1. Introduction
The zetaform plan form of Start Bay curving slightly more in the south than to the north is the product of geological control, combining contrasts of rock resistance with structural lineations. Its shape may also be interpreted as an adjustment to incident wave energy, with increasing exposure from south to north/north-east.
The northern section is indented and irregular in plan, due to geologically recent (and continuing) submergence accentuated by differential marine erosion of small-scale variations in rock strength and planes of structural weakness. Between the mouth of the Dart (Photo 1) and Berry Head (Photo 2), the role of wave quarrying is accompanied by sub-aerial denudation in modelling a series of large rocky headlands and small bays. The southern section has a smooth, arcuate plan due to the presence of a near continuous sequence of barrier beaches that have been driven landwards by rising sea level. Former indentations have been eliminated, except for a few minor rocky salients and the headland of Start Point (Photo 3). Thus, these two sectors of coastline represent strongly contrasting responses to marine transgression.
Another highly distinctive feature of the southern and central coastlines, as well as outer Start Bay, is the presence of virtually closed sediment transport systems. Although much of this coastline is undeveloped, the lack of any significant input of sediment to maintain the finite quantity of transportable material provides site-specific shoreline management problems. This problem was first, dramatically, highlighted by the extinction of the coastal village of Hallsands (Photo 4) in the first two decades of the twentieth century due to a disastrously ill-informed decision in favour of offshore/nearshore gravel extraction. This led very quickly to unrecoverable large-scale beach loss and the subsequent destruction of village properties by a succession of storm waves. A more recent manifestation is the erosion of a part of the barrier beach of central Slapton Sands (Photo 15). This near breach has been at least temporarily reinstated, but remains vulnerable to high energy waves and rising sea-level.
A major new source of coastal data is from the Defra-funded National Network of Regional Coastal Monitoring Programmes. The Programmes consist of topographic beach surveys, nearshore bathymetry, aerial photography, lidar, coastal hydrodynamics (waves and tides) and terrestrial habitat mapping. Specifications for data collection are consistent for all regional programmes and the data and analysis reports are made freely available under the Open Government Licence from www.coastalmonitoring.org. The Southwest Regional Coastal Monitoring Programme commenced in 2006. The Lead Authority is Teignbridge District Council, with data collection, analysis and reporting led by a specialist team at Plymouth Coastal Observatory (PCO).
1.1 Shoreline Evolution
The zetaform of Start Bay, although adjusted to the contemporary pattern of refracted waves from the south west and south east, is the product of long-term marine erosion that started during the last interglacial stage of the Pleistocene, when sea-level was approximately +7mOD. The cliffed coastline north of Pilchard Cove is essentially a re-occupation of this ancestral coast, where sediments deposited during succeeding cold climate conditions have been substantially removed. An interglacial cliffline and associated platform is also being exhumed between Hallsands and Start Point. The coastline immediately west of Start Point, particularly between Lannacombe Bay and Prawle Point, retains a well-preserved record of these events (Orme, 1960; Mottershead, 1971, 1982, 1986).
Between Torcross and Pilchard Cove, the degraded slopes that back the Higher Ley, Slapton Ley and Lower Ley lagoons (Photo 5) represent the interglacial cliffline. These are the product of confinement by the landward transgression of a gravel barrier beach, commencing in early to mid-Holocene times as sea-level rose rapidly following the abrupt end of the preceding stage of periglacial climate. This feature, which is composed dominantly of pebbles derived from distant source areas, may once have been continuous with a precursor of the Chesil Beach (Dorset) barrier. As it was driven landwards, perhaps from an initial position some 20 to 25km seaward of the modern coastline, it acquired sediment from the floor of the English Channel and Start Bay. After mid-Holocene times, further landward and upwards migration caused this barrier to be segmented by contact with rocky drainage interfluves, which then converted to headlands as sea-level continued to rise. Former shallow embayments (e.g. the present site of Slapton Ley), and small tidal re-entrants along river channels were eventually blocked by barriers, converting to lagoons. This stage of shoreline evolution is confirmed by seabed sediment stratigraphy in front of modern barriers, as well as in the lagoon basins. An evolutionary time frame has been derived from absolute dating of organic horizons which indicates that the main barrier segments - in particular Slapton Sands - were in their approximate modern positions by about 3,000 years before present. The overall effect of barrier beach migration has been to simplify the originally more indented early to mid-Holocene coastal planform. (Details of barrier origin, development and contemporary behaviour is given in Sections 5.1 and 5.2).
Barrier placement is not evident north of Blackpool Sands. The River Dart estuary is narrow, steep-sided and sinuous - a classic ria - resulting from invasion by Holocene sea level rise. Its confined mouth extends as a buried channel incised more than 40m into bedrock.
An enigmatic feature in the south of Start Bay is the Skerries Bank, which is currently maintained by a virtually closed sediment circulation system. It is not known if this substantial banner bank is entirely the outcome of sediment concentration by waves and tidal currents during the Holocene, or whether it had earlier Pleistocene predecessors. In contrast to the barrier beaches, it is made up of fine sand and shelly debris and is evidently a product of long-term sediment sorting in the nearshore/offshore zones.
Shoreline management is restricted to a few hazard-prone locations. Human agency has had little overall impact on natural coastal evolution, though the rapid destruction of Hallsands village in the early twentieth century is a spectacular exception.
1.2 Hydrodynamic Regime
In addition to the hindcast computations in the regional Shoreline Management Plan (Posford Duvivier, 1998), several sources report wave height and wave period calculations carried out for limited periods of observation. For this reason, their representative value is doubtful. Papers by Holmes (1975a and b) add useful insights into swell wave refraction induced by the Skerries Bank.
Waves approach this coastline from several directions, all of which may be incident over short periods, particularly during the winter. Highest energy waves approach from the east, southeast and northeast; swell waves moving from the southwest or west are strongly refracted by Start Point and the Skerries. The latter is a submarine banner bank and has the effect of focusing wave energy on the sector of coastline between Beesands and Hallsands. It creates a strong alongshore gradient in the nearshore zone, with the highest energy waves incident along the most northerly sector of the Start Bay shoreline, north of Strete.
A hydraulic study carried out in connection with sea defences at Beesands (HR Wallingford, 1991) calculated maximum inshore wave heights, with a 1 in 50 years recurrence, for three points along the Beesands frontage. They were between 2.6 and 3.0m; offshore heights for storm conditions predicted to occur at least once per year were 3.04 to 3.63m. A study undertaken on the durability of proposed rock armour defences at the proximal end of Dawlish Warren quotes a maximum wave height off Berry Head of 5.3m generated by both south-easterly and south-westerly waves monitored between 14 December 1989 and 28 February 1990.
Carr (1974) stated that 50% of significant wave heights in the inshore area opposite Slapton Sands Monument were in excess of 0.16m, but this was based on a few consecutive days of observations under low energy conditions. Carr, Blackley and King (1982) suggest a mean significant wave height of 0.3m for Slapton Sands, but with heights exceeding 0.5m for summer season waves approaching from the south-east and north-east. Their period of observation did not include autumn or winter months. Using a continuous record from a directional buoy located 2km. offshore Slapton village, Ruiz de Alegria-Arzaburu et al. (2010) calculated that for the period April 2007 to September 2008 the mean significant nearshore wave height was 0.65m. During times of dominant easterly waves the mean significant wave height was 0.73m., with a lower value of 0.65m, when waves approached from a southerly direction. Offshore wave heights for fifteen locations offshore Slapton barrier were measured or computed for the period January 1999 to December 2002. The largest, between 5m and 6m, occurred during the storm of January 2001 (Chadwick, et. al. 2005). Ruiz de Alegria-Arzaburu and Masselink (2010) included 27 storm events over a two year period in their analysis of the morphological responses of the Slapton barrier to changes in wave incidence. During easterly storms, modelled mean significant offshore wave heights decreased from 4.0m at Strete to 2.4m at Hallsands; for southerly storms the equivalent values were 2.8m and 1.6m respectively. Longer-term records for this site from the Coastal Monitoring Programme measured nearshore waves using a Datawell Directional Waverider buoy, which is deployed at Start Bay in 10mCD water depth, have been analysed. Between 2007 and 2012 the prevailing wave direction was from the southwest-by-south. Average 10% significant wave height exceedance is 1.34m (CCO, 2012).
Holmes (1975a) reports on the results of a computer analysis of the wave climate of Start Bay, with input of seabed bathymetry and a set of wave ray equations. The program calculated orthogonal paths from deep water towards the coastline and incorporated shoaling and refraction coefficients. Refraction diagrams for waves approaching from the northeast, east and southeast (Holmes, 1975b) reveal that the Skerries Bank is most effective in focusing energy when prevailing swell waves are entering shallow water from the east and northeast. HR Wallingford (1991), however, concluded that waves, especially storm waves, approaching from the northeast experience least modification. The latter study was more empirically based, which may partially explain this conflict of opinion. Holmes (1975a and b) is able to demonstrate that southeast approaching waves create a more even distribution of energy, but the overall refractive effect of the Skerries Bank is to focus maximum incident wave energy on a sector of shoreline between Hallsands and a point immediately south of Beesands. This was independently confirmed by HR Wallingford (1991). Some detailed modification of the distribution of breaking wave energy is also likely to be introduced by variations in wave period.
Mean tidal range varies from 4.3m at Berry Head to 4.6m at Start Point. Inshore tidal currents have low velocities, but may be locally significant in transporting fine sand and biogenic debris further offshore (Dyer, 1975). Robinson (1961) and others (e.g. Hails, 1974; 1975) have argued that the south/south-west moving ebb and north/north-east directed flood currents are responsible for moving sediment in an anticlockwise pathway around the Skerries Bank. Pingree and Maddock (1983) have modelled a strong tidal vortex in the vicinity of, and directed away from, Start Point. Here, current velocities may exceed 2.0ms-¹ during a part of the tidal cycle. Further detail on the dynamics and transport potential of these currents is given in Section 5.5.
2. Sediment Inputs
2.1 Cliff & Shoreline Platform Erosion
Introduction
With the principal exception of the coastline between Torcross to north of Strete, most of this unit is characterised by cliffs and narrow, discontinuous shoreline platforms developed in a succession of slates, grits, siltstones and massive mudstones with an approximate east to west strike. Between Start Point to just south of Tinsey Head, hornblende, chlorite, mica and quartz schists provide a distinctive ground-forming substrate. Cliff morphology reflects locally very variable lithological and structural controls whilst mean cliff height shows an overall decline from north to south. Platform development tends to be limited to the sites of minor headlands, with boulders and other coarse debris from occasional cliff falls often partly or wholly concealing these features. Posford Duvivier (1998) concluded from analysis of the movement of mean high and low water, 1886-1996, that rates of cliff recession were slow, averaging some 0.25 to 0.30m per year. Between Sharkham and Blackstone Points, the position of the cliff top has been virtually static over the past century.
Analysis of Coastal Monitoring Programme 2007 and 2012 datasets indicates that cliff input is minimal. There is a lack of detailed study information regarding cliff sediment composition and proportions of sediment grain sizes; combined with the low volumes of cliff input yielding shingle or sand grade beach material, this means it has not been possible to quantify inputs from the cliffs, nor annual average cliff recession rates. No significant landslides or cliff erosion events have been measured since the Southwest Programme commenced in 2006. (See PCO Annual Survey Reports for further details).
In the extreme south of this unit, compressed but highly fissile schists produce a rugged coastal slope with basal cliffing. From south of Tinsey Head (Photo 6) to North Hallsands, deeply weathered and substantially decomposed hornblende and chlorite-rich schists occur. Between Greenstraight and Start Point, relatively more resistant mica and quartz schists replace these. Above exposed bedrock, Head deposits (formed by downslope mass movement under former periglacial condition) support a concave slope of 10-15° and partly conceal elevated shore platforms.
In a series of papers, Mottershead (1971; 1981; 1982; 1989; 1997 and 2000) and Mottershead and Pye (1994) have applied both quantitative and qualitative methods of field measurement to the analysis of coastal weathering of the schist outcrop between Start Point and the Kingsbridge Estuary, west of this unit. Spray corrosion due to salt crystallisation is identified as a potent weathering process, exploiting both micro and macro bedrock structures and effecting downwearing at rates of between 0.3 and 0.6mm per year. Weathering products are rapidly removed by solution, overland flow and wind and do not contribute in any significant way to the local beach budget. However, where Head and Raised Beach deposits are exposed to wave erosion, a wide variety of clast shapes and sizes are released to the cliff base and account for much of the sediment composition of local beaches. It is probable that the geomorphological regime that prevails west of Start Point also operates to the north of it. However, because of the significant contrast in exposure and aspect, both hydraulic forces and weathering processes occur at different rates in these two areas. Former frost weathering may also account for the more jagged outline of parts of this coastal slope. Wareham et al. (1991), in an investigation of ground conditions at Start Point lighthouse identify structural damage ascribed to wedge failure and displacement along a planar failure surface. This may be the result of basal slope steepening due to wave erosion.
