Managing the Coast

Sandsfoot Castle, built in 1579 in the centre of a field, and surrounded by a moat, according to some writers was gradually undermined by the sea. By the time of this early twentieth century photograph, parts of its wall had already been lost (Dorset County Museum).

However, with the growth from the mid-nineteenth century of the resort and cliff-top settlements at Swanage, Bournemouth, Southbourne and Highcliffe, cliff retreat and active landsliding became a problem. Construction of sea walls (or promenades) cut off the supply of sand from the retreating cliffs and beaches began to diminish. The new resort of Southbourne–on-Sea was built and destroyed within 30 years. In contrast, Lyme Regis continues to rest on an active and extensive landslide.

The Royal Commission on Coast Erosion and Afforestation reported in its Final Report (1911) that, although in Dorset and Hampshire there had been a net gain of land from the foreshore, there had been a net loss of foreshore to the sea. Put another way the foreshore was becoming steeper. Although later writers all comment on the Royal Commission’s conclusion that “far more land has been gained by accretion and artificial reclamation in recent years than has been lost by erosion” (Final Report : Vol. III, 79), the potential problem of the steepening foreshore is not recognised until the 1970s (May 1973; Bird and May 1976). The problem of increasing loss of beaches was missed by coastal engineers and, even after the 1953 North Sea floods, it was also ignored, especially as the effects of storm surges were emphasised.

The Minutes of the Royal Commission record that considerable concern was expressed about the effects of removal of shingle from the beaches in West Dorset and the loss of land at Highcliffe in the East.

Shingle being removed from the beach at West Bay in the early twentieth century (Dorset County Museum).

During the same period as the surveys carried by the Ordnance Survey for the Commission, only one Dorset local authority (Swanage) had a loan (£4600) sanctioned by the Local Government Board for coastal works. The Ordnance Survey had, however, identified erosion at Portland and Weymouth and throughout the cliffs between Christchurch Bay and Poole Harbour.Nevertheless, development accelerated on the cliffs between Canford Cliffs and Southbourne.

Today Shoreline Management Plans for the Dorset coast provide detailed summaries of the protective structures and the forms and processes along this coast. Current observations of coastal change east from Portland Bill are on the Channel Coast Observatory site (http://www.channelcoast.org). May and Hansom (2003) summarise the coastal geomorphology for much of the coast west of Poole Harbour.

Landslides affect much of the Dorset coast, and other parts are affected by coast retreat and coastal flooding. There are major landslide complexes at Lyme Regis, Black Ven, Charmouth, Stonebarrow Hill, Golden Cap, Abbotsbury Castle, Portland, Osmington, White Nothe, Lulworth, Worbarrow, Gad Cliff, Houns Tout, St Aldhelm’s Head, Durlston, Ballard Down, and Highcliffe.

Smaller features exist along other parts of the coast. Some landslides are very active and great efforts have been made to control them.

A former road between Lyme Regis and Charmouth has been carried away by the landslides of the Spittles and Black Ven landslide system (Dorset County Council).

Several of the present-day landslides had roads or tracks across them in the nineteenth century (and often much earlier) suggesting at least temporary periods of stability. These undercliffs were also often well covered by shrubs and trees.

The most detailed investigations have occurred at Lyme Regis, Seatown, West Bay, Portland and Swanage, where landslides threaten many properties. The most comprehensive geomorphological investigations have been around Lyme Regis, including the very large landslides to the west in Devon. Built upon extensive landslides, the town itself has had the most intensive investigations, probably in Europe.

As early as the sixteenth century, the inhabitants of Lyme Regis expressed their concern that the sea was taking away the town.

“…the banks and walls of the main coast, whereupon your said town is situate is now in such wise undermined broken consumed and wasted…..so that divers dwelling houses be overthrown consumed and utterly lost and many more like to be in short time unless speedy remedy be had…”
Mayor and Burgesses of Lyme Regis 1533 petition to the Lord Chancellor,
Thomas More, quoted by Royal Commission on Coast Erosion and Afforestation

Their plea for help to prevent Lyme Regis from gradually falling into the sea was rejected. Despite this, people continued to build on its landslides. Like their later counterparts to the east in Bournemouth and Southbourne, they preferred to attempt to hold back the sea and prop up the landslides.

