Seismicity, gravity anomalies and lithospheric structure of the Andaman arc, NE Indian Ocean

Tectonophysics 460 (2008) 248–262

Contents lists available at ScienceDirect

Tectonophysics j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / t e c t o

Seismicity, gravity anomalies and lithospheric structure of the Andaman arc, NE Indian Ocean M. Radhakrishna a,⁎, S. Lasitha b,1, Manoj Mukhopadhyay c a b c

Department of Earth Sciences, Indian Institute of Technology Bombay, Powai, Mumbai 400 076, India Department of Marine Geology and Geophysics, Cochin University of Science and Technology, Fine Arts Avenue, Cochin-682 016, India Department of Earth and Environmental Sciences, Faculty of Science, Kuwait University, P.O. Box: 5969, Safat-13060, Kuwait

a r t i c l e

i n f o

Article history: Received 10 January 2008 Received in revised form 14 August 2008 Accepted 22 August 2008 Available online 5 September 2008 Keywords: Andaman arc Slab-Residual Gravity Anomaly Mantle Bouguer Anomaly Lithosphere structure Seismicity Indian Ocean

a b s t r a c t The Andaman arc in the northeastern Indian Ocean defines nearly 1100 km long active plate margin between the India and Burma plates where an oblique Benioff zone develops down to 200 km depth. Several easttrending seismologic sections taken across the Andaman Benioff Zone (ABZ) are presented here to detail the subduction zone geometry in a 3-D perspective. The slab gravity anomaly, computed from the 3-D ABZ configuration, is a smooth, long-wavelength and symmetric gravity high of 85 mGal amplitude centering to the immediate east of the Nicobar Island, where, a prominent gravity “high” follows the Nicobar Deep. The Slab-Residual Gravity Anomaly (SRGA) and Mantle Bouguer Anomaly (MBA) maps prepared for the Andaman plate margin bring out a double-peaked SRGA “low” in the range of −150 to −240 mGal and a wider-cumlarger MBA “low” having the amplitude of −280 to − 315 mGal demarcating the Andaman arc–trench system. The gravity models provide evidences for structural control in propagating the rupture within the lithosphere. The plate margin configuration below the Andaman arc is sliced by the West Andaman Fault (WAF) as well as by a set of sympathetic faults of various proportions, often cutting across the fore-arc sediment package. Some of these fore-arc thrust faults clearly give rise to considerably high post-seismic activity, but the seismic incidence along the WAF further east is comparatively much less particularly in the north, although, the lack of depth resolution for many of the events prohibits tracing the downward continuity of these faults. Tectonic correlation of the gravity-derived models presented here tends to favour the presence of oceanic crust below the Andaman–Nicobar Outer Arc Ridge. © 2008 Elsevier B.V. All rights reserved.

1. Introduction The Sumatra mega-thrust earthquake of 26 December 2004 (Mw ∼ 9.3) and its chain of aftershocks ruptured nearly 1300 km long segment of the plate margin along the Andaman and Sumatra trencharc region (Ammon et al., 2005; Ishii et al., 2005). Such an abnormally large seismic activity has brought to fore, the seismicity of the Indonesia arc system and its extension into the Andaman–Nicobar region. The segment of Andaman–Sumatra arc is characterized by oblique motion between the Indo-Australia and Burma–Sunda plates with predominantly thrust motion in the trench/fore-arc region and strike-slip motion in the back-arc region. The latter motion is mainly taken up by the ridge-transform system in the Andaman Sea and along the Sumatran fault in mainland Sumatra. The ridge-transform motion in Andaman Sea is believed to continue further north into Burma to connect to the Shan–Sagaing fault; together they form an important

⁎ Corresponding author. Formerly at: Department of Marine Geology and Geophysics, Cochin University of Science and Technology, Cochin, India. Tel.: +91 22 2576 7261; fax: +91 22 2576 7253. E-mail address: [email protected] (M. Radhakrishna). 1 Presently at: National Center for Antarctic and Ocean Research, Goa, India. 0040-1951/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2008.08.021

transitional tectonic link between the Eastern Himalaya to the north and the Sunda arc to the south (Fig. 1). All sizeable earthquakes in the Andaman region relate to subduction of the Indian plate presently occurring up to intermediate depths, though, noticeably, no large earthquake of M ≥8.0 has so far been reported from this region, the largest was 1881 Mw ∼ 7.9 event in the historical records (Ortiz and Bilham, 2003). However, the proximity of the Sumatran mega-earthquake of 26 December 2004, and the rupturing produced by this earthquake in the Andaman arc provide ample reason to cast a re-look into ABZ for delineation of its salient details, particularly, the variations in the configuration of the Benioff zone along and across the arc, active faults, current deformation of the accretionary wedge and anomalous subsurface mass distribution resulting from plate subduction. Characteristics of the rupture zone due to the Sumatran 2004 mega-thrust event and the co-seismic slip propagation northward in the Andaman–Nicobar Islands region have been studied in some detail from the seismic as well as geodetic measurements (Ammon et al., 2005; Lay et al., 2005; Banerjee et al., 2005; Catherine et al., 2005; Earnest et al., 2005; Jade et al., 2005; Gahalaut et al., 2006; Chlieh et al., 2007) and suggest significant variation in the rupture pattern along the arc. Multi-wave speed tomography carried out by Kennet and Cummins (2005)

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

249

Fig. 1. Shaded relief map of the Andaman arc showing various tectonic elements related to subduction of the Indian plate and opening of the Andaman Sea. Tectonic elements are adopted from Curray (2005). OCT— Ocean–continent transition, WAF— West Andaman Fault, NSR— North Sumatra Ridge, ASC— Andaman Spreading Center, SEU— Seulimeum strand of Sumatran fault, OWAF— Old West Andaman Fault. Filled triangles indicate location of volcanoes; Bold stars are the location of recent mega-thrust earthquakes (26 December 2004 and 28 March 2005) in offshore Sumatra. Numbered lines in the offshore indicate magnetic anomaly identifications (after Liu et al., 1983). Rectangle represents the study area; Profiles AA′ to EE′ are the gravity traverses used for delineating lithospheric structure.

