VOL. 79, NO. 30
JOURNAL OF GEOPHYSICAL
RESEARCH
OCTOBER 20, 1974
Temporal Variability of SuspendedMatter in Astoria Canyon
WILLIAM S. PLANK, J. RONALD V. ZANEVELD, AND HASONG PAK
Schoolof Oceanography,OregonState University,Corvallis, Oregon 97331
Suspended
matter in Astoriacanyonwasmonitoredby meansof an in situnephelometer
and by means
of light-scatteringand particle concentrationmeasurementsperformed aboard ship on water samples.
Nephelometerprofilesobtainedalongthe axis of the canyonin February and April 1973indicatethat the
canyonis dividedinto two distinctzones:a nearshorezonein whichthe suspensoid
distributionundergoes
largechangesand an offshorezonein whichthe distributionvariesto a muchlesserdegree.A 15-hour
time seriesof light-scatteringand particleconcentrationprofilesat a depth of 1100m in the canyonshows
extensive
and rapid changes
in suspended
matterconcentration
at severaldepths.The effectof nonsteady
state distributionsof suspendedmatter on calculationsof the coefficientof eddy diffusivity is examined
and shown to be an important considerationin a submarinecanyon.
The extent to which observed distributions of suspended
matter vary with time is important relative to their rolesas indicators of abyssalcirculation and sedimentdispersal,and at
the present time, information on the time variability of
pended matter within the canyon. A general survey of the
bathymetry and geology of the canyon has been done by
Carlson[1967]. His studiesof sedimentationin the canyonindicatethat major erosionalactivityhasceasedin the canyonas
suspended
particulates,
especially
in the deepocean,islimited. a resultof the increasingdistanceof the canyonheadfrom the
In estuariesand on the continentalshelfa greatdeal of work shoreline.The canyon, at present,appearsto be filling up.
has been done concerningthe distribution and transport of
EXPERIMENTAL METHODS
sediments[Swiftet al., 1972].However,observationsof changSeveral different techniqueswere used in our studiesof
ing distributionsof suspendedmatter, which are an integral
part of the problemof sedimenttransport,havebeenrare. The suspendedmatter. On earlier cruises,water sampleswere
responseof suspendedsediment to tidal effects has been collected in plastic National Institute of Oceanography
studiedbySchubel
[1969,i971], Wasowski
[1974],andothers. bottles, and laboratorymeasurements
were made of light
Overthecohtinental
shelf,Rodolfo
etal. [1971]haveobserved scattering
by usinga Brice-Phoenix
light-scattering
photomthe responses
of suspendedmatter to a hurricane,and Harlett
[1972], in a study of sediment transport on the shelf off
Oregon, reported severaltime-seriesmeasurementsof light
transmissionin depths of 100-200 m.
The data presentedin this report resultedfrom our effortsto
determinehow the observeddistributionsof suspendedparticulatesrespondto the regimesof wind, surfaceand internal
waves, and tides and currentsin deep oceanicwaters. Our initial efforts in this study were to determine.theextent in time
and spaceof changesin the concentrationof suspended
matter
in the nearshore region and to extend these measurements
eter and of particle concentration and size distribution
by usinga Coultercounter.On later cruises,in situprofilesof
light scatteringweremeasuredby usinga nephelometer
constructedby the optical oceanography
group at OregonState
University.The instrument,which hasbeendescribedin detail
previously[Planket al., 1972],measures
the intensityof light
scatteredat 45ø at dept•hsup to 10,000metersand providesa
deck readout of the light-scatteringsignal along with
temperature and depth signals. A surface-actuatedrosette
samplerallowsthe collectionof water samplesat any depth.
