Purbeck: Redcliff Point to Durlston Head

Introduction - References Map

This is an internationally celebrated coastline, recently declared a UNESCO World Heritage Landscape for its Jurassic/Cretaceous geology and coastal geomorphology. Textbooks on physical geography, at all levels and in many countries, use the landform sequence between Warbarrow Tout and Bat's Head to derive an evolutionary model of coastline development. The elegance and simplicity of this approach has been its main strength. In reality, each 'stage' is a function of site-specific-interaction between pre-existing fluvial relief; Pleistocene climate and environmental change, in particular former periglacial weathering and mass movement; Holocene sea-level rise and neotectonics (Brunsden and Goudie, 1997; Bird, 1996; Canning and Maxsted, 1979; Mottram, 1972).

The geological structure, rock sequence and range of rock lithologies are the most influential factors determining the several dramatic changes in cliff height and morphology along this 55km coastline. This relationship has been described in detail in several textbooks, monographs and papers (e.g. Arkell, 1947; Davies, 1956; May, 1971; Bird, 1996; King, 1976; Perkins, 1977; House, 1958, Nowell, 1997). Basically, the coastline trends nearly parallel to the approximately east to west axis of the Weymouth-Purbeck monocline. Compressional structures increase in intensity between the eastern unit boundary at Durlston Head (Photo 1) andits western counterpart at White Nothe (Photo 2), resulting in progressive westwards narrowing of the outcrop areas of the main rock groups that occur south of the Chalk. Rock dips are nearly vertical, with some overfolding, between Bat's Head and Lulworth Cove (Photo 3), but decline to low angles east of St Aldhelm's Head (Photo 4). Subsidiary fold, fault and other tectonic structures are responsible for several local complexities of rock outcrop pattern, especially west of Ringstead Bay. The Cenomanian overstep between Bat's Head and White Nothe where Chalk overlies softer strata is responsible for the stratigraphical succession that has promoted large-scale semi-rotational slope failure. The behaviour of the different rock types under compressional stress has created a variety of structures that contribute to differential rock resistance to marine and sub-aerial processes. Examples include the brittle fracture of the incompetent Purbeckian strata, best appreciated in the exposed section of the drag fold ("Lulworth Crumple") at Stair Hole; and several thrust planes in the Chalk producing brecciation and crushing (Nowell, 1997).

It is the view of several researchers that cliff morphology is more the product of long-term sub-aerial denudation than marine processes, although the latter function to remove fallen debris from cliff toes (e.g. May, 1971; Bird, 1996; Brunsden and Goudie, 1997; Allison, 1986 and 1989; Jones, et al., 1984). This is readily demonstrated at the various sites of active mass movement. Almost the entire range of failure mechanisms operate, at widely varying frequencies and magnitudes (Allison, 1986; Brunsden and Goudie, 1997). The major landslip sites, e.g. White Nothe; Emmett's Hill/St Aldhelm's Head (Photo 4) and Houns-tout cliff (Photo 5) have not received the same level of documentary research and process monitoring as those further west on the Dorset coast. Most mechanisms of failure, including toppling, shallow sliding and rock fall, are not yet fully understood (Brunsden, 1996).

The wave climate of this coastline is poorly understood. A recent modelling study, based substantially on hindcasting using regional wind records (1975-1996) concludes that there is a strong contrast in exposure to wave energy between east and west sectors (HR Wallingford, 1998). The dominant wave directions are from the south-west and south-south-west, thus the coastline west of White Nothe is relatively protected by the presence of the "Isle" of Portland and offshore banks and shoals, such as the Shambles and Adamant shoals (HR Wallingford, 1994; Bastos and Collins, 2002). These latter features, and the Lulworth Banks, also set up nearshore wave refraction. Maximum extreme wave heights are predicted to occur offshore Kimmeridge Bay, which is open to a fetch of over 8,000km (extending to the Venezuelan coast of South America). The decline in incident wave energy eastwards from Kimmeridge is only slight, but is modified by changes in coastline orientation, such as the north to south trend of the west-facing shoreline of the St Aldhelm's promontory. Deep water waves are transformed as they enter the nearshore zone by offshore bathymetry. This, and the presence of inshore reefs and ledges at various locations e.g. Ringstead Bay (Photo 6) and around Kimmeridge (Photo 7), sets up complex local wave refraction. HR Wallingford (1998) state that mean significant wave height in Ringstead Bay is 0.5m; this is confirmed by Bastos and Collins (2002) based on hydrodynamic modelling. There are no records of storm surge conditions, but HR Wallingford (1998) have modelled extreme significant wave heights, for a range of return frequencies, for Ringstead Bay and the entrance to Lulworth Cove. Thus, for a 1 in 1 year recurrence, heights are calculated as 3.1m and 4.2m, respectively. For 1 in 10 years, expected extreme values are 3.9m and 5.1m; and for 1 in 50 years, 4.5m and 5.6m.

Ringstead Bay was one of the locations for which wave modelling exercises were undertaken as part of the DEFRA Futurecoast Project (Halcrow, 2002). An offshore wave climate was synthesised based on 1991-2000 data from the Met Office Wave Model and then transformed inshore to a prediction point in Ringstead Bay at -2.93m O.D. Potential sensitivities to likely climate change scenarios were then tested by examining the extent to which the total and net longshore energy for each scenario varied with respect to the present situation. Results suggested that a one to two degree variation in wave climate direction could result in a 3.5% to 7% variation in longshore energy and confirmed that the Bay was significantly more sensitive to this factor than many south locations. Wave energy at Ringstead was also found to be especially sensitive to sea-level rise and storminess. It may be that at higher water levels the dissipative function of the protective nearshore reefs becomes reduced significantly.

Tidal currents are comparatively weak except at the major Durlston and St Aldhelm's headlands (Bruce and Watson, 1998; Mouchel, 1998) and also where flows are confined between the shoreline and nearshore/offshore reefs, as at Ringstead Bay (HR Wallingford, 1998). Current vectors are parallel to the coast and reverse before and after high water. A slight net westwards flow, towards Weymouth Bay, operates west of Warbarrow Tout. St Aldhelm's Head, creating a tight clockwise movement in the sea area immediately west of this promontory, sets up a tidal vortex.

Sediment transport is therefore dominantly wave-induced. Despite active cliff erosion, there is a shortage of mobile littoral sediment on most shorelines so that actual rates of sediment moved by longshore drift are well below their theoretical potential (HR Wallingford, 1998). The dominant net direction of littoral transport is west to east, consistent with the direction of approach of dominant wave fronts. However, there is some evidence of a drift divide in the vicinity of the headland of White Nothe. Beaches are of variable dimensions and composition, mostly conditioned by locally available inputs. Numerous "pocket" bays, coves and other re-entrants, separated by headlands, interrupt transport and cause coarser sediment to be trapped within discrete compartments. This, together with the presence of deep water in front of some cliffed sections, inhibits long distance longshore transport. Many pathways are very short as a result of changes in coastline orientation. There is substantial onshore to offshore removal of fine sediment in suspension, resulting from dispersal of much clayey landslide debris and both inter-tidal and sub-tidal shoreface erosion (Posford Duvivier, 1999). Quantities are a function of the lithology of substrate materials, wave and tidal energy, shoreface width, water depth and sea bed relief.

2. SEDIMENT INPUTS - F1 References Map

2.1 Marine Inputs

F1 Ringstead Bay

A reservoir of may exist upon the bed of Ringstead Bay, which can provide an intermittent or pulsed cross-shore input to the local beach. It is uncertain, however, whether it constitutes a net gain, or an exchange of materials with the beach. The beach certainly experiences episodic losses of gravel during storms so that any onshore feed could represent a return of existing beach material. It is thought that material is retained landward of a series of confining reefs located some 100m 150m offshore of the beaches. However, gaps in the confining reefs may permit some permanent losses involving net offshore transport of coarse sediment, moved by tidal and/or wave-induced rip currents. HR Wallingford (1998) note that shore parallel tidal stream velocities can be considerably greater in Ringstead Bay than is usual for the Purbeck and Weymouth Bay shorelines.

2.2 Fluvial Input - FL1 FL2 References Map

The only sediment inputs of any significance are: (1) Osmington stream in the western part of this sector, and (ii) the stream draining into Chapman's Pool, West of St Aldhelm's Head. Other streams, such as Winspit, drain very small catchments and contribute negligible quantities of material.

