Borneo

ARTICLE IN PRESS

Marine and Petroleum Geology 25 (2008) 606–624 www.elsevier.com/locate/marpetgeo

Seismic images of a collision zone offshore NW Sabah/Borneo Dieter Frankea,, Udo Barckhausena, Ingo Heydea, Mark Tingayb, Nordin Ramlic a

Federal Institute for Geosciences and Natural Resources (BGR), Stilleweg 2, 30655 Hannover, Germany School of Earth & Environmental Sciences, University of Adelaide, Geology and Geophysics, SA 5005, Australia c Petroliam Nasional Berhad (PETRONAS) Kuala Lumpur City Centre, 50088 Kuala Lumpur, Malaysia

b

Received 17 September 2004; received in revised form 8 November 2007; accepted 20 November 2007

Abstract Multichannel reflection seismic data from the southern South China Sea, refraction and gravity modelling were used to investigate the compressional sedimentary structures of the collision-prone continental margin off NW Borneo. An elongated imbricate deepwater fan, the toe Thrust Zone bounds the Northwest Borneo Trough to the southeast. The faults separating the individual imbricates cut through post-Early Miocene sediments and curve down to a carbonate platform at the top of the subsiding continental Dangerous Grounds platform that forms the major detachment surface. The age of deformation migrates outward toward the front of the wedge. We propose crustal shortening mechanisms as the main reason for the formation of the imbricate fan. At the location of the in the past defined Lower Tertiary Thrust Sheet tectonostratigraphic province a high velocity body was found but with a much smaller extend than the previously defined structure. The high velocity structure may be interpreted either as carbonates that limit the transfer of seismic energy into the sedimentary layers beneath or as Paleogene Crocker sediments dissected by remnants of a proto-South China Sea oceanic crust that were overthrust onto a southward migrating attenuated continental block of the Dangerous Grounds during plate convergence. r 2007 Elsevier Ltd. All rights reserved. Keywords: Northwest Borneo margin; Sedimentary succession; Seismic data

1. Introduction The NW Borneo continental margin lies within a broad zone of lithospheric deformation at the boundary between Borneo to the south, the Sulu Sea and Celebes Sea regions to the east, and the South China Sea to the northwest (Fig. 1). A striking onshore structural element of this wide and highly complex deformation belt is the CrockerRajang mountain belt which extends along the central part of Borneo, from Sabah to central-south Sarawak (Fig. 1; Hamilton, 1979; Benard et al., 1990; Hutchison et al., 2000). It is generally assumed that this mountain belt formed as an accretionary complex during south-directed subduction of a proto-South China Sea (Hamilton, 1979; Hall, 1996; Hutchinson, 1996; Meng, 1999; Pubellier et al., 2003). The closure of the proto-South China Sea began at 44 Ma (e.g. Hall, 1996, 2002). Collision of the protoSouth China Sea crust is believed to have first occurred in Corresponding author. Tel.: +49 511 643 3235; fax: +49 511 643 3663.

E-mail address: [email protected] (D. Franke). 0264-8172/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpetgeo.2007.11.004

the SW (Luconia Shoals) during the Late Eocene and commenced progressively later towards the NE until the Early Miocene (Hutchinson, 1996). A large deepwater foreland fold and thrust belt, 100 km wide is present adjacent to the NW Borneo Trough (Figs. 1 and 2). The origin of the 5–8 km thick imbricate fan in this Thrust Zone remains uncertain. Hinz et al. (1989) and Ingram et al. (2004) attribute the development of the Thrust Zone to crustal shortening while Tan and Lamy (1990), Hazebroek and Tan (1993) and Hutchinson (2004) interpret the thrusts as gravitational induced compressive deformation at the toe of the Tertiary delta system from NW Borneo. An important argument favouring the latter interpretation is that oceanic spreading in the South China Sea basin, which is commonly interpreted to be coincident with the plate convergence at the NW Borneo continental margin, ceased in the Early Miocene or early Middle Miocene (Barckhausen and Roeser, 2004; Taylor and Hayes, 1980; Briais et al., 1993). However, the crustal shortening hypothesis is strengthened by the observation of Miocene and Pliocene (post oceanic spreading) fault reactivation and inversion

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

throughout the shelf NW Sabah and North GPS-derived 4 cm (?) (Ingram et al., 2004)

and onshore regions of Brunei, Sarawak (Morley et al., 2003), the convergence rate in NW Borneo and presently NW–SE orientated

Fig. 1. Nomenclature and plate tectonic framework of SE Asia. The study area is indicated with a white rectangle.

607

maximum horizontal stress orientation in NW Borneo (Tingay et al., 2005). The northeastern part of the Thrust Zone shows a more compressional tectonic style. Hinz et al. (1989) first proposed that this zone is dominated by two Major Thrust Sheets, with a width between 25 and 70 km, which are superimposed one on top of the other. An older thrust system, called Major Thrust Sheet System, overrides the younger Lower Thrust Sheet System. Although evidence for the suggested structure relies solely on a change in the seismic reflection pattern this interpretation is widely accepted (Tan and Lamy, 1990; Rice-Oxely, 1991). However, while Hinz et al. (1989) suggested movements related to plate convergence as origin it has in the following been interpreted as a nappe of allochthonous masses of Crocker or equivalent pre-Middle Miocene sedimentary rocks that resulted from gravity sliding when the Crocker-Rajang fold-thrust belt was uplifted (Fig. 1; Rice-Oxely, 1991; Hazebroek and Tan, 1993; Madon, 1999). In this paper, we investigate the origin and evolution of the deepwater NW Borneo Thrust Zone using gravity, reflection and refraction seismic data. New seismic data acquired with a 6 km long streamer clearly images the shape and characteristics of the steep faults separating the individual thrust wedges and the underlying de´collement. These data allow evaluation of possible mechanisms leading to the evolution of the toe Thrust Zone. Seismic

Fig. 2. Reflection seismic lines in the study area (BGR86 and BGR01surveys) and interpreted sketch. The location of the example seismic sections shown and discussed in the following is indicated. The top of the individual toe thrusts is shown as grey lines and the numbers correspond to those shown in the example seismic sections in Figs. 3–5. Note the difference in the structural style from the southwestern Toe Thrust Zone to the northeastern Compressed Fold Belt. The extension of a high velocity body as derived from gravity/refraction modelling is indicated. Its western end coincides with the previously defined ‘‘Major Thrust Sheet’’ (e.g. Hinz et al., 1989) but the width is much smaller than in earlier interpretations (see text). Adjacent to the east and only sparsely covered by data from this study the tectonostratigraphic provinces Outboard Belt and Inboard Belt are located (Rice-Oxely, 1991). These structurally complex areas are characterized by compressional folding and strike-slip faulting associated with mobile clay movement (Tan and Lamy, 1990; Rice-Oxely, 1991; Hazebroek and Tan, 1993).

ARTICLE IN PRESS 608

D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

refraction and gravity data are also well suited to test the hypothesis that two major thrust sheets are superimposed one on top of the other in the northeastern Thrust Zone (e.g. Hinz et al., 1989; Madon, 1999). 1.1. The NW Borneo Trough and the deepwater thrust belt The NW Borneo Trough has sometimes been interpreted as seabed expression of a presently inactive subduction trench developed during Paleocene subduction along the NW Sabah margin (Figs. 1 and 2; e.g. Hamilton, 1979; Tan and Lamy, 1990). However, the somewhat different orientation of the Borneo Trough with respect to the strike of the Crocker-Rajang fold-thrust belt and its abrupt termination against the Luconia Shoals in the SW are not easily reconciled with this interpretation (Fig. 1). Moreover, Hinz et al. (1989) suggested that the NW Borneo Trough is not underlain by a southward subducting oceanic crust but that the lower plate is comprised of continental crust similar in structural style to that of the Dangerous Grounds continental terrane. Under the assumption that the NW Borneo Trough is not a subduction trench, it may thus be either interpreted as a sediment-starved foreland trough (Hazebroek and Tan, 1993; Milsom et al., 1997) downfaulted by thrust front loading of the Dangerous Grounds continental terrane (Madon et al., 1999) or as an asymmetric sedimentary sag basin that developed by uniform-sense simple shearing on the proto-South China Sea continental margin (Schlu¨ter et al., 1996). The deepwater fold and thrust belt adjacent to the NW Borneo Trough comprises thick wedges of imbricated sediments (light grey in Fig. 2). Five to seven elongated imbricate thrust fans show a similar strike of 030–0451N (Hinz et al., 1989). In contrast, the margin northeast of about 61450 N, 1151E is dominated by narrow and steep thrusts and the sedimentary successions show a predominantly chaotic seismic facies. We follow earlier descriptions of the NE margin (Tan and Lamy, 1990; Rice-Oxely, 1991) and refer to the region as ‘Compressed Fold Belt’ (Fig. 2). Landwards of the complexly deformed imbricate thrust fan a huge thrust sheet was interpreted at the NE portion of the margin (termed ‘Major Thrust Sheet System’ by Hinz et al., 1989). An Early Tertiary thrust system was proposed to override a younger, Late Tertiary Thrust Sheet System. Earlier estimates placed the width of the Major Thrust Sheet at between 25 and 70 km. Our new data indicate a significant smaller width for this structure and the question is raised whether the interpretation of an allochthonous nappe is still valid.

