Geomorphology 290 (2017) 265–276
Contents lists available at ScienceDirect
Geomorphology
journal homepage: www.elsevier.com/locate/geomorph
Shoreline resilience to individual storms and storm clusters on a meso- MARK
macrotidal barred beach
Donatus Bapentire Angnuurenga,b,c,⁎, Rafael Almarb, Nadia Senechala, Bruno Castellea,
Kwasi Appeaning Addoc, Vincent Marieua, Roshanka Ranasinghed,e,f
a UMR EPOC, University of Bordeaux/CNRS, Bordeaux, France
b UMR LEGOS, University of Toulouse/CNRS/IRD/CNES, Toulouse, France
c MAFS/Remote Sensing Laboratory, University of Ghana, Accra, Ghana
d UNESCO-IHE, Delft, The Netherlands
e Harbour, Coastal and Offshore Engineering, Deltares, Delft, The Netherlands
f University of Twente, Enschede, The Netherlands
A R T I C L E I N F O A B S T R A C T
Keywords: This study investigates the impact of individual storms and storm clusters on shoreline recovery for the meso-to
Storm clusters macrotidal, barred Biscarrosse beach in SW France, using 6 years of daily video observations. While the study
Beach erosion and recovery area experienced 60 storms during the 6-year study period, only 36 storms were analysed due to gaps in the
Sandbar video data. Based on the 36 individual storms and 13 storm clusters analysed, our results show that clustering
Extreme events impact impact is cumulatively weak and shoreline retreat is governed by the first storms in clusters, while the impact of
Open beach
Short-term morphodynamics subsequent events is less pronounced. The average post-storm beach recovery period at this site is 9 days,
consistent with observations at other beaches. Apart from the dominant effect of present storm conditions,
shoreline dynamics are also significantly affected by previous storm influence, while recovery is strongly
modulated by tidal range and the bar location. Our results reveal that not only is the storm energy important but
also the frequency of recurrence (storms result in greater retreat when time intervals between them are longer),
which suggests an interaction between short storm events and longer-term evolution.
1. Introduction is still somewhat unclear (Ranasinghe et al., 2012; Pianca et al., 2015).
The fact that beaches eventually recover to their pre-storm state means
Sustainable management of coastal resources requires a thorough that the beach response does not only depend on storm conditions but
understanding of the processes that drive changes in the shoreline also on other factors such as sea level and its chronic behavior (Zhang
location. The shoreline is a highly dynamic interface between land and et al., 2002), the previous beach state (Wright et al., 1985; Grasso et al.,
ocean and is thus affected by various forces operating at different 2009; Yates et al., 2009) and/or previous wave conditions (Davidson
spatio-temporal scales. Shoreline evolution is to a large extent governed et al., 2013; Splinter et al., 2014b).
by meteorological and oceanic conditions: waves, tides, currents and Given that individual storms can result in dramatic shoreline
atmospheric conditions (wind, inverse barometer). It is generally changes, storms can be treated as outliers (Zhang et al., 2002). Despite
assumed that wave breaking is the main driver of coastal evolution their large impact, storms are considered independent from long-term
but its role may be strongly modulated by other factors. For example, evolution and described separately because of their transient influence
on the lower part of the beach (the beach margin beneath the water due to rapid post-storm recovery. Zhang et al. (2002) support the
surface from the shoreline), a storm may have more erosive impact at assertion of Douglas and Crowell (2000) that the most practical option
low tide than at high tide. Although many studies have focused either is to remove such events from long-term evolution studies. In contrast,
on simple or complex paradigms of shoreline evolution from Wright Fenster et al. (2001) and Genz et al. (2007) observed that individual
and Short's (1984) beach state classification method, as well as more storms should not be excluded, and that including their contribution
complex cross-shore equilibrium models (Yates et al., 2009) and a mix could potentially improve the prediction of long-term shoreline evolu-
of cross-and alongshore-based models (Morton et al., 1993; Hansen and tion. The short-term shoreline changes induced by storms are generally
Barnard, 2010), the response to perpetually changing forcing conditions characterized by a rapid erosion followed by a slower post-storm
⁎ Corresponding author at: UMR EPOC, University of Bordeaux/CNRS, Bordeaux, France.
E-mail address: donatus.angnuureng@ucc.edu.gh (D.B. Angnuureng).
http://dx.doi.org/10.1016/j.geomorph.2017.04.007
Received 7 August 2015; Received in revised form 24 February 2017; Accepted 6 April 2017
Available online 09 April 2017
0169-555X/ © 2017 Elsevier B.V. All rights reserved.
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
recovery and are influenced by storm characteristics (e.g. energy and Changes in sandbar location due to varying wave conditions have
duration, individual versus sequences of storms, or storm clusters; see, been widely documented (e.g. Wright et al., 1985; Lippmann and
among others, Yates et al., 2009; Karunarathna et al., 2014; Coco et al., Holman, 1990; Gallagher et al., 1998; Castelle et al., 2007a, among
2014; Senechal et al., 2015). others). Bar decay can result in its inability to offer protection during
Investigations on storm impact have mainly followed two ap- storms, leading to intensified upper beach erosion (Castelle et al.,
proaches; 1) non-cumulative analyses (e.g. Frazer et al., 2009; Coco 2007b), and alongshore irregularities of sandbar crest can force a
et al., 2014; Splinter et al., 2014a) which take individual storms as template of onshore wave field resulting in localized upper beach
independent events and show that frequent storms or storm sequences erosion (Thornton et al., 2007; Castelle et al., 2015). At barred beaches
do not have a persistent influence on longer term shoreline evolution with large tidal ranges, it is observed that both the sandbar and the tide
and; 2) cumulative storms analysis (e.g. Ferreira, 2005; Karunarathna modulate onshore wave breaking intensity and control morphological
et al., 2014) which shows that storm sequences enhance erosion. The changes (Almar et al., 2010; Ba and Senechal, 2013; Stokes et al.,
latter has been further evidenced recently by equilibrium-based semi- 2015). Although shoreline and sandbar changes have been studied
empirical shoreline models (e.g. Yates et al., 2009; Davidson et al., rather extensively (e.g. Lippmann and Holman, 1990; Hansen and
2013; Castelle et al., 2014) with the rate of shoreline migration under a Barnard, 2010; van de Lageweg et al., 2013), changes have been studied
storm depending on the disequilibrium between the storm energy and mostly as discrete events (except in few studies such as van de Lageweg
previous beach state, and the beach constantly trying to reach a new et al., 2013 at an embayed beach) and their combined effect on storm
equilibrium under varying waves. The equilibrium approach raises the impact and beach recovery evolution is still uncertain.