FL1 Osmington River (see introduction to fluvial inputs)

Rendel Geotechnics and the University of Portsmouth (1996) estimate an annual delivery at Osmington of 63 tonnes of bedload and 192 tonnes of suspended load. During an extreme rainfall event on 18.07.55 (most intense rainfall ever recorded in England at nearby Martinstown) an exceptional stream discharge was produced, causing deep scouring of the bed and supply of "many tons of boulders and other debris into the sea" such that a temporary fan or delta was formed across the foreshore (Arkell, 1955).

FL2 Chapman's Pool (see introduction to fluvial inputs)

A stream draining into this embayment contributes small quantities of sand and clay. During extreme flow events it can flood its narrow valley floor and supply small quantities of gravel to the shore.

2.3 Cliff, Shore Platform and Shoreface Erosion - E1 E2 E3 E4 E5a E5b E6 E7 E8 E9 E10 E11 E12 References Map

E1 Redcliff Point to Bran Point (see introduction to coastal erosion)

The stratigraphical sequence of clays, sandstones and oolitic limestones is affected by local folding and faulting. Between Redcliff Point and Shortlake there is an upfaulted block, whilst other more minor faults create local aquicludes. Active cliff instability characterises the cliffline between Black Head and Bran Point, with several mudslides, mudflows and translational failures (Bird, 1996). At Black Head, a complex of mudslides which developed on the Kimmeridge Clay between 1910 and 1914 were triggered by excess groundwater supply from infaulted permeable Greensand and Chalk behind the cliff face (Arkell, 1947). This continues to be an active area of slope instability, possibly subject to some recent intensification due to cliff toe erosion (Mouchel, 1998). Overall, between Black Head Osmington Mills, Hannah's Ledge and Bran Point, some 1,400m of this frontage is occupied by mudflows and slides; indeed the Black Head is reputed to have produced the biggest single mudflow event recorded in the UK during the twentieth century (Bird, 1996). Instability at this point extends at least 600m inland. Peter Brett Associates (1995) note that mudsliding at Osmington Mills may be episodic, with periods of 10 to 15 years of alternating quiesence and renewed activity. Spalling, toppling and gulleying are also active processes, though the latter is more dominant on the Oxford Clay cliffs between Redcliff Point and Black Head. Photo 8 illustrates a medium size cliff failure within Kimmeridge Clay on the Eastern flank of Black Head that has generated a mudslide that surged across the foreshore.

Mass movement accounts for the detachment of oolite blocks from the upper cliffs, particularly in the vicinity of Osmington. These accumulate at the toe of the marine cliff and provide some protection against wave erosion (Peter Brett Associates, 1995). Corallian limestone ledges, which outcrop on the lower foreshore and offshore, also modify incident wave energy. There are several estimates of rates of cliff recession, ranging from 0.05 to 0.40ma^-1 at Black Head (May, 1966) to 0.11 to 2.2ma^-1 at Osmington Mills (Peter Brett Associates, 1995). Mouchel (1998) and Posford Duvivier (1997; 1999) calculate a mean rate of cliff retreat of 0.5ma^-1 at Osmington for 1970-96 with a mean rate of 0.46a^-1 for this shoreline as a whole. Most researchers agree that erosional loss from cliff retreat has increased since the 1930s and 1940s along the eastern part of this sector. Using both OS maps and aerial photography, Peter Brett Associates (1995) carried out a detailed study of cliff erosion for the headland in front of the settlement at Osmington Mills. This revealed a mean rate of recession of 1.1ma^-1, 1946-71 and 0.8ma^-1, 1983-95. Cliff top erosion was marginally faster than cliff toe retreat. Although this may appear to contradict the observation of accelerating erosion, these figures conceal wide temporal and spatial variability for the various cliff profiles that were surveyed.

Posford Duvivier (1997, 1999) have estimated total sediment yield from cliff erosion of 40,000m³a^-1, and a rate of inter-tidal (shoreface) erosion yield of some 18,000m³a^-1 of fine sediment at an assumed rate of vertical abrasion of up to 9mm. Nearshore/offshore platforms eroded into clay are presumed to be the principal areas of supply. Most, of yield from both sources is clay that is transported offshore in suspension, although some limestones are derived from the Corallian together with some flint gravels from a mostly thin mantling of superficial deposits.

E2 Bran Point to White Nothe (Ringstead Bay) (see introduction to coastal erosion)

Ringstead Bay (Photo 6) is cut into a series of clay dominated lithological units, although rates of cliff recession are relatively modest. May (1966) calculates 0.13 to 0.64ma^-1 (with 0.4ma^-1 as a mean) for 1880-1963, whilst Mouchel (1998) concludes that rates have been between 0.34 and 0.6ma^-1 since 1970. These rates are likely to reflect the role of a series of offshore ledges composed of resistant Corallian limestones outcropping along the foreshore and nearshore bed parallel to the shoreline, that dissipate a part of the incoming wave energy (Cole, 1995; Bird, 1996; HR Wallingford, 1998). A modest beach nourishment comprising 25,000 cubic metres of marine dredged fill, undertaken in 1996, may also have affected recent rates of recession, which appeared to have accelerated since the late 1940s in comparison to the preceding 60 years (May, 1966). To the east of Ringstead Bay is the large deep-seated multiple rotational landslip, which extends eastwards some 2km around White Nothe headland. The 160-170m high cliffs have a characteristic form induced by the Chalk and Greensand overstep of underlying Gault and Kimmeridge Clay (Bird, 1997). This is probably a long-established failure complex, which experiences infrequent but high magnitude falls and slides. There are no measurements of changes in the position of the cliff toe, although toe erosion and release of landslide debris is clearly an ongoing process (Photo 2). Posford Duvivier (1999) calculate a sediment yield of 15,000m³a^-1, but the basis for this figure is uncertain. Some 1,200m³ is flint and chert from the Upper Greensand and Chalk, that contributes to the gravel beach in Ringstead Bay; the rest is fine grained and moves offshore.

E3 White Nothe to Bat's Head (see introduction to coastal erosion)

The most westerly sector is a continuation of the White Nothe landslip complex. Further east, vertical Chalk cliffs dominate. The "needle eye" arch of Bat's Head is evidence of long-term recession. This coastline is more exposed to high energy waves, propagated across an extensive fetch, than the cliffline westwards. May (1966) gives an average retreat rate of 0.22ma^-1 (1890-1960), but for Middle Bottom, west of Bat's Head, May and Heaps (1985) calculate a lower rate of 0.05ma^-1 (1882-1975). This is probably due to the presence of infrequent rockfalls, which offer temporary toe protection prior to their removal by abrasion and solution processes. Posford Duvivier (1997; 1999) calculate that shoreface erosion is at a rate of 4mma^-1 and that total sediment supply from cliff erosion, east to Dungy Head, is 25,000m³a^-1.

E4 Bat's Head to West Face of Durdle Door (see introduction to coastal erosion)

The vertical Chalk cliffs of this sector reach a maximum elevation of 120m at Swyre Head, and truncate three "hanging" dry valleys. As with E3, localised and short-lived rockfalls occur as a result of block detachment from joint and other parting planes. A recession rate of between 0.2 and 0.46ma^-1 (1880-1975) is calculated by May (1960) and May and Heeps (1985) for the Chalk cliffs, but is likely to be faster along the slumped and gulleyed outcrop of the Wealden clays and sandstones forming the Durdle Door isthmus (Photo 9). The 'Door' itself (a world famous arch) is evidence of high wave energy, but the cliffline to the west is screened by a submarine rock ridge of Portland Limestone, parts of which are exposed at low water. Immediately west of the Durdle promontory is a set of caves developed in the crush breccia of a major thrust structure. It is not clear if these have a marine origin, as they are now inaccessible to wave action; they may be the product of groundwater seepage and/or solution weathering. Cliff top recession is faster than at the cliff base across the truncated "hanging" valleys, due to the exposure of Coombe Rock infill (Bird, 1996). This is actively denuded by weathering and gulleying. Sub-aerial slumping and gulleying of the Wealden slopes is almost continuous, particularly during the winter. Surprisingly, no attempt at measuring the rate of denudation of this fragile landform facet has been undertaken.

