Sustainability
of Rives-Manche Shingle Beaches
Funded by Interreg II
Deans
Place Hotel, Alfriston
30 October
Pierre
Bérard, University of Rouen
Stéphane
Costa, University of Caen
Robert
Davidson, University of Caen
Daniel
Delahay, University of Rouen
Uwe
Dornbusch, University of Sussex
Yolanda
Foote, University of Sussex
Dennis
Forsyth, Newhaven Planning Committee
Graham
Keane, Wealden District Council
France
McLester, University of Sussex
Cherith
Moses, University of Sussex
Julie
Pagney, University of Rouen
Richard
Peckham, Possford Duvivier Consultants
David
Robinson, University of Sussex
Ray
Traynor, South Downs Coastal Group, Arun District Council
Letitia
Tual, East Sussex County Council
Rendel
Williams, University of Sussex
Paul
Winfield, Possford Duvivier Consultants
1a.
Cherith Moses welcomed everyone and was very pleased that the adverse weather
conditions had not prevented them from attending the meeting. The previous
night had seen the worst storms on the Sussex coast for ten years but despite
this the French research team arrived only two hours later than planned.
Apologies had been received from many expected who were unable to attend due to
flooding on roads or had commitments to monitoring erosion and flooding.
Daniel
Delahay, Project Leader, Geographer with expertise in flood risk, coastal
dynamics, Geographic Information Systems and natural risk management.
StéphaneCosta,
Scientific Co-ordinator, Geographer with expertise in natural risk in the
coastal environment with particular reference to the Normandy and Picardie
coasts.
Franck
Levoy, Geologist, University of Caen
Robert
Davidson, Engineer, University of Caen
Benoit
Laignel, Geologist, University of Rouen
Julie
Pagny, student, University of Rouen
Pierre
Bérard, student, University of Rouen
Uwe Dornbusch, Postdoctoral Research Fellow and Scientific
Co-ordinator, Geographer with expertise in natural systems analysis and also
map and aerial photograph interpretation.
Cherith
Moses, Project Leader, Geographer with expertise in limestone coasts and
coastal erosion dynamics.
Rendel
Williams, Geographer with expertise in Quaternary environments and rocky coast
erosion.
David
Robinson, Geographer with expertise in land management and shore platform
development.
P.W.
With reference to StéphaneCosta’s point that there is a need to monitor the
behaviour of shingle overlying sand as well as monitoring shingle lying
directly on the rock platform: ‘do the two have different behaviour patterns
under wave attack?’ He also noted that deep shingle behaves in a very different
way to shallow shingle.
S.C.
All beaches on the stretch of coastline monitored for this project are natural
and have not been replenished in the way that some E. Sussex beaches have been.
They therefore have very different behaviour characteristics.
P.W.
Seaford, Eastbourne and Pevensey beaches have all been recharged with material
that has a D50 that is smaller than the original shingle and
therefore behaves in a different way. The need for stringent quality control is
very apparent.
S.C.
Concrete groynes are used but the French coast is less urbanised than the E.
Sussex coast with settlements most commonly found only in the valleys, leaving
the chalk cliffs exposed rather than protected with engineering structures.
There is also some recycling of shingle on beaches, for example at Criel sur Mer (near Le Tréport)
on
the French coast some 6 - 7000m3yr-1 is recycled but this
is low compared to the 30000 m3yr-1 at Seaford alone.
U.D.
Shingle moves from west to east along the Sussex coast. The Newhaven breakwater
traps shingle so that no natural shingle input occurs at Seaford. The output of
shingle from Seaford to the east is largely inhibited by the terminal groyne.
The wave induced longshore currents in Seaford bay have an eastward and
westward component leading to shingle movement away from the approximate centre
of the beach to either side so that the material has to be artificially
transported back to the centre. Some shingle may be lost offshore, some may
wear down by abrasion on the beach itself and small amounts may move around
Seaford Head. There is a critical sediment size for movement around
obstructions such as harbour arms. For example, at Newhaven it is evident that
shingle is trapped to the west of the harbour whilst sand can move around and
is deposited on the east side.
S.C.
The chalk is slightly different and the flint size produced is slightly bigger
in France than that on the Sussex coast. Flint size changes along the French
coast depending upon the chalk type.
