GSA Bulletin; January 2008; v. 120; no. 1-2;
p. 156-178; DOI: 10.1130/B26212.1
© 2008 Geological Society of America
Evolution of the offshore western Gulf of Corinth
Rebecca E. Bell
,1,
Lisa C. McNeill1,
Jonathan M. Bull1 and
Timothy J. Henstock1
1 National Oceanography Centre, Southampton, University of Southampton Waterfront Campus, European Way, Southampton SO14 3ZH, United Kingdom
 |
ABSTRACT
|
|---|
The young Gulf of Corinth rift in central Greece is an ideal place to study processes occurring during the initiation and early stages of continental extension. At the east end of the 100-km–long E-W rift, Holocene extension on N-dipping faults equates to geodetic extension rates. At the western end, however, estimated extension rates on N-dipping onshore faults account for only 
20%–40% of the geodetically measured rates (10–15 mm/yr). We use high-quality, multichannel, seismic reflection and swath bathymetry data to investigate the tectonics of the western Gulf and quantify the contribution of offshore fault activity toward total-rift extension. Five major offshore faults generate the variable basement topography, and a ca. 0.4-Ma unconformity separates stratigraphy into two main packages. Basin-fill geometry indicates that during the early stages of subsidence of the offshore western Gulf, S-dipping faults were the dominant border-fault structure. This is still the case at the western end of the western Gulf; however, elsewhere, N-dipping faults on the south margin are now relatively more dominant and cause stratigraphy to tilt south. The temporally and spatially varying rift structure contrasts with the simple, consistently N-dipping, border-fault, half-graben geometry of the east Gulf of Corinth. Combined west Gulf offshore and onshore Holocene fault extension rates total 5–14.5 mm/yr, within the range calculated geodetically, and render enhanced slip on a low-angle detachment surface beneath the western Gulf unnecessary.
Key Words: Gulf of Corinth • normal fault • seismic reflection • seismic stratigraphy • rift evolution
|
INTRODUCTION
|
|---|
The style of extension and strain distribution during the initiation and early stages of intracontinental rifting is important for understanding the eventual transition to ocean spreading and plate margin development. Few active extensional systems record less than a few million years of fault history, allowing for the study of early rifting processes. The Gulf of Corinth, a high strain band in the active Aegean region, measures 100 km x 20 km and records a Pliocene-Recent history of N-S extension, against the structural grain of the Pindos mountain chain (McKenzie, 1972; Roberts and Jackson, 1991). It is significantly smaller and younger than other examples of continental extension, like the more diffuse 800-km–wide Basin and Range province (ca. 20 million years old, Hamilton, 1987), the extensive East African rift system (extension began ca. 32 Ma; Omar and Steckler, 1996), and the Baikal rift (extension began ca. 35 Ma; Mats, 1993) Extensional deformation within central Greece is thought to be related to some combination of: back arc extension due to subduction at the Hellenic Trench (McKenzie, 1972; Doutsos et al., 1988); westward propagation of the North Anatolian fault (Taymaz et al., 1991; Armijo et al., 1996); and gravitational collapse of lithosphere thickened in the Hellenide orogeny (Jolivet, 2001).
The Gulf of Corinth rift structure has often been generalized as an E-W striking, asymmetric half graben, with N-S extension controlled by a series of N-dipping normal faults along the southern margin together with minor S-dipping antithetic faults (e.g., Roberts and Jackson, 1991; Armijo et al., 1996; Fig. 1). The geology of the steep, south coast footwall mountains results from the exhumation of Gilbert-type fan deltas due to northward propagation of fault activity (Ori, 1989; Dart et al., 1994; Gawthorpe et al., 1994) and marine terraces superimposed on the current master fault footwall (McNeill and Collier, 2004). Deformation and fault activity within the Gulf of Corinth area, over varying timescales, has been assessed through the analysis of instrumentally recorded recent seismicity (e.g., Jackson et al., 1982), historical seismicity (e.g., Ambraseys and Jackson, 1990), paleoseismology (e.g., McNeill et al., 2005a), and geomorphology. Footwall uplift and fault-slip rates have been estimated by the identification and dating of uplifted marine sediments (Pirazzoli et al., 1994; Armijo et al., 1996; Stewart and Vita-Finzi, 1996; McNeill and Collier, 2004).
View larger version (36K):
[in this window]
[in a new window]
|
Figure 1. Framework of the Gulf of Corinth. The locations of major onshore and offshore faults are taken after Stefatos et al. (2002), Leeder et al. (2005), McNeill et al. (2005b), and this study. Topography is from the Shuttle Radar Topography Mission (http://srtm.usgs.gov), and bathymetry is reproduced from the GEBCO Digital Atlas published by the British Oceanographic Data Center on behalf of Intergovernmental Oceanographic Commission (IOC) of UNESCO and the International Hydrographic Organization (IHO), 2003. Inset—Summary of Aegean regional tectonics. Arrows are north coast GPS velocity vectors for a stationary southern coastline, from Clarke et al. (1998) (Fig. 16).
|
|
Pleistocene south coast uplift rates correlate to possible fault-slip rates of the order 4–11 mm/yr (e.g., Armijo et al., 1996; McNeill et al., 2005a). Slightly higher uplift rates are obtained from Holocene base level markers (e.g., wave cut notches, Pirazzoli et al.,1994), suggesting a recent increase in deformation rate (e.g., Leeder et al., 2003; McNeill and Collier, 2004; Pirazzoli et al., 2004). Current regional extension rates have been assessed from Global Positioning System (GPS) data (e.g., Davies et al., 1997; Clarke et al., 1998; Briole et al., 2000) and are found to increase from <5 mm/yr in the east, to >10–15 mm/yr in the west (Fig. 1). In the western Gulf, in particular, there is a discrepancy between the extension rates predicted based on onshore fault activity and that observed across the rift. An explanation proposed for this discrepancy is that offshore faults could contribute significantly to the extension (Brooks and Ferentinos, 1984; Stefatos et al., 2002; Moretti et al., 2003; Sachpazi et al., 2003; McNeill et al., 2005b). It has been suggested that the master fault currently controlling rift geometry changes from being a south coast, N-dipping fault in the east of the rift, to a S-dipping, offshore fault in the west (Stefatos et al., 2002; Sachpazi et al., 2003; McNeill et al., 2005b). Such changes in the polarity of basin segments along strike in a rift system are also common in the East African rift (Rosendahl, 1987). Alternatively, slip could be occurring on a deep, low-angle fault, allowing for reconciliation with geodetic extension rates. Sorel (2000) suggests a low-angle, N-dipping detachment can be mapped on the south coast, a view that may or may not be supported by the location of microseismicity (Rigo et al., 1996; Gautier et al., 2006). The importance of high- and low-angle fault activity in producing the geometry of rift systems has been debated at other rift zones, such as the Basin and Range province and Papua New Guinea (Weissel et al., 1982; Taylor et al., 1999).