Contemporary reports show that in both the sixteenth and eighteenth centuries, there was a large continuous beach along the Lyme Regis shore. The Cobb, which protected the harbour, was not connected to the mainland and shingle was able to travel eastwards without significant interruption. However, by the late twentieth century, the beach was affected by three related conditions:

The alongshore movement of sediment (mostly shingle) was interrupted first when the Cobb was connected to the land in the 1750s. Monmouth Beach built up on the western side of the Cobb. The supply to Monmouth Beach was itself reduced when the Humble Point landslip to the west interrupted longshore transport in the 1840s.

Aerial photo of Lyme Regis showing the Cobb, the build-up of shingle to the west on Monmouth Beach and the groynes and seawall of Marine Parade. Note also how far back the sea has cut into the old landslides east of the Cobb compared to the west (Dorset County Council).

Monmouth beach continued to build up against the Cobb but at the expense of the western part of the beach. Throughout the Dorset coast, beaches have been periodically fragmented by the landslides which have spread across the beach and foreshore. Beach sediment accumulates for a time and then, as the landslide toe is eroded away, beach material is released to travel eastwards. Headlands also separate beaches as the shoreline retreats under the effects of erosion and sea level rise. Construction of harbours, groynes and sea walls has also increased the fragmentation of the beaches. Although the volume of material in the beaches has been declining naturally, human activity has also reduced the supply. Shingle was collected from many of the beaches for industrial use and for pebbledash on houses. As cliffs are protected against erosion, the supply of sand or shingle is cut off and beaches decrease in volume.

At places such as Lyme Regis, these effects have all reduced the volume of beaches needed to prevent erosion of the toe of the inherently unstable slopes of the old landslides. In addition, 19th century quarrying of limestone from the shore platforms as well as the lower cliffs (for example in the East Cliff area) made wave attack easier and weakened the lower cliffs. A similar effect occurred at Hengistbury Head when ironstone boulders were removed. However, at Lyme Regis, the landslides which extend over a kilometre inland are not modern in origin but rather reactivated features from the last interglacial, as are the landslides at The Spittles and Black Ven (Brunsden and Chandler 1996).

The landslides between Lyme Regis and Charmouth have probably been more comprehensively investigated and the complex inter-relationships between active mass-movements, marine erosion and beach development better understood than almost any other coastal landslides. The basal shear surfaces of the principal landslide units are strongly controlled by strata within the Lias, especially as they dip gently south-eastwards, giving landslide blocks an enhanced tendency to move seawards. Many of the landslides are at or close to failure, according to the stability analyses carried out as part of the Lyme Regis Environmental Investigations. The slides have four main forms:·

Mudslides: ‘in which masses of softened argillaceous, silty or very fine sandy debris advance chiefly by sliding on discrete boundary shear surfaces in relatively slow moving, lobate or elongate forms’ Brunsden 1984).

Translational block slides ‘characterized by markedly non-circular slip surfaces forming a combination of a steep, curved or planar rearward part and a flatter sole ‘( Hutchinson 1988).

Deep-seated compound slides, in which the slip surfaces are many metres below the surface. The slides comprise many such surfaces and often give rise to a typical bench scarp morphology.

Shallow rotational or translational failures which although shallow can produce significant changes in the land surface. They are common in the superficial deposits of landslide debris and in the cliffs under direct attack by the sea

As the landslides system of Black Ven and the Spittles expands westwards, eastern parts of Lyme Regis may be increasingly threatened by reactivation of the older landslides.

Finally, even when the cliffs and landslides have been protected, the structures and beaches need to be maintained to ensure that walls are not undermined or outflanked by erosion. By the end of the 20th century, it was evident that at Lyme Regis this was a major problem. For example, the eastern end of the sea wall at East Cliff was being outflanked by marine erosion and the foundations of the Marine Parade sea wall were seriously undermined.

The seabed and shore platform in front of the sea walls at Lyme Regis have undergone considerable erosion and lowering over the last two centuries, such that the sea walls’ exposure to wave attack has increased (Dorset County Museum).