importance of the so-called beach ‘memory effect’ that suggests beach The above discussion highlights that there are still many knowledge
response depends on the antecedent wave conditions. The transient or gaps regarding shoreline resilience to storms at meso- to macrotidal
persistent effects of individual storms and storm clusters is still a subject beaches. This study aims to take a first step towards addressing some of
of debate and discrepancies on storm impact characterization still exist these knowledge gaps. Specifically, the objective of this study is to: (a)
(e.g. Dolan and Davis, 1992; Mendoza et al., 2011; Splinter et al., quantify shoreline resilience to individual storms and storm clusters, (b)
2014a; Senechal et al., 2015). investigate the influence of tide and sandbars on shoreline position, and
To understand the persistence of storms and beach resilience to (c) estimate the post-storm beach recovery duration. To achieve this
different storm recurrence intervals and intensities, a good under- goal, six years (2007–2012) of daily video observations at Biscarrosse, a
standing of post-storm beach recovery conditions and duration is barred meso- to macrotidal beach, are analysed. In Section 2, the study
crucial. Beach recovery from storms depends on the severity of the site and video methods are described. Section 3 presents the results on
event(s) and on how far the sediment has been transported offshore the shoreline response to storms at timescales from days to years, with
(Corbella and Stretch, 2012). With high frequency (daily) video data, an emphasis on the influence of storm recurrence and the modulation
post-storm recovery durations of 5 to 10 days have been reported by played by tidal range and sandbar. The role of tide on shoreline
Ranasinghe et al. (2012) for the microtidal Palm beach, Australia and response to storms and the importance of the frequency of recurrence
Duck beach, USA. However, the recovery duration is yet to be of storms on shoreline resilience are discussed in Section 4 and, finally,
investigated at high-energy meso- to macrotidal beaches, although conclusions are presented in Section 5.
Senechal et al. (2015) postulated that recovery at these latter types of
beaches could be rapid. This could be due to the presence of ‘usual’ 2. Methods
winter storm conditions. To date, different diagnostics have been used
to quantify beach recovery in various studies (e.g. Maspataud et al., 2.1. Field site
2009; Corbella and Stretch, 2012; Ranasinghe et al., 2012) and an
objective means of comparing and contrasting these different estimates Biscarrosse beach, located in the SW France (Fig. 1), is exposed to
is yet to be identified. long and energetic waves originating mainly from the W-NW. The mean
Although it is widely accepted that the shoreline is mostly affected annual offshore significant wave height Hs is reported as 1.4 m with an
by waves, the influence of tidal range and sandbar location cannot be associated averaged mean period Tp of 6.5 s (Butel et al., 2002). Waves
overlooked, in particular at barred meso- to macrotidal beaches such as show seasonal variability (Butel et al., 2002): during fall and winter
encountered along the SW France Aquitaine Coast (Castelle et al., seasons (November to March), mean Hs is 1.6 m and Tp is 7.3 s, while
2007a) or at Perranporth in the north-west coast of Cornwall in the UK during spring and summer (April to October) mean Hs is 1.1 m with a
(Stokes et al., 2015). It has been observed that storm events, while shorter Tp of 6 s (Senechal et al., 2015). The tidal range is meso- to
capable of causing large short-term changes in the shoreline, do not macrotidal, with an average value of 2.9 m that increases up to 5 m
singularly account for the overall observed change (Hansen and during spring tide. The average beach slope is about 0.03, and sediment
Barnard, 2010), and wave impact could be negligible with respect to at the site consists of fine to medium quartz sand with median grain
the magnitude of the seasonal signal and the effect of the inter-annual sizes ranging from 0.2 to 0.4 mm (Gallagher et al., 2011).
signals (Pianca et al., 2015). In macrotidal environments, tides are Biscarrosse is an open double-barred beach; the outer bar often
regarded as a primary factor in the control of the hydrodynamic and exhibits crescentic patterns, while the inner bar in the intertidal domain
sedimentary processes of intertidal flats (Davis, 1985; Masselink and commonly exhibits a transverse bar and rip (TBR) morphology with a
Short, 1993; Robin et al., 2007). There is field evidence for the tidal mean wavelength of about 400 m (Almar et al., 2010). Based on three
modulation (attenuation) of incident wave power by the large tidal years of daily video images, Peron and Senechal (2011) also indicate that
range (Robin et al., 2007; Davidson et al., 2008; Guedes et al., 2011) both up-state and down-state transitions were dependent on the previous
which eventually affects the shoreline. Zhang et al. (2002) observed beach state and that no ‘direct jump’ from the reflective state to the
that the combination of large waves with high water levels during five dissipative beach state was observed. They discussed the possibility that
continuous high tides caused the largest recorded dune (upper beach) the presence of the subtidal bar probably explained the persistence of TBR
erosion from Long Island, New York, to Cape Hatteras. This suggests beach state (mean residence time of about 24 days reaching maximum at
that the effect of tides actually depends on the part of beach (upper, 103 days), even during high-energy conditions as reported in other similar
intertidal or lower) being investigated. However, the effect of tides on environments (Almar et al., 2010). Using three years of video observa-
storm impact at meso- to macrotidal sandy shorelines is relatively tions, Senechal et al. (2015) showed that the range of variation of the inner
poorly investigated, although some recent studies suggest the inclusion sandbar location (120 m) at Biscarrosse is two and a half times larger than
of tidal range in shoreline prediction models can be important (Stokes the range of variation of the shoreline and that rapid erosion of the
et al., 2015). shoreline can be observed under moderate conditions.
266
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
Fig. 1. Location of the study site, Biscarrosse beach (SW France), showing the WW3 grid node (triangle) located at−1°30′W, 44°30′ N and Candhis buoy (triangle) at 1°26.8′W, 44°39.15′
N and the video station.
2.2. Video data and Plant, 2001) being the popular SLIM method, a typical approach
where an intensity peak is used as a proxy for the location of the
A shore-based video system (e.g. Lippmann and Holman, 1989, shoreline, and suitable for reflective beaches (Plant et al., 2007).
1990; Holman et al., 1993; Plant and Holman, 1997) was installed at Subsequent methods used color images (or both color and gray), a
Biscarrosse beach in April 2007 by EPOC laboratory (CNRS/University more sophisticated method (Turner et al., 2001; Aarninkhof et al.,
of Bordeaux) in collaboration with the New Zealand National Institute 2003) based on color segmentation, applicable to detecting the shore-
of Water and Atmosphere (NIWA) (see Almar et al., 2009; Senechal line at both reflective and dissipative beaches. In our study, errors have
et al., 2015). The video station contains five color cameras fixed atop been minimized with the manual delineation of the shoreline (Fig. 2e)
the foredune at 26 m above the mean sea level (MSL), although only to ensure a high-quality dataset. At meso- to macrotidal barred beaches,
four camera images (Fig. 2a–d) were in good state during the observa- it is difficult to select the elevation that best represents the overall
tion period of the present study. The system provides three types of intertidal complex morphology, as observed by Castelle et al. (2014).
images every 15 min including processing time: snapshot, cross-shore Following this and to minimize the influence of the complex intertidal
time stacks and 10-min time exposure (or timex) images. Images are zone, shoreline location was defined here for elevations at
merged and rectified on a 1 m × 1 m grid using conventional photo- 0.45 m ± 0.1 m above MSL (Fig. 2) which corresponds to the lowest
grammetric methods (Holland et al., 1997). The transformation be- high tide level, commonly used through video imagery to get daily
tween oblique image and real-world coordinates was achieved using 18 shoreline data at meso- to macrotidal beaches (e.g. Birrien et al., 2013;
ground control points surveyed with a differential GPS (DGPS, cen- Senechal et al., 2015). Due to the absence of a tide gauge at Biscarrosse,
timeter accuracy). The origin (X = 0, Y = 0) of the local coordinate the tide used here was extracted from a reconstructed tidal signal based
system is the camera location oriented along the cross-shore (X) and on tidal harmonics (WXtide software, Flater, 2010) with reference to
alongshore (Y) directions while the vertical Z = 0 origin denotes the the closest point at Arcachon (1°10 W, 44°40 N, Fig. 1), about 30 km
Mean Sea Level (MLS). A region covering beach area of 1200 m from Biscarrosse (after phase-lag correction). Overall, the video-derived
alongshore and 400 m cross-shore is observed (Fig. 2e, f). shoreline dataset covers 1036 days in 6 years, which is 54.2% of the
Commonly used proxies for shoreline position are either based on study period.
