Crustal Structure beneath the Kashmir Basin ... - GeoScienceWorld

1 downloads 0 Views 3MB Size Report
Sep 25, 2017 - Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis by Ramees R. Mir,* Imtiyaz A. Parvez,* Vinod K. Gaur, ...
Bulletin of the Seismological Society of America, Vol. 107, No. 5, pp. 2443–2458, October 2017, doi: 10.1785/0120150334



Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis by Ramees R. Mir,* Imtiyaz A. Parvez,* Vinod K. Gaur, Ashish, Rakesh Chandra, and Shakil A. Romshoo

Abstract

We present a crustal shear-wave velocity model for the intermontane Kashmir valley derived from eight broadband seismic stations located on hard-rock sites surrounding and within the valley. Receiver functions at these sites were calculated using the iterative time-domain deconvolution method of Ligorria and Ammon (1999) jointly inverted with fundamental-mode Rayleigh-wave group velocity dispersion data to estimate the underlying shear-wave velocity structure. The inverted Moho depths were further constrained within 2 km by forward modeling and conform to those independently estimated using H-κ (crustal thickness vs. V P =V S ) stacks. The Moho descends steeply (> 15° NE) beneath the Pir Panjal range from a depth of ∼40 km south of the Main Central thrust to ∼58 km on the southwestern flank of the valley and undergoes an ∼4 km upwarp beneath the valley, thinning the overlying crust. Further northeast (NE) of the valley, the Moho dips gently (∼3°) beneath the Zanskar range. A persistent low-velocity zone at a depth of ∼12–16 km is interpreted as a sheared zone associated with the décollement separating Himalayan rocks from the Indian plate. Relocated seismicity from the International Seismological Centre (ISC, 2013) catalog (1964–2013), together with a small number of local earthquakes recorded by our network, is apparently confined south of the NE edge of the valley, suggesting that the transition of the Indian plate from locked décollement to aseismic creep lies near here. This result is consistent with the geodetic findings that a broad zone of partial seismic coupling exists beneath the Kashmir valley.

Electronic Supplement: Tables of teleseismic events and figures of receiver functions, H-κ stacks, common conversion point (CCP) stacks, shear-wave splitting measurements, robustness estimates, and available focal mechanism solutions.

Introduction The Himalaya and its ∼4:5-km-high Tibetan buttress were formed and are sustained by continuing indentation of Eurasia by the continental Indian plate that began ∼45 million years ago. Global Positioning System (GPS) measurements show that approximately half of the convergence between India and Eurasia is currently absorbed within the Himalaya at rates of close to 20 mm=yr but that convergence rates reduce westward and in Kashmir, amounting to less than 12 mm=yr (Fig. 1; Schiffman et al., 2013; Stevens and Avouac, 2015). The present-day structure of the Himalaya, as revealed by its better-studied central and Nepalese segments (Schulte-Pelkum et al., 2005; Caldwell et al., 2013), *Also at CSIR Fourth Paradigm Institute (CSIR-4PI), NAL Belur Campus, Wind tunnel Road, Bangalore 560037, Karnataka, India.

is characterized by the existence of a north-dipping Main Himalayan thrust (MHT) below which the Indian plate descends beneath southern Tibet. Slivers of the upper surface of the penetrating Indian crust shaved off during successive thrusting were pushed up and southward to constitute the Himalaya. As Tibet continued to drive southward over India and the Himalayan wedge grew in height, the MHT sprang a succession of south-verging fault splays: the Main Central thrust (MCT), which has overthrust the Great Himalayan crystalline over the Paleozoic sedimentary rocks of the Lesser Himalaya; the Main Boundary thrust (MBT), which has brought the latter over the Tertiary sedimentary rocks of the sub-Himalaya; and the southernmost Main Frontal thrust (MFT), which is currently active throughout most of the Himalayan front (DeCelles et al., 2001). These thrust systems,

2443

2444

R. R. Mir, I. A. Parvez, V. K. Gaur, Ashish, R. Chandra, and S. A. Romshoo

73.0°

73.5°

74.0°

74.5°

75.0°

A1´

8000.0

75.5°

76.0°

km 0 KZL

20

N

L2´

40

~11 mm/yr

6000.0

R

HAR MAM

BAB

l ja an

JF

34.0°

BF BTP

rP Pi

NIL

ge

GUL

an

A1

ar

BAR

2000.0

ANG AHR

0.0 THM

ISC P. basin

GND

SOP

sk

4000.0

WUL

2005

n Za

Elevation (m)

34.5°

CS Hazara Syntaxis

RAJ

MC

T

MB T

1967

OF

A0´

33.5°

RF MC

NAU

T

Kashmir basin

MB T

MF

T

L2

MF

T

2013

A0

33.0° 1905

Figure 1. The main figure shows the Kashmir basin bounded on the southwest, most immediately by the Main Central thrust (MCT) and closely followed by the Main Boundary thrust (MBT), which has severely constricted the Lesser Himalaya in the region. On the other hand, the Main Frontal thrust (MFT) is markedly distended toward the plains (Yin, 2006). Circles denoted by ISC (International Seismological Centre; see legend) are recalculated epicenters of M > 2:6 earthquakes from the ISC phase catalog for the period 1964–2013, excluding aftershocks of the 2005 M w 7.6 Muzaffarabad earthquake and those of the 2013 mb 5.7 Kishtwar earthquake, and circles denoted by CS (current study) are earthquakes (0:1 < M < 3) recorded by our network. The concentration of earthquakes along two southwest–northeast (SW–NE)-trending zones on either side of the valley marks high-strain regions potentially segmenting the Kashmir basin from its eastern and western neighbors. Depth distribution of these earthquakes along three profiles (A1–A1′; L2–L2′; A0–A0′) is plotted in Figure 11. The Jhelum fault (JF) marks a north–south (N–S) lineament almost collinear with the Hazara syntaxis, and the three short lines marked OF at the eastern extremity of the Kashmir basin mark a set of normal faults thought to be the result of an upwarping of the Himalaya by a topographic high on the underthrusting Indian plate (Bilham et al., 2013). BF near the valley axis is the mapped Balapora fault (Ahmad et al., 2014), with a recent slip rate of a few mm/yr. The black line crossing the valley southward is one of the seismic-sounding profiles of Kaila et al. (1984), and the south-southwest (SSW)-directed arrow in the upper right marks the Global Positioning System (GPS) convergence vector (Schiffman et al., 2013), with respect to the Indian plate. Dashed lines in the Kashmir basin represent faults taken from Shah (2013). Labeled stars are recent earthquakes in the region relocated using the double-difference method (except the 1905 event; see the Seismicity section for further details). The 1905 early instrumental event of Kangra (M w 7.8; Ambraseys and Douglas, 2004) is also shown. (Inset) The location of the Kashmir basin and Peshawar basin (P. basin) relative to the Hazara syntaxial bend; the inverted triangle represents location of the Nanga Parbat. Topography data are taken from Shuttle Radar Topography Mission (SRTM). The color version of this figure is available only in the electronic edition. which demarcate distinct lithological formations all along the Himalaya, are also present in the northwestern Himalaya (Fig. 1) and closely follow the Hazara syntaxial bend, testifying to the continuity of the crustal shortening process across the bend but modified by the increasing orogenparallel stresses in its vicinity. The present-day kinematics of Himalaya, as deduced from GPS geodesy, has been best constrained for the Central and Nepal Himalaya (Jackson and Bilham, 1994; Bilham et al., 1997; Bettinelli et al., 2006; Ader et al., 2012). The surface velocity field during the interseismic period indicates that the MHT remains locked to the Indian plate between major earth-

quakes, at depths shallower than ∼18–20 km, but below these depths a transition in frictional properties occurs that permits aseismic descent of the Indian plate beneath Tibet. Throughout most of the Himalaya, the locked zone is less than 100 km wide (Vernant et al., 2014). In contrast, the locked zone broadens to ∼150 km in the Kashmir Himalaya and is accompanied by a broadening in the width of the zone of partial seismic coupling (Schiffman et al., 2013; Stevens and Avouac, 2015), indicating differences in the details of the deformation process. In this article, we investigate details of the crustal structure of the Kashmir Himalaya that may contribute to understanding the broadening of the width of the

Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis locked seismogenic zone as well as a broadening of the transition zone of partial seismic coupling. Similar to its analog in the Peshawar basin west of the syntaxis, the Kashmir basin is perhaps an expression of the modified syntaxial deformation processes (Burg and Podladchikop, 2000). The Kashmir basin is a northwest–southeast (NW–SE) oriented elongate depression about 140 km long and 50 km wide in the Permo-Carboniferous Pir Panjal volcanics of the northwestern Great Himalaya. The thickness of the Pir Panjal basalts decreases from the southwestern edge of the valley, where it is ∼3000 m to less than 300 m on its northeastern edge in the Zanskar range (Shellnutt et al., 2014), with a few exposures within the valley. The basin was created by the thrusting of the southwestern Pir Panjal range over the Lesser Himalaya along the underlying MCT and is filled with a variable thickness of Pliocene–Pleistocene sediments to a depth of 1300 m and in parts even deeper. The sedimentary history of the basin, particularly of its prominent unconformities, defines the tectonic history of the Pir Panjal range that bound its southwestern edge. Accordingly, the basin was created by the Late Cenozoic rise of this range about 4.5 million years ago, at a rate which has been accelerating over the past 350,000 years to ∼4 mm=yr (Burbank and Johnson, 1982). The first experimental determination of the crustal structure of the Kashmir Himalaya was attempted by Kaila et al. (1978, 1982, 1984) using wide-angle deep seismic soundings (DSS) along three profiles, one of which crosses the valley from northeast to southwest (NE–SW), is shown in Figure 1. Kaila et al. (1984) found that the Moho beneath Nanga Parbat, which is ∼90 km west-northwest of the valley (Fig. 1), lies at a depth of 60 km, which is consistent with the automated measurements of EarthScope Automated Receiver Survey for network XG95 that was set up in the 1990s (Crotwell and Owens, 2005). Kaila et al. (1984) found the Moho depth at Babareshi (BAB, Fig. 1) on the SW edge of the valley near Gulmarg to be 51 km, the depth near Srinagar (neighboring HAR) on the northeastern edge to be ∼60 km, and the depth at Kanzalwan (KZL) to be 64 km. They constrained the average crustal P-wave velocity to a depth of 50 km beneath the valley to ∼6:3 km=s and the Moho depth to the south near the MBT to ∼44 km. Kaila et al. (1984) also mapped a prominent reflector at a depth of 10–12 km beneath the MBT. The motivation for the present work arose from the desirability of imaging the crustal structure of the Great Himalaya closest to its western syntaxis and comparing it to that of the region to its east and to the more central parts of the Himalaya. Also, the potentially higher-resolution mapping of the regional seismicity, which the experimentally determined velocity structure from the local network makes possible, was expected to delineate the high strain zones in the region and constrain the limits of the locked zone for which the width, together with the rate of convergence, yields a measure of its hazard potential.

