tidal and meteorological forcing of sediment transport in ... · pdf filetidal and...

ARTICLE IN PRESS

0278-4343/$ - se

doi:10.1016/j.csr

�CorrespondiWoods Hole Oc

02543, USA. Tel

E-mail addre

Continental Shelf Research 27 (2007) 1510–1527

www.elsevier.com/locate/csr

Tidal and meteorological forcing of sediment transport intributary mudflat channels

David K. Ralston�, Mark T. Stacey

Department of Civil and Environmental Engineering, University of California, Berkeley, CA, USA

Received 1 February 2006; received in revised form 16 August 2006; accepted 9 January 2007

Available online 30 January 2007

Abstract

Field observations of flow and sediment transport in a tributary channel through intertidal mudflats indicate that

suspended sediment was closely linked to advection and dispersion of a tidal salinity front. During calm weather when tidal

forcing was dominant, high concentrations of suspended sediment advected up the mudflat channel in the narrow region

between salty water from San Francisco Bay and much fresher runoff from the small local watershed. Salinity and

suspended sediment dispersed at similar rates through each tidal inundation, such that during receding ebbs the sediment

pulse had spread spatially and maximum concentrations had decreased. Net sediment transport was moderately onshore

during the calm weather, as asymmetries in stratification due to tidal straining of the salinity front enhanced deposition,

particularly during weaker neap tidal forcing. Sediment transport by tidal forcing was periodically altered by winter

storms. During storms, strong winds from the south generated wind waves and temporarily increased suspended sediment

concentrations. Increased discharge down the tributary channels due to precipitation had more lasting impact on sediment

transport, supplying both buoyancy and fine sediment to the system. Net sediment transport depended on the balance

between calm weather tidal forcing and perturbations by episodic storms. Net transport in the tributary channel was

generally offshore during storms and during calm weather spring tides, and onshore during calm weather neap tides.

r 2007 Elsevier Ltd. All rights reserved.

Keywords: Sediment transport; Intertidal sedimentation; Salinity gradients; Tidal inlets; Topographic effects; San Francisco Bay,

California, USA

1. Intertidal sediment transport

Sediment transport is closely linked to both themorphology and ecology of mudflats and marshes.The balance between erosion and deposition con-trols local bed elevation, and bed elevation withrespect to mean sea level feeds back into sedimenta-

e front matter r 2007 Elsevier Ltd. All rights reserved

.2007.01.010

ng author. Applied Ocean Physics and Engineering,

eanographic Institution, MS 12, Woods Hole, MA

.: +15082892587; fax: +15084572194.

ss: [email protected] (D.K. Ralston).

tion through inundation frequency and duration(Friedrichs and Perry, 2001). Sediment transportmechanisms are spatially heterogeneous across theintertidal zone, from vegetated marsh surface tounvegetated mudflats to deeper subtidal channels.However, sediment fluxes among mudflat, marsh,and channel regions are coupled and depend on thehydrodynamic forcing (Yang et al., 2003). Althoughfield studies have documented sediment transportprocesses on mudflats (Christie et al., 1999; Dyeret al., 2000; Le Hir et al., 2000), and in marshes(Leonard et al., 1995a, b; Christiansen et al., 2000),

.

www.elsevier.com/locate/csr

dx.doi.org/10.1016/j.csr.2007.01.010

mailto:[email protected]

ARTICLE IN PRESSD.K. Ralston, M.T. Stacey / Continental Shelf Research 27 (2007) 1510–1527 1511

observations remain limited because of logisticalchallenges with soft substrate, small and variablewater depth, intermittent erosion events, and spatialvariability between channels and shoals (Dyer et al.,2000).

Sedimentation in the intertidal zone depends inlarge part on settling and scour lag asymmetries(Postma, 1961). Sediment advected into theintertidal zone during flood tides can settle anddeposit around high water slack. Bed sedimentspartially consolidate during slack water such thatequivalent ebb velocities may not be sufficient toresuspend deposited sediment. On marsh surfaces,vegetation reduces flow velocity and suppressesturbulence, enhancing settling and suppressingscour (Leonard and Luther, 1995). Deposition isenhanced by flood dominant tidal velocities,when faster flood velocities of shorter duration arebalanced by longer, slower ebbs. Flood dominantvelocities and settling around high water contributeto suspended sediment concentration maximaduring flood tides for a range of shallow tidalsystems: shallow estuaries (Chant and Stoner, 2001;Fettweis et al., 1998), mudflats (Le Hir et al., 2000;Christie et al., 1999), and marshes (Lawrence et al.,2004). On ebb dominant mudflats, the settlingand scour lag effect can generate greater sedimentconcentrations during floods than ebbs (Dyer et al.,2000).

Astronomical tides are not the only source ofenergy for intertidal sediment transport. Windwaves can generate bed stresses much greater thantidal currents alone, but the effects depend greatlyon wind direction and water depth (Le Hir et al.,2000). Because near-bed wave orbital velocities aregreater in shallower depths, mudflat shoals are moreexposed to wave stresses than deeper channels.Wave bed stresses suspend sediment, and tidalflows advect the increased sediment concentrations.Consequently, the timing of winds with respect toebb/flood and spring/neap cycles impacts net trans-port (Leonard et al., 1995b; Le Hir et al., 2000;Yang et al., 2003). Generally, storms (especiallyaround high water) reverse calm weather onshoretransport of sediment and drive net export ofsediment from mudflats (Wells and Park, 1992;Shi and Chen, 1996; Christie et al., 1999; Dyer et al.,2000; Le Hir et al., 2000; Janssen-Stelder, 2000;Yang et al., 2003). While erosive response to stormsoccurs over time scales of days, recovery to pre-storm elevations can take weeks to months (Yanget al., 2003).

Sediment transport on intertidal mudflats can bebroadly summarized as varying between calm,depositional periods when tidal forcing dominatesand stormy, erosional periods when wind waves aresignificant. During stormy periods, export dependsboth on wave stresses that suspend bed material andon water column mixing that inhibits settlingaround high water. However, sediment transportin mudflat channels may be significantly differentthan on adjacent shoals, as channels have strongertidal velocities and are less impacted by waves. As aresult, sediment exchanges between shoals andchannel—mud can be stored in channels afterstorms (Yang et al., 2003) and during neap tides(Christie et al., 1999), and resuspended during calmand spring periods. Contrary to the intertidal zoneas a whole, channels may offer conduits for exportof sediment during calm conditions (Wells et al.,1990; Dyer et al., 2000).