Sheets in the northeastern area. Multichannel reflection seismic (MCS) data were acquired using a digital streamer with a length varying from 4486 to 5982 m (360/480 channels). During the cruise BGR01, 1649 km of reflection seismic lines (4 ms sample rate, 14 s record length) were acquired in addition to the existing 3129 km reflection seismic data (BGR86, Hinz et al., 1989). Two sets of high resolution lines (1 ms sample rate, 7 s record lengths, total 1246 km) with a common centre point were also acquired (Fig. 2). Although these lines were designed predominantly for gas-hydrate studies (amplitude versus offset (AVO) analysis under different emergence angles) that are beyond the topic of this paper, these data were also incorporated into the structural interpretations. Data processing of the BGR01 data included a spherical divergence correction, muting, filtering and an interactive velocity analysis every 3 km. Multiple attenuation was performed by applying both an inside mute and an F-K filtering technique. The seabottom multiple was effectively removed, but later multiples are still present, especially from dipping events. After trace equalisation, the data were migrated in the time–space domain by application of a numerical approximation to the Kirchhoff integral description of the wavefield using smoothed stacking velocities. All the data in this paper are displayed with a mute above the seabottom and an automatic gain control (AGC) with a length of 1000 ms. An additional 185 km long wide-angle seismic reflection/refraction profile with four ocean bottom hydrophone (OBH) stations was shot along multichannel (MCS) seismic line BGR01-07 (Fig. 2). Processing of the refraction seismic data included trace editing and minimum-phase bandpass filtering. A spiking deconvolution was applied with a design gate around the first break, showing a velocity in the range of 6 km/s. Gravity data were continuously acquired with a KSS31 seagravimeter system and instrumental drift was removed based on a tie of the measured gravity to the world gravity net IGSN71. Free-air gravity anomalies were calculated after Eo¨tvo¨s correction and subtraction of the normal gravity (IAG67). The analysis of crossover errors showed a mean accuracy of 1 mGal of the data. Magnetic measurements were carried out with a Marine Magnetics SeaSpy sensor towed at a distance of 300 m from the ship almost continuously during the cruise. A temporary magnetic base station near the city of Kota Kinabalu on the island of Borneo was used to correct the magnetic data for daily variations in addition to standard processing of the data with reduction of the International Geomagnetic Reference Field (IGRF). By this it means the crossover errors could be reduced to 10 nT.

2. Data acquisition and processing

3. Presentation and discussion of selected reflection seismic lines

A seismic survey was carried out in 2001 in the deepwater portions of the NW Borneo margin to evaluate the origin of the ‘Thrust Zone’ in the southwestern study area as well as the origin of the proposed two Major Thrust

Selected profiles are presented in Figs. 3–9 to demonstrate the principal features in our data set. The shape and structure of the ‘Thrust Zone’ (Fig. 2) will be discussed on the basis of lines BGR01-01 (Fig. 3) and BGR01-02 (Fig. 4)

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

609

Fig. 3. Line drawing interpretation of MCS line BGR01-01 showing the main tectonic features of the southern part of the study area. To the SE of the NW Borneo Trough five elongated thrust sheets are imaged. Landward two buried toe thrusts terminate against a major counter-regional growth fault. In the lower panel example seismic sections illustrate the shape of inferred Neogene sediments from a syncline in between toe thrusts 4 and 5 in comparison with those at the toe thrust no. 1. We find progressively younger sediments involved in the thrust wedges when approaching the NW Borneo Trough. The defined seismic marker horizons A, B, C, D and F (see Table 1) are indicated. For location see Fig. 2.

that traverse the southwestern part of the study area in more or less northwest–southeast direction, and the strikeperpendicular NNW–SSE trending line BGR01-10 (Fig. 5). Lines BGR01-06 (Fig. 7) and BGR01-07 (Fig. 9) run at an azimuth of 2821N across the northeastern part of the study area and illustrate the structural style of the Compressed Fold Belt and ‘Major Thrust Sheet’. Herein, we have used a seismic stratigraphy comprised of six marker horizons (A, B, C, D, E and F). This seismic stratigraphy is based on the previously interpreted seismic successions developed for the shelf to NW Borneo Trough sediments (Bol and van Hoorn, 1980; Levell, 1987; Hinz et al., 1989; Schlu¨ter et al., 1996). The timing of each of these marker horizons and their relationship to other studies is listed in Table 1. In the line drawing interpretations, reflections from the middle and lower crust are shown in addition to the marker horizons (Figs. 3, 5 and 7). The crust–mantle transition and undated intra-horizons in the Cenozoic sedimentary cover are presented wherever they are well defined. 3.1. The NW Borneo Trough—sediment sequences and evidence for young volcanism Offshore NW Sabah the Borneo Trough is about 50–65 km wide (Fig. 2). The trough contains well-bedded,

hemipelagig to pelagic sediments (Hazebroek and Tan, 1993) that have been only mildly deformed (Fig. 3, shotpoint (SP): 4200–3350; Fig. 4, SP: 1000–2050; Fig. 5, SP: 1700–2650; Fig. 7, SP: 4100–2750 and Fig. 9, SP: 1825–2850). The most prominent regional seismic unconformity is horizon D which is interpreted as the top of the Nido carbonate unit of inferred Oligocene to late Early Miocene age (compare Figs. 4 and 6; Kudrass et al., 1986; Hinz et al., 1989). This key stratigraphic sequence forms the base of the predominantly Neogene sediments filling the Borneo Trough. The interpreted limestone sequence (reflector D; Nido carbonates) can be followed from beneath the NW Borneo Trough to the Thrust Zone (Figs. 4, 6 and 8). The sediments in the trough are subdivided by horizons A, B and C in more or less conformable sequences that onlap in the west unconformably onto the interpreted carbonate platform of the Dangerous Grounds. Unconformity C (early Middle Miocene) forms the top of a lowfrequency band of parallel reflectors (Fig. 3; lower right) and is interpreted to separate the generally deep marine clastic sedimentation during a pre-early Middle Miocene phase from younger clastic shelf/slope deposits. Unconformity C is the deepwater equivalent of the ‘Deep Regional Unconformity’ (DRU), which is believed to mark a major

ARTICLE IN PRESS 610

D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

Fig. 4. Potential field data (top; dashed is gravity), example seismic section (middle) and interpretation (bottom) of line BGR01-02 (location see Fig. 2) showing the main tectonic features of the southwestern part of the study area. To the SE of the NW Borneo Trough five toe thrusts are imaged and indicated by numbers. The faults separating the individual thrusted bodies obviously terminate at the prominent low-frequency reflector D that was interpreted to represent the top of a unit of carbonates of inferred Oligocene to Early Miocene age. The extent of the example seismic section shown in Fig. 6 (bottom) is indicated.

break in the sedimentary history (Bol and van Hoorn, 1980; Levell, 1987). Unconformity B of Hinz et al. (1989) and Schlu¨ter et al. (1996) is interpreted as a time transgressive erosional horizon that forms the top of the accreted wedge, or what was the top of the individual thrust sheets at the time of deformation. The development of the deepwater fold and thrust belt is believed to have occurred from pre-late Early Miocene off N Palawan (Fig. 1) to late Middle Miocene off SW Palawan and unconformity B has been dated as 14–12 Ma old (Schlu¨ter et al., 1996). However, given the time transgressive nature of this horizon in the study area off NW Sabah, this unconformity might be as young as Late Miocene and thus roughly coincide with the ‘Shallow Regional Unconformity’ (SRU; Levell, 1987; Tan and

Lamy, 1990; Rice-Oxely, 1991). In our interpretation, unconformity B (Late Miocene) was formed by marine erosion because it partly onlaps unconformity C (Fig. 3, lower left). Horizon A represents a narrow band of high amplitude reflectors in a surrounding low-amplitude setting (compare Figs. 3, 4 and 6). Horizon A from this study coincides with Horizon II at the Early/Late Pliocene boundary of Levell (1987). In the interpretation of Hinz et al. (1989), this horizon represents the base of the upper Pliocene sediments. A 7 km wide bathymetric high is observed at the edge of the NW Borneo Trough along line BGR01-10 that is also associated with a high amplitude magnetic anomaly (Fig. 5, SP: 1200). The same structure is also traversed by line

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

611

Fig. 5. Depth converted seismic section (top) and line drawing interpretation (bottom) of MCS line BGR01-10. Along this line four elongated thrust sheets are imaged. Landward of the buried toe thrusts extensional faulting seems to predominate. The defined seismic marker horizons A, B, C, D and F (see text) are indicated. The upper panel shows a depth conversion, based on interval velocities derived from stacking velocities of the sedimentary successions above marker horizon D. Apparently the top of the interpreted carbonate platform (D) that forms the major detachment horizon dips landward. The inset in the upper panel illustrates the slope angles of the seafloor and the detachment horizon. For location of the line see Fig. 2. The extent of the example seismic section shown in Fig. 6 (top) is indicated.

