SEARCH

SEARCH BY CITATION

Keywords:

  • Makran;
  • accretionary prism;
  • décollement;
  • fault activity;
  • subduction zone

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[1] The Makran Subduction Zone has the highest incoming sediment thickness (up to 7.5 km) of any subduction zone. These sediments have formed a wide accretionary prism (∼400 km). Seismicity in the Makran is generally low; however the margin experienced an Mw 8.1 earthquake in 1945 which generated a significant regional tsunami. Seismic reflection data and swath bathymetry data from offshore Pakistan are used to analyze the structure and fault activity of the outer accretionary prism. The outer prism has a simple structure of seaward verging imbricate thrust faults, many continuous for over 100 km along strike. Fault activity is analyzed using basin stratigraphy and fault geometry, revealing a frontal continuously active zone, a central intermittently active zone, and a landward inactive zone. Over 75% of the faults in the seaward ∼70 km of the prism show evidence for recent activity. The décollement occurs within the lower sediment section, but steps onto the top-basement surface in regions of elevated basement topography. Fault spacing (6 km) and taper (4.5°) are comparable to other margins such as S. Hikurangi, Cascadia and Nankai, suggesting that high sediment input is not leading to an unusual prism structure. The décollement is unreflective, which is unexpected considering other prism characteristics predicting a weak surface, and may suggest a potentially stronger décollement than previously predicted. This study provides a significant advance in our understanding of the structure of an end-member convergent margin and demonstrates that systematic analyses of accretionary prism structure can help to elucidate subduction zone dynamics with ultimate relevance to seismogenic potential.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[2] The Makran Subduction Zone is generated by the subduction of the Arabian Plate beneath the Eurasian plate at ∼4 cm/yr offshore Pakistan and Iran (Figure 1). The accretionary prism is the largest accretionary complex in the world [Gutscher and Westbrook, 2009] and is thus an end-member globally, with an exceptionally high incoming sediment thickness (up to 7.5 km) and a low taper of ∼4° [Davis et al., 1983; Kopp et al., 2000]. As well as high sediment input, the Makran is characterized by generally low seismicity, however it did experience a Mw 8.1 earthquake in 1945 [Byrne et al., 1992]. The aim of this study is to interpret the morphology and structure of the outer Makran accretionary prism along a 400 km (along-strike) section offshore Pakistan, including an analysis of fault activity. This interpretation will allow the mechanical properties of the outer prism to be investigated, structural comparisons with other margins to be made, and will contribute to the understanding of its seismogenic potential. This manuscript is the first study of the Makran to systematically assess prism structure from seismic reflection data over such a large area. We evaluate the Makran Subduction Zone in the context of other global subduction zones (e.g., Southern Lesser Antilles, Hikurangi (New Zealand), parts of Cascadia and Sumatra/Sunda), including those with relatively high (>1 km) sediment input and/or large accretionary prisms. These accretionary prisms are also located at plate boundaries with slow to moderate convergence rates (∼3–5 cm/yr), similar to the Makran [DeMets et al., 1990].

image

Figure 1. Location map of the Makran Subduction Zone. Inset shows the regional tectonics, with the Makran located where the Arabian Plate subducts beneath the Eurasian Plate. Main map shows focal mechanisms from the Global CMT Catalog (1976–present) colored by depth. Black dots indicate earthquakes (>Mb 3.3) from ANSS catalog from 1963 to 2011. In addition the 1945 and 1947 earthquakes are indicated by stars - locations from Byrne et al. [1992] and Heidarzadeh et al. [2008]. Thick dashed lines indicate major tectonic boundaries. Fine gray lines indicate the 2D seismic lines used in this study. Higher seismicity is visible in the Ornach-Nal strike slip system than the Makran Subduction Zone.

Download figure to PowerPoint

[3] Fault activity within fold-thrust belts has been suggested to concentrate in the toe region, with activity reducing significantly landward of the deformation front. This suggestion has been supported by analogue [e.g., Davis et al., 1983; von Huene and Scholl, 1991; Gutscher et al., 1996] and numerical modeling [e.g., Willett et al., 1993; Fuller et al., 2006]. Other studies suggest that the distribution of deformation may be more complex, with a wider zone of activity and reactivation of landward thrusts [Hardy et al., 1998; Lohrmann et al., 2003]. Significant out-of-sequence thrusts, often associated with major structural and morphological boundaries, have been identified in margins such as Nankai, Japan and Sunda, Indonesia [Park et al., 2002; Kopp and Kukowski, 2003]. The Makran Subduction Zone, with its high sediment input and wide accretionary zone, provides an ideal region in which to examine prism structure and strain distribution.

[4] For submarine prisms, pore fluid pressure is an important controlling factor on thrust mechanics and the strength of prism materials [Moore and Vrolijk, 1992; Saffer and Bekins, 2006]. Fluids are sourced both from the subducting oceanic crust and the incoming sediment section, and from digenetic dehydration reactions with increasing temperature [Moore and Saffer, 2001]. As the subducting section is subjected to horizontal compression and vertical compaction, pore fluids attempt to migrate upwards. The presence of a thick sediment section with low vertical permeability may impede this fluid release, producing pore fluid pressures of up to lithostatic at the plate boundary [e.g., Moore and Vrolijk, 1992]. Overpressured pore fluid can influence décollement properties, leading to decreased effective basal strength which in turn, according to Coulomb wedge theory, will produce a prism of low taper [Davis et al., 1983; Le Pichon et al., 1993]. Elevated pore pressures have been inferred from seismic velocities at several subduction zones including Barbados and Nankai [e.g., Bangs et al., 1990; Tobin and Saffer, 2009]. Fluid may migrate along the décollement horizon, vertically through the accreted sediment, or along thrust faults depending on the relative permeability of these pathways [e.g., Moore and Vrolijk, 1992; Le Pichon et al., 1993]. The high sediment input experienced by the Makran suggests a potentially high fluid input, which may affect the mechanical properties of the prism.

[5] Historically, aseismic behavior was suggested to occur in the Makran, Hikurangi, Southern Barbados, and Cascadia subduction zones due to the presence of large accretionary prisms at these margins and relatively low background seismicity [Byrne et al., 1988]. Subsequent to this suggestion, evidence of prehistoric earthquakes in the Cascadia region suggested that this margin, rather than being aseismic, produces large earthquakes. The occurrence of the 1945 Makran earthquake, ∼30 km landward of the trench, has proved difficult to reconcile with other features of the Makran. The low taper, high sediment input and wide accretionary prism of the Makran would, according to traditional models [Byrne et al., 1988; Hyndman et al., 1997] suggest that the outer prism is largely aseismic. Prior to the Sumatra 2004 and Tohoku Oki 2011 events, seaward rupture was considered unlikely, however both these events show evidence for rupture farther trenchward than anticipated, and in the case of the Sumatra 2004 event, though the outer prism [e.g., Henstock et al., 2006; Dean et al., 2010; Gulick et al., 2011; Ide et al., 2011].