2445

Data and Method The data for this study were generated by four Güralp CMG-3TD broadband seismic sensors at AHR, BAR, BTP, and GUL (Fig. 1) from June 2013 to June 2014 and by Nanometrics Trillium 120P broadband seismometers at AHR, BAR, GUL, ANG, MAM, HAR, and GND from January 2015 to April 2016. Additionally, we used data from the Incorporated Research Institutions for Seismology (IRIS) station at Nilore (NIL), which operates a Nanometrics Trillium 240s seismometer. Considerable care was taken to locate the seismic stations on bedrock away from cultural sources of noise because the objective of our study was to map significant discontinuities in the subsurface structure by isolating and analyzing the weak P-to-S converted signals at their interfaces, minimally affected by reverberations caused by the basin fill. Three of the eight stations were located in the Pir Panjal range on the southwestern edge of the valley (AHR, BAR, and GUL), three on rock outcrops within the valley (ANG, BTP, and HAR), and two in the Zanskar ranges close to the northeastern margin of the valley (GND and MAM) (Fig. 1). Data at all of these stations were recorded continuously at the rate of 100 samples per second, and time stamped by GPS signals that are accurate to better than 0.1 ms. P-to-S receiver functions (RFs) used in this study are shear-wave signals derived from the refraction of a P wave through a seismic discontinuity and its multiple reflections. They can be isolated by deconvolving the incident P wave from the radial and transverse horizontal-component seismograms, thereby eliminating source and path effects. Under ideal conditions of horizontality, homogeneity, and isotropy, their radial components are expected to exhibit a delayed P-to-S converted signal with respect to the generating P wave, followed by multiples generated between the subsurface interface and the surface, recording nothing on the transverse component. Any signals appearing on the latter, therefore, reveal departure from these ideal conditions. Toward calculating RFs for each station, all seismograms generated by M ≥ 5:5 events in the epicentral range of 30°–100° (Fig. 2) were first visually selected on the basis of their clarity. The epicentral distance longer than 30° makes it realistic to treat the wave arrivals as being planar, and the range provides a workable trade-off between the two contrasting requirements: (1) for strong converted signals which require higher incident angles and, therefore, arrivals from nearer teleseisms, and (2) for high lateral resolution requiring near verticality of the teleseismic arrivals to restrict the lateral domain of their conversion points, that is, arrivals from farther teleseisms. The range also eliminates multiple arrivals generated by events closer than 30°, as well as complex contributions from the core–mantle boundary for events farther than 100°. Radial and transverse RFs were calculated for each selected seismogram using the time-domain deconvolution method of Ligorria and Ammon (1999) with a Gaussian width of 1.6, corresponding to a 0.56-Hz low-pass filter to ensure a minimum acceptable vertical resolution of

2446

R. R. Mir, I. A. Parvez, V. K. Gaur, Ashish, R. Chandra, and S. A. Romshoo

Table 1 Station-Wise Receiver Functions (RFs) with Percentage Fit

CS AE

30°

Station

RFs Selected

Number of RF Stacks

AHR ANG GUL BAR GND HAR NIL (IRIS) BTP MAM

111 97 95 89 62 49 38 35 28

7 7 6 6 5 5 3 4 2

Average Number of RFs/Stack (Rounded to Zero Decimal Places)

Range of Percentage Fit

8 8 7 5 6 5 4 4 4

81–96 80–97 75–96 80–96 76–94 75–94 77–95 73–95 71–94

The number of selected RFs at each station, the number of RF stacks formed out of these, the average number of RFs per stack, as well as the average percentage of signal reproduced for each stack. IRIS, Incorporated Research Institutions for Seismology.

100° Figure 2.

Teleseismic events (from distance 30°–100°) recorded by our network. Triangles at the center denote stations, and circles denoted by AE represent all recorded events, and those denoted by CS (current study) are those used in this study. For a detailed list of events, see Ⓔ Table S1 in the electronic supplement to this article. The color version of this figure is available only in the electronic edition.

∼2:6 km, assuming the λ=2 criteria (Bostock and Rondenay, 1999; Rychert et al., 2007), and an average shear-wave velocity of 3:5 km=s. All radial RFs thus calculated were then tested for their acceptability on the basis of the degree of fit (> 70%) between the product of their convolution with the corresponding vertical seismograms and the radial seismograms. The number of RFs thus selected for further analysis is given, by station, in Table 1, along with their respective ranges of percentage fit. The selected RFs are identified by the locations of their conversion points on the Moho, assumed to be horizontal, called Moho Piercing Points (PPs). The closeness and occasional overlaps of the piercing points corresponding to wave arrivals at different stations provide a check on the consistency of the inverted shear-wave velocity structure beneath a given neighborhood of PPs, irrespective of their source–receiver paths. Figure 3 shows the radial and transverse RFs for ANG in increasing order of ray back azimuths (BAZ) and constitutes a fairly representative sample of the selected RFs for other sites. Most of the earthquakes studied here originate in subduction zones in the western Pacific or along the islands of Sumatra and Java (Fig. 2). For this reason, their azimuthal range is generally limited to the back azimuths between 30° and 120° (Fig. 4). Radial RFs for all stations used for this study are shown in a polar plot (Fig. 4) which exposes this data gap while allowing a ready scrutiny of the coherence between corresponding phases from various azimuths. All events for

which records were analyzed for the present work are shown in Figure 2 (for a detailed list of these events see Ⓔ Table S1 in the electronic supplement to this article). Despite stringent screening, most RFs were found to be quite complex in that substantial energy appears on the transverse RFs (e.g., Fig. 3 for ANG), which is comparable to that of the radial RFs. Further, the Ps phases on the radial RFs, while exhibiting significant variations with azimuth (Fig. 4), have random variations, even among those from closely spaced sources, both in their forms and arrival times. These features clearly indicate that the crust departs from the ideal conditions of being isotropic and laterally homogeneous, that is, the prominent interfaces, such as the Moho, may be dipping and/or the layers anisotropic, or the crust hosts significant scatterers. However, to minimize the effect of random noise caused by scatterers, we sorted the selected RFs into neighborhood sets corresponding to arrivals from closely spaced earthquake sources (bins spanning 10° in back azimuth and 5° in epicentral distance). Coherent RFs corresponding to each bin for which PPs lay within the first Fresnel zone fH  λ=22 − H2 g1=2 , that is, of ∼15 km radius, were then stacked p to obtain a set of representative radial RFs for that site with a N gain in signal-to-noise ratio, where N is the number of RFs stacked. An example of one such set of stacked RFs for AHR is shown in Figure 5a.

Seismicity The Kashmir Himalaya has a poor local instrumental record of seismicity, with just one analog seismograph located in the valley and another near the foothills in Jammu. To gain insight into the background seismicity of the region, therefore, we relocated 1318 M > 2:6 events of the International Seismological Centre (ISC) catalog for the period 1964–2013, lying within 32.5° N, 73° E and 35° N, 77° E, by applying the double-difference method (Waldhauser and Ellsworth, 2000) to the ISC phase data (P and S). The epicenters of these relocated events, excluding the after-

Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis

(a) 322.5 296.4 292.8 286.1 285.9 173.2 151.1 142.3 123.1 122.9 122.1 117.7 117.6 117.0 116.6 116.4 115.2 113.3 112.3 112.2 111.2 110.9 110.6 110.5 110.0 109.9 109.9 107.7 107.6 104.4 102.6 102.0 102.0 102.0 102.0 101.9 101.9 101.8 101.7 101.4 101.3 101.1 101.0 101.0 100.0 99.6 98.6 93.0 90.4 90.3 78.1 75.0 73.0 72.9 65.3 63.9 62.5 62.4 62.4 48.9 37.0 36.6 34.8 34.4 26.1 22.9

P

Ps

PpPs

PpSs+PsPs

(b)

2447

Modeling Receiver Functions H-κ Stacking

Back azimuth (°)

Back azimuth (°)