visual assessment (e.g. the high water line) or datum-based (see Boak Video-derived shorelines are subject to relatively large uncertainties
and Turner, 2005). Datum-based shorelines generally consist of the (Holman and Stanley, 2007). In particular, the shoreline-detection
cross-shore position of a specified elevation contour, such as mean high methods are sensitive to waves and lighting conditions. For instance,
water (MHW), the method chosen in this study. Shorelines derived from the SLIM method by Plant and Holman (1997) is sensitive to variations
video have become increasingly common (Plant and Holman, 1997; in water levels which can scale the effects of both setup and run-up, and
Aarninkhof et al., 2003; Plant et al., 2007; Smit et al., 2007). Different fog can reduce the color signal strength (Aarninkhof et al., 2003).
categories of images have been used to delineate shoreline, with the However, the results of shoreline measured from video have been
first methods based on gray images (Plant and Holman, 1997; Madsen comparable to that of topographic surveys (Holman and Haller, 2013)
267
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
Fig. 2. Illustration of camera view fields (a-d) from oblique 10-min averaged images with manual delineation of e) shoreline (29 Sept. 2008) and f) inner sandbar crest (15 June 2007) on
rectified, merged images.
and the differences have been extensively discussed in previous works intermediate to fair energetic conditions. Several studies have shown
(e.g. Aarninkhof et al., 2003; Plant et al., 2007). In addition to the error that surveyed sandbar crests and those extracted from timex video
related to image rectification estimated here at 1–2 m, an error of 0.5 m images are in good agreement (R2 ~ 0.8; Lippmann and Holman, 1989;
is added for shoreline identification equal to the pixel footprint. The Plant and Holman, 1998). The accuracy of sandbar location also
mean pixel resolution at the shoreline location is about 0.1 m and 0.2 m depends on the rectification error of 1–2 m and due to manual
in the cross-shore and alongshore directions, respectively, which digitization and the pixel footprint of 2 m, tide- and wave-induced
worsen to 1–3 m at the viewed edges. Due to the lack of information artificial shift (van Enckevort and Ruessink, 2001; Pape and Ruessink,
on the actual surf zone bathymetry, the main horizontal uncertainty, 2008; Almar et al., 2010) of 5–10 m. In the sandbar location, the pixel
the wave-induced setup was estimated at 0.35β HsL , with β the upper footprint was poorer (reaching 12 m) at around 400 m from the camera.
beach slope and L the offshore wave length, following Stockdon et al. Considering all of the above-mentioned sources of uncertainty, an
(2006). Aarninkhof et al. (2003) reported that such simplification overall error of 15 m for the sandbar location is calculated for this site,
introduces minor deviations in the wave-induced setup at the shoreline. consistent with that reported at Truc Vert beach (Almar et al., 2010),
The associated setup error on shoreline location is about 6 m (for 30 km north of Biscarrosse beach, and using a similar camera setting.
Hs ~ 6–8 m) considering the average beach slope of 0.03, but ranges
between 2 and 12 m for all the data. For complex submerged
morphological beaches such as Biscarrosse, alongshore variations of 2.3. Storms
wave-induced setup can be established. In our study, this bias is
substantially reduced because shoreline location is estimated out of Wave data for this study were obtained from Wavewatch III model
stormy periods. Given the restraints listed above, we estimate that the (Tolman, 1991) at the grid point facing the beach (1°30′W, 44°30′N,
overall uncertainty on video-derived shoreline location is about 9 m. Fig. 1) in about 70-m water depth, at a 3-h interval over the study
Timex images (Fig. 2f) are used, to average-out high-frequency period (2007–2012). The significant wave height, Hs was further
intensity fluctuations due to individual waves, providing a statistically corrected via linear regression with a directional wave buoy
stable pattern of the breaking (Lippmann and Holman, 1989; van (1°26.8′W, 44°39.15′N) moored in 50-m water depth, following
Enckevort and Ruessink, 2001). The high-intensity bands associated Castelle et al. (2014).
with breaking (see Fig. 2f) are commonly used as a proxy for bar crest The commonly used peak over threshold method (POT) is applied
location (Lippmann and Holman, 1989; Pape and Ruessink, 2008; on Hs to select large wave conditions and identify storms (e.g. Dorsch
Almar et al., 2010; Guedes et al., 2011). There is always a substantial et al., 2008). A 5–10% exceedance Hs is commonly adopted in scientific
error O (1–10 m) when locating the cross-shore position of the bar studies to define storm events (e.g. Dorsch et al., 2008; Rangel-Buitrago
crests (van Enckevort and Ruessink, 2001). This is mostly due to the and Anfuso, 2011; Splinter et al., 2014a; Castelle et al., 2015). In the
translation of the breaking zone resulting from the changes in wave present work, Hs values with a probability of occurrence< 5% are
characteristics and tidal level (Lippmann and Holman, 1989; van considered as major storms, corresponding to an Hs of 3.68 m, also in
Enckevort and Ruessink, 2001). In order to reduce the differences line with Splinter et al. (2014a) and Castelle et al. (2015). A single
between the detected and actual bar crest locations, and to be storm is defined as a continuous period of Hs exceeding this threshold
consistent with previous methodologies (e.g. van de Lageweg et al., (Fig. 3) and lasting at least one tidal cycle (12 h), consistent with
2013; Senechal et al., 2015) images for which Hs > 2.5 m were Senechal et al. (2015) approach and particularly to account for the
discarded. Inner-bar extraction was done at a constant water level of impact of tide. Another key parameter used in the present analysis is the
0.55 ± 0.1 m below MSL. The detection resulted in 411 daily along- Storm intensity I (m2hr), which is defined in several studies (e.g. Dolan
shore-averaged cross-shore sandbar positions< Xb >or lines, which is and Davis, 1992; Karunarathna et al., 2014; Senechal et al., 2015) as
20% of the entire period. the product of the maximum Hs by the storm duration, in line with
The reason for choosing low tide to pick the sandbar location relates annual maxima method. Here, we chose to follow the more time-
to the fact that waves barely break over the inner bar at high tides for integrated definition for I given by Mendoza et al. (2011), which is:
268
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
Fig. 3. Illustration of the method used to select a) storm characteristics, beginning and end of Hs above the threshold (Hs= 3.8 m, 95% exceedance level, shown as horizontal line in
upper plot) and in b) the exceedance level where the 50, 95 and 99% levels are shown. In a), squares are the beginning and end of storms, triangular marks are values greater than the
threshold.