1.1. Geological Setting

[6] The offshore Makran accretionary prism is located in water depths of 750–3000 m. There is no bathymetric trench seaward of the deformation front (∼3000 m depth) due to the high sediment input (Figure 2). The age of the lithosphere in the Arabian Sea has been debated due to the lack of identifiable seafloor magnetic anomalies in the region between the Murray Ridge and the Makran subduction zone [White, 1979]. This absence has led to the suggestion that the oceanic crust in this region was formed either in the Cretaceous ‘quiet zone’, supported by heat flow measurements taken from shallow cores in the Gulf of Oman [Hutchison et al., 1981], or late Jurassic-early Cretaceous [Whitmarsh, 1979], or Paleocene-Eocene [Mountain and Prell, 1990; Edwards et al., 2000], both suggested from plate reconstructions.

image

Figure 2. Structure map of the Makran accretionary prism. Imbricate thrust faults picked from MCS data and correlated with the aid of bathymetry data. Bathymetry data in east modified from Bourget et al. [2010]. Submarine canyon-channel systems in blue. Water depth seaward of the deformation front ∼3500 m. Location of Figures S2 and S3 indicated.

Download figure to PowerPoint

[7] The Makran is bound to the west by the right-lateral Minab Fault and to the east by a triple junction consisting of the left-lateral Ornach-Nal fault system and the transtensional Murray Ridge (Figure 1). These strike-slip systems separate the Makran from regions of continental collision (Zagros and Himalayan). The Baluchistan Volcanic Arc, including the Bazman and Taftan Volcanoes in Iran, is located 400–600 km from the trench [Grando and McClay, 2007]. Onshore, two topographic depressions, the Jaz Murian and the Hamun-i-Mashkel, located between the accretionary prism and the Baluchistan volcanoes are interpreted as forearc basins [Jacob and Quittmeyer, 1979]. These basins are separated by the N-S trending strike-slip Sistan Suture Zone (Figure 1) [Tirrul et al., 1983] which has been suggested to segment the subduction zone into discrete eastern and western zones, based on contrasting levels of seismicity (higher in the east) [Rani et al., 2011]. This study is focused on the eastern offshore Makran subduction zone (east of the Sistan Suture), and on the outer ∼70 km of the accretionary prism.

[8] Mud volcanoes are common onshore, where they appear to be associated with E-W-trending fault zones [Ahmed, 1969] and sourced from the Upper Miocene Parkini mudstones [Schlüter et al., 2002]. Mud volcanoes and gas seeps have also been identified within the offshore prism [von Rad et al., 2000; Kukowski et al., 2001; Wiedicke et al., 2001; Ding et al., 2010].

1.2. Previous Studies

[9] The offshore Makran currently has limited coverage of published seismic reflection data. Early seismic reflection surveys identified the imbricate structure of the accretionary prism [White, 1982; Minshull and White, 1989]. Further data, including seismic reflection and refraction, magnetic, gravity, heat flow and swath bathymetry data sets (offshore both Pakistan and Iran) allowed the thick sediment section and velocity structure to be resolved in more detail but with limited along-strike coverage [Kopp et al., 2000; Kukowski et al., 2001; Grando and McClay, 2007; Ding et al., 2010]. Published core coverage is sparse, however analyses of 13 long piston cores from the trench and prism basins found high sedimentation rates (>2 mm/yr) dominated by interbedded hemipelagite and turbidites [Bourget et al., 2010; Mouchot et al., 2010; Bourget et al., 2011].

1.3. Seismicity

[10] The Makran is seismically relatively quiet compared to other margins such as Chile and Nankai, and more similar to quiescent margins such as Cascadia. However, the region has experienced occasional significant historical earthquakes. The largest recent earthquake in the Makran occurred in 1945 (Figure 1), recently relocated to 24.5°N 63.00°E [Heidarzadeh et al., 2008], placing it only 30 km landward of the deformation front. This was a Mw 8.1 earthquake with a possible shallow-dip thrust source (focal mechanism derived from body waveform inversions of P and SH waves and dislocation modeling of coastal uplift, [Byrne et al., 1992]), interpreted as an interplate event at a depth of approximately 25–30 km. This event generated a tsunami which killed 4000 people locally, with wave heights up to 10 m [Heidarzadeh et al., 2008]. A laterally extensive marine shelly bed in an intertidal lagoon in Oman has been interpreted as the 1945 tsunami deposit due to its shallow stratigraphic depth, although actual dating of the layer was unsuccessful [Donato et al., 2009]. The highest runup is reported to have affected the coast of Pakistan three hours after the earthquake. This time lag, in conjunction with the timing of undersea cable breaks, has been interpreted as evidence for widespread delayed slope failure triggered by the earthquake [Byrne et al., 1992]. A further probable thrust earthquake in 1947 (Mw 7.3) is interpreted as an aftershock to the 1945 event, located 80 km to the northeast [Byrne et al., 1992]. Smaller thrust earthquakes in the eastern Makran appear to also be N-S compressional events, potentially on the plate boundary [Byrne et al., 1992]. The two thrust events illustrated in Figure 1 occurred in 1991 and 1992 with Mw of 5.6 and 5.8 respectively, both at a depth of ∼15 km.

[11] Deeper (>60 km), normal faulting events occur further landward likely within the subducting plate. A swarm of strike-slip events occurred further landward from 1978 to 1980 in the region of the Sistan Suture Zone [Byrne et al., 1992]. These events produced right-lateral strike-slip focal mechanisms, indicating that activity is still occurring along the suture zone. This suggests continued relative motion between the onshore Eastern and Western Makran, consistent with a level of segmentation in the onshore Eurasian plate. Higher levels of background seismicity are recorded in the Murray Ridge and onshore strike-slip systems east and west of the Makran Subduction Zone, than in the subduction zone itself.

2. Data and Analysis

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[12] The multichannel seismic (MCS) reflection data set for this project consists of 6200 km of 2D data collected by WesternGeco® in 1998–1999 (Figure 1). These data were acquired using a 2250 cubic inch air gun array (4 strings) towed at a depth of 6.5 m, and recorded using a 5100 m streamer at 8 m depth with a shot spacing of 12.5 m. The MCS data are 102 fold, have been stacked and Kirchoff migrated and have undergone both pre- and post-stack deconvolution. The record length of these data is 10 s TWT. The data set includes 19 strike lines and 34 dip lines with a line spacing of 12 km. Additional constraints on the incoming section in the west were gained through integration of the Cam-30 MCS Line [Minshull and White, 1989]. Swath bathymetry data covering an area roughly 75 km by 150 km of the western study area (Figure 2) were collected on Leg 2 of Meteor Cruise 74 in 2007 using a Simrad EM120 12 kHz hull-mounted multibeam system, at a speed of between 2.5–12 kts, and subsequently gridded at 50 m [Ding et al., 2010] and made available to this project. Additional bathymetric information for the central and eastern areas has been acquired from published multibeam bathymetry data images [Bourget et al., 2010].

