Sediment transport processes at the head of Halibut Canyon, eastern Canada margin: An interplay between internal tides and dense shelf-water cascading

Marine Geology 341 (2013) 14–28

Contents lists available at SciVerse ScienceDirect

Marine Geology journal homepage: www.elsevier.com/locate/margeo

Sediment transport processes at the head of Halibut Canyon, eastern Canada margin: An interplay between internal tides and dense shelf-water cascading Pere Puig a,⁎, Blair J.W. Greenan b, Michael Z. Li c, Robert H. Prescott d, David J.W. Piper c a

Institut de Ciències del Mar (CSIC), Passeig Marítim de la Barceloneta, 37-49, 08003 Barcelona, Spain Fisheries and Oceans Canada, Bedford Institute of Oceanography, P.O. Box 1006, Dartmouth, Nova Scotia B2Y 4A2, Canada c Geological Survey of Canada (Atlantic), Bedford Institute of Oceanography, P.O. Box 1006, Dartmouth, Nova Scotia B2Y 4A2, Canada d Prescott and Zou Consulting, 6 Glenn Dr., Halifax, Nova Scotia B3M 2B9, Canada b

a r t i c l e

i n f o

Article history: Received 1 December 2012 Received in revised form 6 May 2013 Accepted 12 May 2013 Available online 19 May 2013 Communicated by: Dr. J.T. Wells Keywords: submarine canyon sediment transport internal tide dense shelf water cascade gravity flow

a b s t r a c t To investigate the processes by which sediment is transported through a submarine canyon incised in a glaciated margin, the bottom boundary layer quadrapod RALPH was deployed at 276-m depth in the West Halibut Canyon (off Newfoundland) during winter 2008–2009. Two main sediment transport processes were identified throughout the deployment. Firstly, periodic increases of near-bottom suspended-sediment concentrations (SSC) were recorded associated with the up-canyon propagation of the semidiurnal internal tidal bore along the canyon axis, carrying fine sediment particles resuspended from deeper canyon regions. The recorded SSC peaks, lasting less than 1 h, were observed sporadically and were linked to bottom intensified up-canyon flows (~40 cm s−1) concomitant with sharp drops in temperature. Secondly, sediment transport was also observed during events of intensified down-canyon current velocities that occurred during periods of sustained heat loss from surface waters, but were not associated with large storm waves. High-resolution velocity profiles throughout the water column during these events revealed that the highest current speeds (~1 m s−1) were centered several meters above the sea floor and corresponded to the region of maximum velocities of a gravity flow. Such flows had associated low SSC and cold water temperatures and are interpreted as dense shelf water cascading events channelized along the canyon axis. Sediment transport during these events was largely restricted to bedload and saltation, producing winnowing of sands and fine sediments around larger gravel particles. Analysis of historical hydrographic data suggests that such gravity flows are not related to the formation of coastal dense waters advected towards the outer shelf that reached the canyon head. Rather, the dense shelf waters appear to be generated around the outer shelf, where convection during winter is able to reach the sea floor and generate a pool of near-bottom dense water that cascades into the canyon during one or two tidal cycles. A similar transport mechanism is likely to occur in other submarine canyons along the eastern Canadian margin, as well in other canyoned margins where winter convection can reach the shelf-edge. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Submarine canyons are morphological features that are found on many continental margins, acting as preferential conduits for transport of sediment from continental shelves towards deep-sea environments. During Plio-Pleistocene lowstands of sea level, sediment–gravity mechanisms (i.e. turbidity currents, debris flows) dominated transport through submarine canyons, funneling large volumes of terrigenous sediment to deeper parts of the continental margins (Shanmugam et al., 1985; Piper and Normark, 2009). Although Holocene sea-level rise has reduced drastically the supply of sediments to submarine canyons, it is widely ⁎ Corresponding author. Tel.: +34 93 230 9518; fax: 34 93 230 9555. E-mail address: [email protected] (P. Puig). 0025-3227/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.margeo.2013.05.004

recognized that canyons at present continue to be preferential conduits for the transfer of sediments from the shelf to the deep ocean. During the last decades, several studies have provided information about contemporary sediment-transport processes acting within submarine canyons by means of analysis of combined currents and suspended-sediment concentration data. Most of them have been conducted using moored instruments placed at several heights above the sea floor, although few studies also have involved bottomboundary-layer measurements. Storm-induced sediment gravity flows (Xu et al., 2002; Puig et al., 2003, 2004a; Xu et al., 2010; Martín et al., 2011; Masson et al., 2011; Mulder et al., 2012), enhanced off-shelf advection during storms (Carson et al., 1986; Martín et al., 2006), hyperpycnal flows and failures from recently deposited fluvial sediments (Khripounoff et al., 2012), dense shelf water cascading (Canals et al.,

P. Puig et al. / Marine Geology 341 (2013) 14–28

15

along the margin as icesheets crossed the shelf during major glaciations since 0.5 Ma. The general style of Quaternary deposition on the slope is principally from proglacial plume fallout deposits (mud with dispersed ice-rafted debris) with occasional thin turbiditic sand beds (Piper et al., 2005). Canyon floors locally record rare Holocene turbidity current deposits (Savoye et al., 1990; Jenner et al., 2007). The most remarkable sediment transport event in this margin occurred during the 1929 Grand Banks earthquake, which triggered a slump that evolved into a large turbidity current and broke a series of submarine cables and created a >150 km3 sandy turbiditic deposit (~1 m thick) in the Sohm abyssal plain (Heezen and Ewing, 1952; Piper et al., 1988; Mosher et al., 2010). Armitage et al. (2010) conducted a detailed analysis of 3D seismic reflection data with 2D high-resolution seismic-reflection profiles and shallow sediment cores in Halibut Canyon, off southwest Newfoundland (Fig. 2). The contemporary sediment transport processes operating on this canyon, however, could not be properly discerned, and the shelf-to-canyon sediment transport was attributed

100 200

2006; Ogston et al., 2008; Puig et al., 2008), and trawling-induced resuspension (Palanques et al., 2006; Martín et al., 2007; Puig et al., 2012) largely dominate the present-day transfer of sediment through canyons. Additionally, internal waves periodically resuspend ephemeral deposits within canyons and contribute to disperse particles or retain and accumulate them in specific regions (e.g., Gardner, 1989a, b; Puig et al., 2004b; de Stigter et al., 2007). The eastern Canadian continental slope is in places deeply dissected by a dendritic network of submarine canyons, which are separated by steep intercanyon ridges (e.g. Hesse et al., 1999; Mosher et al., 2004; Fig. 1). A late Pliocene lowstand of sea-level resulted in increased glacio-fluvial sediment transport to the slope and the initiation of widespread canyon cutting (Piper et al., 1987). Canyon incision increased during Early Pleistocene, resulting from either longer episodes of sea-level lowstand or an increase in sandy sediment delivered to the shelf break, although the overall style of sedimentation continued to be prodeltaic (Piper and Normark, 1989). The present day submarine canyons are a product of Quaternary ice-related processes that operated

Newfoundland

48o N

Sta. 27

el han n Nickerson Bank

nC

200

100

Ava lo

47 N

0 20 0

10

50

o

50

100

Green Bank

dd ock

St. Pierre Bank

Ha

Halibut

46o N

1 00

Ch an ne l

Channe

l

Buoy C44251

The Grand Banks of Newfoundland 0

Fig. 2

10

100

50

45 N

n ntia ure el La hann C

o

RALPH 100

200 500 1000 10 2000

56o W

0

20 100 500 0 0 3000

1929 epicentre

55o W

20

00

54o W

Buoy C44138

53o W

52o W

Fig. 1. Bathymetric map of the eastern Canada continental margin south of Newfoundland showing the position of RALPH deployment (star), the two meteorological buoys used in this paper (circles), the epicenter of the 1929 earthquake (diamond) and the long-term monitoring station 27 (triangle). Rectangle shows the canyon head area, enlarged in Fig. 2.

