Marine Geology 284 (2011) 217–229

Contents lists available at ScienceDirect

Marine Geology
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / m a r g e o

Equilibrium responses of cliffed coasts to changes in the rate of sea level rise
Andrew D. Ashton a,⁎, Mike J.A. Walkden b, Mark E. Dickson c
a
Coastal Systems Group, Department of Geology and Geophysics, Woods Hole Oceanographic Institution, 360 Woods Hole Rd MS #22, Woods Hole, MA 02543, USA
b
Royal Haskoning, Stratus House, Emperor Way, Exeter, Devon, EX1 3QS, UK
c
School of Environment, The University of Auckland, Private Bag 92019, Auckland, New Zealand

a r t i c l e i n f o a b s t r a c t

Article history: Basic formulae have long been used to predict the effects of sea-level rise on coastal recession; for instance,
Received 17 September 2010 the geometric ‘Bruun rule’ (and its modifications) has often been applied to sandy coasts, both low-lying and
Received in revised form 21 January 2011 steep. However, the behavior of rocky coasts, whether strongly or poorly lithified, should be significantly
Accepted 31 January 2011
different than that of sandy coasts given that rocky coast evolution depends upon the irreversible breakdown
Available online 9 February 2011
of rock, whereas sandy and depositional systems are controlled by the transport (and related transport
Communicated by J.T. Wells gradients) of mobile sediment. Here, we investigate the basis of a modeled relationship which suggests (with
a number of caveats) that the equilibrium soft-rock cliff recession rate can be estimated by the square root of
Keywords: the relative change in sea-level rise rate. Although this relationship was derived using the numerical model
cliffed coasts SCAPE (Soft Cliff And Platform Erosion), which simulates a broad soft-rock cliffed coastal system driven by
rocky coasts stochastic environmental forces, here we show that a simplified modeling approach also reproduces the
coastal erosion relationship. We then extend this approach to develop a general theoretical framework within which it is
sea-level rise possible to consider the potential responses of the different types of cliffed coasts to changes in the rate of sea
Bruun rule
level rise. Although a wide variety of processes affect different coastal settings, this framework demonstrates
dynamic equilibrium
how the strength and the nature of feedbacks within cliffed system control their response to sea-level rise.
This suggests that cliffed environments controlled by different processes can still respond in similar ways to
changes in the rate of sea-level rise. Most rocky coasts would be expected to respond as a damped, or ‘negative
feedback’ system between the extremes of a ‘no feedback’ system that is unresponsive to sea-level rise rate
and an ‘instant response’ system characterized by a linear response similar to the Bruun rule. This framework
suggests that a potential ‘inverse feedback’ case could also exist, in which increased rates of sea-level rise
reduce the rate of coastal recession. In almost all cases, it is apparent that cliffed coast response to sea-level
rise depends not only upon the total elevation change of sea level, but on the rate of the sea-level rise. These
theoretical investigations and the classifications presented provide a framework to understand the behavior
of systems affected by a wide array of processes, and provide expectations that can be tested using more
complex models of cliffed coast evolution in a variety of environments, whether sandy or rocky, hard or soft.
© 2011 Published by Elsevier B.V.

1. Introduction
How coastlines respond to sea-level rise remains one of the
cornerstone challenges confronting coastal geomorphologists and
engineers (Dubois, 2002). It has long been conceptualized that, as a
result of the dynamic processes occurring at the coast, rising sea levels
cause significantly more coastline change than passive inundation
alone. However, we have relatively limited physical understanding of
exactly how coastal regions respond to changes in sea level, and there
remains significant uncertainty regarding any predictions of the
response of coasts to the sea-level accelerations predicted for the
upcoming century and beyond (e.g. Meehl et al., 2007; Rahmstorf,
2007). This lack of predictive ability arises in part from the complexity
⁎ Corresponding author. Tel.: +1 508 289 3751
E-mail address: aashton@whoi.edu (A.D. Ashton).
of the coastal environment, but also because predicted rates of sea-
level rise over the next century are unprecedented during historic
times, making it difficult to extrapolate historic measurements into
the future. Coastal managers rely on predictions of future shoreline
position to assess hazards and risk, predictions which may be based
on experience, statistics (e.g. Hall et al., 2002; Hapke and Plant, 2010),
or process understanding (e.g. Trenhaile, 2010). At present, our lack of
knowledge of the relationships between the rates of sea-level rise and
coastal recession leaves significant uncertainty in all future
predictions.
The most commonly applied relationships between sea-level rise
and shoreline recession follow the concepts of equilibrium profile
geometry and sediment conservation, as introduced by Bruun (1962)
and most often applied to low-lying coasts. ‘Bruun rule’ relationships
are based upon mass conservation principles and might apply on
cliffed coasts composed of cohesive but unconsolidated sediments
(e.g. sand-gravel cliffs or bluffs). However, different relationships

0025-3227/$ – see front matter © 2011 Published by Elsevier B.V.