[13] MCS data have been interpreted using IESX (Schlumberger GeoFrame®). Horizons were picked using a combination of auto and manual picking, and faults manually picked and where possible correlated between lines. The stacking velocities associated with this data set are not used in this study; however published stacking velocity data from the Cam-30 line and coincident refraction lines can be incorporated allowing localized depth conversion and regional depth estimates. Kopp et al. [2000] recorded OBH data and reported sediment velocities of between 1.8 km/s at the seabed to 4.4 km/s just above the basement on the oceanic plate in the western study area. They note a probable weak velocity contrast between the basal sediments and the top of the oceanic basement landward of the deformation front. Interval velocities from stacking velocities of the Cam-30 reflection line suggest high sediment velocities, of up to 4.5 km/s, in the deepest accreted sediment, and up to 4 km/s in the incoming section [Minshull and White, 1989].

3. Results

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[14] We describe outer accretionary prism fault structure and activity, décollement position and properties, prism taper analysis and stratigraphy of the incoming section.

3.1. Accretionary Prism Structure

[15] Seafloor morphology is dominated by E-W trending ridges separated by narrow sedimentary (piggyback) basins and cut by canyon-channel systems (Figure 2). The frontal ridge (formed by the frontal thrust) is a significant bathymetric feature, with a vertical relief of up to 1000 m and local gradients of over 30°. The frontal and subsequent ridges are anticlinal fault propagation folds which express the imbricate thrust structure of the accreted sediments, and can be used to aid along-strike fault correlation. Where a bathymetric ridge is not evident, correlation is more difficult, but can still be achieved using surrounding features, e.g., fault position relative to the hanging wall ridges of adjacent faults, similarities in fault structure and extrapolation of fault trend. The current correlations are conservative; therefore the length of some faults, particularly for those further landward, is likely underestimated (due to increased slope sediment cover and disruption by submarine canyons). Increased slope sediment blankets the accreted sediment and reduces the bathymetric expression of the ridges, thus making their bathymetric trace harder to follow. The frontal thrust can be confidently followed continuously for 250 km (Figure 2). The deformation front is intersected by the Murray Ridge and Little Murray Ridge (north of and parallel to the Murray Ridge but with minimal bathymetric expression) [White, 1983] (Figure 2).

[16] Three reflection seismic lines (Figures 3, 4, and 6) are presented to represent the structure of the accretionary prism, accompanied by two additional seismic lines presented is Figures S2 and S3 in the auxiliary material. The imbricate thrusts are exclusively seaward vergent (landward dipping) with occasional secondary backthrusts (Figure 3). The fault propagation anticlines of the thrusts often contain small bending-moment normal faults in the hinge zone, driving collapse of the ridge slopes. The thrust faults dip at 20–30° and the majority strike E-W (080–090°). In the easternmost Makran there is a slight rotation in fault strike to the NE, following the trend of the deformation front as the Murray Ridge impinges. In the west and central regions of the study area a simple imbricate fault structure dominates, whereas fault structures become more complex in the east, with more antithetic backthrusts and secondary faults (Figure 4). Fault spacing fluctuates along the margin, varying from 2.2 km–18.6 km (Figure 5), however median fault spacing remains fairly consistent at ∼6 km, indicating that fault spacing increases are balanced by corresponding decreases. In the central region (overlying the obliquely subducting Little Murray Ridge) the second thrust steps inward relative to the frontal thrust (spacing >18 km) and the distance between the 2nd and 3rd thrusts is significantly reduced (2.2 km) (Figure 6). By the 3rd and 4th thrusts in this area, spacing has returned to the average value of ∼6 km. The Little Murray Ridge at this location is approximately 30 km wide (measured perpendicular to the margin) where it intersects the deformation front. There is no evidence of the proposed strike-slip Sonne Fault crosscutting the forearc, however the obliquity of this lower plate feature may preclude its clear imaging [Kukowski et al., 2000]. Though the subducting basement ridges produce topography on the downgoing plate there does not appear to be any large scale segmentation (breaks in structure along strike, significant changes in structural style etc.) of the upper plate imbricate thrusts in response to subducting topographic features. This may be due to the smoothing effect of the thick sediment section overlying these basement ridges.

image

Figure 3. Seismic line 116 from the west of the study area (location in Figure 2) (a) without interpretation and (b) with interpretation. Simple imbricate structure visible, with an average fault spacing of ∼6 km. Increased sediment blanketing can be seen toward the north. Incoming sediment thickness of 7.0 km. The two major sedimentary units (A and B) are labeled. Width of view ∼70 km. Vertical exaggeration ×2.3 at seafloor. Location in Figure 2.

Download figure to PowerPoint

image

Figure 4. Seismic line 160 from the east of the study area (a) without interpretation and ( b) with interpretation. Increased sediment input and channel structures visible. Incoming sediment thickness of 4.5 km. In this region the entire incoming sediment section is made up of Unit A. Width of view ∼70 km. Vertical exaggeration of ×2.5 at seafloor. Location in Figure 2.

Download figure to PowerPoint

image

Figure 5. Plot of average (red) and median (blue) fault spacing for each line, with range of values indicated by black error bars. The range of values increase over the Little Murray Ridge. Average fault spacing along the margin is 6 km (gray dashed line).

Download figure to PowerPoint

image

Figure 6. Seismic line 136 (over Little Murray Ridge (a) without interpretation and (b) with interpretation. Width of view is ∼70 km. Sediment thickness of 2.5 km. Imbricate faults show more variable spacing and sole out onto the top-basement surface. 100% of the incoming section is accreted. Vertical exaggeration ×2.3 at seafloor. Location in Figure 2.