In addition to the light-scattering and particle measurements, temperature and conductivity were determined
It was felt that Astoria canyon would be an advantageous with a Geodyne CTD.
site for this study for several reasons.It has been suggested
DATA
often [Moore, 1969;Lyall et al., 1971]that submarinecanyons
Figure I showsthe bathymetry of the Astoria canyon
are sitesof 'channelization'of sedimenttransport and perhaps
region and the location of the stationsat which measurements
of bottom water flow, and there have been numerous studiesof
bottom currents in submarine canyons [Fenner et al., 1971; were made. Thesestationsrangedin depth from 500 m, 28 km
Shepardand Marshall, 1973;Keller et al., 1973], all of which (15 n. mi.) off the mouth of the Columbia River, to 1900 m,
have shownthat bottom currentsof significantmagnitudeex- 106 km (57 n. mi.) off the coast. The shalloweststation was
ist in canyons and that flow reversals are common. One locatednearthe beginningof the canyon,whererelief between
offshore.
floorandthenearbyshelfwasabout200m. The
wouldthusexpectthat con•litions
wouldbe favorable
for thecanyon
deepeststationwas locatednear the point wherethe canyon
reachesthe Astoria fan and becomesa deepseafan valley.The
Astoria canyon is located off the mouth of the Columbia stationswere intendedto be locatedalong the axisof the canRiver, a sourceof large amounts of particulate matter [Gross yon, however,the strongsurfacecurrentsin the area and the
and Nelson, 1966; Pak et al., 1970; Conomosand Gross, 1972]. narrownessof the canyonfloor (as little as 0.6 km) made it
Also, storm activity and large wavesare not uncommon,a fact difficultto consistentlyachievethis aim. Thesedifficultiesacin depthsat stationsof
well known to oceanographersand others in this area. count for the lack of correspondence
the same number on different cruises.
All these factors seemedto indicate a high probability of
The data presentedhere were collectedon three different
measurable short-term changes in the distribution of susresuspensionand advective and diffusive transport of par-
ticulates.
cruises in 1972 and 1973. In November
Copyright¸ 1974by the AmericanGeophysicalUnion.
,
4536
1972 a 15-hour time
seriesof light scatteringand particlecontentwasobtainedin a
PLANK ET AL.: SUSPENDEDMATTER IN THE OCEAN
$0'
20'
4537
125• 00'
$0'
/
ß--
124•00'
20'
5
COLUMB
RIVER
I
6 4 :3 2:
'/
•
IO
{•
12
II
oo'
9
o•KILOMETERS
•o 2o
I
ß
I
ß
40'
40'
:50'
125e00'
::50'
124e00 '
Fig. 1. Bathymetryand station locationsin the vicinity of Astoria canyonafter Carlson[1967].
terestingin severalrespects.We can seethat, especiallyfor the
stationsnearer shore,there is a great deal of variation between
the profilestaken in February and thosetaken in April. Some
of the more pronounceddifferencesincludethe light-scattering
maximums at 200-350 m at stations 5, 6, and 7 in February
closest to the bottom. The date and time each station was octhat do not appear in April and the intense maximum at
cupiedare shownbelow the profilesso that the degreeof non- 100-200 m at stations1, 3, 4, and 5 in April, which is presentto
synopticityof the data can be evaluated.The profilesrepresent someextent at stations2 and 3 in February but is not nearly as
the relative magnitude of the logarithm of the value of the well defined then. At the majority of the stationsthe intensity
volume-scatteringfunction fi at 45ø from the forward direc- of light scatteringincreasesas we approach the bottom, but
tion. Although the instrumentcalibrationdid not changefrom the magnitudeof the increase,as well as the depthrangeover
station to station, it was not calibrated in absolute values of
which the increasetakes place, varies a great deal. The most
fi(45), so that we can really only comparethe shapesof the striking differenceis at station 7 where in February there is acurves.
350-m-thick layer of high-intensityscatteringadjacent to the
Figures 3 and 4 present the results of the time-series bottom but in April the layer is only about 50 m thick and the
measurementof light scatteringand particle content. These vertical gradient of light scatteringis extremely high. These
measurements were made between stations 7 and 8 as close to
observed differencesbetween the February and April disthe axis of the canyonas possiblein approximately1100m of tributions of light scatteringare indicationsof the nonsteady
water. During the time seriesof 6 nephelometerlowerings, state nature of the regimesof advectiveand diffusiveenergy
which lasted from 0730 and 2230 on November 20, 1972, the
which, along with particle settling and the strengthof the
ship drifted 3.9 km (2.1 n. mi.) in a westerlydirection,which various particle sources, determines the distribution of
drift accountsfor the increasingdepth during the time on sta- suspendedmatter.
