The Geomorphometry of Hatteras Transverse Canyon, Hatteras Outer Ridge and Environs: A close-up view

TitleThe Geomorphometry of Hatteras Transverse Canyon, Hatteras Outer Ridge and Environs: A close-up view
Publication TypeJournal Article
YearSubmitted
AuthorsGardner, James V.
Secondary AuthorsArmstrong, Andrew A.
Tertiary AuthorsCalder, Brian R.
JournalGeomorphology
PublisherElsevier

Previously unknown features in Hatteras Transverse Canyon and environs were recently mapped during multibeam surveys of almost the entire U.S. Atlantic continental margin. Significant new features include (1) extensive landslide scarps on the walls of Hatteras Transverse and Hatteras Canyons, (2) an area of at least 7 landslide deposits that block lower Hatteras Transverse Canyon, (3) down canyon of the landslide deposits, a large depositional feature rises 100 m above the uppermost Hatteras Fan that has buried the transition from the mouth of Hatteras Transverse Canyon to uppermost Hatteras Fan, (4) several knickpoints in the channel thalwegs of both Hatteras Transverse Canyon and Hatteras Canyon, one 40 m high, that suggest both canyon channels are out of equilibrium and are in the process of readjusting, either to the channel blockage by the extensive landslide deposits or by a readjustments to the increased sedimentation during the last eustatic lowstand, (5) a large area of outcrop on the lower margin between Pamlico and Hatteras Canyons that previously has been interpreted as an area of slumps, blocky slide debris and mud waves and (6) headward erosion in the head region of Hatteras Transverse Canyon where it has intercepted the lower reaches of Albemarle Canyon as well as in a small side channel that has eroded into Hatteras Outer Ridge. The newly discovered features add a new level to understanding of the recent processes that have profoundly affected Hatteras Transverse Canyon, Hatteras Canyon and, to a lesser degree, Hatteras Outer Ridge. 

Full Text

Abstract  

Previously unknown features in Hatteras Transverse Canyon and environs were recently mapped during multibeam surveys of almost the entire U.S. Atlantic continental margin.  Significant new features include (1) extensive landslide scarps on the walls of Hatteras Transverse and Hatteras Canyons, (2) an area of at least 7 landslide deposits that block lower Hatteras Transverse Canyon,  (3) down canyon of the landslide deposits, a large depositional feature rises 100 m above the uppermost Hatteras Fan that has buried the transition from the mouth of Hatteras Transverse Canyon to uppermost Hatteras Fan, (4) several knickpoints in the channel thalwegs of both Hatteras Transverse Canyon and Hatteras Canyon, one 40 m high, that suggest both canyon channels are out of equilibrium and are in the process of readjusting, either to the channel blockage by the extensive landslide deposits or by a readjustments to the increased sedimentation during the last eustatic lowstand, (5) a large area of outcrop on the lower margin between Pamlico and Hatteras Canyons that previously has been interpreted as an area of slumps, blocky slide debris and mud waves and (6) headward erosion in the head region of Hatteras Transverse Canyon where it has intercepted the lower reaches of Albemarle Canyon as well as in a small side channel that has eroded into Hatteras Outer Ridge.  The newly discovered features add a new level to understanding of the recent processes that have profoundly affected Hatteras Transverse Canyon, Hatteras Canyon and, to a lesser degree, Hatteras Outer Ridge. 

1. Introduction

Hatteras Transverse Canyon and Hatteras Outer Ridge are two distinctive bathymetric features of the middle U.S. Atlantic lower continental margin.  Work in the 1960s identified these features as somewhat anomalous and various interpretations derived from single-beam echosounder and subbottom-profiler data were offered to explain their character.  Improvements of multibeam echosounders (MBES) since the early 1990s have provided increasingly higher-resolution digital terrain models (DTM) of the bathymetry of the deep seafloor.  The new multibeam data allow a three-dimensional quantitative analysis (geomorphometry) of this seascape.  ‘Geomorphometry’ is a term seldom used by the marine geology seafloor-mapping community when describing features of the seafloor, perhaps because the term is defined as the science of quantitative land surface analysis (Pike, 1995).  However, the definition of geomorphometry, when the land reference is taken senso lato, accurately describes seafloor-mapping studies based on bathymetry data produced by multibeam echosounders.  Consequently, the term geomorphometry has been used here.  The focus of this paper is to use the geomorphometry of Hatteras Transverse Canyon, Hatteras Outer Ridge and environs to describe the present canyon and ridge and the processes that have modified them. These DTMs, together with simultaneously collected co-registered acoustic backscatter, allow a better understanding of the processes that formed and modified the present seafloor than was provided by the older data.  The area that includes Hatteras Transverse Canyon and Hatteras Outer Ridge (Fig. 1) was completely surveyed during mapping of the bathymetry of the entire U.S. Atlantic continental margin between the 1000 and 5500 m isobaths as part of the U.S. Law of the Sea Extended Continental Shelf project (Gardner et al., 2006).  The present study, based on these MBES data, investigates the geomorphology of the Hatteras Transverse Canyon-Hatteras Outer Ridge area at a high resolution (100 m/pixel) as well as describes various smaller-scale features that have until now not been identified from this area. 