In this paper, we interpret the geometry of major offshore faults in the western Gulf and assess their contribution to synrift deformation, building on the work of Stefatos et al. (2002), Sachpazi et al. (2003), and McNeill et al. (2005b), and we incorporate current knowledge of onshore structures as described in the previous section. We will analyze western Gulf basin structure, estimate fault-slip rates, and investigate the evolution of the western rift in space and time. In this way, we hope to contribute to a general understanding of the relationship between individual faults and their impact on stratigraphy and help to address how common half-graben versus graben structures and high- versus low-angle faulting is in this and other young continental rifts. This will be achieved through the stratigraphic analysis of high-resolution, seismic reflection profiles.
The Gulf is connected to the Mediterranean by the Rion sill at its western end, which lies 60–70 m below present-day sea level (Perissoratis et al., 2000) (Fig. 1). This 2-km–wide erosional terrace is probably composed of Early Pleistocene or Pliocene strata (Chronis et al., 1991). Holocene deltaic sequences up to 15–20 m thick bank against the sides of the terrace (Perissoratis et al., 2000). The lack of recent fault offset on the terrace suggests that there has been very little change in its absolute depth, at least throughout the Late Pleistocene (Perissoratis et al., 2000). Stratigraphic evidence presented by McNeill et al. (2005), which has been developed by our study, supports a lack of vertical movement on the Rion sill. Based on sedimentology at the eastern end of the Gulf (Collier and Thompson, 1991) and the projection of Late Quaternary uplift rates, there has not been a marine connection during episodes of sea-level lowstand at the Corinth isthmus since at least 450–800 ka (uplift rates of 0.2–0.3 mm/yr from Collier et al., 1992; Leeder et al., 2005). As such, during recent sealevel lowstands, the Gulf has become an isolated lake, predominantly controlled by the level of the Rion sill. This oscillation between lake and open-ocean environment produces sedimentary cycles sensitive to sea-level fluctuation.
This variation in environment leads to a rich and detailed stratigraphic record of tectonism and climate variation. Coring in the central Gulf (Moretti et al., 2004) and Alkyonides Gulf (Collier et al., 2000) shows sediments were deposited in both marine and lacustrine environments, thus supporting a hypothesis of changing environment. The seismic stratigraphy of subsurface sediments within the central Gulf suggests a cyclicity in sediment properties controlled by 100-ka sea-level cycles (Perissoratis et al., 2000; Sachpazi et al., 2003; Leeder et al., 2005; Lykousis et al., 2007).
|
METHODOLOGY AND DATA
|
|---|
Seismic Reflection and Swath Bathymetry Data
High-resolution, multichannel seismic (MCS) reflection profiles and multibeam swath bathymetry were collected in the western Gulf on the MV Vassilios in July 2003 under the direction of the Universities of Southampton, Patras, and Leeds. Seismic reflection data were collected using a 150- to 2000-Hz sparker source and a 60-channel, 1-m group spacing streamer. Ten NS profiles were collected across the rift, together with a 40-km, E-W tie line (Fig. 2). Penetration down to 1–1.5 s two-way travel time (TWTT) allowed basement imaging, except in the deepest parts of the basin. This data set has been supplemented by published and publicly available data to investigate basement depth in the deep basin (e.g., Goodliffe et al., 2003; Sachpazi et al., 2003; Zelt et al., 2004). Swath bathymetry coverage is outlined in Figure 2 and was collected using a Reson Seabat 8160, 50-kHz, multibeam echo sounder.
View larger version (36K):
[in this window]
[in a new window]
|
Figure 2. Extent of the swath bathymetry (dotted) and seismic reflection profiles (solid lines) collected by the MV Vassilios in 2003 (McNeill et al., 2005b). Sedimentation-rate calculations at the labeled sites are given in Table 3. AIG—Aigion fault; WEF—West Eliki fault; EEF—East Eliki fault; DER—Derveni fault.
|
|
Depth Conversion
To make quantitative assessments of fault activity, we have depth converted some parts of the MCS data. Depth conversion was conducted based on a velocity profile compiled from typical interval velocities used during data stacking, velocity estimates for the Alkyonides Gulf (Collier et al., 2000), direct measurements from shallow sediments (Moretti et al., 2004), and generalized sediment curves from Hamilton (1979, 1980). Using our velocity curve, sediments with a TWTT of between 0–0.5 s and 0.5–1 s below the seafloor have estimated average velocities of 1.5–2.0 km/s and 2.0–2.5 km/s, respectively.
|
STRATIGRAPHIC ARCHITECTURE
|
|---|
To extract quantitative tectonic information, it is important to establish the stratigraphic framework and, where possible, estimate ages of horizons. The offshore stratigraphy exhibits a distinct character in the shelf and upper slope regions when compared to the deeper Gulf due to the more direct impact of sea-level fluctuations.
Shelf and Upper Slope Stratigraphic Architecture: The Eratini Sub-Basin
A local, fault-controlled depocenter, known as the Eratini sub-basin, was previously identified offshore Eratini (McNeill et al., 2005b; Lykousis et al., 2007; Fig. 3). The basement-sediment contact dips south at 15–18°, and the sediment wedge thickens toward the N-dipping North Eratini fault (NEF, after McNeill et al., 2005b). The sediments can be divided into two major packages separated by an unconformity (horizon U, Fig. 3B). We have identified five units within the well-stratified basin sediments above horizon U, based on reflection geometry and termination and likely sea-level control (units I to V, Fig. 3B and Table 1). Near the northern shallow margin of the sub-basin, we interpret four distinct clinoform (sloping depositional surface) packages, each located progressively northward (arrowed on Fig. 3B). The steeply dipping clino-forms within each package flatten basinward to become parallel to shallow-dipping, continuous reflections.