16

P. Puig et al. / Marine Geology 341 (2013) 14–28

45o 20'

45o 10' 10 0 15 0 200

45o 00'

10

0

150 0 20

500

200 0

44o 50'

1000

25

44o 40'

00

1500

55o 40'

55o 30'

55o 20'

55o 10'

55o 00'

Fig. 2. High-resolution bathymetry from Halibut Canyon showing the position of RALPH deployment (star) and the positions of CTD profiles from ISDM data inventory conducted in this area during winter (dots). CTDs in water depths deeper than 300 m are not shown. Profiles illustrated in Fig. 11 are shown as black dots. Bathymetric data from regions deeper than 400 m come from a multibeam survey (see Armitage et al., 2010), while shallower bathymetry comes from the interpolation from the raw soundings used to create the hydrographic charts of this region.

to storms and tidal currents. Bottom-boundary-layer measurements on the adjacent continental shelf (Green Bank) indicated that tidal current alone during fair-weather conditions is not strong enough to cause significant sediment resuspension and that wave oscillatory flows and wind-driven currents during storms are responsible for most of the sediment resuspension and transport (Li et al., 2011). However, the exact mechanisms involved in the shelf-to-canyon sediment delivery in this margin are still unknown. The main objective of this paper is to present detailed bottomboundary-layer time-series observations recorded at the head of Halibut Canyon and identify the contemporary sediment transport processes that contribute to the sediment dispersal along its canyon axis. These measurements also represent a data set that can provide insights into the dominant mode of sediment transport in the submarine canyons of the eastern Canadian margin during present conditions, as well in other submarine canyons elsewhere incised in glaciated continental margins. 2. Methods As part of the ‘Nearbed Wave and Current Forcing and Sediment Dynamics’ project at the Geological Survey of Canada, the bottom-

boundary-layer quadrapod RALPH was deployed during winter 2008–2009 at the head of Halibut Canyon, downslope of Halibut Channel, at a depth of 276 m (Fig. 1). The continental slope in this region is cut by a dendritic system of deep canyons with sharp intercanyon ridges. More precisely, RALPH was placed in the head of the western branch of Halibut Canyon, here informally termed West Halibut Canyon (Fig. 2). The heads of these canyons are approximately at 150–200 m below sea-level, canyon walls are steep (up to 40°) and gullied, and canyon widths vary between 1.5 and 4.0 km (Armitage et al., 2010). Deeper downslope across the margin, Halibut Canyon, Haddock Canyon and Green Canyon coalesce into Grand Banks Valley at 2500 m depth. RALPH is an autonomous bottom boundary layer quadrapod frame equipped with various sensors for observing and recording hydrodynamic and sediment-transport parameters on the seafloor such as waves, currents, turbulence, suspended sediment concentration, and bedforms (Heffler, 1996; Li and Heffler, 2002). The sensors mounted on the quadrapod during this deployment, and their heights above the seafloor and sampling strategy are listed in Table 1. RALPH was deployed on 12 December 2008 and recovered on 11 March 2009, but data collection was planned for only one month and the duration of the various time series was limited by battery

P. Puig et al. / Marine Geology 341 (2013) 14–28 Table 1 List of sensors mounted on the quadrapod RALPH during the winter 2008–2009 deployment at the head of the West Halibut Canyon with sampling strategy and heights above the seafloor. Sensors

Sampling strategy

Height above bottom

RALPH pressure transducer (500 psi) RALPH compass

30 min burst at 2 Hz every 60 min 30 min burst at 2 Hz every 60 min 30 min burst at 2 Hz every 60 min 30 min burst at 2 Hz every 60 min 3 min burst at 1 Hz every 60 min 15 min burst at 2 Hz every 60 min 3 min burst at 4 Hz every 60 min 1 min profile at 1 Hz every 3 min 1 min profile at 1 Hz every 3 min 10 s clip and 1 photo every 30 min

156 cm

RALPH tilt/roll sensor 6 D&A optical backscatter sensors (OBS) 2 Alec electromagnetic current meters Nobska travel-time current meter Mesotech 1 MHz acoustic backscatter sensor (ABS) Nortek 2 MHz ADCP (downward looking) RDI 300 kHz ADCP (upward looking) Sony HD-SR8 Camera (downward looking)

152 cm 152 cm 13, 34, 52, 72, 92, 111 cm 30, 50 cm 100 cm 132 cm (1 cm cell size) 130 cm (15 cm cell size) 175 cm (4 m cell size) 133 cm

capacity or instrument failures. The main RALPH logger, which controls the pressure, compass, tilt-roll and optical backscatter sensors (OBS) lasted 20 days. The 1 MHz acoustic backscatter sensor (ABS), which was also controlled by the main RALPH logger, lasted only 10 days due to a data storage issue. The Nobska travel-time current meter and the Nortek 2 MHz downward-looking acoustic Doppler current profiler (ADCP) logged 33 days. The Teledyne RDI 300 kHz upward-looking ADCP recorded for 37 days. The downward-looking camera collected images during 89 days, covering the entire deployment. The downward-looking ADCP successfully recorded profiles of 1-minute-mean velocity in 10 bins every 3 min. The bin size was 15 cm and only the first 6 bins provided valid data due to interferences of the sea floor in the remainder. The upward-looking ADCP also recorded profiles of 1-minute-mean velocity every 3 min in 35 bins of 4 m, covering half of the water column, from approximately 270 m water depth (i.e. 6 m above the bottom) to 130 m water depth (i.e. shelf-break depths). Some of the velocity profiles, however, did not meet the criteria for good quality in the bins furthest from the transducer head because of a low signal-to-noise ratio. This is not uncommon for these types of acoustic profilers that depend on sufficient numbers of scatterers in the water column to produce a return signal. The single point velocimeters mounted on the quadrapod (i.e., the 2 stand-alone Alec electromagnetic current meters and the self-contained Nobska current meter) sampled in burst mode at 1 and 2 Hz respectively, and the recorded data was averaged hourly (see details in Table 1). RALPH's bottom-looking camera showed that the surface sediment within the canyon head was a mixture of sand and gravel partially covered by finer sediments. However, bulk samples were not taken at the deployment site for calibration purposes. Instead, OBS and ABS were calibrated in the laboratory using sediments with similar granulometric characteristics to provide suspended sediment concentrations (SSC). The isolated spikes in the OBS records that were not observed simultaneously in all devices were removed from the series. These spikes were attributed to interferences of the optical measurements by living organisms (fish and shrimps), which were observed in most of the video footages. In situ temperature was recorded independently by all five velocity sensors mounted on the tetrapod and also by RALPH's internal thermistor, which unlike the others was not in direct contact with seawater. Wind and wave conditions in the study area during the deployment period were recorded by two buoys from Environment Canada