doi:10.1016/j.margeo.2011.01.007
218 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229
should be expected to represent the evolution of cliffed shores, in
particular rocky and consolidated coasts whose evolution is controlled
by break down of in situ material through erosion and weathering.
Naylor et al. (2010) emphasize the need for general theoretical sea-
level-rise relationships for rocky coasts, using the general definition of
rocky coasts being predominantly erosional landforms.
Here, we investigate the potential equilibrium responses of cliffed
coasts to an increased rate of sea-level rise. First, we focus on a
recently proposed relationship between sea-level rise rate and
erosion rate by Walkden and Dickson (2008), appropriate for soft
rock coast evolution. The second part of the paper investigates cliffed
coasts more broadly, using the characteristic erosional responses to
changes in sea-level rise rate to compare and contrast the evolution of
sandy, consolidated, soft rock, and hard rock coasts.
1.1. Geometric sea-level rise relationships
The evolution of a coast comprised of unlithified, non-cohesive, or
‘sandy’ sediment is controlled by gradients in sediment transport,
both along- and across-shore. Observations suggest that, over long
timescales, the cross-shore shoreface, extending to the depth where
waves no longer affect sediment transport, attains an equilibrium
shape (Dean, 1991). Assuming that no sandy sediment is lost or
gained from the active profile in the cross-shore direction and that
the shoreface maintains an equilibrium shape, the conservation of
mass leads to a relationship where the landward rate of movement of
a
ΔY ΔZ
ocean
αsf
land
b
αr
land
c d
αr
ΔY
ΔZ
ocean
αsf
land
Fig. 1. Geometric relationships of coastal recession, representing (a) the original ‘Bruun rule’, (b) a modification for a barrier environment, and (c) application to a steep coast. This
relationship may also apply to rocky coasts when the entire active profile, (d), is covered by mobile sediment.
the shoreline, ε (distance/time), is a linear function of the sea-level
rise itself (S, distance/time):
S
ε= ; ð1Þ
αe
as proposed by Bruun (1962) (Fig. 1a). Here we take αe to represent
the effective slope controlling sea-level rise response.
Which geometric boundaries accurately represent the effective
slope, αe, has been an issue of significant debate. Although some
applications (Bruun, 1962; Zhang et al., 2004) suggest that the
effective slope should be the shoreface slope, αsf (Fig. 1), or, for low-
sloped coasts, some other modification accounting for barrier
geometry (Dean and Maurmeyer, 1983), recent analytic research
provides a rigorous demonstration that long-term equilibrium
requires that the coastal recession rate follow the regional upland
slope, αr (Wolinsky, 2009; Wolinsky and Murray, 2009) (Fig. 1b,c).
Equilibrium sea-level response based upon mass conservation must
follow the trajectory of regional slope—any other response is a
transient.
1.2. Application of geometric relationships to cliffed coasts
The Bruun rule has also been adapted and applied to cliffed coasts
(Fig. 1c). Numerical experiments by Roy et al. (1994) demonstrate
that regional slope controls the long-term trajectory of sandy coastal
ΔY ΔZ
ocean
αsf
αr
αr
ΔY
ΔZ
ocean
αsf
sand
rock
 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229
evolution. Bray and Hooke (1997) modified Eq. (1) to account for the
fraction of cliff material that may be too fine-grained to be compatible
with the beach sediment:
S by local sea-level effects (e.g. Peltier, 2004; Mitrovica et al., 2009).
ε= ; ð2Þ
Pαc
where P is the proportion of compatible littoral material in the cliff
and the effective angle, αc, is determined by the cliff height and
shoreface geometry (Fig. 1c) (see also Dean and Maurmeyer, 1983).
Bray and Hooke (1997) clarify that this relationship assumes that
there is always sufficient sediment in the active profile. Hence, Eq. (2)
is not intended for a truly rocky coast but it could apply to a coast
composed of rocky cliffs if the active, wave-affected profile is entirely
blanketed by mobile sediment, including the cliff toe (e.g. Fig. 1d).
Although Eq. (2) may reasonably apply over the short term, it does
not represent a long-term equilibrium. Wolinsky and Murray (2009)
demonstrate how, over time, if the slope of the shoreline trajectory is
less than that of the upland slope, the cliff height must increase. As this
height increases, sediment contributions from the cliff for a given
landward movement of the shoreline increase, eventually compensat-
ing for the sediment deficit. The eventual long-term equilibrium
response for a sandy system follows the regional upland slope where
the effective slope, αe, is the regional upland slope, αr, regardless of the
fraction of compatible sediment in the cliff face. If there is no constant
regional upland slope, an equilibrium state does not exist for a
sediment-dominated coast. This contrasts with rocky, low-sediment-
volume coasts, which can have a rate-based equilibrium response
independent of regional upland slopes, as discussed below.
1.3. Models of rocky coast response to sea-level rise
Although application of equilibrium geometry relationships may
be useful in sediment-rich environments, the underlying concepts of
these linear ‘Bruun-type’ rules are not transferable to most rocky
coasts. If the sediment produced by erosion of a rocky coast is either
too fine-grained to remain within the wave-dominated coastal zone
or is quickly removed by alongshore sediment transport, then
shoreline recession is dominantly controlled by the rate of removal
or conversion of rock itself and not by sediment transport gradients or
sediment conservation. Also, unlike sandy coasts, rocky profiles
cannot accrete. As an analog to terrestrial environments, rocky coasts
are either detachment-limited or weathering-limited, whereas sandy
coasts are transport-limited (Howard, 1994a,b). Because of these
differences, there is little reason to expect rocky coasts to attain an
equilibrium geometry independent of sea-level rise or erosion rate, as
supposed by Bruun-type relationships.
A number of models of hard rock coast evolution have been presented
(e.g. Sunamura, 1977, 1992), some of which consider the coastal response
to sea-level rise. Sunamura (1988) provided one such model discussed
later within this paper. Cinque et al. (1995) used a phenomenological
model to investigate the morphologic interactions between changes in
sea-level rise rate, erosion rate, and antecedent geometry. Investigating
the importance of inheritance in shore profile evolution, Trenhaile (2001)
modeled the evolution of rocky coasts exposed to regimes of differing sea
levels. In these experiments, sea level was held constant at different
relative levels to the coast, and the effects of rise itself were not specifically
investigated. These modeling results suggest that previous sea-level high-
stands may overprint subsequent high-stands, with the possibility that
the modern coast is a palimpsest of previous conditions. More recent
modeling of clay coasts (Trenhaile, 2010) investigates the response to a
100-year sea-level rise after thousands of years of stable sea level. The
formation and persistence of marine terraces due to Pleistocene sea-level
oscillations was studied using a simple model by Anderson et al. (1999),
who also explored the importance of terrestrial processes in eventually
degrading preserved terraces.
219
In locations where coastal cliff recession rates are recorded over
historic times, it is difficult to determine what portion of those changes are
due to ongoing processes and which may be attributable to the 20th
century acceleration in sea-level rise (Church and White, 2006) mitigated
Walkden and Dickson (2008) used numerical simulations to suggest a
relationship for estimating how accelerated sea-level rise could increase
recession of soft rock coasts; Appeaning Addo et al. (2008) demonstrated a
practical application of this relationship along the cohesive cliffed coast of
Ghana near Accra. This model-derived relationship is discussed in more
detail later in the paper. An understanding of these responses will be
important for coastal managers to plan for coastal change over the coming
decades and century.
1.4. Outline
In this paper, we examine the potential responses of cliffed coasts
to changes in sea-level rise in two distinct manners. First, we
specifically examine the relationship presented by Walkden and
Dickson (2008) derived using the SCAPE numerical modeling tool.
We demonstrate that this relationship is robust in that it may also be
obtained using a simpler modeling approach, and, moreover, can be
derived analytically from the underlying principles. This analysis
pertains only to the process formulations derived using the SCAPE
model, which is applicable to soft-rock coasts; however, the analysis
demonstrates that model simplification can be a powerful tool in
understanding behavioral responses from underlying processes,
providing theoretical insight and basic predictions (Carpenter, 2003;
Murray, 2003). Expanding upon these insights, the second portion of
the paper applies a temporal scaling framework to different types of
cliffed coasts, revealing the potential forms and end-member
examples of the relationships between sea-level rise and cliffed
coast erosion. Using this framework, we are able to investigate
previously presented erosion relationships, demonstrating many
cliffed environments can respond to sea-level rise in similar
manners, even if the environments and the driving processes are
significantly different.
2. Predictions from the SCAPE numerical model
2.1. SCAPE model overview
The Soft Cliff And Platform Erosion (SCAPE) numerical modeling
tool incorporates a hybrid of process-based and behavioral modules.
The model is able to represent many coupled alongshore profiles (e.g.
Dickson et al., 2007a), but here we focus on the evolution of a single
two-dimensional cross-shore profile. A map of the main modules can
be seen in Fig. 2a, along with a schematic of the profile and an erosion
calculation; details of the model are provided by Walkden and Hall
(2005) and Walkden and Dickson (2008). As this model focuses on
long-term coastal evolution, spanning temporal scales of years to
millennia, and includes a broad system (including platform, beach,
waves, tides, cliff, and engineering interventions), the process
descriptions are relatively abstract.
Erosion of the profile is calculated with the formulation:
dy 13 = 4 3 = 2 −1
=H T K f1 ðf3 ðt Þ−zÞ tanðf2 ðzÞÞ; ð3Þ
dt
where horizontal and vertical dimensions are y and z respectively
(Fig. 1), t is time, Hb is the breaking wave height, T is the wave period,
and K is a calibration term representing rock strength and several
hydrodynamic constants (units m9/4 s2/3, see Kamphuis, 1987, and
Walkden and Hall, 2005). The dimensionless distribution of soft rock
erosion under a breaking wave field, f1, was derived by Walkden and
Hall (2005) from physical model tests of Skafel (1995). The tidal
220 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229

b
3

2 MHWS
Height above Mean Sea Level (m)

1 Integration through
tidal cycle

0 MSL
Integrated
erosive
−1 potential

Erosion under surf zone (f1)

−2 Shore profile (f2) MLWS

Tidal oscillation (f3)
Depth at wave breaking

Fig. 2. (a) Schematic of the processes and interactions represented within the SCAPE model. (b) Sample of the integrated wave erosion potential over a tidal cycle for given wave
characteristics. Figures after Walkden and Dickson, 2008.

