Você está na página 1de 15

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 115, XXXXXX, doi:10.

1029/2009JB007009, 2010

Seismic anisotropy beneath the Indian continent from splitting 2 of direct S waves
1 3

Dipankar Saikia,1 M. Ravi Kumar,1 Arun Singh,1 G. Mohan,2 and R. S. Dattatrayam3

4 Received 26 September 2009; revised 2 June 2010; accepted 24 August 2010; published XX Month 2010. 5 [1] In this study, we investigate the shear wave splitting beneath 39 broadband seismic 6 stations located over various tectonic units of the Indian shield using direct S waves 7 from deep to intermediate depth earthquakes. The delay times between fast and slow 8 axes of anisotropy do not reveal anomalous values that reflect contributions from 9 midmantle or the mantle transition zone. The fast polarization azimuths are generally 10 consistent with the results from SKS splitting that indicate a predominance of plate 11 motion related strain in the Indian continent. Splitting results at station HYB, which was 12 hitherto considered isotropic, indicate significant anisotropy, with a NE oriented fast 13 axis. Although a general trend of SH leading SV is indicated by most of the 14 measurements, invoking radial anisotropy to explain this phenomenon appears farfetched 15 since the near vertical incidence of the S waves in the usable distance range of 4080 16 makes them more sensitive to azimuthal anisotropy underneath the station. Forward 17 modeling of the dependence of the splitting parameters on incoming polarization 18 assuming horizontal axis of symmetry brings out two layers of anisotropy beneath the 19 Indian shield with fast axis azimuths oriented N15E in the bottom layer and N60E in 20 the top layer. While stronger anisotropy (dt 1 s) in the bottom layer is related to 21 asthenospheric flow, a weaker one (dt 0.4 s) in the upper layer could represent 22 anisotropy frozen in the lithosphere. 23 Citation: Saikia, D., M. Ravi Kumar, A. Singh, G. Mohan, and R. S. Dattatrayam (2010), Seismic anisotropy beneath the Indian 24 continent from splitting of direct S waves, J. Geophys. Res., 115, XXXXXX, doi:10.1029/2009JB007009. 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44

1. Introduction
[2] Seismic anisotropy within the Earths interior can be generally explained invoking a hexagonal axis of symmetry oriented either in a horizontal or vertical direction. Seismologists often refer to the former as azimuthal anisotropy and the latter as radial anisotropy or Vertically oriented transverse isotropy (VTI) in its simplest form. Vertically propagating core refracted phases like SK(K)S that are widely used to characterize anisotropy are sensitive only to azimuthal anisotropy since these waves do not exhibit splitting in a medium with radial anisotropy. While numerous studies of shear wave splitting using core refracted phases have established azimuthal anisotropy as an intrinsic property of the Earth [e.g., Vinnik et al., 1989; Silver and Chan, 1988], results from globalscale surface wave studies [Gung et al., 2003] reveal that radial anisotropy is also pervasive in the upper mantle beneath most cratons, in the depth range of 250400 km. In situations where anisotropy has an orthorhombic axis of symmetry, splitting can occur for any propagation direction albeit producing less azimuthal variaNational Geophysical Research Institute, Hyderabad, India. Department of Earth Sciences, Indian Institute of Technology Bombay, Mumbai, India. 3 India Meteorological Department, New Delhi, India.
2 1

Copyright 2010 by the American Geophysical Union. 01480227/10/2009JB007009

tions in splitting delay time (dt) and fast polarization azimuth (F), than those produced by hexagonal symmetry. [3] In contrast to the core refracted phases, the direct S phases are sensitive to both radial and azimuthal anisotropy since they travel a considerable length of their path horizontally in the midmantle. Also, their incidence angles in the upper mantle corresponding to the epicentral distances suitable for shear wave analysis may deviate significantly from vertical. Although direct S phases are more abundant, energetic and easy to identify on the three component seismograms, their utility for shear wave splitting measurements is less known compared to the core refracted phases. Unlike the converted phases like SKS and SKKS which do not preserve the source side anisotropy, the direct S phase is sensitive to anisotropy lying anywhere along the raypath, and it is often difficult to separate the source, receiver and midmantle contributions. One common means of avoiding the near source effects is to restrict the usable waveforms to those from earthquake sources deeper than the olivine stability level. If proper care is taken to avoid contamination from shear coupled waves, the usable window for direct S waves extends from 40 to 80. Splitting times that are unusually larger than the global average of 1 s [Silver, 1996] are attributed to anisotropy within the midmantle or the mantle transition zone. [4] Recent results of SKS splitting from the Indian region establish the anisotropic nature of the mantle underneath this
1 of 15

45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71

XXXXXX

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Figure 1. Distribution of broadband seismic stations (inverted triangles) in the Indian shield region superimposed on the different geological provinces indicated in the legend. SGT, Southern Granulite Terrain; EDC, Eastern Dharwar Craton; WDC, Western Dharwar Craton; CSZ, Chitradurga Shear Zone; NSL, Narmada Son Lineament; EG, Eastern Ghats; BC, Bhandara Craton; SC: Singhbum Craton; VB, Vindhyan Basin; DAFB, DelhiAravalli Fold Belt; GG, Godavari Graben; CB, Cuddapah Basin; SP, Shillong Plateau; WG, Western Ghats; FD, Foredeep; 1, Moyar; 2, Bhavani; 3, PalghatCauvery; 4, AchanKovil Shear Zones.
72 73 74 75 76 77 78 79

fast moving plate, similar to the other Precambrian shield regions [Kumar and Singh, 2008, Singh et al., 2006, 2007]. Surprisingly, station HYB situated on the Eastern Dharwar Craton in the south Indian shield has not shown any evi-

dences of anisotropy, which prompted some of the earlier researchers to consider the Indian plate as isotropic [Chen and Ozalaybey, 1998; Barruol and Hoffmann, 1999]. As a continuation of our previous study of the Indian shield

2 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109

meters on incoming polarization in terms of two layers of anisotropy, (3) inferring the existence of radial anisotropy (if any) in the upper mantle beneath India, (4) searching for possible evidences for midmantle anisotropy similar to that found for Australia, and (5) deciphering the nature of anisotropy at station HYB.