In estuaries, the salinity distribution impactssediment transport through baroclinic circulationconvergence (e.g., Schubel, 1968) or through tidalasymmetries in velocity (Jay and Musiak, 1994) andturbulent mixing (Geyer, 1993). Sediment depositsand suspended sediment maxima are typically foundnear the limit of the salinity intrusion, either at lowsalinity (Geyer, 1993; Burchard and Baumert, 1998;Chant and Stoner, 2001; Sanford et al., 2001) or inthe fresh region immediately upstream of the saltfield (Uncles and Stephens, 1993). In some estuaries,maximum sediment concentrations have been foundtoward the middle of the salinity distribution due tolateral exchange with supplies of erodible bedsediment (Geyer et al., 1998; Blake et al., 2001).As we present in these observations, the salinity fieldis closely linked to the suspended sediment distribu-tion in a shallow mudflat channel, with peaksediment concentrations consistently in the middleof the salinity gradient, albeit over a much narrowerspatial region than in most estuaries. Observationsof a shallow channel in the Tavy Estuary (UK) alsofound maximum turbidity corresponded with themiddle of a relatively compact salinity field (Unclesand Stephens, 2000). Although the Tavy channelwas deeper, sediment and salinity patterns therewere in many ways similar to these observations—the water column was well mixed and highly turbidearly during flood tides, and strongly stratified withdecreased near-bed suspended sediment during ebbs.

There has been relatively little research onsediment transport in channels through intertidalregions (Wells and Park, 1992; Shi and Chen, 1996;

ARTICLE IN PRESSD.K. Ralston, M.T. Stacey / Continental Shelf Research 27 (2007) 1510–15271512

Lawrence et al., 2004). This work documentssediment transport in a tributary mudflat channelin Central San Francisco Bay during the winter wetseason, detailing both calm and stormy periods.Because of longitudinal salinity gradients generatedby seasonal precipitation, the dominant hydrody-namic feature in the channel is a tidal salinity front(Ralston and Stacey, 2005a). In this paper, wequantify factors controlling sediment transport intributary channels through the intertidal zone. Weconsider how hydrodynamic and estuarine processesat the tidal salinity front impact suspended sedimentconcentrations in mudflat channels. In addition, wecompare calm weather conditions with intermittentmeteorological events with increased wind andprecipitation, considering how this balance betweendominant tidal and meteorological forcing periodsmodulates net sediment flux in the mudflat channel.

2. Site description

The observations presented here were collectedduring field experiments in a mudflat and salt marshregion at the University of California’s RichmondField Station (RFS) in Central San Francisco Bay(Fig. 1). In the experiments, we deployed near-bedinstruments to measure velocity and scalar concen-

Fig. 1. Intertidal mudflat bathymetry at Richmond Field Station (RFS

south of San Pablo Bay (SPB). Instrument frames (A and B) and brea

trations on the mudflats and in the subtidal channelspassing through mudflat and marsh. The channelsflow out of small, mostly urban watersheds, and wefocus on the western channel, Meeker Slough,which has an above-ground length of 2.2 km anddrains roughly 8 km2. Tidal range at the site is about2m, so most of the region shown in the figure isexposed at lower low water (LLW). Water continuesto flow down the channels after the tide has recededoff the mudflats, with discharge in Meeker Sloughestimated between 0.05 and 0:4m3=s duringthe study.

The dominant seasonal variability in the region isMediterranean, with dry summers and wet winters.The discussion here focuses on conditions duringthe wet winter months when precipitation runoffenters the marsh from upstream tributary channels.During the winter, storm systems pass throughthe Bay Area with a periodicity of 5–10 days andbring precipitation and strong winds from thesouth or southeast (Conomos et al., 1985). Pre-cipitation between December and March accountsfor roughly 70% of the annual total in the BayArea, and helps to create a gradient in salinitybetween relatively fresh water draining down thetributary channels and much saltier water in SanFrancisco Bay. The salinity transition occurs over a

, %) on the eastern shore of Central San Francisco Bay (CB) and

kwaters ð’Þ are marked; salt marsh area is white.


narrow region, creating a strong longitudinaldensity gradient that persists throughout the wintermonths. Because of its location and open exposureCentral San Francisco Bay, wave conditions at thefield site are particularly sensitive to winds from thesouth that are commonly associated with winterstorms.

3. Methods

In particular, we focus on data collected fromDecember 19, 2003 to January 3, 2004, wheninstruments were deployed at two locations alongthe channel, downstream where the channel passedthrough mudflats (Frame A), and upstream in thevegetated marsh (Frame B) (Fig. 1). Channel depthwas roughly constant through the study reach; thebed elevation at Frame A was about 0.3m aboveMLLW, and at Frame B was 0.1m above MLLW.On each frame, acoustic Doppler velocimeters(ADVs), conductivity–temperature–depth sensors(CTDs), and optical backscatter sensors (OBSs)were mounted at several elevations near the bed. OnFrame A (mudflat channel) the three ADV samplevolumes were located 0.10, 0.25, and 0.40m abovethe bed, and the ADVs sampled at 16Hz for 5minbursts every 15min (0.25m) or 30min (0.10 and0.40m). CTDs were fixed 0.10, 0.35, and 0.70mabove the bed, with the two lower instrumentssampling every 30 s and the upper CTD samplingevery 15min. OBSs were mounted 0.10 and 0.35mabove the bed, sampling every 30 s and 15min,respectively; in post-processing, data from the lowerOBS were despiked and filtered with a 2 1

2min

moving average.On Frame B (marsh channel), two ADVs were

0.25 and 0.40m above the bed, sampling at 16Hzfor 5min bursts every 30min. Frame B had CTDs0.10 and 0.70m above the bed; the lower CTDsampled every 7.5min and the upper CTD every15min. OBSs were mounted 0.10, 0.20, and 0.40mabove the bed and sampled every 7.5min; a fourthOBS was set 0.70m above the bed and sampledevery 15min. All of the instruments collected datafor two weeks, but 10 days into the experiment(December 29) a large precipitation event signifi-cantly increased flow in the channel and dislodgedthe frames. We limit our analyses to time before thestorm.

Optical backscatter measurements are sensitive toparticle size (Bunt et al., 2002), so the OBSs werecalibrated using suspended sediment from the field

site. Bed sediments on the mudflats ranged fromclay and silt to very fine sand, but suspendedmaterial was dominated by the finer fraction(Ralston, 2005). Laboratory and field calibrationswith suspended sediment were used to convertmeasured OBS voltages to the reported suspendedsediment concentrations (SSC). On Frame A, thelower OBS reached full scale at a sedimentconcentration of 1240mg/L, while full scale forthe upper OBS was 200mg/L. Calibrations for theinstruments on Frame B relied on previous workusing sediment from Northern San Francisco Bayrather than samples from RFS, so conversion tosuspended sediment concentrations was less robust.For this reason, and because the channel geometryis simpler at the downstream frame, we focus theseanalyses on data from Frame A.