At the site of the abandoned village of Hallsands (Photo 4), the upper rock platform, at 6-8m O.D. is dissected by gullies and small ravines that exploit areas of deep weathering. These were formerly infilled with raised beach debris, and periodically contain material from rearward cliff degradation. This is removed by wave action and contributes a small input of coarse clastic material to the local beach system. This process appears to have accelerated in recent years, necessitating closure of this site to public access in 1996. On the lower rock platform, temporarily protected by an emergency seawall constructed in 1904, basal wave notching up to 25m in width has developed (Kalaugher and Grainger, 1991; Mottershead, 1986). Robinson (1961) reported nearly 7m of cliff top retreat between 1907 and 1960 at the Coastguard Cottages. Posford Duvivier (1998) calculated that the prevailing rate of shoreline recession here is 1.0m per year, but presumed that most erosional debris is too fine to be retained in the littoral transport system. Hydraulics Research Station (1963) reported 20 feet of cliff top recession in the vicinity of Start Point, 1907-1960, i.e. less than 0.10m per year. However, the methods of analysis used by these two studies differ in both approach and level of detail, so it is probably unwise to assume a progressive increase in erosion rates over the twentieth century. Cliffs occur at Tinsey Head (Photo 6) and Limpet Rocks, south of Torcross (Photo 7), cut into sandstones, slates and mudstones of the Meadfoot Beds. They appear morphologically stable, with rates of cliff top recession in the order of 0.2 to 0.3m per year (Posford Duvivier, 1998). There is some evidence of past shallow landslipping at Limpet Rocks.
Cliff development recommences north of Strete Gate, but between here and Pilchard Cove, the coastal slope is removed from direct wave erosion by the presence of the most northerly segment of the Start Bay barrier beach system (Photo 8). It is therefore interpreted as a relict form inherited from the +5 to +8m high sea-level of the last interglacial.
Between Asherne and Blackstone Point, cliffs alternate with a sequence of coves and small bays; the latter coincide with local reductions in rock strength within the complex lithological and structural character of the Dartmouth Slates. In places, north of Blackpool Beach, near vertical cliffs are directly controlled by the disposition of cleavage planes. Cliff height reduces progressively southwards, but there is little evidence for active erosion. Posford Duvivier (1998) calculate a mean rate of cliff top recession of 0.3m per year (1886-1996). An exception to this long-term stability is the northern headland enclosing Blackpool Sands, where there is a tendency for recurrent slips and falls (Hydraulics Research Station, 1963). Debris includes overlying Head as well as rock material, but storage time is short. A rock revetment was constructed in 1992 in an attempt to address the basic cause of slope instability, i.e. basal wave erosion. This reduces yet further the insignificant input to the local beach transport system provided by output from this erosional site. Slightly higher than average retreat rates are associated with diabase intrusions, e.g. at Leonard's Cove.
Mottershead (2000) examined weathering rates of local stone used in the construction of local fortifications during the 1530s to 1540s. These included three structures in Dartmouth, built with resistant siltstones and shales from the Dartmouth Slates series. Although in protected sites near the estuary entrance, a median rate of surface reduction of 9.3mm/century was determined. The precise weathering processes were not determined, but comparable, possibly higher, rates may apply to some open coast exposures of the natural bedrock north and south of the estuary mouth. However, cliff top position between Blackstone Point and Inner Froward Point (the opposing headlands delimiting the Dart entrance) shows no movement over the past century (Posford Duvivier, 1998).
Between Inner Froward Point and Berry Head, most of the coastline is cliffed, with cliff heights of up to 130m (Photo 9). Several coves and larger embayments (Pudcombe Cove, Long Sands, Man Sands and St Mary's Bay) are defined by strong headlands, with incipient shore platform development. Offshore rocks and reefs (e.g. Cod Rocks) record long-term coastal recession. The position of the cliff top has been stable over the past 100 years at most locations. Posford Duvivier (1998) calculate a mean rate of retreat of 0.2 to 1.0m per year between Berry Head and Sharkham Point; and 0.1m per year between Sharkham Point and Inner Froward Point.
Variations in planform of this coastline reflect spatial changes in the relative erosional resistance of the predominant grits, shales and slates, interrupted by igneous sills and dykes. The latter are responsible for local salients. Differential erosion is particularly evident where faulting has juxtaposed Devonian limestones and slates, e.g. at and south of Berry Head. There are several sites of past or current cliff instability evidenced by falls and slides. These include currently inactive slips in the 'double' bay between Durl and Berry Heads (Posford Duvivier, 1998) and deeper-seated rectilinear and rotational slides on the north side of Sharkham Point and at Southdown Cliff - see Photo 10 (Doornkamp, 1988). In the latter case, a series of backtilted, arcuate shaped blocks, forming a virtual morphological stairway, and well-defined tension cracks at the cliff top, indicate both historical and continuing instability.
The most thoroughly investigated site of cliff mass movement is St Mary's Bay, Brixham - see Photo 11 (Derbyshire, Page and Burton, 1975; Derbyshire, Cooper and Page, 1979). In the north of this complex, there is evidence of contemporary rocksliding, talus and soil creep; in the central section, small-scale slumping, rockfalls and slab failures predominate, whilst the southern sector is dominated by rotational shear sliding and earth-flow over a thick silty clay regolith. Through analysis of air photographs (1946-1954) detailed field mapping of slope facets and site monitoring of morphological and vegetation changes from 1966 to 1974, Derbyshire et al. (1975; 1979) concluded that larger-scale movements were triggered by immediately preceding periods of prolonged rainfall. However, basal oversteeping by wave erosion, especially waves from the east and south-east sectors, provided the crucial longer-term preparatory factor. Failure events were most likely to occur when there was a coincidence of high soil and groundwater tables, high tide and a period of exceptional wave energy. The latter is important as a means of removal of basal debris, some of which is retained within the local pocket beach within the bay. Overall, this cliff and coastal slope system has maintained an equilibrium form, with basal undercutting and shear sliding in approximate balance. However, between 1946 and 1976, there was approximately 14m of cliff top retreat, giving an annual rate of loss which is substantially higher than for this coastline as a whole. Hydraulics Research Station (1963) examined evidence for cliff instability at St Mary's Bay in the period 1890-1960. They concluded that failure events were episodic, but did not identify causes. The position of Mean High Water retreated approximately 14m, 1880-1905, providing some earlier evidence of the importance of debris store removal and basal erosion of bedrock. At both St Mary's Bay and Sharkham Point, rotational failure is promoted by the presence of Head overlying bedrock, and slip surfaces may occur at their mutual interface. A significant proportion of beach material therefore derives from relict periglacially weathered and gravity-transported debris.
At Berry Head (Photo 2), some local morphological diversity may be due to karstic pits and depressions in the limestone outcrop forming this headland. These have been opened up, at least in part, and modified by wave-induced cliff retreat (Perkins, 1971). Potential also exists for the exposure of sub-surface fossil phreatic caves, as they are intersected by cliff retreat.
3. Littoral Transport
Introduction
Analysis of Coastal Monitoring Programme 2007 and 2012 topographic baseline and lidar datasets have quantified sediment transport rates and directions. Numerous studies report observations on directions, and a few estimate rates and quantities, of littoral transport. The overall consensus is that there is a sequence of headland delimited sub-cells, with some exchange of sediment, constituting together a virtually closed transport macro cell between the absolute boundaries of Start Point and Berry Head. There is seasonal, and perhaps longer-term, exchange between beaches and the nearshore zone (Section 5.2), but no convincing evidence of any net offshore to onshore transport under contemporary hydrodynamic conditions (Section 4). This is in contrast to the Holocene history of barrier evolution (Section 4), although some tendency towards continuing landward translation might be inferred from beach morphodynamics in recent decades (Section 5.2). The following account examines the evidence for bedload movement of the predominantly coarse sediments (coarse sand, granules and gravel) moved by beach drift and breaker zone/nearshore transport in each of the major sub-cells. Net transport pathways are subject to frequent reversal because of rapid changes of incident wave direction. Gross rates and quantities of sediment movement are therefore substantially higher than net rates quoted, but are currently unquantified.
Job (1987) suggests that net northwards movement occurs between (a) Start Point (Photo 3) and Tinsley Head (Photo 6); and (b) Beesands (Photo 12) to Limpet Rocks (Photo 7) when refracted waves from the southwest and southeast prevail. This, however, is reversed under conditions of northeasterly approaching waves. Rates may be highest under storm wave conditions (HR Wallingford, 1991; South West Water Authority, 1979), particularly between Tinsey Head and Beesands where wave energy is focused by refraction due to the Skerries Bank offshore. Beaches tend to widen from south to north, and are widest immediately south of the Hare Stone, Tinsey Head and Limpet Rocks. Coastal Monitoring Programme data supports a net south to north pathway, with some interception of drift at intervening headlands (although quantities moved are modest). The mean grain size at Beesands is slightly smaller than that further north, at Torcross. Between Tinsey Head and Limpet Rocks analysis of Coastal Monitoring Programme data indicated a lower rate, in the order of 1m³ per year of beach grade material. This is a reduction from the 2004 estimated rates of 3-10,000m³ per year. Coastal Monitoring Programme data suggests that in the context of finite beach volumes actual long term drift rates are lower than the theoretical values estimated, as there is no evidence of either massive build up against the headlands or depletion of beaches with material moved north to Slapton Sands. It is probable that decadal or longer term variations in wave climate result in gravel redistribution either northward or southward, but overall the net drift is northward at a low rate and volume.
None of the headlands on this sector are substantial features; they are moderately defined salients made up of rock platforms. Nonetheless, it would appear that they successfully confine longshore sediment movement, and are only by-passed under high wave energy conditions (Job, 1987). Each therefore defines a discrete local transport sub-cell. Visual evidence of coarser grades of sediment adjacent to the updrift side of each salient may provide some further indirect proof of their role as transport discontinuities.
It is uncertain if the catastrophic loss of beach and nearshore gravel in the early years of the twentieth century at Hallsands has had a subsequent perturbating effect on the littoral transport system. Quantities available for movement have obviously diminished, but Job (1987) has revealed that, up to the 1890s, the dominant pathway between Tinsey Head and Hare Stone operated from north-to-south. Beach sediment loss and net drift reversal thus coincided, but without any evident causal relationship.
Given that a weak net northwards pathway has prevailed in each of the sub-cells of this unit for the last century, and that beaches have shown some net loss of volume in recent decades (Section 5.2), it may be reasonable to infer a small offshore transfer of coarse sediment, generally into the nearshore zone rather than fully offshore. The modest recovery, after 1903, of the beach between Hallsands and Start Point (Robinson, 1961) however, indicates a reverse transfer. However, this phase occupied at least 30 years, so sediment feed must be small. The source(s) of re-supply are not precisely known, and could be linked to gravel mobilisation induced by dredging up to 1902.
The consensus view is that Slapton Sands (a misnomer, as the dominant constituent is fine gravel) is a closed transport system lacking any significant fresh inputs and providing little output to adjacent units. Drift occurs intermittently, with an arguably balanced budget of northerly and southerly-directed movements. This is a function of changes in wave approach; waves arriving from the southwest are substantially refracted by Start Point, and further energy losses are due to the presence of the Skerries Bank. Wave energy, and thus sediment flux, is least along the central sector of Slapton Sands (Morey, 1976). It is probable that highest rates of littoral drift occur when north-easterly waves predominate, but they have a small annual frequency compared with waves from the south-east and south-west. Job (1987, 1993) reports on the results of a short-term tracer experiment, which demonstrated that net north to south transfer occurred under sustained south-easterly waves, but was reversed over longer periods by waves from the south and south-west. Given the 'fossil' nature of this barrier beach (absence of natural replenishment from offshore sources), long-term unidirectional transport would result in its eventual updrift thinning and excess downdrift accumulation. This conclusion is supported by Ruiz de Alegria-Arzaburu et al. (2010). From analysis of 2012 aerial photography and 2007-11 topographic data collected through the Coastal Monitoring Programme, the beach is wider in the northeast, suggesting long-term unidirectional transport from southwest to northeast. The long-term survival of Slapton Ley would appear to be evidence that significant long term unidirectional transport has not prevailed in the past, although the beach has thinned in recent decades and is becoming at risk of breaching (Halcrow 2002) - see also Section 5.2. Between October 2006 and October 2007 the central and southern sectors of the Slapton barrier retreated between 5 and 10m, losing a large, unquantified, volume of sediment; during the same period the northern sector increased in width by approximately 10m, thus demonstrating strong net northwards drift during this period of energetic waves (Masselink and Buscombe, 2008).