BGR01-06 where it is associated with a high gravity and a strong negative magnetic anomaly (Fig. 7, SP: 4362–4391). The structure is covered by a thin sedimentary cover of inferred post-Late Pliocene strata but the absence of Miocene and Pliocene strata on top or at the edges indicates a young age for the structures formation. On line BGR01-08 another bathymetric high clearly reflects in gravity and magnetic data (but not shown herein). Less well expressed in the potential field data is a second, deeper buried bathymetric high on line BGR01-06 within the NW Borneo Trough (Fig. 7, SP: 2900–3300) and a similar, about 8 km wide high along line BGR01-02 (Fig. 4, SP: 1700). The sub-parallel bedded Miocene to Quaternary sediments onlap these structures (Fig. 7, SP: 2900–3300; Fig. 4, SP: 1700) indicating an earlier formation with respect to the topographic highs found along lines BGR01-10, -06 and -08. Kudrass et al. (1986) recovered porphyritic basalt from isolated seamounts at two locations off northern Palawan (Fig. 1). The position of the seamounts at the western end of the Borneo Trough is geologically comparable to the observed structures in this study. Kudrass et al. (1986) reported K–Ar ages of 2.7 and of 0.47 Ma for the basalt samples. From the high impedance contrast, resulting in high reflection amplitudes, the high amplitude potential field anomalies and the confirmed volcanism in a geologically comparable setting off Palawan we infer that the bathymetric highs found along lines BGR01-10, -06, 02 and 08 represent volcanoes. Assuming

the seismostratigraphic concept being right, the structures formed beginning in pre-Mid Miocene (onlap of sediments) to Pliocene/Quaternary time (no onlap, thin cover of postLate Pliocene sediments). Apart from the position of the inferred volcano in the NW Borneo Trough on line BGR01-02 the structures beneath the inferred carbonate platform (Horizon D) are well expressed and exhibit extensional deformation manifest by numerous normal faults and the formation of horst and graben structures (Fig. 4). The tilted fault blocks at the western ends of lines BGR01-01 (Fig. 3) and BGR01-02 (Fig. 4) probably represent Triassic to Cretaceous sediments, similar to those dredged from the Reed Bank and Dangerous Grounds (Kudrass et al., 1986). The regional seismic unconformity between Mesozoic and Tertiary rocks is labelled F in the seismic profiles (Figs. 3–5; Schlu¨ter et al., 1996). The end of extension and concomitant block rotation is marked by a distinct seismic pattern that represents the Nido carbonates (reflector D). The basement reflector (F) can be followed along lines BGR01-01, -02 and -10 from the Dangerous Grounds continental terrane beneath the Thrust Zone, where it is observed at two-way times (TWT) of 9 s (Figs. 3–5). 3.2. The Thrust Zone in the southwestern study area The Thrust Zone (Fig. 2) comprises up to five major individual listric thrust sheets, separated by cave-upwards

ARTICLE IN PRESS 612

D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

Fig. 6. Example seismic sections from lines BGR01-10 (top) and BGR01-02 (bottom) from the southwestern part of the study area illustrating the structural style of the toe Thrust Zone (see Fig. 2). The faults separating the individual imbricates ramp down to (or at least close to) the marker horizon D, indicating that the interpreted carbonate platform represents a major detachment surface. For location see Figs. 2, 4 and 5.

listric subsidiary faults (Figs. 3–6). The stratification within the individual toe thrusts is sub-parallel and the individual faults generally steepen overall towards the hinterland (Figs. 3–5). Synclines in between the individual thrust sheets are filled with slightly disturbed sediments of inferred Neogene–Quaternary age (Fig. 3, lower panel). Although the coverage by seismic lines is quite dense, we found no distinct indication for tear faults segmenting the individual listric thrust sheets. The geometry of an imbricate thrust fan alone does not indicate whether the development of a given imbricate system progressed toward the foreland or the hinterland (Twiss and Moore, 1992). However, stratigraphic information can provide additional evidence. We find progressively younger sediments involved in the thrust wedges in the foreland, closer to the Borneo Trough (Figs. 3 and 4; thrust wedges nos. 1 and 2) than in the hinterland (Figs. 3 and 4; thrust wedges nos. 3, 4 and 5). While the synclines in

between the individual thrust sheets are generally filled with only slightly disturbed sediments of inferred Neogene–Quaternary age, these sedimentary sequences are involved in the folding of the first thrust. Fig. 3 shows the data from one of the synclines in between two thrust wedges in comparison with the data from the Borneo Trough. While the strata in the centre of the syncline is only mildly deformed, the area of the eastern Borneo Trough was (and is possibly still) affected by tectonism. A similar style of deformation is visible on line BGR01-02 at approximately SP 2000 (Fig. 4). The listric faults in the imbricate system close to the thrust front (NW Borneo Trough) cut nearly to the seabed while the faults in the hinterland terminate upward within the stratigraphic section. This suggests that the thrust wedges in the hinterland were in their present shape prior to the deposition of the sediments in the synclines, and that younger thrust fans developed successively basinwards.

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

613

Fig. 7. Potential field data (top; dashed is gravity) and line drawing interpretation of MCS line BGR01-06 showing the main tectonic features of the northeastern part of the study area. The interpreted volcano west of the NW Borneo Trough has a clear signal in the magnetic and gravity data while the interpretation of the other one in the NW Borneo Trough is less clear (see text). The example seismic sections below illustrate the shape of the inferred volcanoes and the up-bending of sedimentary layer boundaries B and A. This indicates that the structures built up after the formation of horizon A of inferred Middle Pliocene age. Landward of the NW Borneo Trough only two toe thrusts are imaged and define the Compressed Thrust Zone (see Fig. 2). The chaotic seismic pattern east of the Compressed Thrust Zone was explained previously to be caused by allochthonous masses of Crocker, Chert-Spilite, or equivalent pre-Middle Miocene rocks that were overthrust onto a southward migrating NW Sabah Platform (Hinz et al., 1989) or resulted from gravity gliding due to the uplift of the Crocker-Rajang mountain belt (Tan and Lamy, 1990; Rice-Oxely, 1991; Hazebroek and Tan, 1993). The previously defined width of this feature was almost 70 km at the location of seismic line BGR01-06. The defined seismic marker horizons A, B, C, D and F (see text) are indicated. For location see Fig. 2.

The most prominent reflector, horizon D can be identified from the Borneo Trough to beneath the toe Thrust Zone for at least 50 km, where it is observed at TWT of between 7 and 8 s (Fig. 6). The thrust faults predominantly sole out at this horizon D and the subparallel stratification of the sediments incorporated in the individual toe thrusts can be followed from the surface to this level. These results confirm the interpretation of Hinz et al. (1989) that the carbonates (horizon D) form the major detachment surface for the movements that resulted in the formation of the Thrust Fold Belt. At least two buried ridges are present landward of the thrusts that are not obviously bounded by faults (Fig. 3, SP: 1800–2200; Fig. 4, SP: 3200–3600 and Fig. 5, SP: 3700–4100). These elongate structures are indicated as buried toe thrusts (Fig. 2). Hinz et al. (1989) describe these ridges as piercement of a buried compressional duplex

structure while Hazebroek and Tan (1993) interpret these structures to represent older anticlines associated with claydiapiric ridges. According to these authors, the thrustplanes of older thrusted anticlines steepened as the delta developed. An earlier interpretation of Bol and van Hoorn (1980) favoured compressional stress, possibly related to strike-slip faulting in the basement as cause for these ridges. We observe a general trend of steepening of the anticlines in the Thrust Zone from west to east (Figs. 3–6) and thus propose that the ridges represent thrusted anticlines that steepened as the compressional forces involved younger sediments, closer to the NW Borneo Trough. In addition, we suggest an influence of active growth of shale diapirs, which are commonly observed in association with the shale-prone Middle Miocene sequence (Sandal, 1996). The elongate geometry of the structures indicates that the growth of shale diapirs may have

ARTICLE IN PRESS 614

D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

Fig. 8. Example seismic sections from lines BGR01-07 (top) and BGR01-06 (bottom) from the northeastern part of the study area illustrating the Compressed Thrust Zone and the thrust sheet. The width of a high velocity body (thrust sheet; see text) along seismic line BGR01-07 could be estimated by refraction modelling to be less than 20 km. For location see Figs. 2, 7 and 9.