Meteorological data during the experiment werecollected from nearby stations. The hourly precipi-tation record combines data from three localstations operated by the California Department ofWater Resources. Wind and air temperature wererecorded at a meteorological station at RFSoperated by the Bay Area Air Quality ManagementDistrict. Finally, tidal stage data came from theNOAA station at Richmond (#9414863). The mapof the intertidal and nearby subtidal bathymetry(Fig. 1) was generated based on depth measure-ments during the surveys of the study area with aboat-mounted acoustic Doppler current profiler(ADCP). Measured depths were converted to bedelevations by referencing to tidal stage at the NOAAstation. Elevations are relative to NAVD 88, butthey are effectively equivalent to elevation aboveMLLW, since it is 0.002m below the NAVD 88datum. Unshaded regions in the map are saltmarshes that were inaccessible to ADCP transecting.

4. Results

During the study, tidal forcing was mixed semi-diurnal with amplitude of about 2m (Fig. 2a).Because of the asymmetric tides, the mudflats wereexposed diurnally each LLW; water covered muchof the intertidal region at higher low waters (HLW).This distinction is important when describingconditions on the mudflats, so in the followingdiscussion the diurnal tides are subdivided into fourperiods: an initial strong flood from LLW to LHW,a weak ebb toward HLW, a weak flood towardHHW, and a strong flood toward LLW at the endof the inundation.

ARTICLE IN PRESS

Fig. 2. Suspended sediment observations in the mudflat channel. (a) Water surface in the channel, tidal stage at NOAA station, and

precipitation; (b) wind speed; (c) wind direction; and (d) suspended sediment concentration 0.10 and 0.35m above the bed at Frame A.

Forcing periods shown in greater detail are noted above (a): calm conditions (Fig. 4), wind from the south (Fig. 7).

D.K. Ralston, M.T. Stacey / Continental Shelf Research 27 (2007) 1510–15271514

Suspended sediment concentrations during theexperiment varied from background levels of about25mg/L to over 1000mg/L during storms (Fig. 2).More typically, the diurnal tidal cycle variability inSSC yielded peaks of about 300mg/L. Sedimentconcentrations depended on the hydrodynamicforcing and the available sediment supply, so severalsmall wind and precipitation events during thedeployment impacted salinity and suspended sedi-ment in the channel. During calm weather, tidalinundation of the mudflats and marsh drove thedominant diurnal pattern in SSC, but intermittentincreases in sediment concentration were due toprecipitation and wind.

Salinity and buoyancy dynamics during theDecember 2003 experiment were typical ofwinter wet season conditions. Salinity varied tidallybetween about 28 psu and nearly fresh (Fig. 3a).The longitudinal salinity gradient ðqS=qxÞ wasgreatest near the tidal salinity front at the beginningof each inundation, from 20 to 40 psu/km(Fig. 3b). The tidal signal for qS=qx correspondedwith the buoyancy frequency: N2 ¼ �ðg=r0ÞðDr=DzÞ, where Dr is the density difference bet-ween instruments 0.10 and 0.35m above the bed(Fig. 3c). Stratification depended on qS=qx

through tidal straining; advection of the salinitygradient increased N2 during ebbs, and straining

ARTICLE IN PRESS

Fig. 3. Salinity and stratification. (a) Salinity 0.10, 0.35, and 70m above the bed on Frame A; (b) longitudinal salinity gradient ðqS=qxÞ

between Frames A and B, 0.10m above the bed; (c) stratification ðN2Þ between the bottom two salinity sensors on Frame A, with instances

of unstable stratification ðN2o0Þ marked in white.

D.K. Ralston, M.T. Stacey / Continental Shelf Research 27 (2007) 1510–1527 1515

destabilized the water column and reduced N2

during floods.The diurnal inundation period of the intertidal

zone helped create and preserve a very strong along-channel density gradient through the wet season.Although dispersion reduced the salinity gradientthrough each tide, a sharp front was regenerated atthe beginning of each strong flood. Stratificationdue to tidal straining significantly damped turbulentmixing during ebbs, creating tidal asymmetries inshear, Reynolds stress, and eddy viscosity. A morecomplete discussion of the hydrodynamics can befound elsewhere (Ralston and Stacey, 2005a,b), butthe basic framework is important for consideringsediment transport mechanisms.

4.1. Tidal cycle sediment transport

Suspended sediment concentrations in the mud-flat channel depended both on regular tidal forcingand on intermittent perturbations driven by meteor-ological events. During calm weather when meteor-ological forcing was absent or weak, tidal sedimenttransport was periodic with each inundation andwas tied to advection of the salinity front. We focuson the longest continuous period during thedeployment with calm winds and no rain, roughlyfour days of data following moderate precipitationon day 358 (Fig. 4). For orientation within the tidalcycle, water surface elevation, salinity, and velocityare plotted along with suspended sediment.

ARTICLE IN PRESS

Fig. 4. Conditions in mudflat channel during calm weather conditions. (a) Water surface in the channel, tidal stage, and precipitation; (b)

salinity 0.10, 0.35, and 0.70m above the bed on Frame A; (c) along-channel velocity 0.10, 0.25, and 0.40m above the bed; (d) suspended

sediment concentration 0.10 and 0.35m above the bed.


During each inundation, the salinity front ad-vanced up the channel during the strong flood andbrought a spike in SSC along with the increase insalinity (e.g., t�358:9). Sediment concentrationsdropped to much lower levels after the front passedthe instrument frame. Sediment concentrationsremained low until midway through weak ebbswhen SSC steadily increased and salinity steadilydecreased (e.g., t�359:25). Local maxima in SSCoccurred around HLW, and SSC decreased when thesalinity front advected back upstream during weakfloods. Finally during strong ebbs, SSC rose as thesalinity front passed the instruments, with maximain the middle of the gradient (e.g., 359.7). Sediment

concentrations decreased as the marsh continued todrain around LLW until the following strong floodelevated sediment concentrations. Maximum velo-cities during strong floods and strong ebbs weresimilar, about 0.4 cm/s. However, suspended sedi-ment concentrations did not directly depend on localvelocities, since peak sediment concentrations typi-cally preceded peak flood velocities, and concentra-tions during inundating floods were much greaterthan during receding ebbs despite similar velocitymagnitudes. Similarly, diurnal maxima in suspendedsediment were not strictly tied to tidal stage, butrather were most closely depended on advection ofthe salinity field up the channel.