Download figure to PowerPoint

3.2. Sediment Input and Stratigraphy of the Incoming Sediment Section

[17] The accretionary prism is cut by submarine canyon-channel systems transporting sediment to the trench (from the Pakistan coast). Trench-parallel channel systems are also visible in the bathymetry data in the eastern trench region flowing from east to west; seaward extensions of significant fluvial systems such as the Hab and Bela sourced in the Himalayan collision zone. Stratigraphy of the incoming oceanic Arabian Plate has been interpreted from the MCS data; however a shortage of local borehole data prevents detailed lithological interpretation [Clift et al., 2002]. The thickest input section (7.5 km) is in the west of the study area. In the central region, the elevated basement of the Little Murray Ridge [White, 1983] locally reduces sediment thickness (to as little as <1 km) and in the east the topography of the Murray Ridge flank reduces sediment thickness to approximately 5 km (Figure 7). The Murray Ridge may act as a barrier to northward sediment transport and its evolution exerts a significant control on the stratigraphy of the input section.

image

Figure 7. Grid of the depth to top-basement surface in time. The basement is dominated by NE/SW trending ridges: The Murray Ridge and Little Murray Ridge. The deepest basement occurs in the west at 3 s (∼7.5 km) beneath the seabed.

Download figure to PowerPoint

[18] The top basement surface has been picked as the first high amplitude, positive polarity reflector at the base of the sediment pile, and by the transition into chaotic reflections. The sediment input section can be divided into two units separated by a widespread N-dipping unconformity (Figure 8). The lower unit (Unit A) dips to the north, onlapping the basement, and can be correlated along the length of the deformation front. Reflectors are sub-parallel, closely spaced and laterally continuous. Reflector amplitude is variable with several high amplitude horizons visible, as well as some transparent sections. Channel-levee structures and normal faults are also visible in this unit, particularly in the east (Figure 9). Unit A has a fairly constant thickness of ∼4 km (assuming a velocity of 2.5 km/s from Fruehn et al. [1997], with localized thinning over the Little Murray Ridge. The upper unit (Unit B) contains laterally continuous horizontal reflectors and does not show any evidence for normal faulting. It forms a wedge-shaped, horizontally bedded unit, thinning to the south and east and thickening to the north and west, with a maximum thickness of 3.5 km (velocity model from Fruehn et al. [1997]), and is absent in the easternmost lines. The regional unconformity separating the two units may either be due to subduction-related flexure, uplift of the Murray Ridge, or a combination of the two. The geometry of the upper unit, determined from contoured isopachs, trends SW-NW and parallel to the Murray Ridge, supporting a Murray Ridge related origin [White, 1983].

image

Figure 8. Zoom of the incoming section from Line 136. Three units can be identified: oceanic basement, and two sedimentary units (A and B). Width of view 17 km. Vertical exaggeration ×2.3 at seafloor. Location in Figure 6b.

Download figure to PowerPoint

image

Figure 9. Buried channel structures in Unit A in the east of the study area (Line 160). Incising channels are also visible affecting the uppermost sediments. The entire incoming section is composed of Unit A in this region. Width of view 37 km. Vertical exaggeration ×2.3 at seafloor. Location in Figure 4.

Download figure to PowerPoint

[19] Previous studies have also identified two units in the incoming section. The lower unit in the west of our study area has been interpreted as a turbidite unit derived from Himalayan uplift, with alternating hemipelagic and turbidite layers deposited in a proto-Indus fan [Fruehn et al., 1997; Schlüter et al., 2002]. The upper unit has been previously described as a mixed suite of turbidites and hemipelagics derived both from the east and the onshore Makran to the north [Schlüter et al., 2002] with deposition occurring from the late Miocene to present. The occurrence of modern trench-parallel channels in the eastern Makran shows that there is clearly a component of sediment transport from the east, not solely from the onshore Makran to the north. There is also evidence for small channels flowing northward into the abyssal plain from the northern flank of the Murray Ridge (Figure 2).

3.3. Gas Hydrate Distribution

[20] A BSR (bottom simulating reflector) is visible across much of the prism indicating the presence of widespread gas hydrate (Figure 3a). In the west and central regions of the study area the BSR is fairly ubiquitous; however in the east its distribution is greatly reduced. The BSR occurs at an average depth (TWT) of 500 ms (∼500 m based on the velocity model of Minshull and White [1989]), consistent with the interpretations of Grevemeyer et al. [2000] from the west of the study area. The BSR does not appear to be disrupted where it crosses imbricate thrusts, as suggested by Minshull and White [1989], however further analysis of BSR properties and its relationship to faults and fluid flow is ongoing.

3.4. Décollement Position and Properties

[21] The décollement beneath the outer prism is in general not reflective, however, its location can be inferred from the geometry and downward projection of thrust faults (i.e., where they sole out), the identification of the undeformed subducting sediment, and from the thickness of the accreted hanging wall sediment section. The resulting picks are the result of extensive interpretation correlating between all dip and strike lines generating a consistent and correlated interpretation. From our careful analysis of the entire data set, although we specifically looked for such reflectors, we have not identified a local negative-polarity high-amplitude reflector which seems a more likely candidate for the décollement surface than other surrounding horizons. There is a level of uncertainty associated with locating the décollement surface within the sediment, as the boundary between accreted and underthrust sediment is, in most cases, not a visible sharp boundary. It is also possible that the décollement itself may not be a discrete horizon, rather a zone of sheared sediment. Estimating and integrating uncertainties from these methods, the interpreted décollement positions can be expected to have an error of +/−0.5 sTWT (∼600–700 m). The position of the décollement surface can be estimated to 25–40 km landward of the deformation front, beyond which the thick, deformed sediment section precludes the identification of deeper structure and stratigraphy. The proportion of accreted versus underthrust sediment can be established by extrapolating the likely décollement position into the incoming section and analyzing its variation along strike.

[22] The position of the developing décollement within the input section shows significant variations along strike (Figure 10). In the western survey area (Figure 3), the décollement surface dips north, parallel to the basement reflector, accreting 70–80% of the incoming section (∼4.3 km). In this area the décollement surface is located within the lower sedimentary unit (Unit A), while all of Unit B is accreted. This agrees with previous studies [Fruehn et al., 1997; Kopp et al., 2000; Grando and McClay, 2007], all of which interpreted a décollement within the lower sedimentary unit, and from the mass balance calculations of Platt et al. [1985]. To the east of the Little Murray Ridge (Figure 4) the décollement is again located within the Unit A and approximately 70–80% (∼3–4 km) of the section is accreted. Thus, away from topographic basement anomalies, the décollement consistently occurs ∼0.5–2 sTWT (∼1.5–3 km) above basement, within the lower sedimentary unit.

image

Figure 10. Décollement-basement isopach generated by gridding basement and décollement horizons and calculating thickness. Red indicates zero thickness between the two horizons, i.e., no underthrust sediment. Green/blue colors indicate underthrust sediment thickness in TWT. Maximum thickness of underthrust sediment is 3–4 km in the westernmost area (assuming a sediment velocity of 4 km/s from Kopp et al. [2000]). Locations of Figures 3, 4, 6, S2, and S3 shown.