2. Splitting Analysis
[5] Direct S waves recorded by 39 broadband seismic stations located in different geologic provinces of the Indian shield are used in this study (Figure 1). These stations have been in operation for a sufficiently long period of time to produce good quality S waveforms. The distribution of the stations over the geological provinces, their deploying institutes and the approximate start time are described by Kumar and Singh [2008]. All the events used in this study are from earthquakes with magnitude 5.5 in the epicentral distance range of 35 to 80, primarily from subduction zones like the Banda Sea, southern part of JavaSumatra and the Philippines Sea. In order to minimize the contribution from source side anisotropy, the events are restricted to those having focal depths greater than 200 km, thus limiting the backazimuthal coverage to within 40140 range (Figure 2). [6] Although direct S waves are easy to identify on a seismogram, they are often contaminated by some energetic

Figure 2. Earthquakes (stars) with focal depths >200 km in the epicentral distance range of 4080 from the stations (inverted triangles) used for measurement of shear wave splitting. Lines joining them show the great circle paths followed by the seismic waves.
80 81 82 83 84

anisotropy using core refracted phases, the present study using direct S phases is aimed at (1) comparing the anisotropic parameters obtained from S phases with those derived from SKS splitting, including a few new measurements from this study, (2) modeling the dependence of splitting para-

Figure 3. Schematic diagram showing the raypaths of shear coupled phases like (a) SsPmp, Sp, Ss and (b) PL that arrive in close proximity of the direct S waves. Solid and dashed lines represent the P and S raypaths.
3 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Figure 4. Example waveforms of radial and transverse components at six different stations showing shear coupled phases after the arrival of the direct S wave. Hypocentral parameters of the earthquakes corresponding to the waveforms are indicated in the top left corner of the waveform pairs along with their epicentral distance and back azimuth.

110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136

phases, which arrive a few seconds after the S arrival, as also seen in the waveforms that are used in our analysis. These phases could be either shear coupled P, PcS or free surface reflections of the S phase like SsPmP which arrive 613 s (considering the Moho depth as 35 km) after its arrival, for epicentral distances in the range of 3580 [Heintz, 2006; Wookey and Kendall, 2004]. As predicted by the IASP91 Earth model, the PcS phase should come close to the direct S phase for epicentral distances around 40. There are a variety of ways through which teleseismic SV waves can be coupled to a P wave. A schematic diagram of the raypaths of some of the prominent shear coupled phases are shown in Figure 3. The Sp phase, a PtoS conversion from the Moho, arrives at the station as a precursor to the direct S arrival [Jordan and Frazer, 1975]. On the other hand, the SsPmP phase is generated by the incidence of an S wave at the crust mantle boundary, its subsequent refraction through the crust and conversion to a P wave upon free surface reflection before finally getting reflected postcritically from the Moho to reach the station as a P wave [Langston, 1996; Zandt and Randall, 1985]. This phase is observed at both regional and teleseismic distances with a travel time cross over that depends upon the depth and distance of the source. For regional distances, the SsPmP phase arrives before the direct S and for teleseismic distances it arrives after the direct S arrival [Langston, 1996]. Constructive interference of the rays generated from a direct S wave through mul-

tiple post critical reflection produces another kind of phase called the shear coupled PL or SPL (Figure 3b), first observed and theorized by Oliver [1961] [Zandt and Randall, 1985; Poupinet and Wright, 1972; Baag and Langston, 1985]. This phase sometimes appears as a long wave train after direct S arrival for both regional and teleseismic distances. [7] Figure 4 shows some examples of waveforms from the Indian shield stations where phases with nonnegligible energy can be clearly seen after the direct S phase. Treating the waveforms with a lowfrequency filter makes these phases difficult to identify and increases the possibility of their leakage into the direct S wave analysis window. Inclusion of shear coupled phases often alters the delay time and fast polarization measurements significantly and decreases their reliability. An example of such a filtering effect using good quality waveforms recorded by station TRVM corresponding to an event at 7.83N, 122.59E at a depth of 265 km is depicted in Figure 5. In this case, the data are windowed around the direct S phase using the predicted travel time from IASP91 Earth model and taking the window length as 100 s. In the first measurement, the data are filtered with a Butterworth filter having corners at 1 and 50 s, while in the second case the corner frequencies are chosen at 8 and 50 s (lowfrequency filter). It can be seen that the highfrequency filter which excludes a small peak after the direct S phase produces a wellconstrained (with

137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163

4 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Figure 5. Example demonstrating the influence of shear coupled phases on the estimation of splitting parameters due to their inclusion in the window used for analysis. Results obtained using a Butterworth filter with corners at (a) 1 and 50 s and (b) 8 and 50 s. (left) In each case, the top two traces show the radial and transverse components uncorrected for anisotropy, and the bottom two traces represent the components in the direction of initial polarization and orthogonal to it after correcting for anisotropy. (middle) Fast and slow waveforms and their particle motions (left) before and (right) after correction. (right) The contour plot of energy. Lines labeled A and F indicate the start and end of the shear wave analysis window, while line labeled S represents the S wave pick.
164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197

small error) measurement with a delay time of 1.15 s and a fast direction of 1. On the other hand, inclusion of the shear coupled phase results in a larger delay time of 2.75 s and a fast direction of N22E, with a relatively larger amount of error. This is because the lowfrequency filter smears the shear coupled phase, thus resulting in a drastic change in the measured splitting parameters. To avoid contamination of the analysis through inclusion of these shear coupled waves, we tried to perform the analysis using the raw (unfiltered) data, which was, however, not possible for the whole data set. Whenever filtering became necessary, a series of Butterworth filters were used to check the stability of the results. The filters were varied by changing one end from 1 to 5 s while keeping the other end fixed at 50 s. After processing all the available waveforms we finally retained only 105 with highest quality, based on the following quality criteria: (1) Clarity of the radial and transverse compo-

nents of the waveforms, (2) roughly elliptical particle motion prior to and linear particle motion after correcting for anisotropy, (3) similarity of the fast and slow components after correction, and (4) unique, welldefined error surface. Moreover, for the waveforms where the effects of shear coupled P wave are not quite clear, only those results which are stable for at least four different filters are included in this study. [8] The widely used shear wave splitting technique of [Silver and Chan, 1991] has been recently automated [Teanby et al., 2004] to reduce the subjectivity of the splitting measurements arising due to the choice of the window used for the analysis. In this scheme, the grid search is performed over a range of window lengths to obtain an optimum window for the splitting correction. A simple search for the window that gives the lowest error bars in the measurements is not a good criterion, since an unstable result that is

5 of 15

XXXXXX
t1:1 t1:2 t1:3 t1:4 t1:5 t1:6 t1:7 t1:8 t1:9 t1:10 t1:11 t1:12 t1:13 t1:14 t1:15 t1:16 t1:17 t1:18 t1:19 t1:20 t1:21 t1:22 t1:23 t1:24 t1:25 t1:26 t1:27 t1:28 t1:29 t1:30 t1:31 t1:32 t1:33 t1:34 t1:35 t1:36 t1:37 t1:38 t1:39 t1:40 t1:41 t1:42 t1:43 t1:44 t1:45 t1:46 t1:47 t1:48 t1:49 t1:50 t1:51 t1:52 t1:53 t1:54 t1:55 t1:56 t1:57 t1:58 t1:59 t1:60 t1:61 t1:62 t1:63 t1:64 t1:65 t1:66 t1:67 t1:68 t1:69 t1:70 t1:71