4.1.1. Advection and dispersion at the salinity front

Around LLW, the salinity gradient betweenfresher water upstream of the marsh and saltierwater downstream in the bay occurred over avery narrow region (�200m), creating largeqS=qx. In this transition zone, sediment concentra-tions were greatly elevated over the backgroundconcentrations, roughly 10 times the backgroundSSC of 25–30mg/L. To consider the spatialdistributions of salinity and suspended sediment,we convert salinity and sediment time series tofunctions of distance along the channel axis byintegrating velocity in time: x ¼

R t

0 uðtÞdt. To do so,we assume that flow is dominantly along thechannel axis, and that the time scale for longitudinaladvection is short compared with the time scalefor vertical mixing (Ralston and Stacey, 2005b).The spatial distributions of salt and sediment areplotted relative to the center of each salinity front,

Fig. 5. Salinity and suspended sediment front during calm conditions. (

ebbs; (b) spatial distribution of suspended sediment during strong floo

calculated from time series at Frame A assuming advection by the local v

of the 15 psu isohaline, and dashed lines represent analytical dispersion w

arrows indicate progression of time/distance.

defined here as the location of the 15 psu isohaline(Fig. 5).

Three days of strong floods and strong ebbsduring the calm weather period have very similarstructures. As tidal stage rises from LLW, thetransition zone between fresh and salty wateradvects up the channel and across the mudflats,passing the instruments with a narrow spike insediment concentration. When the transition regionagain advects past the instruments during the strongebb, maximum sediment concentrations are muchlower than during the strong flood. During bothinundating and receding tides, the sediment dis-tribution is approximately Gaussian and is centeredat the salinity front. The link between suspendedsediment and the salinity field can be seen moreclearly by plotting of sediment concentrationagainst salinity (Fig. 5c). Sediment concentrationswere greatest in the middle of the salinity front,

a) Spatial distribution of salinity during strong floods and strong

ds and strong ebbs. The salinity and sediment distributions were

elocity; each day’s observation was plotted relative to the location

ith Kx ¼ 8m2=s; (c) Suspended sediment as a function of salinity;


between 15 and 20 psu. Similarly, during interveningweak ebbs and floods (not shown), sedimentconcentrations increased as the salinity front regionadvected toward the instrument frames. The entirefrontal region did not pass the instruments duringthe weak ebbs and weak floods, so maximumsediment concentrations occurred near HLW slack.

The spatial spread of the transition zone betweensalty and fresh water through each inundation wasthe result of tidal dispersion that decreased thesalinity gradient and suspended sediment concen-trations in the front. Dispersion of the salinitygradient was due spatial correlations in salinity andvelocity, both vertically because of periodic strati-fication and laterally due to differential advectionbetween the channel and shoals (Ralston andStacey, 2005a). An analytical model assumingalong-channel advection and dispersion of a salinitystep function between bay and marsh waterrepresents the spread of the front:

Sðx; tÞ ¼S0

21þ erf

x� xcffiffiffiffiffiffiffiffiffiffi4Kxtp

� �� ,

where S0 is the initial magnitude of the step and xc

advects with the center of the front. The salinitydata were fit with an along-channel dispersivity (KxÞ

of 5–10m2=s.Similarly, evolution of the along-channel sus-

pended sediment distribution with time can beidealized by advection and dispersion of an initialsediment spike:

Cðx; tÞ ¼M0ffiffiffiffiffiffiffiffiffiffi4Kxtp exp �

ðx� xcÞ2

4Kxt

� �,

where M0 is the initial mass, and xc advects with thecenter of the front. The spread of the spatialdistribution as measured by the variance ðs2Þprovides an estimate of the along-channel dispersiv-ity: Kx ¼ ðs22 � s21Þ=2ðt2 � t1Þ (Fischer et al., 1979).We measure the spread of the suspended sedimentpeak by calculating the second moment of thealong-channel concentration distribution. Becausethe instrument frames were relatively close to thelow tide elevation, passage of the salinity front andsediment spike around low water was quicklyfollowed by the returning flood tide. To completelycapture the return to background suspended sedi-ment concentrations away from the salinity front,we measure the width of the sediment spike on thesalty side of the front. Using the change in varianceduring the inundation period between strong floodand strong ebb, the longitudinal dispersivity for

suspended sediment was calculated to be 5–8m2=s.This estimate is consistent with the longitudinaldispersivity of salt based on the spread of thesalinity gradient.

Thus, suspended sediment variability in the mud-flat channel during calm weather was predominatelydue to advection and dispersion of a pulse ofsediment centered at the tidal salinity front. Neterosion and deposition in the intertidal zone appearto be small contributions to the suspended sedimentbudget at tidal timescales compared with advectionand tidal dispersion. We consider integration of thesetidal processes over longer periods in Section 5.

4.1.2. Secondary sediment maxima during floods

In both the time series and the spatial distribu-tions, the primary spike in suspended sedimentduring the strong flood was followed by a second,but distinct, local maximum. The secondary peakfollowed approximately 50min after the maximumsediment concentration (Fig. 4d); this correspondedwith a spatial lag of 500–700m (Fig. 5b). Thesecondary maxima occurred behind the salinityfront every day, and we focus on one such sequenceduring a strong flood (Fig. 6). In this case, thesecondary peak (t�361:025) was roughly half thesediment concentration in the salinity frontðt�360:99Þ. The secondary maximum was distinctfrom the salinity front, and between the two peaks,bay water advected past the instruments withrelatively constant near-bed salinity and velocity.However, near-bed water temperature during thisperiod mirrored the variability in suspended sedi-ment (Fig. 6c). Temperature decreased rapidly whensediment concentrations increased, and slowly roseas sediment concentrations decreased.

During the field deployment, inundation of themudflats began each night when air temperaturewas significantly lower than the water temperature(Fig. 6c). As a result, near-surface water in contactwith the night air cooled more rapidly than near-bedwater. Periods of strong vertical mixing redistrib-uted cool near-surface water downward as well assuspended sediment. Mixing at the salinity frontproduced the initial increase in suspended sedimentand decrease in near-bed temperature. However,immediately behind the salinity front, the watercolumn was salinity stratified and turbulent mixingwas suppressed (Fig. 6d). Only when stratificationdecreased did vertical mixing increase, yielding bothincreased suspended sediment and decreased near-bed temperature.

ARTICLE IN PRESS

Fig. 6. Detail of secondary peak in suspended sediment during strong floods. (a) Depth and tidal stage for the entire inundation, with the

period in the lower panels highlighted. (b) Salinity 0.10 and 0.35m above the bed; (c) temperature 0.10 and 0.35m above the bed along

with air temperature; (d) stratification; (e) suspended sediment 0.10m above the bed.