Download figure to PowerPoint

[23] In the central study area, the incoming section is affected by the discontinuous basement highs of the Little Murray Ridge. This basement topography clearly influences the position of the décollement and the proportion of accreted sediment. Over the basement highs the décollement lies at the sediment-basement interface (Figure 6). The décollement does not appear to structurally step down along strike; rather it abuts and then follows the top-basement surface. In the landward prism it is possible that the décollement steps off the top basement surface and back into the sediment section, though this is difficult to interpret due to imaging limitations of this deeper section. In this region (∼150 km along strike section of the margin) 100% of the incoming section (1–3.5 km) is currently being accreted at the deformation front (Figure 10).

3.5. Prism Taper Analysis

[24] Surface slope values can be combined with the subducting plate dip to establish prism taper, and then used to elucidate information about the basal and internal frictional properties of the prism [Davis et al., 1983]. A low taper suggests low basal friction relative to internal strength [Dahlen, 1990].

[25] Slope and basement dip values were measured (using local depth conversions from velocity data for the latter) from the deformation front to 35 km into the prism (Table 1). The top-basement pick at the deformation front is clearly visible so has minimal error associated with it, however the landward picks are more difficult and so introduce an error of ±2.5%. There is additional uncertainty associated with the average velocities used for depth conversion which introduce an error of ∼10%. The frontal 35 km is the maximum prism width covered by all of the lines allowing a comparison between consistent portions of the prism on each line. For completeness, taper along the lines which covered larger widths of prism (up to 70 km) was also calculated and gives a similar but slightly lower average taper of 3.7°, compared to 4.5° for the frontal region. Basal dip beneath the outer Makran prism is shallow; with an average value of 3.2° (Table 1). Our basement dips in the western region of 1.3°–2.7° agree with results from refraction data of White and Loudon [1982]. Slope values are consistently low along strike, with a range of 0.7°–3°, and average of 2°. Excluding regions of anomalous oceanic basement topography, the average taper is 4.5°. This taper value is slightly higher than previous estimates of ∼3.6° [Davis et al., 1983], though this value was cited from White and Ross [1979], which focused on the western (Iranian) part of the accretionary prism. Our taper value however is similar to the ∼5° calculated from the slope values of Kukowski et al. [2001] and basement dip values of Kopp et al. [2000] Higher (>5°) taper values are found in the central study area where the Little Murray Ridge is locally subducting (Lines 142 and 144 have both the highest slope and basal dip values), and to a lesser extent in the east where the Murray Ridge flank increases basement dip. The taper values for these lines are therefore elevated due to basement topography.

Table 1. Slope and Basement Dip Values Calculated for the Frontal 35 km of Prism for Each 2D Seismic Linea
LineSlope (deg)Basement Dip (deg)Taper (deg)
  • a

    Italic values are excluded from regional average due to proximity to Little Murray Ridge (LMR) and resulting anomalously high (>5°) basement dips. Line numbers increase from west to east.

982.23.55.7
1000.93.84.7
1021.02.23.2
1041.01.82.8
1060.71.82.5
1082.42.34.7
1102.12.64.8
1121.71.63.3
1141.92.03.9
1161.72.44.2
1181.42.74.0
1202.02.34.3
1221.72.84.5
1241.53.24.8
1262.13.45.4
1282.05.67.7
1302.65.07.7
1322.83.66.4
1341.85.57.3
1362.64.06.6
1382.73.96.6
1402.62.85.4
1423.07.310.2
1442.86.69.4
1462.16.08.1
1481.62.23.8
1502.62.75.3
1521.51.32.7
1541.81.73.6
1561.81.43.3
1582.41.64.0
1602.32.75.0
1623.03.76.7
1642.03.55.6
Average2.03.25.2
Average excluding LMR values  4.5

3.6. Fault Activity Analysis

[26] Piggyback basins in the hanging wall of a thrust fault contain sediment which onlaps the deformed, accreted hanging wall sediments of each fault. These basins contain growth packages, the geometry of which is related to fault activity. The sediments within the Makran piggyback basins consist of turbidites and debrites from the failure of adjacent ridges, sediment transported by canyon channel systems and hemipelagic sedimentation [Bourget et al., 2010]. These basins are generally 1–7 km in width (across strike), and can extend for over 50 km along strike. Many piggyback basins contain >2 sTWT (∼2 km) of sediment.

[27] Unconformities present in the piggyback basins are the result of changes in the balance between sedimentation and tectonic activity, and may indicate changes in fault activity, changing fault position/geometry, or the initiation of a new thrust. Examples from the Kumano transect of the Nankai Subduction Zone illustrate that slope failure deposits may also create onlapping relationships and reflector terminations within piggyback basin sediments [Strasser et al., 2011], generating complex stratigraphy. Fanning, wedge shaped growth packages of sediment are likely to indicate periodic or continuous syn-sedimentary activity on the underlying thrust fault which progressively tilts the piggyback basin landward (Figure 11). Conversely horizontal recent sediments indicate minimal current fault activity. These classifications may be complicated by interactions between activity on adjacent faults, for example displacement on the frontal thrust may increase the dip of sediments in the piggyback basin of the second thrust. Piggyback basins have been identified on dip lines and the stratigraphy interpreted and correlated along strike (where possible) to classify faults into currently active versus inactive and continuously versus intermittently active during the depositional history of the basins.

image

Figure 11. Schematic diagrams of active and inactive faults showing interaction between overlying sediment and underlying structure.

Download figure to PowerPoint

[28] Applying these methodologies, over 75% of faults (from the deformation front to ∼70 km landward) show evidence for recent activity (e.g., Figure 12), including some over 50 km from the deformation front. Inactive faults (no evidence for recent activity) are almost exclusively located toward the north of the study area (e.g., Figure 13). Clear variations in fault activity distribution are also observed along strike (Figure 14). In the easternmost Makran prism fault activity is limited to within 30 km of the deformation front (generally up to the ∼4th thrust), and on the easternmost line only the frontal thrust shows evidence for recent activity. This contrasts with the western study area where up to 9 sequential prism faults appear to be currently active. These results show that in the west and central parts of the study area, accretionary prism strain is currently being distributed over a ∼50 km zone of the outer prism, rather than on a single, or small number of thrusts as in the east.

image

Figure 12. Example of an active fault. Piggyback basin shows sequential packages of fanning sediment indicating continuous activity. Location in Figure 3.

Download figure to PowerPoint

image

Figure 13. Example of an inactive fault. Thrust is buried beneath layers of undeformed sediment. Location in Figure 3.

Download figure to PowerPoint

image

Figure 14. Plot of fault activity along/across the accretionary prism. Red faults show evidence for recent activity, black faults do not. Dashed faults show intermittent activity and those with solid lines appear to experience more continuous activity. The prism is divided into three zones based on this analysis: 1) Continuously active since initiation (orange), 2) Active with periods of quiescence (purple), and 3) Inactive(green).