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Table 1. Individual Splitting Parameters at All the Stations Along With Their Uncertainties
Station AGWD AKL AKL AKL BHUJ BHUJ BOM BOM KAND KAND KARD KARD KARD KARD KARD KARD KARD KHER KOY MPAD MULG PUNE PUNE PUNE SKP SKP TANA TANA TANA WAG KOD KOD MDRS PCH TRVM TRVM TRVM KGF KGF KGF HYB HYB HYB HYB HYB HYB HYB HYB HYB DHD DHD DHD DHD MNGR MNGR MNGR KGD VISK VISK VISK VISK VISK BWNR Station Lat 23.96 20.70 20.70 20.70 23.25 23.25 18.90 18.90 20.90 20.90 17.31 17.31 17.31 17.31 17.31 17.31 17.31 20.81 17.41 20.48 19.17 18.53 18.53 18.53 17.00 17.00 21.58 21.58 21.58 17.34 10.23 10.23 13.07 10.53 8.51 8.51 8.51 13 13 13 17.41 17.41 17.41 17.41 17.41 17.41 17.41 17.41 17.41 15.30 15.30 15.30 15.30 12.94 12.94 12.94 17.66 17.72 17.72 17.72 17.72 17.72 20.29 Station Lon 71.93 77.02 77.02 77.02 69.65 69.65 72.81 72.81 71.10 71.10 74.18 74.18 74.18 74.18 74.18 74.18 74.18 73.02 73.75 73.81 73.30 73.85 73.85 73.85 73.71 73.71 71.97 71.97 71.97 73.8 77.47 77.47 80.25 76.35 76.96 76.96 76.96 78 78 78 78.55 78.55 78.55 78.55 78.55 78.55 78.55 78.55 78.55 75.04 75.04 75.04 75.04 74.83 74.82 74.82 80.69 83.33 83.33 83.33 83.33 83.33 85.85 Event Date (mm/dd/yy) 12/19/04 07/15/06 06/06/04 10/27/00 04/21/07 12/12/06 01/15/06 10/27/00 01/15/06 02/05/05 03/09/07 07/16/07 09/09/06 11/14/06 02/26/06 01/15/06 07/15/06 01/15/06 11/12/03 01/07/02 12/05/99 03/09/07 04/21/07 12/12/06 07/27/03 08/02/02 04/21/07 11/14/06 12/12/06 10/04/03 10/04/03 01/29/04 07/16/07 01/15/06 01/15/06 12/12/06 12/27/06 02/16/01 03/09/07 07/16/07 07/14/92 01/15/06 02/16/01 05/26/03 06/09/00 07/01/03 09/26/97 10/27/00 12/22/96 01/29/04 05/26/03 02/01/02 07/27/03 01/15/06 04/21/07 10/26/06 01/29/04 12/12/06 12/27/06 07/15/06 02/05/05 01/15/06 03/02/05 Event Lat Event Lon Depth (km) Baz (deg) 99.6 111.9 123.1 71.2 96.6 101.5 113.9 69.6 114.1 99.5 49.1 56.9 114.5 109.3 111.4 113.7 108.1 115.2 61.3 77.6 65.9 49.8 97.4 100 45 65.5 97.3 110.5 101.5 111.2 108 90.5 53.7 109.9 108.5 92.7 97.7 115 46.3 54.1 111.2 116.5 119.3 97.9 64.6 101.4 110 69.5 49.3 94.1 94.6 45.8 44.2 111 96.6 307.7 98.3 104.1 101.6 113.6 102.7 120.1 116.8 dt (s) 1.65 0.1 0.5 0.04 1.3 0.35 1.9 0.18 0.95 0.28 0.8 0.19 0.95 0.08 1.5 0.3 0.85 0.09 1.1 0.28 0.6 0.09 1.4 0.36 0.55 0.15 0.8 0.1 0.7 0.1 1.4 0.01 0.65 0.08 0.85 0.14 0.9 0.05 0.4 0.06 0.7 0.08 0.5 0.09 0.9 0.18 0.85 0.08 0.85 0.32 0.75 0.08 1 0.21 0.45 0.08 0.55 0.18 0.6 0.05 0.8 0.09 1.1 0.05 0.65 0.11 0.65 0.1 1.15 0.05 0.85 0.09 0.85 0.28 0.9 0.05 0.5 0.18 1.85 0.3 0.75 0.15 0.65 0.08 0.7 0.14 0.8 0.09 1.1 0.08 0.5 0.04 0.7 0.09 2.05 0.2 1.8 0.16 1 0.06 1.25 0.03 0.8 0.08 0.55 0.1 0.8 0.06 1.15 0.11 1.35 0.14 1.5 0.08 1.6 0.47 0.85 0.14 0.85 0.09 0.8 0.11 0.95 0.03 0.85 0.1 F (deg) 25 1.5 6 5.5 24 17.5 25 19.25 49 12 33 7 28 4.75 25 13.5 35 4.75 23 9.75 39 6 62 7.75 8 17.5 5 9.25 16 9.5 3 1.75 16 8.75 37 12.25 43 2 77 10.5 3 4.25 52 13.75 5 11.75 21 6.75 1 15 23 7.25 59 11.25 81 13.5 61 14.75 9 5.25 33 8.5 15 4.5 33 7.25 32 4 1 2.25 49 8 11 15.75 31 4.75 12 16.75 38 5.25 35 12.75 23 8.75 29 13.25 14 7.5 12 6.75 14 5.5 39 11.25 9 10 20 4.25 31 4.75 15 2.25 29 3.5 40 11 4 6.75 6 2.25 24 6.25 30 6.25 34 8.25 19 8 48 2.5 35 7.25 28 3 37 12.25 Pol (deg) 74.6 105.9 147.1 96.2 145.6 134.5 85.9 44.6 79.1 76.5 88.1 5.1 122.5 104.3 95.4 110.7 92.1 78.2 104.3 0.6 62.9 101.8 92.4 121 44 88.5 156.3 29.5 40.5 102.2 75 75.5 86.7 141.9 107.5 43.7 108.7 84 34.3 16.1 76.2 93.5 90.3 83.9 76.6 87.4 71 78.5 69.3 63.1 79.6 74.8 84.2 115 90.6 283.7 68.3 70.1 120.6 65.6 67.7 92.1 79.8

Deccan Volcanic Province (DVP) 7.38 123.78 576 4.45 126.17 368 6.04 113.11 579 26.27 140.06 388 3.55 151.27 407 3.73 124.68 214 7.83 122.60 265 26.27 140.46 388 7.83 122.60 265 5.29 123.34 525 43.22 133.53 441 36.81 134.85 350 7.21 120.11 572 6.42 127.98 352 6.99 125.11 526 7.83 122.60 265 4.45 126.17 368 7.83 122.60 265 33.17 137.07 385 18.96 144.96 599 29.92 138.69 449 43.22 133.53 441 3.55 151.27 407 3.73 124.68 214 47.15 139.25 470 29.28 138.97 426 3.55 151.27 407 6.42 127.98 352 3.73 124.68 214 7.05 125.41 533 Southern Granulite Terrian (SGT) 7.05 125.41 533 6.29 126.94 210 36.81 134.85 350 7.83 122.60 265 7.83 122.60 265 3.73 124.68 214 5.75 154.42 375 Eastern Dharwar Craton (EDC) 7.16 117.49 521 43.22 133.53 441 36.81 134.85 350 4.71 125.43 477 7.83 122.60 265 7.16 117.49 521 6.76 123.71 566 30.49 137.73 563 4.53 122.51 635 5.39 128.99 254 26.27 140.46 388 43.21 138.92 227 Western Dharwar Craton (WDC) 6.29 126.94 210 6.76 123.71 566 45.46 136.72 356 47.15 139.25 470 7.83 122.60 265 3.55 151.27 407 38.67 15.40 217 Godavari Graben (GG) 6.29 126.94 210 Eastern Ghats (EG) 3.73 124.68 214 5.75 154.42 375 4.45 126.15 368 5.29 123.34 525 7.83 122.60 265 6.53 129.93 202