variation in water level, f3, is represented as a sinusoid about mean sea that the models produced steeper profile shapes with higher rates of
level, z. Within the model, erosion is linearly dependent upon the local sea-level rise (Fig. 3b), a phenomenon attributed to a shortening of
slope, f2, as suggested by Kamphuis (1987) based upon the wave the period over which any elevation in the rock profile is exposed to
energy dissipation rate in the surf zone. Sea level rise is implemented wave attack.
as a shifting frame of reference.
2.3. Sea-level rise relationship
2.2. Development of equilibrium profiles
Surprisingly, SCAPE simulations run across a wide range of model
Whereas previous studies utilizing the SCAPE model were parameter space, including variations in wave height, period, tidal
generally site-specific (Walkden and Hall, 2005; Dickson et al., range, and rock strength, revealed that a simple expression could be
2007a), more recently, Walkden and Dickson (2008) applied SCAPE used to relate the rate of sea-level rise and the equilibrium recession
to explore the dynamics of sea-level rise over centennial scales for rate (Walkden and Dickson, 2008):
generic soft rock and rock/beach shore profiles (Fig. 3). In particular,
they explored equilibrium conditions, i.e. the profile shapes and sffiffiffiffiffi
S2
recession rates associated with model shores that had sufficient time ε2 = ε1 ; ð4Þ
to fully adapt to a changed rate of sea-level rise. They demonstrated S1
 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229 221

a profile development b equlibrium profile shape

4

2
elevation above MSL (m)

elevation above MSL (m)

0
0

−2
−5 S (mm yr-1)

−4
2

−6 4
−10
6
16 14 12 10 8
−8
0 2 4 6 8 10 0 500 1000 1500
offshore distance (km) offshore distance (m)

c 3
d 1

2.5
0
elevation above MSL (m)
2
−1
ε2
1.5
ε1
−2
1

Parameter test results −3

0.5
ε2/ε1= (S2/S1)½

0 −4
0 2 4 6 8 0 0.002 0.004 0.006 0.008 0.01
S2/S1 Slope / S½

Fig. 3. SCAPE model results for a low-volume beach. (a) Shore profiles plotted relative to a rising mean sea-level, showing self-parallel retreat and development of (b) equilibrium
profiles whose slope depends on the rate of sea-level rise. (c) Relative equilibrium erosion rate as relative sea-level rate changes for ensemble model runs with different parameters
for forcing conditions and rock properties (points), plotted with fit of ε2 = ε1 = ðS2 =S1 Þ =2 (r2 = 0.96) (see Walkden and Dickson, 2008). (d) Slope divided by square root of sea level
1

rise for all parameter tests plotted in (c).

predicting that the future equilibrium recession rate (ε2) of a low-
volume soft-rock coast is a simple function of the prior equilibrium
recession rate (ε1) and the prior and future rates of sea-level rise (S1
and S2 respectively; Fig. 3c). This expression was found to hold for
profiles that included a beach whose volume was below a threshold
level (determined to be b30 m3/m for the base model parameters).
The relationship is therefore not expected to hold if alongshore
gradients in sediment transport resulted in a thick veneer of beach
sediments over the rock profile.
3. Obtaining the SCAPE sea-level rise relationship through simpler
approaches
The SCAPE modeling tool was developed to represent specific
coastal sites to support coastal management, and consequently it
provides descriptions of a reasonably large number of processes
and interrelationships within a complex coastal system (Fig. 2). The
multiple processes within SCAPE, its feedback pathways, and the
fact that Eq. (3) represents a combination of analysis, empiricism,
and laboratory data, tend to obscure the origin of the strikingly
simple relationship given by Eq. (4). Here, we present a simpler
numerical model, based upon the underlying SCAPE model
relationships, that also yields the square root dependence in Eq.
(4). Further simplifying the problem with an analytic approach not
only also reproduces Eq. (4), it provides a general approach that we
later use to investigate other types of equilibrium cliffed coast
responses to sea-level rise.
3.1. Simplified numerical model
The simplified model adopts the fundamental principles of
SCAPE as applied to generate Eq. (4). Foremost, it assumes a low
sediment condition where beach sediment does not affect the
evolution of the profile. The simple numerical model consists of a
single coastal profile discretized at equal elevation steps (Fig. 4).
Each elevation point records the location of the rocky coast in the
cross-shore (y) direction (here spaced at 0.1 m). Whereas the
SCAPE model uses stochastic wave inputs, an oscillating tidal cycle
(Fig. 2b), and a non-uniform erosion relationship (f1), these effects
are simplified into a ‘zone’ of wave erosion, centered around sea
level and with set vertical range (2 m in all simulations) and
‘erosive potential’, E (m/year, with a value of 20 used here). This
function resembles the tidally integrated erosion function arising
from SCAPE simulations (Fig. 2b), with the vertical range of erosion
representing both the tidal range and incoming wave height
variability. (Erosion functions with different shapes were tested,
222 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229

ε segment of the cliff is exposed to the same conditions. The erosion

relationship from Eq. (3) does not prescribe a long-term equilib-
rium—if sea level is held constant, the rate of cliff erosion decreases
vertical cliff over time, but does not stop (as shown by Walkden and Hall, 2005,
maintained E and Walkden and Dickson, 2008). If there is no sea-level rise,
zone of active ‘equilibrium’ geometry is not attained. While this behavior is
erosion S (mm/yr)
similar to findings for other soft and hard rock environments, it
stands in contrast to Bruun-type relationships for sandy coasts,
msl erosion which suggest that no coastal recession occurs if sea level remains
constant.

abandoned 3.3. Origin of the square root relationship

platform
The emergence of similar dynamic equilibrium geometries and
trends in erosion rates in both models suggests a similar underlying
cause. We simplify this problem further using a single hinge scenario
(Fig. 6)—we envision a position in the shoreline that will be exposed
z to a zone of wave attack that migrates upwards with the rate of sea-
level rise, S. Below this zone, there is presumed to be no wave attack
y and therefore the bottom location (y0 at elevation z0) does not move
Fig. 4. Schematic of the domain for the simplified numerical model demonstrating a over time. The velocity of the upper hinge point of the profile (y1 at
discretized profile subjected to a zone of erosive potential migrating upwards at a elevation z1) is the instantaneous erosion rate (ε̂):
prescribed sea-level rise rate.
dy1
= −ε̂; ð6Þ
dt

which we assume is linear with slope, as in Eq. (3). Further assuming

resulting in only minor behavioral differences). Within this zone of that the location of the subaerial cliff above the zone of wave attack is
wave attack, or tidally averaged surf zone, elements within the rock controlled by the rate of erosion of the platform, oversteepening and
profile are eroded at a rate proportional to their slope, collapse would make the overhanging cliff erode at the same the rate
as y1. This framework also assumes again that this is a physical-
 
dy dy erosion-limited system, and that the cliff does not contribute
= E tan ; ð5Þ
dt dz significant beach-building sediment. Although the cliff itself is steep,
equilibrium platform geometries have relatively shallow slopes, and
we assume:
a simplification of Eq. (3) based upon the linear slope dependence on
erosion as suggested by Kamphuis (1987). Points above the zone of dz ðz −z Þ
wave attack are eroded to stay landward of the top of the wave ε̂ = K =K 1 0 : ð7Þ
dy ðy0 −y1 Þ
platform, maintaining a (arbitrarily) vertical cliff. As in the SCAPE
model, this abstraction represents the long-term collapse of over-
Combining Eqs. (6) and (7) yields:
hanging cliffs. Below the zone of wave attack, the profile does not
change within this model framework—portions of the rocky profile ðy0 −y1 Þdy = K ðz1 −z0 Þdt; ð8Þ
below the zone of wave attack are abandoned relict slopes. The model
is run by raising sea level at a constant rate, S (mm/y), varied between which can be integrated over some characteristic time period, τ,
simulations.
2
ðy0 −y1 Þ = K ðz1 −z0 Þτ; ð9Þ
3.2. Numerical model results
yielding (assuming y1(0) = y0 = 0) a general relationship that the
Similar to the SCAPE model, a dynamic equilibrium profile of amount of recession of a coastal segment increases as the square root
erosion emerges within the simple numerical model (Fig. 5). The of the time exposed,
erosion rate, measured as the velocity of the cliff retreat, increases as pffiffiffi
the rate of sea-level rise increases, but with a diminishing increase in Δy∝ τ: ð10Þ
erosion for higher sea-level rise rates. Not only do the general
behaviors agree, the simple model exhibits the exact relationship As the profile flattens, the rate of recession decreases until the zone
between relative erosion rate and sea-level rise rate from Eq. (4) of erosion rises above z1 and erosion at this location ceases. The time-
(Fig. 5c). As sea-level-rise rates increase, the equilibrium erosion integrated erosion rate is the accumulated erosion divided by the time
shape becomes steeper, as evidenced by the slope of the abandoned of exposure:
platform (Fig. 5b).
Therefore, a simple numerical abstraction of the SCAPE model pffiffiffi
Δy τ 1
components can lead to the same general behavioral response to ε∝ ∝ ∝ pffiffiffi : ð11Þ
Δt τ τ
changes in sea-level-rise rate observed in the full simulations. For
the simple model, there is no scatter about the square-root
relationship, suggesting that the (small) scatter around the fit for By definition, the time of exposure, however, is inversely related to
the SCAPE model (Fig. 3b) may arise from the stochastic driving of the rate of sea-level rise,
the model, and not a predicted variance about this relationship. In
these (and the SCAPE) results, it is useful to note that the 1
τ∝ ; ð12Þ
equilibrium slope is truly dynamic, resulting from the time each S
 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229 223