Download figure to PowerPoint

[29] The faults are also classified into whether they have experienced continuous or intermittent activity. The seaward, younger thrusts (within 30 km of the deformation front/ the 3rd–4th thrust) tend to demonstrate more continuous activity than the older, landward thrusts, where some appear to have had a period of quiescence (represented by undeformed, parallel strata) between active periods (Figure 14). This across strike variation in activity suggests that a given thrust fault is likely to be continuously active until ∼30 km from the deformation front, where periods of quiescence are interspersed with renewed activity. In cases where there is a variation in activity along short sections of a thrust, the most common behavior has been plotted. Variations in activity along strike of individual faults are generally related to lower activity at the lateral tips of the fault, where displacement may be reduced, while increased displacement/activity occurs in the center of the fault.

[30] There appears to be a correlation between fault length and level of activity (Figure 14). This is likely due to a combination of factors. The length over which a fault can be correlated increases with the bathymetric expression of the fault, due to increased recent activity and associated anticlinal growth. Faults of shorter length tend to have small or absent bathymetric expressions, related to their lack of recent activity and/or smaller displacements, and are therefore more difficult to correlate along strike creating the potentially artificial correlation between reduced fault length and inactivity.

4. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

4.1. Prism Structure and Influence of the Incoming Oceanic Plate Section

[31] From these analyses of the Makran margin, the outer prism appears to be relatively structurally homogeneous across and along strike at a regional scale (with the exception of local changes over the Little Murray Ridge). The structures seen across the margin are simple, with a consistent imbricate structure, which may be partly a function of the normal convergence [Kukowski et al., 2001]. There is no evidence for large scale segmentation of the upper plate, changes in fault vergence, or major out-of-sequence thrusts and, with the exception of the region overlying the Little Murray Ridge, fault spacing and taper are very consistent along the margin. The extensive frontal thrust (length ≥250 km) further illustrates the continuity of the prism structures. Comparisons between accretionary margins by McAdoo et al. [2004] highlight the unusual simplicity and along-strike continuity of the Makran structure when compared to other margins such as Cascadia and Nankai. There is however a gradual change in prism characteristics toward the eastern extent of the subduction zone: with increasing proximity to the Murray Ridge and the Eurasian/Arabian/ Indian triple junction, the deformation front begins to rotate to the north and strain becomes more focused in the outermost prism. The subduction of the Little Murray Ridge also causes local changes in fault spacing and reduces the incoming sediment thickness to less than 1 km.

[32] To investigate how the thick accreted sediment section of the Makran affects two of its structural parameters (taper and fault spacing), the results from this study have been compared with those from other margins with significant accretionary prisms: Lesser Antilles, North and South Hikurangi, Cascadia and Nankai (Table 2). What is evident from this comparison is that, despite its high sediment input, the taper and fault spacing of the Makran are comparable to these other margins, suggesting that the high incoming sediment thickness of the Makran is not generating unusual structural behavior in the offshore prism and is not the main control on structural development. The Makran, with an accreted sediment thickness of ∼6 km, has similar fault spacing to N Cascadia (Washington) where the accreted section is only 2–3 km. However, there does appear to be a potential lower cut-off of <1 km accreted sediment, below which the fault spacing drops to low values (3–4 km) such as in Nankai (Muroto) and possibly North Hikurangi (Table 2). An alternative interpretation for fault spacing in the Makran is proposed by Ding et al. [2010], who suggest that the primary fault spacing is 12 km, and that faults which form within this across-strike span are secondary, out-of-sequence thrust faults. We find no categorical evidence from our data set for such regular out-of-sequence thrusting, or any clear distinction between ‘primary’ and ‘secondary’ thrusts, however uncertainties in the exact sequence of faulting should be noted as a possible source of error in the fault spacing measurements as a function of their spacing at initiation.

Table 2. Comparison of Structural Parameters Calculated for the Makran From This Study, With Published Data From Selected Other Marginsa
 Max Accreted Sediment Thickness at Trench (km)Taper Values (°)Average Fault Spacing (km)
Surface Slope (α)Basement Dip (β)Taper (α + β)
  • a

    Where fault spacing is ambiguous, spacing between thrust anticlines has been used.

Makran (this study)62 (avg.)1.5–2.5 (avg.)3.5–4.56
Southern-Central Lesser Antilles (13.5°N) [Bangs et al., 1990]52138
Hikurangi (South-Central) [Barnes et al., 2010]2–413.045–6
North Cascadia (Washington) [Adam et al., 2004]221.83.86
Hikurangi (North) [Barker et al., 2009]0–13811∼3 (frontal 3 thrusts)
Nankai (Muroto) [Gulick et al., 2004]0.71.51.63.14 [Ikari and Saffer, 2011]

[33] The average prism taper of the Makran, without the effects of the Little Murray Ridge is 4.5°. This classes it as a low taper margin among global subduction zones, comparable to Southern-Central Lesser Antilles, Southern Hikurangi and Cascadia-Washington [Lallemand et al., 1994]. According to Coulomb tapered wedge theory, the low taper of the Makran is likely to indicate either a weak basal surface, or high internal strength [Davis et al., 1983]. However, analogue models have suggested that accretionary prisms which experience very high sediment input may have mechanical differences to classic Coulomb wedges and may be prevented from reaching the critical taper their basal friction would predict, forming sub-critical wedges [Storti and McClay, 1995]. Therefore the Makran may represent an example of a sub-critical wedge which is actively deforming in response to high levels of frontal accretion. This mechanism has also been suggested for the Southern Hikurangi margin, which experiences a similarly low taper (4°) and high rates of frontal accretion [Barnes et al., 2010].

4.2. Fault Activity Analysis

[34] Fault activity analysis indicates that much of the outer Makran accretionary prism is currently deforming, with strain distributed over a ∼50 km wide region of the prism, and up to 9 imbricate thrusts. Our analysis suggests that a given thrust in this environment will be continuously active until approximately 30 km from the deformation front (3rd–4th thrust), when it will begin to experience periods of inactivity, and then will likely become fully inactive ∼50–60 km from the deformation front. It is not possible here to establish the activity of the onshore portion of the prism, though it has been suggested that active folding and faulting also occurs in this region [Farhoudi and Karig, 1977; Platt et al., 1988]. Our analysis divides the outer prism into three zones across strike, a seaward continuously active zone, a central intermittently active zone and a landward predominantly inactive zone (Figure 14). Detailed studies of fault activity in active accretionary prisms are relatively rare; however published analysis of the Nankai Muroto 3D volume indicates that the majority of prism thrusts retain some level of activity seaward of the out-of-sequence thrust (∼35 km from the deformation front), with increased recent relief on the frontal 5–6 thrusts [Gulick et al., 2004]. As a further example, in the onshore imbricate fault system of Taiwan, GPS data indicate that maximum deformation is concentrated in a frontal zone ∼15 km wide, but that moderate deformation continues across the prism [Chang et al., 2003].