Carr (1974) also conducted two brief tracer experiments, using clasts from Chesil Beach (mean diameter of 2.40 phi). During higher energy wave states in February, movement was initially from north to south, but was overall dominantly northwards. In October, initial net movement was in the order of 150m southwards, but thereafter nearly 1,000m of northwards transport was measured. The weight of evidence is thus in favour of a weak, net northwards, movement. This is confirmed by Posford Duvivier (1998), who calculated a gross drift rate of 60,000m³a-¹ for this sector. Analysis of 2007-11 baseline topographic data collected through the Coastal Monitoring Programme confirmed net northwards transport of 1-3,000m³ per year, although increasing northwards to approximately 5,000m³ per year at Strete Gate where there appears to be a weak, drift reversal, dependent on prevailing wave direction and conditions. These rates are a reduction from the 2004 estimated 10-20,000m³ per year. Chadwick et al. (2005) observed that the direction of net longshore transport for the entire length of the barrier varied each year between 1999 and 2002, but with a mean northwards movement of 75,000m³ annually (maximum of 150,000m³ per year equivalent for short periods). These latter figures were derived from the application of several theoretical transport rate formulae and with reference to a sequence of plots of shoreline positions. Net northerly transport is also clearly indicated by Ruiz de Alegria-Arzaburu et al. (2010), using different methodology - see Section 5.2 on Beach Morphodynamics. Although Posford Duvivier (1998) state that there is no convincing evidence of any significant by-passing of Limpet Rocks to feed the beach at Torcross and further north, accretion to the west of Limpet Rocks, and an erosive trend between Torcross Beach and further north was observed in the 2007-11 topographic data collected through the Coastal Monitoring Programme, implying a partial littoral boundary at Limpet Rocks. HR Wallingford (1991) suggested that there may be periodic movements of sediment around this salient, but were non-committal on quantities or conditions promoting transfer. Recent behaviour of the beach at Torcross (Photo 5) implies that cross-shore sediment movement is more important than longshore.
Job (1993) has produced some evidence that the net pathway during the nineteenth century was southwards, pointing to the exceptional height and width of the beach at Hallsands (in 1895) as possible evidence of successful by-passing of current sub-cell boundaries along the entire length of the Start Bay barrier beach system. If a net northwards direction now prevails (Chadwick et al., 2005), it might be evident from patterns of longshore grading of mean particle size and shape. Progressive downdrift reduction in mean particle size, for example due to abrasion, might be expected. This is confirmed by Carr (1974) and Gleason et al. (1975), but both sources report frequent short-term and short-distance reversals of longshore grading due, apparently, to rapid variations in prevailing wave approach. The 3km of Slapton Sands that is potentially available for unimpeded transport does not, therefore, show any consistent pattern. Job (1993) argues, from his own experimental data, that longshore grading reveals a small reduction in mean particle size southwards. This may be an artefact of specific experimental conditions, because Carr (1974) was able to demonstrate that maximum longshore transport distances recorded by monitored pebbles were consistently associated with the smallest particle sizes (it is the inverse of the relationship determined from corresponding experiments on Chesil Beach, Dorset). This should, in theory, support a northwards reduction in particle size, a trend that is observed by Chadwick et al. (2005). However, Carr (1974), Gleason et al. (1975) and Carr et al. (1982) state clearly that their data provides a lack of statistical correlation between clast travel distances and prevailing wave parameters. This would suggest non-equilibrium between beach dynamics, littoral transport and hydraulic processes, possibly because of the influence of antecedent beach morphology and sediment grading. None of the above authors is able to demonstrate any cross-shore sorting, though Latham et al. (1998), working at a site on Torcross beach, measured a slight tendency to coarsening from foreshore to berm crest. Chadwick et al. (2005) state that there are contrasting patterns of particle sorting above mean tide level for several parts of Slapton beach, but do not offer further detail or any specific explanations.
Huntley and Bowen (1975 a and b) report results from an experiment involving the introduction of a dye tracer inside the surf zone at a site on Slapton Beach. This revealed a slow northwards movement inside the breakers, but a southwards drift outside. The latter was probably induced by tidal flow. Weak rip currents were also apparent, thus setting up a small-scale, repetitive cell-like circulation pattern that was indirectly confirmed by beach cusp spacing. This, however, is most likely to involve only suspended sediment.
Ruiz de Alegria-Arzaburu and Masselink (2010) undertook an assessment of three-dimensional morphological response of the Slapton barrier beach to two characteristic storm types- the first generating swell waves approaching from the south, the second for wind waves from the east. This study included 27 individual storms over a two year period, with the first type having a higher frequency of occurrence. Southerly waves caused accretion of the supratidal zone and erosion of the intertidal profile, with more loss of material in the southern sector than the north. Conversely, easterly waves induced supratidal erosion and intertidal accretion along all parts of the barrier, with an overall increase in beach volume. During intervening periods of moderate energy waves, a small supratidal berm was created, resulting in some gain of volume i.e. the sediment budget during these periods was positive. Opposing longshore energy fluxes operated during those periods when incident waves associated with the two storm types were effective. A net northwards directed component of transport prevailed during southerly storms and a southwards component during easterly storms. During the period of data collection the northern sector widened by up to 50m and the middle and southern sectors retreated by over 40m for separate intervals of a few months. The largest volume of losses from erosion occurred near the midpoint of the barrier, which were up to 50,000m³; however, net volume change for the barrier as a whole was less than 7,000 m³. Longshore transport rates were derived from indices of morphological change, and indicated that some 30,000 m³ were transferred northwards over the survey period, thus confirming that net littoral drift was to the north. This study was able to demonstrate that the morphological impact of storms was directly related to the directions of wave approach, storm duration and the antecedent conditions of beach form. It corroborated previous research (Austin and Masselink, 2006) that revealed how beach dynamics in the southern sector of Slapton Sands are principally controlled by cross-shore sediment exchange, whilst longshore transport dominates morphological variability along the central and northern sectors. Analysis of Coastal Monitoring Programme data suggested that there was generally a loss of sediment in the order of 6,000m³ per year into the sub-tidal nearshore zone rather than further offshore, and available for cross-shore transport along the entirety of Slapton Sands frontage.
Ruiz de Alegria-Arzaburu et al., (2010) conclude using canonical correlation analysis, that there is a strong positive relationship between wave hydrodynamics and both alongshore and cross-shore morphological variation revealed from a continuous 18 month programme of weekly field measurements for a set of shoreline plans and profiles covering 2.5 km of the northern sector of the Slapton barrier. Wave climate statistics were obtained from an offshore wave rider buoy located in 10m water depth, whilst shoreline plans were derived from video imaging. Morphological variation during this period was a mean of 18m (8m of erosive and 10m of accretionary displacement). During easterly storms cross-shore exchange was dominant, with supratidal erosion contributing intertidal deposition; alongshore sediment transport was more strongly promoted when southerly storms prevailed, resulting in net erosion of the intertidal area of the southern sector of the study site and accretion to the north.
Several authors have noted that the evident widening and thickening of the barrier beach northwards to Pilchard Cove might be taken as indirect evidence of net northwards longshore transport and long-term retention (storage) from south to north (Photo 8). Coastal Monitoring Programme data supports the net northwards transport. Relic cliffs behind the beach provide further evidence of long-term accumulation. There may also be some evidence for an increase in recent decades of the frequency of small "pulses" of gravel moving onto and along this beach system (Job, 1993). If so, it is difficult to link this behaviour with the Slapton Sands system updrift.
Analysis of Coastal Monitoring Programme 2012 aerial photography and 2007-11 topographic data indicates the beach is wider in the northeast, suggesting long-term unidirectional transport from southwest to northeast, with drift rates along Strete Gate of 3-10,000m³ per year, reducing to 1-3,000m³ per year at Strete. There appears to be a potential drift reversal at between Blackpool Sands and Pilchard Cove with some accumulation observed at Strete Gate. Reversals of net drift within the confined bay beach of Blackpool Sands (Photo 13) can be correlated with prevailing waves, but there is no evidence of any gains from, or losses to, adjacent beaches (Hydraulics Research Station, 1963; Perkins, 1971). Posford Duvivier (1998) calculate potential drift of 77,000m³ per year for this sector. This apparently includes both northwards and southwards movement, so net northwards transport is well below this rate. North of Pilchard Cove, limited sediment supply restricts quantities moved longshore.
Littoral transport is confined to a series of pocket beaches within coves and small bays e.g. Man Sands (Photo 14), many of which are relatively well sheltered from wave action. Analysis of Coastal Monitoring Programme 2012 aerial photography and 2007-11 topographic data indicates a southward drift of material, opposite to the drift direction presented proposed in 2004. Hydraulics Research Station (1963) and Posford Duvivier (1998) described evidence for seasonal drift reversals linked to beach volume changes. It would appear that each beach is effectively independent, although some net northwards routing of sediment along this sector is possible under exceptional (and relatively infrequent) high wave energy conditions (Derbyshire et al., 1975). Berry Head is an absolute transport barrier for the longshore movement of coarse sediment.
4. Sediment Outputs
The beaches of this frontage occupy relatively self-contained bays or sub-cells and tend not to permanently lose significant quantities of their material by transport offshore or alongshore. Outputs of beach sediment have occurred by the following processes:
4.1 Nearshore Dredging (Hallsands)
An estimated 300,000m³ of gravel was dredged from the submerged inshore subtidal shoreface between 1897 and 1902 to provide aggregate for dock construction at Devonport. An alternative estimate of 1,600 tonnes per day over 5 years giving a total of 1.8 million m³ has also been made. Undermining of sea walls was recorded in 1900 and by 1903, beach levels had dropped between 4m and 6m and mean high water mark had moved landwards some 25-30m. It was concluded that beach material was drawn down into the large depressions created by gravel extraction. Contemporary concern forced the cessation of dredging operations in 1902, and then the village was progressively destroyed during a series of storm events culminating in severe losses in the storm of 26th January 1917 (Worth, 1904, 1909, 1924; Tanner and Walsh, 1978 and 1984; May (2003). Further details are given in Section 5.2.
4.2 Beach Mining
Deliberate removal of gravels and sand was carried out up to the 1930s by a licensed commercial operation between Scabbacombe Head and Berry Head. Unfortunately, there are no immediately available details of locations (thought to include Long Sands and Man Sands) and amounts taken, nor when this practice started. Beach mining was also undertaken at Blackpool Sands between c. 1870 and 1908, when material was taken for road making.
4.3 Beach Gravel Abrasion Loss
Within each of the beach transport sub systems between Hallsands and Blackpool Sands, the potential for loss of volume through wave-induced clast attrition and abrasion is high. Latham et al. (1998) used exotic clasts, comparable in density and shape but slightly larger than indigenous material, to assess abrasion loss at a site near Torcross. They calculated that it would require 2,300 years to produce a 90% volume loss of flint and chert clasts. This is substantially longer than for loss rates determined by comparable field tests and controlled experiments performed by the BERM project for gravel beaches in East Sussex (see Volume 5 of this study). However, wave energy along Start Bay is significantly less than it is at the coastline between Brighton and Beachy Head. The results described by Latham, et al. (1998) were based on a very short period of data collection, whilst those of the BERM investigation are more representative of a range of dynamic conditions promoting particle wear.
The only other possible source of beach material loss involves deliberate removal carried out up to the 1930s by a licensed commercial operation between Scabbacombe Head and Berry Head. Unfortunately, there are no available details of locations and amounts taken, nor when this practice started.
5. Sediment Stores and Sinks
5.1 Origin and Composition of the Start Bay Barrier Beach System
This swash-aligned coarse clastic barrier system extends some 9km from Hallsands to Pilchard Cove, but is segmented by rocky salients into several quasi-independent units. It is composed primarily of granule-sized material, with fine gravels and interstitial sand with variable cross-shore size grading. The majority of clasts do not derive from immediate up or down drift exposures of bedrock. Morey (1976; 1980; 1983); Mottershead (1986); Mercer (1966) and Job (1993b) calculate that over 80% of clasts are composed of flint or chert within the Slapton Sands section; this percentage may decrease slightly to the south, at Hallsands. As there are no local sources of flint or chert, it is presumed that this material derived originally from either Chalk or Palaeogene outcrops on the floor of the English Channel several kilometres seawards of Start Bay( May, 2003). This can only imply that it was incorporated into an ancestral barrier beach forced progressively landwards by rising sea-level during the mid-Holocene period. The remaining 20%, or so, of clasts are of varied lithology, principally slate; felsite; rhyolite and quartzite (Hails, 1974; Mercer, 1966). A proportion of this material is likely to derive from the erosion of the local cliffline, e.g. quartzites from schists in the extreme south of Start Bay. Some has been contributed from the catchment drained by the River Dart and other small streams now discharging into lagoons. An unknown, but probably small, percentage is from seabed exposures of Permian breccias. The composition of the detached barrier beach that constitutes Blackpool Sands is distinctive, with only just over 50% exotic material and the remainder relating to locally available lithologies. However, much of the locally-derived sediment is slate, which experiences relatively rapid abrasional wear compared with other constituents. This fact may also help to explain why median grain size at Blackpool Sands is smaller than it is on Slapton Sands (Hydraulics Research Station, 1963).
