Storm influence on SPM concentrations in a coastal turbidity maximum area with high
anthropogenic impact (southern North Sea) Michael Fettweis1, Frederic Francken1, Dries Van den Eynde1, Toon Verwaest2, Job Janssens2, Vera Van
Lancker1
1Royal Belgian Institute for Natural Science (RBINS), Management Unit of the North Sea Mathematical Models
(MUMM), Gulledelle 100, 1200 Brussels, Belgium 2Flanders Hydraulics Research, Berchemlei 115, 2140 Antwerp, Belgium
M. Fettweis (corresponding author) (e-mail: [email protected], Tel.: +32-2-7732132, Fax: +32-2-
7706972),
Abstract
Multi-sensor tripod measurements in the high-turbidity area of the Belgian nearshore zone (southern North Sea)
allowed investigating storm effects on near bed suspended particulate matter (SPM) concentrations. The data
have shown that during or after a storm the SPM concentration increases significantly and that high concentrated
mud suspensions (HCMS) are formed. Under these conditions, about 3 times more mass of SPM was observed
in the water column, as compared to calm weather conditions. The following different sources of fine-grained
sediments, influencing the SPM concentration signal, have been investigated: wind direction and the advection
of water masses; the previous history and occurrence of fluffy layers; freshly deposited mud near the disposal
grounds of dredged material, navigation channels and adjacent areas; and the erosion of medium-consolidated
mud of Holocene age.
Based on erosion behaviour measurements of in-situ samples, the critical erosion shear stresses have been
estimated for different cohesive sediment samples outcropping in the study area. The results have shown that
most of the mud deposits cannot be eroded by tidal currents alone, but higher shear stresses, as induced by
storms with high waves, are needed. Erosion can however occur during storms with high waves. Data suggest
that in order to obtain very high SPM concentrations near the bed, significant amounts of fine-grained sediments
have to be resuspended and/or eroded. The disposal grounds of dredged material, navigation channels and
adjacent areas with freshly deposited mud have been found to be the major source of the fine-grained sediments
during storms. This result is important, as it suggests that dredging and the associated disposal of sediments have
made available fine-grained matter that contributes significantly to the formation of high SPM concentrations
and high concentrated mud suspensions.
1
1 Introduction
Improving our understanding of suspended particulate matter (SPM) variability in nearshore areas is essential for
a more sustainable exploitation of the marine environment, to better understand the human footprint and to
develop marine policies complying with international agreements aiming at sustainable development. Tides,
waves, medium-scale meteorological events, biological processes affect SPM concentration in near shore areas.
Coastal beds are usually composed of heterogeneous particles often consisting of sand- and clay-sized material.
The mutual interaction of cohesive and non-cohesive sediments during erosion influences SPM concentration (Le
Hir, et al, 2007). The SPM concentration is further controlled by the seasonal variability in the supply of fine-
grained sediment, the remote or local availability of fine sediments, advective processes, erosion, deposition, and
human activities (Velegrakis et al., 1997; Bass et al., 2002; Schoelhamer, 2002, Chang et al., 2007). The
deepening of channels and construction of ports increases deposition of fine-grained sediments and has as
consequence an increase of maintenance dredging and a local and instantaneous increase of SPM concentration
in and around the disposal site (Truitt, 1988; Collins, 1990; Van den Eynde and Fettweis, 2006; Wu et al., 2006)
followed by segregation between fine- and coarser-grained sediments on the disposal site (Du Four and Van
Lancker, 2008).
The aim of this paper is to present and discuss SPM concentration variations in a shallow coastal turbidity
maximum area, generated by medium-scale meteorological events as also wave heights. Waves have an
important impact on cohesive sediment transport processes on continental shelves (e.g. Green et al., 1995;
Cacchione et al., 1999; Traykovski et al., 2007; Shi et al., 2008). During these events high concentrated mud
suspensions (HCMS) have been measured, which can be formed by settling of suspended matter or fluidization
of cohesive sediment beds (Maa and Mehta, 1987; van Kessel and Kranenburg, 1998; Li and Mehta, 2000).
Fluidization or liquefaction of mud layers occurs, if the stress caused by wave pressure exceeds the yield stress of
the sediments (Silva Jacinto and Le Hir, 2001). The high concentrated mud suspensions (HCMS) and fluid mud
deposits are easily transported and may result in very high sediment fluxes.
In the vicinity of Zeebrugge (southern North Sea), long-term SPM concentration time series have been
collected using a benthic boundary layer tripod: these form a unique dataset for the investigation of storm
influences on SPM concentration. The measuring locations are situated in a very energetic area that is amongst
the most turbid in the North Sea. The origin of the suspended matter in the southern North Sea is mainly from
the inflow of fine-grained sediments through the Dover Strait, generated by a flood-dominated and wind-induced
residual water transport towards the North Sea (Eisma, 1981; van Alphen, 1990; Lafite et al., 2000). Fettweis et
al (2007) point out that this SPM flux cannot solely be responsible for the often very high SPM concentration in
the area. They suggest that erosion of the nearshore medium consolidated Holocene mud deposits contributes to
the sediment balance, but comment that other sources of fine-grained sediments might exist. These sources could
be related to human activities as most of the nearshore areas in the southern North Sea have a long history of
human impact (e.g. coastal defence works, port construction, dredging and disposal of sediments). Verwaest
(2007) stresses the impact of dredging and disposal activities on SPM concentrations.
The cohesive sediment processes associated with storms in a coastal high turbidity zone are not well
documented; the present paper is therefore an attempt to assess, based on field measurements and numerical
model results, the hydrodynamics, sediment dynamics and bed erosion processes during extreme meteorological
2
events. These data will help to better understand the variability of SPM concentration and its relation to erosion,
resuspension and transport of sediments.
2 Regional settings
The study area is situated in the southern North Sea, more specifically in the Belgian-Dutch nearshore zone. The
depth is generally between 0-20 m below MLLWS (Mean Lowest Low Water Spring), except in the mouth of the
Westerschelde estuary where depths can reach more than 20 m below MLLWS (Fig. 1). The tidal regime is semi-
diurnal, with tidal ranges that diminish towards the northeast. The mean tidal range at Zeebrugge is 4.3 m and 2.8
m at spring and neap tide, respectively. The tidal currents are generally flood-dominant dominant (towards the
northeast), as also the residual water transport. The maximum current velocities are more than 1 m s-1. The winds
are dominantly from the southwest and the highest waves occur during north-western winds (Fettweis and Van
den Eynde, 2003). SPM forms a turbidity maximum between Oostende and the mouth of the Westerschelde.
