BARDEX (Barrier Dynamics Experiment): Taking the Beach - e-Geo

Journal of Coastal Research
SI 56
158 - 162
ICS2009 (Proceedings)
Portugal
ISSN 0749-0258
BARDEX (Barrier Dynamics Experiment): Taking the Beach into the
Laboratory
J. Williams†, G. Masselink†, D. Buscombe†, I. Turner‡, A. Matias, Ó. Ferreira, A. Bradbury●, N. Metje§, L.
Coates§, D. Chapman§, C. Thompson●, T. Albers○, S. Pan□
†
School of Geography
University of Plymouth
Plymouth, PL4 8AA, UK.
[email protected]
[email protected]

CIMA, Universidade do
Algarve, 8000 Faro,
Portugal.
[email protected]
[email protected]
‡
●
Water Research Laboratory
University of New South Wales
Sydney, NSW 2052, Australia.
[email protected]
Nat. Oceanography Centre
Southampton SO14 3ZH, UK.
[email protected]
[email protected]
§
School of Civil Engineering,
The University of
Birmingham,
Birmingham, B15 2TT, UK.
[email protected]
[email protected]
[email protected]
○
Technische Universität
Hamburg-Harburg,
D-21071 Hamburg, Germany.
[email protected]
□
School of Engineering,
University of Plymouth,
Plymouth, PL4 8AA, UK.
[email protected]
ABSTRACT
WILLIAMS, J., MASSELINK, G., BUSCOMBE, D., TURNER, I., MATIAS, A., FERREIRA, Ó, METJE, N., COATES, L.,
CHAPMAN, D., BRADBURY, A., THOMPSON, C., ALBERS, A. & PAN, S., 2009. BARDEX (Barrier Dynamics
Experiment): taking the beach into the laboratory. Journal of Coastal Research, SI 56 (Proceedings of the 10th
International Coastal Symposium), 158 – 162. Lisbon, Portugal, ISSN 0749-0258.
Although relatively common features, few laboratory studies have examined the dynamic response of gravel
beaches and barriers to both tides and waves. To address this a prototype gravel barrier (5 m wide and 4 m high
with seaward and lagoon facing slopes of 1:8 and 1:4) composed of sub-rounded gravel ( D50=10 mm) has been
studied in the Delta flume. Detailed hydrodynamics and beach morphology were measured using video, buried
PTs, ECMs and closely spaced bed location sensors on a scaffold frame spanning the entire barrier. Additional
measurements were obtained from instruments on an offshore frame. A series of systematic tests were
undertaken using pumps to change water levels on the seaward (hs) and lagoon (hl) sides of the barrier. These
included: 1) hydraulic conductivity tests (hs and hl levels were varied); 2) tests to assess the impact of waves (hs
= 2.5 m with variable hl and waves of 1m and periods 5-7 s; 3) tests examining the effect of tides (tidal
simulation by varying hs from 1.75 m to 3.25 m with hl at high, medium and low levels and 1 m random and
regular waves with periods 3 s, 5 s and 7 s); and 4) overwash tests (tidal simulation with variable hl and random
waves of height ca. 1 m and periods 4.5s, 6s, 7s and 8s). This paper is intended primarily to make the
community aware of the experiments and to describe the objectives and methods used.
ADITIONAL INDEX WORDS: gravel barrier, field-scale laboratory experiment, tidal simulation, overwash,
morphodynamics.
INTRODUCTION
Most of the world’s gravel beaches are found in meso- to
macro-tidal settings, where tidal effects on beach
morphodynamics cannot be ignored (MASSELINK AND SHORT,
1993) and where beach porosity can exert a significant influence
on morphodynamic behaviour. However, most previous largescale and small-scale laboratory flume experiments have used a
fixed mean water level to study the response of gravel beaches to
waves (e.g. ROELVINK AND RENIERS, 1995; BLANCO, 2002).