Salinity stratification behind the flood front was aregular feature during the wet season. Differentialadvection of the salinity front between channel andshoals created lateral density gradients and forcedlateral baroclinic circulation, with flow out of thechannel near the bed and convergent at the surface(Ralston and Stacey, 2005a). The lateral exchangeenhanced stratification behind the front, and helpedcreate the secondary peak in suspended sediment.When the water column was stratified behind thefront, near-bed flow was relatively warm with lowsuspended sediment concentrations. When the water

column mixed vertically, near-surface water thatwas cooler and had higher sediment concentrationsmixed downward toward the sensors.

The instruments were positioned in the center ofthe channel, so we cannot directly address thesource of the sediment in the secondary maximum.The secondary peak typically occurred when waterdepth in the channel was about 0.7m, and at thisdepth the shoals adjacent to the instrument frameswere shallowly inundated. Because of the closeproximity to the salt marsh, the channel banks nearthe instrument frame were narrower and deeper


than farther downstream on the mudflats. At thetime of the secondary peak, shallow flow on theshoals provided a source of cool, sediment-ladenwater for advection into the channel, and mixing atthe relatively steep channel banks could havecontributed to the breakdown of stratificationbehind the front. Without measurements acrossthe bathymetric gradient, though, we cannotdetermine the channel–shoal exchange. Sedimentconcentrations in the secondary maxima weremuch lower than in the salinity front, and tidaldispersion smoothed the gradient between the twopeaks by the time the tide receded during strongebbs.

4.2. Episodic forcing

Description of the tidal cycle sediment transportin the previous section focused on a period ofrelatively calm weather, with light winds and littleprecipitation. Winter storms perturb the tidal cyclepattern by altering both the hydrodynamics and thesediment supply in the intertidal zone. Winterstorms greatly increased suspended sediment onthe mudflats—the highest sediment concentrationsduring the observations were during and immedi-ately after storms (Fig. 2). Storms added sedimentto the water column both by eroding bed materialby wind waves and by delivering sediment withprecipitation runoff down the tributary channels.Increased freshwater discharge due to runoff afterstorms also strengthened qS=qx (Fig. 3), impactingthe buoyancy dynamics in the channel. Theresponse of the system to these episodic events canbe considered in the context of the tidal cyclepattern, with changes in buoyancy and sedimentsupply as perturbations to the calm weatherexchange.

4.2.1. Wind: sediment suspension

In shallow water, wind waves can create sig-nificant bed stresses that erode sediment and mix thewater column vertically. Wind waves suspendsediment both on the mudflats and in the shallowsubtidal region offshore, increasing sediment con-centrations advected across the intertidal zoneduring floods. Winter storms bring winds from thesouth, but prevailing winds are generally from thenorth. The mudflats at the field site face south, andare sheltered by land, marsh, or breakwaters on theother three sides. Consequently, sediment transportresponse to wind depended greatly on wind direc-

tion. Winds from the north and east had verylimited fetch to generate wind waves. The channelorientation was approximately east–west as itentered the mudflats, so flow there was moderatelysensitive to wind along this axis. For example, windfrom the east pushed water up the channel aroundlow water, increasing the depth, salinity, andsuspended sediment at the instruments (Fig. 2, briefincrease in water depth at t�354:6).

Only winds from the south significantly alteredsuspended sediment patterns in the channel. Wefocus on conditions in the channel during one stormperiod (Fig. 7). Although strong winds from thesouth began around HHW, sediment concentrationsadvecting past the instruments during the ebb wererelatively unaffected because flow in the marshremained sheltered by breakwaters and vegetation.However, wind waves suspended sediment offshorein the shallow subtidal region, and during thefollowing flood, maximum sediment concentrationsat the front were more than twice the concentrationstypical of calm weather. The elevated sedimentconcentrations preceded the precipitation, so thelikely source was bed erosion by wind wavesrather than terrestrial runoff or direct impact ofraindrops on exposed mudflat. Suspended sedimentconcentrations in the channel remained elevatedthroughout the strong southerly wind, but localmaxima still coincided with the salinity gradient.Shortly after the winds dropped, sediment concen-trations decreased to levels more typical of the tidalforcing.

4.2.2. Precipitation: buoyancy and sediment supply

In addition to wind, winter storms broughtprecipitation. Runoff increased freshwater dis-charge in the tributary mudflat channels, andsupplied both buoyancy and terrestrial sediment tothe intertidal region. For example, moderate pre-cipitation on day 358 altered both salinity andsediment conditions in the channel (Fig. 8). Early inthe strong ebb, suspended sediment concentrationswere low and the water column was salinitystratified (t�358:58). However, with the storm,precipitation and wind from the south intensified,and near-bed sediment concentrations increasedrapidly (t�358:595, #1). Fresh, sediment-ladenrunoff mixed vertically downward from the surface,increasing near-bed sediment concentrations anddecreasing near-bed salinity. The near-surface layerhad been warmed by the afternoon air, so when itmixed downward, near-bed temperature increased.

ARTICLE IN PRESS

Fig. 7. Conditions in mudflat channel during winds from the south. (a) Water surface in the channel, tidal stage, and precipitation; (b)

salinity 0.10, 0.35, and 0.70m above the bed; (c) wind speed and direction; (d) suspended sediment concentration 0.10 and 0.35m above

the bed.


Mixing of sediment, salinity, and temperature overthe water column was due to wind waves shoalingand breaking on the shallow mudflats.

After a brief period when sediment concentra-tions were high and the water column was verticallymixed, near-bed suspended sediment rapidly de-creased at the same time that the water columnrestratified (t�358:605, #2). Near-bed suspendedsediment decreased, salinity increased, and tem-perature decreased, all corresponding with flowreversal near the bed as the lowest ADV returned toflooding velocities while the surface layer continuedto ebb. The flow reversal was driven by dense baywater advected up onto the mudflats by the windfrom the south, as the wind-setup in combination

with increased freshwater discharge significantlystrengthened the baroclinic pressure gradient inthe channel. As bay water advected upstreamtoward the marsh, near-bed sediment concentra-tions decreased and the fresher, sediment-ladenrunoff was again confined to a surface layer.Sediment concentration near the bed remainedrelatively low until much later in the ebb, whenthe entire water column was again ebbing and thechannel mixed vertically due to falling water leveland increasing velocity (t�358:64, #3). Upstream inthe marsh at Frame B, conditions were similarduring this period—the water column restratifiedmidway through the ebb and near-bed flow re-versed, and near-bed sediment concentration

ARTICLE IN PRESS

Fig. 8. Conditions in mudflat channel after moderate precipitation. (a) Depth and tidal stage for the entire inundation, with the period in

the lower panels highlighted; (b) Salinity 0.10 and 0.35m above the bed; (c) temperature 0.10 and 0.35m above the bed; (d) along-channel

velocity 0.10, 0.25, and 0.40m above the bed; (e) suspended sediment 0.10 and 0.35m above the bed.


decreased simultaneously. During the storm, thecombination of increased discharge, wind-drivensetup, and wind wave mixing altered the hydro-dynamics, while terrestrial runoff increased theavailable supply of suspended sediment in thechannel.