It seemslikely that the pronouncedmaximumsat depths
tion. Figure 3 showsthe profilesof fi(45) as determinedby in
vitro measurementswith the Brice-Phoenix light-scattering between 100 and 350 m, as noted above, are related to erosion
photometeron water samplescollectedby hydrographiccasts and near-bottomtransporton the shelf,which in this area exand also correspondingprofiles of the total number of par- tendsoffshoreabout 56 km (30 n. mi.) and lies at a depth of
ticles per milliliter greater than 2.2 •tm as determined by about 180 m near the shelf break. The light-scatteringmaximumsat depthsequal to or lessthan the surroundingshelf
Coulter counter measurementson the same samples.
In Figure 4 the valuesof particlecount havebeencontoured depth may representsedimenterodednearbyor further into presenta temporal crosssectionof suspendedparticulate shoreat shallowerdepthsand then transportedinto the axisof
variation. The cross section is also to some extent spatial the canyonby offshoreor longshore
currents.The maximums
at depthsgreaterthan the surroundingshelfdepthmay repbecauseof the drift of the ship during the experiment.
resent material similarly eroded and transported but of
depthof 1100m on the canyonaxis,and in February and April
1973 light-scattering profiles were obtained at a series of
stationsalong the canyon axis. Figure 2 showsthis seriesof
light-scatteringprofiles. The relative geographicposition of
eachprofileis indicatedby the point at whichthe profilecomes
,
DISCUSSION
ß
The two sets of nephelometerprofiles in Figure 2 are in-
sufficient
size to have settled somewhat
trapped in the canyon.
and thus become
4538
PLANK ET AL..' SUSPENDEDMATTER IN THE OCEAN
o
o
•
0
o
-'
I
o
o
E
oo
x
o
0
i
I
o.
'•H
o
'Hld30
,•.
o
o.
-
ßtull ' H.Ld':i0
u"!.
-
o.
PLANK ET AL.' SUSPENDED MATTER IN THE OCEAN
LIGHT SCATTERING,/• (45), (m-ster)-I x I0 -s
o
o
5
I0
15
!
!
!
0
5
I0
!
i
15
0
4539
PARTICLE COUNT, (ml)-' x 102
5
I0
15
5
I0
15
i
!
!
0
'
5
I0
[
!
15
0
5
I0
15
!
i
i
0.5
05
,)
E
~ LiJ 1.0
0730
PST
I0 bj
--
1030
1525
1650
1925
2230
--
LIGHT
----
SCATTERING
PARTICLE
CRUISE
COUNT
Y7211C
NOV 15- 20,
1.5
I
1972
I
Fig.3. Profiles
ofthevolume-scattering
function/5(45)
in(msr)
-• X 10-3andthenumber
ot'particles
permilliliter
larger
than
2.2tzm
atthe
time-seriesstation in Astoria canyon.
The variationsin the shapeof the light-scattering
profileat
station7 are undoubtedlya resultof thechangingnatureof the
fieldsof diffusiveand advectiveenergywithinthe canyon.The
near-bottomthick layerof high-intensity
light scatteringpresent in the Februaryprofilemay be a resultof high-energy
vertical diffusionin the lower part of the water column.On the
other hand, the thin layerwith alarge verticalgradientat this
same location may indicate bottom currents of sufficient
TIME
o
ooo
I
OF
,oo
velocity to erode sedimentsbut vertical diffusion that is insufficientto transport the material upward.
Along with the February to April variationsin the profiles
at stationsI through 8, another strikingthing we noticeabout
thesedata is the lack of temporal variation in overall shape
betweenthe profilestaken at stations9-12. Although there are
small-scale differences, light scattering appears to have
changed very little between February and April and, in
OBSERVATION
1600
•_
.............................................
_...!,
...........
,.............................................
i ..2000
_
0.2
0.4
, '.-.....:.:.:.:.:.-.'.-
ß•
0.6
.'.'.'.'.'.'.'.'.'.'.-.'.'......
a.
0.8
...........ß•.ßß•'ßßß••ßß•
ßß :::::::::::::::::::::
"'.•.ßo.
====================================
....
ß.'.'.':'.'.':':':':':':':-:-:-:-:-:.:-:.:.:-:.:-:.:.:.:-:.'.'.'"
:
...'.,.
ß
..'',:',::,::.:.:.:.:.:.:.:.:.:.:.:.