Although a few studies have attempted to examine the response of
gravel beaches to waves and tides (e.g. TRIM ET AL., 2002), the
experiments are subject to scaling problems and the beaches used
are emplaced on impermeable ramps at the end of the test
facilities. Such experiments fail, therefore, to simulate important
aspects of natural gravel beach hydrology. Moreover, many gravel
beaches (with a hydraulic conductivity greatly exceeding that of
sand beaches) are barrier beaches which front and protect low-
lying coastal areas (lagoons, estuaries, and coastal plains) from
coastal flooding. Being subjected to relative changes in water level
on both their seaward and landward sides, hydraulic gradients are
likely to be an important element governing their dynamics and
stability.
During extreme events, gravel barrier overwashing can
sometimes lead to breaching and contribute over time to largescale roll-back. This essentially 2D response differs significantly
from sandy barriers where weak coastal dune sections frequently
provide foci for destructive wave action. In this case, the overwash
process is highly 3D and thus not amenable to study in a flume.
Although there have been a limited number of attempts to simulate
storm conditions at prototype scales, the post-storm recovery
processes acting to restore the beach have not been simulated. It is
known that overwashing plays an important role in the reestablishment of the pre-storm profile, but our understanding of
the processes by which this is achieved remains incomplete.
Providing some scaling issues can be resolved, there is therefore a
great deal that can be learned from a series of controlled large-
Journal of Coastal Research, Special Issue 56, 2009
158
BARDEX (Barrier Dynamics Experiment)
scale experiments in which a gravel barrier is subjected both to
simulated tidal motion and a range of wave conditions.
This paper describes research undertaken in the large Delta
flume facility, operated by Deltares, the Netherlands, during June
and July 2008. The aim of the research was to investigate the
hydro-, morpho-, and hydrodynamics of a near-prototype scale
open-coast tidal beach composed entirely of medium gravel, and
backed by a lagoon. One fundamental objective was to obtain data
required to parameterise and understand gravel beach sediment
transport processes and pathways, as well as the geotechnical and
hydraulic properties and processes within the beach. For that
purpose experiments were undertaken in a range of still water and
lagoon levels, with time-varying still water level used to simulate
tides. Here the simulated wave conditions were representative of
relatively calm seas. Detailed measurements were taken of: the
nearshore flow field and sub-tidal bedforms; swash
hydrodynamics; and beach/bed-levels. To address the problem of
parameterising and modelling the incipient conditions for natural
overwashing and barrier failure during storms, another set of
experiments used wave conditions typical of rough seas and tidal
modulation. Detailed measurements were taken of nearshore flow
fields and profile change as the barrier underwent progressive
decay until failure occurred.
BARDEX OBJECTIVES
BARDEX activities were split into five workpackages (WPs).
Although each had specific objectives, all WPs aimed to produce
high-quality data sets for process studies and for model calibration
and testing. WP1 (barrier stability and over-wash) aimed to
improve understanding of overwash sediment transport and the
threshold conditions for overwash occurrence. Work included:
verification of wave and tide conditions to initiate overwash flow
and overwash sedimentation; a study of barrier water level
influences on overwash flow velocity; overwash induced
morphological changes under flood and ebb tides; establishment
of threshold conditions for overwash crest accretion/erosion;
determination of critical conditions for barrier breaching. WP2
(barrier groundwater) aimed to investigate the role of backbarrier lagoon levels on the dynamic groundwater profile through
the barrier and to assess whether varying groundwater levels may
induce differing morphological response at the beach face.
Specific research included: effects of lagoon and seaward water
levels on the beach groundwater profile; effects of changes in
beach groundwater profile on erosion and accretion processes.