SPM concentration in the coastal zone varies between minimum 20-70 mg l-1 and maximum 100-1000 mg l-1
during calm meteorological conditions at 3 m above the sea bed; lower values (<10 mg l-1) occur in the offshore
area (Fettweis et al., 2007).
The sea bed consists mainly of Quaternary sandy deposits. Most of them have been deposited during the
Holocene transgression; they are grouped today in sand banks, which form the thickest accumulation of
Quaternary deposits and the most prominent sea bed feature in the area (Le Bot et al. 2005). The nearshore
deposits consist of mainly fine sands with variable mud content. The cohesive sediments (mud, clay) occur
mainly in the eastern nearshore part of the Belgian Continental Shelf and are characterised by a particular
rheological and/or consolidation state. Four different types are distinguished, namely Eocene clay, Holocene
mud (±3000 yr BP), modern mud (<100 yr BP) and freshly deposited mud (Fettweis et al., 2009). Generally, the
freshly deposited mud occurs as thin fluffy layers or locally as gradually soft consolidated thicker packages
(±0.2-1 m). These thicker soft mud layers consisting also of modern mud deposits, are limited to the area around
the disposal place of dredged material near Oostende and the port of Zeebrugge. The Holocene deposits,
extending over most of the foreshore area, consist of medium-consolidated mud with intercalation of thin more
sandy horizons; they are often covered by sand layers (order of cm to dm) or fluffy layers of a few cm thick. In
the offshore swales, the thickness of the Quaternary cover is locally less than 2.5 m; in these areas Eocene
outcrops (clay) are to be expected (Le Bot et al., 2005).
The port of Zeebrugge and its connection to the open sea (Pas van het Zand, Fig. 1) as well as the navigation
channel towards the Westerschelde estuary (Scheur, Fig. 1) are efficient sinks for cohesive sediments. To
conserve the maritime access to the coastal harbours and to the Scheldt estuary dredging is needed. Maintenance
dredging amounts today to about 8.6 million tons of dry matter yearly and capital dredging to 2.8 million tons of
dry matter (averages over 1997-2006), see Lauwaert et al. (2008). About 45% of the maintenance dredging is
carried out in the navigation channels and 55% in the harbours. 93% of the dredging in the harbours is from the
port of Zeebrugge. Half of the matter from the navigation channels is dredged in the access channel towards
Zeebrugge (Pas van het Zand). The dredged matter from the navigation channels consist of 70-85% of mud;
from the harbours this amounts to more than 95%.
3 Materials and Methods
3.1 Instrumentation and measuring sites
Data collection was conducted between April 2005 and February 2007 using a tripod at two sites (Fig. 1). Both
3
sites are characterised by the occurrence of near-bed Holocene medium-consolidated mud, albeit covered with an
ephemeral slightly muddy fine sand layer with a median grain size of about 170 µm. The thickness of the sand
layer increases towards the shore. The water depth is 9 m MLLWS at the MOW1 site and about 5 m MLLWS at
the Blankenberge site (Fig. 1). The tripod measuring system was developed to monitor SPM concentration and
current velocity. Mounted instruments include a SonTek 3 MHz ADP, a SonTek 5 MHz ADV Ocean, a Sea-Bird
SBE37 CT system, two OBS and two SonTek Hydra systems for data storage and batteries (Table 1). Three
periods have been selected for further analysis to assess storm influences (Table 2).
Field calibration of the OBS sensors have been carried out during 5 tidal cycles between April 2005 and May
2006 at site MOW1. A Niskin bottle was closed every 20 minutes, thus resulting in about 40 samples per tidal
cycle. Three sub samples were filtered on board of the vessel from each water sample, using pre-weighted filters
(Whatman GF/C). After filtration, the filters were rinsed once with Milli-Q water (±50 ml) to remove the salt,
and dried and weighted to obtain the SPM concentration. A linear regression between all OBS signals and SPM
concentrations from filtration (about 240) was assumed.
3.2 Calculation of wave- and current induced shear stress from ADV
The high frequency ADV measurements (25 Hz) permit to decompose the velocity in terms of a mean and a
fluctuating part. Several studies report on the possibility to estimate shear stress from the second moment
(turbulence) statistics (Pope et al., 2006; Verney et al., 2007; Andersen et al., 2007). The estimation is based on
the calculation of the turbulent kinetic energy (TKE), which can be obtained from the variance of the velocity
fluctuations. The shear stress has found to be proportional to the TKE through τ = C×TKE, where C=19 was
adopted as proposed by Stapleton and Huntley (1995) and Thompson et al. (2003). This linear relationship will
however fail in the presence of waves. The inertial-dissipation method uses the spectrum of velocity components
and allows to apply a correction for the advection by waves (Trowbridge and Elgar, 2001; Sherwood et al.,
2006). In this case the vertical component was used as it is least contaminated by waves. ADV Ocean data were
discarded when the signal to noise ratio dropped below 15 dB and the correlation coefficient was lower than 0.8.
The data were then transformed into a power density spectrum using a Fast Fourier Transform with a Hanning
window and the spectral density, Eww, was normalized such that the integral over the spectrum yielded the
variance over the burst. Afterwards it was transformed to Eww ω5/3 (2π)-1, with ω the radial frequency, as the
turbulent dissipation, ε, scales with Eww ω5/3 (2π)-1 in the inertial subrange, defined as the range between 1 and 2.5
Hz (Trowbridge and Elgar, 2001). By calculating the mean over this range an estimate of Eww ω5/3 (2π)-1 was
obtained and ε calculated using 3 2
2 35 3 1255wwE Uε ω α⎡ ⎤⎛ ⎞= ⎜ ⎟⎢ ⎥⎝ ⎠⎣ ⎦
where α is the empirical Kolmogorov constant and U is the burst averaged velocity. To correct for the presence
of waves the model uses a function I to correct ε, see Trowbridge and Elgar (2001) and Sherwood et al. (2006).
The shear stress at elevation z can then be obtained using τ = ρ (ε κ z)2/3, where κ is the Von Kármán constant
and ρ the water density.