17

(Fig. 1). Buoy C44251, located on Nickerson Bank in 71 m water depth off the south Newfoundland coast, and buoy C44138, located on the continental slope southwest of The Grand Banks in 1500 m water depth. Wind speed and direction, significant wave height, air temperature and sea surface temperature parameters were obtained directly from the standard meteorological data file. The sensible heat flux was computed from the buoy measurements (wind speed, dry-bulb air temperature, air pressure and sea surface temperature) using the MATLAB® air-sea toolbox (http://woodshole.er.usgs.gov/ operations/sea-mat/air_sea-html/index.html). The latent heat flux could not be calculated properly due to the lack of relative humidity measurements on the buoys. A hydrographic (i.e., CTD) profile was conducted at the time of RALPH deployment to characterize the properties of the water column within the canyon. Unfortunately, due to the fact that the recovery cruise took place on a vessel of opportunity, no CTD cast was conducted at the time of the instrument retrieval, which was almost two months after the time series record elapsed. Additionally, historical hydrographical profiles from the Grand Banks shelf south and east of Newfoundland, including the region of Halibut Channel and upper Halibut Canyon, were retrieved from the data inventory from Fisheries and Oceans Canada Integrated Science Data Management (ISDM) to determine the typical water properties of the study region during winter and their evolution throughout time. 3. Results 3.1. Forcing conditions The temporal evolution of wind vectors, significant wave heights, air and sea surface temperature measured by the meteorological buoys during the study period, and the computed sensible heat flux based on those parameters are shown in Fig. 3. Several storms occurred related to either dominantly southerly or northerly winds. Three major southerly storms occurred with significant wave heights above 6 m, with the peaks of the storms centered on 24 Dec 08, 3 Jan 09 and 15 Jan 09. Average wave periods during these storms (not shown) fluctuated around 12 s, with maximum periods of 16 s. Northerly storms did not generate large waves due to the short fetch, but they brought cold (and presumably dry) air to the study region. On three occasions, 20–21 Dec 08, 27–28 Dec 08 and 17–18 Jan 09, the air temperatures near the coast (Nickerson Bank) were below − 5 °C, while offshore (SW Grand Banks) they reached values below 0 °C. Sea surface temperatures in both meteorological buoys showed a continuous descending trend through the study period, with some marked sharp drops ranging between 1 and 2 °C. These drops were mainly observed on the coastal buoy on 20 Dec 08, 26 Dec 08 and 12 Jan 09, and on the offshore buoy on 21 Dec 08 and on 16 Jan 09, although the sea surface temperature record showed a continuous decreasing trend throughout the observational period. The sensible heat flux plots show the same temporal evolution as the air temperatures records. During most of the deployment, heat fluxes displayed negative values with occasional slightly positive fluxes of b100 W m−2 associated with persistently southerly winds bringing warm air to the region. Negative heat fluxes during northerly storms were below − 150 W m−2 for several days and reached a minimum value of − 290 W m−2 SW of Grand Banks on 20 Dec 08. 3.2. Bottom boundary layer measurements The temporal evolution of the near-bottom temperature, current speed and SSC at various levels above the sea floor in the bottom boundary layer is illustrated in Fig. 4. Temperature sensors were available in various devices mounted on RALPH and showed the same records with slight offsets in absolute values due to calibration issues. Near-bottom temperature at the deployment site during the

P. Puig et al. / Marine Geology 341 (2013) 14–28

Dry bulb air temp. (oC)

Sig. wave height (m)

Wind (m s-1)

18

20 0 -20 10 8 6 4 2 0 15 10 5 0 -5 -10

Sea surface temp. (oC) Sens. heat flux (W m-2)

Nickerson Bank

15

100

SW Grand Banks

10

5

0

0 -100 -200 -300

13-Dec-08

20-Dec-08

27-Dec-08

3-Jan-09

10-Jan-09

17-Jan-09

Fig. 3. Time series during the study period of the wind velocity vectors, significant wave height, air temperature, sea surface temperature measured by the two meteorological buoys, and the sensible heat flux computed from these parameters. Wind data only shown from the offshore buoy (SW Grand Banks). The wind vectors indicate the direction from which the wind blows.

study period ranged from 8 to 0 °C and oscillated mainly at semidiurnal tidal frequencies. Tidal temperature oscillations were generally ~ 1 °C and each tidal cycle displayed a gradual increase when currents were directed down-canyon and a sharp drop associated with a rapid shift to the up-canyon flow (not shown in Fig. 4). Larger drops in temperature (~2 °C) were observed occasionally, mainly during the periods of cold air temperatures and negative sensible heat fluxes. The largest and sharpest drops in temperature (~ 4 °C) were observed towards the end of the deployment period in four consecutive tidal cycles from 13 to 14 Jan 09. Currents in the bottom boundary layer (BBL) also fluctuated at semidiurnal tidal frequencies, being preferentially oriented along the canyon axis. The fluctuations in the velocity magnitude and direction in the first bin of the upward-looking ADCP, at 6 m above the bottom (mab), agreed very well with the other current meters on RALPH. During periods of normal conditions, peak speeds were often > 40 cm s−1 at 6 mab, > 30 cm s−1 at 1 mab and >20 cm s−1

at 0.3 mab, with higher velocities being recorded during up-canyon flows (not shown in Fig. 4). Several periods of high current velocities were observed in the records with maximum speeds ~ 1 m s−1 at 6 mab and ~ 60 cm s−1 at 1 and 0.3 mab. Current velocities during these events were also tidally dominated, but the maximum speeds were directed down-canyon (not shown in Fig. 4). They occurred on 21 Dec 08, 27 Dec 08 and 15–16 Jan 09 during or immediately after northerly storm periods characterized by large drops in air temperature and sensible heat flux, and not during the peak of southerly storms that caused large surface waves (Figs. 3 and 4). Periodic increases of near-bottom SSC at the canyon head were observed sporadically. There is a general correlation between the down-canyon high current events and some of the SSC peaks (e.g., 21 Dec, 24–25 Dec, and 27 Dec 08), although only small increases of ~ 100 mg l−1 were recorded in these events. Peaks of highest SSC mainly occurred during periods of low current velocities (e.g., 13–15 Dec 08). These increases appeared as spikes that lasted

Temp. 1.5 mab (oC)

P. Puig et al. / Marine Geology 341 (2013) 14–28

19

10 8 6 4 2

Nortek

0

RDI

Speed 6 mab (cm s-1)

100

Speed 1 mab (cm s-1)

100

Speed 0.3 mab (cm s-1)

-2

100

80 60 40 20 0

80 60 40 20 0

80 60 40 20 0

SSC (g l-1)

0.4 0.3

0.52 mab 0.13 mab

0.2 0.1 0

13-Dec-08

20-Dec-08

27-Dec-08

3-Jan-09

10-Jan-09

17-Jan-09

Fig. 4. Time series of BBL measurements at the head of West Halibut Canyon showing near-bottom temperature records from two independent devices, current speed at 6 mab recorded by the upward-looking ADCP, at 1 mab recorded by the downward-looking ADCP and at 0.3 mab recorded by the Alec electromagnetic current meter, and SSC records from two OBSs at 0.13 and 0.52 mab.

less than 1 h (i.e. the sampling interval) and were recorded simultaneously in all OBS sensors. Estimated SSC at 0.13 mab during these spikes was generally ~ 200 mg l−1, although sporadically reaching > 300 mg l−1. 3.3. Water column current profiles As described for the BBL measurements, the current magnitude and direction throughout the water column, from 270 m to 130 m water depth, were influenced by the semidiurnal tide, but showed a quite complex hydrodynamic pattern with a complicated vertical structure, alternating periods of high velocities at mid-water depths with others much closer to the bottom (Fig. 5). In general, during

the periods of enhanced near-bottom (i.e. BBL) down-canyon current velocities, and in particular during the 21 and 27 Dec 08 events, the highest velocities (~ 1 m s−1) were observed higher in the water column, but centered between 10 and 40 mab, with weaker velocities higher in the water column. Detailed analysis of the vertical structure of the current velocity profile indicates that these profiles corresponded to the maximum velocity region of a gravity flow. The velocity diminishes toward the bottom in the wall region within the frictional boundary layer, and upwards into the water column within the jet region. Maximum current velocities during the preceding or subsequent tidal cycle were recorded much higher in the water column and generally were centered at shelf-break depths (130–150 m) or occurred as isolated mid-waters jets (Fig. 5). Other minor gravity flows showing similar