U.D.
and S.C. The French coast is probably more sheltered but it would be
interesting to compare the inshore wave climate of the two. (Discussion,
however, revealed that this would not be possible on the East Sussex coast
because there is no monitoring station for measuring wave direction.)
U.D.
Yes, the results presented are preliminary and show unexpected results that contradict
current research literature. A series of experiments are now being designed to
test these results for both the E. Sussex and French shingle.
P.W.
The results of these experiments will be of particular use to those involved in
beach replenishment. Would it be possible to run some tests on dredged shingle
that is currently used for beach replenishment in order to test its durability?
U.D.
Yes, we will be happy to receive samples for laboratory testing.
U.D.
That would be valuable but is beyond the scope for this current Interreg
project.
§
CEE : 50 %
§
Conseil Régional de la Haute Normandie25
%
§
- Conseil Général de la Seine
Maritime 25
%
Responsable du programme ROCC (BRGM et
Université du Havre)
- Conseil Régional de la Haute Normandie
- Conseil Général de la Seine Maritime
- Direction Régionale de l’Environnement
(Haute-Normandie)
-
Direction Départementale de
l’Equipement
French scientific team
Responsables (Leaders)
D. Delahaye
(Professor, University of Rouen ; Laboratary of MTG)
S. Costa
(Professor, University of Caen ; Laboratory of Géophen)
- F. Levoy
(Professor ; Department of Geology)
- J. Pagny
(student ; Laboratory of Géophen)
- P.G. Bérard
(student ; Laboratory of Géophen)
- S. Costa
(Laboratory of Géophen)
- B. Laignel
(Professor ; Department of Geology)
The European
Community finances fifty percent of this programme. The other fifty percent is
provided by the Conseil Régional de Haute-Nomandie and the Conseil Général de
la Seine Maritime (twenty five percent respectively)
From
the University of Caen we have also:
·
Mister R. Davidson (study engineer of the geophen
Laboratory)
·
Misses. J. Pagny and Mister P.G. Bérard, students of
the University of Caen working along the Normandy coast.
·
Mister F. Levoy, professor in the geology department.
From
the University of Rouen we have:
·
Mister B. Laignel, professor in the geology department
·
Mister D. Delahaye,
Our
study area lies on nearly one hundred kilometres between Antifer and Le
Tréport. Geologically, this area corresponds to the north occidental end of the
Parisian sedimentary basin.
As for the East Sussex, this basin ends abruptly in cliffs, which have
an average height of seventy meters and are cut principally in the upper
Cretaceous Chalk-with-Flint.
The
erosion agent exploits a less resistant structural component causing retreat of
the cliff.
These
cliffs are cut by valleys whose bottom is under the highest watermark of spring
tide. Yet, the population have chosen to settle in these low areas.
Under the dominantly Westerly wave action, the shingle longshore drift
moves from Antifer to the Baie de Somme. Piers disrupt this sedimentary system.
During storms, the shingle bar resulting from cliff erosion and which
protects the residential low coast can be flooded.
This phenomenon is alarming because it has been occurring more
frequently (map from 1990 showing how storm surge has affected several towns)
This surge extends very far into the towns and it appears that the shingle bar
no longer plays a protective role against the storm waves.
Do
such facts reveal a sea level rise, changing characteristics of storms, or
simply, the results of human action on the beaches’ sediment budget?
One
reason may be a progressive depletion of the sedimentary stock inherited from
post Ice Age.
This depletion is intensified by human action, which has modified the
volume, the distribution, and the advance of the shingle.
Indeed: -
· Since the beginning of this century, the population quarried fifty
percent of the shingle stock from between Antifer and le Tréport.
·
The harbour piers stop the shingle longshore drift,
causing deposition on the pier and a shingle deficit down stream thus
increasing the coastal erosion.
·
The longitudinal defensive engineering limits the
degree of freedom, or shingle bar morphological adjustment, to the storm wave
resulting in overflowing.
·
Can shingle production from the chalk
cliff erosion make up for the sedimentary deficit? If yes, over what time
period?
·
What kind of coastal defence policy must
we establish to minimise this sedimentary crisis?
