Architectural Changes and Preferential Sand Deposition in a Confined Channel–Levee System Forced to Surmount a Ridge Crest
- File PDFChapter PDF
Efthymios K. Tripsanas, Eddy Lee, Willem Hack, R. Craig Shipp, 2012. "Architectural Changes and Preferential Sand Deposition in a Confined Channel–Levee System Forced to Surmount a Ridge Crest", Application of the Principles of Seismic Geomorphology to Continental Slope and Base-of-Slope Systems: Case Studies from SeaFloor and Near-Sea Floor Analogues, Bradford E. Prather, Mark E. Deptuck, David Mohrig, Berend Van Hoorn, Russell B. Wynn
Download citation file:
A buried Quaternary channel–levee system (CLS) with unique architectural characteristics was identified and studied through 3-D seismic-reflection data on the lower continental slope of the NW Niger Delta. The CLS runs in an elongated topographic depression, created by the edges of two thick (100 m) mass-transport deposits (MTDs). Because of this confinement, the CLS is forced to run over a ridge 20–70 m tall. More than 3.5 km updip from the ridge, the CLS is ∼ 4 km wide, and displays well-developed outer levees and multiple, strongly meandering channel forms (up to 100 m deep). The depth of the channel forms suggests a minimum thickness of 100 m for the largest flows. An erosional valley 3.5 km long and 2 km wide that cuts 50 m down to the substrate develops immediately updip and downdip from the ridge. A plateau 6 km long and 2 km wide is observed at the top of the ridge. On the plateau, the CLS is ∼ 4 km wide, and consists of thin levees and multiple, strongly meandering channel forms. Packets of high-amplitude reflections (HARs) are widespread at the base of these channel forms, and much thicker (up to 80 m) compared to the HAR packets observed in the CLS updip and downdip from the ridge (up to 40 m thick). Gamma-ray logs indicate that these HARs represent thick channel-sand deposits. These observations indicate that updip from the ridge, turbidity currents went through a hydraulic jump and developed a turbulent bore, which prevented deposition and enhanced erosion. The flows were thinned and spread on the plateau, resulting in extensive sand deposition. On the steep downdip flanks of the ridge, the flows accelerated and became erosional. This interpretation is consistent with flume studies showing that flows are able to surmount and transfer their suspended material over an obstruction with relief less than half the height of the flows.
Sand deposition in channel–levee systems (CLSs) has long been in the focus of research due to their potential to act as excellent oil reservoirs (Posamentier and Kolla, 2003). Significant progress has been made in the last decade in the understanding of their depositional elements through the acquisition and interpretation of 3-D seismic-reflection data (Deptuck et al., 2003; Deptuck et al., 2007; Wynn et al., 2007).
Morphologic irregularities along the thalweg of CLSs are common, and include “knickpoints”, steep steps along the thalwegs of the channel (Pirmez et al., 2000; Kneller, 2003; Heiniö and Davies, 2007). The general behavior of gravity flows on these knickpoints is upstream enhanced erosion and downstream enhanced deposition, in an effort for the CLSs to achieve an equilibrium profile (Pirmez et al., 2000; Ferry et al., 2005).
Mass-transport deposits (MTDs) (e.g., Hubbard et al., 2009), fold-and-thrust belts (e.g., Heiniö and Davies, 2007), and diapirs (e.g., Gee and Gawthorpe, 2006) are dynamic forms of morphologic irregularities, which occur before, during, and after the formation of CLSs and produce obstructions along the pathway of gravity flows. Upon impact on an obstruction, flows can either bypass it, be diverted around it to produce avulsion points, or become ponded and trapped, depending on the height of the obstruction and the rheological properties (grain size, density, velocity, thickness) of the flows (Kneller, 1995; Bursik and Woods, 2000; Kneller and Buckee, 2000; Kneller, 2003; Al Ja’Aidi et al., 2004; Lamb et al., 2006; Khan and Imran, 2008).
Despite the significant effect of obstructions on the rheology of gravity flows and in extension to the depositional elements of CLSs, there have been very few case studies dealing with this issue (e.g., thrust folds in Niger Delta by Heiniö and Davies, 2007; diapirs in Ferry et al., 2005, and Gee and Gawthorpe, 2006). This paper is focused on the study of a buried Quaternary CLS on the lower continental slope of the western Niger Delta (Figs. 1, 2). This CLS develops in the topographic depression between two complex MTDs, and, downdip, climbs over a ridge, which is formed by the B thrust fold. This peculiar setting offers a rare opportunity to study the effects of pre-existing morphology on the construction of CLSs and to infer changes in the rheological regime of turbidity currents as they encounter a morphologic obstruction. Important conclusions on sand deposition along channel pathways are also anticipated to be drawn through this study.