5. Analysis of sediment fluxes

Because of particle settling, suspended sediment isnot necessarily conserved in the water column.Suspended sediment at the field site consisted

predominantly of clays and fine silt, with dominantparticle diameters around 10mm (Ralston, 2005).Fine cohesive sediments like these tend to aggregateinto large, low density flocs that settle faster thanindividual particles. Flocs on the order of 100mmdiameter are common in San Francisco Bay, butsizes can range from 10 to 1000mm (Kranck andMilligan, 1992; Gartner et al., 2001). Field estimatesof floc settling rates found velocities of 0.1–0.5mm/sfor floc diameters of about 150mm in the coastalocean off Northern California (Sternberg et al.,1999; Hill et al., 2000).


Assuming settling velocities of 0.1–0.5mm/s anda water depth of 2m, the time scale for settling toremove all suspended sediment from the watercolumn is between 1 and 5 h, significantly less thanthe daily inundation period. However, total sedi-ment in the salinity front did not substantiallychange between inundating and receding tides (Fig.5b). Although dispersion spread the initial sedimentspikes spatially, the integral of near-bed suspendedsediment concentration remained largely unchangedbetween strong floods and strong ebbs. Given theshallow depth, settling during the diurnal inunda-tion could remove most suspended sediment from aquiescent water column. However, turbulent mixing

Fig. 9. Rouse number and cumulative sediment flux. (a) Water surfa

highlighted gray; (b) Rouse number, P ¼ ws=ðku�Þ; (c) cumulative sedim

positive slope for onshore transport and negative slope for offshore tra

in the mudflat channel appears sufficient to keepmost of the sediment at the salinity front insuspension.

The ratio of the settling velocity to the frictionvelocity provides a scale for the balance betweenmixing and particle settling. Specifically, the Rousenumber is P ¼ ws=ku�, where ws is the settlingvelocity, u� is the friction velocity, and k is vonKarman’s constant (0.4). If we set the frictionvelocity equal to the square root of the Reynoldsstress at the lowest ADV (u��

ffiffiffiffiffiffiffiffiu0w0p

10 cmÞ, andassume ws�0:5mm=s, then the resulting Rousenumbers were much less than one for most of theexperiment (Fig. 9b). For Rouse numbers much less

ce, tidal stage, and precipitation, with periods of precipitation

ent flux in the channel from the beginning of the deployment, with

nsport.


than 1, sediment is suspended throughout the watercolumn; for Rouse numbers much greater than 1,settling dominates vertical mixing and suspendedsediment is concentrated near the bed. The Rousenumber has been filtered with a 3-h moving average,assuming this is a relevant time scale for settlingacross the water column. The time series modulatedwith the spring–neap variability in tidal forcing andturbulent mixing. Rouse numbers were lowestduring the middle of the deployment around day358 when tidal amplitudes were greatest; addition-ally, wind and freshwater discharge increased near-bed stresses around this time. Rouse numbersincreased during the calm weather period, coincid-ing with lower tidal amplitudes during the neap.

Although sediment transport in the channel atdiurnal timescales is dominated by tidal advectionand dispersion, we can quantify small differences intransport between tidal cycles by calculating cumu-lative sediment flux past the instrument frame.Velocity and sediment concentrations were mea-sured at fixed elevations in the center of the channel,so we must make several simplifying assumptions toestimate the mass flux. We assume flow is dom-inantly along the channel axis, uniform across thechannel, and that the velocity profile is logarithmic.We also assume the vertical distribution of sus-pended sediment follows the Rouse profile:

CðzÞ ¼ CðaÞh� z

z

a

h� a

� �P=b

,

where concentration at elevation z is relative to theconcentration at reference elevation a; h is the waterdepth, P is the Rouse number, and b is a coefficientassumed equal to 1. Taking the OBS at 0.10mabove the bed for the reference sediment concentra-tion, we multiply the resulting sediment profile bythe velocity profile calculated from the ADVmeasurements to get the sediment flux through time.

Keeping in mind the limitations of these assump-tions, the resulting estimate of cumulative sedimentflux shows effects of the episodic storms and of thespring–neap transition in turbulent mixing (Fig. 9c).The cumulative flux is relative to the arbitrary zeroof the beginning of the deployment, so periods withpositive slopes indicate net onshore transport whileperiods with negative slopes have net offshoretransport. The largest magnitude shifts in sedimentflux are associated with precipitation, with greatlyincreased export associated with storms during themiddle (days 357–358) and at the end (day 363) of

the observations. Between the storms, the directionof net transport in the channel depended on thestrength of the tidal forcing. Early in the deploy-ment, cumulative transport was generally offshore(slope o0); this corresponded with a period ofspring tides, but the effect of increased tidalamplitude may have been enhanced by wave stressdue to moderate winds from the east. During thecalm weather period with relatively weak tidalforcing, net transport in the channel was slightlyonshore (slope 40). Onshore transport during thecalm period corresponds with higher Rouse num-bers, with sediment deposition more likely whenReynolds stresses decreased due to weaker neaptides and stronger periodic stratification.

6. Discussion

Based on the calm weather tidal forcing and theepisodic perturbations by storms, we can begin todescribe sediment transport in the mudflat channelduring the wet season. Maximum sediment concen-trations typically corresponded with the location ofthe sharp salinity gradient between fresh dischargeand saltier bay water. Through each daily inunda-tion, the region of suspended sediment and salinitygradient advected across the intertidal zone anddispersed longitudinally.

A conceptual model for sediment transport in thissystem incorporates both the tidal and episodicforcing regimes. With each winter storm, freshwaterrunoff delivers a pulse of sediment down thetributary channel to the intertidal and nearbysubtidal areas. The bathymetry is such that themudflat channel becomes progressively shallowerand eventually disappears where it intersects thesubtidal zone, so flow from the channel expandsradially and velocities decrease where the channelenters the bay. As a result, some of the suspendedsediment ebbing down the channel around lowwater is likely to deposit immediately offshore of themudflats. This store of sediment is then available forresuspension in the salinity front at the beginning ofthe following strong flood.