The stacked RFs were analyzed to obtain a first estimate on the mean crustal thickness and the V P =V S beneath their corresponding clusters using the H-κ stack formed by weighting the predicted arrival times of the Ps, PpPs, and the PpSs + PsPs phases by 0.7, 0.2, and 0.1, respectively (Zhu and Kanamori, 2000; Eagar and Fouch, 2012). An error ellipse was generated using the error-estimation procedure of Eaton et al. (2006). Thus, we obtained the average Moho depths and crustal V P =V S values at selected stations (e.g., Fig. 5b and Ⓔ Figs. S2, S4, S6, S8 and S9), corresponding to the respective stacked RFs, assuming an average crustal compressional velocity of 6:3 km=s (Table 2; see Ⓔ the electronic supplement for a triangulated map of the Moho depths). The latter figure was constrained by the results of Kaila et al. (1984), based on their DSS results along a profile across the northern part of the valley and also consistent with the inverted shearwave velocity profile which was subsequently estimated. The V P =V S beneath various sites in the valley was found to vary between 1.62 and 1.84, corresponding to a Poisson’s ratio of ∼0:2–0:3. These values encompass the Poisson’s ratio for flood basalts which lie in the 0.25–0.28 range (Christensen, 1996; Brocher, 2005) and that of sialic rocks in 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 the 0.24–0.27 range (Christensen, 1996). Time (s) Time (s) The estimated values are consistent with the mapped presence beneath the valley Figure 3. The (a) radial and (b) transverse receiver functions (RFs) with increasing ray back-azimuths for ANG, which is a fairly representative sample of RFs calculated floor of a variable thickness of the mafic for other sites. Moho-converted Ps phases on the radial RFs are prominent from 23° to Pir Panjal rocks (Farooqi and Desai, 1974; 100° back-azimuth (°BAZ) and again from > 110° BAZ. The multiple PpPs and PpSs + Bhatt, 1982; Shellnutt et al., 2014) underPsPs phases which are not as continuous have been marked by dashed lines. Note the lain by a sialic crust. Estimates of the Moho data gap of ∼95° from 170° to 285° BAZ in both radial and transverse RFs. The color depth from the H-κ stack are not expected version of this figure is available only in the electronic edition. to be robust, especially in an inhomogeneous crust. However, we attempted to calculate these as shocks of the 2005 Muzaffarabad and those of the 2013 a first estimate, using only a selected set of RFs that exhibited Kishtwar events (via Reasenberg, 1985; Wiemer, 2001), recognizable multiple phases. We do not use the H-κ stacking are plotted in Figure 1, together with the epicenters of results to model the crustal structure beneath the valley because 215 M > 0:1 events recorded by our recent network. Because of the availability of the better-constrained RF inversions, but of the larger regional spread of the ISC events, we used the their correspondence with the Moho depths obtained from the conjugate gradient method (Paige and Saunders, 1982) to inverse solutions enhances our confidence in these values. estimate epicenters and depth for which average errors are ∼5 km in epicentral location and ∼8 km in depth. For the Joint Inversion local earthquakes, however, we used the singular-value decomposition method for which average errors are ∼5 km To better constrain the crustal structure beneath the valin epicentral location and ∼10 km in depth. ley, we inverted all the stacked RFs from each of the eight 322.5 296.4 292.8 286.1 285.9 173.2 151.1 142.3 123.1 122.9 122.1 117.7 117.6 117.0 116.6 116.4 115.2 112.3 112.2 111.2 110.9 110.5 110.0 109.9 109.9 107.7 107.6 104.4 102.6 102.0 102.0 102.0 102.0 101.9 101.9 101.8 101.7 101.4 101.3 101.1 101.0 101.0 100.0 99.6 98.6 93.0 90.4 90.3 78.1 75.0 73.0 72.9 63.9 62.5 62.4 62.4 48.9 37.0 36.6 34.4 26.1 22.9

2448

R. R. Mir, I. A. Parvez, V. K. Gaur, Ashish, R. Chandra, and S. A. Romshoo 0° 0° BAR 15

0° 0° HAR 15

300 °

120 °

time (s)

300 °

300

°

120

time (s)

°

120

°

120

°

°

300

0° 24

°

0 24

°

90°

270°

90°

270°

120

° 60

° 60

°





24

24

120

90°

270°

90°

270°

time (s)

120 °

time (s)

°

° 60

60

°



120

24



time (s)

90°

270°

90°

24

time (s)

°

°

300 °

60

60

270°

°





300

90°

24

24

°

30°

270°

90°

270°

300

time (s)

33

° 60

300 °

30°

° 60

time (s)

(2000). The joint inversion has the advantage of exploiting the sensitivity of RFs to short-wavelength variations and 10 10 at the same time of blunting its inherent 5 5 velocity-depth trade-off, with the con0 0 straints imposed by SWD data on the average crustal shear-wave velocity. An important consideration, however, is to choose an appropriate value for the influence parameter, p, which respectively 210 210 ° ° 0 0 measures the relative weights of the 5 5 ° ° 1 1 180° 180° SWD data and RFs in the inverted solu0° 0° 3 3 tion. We, therefore, tested the effect of ° ° 0 0 0 0 ° ° 33 33 BTP 15 GUL 15 various values of p on the fitness of the inverted models to the respective RFs 10 10 and SWD by varying p from 0 to 1 in steps 5 5 of 0.1. The solution converged after 32 0 0 iterations (Fig. 6 for AHR) and reproduced the observed RF data well with a 96% fit when p was increased from 0 to 0.1, but further increases in p made little difference. We, therefore, adopted the value 210 210 0° 0° of p  0:1 (10% weight of SWD and 5 5 ° ° 1 1 180° 180° 90% of RF) for inverting all RFs for the 0° 0° 30° 30° ° ° Kashmir valley, which, while ensuring a 0 0 33 33 AHR 15 ANG 15 close-enough fit with the surface-wave data, maximizes the weight of the RFs 10 10 in the inverted solutions in recognition 5 5 0 0 of their higher lateral resolution. The inversion process for all RFs was initiated using the basic earth model ak135 (Kennett et al., 1995), composed of a stack of four 1-km-thick surface layers underlain by 48 layers each of 2 km thickness and a 210 210 0° 0° constant shear-wave velocity equal to the ° ° 15 15 180° 180° upper-mantle velocity of 4:48 km=s. This 0° 0° 30° 30° ° ° choice was made to let the inversion proc0 0 33 33 MAM 15 GND 15 ess freely exploit the data to condition the 10 10 crustal structure. Then, we examined the 5 5 variations, if any, that a different starting 0 0 velocity might produce in the inverted solutions using two end values of V S equal to 4.68 and 4:28 km=s, but, finding little notable variation (see Ⓔ Fig. S15), we used the average V S value of 4:48 km=s for further analysis. The resulting inverted 210 210 0° 0° ° ° 15 15 shear-wave velocity structure was tested 180° 180° through forward modeling by simplifying it to 4–6 layers, each with a constant or Figure 4. The polar plots of selected radial RFs for each of the eight stations, which linearly varying velocity, and by varying allows ready scrutiny of the coherence between corresponding phases with azimuth as well as the azimuthal availability of earthquake sources. The color version of this figure the Moho depth up to 6 km in steps is available only in the electronic edition. of 2 km. The forward models showed that, for Moho depths departing by more than 2 km from the stations jointly with 1D surface-wave dispersion (SWD) data jointly inverted value, the mismatch between the calculated obtained from Mitra et al. (2006), using the package of Herrand observed values was quite significant (see Ⓔ the elecmann (2013), which follows the method of Julia et al. tronic supplement). The shear-wave velocity structure thus 33

Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis

(a)

(b)

RF Amplitude

Ps

PpPs

2.0 AHR H = 53.0 ± 1.1 = 1.78 ± 0.03

PpSs+PsPs

DEL°

1.9

63.71

53.4

62.51

54.0

62.50

54.3

62.45

54.1

62.35

53.7

62.23

54.2

-0.05

62.20

54.3

-0.10

2.0

0.15 0.10 0.05

κ

Back azimuth (°)

LVZ

2449

1.8

0.00

-0.15

1.7

1.0 0.0

1.6

−1.0 0.0

5.0

10.0

15.0

20.0

25.0

30.0

35.0

Time (s)

30

40

50

60

70

H (km)

Figure 5. (a) The selected RFs from AHR for back azimuth of ∼62°–63° and delta (DEL) of ∼53°–54° corresponding to their Mohopiercing points (PPs; latitude 33.73°, longitude 74.98°) within the valley and confined in the first Fresnel zone. The bottom trace is their stacked representative, along with its 1σ bounds. (b) H-κ stack of the same set delineating the crustal thickness beneath the corresponding point in the valley. The color version of this figure is available only in the electronic edition. obtained beneath some selected groups of piercing points along the SW and the NE margins of the valley, as well as within it, are shown in Figure 7, whereas those for other stations are contained in Ⓔ Figures S3, S5, S7, S10 and S18. Limitations of the Study Radial RF stacks for all eight stations show considerable complexity (Fig. 4), essentially contributed by intracrustal features, because the distinguishing mark of shallow sediments, which is to broaden and add a secondary peak to the main P phase, was found in only one set of RFs from GUL for 20°–90° back azimuths (see Ⓔ Fig. S11). First, their transverse components show substantial energy, almost comparable to that on the radial. Second, the Ps phase (Fig. 3) shows significant variability with azimuth, both in their forms as well as in times of arrivals. These observations clearly require that the underlying earth structure is either laterally inhomogeneous or anisotropic or both (Fontaine et al., 2013). This is also corroborated by the apparently incoherent arrivals on the radial, some of which virtually swamp the usually prominent Ps phase from the Moho and its multiples. Although departures from ideal conditions are theoretically expected to exhibit distinctive patterns, depending on the nature of inhomogeneity, their attribution to any recognizable pattern was difficult to establish in the case of the Kashmir RFs. Thus, for example, a distinctive result of scattering as an extended exponentially decaying coda after notable arrivals could not be discerned, although the often-observed loss of coherence among phase arrivals at a station from event to event suggests the pervasive effect of scattering. Similarly, while a dipping interface may be identified by the center of asymmetry of the tangential phases generated by it, the lack of seismic signals from a wide range of

azimuths made such a test impossible, even as other evidence points to a gently dipping Moho beneath the valley, which should therefore contribute little to signal complexity. Most likely, therefore, the observed complexity of RFs reflects the combined effect of scattering and crustal anisotropy, although we have been unable to resolve any definitive signatures of these in the Kashmir valley crust. Of the several such studies along the various Himalayan segments (Caldwell et al., 2013), only one in the Nepal Himalaya (Schulte-Pelkum et al., 2005) succeeded in modeling anisotropy of an upper crust shear zone, aided by a dense network of seismic stations. A search for split shear-wave phases in the seismograms, as well as RFs to investigate anisotropy in the study region, however, led to only one successful example of the former (SKS) at BAR, indicating a fast direction along the 55° azimuth, almost perpendicular to the valley. This being the lone result from this experiment pertaining to the regional mantle, it is not discussed further in this article. A brief description of the test and related figures, however, appear in Ⓔ the electronic supplement (see Fig. S14). Despite the complexity of RFs, however, the Moho Ps phase can be identified on most of these as a coherent positive signal through a reasonable range of back azimuths, such as those for ANG (Fig. 3), at around 7 s between 15° and 99° and again between 109° and 286°. But neither of the two PpPs or the (PsPs + PpSs) multiples can be coherently traced. There is, however, a strong negative phase at ∼2 s on the radial between the azimuths of 15° and 92°, indicating a low-velocity conversion interface and an equally strong positive phase at about the same time step between the back azimuths of 90° and 190° on the transverse RF, likely implying anisotropy of the causative interface (Schulte-Pelkum et al., 2005).