Geologic Setting and Database
The Niger Delta is located in western equatorial Africa (Fig. 1), covers an area of 140,000 km3, and is up to 12 km thick (Damuth, 1994). The delta is currently undergoing gravitational collapse, which is characterized by the presence of three structural domains: (1) an extensional domain dominated by growth faults on the upper continental slope, (2) a translational domain characterized by mud diapirism on the middle continental slope, and (3) a compressional domain consisting of fold-and-thrust belts on the lower continental slope (Morley and Guerin, 1996; Morgan, 2004; Heiniö and Davies, 2007).
The study area is located on the lower continental slope of the Niger Delta at a water depth of 3200–3600 mbsl. Two thrust folds are recognized in this area, the A and B thrust folds (Fig. 1C). The A thrust fold is visible on the present seafloor in the form of a distinct ridge, whereas the seafloor around the B thrust fold is completely healed. The 250 CLS described in this paper has a NE– SW orientation and is covered by 250–300 m of sediment (Figs. 2, 3, 4). The exact age of the 250 CLS is not known, and is assumed to have been formed during the Quaternary.
Reflection-coefficient 3-D seismic-reflection data used in this study cover an area of 332 km2. The data are sampled at 1 ms intervals, with a line spacing of 12.5 m for both the in-line and cross-line direction. The dominant frequency of the upper 1 s of data is approximately 45 Hz, which indicates an approximate tuning thickness of 12.5 m. The time-to-depth conversion of the 3D seismic-reflection data was carried out using a proprietary velocity cube provided by Shell international E&P Inc.
Stratigraphy and Geomorphologic Characteristics
The 250 CLS is placed in a seismic packet that is enveloped by the well-defined seismic reflections R100 and R300 (Fig. 2). The gentle inclination (0.5–1.1°) of the basal R300 reflection in the southwest of the study area is interrupted by a NW–SE-oriented ridge that has been formed above the B thrust. This ridge is 19 km long and 5 km wide, produces a relief up to 80 m, and is characterized by steep flanks (up to 7°) (Fig. 2). An almost flat area (6 km long and 2 km wide), called the B plateau, develops at the top of this ridge. This plateau has its highest topographic point at the center of the 250 CLS, and gently inclines (< 0.5 %) to the north and south (Fig. 5C). Reflections in the R100–R300 seismic packet show onlap on the updip and downdip sides of the B ridge, suggesting that this obstruction was present during deposition (Figs. 2B, 3E). The formation of the R100–R300 seismic packet started with the deposition of two successive, complex, MTDs, up to 100 m thick, named as MTD 1 and 2 (Fig. 3A, B). MTD 1 occurs in the southern half of the study area, and MTD 2 in the northern part. The south side of MTD 2 covers the northern edge of MTD1and gradually pinches out (Fig. 5A, B). The contact area between the two MTDs is expressed by a morphologic depression ∼ 60 m deep. The presence of pressure ridges in the MTDs immediately updip from the B ridge indicates that both mass movements collided with this obstacle (Fig. 3E). The seismic-reflection profile in Figure 5C shows that parts of both MTDs were able to overflow the southern and northern part of the B plateau, forming a shallower (∼ 40 m) morphologic depression between them. This, in combination with the absence of any incoherent reflections at the topographic peak of the B ridge (Figs. 2A, 5C), suggests that this obstruction was able to deflect in some degree MTDs 1 and 2.
A drape (40–60 m thick) consisting of intercalated chaotic lenses (< 20 m thick) and moderate- to high-amplitude, parallel to subparallel, continuous reflections occurs on top of MTD, 1 and 2 (Fig. 3E). This drape is interpreted to consist of thin MTDs interbedded in hemipelagic sediments (Fig. 3E) and has been partially to totally removed in several parts of the study area through erosion by gravity flows and mass movements (Fig. 5). Updip from the B ridge, the 250 CLS develops in the morphologic depression forming between the two underlying MTDs (Figs. 3C, 5A, B). The CLS crosses the B ridge due to its morphologic confinement, and continues its course to the southeast (Figs. 2B, 3C). The R100–R300 seismic packet ends with the deposition of a massive MTD that occurs on top of the 250 CLS, locally removing large parts of its levees (Figs. 3D, E, 5). A part of this MTD flows along the thalweg of the CLS.