a profile development b equlibrium profile shape

6

0
4
S (mm yr-1)
elevation above MSL (m)

elevation above MSL (m)

2 2

4
0 −5 6
8
10
12
−2 14
16

−4
−10

−6
3200 3400 3600 3800 4000 0 200 400 600 800
offshore distance (m) offshore distance (m)

c 2
d 1

0
1.5 elevation above MSL (m)

−1
ε2
1
ε1
−2

0.5
−3
Parameter test results
ε2/ε1= (S2/S1)½

0 −4
0 0.5 1 1.5 2 2.5 0 0.002 0.004 0.006 0.008 0.01
S2/S1 Slope / S½

Fig. 5. Results from the simple numerical model of cliff erosion. (a) Shore profiles plotted relative to a rising mean sea-level, showing self-parallel retreat and development of (b)
equilibrium profiles whose slope depends on the rate of sea-level rise. (c) Relative equilibrium erosion rate as relative sea-level rate changes (points), plotted with a fit of
ε2 = ε1 = ðS2 =S1 Þ =2 . (d) Equilibrium slope divided by the square root of sea level rise for marked (‘+’) simulations in (c).
1

thereby,
pffiffiffi
ε∝ S: ð13Þ
ε
Comparing the equilibrium erosion rates of two different sea-level
rise rates would therefore follow the relationship in Eq. (4)
demonstrated by both SCAPE and the simple numerical model.

z1 3.4. Discussion of the simplified relationship

msl E S Below, we place Eq. (4) within a larger context of the range of
possible coastal behaviors for a variety of environments, using the
temporal scaling arguments presented in the previous section. The
analysis presented above, which demonstrates that simplified
z0 approaches can be used to derive the sea-level-rise relationship,
Eq. (4), is applicable to soft-rock coasts consisting of poorly
consolidated sediment with little to no sediment cover. This
assumes that beach sediment volume does not significantly alter

z land ocean the equilibrium rate of cliff erosion (although they may cause
transient changes). This implies either no beach or one of small
y y1(t) y0 volume; note, however, that a beach occupying most of the
intertidal zone could still be considered low-volume (Walkden
Fig. 6. Schematic of ‘panel’ model of cliff erosion. and Dickson, 2008). Although this low-volume condition is most
224 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229
likely to hold true when the cliff consists of fine-grained sediment
that is lost from the system, alongshore transport gradients could
also keep the beach volume sufficiently small (e.g. Valvo et al.,
2006; Dickson et al., 2007a).
As demonstrated by both of the models, Eq. (4) appears to arise
directly from the time integration of a presumed erosion rate that
is linear with the local slope. While such a linear relationship
seems reasonable and is given some credence by analysis of
distributions of erosion rate by Walkden and Hall (2005, see Fig. 5)
and their simulation of realistic shore profiles, the original
derivation of the erosion term in Kamphuis (1987) does not
appear to be directly supported by field or laboratory measure-
ments. Furthermore, the slope term itself within Kamphuis (1987)
has an unusual origin, a remnant of an assumed linear relationship
between the Irabarren number (non-dimensional breaking wave
steepness or surf similarity; Battjes, 1974) and the dissipation rate,
which, in turn, is assumed to scale linearly with the rate of
platform erosion. Therefore, although the linear relationship is
sensible, it should be considered either phenomenological or
parametric. However, this is not a unique aspect of the linear
slope relationship, as most other commonly used erosion relation-
ships are similarly phenomenological or parameterized, as dis-
cussed below.
4. Generalized erosion rate relationships for cliffed coasts
The analysis above helped to demonstrate the underlying origin
of a specific soft-rock coastal evolution relationship originally
derived from the multifaceted SCAPE model. The analysis techni-
ques from above, using temporal scaling and the ratios of erosion
rate as a function of sea level rise rate, provide a useful tool to
explore the variety of cliffed coast behavior. To investigate the suite
of potential responses of a wide variety of cliffed coast environ-
ments, sandy or rocky, hard or soft, to changes in sea-level rise, we
expand upon the simple scaling arguments used above. In all cases,
we consider long-term equilibrium states, and the equilibrium
geometries that would arise in each state.
4.1. Instant response system
At one extreme, a coastal system can be envisaged that
maintains a constant geometry within the active wave region.
This requires that profiles (i) instantaneously attain a shape that is
independent of sea-level rise and (ii) maintain this geometry over
time. This response matches that assumed when Bruun-type rules
are applied to sandy environments—the active zone (in most cases
the shoreface) maintains a constant shape (although in rocky
coasts the shape is created through erosion whereas sandy coastal
profiles can both erode and accrete). If so, the erosion rate does not
depend upon the duration of exposure to wave attack, instead it
follows a relationship such as Eq. (2) where the erosion rate is
linearly dependent upon the sea-level rise rate. Local slope and
percentage of fines are the same, leading to:
ε2 S2
∝ ð14Þ
ε1 S1
for different equilibrium states.
In cliffed environments, such a response could occur at a location
where the profile is covered with mobile sediment and where the cliff
itself consists of easily removable conformable sediment (Fig. 1c,d).
Examples could be coasts with glacial till bluffs (Davis, 1896) or sand-
gravel cliffs found along the coasts of alluvial-outwash plains (Dickson
et al., 2007b). The slope of the shoreline trajectory is controlled by
mass balances, using relationships that are adjusted for the percent of
sandy material in the cliff or bluff (Bray and Hooke, 1997; Wolinsky
and Murray, 2009; Walkden and Hall, 2011). This example represents
a ‘transport limited’ response (Howard, 1994b). For erosional coasts
whose evolution requires removal of coherent rock or clay, this exact
instantaneous response does not seem physically reasonable and it is
difficult to imagine physical analogs where an instantaneous
transition to a zero-erosion geometry would be expected to occur.
On a rocky or cohesive coast, the shoreline trajectory would not be
determined by sediment mass balances (as in a Bruun-type
relationship), but rather by local erosional resistance (i.e. the
resistance of the rock to erosion processes).
4.2. No feedback system
Perhaps the simplest potential system is one in which there are no
feedbacks between the profile geometry and erosion rate. This case
would be represented by a constant zone of erosion or weathering
moving up the cliff profile as sea level rises. Breakdown of any
segment would be proportional to the duration of exposure, thus the
erosion rate is the quantity of erosion divided by the (same) time of
exposure,
Δy τ
ε∝ ∝ ∝1: ð15Þ
Δt τ
The amount of erosion of the ‘no feedback’ case therefore is
independent of the sea-level rise rate,
ε2
∝1: ð16Þ
ε1
(Fig. 7c). The equilibrium profile would have a slope that varies
linearly with the rate of sea-level rise.
This thought experiment emphasizes that feedback is required for
the coastal erosion rate to have any dependence on sea-level rise rate.
A ‘no feedback’ system would not only have little to no sediment
cover, but also a recession rate that is independent of coastal
geometry (a situation unlikely in an environment this is dominated
by physical erosion of rock). It could, however, represent an
environment where the time of exposure controls recession, perhaps
at a coast where chemical, physical, thermal, or biological weathering
dominates, and there is little wave energy (although some mecha-
nism must exist to remove sediment and other weathering products,
e.g. tidal flows).
4.3. Negative/Damped feedback system
The most likely responses of rocky coasts (soft or hard) lie
between the end members of the ‘instantaneous response’ and ‘no
feedback’ cases, where negative feedbacks reduce the rate of cliff
recession as shore profiles are exposed to waves over time (Fig. 7b).
We have explored one such system using the SCAPE model relation-
ships applicable for low-beach-volume soft-rock coasts in which the
evolution is dominated by wave erosion, but most other rocky (hard
or soft) coasts types should fall within this response category. A
negative feedback system only requires that, if sea level were to stay
constant, the erosion rate would slow down over time. If there were
little sediment, this would represent any coast where waves create a
low-gradient platform geometry that effectively dissipates wave
energy. This could either be a case where shore platform downcutting
slows though a slope-related feedback (as in Eq. (3)), or one where
development of a wide platform reduces wave attack on the cliff toe,
as implemented in many models (e.g. Fleming, 1965; Anderson et al.,
1999; Trenhaile, 2005). If significant coarse-grained sediment were
present, the continued growth of a bulwarking beach could also
provide a negative feedback for a mixed rock/sediment coast.
However, coasts dominated by weathering could also behave within
 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229 225