The bathymetry of Hatteras Transverse Canyon (HTC) was first described by Rona et al. (1967) who noted that the course of HTC roughly parallels rather than trends perpendicular to the regional isobaths, as is more typical of the other canyons and canyon channels in the area.  Rona et al. (1967) related the trend of HTC to a deflection of downslope sediment transport by Hatteras Outer Ridge (HOR), a large late Tertiary to Late Pliocene sediment drift.  The bathymetry of Hatteras Outer Ridge was first mentioned by Heezen et al. (1959) and later investigated by Rona et al. (1967), Rona (1969), Asquith (1979), Tucholke and Laine (1982); Mountain and Tucholke (1985) and Locker and Laine (1992).  Rona et al. (1967) suggested the formation of both HTC and HOR, as well as bedforms on HOR (so-called “lower continental rise hills” in the descriptions from the 1960s), were due to current-controlled sedimentation related to the Western Boundary Undercurrent that flows southward over the area (Heezen et al., 1966).  

Prior to the construction of the HOR, a series of submarine canyons funneled sediment down the margin and out onto Hatteras Abyssal Plain.  HOR developed as some of the fine-grained sediments from gravity-driven downslope events were intercepted by a geostrophic current presumed to be the precursor of the Western Boundary Undercurrent.  A shear zone developed between the geostrophic flow and the Gulf Stream that created an area conducive to the formation of a sediment drift that eventually buried the lower reaches of the canyons and submarine fans in this area (Tucholke and Laine, 1982).  By the latest Pliocene, the constructional phase of HOR ended and erosion, rather than deposition, became the dominant process on HOR (Tucholke and Laine, 1982).

Prior et al. (1984) and Pratson and Laine (1989) comment that the present seascape of the entire continental margin in this area is mostly blanketed by only a few meters of Holocene sediment, suggesting to them that the large-scale geomorphology has been effectively unmodified by modern (<5 ka) processes.  The only exceptions mentioned by Prior et al. (1984) are in areas of outcrop on the middle and lower margin and in scarps on canyon walls where modern sediment has been eroded away by recent turbidity-current events. 

2. Methods and Data

The new MBES bathymetry and acoustic backscatter data provide a complete (i.e., 100%) coverage of this area (Fig. 2) with millions of accurately (<0.2% of water depth) determined soundings (depths) of variable density that typically average at ~50 m intervals and coregestered backscatter values spaced at a higher density.  The sounding density allows the construction of a digital terrain model (DTM) with a maximum resolution of 100 m/pixel.  The DTM was constructed from data collected on 4 cruises; two cruises in 2005 (PF05-1 and PF05-2) using a hull-mounted 12-kHz Kongsberg Maritime EM121A MBES system aboard USNS Pathfinder, one cruise in 2008 (KNOX17RR) with a hull-mounted 12-kHz Kongsberg Maritime EM120 MBES on RV Roger Revelle, and a 2012 cruise (RB12-1) with the NOAA Ship Ronald H. Brown equipped with a hull-mounted 12-kHz Kongsberg Maritime EM122 MBES system.  Data from the four cruises were combined into a single DTM and showed no horizontal or vertical offsets where the separate datasets join together.  All three of these MBES systems have active roll, pitch and yaw beam steering.  A full patch test was conducted in the survey area prior to each cruise to ensure attitude sensor offsets were correct.  Sound speed in the water column was calculated from calibrated multiple expendable bathythermographs collected no less than once every 6 hr on each cruise.  Sound speed was integrated into the MBES data acquisition system to compensate for the refraction effects within the water column by ray tracing each sounding to its accurate location on the seafloor.  Navigation on each cruise utilized an inertial motion unit (IMU) interfaced to a differential global positioning system that provided position fixes with an accuracy better than ±5 m.

The Kongsberg Maritime MBES systems are capable of simultaneously collecting full time-series acoustic backscatter that can be co-registered with each bathymetric sounding. When the received amplitudes are properly calibrated to the outgoing signal strength, receiver gains, spherical spreading, and attenuation, the corrected backscatter provides clues to the composition of the surficial seafloor.  However, the 12-kHz acoustic signal undoubtedly penetrates the seafloor to an unknown, but perhaps significant, depth in some areas (Gardner et al., 1991), thereby generating a received backscatter value that is a function of some unknown combination of acoustic impedance, seafloor roughness and volume reverberation.

In addition to the MBES, each ship was equipped with a high-resolution 3.5-kHz subbottom profiler.  The profilers collected continuous subbottom profiles along each cruise track with a maximum penetration of ~50 m.

3. General Description of Continental Margin off Cape Hatteras

The traditional (pre-MBES) description of the continental margin southeast of Cape Hatteras is that of a relatively narrow continental shelf (30 to 60 km), a ~95-km-wide continental slope and a broad ~375-km-wide continental rise that merges with the Hatteras Abyssal Plain at approximately the 5400-m isobath.  The slope and rise were constructed by deposition from turbidity currents (Drake et al, 1968; Pilkey and Cleary, 1968; Emery et al, 1970) and large debris flow deposits (Embley and Jacobi, 1977; Embley, 1980; Twichell et al., 2009) that subsequently have been modified by southward-flowing geostrophic currents that reworked the sediment (Heezen et al, 1966; McCave and Tucholke, 1986).  The slope and rise are dominated by a series of submarine canyons and their braided channel extensions (called here canyon channels), especially Albemarle, Hatteras and Pamlico Canyons (Fig. 2) and their associated canyon channels (Rona et al., 1967; Newton and Pilkey, 1969), but include numerous smaller unnamed canyons and canyon channels.  The canyons and canyon channels traverse downslope until they are intercepted by Hatteras Transverse Canyon located immediately upslope of Hatteras Outer Ridge, a large sediment drift that began to form in the Early Miocene (Mountain and Tucholke, 1985) and has served as an obstacle to downslope flow of turbidity currents until very recently.  Hatteras Outer Ridge has diverted sediments to the southwest that were initially transported directly downslope through the canyons and canyon channels to the Hatteras Abyssal Plain.  The diversion of sediments formed HTC and redirected sediment to a more southerly location on Hatteras Fan (Cleary and Conolly, 1974).  The mouth of Hatteras Transverse Canyon emerges from the southwestern end of Hatteras Outer Ridge where sediment has been diverted to the southeast by the large failure masses of Cape Fear and Cape Lookout Slides (Fig. 2) and eventually emerged onto Hatteras Fan as a series of small distributaries that spread out onto the southern Hatteras Abyssal Plain.