2450

R. R. Mir, I. A. Parvez, V. K. Gaur, Ashish, R. Chandra, and S. A. Romshoo

Table 2 The Moho Depths Obtained from the Joint Inversion of RFs Duly Refined through Forward Modeling as well as those Obtained from H-κ Stacks Station

BTP

BAR

HAR

GUL

AHR

Latitude (°N)

Longitude (°E)

Elevation (m)

BAZ (°)

33.922

75.035

1790

53–55 99–101 110–118 118–122 37–41 55–65 73–75 101–102 118 284–294 20–26 61–67 86–89 109–115 330 37 53–62 72–77 99–106 104–112 138–141 14–15 37–39 62–63 72–77 109–117 141–157 285–294 106–109 109–110 26–36 62–64 101–102 116–118 139–142 26–36 62–72 98–100 101–103 110–113 111–122 285–286 34–39 75–77 110

34.204

34.111

34.051

33.642

74.336

74.810

74.399

74.780

1593

1649

2719

2291

NIL (IRIS)

33.651

73.269

0629

GND

34.261

75.090

2441

ANG

33.730

75.156

1789

MAM

34.010

75.310

2194

hm (km)

60 ± 58 ± 58 ± 56 ± 54 ± 54 ± 56 ± 60 ± 55 ± 58 ± 60 58 ± 62 ± 62 ± 55 ± 56 ± 56 ± 56 ± 55 ± 56 ± 56 ± 56 ± 55 ± 54 ± 54 ± 56 ± 58 ± 58 ± 52 ± 52 ± 60 ± 60 ± 60 ± 62 ± 61 ± 54 ± 54 ± 56 ± 56 ± 58 ± 58 ± 54 ± 60 ± 58 ± 60 ±

4 2 2 2* 2* 2* 2* 4* 2* 2 2 2 2 2* 2* 2* 2* 2* 2* 2* 2* 2* 2* 2* 2* 2 2 2 4 2 2 2 2 2 2* 2* 2* 2 2 4 2* 2 2 4

hz (km)

— 57 ± — — 61.0 ± — 56.5 ± 57.5 ± 54 ± — — 56.0 ± 62.0 ± — — — — — 52.5 ± — 55.0 ± — 55.5 ± 53.0 ± — 57.0 ± 62.0 ± 57.5 ± 52.4 ± 55 ± — 59.5 ± 59.5 ± — 58.0 ± — — — — — — — — — 57 ±

1.7

2.0 2.8 1.9 6.9

3.5 2.2

3.7 3.1 3.6 1.1 2.0 2.7 2.5 0.8 2.1 3.4 1.8 5.5

3.2

V P =V S

— 1.76 ± — — 1.64 ± — 1.74 ± 1.82 ± 1.78 ± — — 1.81 ± 1.62 ± — — — — — 1.84 ± — 1.82 ± — 1.76 ± 1.78 ± — 1.68 ± 1.67 ± 1.76 ± 1.80 ± 1.68 ± — 1.82 ± 1.69 ± — 1.64 ± — — — — — — — — — 1.74 ±

0.04

0.04 0.09 0.04 0.14

0.14 0.04

0.08 0.10 0.06 0.03 0.05 0.06 0.06 0.02 0.06 0.07 0.04 0.09

0.08

Nr

2 7 5 3 9 6 6 9 6 3 3 3 2 8 3 3 10 5 6 7 2 2 11 7 3 9 3 4 3 4 6 5 9 2 3 5 11 4 6 11 7 3 4 3 2

hm refers to the Moho depth obtained by forward modeling on inverted results, and hz refers to that obtained from H-κ stacking. BAZ refers to back azimuths of RFs in degrees, and V P =V S refers to compressional to shear-wave velocity ratio obtained from H-κ stacking. N r refers to the number of RFs in respective stacks. *Moho depths from RFs that pierce the Moho beneath the Kashmir basin.

Results Crustal Structure To mitigate the uncertainties arising from loss of coherence between notable phases of the RFs corresponding to closely spaced source–receiver paths, we inverted their stacks (Table 1) to obtain the crustal structure beneath the corresponding piercing points. The results are summarized in

Figure 7 through a small set of typical inversions. Most of these show a low-velocity layer (∼3 km=s) at a depth of about 12–16 km and a transitional layer of small variable thickness near the Moho. We identify the Moho with its iconic shearwave velocity of 4:1 km=s to maintain consistency with the results of other research along the Himalaya, such as that of Rai et al. (2006) and Hazarika (2007). We also calculated and inverted RFs for the NIL station (Table 2; see Ⓔ Figs. S9

Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis

(a)

VS

(b)

2451

quake sources, yield near-coincident piercing points in respect to diverse ak source–receiver paths. The second part of Figure 7 shows that Moho depths beneath the Pir Panjal range on the western flank of the valley are close to ∼58 km, both in its southern part as well as the northern part, whereas those on the eastern flank are ∼60 km. The gently dipping Moho beneath the valley, however, shows crustal thinning of ∼4 km along the valley axis. This unexpected finding led us to p = 0.0 p = 0.1 undertake extensive forward-modeling p = 0.2 p = 0.6 tests to confirm its reality (see Ⓔ the electronic supplement). Kaila et al. (1984) reported a shallow (c) p = 0.0 depth for the Moho (∼51 km) at Babareshi p = 0.1 p = 0.2 (BAB, see Fig. 1) near the western margin p = 0.6 ob of the valley, deepening northeastward to ∼60 km at its eastern margin near Srinagar. Their values at Kanzalwan (KZL) NE of the valley were close to 64 km, which are similar to our estimates at GND. Similarly, beneath the Pir Panjal range on the western flank of the valley, Kaila et al. (1984) estimated the Moho Figure 6. The results of p-value tests in respect to a stack of RFs for AHR from depth to be 57 km, which is quite close BAZ  289:7° and DEL  40:8°. (a) The jointly inverted shear-wave velocity models to our estimate of 58  2 km from joint for different values of the influence parameter p. (b,c) The predicted surface-wave inversion of two Moho-piercing neighbordispersion (SWD) curve and the synthesized RFs, respectively, by the corresponding inversions. It may be noted that the value of p  0:1 (10% weight of SWD and hoods of wave arrivals at AHR and BAR 90% of RF) satisfactorily fits both the RFs and the SWD, and further increases in p (Fig. 7b, first column). All these results are do not significantly enhance the fit, suggesting it to be the best choice while maximizing summarized in Figure 8, which also shows the contribution of the better-resolved RFs. The color version of this figure is available the Moho depths determined by Kaila et al. only in the electronic edition. (1984), as well as the position of a reflector mapped by them at a depth of about 10 km beneath MBT. and S10), which is located on the western side of the Hazara The PPs of RFs calculated for all stations (assuming a 1D syntaxis south of the MBT. The Moho depth at this station was structure), along with the corresponding Moho depths from found to be ∼52 km, which is much deeper than that in the selected stacks (Table 2), are shown in Figure 9a. sub-Himalaya on the eastern side of the syntaxis. Figure 9b–e shows a selected set of RFs along four profiles The first part of Figure 7 shows how advantage was perpendicular to the valley axis, which signal an advanced taken of the existence of closely lying PPs of wave arrivals arrival of the Moho Ps phase at sites in the valley, pointing from widely different azimuths heading for different stations to a thinning of the crust underneath. to establish the consistency of Moho depths beneath the corA prominent negative phase at ∼2 s, appearing in most responding PPs, obtained from inversion of the respective RFs (Fig. 3), signaling the existence of a low-velocity zone RFs. This test, successfully demonstrated in respect to two (LVZ) at a depth of ∼15 km, is clearly highlighted in their pairs of stations, one from the southern part of the valley and inverted models. We tested the robustness of this inference the other from the northern part, engenders confidence in the by calculating synthetic RFs of a suite of rationalized inassumption that Moho depth beneath a given piercing point verted models for AHR, with and without the LVZ, and by is independent of the disposition of the ray through it, even comparing these with the corresponding ones calculated though individual rays intersecting at the point may vary in from the data. The RF generated for a model derived from their orientations and the corresponding RFs may vary in the inverted one at AHR, but with the LVZ removed, resulted form. This result is particularly useful for investigating the in large misfits of the amplitude of the Moho Ps phase. Furcrustal structure in a setting such as the Kashmir valley ther, in addition to the negative PpPs multiple at 5 s produced where the effect of signal degradation caused by inhomogeby this layer alone, its positive PsPs multiple was found to neities may be significantly mitigated by a judicious choice arrive at ∼7 s, about the same time as the arrival of the Moho of seismograph sites that would, for a given scenario of earthPs phase, which in consequence is broadened (see Ⓔ the

2452

R. R. Mir, I. A. Parvez, V. K. Gaur, Ashish, R. Chandra, and S. A. Romshoo VS (km/s) 2

3

4

5

0

30

45

45

60

60

90

75

BAZ: 110−118 DEL: 60−65 PPLON: 74.51 PPLAT: 34.13

2

3

0 15

90 4

5

(b)

AHR(11)

15 30 45

60

60 BAZ: 036−039 DEL: 72−76 PPLON: 74.88 PPLAT: 33.76

75 90

2.0

3.0

4.0

5.0

2.0

3

2.0

15.0

BTP(2)

30.0

45.0

45.0

60.0

60.0

60.0

75.0

BAZ: 284−285 DEL: 35−40 PPLON: 74.07 PPLAT: 34.26

2.0

3.0

4.0

90.0 5.0

2.0

15.0

AHR(4)

75.0

BAZ: 139−141 DEL: 44−53 PPLON: 75.21 PPLAT: 33.74

3.0

4.0

0.0

15.0

90.0 5.0

15.0

ANG(3)

30.0

30.0

45.0

45.0

60.0

60.0

60.0

90.0

JI

FM

75.0 90.0 IM

5.0 NW

GND(2)

BAZ: 116−118 DEL: 47−49 PPLON: 75.36 PPLAT: 34.15

2.0

45.0

BAZ: 285−294 DEL: 39−43 PPLON: 74.53 PPLAT: 33.72

4.0

3.0

4.0

5.0

0.0

30.0

75.0

3.0

0.0

30.0

0.0

BAZ: 285−286 DEL: 39 PPLON: 74.89 PPLAT: 33.79

5.0

45.0

5

ANG(3)