3.3 Erosion behaviour measurements
Erodibility measurements have been performed on mud samples taken at different locations near the navigation
channels Pas van het Zand and Scheur and near the harbour of Oostende (Fig. 1). Boxcores were subsampled
using cylindrical perspex tubes (diameter 13.5 cm) allowing to retrieve relatively undisturbed mud samples of
4
the first 40 cm (on average) of the sea bed. Using a gamma-ray densitometer, bulk density profiles of these
samples were determined in a non-destructive way. The measurements were performed at the University of
Stuttgart using the SETEG-flume (Kern et al., 1999; Witt and Westrich, 2003): a straight pressure duct with
rectangular cross section. The top of the perspex tubes can be attached to a circular hole in the bottom of the
flume, and by pushing the sediment upwards until its surface is level with the bottom surface of the flume the
erosion behaviour of the top layer of the sediment can be investigated. After erosion of the top layer, the
sediment can be pushed further upwards and the underlying layers can be studied. The critical shear stress for
erosion is determined by visual observation of the onset of erosion by a gradual increase of the discharge. Shear
stresses are calculated from the measured discharge via the Darcy-Weisbach equation, using the Colebrook
formula to determine the roughness coefficient (Streeter, 1996). In addition, the flume is equipped with the so-
called SEDCIA-system, enabling to measure the erosion rate. This system consists of a camera which observes
the time-dependent shift of a series of parallel laser lines projected under a certain inclination angle on the
sediment’s surface, from which the sediment’s volume change –and hence also the mass change, since the bulk
density is known– can be calculated in function of time.
3.4 Hydrodynamic numerical model
The currents, surface elevation and turbulent kinetic energy have been modelled using an implementation of the
COHERENS hydrodynamic model to the Belgian Continental Shelf, termed hereafter OPTOS-BCS. The 3D
model solves the continuity and momentum equations on a staggered sigma coordinate grid with an explicit
mode-splitting treatment of the barotropic and baroclinic modes. A description of the COHERENS model can be
found in Luyten et al. (1999). OPTOS-BCS covers an area between 51°N and 51.92°N in latitude and between
2.08°E and 4.2°E in longitude. The horizontal resolution is 0.24’ (longitude) and 0.14’ (latitude), corresponding
both to a grid size of about 250 m. Boundary conditions of water elevation and depth-averaged currents for this
model have been provided by the operational models OPTOS-NOS (covering the North Sea) and OPTOS-CSM
(covering the North-West European Continental Shelf Model) of the Management Unit of North Sea
Mathematical Models (see www.mumm.ac.be). Pison and Ozer (2003) have described the validation of the
current velocities of OPTOS-BCS using ADCP measurements. The bottom shear stress for currents alone is
calculated using the calculated current velocity in the lowest layer of the model and using a bottom roughness of
0.84 cm. The bottom stress under the combined effect of currents and waves is calculated in OPTOS-BCS with
Bijker’s formulae of 1966 (Koutitas, 1988). The currents are from the hydrodynamic model, whereas the wave
data are from wave buoy measurements (Coastal Service of the Ministry of the Flemish Community).
4 Results
4.1 Tripod data
The data of SPM concentration, water depth, ADV current velocity and bottom shear stress collected at MOW1
in autumn 2005 are presented in Fig 2. Spring tide occurred around December 1 (day 10), neap tide around
November 26 (day 5). The measuring period was characterised by 2 days of calm weather followed by a WNW
storm with wind velocities of more than 16 m s-1 (7 Bf) and significant wave heights of up to 3.5 m in the coastal
zone. The water was pushed up against the coast and low water levels raised with almost 2 m. The SPM
concentrations follow a quarter-diurnal (ebb-flood) signal. Although the currents are flood dominated, no
significant difference in magnitude between ebb and flood SPM concentration peak occurs. A significant
increase in SPM concentration occurred almost immediately after the beginning of the WNW storm (day 3-4).
5
The SPM concentration measured by OBS1 (0.3 meter above bottom, mab) increases to 1-3 g l-1, whereas at 2
mab (OBS2) the SPM concentrations remain less than 0.5 g l-1. The sudden increase in SPM concentration at the
onset of the storm was induced by the exceptional meteorological conditions and partly because the storm
occurred after a calm period and around neap tide. Neap tidal conditions and calm periods favour the deposition
of fluffy layers, as has been observed from bottom samples and from model simulations (Fettweis and Van den
Eynde, 2003). After the WNW storm the SPM concentration decreased in magnitude, however both OBS’s still
measured very high peak concentrations (OBS1: up to 3 g l-1, OBS2: up to 1 g l-1) until the end of the
deployment (1 week after the storm). The high minima in SPM concentration measured by the OBS1 (0.5-1 g l-
1), indicates that a HCMS or fluid mud layer was formed. It is only days after the storm that the SPM
concentrations, measured by both OBS’s, show again similar minima and that the near bed high SPM
concentrations have disappeared.
77 days of data were collected at the Blankenberge site between November 8, 2006 and February 7, 2007.
The data from 7-18 November 2006 and 24 December 2006 – 8 January 2007 are presented in Fig. 3-4. Remark
that no bottom shear stresses are reported for the second period (Fig. 4) as the quality of the ADV data was
insufficient. During both periods different storms have occurred. On 12-13 November, a NW storm generated
significant wave heights of about 2.7 m. The second half of December was characterized by low wind speeds
from mainly W-SW. On December 31, wave heights of nearly 3 m were registered (Fig. 4). The beginning of
January 2007 was characterised by storms from mainly a NE direction. The highest SPM concentrations during
the November 12 storm have been registered only about one day after the storm by OBS1 and about two days
after by OBS2 (Fig. 3). The OBS1 data are characterised by very high minima in SPM concentrations (>0.8 g l-
1). The OBS2 has measured only during a short period after the storm an increase in SPM concentration. This
indicates that vertical mixing was limited. Similar data have been collected during the storm of December 31.
The period before the storm was characterised by low differences in SPM concentration between OBS1 and
OBS2. Stratification in SPM concentration has been observed only at the end of the ebbing tide and during slack
water and is due to settling of suspended particles. The increase in SPM concentration after the December 31
storm is – similar as observed during the November 12 storm - only detected one day after the storm in both
OBS. This increase in SPM concentration occurred during ebb indicating that the suspended matter has mainly
been transported from the NE, i.e. from the mouth of the Westerschelde estuary and in the direction of the wind-
driven and the ebb current. Local resuspension of mud layers was at that time of minor importance.