Deposition offshore during ebbs and suspensionat the flood front are both enhanced by dynamics atthe strong salinity gradient. During ebbs, tidalstraining increases stable stratification, reducesturbulent mixing, and enhances sediment deposition(Geyer, 1993). Periodic stratification is strongerand lasts longer during ebbs of neap tides, whentidal mixing and near-bed turbulence are weaker.


Consequently, deposition rates are greater duringneaps, consistent with the net flux calculations in thechannel.

During floods, tidal straining reduces stratifica-tion. Advection of the salinity gradient can createunstable density profiles that mix near-surface waterdownward and enhance turbulence near the bed.The salinity sensors in the mudflat channel recordedunstable density profiles around passage of the frontduring strong floods (Fig. 3c), and strong verticalvelocities can increase sediment resuspension atsalinity fronts (Largier, 1993). Because qS=qx wasgreatest around low water at the start of theinundating flood, the active turbulence at the frontcould suspend local sediment deposits or couldvertically mix sediment associated with relativelyfresh discharge down the channel at the end of thestrong ebb.

Unfortunately, the instruments were positionedupstream in the mudflat channel and could notobserve creation of the sediment spike in theflooding salinity front. Local bed erosion near theinstrument frames did not significantly contribute tothe sediment maxima in the front, since peaksediment concentrations preceded maximum shearstresses. Instead, the tidal sediment transportpattern may depend on expansion of channelizedflow as it enters the bay and an associated supply ofmobile sediment just offshore of the mudflats.During each inundation the sediment spike advects,disperses, and some small fraction may deposit inthe intertidal zone depending on the tidal and waveenergy. During the calm period of these observa-tions, the tidal forcing regime appeared to bemoderately depositional. Fine sediment that enteredthe system with precipitation runoff prior to thecalm period was gradually redistributed to deposi-tional areas on the mudflats or marsh or wasexported offshore. However, most of the sedimentsuspended at the salinity front simply sloshed backand forth across the intertidal zone, and large shiftsin net transport resulted from storm events.

The mudflat channels at the field site aretributaries of small watersheds, but many tidalchannels around San Francisco Bay are created andmaintained strictly by tidal flows (Atwater et al.,1979). Salinity, and therefore sediment dynamics inchannels without freshwater discharge are likelyvery different from conditions described here. Forexample, the seasonal sediment transport patternson extensive mudflats in nearby San Pablo Bayare unlike those at the field site (Krone, 1979;

Thompson-Becker and Luoma, 1985). In San PabloBay, wind forcing dominates transport duringsummer months, and fine sediment is supplied fromthe Sacramento River offshore of the mudflatsrather than from local runoff. At the field site,tributary channels deliver fine sediment, and south-erly winds have the greatest impact during thewinter. Tidal salinity fronts are important tosediment transport in mudflat channels with fresh-water input, but are not necessarily present in allintertidal regions.

7. Summary

On mudflats at edge of Central San FranciscoBay, sediment transport in a shallow tributarychannel was tied to advection and dispersion of atidal salinity front, but was modulated by wind andprecipitation associated with winter storms. Duringcalm weather, the highest suspended sedimentconcentrations were in the frontal zone thatadvected up the intertidal zone and back out tothe bay each inundation period. Along-channeldispersion of suspended sediment correspondedwith dispersion of the longitudinal salinity gradient.Net sediment transport was slightly onshore duringthe calm weather, as asymmetries in stratificationdue to tidal straining of the salinity front, particu-larly during the relatively weak neap tidal forcing,enhanced deposition. Additionally, a secondarymaximum in suspended sediment during floodsresulted from exchange between the mudflat chan-nel and shoals after the breakdown of stratificationimmediately behind the salinity front.

The regular sediment transport by tidal forcingwas periodically altered by winter storms. Duringstorms, strong winds from the south generated windwaves and temporarily increased suspended sedi-ment concentrations. However, because of thesouth-facing orientation of the field site, the morecommon seasonal winds from the north and north-west had negligible impact on sediment suspensionand transport. Increased discharge down the tribu-tary channels due to precipitation during stormshad more lasting impact on sediment transportconditions, supplying both buoyancy and finesediment to the system. Net sediment transportdepended on the balance between calm weathertidal forcing and episodic storms, and thereforeshould vary with the season (freshwater inputand prevailing wind direction), and the location(tributary or strictly tidal mudflat channels) and


orientation (relative to prevailing winds) of a givenintertidal region.

Acknowledgements

The authors thank Matt Brennan, Jon Fram,Maureen Martin, Deanna Sereno, and Stefan Talkefor field assistance, and are grateful to JohnTrowbridge for helpful discussions. The researchwas funded by National Institutes of Health GrantP42ES0475 from the National Institute of Environ-mental Health Sciences.

References

Atwater, B.F., Conard, S.G., Dowden, J.N., Hedel, C.W.,

MacDonald, R.L., Savage, W., 1979. History, landforms,

and vegetation of the estuary’s tidal marshes. In: Conomos,

T.J. (Ed.), San Francisco Bay: The Urbanized Estuary.

American Association for the Advancement of Science, San

Francisco, pp. 347–386.

Blake, A.C., Kineke, G.C., Milligan, T.G., Alexander, C.R.,

2001. Sediment trapping and transport in the ACE Basin,

South Carolina. Estuaries 24 (5), 721–733.

Bunt, J.A.C., Larcombe, P., Jago, C.F., 2002. Quantifying the

response of optical backscatter devices and transmissometers

to variations in suspended particulate matter. Continental

Shelf Research 19, 1199–1220.

Burchard, H., Baumert, H., 1998. The formation of estuarine

turbidity maxima due to density effects in the salt wedge. A

hydrodynamic process study. Journal of Physical Oceano-

graphy 28, 309–321.

Chant, R.J., Stoner, A.W., 2001. Particle trapping in a stratified

flood-dominated estuary. Journal of Marine Research 59,

29–51.

Christiansen, T., Wiberg, P.L., Milligan, T.G., 2000. Flow and

sediment transport on a tidal salt marsh surface. Estuarine

Coastal and Shelf Science 50, 315–331.

Christie, M.C., Dyer, K.R., Turner, P., 1999. Sediment flux

measurements from a macro tidal mudflat. Estuarine Coastal

Shelf Science 49, 667–688.

Conomos, T.J., Smith, R.E., Gartner, J.W., 1985. Environmental

setting of San Francisco Bay. Hydrobiologia 129, 1–12.

Dyer, K.R., Christie, M.C., Feates, N., Fennessy, M.J., Pejrup,

M., van der Lee, W., 2000. An investigation into processes

influencing the morphodynamics of an intertidal mudflat, the

Dollard Estuary, The Netherlands: I. Hydrodynamics and

suspended sediment. Estuarine Coastal and Shelf Science 50,

607–625.