3.1. Hatteras Transverse Canyon System

Rona et al. (1967) first described the bathymetry of Hatteras Transverse Canyon (HTC) based on about 30 wide-angle single-beam echosounder crossings.  They also produced a map of the lower reaches of Hatteras and Pamlico Canyons where Hatteras Transverse Canyon captures the two canyons.  Subsequent studies of this area using single-beam echosounders and subbottom profilers were made by Rona and Clay (1967), Newton and Pilkey (1969), Cleary and Conolly (1974), Cleary, et al. (1977), Bunn and McGregor (1980), Pratson and Laine (1989), with seismic-reflection profilers by Tucholke and Laine (1982) and with GLORIA long-range sidescan sonar and widely spaced seismic profiles by Popenoe and Dillon (1996).  These studies outlined the general characteristics of HTC but with widely dispersed tracklines and varying qualities of bathymetry.  The new MBES mapping provides a complete and clear view of the bathymetry of HTC with a resolution not achieved by the earlier studies and provides new insight into many of the processes that have affected this area. 

Hatteras Transverse Canyon trends 220˚, roughly parallel to the regional isobaths, with a gently curvilinear 130-km length before it exits out of its confines to form the upper portion of southern Hatteras Fan.  The HTC channel floor descends from water depths of 4718 m at its head to depths >5960 m before it is buried by landslide deposits at ~78 km down channel.  The upper 38.5 km of the channel has a gradient of 0.23˚, then the gradient increases to 0.32˚ for the next 18.3 km after which the channel abruptly steps down and the gradient flattens to 0.01˚ for the remaining 21.5 km before encountering the landslides. The channel width varies for the first 135 km from 0.6 to 4.4 km but then abruptly increases to between 12.3 and 43.8 km wide at ~130 km down-canyon (Fig.3A).  The channel eventually evolves into a series of distributary channels throughout the upper Hatteras Fan.  The incision depth of HTC increases along the canyon length, from 60 m of relief at the head of the canyon to 325 m at 119 km down-canyon (Fig.3B), a point just below the confluence of HTC with Hatteras Canyon.  There are four knickpoints (i.e., abrupt step-downs) of the channel; a 40 m drop occurs at 41.8 km down-canyon at 4978 m water depth, a 10 m drop occurs at 48.5 km down canyon at 5015 m water depth, an 8 m drop at 58.5 km down canyon at 5073 m water depth and a 16 m drop at 61.8 km down canyon at 5081 m water depth (arrows “a” through “d”, respectively, in Figs. 3C and Fig. 4).  From a point 119 km down-canyon to the upper Hatteras Fan, the canyon floor incision rapidly decreases to below the resolution of the MBES.  The upper 17 km of HTC has walls with slopes of 6˚ to 12˚.  The western wall of HTC is generally steeper (6˚ to 8˚) from 17 to 34.8 km down-canyon compared to the eastern wall (2˚ to 4˚).  Up to this point (51.8 km from the head), the canyon walls show little evidence of landslide scarps.  However, down-canyon beyond 51.8 km, both walls are heavily eroded with landslide scarps (Fig. 4).

The head of HTC has almost 100 m of relief with headward erosion that has migrated towards Albermarle Canyon channel (Fig. 4).  Much of the older literature called Albermarle Canyon and channel “Albermarle Transverse Canyon”; however, its trend clearly is not transverse to the isobaths except for the last 15 km (Fig. 2).  The upper 5 km of HTC forms a loop around a central bathymetric high that stands 70 m above the eastern side of the canyon floor and 40 m above the western floor of the loop.  The high within the loop is 3 km by 0.9 km in plan view and is oriented N-S.  The northern end of the high is aligned with the end of Albermarle Canyon channel where it enters HTC.

A small 11 km long side channel enters from the east at 16.5 km from the head of HTC (Fig. 4).  This small side channel trends S70W and descends 140 m with a slope of 0.7˚ from its head to its confluence with HTC channel.  No step in depth occurs from the small channel floor to the main floor of HTC.  The side channel widens towards its head, from 0.9 km at the confluence with HTC to 2.4 km at its head, giving the appearance of headward erosion with 40 m of relief at the headwall.  The small side channel has a maximum incision of 90 m that occurs at the confluence with HTC. 