250

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

indicates that rupture propagation was apparently controlled by changes in geometrical configuration of the subducting slab, its physical properties or the barriers related to the slab. Curray (2005) suggested strong heterogeneity in the crust and upper mantle as the lithosphere in the Andaman Sea region underwent complex tectonic deformations during the Neogene–Quaternary period. Previous investigations on seismicity in the subducting as well as the overriding plates suggest variations in interplate coupling from north to south along the Andaman arc (Dasgupta and Mukhopadhyay, 1993; Guzman-Speziale and Ni, 1996). Here, we present results of geophysical modeling for five lithospheric sections covering the entire Andaman arc in relation to known differences in seismogenic behaviour observed for the subduction zone, both along and across the arc. Dynamics of the Andaman subduction zone makes the arc vulnerable to largescale rupturing; the 1941 Andaman earthquake or Dec. 2004 Sumatra earthquake is illustrative example of this; although, it is uncertain if the arc itself is actually capable of producing mega-thrust events analogous to the size of Dec. 2004 Sumatran earthquake. 2. Regional tectonic setting and evolution of the Andaman arc–sea region The Andaman–Nicobar sedimentary islands which evolved during the Oligo-Miocene times (Rodolfo, 1969) form part of the fore-arc sedimentary complex, and west of these islands, the sediments of the Bengal fan are subducted and deformed below the Andaman trench (Curray et al., 1982). The sub-aerial expression of the islands separates the basin from the Bay of Bengal. The Andaman arc is characterized by the presence of east dipping Benioff zone down to about 200 km depth (Mukhopadhyay, 1984). Oblique, but predominantly thrust motion occurs in the Andaman trench with a convergence rate of about 1.4 cm/yr (Lay et al., 2005). The Andaman back-arc spreading ridge-transform system accommodates the remaining plate motion, joining with the Sumatra fault to the south. As a result of highly oblique motion between the Indo-Australian and Eurasian plates, a plate sliver, referred to as the Burma micro plate, has sheared off parallel to the subduction zone from Myanmar to Sumatra (Lay et al., 2005). Several north–south faults and thrusts are known in the Andaman–Nicobar ridge and the adjacent offshore areas, among them, the most significant are the Jarwa thrust developed on the main islands (Roy, 1983) and the WAF, east of the Andaman–Nicobar ridge (Curray et al., 1979); some of these faults/thrusts are clearly seismically active (Mukhopadhyay, 1984). The history as well as the emplacement mechanism of the Andaman ophiolites along its outer arc thrust zones is largely debated in literature: as upthrust oceanic crust due to subduction since late Mesozoic (Curray et al., 1979; Pal et al., 2003; Curray, 2005); to the collision history of Indo-Burma–Andaman (IBA) micro-continent during late Oligocene and initiation of subduction in late Miocene (Acharyya et al., 1989; Acharyya, 2007). The Andaman basin that forms the Andaman back-arc in the Andaman Sea lies between Burma and Sumatra with an average width of 650 km from the Malay Peninsula to the Andaman Nicobar Islands. Oblique subduction of the Indian plate in this region resulted in strikeslip faulting parallel to the trench; back-arc extension and basin formation in the Andaman Sea (Curray et al., 1979), whereas, the Andaman back-arc spreading relates to leaky transform tectonics (Uyeda and Kanamori, 1979). Eguchi et al. (1979) inferred collision of the Ninety East Ridge with the Andaman trench in the middle or late Miocene. According to them, the ridge-trench collision transmitted compressional stresses in the back-arc area and collision of India with Eurasia exerted a drag in the back-arc region that caused opening of the Andaman Sea. The age of this opening is inferred to be about 13 m.y. or in mid-Miocene (Curray et al., 1982). Kamesh Raju et al. (2004) postulated that seafloor spreading started in the Andaman back-arc basin 4 m.y. ago as a consequence of extrusion tectonics which prompted extension and rifting along the plane joining the

Sagaing and Sumatran fault systems. According to Curray (2005), the tectonic scenario developed here in the following sequence: the Mergui basin developed at ∼ 32 m.y, this conjoined the Alcock and Sewell Seamounts at 23 m.y., the East basin formed at ∼ 15 m.y., followed by separation of the Alcock and Sewell rises and formation of the central Andaman basin at 4 m.y. and shifting of the fault onshore from the Mentawai fault to the Sumatra fault system. 3. Analysis of seismicity data In the present study, we compiled the hypocentral and location parameters for all earthquakes of M ≥4.5 in the Andaman arc–sea region between 0 and 16°N latitudes and 90–100°E longitudes from ISC catalogue (ISC, 2001) as well as from NEIC listings corresponding to the period 1900–2005. The data are classified into two parts; one, pre- (1900–2004) and post- (2004–2005) 26 December 2004 megathrust earthquake (hereafter will be referred as pre- and post-tsunami events). From the ISC catalogue, we selectively eliminated those events whose hypocentral parameters are poorly determined. Events prior to 1964 have been compiled from Gutenberg and Richter (1954) and Rothe (1969). Engdahl et al. (1998) relocated large number of events from the ISC and NEIC listing for the period 1964–1995 and prepared a global data set having improved focal depth estimation. The events from this data set falling in the present study region are considered here. We also consider all Harvard CMT solutions available for the Andaman region in addition to several published focal mechanisms from previous investigations (Dasgupta and Mukhopadhyay, 1993; Guzman-Speziale and Ni, 1996; Dasgupta et al., 2003) to study the stress distribution and faulting pattern in the ABZ and its overriding plate. The derived Benioff zone structure is next used to compute the gravity effect for three-dimensional geometry of the slab. 3.1. Configuration of Andaman Benioff Zone (ABZ) Based on available earthquake data up to 1993, Dasgupta et al. (2003) constructed 29 depth sections representing several well-defined blocks in the Burmese–Andaman arc; of these, 11 depth sections belong to the Andaman proper. Here we revise these sections by taking into account the additional data referred above (Fig. 2). The procedure for plotting the sections in a vertical plane and the grid areas represented by the individual sections are the same as adopted by Dasgupta et al. 2003). The dip of the ABZ is mostly sub-vertical— varying between 38 and 53°. In north Andaman (sections: S1–S3), the Benioff zone dip is between 43 and 53°, while, in south Andaman– north Sumatra (sections: S4–S11), the dip angle decreases to 38°–50°. This clearly signifies that the dipping Indian plate does not maintain uniform geometry throughout; rather it is flexed by several transverse faults displaying contortions in the Benioff zone between Andaman and Sumatra. The top surface of the Benioff zone as identified on the sections allows us to prepare a contour map representing the dipping lithosphere. Previous attempts to map the dipping Indian lithosphere below the Andaman arc have been by: using Hypocentral Trend Surface (HTS) analysis (Guzman-Speziale and Ni, 1996), by closer grid bases of the shallow-depth Benioff zone hypocenters (Dasgupta et al., 2003), or by using smooth curves drawn through hypocenters (Khan and Chakraborty, 2005). The Benioff zone sections prepared in the present study are actually a revised version from earlier studies, mainly by Dasgupta et al. (2003), by virtue of inclusion of additional data. An analysis of all 11 seismic sections reveals distinct seismic character of the arc. It is known that the trench is an important tectonic element in the subduction zone environment. The earthquake activity appears to be continuous starting from the trench location and toward east, particularly in south Andaman (S5, S7–S11). The earthquakes are mostly confined to top part of the lithosphere where the downbending of the slab is first initiated. Typically in this region, a seismic gap exists at depths of 90–110 km in the central Andaman arc

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

251

Fig. 2. Sections showing the Benioff zone configuration based on seismic incidence (period: 1900–2004) across the Andaman arc region. Inset shows location of these sections (S1–S11). TA— Trench axis; OAR— Outer Arc Ridge; ASC— Andaman Spreading Center; OCT— Ocean continent transition; SF— Sumatran fault; RF— Renong Fault.

252

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

beneath the Barren Island volcano (Dasgupta and Mukhopadhyay, 1997), where, space–time downward propagation in the occurrence of large magnitude shocks is further evidenced (Dasgupta et al., 2007). 3.2. Focal mechanisms and stress distribution below Andaman arc Seismotectonics, the nature of faulting and stress distribution pattern for the Andaman arc are already well known (Mukhopadhyay, 1984; Dasgupta and Mukhopadhyay, 1993; Ravikumar et al., 1996; Guzman-Speziale and Ni, 1996; Radhakrishna and Sanu, 2002; Dasgupta et al., 2003; Khan and Chakraborty, 2005). On a scrutiny of the best available solutions, here we utilize some 173 focal mechanism solutions: 25 from previous studies and 148 Harvard CMT solutions up to 2004. We use the mean slip angle method (Ravikumar et al., 1996) to categorize these 173 mechanisms. Based on this classification, 32 mechanisms show normal faulting, 104 mechanisms strike-slip faulting and the remaining 37 events show thrust faulting. We have further compiled Harvard CMT solutions pertaining to post-tsunami events during 2004–2005. The mechanism solutions are accordingly grouped into three categories (Fig. 3): i) pre-tsunami shallow upper plate (≤70 km) mechanisms (Fig. 3a), ii) pre-tsunami deeper mechanisms pertaining to the ABZ (Fig. 3b) and iii) post-tsunami mechanisms related to aftershock activity all along the Andaman arc (Fig. 3c). For clarity, we avoided plotting events that show similar faulting pattern in close by localities. The events in south nearer to the trench show strike-slip mechanisms with right-lateral shear motion on easttrending nodal planes indicating the presence of transverse seismic zone across the Andaman trench, whereas, further east, the earthquakes in the Andaman outer arc region show dominantly thrust or low-angle thrust faulting mechanisms with minor strike-slip component, and few normal and pure strike-slip events (Fig. 3a). The Andaman back-arc spreading ridge is characterized by several normal faulting events with NE-SW trending nodal planes or strike-slip mechanism with left-lateral shear on east-trending nodal plane. In south Andaman where the Andaman Spreading Ridge (ASR) meets the WAF, the mechanisms indicate mostly right-lateral shear motion along N–S trending nodal plane. The deeper part of the subducting slab is characterized by the occurrence of down-dip compression (DDC), reverse faulting and strike-slip events (Fig. 3b). The post-tsunami events and their mechanisms (Fig. 3c) are utilized here to understand the structural control on rupture propagation along the arc— refer Section 7.1. The NW-SE trending stress field obtained from average T-axis azimuths along the ASR indicates the direction of back-arc extension which matches well with the spreading direction obtained from magnetic anomalies by Curray et al. (1979). Based on cumulative seismic energy, Ravikumar et al. (1996) observed that the strike-slip events contribute significant amount of energy in the whole of the Burmese and Andaman arcs; indicating large-scale right-lateral slip as the dominant motion. Radhakrishna and Sanu (2002) also noticed the absence of thrust events with large apparent stress values for the Andaman arc; implying that their contribution is not significant. Many strike-slip events in the deeper part of the Benioff zone suggest the presence of transverse faults due to changes in the dip of the Benioff zone (Dasgupta et al., 2003). We consider this observation as highly significant for the Andaman arc in view of the fact that the Indian plate subducts here to intermediate depths (150–220 km). This may imply that the dipping lithosphere is being dragged northward without commonly producing large thrust earthquakes. South of the Nicobar Islands where the arc geometry is more arcuate, this situation rapidly changes to grow into subduction-related thrust mechanisms however. 4. Gravity effect of 3-D subducting slab A general practice in the forward modeling of gravity data is to remove contributions from known sources/structures. It is known that