The landslides affect everyday life in Lyme Regis where West Dorset District Council has advised property owners how to minimise potential landslide movement around their property. For, although the Council deals with the landslides on public land, the responsibility of dealing with the effects of landslide movements rests with the landowner. Because groundwater in the ground is a major cause of land movements, the advice to residents includes:

Make sure gutters and downpipes are not leaking and that drains and ditches are properly maintained;

Avoid loading or cutting into steep slopes as this may destabilise them.” (WDDC press release 3 October 2000)

In contrast on the Isle of Portland there is also a long history of landslides with records as early as the 12th century, but few remedial works have been undertaken. Why? There appear to be three reasons. First, the landslides are different in character dominated by large block falls and failures in the Portland Stone. Second, their locations have largely been avoided as places for construction except on the north end of Portland and, third, the largest areas of landslides are relic in nature and generally were not seen as being a serious threat. Some of the landslides may have been re-activated by the very large quantities of quarrying waste dumped on to them (Brunsden et al 1996). There are five different patterns of landsliding around Portland, summarized in Table 2. The axes of the Shambles Syncline and the Purbeck Anticline control joint systems which run NNE-SSW, N-S and E-W (Goudie and Brunsden 1997). The resultant weaknesses determine the form and behaviour of the landslides, as well as the pattern and success of the quarrying activity.

Table 2 Landslide systems on the Isle of Portland (based on Goudie and Brunsden 1997)

The East Weare landslide of 1734, formed of very large joint-controlled en echelon failures, is the second largest historical landslide in Britain. The toes of some of the slides extend offshore and probably occurred when sea levels were lower and conditions were wetter during the last glaciation (Goudie and Brunsden 1997).

When landslides occurred here as elsewhere they were often attributed to earthquakes or to acts of the Devil, especially when columns of smoke rose from them. They were places to avoid. By the nineteenth century, although these beliefs still affected peoples’ attitudes, the landslides had also become phenomena to visit and be amazed by. For example in 1827 at Burning Cliff just east of Ringstead,

“dry sticks, or any inflammable substance, thrust into one of the apertures in which the fire was perceptible, would kindle immediately, and burn with flame, though none has issued spontaneously since. This interesting appearance was visible five or six days, and would probably have remained so much longer; but many portions of the rock being removed by curious visitants; a large quantity of the upper part of the cliff fell down,.. “'The Complete Pocketbook or Gentleman's and Tradesman's Daily Journal for the year of our Lord 1828. (DRO PE/LR/RE 4)

Some large landslide complexes, for example White Nothe and around Chapman’s Pool, have received very little attention: they are also much less active. Why? Landslides depend upon the rock type and structures, groundwater, removal of material from the foot of the slope. Slopes that are protected from erosion by the sea and well drained are less likely to fail. In many of these landslide systems, the Chalk or Portland Stone upper cliffs provide boulders which are transported by the slides to the cliff-foot. They form a natural rock armouring for the weaker materials there. Some landslides have remained stable because they have not yet been attacked by the rising post-glacial sea. Others have become active as the sea has eroded their toe or groundwater conditions have changed. In others the gradual localised removal of the cliff-foot boulders has allowed parts of the slides to be re-activated.

St Aldhelm’s Head in 1987 shows well the very large accumulations of Portland Stone boulders protecting the Kimmeridge Clay at the foot of the cliffs. The upper debris slope on the right of the image is mainly derived from quarrying at the cliff top. In the background to the left are the more active Houns Tout cliffs where the Kimmeridge Clay forms over half the cliff face

Even without landslides, coastal erosion puts properties and assets at risk. Christchurch, Bournemouth, Poole and Weymouth are almost entirely defended against erosion and flooding by sea walls and beaches retained by groynes. The more rural coast is mostly without man-made defences, but the coastal towns are most at risk from erosion and flooding. Between 1990 and 2002, new or upgraded coast protection schemes were implemented for example at Ringstead, Lyme Regis, Seatown, Mudeford Spit and Charmouth. Particular hazards from flooding by rivers and the sea exist at West Bay , Chiswell, Christchurch and Weymouth. Protection schemes against coastal flooding have been built at Chiswell where the natural structures have been reinforced and Preston where a sea wall has been in place for many years. Christchurch has protection against flooding by the sea and rivers, mainly by the construction of new walls and raising of existing walls. At Hamworthy, a scheme proposed in 2004 would construct flood defences back from the shore. In some locations, however, the long-term sustainable solution may be to allow flooding.