WP3 (swash sediment transport) aimed to improve
understanding of sediment transport processes on the beach face
under accretionary and erosive conditions. Specific research
included: a study of short-lived pressure impacts resulting from
collapsing/plunging breakers and their role in sediment transport;
the morphological response to tidal asymmetry; the role of swash
in morphological change; identification of key hydrodynamic
parameters controlling the magnitude and direction of net
sediment transport over individual swash cycles; investigation of
relationships between step formation and berm development; the
relative roles of incident, sub-harmonic, infra-gravity and
progressive waves in beach face changes. WP4 (nearshore
hydrodynamics and sediment) aimed to examine hydrodynamics
and sediment transport at locations from the swash region to
locations offshore from the barrier. Studies were undertaken in
variable wave conditions to measure bed shear stress and
roughness, to quantify turbulence production; and to measure bed
elevation changes attributable to sediment transport. Using data
from the WPs outlined above. WP5 (integration and modelling)
aimed to develop new practical modelling tools to: (a) predict
barrier morphology from the offshore limit of sediment transport
to the lee side of the barrier; and (b) predict the response of barrier
morphology to beach recharge. It is also supported by field data
(e.g. Slapton Sands, UK) and has the long-term objective of
developing modelling tools to quantify the barrier response to
combined wave action and tides and the barrier response to
increased sediment volume at the beach face. Specifically, this
WP is focused on the integration of experimental data and model
evaluation, development, validation and application. The
modelling activities include: evaluation of existing modelling
tools for gravel beaches and exploration of the relevance of these
tools to the current project; a review of sediment transport
formulations published for gravel beaches; implementation of
sediment transport formulations reviewed for selected models,
(e.g. XBeach); validation of the models; process studies to
improved model representation of overwashing, overtopping,
beach profile evolution, bedforms and effects of water level
modulation.
METHODS
A 4 m high and 50 m wide gravel barrier was constructed in the
Delta flume (Netherlands) with a ‘sea’ and a ‘lagoon’ at either
side (Fig. 1). The grain size distribution of the gravel is shown in
Fig. 2 (D50=10mm). The scale of the Delta flume enabled the
experiments to be conducted with natural gravel, which minimised
the adverse scale effects reported by BLANCO (2002). By raising
and lowering the water level in the lagoon, the groundwater in the
barrier was elevated and lowered relative to the mean sea level
and the effect on beach profile development could be investigated
for a range of wave and water level conditions. The sub-aerial part
of the barrier was extensively instrumented with buried pressure
sensors measuring the beach groundwater, current meters for
recording swash flows and acoustic devices for measuring swash
depths and bed levels (TURNER ET AL., 2008). Additional current
meters and capacitance wires were deployed offshore. The water
levels in the sea and lagoon were maintained at set levels by 4 x
100 l/s pumps and pipe work connections. The system was able to
control the water levels either side of the barrier to a tolerance of
10mm, permitting the simulation of differing sea-level, tide and
beach groundwater conditions. Pump discharges in and out of the
lagoon were recorded and enabled direct determination of flow
rates through the barrier.
The incident wave field was measured using three mechanical
wave followers. These wave height data, in combination with one
velocity measurement, were sufficient to describe the incoming
and outgoing wave characteristics (spectra) accurately. The
distances between the wave followers were periodically adjusted,
depending on the steepness of the waves, as prescribed by the
Mansard and Funk method for derivation of the incident spectrum
from the measured trace. Wave-induced currents were measured
using four 40mm-diameter electromagnetic current meters
(ECMs) deployed from the side wall of the flume. Thirteen Druck
pressure transducers with a 1 bar (10m) range and absolute
accuracy of 0.4% (40mm) were deployed at locations beneath the
barrier. These gave a working accuracy of c.1cm over the timescale of individual wave cycles. Barrier profiles were measured
between each wave sequence using a roller and actuator which
followed the bed profile from an overhead carriage thereby
allowing supra-tidal and sub-tidal profiles to be measured with
identical accuracy. Offshore a frame was used to deploy a Sontek
10MHz Ocean Probe ADV and two 10MHz Nortek Vectrino
ADVs to measure turbulence and a dual 2MHz Marine Electronics
Sand Ripple Profiler (SRP) Sonars to measure in detail bed
morphology.
Journal of Coastal Research, Special Issue 56, 2009
159
Williams et al.
Fig. 1 Plans of the experimental design showing barrier emplacement, pumping system and instrument locations. The photographs show
details of the equipment and Delta flume.