gravity modeling and crustal structure investigations in the island arc–trench areas must consider the geometry of the subducting lithosphere as it induces large-scale mass transfer in the region. They produce rather long-wavelength gravity anomalies at the surface as the descending slab is thermally colder (Minear and Toksoz, 1970), seismically denser (Utsu, 1971) and penetrates into the lighter asthenosphere, producing gravity mass anomalies (Grow, 1973). Crustal seismic sections available for Japan were utilized by Yoshii (1973) to obtain first the residual gravity anomalies after removing the gravity effect of crustal layers and these were next explained due to subducting lithosphere. The gravity effect of the descending lithosphere below ABZ has been studied previously using the ABZ configuration (Mukhopadhyay, 1988; Mukhopadhyay and Krishna, 1991; Kamesh Raju et al., 2007). However, all these studies merely take into account the 2-D-slab contribution, whilst, more realistic modeling requires 3D configuration parameters of the descending slab for calculation of its gravity effect, what is termed as the Slab-Residual Gravity Anomaly (SRGA) (Furuse and Kono, 2003). Notice that an SRGA map basically represents the gravity contributions due to crustal structure variation below an arc–trench. Here we consider the 3-D configuration of the subducting Indian plate derived from the ABZ geometry, drawn along the 11 sections which are discussed in the foregoing (Fig. 2). Generally, high-resolution slab configuration in the trench-arc regions is possible through micro earthquake investigations or other seismological studies. Such data are absent for the Andaman arc. Additional shortcomings arise from the grossly linear disposition of the Andaman–Nicobar Islands aligned north–south and also due to non-availability of azimuthally covered seismic stations data from the surrounding continents. Therefore, the hypocentral distribution of earthquakes provides, as a first approximation, the three-dimensional configuration of the subducting lithosphere in the region. The isodepth contour map constructed on top surface of the subducting slab (Fig. 4) indicates that the depth of the descending slab increases eastward from 80–220 km with maximum penetration depth in the south Andaman Sea. We use the method of Talwani and Ewing (1960) to calculate the gravity anomaly for the 3-D subducting slab below ABZ. For this purpose, we divided the slab into several horizontal cross-sections defined by the depth contours. The gravity effects of these horizontal laminae are next calculated and numerically integrated over depths from top to the deepest point. We consider 5 km × 5 km grids covering the study area for calculating the total gravity effect of the slab. The density contrast between the lithospheric slab and the asthenosphere is considered as 0.065 g/cm3. A similar density contrast has been adopted for other island–arc regions (Furuse and Kono, 2003), based on seismic velocity distribution, studied by Yoshii (1973) for the Japanese islands. The lithospheric plate thickness in the Andaman arc region is considered as 70 km from the surface wave dispersion studies in the northeastern Indian Ocean (Singh, 1990). Furuse and Kono (2003) demonstrate that computed slab anomaly depends on the choice of various parameters such as the density contrast, subducting plate thickness, slab thinning and phase transitions within the slab, out of which, the density contrast and plate thickness have dominant effect. A density contrast of 0.05 g/cm3 was assumed by Grow (1973). In order to evaluate the effect of this parameter, we also calculated the slab contribution for a density contrast of 0.05 g/cm3. A comparison of the slab contribution from the two-density contrasts shows similar anomaly pattern but the amplitude of the gravity high differs nearly by 20 mGal. Here, we finally chose a density contrast of 0.065 g/cm3 as it was the seismically constrained value. However, it should be kept in mind that adopting a higher value would give rise to an upper bound to the crustal thickness that will be discussed in the next section. The effect of phase transitions in the slab may not be significant in the study region as the maximum penetration depth is below 300 km along the Andaman arc. In the absence of any knowledge on the detailed seismic structure of the descending slab, some trade off between the choices of these

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

253

Fig. 3. Earthquake epicenters (1900–2005) and focal mechanisms (Harvard CMT solutions during 1977–2005; Mw ≥5.0) plotted on bathymetric relief map in the Andaman arc and Andaman Sea region. The data are classified into events before (1900–2004) and after (2004–2005) the 26 December 2004 mega-thrust earthquake. A) Plot of shallow (≤70 km) events and mechanisms during 1900–2004. B) Plot of deeper Benioff zone (≥70 km) events and mechanisms during 1900–2004. Numbers adjacent to each mechanism indicate depth of the event. C) Plot of events and mechanisms occurred during 2004–2005 after the mega-thrust earthquake off Sumatra indicating post-seismic deformation. For clarity, mechanisms of major earthquakes are only shown. Details are discussed in the text.

parameters definitely required and can be concluded that slab gravity anomaly is important in delineating realistic crustal models in arc– trench regions. The slab gravity anomaly map computed for the Andaman arc (Fig. 4) reveals a smooth, long-wavelength and symmetric gravity high of maximum 85 mGal centered to the immediate east of the Nicobar Islands, where, the gravity “high” follows the Nicobar Deep. The slab contribution reaches up to ∼20 mGal below the trench axis but gradually diminishes eastward to about 5 mGal at the Malayan margin.