a
instant response
dominated by sediment flux
gradients

ε2 potential environments:
ε1 1 ε2 S2
=
- bluffed coasts
ε1 S1 - high-sediment cliffs

1 S2/S1

b
negative feedback dominated by wave-driven
erosion

ε2 potential environments:
- rocky shore platforms
ε1 1 ε2 =ε1
S2 - cliffs fronted by low-
S1 sediment-volume beaches

1 S2/S1

c wave influence negligible

no feedback no geometric feedbacks
weathering dominated
ε2
=1
ε1 potential environments:
ε2 - hard rock coasts with
ε1 1 no platform
- bioerosion dominant
- low-wave-energy
1 S2/S1 locations

d
inverse feedback
erosion rates increase with
time of exposure

ε2 ε2 =ε1
S2
potential environments:
ε1 1 S1
- bioerosive environments
- reflective cliffs

1 S2/S1
Fig. 7. Schematic of the basic types of potential sea-level-rise relationships for cliffed coasts. Natural examples from (a) Cape Cod, Massachusetts, USA, (b) North Norfolk, UK, (c) near
Cape Coast, Ghana, and (d) Bathsheba, Barbados.

this realm. For example, if increasing sediment depth reduces
abrasion or benthic bio-weathering rates (e.g., Sunamura, 1982;
Valvo et al., 2006), this would also constitute a negative feedback.
4.4. Inverse feedback system
A further end member presents a potentially surprising response
to increased rates of sea-level rise: in an environment where cliff
erosion increases with time of exposure, erosion rates could decrease
as sea-level-rise rates increases (Fig. 7d). Although this example
seems extreme, there are several possible scenarios for this behavior.
One potential example would be a rocky coast controlled by
bioerosion, where notching of the cliff provides increased habitat
which, in turn, increases the rate of further bioerosion (e.g. Focke,
1978). Another potential example would be an environment in which
a scant sediment cover increases abrasion or weathering compared to
a bare rock coast (Robinson, 1977a,b; Sunamura, 1982; Skafel and
Bishop, 1994). This is analogous to ‘roll-overs’ in soil production
226 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229

curves (Anderson, 2002) or bedrock stream erosion rates (Sklar and
Deitrich, 2001) that can lead to interesting feedbacks in terrestrial
environments, including development of bedrock exposures (Strudley
et al., 2006). As another possible mechanism, Wilcock et al. (1998)
note that steep coasts tend to be reflective, and thereby can
experience slower erosion rates than dissipative coasts with breaking
waves. Fast sea-level rise could therefore abandon a platform that
would develop during slower rises (or hiatuses).
In general, these ‘inverse feedback’ systems would have a
tendency to not only create erosional features, such as notches and
platforms, during periods of slow sea level rise, but would also tend to
preserve those features during a sea-level rise acceleration as erosion
would be slower during fast sea-level rise. In this case an increase in
sea-level rise rate could be sufficient to preserve notches from a still-
stand (or periods of slow sea-level rise)—an instantaneous jump in
sea level would not be required for preservation of these features.
5.1. Other basic rocky coast relationships
One commonly used formulation (Trenhaile, 2009; Porter et al.,
2010) suggests that shore platform erosion is proportional to the bed
stress in the surf zone. The derived equation suggests that this stress is
proportional to the square of the shore platform slope. Assuming
small angles and that erosion is proportional to this surf stress yields:
 
dz 2
ε̂ = K ; ð18Þ
dy
which can be integrated as Eq. (7) to give the relationship:
1=3
Δy∝τ : ð19Þ
Following the steps in Eqs. (11)–(14), this leads to:

2=3
4.5. Basic response types ε∝S ; ð20Þ

By inspecting at ratios of equilibrium erosion rates, we are able to
characterize a full range of potential rocky and cliffed coast response
to changes in the rate of sea-level rise (Fig. 7). The general
relationships can be understood assuming a power law relationship
between these ratios,
 m
ε2 S would be expected to tend towards a constant shape, further speeding
∝ 2 ; ð17Þ
ε1 S1
where the exponent, m, is indicative of the response type. (Here
we assume a power law relationship for illustrative purposes; other
functional relationships should exist, such as those presented by
Sunamura and explored below.) The ‘instant response’, or Bruun-type
relationship has an exponent m = 1, whereas m = 0 represents the ‘no
feedback’ case. If a polynomial fit is appropriate, we see that ‘damped
feedback’ cases would lie between these cases and could be described
by any 0 b m b 1; m = 1/2 for the SCAPE relationship. The inverse
feedback case can be represented by m b 0 (Fig. 7 uses m = −1/2 as an
illustration). Eq. (17) suggests one other case could exist, a potential
‘positive feedback’ case, for m N 1. Although such a scenario is possible,
it would be unlikely, as it would require the shore profile to become
increasingly flat for faster sea-level-rise rates. One example could be
permafrost coasts whose evolution is controlled by melting of the icy
matrix (e.g. Mars and Houseknecht, 2007). Generally, the ‘instant
feedback’, linear response relationship can be considered an upper
bound of the equilibrium response of a cliffed coast to changes in the
sea-level-rise rate.
5. Discussion
or Eq. (17) with m = 2/3. As expected, the stronger functional
dependence on slope than in the linear case results in a stronger
negative feedback, and erosion rates are correspondingly less
dependent on the rate of sea-level rise. Typically, however, this
equation is applied with an assumed critical stress that must be
overcome before erosion begins. In this case, the negative feedbacks
will be stronger and, if this threshold is high, the shore platform
the morphologic response towards that of the ‘instant feedback’.
The functional form of Eqs. (3) and (18) is generally the same, so it
is unsurprising that they behave similarly within this context.
Sunamura (1983) presented a rocky coast erosion equation with a
different functional form:
dy −kt
ε= = Ae ; ð21Þ
dt
where A is a background erosion rate (distance/time) and k is a rate
constant (1/time). This relationship was originally proposed to
represent the general behavior that, as a platform widens, its erosion
rate slows down. This relationship can be analyzed for an equilibrium
case using the same steps as in Section 3.3. Integrating Eq. (21) over a
characteristic timescale, τ, assuming an initially vertical cliff wall at
4
ε2 S2
k (mm yr-1): ε1 = S
1
0.5
2
3 8