View larger version (47K):
[in this window]
[in a new window]
|
Figure 3. (A) MCS data from the Eratini sub-basin (location identified in Fig. 2). (B) Four clinoform packages (arrowed) are identified within the basin. The shallowest clinoform within each package marks the end lowstand shoreline, formed before the Rion sill flooded during marine transgression. Unnumbered lines represent interpreted marine to lacustrine transitions. (C) Horizons can be correlated with the eustatic sea-level curve of Siddall et al., (2003) to estimate horizon age. After McNeill et al. (2005b).
|
|
Similar sedimentary architecture was interpreted in this area by McNeill et al. (2005b) and Lykousis et al. (2007) and from other parts of the Gulf (Collier et al., 2000; Leeder et al., 2002; Leeder et al., 2005) and elsewhere in the world, e.g., Lake Edward (McGlue et al., 2006) and Lake Malawi (Scholz and Finney, 1994). Clinoform units are indicative of progradational shoreline deltas formed in response to a falling or static lowstand sea level, forcing the shoreline basinward (e.g., Vail et al., 1977; Gawthorpe et al., 1994; Leeder et al., 2002; Leeder et al., 2005). We interpret the clinoform units in this study as deltaic deposits formed when eustatic sea level dropped below the level of the Rion sill and the Gulf became an isolated lake (Perissoratis et al., 2000; Leeder et al., 2005; McNeill et al., 2005b; Lykousis et al., 2007). The youngest slope break of each clinoform package formed at the depth of the Rion sill.
The slope breaks of each of the preserved youngest delta surfaces (labeled horizons 1–4 in Fig. 3B) lay at average depths of 95, 160, 210, and 260 m below sea level. If the Rion sill has remained at a depth of 60–70 m below sea level (Perissoratis et al., 2000) and the Gulf has remained closed at the Corinth Isthmus during at least the last four lowstands (Collier and Thompson, 1991; Collier et al., 1992), the sill-flooding events may be correlated with a eustatic sea-level curve (e.g., Siddall et al., 2003, Fig. 3C), assuming that there has been no significant erosion. We estimate the ages of end lowstand shorelines, represented by horizons 1, 2, 3, and 4, at ca. 12, 130, 240, and 340 ka, respectively (Fig. 3C and Table 1). If our assumption that the sill has remained static is not correct, this will not greatly affect the age estimates because the eustatic sea-level curve is very steep during transgressions (Fig. 3C). Currently there are no core data in this area to verify our age assignment.
Unit I, between the sea bed and horizon 1, is composed of parallel and continuous reflections that have the same character throughout the sub-basin. The unit has a relatively constant thickness (7 m) and drapes preexisting topography (Table 1). Based on our correlation and that of McNeill et al. (2005b) and Lykousis et al. (2007), this unit is interpreted as Holocene marine sediments. The seismic character of the sediments in this unit is similar to the Holocene veneer seen elsewhere in the Gulf.
Based on our correlation, units II, III, and IV record sedimentation during both marine and lacustrine conditions as sea level fluctuated through 100-ka cycles. As an example, Figure 4 shows a close-up of stratigraphy within unit III. Below horizon 2 (ca. 130 ka), reflections have a low-medium amplitude and a well-developed clinoform character. The oldest clinoforms in the unit down-lap onto a high-amplitude horizon (dashed in Fig. 4B). Below this horizon, reflections show no evidence of clinoform character and generally have a higher amplitude than the clinoform reflections above (Fig. 4B). We suggest that these parallel sediments, extending to horizon 3 (ca. 240 ka), formed by marine aggradation during a sea-level high stand, analogous to the modern deposition of unit I (Fig. 3B).
View larger version (42K):
[in this window]
[in a new window]
|
Figure 4. (A) Close-up of the stratigraphy within the Eratini sub-basin (box in Fig. 3B). (B) Unit III shows a clinoform configuration associated with lacustrine delta progradation during sea-level lowstand. The parallel reflections at the base of the unit are probably highstand and transgressive sediments.
|
|
Maximum sedimentation rates can be estimated for units I to IV at the slope break position, and average values can be deduced for the distal basin, based on the interpreted stratigraphy (Table 1). Sedimentation rates for marine unit I are 0.5–0.6 mm/yr across the sub-basin. Units II to IV show similar sedimentation rates, with a maximum of 0.45 mm/yr and average values of 0.25 mm/yr away from clinoform packages.
Below units I to V, there is a change in seismic character at the horizon designated U (Fig. 3B). This horizon separates well-stratified, clinoform packages from a unit with few traceable reflections that extends down to basement (horizon U to basement in Fig. 3B). If we assume that average sedimentation rates have remained constant, we predict an age of ca. 0.4 Ma for horizon U and ca. 0.5 Ma for the oldest sediments in the basin.
Main Basin Stratigraphic Architecture
Stratigraphy in the main Gulf can be divided into two distinct seismic packages (A and B) (see Figs. 5 and 6). The upper package (A) is well stratified with parallel continuous reflection horizons that show cyclical seismic properties. The deeper package (B) contains few reflections that can be traced for any significant distance (Figs. 5B and 6B). The two packages are separated by an unconformity (horizon U), which is identified throughout the study area as a strongly reflective surface with some degree of angular truncation (Fig. 6B). The same surface was identified in the Eratini sub-basin (Fig. 3), and Sachpazi et al. (2003) have suggested a similar division of stratigraphy farther east within the central Gulf, but an exact correlation between the units in the different areas is unconfirmed.
View larger version (54K):
[in this window]
[in a new window]
|
Figure 5. (A) MCS data from the central western Gulf (location shown in Fig. 2) (B) Seismic stratigraphic interpretation showing the two main sediment packages, A and B, and horizon correlation with the sea-level curve of Siddall et al. (2003) modified for the level of the Rion sill.
|
|
View larger version (93K):
[in this window]
[in a new window]
|
Figure 6. (A) MCS data from the E-W Gulf profile (location shown in Fig. 2). (B) Seismic stratigraphic interpretation showing the two main sediment packages, A and B, and horizon correlation with the sea-level curve of Siddall et al. (2003) modified for the level of the Rion sill.
|
|
Package A contains four cyclical units (I to IV), which we propose can be related to 100-ka climate/sea level cycles (Figs. 5B and 6B). Unit I, between the sea bed and horizon 1, is well stratified with high-amplitude, low-frequency, continuous parallel reflections that diverge at topographic depressions and near growth faults (Figs. 5B and 6B; Table 2). Our E-W tie line crosses the Marion Dufresne long-piston core site, and horizon 1 can be correlated with the location of a lacustrine-marine boundary dated at ca. 12 ka (Moretti et al., 2004). Unit I has been deposited in post–ca. 12-ka marine conditions.