(Should
we build coastal engineering structures that disturb the sediment budget or
should we imagine softer solutions, more environmental such as the by-pass
system or artificial nourishment of beaches?)
·
What is the morphological and sedimentary
adaptation of the shingle beach during a storm and particularly, how does the
sand fraction contribute to the internal structure of the shingle bar.
You
can see here the profile and the sedimentary characteristics of the beach of
Etretat, before a storm at the top, and after a storm at the down (refers to
diagram). You can see also, the influence of the longitudinal engineering
structure. Wave reflection on this structure gives the shingle beach a
particular profile.
To
understand the nature of the shingle beaches of the Rive Manche requires
quantification of their sediment budget.
(refers
to map)
When
we make an inventory of the storms which have created storm surges (in that
case in Fécamp and Etretat) it is possible to find the météo-marines
characteristics (especially the force of wind and the height of tide) at the
moment of the storm surge.
For these beaches, a westerly storm with an average wind force of thirty
knots and a height of tide of eight to nine meters, we know that these
conditions create a storm surge risk. In addition, this diagram shows that for
the same force of wind and same height of tide, we can observe either an
important storm surge or nothing. This means that there is a third factor that
determines the importance of the storm surge – namely the morphology of the
shingle beach.
Finally,
the main object of this program is to establish the sediment budget of the
shingle beaches of the Rives-Manche area.
The
first stage is:
Delimit the system. In Normandy it’s easy because the limits correspond
to the piers, which stop the shingle longshore drift.
For
each sedimentary unity (that is to say between two piers) and for each sub-unit
(that is to say between two principal groynes) it’s necessary to quantify the
existing shingles stock.
·
quantify the input sediments
- the
contribution of the chalk cliff erosion
- quantify the
input longshore drift
- and eventually,
the shingle volumes artificially brought
·
quantify the output sediments
-shingle hoisting
or quarrying
- quantify the
output longshore drift
- and eventually, the output shingles of the system.
(indeed, for example in the harbour of Dieppe, the pebbles can enter on
the harbour access channel where there are dredged and then thrown offshore.
This constitutes a loss to the sediment budget).
1/ the available data
In
1999, we completed a study of the coastal dynamics Normandy and Picardie using
photogrammetry.
In
1995 and 1966, the National Geographic Institute made a special aerial view
mission all along the coast of the Haute-Normandy at 1:2000 scale. This mission
allowed us to make photogrammetric measures.
For each mission, we obtained a numerical restitution of the planimetry
and the altimetry of the top of the cliff in 1966 and 1995 and in 1995 of the
shingle beach. In fact, we used points at each 10 meters along the top of the
cliff and eighty points per hectare on the shingle beaches. For each point we
know the latitude, longitude and altitude.
·
very good precision (one two thousand scale)
·
a small margin of error (plus or minus thirty cm)
With this method we obtain a numeric
restitution of the position of the coastline or the volume of the beaches
allowing us to create a geographic data bank comparable to that of the surveyor
or topographer.
We
can see here an example of this photogrammetric study (refers to map). On this
figure we have some results about the cliff retreat between 1966 and 1995. On
this second figure we can observe the shingle volumes in 1995.
Here,
the longshore drift is stopped by the Fécamp pier, generating a sedimentary deficit,
and an increase of the cliff erosion, and eventually an increase of the storm
surge risk for the adjacent valley.
In
this Interreg program, this geographic data bank stem from the photogrammetric
study will be our work base.
2.1/ For the sediment budget
·
For the stocks estimation:
- We will compare the data of 1995 to the old documents
- We will estimate this year with our GPS (global positioning system) the
volume of the shingle beaches (apart from under the cliff because it is too
dangerous)
·
For the shingle longshore drift
There is not much data for this. The method we used compares shingle
volumes at different periods. This method will allow us to estimate the annual
average longshore drift per area.
With the photogrammetric study we will estimate the cliff retreat
between 1966 and 1995 and this year, we’ll determine (with the GPS) the position
of the top of the cliff so that we can extend the calculation of cliff retreat
up to 2000. This work constitutes a part of the students study.
2.2/ Shingle beach morphological evolution and
cross-shore sand transport between the shingle bar and the low beach.
·
We will make topographic measures of
some beaches using the GPS before, during and after a storm.