The generalized framework of potential equilibrium responses of S2

ε2 = ε1
cliffed coasts to sea-level rise changes suggests that coasts controlled ε2 S1
2
by erosion of rock will most likely act as negative feedback systems. ε1
Although this general behavior has long been understood, exactly
where a particular environment (or erosion formulation) lies within
the transition from ‘no feedback’ to ‘instant response’ systems 1
remains unexplored. We have studied one case based upon the ε2
=1
formulation by Kamphuis (1987) for application to soft-rock coasts. A ε1
host of different environmentally dependent processes transform
soft- and hard-rock coasts, including abrasion, quarrying, piping, and 0
0 1 2 3 4
weathering, with many different equations suggested for each case.
We cannot investigate all of these; however, below we study a few of S2/S1
these equations with different general forms to investigate how these Fig. 8. Equilibrium erosion relationships from the exponential negative feedback
relationships fit within the general framework suggested above relationship, Eq. (21), with varying k values. Dashed lines indicate potential end-
(Fig. 7). member relationships.
 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229 227

t = 0, and substituting the reciprocal relationship between sea-level within the space of the ‘negative feedback’ system, within the bounds
rise and time of exposure, Eq. (12), yields: of the ‘no feedback’ and ‘instant response’ cases (Fig. 8). If the decay is
fast (large k), the system relaxes towards a slowed erosion rate, and
Δy A −kτ
 therefore towards a fixed geometry, quickly within the time of
= 1−e : ð22Þ
Δt τk exposure, acting more like an ‘instant response’ system. A slowly
responding system (small k) does not develop negative feedback
This relationship exhibits asymptotic behavior as, over longer within the characteristic time frame, and its behavior therefore
timescales (as controlled by k), erosion becomes negligible, with the resembles that of the ‘no feedback’ system (characterized by a linear
erosion rate then decreasing as the reciprocal of the characteristic response as the exponential terms become negligible).
timescale.
For different equilibrium conditions, 5.2. Addition of a background erosion rate
 
S2 1−e
−k = S2 Most rocky coasts are affected by a combination of processes. What
ε2
=  : ð23Þ happens when an erosional process dependent upon sea-level-rise
ε1 S1 1−e−k = S1 rate and one independent of sea-level rise rate are working in
concert? Investigating the potential effects of global-warming-
As opposed to Eq. (3), the general form of Eqs. (17) and (20), induced sea-level rise of rocky coasts, Sunamura (1988) suggests a
which are surprisingly independent of local geologic factors, this sea-level-rise-specific erosional relationship, extending from a previ-
relationship suggests that the erosion rate constant, k, explicitly ous quantitative model that approaches the problem with both a
affects the sea-level-rise response of this formulation. Varying k, mechanical and geometric framework (Sunamura, 1978). This
however, demonstrates that the general response types of Eq. (23) fall formulation assumes that the erosion rate consists of two compo-
nents, one due to background erosion (R, distance/time) and one
induced from sea-level rise:
a
4 S
ε2 S2 ε=R+ ; ð24Þ
αp
αpR (mm yr : ε1 = S
-1)