E5a East Face of Durdle Door to Lulworth Cove (Western Entrance) (see introduction to coastal erosion)

Cliff height is variable, reflecting the differential structure and strength of the rock sequence along this sector. The Chalk cliffs in the centre of St Oswald's Bay are 150mOD, whilst the Portland Limestone cliffs that form Dungy Head and the rock rampart protecting the Durdle Door promontory are some 65-80m in height. Cliffs on the intervening outcrop of Wealden sands and clays are as low as 8m. The coastal form reflects the removal of most of the very tightly folded and compressed rock sequence in front of the Chalk, between Durdle Door and Dungy Head (Nowell, 1997). The "chance" survival of the Durdle Door promontory is probably a function of topography, as it marks the position of a former sub-catchment divide (Bird, 1996; Brunsden and Goudie, 1997) Man 'O' War Rocks (Photo 10) are a detached reef of Portland Limestone that absorbs wave energy and has created a subsidiary headland behind. Large boulders and coarse debris stores have accumulated around Dungy Head; offshore rocks, such as Pinion Rock, may be remnants from previous large-scale cliff falls or topples. There have been several falls affecting the Chalk cliffs in recent decades, with cliff top retreat associated with active removal of periglacially weathered debris and the presence of two very deep solution shafts. Mouchel (1998) estimate a retreat rate of 0.07ma^-1 for the cliffs east of the Man'O' War Rocks (1880-1990), although May (1966) gives a figure of 0.2-0.25ma^-1 (1890-1960). Basal notching is evident at various locations (Canning and Maxted, 1979), and cliff fall debris has a short residence time (May and Heeps, 1985). Comparatively very high rates of slope denudation are taking place on the Wealden outcrop. Gully erosion and slumping are active processes on the east-facing side of Durdle Door, whilst immediately north of Dungy Head, shallow translational slides and flows are more characteristic (Jones, et al., 1984; Allison, 1986, 1989). This contrast is likely to be related to differences in exposure to precipitation, temperature changes and wave energy (Canning and Maxted, 1979; Allison, 1986). Posford Duvivier (1999) calculate a shoreface erosion loss of fine grained material of approximately 1,100m³a^-1.

E5b Lulworth Cove (including Stair Hole) (see introduction to coastal erosion)

The rock sequence around the perimeter of the quasi-eliptical planform of Lulworth Cove (Photo 3) has been described numerous times (e.g. Davies, 1956; Arkell, 1947; House, 1958; King, 1976; Bird, 1996; Perkins, 1977, Nowell, 1997). The Chalk forms the steep backwall, whilst the narrow entrance is flanked by overhanging cliffs of Portland Limestone. Much of the lateral expansion of the cove has been confined to the intervening outcrops of Purbeckian Limestone and Wealden clays, marls and sandstones. Both are comparatively erodible, though for different physical reasons (Arkell, 1947; Allison, 1986, 1989, 1990). The geomorphological origin of Lulworth Cove has long attracted attention (e.g. Burton, 1937; Mottram, 1972; Jones, et al., 1984; Horsfall, 1993; Bird, 1996; Brunsden and Goudie, 1997). The classic evolutionary model, that places Lulworth Cove in a sequence of stages starting with Stair Hole and concluding with Durdle or Warbarrow Bay, has been criticised as both simplistic and idealised. The role of fluvial erosion, creating a valley form that was subsequently partly inundated by Holocene sea-level rise is more consistent with local detail (Brunsden and Goudie, 1997). Cliff height and morphological development is influenced by the strong contrasts in rock lithology and structure. Cliff falls and topples are characteristic of both the Chalk and Portlandian Limestone, whilst differential weathering creates the ribbed and recessed form of the Purbeckian Limestone outcrop. As in sector 5a, mass movement on the Wealden beds creates rapid changes in the profile of the coastal slope developed on this lithologically weak unit (Allison, 1986, 1990; Bird, 1997). There is obvious visual evidence of more dynamic cliff change on the exposed west-facing slopes closer to the Cove entrance. The initiation of new mudslides is partially documented in Allison (1986), but there are no reliable quantitative measurements of cliff recession; Allison, (1986) does however record the loss of some 320m² at one site, 1880-1980. Partly submerged small boulder arcs on the east side of Lulworth Cove may record earlier mudslide surges. It is ironic that there should be a dearth of reliable data on cliff behaviour at what is one of the world's most renowned geological/geomorphological coastal sites!

Stair Hole, to the immediate west of Lulworth Cove, is the result of wave energy breaching the Portland Limestone barrier and causing the lateral expansion of a small inlet along the comparatively incompetent brittle and fractured Purbeckian rocks. The north wall of this feature is developed in the Wealden beds, creating a zone of instability extending some 250m inland (Bird, 1996). Brunsden and Goudie (1997) record the pattern of slide tracks and lateral shears that connect a well defined source area to the zone of discharge at the slope base. Here, lobes of clay and silty sand frequently move across and partly conceal the basal boulders of Purbeck Limestone. Landslip activity here is the product of a complex relationship between the mechanisms controlling rates of supply and removal of debris (Jones, et al., 1984; Allison, 1986). Canning and Maxted (1979) consider that the top of the mudslide zone retreated 30-40m between approximately 1880 and 1970. Most researchers (e.g. Allison, 1986, 1990 and Brunsden and Goudie, 1997) note that instability has increased substantially since the mid-1950s. Removal of vegetation, together with trampling due to recreational pressure, may be an important cause (Lulworth Cove receives over 11/2 million visitors annually). It is possible that the rate of backwearing of the adjacent Purbeck Limestone cliffs has declined in the last 30-40 years (Canning and Maxted, 1979; Jones, et al., 1984), thus the supply of coarse debris that protects the base of the Wealden beds opposite the three points where the sea has penetrated the Portland Limestone barrier has diminished. Brunsden (1996) notes that there are several uncertainties regarding mechanisms of mudsliding and soft cliff slope failure on the Dorset coast, citing Stair Hole as an example.

E6 Lulworth Cove (East Entrance) to Mupe Rocks (see introduction to coastal erosion)

The dip of the Portlandian and Purbeckian rocks declines rapidly along this frontage, with the former present only as the complex of Mupe Rocks offshore stacks (Photo 11and Photo 12) at its eastern extremity. Cliff height averages 30m, and cliff form is near vertical due to undercutting of the well-jointed Lower Purbeck Series at their base. Although Allison (1986) identifies two sites of instability, the historical rate of recession is low. May (1966) and Mouchel (1998) calculate minimal erosion for the period 1880-1990, between zero and 0.01ma^-1. At Cockpit Head, adjacent to Mupe Ledges, May (1966) gives a rate of cliff base retreat of 0.23ma^-1. Posford Duvivier (1999) estimate the surprisingly high figure of 1,100m³a^-1 of shoreface erosion.

E7 Mupe Bay to Warbarrow Tout (Warbarrow Bay) (see introduction to coastal erosion)

Cliff height and morphology is precisely controlled by the same shore parallel sequence of strata encountered further west. Maximum height is 170m at Ring's Hill in the eastern sector of the bay, where a wide degradation zone characterises the Chalk cliffs (Nowell, 2000). The cliff line is interrupted at Arish Mell, where a fault-guided valley has almost penetrated the Chalk down to sea-level. Historical recession rates of the main backwall, 1880-1962 were between 0.11 to 0.16ma^-1 (May 1966). Posford Duvivier (1997) calculate an average recession rate of 0.135ma^-1. Cliff top retreat at Ring's Hill has removed slightly more than one half of a late Neolithic hill fort (Flower's Barrow). On the arguable assumption that it was originally built close to the cliff edge, this would give a long-term recession rate of 0.08ma^-1 over some three millennia. Cliff height diminishes to less than 25m on the Wealden Beds, the outcrop area of which is widest on the east side of Warbarrow Bay (Photo 13). Translational slides, flows, semi-rotational slips gullies and linear erosion by a small stream are particularly active over the 350m length of frontage immediately north of Warbarrow Tout (Moore and Brunsden, 1996, Nowell, 2000). Allison (1986, 1990) and Allison and Brunsden (1990) report short-term cliff base recession rates of between 2.5 and 3.0ma^-1 at this location. These high rates are the result of individual slope failures, which rapidly create, redistribute and remove basal debris stores. Mass movement on the Wealden outcrop of Mupe Bay is also active, but this sector is less exposed to wave activity. An average recession rate here is approximately 0.12ma^-1 for the last 100 years (Mouchel, 1998), compared to twice or three times that on the opposite eastern side of the bay. Monitoring of processes within this unit has proved difficult because of MOD-imposed access restrictions, thus the above figures are not considered fully reliable. Posford Duvivier (1999) calculate that 1,400m³a^-1 of fine material is removed from the shoreface outcrop of Wealden beds. Several major rockfalls have occurred from the high Chalk cliffs immediately to the east of Mupe Bay creating major cliff toe debris fans that are rapidly removed by marine erosion. Occasional falls from the Chalk cliffs are rapidly reduced and removed. Posford Duvivier (1997; 1999) calculate a total sediment yield for the bay of 20,000m³a^-1., mostly comprising fine material from the Chalk and Wealden Beds. Small quantities of flints are yielded from the Chalk together with hard pebbles of a variety of lithologies from beds within the Waelden strata. These materials contribute to the narrow gravel beaches in the eastern and western extremities of Warbarrow Bay.