An array of 45 temperature compensated Massa
M300 ultrasonic proximity sensors operating at 95kHz were
mounted at c. 1m from the bed on a frame spanning the barrier
(Fig. 3). Each has a beam angle of 8o giving a measurement
footprint c. 28cm in diameter. The beach frame was also used to
deploy pairs of small ECMs and PTs at 4 streamwise locations.
Their height was adjusted after each experiment. A Sony SSCDC50AP video camera positioned on the profiling gantry high
above the centre of the flume, and facing the waves, was used to
record wave run-up. Images were referenced to ground control
point positions and recorded at 4Hz into Matlab data files
following image orthorectification. These images were
subsequently filtered to remove strong gradients in sunlight across
the flume. Using a local coordinate system, instrument positions
were surveyed using a Trimble 5605 Robotic Total Station. Using
a local network, all logging computers were synchronised from a
GARMIN GPS using TAC32 software. Further details of the
methods and instruments are given by BUSCOMBE ET AL. (2008).
The JONSWAP spectral shape was used to generate random
wave sequences with specified characteristics. These propagated
down the flume in consistent phase to produce the same
hydrodynamic forcing conditions when required. Reflected waves,
as well as low frequency resonant waves were damped using an
Automated Reflection Compensator (ARC). During the tidal runs
a wave-steering signal was devised to avoiding stoppages
for resets of the wave board and filling or draining of the flume.
To achieve this, tides were split into several stages in which the
water was raised or lowered at a near-constant rate with wave
generation on. To avoid the need to reset the wave board during
tidal simulations, the wave steering signal was based on the
average water depth for a given tidal stage. Continuity was
obtained by tapering the design wave conditions to zero at the start
and end of each stage and by concatenation of all the wave signals
from each tidal stage. Although the use of an average water depth
for each tidal stage resulted in wave height statistics that differed
slightly from the design specification due to the continuous
Journal of Coastal Research, Special Issue 56, 2009
160
BARDEX (Barrier Dynamics Experiment)
change in water level, this did not adversely affect the
experiments.
Five test series were undertaken (Fig. 4 for test details). Test
Series A: to determine the in-situ bulk hydraulic conductivity of
the gravel barrier in the absence of waves. Test Series B: to
examine barrier response to waves only; Test Series C: the
response of the barrier to waves with a fixed offshore water level
and varying lagoon levels. Test Series D: barrier response to
waves using simulated tidal cycles; and Test Series E: to examine
the response of the barrier to storm conditions using tidal
simulation and large waves.
Fig. 4 Experimental tests (in chronological order) and associated
wave heights, periods and water levels
Fig. 2 Measured grain size distributions for the barrier gravel.
Fig. 3 Barrier and offshore frames in the Delta flume (left) and the
ultrasonic sensor array looking towards the wave paddle (right).
DISCUSSION
A number of aspects of the BARDEX project are novel. (1)
Laboratory gravel beach research is relatively rare, especially on
this scale (the notable exception being the GWK experiments
reported by BLANCO (2002). (2) The experiments are believed to
be the first combining waves with variation of offshore and lagoon
water levels. (3) The state-of-the-art measurements, including
high-frequency measurements of waves, turbulence, run-up, subtidal, intertidal and supra-tidal bed morphologies, sediment size,
and groundwater table are some of the most detailed ever
undertaken in a laboratory study of a gravel beach. The data set
collected will therefore be of considerable interest to the wider
academic community in the fields of coastal hydrodynamics and
hydraulics, coastal defence and geotechnics, and nearshore
morphodynamics and sediment transport.
Fig. 5 (a) overwashing wave sequence; (b) barrier profile changes
during the overwash experiments (2.5 hours.; Hs = 0.8m; Tp = 8.0;
hs = 3.00 to 3.75m; hl = 3.25m.