Preston - the road north from Weymouth across Lodmoor was regularly covered by shingle washed over the beach crest. This image by E.V. Tanner dated about 1925 shows the old sea defence and the shingle overwash fans (Dorset County Museum).

Coast protection ranges from ‘hard’ engineered solutions such as sea walls, groynes and harbour walls to ‘soft’ protection solutions such as beach replenishment and salt marsh management. A consistent principle of coastal engineering is that the best defence for any coast or engineered structure is a beach which is wide and high enough to absorb wave energy.

Coast protection needs careful design of protection works. They must meet economic, engineering and environmental criteria before they are publicly funded. Because coast protection works are publicly funded, they should be cost-effective. Government funding will not be made available unless the benefits of protection exceed the costs of protection. Although most funding is provided under the Coast Protection Act 1949, most works are only partly funded by DEFRA (usually 45-55%) and so the costs are borne by both the taxpayer in general and the local Council Tax payer.

Funding is based on national priorities, how urgent works are and their economics. So schemes that are urban (including beach management), have a high risk of short-term failure and have high benefit-to-cost ratios are more likely to be funded. Schemes must be based on an understanding of natural processes and, as far as possible, work with those processes, be consistent with shoreline management plans and be environmentally acceptable.

Aerial photo of Bournemouth Pier which shows the beaches strongly aligned to waves from the south-east and indicating a movement of sand and gravel towards Poole. This is not the normal pattern, but has to be taken into account when designing groynes and beach replenishment programmes (Bournemouth Borough Council).

If alongshore sediment transport dominates, the beach volume necessary for protection must be retained. This includes providing a continuous supply of sediment or constructing structures which retain the available sediment. Alternatively, land uses behind the shoreline should be planned and managed in ways which allow for the impacts of major storms or erosion. Most British urban coasts have assets which do not allow for any loss of backshore land. Protection needs therefore to be seawards of the shoreline.

Wooden plank groynes are a well-tried form of beach control structure, but they need to be maintained otherwise beaches can quickly be depleted. In 1930, the groynes at West Bay were retaining very little beach material and the foot of the sea wall was at risk of being undermined (Dorset County Museum).

Many alternatives to traditional ‘hard engineering’ methods of coast protection have been advocated and tried, including offshore breakwaters, surf reefs, rock groynes, beach nourishment, beach de-watering, habitat reconstruction, artificial seaweed banks, managed retreat, setback of existing properties and do-nothing. Many of these methods have been tried around the world. Artificial seaweed banks were tested in the 1960s at Bournemouth based on the observation that kelp beds can be very effective in reducing wave activity. Artificial seaweed was deployed but with little success. Setback of existing properties is sometimes possible. For example, locally, the National Trust has moved beach huts at Studland back from the edge of the rapidly eroding dunes rather than keep attempting to hold the line. Some have succeeded, others have not. Costs are not necessarily lower. They are not always transferable to different tide or wave conditions: the fact that a particular system has been very successful in one location does not mean that it will work effectively at a different site.

In this 1988 image, the beach at Swanage had built up above the level of the groynes. In later years, this beach gradually decreased (Dorset County Museum).

Extreme events, however, can damage, overtop or destroy well-designed and normally effective coast protection. With rising sea levels, existing protection structures and beaches are increasingly likely to be overtopped. Both the costs of protection and the value of property at risk are rising.

The debate about different methods is not a new one. In the early part of the twentieth century, the Royal Commission on Coast Erosion reported that

“ The erosion …would have been far more serious if extensive works of defence had not been constructed by local authorities, railway companies, and others, at great cost, though on the other hand, such works in many places have been responsible for erosion of the neighbouring coasts by interfering with the normal travel of the beach material.” (Royal Commission on Coastal Erosion and Afforestation 1911 p 158).

The division of responsibilities for coast protection amongst local authorities means that administrative boundaries can split natural coastal systems (see Theme 3 Topic 3 Who’s Who in Coastal Management). In the past, this sometimes meant that little account was taken of the natural integrity of beach systems. SMPs are expected to provide a more strategic view of coastal processes and their management. They are prepared by groups of local authorities: DEFRA expects that coast protection proposals are consistent with SMPs.