t2 methods (e.g. using beach state as in Ranasinghe et al., 2012) as it does
I = ∫ H 2s (t) dt not depend on any forcing parameter. However, the method is highly
t1 (1) depended on the amount of shoreline recovery as considered in
where the storm duration D is the time between the beginning t and Corbella and Stretch (2012).1
the end t of each storm. Initiation of a storm t however was defined as A multiple linear regression (Eq. (2)) is used to investigate the role2 1
the time when the three hourly-averaged Hs exceeded the 0.75 quantile of 5 forcing/modulating parameters on shoreline changes during storm
(1.9 m), to be consistent with Masselink et al. (2014), and the end of the (Δ < Xs,i >) and on recovery periods (Tr): current storm energy Ii,
storm t was the time when the three hourly-averaged Hs returned previous storm influence, time between storms, tide range TR and2
below 1.9 m (Fig. 3a). sandbar-to-shoreline distance.
n
2.4. Storm impact Y = co + ∑ CkZk + ε
1 (2)
The Storm impact Δ < X >(in meters) is estimated as the mean where Y is the response variable, Z the predictor or causative variable, ns,i
alongshore averaged cross-shore shoreline migration from the begin- is the number of events (n = 36), co and Ck are the non-standardized
ning to the end of each storm, equivalent to the end point rate method regression coefficients and ε is the residual term. Forcing terms are
(Genz et al., 2007). There can be several ways to define the recovery considered independent. The relative contribution P(Z) of each forcing
duration after each storm. For example, Ranasinghe et al. (2012) used parameter is estimated from the ratio of individual variance to the total
an approach based on the beach states, where the time the nearshore following Eq. (3):
morphology takes to evolve from a post-storm state (e.g. dissipative/ Sk
longshore bar and trough) to its modal state (i.e. the most frequently P (Z ) = 100 (k = 1, 2,…5)S (3)
occurring beach state e.g. rhythmic bar and beach or transverse bar and Y
rip) is defined as the recovery duration. In our study, the recovery where Sk is the variance of CkZk and SY is defined as the sum of
duration Tr (in days) refers to the post-storm period of continuous variances of all causative components SY = ∑
k
1 CnZn to insure a total of
accretion, at the end of which the beach is assumed stabilized. Daily 100%.
average locations are determined from the shoreline location at the end The previous storm influence (S < Xs-i >) is defined as the rate of
of each storm. The length of the period of continuous accretion is previous storm impact (Δ < Xs-i >) with respect to the time interval (Δt
determined by the number of days taken to reach the first maximum in days) between storms (end of a storm and the start of another storm)
migration (recovery) value. This method contrasts with some existing based on the equilibrium concept that shoreline response depends on
269
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
Fig. 4. Time series of a) significant wave height Hs with storm periods (Hs > 3.68 m) in dots, selected 36 storms are marked in large black dots and 13 clusters of storm are marked
circles in gray, b) tidal range TR c) alongshore-averaged shoreline location< Xs >and d) alongshore-averaged sandbar location< Xb > .
the antecedent beach state (Yates et al., 2009) as in Eq. (4): complete tidal cycle (see Section 2.3). In winter (Table 1), 60% of
∆ < X > storms recur within 10 days, while in summer this occurrence isS < Xs−i > = s−i∆t (4) observed to be sparse, with storms recurring on average within
100 days. The year 2011 recorded the lowest number of storms, with
9 storms causing only limited erosion. It is also seen that the standard
3. Results deviation of storm energy is large in winter, which drives the
temporality of the observed shoreline response (Fig. 5a, thin dot solid
Fig. 4a shows that wave regime has large seasonal variations, with line) and sandbar (Fig. 5a, line marked circles). The largest number of
low and high energy in summer and winter, respectively, with Hs storms and most extreme (Hs > 5 m, defined as the 99% percentile,
ranging from< 1 m to 9 m. Fig. 4c shows that the alongshore-averaged Table 1) are observed in 2008 (25%), 2009 (25%) and 2010 (22%),
shoreline location< Xs >also follows a seasonal cycle with most which induced a large total erosion particularly in 2009 (Table 1).
onshore (85 m) and offshore (150 m) position in winter and summer, Individual storms result in a wide range of shoreline impacts (Table 1),
respectively. In Fig. 4d, the alongshore-averaged sandbar location< from large erosion (−21 m) and sometimes to accretion (+14 m). The
Xb >shows a large variability (range of 110 m), varying between immediate cause of this (uncommon) accretion during storms is
212 m to 322 m with outermost location in winter and a less marked unknown, but sediment input from dune erosion is one possible
seasonal cycle. On average, the sandbar-to-shoreline distance< mechanism for upper beach accretion (van Gent et al., 2008). The
Xb > − < Xs > is 162 m, but it can be larger (227 m) or smaller mean storm impact on the shoreline throughout all storms is an erosion
(102 m) during large (winter) and weak (summer) wave conditions, of 8.7 m (σ= 8.9 m).
respectively.
3.2. Modulation of storm impact and recovery by previous events, tides and
3.1. Characteristics of individual storms and morphological impact sandbar presence
In total, 60 storms were identified during the 6-year study period. Storm impact on the shoreline is often quantified separately from
However, due to gaps in the video data, which precluded the derivation the influence of sandbars and tides, except for some recent attempts
of shorelines, 24 of these storms were discarded from further analysis. (e.g. Senechal et al., 2015; Stokes et al., 2015). Here, the relative
The analysis presented herein therefore is based on the remaining 36 contribution of the current and previous storms, tides and sandbars, are
storms (Fig. 4a, marked black). The mean peak storm wave height is investigated together through a multiple linear regression (described in
4.9 m (standard deviation σ= 1.04 m) with the mean wave height Section 2.4). Overall, Fig. 6a–b show that a good agreement is found
throughout the storm duration being 4.5 m (σ= 0.8 m). The mean between reconstructed and observed Δ < Xs >and Tr with regression
storm peak wave periods throughout all the storms is 12.15 s coefficients equal to 0.74 and 0.69 (both significant at 95% level),
(σ= 2.16 s) and the mean storm duration 33 h (σ= 32 h). respectively. Even though Table 2 shows that cross-correlation between
The overall average interval between storms is predominantly terms of the multiple linear regression can be substantial (e.g. waves
seasonal (Fig. 5). Storms are more frequent in winter, while occurring conditions vs. sandbar location) showing some physical links, where the
almost throughout the year. In summer, only a few and short storms values are low, the results can nonetheless be used with reasonable
(< 6 h) are observed and usually do not meet the requirement of a accuracy since they are significant at 95% confidence level. Fig. 6c–d
270
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
Fig. 5. Monthly-averaged characteristics of a) shoreline< Xs > , solid lines indicate monthly standard deviations and sandbar locations< Xb > , dash lines mark standard deviations; b)
recurrence interval between storms; and c) average storm energy I (m2hr) per month. Shaded areas around lines indicate the monthly standard deviation.
3.3. Storm sequences
Table 1
Storm characteristics from 2007 to 2012 of maximum Hs (m), percentage (%) of extreme Fig. 7 shows an ensemble-averaged analysis of the evolution of the
storms (Hs > 5.0 m, 99% threshold), annual average storm impact Δ < Xs,i >(m), and
annual average storm duration (in days). sandbar and shoreline location during the post-storm recovery period.