6 of 15

XXXXXX
t1:72 t1:73 t1:74 t1:75 t1:76 t1:77 t1:78 t1:79 t1:80 t1:81 t1:82 t1:83 t1:84 t1:85 t1:86 t1:87 t1:88 t1:89 t1:90 t1:91 t1:92 t1:93 t1:94 t1:95 t1:96 t1:97 t1:98 t1:99 t1:100 t1:101 t1:102 t1:103 t1:104 t1:105 t1:106 t1:107 t1:108 t1:109 t1:110 t1:111 t1:112 t1:113 t1:114 t1:115 t1:116 t1:117 t1:118 t1:119 t1:120 t1:121 t1:122 t1:123 t1:124 t1:125 t1:126 t1:127 t1:128 t1:129 t1:130 t1:131 t1:132 t1:133 t1:134 t1:135

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Table 1. (continued)
Station NGP NGP NGP NGP NGP NGP NGP NGP BHPL JBP JBP JBP REWA REWA BLSP BOKR BOKR BOKR BOKR BOKR DHR ALB NDI NDI NDI NDI NDI NDI NDI NDI NDI NDI SMLA SMLA SMLA JPA MND MND TEZ TEZ TEZ TEZ AGWD BHK DHRM DHRM MPAD MPAD SNCH SONT SONT TANA TANA Station Lat 21.15 21.15 21.15 21.15 21.1 21.1 21.1 21.1 23.48 23.88 23.88 23.88 24.57 24.57 22.13 23.79 23.79 23.79 23.79 23.79 24.01 25.28 28.58 28.58 28.58 28.58 28.58 28.58 28.58 28.68 28.68 28.68 31.13 31.13 31.13 26.22 25.92 25.92 26.62 26.62 26.62 26.62 23.96 31.41 32.24 32.24 20.48 20.48 24.75 22.38 22.38 21.58 21.58 Station Lon 79.05 79.05 79.05 79.05 79.06 79.06 79.06 79.06 77.42 79.88 79.88 79.88 81.34 81.34 82.13 85.89 85.89 85.89 85.89 85.89 72.85 81.54 77.22 77.22 77.22 77.22 77.22 77.22 77.22 77.22 77.22 77.22 77.17 77.17 77.17 90.57 90.68 90.68 92.78 92.78 92.78 92.78 71.93 76.41 76.30 76.30 73.81 73.81 71.76 71.85 71.85 71.97 71.97 Event Date (mm/dd/yy) 10/04/03 01/15/06 07/15/06 09/04/05 03/09/07 04/21/07 09/09/06 10/26/06 11/14/06 12/17/02 12/21/02 01/29/04 10/04/03 07/01/03 01/15/06 01/15/06 02/05/05 07/15/06 03/09/07 12/12/06 11/12/03 10/03/02 01/07/02 06/06/04 01/15/06 04/19/05 01/27/06 07/03/01 11/17/02 04/29/08 07/15/06 12/12/06 03/28/06 04/29/08 07/15/06 12/21/02 06/06/04 10/04/03 01/07/02 02/07/01 02/15/00 07/03/01 11/21/04 02/18/09 08/28/08 12/09/08 12/05/01 12/23/01 09/28/04 09/06/04 03/19/05 01/05/07 03/18/08 Event Lat 7.05 7.83 4.45 3.00 43.22 3.55 7.21 38.67 Event Lon Depth (km) Baz (deg) 116.7 119.5 113.4 106.8 50.6 100 120.7 303.2 114.8 119 107.4 103.4 120.4 110.1 122.5 126.9 111.2 120.4 51.5 112.1 65 130.4 82.6 129.3 122.7 72.8 116.4 80.5 49.8 117.5 117 109.9 72.5 118.7 118.3 118.6 142.6 129.9 87.7 42.4 89.1 85 306.4 106.7 271.7 102.6 210.1 100.4 204.9 217.8 104.8 24.8 112.5 dt (s) 1.05 0.05 1.2 0.06 1 0.04 1.1 0.1 0.9 0.15 1.1 0.38 0.8 0.23 0.85 0.13 0.4 0.1 0.85 0.08 0.95 0.08 0.95 0.04 0.95 0.04 0.8 0.14 1.1 0.15 0.85 0.1 0.95 0.14 1 0.08 0.45 0.08 1.6 0.34 0.5 0.05 0.55 0.1 1.75 0.08 1.55 0.01 1.25 0.04 0.9 0.04 0.85 0.09 1.3 0.28 0.9 0.16 0.9 0.13 1.55 0.05 1.2 0.11 0.5 0.21 0.4 0.13 0.9 0 0.8 0.04 0.8 0.06 1.05 0.1 0.45 0.01 1.05 0.03 0.6 0.05 2.65 0.33 0.95 0.1 0.7 0.27 0.95 0.2 0.75 0.36 0.95 0.16 0.6 0.11 1.25 0.11 1.25 0.26 1.15 0.26 0.9 0.2 0.5 0.18 F (deg) 57 2.25 45 2.25 47 3 44 84 5.5 43 15.5 72 12 16 6.75 6 14.5 60 2.25 51 3 54 4.5 47 1.75 36 7.5 11 9.5 61 8.75 46 6.75 45 5.25 40 6.75 62 10.75 63 7.00 73 14 20 5.5 35 3 33 4.25 50 3.5 39 6 12 8 14 13.75 1 6.75 33 5.75 35 6.75 14 21.5 4 13.25 9 0.75 28 4 64 1.75 39 4 47 1.75 68 3.75 38 7.25 55 5.5 22 6 31 16.75 28 10.5 23 17.25 28 11.75 79 8.5 73 1.75 26 5.5 17 5 35 7.7 47 14.5 Pol (deg) 59.7 74.5 66.4 102.8 33.4 57 48.7 287.2 108.8 59 56.4 49.4 73.4 74.1 111.5 65.9 65.2 75.4 91.5 50.1 2 203.4 62.6 94.3 89.7 22.8 77.4 92.5 63.8 118.5 84 74.9 58.5 122.7 109.3 90.6 78.6 90.9 40.7 25.6 51.1 30

