Suspended sediment fluxes and transport processes in the ... et al (MG2006).pdf · Suspended sediment fluxes and transport processes in the Gulf of Lions submarine canyons. The role

(2006) 43–61www.elsevier.com/locate/margeo

Marine Geology 234

Suspended sediment f luxes and transport processes in the Gulf ofLions submarine canyons. The role of storms and

dense water cascading

Albert Palanques a,⁎, Xavier Durrieu de Madron b, Pere Puig a, Joan Fabres c,Jorge Guillén a, Antoni Calafat c, Miquel Canals c, Serge Heussner b, Jerôme Bonnin b

a Institut de Ciències del Mar (CSIC), Passeig Marítim de la Barceloneta, 37-49. E-08003 Barcelona, Spainb CEFREM, CNRS, UMR 5110, University of Perpignan, 52, Avenue Villeneuve, 66860 Perpignan cedex, France

c GRC Geociències Marines, Dept. d'Estratigrafia, Paleontologia i Geociències Marines, Universitat de Barcelona, E-08028, Barcelona, Spain

Accepted 5 September 2006

Abstract

Contemporary suspended sediment transport was studied in seven submarine canyons of the Gulf of Lions (GoL). Current metersequipped with turbidity sensors were moored 4 m above bottom at 300 m depth in the canyon axis from November 2003 to May 2004.Sediment transport events were monitored and studied in relation to forcing conditions. There was a large flood in early December, duringwhich discharges from all of the coastal rivers increased bymore than one order of magnitude. A smaller flood of the Rhone River occurredlater in mid-January followed by a persistent high river discharge that lasted until mid-February. There were also several E–SE storm eventsduring the measurement period, two of them causing large swell, one in early December (max Hs: 8.4 m), coinciding with the major riverflood, and one in late February (maxHs: 7m) during a period of lower river discharge.Most of the variability in down-canyon current speedswas linked to strong downwelling induced byE–SE storms and to cascading of dense shelfwater induced byNandNWwinds. The intensityand timing of these processes strongly varied spatially. Eastern storms generated higher waves and induced stronger downwelling in thewestern than in the eastern sector of the GoL. Shelf dense water cascading events were enhanced during eastern storms and they were alsomore intense in the western sector of the GoL. These events occurred frequently from January to May along the western canyons and fromFebruary to April along the eastern canyons. From February to April they occurred simultaneously in all the canyons.

Sediment transport was mainly down-canyon and mostly concentrated during storm-induced downwelling and dense watercascading events. During the high river discharge season most of the newly-supplied and resuspended sediment remained stored onthe shelf. However, during the late February storm and cascading event, the stored sediment was quickly resuspended and transported,mainly through the westernmost submarine canyon (Cap de Creus), following a flushing pattern. Later, although minor storms andshelf dense water cascading hardly increased suspended sediment concentrations within the canyons, they induced strong near-bottomcurrent velocities, increasing down-canyon sediment fluxes and generating small but significant sediment transport events. Most ofthe shelf-canyon transfer took place during the late February storm and cascading event through the Cap de Creus canyon where thenet suspended sediment flux was between one and two orders of magnitude higher than in the other canyons.© 2006 Elsevier B.V. All rights reserved.

Keywords: suspended sediment transport; submarine canyon; Gulf of Lions; storm events; cascading events

⁎ Corresponding author. Tel.: +34 932309500; fax: +34 932309555.E-mail address: [email protected] (A. Palanques).

0025-3227/$ - see front matter © 2006 Elsevier B.V. All rights reserved.doi:10.1016/j.margeo.2006.09.002

mailto:[email protected]

http://dx.doi.org/10.1016/j.margeo.2006.09.002

44 A. Palanques et al. / Marine Geology 234 (2006) 43–61

1. Introduction

About 90% of the sediment generated by erosion onland is deposited on the continental margins. Oncesediment particles are discharged by rivers into themarine system, they are dispersed under the combinedaction of currents and waves. Forcing conditions aregoverned by local hydrodynamics, meteorology, climateand bottom morphology. Subsequently particles may beaffected by several transport, deposition, and resuspen-sion cycles and are distributed over the continental shelfand slope before being more definitively deposited.Because of limitations on present observational techni-ques for studying the episodic and three-dimensionalcharacter of particulate transfer, the transport mechan-isms, sediment fluxes and ultimate fate of this material arepoorly constrained. Several projects, such as AMASEDS,SEEP, ECOMARGE, OMEX, MATER, EUROMARGEand STRATAFORM, (Nittrouer and DeMaster, 1986;Biscaye et al., 1988; Monaco et al., 1990a; Biscaye andAnderson, 1994; van Weering et al., 1998; Nittrouer,1999; Wollast et al., 2001), have studied continent-oceantransfer during the last few decades, focusing mainly oncontinental margins that in some cases included subma-rine canyons.

One of the questions we still do not know well is how-and how much-sediment entering the ocean is transferredoffshore through submarine canyons. Submarine canyonsare physiographic structures that favor seaward sedimenttransport, funnelling sediment as preferential conduitsconnecting the shelf to the slope and open sea (Drake andGorsline, 1973; Hickey et al., 1986; Gardner, 1989a;Durrieu de Madron, 1994; Puig and Palanques, 1998). Awidespread concept nowadays is that particulate mattertransferred from the shelf through submarine canyonscomes directly from continental runoff and/or resuspension(e.g., during storms) and coastal biological productivity(Thunnell, 1998; Boyd and Newton, 1999). During theeighties, several canyon studies defined important dynam-ics and sedimentary patterns in them (Hickey et al., 1986;Gardner, 1989a,b), but they did not elucidate sufficientlythe processes controlling shelf-canyon transfer. Sedimenttrap experiments conducted in NWMediterranean canyonsrevealed that vertical particle fluxes reached maximumvalues after storms or following increases in the outflowfrom a nearby river, (Monaco et al., 1990b; Puig andPalanques, 1998; Monaco et al., 1999). However, therestricted resolution of sediment trap sampling, the limitedinformation about forcing conditions and the lack of directsimultaneous measurements on the shelf and at the head ofthe canyons prevented such transfers from being corrob-orated, quantified and explained in detail.

The link between river discharge and/orwindstorms andparticle transfer has been identified in several other settingsas is the case for canyons located in areas of low waveenergy, narrow shelves and connected to rivers, such as theGuadiaro Canyon in the Alborán Sea which can be do-minated by floods (Palanques et al., 2005a). Submarinecanyons that are deeply incised into the continental shelf arevery sensitive to storms such as theMonterrey Canyon (Xuet al., 2002), the Palamós Canyon (Palanques et al., 2005b;Martín et al., 2006), Kao-Ping Canyon (Liu et al., 2002),Hudson Valley (Harris et al., 2003) and Nazare Canyon.Other canyons connected to rivers and exposed to veryenergetic conditions are also dominated by storms, as is thecase of the Eel Canyon, where shelf and canyon transferprocesses were studied in detail (Puig et al., 2003).