[35] The thick input section of the Makran, the majority of which is accreted, and the moderate convergence rate (∼4 cm/yr) imply that frontal accretion rates will be high. This may require enhanced deformation across a wider region of the prism relative to other margins. Diffuse fault activity may also be a result of low basal friction: analogue models [Koyi et al., 2000] suggest that fold-thrust belts which form over a low basal friction décollement are more prone to sustained activity over a broad region than those which form over a high basal friction surface. It is also important to consider the scale of the Makran subduction zone. At ∼400 km wide, it is the widest accretionary prism on earth, and therefore may be expected to have a wider active prism toe than narrower prisms if the active prism occupies a certain fraction of its total width. The division of the accretionary prism into three discrete zones of activity is similar to what is produced by analogue models [e.g., Hardy et al., 1998; Lohrmann et al., 2003]. The intermittent activity of the central zone thrusts may be in response to changes in accretion or sedimentation rate, for example an increase in the rate of frontal accretion may trigger increased deformation as the prism deforms to reach/maintain its critical taper [Lohrmann et al., 2003]. The reduction in fault activity in the east of the study area may be due to increased sediment blanketing (as observed from our MCS data), which has been suggested by analogue models to inhibit fault reactivation through increasing the normal stress on the fault [Storti and McClay, 1995; Hardy et al., 1998; Bigi et al., 2010]. The proximity of the Murray Ridge triple junction and onshore strike-slip systems may also act to reduce fault activity in the east.

4.3. Response of the Décollement to Subducting Basement Topography

[36] The position of the décollement affects many aspects of subduction zone dynamics [Moore, 1989; von Huene and Scholl, 1991; Saffer and Bekins, 2006]. The proportion of sediment subducted is important for global sediment budgets while the stratigraphic position of the décollement will affect its shear strength, and hence its development and seismogenic behavior. For the majority of the E. Makran Subduction Zone, the décollement is located within the sediment section, with along-strike variations in the thickness of underthrust sediment as shown in Figure 10. In the central section where the Little Murray Ridge intersects with the deformation front, the décollement forms at the basement-sediment interface, accreting the entire (albeit thinner) sediment section. As a result, 100% of the incoming sediment is accreted in this region. The same phenomenon has been observed on the Central Hikurangi margin where the décollement position changes from within the sediment to the basement-sediment interface where a seamount is being subducted [Barnes et al., 2010; Bell et al., 2010].

4.4. Implications for Décollement Properties and Prism Mechanics

[37] The Makran can be characterized as a fluid and sediment rich margin, supported by its low prism taper (4.5°), high sediment input (up to 7 km), and evidence for fluids. The identification of a BSR, fluid seeps, bubble plumes and regions of high backscatter in side scan sonar data (likely seafloor carbonate crusts at cold seep sites) across the prism suggest that although dewatering must be significant due to compaction in the input section and accretion in the outermost prism, significant fluid and fluid flux persist throughout the prism [Spiess et al., 2007; Ding et al., 2010]. These fluids also form mud volcanoes both offshore and onshore, suggesting a degree of overpressure [Schlüter et al., 2002]. Velocity-porosity conversions along the Cam-30 Profile [Fruehn et al., 1997] showed no significant evidence for overpressure in the incoming section, but increased porosity and slight inferred overpressure at depth landward of the deformation front. Similar conclusions were reached by Fowler et al. [1985] who noted from sonobuoy data that a high degree of tectonic consolidation occurs at the prism toe, where the incoming section is initially deformed. The normally compacted nature of the incoming sediment section [Fruehn et al., 1997] may be due to the presence of permeable layers in the sediment which facilitate vertical fluid loss and compaction, enhanced by the presence of widespread normal faults. It is important to note that overpressure measurements from velocity models are for the bulk sediment volume, and will therefore not resolve localized overpressure variations at a discrete horizon (e.g., the décollement).

[38] The potential effects of a thick sediment section, such as in the Makran Subduction Zone, on pore pressures are debated. On the one hand, thicker sediment is likely to suggest a higher sedimentation rate, which could increase pore pressures if sedimentation rates prevent sufficient fluid expulsion, alongside reduced vertical permeability due to increased drainage length [Saffer and Bekins, 2006]. Conversely high sedimentation may indicate a higher proportion of coarser sediment, with likely increases in overall permeability (although not necessarily vertically) and hence reduced pore pressure. It is likely that reduced vertical permeability and insufficient fluid expulsion dominates, as margins with thick incoming sections (e.g., Southern Lesser Antilles, South Hikurangi) tend to be characterized by low taper angles indicative of high basal fluid pressures [Dahlen et al., 1984].

[39] Low taper, fluid-rich subduction zone forearcs are often inferred to have weak basal surfaces, commonly observed as high-amplitude, negative polarity décollement reflectors in seismic data suggesting reduced bulk density, increased porosity and likely increased pore fluid pressure [e.g., Moore et al., 1995; Zhao et al., 2000; Bangs et al., 2004]. Prism fault geometry analysis of Kukowski et al. [2001] suggest that the décollement in the Makran is >3 times weaker than the overlying prism sediments, however they also note that the slight observed overpressure [Fruehn et al., 1997] may only account for ∼30% of the inferred décollement weakness, with the remainder a function of lithology or fracturing. The observed overpressure may also be distributed through the lower sediment section, rather than being concentrated at a particular interface (such as the décollement).

[40] The décollement in the E Makran appears to be unreflective in seismic reflection data across the study area, which is unexpected considering other prism characteristics which may suggest a weak décollement. Kopp et al. [2000] point out a high amplitude horizon deep in the input section in the west of the study area on one seismic profile but this is a localized feature and there is no clear indication that this represents the pre-décollement horizon. High amplitude, negative polarity reflectors may either be indicative of a discrete décollement horizon with specific properties, or be due to an impedance contrast between the accreted and underthrust sediment, and ordinarily these two cannot be distinguished. On the Muroto Nankai transect fluid overpressure and associated reduction in effective stress in the underthrust section were observed from drilling and direct sampling up to 20 km landward of the trench, in a region of high décollement reflectivity in seismic data [Tobin and Saffer, 2009].