Singhbum Craton (SC) 125.41 533 122.60 265 126.17 368 123.07 444 133.53 441 151.27 407 120.11 572 15.40 217 NarmadaSon Lineament (NSL) 6.42 127.98 352 7.00 125.41 493 4.98 123.16 602 6.29 126.94 210 125.41 533 7.05 4.53 122.51 635 Bhandra Craton (BC) 7.83 122.60 265 7.83 122.60 265 5.29 123.34 525 4.45 126.17 368 43.22 133.53 441 3.73 124.68 214 DelhiAravali Fold Belt (DAFB) 33.17 137.07 385 Foredeep (FD) 7.53 115.66 316 18.96 144.96 599 6.04 113.11 579 7.83 122.60 265 29.64 138.89 426 5.47 128.13 397 21.64 142.98 290 47.82 146.21 397 6.11 127.48 405 4.45 126.17 368 3.73 124.68 214 Himalaya (HIM) 31.71 137.75 401 6.11 127.48 405 4.45 126.17 368 Assam Valley (AV) 4.98 123.16 602 6.04 113.11 579 7.02 125.43 531 18.96 144.96 599 52.75 153.85 427 17.67 145.40 522 21.64 142.98 290 New SKS Measurements 15.68 61.70 14 27.42 176.33 25 0.25 17.36 15 16.01 168.14 228 52.61 18.35 10 9.61 159.53 16 52.51 28.02 10 55.37 28.98 10 21.89 179.55 598 55.76 156.06 18 29.25 177.44 25

198 199 200 201 202 203

sensitive to small window changes may be selected. As an alternative, a search is made for measurements that are stable over many different analysis windows, achieved by varying the window length and looking for plateaus in dt and F. These plateaus indicate stable splitting measurements; so that once the appropriate plateau has been iden-

tified, the window that gives the smallest error can be selected. Initially, the S wave is picked manually and then a set of windows with different start and end times are defined around the picked time. Subsequently, measurements are performed for each window length to find the polarization azimuth of the fast moving shear wave (F) and

204 205 206 207 208 209

7 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Figure 6. Individual splitting parameters at broadband stations on the Indian continent obtained using S (open bars) and SKS (solid bars) waveforms. Grey circles represent stations where new SKS splitting measurements are made in this study. Lengths of the bars are proportional to the delay time. Arrows represent APM directions of the Indian plate in a no net rotation frame [DeMets et al., 1994].

210 211 212 213 214 215 216 217 218 219 220 222 223 224 225 226 227

splitting time (dt), by adopting the eigenvalue method. Then, a cluster analysis is applied to find the best stable solution. Applying this method to the direct S wave data from 39 broadband seismic stations in India, a total of 105 new nonnull measurements are obtained in this study. In addition, 10 new SKS splitting measurements are obtained from 7 numbers of stations. The individual measurements are listed in Table 1 along with the event and station information and plotted in Figure 6 superimposed on the results from SKS splitting.

3. Results
Splitting Parameters [9] Histograms of the delay times between fast and slow axes (Figure 7a) tend to cluster around 0.9 s, which is close to the global average for continental shield regions and also similar to the values previously obtained for India from SKS splitting. None of our measurements of delay times made using data from deep and interme-

221 3.1.

diate depth earthquakes yield a splitting time 3 s. Only 10 out of 105 measurements show a delay time that is greater than 1.5 s. Two delay time values greater than 2 s; one at station HYB (2.05 s) and the other at TEZ (2.65 s) are from events with depths of 388 and 290 km, respectively, from the Philippine Sea region. Since anisotropy in subduction zones may exist even at depths greater than 500 km [e.g., Fouch and Fischer, 1996], these higher values of dt may be ascribed to possible contamination from source side anisotropy. [10] Most of the fast axis azimuths tend to be aligned in the NE direction, close to the absolute plate motion direction of the Indian plate, followed by a NS trend (Figure 7b). As also revealed by the SKS splitting measurements [Kumar and Singh, 2008] the fast axis azimuths do not reveal a coherent pattern across the continent. The NE trend is predominant at stations TEZ, MND, JPA (northeast India), NDI (IndoGangetic basin), NGP, REWA, JBP (close to the Narmada son lineament), BLSP (Bhandara craton), BOKR (Singhbum craton), KGD, KGF, HYB (Eastern Dharwar

228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247

8 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Figure 7. Histograms of (a) delay times and (b) fast polarization azimuths at all the stations showing a predominance of splitting times at 0.9 s and fast polarization azimuths around N40E close to the APM direction of the Indian plate.
248 249 250 251

Craton), DHD (Western Dharwar Craton), VISK, BWNR (Eastern Ghats) and KOD (Southern Granulite Terrain). In contrast to the NE trend, twenty eight measurements from 14 stations (AGWD, AKL, BOM, KAND, KARD, KHER,

KOY, MULG, MPAD, PUNE, SKP, TANA, UTWD and WAG) situated on the Deccan Volcanic Province yield either north south or northwest fast polarization azimuths. A single measurement from the station MNGR in the western

252 253 254 255

Figure 8. Same as Figure 5 except example showing dependence of splitting parameters on the initial polarization directions for station HYB.
9 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322

shows that the average value of pol is 80 showing a general trend of SH leading SV. [12] Previous studies [Chen and Ozalaybey, 1998; Barruol and Hoffmann, 1999] using SK(K)S phases reported null splitting at HYB implying an isotropic upper mantle beneath this station. However, splitting measurements using direct S phases have indicated significant anisotropy, with NE oriented fast axis azimuths, consistent with the results from most other stations. Calculation of the fast polarization for all the seven events in SV/SH coordinate system gives a value close to 90, implying that the faster quasi shear wave is actually SH. 3.3. Two Layers of Anisotropy [13] In the presence of two or more layers of anisotropy, it has been theoretically shown that the apparent splitting parameters obtained by assuming a single anisotropic layer in the presence of multiple layers, exhibit a systematic 90 periodicity as a function of source polarization, which is equivalent to back azimuth for the SKS/SKKS phases [Silver and Savage, 1994; Rmpker and Silver, 1998]. Augmenting the SKS/SKKS results with those obtained from direct S waves is very advantageous to model two layers of anisotropy. This is due to the fact that S wave splitting parameters represent a wide range of source polarization directions (being governed by the faulting mechanism) in contrast to the SKS splitting parameters which have very limited back azimuthal coverage. In order to ascertain whether the splitting parameters obtained in this study indicate two layers of anisotropy, the fast axis azimuths and delay times at all the stations derived from S wave splitting are plotted with

Figure 9. Geometry of propagation of the SV and SH components of a direct S wave. Star denotes the hypocenter of the earthquake, and triangle denotes the station.
256 257 258 259 260 261 262 263 264 265 266 267