The Chalk cliffs descend to sea-level at Arish Mell, where they truncate a large dry valley. An embayment is developing here, partly due to the exposure of weakly consolidated alluvial and colluvial infill.

E8 Warbarrow Tout to Hobarrow Bay (see introduction to coastal erosion)

This coastline is characterised by high (140m) Portland Stone and Portland Sand cliffs of Gad Cliff (Photo 14) and the 'pocket' bays of Pondfield Cove and Brandy and Hobarrow Bays. Allison (1986, 1990) and Allison and Kimber (1999) note the presence of areas of instability where the basal Kimmeridge Clay has been exposed. At Gadd cliff, a landward dipping caprock of Portland Stone forms a steep upper cliff that fails by rockfall creating a debris accumulation slope below. As the toe of the slope is eroded by wave action a protective apron of hard Portland Stone boulders is left remaining at the shoreline and this together with the landward dip of outcropping strata helps to explain the generally slow recession og this cliff sector. May (1966) Posford Duvivier (1997) and Mouchel (1998) calculate a spatially variable rate of cliff recession of 0.01 to 0.15ma^-1 for the period 1880-1990 with fault structures determining specific areas of weakness (Nowell, 2000). Rockslides and falls are active on the seaward side of Warbarrow Tout headland, for which Allison (1986) calculates recent retreat of 0.11 to 0.16ma^-1. Posford Duvivier (1999) determine an annual yield of fine sediment of 15,000 m³ from the cliff and 4,500m³, from shore platform and shoreface areaassuming vertical corrasion at 4mma^-1.

E9 Broad Bench to Hen Cliff (Kimmeridge Bay) (see introduction to coastal erosion)

The two main characteristics of this short sector are low, 5-20m, cliffs and a wide inter-tidal shore platform developed upon thin limestone or calcareous mudstone seams within the Kimmeridge Clay (Photo 7). Shales and clays are interbedded with a series of limestone bands ("Cementstones") that are responsible for the widest and most resistant elements of the inter-tidal ledges and platforms (Arkell, 1957; Davies, 1956; Bird, 1996; Mouchel, 1998). Rates of cliffline recession are high compared with much of the westward coastline. For Kimmeridge Bay, which is the result of long-term differential erosion, May (1966) gives a rate of 0.38 to 0.45ma^-1, approaching 0.8ma^-1 where platform relief and width is least (Mouchel, 1998). Erosion is less between Clavel Tower and Hen Cliff (0.13 to 0.20ma^-1, May, 1966), although exposure to prevailing south-westerly waves is actually slightly greater. The reason for this is uncertain, but is probably related to a change in the lithology of the Kimmeridge Clay. Although not reported in the literature, the fastest rate of cliff recession would appear to be in the vicinity of Gaulter Gap, where a stream is sharply incised. A second World War defence structure at this point is now some 12m in front of the cliffline, whilst cliff top denudation now threatens the oil well, located here in 1958. Very rapid physical weathering occurs where flaking of paper shales creates redistributed talus, e.g. near the boathouse site. Kimmeridge Bay is the most important source of sediment input east of White Nothe; shoreface abrasion of clay and shale outcrops, including platform areas, may remove between 1,200-7,500m³a^-1, according to Posford Duvivier (1999). This figure may seem a low estimate, but sediment may be trapped, and thus re-circulate, between nearshore ledges.

E10 Hen Cliff to Egmont Point (see introduction to coastal erosion)

The Kimmeridge Ledges between Clavell's Hard and Egmont Bight, which extend seawards at an oblique angle to the general coastline trend, promote wave energy dissipation and complex local refraction. Thus the 35-50m high cliffs of this remote sector are comparatively stable; a retreat rate of 0.09ma^-1 (1870-1985) is proposed by Posford Duvivier (1997). However, the Ledges form part of a wide shoreface abrasion platform that are evidence of long-term recession. Posford Duvivier (1999) estimate a yield of fine grained sediment from the shoreface of approximately 7,500m³a^-1. In most respects, this sector is poorly documented.

E11 Houns-tout Cliff - Chapman's Pool - St Aldhelm's Head (West) (see introduction to coastal erosion)

Egmont Point forms an integral part of the large scale unstable coastal slope complex of Houns-tout (Photo 5). The latter has a well defined upper free face of Portland Limestone, but the Kimmeridge Clay at the base has promoted a sequence of slides, slumps and falls. The cliff morphology west of Chapman's Pool has been substantially modified since the early 1970s by several large falls (Bird, 1996). Boulder accumulations provide some basal armouring, but fines are rapidly removed offshore. The recess of Chapman's Pool (Photo 15) is due in part to a local increase in the height of the Kimmeridge Clay outcrop. This is due to an upfold that accounts for the abrupt change in coastline orientation to the east, thus creating the prominent St Aldhelm's (sometimes St Alban's) Head. Because of the frequency of rockfalls and slides delivering material to the toe, the shoreline beneath Houns-tout experiences some temporary advances of mean high water followed by longer term recession, thus map analysis indicates only a slow overall retreat. May (1966) suggests a maximum of 0.10ma^-1, 1880-1960, whilst Mouchel (1998) calculate almost no net recession, 1890-1990. In reality, actual erosion rates involved in removing basal debris supplied by cliff falls are probably at least ten times higher than this, but material eroded is replaced by renewed debris slips. Chapman's Pool has retreated at a marginally higher rate, approaching 0.20ma^-1 (May, 1966). A small contribution comes from fluvial erosion via the deeply incised West Hill stream, but most of this loss is the result of wave attack of the cliff base. Cliff elevation between Chapman's Pool and St Aldhelm's Head (Photo 4) is over 120m (Emmett's Hill). This west-facing slope complex is the product of deep-seated semi-rotational failure (Allison, 1989). The rock sequence of impermeable Kimmeridge Clay at the base and relatively permeable and porous Portland Sand and Portland Limestone above provides appropriate hydrogeological conditions for slope instability. In addition there is direct exposure to the highest wave energy received along any part of this coastline (May, 1971). A series of major failures, possibly originating in the early or mid-Holocene (Allison, 1986; Allison and Kimber, 1999) have contributed to the substantial mid-slope and basal store of boulders. These originate from the uppermost free face (backscar) of Portland Limestone, but the failure surface may start in the underlying Portland Sands and translate upslope. It is not, however, transmitted downslope (Jones, 1980; Jones, et al., 1984). This is apparent from visual monitoring of the development of instability since 1978. Posford Duvivier (1999) estimate that the shoreface, which has an average depth of 15m, yields some 1,300m³a^-1 of fine grained material due to wave abrasion. Despite a width of between 100 and 250m, this platform-like surface is only occasionally fully exposed, hence this relatively small quantity of sediment loss.

E12 St Aldhelm's Head (East) to Durlston Head (see introduction to coastal erosion)

High vertical cliffs (30-38m) of Portland Limestone that plunge directly into deep water front virtually the entire length of this coastline (Photo 16). Local diversity is provided by the mouths of the Winspit and Seacombe Valleys, the latter usually dry, and small platforms and benches, e.g. Dancing Ledge. Cliff morphology is largely controlled by sets of rock joints orientated at high angles to the near-horizontal bedding planes. Basal undercutting and cave formation are recurrent features due to solutional weathering and bioerosion as well as wave corrasion (Bird, 1996). However, coastal quarrying and smuggling, between the mid-eighteenth and early twentieth centuries, account for substantial enlargement of several ledges, platforms and caves (e.g. Blacker's Hole, Dancing Ledge, West Man Quarry and Tilly Whim Caves, close to Anvil Point). Rates of natural cliff line recession are virtually zero for the period of reliable map evidence (May, 1966; Mouchel, 1998). Where former quarrying near cliff top sites has weakened rock coherence rates might be higher. However, this may be counter-balanced by the dumping of quarrying waste, artificially creating protective talus slopes in a few locations, e.g. Seacombe Cliff. As the shoreface is not less than 20m deep, and is at or below average wave base, erosional losses are likely to be small; thus the figure of 12,000m³a^-1 given by Posford Duvivier (1999) is difficult to accept. However, solutional loss is active, though not quantified.