Landslides and landslide scarps occur along some, but not all, sections of HTC (Fig. 5).  The channel floor of HTC, from its head to a point 54.8 km down-canyon, is very flat with relief just at or below the resolution of the 12-kHz MBES (<10 m at these depths) and the canyon walls in this section show no signs of landslide scarps.  The next 23.8 km of the canyon floor continues with little relief but the canyon walls have extensive landslide scarps (Fig. 5).  The series of landslide scarps on the east wall above the confluence with lower Hatteras Canyon are 21.8 km long with 300 m of relief.  The opposite west wall has a series of smaller scarps with maximum relief of ~150 m.  The landslides in this area have exposed at least three levels of apparent outcrops with surface dips that parallel the regional slope of the seafloor (labels 1, 2 and 3 in Fig. 5).  The largest section of landslide scarps and landslide deposits on HTC occur 5.5 km below the confluence with lower Hatteras Canyon, 83 km from the head of HTC (area between white arrowheads in Fig. 5).  This landslide section has extensive landslide scarps on both canyon walls, although the landslide scarps on west wall in this section have not exposed outcrops.  Landslide deposits extend for 24 km down-canyon and have accumulated as high as 25 m above the featureless channel floor above and below the landslide deposits, effectively blocking any present flow down-canyon.  At least 7 large landslide scarps and several smaller scarps can be identified on the channel walls along with their associated landslide deposits on the channel floor (Fig. 6).  The total volume of landslide deposits in this section of the channel is at least 20 km3.  The headwalls of the landslide scarps on the canyon walls have 50 to more than 100 m of relief with extensive debris aprons at their bases.  

3.2. Lower Hatteras Canyon

Lower Hatteras Canyon (LHC) trends directly downslope and enters HTC from the north-northwest at a 70˚ angle with no step in floor depths (Fig. 4).  Compared to Hatteras Transverse Canyon, LHC is narrow with channel widths that increase from only 200 to 500 m along the lower 300 km of the canyon.  An incised channel thalweg, that abruptly disappears 23 km up-canyon from the junction of LHC with Hatteras Transverse Canyon, decreases in incision depth down channel from 100 to 40 m.  LHC in general has downcut from less than 100 m to as much as 260 m into the lower continental margin.  The last 35 km of the channel profile of LHC has an overall gradient of 0.4˚ but with three knickpoints that occur at 5, 15 and 35 km up-canyon from the point where LHC joins Hatteras Transverse Canyon.  The knickpoints drop the channel floor 55, 30 and 20 m in water depths of 5080, 4942 and 4753 m, respectively (profile C-B in Fig. 4).  Landslide scarps occur only on the lower 22.6 km of the canyon but there is no evidence at the resolution of the MBES data of landslide deposits on the floor of the canyon.

3.3. Upper Hatteras Fan

Hatteras Transverse Canyon leads directly out onto Hatteras Fan but only the upper part of Hatteras Fan was mapped with MBES.  Upper Hatteras Fan (UHF) does not have a typical morphology of submarine fans because it is confined between Hatteras Outer Ridge (HOR) on the north and two large gravity slides, Cape Lookout (CLS) and Cape Fear Slides (CFS), on the south (Fig. 7).  For example, levees typically found on upper fans are not found associated with UHF.  Other unusual aspects of UHF include a large depositional feature that rises as much as 100 m above the main channel floor immediately downslope from the large landslide deposits discussed above (Fig. 8) that includes an area of large-scale stepped surfaces (Figs. 8A and B).  The large depositional feature is 120 km long and varies between 9 and 26 km wide with an average height above the channel floor of 50 m that represents ~110 km3 of material that blocks the main channel of the upper fan.  The deposit is clearly not related to the large landslide deposits found upslope as shown by the ~30 km downslope offset of the center of the deposit from the southern end of the landslide deposits and by a 70-m-deep bathymetric low between the landslide deposits and the depositional feature (Fig. 8C).  The stepped section of the depositional feature has ~20 m of relief between successive relatively flat surfaces that are 1.5 to 5 km in length with arcuate plan shapes that are convex downslope.  The steps occur on a slope of 0.2˚ and abruptly die out on a featureless slope of 0.02˚.

A series of small, very shallow distributary channels, best seen in the acoustic backscatter, can be followed for at least 73 km immediately downslope of the stepped surface of the upper fan (Fig. 7). The channels are less than ~3 m deep (near the resolution of the MBES at these depths) and less than 500 m wide.  The distributaries are confined to the northern edge of the upper section of the depositional feature but splay out to occupy about a third of the width of UHF where the depositional feature ends and they continue to splay out downslope.  These must be what Popenoe and Dillon (1996) misinterpreted as sediment waves from GLORIA imagery of the UHF. 

3.4. Hatteras Outer Ridge

Only the southern half of Hatteras Outer Ridge is discussed here because that is all that has been mapped with a MBES system.  Hatteras Outer Ridge (HOR) is a large (>19,000 km3) sediment drift immediately downslope (east) of Hatteras Transverse Canyon (Fig. 1) that has buried the lower reaches of Albermarle and Hatteras Canyons and the upper parts of their associated fans, as well as buried several sections of lower canyons farther to the northeast towards Hudson Canyon (Tucholke and Laine, 1982).  HOR was constructed during the Miocene to Late Pliocene (Sheridan et al., 1978; Tucholke and Laine, 1982) and is at least 700 m thick. Hatteras Outer Ridge is at least 600 km long with a maximum width of ~150 km.  The crest of the southern half of HOR trends N42˚E for the first 158 km from the southwestern end, then the crest changes trend to N53˚E for at least another 100 km (Fig. 9).  Water depths of the crest in the southern half of the ridge range from ~4700 m at the southwest end to a minimum depth of 4360 m at a point 222 km to the northeast along the crest.  The water depth of the crest slowly deepens to 4473 m northeastward at the end of the MBES data.  The southern half of the ridge is asymmetrical in NW-SE cross section with slopes of 0.9˚ on the west side but with slopes of ~0.5˚ on the east side and the west flank is only ~10 km wide compared to as much as 140 km wide for the east flank because of the differences in water depths of the two flanks.

