 90 (2014) 1–14
Das Gupta, A., Biswas, A.K., 2000. Geology of Assam, vol. 170. Geological Society of
India, Bangalore, India.
De, R., Kayal, J.R., 1990. Crustal P-wave velocity and velocity–ratio study in
Northeast India by a microearthquake survey. Pure Appl. Geophys. 134, 93–108.
Devi, E.U., Kumar, P., Kumar, M.R., 2011. Imaging the Indian lithosphere beneath the
Eastern Himalayan region. Geophys. J. Int. 187, 631–641.
Evans, P., 1964. Tectonic framework of Assam. J. Geol. Soc. India 5, 80–96.
Gaur, V.K., Bhattacharya, J.C., 1983. Gravimetric determination of the shape of
Mohorovicic discontinuity in Peninsular and north-eastern India. Paper
presented at I.U.G.G. General Assembly at Hamburg, Germany, August 15–27.
Gokarn, S.G., Gupta, G., Rao, C.K., Selvaraj, C., 2002. Electrical structure across the
Indus Tsangpo suture and Shyok suture zones in NW Himalaya using
magnetotelluric studies. Geophys. Res. Lett. 29. http://dx.doi.org/10.1029/
2001GL014325. issn: 0094- 8276.
Gokarn, S.G., Gupta, G., Walia, D., Sanabam, S.S., Hazarika, N., 2008. Deep geoelectric
structure over the Lower Brahmaputra valley and Shillong Plateau, NE India
using magnetotellurics. Geophys. J. Int. 173, 92–104.
Harinarayana, T., Abdul Azeez, K.K., Naganjaneyulu, K., Manoj, C., Veeraswamy, K.,
Murthy, D.N., Rao, Prabhakar Eknath., Prabhakar Eknath Rao, S., 2004.
Magnetotelluric studies in Puga valley geothermal field, NW Himalaya,
Jammu and Kashmir, India. J. Volcanol. Geothermal Res. 138, 405–424.
Herrmann, R.B., Ammon, C.J., 2004. Surface waves, receiver functions and crustal
structure, in Computer Programs in Seismology, Version 3.30, Saint Louis
University.
Kayal, J.R., 2001. Microearthquake activity in some parts of the Himalaya and the
tectonic model. Tectonophysics 339, 331–351.
Kayal, J.R., 2008. Microearthquake Seismology and Seismotectonics of South Asia.
Springer, pp. 273–275.
Kayal, J.R., Zhao, D., 1998. Three dimensional seismic structure beneath Shillong
Plateau and Assam Valley, northeast India. Bull. Seismol. Soc. Am. 88, 667–676.
Kayal, J.R., Arefiev, S.S., Baruah, S., Hazarika, D., Gogoi, N., Kumar, A., Chowdhury,
S.N., Kalita, S., 2006. Shillong Plateau earthquakes in Northeast India region:
complex tectonic model. Curr. Sci. 91, 109–114.
Kayal, J.R., Arefiev, S.S., Baruah, S., Tatevossian, R., Gogoi, N., Sanjoum, M., Gautam,
J.L., Hazarika, D., Borah, D., 2010. The 2009 Bhutan and Assam felt earthquakes
(Mw 6.3 and 5.1) in Northeast Himalaya Region: the active fault. Geometrics
Nat. Hazards Risk 1 (3), 273–281.
Kayal, J.R., Arefiev, S.S., Baruah, Saurabh., Hazarika, D., Gogoi, N., Gautam, J.L.,
Baruah, Santanu., Dorbath, C., Tatevossian, R., 2012. Large and great
earthquakes in the Shillong plateau–Assam valley area of Northeast India
Region: pop-up and transverse tectonics. Tectonophysics 532–535, 186–192.
Kumar, M.R., Raju, P.S., Devi, E.U., Saul, J., Ramesh, D.S., 2004. Crustal structure
variations in northeast India from converted phases. Geophys. Res. Lett. 31 (1–
4). http://dx.doi.org/10.1029/2004GL020576.
Langston, C.A., 1977. Corvallis, oregon, crustal and upper mantle receiver structure
from teleseismic p and s waves. Bull. Seismol. Soc. Am. 67 (3), 713–724.
Langston, C.A., 1979. Structure under Mount Rainer, Washington, inferred from
teleseismic body waves. J. Geophys. Res. 84, 4749–4762.
Ligorria, J.P., Ammon, C.J., 1999. Iterative deconvolution and receiver function
estimation. Bull. Seismol. Soc. Am. 89, 1395–1400.
Mitra, S., Pristley, K., Bhattacharya, A., Gaur, V.K., 2005. Crustal structure and
earthquake focal depths beneath northeastern India and southern Tibet.
Geophys. J. Int. 160, 227–248.
Molnar, P., England, P., 1990. Late Cenozoic uplift of mountain ranges and global
climate change: chicken or egg? Nature 346, 29–34.
Nandy, D.R., 2001. Geodynamics of Northeastern India and the Adjoining Region.
ACB publication, Calcutta, p. 209.
Nayak, G.K., Rao, V.K., Rambabu, H.V., Kayal, J.R., 2008. Pop-up tectonics of the
Shillong plateau in the great 1897 earthquake (Ms 8.7): insights from the
gravity in conjunction with the recent seismological results. Tectonics 27, 1–8.
http://dx.doi.org/10.1029/2006TC002027.
Oldham, R.D., 1899. Report on the great earthquake of the 12th June 1897.
Geological Survey of India Memoir 29, pp. 1–379.
Owens, T.J., Zandt, G., Taylor, S.R., 1984. Seismic evidence for an ancient rift beneath
the Cumberland Plateau, Tennessee; a detailed analysis of broadband
teleseismic P- waveforms. J. Geophys. Res. 89, 7783–7795.
Rai, S.S., Prakasam, K.S., Agarwal, N., 1999. Pn wave velocity and Moho geometry in
north eastern India. J. Earth Syst. Sci. 108, 297–304.
Rajasekhar, R.P., Mishra, D.C., 2008. Crustal structure of Bengal Basin and Shillong
plateau: extension of Eastern Ghat and Satpura Mobile Belts to Himalayan
fronts and seismotectonics. Gondwana Res. 14, 523–534.
Rajendran, C.P., Rajendran, K., Duarah, B.P., Baruah, S., Anil, Earnest., 2004.
Interpreting the style of faulting and paleoseismicity associated with the
1897 Shillong, northeast India, earthquake: implications for regional tectonism.
Tectonics 23. http://dx.doi.org/10.1029/2003TC001605, 1-1.
Rudnick, R.L., Gao, S., 2003. Composition of the continental crust. Treat. Geochem. 3,
1–64.
Sambridge, M.S., 1999. Geophysical inversion with a neighbourhood algorithm: I.
Searching a parameter space. Geophys. J. Int. 138, 479–494.
Sengupta, S., 1966. Geological and geophysical studies in the western part of the
Bengal basin, India. Am. Assoc. Pet. Geol. Bull. 50, 1001–1017.
Sitaram, M.V.D., Yadav, D.K., Goswami, K., 2001. Study on crustal structure beneath
Arunachal Himalaya and Assam. J. Geol. Soc. India 58, 285–301.
Verma, R.K., Mukhopadhyay, M., 1977. An analysis of the gravity field in
northeastern India. Tectonophysics 42, 283–317.
Zhu, L., Kanamori, H., 2000. Moho depth variation in southern California from
teleseismic receiver functions. J. Geophys. Res. 105, 2969–2980.