Fig. 7b shows that while wave height is decreasing after the storm, the
Number of storms Hsmax Hsmax > 5 m Δ < Xs > Duration shoreline continuously migrates offshore (3.7 m/day) before it reaches
(m) (%) (m) (days) stabilization after 9 days (on average), which can be used as an average
2007 7 4.7 11 6.4 2.7 estimate for the post-storm recovery duration at Biscarrosse. This post-−
2008 18 5.0 25 −8.0 3.4 storm recovery duration is different from the time interval between
2009 15 5.0 25 −12.4 4.0 storms; whereas the interval between storms could comprise both
2010 13 4.7 22 −6.5 2.5 accretion and erosion, Tr is purely continuous accretion. Interestingly,
2011 9 4.8 4 3.0 2.7 while the shoreline is observed to stabilize in 9 days on average, the
2012 11 4.5 11 −10.3 4.5
mean 12 4.8 16 −7.0 3.3 sandbar continuously migrates onshore under persistent moderate wave
conditions, indicating a longer recovery but also a post-storm onshore
migration that is likely to end up with the bar welding to the upper
beach under persistent calm conditions, in line with downstate beach
transition schemes (Wright and Short, 1984; Ranasinghe et al., 2004).
shows the contribution (in %) of each term on the total Δ < Xs >and Based on this recovery duration, storm clusters are defined as a
Tr variance, together with their confidence level. Fig. 6c indicates that group of storms recurring in< 10 days. 13 such clusters are identified
storm impact depends predominantly (55%) on current storm energy. It within the 6-year period, with at least one per year. The overall impact
is a common outcome that wave conditions dominate the shoreline of clusters on shoreline location ranges from no substantial change to
response during storms (e.g. Yates et al., 2009; Davidson et al., 2013; Table 2
Castelle et al., 2015), with large intensities (i.e. D and/or Hs) resulting Cross-correlation coefficients between terms used in the multiple linear regression
in large impacts on shoreline, but here we show that previous analysis for storm impact Δ < Xs >(upper white panel) and recovery Tr (lower gray
panel) (Section 3.2).
conditions have a substantial role (37%) while modulation by tide
and sandbar play only a minor role (8% for tide and sandbar Waves Previous Sandbar Tide
altogether). In contrast, during recovery (Fig. 6d), it is almost the conditions
reverse: while current and previous wave conditions have a secondary Waves 0.09 –0.35 –0.24
importance (15% and 13%, respectively), tide and sandbar contribu- Previous conditions 0.13 0.15 0.10
tions rise to 45 and 23%, respectively. This clearly shows a different
Sandbar 0.01 0.04 0.27
forcing control on beach response between energetic, eroding condi-
tions and recovery periods. This suggests that in times of low energy, Tide 0.03 –0.13 –0.07
shoreline retreat could be reduced, as was observed in 2011.
271
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
Fig. 6. Multiple linear regression analysis for Δ < Xs >(left) and Tr (right). a) and b) illustrate the comparison between observed and reconstructed variables. Thick solid and thin solid
lines are 1:1 and linear regression (forced through the origin), respectively, while dashed lines indicate the 95% confidence levels. Lower panels c-d) describe the percentage of
reconstructed signal explained by each component during storms and recovery, respectively. Error bars show the 95% confidence levels.
16 m of recession. The cluster with the largest number of storms that of the sandbar, although statistically significant at the 95%
observed in Nov–Dec 2009 with a total energy of 7133 m2hr resulted confidence level, is weak (correlation coefficient ~ 0.2) in comparison
in 14 m erosion. However, a smaller cluster of 2 events with a to storm intensity and previous storm influence. Though it has been
cumulative energy of 5573 m2hr resulted in 11 m shoreline retreat, as observed elsewhere (e.g. Robin et al., 2007; Davidson and Turner,
this cluster includes the longest storm lasting 12 days in January 2009. 2009) that spring tides might enhance storm impact of the upper beach,
Fig. 8 shows the impact of storms Δ < Xs >ranked from one to five it is hard to conclude with our dataset. The shoreline proxy used in this
in the clusters. Note that the storm numbering here only depends on the study could also have an impact on the contribution of the tides during
occurrence sequence of the individual storms in the cluster, which storms. Similarly, the inner sandbar has only a limited influence on
means the first storm is not necessarily the most energetic. It is clear shoreline retreat, though one could expect that the closer the sandbar is
that the storm impact within a cluster decreases with storm rank. The to the shoreline, the more the inner sandbar will be coupled to the
influence of previous storms and the importance of recurrence is shoreline and plays its sheltering effect. For example, by limiting
discussed in the next section. incoming wave height (Masselink et al., 2006; Davidson and Turner,
2009; Almar et al., 2010) through breaking over the shallow crest. This
could result in large variability in the cross-shore shoreline location.
4. Discussion Given that this is a double-barred beach, a coupling between the inner
and outer sandbars could influence the effect of the inner bar on the
4.1. Role of sandbars and tides in modulating storm impact and recovery shoreline. For the post-storm period, Fig. 6d shows that both tide and
sandbar location affect substantially the recovery values Tr thereby
Results in Section 3.2 (Fig. 6c) show that the influence of tide and modulating the recovery duration.
sandbar during storms on the shoreline is not substantial in comparison The use of a linear regression for possible non-linear relationships
with present storm intensity and previous storm influence. The between the various parameters is foremost to identify the predominant
influence of tide and sandbar contribute 8% in total. The influence of parameters. To account for the linearity in the multiple regression
tide is not statistically significant (correlation coefficient < 0.1) and
272
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
Fig. 7. Ensemble-averaged (over 36 storms) evolution during post-storm recovery period for a) Hs, b) shoreline location< Xs >and c) sandbar location< Xb > from their location at the
end of the storm. Shaded zone stands for standard deviation.
Fig. 8. Cluster of storms. a) Cumulative storm impact and b) number of storms taken into account as a function of their rank in the cluster. Circles and triangles in a) illustrate average and
individual values, respectively. In a) offshore direction is traced by more positive values.
273
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
method used, several parameters were tested including the relative tide maximize coastal erosion at the scale of a winter season. This is what
range (RTR= TR/Hs) and hydrodynamic forcing index (HFI, Almar was observed along the west coast of Europe during the 2013/2014
et al., 2010). As the RTR increases, wave action becomes strongly winter along the west coast of Europe (Masselink et al., 2016) and
controlled by tidal level and sandbar location, or most probably a during the highest WEPA index recorded over at least the last 75 years
combination of both. Under such moderate wave conditions, a large (Castelle et al., 2017), and more recently during the 2015/2016 winter
tidal range will result in reducing the occurrence of surf-zone processes along the west coast of the US for one of the largest El Nino over the last
at the upper beach, and thus increase recovery duration. It will also 145 years (Barnard et al., 2017).