occurred along/or in association with pre-existing thrust faults. The eastern ends of the lines BGR01-01, -02 and -10 (right side in Figs. 3–5) are close to the position of a regional wrench fault (Morris Fault) interpreted by RiceOxely (1991) and Tan and Lamy (1990). The proposed fault marks the transition from the Inboard and Outboard Belts of the shelf region (Fig. 2). The data indicate a high structural complexity, possibly confirming the interpreted wrench mechanisms along this fault. However, we suggest the interpretation of counter-regional growth faults in the area covered by our seismic lines. On line BGR01-01 at SP 1650 (Fig. 3), line BGR01-02 at SP 3720 (Fig. 4), and line BGR01-10 at SP 4200 (Fig. 5), the positions of interpreted counter-regional growth faults, in parts possibly inverted by shale intrusions, are marked at the eastern end of the Thrust Zone. The complexity of the

deeper parts of the counter-regional growth faults may be explained by intrusions of overpressured mobile shales associated, in part with positive inversion of gravity structures, as suggested by Morley et al. (2003) in the Brunei region. Widely distributed bottom simulating reflectors (BSRs) are an indirect indicator for the presence of gas hydrates in the study area. The BSRs were identified unambiguously when they transect reflectors from strata and, in addition, show the characteristic polarity reversal with respect to the seafloor reflection. The BSRs have been observed with varying reflection amplitudes in water depths between 1100 and 2800 m in the post-Miocene sediments, mainly at the top of the imbricate thrusts. The BSR depths below seafloor vary between 250 and 350 m. Fig. 3 (bottom) shows examples of BSRs and interpreted BSRs are indicated in Figs. 4–6 and 8.

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

615

Fig. 9. Refraction seismic and gravity derived velocity–depth model (top) and reflection seismic line BGR01-07 (bottom). The numbers indicate P-wave velocities derived from traveltime modelling. Superimposed on the reflection seismic line are the refraction model-boundaries converted to two-way traveltime. We found two solutions explaining the data from the deep crust and Moho (models 1 and 2; see text). The mid-crustal horizon and the Moho are from refraction model 2. The respective boundaries for model 1 are dashed. For location see Fig. 2.

3.3. The Compressed Fold Belt in the mortheastern study area Lines BGR01-06 (Fig. 7) and BGR01-07 (Fig. 9) provide information on the northeastern study area, which includes the Compressed Fold Belt and the feature termed ‘‘allochthon’’ or ‘‘Major Thrust Sheet’’. In the Compressed Fold Belt (Fig. 2), the surface expression of the imbricate fan is limited to two to four individual thrusts (Fig. 8). The individual thrusts are narrower and steeper in comparison to the southern Thrust Zone. In addition, the faults appear to have a less concave geometry. This indicates that the compressional forces that affected the Compressed Fold Belt were greater than in the southwestern study area. The ‘‘Major Thrust Sheet’’ feature lies to the east of the Compressed Fold Belt. The boundary between the Compressed Fold Belt and the ‘‘Major Thrust Sheet’’ has been interpreted as a wrench structure (Tan and Lamy, 1990). However, this boundary offsets

horizon B by more than 1 s (TWT), indicating that the eastern boundary of the Compressed Fold Belt is a major thrust fault (Fig. 8). Previous authors have interpreted the ‘‘Major Thrust Sheet’’ feature to have a maximum width of 70 km (Hinz et al., 1989; Tan and Lamy, 1990; Rice-Oxely, 1991; Hazebroek and Tan, 1993; Madon et al., 1999). However, lines BGR01-06 and -07, which cross the previously defined ‘‘Major Thrust Sheet’’ at similar locations to previous surveys indicate a much smaller structure—only 20 km wide (Figs. 7–9). Indeed, this much smaller size, and the following physical analysis, raises some doubts over the previously hypothesised origin of this structure (see the ‘‘Modelling the refraction seismic and gravity data’’ section). The data collected for this study confirms the chaotic seismic facies of sedimentary sequences below marker horizon B, although the proposed boundary to the ‘‘Lower Thrust Sheet’’ is either not observed or absent (Fig. 8, Hinz et al., 1989).

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

616

Table 1 Horizons of this paper and their relationship to other studies Horizon

Time

Unconformities mentioned in other papers

Properties and identifications

Pleistocene

Horizon I (Levell, 1987)

Uplift associated with tensional faulting over the crests of many structures (mild onlap)

A

Early/Late Pliocene boundary 3.6 Ma

Unconformity A (Hinz et al., 1989)

B

Late Miocene 10 Ma

Shallow regional unconformity (SRU); unconformity B of Schlu¨ter et al. (1996) Upper intermediate unconformity (UIU) (Levell, 1987) Lower intermediate unconformity (LIU) (Levell, 1987)

Truncation or mild onlap associated with the formation of open anticlines and synclines with a general NW–SE orientation (compressional tectonic) This erosional surface remained on the shallow shelf above sea-level until Pliocene times (Rice-Oxely, 1991) (onlap, erosional)

C

D

E F

Early Middle Miocene? 16 Ma

Oligocene to late Early Miocene 17–33 Ma Early/Middle Oligocene 32 Ma Mesozoic/Tertiary basement

Horizon II (Levell, 1987)

Reduced rates of relative sea-level rise in the early Late Miocene. (onlap, erosional) A marine onlap unconformity. Increase in the rate of relative sea-level rise (late Middle Miocene)

Unconformity C/6 (Hinz et al., 1989; Schlu¨ter et al., 1996) Deep Regional Unconformity (DRU) (e.g. Levell, 1987)

Underlies the wedge of deformed thrust rocks

Horizon D (Kudrass et al., 1986; Hinz et al., 1989; Schlu¨ter et al., 1996)

Top of a unit of carbonates known from the Dangerous Grounds

Unconformity E (Schlu¨ter et al., 1996) Unconformity F (Schlu¨ter et al., 1996)

Base of the carbonates

The new data clearly image a change in the detachment topography from the southwestern towards the northeastern margin. The top of the inferred carbonates (horizon D) in this region is not as smooth as the detachment surface in the southwest of the Thrust Zone. A dissection of marker horizon D is distinct along line BGR01-06 (Figs. 7 and 8, bottom). The irregular horizon can be unequivocally followed beneath the Compressed Fold Belt and partly beneath the ‘‘Major Thrust Sheet’’ (Fig. 8, bottom). We suggest that the dissection of horizon D is caused by higher compressional forces affecting the northeastern margin resulting in thrust faults in the formerly smooth detachment horizon. Alternatively, it is possible that the compressed style of the sediments resting on the subsiding lower plate results purely from the relief in the detachment topography. However, regardless of origin, we observed that the amount of dissection increases towards the northeast and the seismic pattern suggests a higher sole-out level (time equivalent to 0.5–1.0 s (TWT)) for the individual thrusts, rather than horizon D itself (e.g. line BGR01-07; Fig. 8, top). The basement reflector (F) can be only occasionally interpreted along the lines traversing the Compressed Fold Belt (Figs. 7–9). The easternmost parts of the lines BGR01-06 and -07 show a complex structural pattern comprising extensional, thrust and wrench faulting (Figs. 7 and 9). As in the SW imbricate zone, the complexity of the deeper parts of the

Angular and somewhere erosional unconformity which separated the deep marine shales and turbidity sands of the Late Eocene to Early Miocene from the clastic shallow sediments of post-early Middle Miocene (onlap)

Unconformity between Mesozoic and Tertiary rocks

faults may be explained by intrusions of overpressured mobile shales (Morley et al., 2003). 4. Modelling the refraction seismic and the gravity data Refraction seismic and gravity data were modelled to elucidate further information on the sedimentary successions and the nature of the ‘‘Major Thrust Sheet’’. Primary P-phases were identified in the refraction seismic data and the arrival times were interactively picked. We calculated synthetic traveltimes and compared them to the picked values using a 2-D FD-Vidale forward modelling algorithm. The known bathymetry and the depth to the acoustic basement from the reflection seismic data formed the minimum-parameter/minimum-structure starting model (Zelt, 1999). The layer velocities for the starting model were adopted from the reflection seismic derived interval velocities. In accordance with the interpreted reflection seismic data more layers were added successively when the wide-angle data made them necessary. In the case that the data could not be fitted with this approach additional velocity points were added. The refraction modelling was an iterative process between seismic velocity and density modelling, whereby standard velocity–density relations were used (Ludwig et al., 1970; Christensen and Mooney, 1995). Every iteration in the refraction model was checked for agreement with the shipboard gravity data. The