0.25 1

1 where αp is the offshore platform slope. In this instance, the cliff

3 4 response to sea level rise is assumed to be instantaneous (Sunamura,
1988), and therefore is expected to behave linearly with sea-level rise.
Bray and Hooke (1997) offer a similar formula, with a background
ε2 recession superimposed upon a sea-level-rise term, with an important
2
ε1 distinction that this background recession is assumed to be caused by
sediment deficits from alongshore sediment transport gradients, not
from rock erosion. In general, the incorporation of a background rate
of erosion term independent of S seems particularly warranted in the
1
ε2 case of rocky coasts, which would be expected to erode and weather
=1 even if sea level does not rise.
ε1
Assuming that the platform geometry remains constant (i.e.
0 constant αp), this relationship can be placed in terms of a ratio of
0 1 2 3 4 erosion rates,
S2/S1
ε2 R + S2 = αp
b = : ð25Þ
ε1 R + S1 = αp
4
kpR (mm½ yr-½):
The ratio of equilibrium erosion rates will be site-dependent,
0.25
1
controlled by the local geometry and the background erosion rate. If
3 4 sea-level rise adds little relative shoreline recession (small αpR), the
combined behavior resembles that of the ‘no feedback’ system—the
S2 background erosion rate dominates the contributions from sea-level
ε 2 = ε1
S1 rise (Fig. 9a). As the sea-level rise component outpaces the
ε2
2 background recession, the response approaches that of the ‘instanta-
ε1
neous feedback’ system. The background erosion contribution
becomes inconsequential. This relationship improves upon others as
it allows for a non-zero rate of recession for a steady sea level.
1 Interestingly, it also suggests that knowledge of the platform
ε2 geometry and the local rate of sea-level rise could be combined to
=1
ε1 determine the background erosion rate using Eq. (24).
The assumption of a constant αp represents an oversimplification;
0 the original Sunamura (1988) model does not assume a constant
0 1 2 3 4
S2/S1 platform slope. For illustration purposes, we can apply Eq. (24) to the
soft rock coast case explored in the first part of the paper, using the
Fig. 9. Erosion relationships including a superimposed background erosion rate for (a) a derived relationship between profile slope and S from the SCAPE (and
constant platform geometry, Eq. (25), and (b) negative feedback system, Eq. (27). simple model) simulations,
Dashed lines represent other potential relationships presented in Fig. 7. Respectively,
αpR and kpR represent the background erosion rate as a proportion of the recession due pffiffiffi
to sea-level rise for rate S1. αp = kp S: ð26Þ
228 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229
where the constant kp is approximately 6.2 (yr1/2 m− 1/2) for the
parameters used in the model simulations shown here. Substituting
into Eq. (25) yields:
pffiffiffiffiffi
ε2 R + S2 = kp
= pffiffiffiffiffi : ð27Þ
ε1 R + S1 = kp
This hypothetical relationship represents a transition from a ‘no
feedback’ system to a ‘negative feedback’ system (Fig. 9b), with an
upper bound constrained by Eq. (4). Both of these illustrative cases
demonstrate that the addition of a background erosion term reduces
the predicted sensitivity of coastal recession on sea-level rise, in both
cases reducing the slope of the response (Fig. 9a, b), making the
system behave more like a ‘no feedback’ system. For a rocky coast, the
response rate determines if the behavior is more akin to an instant
response or no feedback case.
5.3. Explicit dependence upon the rate of sea-level rise
The basic relationships for cliff erosion suggest that, in a system
where beach sediments are limited and where the rate of rock profile
erosion is dependent on the profile slope, an increase in the rate of
sea-level rise does not produce a linear increase in rock recession rate.
This is a significant difference with Bruun-type relationships that
predict a linear relationship with sea-level rise. However, there is
another significant difference between rocky coast responses pre-
sented here and Bruun-type rules—if a 1 m sea level rise took one year
or 1000 years, the Bruun rule would predict the same total landward
movement of the shoreline. On the other hand, the relationships
derived for rocky or cohesive coasts depend on the rate of sea-level
rise, not just the total rise. This applies even for rocky coast ‘instant
feedback’ systems—although the response is linear, it is time-
dependent and the ‘effective slope’ is unrelated to the upland
geometry.
Even though some rocky coasts could behave in a manner similar
to sandy coasts as ‘instant response’ systems (Figs. 8 and 9a),
application of Bruun-type rules to a cliffed coast must be made with
care. Bray and Hooke (1997) explicitly state that their application of
the Bruun rule to a cliffed coast is based on an assumed sediment
abundance and the conservation of mass. Such modifications of the
Bruun rule could only be applied to sandy, transport-limited systems,
not to low-volume, weathering-limited environments.
5.4. Erosion rate and erosion formulations
The approaches here are abstract, and specific rates of cliff erosion
are not addressed. It is important to recognize that relationships
regarding how a system could respond to a change in the rate of sea-
level rise are generally independent of the rate of sea-level rise itself.
For example, although a bluffed coast with an ‘instant response’
behavior may be highly sensitive to sea-level rise, the recession rate of
the coast itself could be orders of magnitude lower than that of a
friable rocky coast exposed to large waves which responds to sea-
level rise with a ‘no feedback’ response.
Although the investigations presented tend to significantly oversim-
plify complex coastal systems affected by myriad processes, the results
emphasize the need for careful considerations of the underlying process
equations used in all rocky coast evolution models, particularly as rocky
coast erosion formulations are typically ‘ad hoc’, heavily parameterized,
and difficult to test over the timescales of interest. This applies to models
at all levels of complexity. Because the formulae we use are simplifications
and meant to apply over long timescales, we suggest great care be used
before considering their application in management applications—the
same care that should be applied before use of the Bruun rule over all
management timescales.
5.5. Time to develop equilibrium
Another important consideration is that the relationships investi-
gated here assume the attainment of a dynamic equilibrium shape, i.e.,
they describe erosion after adequate time has passed for the profile to
fully develop in relation to current conditions. The relationships do not
describe recession during transients, such as during a rapid acceleration
of sea-level rise as predicted over the upcoming century and beyond. For
the types of relationships we present here, a first-order estimate for the
full equilibrium transition is the time it takes for sea level to traverse the
effective zone of erosion, Ze (which is determined by the tidal range in
conjunction with wave variability). An approximation for the transient
timescale is Ze/S: for a 2 m erosion zone, it should take 1000 years for a
2 mm/year rise to fully traverse this zone and 200 years for a 10 mm/
year rise. Therefore, coastal management timeframes are shorter than
the transients of rock profile response and are certainly less than the
time it takes for sea-level-rise rates to equilibrate. Therefore, care should
be taken with direct application of the formulations presented,
particularly over shorter temporal scales.
6. Conclusions
Shore response to accelerated sea-level rise is one of the most
important problems facing coastal geomorphology. For cliffed coasts,
we have characterized these responses by analyzing the feedback
relationships between erosion and shore profile shape. By studying
equilibrium conditions, and normalizing posterior recession rate and
sea-level rise with prior conditions, we provide a theoretical
framework to understand the future response of a wide variety of
cliffed coast environments (consolidated or unconsolidated, soft- or
hard-rock) to increased sea-level rise rates.
The general type of response to sea-level rise changes will be
determined by the coast type, environmental drivers, and dominant
processes (Fig. 7). The most common behavior of cliffed coasts is likely
to be that of a ‘negative feedback’ in which an increase in sea-level-
rise rate results in an increase in the profile slope. While this increased
slope indicates a faster local erosion rate, it also reveals a quicker
abandonment of the active platform. Although negative feedback
behavior has long been noted and suggested for rocky cliffs, the
framework presented here demonstrates how the relative strength
and timescales of the negative feedback could make coasts either
respond instantaneously to sea-level rise (i.e. with a fixed geometry)
or apparently not at all (i.e. with little feedback). Environments
dominated by weathering should have little response to increasing
rates of sea-level rise, with the potential of even an inverse feedback
where erosion rates could actually decrease as sea-level-rise rates
increase. The implications of this inverse response should be studied
further, as this behavior could affect the interpretation of preserved
coastal erosion features from previous sea-level highstands or
hiatuses.
The relationships between sea-level-rise rate and erosion pre-
sented here can be useful given the need for rapid assessment of the
coastal response to sea-level rise. However, the duration of the
transient stage constrains specific application of the equilibrium
relationships. As described above, this stage should likely last on the
order of centuries—in almost all cases this period extends beyond that
of concern to coastal managers, particularly as sea level is expected to
undergo significant accelerations over the coming centuries. Over
sub-equilibrium timescales, the equilibrium relationships, however,
offer conservative estimates of the response to increased rates of sea-
level rise. The most conservative limit would be a linear, no-feedback
response. Coasts with dynamic feedbacks should be expected to have
a damped response. For example, broader-scale coastal management
studies such as the UK Shoreline Management Plans (e.g. Royal