Within the framework of the EUROSTRATAFORMproject, one key objective was to investigate transfer fluxesand processes in submarine canyons. The Gulf of Lions(GoL) was chosen to study a river-dominated continentalmargin. The objectives of this paper are: 1) quantify theshelf–slope suspended sediment transport through GoLsubmarine canyons during winter and spring (the flood andstorm season), 2) study the temporal and spatial variabilityof suspended sediment fluxes, 3) identify themajor transferevents and 4) characterize the processes that controlsuspended sediment transport through submarine canyonsin this area.

2. Regional setting

The Gulf of Lions is a micro-tidal continental andriver-dominated margin that extends from Cap Croisettein the northeast to Cap de Creus in the southwest(Fig. 1). This margin is fed by several rivers flowingfrom different hydrographic basins. The most importantis the Rhone River, which comes from the Alps andflows into the eastern part of the GoL. Its suppliesrepresent about 90% of the total freshwater inputs of theGoL. The Vidourle, Herault and Orb rivers come fromthe Massif Central and discharge into the central part ofthe GoL, whereas the Agly, Tet and Tech rivers flowfrom the Pyrenees and discharge into the western part ofthe GoL (Fig. 1). The major river discharge periods arein spring and autumn.

The different wind regimes determine the circulationand the transport of suspended sediment on the shelf.Northerly (Mistral) and northwesterly (Tramontane)winds, and S– SE wind (Marin) are the predominantwind regimes in the GoL. The Mistral and Tramontaneare more frequent in winter and spring and inducedistinctive and opposite circulation cells on the shelffavoring the intrusion of slope waters in the eastern and

Fig. 1. Map of the Gulf of Lions (north-western Mediterranean) showing 1) the rivers discharging into it, 2) the shelf mud belt, 3) general bathymetry andGoL submarine Canyons and 4) the position of the moorings installed at the head of seven submarine canyons. CC: Cap de Creus submarine canyon; LD:Lacaze-Duthiers submarine canyon;AU:Aude submarine canyon; HE:Herault submarine canyon; PR: Petit Rhone submarine canyon; GR:Grand Rhonesubmarine canyon, PL: Planier submarine canyon. Cre: Cap de Creus. Cro Cap Croissette.

45A. Palanques et al. / Marine Geology 234 (2006) 43–61

central parts of the shelf and export of shelf water at itssouthwestern end (Estournel et al., 2003; Petrenko et al.,2005). Furthermore, these cold, dry winds are respon-sible for the strong cooling and homogenization of theshelf water column in winter, which facilitate densewater formation. Due to the short fetch, they generateonly small waves (significant wave height b2 m, peakperiod b6 s) on the inner shelf. Conversely, southeasternand eastern Marin wind events, although rare and brief,are associated with large swell (significant wave heightup to 10 m, peak period up to 12 s) and a significant risein sea level along the coast. These episodes are often

simultaneous with river floods as the transport of humidmarine air over the coastal relief promotes heavyprecipitation.

As a result of the large river inputs and the circulation inthe GoL, the distribution of the suspended particulatematter concentration shows a strong seaward decreasinggradient during the entire year (Durrieu de Madron et al.,1990; Durrieu de Madron and Panouse, 1996; Heussneret al., 2006-this issue). The distinct surface turbid layer,which extends over the shelf and slope, is essentiallycomposed of biological material. Conversely, the bottomturbid layer, which extends to the shelf-edge, is mostly

Fig. 2. Mean daily water discharge of the Gulf of Lions rivers from November 2003 to May 2004 grouped by watersheds.


lithogenic. Dispersal of riverine sediment, together withresuspension of fine sediment by waves and its subsequenttransport by the shelf circulation leads to the formation of amud belt along the inner andmid-shelf (Aloïsi et al., 1976).

Along the GoL slope, the Northern Current flows as apart of the cyclonic circulation of the western Mediter-ranean basin (Millot, 1999). It forms a density front thatseparates the low-salinity shelf water from themore salineopen sea water influencing the shelf–slope exchanges(Durrieu deMadron et al., 1999). Lapouyade and Durrieude Madron (2001) further estimated that the horizontaltransport of suspended particulate matter advected by thepermanent along-slope current increases significantlydownstream between the entrance and the exit of theGulf as a result of the shelf export. The particulate matterbudgets indicated a greater export from the shelf in winterthan in summer due to enhanced shelf–slope exchangeprocesses, in particular the cascading of cold dense waterfrom the shelf.

In the GoL slope, the main morphological character-istic is the numerous submarine canyons incised into it(Fig. 1). Downward particulate fluxes have been

intensively studied in some of these canyons. Particlesfluxes were measured by near-bottom (30mab) sedimenttraps at depths from 500 to 1000 m in the Grand-Rhoneand Planier Canyons in the northeastern part of the Gulfand in the Lacaze-Duthiers Canyon in its southwesternpart (Monaco et al., 1990b; Monaco et al., 1999;Heussner et al., 2006-this issue). These studies showedthat: 1) downward particle fluxes varied seasonally, withmaximum fluxes in winter; 2) the summer/winteralternation appeared to correlate with that of majorsources of particulate matter (river discharge) andphysical forcing (winter convection, along-slope currentvariability); 3) fluxes inside the canyons were larger thanthose on the open slope, suggesting a preferentialtransport of material through the canyons; and 4) fluxesincreased significantly westward (with annual meanfluxes in the Lacaze-Duthiers Canyon being about 3times greater than in the Grand-Rhône and PlanierCanyons), suggesting a preferential export through thesouthwestern part of the Gulf.

However, up to now temporal series of suspendedsediment fluxes in the GoL submarine canyons had

Fig. 3. Significant wave heights (Hs) recorded in different coastal areas of the Gulf of Lions from November 2003 to May 2004.


never been studied, and even a simultaneous record ofsuspended sediment fluxes in several submarine canyonshad not been carried out in this or in any continentalmargin before.

3. Material and methods

Contemporary suspended sediment transport at sevensubmarine canyon heads in the Gulf of Lions was studiedby deploying moorings at 300 m depth in sevensubmarine canyons: the Cap de Creus, Lacaze-Duthiers,Aude, Herault, Petit Rhone, Grand Rhone and Planiersubmarine canyons (Fig. 1). On each of these moorings,one Aanderaa RCM9/11 current meter equipped with aturbidimeter (OBS) was installed 4 m above bottom fromNovember 2003 to May 2004. This period was dividedinto two consecutive deployments. The sampling intervalof the current meters was set to 20 min. Temperature andconductivity sensors were calibrated using contemporaryCTD measurements. Turbidity data recorded in FTU wereconverted into suspended sediment concentrations follow-ing the methods described in Guillén et al. (2000).