5. Gravity anomaly maps 5.1. Free-air anomaly The gravity field of the Andaman arc was first described by Peter et al. (1966), and subsequently by Mukhopadhyay (1988) and Mukhopadhyay and Krishna (1991). These were based on the available ship track gravity data for the region, where, large data gaps still create hindrance for detailed gravity modeling. To partially overcome this

254

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

Fig. 4. Contour map showing the 3-D-slab gravity effect (in mGal) calculated from the Benioff zone depth sections shown in Fig. 2. Thick numbered dashed lines represent top surface of the Benioff zone (in km) below the Andaman arc. Location of the trench to the west of Andaman–Nicobar Islands is indicated with hatched line.

difficulty, the 2-minute grid GEOSAT gravity anomalies (Sandwell and Smith, 1997) are found to be useful to study gravity field as they provide fairly uniform coverage. Such anomalies have a resolution up to about 23 km wavelength and are accurate within 5–10 mGal even in rough bathymetric areas. As a result, these are considered very useful in regional tectonic studies (Marks, 1996; Radhakrishna and Searle, 2006). Based on such data, the Free-air anomaly map prepared for the Andaman arc–sea, with superposed tectonic elements (Fig. 5a), reveals a broad and distinct gravity low of as much as −200 mGal over the Andaman arc. Near 8.5°N, the low anomaly shifts towards east into the Nicobar Deep. Hugging this low, two strong N–S trending highs can be seen over the Invisible Bank and the West Sewell ridge. Another gravity low with values around −40 to −60 mGal occurs locally along the Andaman trench. Sandwiched between these is the outer arc sediment accretionary prism comprising of Andaman, Nicobar and Nias Islands, with their attendant gravity “high”. The Ninety East Ridge (NER) is characterized by a large positive gravity field striking north where Free-air anomalies reach up to +60 mGal. The large gravity high can be seen extending from south to the north as far as 10°N, where the ridge has sub-bottom topography but the gravity high gets gradually subdued further north (Mukhopadhyay and Krishna, 1995), as there, the NER rapidly plunges below the Bengal Fan sediments and runs oblique to the Andaman arc (Curray et al., 1982). Consequently the Andaman trench gravity low near 8°N is seen disrupted by the positive gravity field of the ridge suggesting partial subduction of the ridge below Andaman trench. The Andaman arc gravity “low” is consequently displaced to the east in the Andaman Sea (Fig. 5a). This has been interpreted by Subrahmanyam et al. (2007) as due to the effect of convergence of NER upon the Andaman arc. East of the arc, in the Andaman basin, the gravity value displays substantial variation (−20 to 20 mGal) locally over the ASR,

whereas, both the Alcock and Sewell rises commonly describe a gravity high of amplitude 60 mGal. In the east basin, the Mergui ridge is an area of large and linear gravity “high” of 50 mGal. This shelf-edge positive gravity field can be seen all along the Malayan shelf, which is quite characteristic for ocean–continent crustal transition. 5.2. Slab-Residual Gravity Anomaly (SRGA) The computed slab anomaly (Fig. 4) has been subtracted from the Free-air anomaly map presented in Fig. 5a in order to obtain the SRGA map. For this purpose, both data sets were converted into a 5 km × 5 km grid. The resultant SRGA map (Fig. 5b) illustrates the slab contribution coming mainly from layers below 70 km depth in the asthenosphere, the SRGA should therefore be attributed to mass anomalies within the lithosphere i.e., the crust and the lithospheric mantle. The SRGA map essentially shows similar anomaly pattern as the Free-air anomaly map with an important distinction that anomalies change. The anomaly amplitude decreases in certain areas, such as the Mergui Terrace or the Sewell rise, whereas; the anomaly amplitude for the Andaman arc proper clearly enhances. As the slab anomaly computed here affects only the immediate surroundings of the arc, the regions seaward of the trench along NER or the gravity field of the Malayan shelf are not significantly affected. We next consider this map for modeling the crustal and sub-crustal mass anomalies along a number of profiles transecting the Andaman arc. 5.3. Mantle Bouguer Anomaly (MBA) The Free-air anomaly map is dominated in general by the gravity attraction of the density contrast at the seafloor. The subsurface

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

255

Fig. 5. Tectonic map with gravity anomaly overlay for the Andaman arc–sea: A) Free-air anomaly map obtained from satellite-derived (GEOSAT) gravity data (after Sandwell and Smith, 1997). Axis of the gravity low is indicated as a thick dashed line. B) Slab-Residual Gravity Anomaly (SRGA) deduced by subtracting the 3-D-slab effect (Fig. 4) from the Free-air anomaly map and C) Mantle Bouguer Anomaly (MBA) map of the region after subtracting the slab effect. Details are discussed in text.

density structure can be investigated by applying mantle Bouguer correction following the method of Prince and Forsyth (1988). The predicted gravity signal of the seafloor–water interface and crust– mantle interface assuming a constant thickness crustal layer of 6 km is obtained using the Fourier transform method of Parker (1972). This

gravity effect when subtracted from the Free-air gravity will give rise to Mantle Bouguer Anomalies (MBA), a useful tool to decipher crustal thickness variations. The bathymetry data required for this purpose is considered from ∼1–2 km interval digitized GEBCO database which is a compilation of accumulated ship track data from many sources (refer

256

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

for instance; Subrahmanyam et al., 2005 for a comparative study between GEBCO bathymetry with ship track bathymetry data). Excellent correlation of Free-air anomalies with respective morpho-

tectonic features in the Andaman arc–sea region renders further justification for computation of MBA. However, we refrain ourselves from subjecting this map for any quantitative interpretation.

Fig. 6. Gravity-derived lithospheric configuration along the profiles AA′ and BB′ in the northern most Andaman arc region (Fig. 1 for location). The post-tsunami events shown in Fig. 3c are projected on to the crustal section to understand the relation of crustal structure and seismogenic behaviour. Filled circle— smaller events having mb b5.0, Triangle— events mb ≥5.0 and Star— focal mechanisms of large events (Mw N5.5) with type of faulting shown Thrust (T), Normal (N), Strike-slip (S). ANR— Andaman Nicobar Ridge; WAF— West Andaman Fault; ASC— Andaman Spreading Center. Right-lateral shear motion is indicated for WAF.

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

Gravity and bathymetry data for the Andaman arc–sea region were interpolated onto 4 km grid interval, excluding the very shallow shelf or the land positions which are eliminated from the purview of the MBA computation. The computation involves the following standard density values: water layer (1.03 g/cm3), crust (2.7 g/cm3) and mantle (3.3 g/cm3). Though the compiled seismic reflection records in the

257

region are used by Curray (2005) for constructing the basement topography in many areas, it is not possible to derive an unambiguous sediment isopach map down to 4 km-interpolated grid, hence this layer was not considered in the present computation. The MBA was then computed by removing the net mantle Bouguer correction from the Free-air anomaly at the respective grid points (Fig. 5c). The deeper

Fig. 7. Gravity-derived lithospheric configuration along profiles CC′ and DD′ in the central Andaman arc region. Other details are as in Fig. 6. Smaller events (mb b 5.0) are not shown in DD′ as there are too many such events. ND— Nicobar Deep; EB— Eastern Basin.

258

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

slab effect was also subtracted at these grid points, so that, the slabresidual MBA essentially reflects the crustal thickness variations or density heterogeneities. The MBA map shows a high anomaly coinciding with NER, a broad and continuous gravity low zone between the Andaman trench and arc, gravity high over the Andaman spreading ridge flanked by minor negative anomalies over the Alcock and Sewell rises, and large negative anomalies in the eastern part of study region over the Malayan shelf. While, low MBA areas reflect either the crustal thickening or presence of low-density crustal material, the positive MBA values indicate crustal thinning or densification of the crust. It may be noted that crustal thickness in the back-arc ridges shows systematic variation with arc proximity rather than the ridge-spreading rate or subduction effects (the latter producing the so-called mantle wedge composition), thus giving rise to opposed trends in crustal thickness and spreading rate. Such a situation is believed to exist with the Lau Basin spreading center (Martinez and Taylor, 2002; Martinez et al., 2006). Therefore, the across-basin gravity gradient due to slab and subduction effects must be removed for detailed interpretation of MBA map. Here, we utilize the observed MBA map to further refine the seismically derived gravity modeling for the Andaman arc. 6. Gravity modeling In order to delineate the crustal and lithospheric mantle configuration along and across the Andaman arc, five regional gravity traverses, AA′ through EE′ (Fig. 1 for location) are chosen from the 11 seismological sections considered in Fig. 2, to cover different segments of the arc starting from NER on the west and extending to the Malayan shelf towards east. Figs. 6–8 illustrate the changes in SRGA and MBA along these profiles. The SRGA values together with the bathymetry, constrained by tectonic and sediment thickness details along these profiles, are interpreted in terms of 2-D-lithosphere structure underlying the Andaman arc–sea.