Characteristics of the 250 CLS
The 250 CLS has an ENE–WSW orientation and consists of complex, meandering channel forms (cf. Deptuck et al., 2003; Deptuck et al., 2007), which are confined to the north and south by well-developed outer levees that are 20–160 m thick (Figs. 4, 5). This CLS is characterized by limited vertical aggradation. Successive channel forms develop in a specific zone through the plug-and-cut migration process described by Deptuck et al. (2007). No high-amplitude reflections (HARs) exist at the upper part of the CLS, indicating that its final filling resulted from the deposition of dilute, muddy turbidity currents (Deptuck et al., 2003; Deptuck et al., 2007).
The architectural elements of the 250 CLS vary significantly along its course, and consist of five discrete sectors (Figs. 4, 5). Sector 1 occupies the most updip part of the CLS in the study area, is 3–5.5 km wide (area bounded by the outer levees), and consists of complex, meandering channel forms (20–100 m deep and 0.2– 1 km wide) that cut through each other, forming multiple, complex pseudo-meander cut-off features (cf. Deptuck et al., 2007) (Figs. 4, 5A). The outer levees are up to 160 m thick. Erosion at the base of this sector is in general less than 20 m deep. High-amplitude reflections are commonly observed at the base of the 250 CLS (Figs. 4, 5A), and they are interpreted as channel-sand deposits, based on their curvilinear shape in map view. The present-day average gradient of the 250 CLS in sector 1 is 1.2° (2%) (Fig. 6).
Sector 2 occurs directly updip from the B ridge (Fig. 4) and passes from sector 1 through a steep step (4° or 7%: Figs. 2, 6). In this sector, the 250 CLS is 2 km wide and is characterized by an erosional base that has cut by as much as 50 m down into the underlying substrate (Figs. 4, 5B). Channel forms in this sector are 50–150 m deep and 0.5–1 km wide, and almost straight. The outer levees are up to 140 m thick. Rotational slides at the edges of the channel forms are common. The present-day average inclination of the 250 CLS in this sector is ∼ 1.2° (2%) (Fig. 6). HARs develop only locally at the floor of the valley (Fig. 4). The updip height of the B ridge along the thalweg of the 250 CLS is 70 m.
Sector 3 occupies the plateau on the B ridge (Fig. 4), is 3–4.5 km wide, and consists of complex, meandering channel forms (20–60 m deep and 0.1–1 km wide) that cut through each other, forming multiple, complex, pseudo–meander cut-off features, similarly to sector 1. The outer levees in sector 3 are much thinner (∼ 20 m) than its inner levees, which can be up to 80 m thick (Fig. 5C). Widespread HARs at the base of the 250 CLS are characteristic in this sector (Fig. 4), and their amplitude is an order of magnitude greater compared to the HARs observed in the other sectors. In addition, the thickness of the HAR packages in this sector are much thicker (up to 80 m) than in the other sectors (< 40 m) (Figs. 2A, 5C). Gamma-ray and resistivity logs from an exploration well in this sector indicate that these HARs represent thick sand deposits with gas hydrates (Lee et al., unpublished internal report).
Sector 4 is located on the southern flank of the B ridge, is ∼ 2 km wide, and is characterized by a steep slope of ∼ 4° (7%) (Figs. 2B, 4, 6). Channel forms (40–80 m deep and 0.3–1 km wide) in this sector are straight and strongly erosional, with downcutting of as much as 60 m into the underlying substrate (Fig. 5D). Failures at the side walls of the channel forms are common. The outer levees have a thickness up to 80 m. Incoherent, low-amplitude reflections covered by a high-amplitude reflection characterize the plugging of the channel forms.
Sector 5 of the 250 CLS is characterized by a low present-day gradient of 0.5° (1%) and is 2–4 km wide, and its outer levees are up to 100 m thick. Complex, meandering channel forms (50–80 m deep and 0.2–1 km wide) forming pseudo–meander cut-off features comprise the valley of the CLS (Figs. 4, 5E). HARs at the floor of channel forms are common. Failures at the channel walls are also common. A splay channel develops at the contact between sectors 4 and 5 (Figs. 4, 5E).