A few conclusions can be drawn from the three measurement periods:
1. There is a dominant quarter-diurnal (ebb-flood) signal in the SPM concentration time-series. The spring-
neap tidal signal can be identified clearly during calm meteorological conditions;
2. Considerable variations in SPM concentrations exist during a tidal cycle: maximum concentrations were
sometimes up to 50 times higher than the minimum concentrations;
3. The very high SPM concentrations measured near the bed are related to storm periods; our data suggest that
HCMS occur near the bottom in the coastal turbidity maximum of the Belgian-Dutch nearshore zone; and
4. Wind-driven advection can have a significant influence on SPM concentration.
4.2 Erosion behaviour measurements
In Fig. 5 the measured depth profile of the bulk density, ρ, and the critical shear stress for erosion, τce, of two
cores with Holocene mud and one with freshly deposited mud are shown (location is indicated in Fig. 1). The
6
profiles show that a high variability of τce can be observed in muddy sediments. The cores containing Holocene
mud are both characterised by a 5-10cm thick surface layer of freshly deposited mud above soft to medium-
consolidated mud. The Holocene mud deposits are characterised by intercalations of thin sandy layers. The top
layer has a τce ranging from 0.5 to 2.5 Pa. The Holocene mud layers are very strong, with a τce up to 13 Pa, while
the intermediate sandy layers exhibit much lower values (~1 Pa). The core with freshly deposited mud has been
collected in the navigation channel. The upper 45 cm are not consolidated and have a τce of 1-4 Pa. A total of 35
sediment samples have been taken in the nearshore zone from which the following observations can be made:
• Generally, the top layer has low values for both τce (0.5 to 1 Pa) and ρ (fluffy layers);
• The sediment stability of freshly deposited mud strongly increases in the first few cm. A value of ~4 Pa is
reached at a depth of a few cm;
• Only in some cores a positive correlation between τce and ρ exist;
• The presence of shells has a negative effect on the stability of muddy sediments; and
• Sandy fractions always exhibit low to medium stability (maximum τce: ~4 Pa).
4.3 Hydrodynamic model results
The numerical model has been used to simulate the bottom shear stress during autumn 2005 (22 November – 5
December). In Fig. 2 the model results are compared to the shear stresses derived from ADV measurements. The
maximum bottom shear stress is shown in Fig. 6 with and without wind and wave effects. The figure shows that
the bottom shear stress increases significantly to values of more than 12 Pa. Without meteorological effects the
maxima, occurring during spring tide in the channel towards the Westerschelde (Scheur), reaches about 2.5 Pa.
The difference between model results and shear stress results from ADV can be ascribed to inaccuracies in
calculating near-bed shear stress from ADV data and to the fact that the bottom roughness in model is not well
known and remains a calibration parameter.
5 Discussion
The formation of HCMS in wave-dominated areas is well documented in the scientific literature (de Wit and
Kranenburg, 1997; Winterwerp, 1999; Li and Mehta, 2000). The occurrence of fluid mud or HCMS on many
continental shelves is associated with wave or current-driven sediment gravity flows off high-load rivers (Wright
and Friedrichs, 2006). However, the origin of the suspended matter in the southern North Sea and in the Belgian-
Dutch nearshore zone has been mainly ascribed to the inflow of fine-grained sediments through the Dover Strait
(Gerritsen et al., 2000), as no high-load rivers exist in the area (Fettweis et al., 2007). This SPM flux amounts to
about 34×106 t per year, from which about 50% is transported along the continental side (i.e. France-Belgian-
Dutch nearshore area), see Fettweis et al. (2007). Due to the high tidal energy permanent layers of freshly
deposited to very soft consolidated cohesive sediments occur in the Belgian-Dutch nearshore area only in some
protected hydrodynamic environment such as ports, navigation channels and disposal grounds of dredged
material. However, thin fluffy layers, temporarily present, do occur over large areas. The fluctuation of SPM
concentration with time is complex and it is not always straightforward to identify the origin of some of the
variations. The relation between tidal amplitude and tide-averaged SPM concentration for the MOW1 data is
shown in Fig. 7. The low correlation points to the fact that the neap-spring tidal signal is overprinted by other
processes, such as high wave conditions and wind-induced long- and cross-shore advection.
Using SeaWiFS images the average mass of SPM over the period 1997-2004 in the turbidity maximum area
has been calculated as about 1×106 t. Variations of the order of 30% occur between spring and neap tides, as also
7
of the order of 40-60% in-between seasons. Based on the tripod measurements, the SPM mass during storm
conditions has been estimated as 3-5×106 t in the same area. An important amount of fine-grained matter has
thus to be resuspended, eroded or transported during a storm. Below, some points are discussed in more detail to
better identify the possible sources of fine-grained sediments and the processes that result in the significant
increase in SPM concentration during storms.
5.1 SPM transport and wind-driven advection
The effect of winds on SPM concentration is variable and depends also on the wind direction and the availability
of muddy sediments. Along- or cross-shore advection, enhanced during winds from the SW/NE or NW/SE,
respectively, transports water masses with low SPM concentration and higher salinity to the measurement
location. During periods with high salinity variations during a tide, a negative correlation between salinity and
SPM concentration exists, see Fig. 8. High salinities and low(er) SPM concentrations are associated with
flooding and low salinities and high(er) SPM concentrations with ebbing tide. The data show that during the first
3 weeks of January 2007, SW winds prevailed resulting in advection of high salinity and low turbid water from
the English Channel towards the southern North Sea. At the end of January, the wind direction changed towards
S-SE and low salinity, high turbidity water originating from the Schelde estuary dominated the signal. Our data
show that high SPM concentrations are often more closely related to advection (Velegrakis et al., 1997; Blewett
and Huntley, 1998) rather than instantaneous bed shear stress (Stanev et al., 2009). This confirms the idea that
the Belgian coastal area can be seen as a congestion in the residual SPM transport of the southern North Sea
rather than an important source of sediments (Fettweis and Van den Eynde, 2003).