6.1. Sediment thickness and nature of basement Sediment thickness values for a large number of control points on the basement underlying the profiles AA′ to EE′ are taken into account for modeling the deeper structure. These discrete thickness values along with the data on crustal velocity for areas west of the Andaman Islands are taken from Curray et al. (1982), but for the deeper part of the trench, these values are projected from the NGDC sediment thickness grid. Data on sediment thickness and basement for the Andaman Sea are compiled from available sources (Peter et al., 1966; Weeks et al., 1967; Rodolfo, 1969; Curray et al., 1982; Curray, 2005), which together provide a reasonable picture on basement structure along these profiles. An important conclusion derived from seismic data as well as litho-stratigraphic information collected from commercial drilled wells and dredged rock samples in the Andaman arc–sea region is that the Andaman basin is underlain by oceanic crust with a thick pile of volcanic layer on top below Alcock and Sewell rises and the east basin (Curray, 2005). Further east, the ocean–continent transition occurs at the Malayan shelf which continues northward into Shan plateau as well as southward into Sumatran fault system (Fig. 1). The Mergui basin in the southern Andaman Sea is underlain by rifted continental crust which most likely extends further west to the North Sumatra Ridge (NSR). Contrasting views are expressed by different workers regarding the nature of crust, continental vs. oceanic, underlying the Andaman–Nicobar Outer Arc Ridge: (a) Acharyya (2007) postulates continental crust at this location on the basis of structural and stratigraphic reconstruction of onshore geological data. (b) On the other hand, Curray (2005) inferred oceanic crust by using a reversed seismic section in the offshore Indo-Burman Ranges (where basement velocities range here between 6.3 and 6.9 km/s). Based on well data as well as multi-channel seismic reflection lines in and around Andaman Islands, Roy (1992) observed the presence of a highly complex basement structure consisting of folded thrust packets successively under thrust from east to west, a tectonic mélange structure in the Cretaceous section embodied in the oceanic basement or ophiolites.

Fig. 8. Gravity-derived lithospheric configuration along profile EE′ in the southern part of Andaman arc. Other details are as in Figs. 6 and 7. MR— Mergui Ridge.

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

Further south, in the offshore Sumatra, seismic refraction data of Kieckhefer et al. (1980) suggests the presence of nearly 13 km thick sediment wedge of 4.7–4.9 km/s overlying the oceanic basement and an oceanic layer 3 of 6.6–7.5 km/s in the outer arc Nias Island, a 6.5– 6.8 km/s layer inferred as a lower continental crust or oceanic layer 3 below fore-arc basin and the Sumatran coast. Gravity modeling in the offshore Sumatra indicates that the outer arc ridge is underlain by a high density material containing oceanic crustal fragments, whereas, the continental crust may be present below the fore-arc basin (Kieckhefer et al., 1981). In the light of above geophysical observations, the inference of continental crust below NSR, further east of the outer arc is meaningful, whereas, we tend to believe that the outer arc Andaman Nicobar Ridge may be underlain by oceanic crust. 6.2. Interpreted gravity models The 2-D-forward gravity modeling presented here for the Andaman arc–sea delineates the lithospheric structure based upon the available seismic control. As it can be seen that seismic data control is reasonably high along the profiles, consequently these are held fixed to infer structure in the surrounding areas for modeling purpose. The modeling is carried out using the USGS SAKI program. Mukhopadhyay and Krishna (1991) estimated the rock density values from seismic velocities for north part of the NE Indian Ocean. Their values suggest densities of 2.4 g/cm3 and 2.6 g/cm3 for upper and lower parts of the sediments where sediment thickness exceeds by 5.0 km. We adopted here the similar density structure for modeling. The densities for other rock types in the region are: volcanics (2.6 g/cm3), oceanic crust (2.9 g/cm3), and continental/transitional crust below Malayan shelf (2.85 g/cm3). Mukhopadhyay and Krishna (1991) also inferred a 60 km wide lowdensity column below the volcanic arc/spreading ridge. We adopted a density of 3.29 g/cm3 for this column below spreading ridge, with lithosphere and asthenosphere having respective densities of 3.3 g/cm3 and 3.235 g/cm3. Some salient features of gravity anomaly variations and structural details from model studies are discussed below. The SRGA values in each profile show two broad lows, one associated with the trench (−60 to −100 mGal) and the other related to the Andaman arc along the ABZ (−150 to −240 mGal). The slab-residual MBA values show a distinct low of the order of −280 to −315 mGal correlating with the Andaman arc. A comparison of the SRGA and MBA along these profiles indicates that the double-peaked SRGA associated with the Andaman trench-arc system shows up as a broad (250– 300 km wide) MBA low. Further east over the Mergui Terrace, the MBA values are in the order of −155 to −165 mGal. It is interesting to notice that while the SRGA are used to derive the lithosphere structure in the gravity modeling, the interpreted crustal configuration for profiles AA′–EE′ closely resembles the shape of MBA, giving further credence that MBA reflects crustal thickness variations both along and across the arc. The gravity models suggest for a thinner oceanic crust and 3–6 km of underplated material below the NER, whereas, 40–47 km thick upper lithospheric layer consisting mainly of oceanic crust underlies the Andaman Nicobar Ridge (ANR). The presence of thicker and wider zone of oceanic crust may be due to accretion of oceanic crust to the southeast Asian plate during westward shifting of subduction front since Cretaceous time (Roy, 1992). Roy et al. (2008) speculate that subduction along the arc has probably remained active since the Cretaceous, where, repetitive upthrusting of accretionary wedges ultimately formed the ANR. Within such framework of subduction, large variations in the thickness of oceanic crust can be observed if imbricate faulting extends downward into the oceanic layer 3, thereby; portions of oceanic crust can actually get incorporated into the accretionary prism (Kieckhefer et al., 1980, 1981). A thicker and wider magnetic crust (Purucker and Ishihara, 2005) observed along the outer arc ANR correlate with the MBA low. We can only surmise that oceanic crust forms parts of the upper lithospheric layer but deep

259

crustal seismics are needed to ascertain the relative thickness of the oceanic crust at this crucial position. Below Alcock rise, the crustal layer may be as thick as 15–20 km whose top layer (3–5 km) is volcanic (profiles AA′ and BB′); however, this quickly thins out on reaching the ASC. The models show different proportions of thinning below the ASC from north to south (profiles AA′–CC′). The profiles DD′ and EE′ crossing the Eastern Basin shows that the basin is underlain by 12–18 km thick oceanic crust with wide spread volcanic layer on the top. An extended and further thinned continental crust of thickness 15–25 km is inferred by gravity modeling for the Mergui Terrace; this evidently has been created in consequence to crustal transition under the Terrace. The stretched continental crust is thicker in the north but is distinctly thinner in the south Andaman Sea. This is probably because of additional stretching the crust encountered by extension in the Mergui basin (Curray, 2005). 7. Discussion From the gravity models it is apparent that a wider-cum-larger MBA “low”, of amplitude −280 to −315 mGal, demarcates the ABZ. The SRGA are also low in the range of −150 to −240 mGal in this region. Over the Mergui Terrace the MBA values are largely negative of the order of −155 to −165 mGal suggesting what order of gravity variation is to be expected from mantle sources where the crust is barely thicker than 6 km. Notice that as compared to the Malayan margin, the Andaman subduction zone is associated with MBA what is distinctly more negative by 75–120 mGal. Grevemeyer and Tiwari (2006) also find that Bouguer gravity anomalies correlate well with the occurrence of large mega-thrust earthquakes in the Sunda subduction zone; negative anomalies mark segments characterized by larger earthquakes, while positive anomalies indicate lower seismic potential. In direct contrast to the Andaman subduction zone, the NER is associated with a positive gravity bias (Mukhopadhyay and Krishna, 1995). The NER is believed to be the Kerguelen hotspot trace and has a volcanic emplacement below it. The Moho depth below the NER crust is more than the adjacent oceanic crust. The study reveal that east of the NER sediment thickness increases towards the Andaman–Sunda trench slope, where the sediments are accreted and are much deformed, with thickness of the order of 9–10 km as also observed by Curray et al. (1982) from seismic reflection and multi-channel seismic data. Based on flexural-cum-forward gravity modeling for NER, Subrahmanyam et al. (2007) suggest that the ridge is at its starting phase of collision with the Andaman arc. Tomographic imaging of the upper mantle by Miller and Lee (2008) revealed an anomalously low P-wave signal at 60–160 km within the subducting slab in the Andaman arc that probably resulted from chemical modification of the lithosphere due to subduction of NER below the arc. The MBA map discussed above presents gravity signature of the subduction process. The seismic refraction studies by Moore and Curray (1980) and Kieckhefer et al. (1980) in the Sunda fore-arc region indicate a thick crust of 40 km in the fore-arc region which thins towards Sumatra mainland to 25 km. From the models (AA′ to EE′), it can be seen that the crust exceeds thickness of 40 km at many places under the ANR. Crustal thinning is observed below the back-arc spreading ridge which bisects the Andaman volcanic arc; the latter is constituted by the Barren, Sewell and Alcock seamounts. The penetration depth of the ABZ largely increases from about 150 km to 240 km along profiles AA′ to EE′ in north–south direction. The lithospheric models presented here bring out the geometry and structural heterogeneities of the Burma micro plate. 7.1. Relationship between structural configuration and post-seismic deformation It is known that the tectonic and structural elements play a crucial role in governing the nucleation, growth or the arrest of rupture