Suspended sediment fluxes in the studied canyonheads were calculated for each mooring site. Instanta-neous sediment fluxes were obtained by multiplying thecurrent speed by the SSC. The instantaneous fluxes werealso calculated for the along-canyon and across-canyoncurrent components. Averaged over time, these give thenet across- and along-canyon suspended-sedimentfluxes. From the resultant vector of those flux compo-nents, the estimated magnitude of the net horizontal fluxof suspended sediment and the flux direction at each sitewere obtained.

CTD and light transmission sections were performedalong canyon axes inNovember 2003, February 2004 andMarch 2004. In this paper, we selected the Cap de Creuscanyon transects recorded along the axis from the inner-shelf to 500 m depth. The CTD was a Sea Bird CTD 25with a Sea Tech 25 cm path length transmissometer and abottom-triggered mechanism for collecting a water sam-ple very near the bottom (about 50 cmab).

River water discharges were measured by gaugingstations close to the river mouths and supplied by the“Compagnie Nationale du Rhone” and the “Banque

Fig. 4. Significant wave heights (Hs) recorded in different sectors of the GoL and river discharges of the Rhone River and selected rivers of the centraland westerns drainage basins from November 2003 to May 2004. Major forcing events (waves higher than 7 m or Rhone river discharge higher than5000 m3 s−1) are shown as dark shaded areas. Minor forcing events (waves between 2.5–4 m height or Rhone River discharge between 3000 m3 s−1

and 5000 m3 s−1) are shown as light shaded areas. These curves are presented in some of the following figures to show forcing conditions.


HYDRO” of the French Ministry of Environment. Waveheight was measured with an ADCP next to the Tet Rivermouth (western coastal sector) and bywave buoys offshoreofSete(centralcoastalsector)andCamargue(easterncoastalsector) (data fromCETMEF).

4. Results

4.1. River discharge and wave climate

Fig. 2 shows the water discharge of the rivers flowinginto the GoL during the deployment period, grouped bydrainage basins (Pyrenees in the West, Massif Central inthe centre and Alps in the East). It is important to noticethat the Rhone River discharge was at least one order ofmagnitude higher than that from the other rivers.However, the cumulative discharge of all the other riversduring general flood events, although lower than that fromthe Rhone River alone, can be significant. The mostimportant flood occurred on December 4th, 2003 andaffected all the rivers of the GoL. During this flood, the

maximum Rhone river (eastern drainage basin) dailydischarge reached 9346 m3 s−1 (a flood with a 75 yearrecurrence interval), the rivers of the central drainagebasin discharged between 580 and 660 m3 s−1 and therivers from the western drainage basin between 280 and550m3 s−1. The total discharge of all the small GoL riversduring this event was about one third of that from theRhone. In mid-January, there was another importantflood in the Rhone River, during which dischargeincreased up to 3800 m3 s−1. This river supplied morethan 2000 m3 s−1 daily between January 13th andFebruary 6th, maintaining more than 3000 m3 s−1 for10 days. In late February a smaller flood affected most ofthe rivers and the water discharge reached 2900 m3 s−1 inthe Rhone River, between 210 and 320 m3 s−1 in the Orb,Herault and Aude Rivers, but only some tens of m3 s−1 inthe other rivers. Other floods had a more local effect,occurring only on rivers of different drainage basins. Inmid-January, late January, mid-April and early May therewere river discharge increases (up to 450 m3 s−1) in thewestern drainage basin rivers, and in mid-March, early

Fig. 5. Cross-margin sections of the shelf and Cap de Creus Canyon axis between the inner shelf and 500 m depth showing the distribution of watertemperature, salinity and beam attenuation (Cp) in November 2003 (A), February 2004 (B) and March 2004 (C). Density distribution is superimposedin contour lines.


April and early May there were increases in the centraldrainage basin rivers.

Suspended sediment discharges for the Rhone Riverwere estimated using the log–log relationship betweentotal suspended matter and water discharge establishedby Sempéré et al. (2000). The solid discharge during theexperimental period (November 2003–May 2004)reached about 7 Mt, but most of it was input duringfloods. Considering a threshold of 3000 m3 s−1 forflood, the December 2003 flood supplied about 5 Mtand the January 2004 flood supplied about 0.5 Mt.

The larger waves were generated by easterly stormsand the highest waves were recorded in the westernsector. Fig. 3 shows this trend in westward increase inwave height during these storms. There were two majorstorms, during which significant wave heights reached8.4 m (December 2003) and 7 m (February 2004) in thewestern sector which recurrence interval is about 50 and10 years respectively (Puertos del Estado). During thesetwo storm events, waves reached significant heights ofabout 5.5 m and 3.5 m in the central and in the eastern

sector of the GoL, respectively. Other minor storms tookplace in mid-November, early and mid-December, earlyFebruary, mid-March, early and mid-April and earlyMay. The maximum significant wave height of most ofthese minor storms was between 2.5 and 4 m andoccurred in the central sector (Fig. 3).

Thus, the two major events during the deploymentperiod were: 1) a very brief (9 h) and strong eastern stormassociated with a major flood in all rivers in the GoL, inearly December 2003 and 2) a longer (3 days) and strongeastern storm event in late February 2004 with moderatefloods in most rivers (Fig. 4).

The minor events corresponded to 1) a relativelyimportant flood in the Rhone River at mid-Januaryfollowed by a period of continued high Rhone dischargeduring 24 days, but without significant wave activity andto 2) small storms events that generated the highestwaves in the central GoL sector and caused waterdischarge increases in the rivers from the centraldrainage basin and in the Rhone River. They occurredinmid-November, early andmid-December, mid-March,

Fig. 6. Temporal series of near-bottom temperature at the heads of the monitored submarine canyons from East (above) to West (below) fromNovember 2003 to May 2004. Cascading events in grey bands. Thick dashed lines represent major E–SE storms and narrow dashed lines representmoderate storms that generated temperature fluctuations. River discharge of selected rivers from the western, central and eastern GoL drainage basins.R. d.: river discharge.


early April and early May. 3) An event in mid-April thatgenerated wave increases in the central and western GoLsector and water discharge increases in the rivers fromthe western drainage basin.

4.2. Hydrography

The seasonal variability of the hydrographic struc-tures during the deployment period was greater in thesouthwesternmost canyons. In the Cap de Creus Canyonthe following structures were measured:

In November, prior to the first major storm, the watercolumn was strongly stratified with warm water(TN15 °C) in the upper layer and less saline water(Sb38.1) on the inner shelf (Fig. 5A). Suspendedparticulate concentration was higher in the core of freshercoastal water and around the shelf break area between 100and 300 m depth, where an intermediate nepheloid layerdetachment took place.