2.4 Beach Replenishment - N1 References Map

N1 Ringstead

A beach nourishment comprising 25,000 cubic metres of marine dredged fill retained by a terminal rock groyne was undertaken in 1996 (Photo 6). The scheme was developed following beach losses and accelerating cliff recession over the 1980s and early 1990s that was beginning to threaten properties (S W and H Pattison Ltd. 1995; HR Wallingford, 1994). Progression of the scheme raised a number of important earth science conservation issues as explained in Section 6.

3. LITTORAL TRANSPORT - LT1 LT2 LT3 LT4 LT5 LT6 LT7 LT8 References Map

HR Wallingford (1998) identifies 11 transport sub-cells, but with only conjectured linkage between several of them. Based on field observations (over some 25 years), the following account recognises 8 discrete sub-cells. However, it can only be a provisional sub-division, due to absence of reliable and quantified data.

LT1 White Nothe to Redcliff Point (see introduction to Littoral Transport)

Net drift is east to west, with an apparent transport divergence in the vicinity of White Nothe (HR Wallingford, 1998). Weak easterly drift has been observed in Ringstead Bay when waves are approaching from the south-south-west (Mouchel, 1998), thus there may be periodic reversals of movement. Transport is mostly of gravel and coarse sand, with gravel more dominant in Ringstead Bay. This is evident from the composition of beaches, though recharge at both Osmington Mills (East) and Ringstead in the late 1990s will have modified particle size-range distribution. Rates and volumes of transport may be substantial over short periods, particularly after beaches have experienced drawdown due to exceptionally high energy waves. This was observed at Ringstead Bay in 1989-90 (Pattison S. W. and H., 1995; Mouchel, 1998) and may be due to the trapping of sand and gravel between the foreshore and several sub-parallel nearshore rock ridges and reefs (Photo 6). These are composed of relatively resistant Corallian strata and are only exposed at maximum low water (Jolliffe, 1976). This is the probable explanation of the lack of recovery of the reflective beach form of eastern Ringstead Bay following severe winter storms in 1989-90, and thus the need for their subsequent renourishment. Under average wave climate conditions, this natural offshore breakwater dissipates incident waves, thus suppressing potential drift rates and volumes. Whilst there is evidence of some bypassing of promontories at Bran Point, Osmington and Black Head, it is uncertain if sediment is moved past Redcliff Point where drift is intercepted by prominent boulder aprons. Coastal erosion and mudflows provide supplemental sources of mostly sand sized sediment, but most of the gravel supply must derive from the degradation and recession of the White Nothe landslides. There may be a reservoir of gravel upon the nearshore bed in Ringstead Bay, which provides cross-shore pulsed supply to the local beach as well as the littoral transport pathway. However, gaps in the confining reefs may promote some net offshore movement of coarse sediment, moved by tidal and/or wave-induced rip currents. HR Wallingford (1998) note that shore parallel tidal stream velocities can be considerably greater in Ringstead Bay than is usual for the Purbeck and Weymouth Bay shorelines.

Arkell (1947) has noted the absence of Corallian Limestone clasts east of central Ringstead Bay, which he takes as evidence of net east to west longshore movement. He also reports that the coarsest sediment particles accumulate on the eastern sides of headlands, with size grading occurring immediately downdrift of these partial barriers to movement. Mouchel (1998) infers that this pattern is most apparent during periods of beach accumulation under constructive wave action. Waves approaching from the south-east may be the most effective, implying (from their comparative infrequency) that significant drift movement only takes place when forcing conditions are favourable.

Posford Duvivier (1999) calculate the following rates of annual sediment yield from shoreface erosion: Kimmeridge Clay, 3,000-8,000m³; Oxford Clay, 25,000m³; Corallian Beds, 10,000m³.

LT2 White Nothe to Durdle Door (see introduction to Littoral Transport)

Immediately east of the inferred transport divide at White Nothe, net longshore drift along this section is eastwards. However, high potential rates and volumes of movement are not achieved because of limited supply of sediment. A wide, steep swash-aligned gravel beach has accumulated against the western side of the Durdle Door promontory (Photo 9), which functions as a local drift boundary. The well-worn form of most of the clasts indicates constant attrition and abrasion. Although fully exposed to south-west waves, an offshore reef of Portland Limestone absorbs some of their energy. Small parts of this latter feature are exposed during both the rising and falling tides. The well-sorted nature of this coarse clastic beach infers selective net offshore loss of finer grade material. Posford Duvivier (1999) calculate that the erosion of Chalk cliffs along this sector yields 200-400m³ of coarse debris, and some 25,000m³ of fines, annually. Much of this material derives from rock falls, which have a short residence time; an example, in 1983, is described by May and Heeps, 1985. Cross-shore transport is indicated by rapid changes in beach profile form, particularly the construction and destruction of multiple sets of berms.

LT3 Durdle Door to Lulworth Cove (including within the cove) (see introduction to Littoral Transport)

A mixed gravel and coarse sandy beach extends from the eastern (lee) side of the Durdle promontory to Dungy Head. From here, to Lulworth Cove entrance, there is a discontinuous accumulation of boulders; these derive from occasional rock falls from the Portland Limestone cliffs. The presence of the inshore reef of Man'O' War Rocks (Photo 10) immediately east of Durdle Door dissipates much potential wave energy and creates a wide cuspate beach modelled by waves created by the very local fetch. Sets of beach cusps are a characteristic feature here (Canning and Maxsted, 1979). There are no calculations of potential shoreface erosional losses, whilst HR Wallingford (1998) estimates a minimal rate of eastwards littoral drift of approximately 220m³a^-1. This is partly due to very limited supply of sediment stable on the beaches of this sector; and partly to a complex wave climate that is affected by the presence of the offshore Corallian Limestone outcrop of the Lulworth Banks (May, 1990). Occasional rock falls are detached from the high, solutionally weathered Chalk cliffs, but are rapidly broken down by abrasion and solution (May and Heeps, 1985). Fine material is presumed to move offshore. A coarse boulder beach is trapped within Stair Hole, mostly the product of sub-aerial breakdown of Purbeckian Limestones.

LT3a Lulworth Cove (see introduction to Littoral Transport)

There is a well developed gravel beach around the perimeter of the cove (Photo 3), which widens eastwards. It has been described by numerous authors (e.g. Bird, 1996; Canning and Maxted, 1979), but there is no quantitative data on which to base a morphodynamic assessment. Refracted waves pass through the cove entrance, creating swash-aligned beach berms; numerous student projects have measured clast size and shape distribution, suggesting that there may be a weak net easterly drift. However, drift reversal appears to take place in the eastern part of the cove, where a storm beach profile sometimes develops in late winter. The well rounded nature of most Chalk and limestone clasts indicates abrasion of material confined to the cove. As there is no evidence of either output or input of coarse material, Lulworth Cove must therefore be regarded as a sediment sink. Fine material, mostly derived from sub-aerial denudation of the Wealden clays and sands, is probably removed offshore in suspension.

LT4 Lulworth Cove (Eastern Entrance) to Mupe Rocks (see introduction to Littoral Transport)

With very little supply of material from the steep, plunging cliffs of this sector, littoral transport is virtually zero. The shoreface is also very restricted in width, yielding less than 1,000m³a^-1 of fines (Posford Duvivier, 1999). This is presumed to move offshore.

LT5 Warbarrow Bay (Mupe Rocks to Warbarrow Tout) (see introduction to Littoral Transport)

Arkell (1947) inferred, from patterns of sorting of beach particle size and composition, a complex pattern of littoral transport, with larger particles moving towards the centre of Warbarrow Bay. Smaller clast sizes were, he considered, selectively moved both east and west. However, both HR Wallingford (1998) and Mouchel (1998) observe that beaches are swash aligned and adjusted to wave refraction. Coarse sand and gravel beaches have accumulated in Mupe Bay and north of the headland of Warbarrow Tout reflecting local supply from the degradation of the Wealden Clay cliffs immediately behind. The central Chalk cliff section of Warbarrow Bay is backed by Chalk debris rather than beaches at the toes of high cliffs so that any actual drift is likely to be low. A progressive increase in beach width eastwards from Arish Mell is taken to infer net littoral transport in the same direction, as coarse sediment is apparently unable to by-pass Warbarrow Tout. The potential for longshore movement is limited, and actual rates are low. Warbarrow Bay is therefore a self-contained transport sub-cell, although some onshore to offshore movement of fine grained sediment is likely. This is estimated at 1,500m³a^-1 (Posford Duvivier, 1999). Approximately 100m³a^-1 of flint shingle derives from falls and topples from the Chalk cliffs east and west of the truncated valley of Arish Mell (May and Heeps, 1985; Posford Duvivier, 1999) and additional hard gravels are supplied by erosion of a series of thin pebble beds from within the Wealden formation.