Dharwar craton south of the DVP and three measurements from the only station SMLA located in the Himalayan foothills also produce an average NS fast direction. Six measurements from the three stations KOD, PCH and TRVM in the Southern Granulite Terrain produce scattered fast axis azimuths. Two measurements from the station KOD yield a nearly NE fast direction while other measurements from the stations TRVM and PCH produce NS and NW fast direction, respectively. Also, dependence of the splitting parameters on the incoming polarization of the S waves is suggestive of multiple layers of anisotropy beneath India (Figure 8). Evidence for Radial Anisotropy? [11] For shear waves that are propagating horizontally, radial anisotropy with a vertical axis of symmetry is manifested in terms of faster velocity of the horizontally polarized S wave in comparison with the vertically polarized wave. For a near vertical incidence at the station, the direction of the SV component of the shear wave can be considered to be in the direction of the back azimuth as a 2D approximation, although the SV/SH vibration plane is not exactly horizontal; the deviation from horizontal being equal to the incidence angle of the wave (Figure 9). Assuming that the back azimuth coincides with the direction of the SV vibration we subtracted the measured fast polarization direction from the back azimuth of the corresponding events to obtain the fast polarization in the SV/SH (radial/transverse) coordinate system. This quantity termed as pol by earlier workers is considered as a diagnostic of radial anisotropy, since it serves as a diagnostic to discriminate whether the SH component is faster than the SV component [Heintz, 2006]. A value of 90 for pol implies that the SH wave coincides with the fast axis and hence leads SV while a value of 0 corresponds to a situation in which the SV polarized wave leads SH. The values of pol corresponding to the individual splitting measurements are presented in the last column of Table 1. A polar histogram plot (Figure 10)

268 3.2. 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292

Figure 10. Polar diagram of the fast polarization direction (pol) in the SV/SH coordinate system obtained by subtracting the fast axis azimuth from the back azimuth of the event. A value of 90 for pol indicates that the SH component leads SV, which is evidence for radial anisotropy for horizontally propagating S waves. Labels on the radial axis denote the scale used to represent the number of measurements.

10 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Figure 11. Plot showing the variation of (a) delay time (dt) and (b) fast axis azimuth (F) estimated at the stations, with respect to the source polarization direction. The measurement errors are shown as vertical bars. Red circles indicate the splitting parameters obtained from S, and blue triangles represent those obtained from SKS waves.
323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396

respect to the source polarization direction (Figure 11). This data set is supplemented with the previously published SKS splitting results [Kumar and Singh, 2008; Heintz et al., 2009] and the few new ones obtained from this study. Although the delay times do not show a discernible trend, the fast axis azimuths do reveal a near 90 periodicity as a function of source polarization suggesting at least two layers of anisotropy. It has been opined that in situations where the fast axis of symmetry in any of the layers is tilted, like in the lithospheric portion of a paleosubduction environment in Central Europe, the azimuths of fast axes do show a deviation from the 90 periodicity [Plomerov et al., 1998]. [14] In view of the continentalscale nature of the present study and the scarcity of data at individual stations, we grouped the stations sampling the same geological province to forward model the dependence of the splitting parameters on source polarization in terms of two layers of anisotropy. The modeling results (Figure 12) for three regions, namely, the Dharwar craton, the Bhandara craton, and the Indo Gangetic plain combined with Himalayan foothills, suggest that two layers of anisotropy are required to satisfactorily explain the splitting measurements. Favorable models for all these regions indicate that the lower layer with a NNE oriented fast axis azimuth produces a 1 s delay time, while the upper layer has a weak anisotropy (dt close to 0.4 s) with a fast axis close to 60. However, such a modeling was not possible for the other regions since the measurements are either insufficient or scattered as a function of the source polarization, thus inhibiting the modeling.

4. Discussion
[15] Previous results of SKS splitting suggest plate motion dominant strain in the Indian shield with the variations in fast azimuths being influenced by the basal topography of the lithosphere and edge flow [Kumar and Singh, 2008]. Evidence for anisotropy frozen in the lithosphere was found at some locales in the vicinity of continental shear

zones in south India. This study using direct S phases is in conformity with a predominance of APM related strain in the Indian shield, indicating a similar (asthenospheric) source of anisotropy for both SKS and direct S. Similarity of the splitting parameters derived from SKS and S phases (Figure 6) also indicates negligible lower mantle anisotropy [Heintz, 2006]. [16] Although the direct S waves analyzed in this study reveal a general trend of SH leading SV, directly interpreting this observation in terms of radial anisotropy, as previously reported for Australia [Heintz, 2006] could be ambiguous since the scenario of SH leading SV that is consistent with vertically oriented transverse isotropy is only relevant for horizontally propagating shear waves. This criterion is not fully met by the direct S waves traversing the upper mantle beneath the receiver since their raypaths are much closer to vertical than horizontal. For paths through the upper mantle, SKS/SKKS phases have near vertical incidence angles of 1015, while direct S phases in the distance range of 40 80 have incidence angles up to 35 (Figure 13). Further, azimuthal anisotropy can also produce a situation where SH leads SV depending on the angle between the fast axis azimuth and the incoming source polarization. Considering the low sensitivity of S waves to possible effects of radial anisotropy near the receiver (due to their near vertical ray geometry), it becomes imperative to explore other possibilities to explain the observed nature of anisotropy. The difference in the patterns of S and SKS splitting at the same station (HYB for example) might reflect the different propagation paths and response to the anisotropic properties of the medium that they sample. A model for upper mantle anisotropy that can explain both sets of observations might involve a difference in wave speed between vertically and horizontally polarized waves and also presence of anisotropy at much deeper levels. [17] Although the upper mantle deeper than 200 km and that from 600 km to the core mantle boundary is generally considered to be isotropic [Silver, 1996; Savage, 1999],

11 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Figure 12. (left) Delay time and (right) fast axis azimuth plotted as a function of incoming polarization direction for three regions indicated. Red circles denote S measurements, and blue triangles denote SKS measurements. Apparent splitting parameters estimated using a two layer model are plotted as dashed lines. The F and dt values of the top (subscript 1) and bottom (subscript 2) layers used for forward modeling are indicated.
12 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX

Figure 13. Raypaths of S waves in the epicentral distance range of 3080 and SKS waves in the distance range of 85115 registered at a seismic station shown as an inverted triangle.
397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436

global and regionalscale studies do provide evidences for weak anisotropy (12%) within the mantle transition zone (MTZ). However, the nature and possible mechanisms that generate this anisotropy still remain ambiguous. Montagner and Kennett [1996] noted that the discrepancy between the reference Earth models derived from body waves and normalmode eigenperiods can be reconciled by the presence of radial anisotropy. Through inversion of the eigenperiods of the Earths normal modes they demonstrate evidence for radial anisotropy in the MTZ. Independent studies based on observations of PSH converted waves from the 660 km discontinuity at stations GRF (Central Europe) and OBN (East European platform) that cannot be explained by a dipping nature of the interface lend further credence to presence of anisotropy within the MTZ [Vinnik and Montagner, 1996; Vinnik et al., 1998]. Modeling of these PtoSH phases together with the SKS/SKKS waveforms brings out clear presence of weak azimuthal anisotropy (2%) localized in a 3040 km layer atop the 660 km boundary. This layer of anisotropy has been interpreted to be caused by horizontal flow at the bottom of the transition zone. [18] Presence of anisotropy in the transition zone advocated through splitting of direct S waves has been contentious. Splitting analysis of local S phases combined with those obtained from teleseismic core phases revealed variable anisotropy in the transition zone beneath subduction zones [Fouch and Fischer, 1996]. While some subduction zones like Sakhlain Islands show evidences for 0.5% anisotropy, results from regions like IzuBonin rule out any anisotropy in the MTZ. A consistent observation that emerged from the splitting of local S waves is that the 410 520 km depth range of the transition zone is intermittently anisotropic due to preferred orientation of the b spinel, which has strong crystal anisotropy; the lower part of the MTZ is largely isotropic. Reporting unusually large delay times up to 6 s, Wookey and Kendall [2004] and Wookey et al. [2002] argue in favor of anisotropy in the vicinity of the 660 km discontinuity between FijiTonga subduction zone and the Australian continent. Also, based on the