Morey (1976, 1980, 1983) has presented a history of the evolution of the Start Bay barrier beach system based on detailed sedimentological and stratigraphical evidence, particularly from the sediment sequence in the Slapton Ley lagoons. He demonstrated that Slapton Sands had developed across a former shallow estuarine embayment backed by a now degraded cliffline, presumed to be of Ipswichian interglacial age. Barrier migration dammed six streams that formerly flowed into the estuary, thus creating the Upper and Lower Ley lagoons. Previously, Slapton Sands was part of a more continuous barrier, backed by wider possibly more continuous lagoons, at one or several stages of lower sea-level. Protruding headlands segmented the barrier as it translated landwards through "rollover", forced by mid Holocene sea-level rise. Other segments were driven against elevated hinterland topography at Beesands, Hallsands, Strete to Pilchard Cove and Blackpool Sands, also forming lagoons within valleys at the first two sites. Morey (1980; 1983) concludes from the stratigraphic record that the Slapton Sands barrier has been in its approximate present position for slightly longer than 3,000 years.
The detailed evidence for this interpretation of the chronology of barrier migration and stabilisation is taken from several published sources. Briefly, these are as follows:
- Relict barrier sediments close to the modern beach have been described by Hails (1974;1975), from vibrocore surveys of shallow seabed material. These are considered to represent the position(s) of former barrier(s) destroyed by overwashing, foreshore erosion or failure of sediment supply. Barrier destruction and reformation might have been a function of the original presence of a "chain" of barrier islands, with tidal passes between them possibly coinciding with stream outlets, or valleys within the pre-existing basement topography. These are recorded on the contemporary sea floor as buried channels, infilled with up to 28m of Holocene sediment (Lees, 1975; Hails, 1975).
- Given the possibility of repeated cycles of barrier growth and breakdown in the mid and late Holocene, lagoons would have varied considerably in their degree of exposure, at times being tidal bays. This is in part confirmed by the absence of freshwater peats - representing lagoonal sedimentation - outcropping in the modern offshore area. The creation of a continuous barrier (infilling of tidal passes) was, however, accomplished before it assumed its modern position; this is evident from freshwater peats and lacustrine clays beneath the gravel of the present beach, and thus proof of its landward transgression. Job (1993) has observed that during this latter part of barrier history, there could have been no new inputs of coarse sediment. A finite "reservoir" has therefore been continuously recycled in an effectively closed transport system since then. A radiocarbon dated peat bed at the base of the Beesands barrier (c. 4760 +50 years B.P.) suggests that this may have been the situation for approximately the last 5,000 years.
- On the assumption (taken from other regional evidence) that mean sea-level was about -5m O.D. at 5,000 years B.P. and -40m O.D. at 10ka B.P., this gives a mid Holocene barrier migration rate of nearly 1km per 1,000 years. During this stage a prototype Skerries Bank might have contributed some input of sediment to the developing barrier/barrier islands, perhaps still retained today as the interstitial sand component. Repetitive units of estuarine minerogenic and organogenic (saltmarsh) sediments beneath Lower Bay deposits (Hails et al., 1975, see Section 5.5) record this process of rapid transgression before the barrier was finally closed. A dated freshwater peat horizon (8,108 +60 years B.P.) on the inner south-west margin of the Skerries Bank does, however, suggest a temporary phase of exclusion of brackish or saline conditions during migration. It might also suggest temporary barrier stability.
- Borehole logs of lagoon sediments in Lower Slapton Ley (Morey, 1976, 1980, 1983; Crabtree and Round, 1967) reveal a sediment stratigraphy that confirms part of the above event sequence. At the base, estuarine clays indicate that the site of the modern lagoon was open, though not fully exposed, to marine conditions. They pass beneath the present barrier sediments and grade into the alluvium infilling the Upper Ley. Marine muds succeed this unit, suggesting greater exposure to marine conditions, possibly caused by barrier breakdown and/or an intensification of wave energy. The succeeding unit consists of organic freshwater silt, grading upwards into fen peat. Pollen content suggests the dominance of reedswamp, but with some vestigial saltmarsh. This is a clear indication of the establishment of an environment transitional between estuarine and lagoonal, implying progressive isolation from the open sea by barrier approach. Intercalated gravel is most likely to be the product of washover events. Radiocarbon dating of the base of the fen peat is approximately 2,900Ka B.P., thus giving a definite date for barrier closure. A date for the top of this peat bed is circa 1850ka B.P., providing about 1ka for its accumulation. A unit of muddy sand abruptly terminates fen peats; it fills small depressions, thickens seawards and may be associated with a gravel fan at the southern end of the Ley. The probable explanation is that this deposit is due to a barrier breach. Its repair is indicated by the succeeding accumulation of terrigenous diatom-rich lacustrine muds, in a water depth of approximately 2m - 3m. Interdigitated gravel sheets within this layer record further washover events, up to historical (medieval?) times. Crest height was apparently raised substantially during the first four centuries A.D., but has reduced subsequently.
The now infilled lagoon at North Hallsands (Greenstraight) provides a similar, though compressed and less complete, stratigraphy. A peat exposed on the foreshore in front of the Hallsands barrier has been given a 'most recent' date of approximately 1.7Ka B.P., and is truncated by a gravel washover deposit.
From this evidence, certain deductions or inferences are possible, namely:
- Gravel was accumulated from a wide area of Start Bay, and well beyond, to feed the barrier system in its initial growth and later transgressive stages.
- Additional supplies of gravel from offshore were no longer available from about 5,000 years B.P., when a closed sediment transport system was established.
- The building of a continuous barrier to replace an earlier aligned 'chain' of barrier islands must have involved changes in morphology if there was no significant further input of gravel from original feeder sources. Passes between islands would have been infilled by their narrowing and elongation, with or without crest elevation. This may be further explained by the fact that wave energy is least along the central sector of Slapton Sands. Thus, once gravel was emplaced, it had less mobility, a condition reinforced by the decline in the rate of sea-level rise after about 5,000 years B.P. The latter date is therefore critical in the development and morphodynamic behaviour of the contemporary barrier system if profile and planform has adjusted to sea-level rise to maintain a near constant nearshore water depth. Over time, the barrier system as a whole has switched from drift to swash alignment, thus reducing throughput of sediment moved by littoral transport, and increasing morphological stability.
- Approximately 20% of clasts in the modern barrier frame have a local derivation, although perhaps less than 10% have been transferred to it via littoral transport. Nonetheless, this input represents a potential for modest volume increase, unless abrasion, or some other loss, has offset it.
- Both the coarse clastic barrier beaches and lagoonal sediments may be regarded as sinks, although exchange between the immediate nearshore shoreface and inter-tidal beach qualifies as storage.
Subsequent to the above synthesis, research by Massey and co-authors has added further details and some modifications of understanding of inferred Holocene relative sea-level recovery and barrier evolution. This work (Massey, et al. 2006a, b and c); Massey and Taylor (2007) and Massey et al. (2008) employs a wide range of field and analytical techniques not previously applied to the Slapton barrier system, including electrical resistivity tomography, sediment coring to depths of between -14 and -3m OD, and the resolution of biostratigraphy using indicator foraminifera. Calculations of post-depositional autocompaction of sediments, particularly organic facies, is applied to the interpretation of the sediment stratigraphy inferring regional and local sea-level change. Early Holocene sea-level rise was in the order of 5m/Ka, during which stage the basal substrate underlying the present day barrier at depths of up to -14mOD was covered by peat and fine grained minerogenic sediments; these also infilled river channels and overlay adjacent marsh bordering shallow lagoons behind an ancestral barrier. Between 9,000 and 6,000 years BP sea-level rose from -17.0 to -6.2mOD, with a recovery of about 8m between 7,000 years BP and the present. Intertidal conditions persisted until approximately 4,400 years BP when relative sea-level stood at -5mOD. Since 3,000 years BP sea-level rise has been at a rate of about 1m each millennium. At this stage, i.e. commencing at about 4,500 years BP and continuing to 3,000 years BP the modern barrier was fully constructed and up to 10m of sediment choked pre-existing tidal passes, thus closing off back barrier lagoons from saline marine influences and initiating freshwater marsh environmental conditions. These then progressively infilled with organic clays and silts. Evidence from Blackpool Sands to the north of the Slapton barrier, where the basal surface is at-6m.OD and overlain by a sequence of clastic and organic sediments, relative sea-level rise was 13m. (+/- 5m) between 9,000 and 7,000 years BP and 8.0m (+/- 1m.) subsequently. Here, sea level has been recovering at a steady rate of approximately 1m each 1,000years since mid Holocene times. Terrestrialisation of the back barrier lagoon at Blackpool Sands commenced around 4,500 years BP.
5.2 Beach Morphodynamics
Most available information relates to components of the barrier beach system between Hallsands and Blackpool Sands, and is largely derived from academic research and work carried out in connection with the installation and upgrading of defence and protection works. There is comparatively little monitoring data, a fact that makes it difficult to present a coherent analysis of contemporary, recent and historical morphodynamic behaviour of individual beaches. However, collaborative academic work on beach morphodynamics by several researchers from the University of Plymouth has involved detailed repetitive surveys of beach planform, profiles, sediment composition and budgets for the main sectors of Slapton Sands for continuous periods of up to two years. Appropriate details are provided below and also in Section LT2, where they provide further direct and indirect insights into longshore transport rates and directions.
Hallsands
The catastrophic erosion of the Hallsands barrier beach after 1900 is, perhaps, the most notorious example from the British Isles of shoreline mismanagement (May, 2003). Documentary evidence of this beach prior to 1897 indicates that its backshore level was approximately coincident with the now fully exposed upper rock-cut platform at 6 to 8mOD - see Photo 4 (Mottershead, 1986; Tanner and Walsh, 1978 and 1984). It was sufficiently high and stable to protect Hallsands village from the 1891 storm surge event, which elsewhere along the South Devon coastline created severe beach drawdown (Job, 1993). An estimated 300,000m3 of gravel was dredged from the submerged inshore face of the barrier between 1897 and 1902 to provide aggregate for dock construction at Devonport. An alternative estimate of 1,600 tonnes per day over five years (1897-1903) giving a total of 1.8 million has also been made (see http://www.saveslaptoncoastroad.uk/history). Undermining of sea walls was recorded in 1900 and by 1903 beach levels had dropped between 4m and 6m and mean high water mark had moved landwards some 25-30m. All authorities presume a direct cause: effect relationship, and argue that dredging removed a nearshore sediment store that could not be naturally replenished. Beach material was drawndown into the large depressions created by gravel extraction. Job (1993), however, states that there is some evidence of falling beach levels after 1894, but this may have been a component of a natural cycle of fluctuation as there was some natural and gradual beach rebuilding north of the site of the village at least up to 1907. Contemporary concern forced the cessation of dredging operations in 1902, and between then and 1917 the village was progressively destroyed during a series of storm events culminating in severe losses in the storm of 26th January 1917 (Worth, 1904, 1909, 1924; Tanner and Walsh, 1978 and 1984; May (2003) and also see http://www.saveslaptoncoastroad.uk/history). This dramatic sequence of events demonstrates that this beach (and, by implication, others northwards to Strete) is not supplied by offshore to nearshore sediment transport - i.e. that the offshore sediment transport system of Start Bay is effectively decoupled from its marginal beaches. This realisation helped to generate the concept of Holocene barrier morphogenesis (Worth, 1909; Ward, 1922) and emphasised that the modern beach is effectively a "fossil" store. Events were thus a total contradiction of the assumption of the Admiralty (who commissioned the dredging operation) that there would be rapid recovery of any drawdown of beach levels and losses of volume.