In February, before the second major storm, the watercolumn was weakly stratified due to the strong cooling

and mixing of surface water by the sustained cold anddry northerly winds (Fig. 5B). Shelf water became colder(Tb12.5 °C) and still less saline (Sb38.1) than theoffshore and slope waters. A 20–30m thick bottom layerof rather dense and turbid water extended over the outershelf and detached at the shelfbreak. The density of thistongue was close to that of the waters in the uppercanyon.

InMarch, after a minor storm event (Hs about 3 m), thecooling of the surface water reached its maximum(Tb12 °C). Shelf water was well mixed and denser thanthe off-shore surface and intermediate waters (Fig. 5C). Aturbid tongue of shelf dense water was traced in the upperpart of the canyon reflecting an on-going cascading event,as the leading edge of the tongue had not reached itsneutral density level, between 400 and 500 m.

4.3. Near-bottom temperature time series at canyon heads

Fig. 6 shows temperature time series recorded by theinstruments moored at the canyon heads during the study


period. Near-bottom temperature maintained relativelyconstant values (around 13.4 °C) at most canyon heads inNovember and December 2003, with some brief butstrong positive anomalies (temperature increases of up to2.1 °C) in the western and central canyons, coincidingwith the early December storm and flood event (Fig. 6). Inthe Cap de Creus Canyon, temperature during this eventincreased up to about 15.5 °C, which was the temperatureof the inner and mid-shelf water in that season (Fig. 5A).Between the last week of January and the third week ofFebruary repetitive negative anomalies were recordedstarting in the westernmost canyons (Cap de Creus andLacaze-Duthiers) and progressing towards the centralones (Aude andHerault). In late February, coincidingwiththe major storm event, and inMarch negative temperatureanomalieswere recorded in all canyons. The strongest andmost frequent and prolonged negative anomalies occurredin the westernmost submarine canyons (Cap de Creus andLacaze-Duthier) during February and March (with dropsof 1.6 °C). In Cap de Creus canyon temperature decreasedto 11.8 °C and 11.6 °C in late February and mid-March

Fig. 7. Temporal series of near-bottom current speed at the heads of the mNovember 2003 to May 2004. Cascading events in grey bands. Thick dashedmoderate storms that generated current fluctuations. River discharge of selecR. d.: river discharge.

respectively, which were about the inner and mid shelfwater temperatures during each of those events (Fig. 5Band C). The intensity of the temperature drops decreasedin most central canyons (Petit Rhone and Herault) andincreased slightly in the eastern canyons (Grand Rhoneand Planier). Between April and May these anomaliesceased to occur progressively westwards and inMay theyonly occurred in the western submarine canyons (Fig. 6).Finally it is important to notice that major temperatureanomalies were associated with eastern storms, especiallythose during early December and late February.

4.4. Near-bottom current time series at canyon heads

At the monitored canyon heads, near-bottom currenttime series showed high temporal and spatial variability(Fig. 7). During the deployment period, large andrepetitive speed fluctuations often occurred simulta-neously with the temperature anomalies described in theprevious section. Maximum current speeds peaked up to60–80 cm s−1 in the westernmost submarine canyons

onitored submarine canyons from East (above) to West (below) fromlines represent major E–SE storms and narrow dashed lines representted rivers from the western, central and eastern GoL drainage basins.


(Cap de Creus and Lacaze-Duthier) and 30–50 cm s−1 inthe submarine canyons of the central and eastern slope ofthe GoL (Fig. 7). Many of the most important events ofincreased current speed occurred during eastern storms,and they often were maintained several days after thesestorms. Currents were mainly oriented along-canyon,with fluctuating up- and down-canyon motions, but thenet transports were down-canyon (Fig. 8). During thestorms and the increased current speed events, thedirection was always down-canyon.

Current variability between November and lateJanuarywas characterised by high frequency fluctuationswith peaks of current speed lasting for only a few hoursas for example the increase associated with the majorDecember event that lasted only for about 9 h and wasrecorded only in the western and central slope canyonsalong with the positive temperature anomaly. There weresignificant current increases in the Petit Rhone andGrand Rhone Canyons during December and Januarythat were not associated with temperature changes.

Conversely, from the end of January onwards theperiods of increased current speeds lasted up to severaldays, and the longest and most intense were those thatstarted during easterly storms and were maintained afterthose storms abated. Between the end of January and May,the spatial and temporal distribution of current speedincreases generally matched well with the strongertemperature fluctuations. Very strong current speed

Fig. 8. Polar diagrams of near-bottom current speed at the heads of the monlines indicate the direction of the canyon axis at the mooring sites.

increases occurred in all the submarine canyon headsduring the late February,mid-March and earlyApril events.

4.5. Near-bottom suspended sediment concentrationtime series at canyon heads

Time series of suspended sediment concentration (SSC)recorded 4 mab show several peaks at the canyon headsduring the deployment period (Fig. 9). The higher valuestook place in the Cap de Creus Canyon, where near-bottomSSC reached 48 mg l−1 during the early December eventand more than 68 mg l−1 during the late February event.During a 10-h period of this latter storm, turbidity wasabove the sensor range, pegging its signal and therefore, themaximum value of 68 mg l−1 should be considered as aminimum value during the sensor saturation period. In theother canyons, SSC peaks were lower and the absolutemaximum values did not always occur during the Februaryevent. In the Lacaze-Duthier, Herault and Petit RhoneCanyons, SSC during the early December event (peaks of9.4, 6.0, 8.3mg l−1 respectively) increasedmore than twicethan during the February event (peaks of 3.7, 2.9 and3.9 mg l−1 respectively). In the Aude Canyon, the maxi-mum SSC value was recorded in February (5.5 mg l−1)and no increase was recorded during the Decemberevent. There were other SSC increases affecting thewestern and central canyons during the mid-April event.The other minor storms events occurring in November,

itored submarine canyons from November 2003 to May 2004. Dashed

Fig. 9. Temporal series of near-bottom turbidity (SSC in mg l−1) at the heads of the monitored submarine canyons from East (above) to West (below)from November 2003 to May 2004. Cascading events in grey bands. Thick dashed lines represent major E–SE storms and narrow dashed linesrepresent moderate storms that generated temperature and currents fluctuations. River discharge of selected rivers from the western, central andeastern GoL drainage basins. R. d.: river discharge.


mid-December, mid-March and early April producedpractically no significant SSC increase, excluding a fewsmall, local peaks. The eastern canyons, Planier and theGrand Rhone, did not show any significant SSC increaseduring the study period.