The surface of HOR is covered with bedforms (Fig. 9) that are described in older literature as “lower continental rise hills” (Fox et al., 1968; Rona, 1969; Hollister and Ewing, 1972; Benson and Sheridan, 1978; Asquith, 1979) even though Rona (1969) correctly noted that the features are not hills but elongate ridges.  The consensus by the 1980s was that these features are actually modified bedforms, not “hills” in the strict sense (Tucholke and Laine, 1982; Mountain and Tucholke, 1985), although the name “continental rise hills province” remained in later literature (e.g., Pratson and Laine, 1989; Locker and Laine, 1992).  Fox et al. (1968) studied a small area of the bedforms in detail and concluded they are relict antidunes.  The MBES data clearly delineate the bedforms and several subbottom-profiler records were collected across the bedforms during the MBES cruises that provide insights into their structure.  Individual bedforms range in length from a few 10s of km to more than 200 km and their crests rise from 12 to greater than 100 m above the surface of HOR (Fig. 10A).  Wavelengths of the bedforms range from 2 to ~11 km.  There is no correlation of bedform wavelength to crest water depth or bedform height (Fig. 10A-C).  The bedform crests have a strong east-west orientation that is 30 to 50˚ off the general trend of the margin in this area (Fig. 10D).  Individual bedform shapes vary from nearly symmetrical to sharply asymmetrical (Fig. 11) although there are no areas with one shape that predominates over the other.  Sections of the bedform field in the middle of HOR have been modified by channelized sediment within the Norfolk-Washington Canyon system that has been directed downslope and out onto the sediment drift and has partially filled in the bedform troughs (Fig. 12), as first noted by Rona (1969).  The troughs of the bedforms in the path of this downslope transport have been filled with acoustically strong, horizontally stratified sediment (first noted by Fox et al., 1968 and amplified by Tucholke and Laine, 1982).  Sediment transported from the east has also filled bedform troughs on the southeastern edge of HOR (Fig. 12).

3.5. Nearby Features

3.5.1. Cape Fear Slide

Cape Fear Slide (CFS) is a major landslide that stretches more than 390 km downslope from a crown scarp at the 2411 m isobath.  The landslide was first described by Embley (1980) and later by Popenoe et al. (1993), Popenoe and Dillon (1996) and Lee (2009).  The age of the Cape Fear event is 14.5 to 9 ka (Paull et al., 1996; Rodriguez and Paull, 2000).  The distal toe of CFS is an abrupt 15 m wall with a slope of ~3˚.  The toe of the slide is 76 km wide at its widest.  The northern margin of the CFS overlaps Cape Lookout Slide (discussed below) and is clear evidence that the Cape Lookout Slide occurred some time before the CFS event (Popenoe and Dillon, 1996) (Fig. 13). 

3.5.2. Cape Lookout Slide

Only a section of exposed distal Cape Lookout Slide (CLS) is seen in the mapped area, located north of and partially overridden by the CFS (Fig. 13).  This section of the CLS is part of a much larger slide feature that originated on the lower slope (Popenoe and Dillon, 1996).  There is ~10 m of relief where the younger CFS overlaps the CLS. Although all that can be said of the age of the CLS is that it predates the CFS, the significance of the CLS to the present study is that it forms the present western boundary of upper Hatteras Fan.  The southern flank of upper Hatteras Fan is now buried beneath the CLS but the eastern flank of the CLS certainly now and perhaps in the past has diverted the lower reaches of Hatteras Transverse Canyon towards the south and out onto the upper Hatteras Fan. 

4. Discussion

Hatteras Transverse Canyon occurs in an area that includes the distal reaches of several major canyons, a large sediment drift and two large mass-failure deposits.  The general relationships of these features to Hatteras Transverse Canyon were described in previous studies and the present study does not contradict most of those earlier conclusions.  However, the new multibeam bathymetry and backscatter data do add a new level of detail that reveals mesoscale (i.e., ~100-m horizontal and ~5-m vertical) processes that have not previously been described for this area.  Newly described features include extensive landslide scarps on the walls of HTC, large landslide deposits that completely block HTC, several knickpoints in most of the channel floors and a large area of outcrop between Pamlico and Hatteras Canyons. 

Landslide scarps occur on both walls of HTC and the lower reaches of LHC but for most of their length, the canyon floors show no evidence of landslide deposits.  The landslide deposits that were initially deposited on the HTC floor upstream of the confluence with LHC must have been reworked and transported through the lower reaches of the canyon and out onto the area below the confluence.  However, immediately down canyon from the confluence of HTC and LHC, a large landslide complex completely blocks the lower canyon.

The composite landslide deposits are located in the immediate area of extensive landslide scarps on the wall of HTC below the confluence of HTC and LHC.  The landslide deposits are 24 km long and as much as 25 m high and completely fill the channel in the landslide zone.  The lack of landslide deposits on the floor of HTC and LHC above the confluence of HTC and HC suggests that the landslide scarps on the walls of HTC and LHC and their associated deposits are either older than the composite landslides farther down canyon or that the landslide deposits were resuspended, transported down canyon and accreted to the composite landslide deposits below the confluence of HTC and LHC.  Alternately, the landslide deposits may be older than the composite landslide deposits and make up part of the large depositional feature found farther downslope.  There is no evidence on the surface of the landslide deposits that a channel has begun to incise through the deposits.  This suggests that either 1) no post-landslide turbidity currents have transited through HTC or 2) any significant-sized post-landslide turbidity currents that flowed down HTC and LHC were either blocked, or perhaps were dissipated, by the landslide deposits.  Regardless, the distributary channels on the upper Hatteras Fan are inactive channels that formed prior to the blockage by the landslide deposits.  Lamont Doherty Earth Observatory piston core VM22-2 (Fig. 5) was collected from HTC is at the southern end of the landslide deposits.  The top of the core consists of 1.72 m of medium olive gray hemipelagic sand, with a 10-cm-thick interbed of foram marl ooze, that overlies a foram marl ooze.  Eastward core 10365 (Fig. 5) was collected from HTC up-canyon from the landslides and is described as having a surface sand layer greater than 0.8 m thick (Field and Pilkey (1971).  The lack of hemipelagic sediment at the top of the core 10365 suggests that HTC has been an active conduit for sand transport up until very recently.  The surface sand layer in the core 10365 up-canyon from the landslides, together with the surface sand layer of core VM22-2 within the landslide deposits, suggests that the landslides within HTC are probably Holocene in age and not Pleistocene or older, otherwise one would expect a thin hemipelagic cap on the cores.