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

elongated structure of the free-air gravity anomalies in SW–NE direction suggests that the forward modelling of the free-air anomalies in two dimensions can be performed with a suitable degree of accuracy. The final velocity model was gridded with a spatial increment of 175 m and the geometry of this model superimposed on the reflection seismic data (Fig. 9). The density model and calculated gravity provided a good fit to the observed gravity (Fig. 10) as for methodology specifications by default, a constant gravity value for each density model is subtracted from the calculated gravity to make the measured and the calculated gravity values comparable. To illustrate the gravity effect of the sea bottom bathymetry a model with a uniform density was calculated (sea water density ¼ 1030 kg/m3). The density was chosen to 2200 kg/m3 representing usual oceanic sediments. The calculated gravity effect of this model (dotted line in Fig. 10A) roughly explains the measured gravity data. The detailed match of the shipboard free-air gravity anomalies with the calculated anomalies within a few mGal was obtained by modelling the structures of the refraction model with densities derived from standard velocity–density relations. The velocities and the corresponding density values are summarised in Table 2. 4.1. Refraction derived velocities of the sedimentary succession The modelled sedimentary succession consists of three main layers that are in accordance with both the traveltimes of the refraction seismic and distinct horizons in the MCS data in the western part of the model. The uppermost layer was modelled to have a velocity of vp ¼ 2.0 km/s that slightly increases to 2.1 km/s in the easternmost section, where the modelled thickness of this layer also increases (Fig. 9; distance 160–180 km). The Nafe-Drake derived density of this layer is 2000 kg/m3. The base of this layer approximately matches horizon A. The base of the following layer 2 with a velocity of 2.3–2.4 km/s (r ¼ 2300 kg/m3) in the Borneo Trough and the Compressed Thrust Zone area coincides roughly with reflection seismic horizon B (Fig. 9; distance 50–120 km). Layer 3 was modelled as having a velocity of 3.1–3.2 km/s (r ¼ 2380–2400 kg/m3) in the area of the Borneo Trough. However, in the eastern part of the model, coinciding with the previously termed ‘‘Major Thrust Sheet’’, the modelled velocity and density were found to be anomalously high (Hinz et al., 1989; Tan and Lamy, 1990; Hazebroek and Tan, 1993; Figs. 9 and 10). This high velocity aspect of the structure, and its implications on the origin of this feature, will be discussed in more detail in the following chapter. The fourth layer in the model, with a velocity of 4.0 km/s, represents the Oligocene to Lower Miocene carbonate platform sequence (Nido carbonates). The fill of the half grabens between tilted fault blocks shows a modelled velocity of 3.6 km/s (r=2420 kg/m3). The base of layers 4 and 5 of the refraction model show a reasonable fit with

617

horizons D and F, respectively, as interpreted from the MCS data (Fig. 9). The thin carbonate layer could not be resolved in the gravity data and thus was not considered in the density model. In the eastern part of the line, where the MCS data show a chaotic pattern, a high velocity and density (4.3 km/s; 2550–2600 kg/m3) body is modelled with a thickness of 7.5 km. This may indicate a me´lange of Rajang Group fold-thrust belt and/or Crocker sediments. It is also possible that parts of the upper crust of the subsiding Dangerous Grounds plate are incorporated into this me´lange. The top of the upper crustal layer (labelled F in Fig. 9, bottom) represents the unconformity between Mesozoic rocks and Tertiary rocks, such as the siltstones, black shales, gabbro and volcanics dredged from the Reed Bank and Dangerous Grounds from below a porous packstone of Late Oligocene to Early Miocene age (Kudrass et al., 1986; Schlu¨ter et al., 1996). The faulted blocks with an equivalent density for the upper crust of 2650 km/m3 were fitted with a P-wave velocity of 5.7–5.8 km/s. The model fits the half graben structures identified in MCS data quite well in the central and western part of the line (Fig. 9, bottom). The depth and shape of the sedimentary marker horizons and the Nido carbonates are constrained by gravity and refraction data and correlated with structures identified in the MCS data. However, the deep crust and the Moho were modelled using mainly the gravity and in parts refraction data. The depth and shape of the deep parts of the model are therefore only tentatively shown to complete the picture. Hence, an error of some 2–4 km must be taken into account. 5. Discussion 5.1. Is there a ‘‘Major Thrust Sheet’’? The modelled velocity of sedimentary layer 3 increases from 3.1–3.2 km/s to 3.45 km/s at a distance of 120 km (Fig. 9; between stations OBH3 and OBH4). These two OBH stations show distinctly higher first break velocities of 3.45 km/s in their intersecting part (Fig. 11). To account for these data, a structure with a velocity of 3.45 km/s and extending from 120 to 135 km offset was modelled on top of layer 3. In the density model, a 15 km wide and 3 km thick body of higher density (r ¼ 2430 kg/m3) needed to be included. The outer branches of OBH3 and OBH4 indicate velocities of only about 2.85–2.9 km/s. Hence, the need to use a slightly higher velocity of 3.1 km/s to model these observed breaks suggests a shallower position of the high velocity body when compared to the average 3–4 km depth of the top of this layer to the WNW and ESE of this body. This results in a down-dip path for the refracted waves on both sides of the body (Fig. 11). The western limit of this high velocity body coincides with the western end of the previously termed ‘‘Major Thrust Sheet’’ or ‘‘allocthon’’ (Hinz et al., 1989; Rice-Oxely, 1991;

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

618

Distance (km)

A

0

10

20

30

40

50

60

70

80

90

100 110 120 130 140 150 160 170 180

Gravity (mGals)

80 40 0 -40 =Observed,

-80

B

0

=Calculated,

D=2000

D=1030

5 Depth (km)

=Sea Effect

=Error

D=2000 D=2380

D=2420

upper crust

10

D=2430

D=2550 D=2600

D=2650

15 lower crust

D=2930

20 25

D=3350

upper mantle

gravity model

D=2650

30 0

C

D=2300 D=2400

10

20

30

40

50

60

70

80

90

gravity model refraction model density value (kg/m3)

100 110 120 130 140 150 160 170 180

0

Depth (km)

5 10 15 20 25 BGR01-07R

ray coverage 30

D

0

5 Time (s)

2

calculated traveltime

3

4

5

measured traveltime

10

15

20 0

10

20

30

40

50

60

70

80 90 100 110 120 130 140 150 160 170 180 Distance (km)

Fig. 10. Refraction and gravity modelling results: (A) shows the general fit of the measured and calculated gravity from the density model shown in (B). The dotted curve illustrates the sea bottom effect (seafloor topography of density 2200 km/m3). Superimposed on the density model (dashed lines) shown in (B) are the boundaries from the resulting refraction model (solid lines). The numbers denote density values in kg/m3. The high velocity/density body is shown in grey. In (C) the ray coverage from the refraction model is shown and in (D) the measured and calculated traveltimes from the refraction model (compare Fig. 9) are presented.

Hazebroek and Tan, 1993; Madon et al., 1999). Hinz et al. (1989) first proposed that this province is defined by two Major Thrust Sheets, with a width between 25 and 70 km, stacked over a southward migrating Dangerous Grounds continental basement. An older thrust system, called Major

Thrust Sheet System, overrides the younger Lower Thrust Sheet System. Several authors followed the interpretation of a Major Thrust system (Tan and Lamy, 1990; RiceOxely, 1991; Hazebroek and Tan, 1993) and proposed that the thrust sheet represents a nappe that resulted from

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

619

Table 2 Velocities and corresponding densities from refraction seismic and gravity modelling Layer

Interpretation

P-velocity (m/s)

Density (kg/m3)

1 2 3

Sea water Plio-/Pleistocene sediments (A) Late Miocene–Pliocene sediments (B)

1500 1900–2100 2300–2400

1030 1900–2000 2300

4

Early Middle Miocene–Late Miocene sediments (C) Thrust sheet

3100–3200 3450

2380–2400 2430

5 6 7

Oligocene to late Early Miocene carbonate platform (D) Pre-Miocene sediments Mesozoic/Tertiary basement

4000 3600 5500–5800

2550 2380–2420 2650–2720

8

Lower crust Upper mantle

6500 8000

2930 3350

The velocity density relations are in accordance with the Ludwig et al. (1970) curve for the sedimentary rocks and with Christensen and Mooney (1995) for the basement rocks.

Fig. 11. Refraction seismic data from OBH3 and OBH4 displayed in reduced traveltime at Vred ¼ 6.000 m/s. In the intersecting part (i.e. at a distance from 115 to 128 km) the first breaks show distinctly higher velocities in comparison with the outer sides to the west and to the east; consequently a high velocity body was modelled in the refraction model (cut-out shown in the lower panel).

gravity sliding associated with the uplift of the RajangCrocker fold-thrust belt (Fig. 1). The higher velocity, density and reflection seismic characteristic are in accordance with earlier interpretations proposing that this structure consists of rocks of the Crocker Accretionary Prism overlain by a thin cover of younger deposits (Hinz et al., 1989; Rice-Oxely, 1991; Hazebroek and Tan, 1993; Madon et al., 1999). However, the refraction data (Fig. 11) demonstrate that the extent of the structure is much smaller

than previously predicted with an extent of only 20 km, rather than previously interpreted 70 km extent (Tan and Lamy, 1990; Rice-Oxely, 1991; Hazebroek and Tan, 1993; Madon et al., 1999; Fig. 11). Furthermore, the refraction seismic data require a modelled lower velocity and a layer dipping landward to the east of 135 km distance (Fig. 9; distance 135–150 km) to account for the difference in the first breaks from the left and right side of OBH4 (Fig. 11, upper right panel).