Haskoning, 2011) utilize, with care, Eq. (4) rather than applying more
 A.D. Ashton et al. / Marine Geology 284 (2011) 217–229
detailed studies, for which direct modeling of the transient is required
(e.g. Royal Haskoning, 2010).
The approaches here oversimplify the broad spectrum of processes
that affect individual coasts, yet they shed light on dominant process.
These simplifications can also be useful for understanding of more
detailed models and complicated environments. As coastal evolution
models for rocky coasts become increasingly complex, it is important
to have basic theoretical expectations of how the constituent
equations behave. Simple relationships, such as Eq. (4), provide
straightforward and much needed means of estimating how quickly
sections of the world's coastline could retreat in the future as sea-level
rise accelerates. The analyses presented here offer a first step towards
development of such simple theoretical benchmarks for future
prediction. Work is now needed to characterize the transient stage
to further improve the information available to coastal managers.
Acknowledgements
Partial funding of this research was provided by the Coastal Ocean
Institute of the Woods Hole Oceanographic Institution and NSF grant
CNH-0815875. Reviews by Robert Anderson, Peter Adams, and an
anonymous reviewer considerably improved this paper.
References
Anderson, R.S., 2002. Modeling the tor-dotted crests, bedrock edges, and parabolic
profiles of high alpine surfaces of the Wind River Range, Wyoming. Geomorphology
46 (1–2), 35–58.
Anderson, R.S., Densmore, A.L., Ellis, M.A., 1999. The generation and degradation of
marine terraces. Basin Research 11, 7–19.
Appeaning Addo, K., Walkden, M., Mills, J.P., 2008. Detection, measurement and
prediction of shoreline recession in Accra, Ghana. ISPRS Journal of Photogrammetry
and Remote Sensing 63 (5), 543–558.
Battjes, J.A., 1974. Computation of setup, longshore currents, runup and overtopping
due to wind generated waves. Delft University of Technology, Civil Engineering
Department Report 74-2, p. 244.
Bray, M.J., Hooke, J.M., 1997. Prediction of soft-cliff retreat with accelerating sea-level
rise. Journal of Coastal Research 13 (2), 453–467.
Bruun, P., 1962. Sea-level rise as a cause of shore erosion. Proceedings of the ASCE,
Journal of the Waterways and Harbors Division 88, 117–130.
Carpenter, S.R., 2003. The need for fast-and-frugal models. In: Canham, C.D., Cole, J.J.,
Lauenroth, W.K. (Eds.), Models in Ecosystem Science. Princeton University Press,
Princeton, NJ, pp. 455–460.
Church, J.A., White, N.J., 2006. A 20th century acceleration in global sea-level rise.
Geophysical Research Letters 33, L01602, doi:10.1029/2005GL024826.
Cinque, A., De Pippo, T., Romano, P., 1995. Coastal slope terracing and relative sea-level
changes: deduction based on computer simulations. Earth Surface Processes and
Landforms 20, 87–103.
Davis, W.M., 1896. The outline of Cape Cod. Proceedings of the American Academy of
Arts and Sciences 31, 303–332.
Dean, R.G., 1991. Equilibrium beach profiles: characteristics and applications. Journal of
Coastal Research 7 (1), 53–84.
Dean, R.G., Maurmeyer, E.M., 1983. Models for beach profile response. Handbook of
Coastal Processes and Erosion. CRC Press, pp. 151–165.
Dickson, M.E., Walkden, M.J.A., Hall, J.W., 2007a. Systemic impacts of climate change on an
eroding coastal region over the twenty-first century. Climatic Change 84 (2), 141–166.
Dickson, M.E., Hicks, D.M., Coco, G., 2007b. A Holocene-scale numerical model of shore
profile evolution on an eroding alluvial fan coast. Journal of Coastal Research 50,
821–826 ICS2007 (Proceedings) SI.
Dubois, R.N., 2002. How does a barrier shoreface respond to a sea-level rise? Journal of
Coastal Research 18 (2), iii–v.
Fleming, N.C., 1965. Form and relation to present sea level of Pleistocene marine erosion
features. Journal of Geology 5, 799–811.
Focke, J.W., 1978. Limestone cliff morphology on Curacao (Netherlands Antilles), with
special attention to the origin of notches and vermetid/coralline algal surf benches
(“cornices”, “trottoirs”). Z. Geomorph. N.F. 22 (3), 329–349.
Hall, J.W., Meadowcroft, I.C., Lee, E.M., van Gelder, P., 2002. Stochastic simulation of
episodic soft coastal cliff recession. Coastal Engineering 46, 159–174.
Hapke, C., Plant, N., 2010. Predicting coastal cliff erosion using a Bayesian probabilistic
model. Marine Geology 278, 140–149.
Howard, A.D., 1994a. A detachment-limited model of drainage basin evolution. Water
Resources Research 39 (7), 2261–2285.
Howard, A.D., 1994b. Badlands. In: Abrahams, A.D., Parsons, A.J. (Eds.), Geomorphology
of Desert Environments. Chapman and Hall, London, pp. 213–242.
Kamphuis, J.W., 1987. Recession rate of glacial till bluffs. Journal of Waterway, Port,
Coastal and Ocean Engineering 113 (1), 60–73.
229
Mars, J.C., Houseknecht, D.W., 2007. Quantitative remote sensing study indicates
doubling of coastal erosion rate in past 50 yr along a segment of the Arctic coast of
Alaska. Geology 35 (7), 583–586.
Meehl, G.A., et al., 2007. Global Climate Projections, Climate Change 2007: The Physical
Science Basis. Contribution of Working Group I to the Fourth Assessment Report of
the Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge, UK.
Mitrovica, J.X., Gomez, N., Clark, P.U., 2009. The sea-level fingerprint of West Antarctic
collapse. Science 323 (5915), 753.
Murray, A.B., 2003. Contrasting the goals, strategies, and predictions associated with
simplified numerical models and detailed simulations. In: Iverson, R.M., Wilcock, P.
R. (Eds.), Prediction in Geomorphology, AGU Geophysical Monograph 135,
Washington, D.C, pp. 151–165.
Naylor, L.A., Stephenson, W.J., Trenhaile, A.S., 2010. Rock coast geomorphology: recent
advances and future research directions. Geomorphology 114 (1–2), 3–11.
Peltier, W.R., 2004. Global glacial isostacy and the surface of the ice-age earth: the ICE-
5G (VM2) model and GRACE. Annual Review of Earth and Planetary Sciences 32
(1), 111–149.
Porter, N.J., Trenhaile, A.S., Prestanski, K.J., Kanyaya, J.I., 2010. Shore platform
downwearing in eastern Canada: micro-tidal Gaspe, Quebec. Geomorphology 116
(1–2), 77–86.
Rahmstorf, S., 2007. A semi-empirical approach to projecting future sea-level rise.
Science 315 (5810), 368–370.
Robinson, L.A., 1977a. The morphology and development of the northeast Yorkshire
shore platform. Marine Geology 23 (3), 237–255.
Robinson, L.A., 1977b. Marine erosive processes at the cliff foot. Marine Geology 23 (3),
257–271.
Roy, P.S., Cowell, P.J., Ferland, M.A., Thom, B.G., 1994. Wave-dominated Coasts. In:
Carter, R.W.G., Woodroffe, C.D. (Eds.), Coastal Evolution: Late Quaternary Shoreline
Morphodynamics. University Press, Cambridge, pp. 121–186.
Royal Haskoning, 2010. Soft Cliff And Platform Erosion (SCAPE) Modeling of the Drigg
Coast in Cumbria, on behalf of LLWR Ltd.
Royal Haskoning, 2011. Cornwall and Isles of Scilly Shoreline Management Plan, on
behalf of the Cornwall and Isles of Scilly Coastal Advisory Group.
Skafel, M.G., 1995. Laboratory measurement of nearshore velocities and erosion of
cohesive sediment (till) shorelines. Coastal Engineering 24 (3–4), 343–349.
Skafel, M.G., Bishop, C.T., 1994. Flume experiments on the erosion of till shores by
waves. Coastal Engineering 23 (3–4), 329–348.
Sklar, L.S., Deitrich, W.E., 2001. Sediment and rock strength controls on river incision
into bedrock. Geology 29, 1087–1090.
Strudley, M.W., Murray, A.B., Haff, P.K., 2006. Emergence of pediments, tors, and
piedmont junctions from a bedrock weathering-regolith thickness feedback.
Geology 34 (10), 805–808.
Sunamura, T., 1977. A relationship between wave-induced cliff erosion and erosive
force of waves. Journal of Geology 85 (4), 613–618.
Sunamura, T., 1978. A mathematical model of submarine platform development.
Mathematical Geology 10 (1), 53–58.
Sunamura, T., 1982. A wave tank experiment on the erosional mechanism at a cliff base.
Earth Surface Processes and Landforms 7, 333–343.
Sunamura, T., 1983. Processes of sea cliff and platform erosion. In: Komar, P.D. (Ed.),
Handbook of Coastal Processes and Erosion. CRC Press, Boca Raton, Florida, USA.
Sunamura, T., 1988. Projection of future coastal cliff recession under sea level rise
induced by the greenhouse effect: Nii-jima Island, Japan. Transactions, Japanese
Geomorphological Union 9 (1), 17–33.
Sunamura, T., 1992. Geomorphology of Rocky Coasts. John Wiley & Sons, Chichester,
West Sussex, UK.
Trenhaile, A.S., 2001. Modeling the effect of late Quaternary interglacial sea levels on
wave-cut shore platforms. Marine Geology 172 (3–4), 205–223.
Trenhaile, A.S., 2005. Modelling the effect of waves, weathering and beach
development on shore platform development. Earth Surface Processes and
Landforms 30 (5), 613–634.
Trenhaile, A.S., 2009. Modeling the erosion of cohesive clay coasts. Coastal Engineering
56, 59–72.
Trenhaile, A.S., 2010. Modeling cohesive clay coast evolution and response to climate
change. Marine Geology 277, 11–20.
Valvo, L.M., Murray, A.B., Ashton, A.D., 2006. How does underlying geology affect
coastline change? An initial modeling investigation. Journal of Geophysical
Research, Earth Surface 111 (F02025), doi:10.1029/2005JF000340.
Walkden, M., Dickson, M., 2008. Equilibrium erosion of soft rock shores with a shallow or
absent beach under increased sea level rise. Marine Geology 251 (1–2), 75–84.
Walkden, M.J.A., Hall, J.W., 2005. A predictive Mesoscale model of the erosion and
profile development of soft rock shores. Coastal Engineering 52 (6), 535–563.
Walkden, M.J.A., Hall, J.W., 2011. A Mesoscale Predictive Model of the Evolution and
Management of a Soft Rock Coast. Journal of Coastal Research 27 (3), 529–543.
Wilcock, P.R., Miller, D.S., Shea, R.H., Kerkin, R.T., 1998. Frequency of effective wave
activity and the recession of coastal bluffs: Calvert Cliffs, Maryland. Journal of
Coastal Research 14 (1), 256–268.
Wolinsky, M.A., 2009. A unifying framework for shoreline migration: 1. Multiscale
shoreline evolution on sedimentary coasts. Journal of Geophysical Research,
doi:10.1029/2007JF000855.
Wolinsky, M.A., Murray, A.B., 2009. A unifying framework for shoreline migration: 2.
Application to wave-dominated coasts. Journal of Geophysical Research,
doi:10.1029/2007JF000856.
Zhang, K., Douglas, B., Leatherman, S., 2004. Global warming and coastal erosion.
Climatic Change 64 (1–2), 41–58.