·
At the same time, we will measure
the current on the beach and inject fluorescent sand on the low beach to
estimate the sand exchanges between the shingle bar and the low beach (again
before, during and after a storm).
Some results
(refers to figure) We estimated, between 1966 and 1995, the cliff
retreat on several sites. We can observe that there is an important spatial
variability of the rate of erosion and consequently an important variation of
the shingle volumes produced.
Regarding the question of the capacity of the cliff retreat to make up
the actual sedimentary crisis:
On a test
site (Dieppe/le Tréport) we estimated, using historic documents, the nineteen
forty-seven shingle volumes. If we imagine that at this date the beach sediment
budget was in equilibrium, this means that it would take ninety years with only
cliff erosion to establish the 1947 shingle volume.
Investigative
approaches and preliminary findings: an English perspective
Shingle
has been moving along the Sussex coast for centuries as evidenced by its volume
fluctuations at specific places such as The Crumbles. This movement is a
natural process in response to the energy input (waves) and its direction, the
geometry of the coast and the materials forming the coast.
Shingle
becomes an issue when the amount in one area increases, as for example across
harbour entrances, or when the amount decreases and low-lying areas behind the
shingle beach become vulnerable to flooding. This is even more when man
interferes with the natural processes in an attempt to halt the shingle
movement locally by the constructions of groynes. Trapping of shingle impedes
longshore transport causing a reduction of shingle input further along the
coast leading to beach erosion. The best example is probably the construction
of the Newhaven harbour entrance breakwater and the following shingle and sand
depletion on Seaford beach.
Along
the Sussex coast “erosion of shingle beaches is widespread” (Lewes District
Council 1996, page6) and so a decline in shingle beaches is thought to be the
more prominent of the two basic problems. To further investigate this problem
we need to quantify the erosion and shingle movement, especially movement out
of the system.
Shingle
can be contributed to the coast by rivers from the hinterland, from offshore
and from cliff erosion (reworked sediments or from primary sources such as the
Chalk). Apart from these natural processes shingle has been brought into the
coastal zone as artificial beach nourishment or as a by-product of other
coastal defence work (see below).
The
second process is difficult to investigate especially with regard to past
supplies. Presently there seems to be little or no shingle moving into Sussex
from offshore. However, shingle deposited on the Channel bed in glacial times
is likely to have moved towards the coast with rising sea level so that part of
the present shingle found on beaches might have come from this source.
The
third process is the erosion of the cliffs and the shore platform, especially
the Chalk cliffs. In calculating the amount of shingle injected from the cliffs
two problems have to be addressed. Firstly, the amount of cliff retreat along
different stretches needs to be estimated and secondly the amount of flint
within the Chalk has do be determined.
So
far only broad estimates of the amount of flint within the Chalk have been
given in the literature. One figure is 5% (a figure attributed to R. Mortimore,
University of Brighton) without taking into account the varying contribution of
Chalk beds with varying flint content to each stretch of coastline. The other
is attributed to Posford Duvivier Consultants (1993) which gives a figure of 2%
for the Seven Sisters. These figures are quite different from each other and no
methods are given for deriving these figures. Given that the flint content is
likely to differ between Chalk beds and that the various Chalk beds are exposed
to a different extent along different stretches of the coast a more detailed
survey is clearly necessary, including the flint eroded from the Chalk shore
platform too.
The
cliff retreat rate is the second variable in the volumetric calculation and
needs to be determined on a similar scale as the flint content. Difficulties
exist with the varying quality of former maps and air photograph coverage has
so far only extended into the 1950s. Data supplied by the Environment Agency
together with air photos made by the German Luftwaffe and now held in the
US-Archives which we have ordered will make it possible to extend
high-resolution records back into the early 1940s.
That
the cliff retreat rate as well as the flint content needs to be determined as
accurately as possible can be seen from the example calculations in (Table 1).