As the work was only undertaken in the summer of 2008 it is
not yet possible to present detailed results from the BARDEX
experiments and here only initial results from the overwash
experiment are presented as an example. The present exception to
this is the more detailed treatment of results from WP2 given by
MASSELINK ET AL. (2009, this volume).
Fig. 5 shows the barrier profile before, during and after the
BARDEX overwash experiments. The barrier crest has been
displaced c. 10m with accompanying lowering of the crest height,
erosion of the seaward side and accretion in the lagoon.
Interestingly a series of well-developed bedform-like features
were present on the beach face at the end of the study (shown also
Journal of Coastal Research, Special Issue 56, 2009
161
Williams et al.
in the inset photograph in Fig. 5). Although these have been
observed in the field, data on their characteristic and behaviour is
scarce.
Volumetric change of the crest and the overwash region during
the final test series E9 are shown in Fig. 6. The volumes are wellbalanced, indicating that the erosion of the crest results in the
accretion in the overwash region. It is also of interest to note a
near constant rate of volumetric change during the runs with
constant wave-tide conditions (sea level at 3.75 m). Crest build-up
during the last two tests (17 and 18, Fig. 6) is also apparent.
MASSELINK, G.; WILLIAMS, J. J. AND TURNER, I. L., 2009.
Large-Scale Laboratory Investigation into the Effect of the
Beach Groundwater Table on Gravel Beach Morphology.
Journal of Coastal Research, Special Issue 56, this volume.
ROELLEVINK, J.A. AND RENIERS, A.J.H.M., 1995. LIP11 D
Deltaflume experiment: dataset for profile validation. Report
H2130 Delft Hydraulics, Delft, The Netherlands.
TRIM, L.K.; SHE, K., AND POPE, D.J., 2002. Tidal effects on
cross-shore sediment transport on a shingle beach. Journal of
Coastal Research, SI 36, 708-715.
TURNER, I.L.; RUSSELL, P.E., AND BUTT, T., 2008.
Measurement of wave-by-wave bed-levels in the swash zone.
Coastal Engineering, doi:10.1016/j.coastaleng.2008.09.009.
ACKNOWLEDGEMENTS
The data reported here were collected in the Delta flume
(Netherlands) as part of the EU-funded BARDEX project
(HYDRALAB III Contract no. 022441 (RII3), Barrier Dynamics
Experiments). We would like to thank all BARDEX collaborators
for their contributions, but in particular Celia Swinkels for
managing the project and the staff of the Delta Flume for their
high-spirited and excellent assistance.
Fig. 6 Results from test series E9 showing volumetric changes in
barrier sediments.
CONCLUSIONS
Experiments in the Delta flume, with a near field-scale gravel
barrier, have for the first time examined barrier response to
different water levels either side of the barrier in the presence of
waves and to overwash events. The use of metered high capacity
pumps has allowed water levels to be held at different relative
level either side of the barrier and enabled direct measurement of
hydraulic conductivity. Data acquired in the experiments will
assist in the development and testing of models to simulate crossshore morphological change on gravel beaches and in improving
understanding of processes. This in turn will assist in the design of
artificial nourishment schemes and in the management of natural
gravel barrier systems threatened by erosion and/or rising sea
level. The BARDEX data set will be made available to bona fide
researchers in the future to assist in progressing knowledge and
prediction of gravel beach and barrier behaviour.
LITERATURE CITED
BLANCO, B., 2002. Large Wave Channel (GWK) Experiments on
Gravel and Mixed Beaches. Experimental procedure and data
documentation. Report TR-130, HR Wallingford.
BUSCOMBE, D.; WILLIAMS, J.J. AND MASSELINK, G., 2008.
BARDEX: Experimental procedures, technical information
and data report. School of Geography, University of
Plymouth, UK, 164pp., unpublished manuscript1.
MASSELINK, G., AND SHORT, A., 1993. The influence of tide
range on beach morphodynamics: a conceptual model. Journal
of Coastal Research, 9, 785-800.
1
Now available from University of Plymouth, subject to a handling fee.
Journal of Coastal Research, Special Issue 56, 2009
162