:::::::::::::::::::::::::::::
::::::::::::::::::::::::::::::
1.0
_
1.2
Fig. 4. Contoursof particlecountdata from Figure3 duringtime-series
measurement
in Astoriacanyon.Arrowsindicatetimesof
nephelometerlowerings.
PLANK
ETAL.:SUSPENDE[•
MATTER
INTHEOCEAN
454O
general, it seemsthat the incr6asein light scatteringnear the
bottom
seems less marked
than in the stations closer to shore.
As a result,it appearsreasonableto divide the canyoninto two
distinct regionsthat reflect the observedlight-scatteringdistribution and the probable regimesof advectionand mixing.
Zone I (between•tations I and 9). This is a zone of high
temporalvariability in the processes
of advectionand diffusion
leading to large spatial and temporal gradientsin suspended
matter. Bottom erosion and near-bottom turbulent transport
are likely. Flow is probably channelizedto a large extent.
Zone 2 (beyondstation 9). This is a 'quieter' zone where
propertiesare more stable.Erosion of bottom sedimentsis less
likely in this zone, althoughthere is still somevariability in the
suspensoidsresulting from nonsteady state advection and
horizontal and vertical mixing.
By examining the bathymetry of the canyon we can find a
possibleexplanationfor this division. Shorewardof station 9,
bathymetric profiles acrossthe canyon show a deep(up to 500
m) canyon with steep sides. Water motions here would be
restricted laterally and the likelihood of strong horizontal
currents and vertical mixing greater. Beyond station 9 the
shapeof the canyonchanges.The heightof the canyonwallsis
less(<300 m) and the slope of the walls more graqlual.The
effectis to allow more lateral spreading,whichmay reducethe
velocityof axial currentsand decreasethe intensityof vertical
gradients within the canyon. The characteristics of the
suspended sediment distribution in zone I are also undoubtedly related to the nearnessof sedimentsources(Columbia River, surf zone) and the greater intensityof surfacewave
effects.
We can get someidea of the magnitude and rapidity of the
changesin the distribution of suspendedmatter by examining
the resultsof the time-seriesmeasurementsof light scattering
and particle count in Figures 3 and 4. During a time period of
15 hours the measurementsof both of these properties increasedby a factor of about 1.9 near the bottom, and it can be
seenthat a turbid layer about 300 m in thicknessadjacentto
the bottom has appeared in this short time. Also an intermediatemaximum at 800-900 m depth has completelydisappeared,and a maximum at 300-400 m depth hasdiminished
greatly in intensity.
Previousinvestigators
have indicatedrelationshipsbetween
suspended matter and canyon bottom currents or
meteorologicalconditions [Rodolfoet al., 1971; Shepardand
Marshall, 1973].Shipboardrecordsof wind and seastateshow
that there was a fairly rapid rise in wind speed(from 5 to 10
m/s in about 18 hours)and in waveheight(from about 1.3 to
2.7 m) about 60 hours prior to the beginningof the time
series.These strong winds blew for about 36 hours and then
decreased to less than 2.5 m/s. At the time the measurements
about 1.8 cm/s. It seemsunlikely thereforethat our observed
rise in particle concentration near the bottom is the direct
result of the meteorological
conditionsobservedduringthe
cruise.It is possible,of course,that we couldbe observing
the
resultsof somepreviousstorm,however,considering
the lack
of spatial coherencein the canyon bottom currentsobserved
by Shepardand Marshall [1973] and the time variabilitywe
have observedin suspensoid
distributions,it is probablyunrealisticto think of 'patches'of sedimenttravelingdown the
lengthof a canyon,and we shouldrather think of a continuous
processof erosion,mixing, advection,redeposition,etc. What
weseethenat a depthof 1000m (35 n. mi.) offthecoastmay
be poorly correlatedwith recentmeteorologicalevents.
Measurementsof temperatureand salinitymadecoincident
with measurements
of light scatteringand particlecountgave
no indication of the changestaking place below the surface
layers.
Severalauthorshaveusedobservedprofilesof light scattering or particleconcentrationto computecoefficients
of vertical
eddydiffusivity[Ichiye,1966;EittreimandEwing,1972;Ichiye
et al., 1972],and it hasbeennecessary
becauseof a lack of data
on temporalvariabilityto assumesteadystatein thesecases.