Visual observations between 1920 and the early 1950s indicated apparent stability of most sectors of this severely diminished beach, which had been lowered by almost 2m during the 1917 storm. Robinson (1961) resurveyed the 6 profiles selected by Worth to record beach erosion (Worth, 1904; 1924). His results indicate an ongoing process of sediment depletion and lowering between 1917 and 1956, except in the area south of the village site. Here, beach levels had recovered to where they were in 1903. Over a one year period (1956-7), small magnitude seasonal fluctuation of levels was apparent, but localised net lowering occurred after several weeks of easterly winds and waves in 1960. Robinson (1961) ascribed this absence of recovery to the low potential for littoral transport due to (a) shallow angles of obliquity of wave approach along this frontage, and (b) the compartmentalisation of the beach between Start Point and Hallsands into discrete, small-scale, sub-cells lacking any mutual exchanges. Since the early 1960s, Job (1993) states that there has been further net loss of beach volumes, estimated by him to be approximately one-third of the quantity present in 1959. Posford Duvivier (1998) note that a marked phase of beach erosion commenced in 1996, accelerating the retreat of the cliff in front of the platform site of the abandoned village and exposing a rock-cut surface below it. This continues a trend apparent since at least 1905, as evidenced by the landwards movement of both mean high and mean low water marks (Robinson, 1961; Mottershead, 1986; Halcrow, 2002). The emergence, and isolation, of Wilson's Rock over this period provides strong visual proof, as prior to about 1901 it was concealed beneath beach deposits. Since at least the early 1960s, gravel on the foreshore has been confined to patches, with extensive exposure of fine to medium sand. The latter is presumed to represent the original beach foundation, as there is no apparent offshore to onshore or longshore supply of sandy sediments at present (Hydraulics Research Station, 1963).
Overall, it would appear that the beach in front of Hallsands maintained a quasi-equilibrium form from the 1920s to the early 1950s following the massive losses in the early years of the twentieth century. Thereafter, it has shown more fluctuation, with a net loss of volume up to the present. The beach between Hallsands and Start Point has been comparatively stable, having recovered from losses between 1897 and 1903 over the succeeding 30 or 40 years. Nearshore to onshore transfer is implied, possibly utilising some material lost further north or even mobilised by dredging.
Beesands
The beach at Beesands is another segmented barrier structure that confines a lagoon (Widdicombe Ley) due to damming the mouth of a small stream. Its natural composition is fine to medium grade gravel (median long axis of 8mm), fining slightly in a northwards direction (Hydraulics Research, 1991). Its morphodynamics are primarily controlled by refracted swell waves that approach this shoreline at high oblique angles.
Posford Duvivier (1998) and Halcrow, (2002) reported 30m of shoreward movement of mean low water since 1885, but over at least the past 100 years this beach system has shown considerable fluctuation in form and volume. Foreshore width has diminished since the late 1920s or mid-1930s (Job, 1993), although in general inter-tidal beach morphology was relatively constant up until the early 1970s. Since then beach lowering, narrowing and steepening has been the dominant trend (South West Water Authority, 1979), with seasonal fluctuations of up to 5m (Hydraulics Research, 1991; Posford Duvivier, 1998). This annual variability is especially evident at the southern end, where easterly and south-easterly winter storms have caused severe drawdown. This has necessitated a progressive series of coast protection measures for the village of Beesands, which is built on the back barrier beach (HR Wallingford, 1991; NRA, 1991). Rip-rap boulders placed in 1990 were subject to failure by undermining shortly after construction, and were replaced by a rock armour revetment and wave return wall in 1992. These structures are designed to inhibit the natural tendency of combined beach steepening and retreat; the latter is illustrated by 3m of lowering and 2m of crest cut back during a storm in February 1990. At Beesands Cellars, which remains partly unprotected, beach crest erosion was in the order of 10m per year between 1994 and 1997. This rate may have been affected by the seawall at Beesands village. The gabions and rock armour installed here by local residents in 2002 can only provide a temporary modification of longer-term beach behaviour (Hydraulics Research Station, 1963 and Posford Duvivier, 1998). The same must apply to the gabions protecting the barrier beach in front of Widdicombe Ley (Photo 12).
Torcross
Forming the extreme southern end of Slapton Sands (Photo 5), some 35m of retreat of mean low water level occurred between 1908 and 1979 (South West Water Authority, 1979). During this period, and continuing up to the early 1990s, beach steepening has been a consistent trend. Job (1993, 1994) states that width was at a maximum in the late 1930s. Overtopping and damage to the village occurred in December 1978 and January 1979 in association with easterly waves of up to 5.5m in height (Mottershead, 1986). Severe damage to the seawall (built in 1944) and village properties also occurred earlier, in 1951 and 1954, when easterly/south-easterly waves caused catastrophic beach drawdown. South West Water Authority (1979) tentatively ascribed beach steepening to the selective loss of fine material, a factor also identified by Hydraulics Research Station (1963). It would, of course, be a direct response to incident waves, although the reconstruction of the wave return wall and addition of a rock revetment (1980-1982) may have added to this effect, in more recent years, by promoting reflective scour. Since the mid-1990s, winter drawdown and steepening has been compensated by summer accretion, thus maintaining equilibrium. This would suggest sediment exchange between the inter-tidal beach and adjacent nearshore zone, as longshore transport from updrift sources is negligible. There have been episodic losses (subsequently recovered), such as 20m of recession of mean low water in the winter of 1982-1983 (Hydraulics Research, 1991). However, it is the conclusion of Chadwick et al. (2005) that there has been a modest net loss of beach sediment above mean high water at Torcross between 1972 and 2003, as well as overall net retreat of the barrier at this location since the beginning of the last century.
Slapton Sands
Apart from an embankment wall and recently introduced rock armouring near the Middle Car Park, this is an unprotected barrier beach that widens progressively northwards between Torcross in the south and Strete some 3.5km to the north. Over the past 180-200 years, overall morphodynamic stability has been punctuated by phases of erosion and crest retreat. Several shallow temporary breaches, or areas of significant crest lowering by overwashing have occurred (e.g. in 1917), but the most recent major breach appears to have been in 1824. Mean low water has slowly retreated, at a rate of approximately 0.8m per year from 1886 to 1996, over the central and southern segments, whilst the position of mean high water has retreated by up to 20m (Posford Duvivier, 1998; Halcrow, 2002). Slapton Ley is reported to have overflowed the barrier crest during most winters prior to the construction of the A379 in 1856, creating shallow depressions similar to wave-induced crest erosion by overwashing (Mercer, 1966). Regulation of the discharge of the Ley, at Torcross, has subsequently prevented this occurrence.
Severe erosion, resulting in increased profile concavity due to crest retreat and foreshore lowering has occurred in recent years in the vicinity of the Middle Car Park, particularly during the winters of 1995 and 2000-01 (Chell, 2002). In the latter event, notably between 11th and 12th January 2001, up to 5m of crest cut-back landwards occurred, under the influence of north-easterly waves, over a section length of nearly 1,000m. (refer to Chadwick et al. (2005) for an account of the hydrodynamic conditions of this event, which had a 1 in 25 year return period.) This undermined a portion of the A379, necessitating a three-month long road closure, and posing a risk of breaching. Less severe events had occurred previously at the same location in 1892 and 1953, but on this latter occasion the incipient breach was sealed by a combination of bulldozing and subsequent constructive wave action. Emergency protection was provided by some 3500 tonnes of rock armour placed along 250m of the worst affected road frontage. The undermined section of road was then re-routed some 21m inland and the new section of road was opened in February 2002. The emergency boulders and the undermined road were removed in summer 2002. In January 2003, two shingle bastions were constructed, one at each end of the segment of realigned road. Some 12,000 tonnes of shingle taken from the accreting beach at Strete Gate was delivered to form the bastions. The aim was to provide additional material to contribute to the beach crest should another episode of overwashing occur. The problem is that without intervention the beach would naturally migrate landward and eventually breach, severing the road and potentially altering the regime of Slapton Ley (Chadwick et al., 2005). However, the beach itself is an important geomorphological feature and it is argued that it should remain free to respond to wave action and rising sea-level. The authorities and bodies that have responsibilities associated with this issue have formed the Slapton Line Partnership to collectively determine the best and most sustainable long-term option(s) for dealing with future erosion, its effect on the road, the environment and the local community.(Scott Wilson, 2004). It is uncertain whether the central section of Slapton Sands has yet entered a phase of instability and imminent breakdown, or whether conditions could improve given prolonged north-easterly wave approach and southward drift. Certainly, without fresh gravel inputs landward recession, at an estimated current rate of 0.5m per year, and eventual breaching into the Ley are probable in the medium term. Pethick (2000) and Orford (2001) confirm that recent behaviour is consistent with barrier response to storm wave impact superimposed on rising sea-level - somewhere in the order of 0.8-1.0m retreat per 1mm of sea-level increase per annum. The specific cause of breaching is nearshore storm wave focusing during a storm surge event and is likely to be episodic. Chadwick et al. (2005) model the conditions that would promote a future breach event, informed by the impact of a major storm that occurred in October 2004; they calculate that they would have a recurrence interval of 1 in 25 years. Crestal low points may be either naturally infilled or persist according to the effectiveness of refracted constructive swell waves during intervening periods. The intervals available between storms for recovery are critical because potentially damaging overwashing will exploit areas of low crest. Given that this is a swash-aligned barrier structure, sediment exchange between the beach and the nearshore zone is critical to beach stability. Armouring will inhibit the natural tendency for landward translation, through the "rollover" process, but may create a form of outflanking beyond the terminal points of barrier strengthening where the beach would remain free to migrate landward (Orford, 2001). This, in turn, may create a sequence of discrete transport sub-cells where there was previous continuity and generate weaknesses that could be exploited as potential sites of breaching. Such understanding has led to removal of emergency armouring from the beach. Slapton Ley Field Studies Centre has carried out semi-routine monitoring of ten cross- sectional beach profiles between Strete Gate and Torcross since 1972. Unfortunately, this is not a continuous record, but analysis of profile change reveals that the beach foreshore over the central sector of Slapton Sands has been dynamic. Job (1994), analysing selected profiles for the period 1971-1986, notes a consistent trend for summer (swell wave) construction and winter drawdown. This suggests that a substantial exchange occurs between the foreshore and nearshore zones and that during this time there was an approximate adjustment of beach form to forcing factors. More detailed examination indicates a complex response of both advance and retreat of the basal element of the foreshore over periods of a few weeks to a few months. Profile shapes were consistently concave or convex-rectilinear, the latter normally characterising periods of recovery following previous drawdown. Surveys taken in August 2001 along the sector affected by crest retreat the previous winter showed responses varying from full post-storm recovery to the persistence of a well-defined upper crest scarp and low elevation foreshore. These surveys, and earlier work (e.g. Mercer, 1966) indicted main beach slopes varying between 6 degrees and 11 degrees but with berm segments up to 28 degrees (Doe, 1994). Chadwick et al. (2005) note that there has been substantial variability of beach width and crest height between 1972 and 2003, but with net accretion above mean high water at Strete Gate and net loss of beach volume at Torcross. Job (1993) argued that the beach changes between 1971 and 1986 justify the recognition of four forcing wind/wave environments, as follows:
- Offshore light winds generate low (flat) breakers that create a steep profile, high backshore berms and well-sorted gravel;
- South and south-easterly winds are associated with steep plunging breakers that are usually oblique to the shoreline; they produce profile drawdown and flattening (crest height reduction);
- East and north-easterly winds, which are normally at least gale force, create high steepness waves that result in substantial net beach accretion, overall profile flattening and loss of well-defined concavity;
- Easterly onshore winds, giving plunging breakers that promote net offshore transport and profile flattening with an absence of berm construction.
The above are, at best, generalisations that tend to be characteristic of the seasons during which the highest frequency of each type of prevailing winds and waves is experienced. Further quantitative insight into the dynamic relations between beach morphodynamics and storm duration; wave approach and wave dimensions is provided by Ruiz de Alegria-Arzaburu and Masselink (2010) and Ruiz de Alegria-Arzaburu et al. (2010)- refer to text for section LT2.This work derives from continuous monitoring of simultaneous morphological and hydrodynamic conditions. See also Chadwick et al. (2005) for further comments on post-1986 profile and hydrodynamic data acknowledged earlier in this section.
Both width and crest elevation of the entire Slapton barrier beach system declines southwards. At Pilchard Cove, mean crest elevation is 8.9m above mean HWM, reducing to 5.5m above mean HWM at Torcross. Width reduces from 110m to 83m between the same two points, although the maximum recorded width between average low water and the mean position of Slapton Ley shoreline, at the Monument Car Park, is 136m. These distances are considerably greater if measured from maximum low water of spring tides - some 450m in the central sector, for example. Chadwick et al. (2005), using a combination of sequential Ordnance Survey maps, air photo cover and Lidar imagery, state that the shoreline at Strete advanced almost 45m between 1890 and 2003 whilst conversely there was some 30m of retreat at Torcross over the same period of time. For the barrier beach as a whole, they calculate retreat of the crest during the twentieth century was at an average of 0.1m per year. Carr (1974) established that there was a north to south coarsening of median grain size, but Carr, Blackley and King (1982) were unable to discern any clear statistical correlations between beach height/width variation; particle shape and size and incident wave energy.