5.2 Holocene mud as SPM source
The largest reservoir of fine-grained sediments in the nearshore area consists of the medium-consolidated
Holocene mud (bulk density >1500 kg m-3). The area where these mud deposits occur in the first meter of the
seabed is ~744 km² (Fettweis et al, 2009). Erosion behaviour measurements (see above) confirm that
consolidated cohesive bed layers are difficult to erode by only fluid-transmitted stress. The maximum bottom
shear stress during calm periods and during spring tide is about 4 Pa (see Fig. 2-4 and 6). The sediments of the
mud fields can thus not be eroded under calm meteorological conditions. Near bed shear stresses derived from
the ADV data amount up to 40 Pa during the storm of November 2005 at MOW1 (Fig 2) and of December 2006
(Fig. 3). These values indicate that erosion of Holocene mud by fluid-transmitted shear stress can occur and that
SPM could have been released into the water column from the Holocene mud fields under storm conditions.
Using the Ariathurai-Partheniades formulation (Ariathurai, 1974) for erosion of cohesive sediments, a mean bed
shear stress of 12 Pa, a critical erosion shear stress of 10 Pa and an erosion rate of 0.5 g m-2 s-1 the mass of
Holocene mud that can be eroded amounts to 864 t km-2 day-1. If all the Holocene mud would outcrop then about
0.6×106 t per day could have been eroded by horizontal tidally- and wave-induced shear stress during the storm.
This is a maximum estimate as large parts of the Holocene mud fields are probably covered by thin layers of
sandy sediments. Most probably, erosion of Holocene mud by entrainment represents only a minor source of
suspended sediments during storms.
Still, in literature, it is well documented that a cohesive mud bed may be eroded and fluidised by waves (Maa
and Mehta, 1987; De Wit and Kranenburg, 1997; Li and Mehta, 2000; Silva-Jacinto and Le Hir, 2001). The
effect of water waves on a cohesive, deformable bed can be described by a pressure wave, inducing normal and
shear stresses in the bed. These stresses modify the strength of the bed and thus also the erodibility of the
8
sediments. Silva Jacinto and Le Hir (2001) explain the observed liquefaction and failure of consolidated mud
layers along sandy layers by waves due to pumping effects. Mud pebbles are an indication of this type of
erosion, they have been observed regularly in the area of investigation (Fettweis et al., 2009). However, due to
insufficient data availability, the mass of mud pebbles or of suspended matter originating from wave induced bed
failure cannot be quantified yet.
5.3 Mixed sediments as SPM source
Other important erosion mechanisms are due to the mutual interaction of cohesive and non-cohesive sediments
(Le Hir et al., 2007): one can distinguish between sand grains moving on a cohesive substrate and erosion of
mixed sediments. Thin sand layers on top of Holocene mud layers and mixed sediments have often been
observed in the turbidity maximum area. Williamson and Torfs (1996) were the first to show that the addition of
mud to a sandy bed increases the sediment shear strength. Van Ledden et al. (2004) argue that a transition in
erosion behaviour can be expected when the bed changes between cohesive and non-cohesive properties, and
when the network structure is formed by another sediment fraction. Mixed sediments could therefore act as a
reservoir for fine-grained material that is only resuspendable under more extreme meteorological conditions.
However, on the basis of existing data, we cannot quantify the amount of fine sediments that is released as SPM
due to the erosion of mixed sediments.
5.4 Freshly deposited mud as SPM source
The movement of recently deposited sediment in the wave boundary layer exposes fine-grained sediment to
resuspension when shear stresses become sufficiently high. These resuspended sediments are then transported
further by waves and currents. The numerical model results indicate that, under normal conditions, the bed shear
stress in the navigation channels and at location MOW1 or Blankenberge is lower than 4 Pa (Fig. 6). The critical
erosion shear stresses of in-situ samples in the navigation channels and around Zeebrugge and the disposal
ground of Oostende are, below the fluffy surface layer, generally higher than 4 Pa. The deposits of fresh mud
below the fluffy layer in these areas forms thus a reservoir of SPM that will only be resuspended during periods
with high shear stresses e.g. caused by storms. The fluffy surface layer having thicknesses of a few centimetres
and a τce of <1 up to ~4 Pa, will be fully resuspended during periods with higher stresses (storms, spring tides).
Generation of fluffy layers can occur during periods with low stresses (neap tides).
The total surface of the area where freshly deposited mud is found is not precisely known. The surface of the
navigation channels, where mud is dredged, equals to ±15 km². Mud with similar erosion behaviour has also
been sampled near the disposal ground of dredged material near Oostende and around Zeebrugge. Based on bed
samples, the surface of these deposits has been estimated as 30 km². The bulk density of these sediments
amounts to 1200-1400 kg/m³. If we assume a thickness of 20 cm very soft mud in these areas, then the total mass
of mud available for resuspension equals 1.8-3.6×106 t. This is of the same order of magnitude as what has been
estimated to be in suspension during storms (see above). The MOW1 and the Blankenberge site are both situated
in the vicinity of the navigation channel towards Zeebrugge (Pas van het Zand) and the Westerschelde (Scheur).
The data suggest that an important part of the HCMS, measured at both sites, could have been resuspended from
the very soft mud deposits in the navigation channels and adjacent areas.
6 Conclusion
Measurements have been collected at two locations in the vicinity of the port of Zeebrugge and its navigation
channels. Between autumn 2005 and winter 2007, three stormy periods have been selected with similar wave
9
conditions. The data have shown that during or after a storm, the SPM concentration increases significantly and
that HCMS have been formed. SPM concentration is clearly related to high waves and winds. The wind direction
and the advection of water masses, the previous history and the availability of fine-grained sediments in fluffy
layers, the very soft mud deposits around navigation channels, and the erosion of medium-consolidated mud of
Holocene age influence the SPM signal. The data suggest that for the generation of very high SPM
concentrations near the bed, significant amounts of fine-grained sediments have to be resuspended and/or
eroded. The navigation channels and other areas with soft mud have been found to be the major source of the
fine-grained sediments during storms. This result is important as it suggest that the deepening of the navigation
channels has made available fine-grained matter that contributes significantly to the formation of high
concentration mud suspensions. This suggests that HCMS were probably less frequent in the past when
anthropogenic activities where limited.