Given the time serieswe have observedwe can attemptto
determine the effect of the time-dependentnature of the
observed distributions. Since in this case we lack data on the
horizontaldistributionof suspended
matter and the velocity
field,we•hallassume
thatonlyverticalprocesses
areat work.
If we disregard the effectsof horizontal advection and diffu-
sion,we will obtainan upperlimit on the magnitudeof vertical
eddy diffusivity.We may then useas the diffusionequation
0-• A,•z- -- •zz[wP]- Ot
where P is the concentrationof a property. In this casewe use
light-scatteringvaluesthat are assumedto be proportionalto
the concentration of suspendedmatter. There is some experimental justification for this assumption [Beardsleyet al.,
1970].In this case,w will refer only to the settlingvelocityof
the particles;we assumevertical advection, which is included
in this term to be negligible.Also, sincew varieswith particle
size and density,we will choosea singlesettlingvelocityfor a
particle of averagesizeand representativedensity.The effect
of the useof the completeparticlesizedistributionin solutions
to the diffusion equation has been consideredby Zaneveld
[1972],and it seemslikelythat the useof a singleaverageparticle size is adequatefor our purposeshere.
In order to determine the effect of the time-dependent
nature of the particledistributionwe first assumesteadystate.
The diffusion equation then becomes
were taken the wind was againin the processof increasingto
W/Az = 1/P OP/Oz
above 10 m/s. It is possiblethat the increasein suspended
matter adjacentto the bottom is the resultof sedimenterosion In ordertocalculate
Azweassume
a settling
velocity
of 1.8X
in shallowwater and subsequent
transportationby advection 10-3 cm/s, which is givenby Gibbset al. [1971] as the settling
to the site of the measurements. If we assume that the increase
velocity for a particle of diameter 5 tzm and specificgravity
in light scatteringnear the bottom is dueto sedimenterodedin 2.65. We obtain valuesfor •9P/Oz and P by assumingthat A,
the surf zone by the rapid increasein wind and waveheight and w are constant over a layer 50 m thick. We choosethe
and then carrieddownthe canyonby a bottom current,we can profile taken at 1525PST as being representativein shape,and
estimatethe speedof this currentat about 25 cm/s (56 km in we determineP at 25 m abovethe bottom. We then useAp/Az
60 hours).We mightcomparethis to the net transportfigures for this layer as •9P/Oz. Note that the unitsof P cancel.Using
for bottom currentsin canyonsoff southernCalifornia and valuesthusdeterminedfrom the profile at 1525PST, we obtain
Baja Californiareportedby ShepardandMarshall [1973]that A• = 17 cmø'/s.
rangedfrom 0 to 123m/h with an averageof only 67 m/h or
If we do not assumesteadystatebut we do assumethat A,
PLANK ET AL.' SUSPENDEDMATTER IN THE OCEAN
and w are constant, the diffusion equation becomes
O•P
OP
A• Off w Oz
OP
Ot
Here •9P/•9t is determinedby plotting valuesof •(45) at 25
m off the bottom
versus t for the entire time series and measur-
4541
Gross, G. M., and J. L. Nelson, Sediment movement on the continental shelf near Washington and Oregon, Science,154, 879-885, 1966.
Harlett, J. C., Sedimenttransporton the northernOregoncontinental
shelf, Ph.D. thesis,Oregon State Uni¾., Corvallis, 1972.
lchiye, T., Turbulent diffusionof suspendedparticlesnear the ocean
bottom, Deep Sea Res., 13, 679-685, 1966.
lchiye, T., N.J. Bassin,and J. E. Harris, Diffusivity of suspended
matter in the CaribbeanSea,J. Geophys.Res., 77(33), 6576-6588,
ing the slopeof the tangentto this curve at 1525 PST. Then
1972.
•9P/3z and •92P/•9z
2 are againmeasuredfrom the 1525profile Keller, G. H., D. Lambert, G. Rowe, and N. Staresinic, Bottom
currentsin the Hudson Canyon, Science,180(4082), 181-183, 1973.
for the layer 50 m abovethe bottom,and w is againassumedto
be 1.8 X 10-8 cm/s. In the nonsteadystatecasethen, Az = 500 Lyall, A. K., D. J. Stanley,H.,N. Giles,and A. Fisher,Jr., Suspended
sediment and transport at the shelf-break and on the slope, Mar.