4.0

30.0

90.0 4

15.0

BAR(3)

75.0

BAZ: 036−037 DEL: 75 PPLON: 74.5 PPLAT: 34.17

3.0

VS (km/s)

0.0

15.0

GUL(5)

2

45

90

5

0

30

75

4

VS (km/s)

0.0

15

BAR(6)

30

75

3

Depth (km)

Depth (km)

15

Depth (km)

2 0

Depth (km)

(a)

VS (km/s)

VS (km/s)

BAZ: 034−036 DEL: 76−77 PPLON: 75.26 PPLAT: 33.85 LVZ

75.0 90.0

MAM(4)

BAZ: 034−039 DEL: 73−77 PPLON: 75.43 PPLAT: 34.14

SE

Moho

E

W

Figure 7. The jointly inverted and forward-modeled shear-wave velocity structure beneath selected clusters of PPs corresponding to tightly binned source locations. In each case, inversion was initiated with the fixed shear-wave velocity (4:48 km=s) model shown by the dashed thin line (denoted by IM [initial model]). Jointly inverted (JI) results are shown by the solid thin line, forward models (FM) by solid thick lines, and uncertainty bounds on the Moho by the dashed thick line. Arrows denote the depth of the low-velocity zone (LVZ) and the Moho (see legend). (a) Two pairs of inversions: the first for BAR and GUL in respect of RFs from BAZ ∼114° and ∼36°, respectively, and the other for AHR and ANG in respect to RF stacks from BAZ ∼37° and ∼285°, respectively. Despite wide differences in source locations, each pair corresponds to close PP and yield similar Moho depths. (b) Two rows of inversions for sites from the southern side of the valley to the northern side, and three columns of inversions from the western flank of the valley through its middle to the eastern flank. The Moho depths yielded by them vary from ∼58 km under the Pir Panjal range on the SW to ∼62 km under the eastern margin with a Moho upwarp in the valley at ∼54 km. The color version of this figure is available only in the electronic edition. electronic supplement). The broadening of the Moho Ps phase seen on most RFs is therefore most likely brought about by the superposition of the PsPs multiple generated by the LVZ and suggests itself as authentic evidence for the existence of this layer.

predominantly thrust solutions, except for a very few along its western border that show a strike-slip mechanism (Ⓔ Fig. S19). The latter, delineating a narrow approximately north–south (N–S) zone, suggests a possible slip boundary that segments the Kashmir Himalaya just east of the Hazara syntaxis.

Seismicity We also used the inverted velocity structure to determine the hypocenters of 215 M > 0:1 events located by our network as well as to recalculate those of the 1318 M > 2:6 events reported in the ISC catalog for the period 1964–2013, excluding aftershocks of the 2005 Muzaffarabad earthquake and of the 2013 Kishtwar earthquakes. These were laterally constrained within 5 km and vertically constrained within 8 and 10 km, respectively, for the ISC and locally recorded events. Their plot in Figure 1 clearly delineates two NE–SW-directed narrow zones of high strain which apparently bound the rupture zone of the 1555 M w 7.56 event (Ambraseys and Douglas, 2004) and is thought to constitute a current seismic gap. Published fault-plane solutions of 73 moderate (4:6 ≤ M ≤ 7:6) earthquakes in the region (ISC, 2013) that occurred on either side of the Kashmir basin show

Discussion and Conclusion In the northwestern Himalaya, Indo-Eurasian convergence over the past million years has been mostly accommodated by the three Main Cenozoic thrusts, the MFT, the MBT, and the MCT (Vassallo et al., 2015), as elsewhere in the Himalaya. Individual roles of these thrusts in relieving pentup strain at millennial time scales, however, are likely to be quite different from those along the central Himalayan arc because of their greatly altered geometrical dispositions. First, the width of the Lesser Himalaya between the MCT and the MBT SW of the Kashmir basin decreases to < 30 km compared with ∼100 km in the central Himalaya (Fig. 1). Correspondingly, the sub-Himalaya between the frontal thrusts and the zone of microseismicity that signifies the onset of incomplete seismic coupling on the MHT is unusually widened to

Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis L2

L2´

2453

NE Zanskar

Depth (km)

BF

Kashmir basin

KB

RF

MFT

MBT MCT

Pir Panjal

Elevation (km)

from the frontal faults in the south to the Karakoram in the north, is called for to 5 quantify the seismogenic potential of the 4 various thrusts involved in accommodating 3 2 crustal shortening in the Kashmir Hima1 ? ? laya. This article makes a contribution to0 LVZ ? ward delineating the first-order structure ? ? −20 beneath the Kashmir basin and the northern −40 limit of the locking line in the region, which CS ISC MOHO −60 is critical to evaluating its hazard potential. 0.1< M ≤1 1< M ≤2 In the study region, the Moho near the −80 2< M ≤3 3< M ≤4 AHR HAR 4< M ≤5 5< M ≤6 GUL GND MBT was shown to lie at a depth of −100 −192 −144 144 −48 0 48 −96 96 ∼44 km by Kaila et al. (1984) and more Distance (km) recently by Wanchoo et al. (2014), who determined its depth to be ∼42 km at RaFigure 8. Cross section of the Kashmir valley beneath L2–L2′ (Fig. 1). Arrows on jouri (RAJ), which lies close to the Moho, the surface denote the surface expressions of major faults (Fig. 1). Inverted triangles in west of the valley. Moho depths in other the subsurface are depths to a prominent low-velocity horizon required by the inversions of RFs, whereas stars and diamonds, respectively, denote the Moho depths from Kaila segments of the Himalaya have also been et al. (1984) and from the current study (CS). Pentagons represent a reflector mapped by shown to have similar values near the Kaila et al. (1984), located in conjunction with our LVZ. Error bounds on the Moho were MBT: ∼44 km in Nepal (Schulte-Pelkum obtained from forward modeling. All depths are below sea level after correcting for the et al., 2005; Caldwell et al., 2013) and elevation of stations. The vertical line marks the valley axis (Fig. 1) and serves as a ∼42 km in NE Himalaya (Acton et al., reference for reckoning distances on either side. The ISC- and CS-denoted circles respectively mark the hypocenters of earthquakes from ISC and those recorded by our 2011), with a gentle dip beneath the range. network and are scaled by size (see legend). Triangles mark the station locations posiHowever, our results show that in the tioned according to their elevation along L2. Note that the Moho has a steep descent Kashmir Himalaya, the Moho dips steeply from ∼44 km beneath the MBT to ∼58 km beneath the Pir Panjal range within ∼35 km, northeastward (> 15°) from the MBT tobecoming gentler further NE across the valley, with a slight upwarp in the valley, sugward the Kashmir valley to a depth of gesting thinning of the crust. Squares with error bounds are the selected Moho depths from H-κ stacking algorithm. The color version of this figure is available only in the ∼58 km at its southwestern edge but beelectronic edition. comes gentler beneath the valley with an ∼4 km upwarp, thereby thinning the crust. At the northeastern edge of the valley flanked by the Zanskar ∼70 km. Furthermore, crustal thickness beneath the 3.5 km range, the gently dipping Moho (> 3°) lies at a depth of elevation contour, that acts as a proxy for the northern limit ∼60 km. Although the NE dip of the Moho beneath the valof interseismic coupling in the Himalaya (Avouac, 2003, ley is the general result of India underthrusting Tibet, the re2015), is the largest in the NW (Fig. 10) where the Indomarkable thinning of the underlying crust by ∼4 km within Tibetan convergence undergoes a substantial decrease in the narrow confines of the Kashmir valley may be due to the velocity to ∼12 mm=yr, compared with ∼18–20 mm=yr to response of the basin to local extensional stresses in an overits east (Schiffman et al., 2013). all compressional framework. The 1905 M w 7.8 Kangra earthquake east of the Kashmir The most notable feature of the inverted shear-wave Himalaya most likely ruptured a northeastward dipping velocity models of the Kashmir Himalaya, after the Moho, 100 × 50 km2 plane, up-dip from the locking line in the reis the presence of an LVZ at a depth of ∼12–16 km, indicated gion (Wallace et al., 2005). The 2005 M w 7.6 Muzaffarabad predominantly by a negative phase on the RF between 2 and earthquake to its west also ruptured a northeastward dipping 3 s. Although most of the inverted models show the existence but steep (100 × 20 km2 ; dip ∼29°) plane eastward from the of an LVZ, at some stations the corresponding phase either hairpin bend in the MBT at the Hazara syntaxis (Avouac et al., suffers a time shift, changes sign, or almost disappears for 2006). Unfortunately, rupture parameters of the only major certain azimuths, possibly due to anisotropy (Schulte-Pelearthquake that occurred SW of the Kashmir basin in 1555 kum et al., 2005) and/or scattering. Thus, for example, at (M w 7.56; Ambraseys and Douglas, 2004), spanning the AHR, this phase at 1.5–3 s is strongly indicated for back region between the meizoseismal zones of the above two azimuths from 14° to 90°, arrives ∼0:4 s earlier for 90°–150°, earthquakes, are too poorly estimated to shed any light on the and virtually disappears thereafter, except for the 173° back seismogenic potential of the main thrusts. Recently, Vassallo azimuth at which it changes sign (see Ⓔ Fig. S12). At GND et al. (2015), using chronology of the geomorphic markers, on the eastern flank of the valley, this phase of the RFs, showed that deformation in the region during the late which arrives much earlier, appears to be weak for most aziQuaternary has been localized south of the MBT in the submuths, nonexistent for some, and possibly suffers a reversal Himalaya. Clearly, a comprehensive investigation of the strucbeyond 140° back azimuth. An intriguing aspect of this phase on the GND RFs is its rather early arrival at ∼1 s (see Ⓔ ture and deformation rates of the various arc-parallel sectors, SW

2454

(a)