Acknowledgement
This study was partly funded by the by the Belgian Science Policy within the framework of the QUEST4D
project (SD/NS/06A) and by the Maritime Access Division of the Ministry of the Flemish Community in the
framework of the MOMO project. G. Dumon (Coastal Service, Ministry of the Flemish Community) made
available wave measurement data. We want to acknowledge the crew of the RV Belgica, Zeearend and Zeehond
for their skilful mooring and recuperation of the tripod. The assistance of marine divers during recuperation of
the tripod on February 7 2007 is also acknowledged. The measurements would not have been possible without
technical assistance of A. Pollentier, J.-P. De Blauwe and J. Backers (measuring service of MUMM, Oostende).
We also are grateful to G. Voulgaris, for helping us out with the inertial-dissipation method.
7 References
Andersen, T. J., Fredsoe, J., Pejrup, M. 2007. In-situ estimation of erosion and deposition thresholds by Acoustic
Doppler Velocimeter (ADV). Estuarine, Coastal and Shelf Science, 75, 327-336. doi:
10.1016/j.ecss.2007.04.039
Ariathurai, C. R. 1974. A finite element model for sediment transport in estuaries. University of California,
Davis.
Bass, S.J., Aldridge, J.N., McCave, I.N., Vincent, C.E. 2002. Phase relationships between fine sediment
suspensions and tidal currents in coastal seas. Journal of Geophysical Research 107(C10), 1-14.
Blewett, J., Huntley, D.A. 1998. Measurement of suspended sediment transport processes in shallow water off
the Holderness Coast, UK. Marine Pollution Bulletin 37, 134-143.
Cacchione, D.A., Wiberg, P.L., Lynch, J.F., Irish, J.D., Traykovski, P. 1999. Estimates of suspended-sediment
flux and bedform activity on the inner portion of the Eel continental shelf. Marine Geology, 154, 83-97.
Collins, M. 1990. The behaviour of cohesive and non-cohesive sediments. In: Alzieu, C., Galenne, B. (Eds.),
Proceedings of the international seminar on the environmental aspects of dredging activities. Port Autonome
de Nantes Saint-Nazaire, Nantes, 15–32.
Chang, T.S., Flemming, B., Bartholomä, A. 2007. Distinction between sortable silts and aggregated particles in
muddy intertidal sediments of the East Frisian Wadden Sea, southern North Sea. Sedimentary Geology, 202,
453-463. doi:10.1016/j.sedgeo.2007.03.009
de Wit, P.J., Kranenburg, C. 1997. The wave-induced liquefaction of cohesive sediment beds. Estuarine, Coastal
and Shelf Science, 45, 261-271.
10
Du Four, I., Van Lancker, V. 2008. Changes of sedimentological patterns and morphological features due to the
disposal of dredge spoil and the regeneration after cessation of the disposal activities. Marine Geology, 25,
15-29. doi: 10.1016/j.margeo.2008.04.011
Eisma, D. 1981. Supply and deposition of suspended matter in the North Sea. In: Holocene Marine
Sedimentation in the North Sea Basin. (Nio, D.D., Shuttenhelm, R.T.E., van Weering, T.C.E., eds.). IAS
Special Publication, 5, Blackwell Scientific, 415-428.
Fettweis, M., Van den Eynde, D. 2003. The mud deposits and the high turbidity in the Belgian-Dutch coastal
zone, Southern bight of the North Sea. Continental Shelf Research, 23, 669-691. doi: 10.1016/S0278-
4343(03)00027-X
Fettweis, M., Nechad, B., Van den Eynde, D. 2007. An estimate of the suspended particulate matter (SPM)
transport in the southern North Sea using SeaWiFS images, in-situ measurements and numerical model
results. Continental Shelf Research, 27, 1568-1583. doi: 10.1016/j.csr.2007.01.017
Fettweis, M., Houziaux, J.-S., Du Four, I., Van Lancker, V., Baeteman, C., Mathys, M., Van den Eynde, D.,
Francken, F., Wartel, S. 2009. Long-term influence of maritime access works on the distribution of cohesive
sediments: analysis of historical and recent data from the Belgian nearshore area (southern North Sea). Geo-
Marine Letters, 29(5), 321-330. doi: 10.1007/s00367-009-0161-7
Gerritsen, H., Boon, J.G., van der Kaaij, T., Vos, R.J. 2001. Integrated modelling of suspended matter in the
North Sea. Estuarine, Coastal and Shelf Science, 53, 581-594. doi: 10.1006/ecss.2000.0633
Green, M.O., Vincent, C.E., McCave, I.N., Dickson, R.R., Rees, J.M., Pearson, N.D. 1995. Storm sediment
transport: observations from the British North Sea shelf. Continental Shelf Research, 15, 889-912.
Günther, H., Rosenthal, W., 1985. The hybrid parametrical (HYPAS) wave model. In: Ocean wave modelling,
Swamp group. Plenum Press, New York, 211-214.
Kern, U., Haag, I., Schürlein, V. Holzwarth, M., Westrich, B. 1999. Ein Strömungskanal zur Ermittlung der
tiefenabhängigen Erosionsstablität von Gewässersedimenten. Wasserwirtschaft, DWA, 89(2), 72-77. Koutitas, C.G., 1988. Mathematical Models in Coastal Engineering. Pentech Press, London, UK, 156pp.
Lafite, R., Shimwell, S., Grochowski, N., Dupont, J.-P., Nash, L., Salomon, J.-C., Cabioch, L., Collins, M., Gao,
S. 2000. Suspended particulate matter fluxes through the Strait of Dover, English Channel: observations and
modelling. Oceanologica Acta, 23, 687-700.
Lauwaert, B., Bekaert, K., Berteloot, M., De Brauwer, D., Fettweis, M., Hillewaert, H., Hoffman, S., Hostens,
K., Mergaert, K., Moulaert, I., Parmentier, K., Vanhoey, G., Verstraeten, J. 2008. Synthesis report on the
effects of dredged material disposal on the marine environment. BMM, ILVO, AK & AMT rapport,
BL/2008/01, 109pp.
Le Bot, S., Van Lancker, V., Deleu, S., de Batist, M., Henriet, J.-P., Haegeman, W. 2005. Geological and
geotechnical properties of Eocene and Quarternary deposits on the Belgian continental shelf: synthesis in the
context of offshore wind farming. Netherlands Journal of Geosiences, 84, 147-160.