Reflections in the upper part of unit II have lower amplitudes and higher frequencies than those in unit I (Figs. 5B and 6B; Table 2). The shallowest sediments of unit II have been sampled (Moretti et al., 2004) and correspond to lacustrine sedimentation prior to Rion sill flooding ca. 12 ka. From the limited core data available, it appears that a change in seismic character from high amplitude/low frequency in unit I, to low amplitude/high frequency in unit II, reflects a change in Gulf conditions from marine (high-stand) to lacustrine (lowstand). This seismic character variation has also been identified by Perissoratis et al. (2000), who postulate that during glacial lowstands the frequency of turbidite deposition may increase in the main Gulf due to the emergence of unstable margin slopes.
These two seismic facies types can be distinguished in older, underlying units. We interpret horizons 2 and 3 as transitions between lacustrine and marine conditions (Figs. 5B and 6B). These horizons have been correlated with the eustatic sea-level curve of Siddall et al. (2003), yielding age estimates of ca. 130 and 240 ka for horizons 2 and 3, respectively (Table 2). decreasing to 0.5–0.55 mm/yr in the center of the study area. For the former, values are within the range of rates suggested elsewhere in the central Gulf over the last ca. 20 ka (1.2 mm/yr, Moretti et al., 2004). The sedimentation rates estimated in this study are therefore realistic (giving increased confidence in age assignment to horizons). Extending these sedimentation rates gives age estimates of ca. 0.3–0.42 Ma for the unconformity surface, horizon U (Figs. 5B and 6B; Table 3).
We estimated sedimentation rates within the main Gulf for undisturbed sediments away from major faults, over different time periods, based on our age estimates (Table 3). Weighted average sedimentation rates of 1.0–1.2 mm/yr are obtained in the eastern part of the western Gulf, Very few traceable reflectors are observed in package B, and all have low amplitudes when compared to package A. Especially in the eastern study area, the unit is partly below the depth of seismic recording, but published data have been integrated into this study, thus allowing further interpretation (e.g., Goodliffe et al., 2003; Sach-pazi et al., 2003; Clement et al., 2004; Zelt et al., 2004). Analysis of published data and extrapolation of our average sedimentation rates yield an age for the oldest sediments within the deepest western basin of ca. 1.5 Ma.
Correlation between Eratini Sub-Basin and Main Gulf Stratigraphic Units
Sedimentation and seismic stratigraphy in the marginal Eratini sub-basin and in the main Gulf are both controlled by Quaternary sealevel (climatic) variations (Tables 1 and 2). The unconformity between well-stratified and poorly stratified sediments is seen in both the stratigraphy of the northern Eratini sub-basin and in the main Gulf. Extrapolating estimated sedimentation rates in both environments yields a similar age of the unconformity, horizon U, of ca. 0.4 Ma. The age of the basement-sediment contact in the deepest parts of the Gulf is estimated at ca. 1.5 Ma, which contrasts with the younger age (ca. 0.5 Ma) we determined for the Eratini sub-basin and elsewhere on the northern margin in general (Table 3).
|
BASIN GEOMETRY
|
|---|
Basement Structure
Interpretation of the MCS and swath bathymetry data has led us to identify five dominant active offshore faults—the South Eratini, North Eratini, East Channel, West Channel, and Akrata faults (Fig. 7; see also Stefatos et al., 2002; McNeill et al., 2005b; Lykousis et al., 2007). These faults, with throws of >0.5 s TWTT (>450 m), are seen to dissect basement and are responsible for the observed basin geometry, together with the East and West Eliki, Derveni, and Aigion faults along the southern coastline. Figure 7 also shows other major faults that are buried and no longer active or have throws <0.5 s TWTT. Minor faults that do not displace basement, have limited or no surface expression, cannot be traced between profiles due to lengths <5 km, or have unfavorable trends, are not shown on Figure 7. Basement is clearly imaged in the seismic data of this study down to 1.5 s TWTT.
View larger version (40K):
[in this window]
[in a new window]
|
Figure 7. Interpreted major faults in the western Gulf of Corinth. Positions of faults are constrained by the seismic profiles and swath bathymetry used in this study and that by McNeill et al. (2005b), and are supplemented by interpretations from Zelt et al. (2004) and Goodliffe et al. (2003). AIG—Aigion fault; WEF—West Eliki fault; EEF—East Eliki fault; DER—Derveni fault; WCF—West Channel fault; ECF—East channel fault; SEF—South Eratini fault; NEF—North Eratini fault; AKR—Akrata fault.
|
|
Other published and publicly available data sets (Zelt et al., 2004; seismic images from the R/V Maurice Ewing cruise EW0108/2001 accessed through the Marine Seismic Data Center at http://www.ig.utexas.edu/sdc/) were incorporated to produce a time-structure contour map of the sediment-basement contact and an isopach of the total sediment thickness (Figs. 8A and 8B, respectively). The basement structure of the western Gulf can be broadly separated into a major depocenter to the south (>0.8 s below the seafloor, containing up to 2 km of sediments, Fig. 8B) and a shallower region to the north, controlled by the traces of the West Channel and East Channel faults. The basement-sediment contact in the hanging wall of the S-dipping West Channel fault-East Channel fault gradually increases in depth to the east, reaching 2.8 km, with the surface dipping 12–18° N toward the fault planes (Fig. 8A). The basement surface between the East Channel fault and the Akrata fault deepens adjacent to both fault structures with a central structural high (Fig. 8A–B).
View larger version (46K):
[in this window]
[in a new window]
|
Figure 8. (A) TWTT depth to the basement-sediment contact and (B) TWTT sediment thickness isopach. Produced using the basement interpretations of our MCS study and additional basement interpretations from deep seismic reflection profiles from the R/V Maurice Ewing cruise EW0108/2001 (Zelt et al., 2004; seismic images accessed through the Marine Seismic Data Center at http://www.ig.utexas.edu/sdc/).
|
|
North of the West Channel-East Channel fault trace, basement depth is typically shallower than 0.4 s TWTT (0.45 km) below the seafloor (Fig. 8B), with localized fault-controlled highs and lows. The region between the eastern tip of the West Channel fault and western tip of the East Channel fault has been warped into a relay ramp structure. Tie lines crossing this region suggest that the relay ramp has been breached by a connecting fault that now links the West Channel fault and East Channel fault (Fig. 8 and seen in Fig. 6), but the precise orientation of this fault cannot be resolved.