R. R. Mir, I. A. Parvez, V. K. Gaur, Ashish, R. Chandra, and S. A. Romshoo 74.0°

75.0°

74.5°

75.5°

geological situations: presence of subducted sedimentary cover (Guo et al., GG GG G 60G GG 2009) of fluids (Caldwell et al., 2013) or G GG BBBBBBB 54 G G GG of a sheared zone (Schulte-Pelkum et al., B B B B B 60 BBBB B H G 60 G BBB L0´ HH GND H 56 BB 2005). The first can be clearly excluded H B H 54 55 G G B B HH 58 GG H B 54 BAR H 60GGG BBBBBB HH B GGGG G GG G beneath Kashmir on the basis of geologiB H G GG G G M M 59H H H G 62GGM GGG56 55 BBBBB G M M M M G G G G BB B GG G H M G 60 BB B BB54GG H G G G G G HAR cal mapping reported by several investigaM 60 H G HH G G HHH 58 H M M M GG58 BGUL BB56 G B M B B BB H H BB G G GGG GGG GG G tors (Farooqi and Desai, 1974; Bhatt, B H H H H 58 B GGG G 56 B B BBB B H HHH MAM B G G 34.0° M B GG GGG B B MM G BB M G G GGG GG 1982; Shellnutt et al., 2014; Vassallo et al., G G B B H HH G M M H BTP M M G BBBB B 58 B H G 2015) who conclude that the Kashmir val56 G B H F 58 AA ABAABAAA G G A BBBM AB G ley floor consists of Pir Panjal volcanics 56 AA A A AM M A A A A G A A 54 L3 A AM 54 A AA B B A AA overlying the metasedimentary Lesser HiB M AA AAAAA AAAA AA 54 A A A 56 A ANG A A A° AHR G° GND A A A A 58 A AA AA AAA A A 56 A A A malaya. A decisive discriminant between 54 A A A 54 A AA AA A ANG G GUL A AAAAA 58 AHR A B° BAR H HAR AAA A A AA A AA AA the presence of fluids or a sheared zone A AAA A56 A B BTP M MAM AAA AAA AA A A can possibly be provided by the absence A A AAA A A A MC AA T AA A 33.5° A or presence of anisotropy in this LVZ. A A O A F A A AAAA L2 57 Although this was possible in one of the MB L1 several Himalayan segments investigated T L0 by broadband seismology (SchultePelkum et al., 2005), it was rendered 0.0 1000.0 2000.0 3000.0 4000.0 5000.0 6000.0 7000.0 8000.0 Elevation (m) infeasible in our case because of the ubiquitous presence of scatterers in the (b) (c) subsurface. Nabelek et al. (2009) suggest that the low velocities at a depth of 10–15 km in the Lesser Nepal Himalaya are caused by water released by the sediment cover of the Indian plate that underthrusts the Lesser Himalaya and facilitates low-angle thrusting by induced overpressure of trapped fluids in the fault gouge. On the (d) (e) other hand, Caldwell et al. (2013) argue that thrust planes such as the MHT are generally characterized by a negative impedance contrast because the overthrusting material is usually composed of older deeply exhumed rocks of higher velocity. Indeed, they stress the point that certainty in measuring the precise depth of Figure 9. (a) Map view of the Kashmir valley bounded to the SW by the Pir Panjal MHT depends on the assumption that it is range and to the NE by the Zanskar range. The four sections across the valley highlight the form of RFs and, in particular, the crustal thinning underneath the valley. The PPs are characterized by a peak negative ammarked by an initial letter of their respective station name; stations having same initial plitude. letter are differentiated by a degree symbol (see legend). (b)–(e) RFs calculated for variIn Kashmir, the MCT that raised the ous sets of PPs along four profiles crossing the valley from SW to NE, marked in (a). All southwestern Pir Panjal range to create four profiles are measured from the valley axis (vertical line). The BF is the geologicallythe valley passes beneath the valley and inferred Balapora fault (Ahmad et al., 2014) paralleling the valley axis to its southeast (SE). The color version of this figure is available only in the electronic edition. merges into the décollement to its NE. The small velocity contrast between the volcanics (V S ∼ 3:2–3:5 km=s) and the underlying metasediFig. S12), indicating a northeastward shallowing of the LVZ mentary rocks (V S ∼ 3:59–3:6 km=s) makes it difficult to associated with a thickening of the underlying Indian crust. determine how this thrust plane is related to the relative disThe existence of low-velocity layers at similar or positions of these formations, or if we have any independent slightly greater depths in the Himalayan crust have been evidence of the presence of fluids through their highreported from Sikkim (Acton et al., 2011), Nepal (Schulteconductivity signature. Notwithstanding this uncertainty, Pelkum et al., 2005; Nabelek et al., 2009), and the Central the low velocities at a depth of ∼12–16 km beneath the valHimalaya (Caldwell et al., 2013). Low-velocity layers at ley signify markedly different physical properties, and we infer them to be related to the processes that facilitate seismic ∼10 km depth within the crust may represent one of the three L3´

L2´ °

°

°

L1´

°°

°°

°

°

° °

°

° ° °° °

°

°°

°

°

°°

°°

°

° ° ° °° °

°

° °

° °° ° °

°

°°

°

°

° °°

°

°

°

°

°°

°

°

° ° ° °°

°

°° °

°

°

° °

°°

°

°°

° ° ° °°

°

°°

°

° °° °

° °° ° °° ° ° °° °

°

°

°

°°

° ° °° ° ° ° ° ° ° ° °°

°

°

°

°

° °

Elevation (km)

°

°

° °

° ° °

°

°

° ° °° ° °

°

° °

° ° °

Elevation (km)

° ° °

°

°

°

°

°° °° ° ° °° °

Elevation (km)

°°°

Elevation (km)

°°

°

34.5°

Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis

2455

cally-inferred Balapora fault (BF). Earthquakes in the region are found to occur throughout the crust, as reported for other 60 parts of the Indian crust (e.g., Maggi et al., 34˚ 58 2000; Hetenyi et al., 2006), showing that A the whole of the Indian crust is seis32˚ 56 mogenic. The meizoseismal zone of the Muzaf54 30˚ Kashmir Valley farabad earthquake (Figs. 1 and 11a) exB hibits a bimodal distribution of events in C 50 28˚ depth, being more concentrated in the 57 50 54 D E depth ranges of 10  5 and 30  5 km. G F This feature, which is distinctive of the 26˚ 72˚ 74˚ 76˚ 78˚ 80˚ 82˚ 84˚ 86˚ 88˚ 90˚ 92˚ 94˚ 96˚ boundary region between the Kashmir segment of the Himalayan arc and that Figure 10. The location of the Kashmir valley fronted by the Pir Panjal range of the to its NW, may be caused by the interplay Great Himalaya adjoining the Hazara syntaxial bend. Lines across the range mark the various broadband seismic profiles from the western Himalaya to the eastern. A, current between stresses along this boundary and study; B, Rai et al. (2006); C, Gilligan et al. (2015); D, Nabelek et al. (2009); E, Schultethose arising from convergence across the Pelkum et al. (2005); F, Acton et al. (2011); and G, Hazarika (2007). Numbers (depths in changing strike of the orogen. Beneath the kilometers) indicate both eastward and westward thickening of the crust beneath the Kashmir basin (Fig. 11b), most events are 3500-m-high contour that approximately follows the northern edge of the locked Main concentrated between 15 and 40 km depth Himalayan thrust (MHT; Avouac, 2003, 2015). The color version of this figure is available only in the electronic edition. in a 30-km-wide linear band, placing them entirely within the underthrusting Indian plate below the LVZ, which we identify with the MHT. This slip. This inference is supported by the geometry and thickabsence of seismicity above the inferred MHT implies that ness of the MHT mapped by Schulte-Pelkum et al. (2005) in the overthrusting Himalayan layer is rigid but does not shed the Nepal Himalaya on the basis of their identification of an any light on its degree of coupling with the Indian crust. ∼6-km-thick anisotropic layer stretching from a depth of However, the absence of seismicity in the regions NE and ∼8 km under the foothills to about 20 km south of the High SW of the valley suggests that the valley spans the transiHimalaya, arguing that the anisotropic fabric is caused by tional region NE of which the Indian plate slides steadily into shear on the décollement. Schulte-Pelkum et al. (2005), Tibet but is locked to the Lesser Himalaya to its SW. The who had been led to suspect the presence of an anisotropic locus of this seismicity is spatially offset ∼15 km to the SE layer by azimuthal variations in polarity reversals of the conof the center of the ∼25-km-wide transition zone derived verted phases from this horizon, were able to confirm its from the GPS velocity field by Schiffman et al. (2013) existence by reproducing this result when this layer was but corresponds more closely to the transition derived by modeled as having 20% anisotropy. Stevens and Avouac (2015), who used a combination of geoIn the Central Himalaya west of Nepal, Caldwell et al. detic and microseismic data to provide additional constraints. (2013) also image the MHT at a depth of ∼10 km on the Both of these geodetically derived models suggest wider basis of a negative impedance contrast, which they ascribe zones of incomplete seismic coupling than does the narrow to the presence of ponded fluids beneath the décollement, zone of microseismicity depicted in Figure 11b. Finally, the indicated by the existence of a nearly coincident horizon seismicity in the valley along its axis shows a concentration of high electrical conductivity (Rawat et al., 2014). Although below a depth of ∼15 km, marking a region of relatively we lack other evidence, such as anisotropy or electrical conhigher stresses. Southeastward of the valley, at longitude ductivity, we consider it most likely that the LVZ identified 75.75° E near Kishtwar (Fig. 11c), seismicity becomes difby RFs in the valley represents the MHT for the reasons disfuse and somewhat shallower, indicating a thickening of the cussed above, and the reflector mapped by Kaila et al. (1984) seismogenic crust beneath the valley and further to its at a depth of ∼10 km beneath the MBT marks its southward NW (Fig. 11). up-dipping extension. In conclusion, we determine the following features of The background seismicity clearly demarcates the Kashthe crust beneath the Kashmir valley that had not been remir Himalaya by high-strain zones on either end of the valley ported earlier. First, we present the first shear-wave velocity (Fig. 1), which segment it from its eastern and western neighstructure of the Kashmir crust, which allows one to map the bors. This segment corresponds to a current seismic gap beMoho beneath the valley and greatly improve the location of tween the inferred 1905 Kangra and the 2005 Muzaffarabad earthquakes in the region from the knowledge of wellrupture zones. Events recorded by our network over a fiveconstrained crustal velocities. These results show that the month period point to a concentration of microseismicity Kashmir Himalayan crust is thicker than in the Central Himalaya (Fig. 10), and the Moho beneath the region, while (Figs. 1 and 11) along the valley axis, NE of the geologi36˚

3.5k Altitude Faults Seismic profile across Himalaya−Tibet

Hazara Syntaxis

MF

T

2456

R. R. Mir, I. A. Parvez, V. K. Gaur, Ashish, R. Chandra, and S. A. Romshoo A1−A1′

(a)

L2−L2′

(b)

Distance (km) −84

−42

Distance (km)

Distance (km) 0

−168−126 −84−42 0

42 84 126

−42

0

42

0

0

−30

−30 −60

−60 ISC

CS −90

−90

Number of earthquakes 10

20

30

40

Number of earthquakes 0

5

10

Number of earthquakes 15

0

5

10

20 40

60

60

40

Depth (km)

20

Depth (km)

15 0

0

0

Depth (km)

Depth (km)

−126

A0−A0′

(c)

ISC

CS

Figure 11.