SMPs identified future strategies for coast protection by allocating management units to one of the following key categories:

Advance the line: i.e. carry out works, reclamation or replenishment which moves the shoreline seawards of its present position

Retreat the line: accept that retreat is occurring and take steps to carry out managed retreat of the shoreline and the associated assets.

Hold the line: ensure that the existing alignment is maintained. This usually means new coast protection works or more commonly, maintenance or enhancement of the existing works

Do nothing: allow the natural processes to continue regardless of the potential losses of land or property. The costs of protection will outweigh the benefit to be gained or there is likely to be significant environmental damage as a result of carrying out the works.

Any proposal has, however, to have planning permission from the local authority, as well as permission from the Crown Estates for structures which extend on to the seabed.

Often combinations of different methods are used to provide resistant structures, to protect them and to build up attractive recreational beaches. At the beginning of the twentieth century, Sandbanks was just that – an area of sand dunes. There were two coastguard cottages built in 1850 and a few wooden holiday homes. During the 1910’s, the landowner, the Wimborne Estates, sold a small number of plots of land and the first permanent homes were built. In 1929, the whole of the beach and present recreation ground (just over 5.2 ha) were purchased from the Estate for the sum of £13! The area now has some of the most expensive residential properties in Britain and these need to be protected against the risks of erosion and flooding.

At the mouth of Poole Harbour, there are works to protect
· the frontage of Brownsea Island,
· the shore of Sandbanks,

· a training bank to regulate the flow in and out of the harbour as well as
· private structures to defend some of the properties at North Haven

The works here include sea walls, rock groynes, beach nourishment and habitat reconstruction to rebuild dunes at Sandbanks (Dorset County Council).

This coastline was recognised by the Royal Commission on Coast Erosion (1911) as an eroding coast, but for the previous 300 years there had been much more concern about the mouth of Christchurch harbour. During the nineteenth century, there was a similar debate about the mouth of Poole Harbour. Both involved ways of making the harbour more accessible and in later years included many plans to construct very long jetties and in Poole’s case to extend the quays and rail access (see Theme 2 Topic 4 The Influence of the Sea).

In 1676, Andrew Yarranton put forward a scheme to improve Christchurch harbour. Because of “the ground Tide about the point, which had carried and lodged the Sands, so that it had choked up the Harbour” he proposed cutting a direct deep water channel from the harbour to avoid the existing difficult narrow entrance –“The Run”. By 1698 the channel had been cut through Mudeford spit about 600 metres from Hengistbury Head and was in use. A jetty on the northern side of the channel used ironstone boulders taken from the Head. With sand moving from the south (probably in increased amounts because of the newly exposed cliffs from which he had retrieved the boulders), the channel became blocked, despite the river’s flow and dredging, and was abandoned in the 1730s after several storms. The jetty is still visible as Clarendon or Long Rocks.

In 1762, Smeaton commented on the severe erosion of the Head. He planned to improve access into Christchurch harbour and aid navigation from Christchurch to Salisbury via the river Avon. Unlike Yarranton, however, he intended to build two jetties, the southernmost first, using the ironstone boulders. The estimated cost was about £6000, but the scheme was never carried out. Fortunately, this meant that further accelerated erosion was avoided – or at best delayed for another 100 years. For by the 1840s, the ironstone boulders had a different value, as they could be used both as ballast for colliers carrying coal from South Wales and as a source of iron for iron making.

A local coal merchant, George Holloway, removed ironstone ‘doggers’ from the beaches on the south side of Hengistbury Head as well as dredging offshore. A previously somewhat slow rate of erosion accelerated. Large quantities of sand were carried into Christchurch Bay and the spit extended rapidly towards Highcliffe. At the same time, the rapid retreat of the headland began a process of re-alignment of the beaches and cliffs of the eastern shore of Poole Bay. Although removal of the boulders from the shore was stopped in 1856, the accelerated erosion continued until the construction of the Long Groyne in 1938.

Often the only measurement of coastal change came from losses of buildings or other structures. At Double Dykes, the shoreline retreated by almost 50 m in the late nineteenth century, according to this image (Bournemouth Borough Council).