20

P. Puig et al. / Marine Geology 341 (2013) 14–28

Fig. 5. Upward-looking ADCP current speed (top) and direction (bottom) as a function of depth and time throughout the entire deployment (ranges indicated by color bars). White areas within the range of the ADCP indicate no data due to low signal to noise ratio. The canyon axis orientation is N–S. The time range of Figs. 6 and 8 are indicated by the arrows at the top of plot.

vertical velocity profiles were observed throughout the record (e.g., 31 Dec 08 or 08 Jan 09) although maximum velocities were weaker (60–80 cm s− 1) and generally located farther away from the sea floor (Fig. 5). Towards the end of the record, between 10 and 14 Jan 09, numerous tidally dominated intensified flows directed to the south (i.e. down-canyon) were observed at shelf-break depths, reaching high velocities down to 240 m depth but without creating any near-bottom current increase (Fig. 5). After this period, and coinciding with a southerly storm on 15 Jan 09 (Fig. 3), current velocities at shelf-break depths decreased and were directed towards the north (i.e. up-canyon) for several hours. After the passage of the storm, two consecutive gravity flows occurred during 15–16 Jan 09, which accounted for the highest velocities of the records (125 cm s−1). In the first one, the highest velocities were observed in the lowermost 20 m of the current profile, whereas in the second one on 16 Jan, the highest velocities were observed higher up in the water column around 220 m depth, but high values were observed in most of the sampled water column, from 270 to 160 m depth (Fig. 5). 3.4. Internal tides The peaks of high SSC recorded during periods of low current velocities occurred at the time of the sharp drops in water temperature during the sudden change from down-canyon to up-canyon current direction, indicative of resuspension by internal tides. To illustrate this process in detail, a zoom of the first two days of the record, when SSC peaks were more frequent, is shown in Fig. 6. Overall, the currents were quite weak throughout the water column during this period of the deployment, although the semidiurnal tidal

cycles can be clearly distinguished in the upward-looking ADCP records, showing alternating down-canyon and up-canyon currents. Higher currents were observed at shelf-break depths (130–150 m) and occasionally at deeper levels, but near the bottom, a noticeable intensification of the currents up to ~40 cm s−1 was observed in the lowermost 20 m at the time of the up-canyon propagation of the cold water internal bore, accompanied by a sharp drop in temperature of ~1 °C to values ~5 °C. This low temperature persisted for several hours after which temperature gradually rose to previous higher values ~6 °C. The associated SSC spike was well recorded by the OBSs during the two first internal bores, but the optical sensors missed the third one on 14 Dec at 09:30 UTC, when the bottom intensified up-canyon flow was more evident. Nevertheless, the bottom-looking camera, which took pictures every 30 min, did capture the increase of water turbidity associated with the passage of the tidal bore (Fig. 7). Looking at the three consecutive pictures of the seafloor, the one collected at 09:00 UTC showed a clear image of the seafloor, while the one taken right at the passage of the up-canyon tidal bore at 09:30 UTC showed turbid water. At 1000 UTC, the water was clear and the bottom could be seen again, although it was less transparent than an hour before. Given the short-lived character of this transport mechanism (less than 30 min), no such clear example was seen in the rest of the camera records, although in many pictures the water was slightly turbid or the cobbles suddenly appeared coated with fine sediments that were rapidly grazed by benthic organisms, as seen in the 10 s video clips. Although OBS records reached maximum SSC > 200 mg l−1, visual observations from the bottom-looking video camera suggest that these concentrations may be overestimated due to the coarse grained sediment used to calibrate them, and therefore be caused by the higher response of OBS to fine particles in suspension. Video images

P. Puig et al. / Marine Geology 341 (2013) 14–28

21

Fig. 6. Time series evolution from a period of low background velocity during which up-canyon transport of fine sediment by semidiurnal internal tides occurred. Upward-looking ADCP current speed and direction as a function of depth and time are plotted in the top and middle panels (ranges indicated by color bars). Temperature measured at the ADCP transducer head (red line) and SSC at 0.13 mab (blue line) are plotted in the bottom panel. The canyon axis orientation is N–S. The time ranges of pictures in Fig. 7 are indicated by the arrows at the top of plot.

also suggest that particles transported up-canyon by the internal tidal bore are mostly flocculated, since large individual aggregates could be identified easily.

3.5. Gravity flows The intensified down-canyon current events observed in the BBL measurements (Fig. 4) and at several meters above in the water column (Fig. 5) were clearly linked with periods of cold air temperatures and sustained heat loss and not with periods of major wave storm events (Fig. 3). Looking in detail at the velocity profile structure of these events, they corresponded to gravity-driven density flows that occurred just during one or two tidal cycles. The detailed time series of the gravity flow on 21 Dec 08 are illustrated in Fig. 8. The maximum velocities were up to ~ 100 cm s−1 and lasted 6 h (i.e. the down-canyon tidal phase), with the central position of the jet located between 10 and 40 mab. Maximum velocities in the preceding and following tidal cycle were observed much higher into the water column, located around the shelf break depth (130–150 m) and also directed southwards (i.e., offshore and down-canyon). The same sequence of tidally dominated down-canyon flows was observed in the gravity flow on 27 Dec 2008 (Fig. 5), with maximum speeds around the shelf break depth in the previous and following tidal cycle. The two consecutive events in 15–16 Jan 09 showed a more complicated velocity profile structure (Fig. 5), although both showed a well-defined jet and wall region characteristic of gravity flows.

During the 21 Dec 08 gravity flow, a larger drop in near-bottom temperature compared to the ones created by internal tides, was recorded, while the increase in the SSC was lower (Fig. 8). The temperature time series indicates a very stable environment prior 19:00 UTC on 20 Dec 2008 with the water temperature rising slowly from 5.7 to 6.1 °C and then abruptly dropping to 2.7 °C. This drop appears to be related to the off-shelf and down-canyon flow that had the highest velocities up in the water column around shelf-break depths. Following this, the temperature slowly rose to 6.5 °C and then dropped again abruptly to about 2.6 °C as the near-bottom gravity flow with current speeds in excess of 100 cm s−1 was observed with the ADCP (07:00–13:00 UTC on 21 Dec 2008). This is followed by a second drop in temperature to 2.3 °C and a very sharp rise in temperature as the gravity flow was terminated at 16:00 UTC. The drop in temperature again at about 22:00 UTC seems to be linked to an increase in current speed in the water column near the top of the ADCP range during the following off-shelf tidal cycle at shelf-break depths (Fig. 8). SSC at 0.13 mab increased following the temperature fluctuations. SSC during the first temperature drop reached ~ 50 mg l−1, while at the beginning of the gravity flow SSC was slightly higher and reached ~ 75 mg l−1, decreasing progressively through time (Fig. 8). These low SSC responses to the strong current velocities are a consequence of the difficulty of resuspending the coarse grained surface sediments on the outer shelf and within the upper canyon head and to the dilution of the easily resuspendable fine particles up into the water column. Nonetheless, seafloor photos and video footage showed that this process contributed to the

22

P. Puig et al. / Marine Geology 341 (2013) 14–28

Fig. 7. Seafloor pictures taken on 14 Dec at 09:00 UTC (top) 09:30 UTC (middle) and 10:00 UTC (bottom) showing the turbidity increase produced by an up-canyon tidal bore. The down-canyon direction is towards the upper-right corner of the picture. See text for discussion. Video clips from the same sampling intervals are available as supporting material in the web version of this article.