The North Eratini fault has caused subsidence and 8° tilting of the basement toward the south and production of the Eratini sub-basin against the uplifted horst footwall block (Fig. 8 and shown in Fig. 3). North of the Eratini sub-basin, on the shelf, the basement platform is subhorizontal at a depth of 100 m. Activity in the hanging wall of the South Eratini fault has resulted in tilting and deepening of the basement surface northward. Thus, the western Gulf of Corinth basement topography is highly variable and is clearly controlled by major active normal faults.
Fault Architecture
The North Eratini fault and South Eratini fault have a similar length of 15 km and overlap completely, producing a prominent topographic basement horst between the fault traces (Figs. 7, 8, 9, and 10). The North Eratini fault dips N at 60° (producing the Eratini sub-basin in its hanging wall) and the South Eratini fault dips S at 50–60°, exposing steep basement fault scarps that are visible in the bathymetric data (Fig. 7). Toward the western fault tips, the basement horst is buried (Fig. 9), and it reaches a minimum depth below sea level of 150 m in the fault center (Fig. 10).
View larger version (52K):
[in this window]
[in a new window]
|
Figure 9. (A) MCS data from the west of the study area (location in Fig. 7). (B) Structural interpretation of the sub-bottom structure and inset line drawing showing basement depth for this location from Figure 8.
|
|
View larger version (57K):
[in this window]
[in a new window]
|
Figure 10. (A) MCS data from the center of the study area (location in Fig. 7). (B) Structural interpretation of the sub-bottom structure and inset line drawing showing basement depth for this location from Figure 8.
|
|
To the east, the East Channel fault is observed in the seismic profiles as a basement-displacing fault that creates a sediment sea-bed scarp (Figs. 8 and 11). The fault dips 60–70° S in shallow sediments decreasing to 45–55° in basement (data from Goodliffe et al., 2003; Zelt et al., 2004). The East Channel fault overlaps the eastern tip of the South Eratini fault by 5 km (Figs. 7 and 8).
View larger version (54K):
[in this window]
[in a new window]
|
Figure 11. (A) MCS data from the east of the study area (location in Fig. 7). (B) Structural interpretation of the sub-bottom structure and inset line drawing showing basement depth for this location from Figure 8.
|
|
The West Channel fault (equivalent to the Sub Channel Fault of McNeill et al., 2005b) overlaps the South Eratini and North Eratini faults, lies 3 km south of the South Eratini fault, and dips S at 45–60° within shallow sediments (Fig. 10). This fault controls the position of the central axial channel for at least 15 km and overlaps the tip of the East Channel fault by 2 km (Fig. 7). This fault has also been interpreted by Stefatos et al. (2002) and is clearly visible with significant basement offset in deeper seismic profiles (e.g., Goodliffe et al., 2003; Zelt et al., 2004; Fig. 8).
The Akrata fault (also interpreted by Stefatos et al., 2002) produces steep sediment slopes along the southern margin of the Gulf and causes significant basement offset for 15 km (Figs. 8 and 11, supplemented by data from Goodliffe et al., 2003; Zelt et al., 2004). The Akrata fault dips N at 60° and lies 4 km north of, and may be a splay of the major East Eliki fault.
In addition to these major, active normal faults, a number of more minor, basement and seafloor-displacing faults also exist. West of Eratini, short (ca. 5-km) fault segments dissect the basement into three horst blocks and graben, trending roughly NE-SW (Figs. 7 and 9). The Trizonia subaerial horst block (Fig. 1) is probably one of these local horst-graben systems. Just south of the Galaxidi peninsula, a series of 3- to 10-km–long faults offset the basement surface (Fig. 7 and N end of Fig. 11).
Slip Rates
North Eratini Fault
The stratigraphy and preservation of four, lowstand, clinoform deltas within the Eratini sub-basin (Figs. 3 and 4) is a function of the balance between sediment influx, eustatic and local sea-level fluctuation, and tectonic activity (namely fault-controlled subsidence). The slope-break positions of the clinoform units (e.g., Fig. 3) are useful because they are believed to form in 0–10 m of water (M. Leeder, 2006, personal commun.) and are thus good paleosea-level indicators. Figure 12A shows the average depths to each of the four lowstand delta slope breaks in Package A (Figs. 3 and 13A), and the vertical subsidence for each of the horizons is given in Table 4 (assuming lowstand level was controlled by a relatively static Rion sill). The subsidence at the fault plane itself, rather than at the slope-break positions (2.5 km from the fault), is thought to be 10% higher, given the best-fit displacement decay curves of Armijo et al. (1996), and we have used this value in our calculations. To estimate the throw on the fault from the subsidence component, we can apply an uplift: subsidence ratio to determine total slip rate. Applying the methodology of McNeill et al. (2005b), and based on a review of the available literature and long-term uplift: subsidence estimates in the Gulf and measured fault dip, we have used a ratio of 1:1.2–2.2. This yields a late Quaternary to Holocene slip rate on the North Eratini fault of 0.9–1.8 mm/yr and 2–6.7 mm/yr for the Holocene (Table 4 and Fig. 13B). Due to the young age (ca. 12 ka) and shallow depth of C1, the uncertainty in the slip-rate calculation is higher than for the other shorelines, yet we are confident that subsidence has significantly accelerated in the Holocene.
View larger version (25K):
[in this window]
[in a new window]
|
Figure 12. (A) Comparison of the average depth of buried clinoform slope breaks, C1–C4 (Fig. 3) with the eustatic sea-level curve of Siddall et al. (2003). (B) Subsidence rates at the fault plane averaged within each time step between shoreline formation. Results suggest that subsidence was relatively constant within the Late Pleistocene but has increased within the Holocene, assuming sill depth, lowstand level, and water depth of formation are approximately constant and accurate.
|
|
View larger version (30K):
[in this window]
[in a new window]
|
Figure 13. (A) Subsided shoreline, total basement throw, and stratigraphic offset measurement methods used in the determination of estimated slip rates for different time periods. (B) Summary of the estimated slip rates, in mm/yr, for each fault over different time periods, using the three measurement methods described in (A). Slip rates in white boxes have the highest confidence level; those in light-gray boxes are associated with uncertainty in unknown paleotopography, and those in black have been determined using the interpolated basement structure of Figure 8.