Table 1: Flint injection into the coastal area from Chalk-cliff retreat
in East Sussex, example calculations
|
Cliff Length |
Average Cliff Height
|
Average Annual Cliff Retreat |
Average Flint Content |
Annual Flint Injection |
|
25km |
45m |
.4m |
5% |
22.500m³ |
|
25km |
45m |
.4m |
2% |
9.000m³ |
|
25km |
45m |
.5m |
5% |
28.125m³ |
|
25km |
45m |
.5m |
2% |
11.250m³ |
|
Phase
III |
200.000m³
from cliff trimming |
5% |
10.000m³ |
|
Though
the natural cliff retreat is probably the main contributor to flint input
artificial input has taken place in the past. Artificial shingle input can be
quite large and includes, apart from the beach nourishment schemes, material
dumped from other coastal works. During the construction of the seawalls
between Brighton and Peacehaven the material from the cliff trimming was dumped
on the shore platform amounting to 200,000m³ of chalk for Phase 3 alone (a
stretch of 670m, Stammers 1982). It very likely that material from the previous
construction phases has been dealt with in a similar manner.
Generally
the sealing off of cliffs through the construction of sea walls, such as those
along the stretch from Brighton to Peacehaven, will reduce the natural flint
input.
With regard to the shingle output BERM will concentrate on natural
processes taking into account the amount removed by commercial and government
agencies.
The
literature perspective:
Flint
(and its paler counterpart chert) are assumed to be quite resistant to abrasion
in a coastal environment (e.g. Bray 1997) and early experiments in a wave simulator
state that ”rounded chert is ten times as resistant as quartzite” (Kuenen 1964:
29). It is suggested that on a beach with long periods of burial it would take
“a thousand years for chert to form an ellipsoid” (Kuenen 1964: 42). Latham et
al. (1998) show a negligible weight loss for rounded flint of pebble size in a
tumbler experiment over 200,000 tumbler revolutions. Bray (1997: 1041) found
that “freshly supplied angular cliff gravel suffers an approximate 10% loss
within the first year on the beach, whilst well rounded pebbles are abraded
very slowly”. Past research therefore suggests that flint is a highly durable
material that is abraded at extremely slow rates.
However,
even next to fresh cliff falls one rarely finds freshly broken material on the
beach suggesting that it is quickly rounded. This observation calls into
question the supposed durability of flint.
Experiments:
Preliminary
experiments by BERM with a rock tumbler with a diameter of 22cm at 28rpm seem
to paint a quite different picture. Well rounded flints taken from a beach east
of Peacehaven have been tumbled for over 300 hours (over 12 days) causing and
overall reduction in weight of the order of 20%.
Angular,
freshly broken flints with razor sharp edges taken from a recent cliff fall
have had their edges broken within minutes in the tumbler became appreciably
rounded after about 100 hours.
The
products of attrition are silt-sized particles or a rock flour that would be easily
washed away with even the weakest currents. The larger fragments chipped of the
of edges in the first minutes generally weigh less than 0.1g and can have any
form though platy forms dominate.
Although
the experiment is essentially a pilot study flint attrition seem to depend on a
variety of factors, further experiments are unlikely to show significantly less
abrasion.
Certainly,
the experimental results need to be calibrated to reflect attrition in the
field, but the lack of sharply angular shingle next to recent cliff falls and
the field observation that freshly broken flint loses its sharpness over night
(i.e. in one tide), indicate that flint is less durable than is commonly
supposed.
The
lack of adequate data about the input of shingle into the coastal system and
the surprising preliminary results about possible rates of attritionwill take
the study of a shingle budget an important step beyond shingle budgets based
mainly on longshore drift.
Cited
literature:
·
Bird, C. F. 1996: Lateral grading of beach sediments:
a commentary. Journal of Coastal Research 12(3): 774-785
·
Bray, M. J. 1997: Episodic shingle supply and the
modified development of Chesil Beach, England. Journal of Coastal Research,
13(4): 1035-1049
·
Kuenen, P.H. 1964: Experimental Abrasion: 6. Surf
action. Sedimentology 3: 29-43
·
Latham, J.-P. , J.P. Hoad and M. Newton 1998: Abrasion
of a series of tracer materials on a gravel beach, Slapton Sands, Devon, UK. Advances
in Aggregates and Armourstone Evaluation. Geological Society, London,
Engineering Geology Special Publication, 13: 121-135
·
Stammers, R.L. 1982: Coastal defence engineering in
East Sussex – Part II. Municipal Engineer, 109: 353-359