The latter paper reports on the degree of conformity of Slapton Sands with general gravel beach morphodynamic models. Beach cross sections, 200-300m apart, were resurveyed every month between September 1971 and September 1972. Results indicated overall morphological stability, although monthly profile "sets" were consistently either erosional or accretionary. Net annual volume changes (averaged for all sections) were slight, varying from +200m³ to -186m³. Several short-term but locally significant accretion events during winter months complicated the overall rather weak tendency for winter erosion and summer accretion. Winter period gains contributed approximately 60% more sediment gain than during summer season construction, and had a cumulative effect. By contrast, summer accretion was rapidly removed. The overall absence of obvious "model" conditions was put down to:
- Absence of bay planform equilibrium, with the two main confining headlands tending to cause a "piling up" of sediment at the northern and southern limits of the beach system (although the effect was modest because of low littoral transport rates).
- A steep beach gradient, more characteristic of coarse gravel rather than the coarse sand/granules/fine gravel that dominate the sediment frame.
- A wide range of wind and wave directions, preventing the establishment of fully stable morphodynamic conditions. This was indicated by greater variability of profile form, and probably volume, within the summer and winter seasons than between them. Incident wave energy is apparently more directly affected by the refraction effects created by offshore bathymetry than by wave fetch.
Further research indicating differential patterns of erosion and accretion in relation to forcing factors is reported in Austin and Masselink (2006); Ruiz de Alegria-Arzaburu and Masselink (2010) and Ruiz de Alegria-Arzaburu et al. (2010). This work is summarised in the preceding section (LT2) relating to littoral transport along the Slapton Barrier System.
However, Gleason, Blackley and Carr (1975) report a possibly significant correlation between wave However, Gleason, Blackley and Carr (1975) report a possibly significant correlation between wave approach direction and mean particle size for various positions on this beach.
The Slapton Barrier system has provided sites for a number of short-term experimental investigations into several aspects of gravel beach morphodynamics. Huntley and Bowen (1975a) observed distinctive sets of cusps generated by swash interaction across the steep intertidal beach face in the presence of edge waves with a longshore wavelength of 32m and a period twice that of incident waves. Horn and Li (2006) concluded that hydraulic conductivity, friction factors and the ratio of swash run-up and backwash sediment transport rates were critical factors controlling beach profile changes. As hydraulic conductivity increased so did upper beach face accretion, whereas an increase in friction reduced run-up elevation (swash infiltration) and thus berm development. In a related study, Austin and Masselink (2006a) carried out cross-shore measurements of swash depth and flow velocity together with groundwater elevation over the mid-point of the beachface that demonstrated that infiltration across the upper swash zone causes backwash discharge to be reduced by approximately 50%. The upper beachface remains unsaturated during each tidal cycle, allowing for accretion (i.e. dominant onshore sediment transport.) The same authors (Austin and Masselink, 2006b) also undertook short-term measurements of sediment transport in the swash zone in relation to changes in incident waves. Transport was confined to the swash zone when inshore wave height did not exceed 0.4m and wave period was less than 6 seconds; the intertidal beach face was steepened by net onshore movement of sediment during the flood stage of each tidal cycle and flattened by the dominance of offshore transport during the ebb. They observed that the presence of a beach step had the effect of preserving the reflectivity of the beach face when higher energy waves prevailed, and that under these conditions the depth of closure extended to a depth of at least 2m in the nearshore zone. As an average of conditions prevailing during the period of observation, the sediment load of the backwash was close to 0.6 of the swash load. A subsequent study (Austin et al., 2011) was concerned with the generation of alongshore currents in the swash zone. During the experimental period prevailing wave heights were between 0.3 and 1.0m and wave periods from 4 to 9 seconds. Under these conditions it was determined that alongshore flows were unidirectional and continuous and that swash flow was highly oblique and strongest in the upper swash. Infiltration reduced cross-shore backwash flow velocities. Austin and Buscombe (2008) located their research into the morpho-sedimentary evolution of the gravel beach step in relation to wave hydrodynamics, tidal cycles and sediment transport at sites on Slapton Sands. This feature is an accretionary form that is positively linked to tidal stage but with a relaxation time different to that of the main beach berm. It would appear to be initiated by a change in incident breaker type, from plunging to surging. Nearshore sediment transport is closely related to wave group frequency.
Strete Gate to Pilchard Cove
A wide beach fronting vegetated relic cliffs characterises this northern beach sector (Photo 8). Although a functional component of the Slapton Sands barrier, it has shown some distinctive behaviour. Hydraulics Research Station (1963) calculate 40m of seawards advance of Mean High Water, 1890-1953, indicating significant net accretion, in the vicinity of Strete Gate. This trend continued between 1972 and 1980 (Job, 1994) and 1983-1986. More recently, recession has been dominant, including 4m during the winter of 1982-1983 and 15m between 1986 and 1994 (Doe, 1994). It is uncertain if this is a function of changes in incident wave energy and/or net longshore sediment transport.
Blackpool Sands
Compared to Slapton Sands, Blackpool Sands (Photo 13) is a small headland-confined beach although it is probably a segmented residual of the original barrier that has been driven landwards (Massey, et al., 2008). Mean particle size is slightly smaller than for equivalent barrier beaches at Hallsands, Beesands and Slapton Sands. Hydraulics Research Station (1963) note that beach levels fluctuated considerably over the period 1890-1960, with a tendency for almost all sediment to be periodically removed either offshore or to the western end of the beach under both north and south-easterly storm waves. Losses were incurred between c. 1870 and 1908, when material was mined for road making; it is presumed that recovery has occurred since, suggesting a transfer from the nearshore to the beach. Merefield (1984) determined that 6% of beach sand here is composed of carbonate derived from communited shell debris. This may confirm the operation of this supply pathway, although most organic material is transported as suspended load.
Posford Duvivier (1998) tentatively suggest some ongoing loss of beach volume since the mid to late 1980s, necessitating the construction of a low sea wall, revetment and embankment as defence against localised erosion and flooding. Retreat of mean low water is calculated as 0.8 to 0.9m per year, 1886-1996, but has probably occurred episodically (Posford Duvivier, 1998). Major sediment losses, almost always during winter, are relatively rapidly recovered, but mid-point pivoting may have occurred in recent years. This is based on evidence of net recession in the north-east of the bay and net advance in the south-west. Perkins (1971) suggested that storms selectively transport sediment, with waves from the east/south-east moving both sand and gravel to the western end of the bay, and the reverse for high energy swell waves from the south-west.
Kelly's Cove to St Mary's Bay
Most of the enclosed "pocket" beaches along this indented coastline contain small quantities of fine to coarse sand, with a small component of granules. Posford Duvivier (1998) indicate that beach levels fluctuate seasonally, but there are no discernible trends of recent erosion or accretion. Hydraulics Research Station (1963) reported that there was retreat of mean low water at Long Sands, 1890-1937 but a small advance at Man Sands (Photo 14) up to 1960. There is undocumented knowledge that sand extraction occurred at both sites, for road making and industrial processing, for an unknown period in the early to mid-twentieth century.
5.3 Lagoons
The Start Bay Barrier Beach has confined several lagoons between its former, and present, positions and the degraded slopes of the Ipswichian interglacial cliffline (see Section 5.1 for summary details of Holocene evolution). The major example is Lower Slapton Ley, behind Slapton Sands (Photo 5); others include Widdicombe Ley, Beesands (Photo 12); the Upper (Higher) Ley at Slapton and a small coastal pond at Blackpool Sands. A former lagoon at North Hallsands-Greenstraight - has been infilled by natural alluviation; Slapton Upper Ley is a residual form of a formerly more extensive lagoon, also largely infilled by alluvial deposition.
All are currently sediment sinks occupied by lacustrine, alluvial and colluvial material overlying sequences of sediments whose stratigraphy provides details on the more recent history of barrier evolution (Section 5.1). The physical and biological environment of Lower Slapton Ley was described in detail in a sequence of papers in the 1960s and 70s, of which van Vlymen (1979) is the most relevant in the present context. Using data for 1973-1977, he examines lagoon hydrology and the contemporary water balance. Mean water depth is 1.8m (but can rise to 2.8m under high discharge conditions of inflowing streams) and hydraulic retention is 18 days. Hydrology is flow-dominated, i.e. in response to stream hydrographs, and therefore indirectly to catchment rainfall. The outfall is an artificial culvert at Torcross, first constructed in 1856 when the A379 was built. As this is commonly blocked by gravel, it maintains the lagoon at an artificially high level, above mean sea-level; it is normally opened during high discharge events, usually in autumn and winter. Thus, seepage through the barrier of Slapton Sands is the main output - indeed, the only significant source of water loss for over half of the year, excepting evaporation. Seepage can be observed on the lower seaward (foreshore) slope of Slapton Sands during low water springs. Lateral seepage is promoted by impermeable lake floor sediments, and occurs at a maximum rate in mid-autumn immediately prior to the opening of the outflow culvert. It is at a minimum in mid-late summer, due to the reduced stream discharge. Van Vlymen (1979) estimated maximum daily loss via seepage to be 0.030 x 10⁶ m³ day-¹. The seawards hydraulic gradient of the free (unconfined) aquifer of the gravel barrier is maintained throughout most tidal cycles because the lakebed is above mean sea-level.
Morey (1976) states that water levels have been rising faster than rates of sedimentation, thus increasing the mean depth. Approximately 85% of the surface area of Slapton Lower Ley is open, with its fringes occupied by Phragmites reeds. Morey (1976, 1980) states that the latter is retreating, but the cause is uncertain.
The lagoon is thought to be vulnerable to breaching due to the finite volume of the protecting Slapton Sands Barrier. Such an event would cause major changes in the lagoon increasing its salinity and potentially leading to tidal scour of a permanent inlet channel through the beach. The possible occurrence of breaching and its effects were discussed briefly by a scoping study by WS Atkins (2002) and are elaborated in the report by Scott Wilson (2004).
It is probable that the current imbalance between water storage and sedimentation has not prevailed for more than a few hundred years; Morey (1976) suggests that the present outline of Lower Slapton Ley cannot be older than about 1,000 years. In earlier millennia, human land use exploitation accelerated rates of erosion of catchment slopes. This may largely account for the virtual elimination of open water in Upper Slapton Ley, whose former extent is now occupied by lacustrine, alluvial and colluvial sediments. Interbedded gravels may be barrier washover fans. The former lagoon at North Hallsands, due to natural barrier damming of the Bickerton stream, has also been infilled by washover and alluvial sediments. However, in this case, there may have been a contribution from human agency. Widdicombe Ley, Beesands, continues to persist and was evidently a larger lagoon basin in previous centuries. There are no reliable records of its hydrology and history, but it has been stable in depth and shape in recent decades.
5.4 River Dart Estuary
The River Dart is a long-established, pre-Quaternary, drainage system, whose erosional history is represented by a sequence of partial planation surfaces. Those below approximately 230m OD may be of marine origin (Brunsden, 1963). There is apparent conformity between raised beach and sub-aerial terrace surfaces at 24m OD at Blackstone Point. The alignment of the modern estuary (Photo 1) is discordant to the geological strike, with no evident relationship to structural axes. Green (1949) and Brunsden (1963) inconclusively discuss the possibility that an ancestral River Dart discharged via an exit at, or in the vicinity of, Brixham. High concentrations of orthoclase feldspar mineral grains occur in deposits in the northern part of Start Bay (Hails, 1974, 1975; Hails, et al., 1975), which probably derive from, the Dartmoor granite, drained by the upper Dart and its tributaries. Samples of bedload from the modern river channel and its estuary are also rich in detrital feldspar (Hails, 1974). Thus, there is a high probability of a long-term, substantial contribution of sediment output from the Dart to the Start Bay offshore sink. This would have functioned throughout the several Quaternary stages of sea-level fluctuation and climate change, in particular when low sea-levels coincided with intense periglacial weathering and mass movement supplying high magnitude summer season river discharge and load transport. Hails (1975) emphasises that this source of supply has greatly diminished since mid-Holocene times, as sediment is now trapped in the deep and narrow drowned ria-type estuary created by rapid early Holocene sea-level transgression. The latter would have invaded via the constricted, bedrock entrenched (now infilled and buried) channel known to reach a maximum depth of -41.2mOD at the present entrance to the Dart (Durrance, 1974). This is the most deeply incised palaeochannel offshore the modern South Devon coastline; it is abruptly terminated by a submerged cliffline at approximately -44mOD, partly concealed by scree-like debris.