260

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

propagation. In order to understand the correlation between crustal geometry and seismogenic behaviour which controls the stress propagation along the arc, the post-tsunami earthquakes (Mw N5.0) and mechanisms (Mw N5.5) were projected and plotted on each model section. Large incidence of post-tsunami seismic activity (aftershocks) along the arc can be seen rupturing the whole plate in plan view, with most of the events confining to either the shallow ABZ along the outer arc or ridge-transform structures in the Andaman Sea and along the WAF (Fig. 3c). According to Gahalaut and Catherine (2008), the aftershocks and post-seismic deformation along the arc follow similar relationship, though with different decay times. The post-tsunami activity shows, dominantly thrust faulting (arc-normal compression) events with few strike-slip and normal events along the outer arc region, normal and strike-slip events in the back-arc region (Fig. 3c). Another characteristic feature of the aftershock activity is the occurrence of the most energetic swarm event ever observed globally (Lay et al., 2005) centered near 8°N during January 2005. A detailed analysis of this swarm sequence indicates intrusion of magmatic dyke in the crustal weak zone and believed to have generated a new rift basin (Mukhopadhyay and Dasgupta, in press). The seismic activity associated with the WAF probably reflects that the stress propagation found passage through the weak plane along the WAF (Kamesh Raju et al., 2007). In fact, the models indicate that the post-tsunami activity exhibited along WAF was remarkably low in the north (AA′ and BB′), rather the aftershock activity for WAF was more intense underlying profiles DD′ and EE′. This probably is a consequence of complex interaction between WAF with shorter segments of the Andaman spreading ridge near the Sewell rise that locally produces normal faulting events (profile DD′). However, very little geophysical data constraint is presently available for determining the configuration of the WAF, although, this serves as an important tectonic link crossing the Andaman Sea. Together with seismics, close-spaced gravity station coverage is required to improve our understanding on the WAF. The plate margin configuration attempted here suggests that together with the WAF, there is a set of sympathetic faults of various proportions slicing the plate margin, often cutting across the fore-arc sediment package which is also responsible for stress propagation (Figs. 6–8). Based on detailed morphological and aftershocks study in northern Sumatra (south of profile EE′), Sibuet et al. (2007) suggest that most of the post-seismic deformation took place along two thrust faults— the lower and median thrusts, in the outer arc. Araki et al. (2006) earlier interpreted some of these (active) thrust faults as splay faults. The present data can not resolve and establish their depth continuity beyond doubts (look at the whole range of seismic clusters at about 33 km depth in most sections, signaling, how poor the focal depth control is). Sibuet et al. (2007) further argue that the WAF and Sumatran fault zone which did not show much post-seismic activity could be potential areas for future rupture. The long-term deformation pattern estimated along the Andaman and Sumatra arc prior to and after the 26 December 2004 mega-thrust earthquake also indicates significantly low post-seismic deformation along these two faults (Lasitha, 2007). Kennet and Cummins (2005) attributed variations in the speed of rupture propagation northward due to changes in bulk modulus property in the subducting slab and the 26 Dec.2004 event is huge enough to overcome multiple physical barriers and rupture the plate boundary for large distances. Most recently, Shapiro et al. (2008) interpreted the slow rupture propagation in the Andaman segment due to weaker seismic coupling resulted from combination of less buoyant subducting plate and the weak back-arc lithosphere. Dasgupta et al. (2007) have noticed that seismic activity of relatively large magnitudes in the Sunda Arc region display a space–time relationship; the bigger events are of increasing focal depth as the time progresses. This observation, although based on only 25 years' of data, offers an interesting possibility to look at the data sets from the viewpoint of seismic precursors for large mega-shocks. It is suggested by some workers (cf. Sieh and Natawidjaja, 2000) that the upper plate boundary plays a key role in defining the size and

nature of such mega-thrust earthquakes. The detailed lithospheric sections presented here lend some support to this hypothesis; although, it requires detailing by better constrained seismicity data in terms of location and hypocentral depth that can be provided only through local seismic network, both on land and OBS. The aftershock pattern in the Andaman region indicates the influence of lithospheric plate boundaries which was confirmed through the gravity models presented here. However, more detailed marine geophysical studies and GPS measurements are necessary to understand the structure and tectonics of this highly complex segment of the convergent plate boundary. 8. Conclusions Some significant results in terms of lithospheric structure and deformation pattern in the Andaman arc region have been obtained based on epicentral distribution of earthquakes, focal mechanism solutions and interpretation of gravity anomalies. The study has brought out the relationship between the crustal structure and postseismic deformation along the arc. These are summarized as below: • Dip of the ABZ varies in general between 43 and 53° in the northern Andaman and between 38 and 50° in south Andaman–northern Sumatra region. • The slab gravity anomaly computed from the three-dimensional ABZ configuration reveals a smooth, long-wavelength and symmetric gravity high of maximum 85 mGal centered to the immediate east of the Nicobar Islands, where, the gravity “high” follows the Nicobar Deep. The slab contribution is ∼20 mGal below the trench but diminishes away from the trench to about 5 mGal at the Malayan margin. • The Slab-Residual Gravity Anomaly (SRGA) and Mantle Bouguer Anomaly (MBA) maps of the region indicate that the double-peaked SRGA low in the range of −150 to −240 mGal and a wider-cum-larger MBA “low”, of amplitude −280 to −315 mGal demarcate the Andaman trench–arc system. The gravity-derived lithospheric structure correlates with the intensely deformed region of the Burma micro plate, where, several significant thrust and strike-slip faulting events are seen within the folded sedimentary wedge of the outer arc and thicker igneous crust. • Large incidence of post-seismic activity mainly confine to the shallow ABZ along the outer arc or ridge-transform structures in the Andaman Sea and along the WAF. The plate margin configuration inferred here further suggests that together with the WAF, there is a set of sympathetic faults of various proportions responsible for stress propagation. The post-seismic activity along WAF is significantly low in the north Andaman Sea. • Lack of depth resolution for many of the events prohibits understanding the downward continuity and correlation. For a more detailed understanding of seismogenesis, we recommend for a focused and concerted study in the Andaman arc region. Acknowledgements The author S.L. thanks the CSIR, New Delhi and Cochin University of Science and Technology for awarding the research fellowship. Critical comments and suggestions from Dr. C. Subrahmanyam, Dr. S.K. Acharyya and Prof. Tom Parsons helped to improve the manuscript. Authors thank Prof. Joseph Curray for many valuable clarifications at the time of revising the manuscript. Maps were prepared with GMT software (Wessel and Smith, 1995). References Acharyya, S.K., 2007. Collisional emplacement history of the Naga-Andaman ophiolites and the position of the eastern Indian suture. Journal of Asian Earth Sciences 29, 229–242. Acharyya, S.K., Ray, K.K., Roy, D.K., 1989. Tectono-stratigraphy and emplacement history of the ophiolite assemblage from the Naga Hills and Andaman Island Arc, India. Journal Geological Society of India 33, 4–18.