Le Hir, P., Monbet, Y., Orvain, F. 2007. Sediment erodability in sediment transport modelling: Can we account
for biota effects? Continental Shelf Research, 27, 1116-1142. doi: 10.1016/j.csr.2005.11.016
Li, Y., Mehta, A.J. 2000. Fluid mud in the wave-dominated environment revisited. In: Coastal and Estuarine
Fine Sediment Dynamics. (McAnally, W.H., Mehta, A.J., eds.), Proceedings in Marine Science, 3, 79-93.
11
Luyten, P. J., Jones, J. E., Proctor, R., Tabor, A., Tett, P., Wild-Allen, K. 1999. COHERENS, a coupled
hydrodynamical-ecological model for regional and shelf seas: User Documentation. MUMM report, Brussels,
Belgium. 911pp. http://www.mumm.ac.be/coherens.
Maa, P.-Y., Mehta, A. J. 1987. Mud erosion by waves: a laboratory study. Continental Shelf Research, 7, 1269-
1284.
Pison, V., Ozer, J. 2003. Operational products and services for the Belgian coastal waters. In: Building the
European capacity in operational modeling, Proc. 3rd Int. Conf. on EuroGOOS (Dahlin, H., Flemming, N.C.,
Nittis, K., Petersson, S.E., eds.). Elsevier Oceanography Series 69, 503-509.
Pope, N. D., Widdows, J., Brinsley, M. D. 2006. Estimation of bed shear stress using the turbulent kinetic energy
approach - A comparison of annular flume and field data. Continental Shelf Research, 26, 959-970. doi:
10.1016/j.csr.2006.02.010
Sherwood, C., Lacy, J., Voulgaris, G. 2006. Shear velocity estimates on the inner shelf off Grays Harbor,
Washington, USA. Continental Shelf Research, 26, 1995-2018.
Shi, J.Z., Gu, W.-J., Wang, D.-Z. 2008. Wind wave-forced fine sediment erosion during slack water periods in
Hangzhou Bay, China. Environmental Geology, 55, 629-638. doi: 10.1007/s00254-007-1013-2
Silva-Jacinto, R. Le Hir, P. 2001. Response of stratified muddy beds to water waves. In: Coastal and Estuarine
Fine Sediment Processes (McAnally, W.H., Mehta, A.J., eds.). Proceedings in Marine Science, 3, 95-108.
Stanev, E.V., Dobrynin, M., Pleskachevsky, A., Grayek, S., Günther. H. 2009. Bed shear stress in the southern
North Sea as an important driver for suspended sediment dynamics. Ocean Dynamics, 59, 183-194. doi:
10.1007/s10236-008-0159-0
Stapleton, R.L., Huntley, D.A., 1995. Seabed stress determination using the inertia dissipation method and the
turbulent kinetic energy method. Earth Surface Processes and Landforms 20, 807-815.
Streeter, V.L. 1996. Fluid Mechanics. McGraw-Hill Book Company, New York, 705pp.
Thompson, C.E.L., Amos, C.L., Jones, T.E.R., Chaplin, J. 2003. The manifestation of fluid-transmitted bed shear
stress in a smooth annular flume - A comparison of methods. Journal of Coastal Research, 19, 1094-1103.
Trowbridge, J., Elgar, S. 2001. Turbulence Measurements in the Surf Zone. Journal of Physical Oceanography,
31(8), 2403-2417.
Truitt, C.L. 1988. Dredged material behaviour during open water disposal. Journal of Coastal Research, 4, 389–
397.
Traykovski, P., Wiberg, P.L., Geyer, W.R. 2007. Observations and modeling of wave-supported sediment
gravity flows on the Po prodelta and comparison to prior observations from the Eel shelf. Continental Shelf
Research, 27, 375-399. doi: 10.1016/j.csr.2005.07.008
Van Alphen, J.S.L.J. 1990. A mud balance for Belgian-Dutch coastal waters between 1969 and 1986.
Netherlands Journal of Sea Research, 25, 19-30.
Van den Eynde, D., 1992. MU-WAVE: An operational wave forecasting system for the Belgian coast. 3rd Int
Workshop on Wave Hindcasting and Forecasting, Montréal, Canada, 313-324.
Van den Eynde, D., Fettweis, M. 2006. Modelling of fine-grained sediment transport and dredging material
dumpings at the Belgian continental shelf. Journal of Coastal Research, SI39, 1564-1569.
van Ledden, M., van Kesteren, W.G.M., Winterwerp, J. 2004. A conceptual framework for the erosion behaviour
of sand-mud mixtures. Continental Shelf Research, 24, 1-11. doi: 10.1016/j.csr.2003.09.002
12
van Kessel, T., Kranenburg, C. 1998. Wave-induced liquefaction and flow of subaqueous mud layers. Coastal
Engineering, 34, 109-127.
Verney, R., Deloffre, J., Brun-Cottan, J. C., Lafite, R. 2007. The effect of wave-induced turbulence on intertidal
mudflats: Impact of boat traffic and wind. Continental Shelf Research, 27, 594-612. doi:
10.1016/j.csr.2006.10.005
Verwaest, T. 2007. Using simple semi-empirical models for integrated assessment of scenarios for a navigation
channel. The case of the port of Ostend, Belgium. Coastal Sediments 2007, New Orleans, USA.
Velegrakis, A.F., Gao, S., Lafite, R., Dupont, J.P., Huault, M.F., Nash, L.A., Collins, M.B. 1997. Resuspension
and advection processes affecting suspended particulate matter concentrations in the central English Channel.
Journal of Sea Research, 38, 17-34.
Williamson, H., Torfs, H. 1996. Erosion of mud/sand mixtures. Coastal Engineering, 29, 1-25.
Winterwerp, J. 1999. On the dynamics of high-concentrated mud suspensions. PhD thesis, TU Delft, The
Netherlands, 172pp.
Witt, O., Westrich, B. 2003. Quantification of erosion rates for undisturbed contaminated cohesive sediment
cores by image analysis. Hydrobiologia, 494, 271-276. Wright, L.D., Friedrichs, C.T. 2006. Gravity-driven sediment transport on continental shelves: A status report.