down-canyon transport of the coarse bottom sediments during the period of higher velocities, being winnowed around the interstices of the gravel and cobbles (Fig. 9). The picture taken by the bottomlooking camera on 20 Dec at 18:30 UTC, just before BBL current increased, showed a sediment surface with large cobbles being partially covered by finer sediments with crawl marks from benthic organisms. At the time of the gravity flow, on 21 Dec 2008 at 08:30 UTC, the water looked turbid but the seafloor was still visible on the

picture. It showed comet marks and flute structures typical of bedload transport and saltation with a clear down-canyon direction (i.e., towards the upper-right corner of the image). After the transport event, at 16:30 UTC, some of these bedforms still persisted on the seafloor and the sediment surface was covered by a coarser lag deposit (Fig. 9). The ABS burst measurements recorded during this gravity flow captured the resuspension plumes created by winnowing of the surface sediment (Fig. 10). Resuspension plumes were randomly

P. Puig et al. / Marine Geology 341 (2013) 14–28

23

Fig. 8. Time series evolution from a period of high down-canyon velocity and sediment transport during a gravity-driven flow. Upward-looking ADCP current speed and direction as a function of depth and time are plotted in the top and middle panels (ranges indicated by color bars). Temperature measured at the ADCP transducer head (red line) and SSC at 0.13 mab (blue line) are plotted in the bottom panel. The canyon axis orientation is N–S. The time ranges of pictures in Fig. 9 and from Fig. 10 are indicated by the arrows at the top of plot.

distributed along the burst record, reached SSC ~ 1 g l−1 near the seabed and were uplifted only a few decimeters away from the bottom, with their concentration diminishing rapidly. At 0.13 mab (i.e., the height of the lowermost optical sensor), SSC measured acoustically by the ABS during the resuspension events was ~ 50 mg l−1, slightly lower than the values measured by the OBS (Fig. 9), but showing a good agreement between both methods. 4. Discussion 4.1. Contemporary sediment transport processes The analyses of the time series collected by RALPH suggest two main hydrodynamic and sediment transport processes acting on the head of Halibut Canyon. The first occurs during periods of moderate current velocities associated with bottom intensified up-canyon flows, which indicates the presence of semidiurnal internal tides affecting the near-bottom SSC along the canyon axis. The second is associated with down-canyon gravity flows occurring during periods of storms with winds from the north, cold air temperatures and sustained heat loss. No transport events seem to occur in this canyon head during southerly storms causing large surface waves. An integrated study conducted by Gardner (1989a, b) in Baltimore Canyon using moored current meters and transmissometers clearly revealed that sediment from the canyon floor between 200 and 800 m was resuspended regularly when energy of internal tides

was focused along the canyon axis. Analysis of time series was used to describe a mechanism consisting of a bore of cold water with a turbulent head moving up-canyon that resuspended sediments, resulting in a sharp peak of SSC at the beginning of the event, followed by a smaller and more continuous increase in SSC that corresponded to the downcanyon advection of the resuspended particles. The same resuspension mechanism was inferred in Hudson Canyon (Hotchkiss and Wunsch, 1982) and observed to occur in the Guadiaro Canyon (Puig et al., 2004b) and in Nazaré Canyon (de Stigter et al., 2007). The data collected at the head of Halibut Canyon suggest the occurrence of the same process, although the sampling interval of the OBSs was too long (60 min vs. the 5 min interval used by Gardner in Baltimore Canyon) to clearly depict the internal wave resuspension mechanism, and only the bursts that coincided in time with the passage of the internal bore captured the observed SSC peaks shown in Fig. 4. Based on the coarse surface sediment at the deployment site, these particles seem to have been resuspended from deeper parts of the canyon, where surface sediments are much finer (Armitage et al., 2010). Comparing the temperature drop down to ~ 5 °C at the time of the passage of the internal tidal bore and the data from the historical CTD profiles, we can estimate that the water was coming from around 500–600 m depth, suggesting a vertical internal tide excursion of ~ 200 m. The low SSC measurements recorded during the down-canyon gravity flows indicate that their increased density cannot be explained by large amounts of sediments in suspension, while the decrease in

24

P. Puig et al. / Marine Geology 341 (2013) 14–28

Fig. 9. Seafloor pictures take on 20 Dec at 18:30 UTC (top) and on 21 Dec at 08:30 UTC (middle) and at 16:30 UTC (bottom), before, during and after a down-canyon gravity flow, respectively. The down-canyon direction is towards the upper-right corner of the picture. See text for discussion. Video clips from the same sampling intervals are available as supporting material in the web version of this article.

temperature observed in the records (Fig. 8) presumably corresponded to the formation of cold dense waters over the shelf that eventually cascaded into the canyon head. Unfortunately, no conductivity measurements were available on RALPH during this deployment to determine the density of the seawater at the time of the gravity flows. However, this interpretation is supported by the fact that the gravity flows were recorded during or immediately after periods of cold air temperature and sustained heat loss that could trigger convection and eventually increase the density of the near-bottom shelf water.

4.2. Dense shelf water cascading on the eastern Canadian margin Dense shelf water cascading (DSWC) is a meteorologically-driven oceanographic phenomenon occurring not only on high latitude continental margins, but also on mid latitude and tropical margins (Ivanov et al., 2004). DSWC is a specific type of buoyancy driven current, in which dense water formed by cooling, evaporation or freezing in the surface layer over the continental shelf descends down the continental slope to a greater depth. The same phenomenon is also

P. Puig et al. / Marine Geology 341 (2013) 14–28

25

Fig. 10. ABS burst record collected on 21 Dec 2008 at 10:00 UTC during a gravity flow showing resuspensión plumes. The dashed white line represents the height of the lowermost OBS sensor (0.13 mab).

referred to as near-boundary convection or also as shelf-slope convection, to be differentiated from the open-sea convection (Killworth, 1983). DSWC sites have been identified in many continental margins around the world, mostly on the basis of cross-margin hydrographic sections (see Ivanov et al., 2004), since time series observations from this oceanographic phenomenon are scarce. The Gulf of Lions continental margin, in the northwestern Mediterranean Sea, is one of the known regions where this phenomenon occurs on an annual basis (Durrieu de Madron et al., 2005) and from where DSWC events have been continuously monitored during the last decades (e.g., Heussner et al., 2006; Palanques et al., 2006; Canals et al., 2006; Puig et al., 2008; Ribó et al., 2011; Palanques et al., 2012). Dense shelf waters are formed along the Gulf of Lions coast by cold and dry northerly winds (Tramontane and Mistral), and are transported along the shelf as a bottom layer towards the southwest, until they cascade preferentially into the Cap de Creus Canyon. These DSWC can be triggered or enhanced during storms events and last for several days or weeks, and the associated strong currents (up to 100 cm s−1) can induce erosion and resuspension of surface sediments on the outer shelf/upper slope (Canals et al., 2006; Ogston et al., 2008; Puig et al., 2008). Detailed observations for DSWC flows also exist in the Adriatic Sea. In this region, dense waters are formed in the northern Adriatic shelf, also by cold and dry northerly winds (Bora), and are advected along the shelf towards the south until cascading in the southern Adriatic Pit, partially via the Bari Canyon (Trincardi et al., 2007; Turchetto et al., 2007; Verdicchio et al., 2007). Although the time series observations at the head of Halibut Canyon indicate the occurrence of DSWC events, the main source of the dense waters on the eastern Canadian shelf is not evident. A potential source would be coastal dense waters generated in the bays of Newfoundland during formation of sea ice, which could be transported near the bottom with the inshore branch of the Labrador Current. The near-shore waters in most of the bays of the Newfoundland coast are ice-covered during the winter months, and therefore the upper layer temperatures and the entire water column at shallow depths are very near the freezing point during this period (Narayanan et al., 1996). This can increase coastal water densities after heat loss and brine rejection and trigger convection, which is known to occur to great depths (deYoung and Sanderson, 1995). These dense waters could be advected near-bottom and channelized through Avalon Channel and Halibut Channel, to eventually reach the head of Halibut Canyon (Fig. 1). Near-bottom current trajectories derived from drifters (Petrie and Anderson, 1983) and current bedforms on the sea floor (Dalrymple et al., 1992) suggest such a pathway. However, the analysis of the historical CTDs in the ISDM data inventory on the Grand Banks shelf south of Newfoundland, and particularly along Halibut Channel, do not show any evidence of a bottom dense layer during winter. Moreover, the continuous monitoring of the hydrographic properties of the water column at long-term monitoring station 27 in the Avalon Channel (triangle in Fig. 1) during winter 2008 indicated that the near-bottom density of the water advected by