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262 Ammon, C.J., Ji, C., Thio, H.K., Robinson, D., Ni, S., Hjorleifsdottir, V., Kanamori, H., Lay, H., Das, S., Helmberger, D., Ichinose, G., Polet, J., Wald, D., 2005. Rupture process of the 2004 Sumatra–Andaman Earthquake. Science 308, 1133–1139. Araki, E., Shinohara, M., Obana, K., Yamada, T., Kaneda, Y., Kanazawa, T., Suyehiro, K., 2006. Aftershock distribution of the 26 December 2004 Sumatra–Andaman earthquake from ocean bottom seismographic observation. Earth Planets Space 58, 113–119. Banerjee, P., Pollitz, F.F., Burgmann, R., 2005. The size and duration of the Sumatra– Andaman earthquake from far-field static offsets. Science 308, 1769–1771. Catherine, J.K., Gahalaut, V.K., Sahu, V.K., 2005. Constraints on rupture of the 26 December 2004 Sumatra earthquake from far field GPS observations. Earth and Planetary Science Letters 237, 673–679. Chlieh, M., Avouac, J.-P., Hjorleifsdottir, V., Song, T.A., Ji, Ch., Sieh, K., Sladen, A., Hebert, H., Prawirodirdjo, L., Bock, Y., Galetzka, J., 2007. Coseismic slip and afterslip of the great Mw 9.15 Sumatra–Andaman earthquake of 2004. Bulletin of the Seismological Society of America 97, S152–S173. Curray, J.R., 2005. Tectonics and history of the Andaman Sea region. Journal of Asian Earth Sciences 25, 187–232. Curray, J.R., Moore, D.G., Lawver, L., Emmel, E., Raith, R., Henry, M., Kieckhefer, R., 1979. Tectonics of the Andaman Sea and Burma. American Association of Petroleum Geologists Memoir 29, 189–198. Curray, J.R., Emmel, F.J., Moore, D.G., Raitt, R.W., 1982. In: Nairn, A.E.M., Stehli, F.G. (Eds.), Structure, Tectonics and Geological History of the Northeastern Indian Ocean. . The Ocean Basins and Margins, vol. 6. Plenum, New York N.Y., pp. 399–450. Dasgupta, S., Mukhopadhyay, M., 1993. Seismicity and plate deformation below the Andaman arc, northeastern Indian Ocean. Tectonophysics 225, 529–542. Dasgupta, S., Mukhopadhyay, M., 1997. Aseismicity of the Andaman subduction zone and recent volcanism. Journal Geological Society of India 49, 513–521. Dasgupta, S., Mukhopadhyay, M., Bhattacharya, A., Jana, T.K., 2003. The geometry of the Burmese–Andaman subducting lithosphere. Journal of Seismology 7, 155–174. Dasgupta, S., Mukhopadhyay, B., Bhattacharya, A., 2007. Seismicity pattern in north Sumatra–Great Nicobar region: in search for precursor for the 26 December 2004 earthquake. Journal of Earth System Sciences 116, 215–223. Earnest, A., Rajendran, C.P., Rajendran, K., Anu, R., Arun, G.M., Mohan, P.M., 2005. Nearfield observations on the co-seismic deformation associated with the 26 December 2004 Andaman–Sumatra earthquake. Current Science 89, 1237–1244. Eguchi, T., Uyeda, S., Maki, T., 1979. Seismotectonics and tectonic history of the Andaman Sea. Tectonophysics 57, 35–51. Engdahl, E.R., van der Hilst, R.D., Buland, R., 1998. Global teleseismic earthquake relocation with improved travel times and procedures for depth determination. Bulletin Seismological Society of America 88, 722–743. Furuse, N., Kono, Y., 2003. Slab residual gravity anomaly: gravity reduction due to subducting plates beneath the Japanese Islands. Journal of Geodynamics 36, 497–514. Gahalaut, V., Catherine, J.K., 2008. GPS measurements of coseismic and postseismic deformation in the Andaman–Nicobar–Sumatra region. International Association for Gondwana Research Conference series 5, 218–219. Gahalaut, V.K., Nagarajan, B., Catherine, J.K., Kumar, S., 2006. Constraints on 2004 Sumatra–Andaman earthquake rupture from GPS measurements in Andaman– Nicobar Islands. Earth and Planetary Science Letters 242, 365–374. Grevemeyer, I., Tiwari, V.M., 2006. Overriding plate controls spatial distribution of megathrust earthquakes in the Sunda-Andaman subduction zone. Earth and Planetary Science Letters 251, 199–208. Grow, J.A., 1973. Crustal and upper mantle structure of the central Aleutian arc. Bulletin Geological Society of America 84, 2169–2192. Gutenberg, G., Richter, C.F., 1954. Seismicity of the Earth and its Associated Phenomena. Princeton University Press, Princeton, NJ, p. 310. Guzman-Speziale, M., Ni, J.F., 1996. Seismicity and active tectonics of the Western Sunda Arc. In: An Yin, T., Mark Harrison (Eds.), The Tectonic Evolution of Asia, pp. 63–84. ISC, 2001. International Seismological Centre. International Seismological Centre, United Kingdom. On-line Bulletin, http://www.isc.ac.uk. Ishii, M., Shearer, P.M., Houston, H., Vidale, J.E., 2005. Extent, duration and speed of the 2004 Sumatra–Andaman earthquake imaged by Hi-Net array. Nature 435, 933–936. Jade, S., Ananda, M.B., Dileep Kumar, P., Banerjee, S., 2005. Co-seismic and postseismic displacements in Andaman and Nicobar Islands from GPS measurements. Current Science 88, 1980–1984. Kamesh Raju, K.A., Ramprasad, T., Rao, P.S., Rao, B.R., Varghese, J., 2004. New insights into the tectonic evolution of the Andaman basin, northeast Indian Ocean. Earth and Planetary Science Letters 221, 145–162. Kamesh Raju, K.A., Murty, G.P.S., Dileep Amarnath, Mohan Kumar, M.L., 2007. The west Andaman fault and its influence on the aftershock pattern of the recent megathrust earthquakes in the Andaman–Sumatra region. Geophysical Research Letters 34. doi:10.1029/2006GL028730. Kennet, B.L.N., Cummins, P.R., 2005. The relationship of seismic source and subduction zone structure for the 2004 December 26 Sumatra–Andaman earthquake. Earth and Planetary Science Letters 239, 1–8. Khan, P.K., Chakraborty, P.P., 2005. Two-phase opening of Andaman Sea: a new seismotectonic insight. Earth and Planetary Science Letters 229, 259–271. Kieckhefer, R.M., Shor Jr., G.G., Curray, J.R., 1980. Seismic refraction studies of the Sunda trench and fore arc basin. Journal of Geophysical Research 85, 863–889. Kieckhefer, R.M., Moore, G.F., Emmel, F.J.,1981. Crustal structure of the Sunda fore arc region west of central Sumatra from gravity data. Journal of Geophysical Research 86, 7003–7012. Lasitha, S., 2007. Geodynamics of the Andaman–Sumatra–Java Trench–Arc system based on gravity and seismotectonic study. Ph.D Thesis, Cochin University of Science and Technology, India. Lay, T., Kanamori, H., Ammon, C.J., Nettles, M., Ward, S.N., Aster, R.C., Beck, S.L., Bilek, S.L., Brudzinski, M.R., Butler, R., DeShon, H.R., Ekstrom, G., Satake, K., Sipkin, S., 2005.