persistent nature of SH leading SV coupled with observations of large delay times (3 s), a rare and unequivocal evidence for radial anisotropy in the transition zone beneath this subduction zone has been presented [Chen and Brudzinski, 2003]. Forward modeling of the observed waveforms requires 0.9% anisotropy with a vertically oriented slow axis, to satisfactorily explain the observed seismograms. This radial anisotropy has been explained invoking presence of cold remnant lithospheric slabs that are stagnant in the transition zone. [19] The results of Wookey et al. [2002] have been contested by Saul and Vinnik [2003] who argued that such anomalous delay times are artifacts of processing and demonstrated using similar and more extensive data sets that the delay times are indeed small and can be explained by a radially anisotropic uppermost mantle. Subsequent analysis of a much more comprehensive data set for direct S waves leads to an average delay time of only 0.9 s at stations deployed across the continent highlighting the lack of evidence for midmantle deformation between the FijiTonga subduction zone and the Australian continent [Heintz, 2006]. The splitting times obtained in this study for the Indian shield being close to 1 s also vindicate lack of anomalous delays arguing against the possibility of midmantle anisotropy, since the possible delay time contribution considering 2% azimuthal anisotropy within the top 110 km of the MTZ would amount to as much as 0.4 s. Also, the delay times do not seem to be correlated with source depth (Figure 14), suggesting that the measurements mostly represent receiver side anisotropy with negligible contribution from the source side. [20] Presence of radial and midmantle anisotropy beneath India being less likely, the splitting observations may be reconciled by a two layer model of azimuthal anisotropy. As reiterated by several workers, the splitting measurements obtained from vertically propagating S and core refracted phases from most Precambrian shields represent the combined effect of fossil anisotropy imprinted in the lithosphere and anisotropy associated with asthenospheric flow, resulting in two layers of anisotropy [Fouch and Rondenay,

437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476

13 of 15

XXXXXX

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562

IndiaEurasia collision or flow in asthenosphere due to absolute plate motion of Eurasia [Gao and Liu, 2009]. The scenario of two layers of azimuthal anisotropy underneath the Indian continent inferred from this study, though fraught with issues like nonuniqueness, seems to be a strong possibility to explain both the S and SKS slitting measurements.

5. Conclusions
[21] Analysis of direct S waves from deep and intermediate depth events in the Fiji Tonga and Japan subduction zones has yielded 105 new splitting measurements at 39 broadband stations sited on various tectonic environments of the Indian continent. The delay times between the fast and slow axes are clustered around 0.80.9 s which is close to the global average for continental shields and also similar to those obtained earlier from SKS splitting. Also, these delay times do not seem to be correlated with source depth, suggesting that the measurements mostly represent receiver side anisotropy with negligible contribution from the source side or midmantle. The fast polarization azimuths are generally consistent with the results from SKS splitting that indicate a predominance of plate motion related strain in the Indian continent. Station HYB, which has otherwise indicated absence of SK(K)S splitting shows significant S wave anisotropy with a NE oriented fast axis azimuth. This can be due to the different propagation paths sampled by the S and SKS waves. Although a general trend of SH leading SV is consistently observed at most of stations we refrain from interpreting this phenomenon in terms of radial anisotropy since (1) S waves with their steep incidence angles beneath the receiver are more sensitive to azimuthal anisotropy and (2) depending on the polarization of the S wave, SH can lead SV even for azimuthal anisotropy. Modeling the variations of splitting parameters in terms of two layers of azimuthal anisotropy reveals a NNE oriented fast axis azimuth in the lower layer and a N60E oriented anisotropy in the upper layer that produce an effective delay time (between the fast and slow waves) of 1 and 0.4 s, respectively. While the anisotropy in the lower layer may be explained by shear at the base of the lithosphere, the fast axis azimuths in top layer could be a manifestation of anisotropy frozen within the Indian lithosphere due to past tectonic events. [22] Acknowledgments. We thank two anonymous reviewers, the Associate Editor, and Editor Robert Nowack for their valuable suggestions. We are grateful to the India Meteorological Department for making the data available and Teanby for the shear wave splitting code. The work has been performed under the Supra institutional project of NGRI SIP 001228 (M.R.K.).

Figure 14. Delay time dt plotted as a function of source depth. Lack of correlation of dt with source depth suggests small or negligible source side anisotropy.
477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502

2006]. The NNE oriented fast axis azimuth for the lower layer suggested by our modeling results may represent shear at the base of the Indian continental lithosphere, as argued to explain the previous SKS splitting measurements [Kumar and Singh, 2008]. In addition, the N60E orientation in the top layer could be a signature of the frozen anisotropy within the Indian lithosphere that was subjected to protracted episodes of deformation. Earlier results obtained through analysis of data in the vicinity of Himalayan collision zone and Tibet lend credence to the observation of two layers of anisotropy. Jointly inverting P receiver functions and SKS particle motion at seismic station NIL in Pakistan Himalaya resulted in two layers of anisotropy at depths of 80160 km and 160220 km [Vinnik et al., 2007]. While the anisotropic direction of N150E found in the lower layer is parallel to the strike of Himalayan mountains belts, the fast axis azimuth of 60 in the upper layer has been attributed to current deformations related to plate motion or an effect of anisotropy frozen from past significant tectonic events. Almost similar observations were reported for station XKS on the Tibetan plateau, where a NESW oriented anisotropy in the top layer was related to the direction of surface movement observed from GPS or due to presence of certain anisotropic minerals like amphibole in the crust and the EW anisotropic directions in the lower layer were interpreted in terms of NS directed compressional stress originated due to

References
Baag, C. E., and C. A. Langston (1985), Shearcoupled PL, Geophys. J. R. Astron. Soc., 80, 363386. Barruol, G., and R. Hoffmann (1999), Seismic anisotropy beneath the Geoscope stations from SKS splitting, J. Geophys. Res., 104, 10,75710,773, doi:10.1029/1999JB900033. Chen, W., and M. R. Brudzinski (2003), Seismic anisotropy in the mantle transition zone beneath FijiTonga, Geophys. Res. Lett., 30(13), 1682, doi:10.1029/2002GL016330. Chen, W. P., and S. Ozalaybey (1998), Correlation between seismic anisotropy and Bouger gravity anomalies in Tibet and its implications for lithospheric structures, Geophys. J. Int., 135, 93101, doi:10.1046/ j.1365-246X.1998.00611.x.