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 4. SLIP RATES FOR THE NORTH ERATINI FAULT DERIVED FROM SUBSIDENCE OF PALEOSHORELIONES IN THE ERATINI SUB-BASIN
|
|
The apparent basement throw of the North Eratini fault provides an additional slip-rate estimation (averaged over 0.5 m.y.), which is a minimum value due to the possibility of erosion of the top of the horst block (Figs. 10 and 13A). Total basement offset of 400 m yields a minimum Quaternary averaged slip of 0.9 mm/yr for the center of the North Eratini fault, similar to the rate derived from clinoform subsidence (Fig. 13B).
South Eratini Fault
No horizons can be correlated between the South Eratini fault hanging wall and footwall (Figs. 10 and 13A). We can only obtain a minimum estimate of average Quaternary fault slip based on the basement throw (Fig. 13A). At the center of the fault, total basement offset is 630 m (Fig. 10), which gives a minimum slip rate of 1.4 mm/yr.
West Channel Fault
The seismic data show that footwall and hanging wall stratigraphy can be traced across the West Channel fault; however, paleotopography at the time of sediment deposition is uncertain (Fig. 10). A significant seafloor scarp related mostly to the axial channel exists today, and it is unlikely that post–ca. 0.4-Ma horizons were continuous across the fault during their initial deposition. Slip rates calculated based on the offset of these horizons (Fig. 13A) are therefore probably a maximum, because some contribution may be due to paleotopography. Average maximum slip rates derived from offset of the ca. 0.4-Ma unconformity and ca. 0.25-Ma horizon are 0.4 mm/yr and 0.5 mm/yr, respectively (Fig. 13B). Slip rates, however, between ca. 0.4 Ma and ca. 0.25 Ma are significantly reduced at 0.15 mm/yr. Little sediment thickening of post–ca. 0.4-Ma stratigraphy occurs across the West Channel fault, suggesting low fault activity since then. Paleotopography created pre–ca. 0.4-Ma results in larger than expected horizon offsets and misleadingly high slip rates when total offset on the ca. 0.4- and ca. 0.25-Ma horizons is considered. Total-basement offset for the West Channel fault yields an average slip rate since ca. 1.5 Ma of 0.6 mm/yr; however, we recognize this value is subject to large uncertainties in basement throw (Fig. 13B). Given the measured slip rates, we can calculate an estimate of the ca. 1.5- to 0.4-Ma slip rate of 0.7 mm/y (Fig. 13B).
East Channel Fault
Although footwall and hanging wall stratigraphy is preserved across the East Channel fault, a scarp of up to 300 m exists in the modern seafloor (Fig. 11). This paleotopography probably existed throughout post–ca. 0.4-Ma deposition, and as such, meaningful slip rates cannot be derived. In a qualitative sense, the thickening of package A is much smaller than that of package B across the East Channel fault (Fig. 11), indicating a decrease in slip rate for the period 0 to ca. 0.4 Ma compared to ca. 1.5 to ca. 0.4 Ma.
Estimates of total basement throw yield average slip rates in the last ca. 1.5 Ma up to 1.2 mm/yr (Fig. 13B). The greater total basement subsidence and higher overall slip rate indicate the East Channel fault is more significant than the West Channel fault.
|
BASIN EVOLUTION
|
|---|
The structure of the western offshore Gulf of Corinth basin appears to be distinct in the east, west, and center, and we will discuss rift evolution in the context of these three regions (Fig. 14). Area 1, to the east, is dominated by two depocenters, one in the hanging wall of the East Channel fault and the other adjacent to the Akrata fault (Fig. 8B). The second region (central, area 2), is more structurally complex with highly variable seafloor and basement topography. In area 3 to the west, the subsurface is broken into a series of NE-SW–trending horst and graben. Figure 14 shows the locations of these areas and gives interpretations of the full MCS data set.
View larger version (24K):
[in this window]
[in a new window]
|
Figure 14. (A–J) Stratigraphic interpretations for the MCS data set in the western Gulf of Corinth. Basement depths are observed directly from the MCS data or derived from the regional grid in Figure 8A (see Basement Structure section for method).
|
|
Eastern Study Area: Area 1
Basin Deformation: Circa 1.5–0.4 Ma (Package B)
There is considerable variation in pre–ca. 0.4-Ma sediment thickness (package B) within area 1 (Fig. 14–C). The thickness of ca. 1.5- to ca. 0.4-Ma sediments increases five-fold between the footwall and hanging wall of the East Channel fault, with the greatest thickness occurring at the east end of area 1 reaching >1250 m (1s TWTT) (Fig. 14A–B). Sediment thickness suggests this fault continues farther east beyond the study area. The onshore East Eliki fault and offshore Akrata fault have together amassed a package B thickness of 900 m (0.75 s TWTT) in their hanging walls (Fig. 14B), and the Akrata fault may be a splay of the East Eliki fault. Package B contains a limited number of traceable, coherent reflectors, but the general northward dip of observable horizons and sediment thickening toward the north (Fig. 14A–B) supports the fact that the offshore, S-dipping East Channel fault is significant at this time and location.
Basin Deformation: Circa 0.4–0 Ma (Package A)
Package A in this area shows less thickness variation than package B. North of Akrata (Figs. 14B–C and 15A) thickening and tilting of post–ca. 0.4-Ma sediments is consistently toward the south. At this location, the N-dipping East Eliki fault on the south margin of the Gulf and the Akrata fault control overall southward tilting and sedimentation post-ca. 0.4 Ma.
View larger version (48K):
[in this window]
[in a new window]
|
Figure 15. (A) Uninterpreted and interpreted MCS data for the boxed region in Figure 14B. Post–ca. 0.4-Ma sediments show a southward tilt toward the East Eliki fault and Akrata fault at the south end of the profile. (B) Uninterpreted and interpreted MCS data for the boxed region in Figure 14F. Sediments to the south show a preferred southward tilt toward the onshore East Eliki fault, whereas those to the north dip toward the S-dipping West Channel fault. Intense minor faulting exists between sediments dipping with a preferred north and south orientation. (C) Uninterpreted and interpreted view of the East Channel fault from the central western gulf (Fig. 14D). The East Channel fault is a complex fault zone with activity pre-ca. 125 ka (horizon 2) being concentrated on a buried fault to the south of the East Channel fault. Note exaggerated vertical scale compared to Figures 15A and 15B.
|
|
Sediment thickness change across the East Channel fault is much less exaggerated than in the period ca. 1.5 to ca. 0.4 Ma. Although thickening occurs across the East Channel fault, into its hanging wall (Fig. 14A–C), and Holocene deformation proves the structure still active, south coast faults (East Eliki fault and Derveni fault?) dominate post-ca. 0.4 Ma, and activity on the East Channel fault has decreased.