It is tempting to correlate the channel knickpoints with the outcrops on the walls of HTC and LHC.  However, the water depths of the knickpoints and outcrops do not form a plane surface or surfaces.  The one exception is the channel knickpoints at 5081 in HTC and at 5080 in LHC (d and e in Fig. 5).  These knickpoints may indeed be the result of excavations of resistant horizontal strata by erosion in the channels (Popenoe and Dillon, 1996).  However, a more likely explanation is that all the knickpoints in the floors of HTC and LHC, which are all up-channel from the large landslide deposits, represent a disruption of the equilibrium profiles of both canyon channels by the landslide deposits and that the two channels are presently headward-eroding to re-establish new equilibrium profiles.  The flat, featureless nature of the channel floors of both HTC and LHC suggests channel entrenchment during knickpoint retreat has yet to begin (Pirmez, et al, 2000; Kneller, 2003; Mitchell, 2004; 2006; Holland and Pickup, 2006; Heiniö and Davies, 2007).  In addition to the knickpoints in the channel, the head of HTC and the side channel of upper HTC also appear to be eroding headward.  The presence of knickpoints also could be an indication that the canyon channels are re-establishing their equilibrium profiles after a period of increased sedimentation after the last glacial lowstand, as first suggested by Embley and Jacobi (1986).  Hatteras Canyon, a tributary channel to HTC, must be adjusting by erosion to a drop in HTC channel thalweg, the principal channel, because there is no suggestion of a step down at the confluence of the two channel floors.  Also, the channel floor of Hatteras Canyon is steeper (0.69˚) upstream of the confluence with HTC than the principal channel floor (0.02˚), which Mitchell (2004) suggests reflects the tributary’s smaller contributing areas.

Newton and Pilkey (1969), Embley and Jacobi (1986), Pratson and Laine (1989) and Popenoe and Dillon (1996) all mention a hummocky area between Pamlico and Hatteras Canyons on the lower margin (Fig. 2a and b) as seen on seismic records and sidescan images.  These authors suggested the hummocky terrain is the result of slumps, blocky slide debris or mud waves.  Popenoe and Dillon (1996) concluded that the so-called hummocky area is an old landslide draped by hemipelagic sediment.  The MBES data show these features are not hummocks (i.e., individual hillocks), blocks or landslide deposits but rather are long (~50 km), low (<15 m) linear ridges that strike N20E (60˚ to the regional slope) and that occupy a somewhat rectangular area and most resemble a broad area of outcrop (Fig. 14).  The outcrop area encompasses ~6500 km2 and is roughly confined between the 4600 and 5150 m isobaths.  This range in water depths coincides with the lower range of effects from the Western Boundary Undercurrent (Schneider et al., 1967).  The high acoustic backscatter of this area (-25  -32 to -36 dB for the surrounding area) is also seen on the walls of HTC and LHC, again suggesting that the high backscatter area is an area of outcrops with perhaps only a very thin sediment cover. 

5. Conclusions

Recent mapping the area of Hatteras Transverse Canyon and environs using high-resolution multibeam bathymetry, co-registered acoustic backscatter and 3.5-kHz subbottom profiler data allows a geomorphometrics analysis of the area.  The detailed digital terrain model generated from these data reveals several previously unknown important features in an area that was extensively explored in the past.  The newly discovered features add a new level to understanding of the recent processes that have profoundly affected Hatteras Transverse Canyon, Hatteras Canyon and, to a lesser degree, Hatteras Outer Ridge.  The new features include:

•  Extensive landslide scarps that are restricted to the walls of lower Hatteras Transverse Canyon and lower Hatteras Canyon.

•  An area composed of at least 7 landslide deposits that clog Hatteras Transverse Canyon just down canyon from the confluence of Hatteras Transverse and Hatteras Canyons.

•  A large depositional feature down canyon from the 7 composite landslides that may represent older landslide deposits.  This deposit rises 100 m above the fan surface and appears to completely cover the transition zone between lowermost Hatteras Transverse Canyon and the uppermost Hatteras Fan.

•  A series of knickpoints in the channel floor of lower Hatteras Transverse Canyon and lower Hatteras Canyon that suggest both canyon channels are in the process of reestablishing new equilibrium profiles, either because of the disruption of a former equilibrium profile by the blockage of the canyons by the extensive landslide deposits or from a period of increased sedimentation during and immediately after the last glacial eustatic lowstand.

•  A large (~6500 km2) area of outcrop on the lower margin that is almost devoid of blanketing hemipelagic sediments.