Erosion , Development and Protection at Highcliffe and Southbourne - Success and Failure

To the east, at Highcliffe, although there was considerable cliff erosion, there was also concern about the lengthening Mudeford spit. In evidence to the Royal Commission on Coast Erosion, the then owner of Highcliffe Castle said that the original site of the house was about two miles (3.6 km) out to sea. According to him, during the early eighteenth century, 2 or 3 poles (10 to 15 metres) of land was lost at a time and as much as 2 or 3 acres (around 1 hectare) annually. This rate of loss, if it is to be believed, is about 7 times faster than the worst losses at Hengistbury Head. He also complained that his land was only worth £1 an acre compared to the £100 per acre he could get for land at Bournemouth. By the early nineteenth century, erosion of the cliffs was rapid and frequent falls occurred (see Theme 2 Topic 4 The Influence of the Sea). When Highcliffe was bequeathed to the Earl of Bute’s fourth son, Charles, the expense and annoyance of the landslips was so intolerable that he sold the estate (Dale 1914).

Louisa, Marchioness of Waterford, having known since childhood at Highcliffe Castle the erosion problem, attempted to reduce the speed of land-loss. She oversaw probably the earliest local construction of a rock groyne formed of limestone and "granite-porphyry" blocks (still in place in 1880), and slabs of the Barton Sands basal "Shell Bed" (Burton 1931). By this time the Run was diverted along the Highcliffe frontage by the extended Mudeford Spit. The structure was intended to divert the extended channel of the Run outward rather than directly preventing marine erosion of the beaches and cliffs (West 1885). Only a few of the stones were visible at low tide by 1931 (Burton 1931), although they may have been buried by the sand which accumulated around the outlet of the Run near Highcliffe Castle.

Louisa also recognised that the continued failure of the clay cliffs was caused by landslides in the clays and had extensive drains placed in the cliffs to drain the water to the beach and so keep the clays dry. It is interesting to record that the techniques of her early coastal defence scheme have been followed in modern cliff drainage and the construction of rock groynes. The Highcliffe Castle protective works were largely a successful response to the serious problem of land loss. At Southbourne , however, speculative development of a new resort was caught out by the changes of the overall pattern of the shoreline as Hengistbury Head was rapidly cut back and early protection of the Bournemouth cliffs began to reduce sand supply.

The Southbourne Land Company began construction of the 0.5 km long Undercliff Parade at Southbourne-on-Sea in 1883. It was opened in September 1885, thus pre-dating the Undercliff Drive at Bournemouth. It cost £15,000. The promenade consisted of a curved roadway 12 m wide and a 5 m pathway along its seaward edge. It stood about 2.5 m above beach level, with foundations from 1.8 to 2.7 m below it. The base of the wall was 2.1 m thick, narrowing to 1.2 m at the top. A 100m long pier was built in 1887.

Houses on Promenade

A terrace of six large houses were built at the foot of the cliffs behind the promenade (Bournemouth Borough Council).
Carriages could drive on to the promenade by descending a gently sloping central approach road.

The sea wall was breached and the pier damaged severely in gales on 28th December 1900 and 3rd January 1901. Repairs to the promenade were possible but the Southbourne Land Company was unable to finance the cost of repairs. Severe gales and heavy seas in the winter of 1902 smashed holes in the promenade and Undercliff Parade and severely damaged the houses. They were demolished later in 1902.The pier was dismantled in 1907. The remains of the sea wall and promenade gradually collapsed and the cliffs were exposed to further erosion.

Houses were built on the cliff top and because erosion continued the story of the impact of coastal change does not stop here. In 1949, just as the Coast Protection Act came into being, there were extended discussions between Bournemouth Council and the Southbourne Cliffs Association (SCA). Nine properties on Southbourne Overcliff Drive were immediately endangered, but the SCA members argued vehemently against having to pay coast protection charges. By February 1954 the cost of various Works Schemes exceeded £1.5 million. As late as 1972, cliff face erosion was still occurring and the council decided to grade and drain the cliffs with further losses of cliff top land. What of the nine houses at risk in 1949? Six have disappeared, the remaining three remain right on the cliff edge.

Summary

The impacts of landslides and coastal change along the Dorset coast have been met differently by different communities and landowners. Disputes about the causes, responsibilities and payment for protection works occur throughout the past four centuries, but the examples described above give only a brief summary of the details.