change the breaking intensity and occurrence over the bar which can
have a direct consequence on the fine threshold between erosion/ 5. Conclusions
accretion and no change, as observed by Almar et al. (2010). Stokes
et al. (2015) observed that at seasonal scale, the inclusion of tide Six years of video-derived shoreline and sandbar locations were
(through a modulation of incoming wave energy) improves the predic- collected at the meso- to macrotidal barred beach of Biscarrosse, SW
tion of shoreline change, and it is expected to even be truer at short/ France. Over 60 individual storms (~15 storms per year) were
event time scales, in particular the post-storm relaxation time. This is identified using 5% exceedance for Hs (Hs > 3.68 m) as the storm
mainly due to the modulation of swash characteristics by tide, as threshold. However, due to the non-existence of shoreline data during
observed by Guedes et al. (2011). 24 of these storms because of poor weather and video malfunctions,
only the remaining 36 storms were investigated in detail. Based on
4.2. Frequency of recurrence of storms and shoreline resilience these, the average storm recurrence is 27 days in winter, with 60% of
the storms recurring within 10 days. This large recurrence shows a
The fact that previous storm influence contributes significantly to strong seasonality in storm occurrence, also reflected in the shoreline
shoreline change, underlines the significance of the so-called beach and sandbar locations. Shoreline retreat is predominantly influenced by
memory effect (e.g. Turki et al., 2012; Reeve et al., 2014) where the current storm (55%), but the previous storm influence also plays a
shoreline response to events depends on the history of beach conditions significant role (37%). The modulating parameters such as the sandbar-
(e.g. see Splinter et al., 2014a). In line with previous studies (Yates to-shoreline distance and tides play only a secondary role (8%) in storm
et al., 2009; Castelle et al., 2014; Angnuureng, 2016), the first winter induced shoreline retreat. Antecedent storm conditions were also
storms drive the most pronounced erosion because the wave energy observed to reduce current storm impact, likely explained by the rapid
disequilibrium and erosion potential are large. During the rest of winter adjustment of the beach to a more energetic state when a storm occurs.
season, even if the beach is often exposed to severe storms, they do not With moderate wave energy during post-storm recovery, the
significantly erode the beach as the disequilibrium energy is smaller. It influence of the tidal range and the sandbar increases (23 and 45%,
should be noted that the correlation coefficient between the preceding respectively), with recovery duration increasing for larger tidal range
storm influence and storm impact was observed to be negative (−0.35, and larger distance between the sandbar and the shoreline. The
significant at the 95% level), which means that if the previous storm presence of a double sandbar on the meso- to macrotidal Biscarrosse
event is larger but closer to the current storm event, the erosion will be beach induces a threshold on wave energy (Almar et al., 2010) at the
less. If storm recurrence is long enough, individual storm impacts shore by height limitation due to breaking over the sandbar which
become independent as the beach has time to recover and reach its pre- modifies onshore wave energy and frequencies: the magnitude of
storm equilibrium. If the interval is sufficiently short, such as for storms shoreline change may be highly controlled by the conjugate effect of
in sequences described in Section 3.3, only the previous storm appears sandbar and tide. These results argue in favor of integrating sandbar
to have a destabilizing effect on the beach while the subsequent storms and tide effects in shoreline equilibrium models, especially the way in
decreasingly impact on the beach. This is consistent with Dissanayake which they influence the complex beach recovery process, which could
et al. (2015) who found at Formby Beach (UK) that the largest erosion substantially improve model performance at longer timescales.
was always observed for the first storm, because the beach had Analysis of post-storm beach recovery shows that the beach recovers
sufficient time to recover fully from the previous storm season. within 9 days of individual storms. With respect to storm clusters, the
Our observations are in line with Coco et al. (2014) and Splinter first storms result in the highest erosion. This agrees with equilibrium-
et al. (2014a) who demonstrated that a sequence of storms does not based approaches where storms are less and less effective in eroding the
necessarily result in cumulative erosion, although frequent sequences beach as the beach progressively reaches a new equilibrium with the
can slightly affect shoreline resilience (Dissanayake et al., 2015). These prevailing wave conditions. The beach response depends on the beach
results suggest a possible relation between beach changes on episodic ‘memory’. These results suggest the existence of beach resilience
and seasonal timescales, and that the frequency of storm recurrence and interactions at different timescales. Beach resilience is weaker when
the storm frequency change over time (e.g. seasonal, interannual, storms are infrequent (in summer) and vice versa in winter. These
climate change) are of some importance in assessing beach equilibrium results also show that a weaker single storm could have a larger impact
and evolution. When storm sequences are frequent in winter, only the than stronger storm occurring in the middle of storm sequence,
first storm causes severe impact while the rest of the storms in the illustrating the key role of the temporal evolution of not only the storm
sequence have minimal effect. Considering variable recurrence fre- intensity but also their frequency of recurrence when considering beach
quency, longer storm intervals enhance storm impact. Storm impact resilience.
will also change if the frequency of storms (i.e. storminess) evolves
under changing climate, or under regional modes of climate variability Acknowledgements
(e.g. on the west coast of Europe the North Atlantic Oscillation, NAO;
northward of 52°N (Hurrell, 1995), and the West Europe Pressure The first author is co-funded by SCAC (French embassy in Ghana)
Anomaly, WEPA, southwards of 52°N (Castelle et al., 2017)). Such and ARTS-IRD programs. Authors acknowledge the Region Aquitaine
natural modes of climate variability can cause outstanding series of for financially supporting the installation of the video system at
storm events such as that observed during the 2013/2014 winter in Biscarrosse. This research has received support from French grant
Western Europe (Masselink et al., 2016), and drive some interannual through ANR COASTVAR: ANR-14-ASTR-0019. RR is supported by
change in beach and sandbar behavior (Masselink et al., 2014). It is the AXA Research fund and the Deltares Harbour, Coastal and Offshore
hypothesized that our findings on the timescale of storms may apply to Engineering Research Programme ‘Bouwen aan de Kust’. BB is supported
the interannual timescales, where a winter with extreme storminess by French “Agence Nationale de la Recherche” through project CHIPO
following a few years of reasonably fair winter wave conditions may (ANR-14-ASTR-0004-01).
274
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
References shoreline. Coast. Eng. 57, 959–972.
Holland, K.T., Holman, R.A., Lippmann, T.C., 1997. Practical use of video imagery in
neareshore oceanographic field studies. IEEE J. Ocean. Eng. 22 (1), 81–92.
Aarninkhof, S.G.J., Turner, I.L., Dronkers, T.D.T., Caljouw, M., Nipius, L., 2003. A video- Holman, R., Haller, M., 2013. Remote sensing of the nearshore. Annu. Rev. Mar. Sci. 113,
based technique for mapping intertidal beach bathymetry. Coast. Eng. 49 (4), 5–95.
275–289. Holman, R.A., Stanley, J., 2007. The history and technical capabilities of Argus. Coast.
Almar, R., Castelle, B., Ruessink, B.G., Senechal, N., Bonneton, P., Marieu, V., 2009. High- Eng. 54, 477–491.
frequency video observation of two nearby double-barred beaches under high-energy Holman, R.A., Sallenger, A.H., Lippmann, T.C., Haines, J.W., 1993. The application of
wave forcing. J. Coast. Res. SI 56 (2), 1706–1710. video image processing to the study of nearshore processes. Oceanography 6, 3.
Almar, R., Castelle, B., Ruessink, G., Senechal, N., Bonneton, P., Marieu, V., 2010. Two Hurrell, J.W., 1995. Decadal trends in the North Atlantic oscillation: regional
and three-dimensional double-sandbar system behaviour under intense wave forcing temperatures and precipitation. Science 269 (5224), 676–679.
and a meso–macro tidal range. Cont. Shelf Res. 30 (7), 781–792. Karunarathna, H., Pender, D., Ranasinghe, R., Short, A.D., Reeve, D.E., 2014. The effects
Angnuureng, D.B., 2016. Shoreline Response to Multiscale Oceanic Forcing from Video of storm clustering on beach profile variability. Mar. Geol. 348, 103–112.