Acknowledgments

This study was inspired by the seminal work by Peter A. Rona on Hatteras Transverse Canyon and Hatteras Outer Ridge.  He had an amazing grasp of the bathymetry and processes of these two features that he visualized from sparse, poorly navigated and often very noisy single-beam data.  It is a tribute to Peter that the general bathymetry of these features are as he described them almost 60 years ago. We acknowledge the extraordinary help and cooperation of the officers and crews of the USNS Pathfinder, RV Roger Revelle and NOAA Ship Ron Brown.  National Oceanic and Atmospheric Administration (NOAA) grants supported the three cruises that collected the multibeam and subbottom data.  All three cruises were in support of bathymetry mapping for the U.S. Law of the Sea Extended Continental Shelf efforts.  All of the bathymetry and backscatter data and several images made from the data are available at http://www.ccom.unh.edu/theme/law-sea/atlantic-margin.  We appreciate the thoughtful and constructive reviews of an earlier version of the manuscript by Larry A. Mayer that lead to substantial improvements.

 

References Cited

Asquith, S.M. 1979, Nature and origin of the lower continental rise hills off the east coast of the United States.:  Marine Geology, v. 32, p. 165-190.

Benson, W.E. Sheridan, R.E., and others, 1978, Initial Reports of the Deep Sea Drilling Project, Volume 44: Washington D.C., U.S. Government Printing Office, 1005 p.

Bunn, A.R. and McGregor, B.A., 1980, Morphology of the North Carolina continental slope, western North Atlantic, shaped by deltaic sedimentation and slumping:  Marine Geology, v. 37, p. 253-266.

Cleary, W.J. and Conolly, J.R., 1974, Hatteras deep-sea fan:  Journal of Sedimentary Petrology, v. 44, p. 1140-1154.

Cleary, W.J., Pilkey, O.H., and Ayers, M.W., 1977, Morphology and sediments of three ocean basin entry points, Hatteras Abyssal Plain:  Journal of Sedimentary Petrology, v. 47, p. 1157-1170.

Drake, C.L., Ewing, J.I., and Stockard, H.P., 1968, The continental margin of the eastern United States:  Canadian Journal of Earth Sciences, v. 5, p. 933-1010.

Embley, R. W., 1980, The role of mass transport in the distribution and character of deep-ocean sediments with special reference o the North Atlantic:  Marine Geology, v. 38, p. 23-50.

Embley, R.W. and Jacobi, R.D., 1977, Distribution and morphology of large submarine sediment slides and slumps on Atlantic continental margins:  Marine Geotechnology, v. 2, p. 205-228.

Embley, R.W. and Jacobi, R.D., 1986, Mass wasting in the western North Atlantic, in Vogt, P.R. and Tucholke, B.E., eds., The western North Atlantic Region:  Boulder, Colorado, Geological Society of America, Geology of North America, v. M, p. 479-490.

Emery, K.O., Uchupi, E., Phillips, J.D., Bowin, C.O., Bunce, E.T. and Knott, S.T., 1970, Continental rise of eastern North America:  American Association of Petroleum Geologists Bulletin, v. 54, p. 44-108.

Field, M.E. and Pilkey, O.H., 1971, Deposition of deep-sea sands:  Comparison of two areas of the Carolina continental rise:  Journal of Sedimentary Petrology, v. 41, p. 526-536.

Fox, P.J., Heezen, B.C. and Harian, A.M., 1968, Abyssal anti-dunes:  Nature, v. 220, p. 470-472.

Gardner, J.V., Field, M.E., Lee. H., Edwards, B.E., Masson, D.G., Kenyon, N., and Kidd, R.B., 1991, Ground truthing 6.5-kHz sidescan sonographs:  What are we really imaging?:  Journal of Geophysical Research, v. 96, p. 5955-5974.

Gardner, J.V., Mayer, L.A., and Armstrong, A., 2006, Mapping supports potential submission to U.N. Law of the Sea, Eos (Transactions, American Geophysical Union), v. 87, 157-159.

Heezen, B.C., Tharp, M., and Ewing, M., 1959, The floors of the oceans I. The North Atlantic.  Geological Society of America, Special Paper 65, 122p.

Heezen, B.C., Hollister, C.D. and Ruddiman, W.F., 1966, Shaping of the continental rise by deep geostrophic contour currents:  Science, v. 152, p. 502-508.

Heiniö, P. and Davies, R.J., 2007, Knickpoint migration in submarine channels in response of fold growth, western Niger Delta:  Marine and Petroleum Geology, v. 24, p. 434-449.

Holland, W.N. and Pickup, G., 1976, Flume study of knickpoint development in stratified sediment:  Geological Society of America Bulletin, v. 87, p. 76-82.

Hollister, C.D. and Ewing, J.I. and others., 1972, Initial Reports of the Deep Sea Drilling Project, Volume. 11, Washington D.C., U.S. Government Printing Office, 1077 p.

Hornbach, M.J., Lavier, L.L., and Ruppel, C.D., 2007, Triggering mechanism and tsunamogenic potential of the Cape Fear Slide complex, U.S. Atlantic margin:  Geochemistry Geophysics Geosystems, doi:10.1029/2007GC001722, 16p.

Kneller, B., 2003, The influence of flow parameters on turbidite slope channel architecture:  Marine and Petroleum Geology, v. 20, p. 901-910.

Lee, H.J., 2009, Timing of occurrence of large submarine landslides on the Atlantic Ocean margin:  Marine Geology, v. 264, 9. 53-64.

Locker, S.D. and Laine, E.P., 1992, Paleogene-Neogene depositional history of the middle U.S. Atlantic continental rise:  Mixed turbidite and contourite depositional systems:  Marine Geology, v. 103, p. 137-164.