Depth distribution of earthquakes along three profiles (A1–A1′; L2–L2′; A0–A0′) marked in Figure 1, with distances reckoned from the valley axis. Circles denoted by ISC represent relocated events from the ISC phase (P and S) data catalog (1964–2013), whereas the ones denoted by CS represent locally recorded events. (a) Note the bimodal distribution of seismicity at ∼10 and ∼30 km depth, and (b) the local earthquakes are focused along the valley axis (also see Fig. 1), which may represent locked portion of the décollement (SchultePelkum et al., 2005) or the beginning of a transition zone. The depth of earthquakes is shallower along (c) A0–A0′ than (b) L2–L2′ and (a) A1–A1′, which is probably due to the increase in thickness of seismogenic crust as we move westward from (c) to (a). The color version of this figure is available only in the electronic edition.

steadily dipping toward the NE, is inflected at the margins of the valley but nearly flat underneath with an upwarp. However, we refrain from overinterpreting this remarkable feature of the Moho in this region, in the absence of other evidence that could have been produced by experimental data on the gravity and thermal fields of the region. Second, we show that the Kashmir crust hosts a lowvelocity layer at a depth of ∼12–16 km, as found in Himalayan segments to the east, possibly representing the décollement, separating the subducting Indian plate and the overlying Himalayan crust in the region, even as it shows an unexpected shallowing near the northeastern margin of the valley. The upper surface of this layer corresponds to a nearby reflector mapped by Kaila et al. (1984), and we interpret it as a plausible candidate for rupture in a future earthquake. Third, using double-difference relocated seismicity in the region, we delineate the existence of two prominent highstrain zones (Fig. 1) almost perpendicular to the Himalayan arc in the region that clearly segments the Kashmir Himalaya from its neighbors, thereby refining the location of the current seismic gap between the 2005 Muzaffarabad earthquake and the 1905 Kangra earthquake. It is possible that this gap last slipped in 1555 or earlier (Kaneda et al., 2008). Finally, from

the pattern of depth distribution of earthquakes along three profiles perpendicular to the Himalaya, including one right through the middle of the valley, we show that the thickness of the seismogenic crust of the Kashmir Himalaya is greater than that to its east and that it increases further to its west around the region of the 2005 Muzaffarabad earthquake.

Data and Resources Nilore (NIL) data were downloaded from Incorporated Research Institutions for Seismology (IRIS; https://www.iris .edu/hq/programs/gsn, last accessed January 2015). Figures were generated with Generic Mapping Tools (Wessel et al., 2013) and Gnuplot (http://gnuplot.info, last accessed December 2011). Data analysis was done using the Seismic Analysis Code (Goldstein and Snoke, 2005) and piercing points at Moho were calculated using the TauP package (Crotwell et al., 1999). Phase data of earthquakes (1964–2013) were downloaded from the International Seismological Centre (ISC, 2013), United Kingdom (last accessed December 2015), and the Shuttle Radar Topography Mission (SRTM) data were taken from http://www.jpl.nasa.gov/srtm/ (last accessed June 2015). The broadband data used in this article are available upon request from the authors.

Crustal Structure beneath the Kashmir Basin Adjoining the Western Himalayan Syntaxis

Acknowledgments This study was supported by the Advanced Research in Engineering and Earth Sciences (ARiEES) project of the Council of Scientific and Industrial Research of the Government of India and greatly facilitated by logistic supports from the former Vice Chancellor of Kashmir University, Talat Ahmad, and several students of the Earth Science Department of Kashmir University. The authors particularly acknowledge the inspiration gained from the pioneering initiative of Roger Bilham, which drew attention to the existence of a wide zone of incomplete coupling on the décollement of the Kashmir Himalaya and for his incisive comments on this article. The authors also acknowledge the painstaking review of this article and critical comments made by Associate Editor Thomas Brocher and other anonymous reviewers and the suggestions made by Walter Mooney, all of which greatly contributed to improving the original article. A temporary loan of four broadband sensors by the Indian Institute of Astrophysics is also gratefully acknowledged.

References Acton, C. E., K. Priestley, S. Mitra, and V. K. Gaur (2011). Crustal structure of the Darjeeling–Sikkim Himalaya and southern Tibet, Geophys. J. Int. 184, 829–852. Ader, T., J.-P Avouac, J. Liu-Zeng, H. Lyon-Caen, L. Bollinger, J. Galetzka, J. Genrich, M. Thomas, K. Chanard, and S. N. Sapkota (2012). Convergence rate across the Nepal Himalaya and interseismic decoupling on the main Himalayan thrust: Implications for seismic hazard, J. Geophys. Res. 117, no. B044403, doi: 10.1029/2011JB009071. Ahmad, S., M. I. Bhat, C. Madden, and B. S. Bali (2014). Geomorphic analysis reveals active tectonic deformation on the eastern flank of the Pir Panjal range, Kashmir valley, India, Arab. J. Geosci. 7, 2225–2235. Ambraseys, N. N., and J. Douglas (2004). Magnitude calibration of north Indian earthquakes, Geophys. J. Int. 159, 165–206. Avouac, J.-P. (2003). Mountain building, erosion and the seismic cycle in the Nepal Himalaya, in Advances in Geophysics, R. Dmowska (Editor), Elsevier, San Diego, California, 1–80. Avouac, J.-P. (2015). Mountain building: From earthquakes to geologic deformation, in Treatise on Geophysics, Second Ed., G. Schubert (Editor-in-Chief), Vol. 6, Elsevier, Oxford, England, 381–432. Avouac, J.-P., F. Ayoub, S. Leprince, O. Konca, and D. V. Helmberger (2006). The 2005, Mw 7.6 Kashmir earthquake: Sub-pixel correlation of ASTER images and seismic waveforms analysis, Earth Planet. Sci. Lett. 249, 514–528, doi: 10.1016/j.epsl.2006.06.025. Bettinelli, P., J.-P. Avouac, M. Flouzat, F. Jouanne, L. Bollinger, P. Willis, and G. R. Chitrakar (2006). Plate motion of India and interseismic strain in the Nepal Himalaya from GPS and DORIS measurements, J. Geodes. 80, 567–589. Bhatt, D. K. (1982). A review of the stratigraphy of the Karewa Group (Pliocene/Quaternary), Kashmir, Man Environ. 6, 46–55. Bilham, R., B. S. Bali, S. Ahmad, and C. Schiffman (2013). Oldham’s lost fault, Seismol. Res. Lett. 84, 702–710, doi: 10.1785/0220130036. Bilham, R., K. Larson, J. Freymueller, and Project Idylhim members (1997). GPS measurements of present-day convergence across the Nepal Himalaya, Nature 386, 61–64. Bostock, M. G., and S. Rondenay (1999). Migration of scattered teleseismic body waves, Geophys. J. Int. 137, 732–746. Brocher, T. M. (2005). Empirical relations between elastic wavespeeds and density in the Earth’s crust, Bull. Seismol. Soc. Am. 95, 2081–2092, doi: 10.1785/0120050077. Burbank, D. W., and G. D. Johnson (1982). Intermontane basin development in the past 4 Myr in the north-west Himalaya, Nature 298, 432–436. Burg, J. P., and Y. Podladchikop (2000). From buckling to asymmetric folding of the continental lithosphere: Numerical modelling and application to Himalayan syntaxis, in Tectonics of the Nanga Parbat Syntaxis and the Western Himalaya, M. A. Khan, P. J. Treloar, M. P. Searle, and M. Q. Jan (Editors), Geol. Soc. Lond. Spec. Publ. Vol. 170, 219–236.