[41] The high sediment velocities at the base of the Makran sedimentary section [Fruehn et al., 1997; Kopp et al., 2000] may imply that there is an insufficient velocity contrast between the accreted and underthrust sediment to form a coherent reflector. The Makran has the thickest accreted sediment section of any accretionary margin, and it is therefore difficult to compare with other localities in this respect, however, seismic reflection data from the Southern Lesser Antilles accretionary prism (where the incoming section is ∼4 km thick) also show little velocity contrast between accreted and underthrust sediment, though the reflectivity of the décollement here is unclear [Bangs et al., 1990]. However, data from the Sumatran margin show similarly high (>4 km/s) velocities to the Makran in the lower sediment section, and do resolve a high amplitude, negative polarity basal surface [Dean et al., 2010].

[42] The Makran therefore represents an intriguing contradiction between evidence for overpressure, and a weak basal surface, with an unreflective décollement. These observations might be explained by a weak décollement which is either fluid rich, but not producing a reflector in the MCS data, or not fluid rich, but weak due to other sediment properties which also do not produce a significant impedance contrast. However, these explanations do not truly provide an adequate explanation of the seismic properties observed when compared with other margins and data sets. These results may instead be explained by a sub-critical taper which therefore may not reflect basal strength, and distributed overpressure through the thick accreted sediment section, rather than at a discrete horizon. In this scenario the décollement in the Makran may be stronger than expected.

4.5. Implications for Seismogenic Potential

[43] The high velocities (<4.4 km/s) observed in the deepest sediments of the Makran accretionary prism may indicate that the sediments at décollement level are sufficiently consolidated to support seismic rupture. These sediment velocities are comparable to those seen at depth in the southern December 2004 Sumatra rupture area [Dean et al., 2010] where shallow/seaward rupture beneath the prism has been suggested. As discussed above, the basal surface itself may also be stronger than expected, therefore increasing the potential for seismogenic rupture beneath the outer forearc, with implications for tsunamigenesis. Our MCS data show that the structures seen in this portion of the Makran are fairly homogenous along strike and show little evidence for large scale segmentation (although the basement topography of the subducting Little Murray Ridge may affect the outermost prism). Integrating all evidence, it therefore seems reasonable to suggest that the low level of seismicity currently seen in the Makran could be a reflection of a long recurrence interval of plate interface earthquakes, and short historical record, rather than simply due to aseismic subduction, and that there may be a significant seismogenic potential. The occurrence of the 1945 and 1947 earthquakes, and reports of previous historical events [Byrne et al., 1992] emphasizes this point. This discussion illustrates how an increased knowledge of the Makran though a systematic, spatially extensive investigation of prism structure, can help to elucidate information regarding seismogenic potential.

5. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[44] 1. The structure of the Eastern Makran Subduction Zone, with the exception of the region immediately overlying the Little Murray Ridge and other local variations, is consistent at a regional scale, with a relatively simple seaward verging, evenly spaced imbricate thrust structure. Despite its high sediment input, the taper and fault spacing of the Makran are comparable to margins with significantly thinner incoming and accreted sediment sections. The high accreted sediment thickness does not appear to affect fault spacing, suggesting that there may be an upper limit on fault spacing in imbricate thrust systems.

[45] 2. Fault activity is distributed across up to 9 imbricate thrusts (over a 50 km region) in the outer prism. The outer prism can be divided into three zones according to fault activity: frontal continuous activity, central intermittent activity and landward inactivity. These zones narrow to the east with increasing proximity to the Murray Ridge and triple junction, with only a single active thrust on the easternmost survey lines. These results suggest that deformation in accretionary prisms may be more widely distributed than previously suggested, especially at low taper/high sediment input margins.

[46] 3. The average taper of the Makran accretionary prism is low (∼4.5°). This taper is likely to be a reflection of low basal strength and overpressured deep sediments; however the Makran may also have developed as a sub-critical wedge, where the prism is actively deforming but unable to reach its critical taper due to high rates of frontal accretion and sediment input.

[47] 4. The décollement is located within the lower sedimentary unit (Unit A) except in the vicinity of the Little Murray Ridge where it steps onto the top-basement horizon, accreting 100% of the incoming sediment section. The décollement is neither high amplitude nor negative polarity, which is surprising as reflective décollements have been observed in the seaward parts of several other accretionary margins, particularly those with low taper and high sediment input, where they have been inferred to represent a weak, overpressured basal surface. Other evidence from the Makran (porosity –depth trends, mud volcanoes, low taper etc.) also supports a degree of overpressure and potentially weak basal surface; however these observations may also be explained in the context of a stronger décollement. If the Makran décollement is stronger than expected, then this may go some way to explaining the presence of historical seismicity in the offshore prism, and have important implications for seismogenic potential.

[48] 5. This study is the first of its kind on the Makran margin to examine accretionary prism deformation over a large along-strike (∼400 km) section and represents one of only a small number of systematic studies of active prism structure over a large proportion of a single subduction margin. Our results indicate how prism structure responds to subducting basement features and sediment input, describes distribution of deformation, and may be important for the assessment of seismogenic potential.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

[49] The authors would like to thank WesternGeco for providing the seismic reflection data. We would also like to thank the crew and scientists of Meteor Cruise 74 (Leg 2) for the collection of the bathymetry data and allowing access to the data, and Tim Minshull for helpful discussion. We thank Oliver Ralph for a useful preliminary analysis of the data. We thank the Associate Editor, and reviewers Sean Gulick and Nini Kukowski for their helpful comments on the manuscript.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Data and Analysis
  5. 3. Results
  6. 4. Discussion
  7. 5. Conclusions
  8. Acknowledgments
  9. References
  10. Supporting Information

Auxiliary material for this article contains three figures and one text file showing two additional reflection seismic lines for the Makran accretionary prism collected by Western Geco in 1998–1999.

Auxiliary material files may require downloading to a local drive depending on platform, browser, configuration, and size. To open auxiliary materials in a browser, click on the label. To download, Right-click and select “Save Target As…” (PC) or CTRL-click and select “Download Link to Disk” (Mac).

Additional file information is provided in the readme.txt.

FilenameFormatSizeDescription
jgrb17250-sup-0001-readme.txtplain text document1KReadme.txt
jgrb17250-sup-0002-fs01.tifTIFF image12908KFigure S1. The line locations.
jgrb17250-sup-0003-fs02.tifTIFF image20088KFigure S2. A 2D seismic reflection line located in the west of the study area.
jgrb17250-sup-0004-fs03.tifTIFF image21318KFigure S3. A 2D seismic reflection line located in the east of the study area.
jgrb17250-sup-0005-txts01.txtplain text document2KText S1. Description of Figures S1–S3 and their interpretation.
jgrb17250-sup-0006-t01.txtplain text document1KTab-delimited Table 1.
jgrb17250-sup-0007-t02.txtplain text document1KTab-delimited Table 2.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.