ARTICLE IN PRESS 620

D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

In our view, the much smaller size of this body raises some questions on the proposed origin of this unit as an allochthonous mass of pre-Middle Miocene rocks that came to its present position by gravity gliding. The limited width and the great distance from the Crocker-Rajang fold-thrust belt are already strange but the shape and orientation of the body in relation to the Crocker-Rajang mountain belt are difficult to explain by gravity gliding. Moreover, we found no indication in the data for the proposed boundary, i.e. a change in the reflection pattern, between the Major Thrust Sheet and the Lower Thrust Sheet. The apparent higher velocity of the body (Fig. 9) and the chaotic seismic facies of the sediments beneath the body (Fig. 8) could alternatively be explained by the presence of a carbonate layer. In this case, horizon B would be interpreted to represent the top of the carbonates in the area where the thrust sheet is indicated in Fig. 8. Carbonates limit the transfer of seismic energy into the layers beneath and show distinct higher velocities in comparison to Neogene sediments. This interpretation is analogous to the Kudat platform (Madon et al., 1999), an N–S trending area in the Inboard Belt, offshore Sabah, some 100 km to the east of the high velocity body (Fig. 2). At the Kudat platform, shallow marine carbonates probably grew on local highs and the platform shows a comparable size as the high velocity body herein. Another possibility is that the high velocity body represents Crocker sediments that were overthrust onto a southward migrating attenuated continental block of the Dangerous Grounds during plate convergence. Sediments of the Paleogene Crocker formation, probably in parts dissected by ophiolites from a proto-South China Sea oceanic crust, are expected to show higher seismic velocities compared to the surrounding Miocene and younger sediments. Furthermore, such a crustal shortening process would be expected to destroy the sedimentary facies and thus provides an explanation for the chaotic seismic facies observed in the underlying sediments. In addition, such a compacted sedimentary body is likely to act as backstop for the progressively transferred sediments from the NW Borneo Trough, and thus may explain the more compressed style of the toe thrust belt in the north compared to the southern part of the area. 5.1.1. The deep crust and the Moho Although the Moho is not as well constrained as the shallower horizons, it is quite clear that the Moho is elevated with respect to typical continental crust (Christensen and Mooney, 1995). The shallow level of the Moho of about 22 km or even less implies that the crust is likely to have been already thin before subduction/collision took place as suggested by Milsom et al. (1997). However, especially the gravity data require an additional thinning from beneath the Dangerous Grounds to beneath the Compressed Thrust Zone (Fig. 9) in the model. It is not possible to construct a viable model with a continuous

downward bulge of the Moho beneath the increasing sedimentary load at the slope, such as would be expected for a subduction/collisional margin. The upper crustal body as well as the sedimentary successions in this eastern part of the model were already modelled with higher densities compared to the trough/slope area. A slight increase in density is not surprising since the load increases in eastward direction. But even values at the upper end of plausible variations, as shown in Fig. 10, are not in accordance with a continuously deepening or, at least, flat Moho. There is a slight deepening of the Moho in the easternmost part of the model but the crust continues to be thin up to the end of the model (Fig. 9, top). 5.2. Is the Thrust Zone part of a deltaic system? The key question for the determination of a proper model for the evolution of the system of five to seven elongated parallel imbricate thrust is the nature and extent of the sole fault beneath the hinterland. In principle, three end-member models can be considered (Fig. 12). A. The toe thrusts result from gravitational delta tectonics with the sole fault returning to the surface either to the west (1) or to the east (2) of the Morris Fault. B. The gravitational collapse of the uplifted hinterland created by the shortening and thickening of the basement rocks (the root model) resulting in compression in the wedge-shaped fold and thrust belt on the margin. C. The imbricate thrusts are caused by crustal shortening in connection with the subduction/collision of the Dangerous Grounds continental terrane with the NW Borneo margin. Hinz et al. (1989) interpreted the thrust system as postEarly Miocene sediments that are thrusted onto the progressively subsiding continental crust of the gradually overridden Dangerous Grounds continental terrane (model C). Tan and Lamy (1990), in contrast, suggest gravity induced sliding along deep de´collement planes induced by differential loading in the more updip parts of the delta as cause for the young Thrust Zone (model A), but also mention that the zone may have been initiated by subduction. Madon et al. (1999) suggest the Thrust Zone to be part of a delta that forms a major depocentre for late Middle Miocene and younger shelf and slope sediments (model A). In these models, the eastern limit of the proposed delta (East Baram Delta) is assumed to coincide with the western limit of the Inboard Belt (e.g. Tan and Lamy, 1990; Rice-Oxely, 1991; Madon et al., 1999). This implies that the belt of shortening (Thrust Zone) as well as the corresponding inboard extensional belt are located within the extent of the seismic lines from this study (Fig. 12A (1)). The deltaic structures are interpreted to be essentially detached from a relative basement underlying the delta systems. Analogues would be thick piles of deltaic

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

621

Fig. 12. Schematic models for the formation of the toe Thrust Zone (modified after Twiss and Moore, 1992). The left sides of the sketches are simplified and schematic, but based on the data from this study while the right sides are solely hypothetical. (A) The zone of shortening is paired by a zone of lengthening either to the west (the ‘‘East Baram Delta’’ hypothesis) or to the east of the Morris Fault. (B) The sole faults of the foreland toe Thrust Zone terminate in a zone of ductile basement deformation. The compression in the toe Thrust Zone is caused by a gravitational collapse of the topographic high created by the shortening and thickening of the metamorphic core. (C) Continental crust is on a down-going plate. The sole fault is the subduction zone.

sediments which have prograded over passive continental margins (e.g. Niger delta, Mississippi delta). 5.2.1. Model A Under the assumption that seafloor-spreading in the South China Sea ended in the Early Miocene (20.5 Ma) and terminated plate convergence at the NW Sabah margin, the apparently young age of the individual toe thrusts affecting late Tertiary, and even Quaternary, sediments forms the major argument for delta toe thrusting. However, at least two preconditions must be fulfilled for the gravitational delta tectonics hypothesis: (1) gravitational sliding must be downslope and (2) an extensional regime is necessary in the hinterland to create normal faults. What we found herein is that the top of the lower plate (carbonate platform; horizon D) forms the detachment for the thrusted sediments (Figs. 4 and 6). This detachment clearly dips towards the continent (Fig. 5). A set of foreland-dipping normal faults landward of the Thrust Zone, necessary for the initiation of a gravity sliding process, is not clearly imaged or absent in the data from this study and also in the literature. A huge wrench fault,

the Morris Fault (that extends towards Brunei as Jerudong Fault) is commonly interpreted at the shelf edge (Tan and Lamy, 1990; Rice-Oxely, 1991; Hazebroek and Tan, 1993; Madon et al., 1999; compare Fig. 2). One might argue that the thrust sheets are the result of ductile flow, or a transfer of forcy, within the sedimentary successions. Then the driving force is provided by the topographic slope and the slope of the detachment is not restricted. The flow of sediment packages could take place on slopes dipping upward in the direction of thrusting. The topographic slope in the study area is about 2% and the decollement dips at 4% towards the hinterland (Fig. 5). However, ductile flow is expected to destroy, at least to a certain degree, a distinct stratification. We find the stratification of the parallel-layered sediments in the Borneo Trough well preserved (although tilted) in the individual imbricates (see Figs. 3 and 6). Another question with ductile flow is raised by the imbricate thrusts itself. Assuming that the movements are mainly caused by ductile flow of no brittle material, why should an imbricate system of toe thrusts develop and not one big sheet in the area of the Borneo Trough?