Central Study Area: Area 2
Basin Deformation: Circa 1.5–0.4 Ma (Package B)
The greatest thickness of pre–ca. 0.4-Ma sediments within area 2 is seen in the hanging wall of the West Channel fault with a maximum thickness of 850 m (0.7 s TWTT), significantly thinner than was observed farther east (area 1) (Fig. 14D–F). The basement-sediment contact in the hanging wall of the West Channel fault and reflections within package B show a dominant tilt north. Package B sediments in the footwall of the West Channel fault and within the hanging-wall region of the South Eratini and North Eratini faults are much thinner, at 180 m (0.15 s TWTT). This suggests that in the time period between the initiation of offshore synrift sedimentation and ca. 0.4 Ma, the dominant structure was the West Channel fault. Its dominance over N-dipping south coast faults, during at least some part of the deposition of package B, since we do not image package B in its entirety, is clear from the N-tilted sediments and half-graben geometry (Fig. 14D–F).
Basin Deformation: Circa 0.4–0 Ma (Package A)
Stratigraphy at the southern end of profiles in area 2 thickens and tilts south, probably due to the activity of the onshore, N-dipping East Eliki fault; this is most obvious in profile Figure 14F and 15B. Package A sediments within area 2 reach thicknesses of 350 m (0.3 s TWTT). Sediments clearly tilt northward toward the West Channel fault, but stratigraphic thickening is minimal due to the axial channel location, which disrupts sediment deposition patterns. Between the West Channel fault and the area more greatly influenced by the East Eliki fault to the south, is a zone of minor faulting (Fig. 15B), which can be traced over a distance of 10 km E-W.
The East Channel fault, which dominated stratigraphy within area 1, dies out westward within area 2 (Fig. 7). The data suggest the East Channel fault becomes a complex zone at its western tip (shown in Fig. 15C). Package A sediments in the hanging wall of the central South Eratini fault tilt and thicken to the north (Fig. 14F), but farther east (Fig. 14D), the sediments reverse their tilt and dip south, suggesting a decrease in activity on the South Eratini fault or an increasing dominance of N-dipping faults to the south. Modern to ca. 0.4-Ma sediments within the Eratini sub-basin tilt and thicken toward the North Eratini fault with local drag folding (Fig. 3).
Western Study Area: Area 3
Basin Deformation: Circa 1.5–0.4 Ma (Package B)
Package B sediments in the far west (area 3) are significantly thinner than to the east (areas 1 and 2). Sediment thicknesses of up to 600 m (0.5s TWTT) are estimated within the West Channel fault hanging wall, but elsewhere thicknesses of <80 m (0.07s TWTT) are observed (Fig. 14G–I). North of the West Channel fault trace, very little sediment is deposited, suggesting minimal subsidence.
Basin Deformation: Circa 0.4–0 Ma (Package A)
The thickness of package A is very similar to that farther east (areas 1 and 2) reaching 350 m (0.3 s TWTT) (Fig. 14G–I). As well as in the hanging wall of the West Channel fault, deposition occurs within two main symmetrical graben, controlled by the western tips of the South Eratini fault and North Eratini fault (and associated splays), together with a number of shorter basement-offsetting faults (Fig. 7). Significant sediment deposition occurs within these graben, which were not active prior to ca. 0.4 Ma. A number of significant Holocene debris flow events are seen within package A in area 3.
|
DISCUSSION
|
|---|
Offshore Western Gulf Fault Activity
Our analysis has led to a quantitative understanding of the timing and magnitude of activity on offshore western Gulf faults. Based on the stratigraphic age and sedimentation rate interpretations discussed earlier in this paper, the West Channel and East Channel faults began accumulating synrift sediments ca. 1.5 Ma, while the South Eratini and North Eratini faults became active much later at ca. 0.5 Ma. From ca. 1.5 to ca 0.5 Ma, West Channel and East Channel faults were the dominant offshore faults, having individual slip rates of 0.6–1.2 mm/yr averaged over this period, and they produced deep subsided depocenters in what is now the center of the western Gulf basin. Their activity during this early stage and locally greater influence than N-dipping south coast faults are clear in the northward basement tilt and northward sediment thickening (Fig. 8). Slip rates on the West Channel fault decrease toward its western tip, with reduced subsidence in the western area 3.
Since the activation of North Eratini and South Eratini faults at ca. 0.5 Ma, slip rates have decreased on the West Channel-East Channel fault system (0.2 mm/yr for the West Channel fault, Fig. 13B). Minimum Late Pleistocene slip rates of 1–1.5 mm/yr are predicted for the North Eratini and South Eratini faults, with slip rates increasing up to 5.5 mm/yr in the Holocene. Post–ca. 0.4-Ma stratigraphy at the east end of the study area shows that N-dipping faults on the southern margin (East Eliki and Akrata faults) became relatively more dominant at ca. 0.4 Ma, resulting in southward tilting and fanning of sediments. The present overall graben geometry of this eastern area results from the net effect of alternating activity on these faults (similar to interpretations by Sachpazi et al. (2003) in the central Gulf). The central part of the study area is only weakly influenced by south margin faults post-ca. 0.4 Ma, and the S-dipping South Eratini fault dominates in maintaining an N-tilted half graben, despite the reduction in activity of the West Channel fault. In the west, the growth of the South Eratini and North Eratini faults and other basement displacing faults has produced large post–ca. 0.4-Ma accumulations of sediment.
Western Gulf of Corinth Rift Evolution
Various tectonic models have been proposed to explain the evolution of the western Gulf of Corinth based on the south coast geology and stratigraphy (e.g., Ori, 1989; Doutsos and Piper, 1990; Sorel, 2000; Rohais et al., 2007). We attempt to incorporate our findings into the framework of these models and move toward a complete scenario for rift evolution. Figure 16 presents schematic cross sections of the dominant faults during different time periods.