Prior to the discussion of the architectural elements of the 250 CLS, an effort to determine the paleotopography on which it developed should be made. The observation of onlap on the updip and downdip sides of the B ridge, in combination with the pressure ridges observed in MTDs 1 and 2, indicates that the B ridge was present during their development (Figs. 2B, 3E). In addition, the healing of the relief updip from the B ridge by the end of the R100–R300 seismic packet, in combination with the complete healing of the present-day seafloor, indicates very limited reactivation for the B thrust fold, if at all, since the formation of MTDs 1 and 2. Based on this, it is safe to argue that the paleomorphology of the 250 CLS was close to that observed in the 3-D seismic-reflection data.
The 250 CLS forms in the topographic depression between MTDs 1 and 2, which manages to extend itself over the B ridge (Figs. 3E, 5A–C), due to the partial northern and southern deflection of MTDs 1 and 2, respectively. On the B plateau, the depth of this trough is ∼ 20 m shallower than on its updip extension. This depth difference is attributed to the topographic influence of the B ridge, which was expressed by a 20-m-tall obstacle along the path of the topographic depression. This is supported by the present-day relief of the base of the 250 CLS, showing that the height of the B ridge that turbidity currents had to overflow was ∼ 70 m (Figs. 2A, 6). If this value is corrected for the 50 m of erosion at the base of the CLS in sector 2 (Fig. 5B), then it is estimated that the obstacle that the flows had to surmount at the initial stages of the CLS was ∼ 20 m.
Architectural Elements and Inferred Flow Rheology in the 250 CLS
The 250 CLS is divided into five sectors with discrete architectural elements, which allow inferences to be made on the rheol-ogy of turbidity currents surmounting an obstacle. Updip from the B ridge (sector 1), the 250 CLS bears many characteristics of a typical deep-sea channel–levee system (Figs. 4, 5A), consisting of well-developed outer levees and strongly meandering channel forms (Posamentier and Kolla, 2003; Deptuck et al., 2003; Kolla et al., 2007). The average slope of the base of the CLS in Sector 1 is ∼ 2% (1.2°). Lamb et al. (2008) estimated that turbidity currents on slopes more than 1% are supercritical for a very wide range (0.001–0.01) of values for the friction coefficient (cf). Based on this, it is concluded that flows within sector 1 were supercritical at the initial stage of the 250 CLS. The slope along the strongly meandering channel forms of the 250 CLS is < 1%, and therefore the flows in them would be subcritical to critical. However, even these flows would become supercritical through their acceleration at the steep step occurring between sectors 1 and 2 (Fig. 7).
In the last 3 km updip from the B ridge (sector 1), the 250 CLS becomes much narrower (2 km wide) and thicker (240 m), and acquires a strongly erosional character, which has resulted in the significant excavation (up to 50 m) of the underlying substrate (Figs. 4, 5A, B). These characteristics indicate that the flows in this interval become thicker and more energetic. The absence of extensive HARs at the base of the channel forms (Fig. 4), compared to the widespread HARs observed in sector 1, also suggests a more energetic nature for the flows in sector 2, preventing sand deposition. These inferred changes in the rheology of the turbidity currents are probably a consequence of a hydraulic jump and the development of a turbulent bore that was caused by their interaction with the local topography (Fig. 7). This interpretation is supported by multiple studies on changes in the flow and depositional behavior of turbidity currents upon their collision with topographic obstructions through laboratory experiments and numerical simulations (Kneller and McCaffrey, 1999; Bursik and Woods, 2000; Brunt et al., 2004; Lamb et al., 2004; Al Ja’Aidi et al., 2004; Toniolo et al., 2006a; Khan and Imran, 2008).
The extension of the 250 CLS on the B plateau indicates that turbidity currents were able to climb the ridge (Fig. 4), the height of which was increased from 20 m initially to 70 m by the erosional action of the flows. Sector 3 is twice as wide (4.5 km) as sector 2, and consists of multiple, complex, strongly meandering channel forms, which have shallower cross-sectional relief compared to the channel forms of the other sectors (Fig. 5C). The outer and inner levees are also thinner (< 40 m thick) than in the other sectors. This, in combination with the presence of widespread and thick (up to 80 m) HARs in this sector interpreted as channel sand deposits (Figs. 2B, 4, 5C), indicates that the flows became thinner and subcritical, and acquired a strong depositional character on the B plateau. This is in contrast with most laboratory and numerical simulations that show that it is immediately updip from an obstruction that depositional rates are higher (Kneller, 1995; Lamb et al., 2006; Al Ja’Aidi et al., 2004; Toniolo et al., 2006b; Khan and Imran, 2008). However, Brunt et al. (2004) demonstrated through laboratory experiments that if the obstruction is less than half of the height of the flow, then turbidity currents are able to transport most of their material over the obstruction. A crude estimate of 100 m for the minimum thickness of the largest flows can be made, based on the relief of the channel forms in sector 1 (prior to the hydraulic jump) (Fig. 5A). The relief of the B ridge during the formation of the 250 CLS ranged between 20 and 70 m tall, and therefore it is concluded that at least the largest turbidity currents in the 250 CLS were able to transport their suspended material over the obstruction. The absence of any deposition on the top of the obstruction in the flume experiments of Brunt et al. (2004) is attributed to its short width. As soon as flows climb over such a narrow obstruction, they begin to accelerate on their downdip flank, preventing them from becoming subcritical and depositional.