Continental Shelf Research, 26, 2092-2107. doi: 10.1016/j.csr.2006.07.008
Wu, J., Liu, J.T., Shen, H., Zhang, S. 2006. Dispersion of disposed dredged slurry in the meso-tidal Changjiang
(Yangtze River) Estuary. Estuarine, Coastal and Shelf Science, 70, 663-672. doi: 10.1016/j.ecss.2006.07.013
13
Table 1: Oceanographic instruments mounted on the tripod and distance in m above bed (mab). The saturation
concentration of the OBS’s was 3.3 g l-1.
Survey Nr ADV ADP OBS1 OBS2
2005/29 0.3 2.2 0.3 1.9 0.8
2006/23 0.2 2.3 0.2 2.2 0.8
2006/27 0.2 2.3 0.2 2.2 0.8
Table 2: Tripod deployments at MOW1 and Blankenberge.
Survey Nr Location Start - End date (dd/mm/yyyy hh:mm) Duration (days)
2005/29 MOW1 22/11/2005 08:11 - 05/12/2005 09:03 13.04
2006/23 Blankenberge 08/11/2006 14:29 – 27/11/2006 09:04 18.83
2006/27 Blankenberge 18/12/2006 10:44 - 07/02/2007 13:17 50.11
14
Figure 1: (A) Yearly and depth-averaged SPM concentration (mg/l) in the southern North Sea, derived from 370
SeaWiFS images (1997-2002), see Fettweis et al (2007a). (B) Bathymetry (m below mean sea level) of the
Belgian coastal area. Indicated are the tripod measuring stations (black dots: MOW1, Blankenberge), the
navigation channel (Pas van het Zand and Scheur) and the location of the box-core samples for erosion
behaviour measurements (white dots). Coordinates are in latitude (°N) and longitude (°E).
15
0 2 4 6 8 10 12 146
11
16D
epth
bel
ow s
urfa
ce (
m)
0 2 4 6 8 10 12 140
1
2
3
4
Wav
e he
ight
(m
)
0 2 4 6 8 10 12 140
0.5
1
Vel
ocity
(m
/s)
0 2 4 6 8 10 12 140
10
20
30
She
ar s
tres
s (P
a)
0 2 4 6 8 10 12 140
1000
2000
3000
Time (days)
SP
M c
onc
(mg/
l)
depth wave height
SPM1
SPM2
measurements model results
Figure 2: MOW1 site, tripod measurements of 22 November – 5 December 2005 (survey 2005/29). From up to
down: depth below water surface (m) and significant wave heights; ADV current velocity (m/s); shear stress (Pa)
derived from the ADV and from the hydrodynamic model and SPM concentration at 0.3 mab (SPM1) and 1.9
mab (SPM2).
16
0 2 4 6 8 10 12 144
9
14D
epth
bel
ow s
urfa
ce (
m)
0 2 4 6 8 10 12 140
1
2
3
4
Wav
e he
ight
(m
)
0 2 4 6 8 10 12 140
0.5
1
Vel
ocity
(m
/s)
0 2 4 6 8 10 12 140
20
40
She
ar s
tres
s (P
a)
0 2 4 6 8 10 12 140
2000
4000
Time (days)
SP
M c
onc
(mg/
l)
depth wave height
SPM1 SPM2
Figure 3: Blankenberge site, tripod measurements of 7-20 November 2006 (survey 2006/23). From up to down:
depth below water surface (m) and significant wave heights; ADV current velocity (m/s); shear stress (Pa)
derived from the ADV; and SPM concentration at 0.2 mab (SPM1) and 2.2 mab (SPM2).
17
0 5 10 154
9
14D
epth
bel
ow s
urfa
ce (
m)
0 5 10 150
0.5
1
Vel
ocity
(m
/s)
0 5 10 150
2000
4000
Time (days)
SP
M c
onc
(mg/
l)
0 5 10 150
1
2
3
4
Wav
e he
ight
(m
)
depth wave height
SPM1 SPM2
Figure 4: Blankenberge site, tripod measurements of 27 December 2006 – 10 January 2007 (survey 2006/27).
From up to down: depth below water surface (m) and significant wave heights; ADV current velocity (m/s) and
mean particle size (µm); and SPM concentration at 0.2 mab (SPM1) and 2.2 mab (SPM2).
18
0
10
20
30
40
50
0 2 4 6 8 10 12τce (Pa)
dept
h be
low
bed
[cm
]
1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,
bulk density [g cm-³]
14
8
Tau-crit
bulk density
0
10
20
30
40
50
0 2 4 6 8 10 12 14τce (Pa)
dept
h be
low
bed
(cm
)
1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8bulk density (g cm-³)
tau-critbulk density
0
10
20
30
40
50
0 2 4 6 8 10 12 14τce (Pa)
dept
h be
low
bed
(cm
)
1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8
bulk density (g cm-³)
Tau-critBulk density
A B C
Figure 5: Depth profile of critical erosion shear stress (τce) and bulk density of box cores. (A) 7 cm of soft mud
above 40 cm of medium-consolidated mud (Holocene) with intercalations of sand, muddy sand and shell layers
(location: near Zeebrugge). (B) 5 cm of very soft mud with fine sand above 40 cm of alternating medium-
consolidated mud and thin muddy sand layers with shells (Holocene). (C) 40 cm of freshly deposited mud from
the navigation channel (‘Pas van het Zand) on top of medium-consolidated mud (Holocene). The sandy and
shelly layers have a lower τce and a higher bulk density.
19
Figure 6: Hydrodynamic model results of maximum bottom shear stress (Pa) during a spring tide without wind
(A) and during the November 2005 storm (B). Coordinates are in latitude (°N) and longitude (°E).
20
Figure 7: Tidally-averaged SPM concentration (OBS2, 2 mab) as a function of tidal amplitude at the
Blankenberge site (November 2006 – February 2007).
21
Figure 8: SPM concentration from OBS2 (2 mab) as a function of salinity. The figures show the effect of
alongshore advection of high saline and low turbid water during January 2007 at the Blankenberge site. (A)
Negative correlation between salinity and SPM concentration during prevailing SW winds (5-7 January). (B)
Two successive periods are shown; the first is characterised by high salinity and low SPM concentration (16-22
January) and the second by low salinity and high SPM concentration (22-31 January). The sudden decrease in
salinity was induced by a change in wind direction from SW to SE.
22