the inshore branch of the Labrador Current in this particular year did not exceed 26.8 kg m−3 (Colbourne et al., 2009). The CTD cast conducted at the time of RALPH deployment indicates that the ambient potential density near the bottom at the canyon head site was ~ 27.3 kg m−3. Therefore, denser shelf waters than the ones advected by the inshore branch of the Labrador Current are required to eventually cascade into the canyon. Among all CTD casts from the ISDM data inventory available in our study area, only a few profiles located at the outer shelf on Halibut Channel (black dots in Fig. 2) showed near-bottom dense waters with similar values as the ones found at the RALPH site (Fig. 11). In fact, these near-bottom dense waters were warmer and saltier than the overlying waters and corresponded to slope waters presumably upwelled onto the outer shelf by internal waves (Sandstrom and Elliott, 1984; de Margerie and Lank, 1986). However, if convection is able to reach the seafloor around the shelf-break region it could increase the density of this upwelled slope water by reducing its temperature. Consequently, once this water is advected off-shelf during the ebb tidal cycle, its excess of density will cause it to progress downslope as a gravity flow and cascade into the canyon. Alternatively, and in absence of upwelled slope water over the shelf, convection alone could also generate a pool of near-bottom dense water along Halibut Channel, which could receive additional inputs from water flowing downslope from Saint Pierre Bank and Green Bank (Fig. 1) into Halibut Channel and advancing progressively offshore until cascading into Halibut Canyon. Although the maximum depth of convection during each of the periods of sustained heat loss from surface waters (Fig. 3) is unknown, it would presumably reach deeper into the water column through time during the winter, as shelf water becomes homogenous (Fig. 11). In that sense, the four sharp drops from 4 °C down to 0 °C observed on 13–14 Jan 09 (Fig. 3) seem to indicate that convection during this particular period reached down to 276 m (i.e., the depth where RALPH was deployed). After these four consecutive convection events, the southern storm on 15 Jan 2009 (Fig. 3) presumably blocked the release of the newly formed dense water pool, which could only cascade off the Halibut Channel in two consecutive gravity flows or DSWC events on 15–16 Jan 09 (Figs. 4 and 5), after the passage of the storm. These DSWC events recorded by RALPH on the West Halibut Canyon potentially also affected the East Halibut Canyon, since both are connected to Halibut Channel (Fig. 2). Similarly, Haddock Canyon, which is located nearby and exposed to the same meteorological regime, could have experienced the same DSWC flows but from near bottom dense shelf waters being released from Haddock Channel (Fig. 1). Observational data suggest that DSWC can occur in eastern Canadian submarine canyons several times during a winter season associated with the arrival of cold winds causing sustained heat loss and triggering convection. A major difference with previously described DSWC regions is that the dense waters are generated around the outer shelf and are evacuated rapidly towards the slope in one or

26

P. Puig et al. / Marine Geology 341 (2013) 14–28

0

Depth (m)

50 12 Dec 2008 11 Jan 2002 6 Feb 1991 14 Feb 1996 14 Feb 1996 25 Feb 1995 25 Feb 1995

100

150

200

R ALPH

RA L P H

RALPH

250 -2

0

2

4

6

8

31

32

Potential temperature (oC)

33

34

35 25

Salinity

25.5

26

26.5

27

27.5

Potential density (kg/m3)

Fig. 11. Vertical hydrographic profile collected at the head of Halibut Canyon at the time of RALPH deployment (black line) together with historical profiles from the ISDM data inventory collected at the outer shelf of Halibut Channel showing upwelled slope waters near the bottom. Locations of historical profiles are shown in Fig. 2 as black dots.

two tidal cycles, instead of being generated on coastal or shallow regions and advected across the shelf for long distances. This type of DSWC process could be more frequent than previously expected and occur on any continental margin where convection could reach the outer shelf. 4.3. Implications for canyon sedimentation Both type of sediment transport processes observed at the head of the canyon during the RALPH deployment contribute to resuspend sediments, preventing the deposition and accumulation of the fine sediments along the canyon axis. In Baltimore Canyon, active internal wave resuspension was observed down to 800 m depth (Gardner, 1989a, b). Based on these previous observations and on the temperature fluctuations observed at 276 m depth in Halibut Canyon, it can be expected that the internal tide resuspension mechanism could be also active further down-canyon, at least down to 500–600 m depth. The short-lived DSWC gravity flows will also have a limited downslope displacement (i.e., until reaching their equilibrium density/depth) and their implications for canyon sedimentation also appears to be restricted to the canyon head region. Halibut Canyon is located downslope of the Pleistocene Halibut Channel ice stream (Shaw et al., 2006), which is the likely source of locally derived coarse-grained glacial outwash seen in the bottom-looking images. Armitage et al. (2010) assumed that some of this coarse grained sediment on the shelf might have been reworked and redirected into the canyon by storms or tidal currents (Li et al., 2011), although direct observations presented in this paper indicate that down-canyon sediment transport during wintertime occurs mainly during DSWC events. As described in earlier sections, this transport is largely restricted to bedload and saltation, producing winnowing of sands and fine sediments around larger gravel particles (Fig. 9) that will be eventually deposited deeper down-canyon. In the Gulf of Lions, DSCW has been described as a mechanism causing the reworking, transport and accumulation of sand within canyon heads. Several interface and piston cores taken in the Bourcart Canyon showed that down to 350 m water depth, the seabed in the

canyon head is blanketed by up to 1.5 m of structureless muddy medium-grained sand with excess 210Pb activity; even though the canyon has not been connected directly to continental sources since the Last Glacial Maximum (Gaudin et al., 2006). These sand beds were interpreted as the product of repeated DSWC flows and termed “cascadites”, to be differentiated from other typical slope deposits. In the light of the observed transport processes at the head of Halibut Canyon it is likely that similar “cascadite” deposits might have been developed in its interior, as well in other submarine canyons along the eastern Canadian margin affected by DSWC. Such deposits will be confined to the canyon head region and limited in depth by the downslope displacement of the gravity flows. Further sediment coring efforts targeting some of the eastern Canadian submarine canyon heads will be required to corroborate this hypothesis. 5. Conclusions Analyses of bottom boundary layer and water column measurements collected at the head of Halibut Canyon during winter 2008– 2009 provided a comprehensive view of the contemporary sediment transport mechanism operating along this submarine canyon. The results presented in this study support the following conclusions: 1) Bottom intensified up-canyon flows ~ 40 cm s−1concurrent with sharp drops in near-bottom temperatures and increases of SSC indicate the presence of semidiurnal internal tide resuspension processes along the canyon axis. This transport process involves the resuspension of fine sediments from deeper canyon regions, which are advected towards the canyon head with the up-slope propagation of the internal tidal bore. 2) Strong down-canyon gravity flows reaching ~ 1 m s−1 occur periodically in the canyon head regions during wintertime. Such gravity flows are associated with low SSC and cold water temperatures and have been interpreted as dense shelf water cascading (DSWC) events channelized along the canyon axis, which contribute to the transport of coarse sediments by bedload and saltation towards deeper canyon regions.