The hard, erosionally resistant rocky headlands that define the entrance to the Dart estuary create stability of form and together with a relative absence of coarse sediment around the entrance exclude any interaction between the estuarine and littoral sediment environments. The estuary penetrates over 20km inland, to Totnes, within a steep sided valley system that has several tributary tidal creeks, e.g. Old Mill, Galmpton, Parsons and Bow Creeks. There is, therefore, a substantial capacity for storage of sediment and the apparent immobility of this material warrants this store being regarded as a sink.
Over half of the estuary, which has an area of 860 ha, is subtidal. Sand and mudflats occur at the base of its steeply sloping, wooded, margins and in creek re-entrants upstream of Dartmouth. Saltmarsh occupies an area of less than 30 ha in the upper estuary and creek heads, being only marginally greater than the extent of sheltered rocky shoreline (Horsman, 1986; Moore, 1988; Joint Nature Conservancy Council, 1993). There are no records of any significant sediment inputs from marsh or tidal flat erosion or slope mass movement. Hard defences, inhibiting erosion, protect the developed frontages of Kingswear and Dartmouth. In the absence of adequate baseline surveys, the geomorphological evidence points to slow, but cumulative, accretion in marginal areas of shallow water. However, low cliffs at the edge of some areas of saltmarsh occur upstream from Stoke Gabriel, indicating the possibility of past erosion. These features appear to have been stable since at least the early 1970s (Moore, 1988; Joint Nature Conservancy Council, 1993). A small, enclosed, shingle beach has accumulated at Dittisham, immediately downstream of a major meander of the Dart channel. This might be, in part, a flood deposit.
The Futurecoast project compiled a series of quantitative data and indices of estuary characteristics (Halcrow, 2002). Its results indicated that the Dart cross sectional area/tidal prism ratio is high and typical of ria type systems where there is a lack of sediment input. The intertidal area ratio is low at 0.36, suggesting that there is a high capacity for further sediment accumulation, should the material be available. It has an ebb dominant tidal flow, with a flow ratio that suggests that a plume may be present at ebb tides on maximum river discharges. The estuary is likely to be partially mixed at low flows, becoming highly stratified at higher flows. It is likely to be a strong sink for fine sediment, and for coarse sediment, if there is any available.
No data relating to discharge below Totnes is available, but maximum river discharges affect flow ratios and create a flood risk zone along the Dartmouth frontage. Aerial photography has revealed a small plume of suspended sediment at the estuary mouth coincident with high discharges and maximum ebb current velocities. Some of this material may contribute to sedimentation within Start Bay, but it is probably of little quantitative significance. There are no records of significant ebb tidal delta formation or tidal and wave driven coarse sediment circulations at the entrance.
5.5 Offshore Morphology, Sedimentation and Sediment Transport
Introduction
The morphology and sedimentology of Start Bay has been described, analysed and partially interpreted in a series of papers, the major ones being Robinson (1961); Hails (1974, 1975); Hails, Kelland and Lees (1975); Kelland (1975) and Kelland and Hails (1972). These sources contain overlapping and repetitive detail; this section attempts to identify the major points.
The seabed morphology of Start Bay is mostly a gently sloping rock shelf, some 3km in width and between 3 and 45m below mean sea-level. Comparatively steep slopes occur in the nearshore area seawards of the limit of beach deposits. Its outer limit is defined by a marked break of slope at approximately -42m OD, orientated roughly parallel to the modern shoreline. This landform, which is closer inshore between Scabbacombe Head and Blackpool Sands, is cut into bedrock and has been interpreted as a relict Pleistocene cliffline. It tends to be concealed in the vicinity of the mouth of the Dart, but elsewhere may be kept clear by tidal scour. It passes beneath the major accretion feature of Start Bay, the Skerries Bank, which has a crudely quasi-parabola shape and a north-east to south-west axis that follows the line of the shelf edge. This well-defined submarine landform occupies the south-central area of the bay and approaches closest to the coastline between Torcross and Start Point. Buried channels, incised into bedrock, extend the courses of rivers and streams that formerly discharged at the early Holocene coastline but now drain via the several lagoons and wetlands created by the emplacement of the Hallsands-Blackpool Sands barrier beach system. Most buried channels are infilled with sediment and are only minor modern seabed features. Exceptions are (i) the buried channel of the River Dart, incised to depths between 35 and 40m, and only partially occupied by sediment, and (ii) the former seaward extension of the Slapton stream, which has created an abruptly truncated "tongue" of deeper water offshore Torcross.
Sediment cover is thinnest in the area between the inner boundary of the Skerries Bank and the shoreline. Bare rock is exposed in a number of zones that occur immediately seawards of the main headlands and rock platforms. Much of the sediment is sand, but it lacks a uniform size distribution. There are several areas of apparent admixture of both fine and coarse sands, but coarsening is apparent towards the seawards edge of the Skerries. Gravel is confined to a few well-defined strips on the inner shelf, but it becomes more widespread below depths of 30-35m. Small areas of sandy muds extend out from the coast at Blackpool Sands and the approximate position of the Middle Car Park of Slapton Sands. Both may crudely trace the alignment of buried channels. The area close to the present mouth of the River Dart is dominated by silty clays and muds (Robinson, 1961).
Hails (1974, 1975); Hails, et al. (1975) and Kelland (1975) report the results of sediment sampling and analysis over a large area of the seabed of Start Bay. They propose a classification into Barrier, Bay, Bank and Buried Channel deposits based on sediment texture, thickness, sub-surface lithostratigraphy and inferred aspects of genesis. In more detail, the main characteristics of each type are:
- Barrier sediments. These occur seawards of the modern barrier beaches to approximately 2-500m offshore. Thickness varies between 5m and 8m, and in places they have been deposited directly over eroded bedrock surfaces (e.g. offshore Beesands). These sediments are dominated by fine to medium sized gravel, but with some interbedded gravelly sand and sand horizons. Lithological composition is very similar to the modern inter-tidal barrier beaches, with 80-85% of clasts consisting of quartz and flint. Mean particle shape varies between well rounded (quartz) and sub-angular (flint). In almost all respects this deposit is virtually indistinguishable from modern barrier beach composition, and is therefore interpreted as relict barrier material, although much may potentially be active in energetic conditions. This conclusion is supported by their morphological arrangement into a set of terrace-like forms. An unknown proportion of material may be involved in beach-nearshore exchange under the contemporary hydrodynamic regime.
- Bay sediments. Composed predominantly of medium to fine grained sands, but with spatially variable proportions of clay, silt, whole and broken shells and gravel, these deposits attain a known maximum thickness of 28m, some 2km south-east of Blackpool Sands. In places, however, they are little more than 1m in thickness. Hails (1975) and Kelland (1975) suggest a subdivision into Upper and Lower units, based on lithostratigraphy, texture and heavy mineral content. The Upper unit is a homogeneous deposit of fine silt to medium sand with lenses of shelly material; it contains glauconite and chlorite grains, indicating probable derivation from the Dart catchment and adjacent Permo-Triassic seabed outcrops. In places, carbonate (shell) content is between 10 and 25% of total composition. The Lower unit, which passes beneath the Skerries Bank, is an intercalated and stratified sequence of estuarine silts, clays and sands, with some gravelly horizons. There is a discernible angular nonconformity between the two units, suggesting a depositional hiatus. The stratigraphy of the Lower Unit is indicative of marine transgression, with the gravel deposits comprising possible relict barrier bases and washover fans representing earlier positions of the migrating barrier.
- Bank deposits. These comprise the bulk of the Skerries Bank sediments, and are difficult to distinguish from Bay sediments in the zone of transition from one type to the other. They unconformably overlie Bay deposits, and have a predominantly coarse shelly sand composition in uppermost layers, underlain by silty, fine/medium sands. Well-defined gravel horizons (at -17, -21 and -30m OD) are strong indicators of the positions of relict barrier beaches within the timeframe of Holocene sea-level rise.
- Buried Channel deposits. These sediments, often partially concealed by later Bay deposits, are mostly gravels, set in a coarse sandy matrix, and contained in well-defined channels 100 to 450m in width. The infill is normally clearly stratified, with variable sequences of silty and shelly sand; gravel and laminated clays. There is insufficient data to provide scope for a full interpretation, but the available stratigraphy suggests relative land: sea level fluctuations superimposed on an overall marine transgression (Kelland, 1975). This has been confirmed by study of benthic foraminfera and macrofauna (Hails et al., 1975). The deposits seawards of central Slapton Sands indicate the possibility of one or two ancestral lagoon/barrier beach systems. This complex pattern of sediment succession and distribution has provided scope for the interpretation of mid to late Holocene sea-level history and environmental change. This is examined further in Section 5.1
This is a classic banner bank whose planform and morphology has been remarkably stable since the early nineteenth century (Robinson, 1961). In this respect it is distinct from the majority of similar features offshore the British coastline, which have tended to migrate landwards over this period. It rests on the outer margin of the gently sloping shelf of central-southern Start Bay at depths of between -11 to -15mOD. Its maximum height is -4.8mOD at mean spring low water, but is an average of -7.5m to -9.0mOD. Its boundaries are relatively well defined, and it is separated from Start Point by a channel that becomes infilled further north. The main axis of the Skerries is approximately north-east to south-west, with approximately the same orientation as the outer shelf break (Hails, 1975).
Robinson (1961) analysed Admiralty hydrographic charts for 1825, 1853, 1921 and 1951 and concluded that there had been very little change in planform over this time period, except at its extremities. The northern tip moved some 50m landwards, 1825-1921, but showed no subsequent change, whilst the southern portion migrated 180 to 230m towards Start Point, over the period from 1825 to 1951. Although there has not been more recent research on post-1951 change, it is apparent that the Skerries Bank is very stable. Analysis of available bathymetry data (collected by the Maritime and Coastguard Agency’s Civil Hydrography Programme) indicates no evidence that material is moving from the inshore boundary of the bank towards the coastline. It would also appear that the Skerries is not receiving sediment inputs from elsewhere in Start Bay, or from further offshore. It would therefore seem that this is a virtually closed sediment transport system. Given that the predominantly sandy and shelly materials of the bank attain a maximum thickness of 18m (Hails, 1974), it is difficult to conclude other than that this large accumulation of sediment is related to previous morphogenetic and sediment transport systems.
Acton and Dyer (1975) argued that if present sea-level were to be lowered by 6m, part of the Skerries would be emergent and act as a linear breakwater to swell waves; these currently are refracted by the bank and arrive at the Start Bay coastline in attenuated form. Thus, the Skerries must have been initiated as a submergent feature seaward of the position of the present day barrier-beach system, giving it a minimum age of 3,000 years. Without its presence, barrier beaches would have migrated more quickly and would probably have made contact with the entire interglacial cliffline. This argument can be applied to relict barriers deduced from gravel horizons in Bay deposits (Hails, 1975, Hails, et al., 1975). Given their positions several hundred meters seaward of the present coastline, and at depths (i.e. at previous lower sea-levels) of up to -30m, the apparent age of the Skerries can perhaps be extended back to the early Holocene. It is possibly older than that, but conclusive evidence is lacking. A minimum age of not less than 6,000-8,000 years is suggested by a radiocarbon dated peat horizon at -1.26m below the bank crest (Lees, 1975). This dates to approximately 8100 ka BP, but erosional scour may have removed overlying younger material. Kelland and Hails (1972) state that the Skerries has always been a submergent form, albeit at shallower depth than at present. This is based on the observation that, in places, carbonate content (broken and whole shells) constitutes 30-35% of total volume. The majority of species represented are likely to have died in situ, rather than being transported from elsewhere in the western English Channel.
Despite stability of plan shape, there is evidence of sediment mobility on the summit and at the margins of the Skerries Bank. Robinson (1961) observed that almost all of the material composing this feature can be potentially mobilised by local tidal current velocities, and quotes supporting evidence of detailed changes in crest shape, as well as long-term migration over the past 180 years. Further indirect evidence of sediment mobility comes from analysis of sediment textures of Upper Bay deposits; these suggest selective winnowing by either or both tidal or wave-induced currents.