Imagery. (PhD thesis) Université de Bordeaux (181 pp.). Lippmann, T.C., Holman, R.A., 1989. Quantification of sandbar morphology: a video
Ba, A., Senechal, N., 2013. Extreme winter storm versus summer storm: morphological technique based on wave dissipation. J. Geophys. Res. 94, 995–1011.
impact on a sandy beach. J. Coast. Res. SI 65, 648–653. Lippmann, T., Holman, R., 1990. The spatial and temporal variability of sandbar
Barnard, P.L., Hoover, D., Hubbard, D.M., Snyder, A., Ludka, B.C., Kaminsky, G.M., morphology. J. Geophys. Res. 95, 11575–11590.
Ruggiero, P., Gallien, T., Gabel, L., McCandless, D., Weiner, H.M., Cohn, N., Madsen, A.J., Plant, N.G., 2001. Intertidal beach slope predictions compared to field data.
Anderson, D.L., Serafin, K.A., 2017. Extreme oceanographic forcing and coastal Mar. Geol. 173, 121–139.
response due to the 2015-2016 el Niño. Nat. Commun. 8, 14365. Maspataud, A., Ruz, M.-H., Hequette, A., 2009. Spatial variability in post-storm beach
Birrien, F., Castelle, B., Dailloux, D., Marieu, V., Rihouey, D., Price, T.D., 2013. Video recovery along a macrotidal barred beach, southern North Sea. J. Coast. Res. 56,
observation of megacusp evolution along a high-energy engineered sandy beach: 88–92.
Anglet, SW France. J. Coast. Res. SI 65, 1727–1732. Masselink, G., Short, A.D., 1993. The effect of tide range on beach morphodynamics and
Boak, E.H., Turner, I.L., 2005. Shoreline definition and detection: a review. J. Coast. Res. morphology: a conceptual beach model. J. Coast. Res. 9 (3), 785–800.
21 (4), 688–703. Masselink, G., Kroon, A., Davidson-Arnott, R.G.D., 2006. Morphodynamics of intertidal
Butel, R., Dupuis, H., Bonneton, P., 2002. Spatial variability of wave conditions on the bars in wave-dominated coastal settings-a review. Geomorphology 73, 33–49.
French Aquitanian coast using in-situ data. J. Coast. Res. SI 36, 96–108. Masselink, G., Austin, M., Scott, T., Poate, T., Russell, P., 2014. Role of wave forcing,
Castelle, B., Bonneton, P., Dupuis, H., Senechal, N., 2007a. Double bar beach dynamics on storms and NAO in outer bar dynamics on a high-energy, macro-tidal beach.
the high-energy meso-macrotidal French Aquitanian coast : a review. Mar. Geol. 245, Geomorphology 226, 76–93.
141–159. Masselink, G., Castelle, B., Scott, T., Dodet, G., Suanez, S., Jackson, D., Floc'h, F., 2016.
Castelle, B., Turner, I.L., Ruessink, B.G., Tomlinson, R.B., 2007b. Impact of storms on Extreme wave activity during 2013/2014 winter and morphological impacts along
beach erosion: Broadbeach (Gold Coast, Australia). J. Coast. Res. SI 50, 534–539. the Atlantic Coast of Europe. Geophys. Res. Lett. 43 (5), 2135–2143.
Castelle, B., Marieu, V., Bujan, S., Ferreira, S., Parisot, J.P., Capo, S., Senechal, N., Mendoza, E.T., Jimenez, J.A., Mateo, J., 2011. A coastal storms intensity scale for the
Chouzenoux, T., 2014. Equilibrium shoreline modelling of a high energy meso- Catalan sea (NW Mediterranean). Nat. Hazards Earth Syst. Sci. 11, 2453–2462.
macro-tidal multiple-barred beach. Mar. Geol. 347, 85–94. Morton, R.A., Leacht, M.P., Painet, J.G., Cardozat, M.A., 1993. Monitoring beach changes
Castelle, B., Marieu, V., Bujan, S., Splinter, K.D., Robinet, A., Senechal, N., Ferreira, S., using GPS surveying techniques. J. Coast. Res. 9 (3), 702–720.
2015. Impact of the winter 2013-2014 series of severe Western Europe storms on a Pape, L., Ruessink, B.G., 2008. Multivariate analysis of nonlinearity in sandbar behavior.
double-barred sandy coast: beach and dune erosion and mega cusp embayments. Nonlinear Process. Geophys. 15, 145–158.
Geomorphology 238, 135–148. Peron, C., Senechal, N., 2011. Dynamic of a meso to macro-tidal double barred beach:
Castelle, B., Dodet, G., Masselink, G., Scott, T., 2017. A new climate index controlling inner bar response. J. Coast. Res. SI 64, 120–124.
winter wave activity along the Atlantic coast of Europe: the West Europe pressure Pianca, C., Holman, R., Siegle, E., 2015. Shoreline variability from days to decades:
anomaly. Geophys. Res. Lett. 44. http://dx.doi.org/10.1002/2016GL072379. results of long-term video imaging. J. Geophys. Res. Oceans 120, 2159–2178. http://
Coco, G., Senechal, N., Rejas, A., Bryan, K.R., Capo, S., Parisot, J.P., Brown, J.A., dx.doi.org/10.1002/2014JC010329.
MacMahan, J.H.M., 2014. Beach response to a sequence of extreme storms. Plant, N.G., Holman, R.A., 1997. Intertidal beach profile estimation using video images.
Geomorphology 204, 493–501. Mar. Geol. 140, 1–24.
Corbella, S., Stretch, D.D., 2012. Shoreline recovery from storms on the east coast of Plant, N., Holman, R., 1998. Extracting morphologic information from field data. Coast.
southern Africa. Nat. Hazards Earth Syst. Sci. 12, 11–22. Eng. Proc. 1 (26). http://dx.doi.org/10.9753/icce.v26.
Davidson, M.A., Turner, I.L., 2009. A behavioral template beach profile model for Plant, N.G., Aarninkhof, S.G.J., Turner, I.L., Kingston, K.S., 2007. The performance of
predicting seasonal to interannual shoreline. J. Geophys. Res. 114, F01020. http:// shoreline detection models applied to video imagery. J. Coast. Res. 23 (3), 658–670.
dx.doi.org/10.1029/2007JF000888. Ranasinghe, R., Symonds, G., Black, K., Holman, R., 2004. Morphodynamics of
Davidson, M.A., O'Hare, T.J., George, K.J., 2008. Tidal modulation of incident Wave intermediate beaches: a video imaging and numerical modelling study. Coast. Eng.
Heights: fact or fiction? J. Coast. Res. 24 (2), 151–159. 51, 629–655.
Davidson, M.A., Splinter, K.D., Turner, I.L., 2013. A simple equilibrium model for Ranasinghe, R., Holman, R., de Schipper, M.A., Lippmann, T., Wehof, J., Minh Duong, T.,
predicting shoreline change. Coast. Eng. 73, 191–202. Roelvink, D., Stive, M.J.F., 2012. Quantification of nearshore morphological recovery
Davis, R.A., 1985. Beach and nearshore zone. In: DAVIS, R.A. (Ed.), Coastal Sedimentary time scales using Argus video imaging: Palm Beach, Sydney and Duck, NC. Coast.