4.6. Near-bottom suspended sediment fluxes

The temporal evolution of instantaneous suspendedsediment fluxes (SSF) varied within a very wide range of

more than 4 orders of magnitude, between 1 mg m−2 s−1

and more than 52 g m−2 s−1 (Fig. 10). The highestinstantaneous SSF were recorded during the two majoreastern storm events in the Cap the Creus Canyon, 31.12 gm−2 s−1 in early December and 52.07 g m−2 s−1 in lateFebruary, which is a minimum value because the turbiditysensor was pegged for 10 h of this event and the flux couldbe higher. Lower SSF increases, reachingmaximum valuesbetween 1 and 5 g m−2 s−1, were also recorded in theLacaze-Duthier, Herault and Petit Rhone Canyons during

Fig. 10. Temporal series of near-bottom instantaneous suspended sediment fluxes at the heads of the monitored submarine canyons from East (above)to West (below) from November 2003 to May 2004. Cascading events in grey bands. Thick dashed lines represent major E–SE storms and narrowdashed lines represent moderate storms that generated temperature and currents fluctuations. River discharge of selected rivers from the western,central and eastern GoL drainage basins. R. d.: river discharge.


the December and February events. The February eventalso caused a significant increase in the Aude Canyon(0.98 g m−2 s−1) and a small increase in the Grand Rhoneand Planier Canyons (0.43 and 0.21 g m− 2 s−1

respectively). During the mid-April event, instantaneousSS fluxes increased significantly in the Cap de CreusCanyon (7.87 g m−2 d−1) and the Herault Canyon (2.13 gm−2 s−1) and slightly in the Lacaze-Duthier, Aude, PetitRhone and Grand Rhone Canyons (between 0.30 and1.10 g m−2 s−1). Small SSF increases occurred during themid-March and early April events in all the submarinecanyons, including the easternmost ones.

The cumulative along-canyon and across-canyonsuspended sediment transport indicates a predominantdown-canyon sediment flux, showing strong increasesduring the eastern storm events and maintainingsignificant transport during increased current speedevents (Fig. 11). During all the sediment transportevents, the highest shelf-canyon transport was through

the Cap de Creus Canyon and the second largest throughthe Lacaze-Duthier Canyon, whereas the lowest trans-port occurred in the eastern canyons. Although the lateFebruary event was the most important in the study areaand in the western canyons, it affected the central andeastern canyons less, where the most important sedimenttransport events were the March or mid-April ones.

During the December event the highest cumulativetransport was about 350 kg m−2 through the Cap de CreusCanyon. During the late February event, the totalcumulative sediment transport was at least one order ofmagnitude higher than during the other events, and themaximum cumulative sediment transport took place in theCap de Creus Canyon (2932 kg m−2). In this canyon, thecumulative sediment transport during the March and mid-April events was of the order of 150–350 kgm−2 (Fig. 11).

The minimum net suspended sediment fluxes duringthe deployment period were measured in the Planier andAude Canyons. In the Herault, Grand Rhone and Petit

Fig. 11. Time-Integrated down-canyon cumulative suspended sediment transport at the heads of the monitored submarine canyons from November 2003to May 2004. Sediment transport is predominantly down-canyon. Cascading events that generated significant sediment transport in grey bands. Thickdashed lines represent major E–SE storms and narrow dashed lines represent moderate storms that generated significant sediment transport. Riverdischarge of selected rivers from the western, central and eastern GoL drainage basins. R. d.: river discharge.


Fig. 12. Net suspended sediment fluxes at the head of the seven studied canyons during the deployment period.


Rhone Canyons, net fluxes were about twice theminimum ones. Net flux in the Lacaze-Duthier Canyonswas 3 to 6 times higher than in the central and easterncanyons and net flux in the Cap de Creus Canyon was 1to 2 orders of magnitude higher than in all the othercanyons (Fig. 12).

5. Discussion

5.1. Controlling factors

In the GoL, strong sediment transport events occurduring the E–SE storms. Strong E and SE wind eventsare relatively infrequent and brief (less than 3 days) butthese episodes are associated with the larger waves on

the shelf, due to their long fetch, and occasionally withheavy rainfall and flash floods, due to the advection ofhumid marine air towards the continent (Ferré et al.,2005). During these events, river plumes and sedimentresuspended by waves are primarily advected along theshelf and exported at the southwestern end of the GoL,as is shown by the MODIS image (Fig. 13) and sketchedin Fig. 14A for the December 2003 event and in Fig.14B for the late February 2004 event. During E–SEstorm events, the cyclonic circulation on the shelfinduces a massive convergence of shelf water at thesouthwestern end of the gulf, which in turn causes ashort but intense downwelling of shelf water on theupper slope (Ulses et al., submitted for publication)along with down-canyon sediment transport.

Fig. 13. MODIS image taken during the December flood and storm event showing river plumes dispersion and resuspension along the shelf and theacross-margin export around the Cap de Creus.


For stratified autumn conditions, this energeticmechanism produces a sudden transport of lighter(warmer and less saline) and more turbid shelf waterdownslope mainly through the western canyon heads.However, buoyancy hinders the vertical displacementand limits the intrusion of shelf water into the canyonheads (few hundred meters depth). An intrusion wastraced during the early December event by the increasesin temperature, current speed, and SSC at 300 m depthmainly within the western canyons and also in the centralcanyons (Figs. 6–8). Maximum cumulative sedimenttransport during this event was about 350 kg m−2

through the Cap de Creus Canyon (Figs. 11 and 14A). Nointrusion was observed in the easternmost canyons.

During winter well-mixed conditions, the densitycontrast between shelf and slope waters is small, evennil and the downwelled shelf water can flow andtransport sediment deeper into the canyons. This was thecase for the late February storm, when the density of theshelf water matched that of the slope water at 300 mdepth. Eastern and southeastern storm events combineseveral elements that favor the export of river and shelfsediment to the slope. They simultaneously promote aresuspension of shelf sediment, a cyclonic transport ofthe shelf water and river plume, and downwellingmainly in the southwestern side of the GoL. Thissituation and the cumulative sediment transport duringthe February 2004 event is shown in Fig. 14B.

Other significant sediment transport events are relatedto the dominant strong Mistral (N) and Tramontane (NW)

winds. Inwinter, these sustained cold and drywinds inducea large cooling and mixing of the shallow inner-shelfwaters and form dense cold waters. The newly formedwater sinks, spreads on themid- and outer shelf as a bottomlayer, and ultimately cascades down the slope as is shownin Fig. 5C. This situation and the cumulative sedimenttransport range during the March–April 2004 events isshown in Fig. 14C. Dufau-Julliand et al. (2004) and Ulseset al. (submitted for publication) showed that dense waterthat accumulates on the shelf is then mostly transported tothe southwest by the cyclonic meso-scale circulation andwinds, and finally escapes the shelf preferentially throughcanyons. However, part of the dense water can be trappedon the concave, or even flat, parts of the outer shelf, andremain there until it is flushed out by another process. Thetemperature decreases and down-canyon current speedincreases recorded in winter were associated with thesetypes of cascading events (Figs. 6 and 7). These events didnot occur simultaneously in all the canyons, and they haddifferent durations, from a few hours to more than 1 week.The duration of the cascading season, defined as the periodof time between the first and the last cascading event,showed a regional eastward decreasing trend (Figs. 6 and7). The cascading season was longer (about 4 months),began earlier (January) and ended later (May) in thewestern canyons and was progressively shorter, beginninglater and ending earlier towards the east. In the easternsubmarine canyons it lasted only about 2 months,beginning in February and ending in April. The strongestcascading intensity occurred in the western submarine