LT6 Warbarrow Tout to Egmont Point (see introduction to Littoral Transport)

A net eastwards drift characterises this sector (HR Wallingford, 1998), but is subdivided into a consecutive series of localised confined transport systems by the sequence of bays and headlands. Only fine material by-passes the latter, thus Brandy, Hobarrow and Kimmeridge Bays effectively function as virtually self-contained units, with small gravel "pocket" beaches. The precise compositions of the latter closely reflect the availability of local material from cliff erosion. HR Wallingford (1998) calculate gross annual drift in the nearshore zone to be in the order of 1,500m³. Tidal velocities are low (< 0.5msec-1), thus all transport is generated by breaking waves. Wave energy, however, is dissipated by the presence of wide shore platforms that are particularly well developed in Hobarrow and Kimmeridge Bays, off Broad Bench (Photo 7) and the Kimmeridge Ledges between Hen Cliff and Egmont Bight. Where the alternating clay, shale and cementstone lithology creates obliquely-trending ridges potential drift rates may be as low as 200m³a^-1. A regionally high rate of shoreface erosion, in the order of 13,200-14,000m³a^-1 (Posford Duvivier, 1999) is due to the predominance of the clay and shale outcrop areas. Much of this material is removed offshore, in suspension, and is added to by locally high rates of cliff erosion where platforms are absent or poorly-developed (e.g. central Kimmeridge Bay). There is a slight increase in the supply of coarse material east of Rope Lake Head, and much of this appears to move eastwards into the confined pocket beach of Egmont Bight where transport is intercepted by the debris lobe of Egmont Point beneath Houns-tout cliff (Photo 5). It is uncertain if Egmont Point is a transport barrier to fine sand, which may therefore feed the next sector downdrift (Chapman's Pool). North of Clavel Tower, in eastern Kimmeridge Bay, there appears to have been some small-scale, private, beach renourishment in the past; previous shale workings have also contributed a legacy of basal debris.

LT7 Egmont Point to St Aldhelm's Head (West) (see introduction to Littoral Transport)

Weak north-easterly longshore drift into the embayment of Chapman's Pool is implied from the permanence of the pocket beach. However, this beach exhibits periodic fluctuations in both width and sediment composition, sometimes possessing a gravel-dominated backshore as well as a more persistent sandy foreshore. This may imply a direct exchange with the nearshore, as dominant high-energy swell waves approach nearly parallel to this coastline and can promote beach drawdown. A well-defined clockwise tidal eddy, set up by the blocking effect of St Aldhelm's Head, exists immediately to the west and offshore. This may create tidal currents capable of moving sand; HR Wallingford (1998) suggest the presence of a small-scale sink, as there is no evidence for transport around this large headland, and submerged offshore ledges are apparently concealed by a considerable accumulation of sand and gravel. Around Egmont Point beneath Houns-tout cliff (entrance to Chapman's Pool) a boulder beach, with several sand traps, represents debris from previous cliff failures (Photo 5). Aprons of Portland Stone boulders are even more pronounced on the west-facing coast of St Aldhelm's Head, where there is little sand and gravel accumulation. The sediment budget of this unit is difficult to quantify, with a possibly locally significant input to both the beach at Chapman's Pool and the offshore sink from the deeply-incised stream that discharges below West Hill. Most of the slopes of this valley are unstable, and may contribute a periodic sediment input. Posford Duvivier (1999) estimate shoreface erosion and offshore removal of fines to be approximately 13,000m³a^-1.

LT8 St Aldhelm's Head (East) to Durlston Head (see introduction to Littoral Transport)

The cliffs along this frontage plunge vertically into deep water, with no beach or foreshore development. Thus, although potential longshore transport is north-east to Seacombe, and east between Seacombe and Anvil Point, lack of available sediment reduces actual transfer to virtually zero. The doninant westerly directed tidal stream off St Aldhelm's Head does not appear to be a significant factor in either near-or offshore sediment movement, although it has not received detailed

4. SEDIMENT OUTPUTS - O1 O2 References Map

4.1 Nearshore and Offshore transport

O1 Westward and South-westward Sand Transport on the Offshore Bed

A general tendancy exists for westward or south-westward transport of sand on the bed for up to 10km seaward of the shore to the west of Warbarrow Bay (Bastos and Collins, 2002). Donovan and Stride (1961) suggest that tidal currents may have sufficient energy to entrain and "sweep" silt and sand and move it offshore. HR Wallingford (1998) has made a general inference that this material could join a southwest-directed tidal current pathway, possibly feeding a sink (The Shambles) east of Portland Bill. This is supported by Bastos and Collins (2002), who investigated sea bed mobility and transport paths in the vicinity of the Shambles.

Since there is very limited direct and reliable information on sediment transport on the seabed, especially the inshore zone within 3km of the shoreline, the following observations of offshore sediments are included to indicate possible locations of sediment stores or sinks on the seabed.

British Geological Survey Offshore Solid Geology and Seabed Sediment maps covering this area at 1:250,000 scale indicate that large areas of the offshore zone, to a distance of 2,000m from the coastline, are free of sediment. There are, however, patches of muddy and sandy gravels, characteristically less than 0.5m in thickness, offshore of Ringstead Bay, White Nothe, Lulworth Cove, the eastern part of Warbarrow Bay and immediately west of St Aldhelm's Head. A small area of thin, gravel dominated, sediment cover exists seawards of Osmington, whilst gravelly sand conceals bedrock over a small area due east of St Aldhelm's Head (Dorset Coast Forum, 1998).

The presence of large boulders scattered across several parts of the nearshore seabed is reported by Donovan and Stride (1961), but there is insufficient detail given to determine any pattern. Boulders are likely to be relict features of former large scale coastal landslips, mudflows, etc. that occurred during, and since, periods of lower sea-level. One recent rockfall event is described by May and Heeps (1985), who note the rapid removal of fine grained sediment and the isolation of individual boulders in the nearshore zone.

Apart from nearshore ledges, the only significant relief features of the seabed are the Lulworth Banks, offshore Lulworth Cove (May, 1990), and a distinct linear 'step' south of St Aldhelm's Head. The first is the result of the outcrop of Corallian Limestone, due to local anticlinal flexure; the second has been tentatively interpreted as an abandoned cliffline associated with a lower Quaternary sea-level (Mouchel, 1998).

O2 Clockwise Tidal Eddy to West of St Aldhelm's Head

A well-defined clockwise tidal eddy, set up by the blocking effect of St Aldhelm's Head upon tidal flows, exists immediately to its west and may create tidal currents capable of moving sand on the bed. HR Wallingford (1998) suggest the presence of a small-scale sand sink, as there is no evidence for transport around this large headland, and submerged offshore ledges are apparently concealed by a considerable accumulation of sand and gravel. There are currently no further details of this process.

4.2 Wave Driven Onshore to Offshore Loss

W01 Wave-driven loss offshore Ringstead Bay

The Ringstead Bay gravel beach has experienced episodic losses of gravel during storms, especially over the winter of 1989-90 (S W and H Pattison Ltd. 1995). It is thought that material is drawn down from the intertidal beach and retained on the nearshore bed by a series of confining reefs located some 100m 150m offshore of the beaches. However, gaps in the confining reefs appear to permit some permanent losses to seaward by tidal and/or wave-induced rip currents. Further monitoring of the beach is required to determine whether (i) material is returned naturally to the beach and (ii) whether losses to this source have continued following the replenishment using 25,000 cubic metres of marine dredged fill, undertaken in 1996. If losses continue, then consideration could be given to the possibility of blocking the gaps in the reefs, although there could be geological and environmental conservation issues associated.