McCave, I.N. and Tucholke, B.E., 1986, Deep current-controlled sedimentation in the western North Atlantic, in Vogt, P.R. and Tucholke, B.E., eds., The western North Atlantic Region:  Boulder, Colorado, Geological Society of America, Geology of North America, v. M, p. 451-468.

Mitchell, N.C., 2004, Form of submarine erosion from confluences in Atlantic USA continental slope canyons:  American Journal of Science, v. 304, p. 590-611.

Mitchell, N.C., 2006, Morphologies of knickpoints in submarine canyons:  Geological Society of America Bulletin, v. 118, p. 589-605.

Mountain, G.S. and Tucholke, B.E., 1985, Mesozoic and Cenozoic geology of the U.S. Atlantic continental slope and rise, in Poag, C.W. ed., Geologic Evolution of the United States Atlantic Margin:  New York, Nostrand Reinhold, p. 293–341.

Newton, J.G. and Pilkey, O.H., 1969, Topography of the continental margin off the Carolinas:  Southeastern Geology, v. 10, p. 87-92.

Pike, R.J., 1995, Geomorphometry-progress, practice and prospect.  Zeitschrift für Geomorphologie, Supplementband 101, p. 221-238.

Pilkey, O.H. and Cleary, W.J., 1986, Turbidity sedimentation in the northwestern Atlantic Ocean basin, in Vogt, P.R. and Tucholke, B.E. (eds.), v. M, The Western North Atlantic region, Geology of North America, Geological Society of North America, Boulder, CO, p.437-450.

Pirmez, C., Beaubouef, R.T., Friedmann, S.J., and Mohrig, D.C., 2000, Equilibrium profile and base level in submarine channels:  Examples from late Pleistocene systems and implications for the architecture of deep-water reservoirs, in Weimer, P., Slatt, R.M., Coleman, J.M., Rosen, N.C., Nelson, H. Bouma, A.H., Styzen, M.J., and Lawrence, D.T., eds., Deep-water reservoirs of the world:  Houston, Texas, Gulf Coast Section SEPM, p. 782-805.

Popenoe, P., Schmuck, E.A., and Dillon, W.P., 1993, The Cape Fear landslide:  Slope failure associated with salt diapirism and gas hydrate decomposition, in Schwab, W.C., Lee, H.J., and Twichell, D.C., eds., Submarine landslides:  Selected studies in the U.S. Exclusive Economic Zone, U.S. Geological Survey Bulletin 2002, p. 40-53.

Popenoe, P. and Dillon, W.P., 1996, Characteristics of the continental slope and rise off North Carolina from GLORIA and seismic-reflection data:  The interaction of downslope and contour current processes, in Gardner, J.V., Field, M.E., and Twichell, D.C., eds., Geology of the United States’ seafloor:  The view from GLORIA, Cambridge, U.K., Cambridge University Press, p. 59-79.

Pratson, L. F. and Laine, E.P., 1989, The relative importance of gravity-induced versus current-controlled sedimentation during the Quaternary along the Mideast U.S. outer continental margin revealed by 3.5 kHz echo character:  Marine Geology, v. 89, p. 87-126.

Prior, D.B., Coleman, J.M., and Doyle, E.H., 1984, Antiquity of the continental slope along the mille Atlantic margin of the United States:  Science, v. 223, p. 926-928.

Rodriguez, N.M. and Paull, C.K., 2000, 14C dating of sediment of the uppermost Cape Fear slide plain:  Constraints on the timing of this massive submarine landslide, in Paull, C.K., Matsumoto, R. Wallace, P.J. and Dillon, W.P., eds., Proceedings of the Ocean Drilling Program, Scientific Results, v. 164, p. 325-327.

Rona, P.A., 1969, Linear “lower continental rise hills” off Cape Hatteras:  Journal of Sedimentary Petrology, v. 39, p. 1132-1141.

Rona, P.A. and Clay, C.S., 1967, Stratigraphy and structure along a continuous seismic reflection profile from Cape Hatteras, North Carolina, to the Bermuda Rise:  Journal of Geophysical Research, v. 72, p. 2107-2130.

Rona, P.A., Schneider, E.D., and Heezen, B.C., 1967, Bathymetry of the continental rise off Cape Hatteras:  Deep-Sea Research, v. 14, p. 625-633.

Schneider, E.D., Fox, P.J., Hollister, C.D., Needham, H.D. and Heezen, B.C., 1967, Further evidence of contour currents in the western North Atlantic:  Earth and Planetary Science Letters, v. 2, 351-359.

Sheridan, R.E., Enos, P., Gradstein, F., and Benson, W.E., 1987.  Mesozoic and Cenozoic sedimentary environments of the western North Atlantic, in Benson, W.E., Sheridan, R.E., et al., Initial Reports of the Deep Sea Drilling Project v. 44: Washington D.C., U.S. Government Printing Office, p. 971-979.

Smith, W.H.F. and Sandwell, D.T., 1997. Global seafloor topography from satellite altimetry and ship depth soundings:  Science v. 277, p. 1957-1962.

Tucholke, B.E. and Laine, E.P., 1982, Neogene and Quaternary development of the lower continental rise off the central U.S. East Coast, in Watkins, J.S. and Drake, C.L., eds., Studies in continental margin geology:  Tulsa, OK, American Association of Petroleum Geologists Memoir 34, p. 295-305.

Twichell, D.C., Chaytor, J.D. ten Brink, U.S. and Buczkowiski, B., 2009, Morphology of late Quaternary submarine landslides along the U.S. Atlantic continental margin:  Marine Geology, v. 264, p. 4-15. 

Region