P. Puig et al. / Marine Geology 341 (2013) 14–28

3) DSWC events occur during periods of cold air temperatures and sustained heat loss triggering convection and not with periods of major wave storm events. The dense shelf waters appear to be generated at the outer shelf and are evacuated into the canyon head during one or two tidal cycles, when the tidal flow is directed offshore. 4) Both oceanographic phenomena (internal tides and DSWC) are likely to occur in other submarine canyons along the eastern Canada margin and elsewhere where winter convection reaches the shelf edge. These are two key mechanisms for keeping fine sediments in suspension and for transporting coarse sediments from the outer shelf towards deeper canyon regions, with the potential to create “cascadite” deposits inside submarine canyon heads.

Acknowledgments This field experiment was funded by the Earth Science Sector Offshore Geoscience Program, and the Program of Energy Research and Development (PERD) through the Nearbed Wave and Current Forcing and Sediment Dynamics project. P. Puig also benefited from a travel grant from the Spanish Ministry of Education (Ref. PR2011-0538) to conduct this study at Bedford Institute of Oceanography. T. Milligan is specially acknowledged for facilitating access to the analyzed dataset. E. King provided consultation and background geological information for the selection of the deployment site. The authors would like to thank A. Robertson, M. Scotney and R. Boyce for their assistance in preparation, deployment and recovery of the instrumentation. This is Earth Sciences Sector contribution number 20120402. Appendix A. Supplementary data Supplementary data to this article can be found online at http:// dx.10.1016/j.margeo.2013.05.004. References Armitage, D.A., Piper, D.J.W., McGee, D.T., Morris, W.R., 2010. Turbidite deposition on the glacially influenced, canyon-dominated Southwest Grand Banks Slope, Canada. Sedimentology 57, 1387–1408. Canals, M., Puig, P., Durrieu de Madron, X., Heussner, S., Palanques, A., Fabrés, J., 2006. Flushing submarine canyons. Nature 444, 354–357. Carson, B., Baker, E.T., Hickey, B.M., Nittrouer, C.A., DeMaster, D.J., Thorbjarnarson, K.W., Snyder, G.W., 1986. Modern sediment dispersal and accumulation in Quinault submarine canyon — a summary. Marine Geology 71, 1–13. Colbourne, E., Craig, J., Fitzpatrick, C., Senciall, D., Stead, P., Bailey, W., 2009. An assessment of the physical oceanographic environment on the Newfoundland and Labrador Shelf during 2008. DFO. Canadian Science Advisory Secretariat Research Document (032 iv + 20 pp.). Dalrymple, R.W., LeGresley, E.M., Fader, G.B.J., Petrie, B.D., 1992. The western Grand Banks of Newfoundland: transgressive Holocene sedimentation under the combined influence of waves and currents. Marine Geology 105, 95–118. de Margerie, S., Lank, K.D., 1986. Tidal circulation of the Scotian Shelf and Grand Banks. Intergovernmental Panel on Energy Research and Development.Department of Fisheries and Oceans, Canada (43 pp). de Stigter, H.C., Boer, W., de Jesus Mendes, P.A., Jesus, C.C., Thomsen, L., van den Bergh, G.D., van Weering, T.C.E., 2007. Recent sediment transport and deposition in the Nazaré Canyon, Portuguese continental margin. Marine Geology 246, 144–164. deYoung, B., Sanderson, B., 1995. The circulation and hydrography of Conception Bay, Newfoundland. Atmosphere-Ocean 33, 135–162. Durrieu de Madron, X., Zervakis, V., Theocharis, A., Georgopoulos, D., 2005. Comments on “Cascades of dense water around the world ocean”. Progress in Oceanography 64, 83–90. Gardner, W.D., 1989a. Baltimore canyon as a modern conduit of sediment to the deep sea. Deep Sea Research Part A 36, 323–358. Gardner, W.D., 1989b. Periodic resuspension in Baltimore Canyon focusing of internal waves. Journal of Geophysical Research 94, 18185–18194. Gaudin, M., Berné, S., Jouanneau, J.-M., Palanques, A., Puig, P., Mulder, T., Cirac, P., Rabineau, M., Imbert, P., 2006. Massive sand beds attributed to deposition by dense water cascades in the Bourcart canyon head, Gulf of Lions (northwestern Mediterranean Sea). Marine Geology 234, 111–128. Heezen, B.C., Ewing, M., 1952. Turbidity currents and submarine slumps and the 1929 Grand Banks earthquake. American Journal of Science 250, 849–873.