Technol.Soc. J., 5(1), 15-27, 1971.
one order of magnitude would have been incurred by an Moore, D. G., Reflectionprofiling studiesof the California continental borderland: Structure and quarternary turbidity basins,Geol.
assumptionof steady state.
Soc. Amer. Special Pap. 107, 142, 1969.
Our evidenceof the time-dependentnature of the distribution of suspendedsediment indicates that assumptionsof Pak, H., G. F. Beardsley,Jr., and P. K. Park, The Columbia River as a
source of marine light-scattering particles, J. Geophys. Res.,
steadystatefor the purposeof inferringdeep-oceancirculation
75(24), 4570-4577, 1970.
are unrealisticin the type of environmentwe haveinvestigated. Plank, W. S., H. Pak, and J. R. V. Zaneveld, Light scatteringand
suspendedmatter in nepheloid layers, J. Geophys.Res., 77(9),
Our work indicatesa need for similar studiesin other deep-
cmUs, and we can see that in this case an error of more than
ocean en•iironments that have often been assumed to be in
steady state.
Acknowledgment.This investigationhas been supportedby contract N 00014-67-A-0369-0007underprojectNR 083-102 with Oregon
State University.
1689-1694, 1972.
Rodolfo, K. S., B. A. Buss,and O. H. Pilkey, Suspendedsedimentincreasedue to hurricane Gerda in continental shelf waters off Cape
Lookout, N. C., J. Sediment.Petrology,4•(4), 1121-1125, !971.
Schubel, J. R., Size distributions of the suspendedparticles of the
ChesapeakeBay turbidity maximum, Neth. J. Sea Res., 4(3),
283-309, 1969.
REFERENCES
Schubel,J. R., Tidal variations of the size distribution of suspended
sedimentat a station in the ChesapeakeBay turbidity maximum,
Beardsley,G. F., H. Pak, K. Carder, and B. Lundgren,Light scatterNeth. J. Sea Res., 5(2), 252-266, 1971.
ing and suspended
particlesin the easternequatorialPacificOcean, Shepard, F. P., and N. F. Marshall, Currents along floors of subJ. Geophys.Res., 75, 2837, 1970.
marinecanyons,
Amer.Ass.Petrol.Geol.Bull.,57(2),244-264,
Carlson, P. R., Marine geologyof Astoria submarinecanyon, Ph.D.
1973.
thesis,259 pp., Oregon State Univ., Corvallis, 1967.
Swift, D. J.P., D. B. Duane, and O. H. Pilkey (Eds.), Shelf Sediment
Conomos,J. T., and M. G. Gross, River-oceansuspendedparticulate
Transport.'Processand Pattern, 652 pp., Dowden, Hutchinson and
matter relations in summer, in The ColumbiaRiver Estuary and AdRoss, Stroudsburg, Pa., 1972.
jacent OceanWaters,editedby A. T. Pruterand D. L. Alverson,pp. Wasowski, S. F., Measurements of turbulent velocities and an ex176-202, University of WashingtonPress,Seattle, 1972.
amination of their effectson mixing and suspensionof particulate
Eittreim, S., and M. Ewing, Suspended
particulatematter in the deep
matter, M.S. thesis, 91 pp., Oregon State Univ., Corvallis, 1974.
waters of the North American basin, in Studies in Physical Zaneveld, J. R. V., Optical and hydrographicobservationsof the
Oceanography,
vol. 2, editedby ArnoldL. GordOn,pp. 123-168,
Gordon and Breach, New York, 1972.
Fenner, P., G. Kelling, and D. J. Stanley, Bottom currents in
Wilmington Canyon, Nature, 229(2), 52-54, 1971.
Gibbs, R. J., M.D. Matthews, and D. A. Link, The relationship
betweenspheresize and settlingvelocity,J. Sediment.Petrology,
41(1), 7-18, 1971.
Cromwell Current between92000' west and the GalapagosIslands,
Ph.D. thesis,87 pp., OregonState Univl, Corvallis, 1972.
(Received April 3, 1974;
acceptedJune 19, 1974.)