ARTICLE IN PRESS 622

D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

Another issue is that the study area might not examine the complete system, i.e. the belt of lengthening pairing the belt of shortening (the toe Thrust Zone) may occur within the Inboard Belt tectonostratigraphic province or even onshore NW Sabah (Fig. 12A(2)). However, the Inboard Belt (Fig. 2) is characterised by N–S to NNE–SSW trending anticlines with steep flanks and faulted crests (Madon et al., 1999). The anticlines are interpreted as transpressional structures (Hazebroek and Tan, 1993) from predominantly Late Miocene strike-slip and compressional movements. Also, there is strong evidence for Late Pliocene compressional tectonics as indicated by numerous crestal faults on the anticlines in the transition zone from the Inboard Belt to the toe Thrust Zone (Fig. 2) (Madon et al., 1999). The onshore area from the NW Sabah coast to the Mount Kinabalu intrusion is defined by predominantly thrust faults and some strike-slip faults (Benard et al., 1990; Tongkul, 1991, 1994). Indications for extensional movements are—apart from the offshore border area to Brunei—not distinct or absent. 5.2.2. Model B and Model C We propose that the crustal shortening models (B, C) fit much better the data and are supported by onshore findings which confirm Pliocene, probably extending into the Pleistocene, compressional deformations (approximately N–S) in Sabah (Benard et al., 1990; Tongkul, 1994; Meng, 1999). The young volcanism in the Borneo Trough (see chapter discussion of selected lines) also indicates tectonic activity. In model B, the basement rocks are shortened by processes other than thrusting. Sole faults of the imbricate thrust belt terminate in the root zone of ductile deformation within the metamorphic core. The gravitational collapse of the thickened and elevated Crocker-Rajang fold belt would be responsible for compression in the toe Thrust Zone. The third possible model (C) is that the toe Thrust Zone reflects the ongoing, or at least lasting into the Middle Pleistocene, plate convergence. The continental crust of the Dangerous Grounds terrane is subducted to some extent and the sole fault is in this model simply the subduction zone. The apparently weak seismicity may support model B. Few earthquakes are reported in the collision area of northeast Borneo since the inception of recording. The Harvard Centroid Moment Tensor (CMT) database shows more or less E–W trending extensional focal solutions in the Sabah onshore region. One large earthquake (Ms 6.0) (see Fig. 2) occurred offshore NW Sabah on July 21, 1930 somewhere in between 7.51N, 116.01E and 7.01N, 114.01E (i.e. in between the Borneo Trough and the Outboard Belt) and another large earthquake that is quite well located to the south of the toe Thrust Zone (February 22, 1992, 5.561N, 114.751E, Mb 5.6). One Mb 5.9 event occurred in the southern prolongation of the previous earthquake at 4.25601N, 115.02551E (April 23, 2004) (all data according to the ISC, On-line Bulletin, 2001). Applying the empirical relation between earthquake size and frequency of occur-

rence to the region at least some 20 Mb 4.6 and 200 Mb 3.6 events would have been expected for the past 30–40 years. The uncertainties in the event localisations and the missing reports for the smaller events point towards a still unsatisfactory density of recording stations in that region and not automatically towards a seismically quiet region. In addition, we do not necessarily expect large earthquakes because the folds separating the individual thrust sheets in the toe Thrust Zone terminate downward at a basal decollement over a widely undeformed and rigid basement. Yeats et al. (1981) pointed out that faults which do not extend downward into rocks of high strength will not be expected to produce large-amplitude ground acceleration due to seismic shaking. Such rocks under near-surface confining pressures are not capable of storing enough elastic strain energy to generate a large earthquake when that strain energy is released instantaneously. Where the de´collement contains rocks of such low strength that deformation may occur plastically under low confining pressure, e.g. in the fold-and-thrust belt in southeast Iran and the Pakistan Salt Range, internal deformation is probably not accompanied by large earthquakes (Berberian, 1981; Seeber et al., 1981; Fielding et al., 2004). Although earlier GPS results in first-approximation indicated that SE Asia moves eastward at 1 cm/year relative to Siberia, there were significant discrepancies in the definition of boundaries of coherent blocks and relative motions (e.g. Michel et al., 2000; Kreemer et al., 2000). Rangin et al. (1999) were the only authors who suggested that crustal shortening at the NW Borneo Trench accommodates remnant E–W distributed motion of the Sundaland/Philippine plate convergence. A higher level of accuracy required to resolve such motion were achieved with long time series of GPS data from a dense network (Simons et al., 2007). These data show that the entire northern tip of Borneo moves at a velocity of 6 mm/year with respect to Sundaland (Simons et al., 2007). According to these authors, the remaining Sundaland/Philippine plate convergence is absorbed at the NW Borneo Trench (Rangin et al., 1999; Simons et al., 2007). Hereby the strike-slip faulting at the south coast of Sabah may either also accommodate E–W shortening at the Celebes basin margin (Rangin et al., 1999) or instead result from the clockwise rotation of the northern tip of Borneo. In our interpretation, the toe Thrust Zone results more likely from crustal shortening rather that from gravitational tectonics. The distinction between the two endmember models for crustal shortening (Fig. 12B and C) solely with data from this study is difficult. There are no indications for multiple sole faults in the southwestern study area as predicted by the root complex model (B) and the detachment or sole fault reaches only to a depth of 10 km along our lines, i.e. the crust is well within the depth range where its behaviour is brittle. This may tentatively indicate a preference for model C. Gravity-related delta tectonics may have influenced the evolution of the toe Thrust Zone, especially the shallower sedimentary succes-

ARTICLE IN PRESS D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

sions, but we assume this contribution to be small. This is in accordance with Ingram et al. (2004) who postulated mainly crustal shortening with a small contribution from gravity-related delta tectonics.

623

during plate convergence, or the top of the high velocity body may represent a carbonate platform that limits the transfer of seismic energy into the layers beneath. Acknowledgements

6. Summary MCS data, refraction and gravity modelling offer the opportunity to investigate the origin and evolution of the collision-prone continental margin off NW Borneo. This paper focuses on the compressional sedimentary structures landward of the Borneo Trough that show a striking change from southwest to northeast. In the southwest, a toe thrust system of five to seven elongated imbricate thrust fans developed which overlap like roof tiles and trend all in the same general direction of 30–451 (Figs. 2, 4 and 6). We found the age of deformation to migrate outward toward the front of the wedge. The faults separating the individual imbricates cut through post-Early Miocene sediments and curve down to the carbonate platform that represents the top of the lower plate. These carbonates represent the major detachment surface for the sediments that fill the Borneo Trough. Evidence for gravitational gliding processes as origin for the young toe thrusts, affecting late Tertiary and even Quaternary sediments, is rare. We found in the data neither indication for a foreland-dipping detachment horizon, nor for a zone of lengthening pairing the toe thrust belt of shortening. Ductile flow is an unlikely mechanism for the formation of an imbricate fan. The shallow shelf (the Inboard Belt tectonostratigraphic province) is by almost all authors consistently interpreted as wrench-prone with few or no seaward dipping normal or growth faults. We propose that the toe thrusts result predominantly from crustal shortening processes. In the northeastern part of the study area, the thrusted sediments show a considerably stronger compressed style (Compressed Thrust Zone, Fig. 2). The compressed toe thrusts are limited to only two to four steep individual imbricate thrusts and we found the proposed detachment surface to be also affected by compression. Landward to the Compressed Thrust Zone follows the previously defined ‘‘Major Thrust Sheet’’. The refraction seismic data indicate a high-velocity body at the western location of this predicted nappe but also limit the extend of this body to only 20 km. Earlier interpretations predicted an extent of the thrust sheet of 70 km at the location of the refraction seismic line. The limited extent of the body and the great distance from the Crocker-Rajang fold-thrust belt led us to question the idea of gravitational gliding of an allochthonous mass of pre-Middle Miocene rocks. The body on top of sediments showing a chaotic seismic facies may be interpreted either as Paleogene Crocker sediments in parts dissected by remnants of a proto-South China Sea oceanic crust that were overthrust onto a southward migrating attenuated continental block of the Dangerous Grounds