Environments. Springer-Verlag, New York, pp. 379–444. Eng. Proc. 1 (33), 24.
Dissanayake, P., Brown, J., Karunarathna, H., 2015. Impacts of storm chronology on the Rangel-Buitrago, N., Anfuso, G., 2011. An application of Dolan and Davis (1992)
morphological changes of the Formby beach and dune system, UK. Nat. Hazards classification to coastal storms in SW Spanish littoral. In: J. Coast. Res., SI64
Earth Syst. Sci. 15, 1533–1543. (Proceedings of the 11th International Coastal Symposium), Szczecin, Poland, pp.
Dolan, R., Davis, R., 1992. An intensity scale for Atlantic coast northeast storms. J. Coast. 0749-0208.
Res. 8 (4), 840–853. Reeve, D.E., Pedrozo-Acuña, A., Spivack, M., 2014. Beach memory and ensemble
Dorsch, W., Newland, T., Tassone, D., Tymons, S., Walker, D., 2008. A statistical approach prediction of shoreline evolution near a groyne. Coast. Eng. 86, 77–87.
to modeling the temporal patterns of ocean storms. J. Coast. Res. 24 (6), 1430–1438. Robin, N., Levoy, F., Monfort, O., 2007. Bar morphodynamic behaviour on the ebb delta
Douglas, B.C., Crowell, M., 2000. Long-term shoreline position prediction and error of a macrotidal inlet (Normandy, France). J. Coast. Res. 23 (6), 1370–1378 (West
propagation. J. Coast. Res. 16 (1), 145–152. Palm Beach (Florida), ISSN 0749–0208).
Fenster, M.S., Dolan, R., Morton, R.A., 2001. Coastal storms and shoreline change: signal Senechal, N., Coco, G., Castelle, B., Marieu, V., 2015. Storm impact on the seasonal
or noise? J. Coast. Res. 17 (3), 714–720. shoreline dynamics of a meso- to macrotidal open sandy beach (Biscarrosse, France).
Ferreira, O., 2005. Storm groups versus extreme single storms: predicted erosion and Geomorphology 228, 448–461.
management consequences. J. Coast. Res. SI 42, 221–227. Smit, M.W.J., Aarnikhof, S.G.J., Wijberg, K.M., Gozalez, M., Kingston, K.S., Southgate,
Flater, D., 2010. www.wXtide32.com (Last accessed on 25/01/2016). H.N., Ruessink, G., Holman, R.A., Siegle, E., Davidson, M., Medina, R., 2007. The role
Frazer, L.N., Anderson, T.R., Fletcher, C.H., 2009. Modeling storms improves estimates of of video imagery in predicting daily to monthly coastal evolution. Coast. Eng. 54,
long-term shoreline change. Geophys. Res. Lett. 36, L20404. 539–553.
Gallagher, E.L., Elgar, S., Guza, R.T., 1998. Observations of sandbar evolution on a Splinter, K.D., Carley, J.T., Golshani, A., Tomlinson, R., 2014a. A relationship to describe
natural beach. J. Geophys. Res. 103, 3203–3215. the cumulative impact of storm clusters on beach erosion. Coast. Eng. 83, 49–55.
Gallagher, E.L., MacMahan, J.H., Reniers, A.J.H.M., Brown, J., Thornton, E.B., 2011. Splinter, K.D., Turner, I.L., Davidson, D.A., Barnard, P., Castelle, B., Oltman-Shay, J.,
Grain size variability on a rip channeled beach. Mar. Geol. 287, 43–53. 2014b. A generalized equilibrium model for predicting daily to inter-annual shoreline
Genz, A.S., Fletcher, C.H., Dunn, R.A., Frazer, L.N., Rooney, J.J., 2007. The predictive response. J. Geophys. Res. Earth Surf. 119, 1936–1958.
accuracy of shoreline changes rate methods and alongshore beach variation on Maui, Stockdon, H.F., Holman, R.A., Howd, P.A., Sallenger, A.H., 2006. Empirical
Hawaii. J. Coast. Res. 23 (1), 87–105. parameterization of setup, swash and runup. Coast. Eng. 53, 573–588.
Grasso, F., Michallet, H., Barthélemy, E., Certain, R., 2009. Physical modeling of Stokes, C., Davidson, M., Russell, P., 2015. Observation and prediction of three-
intermediate cross-shore beach morphology: transients and equilibrium states. J. dimensional morphology at a high-energy macrotidal beach. Geomorphology 243,
Geophys. Res. 114, C09001. 1–13.
Guedes, R.M.C., Calliari, L.J., Holland, K.T., Plant, N.G., Pereira, P.S., Alves, F.N.A., 2011. Thornton, E.B., MacMahan, J.H., Sallenger Jr., A.H., 2007. Rip currents, mega-cusps, and
Short-term sandbar variability based on video imagery: comparison between eroding dunes. Mar. Geol. 240, 151–167.
time–average and time–variance techniques. Mar. Geo. 289, 122–134. Tolman, H.L., 1991. A third generation model for wind waves on slowly varying,
Hansen, J.E., Barnard, P.L., 2010. Sub-weekly to interannual variability of a high-energy unsteady and inhomogeneous depths and currents. J. Phys. Oceanogr. 21, 782–797.
275
D.B. Angnuureng et al. Geomorphology 290 (2017) 265–276
Turki, I., Medina, R., Gonzalez, M., 2012. Beach memory. In: McKee Smith, Jane (Ed.), van Gent, M.R.A., van Thiel de Vries, J.S.M., Coeveld, E.M., de Vroeg, J.H., Van de Graaff,
Proceedings of the 33rd International Conference, World Scientific, Santander, J., 2008. Large-scale dune erosion tests to study the influence of wave periods. Coast.
Spain, . Eng. 55, 1041–1051.
Turner, I., Leyden, V., Symonds, G., McGrath, J., Jackson, A., Jancar, T., Aarninkhof, S., Wright, L.D., Short, A.D., 1984. Morphodynamic variability of surf zones and beaches: a
Elshoff, I., 2001. Comparison of observed and predicted coastline changes at the Gold synthesis. Mar. Geol. 56, 93–118.
Coast artificial (surfing) reef, Sydney, Australia. In: Proceedings of International Wright, L.D., Short, A.D., Green, M.O., 1985. Short-term changes in the morphodynamic
Conf. on Coast. Eng., Sydney, . states of beaches and surf zones: an empirical predictive model. Mar. Geol. 62,
van de Lageweg, W.I., Bryan, K.R., Coco, G., Ruessink, B.G., 2013. Observations of 339–364.
shoreline–sandbar coupling on an embayed beach. Mar. Geol. 344, 101–114. Yates, M.L., Guza, R.T., O'Reilly, W.C., 2009. Equilibrium shoreline response:
van Enckevort, I.M.J., Ruessink, B.G., 2001. Effect of hydrodynamics and bathymetry on observations and modeling. J. Geophys. Res. 114.
video estimates of nearshore sandbar position. J. Geophys. Res. 106 (C8), Zhang, K., Douglas, B., Leatherman, S., 2002. Do storms cause long-term beach erosion
16969–16979. along the U.S. east barrier coast? J. Geol. 110 (4), 493–502.
276