Fig. 14. Scheme of typical circulation and processes during autumn and winter in the GoL and cumulative sediment transport during the main eventsof the study period. A: E–SE winds with stratified fall conditions and cumulative sediment transport during the December 2003 event; B: E–SE windswith unstratified winter conditions and cumulative sediment transport during the February 2004 event; C: Tramontane and Mistral winds (NWand N)with unstratified winter conditions and cumulative sediment transport range during the March–April events.


canyons, located in the sector where the shelf is narrowest,whereas the lowest intensity corresponded to the centralslope canyons, especially the Petit Rhone and the Herault,located where the shelf is widest. The different sourceareas, shelf morphology and circulation may explain thedifferences in each sector.

During the deployment period, the temperaturedecrease events started or were intensified during theeastern storms and they continued after the storms(Fig. 6). The downwelling induced by the eastern stormsprobably triggered and/or enhanced the cascading byflushing the dense water trapped on the shelf.Whereas thestorm events lasted only from a few hours to 2–3 days, thecascading persisted after the storm and could last morethan 1 week. The intensification of the cascading wasmore intense in the westernmost canyons, where theeastern storms produced the largest downwelling.

The asymmetry of the export from the shelf iscontrolled by the dominant along-shelf circulationpattern and also by the shelf morphology, which allows

the escape at the western end where it becomes verynarrow (less than 1 km) and constricted by the Cap deCreus promontory. In addition, higher waves during theE–SE storm events induce stronger sediment resuspen-sion in the western sector and the downwelling andcascading induced by storms and N and NW winds arealso more intense along the western slope. All of thisfavors the dominant shelf sediment export through theCap de Creus zone.

5.2. Sediment transport events and general evolution

The general sediment transport evolution during thedeployment period is a consequence of the different forcingevents and is shown by the along-canyon cumulativetransport (Fig. 11). The early December event wasproduced by a strong eastern storm accompanied by avery significant flood and high river sediment discharges(5 Mt in the Rhone River). This event generated intensesediment resuspension and downwelling of warmer shelf


water, with sharp increases in near-bottom down-canyonsediment fluxes in the western submarine canyons and lesssharp increases in the central canyons (Figs. 11 and 14A).However, this strong eastern storm lasted only about 9 h,causing a sediment transfer through the Cap de CreusCanyon of about 350 kg m−2. The short duration of thestorm probably explained the fact that a very significantpart of the sediment load discharged during the flood andresuspended by the waves remained on the shelf. This isconcordant with the observations made in the Tet Riverprodelta, where there was first sediment erosion during themajor early December storm, and later several centimetresof sediment accumulated at that site after the flood, when aminor storm redistributed the flood material on the shelf(Guillen et al., 2006-this issue). However, the small stormsoccurring after the early December event were not strongenough to cause significant sediment escape offshore, andthe shelf continued to receive additional river sedimentinputs. The more important ones were those discharged bythe Rhone between mid-January and early February whenthe Rhone maintained a high water discharge (reachingabout 4000 m3 s−1 in mid-January), supplying about0.5 Mt of sediment. In early January, late January and mid-February, shelf cold water cascading events generatedstrong near-bottom down-canyon currents and smallturbidity increases in the western submarine canyons.However, some significant sediment export was onlyobserved during the late January cascading event in theCapde Creus Canyon (about 100 kg m−2).

The late February sediment transport event wasproduced by a strong eastern storm that induceddownwelling and amplified cascading by flushing outthe dense water stored on the flat outer shelf. Bothprocesses produced strong near-bottom down-canyoncurrent increases in all the studied submarine canyons,including the easternmost ones, and major turbidityincreases in canyons of the western and central GoLslope (Figs. 11 and 14B). When the late February eventoccurred, there was more “fresh” sediment available onthe shelf to be resuspended and exported than in theearly December event. In addition, the late Februarystorm was longer (three days) than the early Decemberstorm and flushed out most of the material recentlydeposited on the shelf. As a consequence, most of thesuspended sediment transfer during the deploymentperiod took place during the February event through theCap de Creus Canyon, where about 3000 kg m−2 weretransferred down-canyon (Figs. 11 and 14C). After theFebruary event, less resuspendable sediment remainedon the shelf and the following moderate and minorstorms could not resuspend and export significantamounts of sediment. The mid-March, early April and

mid-April sediment transport events were produced bysmall eastern storms that generated downwellingstarting again shelf dense water cascading, whichcontinued some days after the storms. These densewater cascading events and moderate storms hardlyincreased the SSC, but they induced persistent andstrong current velocities, increasing down-canyonsediment fluxes and generating a small but significantsediment transport in all the studied submarine canyons(Figs. 11 and 14C). In the Cap de Creus Canyon, thesediment export during the mid-March, early April andmid-April events was about 200, 150 and 360 kg m−2

respectively, which indicates that during cascadingcombined with moderate storm events, the sedimenttransport was of the same order of magnitude as theexport during the December major storm and floodevent (Figs. 11,14A and C).

The recorded data indicate that the GoL canyons haverecurrent sediment transport events during shelf watercascading events and sporadic strong sediment transportevents during strong eastern storms. Storm-induceddownwelling can be combined with cascading, thusenhancing the sediment transport. When these processesare very intense and occur at the end of or after the floodseason, they can flush the sediment stored on the shelf,mainly though the Cap de Creus Canyon, at thesouthwestern end of the Gulf. Thus, the western andcentral GoL submarine canyons show a sporadic sedimentflush transport pattern, fed by river floods and controlledby strong storms combined with cascading, along with arecurrent and more regular pattern during the cascadingseason. In the easternmost submarine canyons, sedimenttransport events are smaller and mainly associated withshelf water cascading.

The recurrence period of the strongest eastern stormsrecorded in early December 2003 and late February2004 (tens of years) indicates that the dominant role ofthese strong storms takes place only from time to time,during the years when they are produced. On the otherhand, the shelf water cascading occurs each year and itsrole is more recurrent. However, the intensity andactivity of the shelf water cascading can change a lot asobserved by Bethoux et al. (2002). In addition, theimportant Rhone river floods have also several years ofrecurrence, as the 75 years recurrence period of theDecember 2003 flood. Therefore, the interannualvariability in the GoL is very high.