261

The great Sumatra–Andaman Earthquake of 26 December 2004. Science 308, 1127–1133. Liu, C.S., Curray, J.R., McDonald, J.M., 1983. New constraints on the tectonic evolution of the Eastern Indian Ocean. Earth and Planetary Science Letters 65, 331–342. Marks, K.M., 1996. Resolution of the Scripps/NOAA marine gravity field from satellite altimetry. Geophysical Research Letters 23, 2069–2072. Martinez, F., Taylor, B., 2002. Mantle wedge control on back-arc crustal accretion. Nature 416, 417–420. Martinez, F., Taylor, B., Baker, E.T., Resing, J.A., Walker, S.L., 2006. Opposing trends in crustal thickness and spreading rate along the back-arc eastern Lau spreading center: implications for controls on ridge morphology, faulting and hydrothermal activity. Earth and Planetary Science Letters 245, 655–672. Miller, M.S., Lee, C.-T.A., 2008. Possible chemical modification of oceanic lithosphere by hotspot magmatism: seismic evidence from the junction of Ninetyeast Ridge and the Sumatra–Andaman arc. Earth and Planetary Science Letters 265, 386–395. Minear, J., Toksoz, M.N., 1970. Thermal regime of a downgoing slab and new global tectonics. Journal of Geophysical Research 75, 1379–1419. Moore, G.F., Curray, J.R., 1980. Structure of the Sunda trench lower slope off Sumatra from multi-channel seismic reflection data. Marine Geophysical Researches 4, 319–340. Mukhopadhyay, M., 1984. Seismotectonics of subduction and back-arc rifting under the Andaman Sea. Tectonophysics 108, 229–239. Mukhopadhyay, M., 1988. Gravity anomalies and deep structure of the Andaman arc. Marine Geophysical Researches 9, 197–210. Mukhopadhyay, M., Krishna, M.R., 1991. Gravity field and deep structure of the Bengal fan and its surrounding continental margins, northeast Indian Ocean. Tectonophysics 186, 365–386. Mukhopadhyay, M., Krishna, M.R., 1995. Gravity anomalies and deep structure of the Ninety East Ridge north of the equator, eastern Indian Ocean— a hot spot trace model. Marine Geophysical Researches 17, 201–216. Mukhopadhyay, B., Dasgupta, S., (in press). Swarms in Andaman Sea, India— a Seismotectonic Analysis. Acta Geophysica. Ortiz, M., Bilham, R., 2003. Source area and rupture parameters of the 31 December 1881 Mw = 7.9 Car Nicobar earthquake estimated from Tsunami recorded in the Bay of Bengal. Journal of Geophysical Research 108, 2215–2230. Pal, T., Chakraborty, P.P., Gupta, T.D., Singh, C.D., 2003. Geodynamic evolution of the outer arc-fore arc belt in the Andaman Islands, the central part of the Burma–Java subduction complex. Geological Magazine 140, 289–307. Parker, R.L., 1972. The rapid calculation of potential anomalies. Geophysical Journal International 31, 447–455. Peter, G., Weeks, L.A., Burns, R.E., 1966. A reconnaissance geophysical survey in the Andaman Sea and across the Andaman–Nicobar island arc. Journal of Geophysical Research 71, 495–509. Prince, R.A., Forsyth, D.W., 1988. Horizontal extent of anomalously thin crust near the Vema fracture zone from the three-dimensional analysis of gravity anomalies. Journal of Geophysical Research 93, 8051–8063. Purucker, M., Ishihara, T., 2005. Magnetic images of the Sumatran region crust. EOS American Geophysical Union 86 (10), 101–102. Radhakrishna, M., Sanu, T.D., 2002. Shallow seismicity, stress distribution and crustal deformation pattern in the Andaman–West Sunda arc and Andaman Sea, Northeastern Indian Ocean. Journal of Seismology 6, 25–41. Radhakrishna, M., Searle, R.C., 2006. Isostatic response of the Alula–Fartak and Owen transforms in the Eastern Gulf of Aden and the adjoining Arabian Sea. Geophysical Journal International 165, 62–72. Ravikumar, M., Rao, N.P., Chalam, S.V., 1996. A seismotectonic study of the Burma and Andaman arc regions using centroid moment tensor data. Tectonophysics 253, 155–165. Rodolfo, K.S., 1969. Bathymetry and marine geology of the Andaman basin and tectonic implications for southeast Asia. Geological Society of America Bulletin 80, 1203–1230. Rothe, J.P., 1969. The Seismicity of the Earth. UNESCO, Paris, p. 336. Roy, T.K., 1983. Geology and hydrocarbon prospects of Andaman–Nicobar basin. In: Bhandari, L.L. (Ed.), Petroliferous Basins of India. Petroleum Asia Journal, vol. 1, pp. 37–50. Roy, S.K., 1992. Accretionary prism in Andaman fore arc. Geological Survey of India, Special Publication 29, 273–278. Roy, A., Pal, T., Bhattacharya, A., Ghosh, B., 2008. Emplacement of lithotectonic units of accretionary prism and forearc setting in subduction complex— a geodynamic approach in Andaman Islands. International Association for Gondwana Research Conference series 5, 230–231. Sandwell, D.T., Smith, W.H.F., 1997. Marine gravity anomaly from GEOSAT and ERS1 satellite altimetry. Journal of Geophysical Research 102, 10039–10054. Shapiro, N.M., Ritzwoller, M.H., Engdahl, E.R., 2008. Structural context of the Great Sumatra–Andaman Islands earthquake. Geophysical Research Letters 35, L05301. doi:10.1029/2008GL033381. Sibuet, J.C., Rangin, C., Le Pichon, X., Singh, S., Cattaneo, A., Graindorge, D., Klingelhoefer, F., Lin, J.-Y., Malod, J., Maury, T., Schneider, J., Sultan, N., Umber, M., Yamaguchi, H., The Sumatra aftershock team, 2007. 26th December 2004 great Sumatra–Andaman earthquake: co-seismic and postseismic motions in northern Sumatra. Earth and Planetary Science Letters 263, 88–103. Sieh, K., Natawidjaja, D., 2000. Neotectonics of the Sumatran fault, Indonesia. Journal of Geophysical Research 105, 28295–28326. Singh, D.D., 1990. Q-structure beneath the north and central Indian Ocean from the inversion of the observed Love and Rayleigh wave attenuation data. Physics of Earth and Planetary Interior 59, 243–258.

262

M. Radhakrishna et al. / Tectonophysics 460 (2008) 248–262

Subrahmanyam, C., Gireesh, R., Gahalaut, V., 2005. Continental slope characteristics along the tsunami-affected areas of eastern offshore of India and Sri Lanka. Journal Geological Society of India 65, 778–780. Subrahmanyam, C., Gireesh, R., Shyam Chand., Kamesh Raju, K.A., Gopala Rao, D., 2007. Geophysical characteristics of the Ninety East Ridge–Andaman Island arc/trench convergent zone. Earth and Planetary Science Letters. doi:10.1016/j.epsl. 2007.10.016. Talwani, M., Ewing, M., 1960. Rapid computation of gravitational attraction of threedimensional bodies of arbitrary shape. Geophysics 25, 203–225. Utsu, T., 1971. Seismological evidence for anomalous structure of island arcs with special reference to the Japanese region. Review of Geophysics and Space Physics 9, 839–890.

Uyeda, S., Kanamori, H., 1979. Back-arc opening and the mode of subduction. Journal of Geophysical Research 84, 1049–1061. Weeks, L.A., Harbison, R.A., Peter, G., 1967. Island arc system in Andaman Sea. American Association of Petroleum Geologists Bulletin 51, 1803–1815. Wessel, P., Smith, W.H.F., 1995. New version of the Generic Mapping Tools released. EOS, Transactions, American Geophysical Union 76, 329. Yoshii, T., 1973. Upper mantle structure beneath the north Pacific and the marginal seas. Journal Physics of Earth 21, 313–328.