5. SUMMARY OF SEDIMENT PATHWAYS AND BUDGET - References Map

1. The South Purbeck coast is subject to relatively high wave energy, which declines westwards due to protection from the "Isle" of Portland. Tidal currents are weak in the inshore region, but intensify at the major headlands.

2. Much of the shoreline is composed of cliffs, which vary considerably in height and morphology. The latter is controlled by the influences of regional and local tectonic structures, rock lithology and stratigraphic successions. Clay and shale outcrops are dominant in the west and in the Kimmeridge areas. Elsewhere, Calcareous rocks are the major ground forming materials. Several deep-seated landslide complexes result from the combination of interbedded competent and incompetent strata with basal marine erosion. Rockfalls and topples; translational and rotational slides; slumps and mudslides are active cliff-forming processes at many locations. The balance between the different cliff degregation processes is a function of the spatial variabilities of rock structure and exposure to climatic influences and wave energy. It has produced a complex series of headlands and bays along this coast. A classic evolutionary model based on time-space integration of the various coves and embayments between Durdle Door and Warbarrow Bay is widely known, but oversimplified by many earlier texts.

3. This unit provides an important source of sediment from coast erosion. In spite of a large overall input of sediment from cliff and shoreface erosion (200,000m³a^-1), much of the material supplied is either clays or other weakly resistant rock that is rapidly abraded by wave action and transported seaward in suspension. Solution removes a substantial proportion of material detached from Chalk and Limestone cliffs. Some harder limestones are supplied which persist as boulder aprons upon the foreshore providing protection to cliff toes. Only small quantities of hard flint and chert materials are yielded that can form persistent beach gravels, amounting to around 2% of total coast erosion sediment yields.

4. Littoral drift is discontinuous being intercepted by numerous headlands and embayments. A weak littoral transport divide exists in the vicinity of White Nothe, giving net directions of drift westwards towards Weymouth Bay and eastwards towards Durlston Head. The latter is a fixed and absolute transport boundary for coarse sediment. Actual rates and volumes of wave-driven littoral transport are below potential values, due to: (i) trapping of sediment in various embayments and coves confined by hard rock headlands and shoreline debris aprons, (ii) dissipation of wave energy by offshore reefs, ledges and platforms and (iii) poor availability of littoral sediment and/or the presence of deep water adjacent to several sectors of cliffed coastline. Except at the major headlands, tidal currents are of limited significance in the nearshore zone, but in combination with waves they may promote the movement of sand across the offshore seabed. In western parts, sand transport on the offshore bed appears to operate in a net south-westwards direction, feeding the sediment stores of the Adamant Shoal and The Shambles bank.

5. There are relatively few coastal "problems" because this is an undeveloped and partially inaccessible coastline with very limited and localised attempts at shoreline protection. Habitat and Earth Science conservation interests are exceptionally high. Partly because of the absence of population centres and intensive land uses there have been few studies of the hydraulic and morphodynamic regimes. Routine process monitoring is available for only a very few sites. Most knowledge derives from academic research and infrequent geological survey descriptions.

6. KEY COASTAL DEFENCE AND HABITAT ISSUES - - References Map

This coastline is of prime international, as well as national, importance for its geological and geomorphological features and was given UNESCO World Landscape Heritage status in 2001. It also accommodates several nationally significant sub-tidal, inter-tidal shoreline and cliff habitats (Roberts and Jewell, 2000). It is, therefore, comprehensively covered by European and national conservation designations. It is unusual in having two Marine Nature Reserves, at Kimmeridge Bay and off Durlston Head.

Less than 2% of this frontage has been formally defended, and it is not anticipated that this will be extended in the immediate future (Mouchel, 1998). Consequently, there are relatively few opportunities for significant conflicts between habitat or earth science conservation and shoreline management. An exception to this general rule has occurred at Ringstead, where beach loss and accelerating cliff recession over the 1980s and early 1990s was beginning to threaten properties such that a need arose for improvements to existing ad hoc defences. A beach replenishment using 25,000 cubic metres of marine dredged fill and a retaining terminal rock groyne was undertaken in 1996, but its inception initially generated strong opposition from earth science conservation interests for it had a potential to obscure some key Corallian stratotype exposures. The scheme eventually went ahead on the understanding that (i) the replenishment fill could be cleared from the most important exposures and (ii) the beach would be likely to gradually lose material such that obscured exposures would emerge naturally in the future. A key lesson arising was that the earliest possible consultations between coastal engineers and the earth science community should in future be established so that potential issues on this highly sensitive coastline can be identified. The Jurassic Coast Project is now promoting this type of approach for the World Heritage Site as part of its management scheme see: http://www.swgfl.org.uk/jurassic/consult.htm. and Jurassic Coast, (2003).

There is an issue of Ministry of Defence ownership of the coastline between Lulworth Cove and Gad Cliff, which substantially reduces public access. However, it is probable that the exclusion of public pressure since the late 1930s has been a positive benefit for habitat diversity and geological conservation.

7. OPPORTUNITIES FOR CALCULATION AND TESTING OF LITTORAL DRIFT VOLUMES - References Map

Very few estimates of drift are available and those that have been made are potentially unreliable due to the large differences between potential drift and the drift actually occurring that is controlled by sediment availability, frequent headland obstructions and nearshore reefs. Indeed, the discontinuous nature of the shoreline of this unit with its numerous headlands, nearshore reefs, boulder aprons and pocket beaches means that it is unsuited for definitive studies of drift due to its complexity.

There are, however opportunities to study drift occurring on the major pocket beaches e.g. Ringstead and Warbarrow Bays where each bay would appear to operate as a relatively closed system for gravel and coarse sand. A possible approach would be to compare the sediment accretion/depletion (based on profile measurements) against confining headlands within each bay with estimates of transport derived from modelling based on hindcast waves. Potentially, a beach plan shape model could be set up to simulate the beach responses to SW and SE waves that would tend to cause significant beach re-orientations within the confined bays. Major problems could be introduced by the complex nearshore reefs at these locations that would cause shoaling and refraction of waves, potentially setting zones of wave energy focusing and causing hydrodynamic discontinuities in drift.

8. RESEARCH AND MONITORING REQUIREMENTS - References Map

The basis of recommended future monitoring of coastal processes is outlined in Mouchel (1998). Some of the recommendations are in the process of implementation by the Strategic Regional Coastal Monitoring Programme, a consortium of coastal groups working together to improve the breadth, quality and consistency of coastal monitoring in South and South East England (Bradbury, 2001). A Channel Coastal Observatory has been established at the Southampton Oceanography Centre to serve as the regional co-ordination and data management centre. Its website at www.channelcoast.org provides details of project progress (via monthly newsletters), descriptions of the monitoring being undertaken and the arrangements made for archiving and dissemination of data. Monitoring includes directional wave recording, provision of quality survey ground control and baseline beach profiles, high resolution aerial photography and production of orthophotos, provision of new ground survey control and some baseline beach profiles, LIDAR imagery and nearshore hydrographic survey. Not all of these actions are presently planned for this unit. Data is archived within the Halcrow SANDS database system and the aim is to make data freely available via the website.

Foregoing sections have revealed considerable uncertainties relating to sources and volumes of sediment input and both volumes and rates of sediment throughput via longshore transport. Outputs or losses from beaches are largely a matter of inference as none of these components of the regional sediment budget have been quantified. These deficiencies largely reflect the fact that most of this coastline is relatively remote and undeveloped such that there are few coastal defence issues requiring investigation and the practical needs for further research and monitoring might be considered to be a lower priority than for adjoining coastlines. This situation is unlikely to change in the foreseeable future, but the existing knowledge base would benefit from:

1. Dedicated offshore and nearshore wave data recording. Due to the complex nearshore bathymetry, a permanent recording location is recommended within the offshore zone so as to determine a reliable general wave climate. Wave transformation studies could then be undertaken to develop local nearshore wave climates. Ideally, the offshore recording station should be sited away from shelter afforded by the Isle of Portland and correlations could be developed with wave records in Weymouth Bay to examine intermediate conditions. Where precise inshore wave climates are required some temporary wave recording could be considered to calibrate the transformations from offshore. Ideally some nearshore hydrographic surveys would also be needed to provide reliable bathymetric data for the wave transformations.

2. Routine beach profile (including inshore bathymetric survey) and sedimentology surveys, particularly at locations of dynamic change such as between Ringstead and Osmington. This could assist in calculating beach volume changes enabling better understanding of beach responses such as drawdown and re-orientation induced by storm waves and changes in wave direction, respectively. It would also be beneficial for future beach management to study the behaviour of the Ringstead replenished beach.