View larger version (28K):
[in this window]
[in a new window]
|
Figure 16. Maps to show major active faults and possible positions of the coastline during the development of the western Gulf of Corinth, together with schematic cross sections. (A) Pre-ca. 2 Ma, a number of faults were active leading to the subsidence of a wide basin with unknown northern extent. (B) During the Early Pleistocene, the rift was controlled primarily by the Mamousia-Pirgaki fault in the south and the West Channel-East Channel fault system to the north. (C) Within the last ca. 1 Ma, activity has become focused on the Aigion fault/East Eliki fault along the southern margin and ca. 0.5-Ma activity on the northern margin transferred onto the South Eratini/North Eratini faults. South coast rift geometry after Collier and Jones (2003).
|
|
The earliest onshore synrift sediments from the western Gulf of Corinth have not been dated, but they are thought to be similar in age to the oldest sediments found in the Corinth area, ca. 3.6–4 Ma (Collier and Dart, 1991; Rohais et al., 2007). Geological mapping and onshore stratigraphic analysis suggest that the initial rift was wide, with activity distributed over a number of faults including the Kalamita, Kerpini, and Dhoumena faults in the Eliki region (Collier and Jones, 2003; Rohais et al., 2007). The location of the northern boundary of the rift in the Pliocene (>2 Ma) is unknown, but sediments during this time were deposited under alluvial to shallow water lacustrine conditions (Collier and Jones, 2003; Rohais et al., 2007; Fig. 16A).
This period of distributed extension was terminated by the focusing of activity on the Mamousia-Pirgaki fault and deepening of the rift allowing deposition of giant Gilbert fan deltas (Fig. 16B; Dart et al., 1994; Collier and Jones 2003; Rohais et al., 2007) in the Early Pleistocene, ca. 0.8–1.8 Ma (Symeonidis et al., 1987). We have demonstrated that the offshore West Channel-East Channel faults began accumulating synrift sediment in Early Pleistocene time (ca. 1.5 Ma); therefore, these S-dipping faults may have formed the northern boundary of the Pleistocene rift at this time (Fig. 16B). The focusing of extension resulted in a narrowing of the basin compared to the pre–ca. 2-Ma structure.
Extension transferred north onto the Eliki fault systems and other modern south shore faults in the Early-Middle Pleistocene, leaving the Mamousia-Pirgaki fault relatively inactive (Fig. 16C). The magnitude of Eliki footwall uplift (800–1000 m) and Late Pleistocene average uplift rates (1–1.5 mm/yr; McNeill and Collier, 2004) suggest this fault initiated between 0.7 and 1 Ma. Along the northern margin, ca. 0.5-Ma activity transferred from the West Channel-East Channel fault system northward to the South Eratini fault and North Eratini fault (Fig. 16C); since 0.2–0.3 Ma, the Aigion fault has initiated, continuing the pattern of northward fault propagation (De Martini et al., 2004; McNeill et al., 2007).
Across a profile 22°15' E, there are four major faults—the East Eliki fault, West Channel fault, South Eratini fault, and North Eratini fault. Horizontal components of extension for the East Eliki fault are 2–4 mm/yr (McNeill et al., 2005b) and have been calculated in this paper to be 1–3.5 mm/yr for the North Eratini fault during the Holocene. Recent rates on the South Eratini fault and West Channel fault cannot be calculated directly from stratigraphy, but if we assume similar values, the summed horizontal extension component across the western Gulf is 5–14.5 mm/yr. These calculations cover the range of recent geodetic measurements of extension of 10–15 mm/yr and suggest that geological and geodetic rates of deformation can be reconciled with the major faults identified.
Total Gulf of Corinth Structure and Evolution
Comparison of our interpretations with those in the central and eastern Gulf of Corinth highlights important structural and evolutionary differences. In the central Gulf, the early structure was controlled by S-dipping faults (Stefatos et al., 2002; Sachpazi et al., 2003, also imaged in our Fig. 14B). Later, the polarity of the deformation reversed; thus activity focused on N-dipping faults resulting in sediment thickening to the south. Overall, the net basin geometry caused by these two facing, subsiding, half grabens with differing border-fault polarity resembles a symmetric graben. Deformation in the eastern Gulf of Corinth and the Alkyonides Gulf has been controlled by N-dipping faults, thus creating a half graben that shows sediment thickening to the south, throughout the basin history (Stefatos et al., 2002; Leeder et al., 2005) (Fig. 16B). On a length scale of 20 km, the east, central, and west parts of the Gulf of Corinth have developed distinctly throughout rift history.
Faulting at Depth
We have discussed the evolution of the western Gulf rift as observed from the interpretation of onshore geology and offshore basin fill. We have not directly imaged faults below 1.5 km; however, our results can contribute to the frequently debated subject of high- versus low-angle faulting in young continental rifts (Rosendahl, 1987; Taylor et al., 1999). In general, there is limited information on the geometry of Gulf of Corinth faults at depth, and we attempt to combine the various published ideas in the following discussion using Figure 17.
View larger version (27K):
[in this window]
[in a new window]
|
Figure 17. Summary of evidence concerning the geometry and interaction of western Gulf of Corinth faults at depth. Inset shows the position of the transect and locations of Ms >5 seismicity in the area from Ambraseys and Jackson (1990) and Bernard et al. (1997). Faults observed on the southern coastline are from Collier and Jones (2003) and offshore faults from this study. Faults have been projected with dips of 60° down to the probable depth of the brittle-ductile transition as dashed lines. The position of the detachment surface proposed by Sorel (2000) and Rigo et al. (1996) have been projected onto this line. Note that these are distinct features at very different depths. The best-fit modeled fault plane for the 1995 Aigion earthquake from Bernard et al. (1997) is shown.
|
|
Sorel et al. (2000) suggest there is a low-angle detachment surface initiating at the Khelmos fault in the Western Gulf; this detachment projects northward into the offshore basin at 2 km depth (Fig. 17). Collier and Jones (2003) suggest this low-angle fault is actually due to rotation of an initially high-angle fault. They find detailed mapped fault patterns are inconsistent with propagation along a low-angle detachment (Figs. 16A and 16B and discussion above). Taylor et al. (2003) suggest a 10–20° detachment surface offshore at 2.5–4 s TWTT in seismic reflection data. This 3-km detachment corresponds roughly to the sediment-basement contact in the western Gulf; however, a convincing detachment surface at this depth is not clear in this data set where available. It is unknown whether this detachment could link to the unconfirmed Sorel (2000) detachment; however, Figure 17 suggests they may be at different depths.
Low-angle faulting has also been suggested from
TOP
ABSTRACT
INTRODUCTION