27

Heffler, D.E., 1996. RALPH — a dynamic instrument for sediment dynamics. Proceedings Oceans'96, IEEE, Ft. Lauderdale, Florida, USA, September 1996, pp. 728–732. Hesse, R., Klaucke, I., Khodabakhsh, S., Piper, D., 1999. Continental slope sedimentation adjacent to an ice margin. III. The upper Labrador Slope. Marine Geology 155, 249–276. Heussner, S., Durrieu de Madron, X., Calafat, A., Canals, M., Carbonne, J., Delsaut, N., Saragoni, G., 2006. Spatial and temporal variability of downward particle fluxes on a continental slope: lessons from an 8-yr experiment in the Gulf of Lions (NW Mediterranean). Marine Geology 234, 63–92. Hotchkiss, F.S., Wunsch, C., 1982. Internal waves in Hudson Canyon with possible geological implications. Deep-Sea Research 29, 415–442. Ivanov, V.V., Shapiro, G.I., Huthnance, J.M., Aleynik, D.L., Golovin, P.N., 2004. Cascades of dense water around the world ocean. Progress in Oceanography 60, 47–98. Jenner, K.A., Piper, D.J.W., Campbell, D.C., Mosher, D.C., 2007. Lithofacies and origin of late Quaternary mass transport deposits in submarine canyons, central Scotian Slope, Canada. Sedimentology 54, 19–38. Khripounoff, A., Crassous, P., Lo Bue, N., Dennielou, B., Silva Jacinto, R., 2012. Different types of sediment gravity flows detected in the Var submarine canyon (northwestern Mediterranean Sea). Progress in Oceanography 106, 138–153. Killworth, P.D., 1983. Deep convection in the world ocean. Reviews of Geophysics and Space Physics 21, 1–26. Li, M.Z., Heffler, D.E., 2002. Continental shelf sediment transport studies in Canada: theories and recent technology advances. Geoscience Canada 29, 35–48. Li, M.Z., Prescott, R.H., Robertson, A., 2011. Hydrodynamics and seabed stability observations from 2007 Lander deployment on Green Bank, western Grand Banks. Geological Survey of Canada. http://dx.doi.org/10.4095/288035 (Open File 6744, 54 pp.). Martín, J., Palanques, A., Puig, P., 2006. Composition and variability of downward particulate matter fluxes in the Palamós submarine canyon (NW Mediterranean). Journal of Marine Systems 60, 75–97. Martín, J., Palanques, A., Puig, P., 2007. Near-bottom horizontal transfer of particulate matter in the Palamós Submarine Canyon (NW Mediterranean). Journal of Marine Research 65, 193–218. Martín, J., Palanques, A., Vitorino, J., Oliveira, A., de Stigter, H.C., 2011. Near-bottom particulate matter dynamics in the Nazaré submarine canyon under calm and stormy conditions. Deep-Sea Research Part II 58, 2388–2400. Masson, D.G., Huvenne, V.A.I., de Stigter, H.C., Arzola, R.G., LeBas, T.P., 2011. Sedimentary processes in the middle Nazaré Canyon. Deep-Sea Research Part II 58, 2369–2387. Mosher, D.C., Piper, D.J.W., Campbell, D.C., Jenner, K.A., 2004. Near-surface geology and sediment-failure geohazards of the central Scotian Slope. AAPG Bulletin 88, 703–723. Mosher, D.C., Piper, D.J.W., MacKillop, K., Jarrett, K., 2010. Near surface geology of the Halibut Channel region of the SW Newfoundland Slope from GSC data holdings. Geological Survey of Canada. (Open File 6214, 72 pp.). Mulder, T., Zaragosi, S., Garlan, T., Mavel, J., Cremer, M., Sottolichio, A., Sénéchal, N., Schmidt, S., 2012. Present deep-submarine canyons activity in the Bay of Biscay (NE Atlantic). Marine Geology 295–298, 113–127. Narayanan, S., Colbourne, E.B., Stead, P., 1996. Temperature climate atlas for the inshore regions of Newfoundland and Labrador. Canadian Technical report of Hydrography and Ocean Sciences No. (174, 181 pp.). Ogston, A.S., Drexler, T.M., Puig, P., 2008. Sediment delivery, resuspension, and transport in two contrasting canyon environments in the southwest Gulf of Lions. Continental Shelf Research 28, 2000–2013. Palanques, A., Durrieu de Madron, X., Puig, P., Fabrés, J., Guillén, J., Calafat, A., Canals, M., Heussner, S., Bonnin, J., 2006. Suspended sediment fluxes and transport processes in the Gulf of Lions submarine canyons. The role of storms and dense water cascading. Marine Geology 234, 43–61. Palanques, A., Puig, P., Durrieu de Madron, X., Sanchez-Vidal, A., Pasqual, C., Martín, J., Calafat, A., Heussner, S., Canals, M., 2012. Sediment transport to the deep canyons and open-slope of the western Gulf of Lions during the 2006 intense cascading and open-sea convection period. Progress in Oceanography 106, 1–15. Petrie, B., Anderson, C., 1983. Circulation on the Newfoundland Continental Shelf. Atmosphere-Ocean 21, 207–226. Piper, D.J.W., Normark, W.R., 1989. Late Cenozoic sea level changes and the onset of glaciation: impact on continental slope progradation off eastern Canada. Marine and Petroleum Geology 6, 336–348. Piper, D.J.W., Normark, W.R., 2009. Processes that initiate turbidity currents and their influence on turbidites: a marine geology perspective. Journal of Sedimentary Research 79, 347–362. Piper, D.J.W., Normark, W.R., Sparkes, R., 1987. Late Cenozoic stratigraphy of the central Scotian Slope, Eastern Canada. Bulletin of Canadian Petroleum Geology 35, 1–11. Piper, D.J.W., Shor, A.N., Hughes-Clarke, J.E., 1988. The 1929 “Grand Banks” earthquake, slump, and turbidity current. In: Sedimentologic Consequences of Convulsive Geologic Events. In: Clifton, H.E. (Ed.), Geological Society of America Special Paper, 229, pp. 77–92. Piper, D.J.W., MacDonald, A.W.A., Ingram, S., Williams, G.L., McCall, C., 2005. Late Cenozoic architecture of the St. Pierre Slope. Canadian Journal of Earth Sciences 42, 1987–2000. Puig, P., Ogston, A.S., Mullenbach, B.L., Nittrouer, C.A., Sternberg, R.W., 2003. Shelf-tocanyon sediment-transport processes on the Eel continental margin (northern California). Marine Geology 193, 129–149. Puig, P., Ogston, A.S., Mullenbach, B.L., Nittrouer, C.A., Parsons, J.D., Sternberg, R.W., 2004a. Storm-induced sediment gravity flows at the head of the Eel submarine canyon, northern California margin. Journal of Geophysical Research 109, C03019.

28

P. Puig et al. / Marine Geology 341 (2013) 14–28

Puig, P., Palanques, A., Guillén, J., El Khatab, M., 2004b. Role of internal waves in the generation of nepheloid layers on the northwestern Alboran slope: implications for continental margin shaping. Journal of Geophysical Research 109, C09011. Puig, P., Canals, M., Company, J.B., Martín, J., Amblas, D., Lastras, G., Palanques, A., Calafat, A.M., 2012. Ploughing the deep seafloor. Nature 489, 286–289. Puig, P., Palanques, A., Orange, D.L., Lastras, G., Canals, M., 2008. Dense shelf water cascades and sedimentary furrow formation in the Cap de Creus Canyon, northwestern Mediterranean Sea. Continental Shelf Research 28, 2017–2030. Ribó, M., Puig, P., Palanques, A., Lo Iacono, C., 2011. Dense shelf water cascades in the Cap de Creus and Palamós submarine canyons during winters 2007 and 2008. Marine Geology 284, 175–188. Sandstrom, H., Elliott, J.A., 1984. Internal tide and solitons on the Scotian Shelf: a nutrient pump at work. Journal of Geophysical Research 89, 6415–6426. Savoye, B., Cochonat, P., Piper, D.J.W., 1990. Seismic evidence for a complex slide near the Titanic wreck: model of an instability corridor for non-channelized gravity events. Marine Geology 91, 281–298. Shanmugam, G., Moiola, R.J., Damuth, J.E., 1985. Eustatic control of submarine fan development. In: Bouma, A.H., Normark, W.R., Barnes, N.E. (Eds.), Submarine Fans and Related Turbidite Systems. Springer-Verlag, New York, pp. 23–28.

Shaw, J., Piper, D.J.W., Fader, G.B.J., King, E.L., Todd, B.J., Bell, T., Batterson, M.J., Liverman, D.G.E., 2006. A conceptual model of the deglaciation of Atlantic Canada. Quaternary Science Reviews 25, 2059–2081. Trincardi, F., Foglini, F., Verdicchio, G., Asioli, A., Correggiari, A., Minisini, D., Piva, A., Remia, A., Ridente, D., Taviani, M., 2007. The impact of cascading currents on the Bari Canyon System, SW-Adriatic Margin (Central Mediterranean). Marine Geology 246, 208–230. Turchetto, M., Boldrin, A., Langone, L., Miserocchi, S., Tesi, T., Foglini, F., 2007. Particle transport in the Bari Canyon (southern Adriatic Sea). Marine Geology 246, 231–247. Verdicchio, G., Trincardi, F., Asioli, A., 2007. Mediterranean bottom-current deposits: an example from the Southwestern Adriatic Margin. In: Viana, A.R., Rebesco, M. (Eds.), Economic and Palaeoceanographic Importance of Contourite, Geological Society Special Publication 276. Geological Society London, London, pp. 199–224. Xu, J.P., Noble, M., Eittreim, S.L., Rosenfeld, L.K., Schwing, F.B., Pilskaln, C.H., 2002. Distribution and transport of suspended particulate matter in Monterey Canyon, California. Marine Geology 181, 215–234. Xu, J.P., Swarzenski, P.W., Noble, M., Li, A.-C., 2010. Event-driven sediment flux in Hueneme and Mugu submarine canyons, southern California. Marine Geology 269, 74–88.