Sector 4 develops on the downdip slope of the B ridge, and is characterized by inclinations up to 7% (4°) (Figs. 2, 4, 6). The 250 CLS in sector 4 narrows down to 2 km wide and acquires a strong (up to 60 m) erosional character (Fig. 5D). The channel forms in this sector are straight and deeper (20–80 m) than in sector 3 (20– 50 m) (Figs. 4, 5). All of these observations are in agreement with an acceleration of the turbidity currents on a steep slope that resulted in their transformation into erosional and supercritical. The filling of the channel forms in sector 4 consists of a zone of incoherent, low-amplitude reflections covered by a high-amplitude reflection (Figs. 2B, 5D). These acoustic facies are interpreted as MTDs. The presence of failures on the walls of the channel forms support such an interpretation. The high-amplitude reflection on top of the incoherent facies can be interpreted as: (1) the high-density contrast between normally deposited channel sediments and MTDs, and/or (2) sand deposits originating from the failure of sand deposits in sector 3 and/or deposition by turbidity currents.
In sector 5, the 250 CLS acquires architectural characteristics similar to those in sector 1 (Figs. 2, 4, 5E), indicating the building of a CLS in an unconfined slope. At the transition between sectors 4 and 5, the northern outer levee has been breached and a splay channel is developed (Figs. 2, 4, 5E). Such features commonly occur at knickpoints (Pirmez et al., 2000). This splay probably results from the hydraulic jump to which the flows were subjected by the sudden change in gradient (∼ 6%) from sector 4 to sector 5 (Fig. 6).
Although zones of HARs are present and widespread in sectors 1 and 5 (up to 40 m thick), it is only in sector 3 that they acquire their greatest thicknesses (up to 80 m) and amplitudes (Figs. 2B, 4, 5). This observation not only agrees with flume experiments (Brunt et al., 2004) but also argues for enhanced sand deposition on tops of small (less than half the height of the flows) and wide (a couple of kilometers) obstructions. Such obstructions might exist by analogy in many other areas around the world. For example, it is known that gravity flows in the rugged continental slope of the northwest Gulf of Mexico propagate downdip through a fill-and-spill process (e.g., Brunt et al., 2004; Toniolo et al., 2006a; Toniolo et al., 2006b). Gravity flows gradually fill the most updip mini-basin and spill to the down-dip mini-basin when the sill between them has become thin enough. Based on this, preferential sand deposition is inferred to occur above the sills at the latest filling stages of the updip mini-basin.
The acquisition and analysis of 3-D seismic-reflection data from the lower continental slope of the NW Niger Delta has revealed the existence of a buried CLS that is formed in an elongated topographic depression. Along this trough, the CLS manages to climb over a ridge 20–70 m tall.
Updip from the ridge, the CLS becomes strongly erosional, twice as narrow, and straight, indicating the formation of erosional bores, formed upon the collision of the largest (> 100 m thick) turbidity currents on the ridge. Limited deposition of coarse-grained sediments in this area is in agreement with flume experiments that have shown that turbidity currents with thickness more than twice the height of an obstruction can transport most of their suspended material over it.
On the wide top of the ridge, the CLS becomes depositional and strongly meandering, and is characterized by the presence of channel-sand deposits (HARs) that are more extensive and twice as thick (80 m) as in other areas away from the ridge. This indicates that on the flat top of the ridge the flows become subcritical and strongly depositional.
Based on this study, it is revealed that sand distribution across a rugose topography depends not only on the rheological properties of the flows and the height of the topographic obstructions, but also on the width of the obstruction.
We would like to deeply thank David Piper, Charlie Winker, Brad Prather, and Simon Newton for the fruitful discussions, which resulted in the improvement and fulfillment of this paper. We would also like to thank Shell International E&P Inc. and Shell Nigeria E&P Co. for permission to publish this study.