According to the recorded data, the western and centralGoL submarine canyons are storm-dominated systems,but in addition, considering the importance of dense watercascading, they are “storm and cascading-dominatedsystems” and in periodswithout strong storms theywill be

http://dx.doi.org/10.1016/j.margeo.2006.09.018


probably only cascading-dominated systems. The easternGoL submarine canyons are mainly “cascading-dominat-ed systems”. River floods by themselves do not generatestrong sediment transport through canyons, but theygenerate temporal deposits on the shelf that are winnowedand transferred offshore during sporadic strong easternstorms and seasonal cascading events. Similar behaviourhas been observed in the Eel Canyon, where sediment isstored temporarily in shelf depocentres until a strongstorm reaches the area (Puig et al., 2003). Other canyonsprobably also follow this pattern, although each one withits own particularities.

6. Conclusions

The GoL continental shelf stores sediment suppliedby rivers, which can be subsequently resuspended by E–SE storms and exported to the deep sea by wind-driventransport processes, namely storm-induced downwel-ling and dense water cascading, mainly throughsubmarine canyons. Down-canyon sediment transportis enhanced when these processes interact in winter andespecially during the strongest and sporadic E–SEstorms (Hs: 7–8 m). Along the GoL slope the down-canyon suspended sediment transport during E–SEstorms and dense water cascading is notably asymmet-ric, mainly due to stronger wind-induced processes inthe western sector, westward along-shelf circulation,and shelf and slope morphology. Storm-induced down-welling and cascading are more intense in the westernGoL submarine canyons and the cascading season islonger in the western than in the eastern submarinecanyons. Shelf–slope suspended sediment transfer ismore intense towards the western part of the Gulf,particularly through the Cap de Creus submarine canyonthat is the final outlet before the constriction of the Capde Creus promontory. As a consequence of all this,suspended sediment fluxes in the westernmost canyonswere several times to two orders of magnitude largerthan those in the easternmost ones.

During the deployment period, the December 2003flood supplied large amounts of sediment onto the shelfand the associated strong (Hs: 8.4 m) but short easternstorm only exported a small part of it. The Rhone Rivercontinued supplying large amounts of sediment inJanuary and February 2004. The maximum shelf-canyonexport took place during the late February strong easternstorm (Hs: 7 m) associated with shelf water cascadingthat occurred at the end of the flood season, when thefresh and more easily erodible sediment stored on theinner and mid-shelf was flushed away, mainly throughthe Cap de Creus Canyon. In addition, minor storms

(Hs: 2–4 m) and shelf water cascading, increased down-canyon suspended sediment fluxes, mainly by increasingthe down-canyon current speed for several days.

The western and central GoL submarine canyons aresystems dominated by the escape of the sediment storedon the shelf during energetic eastern storms and alsoduring recurrent shelf water cascading events in winter.The suspended sediment transport through the easterncanyons, which receive much lower amounts ofsediment, is hardly affected by eastern storms and ismainly dominated by shelf water cascading. This studycontributes to a better understanding of the shelf-canyontransfer in the GoL and gives a more complete regionalpattern, defining as a new finding the important role ofthe Cap de Creus Canyon.

Acknowledgements

This study was supported by the EUROSTRATA-FORM Project funded by the EU (EVK3-CT-2002-00079, EU Fifth Framework Programme: Energy,Environment and Sustainable Development). We thankthe officers and crew of the R/V “Thetis II” for their helpand dedication during the three cruises. We also thankFrançois Bourrin for the processing of theMODIS image.

References

Aloïsi, J.C., Cauwet, G., Gadel, F., Got, H., Monaco, A., Vile, F., Causse,C., Pagmon, M., 1976. Contribution à l'étude de la sedimentationrécente et de la pollution sur le plateau continental du Golfe du Lionentre Fos et Sète. Bull. Bur. Rech. Geol. Min. 4, 69–83.

Bethoux, P., Durrieu de Madron, X., Nyffeler, F., Tailliez, D., 2002.Deep water in the western Mediterranean: peculiar 1999 and 2000characteristics, shelf formation hypothesis, variability since 1970and geochemical inferences. J. Mar. Syst. 33–34, 117–131.

Biscaye, P.E., Anderson, R.F., 1994. Fluxes of particulate matter on theslope of the southern Middle Atlantic Bight: SEEP-II. Deep-SeaRes. 41, 459–509.

Biscaye, P.E., Anderson, R.F., Deck, B.L., 1988. Fluxes of particlesand constituents to the eastern United State continental slope andrise: SEEP-I. Cont. Shelf Res. 8, 855–904.

Boyd, P.W., Newton, P.P., 1999. Does planktonic community structuredetermine downward particulate organic carbon flux in differentoceanic provinces? Deep-Sea Res. I 46, 63–91.

Drake, D.E., Gorsline, D.S., 1973. Distribution and transport ofsuspended particulate matter in Hueneme, Redondo, Newportand la Jolla submarine canyons. Geol. Soc. Amer. Bull. 84,3949–3968.

Dufau-Julliand, C., Marsaleix, P., Petrenko, A., Dekeyser, I., 2004. 3DModelling of the Gulf of Lion's hydrodynamics (NWMed.) duringJanuary 1999 (MOOGLI3 Experiment) and late winter 1999: WIWformation and cascading over the shelf break. J. Geophys. Res.109, C11002. doi:10.1029/2003JC002019.

Durrieu de Madron, X., 1994. Hydrography and nepheloid structuresin the Grand Rhone canyon. Cont. Shelf Res. 14, 457–477.

http://dx.doi.org/10.1029/2003JC002019


Durrieu de Madron, X., Panouse, M., 1996. Transport de matièreen suspension sur le plateau continental du Golfe du Lion-Situation estivale et hivernale. C. R. Acad. Sci., Ser. IIa, Paris 322,1061–1070.

Durrieu de Madron, X., Nyffeler, F., Godet, C.H., 1990. Hydrographicstructure and nepheloid spatial distribution in the Gulf of Lionscontinental margin. Cont. Shelf Res. 10, 915–929.

Durrieu deMadron,X., Radakovitch,O., Heussner, S., Loye-Pilot,M.D.,Monaco, A., 1999. Role of the climatological and current variabilityon shelf–slope exchanges of particulate matter. Evidence from theRhône continental margin (NWMediterranean). Deep-Sea Res. I 46,1513–1538.

Estournel, C., Durrieu de Madron, X., Marsaleix, P., Auclair, F.,Julliand, C., Vehil, R., 2003. Observations and modelisation of thewinter coastal oceanic circulation in the Gulf of Lions under windconditions influenced by the continental orography (FETCHexperiment). J. Geophys. Res. 108 (C3), 8059.

Ferré, B., Guizien, K., Durrieu de Madron, X., Palanques, A., Guillén, J.,Grémare, A., 2005. Fine-grained sediment dynamics during a strongstorm event in the inner-shelf of theGulf of Lion (NWMediterranean).Cont. Shelf Res. 25 (19–20), 2410–2427.