Fettweis, M., Sas, M., Monbaliu, J., 1998. Seasonal, neap-spring

and tidal variation of cohesive sediment concentration in the

Scheldt Estuary, Belgium. Estuarine Coastal and Shelf

Science 47 (3), 21–36.

Fischer, H.B., List, E.J., Koh, R.C.Y., Imberger, J., Brooks,

N.H., 1979. Mixing in Inland and Coastal Waters. Academic

Press, San Diego.

Friedrichs, C.T., Perry, J.E., 2001. Tidal salt marsh morphody-

namics: a synthesis. Journal of Coastal Research SI 27, 7–37.

Gartner, J.W., Cheng, R.T., Wang, P.F., Richter, K., 2001.

Laboratory and field evaluations of the LISST-100 instrument

for suspended particle size determinations. Marine Geology

175, 199–219.

Geyer, W.R., 1993. The importance of suppression of turbulence

by stratification on the estuarine turbidity maximum.

Estuaries 16, 113–125.

Geyer, W.R., Signell, R.P., Kineke, G.C., 1998. Lateral trapping

of sediment in a partially mixed estuary. In: Dronkers, J.,

Scheffers, M. (Eds.), Physics of Estuaries and Coastal Seas.

Balkema, Rotterdam, pp. 115–124.

Hill, P.S., Milligan, T.G., Geyer, W.R., 2000. Controls on

effective settling velocity of suspended sediment in the Eel

River flood plume. Continental Shelf Research 20, 2095–2111.

Janssen-Stelder, B., 2000. The effect of different hydrodynamic

conditions on the morphodynamics of a tidal mudflat in the

Dutch Wadden Sea. Continental Shelf Research 20,

1461–1478.

Jay, D.A., Musiak, J.D., 1994. Particle trapping in estuarine tidal

flows. Journal of Geophysical Research 99 (C10),

20445–20461.

Kranck, K., Milligan, T.G., 1992. Characteristics of suspended

particles at an 11-hour anchor station in San Francisco Bay,

California. Journal of Geophysical Research 97 (C7),

11373–11382.

Krone, R.B., 1979. Sedimentation in the San Francisco Bay

system. In: Conomos, T.J. (Ed.), San Francisco Bay: The

Urbanized Estuary. American Association for the Advance-

ment of Science, San Francisco, pp. 85–96.

Largier, J.L., 1993. Estuarine fronts: how important are they?

Estuaries 16 (1), 1–11.

Lawrence, D.S.L., Allen, J.R.L., Havelock, G.M., 2004. Salt

marsh morphodynamics: an investigation of tidal flows and

marsh channel equilibrium. Journal of Coastal Research 20

(1), 301–316.

Le Hir, P., Roberts, W., Cazaillet, O., Christie, M., Bassoullet, P.,

Bacher, C., 2000. Characterization of intertidal flat hydro-

dynamics. Continental Shelf Research 20, 1433–1459.

Leonard, L.A., Luther, M.E., 1995. Flow dynamics in tidal

marsh canopies. Limnology and Oceanography 40 (8),

1474–1484.

Leonard, L.A., Hine, A.C., Luther, M.E., 1995a. Surficial

sediment transport and deposition processes in a Juncus

roemerianus marsh, west-central Florida. Journal of Coastal

Research 11 (2), 322–336.

Leonard, L.A., Hine, A.C., Luther, M.E., Stumpf, R.P., Wright,

E.E., 1995b. Sediment transport processes in a west-central

Florida open marine marsh tidal creek; the role of tides and

extra-tropical storms. Estuarine Coastal and Shelf Science 41,

225–248.

Postma, H., 1961. Transport and accumulation of suspended

matter in the Dutch Wadden Sea. Netherlands Journal of Sea

Research 1, 148–190.

Ralston, D.K., 2005. Hydrodynamics and scalar transport in

subtidal channels through intertidal mudflats, Ph.D. Thesis,

University of California, Berkeley.

Ralston, D.K., Stacey, M.T., 2005a. Longitudinal dispersion and

lateral circulation in the intertidal zone. Journal of Geophy-

sical Research 110, C07015.

Ralston, D.K., Stacey, M.T., 2005b. Stratification and turbulence

in subtidal channels through intertidal mudflats. Journal of

Geophysical Research 110, C08009.


Sanford, L.P., Suttles, S.E., Halkma, J.P., 2001. Reconsidering

the physics of the Chesapeake Bay estuarine turbidity

maximum. Estuaries 24 (5), 655–669.

Schubel, J.R., 1968. Turbidity maximum of the northern

Chesapeake Bay. Science 161, 1013–1015.

Shi, Z., Chen, J.Y., 1996. Morphodynamics and sediment

dynamics on intertidal mudflats in China (1961–1994).

Continental Shelf Research 16 (15), 1909–1926.

Sternberg, R.W., Berhane, I., Ogston, A.S., 1999. Measurement

of size and settling velocity of suspended aggregates on the

northern California continental shelf. Marine Geology 154,

43–53.

Thompson-Becker, E.A., Luoma, S.N., 1985. Temporal fluctua-

tions in grain-size, organic materials and iron concentrations

in intertidal surface sediment of San Francisco Bay. Hydro-

biologia 129, 91–107.

Uncles, R.J., Stephens, J.A., 1993. The freshwater–saltwater

interface and its relationship to the turbidity maximum

in the Tamar Estuary, United Kingdom. Estuaries 16 (1),

126–141.

Uncles, R.J., Stephens, J.A., 2000. Observations of currents,

salinity, turbidity and intertidal mudflat characteristics and

properties in the Tavy Estuary, UK. Continental Shelf

Research 20, 1531–1549.

Wells, J.T., Park, Y.A., 1992. Observations on shelf and subtidal

channel flow: implications of sediment dispersal seaward of

the Keum River Estuary, Korea. Estuarine Coastal and Shelf

Science 34, 365–379.

Wells, J.T., Adams, C.E., Park, Y.A., Frankenberger, E.W.,

1990. Morphology, sedimentology and tidal channel processes

on a high-tide-range mudflat, west coast of South Korea.

Marine Geology 95, 111–130.

Yang, S.-L., Friedrichs, C.T., Shi, Z., Ding, P.-X., Zhu, J., Zhao,

Q.-Y., 2003. Morphological response of tidal marshes, flats,

and channels of the Outer Yangtze River mouth to a major

storm. Estuaries 26 (6), 1416–1425.

tidal and meteorological forcing of sediment transport in ... · pdf filetidal and...

Documents