3. A systematic analysis of all cartographical and air photo records of this coastline, to produce geomorphological mapping of cliff and shoreline features and for comparisons of present beach, shore and cliff feature positions with historical data determine rates of change with improved precision.

4. Instrumentation of selected unstable slopes, to determine operative processes and rates of movement of ground materials.

Considering that the requirements for knowledge to support practical coastal defence operations are relatively modest, some of the aforementioned work could be promoted by encouragement of the initiatives of academically based researchers.

9. REFERENCES - References Map

ALLISON R J (1986) Mass movement and coastal cliff development of the "Isle" of Purbeck, Unpublished PhD thesis, Department of Geography, King's College, University of London.

ALLISON R J (1989) Rates and Mechanisms of Change on Hard Rock Coastal Cliffs, Zeitschrift für Geomorphologie, Supp. Bd., 73, 125-138.

ALLISON R J (Ed.) (1990) Landslides of the Dorset Coast. British Geomorphological Research Group Field Guide, 125pp.

ALLISON R J and BRUNSDEN D (1990) Some Mudslide Movement Patterns, Earth Surface Processes and Landforms, 15, 297-311.

ALLISON R and KIMBER O (1999) Distinct Element Computer Modelling of Portland Limestone Coastal Cliffs, in: Allison R J (Ed.) Dorset Revisited: Position Papers and Research Statements, West Dorset Coastal Research Group, 53-62.

ARKELL W J (1947) The Geology of the Country around Weymouth, Swanage, Corfe and Lulworth, Memoir Geological Survey of Great Britain, London HMSO, 386pp.

ARKELL W J (1955) Geological effects of the cloudburst in the Weymouth district, 18th July 1955. Proceedings Dorset Natural History and Archaeological Society,77, 90-96.

BASTOS, A and COLLINS, M (2002) Seabed Mobility and Sediment Transport Pathways Along the Dorset Inner Continental Shelf (Southern UK), Report to SCOPAC, School of Ocean and Earth Science, University of Southampton (Southampton Oceanography Centre), Vol. 1 (text), 93pp; Vol. 2, Atlas (21 Plates)

BIRD E (1996) Geology and Scenery of Dorset, Bradford-on-Avon: Ex Libris Press, 178pp.

BRADBURY, A.P. (2001) Strategic monitoring of the coastal zone: towards a regional approach. Report to SCOPAC, South Downs Coastal Group, South East Coastal Group and Environment Agency, 91p.

PETER BRETT ASSOCIATES (1995) Osmington Mills, Dorset: Coastal Protection Study. Feasibility Report Stage 2. Report to West Dorset District Council, 27pp.

BRUCE P and WATSON G (1998) Tidal Streams Between Portland Bill and St Albans, Boldre Marine Ltd.

BRUNSDEN D (1996) Landslides of the Dorset Coast. Some Unresolved Questions, Proceedings Ussher Society, 9, 1-11.

BRUNSDEN D and GOUDIE A (1997, 2nd Edition) Classic Landforms of the Dorset coast, Sheffield: Geographical Association, 48pp.

BURTON H St J (1937) The Origin of Lulworth Cove, Geological Magazine, 74, 377-383.

CANNING A D and MAXTED K R (1979) Coastal Studies in Purbeck, Swanage: The Purbeck Press, 84pp plus appendices.

COLE D (1995) Ringstead Bay Coast Protection Scheme: Geological Survey. Report to West Dorset District Council, 12pp. and 4 Appendices.

DAVIES G M (1956, 2nd Edition) The Dorset Coast: A Geological Guide, London: A and C Black, 128pp.

DONOVAN D T and STRIDE A H (1961) Erosion of a Rock Floor by Tidal Sand Streams, Geological Magazine, Vol. 98, 393-398.

DORSET COUNTY COUNCIL, (1994) The Dorset Coast Today, Dorchester: Dorset County Council. pp. 3-15.

DORSET COAST FORUM (1998) Towards Policy for Dorset's Coast: Marine Aggregates. Dorchester: Dorset Coast Forum. 8p.

HALCROW, (2002) Futurecoast: research project to improve the understanding of coastal evolution over the next century for the open coastline of England and Wales. Report and CD-ROM produced by Halcrow-led consortiun for DEFRA.

HORSFALL D (1993) Geological Controls of Coastal Morphology, Geography Review, 7(1), 16-22.

HOUSE M R (1958) The Dorset Coast from Poole to Chesil Beach, Colchester: Geologists' Association: Guide No. 22, 21pp.

HR WALLINGFORD (1994) Ringstead Bay Coastal Protection Scheme. Report EX 2821. Report to West Dorset District Council, 8pp.

HR WALLINGFORD (1998) Durlston Head to Portland Bill Shoreline Management Plan. Report EX 3824. Report to Mouchel Consulting Ltd., 9pp., Tables and Figures.

JOLLIFFE I P (1976) Weymouth Bay, Dorset. Coastal Regiment Conditions and Resource Use. Report to Weymouth and Portland Borough Council and Dorset County Council, 73pp.

JONES M E (1980) The Landslips at Chapman's Pool and St Aldhelm's Head, Dorset, Proceedings Dorset Natural History and Archaeological Society, 87, 6-17.

JONES M E, et al. (1984) On the Relationships Between Geology and Coastal Landforms in Central-Southern England, Proceedings Dorset Natural History and Archaeological Society, 106, 107-118.

JURASSIC COAST (2003) Dorset and East Devon Coast World Heritage Site Management Plan, First Revision 2003. produced by Jurassic Coast Dorset and East Devon Coast World Heritage Site. 42p. plus 11 Appendicies. See website at: http://www.swgfl.org.uk/jurassic/

KING M (1976) The Geology of Purbeck, Basingstoke: Globe Education (for Dorset County Council Education Committee), 16pp.

MAY V (1990) Unravelling the Coast off Lulworth, Journal of National Association of Field Study Officers, 15, 19-21.

MAY V J (1966) A Preliminary Study of Recent Coastal Changes and Sea Defences in South-East England, Southampton Research Series in Geography, 3, 3-24.

MAY V J (1971) Poole Harbour and the Isle of Purbeck, in: Clark M J (Ed.), Field Studies in South Hampshire and Surrounding Region, Southampton Branch of the Geographical Association, 90-97.

MAY V J and HEEPS C (1985) The Nature and Rates of Change on Chalk coastlines, Zeitschrift für Geomorphologie NF Suppl. Band, 57, 81-94.

MOORE R and BRUNSDEN D (1996) Physico-Chemical Effects on the Behaviour of a Coastal Mudslide, Géotechnique, 47, 259-278.

MOTTRAM B (1972) Some Aspects of the Evolution of Parts of the Dorset Coast, Proc. Dorset Natural History and Archaeological Society, 94, 21-26.

MOUCHEL CONSULTING LTD (1998) Portland Bill to Durlston Head Shoreline Management Plan. 2 Volumes. Report to Portland to Durlston Coastal Group (Lead Authority: West Dorset District Council).

NOWELL, D.A.G (1997) Structures Affecting the Coast Around Lulworth Cove, and Synsedimentary Wealden Faulting, Proc. Geologists' Association, 108, 257 - 268

NOWELL, D.A.G. (2000) The Geology of Warbarrow, Dorset, Geology Today, 7(3), 71-75

PATTISON, S. W. and H. LTD (1995) Ringstead Bay Coast Protection Scheme. Engineer's Report. Report to West Dorset District Council, 40pp.

PERKINS J W (1977) Geology Explained in Dorset, Newton Abbot: David and Charles, Ltd, 224pp.

POSFORD DUVIVIER (1974) SCOPAC Research Project. Sediment Inputs to the Coastal System. Cliff Erosion. Report to SCOPAC

POSFORD DUVIVIER (1999) SCOPAC Research Project. Sediment Inputs to the Coastal System. Cliff Erosion.. Report to SCOPAC.

POSFORD DUVIVIER (1999) SCOPAC Research Project. Sediment Inputs to the Coastal System. Summary Document. Report to SCOPAC, 41-46.

RENDEL GEOTECHNICS and UNIVERSITY OF PORTSMOUTH (1996) Fluvial Inputs into the Coastal System. Report to SCOPAC.

ROBERTS H and JEWELL S (2000) Soft Cliff Biodiversity along the South Coast of England, Newport: Isle of Wight Centre for the Coastal Environment, 36pp.