Gardner, W.D., 1989a. Baltimore canyon as a modern conduit ofsediment to the deep sea. Deep-Sea Res. I 36, 323–358.

Gardner, W.D., 1989b. Periodic resuspension in Baltimore canyon byfocusing of internal waves. J. Geophys. Res. 94, 18185–18194.

Guillén, J., Palanques, A., Puig, P., Durrieu deMadron, X., Nyffeler, F.,2000. Field calibration of optical sensors for measuring suspendedsediment concentration in the Western Mediterranean. Sci. Mar. 64(4), 427–435.

Guillen, J., Bourrin, F., Palanques, A., Duurrieu de Madron, X., Puig, P.,Buscail, R., 2006-this issue. Sediment dynamics during ‘wet’ and‘dry’ storm events on the Tet inner shelf (SW Gulf of Lions). Mar.Geol. 234, 129–142. doi:10.1016/j.margeo.2006.09.018.

Harris, C.K., Butman, B., Traykovski, P., 2003.Winter-time circulationand sediment transport in the Hudson Shelf Valley. Cont. Shelf Res.23, 801–820.

Heussner, S., Durrieu de Madron, X., Calafat, A., Canals, M.,Carbonne, J., Delsaut, N., Saragoni, G., 2006-this issue. Spatialand temporal variability of downward particle fluxes on acontinental slope: lessons from an 8-yr experiment in the Gulfof Lions (NW Mediterranean). Mar. Geol. 234, 63–92.doi:10.1016/j.margeo.2006.09.003.

Hickey, B.M., Baker, E.T., Kachel, N., 1986. Suspended particlemovement in and around Quinault submarine canyon. Mar. Geol.71, 35–83.

Lapouyade, A., Durrieu de Madron, X., 2001. Seasonal variability ofthe advective transport of particulate matter and organic carbon inthe Gulf of Lion (NW Mediterranean). Oceanol. Acta 24 (3),295–312.

Liu, J.T., Liu, J.K., Huang, J.S., 2002. The influence of a submarinecanyon on river sediment dispersal and inner shelf sedimentmovements: a perspective from grain size distribution. Mar. Geol.181 (4), 357–386.

Martín, J., Palanques, A., Puig, P., 2006. Composition and Variabilityof Downward Particulate Matter Fluxes in the Palamós SubmarineCanyon (NW Mediterranean). J. Mar. Res. 60, 75–97.

Millot, C., 1999. Circulation in the western Mediterranean Sea. J. Mar.Syst. 20, 423–442.

Monaco, A., Biscaye, P.E., Soyer, J., Pocklington, R., Heussner, S.,1990a. Particle fluxes and ecosystem response on a continental

margin: 1985–1988 Mediterranean ECOMARGE experiment.Cont. Shelf Res. 10, 809–839.

Monaco, A., Courp, T., Heussner, S., Carbonne, J., Fowler, S.W.,Deniaux, B., 1990b. Seasonality and composition of particulatefluxes during ECOMARGE_I, western Gulf of Lion. Cont. ShelfRes. 10, 959–988.

Monaco, A., Durrieu de Madron, X., Radakovitch, O., Heussner, S.,Carbonne, J., 1999. Origin and variability of downward biogeo-chemical fluxes on the Rhone continental margin (NW Mediter-ranean). Deep-Sea Res. 46, 1483–1511.

Nittrouer, C.A., 1999. STRATAFORM: overview of its design andsynthesis of its results. Mar. Geol. 154, 3–12.

Nittrouer, C.A., DeMaster, D., 1986. Sedimentary processes on theAmazon continental shelf: past, present and future research. Cont.Shelf Res. 6, 5–30.

Palanques, A., El Khatab, M., Puig, P., Masqué, P., Sánchez-Cabeza, J.A.,Isla, E., 2005a. Downward particle fluxes in the Guadiaro submarinecanyon depositional system (north-western Alboran Sea), a river flooddominated system. Mar. Geol. 220 (1–4), 23–40.

Palanques, A., Garcia-Ladona, E., Gomis, D., Martín, J., Marcos, M.,Pascual, A., Puig, P., Emelianov,M., Guillén, J., Gili, J.M., Tintoré, J.,Jordi, A., Basterretxeab, G., Font, J., Segura, M., Blasco, D.,Monserrat, S., Ruiz, S., Pagès, F., 2005b. General patterns ofcirculation, sediment fluxes and ecology of the Palamós (La Fonera)submarine canyon, north-western Mediterranean. Prog. Oceanogr. 66,89–119.

Petrenko, A., Leredde, Y., Marsaleix, P., 2005. Circulation in astratified and wind-forced Gulf of Lions, NW Mediterranean Sea:in situ and modelling data. Cont. Shelf Res. 25, 7–27.

Puertos del Estado. Extremos máximos de oleaje (Altura significante)Boya de Palamós. www.puertos.es.

Puig, P., Palanques, A., 1998. Temporal variability and composition ofsettling particle fluxes on the Barcelona continental margin. J. Mar.Res. 56, 639–654.

Puig, P., Ogston,A.S.,Mullenbach,B.L., Nittrouer, C.A., Sternberg,R.W.,2003. Shelf-to-canyon sediment-transport processes on the Eelcontinental margin (northern California). Mar. Geol. 193, 129–149.

Sempéré, R., Carriere, B., Cauwet, G., van Wambekek, F., et al., 2000.Carbon inputs of the Rhone River to the Mediterranean Sea :biogeochemical implications. Glob. Biogeochem. Cycles 14 (2),669–681.

Thunnell, R.C., 1998. Seasonal and annual variability in particle fluxesin the Gulf of California: a response to climate forcing. Deep-SeaRes. 45, 2059–2083.

Ulses, C., Estournel, C., Bonnin, J., Durrieu deMadron, X.,Marsaleix, P.,submitted for publication. Impact of storms and dense watercascading on shelf-slope exchanges in the Gulf of Lion (NWMediterranean). J. Geophys. Res.

van Weering, T.C.E., McCave, I.N., Hall, I.R., 1998. Ocean MarginExchange (OMEX I) benthic processes study – Preface. Prog.Oceanogr. 42, 1–4.

Wollast, R., Chou, L., Avril, B., Huthnance, J., 2001. Ocean marginexchange in the northern Gulf of Biscay: OMEX I. Deep-Sea Res.II 48, 14–15.

Xu, J.P., Noble, M., Eittreim, S.L., Rosenfeld, L.K., Schwing, F.B.,Pilskaln, C.H., 2002. Distribution and transport of suspendedparticulate matter in Monterey Canyon, California. Mar. Geol. 181,215–234.

http://www.puertos.es

Documents

Suspended sediment fluxes and transport processes in the ... et al (MG2006).pdf · Suspended sediment fluxes and transport processes in the Gulf of Lions submarine canyons. The role