2457

Caldwell, W. B., S. L. Klemperer, J. F. Lawrence, S. S. Rai, and Ashish (2013). Characterizing the Main Himalayan thrust in the Garhwal Himalaya, India with receiver function CCP stacking, Earth Planet. Sci. Lett. 367, 15–27, doi: 10.1016/j.epsl.2013.02.009. Christensen, N. I. (1996). Poisson’s ratio and crustal seismology, J. Geophys. Res. 101, no. B2, 3139–3156. Crotwell, H. P., and T. J. Owens (2005). Automated receiver function processing, Seismol. Res. Lett. 76, 702–708. Crotwell, H. P., T. J. Owens, and J. Ritsema (1999). The TauP toolkit: Flexible seismic travel time and raypath utilities, Seismol. Res. Lett. 70, 154–160. DeCelles, P. G., D. M. Robinson, J. Quade, T. P. Ojha, C. N. Garzione, P. Copeland, and B. N. Upreti (2001). Stratigraphy, structure, and tectonic evolution of the Himalayan fold-thrust belt in western Nepal, Tectonics 20, 487–509. Eagar, K. C., and M. J. Fouch (2012). FuncLab: A MATLAB interactive toolbox for handling receiver function datasets, Seismol. Res. Lett. 83, 596–603, doi: 10.1785/gssrl.83.3.596. Eaton, D. W., S. Dineva, and R. Mereu (2006). Crustal thickness and V P =V S variations in the Grenville orogen (Ontario, Canada) from analysis of teleseismic receiver functions, Tectonophysics 420, 223–238, doi: 10.1016/j.tecto.2006.01.023. Farooqi, I. A., and R. N. Desai (1974). Stratigraphy of Karewas, India, J. Geol. Surv. India 15, 299–305. Fontaine, F. R., H. Tkalcic, and B. L. N. Kennett (2013). Crustal complexity in the Lachlan orogen revealed from teleseismic receiver functions, Aust. J. Earth Sci. 60, 413–430, doi: 10.1080/08120099.2013.787646. Gilligan, A., K. F. Priestley, S. W. Roecker, V. Levin, and S. S. Rai (2015). The crustal structure of the western Himalayas and Tibet, J. Geophys. Res. 120, 3946–3964, doi: 10.1002/2015JB011891. Goldstein, P., and A. Snoke (2005). SAC Availability for the IRIS Community, Incorporated Research Institutions for Seismology Data Management Center, Electronic Newsletter, available at http://ds.iris .edu/ds/newsletter/vol7/no1/sac‑availability‑for‑the‑iris‑community/ (last accessed January 2013). Guo, Z., X. Gao, H. Yao, J. Li, and W. Wang (2009). Midcrustal low-velocity layer beneath the central Himalaya and southern Tibet revealed by ambient noise array tomography, Geochem. Geophys. Geosys. 10, Q05007, doi: 10.1029/2009GC002458. Hazarika, N. K. (2007). Receiver function analysis of broadband seismic data from various seismic stations and modelling the crustal structure of Northeast India, Ph.D. Thesis, Tezpur University, Assam, India, 182. Herrmann, R. B. (2013). Computer programs in seismology: An evolving tool for instruction and research, Seismol. Res. Lett. 84, 1081–1088, doi: 10.1785/0220110096. Hetenyi, G., R. Cattin, J. Vergne, and J. K. Nabelek (2006). The effective elastic thickness of the Indian Plate from receiver function imaging, gravity anomalies and thermo-mechanical modeling, Geophys. J. Int. 167, 1106–1118. International Seismological Centre (ISC) (2013). On-line Bulletin, International Seismological Centre, Thatcham, United Kingdom, available at http://www.isc.ac.uk (last accessed December 2015). Jackson, M., and R. Bilham (1994). Constraints on Himalayan deformation inferred from vertical velocity fields in Nepal and Tibet, J. Geophys. Res. 99, 13,897–13,912. Julia, J., C. J. Ammon, R. B. Herrmann, and A. M. Correig (2000). Joint inversion of receiver function and surface wave dispersion observations, Geophys. J. Int. 143, 99–112. Kaila, K. L., K. R. Chowdhury, V. G. Krishna, M. M. Dixit, and H. Narain (1982). Crustal structure of Kashmir Himalaya and inferences about the asthenosphere layer from DSS studies along the international profile, Qarrakol–Zorkol–Nanga Parbat–Srinagar–Pamir Himalaya monograph, Boll. Geof. Teor. Appl. 25, 221–234. Kaila, K. L., V. G. Krishna, K. Chowdhury, and H. Narain (1978). Structure of the Kashmir Himalaya from deep seismic soundings, J. Geol. Soc. India 19, 1–20.

2458

R. R. Mir, I. A. Parvez, V. K. Gaur, Ashish, R. Chandra, and S. A. Romshoo

Kaila, K. L., K. M. Tripathi, and M. M. Dixit (1984). Crustal structure along Wular Lake–Gulmarg–Naoshera profile across Pir Panjal range of the Himalayas from deep seismic soundings, J. Geol. Soc. India 25, 706–719. Kaneda, H., T. Nakata, H. Tsutsumi, H. Kondo, N. Sugito, Y. Awata, S. S. Akhtar, A. Majid, W. Khattak, A. A. Awan, et al. (2008). Surface rupture of the 2005 Kashmir, Pakistan, earthquake and its active tectonic implications, Bull. Seismol. Soc. Am. 98, 521–557. Kennett, B. L. N., E. R. Engdahl, and R. Buland (1995). Constraints on seismic velocities in the earth from travel times, Geophys. J. Int. 122, 108–124. Ligorria, J. P., and C. J. Ammon (1999). Iterative deconvolution and receiver-function estimation, Bull. Seismol. Soc. Am. 89, 1395–1400. Maggi, A., J. A. Jackson, K. Priestley, and C. Baker (2000). A re-assessment of focal depth distributions in southern Iran, the Tien Shan and northern India: Do earthquakes really occur in the continental mantle? Geophys. J. Int. 143, 629–661. Mitra, S., K. Priestley, V. K. Gaur, S. S. Rai, and J. Haines (2006). Variation of Rayleigh wave group velocity dispersion and seismic heterogeneity of the Indian crust and uppermost mantle, Geophys. J. Int. 164, 88–98, doi: 10.1111/j.1365-246X.2005.02837.x. Nabelek, J., G. Hetenyi, J. Vergne, S. Sapkota, B. Kafle, M. Jiang, H. Su, J. Chen, B. Huang, and the Hi-CLIMB Team (2009). Underplating in the Himalaya–Tibet collision zone revealed by the Hi-CLIMB experiment, Science 325, 1371–1374, available at http://science.sciencemag.org/ content/325/5946/1371 (last accessed November 2015). Paige, C. C., and M. A. Saunders (1982). LSQR: Sparse linear equations and least squares problems, ACM Trans. Math. Software 8/2, 195–209. Rai, S. S., K. Priestley, V. K. Gaur, S. Mitra, M. P. Singh, and M. Searle (2006). Configuration of the Indian Moho beneath the NW Himalaya and Ladakh, Geophys. Res. Lett. 33, L15308, doi: 10.1029/ 2006GL026076. Rawat, G., B. R Arora, and P. K. Gupta (2014). Electrical resistivity cross-section across the Garhwal Himalaya: Proxy to fluid seismicity linkage, Tectonophysics 637, 68–79. Reasenberg, P. (1985). Second-order moment of central California seismicity, J. Geophys. Res. 90, 5479–5495. Rychert, A. R., S. Rondenay, and K. M. Fischer (2007). P-to-S and S-to-P imaging of a sharp lithosphere-asthenosphere boundary beneath eastern North America, J. Geophys. Res. 112, no. B08314, doi: 10.1029/2006JB004619. Schiffman, C., B. S. Bali, W. Szeliga, and R. Bilham (2013). Seismic slip deficit in the Kashmir Himalaya from GPS observations, Geophys. Res. Lett. 40, 5642–5645, doi: 10.1002/2013GL057700. Schulte-Pelkum, V., G. Monsalve, A. Sheehan, M. Pandey, S. Sapkota, R. Bilham, and F. Wu (2005). Imaging the Indian subcontinent beneath the Himalaya, Nature 435, 1222–1225. Shah, A. (2013). Earthquake geology of Kashmir basin and its implications for future large earthquakes, Int. J. Earth Sci. 102, 1957–1966. Shellnutt, J. G., G. M. Bhat, K.-L. Wang, M. E. Brookfield, B.-M. Jahn, and J. Dostal (2014). Petrogenesis of the flood basalts from the Early Permian Panjal Traps, Kashmir, India: Geochemical evidence for shallow melting of the mantle, Lithos 204, 159–171, doi: 10.1016/ j.lithos.2014.01.008. Stevens, V. L., and J.-P. Avouac (2015). Interseismic coupling on the main Himalayan thrust, Geophys. Res. Lett. 42, 5828–5837.

Vassallo, R., J. L. Mugnier, V. Vignon, M. A. Malik, R. Jayangondaperumal, P. Srivastava, and J. Carcaillet (2015). Distribution of the LateQuaternary deformation in northwestern Himalaya, Earth Planet. Sci. Lett. 411, 241–252, doi: 10.1016/j.epsl.2014.11.030. Vernant, P., R. Bilham, W. Szeliga, D. Drupka, S. Skalita, A. Bhattacharyya, V. K. Gaur, P. Pelgay, R. Cattin, and T. Berthet (2014). Clockwise rotation of the Brahmaputra valley: Tectonic convergence in the eastern Himalaya, Naga Hills and Shillong plateau, J. Geophys. Res. 119, 6558–6571, doi: 10.1002/2014JB011196. Waldhauser, F., and W. L. Ellsworth (2000). A double-difference earthquake location algorithm: Method and application to the northern Hayward fault, California, Bull. Seismol. Soc. Am. 90, 1353–1368. Wallace, K., R. Bilham, F. Blume, V. K. Gaur, and V. Gahalaut (2005). Surface deformation in the region of the 1905 Kangra M w 7.8 earthquake in the period 1846–2001, Geophys. Res. Lett. 32, L15307, doi: 10.1029/2005GL022906. Wanchoo, S. K., D. Powali, S. Sharma, S. Mitra, K. F. Priestley, and V. K. Gaur (2014). Lithospheric structure and earthquakes beneath Kashmir Himalaya, AGU Fall Meeting, Abstract T21B–4587. Wessel, P., W. H. F. Smith, R. Scharroo, J. F. Luis, and F. Wobbe (2013). Generic Mapping Tools: Improved version released, Eos Trans. AGU 94, 409–410. Wiemer, S. (2001). A software package to analyze seismicity: ZMAP, Seismol. Res. Lett. 72, 374–383. Yin, A. (2006). Cenozoic tectonic evolution of the Himalayan orogen as constrained by along-strike variation of structural geometry, exhumation history, and foreland sedimentation, Earth Sci. Rev. 76, 1–131. Zhu, L. P., and H. Kanamori (2000). Moho depth variation in southern California from teleseismic receiver functions, J. Geophys. Res. 105, 2969–2980. Academy of Scientific and Innovative Research (AcSIR) CSIR-Fourth Paradigm Institute (CSIR-4PI) Campus Bangalore 560037, Karnataka, India [email protected] [email protected] (R.R.M., I.A.P.)

CSIR Fourth Paradigm Institute (CSIR-4PI) NAL Belur Campus, Wind tunnel Road Bangalore 560037, Karnataka, India [email protected] [email protected] (V.K.G., A.)

Department of Earth Sciences University of Kashmir Srinagar 190006, Jammu and Kashmir, India [email protected] [email protected] (R.C., S.A.R.) Manuscript received 26 November 2015; Published Online 25 September 2017