For fruitful discussions we expressly thank Karl Hinz, the former head of BGRs geophysical division. We are also indebted to Heinrich Meyer who took over the leadership of the survey. Thoroughly reviews and insightful comments and suggestions on an earlier version of the paper by John Milsom and Robert Hall are kindly acknowledged. Funding for this work was provided by the Federal Institute for Geosciences and Natural Resources (BGR). This is a contribution to project FR 2119/1-1 funded by the Deutsche Forschungsgemeinschaft (German Research Foundation). References Barckhausen, U., Roeser, H.A., 2004. Seafloor spreading anomalies in the South China Sea revisited. In: Clift, P., Kuhnt, W., Wang, P., Hayes, D.E. (Eds.), Continent–Ocean Interactions in the East Asian Marginal Seas, Geophys. Monograph Series, vol. 149, pp. 121–125. Benard, F., Muller, C., Letuzey, J., Rangin, C., Tahir, S., 1990. Evidence for multiphase deformation in the Rajang–Crocker Range (northern Borneo) from Landsat imagery interpretation: geodynamic implications. Tectonophysics 183, 321–339. Berberian, M., 1981. Active faulting and tectonics of Iran, in ZagrosHindu Kush-Himalaya Geodynamic Evolution. In: Gupta, H.K., Delany, F.M. (Eds.), Am. Geophys. Union, Geodyn. Ser., vol. 3, pp. 33–69. Bol, A.J., van Hoorn, B., 1980. Structural styles in Western Sabah offshore. Geological Society of Malaysia Bulletin 12, 1–16. Briais, A., Patriat, P., Trapponier, P., 1993. Updated interpretation of magnetic anomalies and seafloor spreading stages in the South China Sea: implications for the Tertiary tectonics of Southeast Asia. Journal of Geophysical Research 98, 6299–6328. Christensen, N.I., Mooney, W.D., 1995. Seismic velocity structure and composition of the continental crust; a global view. Journal of Geophysical Research 100 (6), 9761–9788. Fielding, E.J., Wright, T., Muller, J., Parsons, B.E., Walker, R., 2004. Aseismic deformation of a fold-and-thrust belt imaged by synthetic aperture radar interferometry near Shahdad, southeast Iran. Geology 32 (7), 577–580. Hall, R., 1996. Reconstructing cenozoic SE Asia. In: Hall, R., Blundell, D.J. (Eds.), Tectonic Evolution of SE Asia, Geol. Soc. Spec. Pub., vol. 106, pp. 153–184. Hall, R., 2002. Cenozoic geological and plate tectonic evolution of SE Asia and the SW Pacific: computer-based reconstructions, model and animations. Journal of Asian Earth Sciences 20 (4), 353–431. Hamilton, W., 1979. Tectonics of the Indonesian region. USGS Professional Paper, 1.078, pp. 1–345. Hazebroek, H.P., Tan, D.N.K., 1993. Tertiary tectonic evolution of the NW Sabah continental margin, In: Geh, G.H. (Ed.), Proceedings of the Symposium on Tectonic Framework and Energy Resources of the Western Margin of Pacific Basin. Geological SocIety of Malaysia Bulletin 33, 195–210. Hinz, K., Fritsch, J., Kempter, E.H.K., Manaaf Mohammad, A., Meyer, H., Mohamed, D., Vosberg, H., Weber, J., Benavides, J.J., 1989. Thrust Tectonics along the continental margin of Sabah, Northwest Borneo. Geologische Rundschau 78, 705–730. Hutchinson, C.S., 1996. The ‘‘Rajang accretionary prism’’ and ‘‘Lupar Line’’ problem of Borneo. In: Hall, R., Blundell, D.J. (Eds.), Tectonic

ARTICLE IN PRESS 624

D. Franke et al. / Marine and Petroleum Geology 25 (2008) 606–624

Evolution of Southeast Asia. Geological Society of London Special Publication, vol. 106, pp. 247–261. Hutchinson, C.S., 2004. Marginal basin evolution: the southern South China Sea. Marine and Petroleum Geology 21/9, 1129–1148. Ingram, G.M., Chisholm, T.J., Grant, C.J., Hedlund, C.A., Stuart-Smith, P., Teasdale, J., 2004. Deepwater North West Borneo: hydrocarbon accumulation in an active fold and thrust belt. Marine and Petroleum Geology 21, 879–887. International Seismological Centre (ISC), 2001. On-line bulletin. Internat. Seis. Cent., Thatcham, UK. /http://www.isc.ac.uk/BullS. Kreemer, C., Holt, W.E., Goes, S., Govers, R., 2000. Active deformation in eastern Indonesia and the Philippines from GPS and seismicity data. Journal of Geophysical Research 105, 663–680. Kudrass, H.R., Wiedicke, M., Cepek, P., Kreuzer, H., Mu¨ller, P., 1986. Mesozoic and Cenozoic Rocks dredged from the South China Sea (Reed Bank Area) and Sulu Sea and their significance for Plate Tectonic reconstructions. Marine and Petroleum Geology 3, 19–30. Levell, B.K., 1987. The nature and significance of regional unconformities in the hydrocarbon-bearing Neogene sequence offshore West Sabah. Geological Society of Malaysia Bulletin 21, 55–90. Ludwig, W., Nafe, J., Drake, C., 1970. Seismic refraction. In: Maxwell, A. (Ed.), The Sea, vol. 4. Wiley-Interscience. Madon, M.B.Hj., 1999. Basin types, tectonostratigraphic provinces, and structural styles. In: The Petroleum Geology and Resources of Malaysia. Petroliam Nasional Berhad (PETRONAS), Kuala Lumpur 1999, pp. 77–111, ISBN 983-9738-10-0. Madon, M.B.Hj., Meng, L.K., Anuar, A., 1999. Sabah basin. In: The Petroleum Geology And Resources of Malaysia. Petroliam Nasional Berhad (PETRONAS), Kuala Lumpur 1999, pp. 501–542, ISBN 983-9738-10-0. Meng, L.K., 1999. Geological setting of Sabah. In: The Petroleum Geology and Resources of Malaysia. Petroliam Nasional Berhad (PETRONAS), Kuala Lumpur 1999, pp. 473–498, ISBN 983-9738-100. Michel, G.W., Becker, M., Angermann, D., Reigber, C., Reinhart, E., 2000. Crustal motion in E- and SE-Asia from GPS measurements. Earth Planets Space 52, 713–720. Milsom, J., Holt, R., Bin Ayub, D., Smail, R., 1997. Gravity anomalies and deep structural controls at the Sabah-Palawan margin, South China Sea. In: Frazer, A.J., Matthews, S.J., Murphy, R.W. (Eds.), Petroleum Geology of Southeast Asia, Geol. Soc. Spec. Pub. no. 126, pp. 417–427. Morley, C.K., Back, S., Van Rensbergen, P., Crevello, P., Lambiase, J.J., 2003. Characteristics of repeated, detached, Miocene-Pliocene tectonic inversion events, in a large delta province on an active margin, Brunei Darussalam, Borneo. Journal of Structural Geology 25, 1147–1169. Pubellier, M., Monnier, C., Maury, R., Tamayo, R., 2003. Plate kinematics, origin and tectonic emplacement of supra-subduction ophiolites in SE Asia. Tectonophysics 392, 9–36.

Rangin, C., Le Pichon, X., Mazzotti, S., Pubellier, M., Chamot-Rooke, N., Aurelio, M., Walpersdorf, A., Quebral, R., 1999. Plate convergence measured by GPS across the Sundaland/Philippine Sea plate deformed boundary: the Philippines and eastern Indonesia. Geophysical Journal International 139, 296–316. Rice-Oxely, E.D., 1991. Paleoenvironments of the Lower Miocene to Pliocene sediments in offshore NW Sabah Area. Geological Society of Malaysia Bulletin 28, 165–194. Sandal, S.T., 1996. The geology and hydrocarbon resources of Negara Brunei Darussalam (1996 revision). Brunei Shell Petroleum Company/Brunei Museum, Syabas Bandar Seri Begawan, Brunei Darussalam, 243pp. Schlu¨ter, H.U., Hinz, K., Block, M., 1996. Tectono-stratigraphic terranes and detachment faulting of the South China Sea and Sulu Sea. Marine Geology 130, 39–68. Seeber, L., Armbruster, J.G., Quittmeyer, R., 1981. Seismicity and continental subduction in the Himalayan arc. In: Gupta, H.K., Delany, F.M. (Eds.), Zagros, Hindu Kush, Himalaya Geodynamic Evolution. American Geophysical Union Geodynamic Series, no. 3, pp. 215–242. Simons, W.J.F., Socquet, A., Vigny, C., Ambrosius, B.A.C., Haji Abu, S., Chaiwat, P., Subarya, C., Sarsito, D.A., Matheussen, S., Morgan, P., Spakman, W., 2007. A decade of GPS in Southeast Asia: resolving Sundaland motion and boundaries. Journal of Geophysical Research 112, B06420. Tan, D.N.K., Lamy, J.M., 1990. Tectonic evolution of the NW Sabah continental margin since the late Eocene. Geological Society of Malaysia Bulletin 27, 241–260. Taylor, B., Hayes, D.E., 1980. The tectonic evolution of the South China Sea Basin. In: Hayes, D.E. (Ed.), The Tectonic and Geologic Evolution of South-East Asian Seas and Islands, Part 1. AGU Geophys. Monograph, vol. 23, pp. 89–104. Tingay, M.R.P., Hillis, R.R., Morley, C.K., Swarbrick, R.E., Drake, S.J., 2005. Present-day stress orientation in Brunei: a snapshot of ‘prograding tectonics’ in a Tertiary delta. Journal of the Geological Society 162 (1), 39–49. Tongkul, F., 1991. Tectonic evolution of Sabah, Malaysia. In: Nichols, G., Hall, R. (Eds.), Proceedings of the Orogenesis in Action Conference, London 1990. Journal of Southeast Asian Earth Sciences 6, 395–405. Tongkul, F., 1994. The geology of Northern Sabah, Malaysia: its relationship to the opening of the South China Sea Basin. Tectonophysics 235, 131–137. Twiss, R.J., Moore, E.M., 1992. Structural Geology. W.H. Freemann and Company, New York, USA. Yeats, R.S., Clark, M.N., Keller, E.A., Rockwell, T.K., 1981. Active fault hazard in southern California: ground rupture versus seismic shaking. GSA Bulletin 92, 189–196. Zelt, C.A., 1999. Modelling strategies and model assessment for wideangle seismic traveltime data. Geophysical Journal International 139, 183–204.