14 of 15

XXXXXX 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612

SAIKIA ET AL.: AZIMUTHAL ANISOTROPY BENEATH INDIA

XXXXXX 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661

DeMets, C., R. G. Gordon, D. F. Argus, and S. Stein (1994), Effect of recent revisions to the geomagnetic reversal time scale on estimate of current plate motions, Geophys. Res. Lett., 21, 21912194, doi:10.1029/ 94GL02118. Fouch, M. J., and K. M. Fischer (1996), Mantle anisotropy beneath northwest Pacific subduction zones, J. Geophys. Res., 101, 15,98716,002, doi:10.1029/96JB00881. Fouch, M. J., and S. Rondenay (2006), Seismic anisotropy beneath stable continental interiors, Phys. Earth Planet. Inter., 158(24), 292320, doi:10.1016/j.pepi.2006.03.024. Gao, S. S., and K. H. Liu (2009), Significant seismic anisotropy beneath the southern Lhasa Terrane, Tibetan Plateau, Geochem. Geophys. Geosyst., 10, Q02008, doi:10.1029/2008GC002227. Gung, Y., M. Planning, and B. Romanowicz (2003), Global anisotropy and the thickness of continents, Nature, 422, 707711, doi:10.1038/ nature01559. Heintz, M. (2006), Midmantle deformation between the Australian continent and the FijiTonga subduction zone?, J. Geophys. Res., 111, B09303, doi:10.1029/2005JB004058. Heintz, M., V. P. Kumar, V. K. Gaur, K. Priestly, S. S. Rai, and K. S. Prakasam (2009), Anisotropy of the Indian continental lithospheric mantle, Geophys. J. Int., 179, 13411360, doi:10.1111/j.1365-246X. 2009.04395.x. Jordan, T. J., and L. N. Frazer (1975), Crustal and upper mantle structure from Sp phases, J. Geophys. Res., 80, 15041518, doi:10.1029/ JB080i011p01504. Kumar, M. R., and A. Singh (2008), Evidence for plate motion related strain in the Indian shield from shear wave splitting measurements, J. Geophys. Res., 113, B08306, doi:10.1029/2007JB005128. Langston, C. A. (1996), The SsPmP phase in regional wave propagation, Bull. Seismol. Soc. Am., 86, 133143. Montagner, J.P., and B. L. N. Kennett (1996), How to reconcile body wave and normalmode reference Earth models, Geophys. J. Int., 125, 229248, doi:10.1111/j.1365-246X.1996.tb06548.x. Oliver, J. (1961), On the long period character of shear waves, Bull. Seismol. Soc. Am., 51, 112. Plomerov, J., V. Babuka, J. ileny, and J. Horlek (1998), Seismic anisotropy and velocity variations in the mantle beneath the SaxothuringicumMoldanubicum contact in central Europe, Pure Appl. Geophys., 151, 365394, doi:10.1007/s000240050118. Poupinet, G., and C. Wright (1972), The generation and properties of shearcoupled PL waves, Bull. Seismol. Soc. Am., 62, 16991710. Rmpker, G., and P. G. Silver (1998), Apparent shearwave splitting parameters in the presence of vertically varying anisotropy, Geophys. J. Int., 135, 790800, doi:10.1046/j.1365-246X.1998.00660.x. Saul, J., and L. Vinnik (2003), Mantle deformation or processing artefact?, Nature, 422, 136, doi:10.1038/422136a. Savage, M. K. (1999), Seismic anisotropy and mantle deformation: What have we learned from shear wave splitting?, Rev. Geophys., 37, 65106, doi:10.1029/98RG02075.

Silver, P. G. (1996), Seismic anisotropy beneath the continents: Probing the depths of geology, Annu. Rev. Earth Planet. Sci., 24, 385432, doi:10.1146/annurev.earth.24.1.385. Silver, P. G., and M. Savage (1994), The interpretation of shearwave splitting parameters in the presence of two anisotropic layers, Geophys. J. Int., 119, 949963, doi:10.1111/j.1365-246X.1994.tb04027.x. Silver, P. G., and W. W. Chan (1988), Implications for continental structure and evolution from Seismic anisotropy, Nature, 335, 3439, doi:10.1038/ 335034a0. Silver, P. G., and W. W. Chan (1991), Shear wave splitting and mantle deformation, J. Geophys. Res., 96, 16,42916,454, doi:10.1029/ 91JB00899. Singh, A., M. R. Kumar, P. S. Raju, and D. S. Ramesh (2006), Shear wave anisotropy of the northeast Indian lithosphere, Geophys. Res. Lett., 33, L16302, doi:10.1029/2006GL026106. Singh, A., M. R. Kumar, and P. S. Raju (2007), Mantle deformation in Sikkim and adjoining Himalaya: Evidences for a complex flow pattern, Phys. Earth Planet. Inter., 164, 232241, doi:10.1016/j.pepi. 2007.07.003. Teanby, N. A., J. M. Kendall, and M. Van der Baan (2004), Automation of shearwave splitting measurements using cluster analysis, Bull. Seismol. Soc. Am., 94(2), 453463, doi:10.1785/0120030123. Vinnik, L., A. Singh, S. Kiselev, and M. R. Kumar (2007), Upper mantle beneath foothills of the western Himalaya: Subducted lithospheric slab or a keel of the Indian shield, Geophys. J. Int., 171, 11621171, doi:10.1111/j.1365-246X.2007.03577.x. Vinnik, L., S. Chevrot, and J. P. Montagner (1998), Seismic evidence of flow at the base of upper mantle, Geophys. Res. Lett., 25(11), 19951998, doi:10.1029/98GL01418. Vinnik, L., and J. P. Montagner (1996), Shear wave splitting in the mantle Ps phases, Geophys. Res. Lett., 23(18), 24492452, doi:10.1029/ 96GL02263. Vinnik, L. P., V. Farra, and B. Romanawicz (1989), Azimuthal anisotropy in the Earth from observations of SKS at GEOSCOPE and NARS broadband stations, Bull. Seismol. Soc. Am., 79, 15421558. Wookey, J., and J. M. Kendall (2004), Evidence of midmantle anisotropy from shear wave splitting and the influence of shearcoupled P waves, J. Geophys. Res., 109, B07309, doi:10.1029/2003JB002871. Wookey, J., J. M. Kendall, and G. Barruol (2002), Midmantle deformation inferred from seismic anisotropy, Nature, 415, 777780. Zandt, G., and G. E. Randall (1985), Observation of shearcoupled P waves, Geophys. Res. Lett., 12, 565568, doi:10.1029/GL012i009p00565. M. Ravi Kumar, D. Saikia, and A. Singh, National Geophysical Research Institute, Uppal Road, Hyderabad, Andhra Pradesh 500 007, India. (mangalampallyravikumar@rediffmail.com) G. Mohan, Department of Earth Sciences, Indian Institute of Technology Bombay, Powai, Mumbai, Maharashtra 400 076, India. R. S. Dattatrayam, India Meteorological Department, Lodhi Road, New Delhi, India.

15 of 15

Você também pode gostar