Although residual current velocities over the main bank crest are low (0.25-0.30ms-¹) both flood and ebb streams have sufficient transport capacity to move sand (Acton and Dyer, 1975). Robinson (1961) presented an explanatory model which identifies an anti-clockwise tidal current circulation dominated by a south/south-west directed ebb current (T1) and north/north-east moving flood current (T2). The ebb is stronger and has a longer duration, which - together with the "average" trajectory of both currents - is used as an explanation of the position and shape of the Skerries Bank. It is presumed that this seabed transport system by tidal streams is effectively closed with no potential for shorewards movement. Robinson (1961) quotes evidence of sand wave asymmetry on the inner margin of the bank to confirm the operation of the ebb tidal stream. He concedes that waves from the south-east or north-east shoaling on the Skerries at or close to low water might entrain sand, but actual movement is by tidal currents. The boundaries of the bank are therefore 'trimmed' and maintained by tidal steams, with the more dominant ebb moving material southwards and eastwards. It is uncertain if the ebb and flood channels inferred by Robinson (1961) are established, stable features for repeat studies have not been undertaken.
Robinson (1961) also implied that there is an auxiliary tidal flow off Start Point that intensifies seaward movement of sand, and which may account for the southern boundary of the Skerries Bank. This was subsequently investigated, using numerical modelling, by Pingree and Maddock (1983). They observe that during peak flood tidal current streaming to the north-east, centrifugal effects and Coriolis force act outwards from Start Point. Sediment already in suspension will therefore be moved towards the Skerries, although this vector is offset (but not reversed) during the succeeding ebb tide. A rotational vortex is generated in the lee of Start Point, which is anti-clockwise when tidal flow is to the north-east. This gyre is generated offshore of Hallsands, resulting in southerly flow between the shoreline and the Skerries Bank, whilst tidal streaming is north-eastwards seawards of Start Point. With reversal of tidal oscillation within the English Channel, flow between Hallsands and the Skerries is still southwards, although the eddy-like motion weakens. Thus, south-directed tidal flow occurs for 10 hours during the tidal cycle, with strong residual flow directed towards, and then away from, Start Point. Bottom stresses ensure no deposition of sediment adjacent to Start Point, but their reduction offshore is not in itself an explanation of the existence of the Skerries Bank. In fact, Pingree and Maddock (1983) state that the distribution of seabed stresses exerted by tidal currents on the southern boundary of the bank are largely a product of its existence rather than an explanation of its formation. They are the outcome of factors such as water depth and seabed roughness affecting frictional forces.
Seawards of about 3km from the Start Bay shoreline, sediment cover thins and a larger area is occupied by a uniformly level (1:400) slope developed in bedrock. This resembles an abrasion surface, as it cuts across a series of folded strata (Kelland, 1975). Dyer (1980) undertook a survey of submarine bedforms in this area, some 1,000m seawards of the outer edge of the Skerries Bank. Rippled sand was mapped, with mean wave heights of 3cm and wavelengths of 20cm. Velocity profiles less than 2m above the seabed were measured to determine sediment entrainment thresholds; these were found to be directly influenced by seabed roughness, but changed during the tidal cycle. Suspended sediment concentrations varied in proportion to excess shear strength. This work therefore identified the presence of a "carpet" of suspended grains up to 10cm above the seabed, but much of this sediment was smaller in size than the grains over which it was moving. This suggested non-local derivation, a point clearly applicable to the sub-category of broken, fine shell debris. There was no clear evidence of any significant movement towards the Skerries Bank.
6. Sediment Budget
Most of the primary research on the Start Bay shoreline (e.g. Robinson, 1961; Hails, 1974, 1975; Hails, et al., 1975; Job, 1987, 1993b, 1994; Kelland, et al., 1972; Morey, 1980,1983; May, 2003; Chadwick et al., 2005) regard both the littoral and offshore sediment transport systems as effectively closed. The first is wave-driven and tidal currents dominate the second. Although little is known about their possible linkages, the offshore system maintains the Skerries Bank that refracts and dissipates waves, providing an element of shelter to the entire Start Bay shoreline.
Littoral System
Littoral transport is confined by the absolute transport barriers of Start Point and Berry Head and is compartmentalised into a sequence of sub-cells defined by rock salients and the entrance of the River Dart estuary. Virtually all mobile sediment is coarse material and is inherited, having been introduced by barrier beach migration during the Holocene. As was first made apparent by the lack of recovery of Hallsands beach in the early twentieth century following massive removal of nearshore gravel deposits, this is a 'fossil' store that cannot be naturally replenished under present conditions. The lithological composition of beach clasts, with over 80% from non-local sources, is further proof of this. The overall littoral transport budget is therefore a function of the recycling of a finite reserve of material. The net pathway of longshore movement is currently northwards, but the rate of transport in this direction is low and the net volume small in comparison to the gross (cumulative) transfers, both northwards and southwards, that occur during an average year characterised by frequent changes in storm intensity and duration and thus the approach direction and energy of incident waves. If a unidirectional drift system were to be maintained, the littoral sediment budget of the southern half of the bay would be strongly negative, as opposed to its currently approximately balanced or stable condition. Thus, the present net northward drift may not be typical of the long term and switches between decadal intervals of weak northward and southward net drift might be anticipated in future. Cross-shore variations in beach morphology, both regular (seasonal) and episodic, imply sediment exchange between inter-tidal beaches and their nearshore shorefaces. There must be a presumption that absolute losses affecting the littoral system (e.g. from clast abrasion and occasional removal to deep water under storm conditions) are compensated by some gains from local cliff and shore platform erosion. It is important to note that the detailed behaviour and volume changes of each of the component sub-cells of the Start Bay barrier system have not been quantified so that only the broad conceptual budget model is understood to date.
Offshore System
The large-scale anticlockwise circulation of fine-grained sediments around, and within, the Skerries Bank by tidal currents is considered also to be virtually closed. The only inputs might come from offshore-directed tidal currents set up by the Start Point gyre and from sand deposits further offshore. The latter source, however, has yet to be demonstrated, as residual currents have pathways that run parallel to the Skerries Bank. This banner bank is therefore best understood as a large sink of sand that is currently maintained in a condition of dynamic equilibrium and characterised by a "steady state" budget. No exchange of sediment between the Skerries and the nearshore zone has been demonstrated, though this apparent decoupling of littoral and offshore transport and budgetary systems requires long-term monitoring before it can be confirmed. Meanwhile, the working assumption must be that large quantities of both coarse and fine sediment are subject to constant re-working and re-circulation in spatially discrete transport systems. The offshore system would appear to be critical to the maintenance of the littoral system, although the latter does not influence the offshore system.
7. Opportunities for Calculation and Testing of Littoral Drift Volumes
Data collected by the Defra-funded National Network of Regional Coastal Monitoring Programmes is pivotal for future improvement in estimating beach change. The Programmes consist of topographic beach surveys, nearshore bathymetry, aerial photography, lidar, coastal hydrodynamics (waves and tides) and terrestrial habitat mapping. Specifications for data collection are consistent for all regional programmes and the data and analysis reports are made freely available under the Open Government Licence from www.coastalmonitoring.org.
The Southwest Regional Coastal Monitoring Programme commenced in 2006. The Lead Authority is Teignbridge District Council, with data collection, analysis and reporting led by a specialist team at Plymouth Coastal Observatory (PCO). Although at present a relatively short-term time series of data has been collected (~ 6 years), longer term Coastal Monitoring Programme data, when combined with other data sets, academic research and historical studies may enable sediment budgets, transport rates and directions to be identified and/or validated in the future.
Estimates of gross and net littoral drift potential derived from numerical modelling based upon wave hindcasting are available at several points along the shoreline from the Durlston Head to Rame Head SMP2 (SDDCAG, 2011). However, these rates remain relatively uncertain because they have not been validated against reliable long estimates of beach volume changes. Uncertainties encountered in applying numerical model studies include:
- Imprecision in the selection of synthetic wave climates in the absence of field validation of inshore waves. The bathymetry of the Skerries Bank in combination with the tendency for significant wave energy to approach from several different direction sectors introduces complexity in the waves actually experienced at the shore;
- The problem of selecting a representative sediment gain size on the mixed sand and fine gravel beaches (sediment mobility is highly sensitive to grain size);
- Uncertainty relating to the extent of bypassing of the rocky headlands of Tinsley Head and Limpet Rocks and their shore platforms that undoubtedly affect transport.
- High uncertainty in estimates of exchanges of gravel between the intertidal barrier beaches and their subtidal shorefaces.
Opportunities are available for testing of littoral drift volumes by means of a thorough examination of the budget of beach sediments, especially that which accumulates against headland obstructions. For this to be feasible, it is important that beach volumes should be monitored and historical volumes are reconstructed (e.g. using existing historical measured profiles, perhaps supplemented by photogrammetrically derived data from historical air photos dating back to the 1940s).
A potentially useful approach with which to investigate some of the uncertainties might be to undertake detailed sediment budget analysis of each of the three main barrier beach sub-cells (i) Hallsands to Tinsley Head (ii) Tinsley Head to Limpet Rocks and (iii) Limpet Rocks to Pilchard Cove. In each case, a gravel barrier occupies a partially enclosed embayment, where it may be possible to infer minimum rates of net drift based on analyses of beach volume change. Careful study of headland bypassing would be required in order to define the outputs from one segment that become inputs to its neighbour. By testing alternative input and output scenarios the recorded patterns of historical beach behaviour could be accounted for. Provision of realistic estimates of sediment inputs and losses would assist interpretations of drift using volume changes and aid the setting up of numerical models of drift. It could reduce reliance upon non-measured assumptions of offshore loss/gain previously employed as a means of explaining the lack of fit between modelled and observed changes in beach morphology.
8. Knowledge Limitations and Monitoring Requirements
Analysis of Coastal Monitoring Programme data between 2006 and 2012/13, combined with other datasets, academic research and historical studies has enabled sediment budgets, transport rates and directions to be identified or verified. However, at certain sites either due to a lack of long-term data, data coverage or sedimentological information (e.g. composition and proportion of beach grade material arising from cliff erosion), quantification of sediment transport rates of gravel and sand has not been possible.
Although this coastline is frequently cited in the national and international literature on coastal geomorphology, understanding is essentially conceptual. This relates, in particular, to the operation of apparently closed littoral and offshore sediment transport cells. This, of course, is illustrated by the causes of the destruction of Hallsands village nearly a century ago. Notwithstanding results from the Southwest Regional Coastal Monitoring Programme, a programme of inter-disciplinary research that focused on the sea bed sediments of Start Bay, in the 1970s, provided some valuable data (Hails et al., 1975) and the Durlston Head to Rame Head SMP2 (SDDCAG, 2011) summarised existing knowledge but there remain several uncertainties concerning the process regime, requiring focus concentrated on the following issues: Quantitative assessment of the wave climate at a series of inshore points along the Start Bay barrier beach. It ideally requires a representative long-term hindcast offshore wave climate based on some 20-30 years of wind data, together with inshore field validation of model studies of effects of refraction and diffraction on waves approaching from different directions. A magnitude-frequency analysis should also be linked to a quantitative study of the recurrence probabilities of extreme water levels. This is considered important for it is storm waves and storm tidal surges in combination that will define barrier overtopping and overwashing criteria.
Recent and contemporary morphodynamic behaviour of the Start Bay barrier beach system, in particular Slapton Sands between Torcross and Strete Gate. Careful study of historical records may help to determine if this beach is moving from stability to instability, and the extent to which recent events of foreshore lowering and crest cut-back are likely to be repeated in the near future. Existing information would need to be supplemented by photogrammetrically derived data on crest positions and beach volumes from historical air photos.
- The gross rates of longshore sediment transport and the budgets of beach sediments are not understood quantitatively, as there is inadequate knowledge of the frequency, duration and magnitudes of reversals of drift direction. Future work should involve analyses of beach budgets and volume changes drawing upon historical analyses and profile monitoring Uncertainties may need to be resolved by sediment tracing, preferably over intervals representing the full range of wave energy conditions. This data would be invaluable in the verification of numerical modelling approaches, and therefore needs to be undertaken alongside simultaneous measurement of inshore waves in the field. To gain accurate knowledge of bypassing between the beach sub-cells, tracer investigations may need to be undertaken at Tinsley Head and Limpet Rocks.
- To understand beach profile changes it is important to have knowledge of the beach sedimentology (grain size and sorting). These can alter significantly along this frontage due to the combination of cross-shore and longshore transport and could also be affected by beach management. Ideally, a one-off field-sampling programme is required to provide baseline quantitative information along this shoreline together with a provision for periodic re-sampling to determine longer-term variability. Results could be compared with earlier sampling reported by Carr (1974), Gleason et al. (1975) and Carr et al. (1982). Such data would also be of great value for future modelling of sediment transport, for uncertainty relating to grain size is often a key constraint in undertaking modelling.
LITERATURE REVIEWPHOTOSMAP
