East Head to Pagham Harbour, West Sussex

1. Introduction

The coastal zone of the Selsey peninsula is an exceptionally complex environment, not least because the well-defined headland of Selsey Bill separates shorelines with different orientations (Photo1). Because of spatial variation in wave climate, and the effects of both planshape and submerged relief on the local tidal current system, the apex of the peninsula functions as a regionally significant boundary between adjacent sediment transport cells. The presence of offshore and nearshore banks, bars, shoals and reefs adds unusual complications to the sediment budgets of each of the several distinct littoral transport sub-systems. Exceptionally rapid erosion over at least the last five millennia has resulted in the submergence of both natural and human-modified coastal landscapes. This legacy has not been fully explored, but has generated considerable speculation over the sequence of coastal evolutionary changes.

At the shoreline, a partially swash aligned shingle storm ridge and sandy lower foreshore extends the length of Bracklesham Bay to the Chichester Harbour Inlet. The eastern side of the Selsey peninsula is fronted by a drift aligned gravel beach. The hinterland is low-lying, but elevated slightly at East Wittering and Selsey. At Medmerry, the hinterland is close to or below mean sea level and is formed of soft alluvial deposits comprising a reclaimed estuary channel. A weak to moderate net shoreline drift transports sediments from the east to west, although actual drift of shingle is presently very low due to the widespread controlling effects of groynes.

It is only within the last 50 years that the majority of this coastline has been protected by formal defences and regulated by other shoreline management practices. Artificial control of beach volumes and sediment transport pathways has not succeeded in achieving conditions of shoreline stability at all points; indeed, there are several critical locations where, in future, it may be necessary to allow for natural shoreline behavioural tendencies and relax management controls (Posford Duvivier, 2001; HR Wallingford, 1995, 1997; Cobbold and Santema, 2001).

1.1 Coastal Evolution - Map

The Selsey coastline is developed in Eocene (principally Bracklesham Group) sandstones and clays, overlain by Quaternary drift deposits. The former provide the substrate beneath the inter-tidal foreshore and are highly erodible; prior to the construction of comprehensive "hard" defences, coastline recession rates were up to 8ma^-1 in places. Last Interglacial (Ipswichian stage) Raised Beach deposits (Photo 2) overlie earlier Quaternary deposits (Reid, 1892; West and Sparks, 1960; West, et al. 1984 and Bates, 1998 and 2000), but their formerly extensive exposure at the shoreline is now restricted to a few localities. Late Devensian or early Holocene loamy silt ('Brickearth') overlies the Raised Beach and provides the substrate to modern soil profiles. These deposits overlie the most recent of a sequence of marine erosional platforms that extend 25km inland. They have been interpreted as the product of successive Middle Pleistocene sea-level transgressions, punctuated by regressive stages and subsequently displaced by neotectonic movements (Bates, 1998, 2000; Hodgson, 1964; Bates, Parfitt and Roberts, 1997). Further detail is contained within the separate Section on the Quaternary History of the Solent.

During the last (Devensian) cold stage, sea level was at least –50 to –60mOD, and the regional shoreline some 5-7km seawards of its present position. Subsequent Holocene sea- level rise has therefore released a substantial quantity of sediment, including gravels derived from the ancestral Coastal Plain composed of Raised Beach, Coombe Rock (periglacial) and River Solent fluviatile (terrace) materials. Some of this continues to be available as scattered deposits on the seabed, but it is thought that much lies stranded offshore within submerged barrier beaches built during several stages of mid to late Holocene sea-level transgression. Wallace (1990 and 1996) has tentatively identified the foundations of several indurated barrier structures, some of which may have originally been independent barrier islands, separated by tidal passes. Others may have become "anchored" to the predecessors of modern reefs, such as the Mixon, and once extended as far eastwards as Bognor. Strongly indurated, stable cobble " pavements" and large boulders of local Eocene and Oligocene rocks located in water depths in excess of 8m offshore the west and south-west coastlines might be the foundations of barriers that were submerged during one or more stages of rapid sea-level rise. Others continued to be driven landwards, probably by storms, to eventually produce modern barrier forms, as at Medmerry and Church Norton beach and spit. Overstepping of relict ebb deltas, banks and shoals, adjusted to earlier sea-level stillstands, is also likely to have occurred. Bone (1996) and Wallace (1990) offer several speculative dates for barrier breaching and reformation. Part of the substantial sediment resource of earlier gravel barriers has been redistributed to modern stores such as the Kirk Arrow spit; the Inner Owers; and the Pagham Harbour spits and tidal delta. Wallace (1990 and 1996) has attempted to fit a chronology of stages of sea-level rise to the apparent evidence of barrier breaching, breakdown and submergence. Evidence of brackish and estuarine sediments offshore Medmerry does suggest the former existence of a back-barrier lagoon that formed during a phase of sea-level stability; however, knowledge of the precise distribution and age of these sediments is insufficient to provide a more specific timeframe for barrier evolution.

Archaeological and sedimentological evidence supports the reconstruction of a continuous tidal creek linking Pagham Harbour with Bracklesham Bay (Heron-Allen, 1911; Millward and Robinson, 1973; Hinchcliffe, 1988; Wallace, 1990 and 1996; Castleden, 1998; Bone, 1996; Thomas, 1998). This may date back at least 2,000 years, perhaps resulting from a major breach of an earlier Bracklesham Bay barrier beach at Medmerry (Wallace, 1990). The Medmerry barrier is believed to have reformed and breached several times during subsequent centuries; at times isolating the Selsey peninsula as an island. Archaeological evidence demonstrates that the coastline was some 2 to 3km seawards of where it is now at about 5,000 years Before the Present (Cavis-Brown, 1910; White, 1934; Wallace, 1967, 1968 and 1996; Aldsworth, 1987; Goodburn, 1987; Thomas, 1998). Coastal erosion over this period must have occurred at a rate at least as fast as that recorded for the nineteenth and first half of the twentieth centuries (May, 1966). Documentary evidence for the medieval period (Bone, 1996) also indicates rapid coastline recession, especially during major storms. The latter probably caused the Medmerry barrier to repeatedly breach and break down, although there is reliable evidence that it was in place in the mid-sixteenth century. Stratigraphy from shallow boreholes into sediments infilling the former tidal creek isolating Selsey (Hinchcliffe, 1988; Wallace, 1990) clearly indicate oscillations between lagoon and brackish water conditions. A barrier spit may have connected Selsey Island with the mainland in the sixth century AD, but was permanently removed by a storm surge of exceptional magnitude in 1048.

Reclamation of some 120 hectares of saltmarsh occupying the tidal channel between Pagham Harbour and Medmerry was achieved when the Broad Rife sluice was built in 1884 (Photo3). This was undertaken in response to back barrier flooding resulting from a large pulse of gravel drift that blocked the Medmerry exit of this stream in 1880 (Bone, 1996). Further temporary blockages occurred in 1918, 1920 and 1924 before stabilisation of its present mouth in 1930.

The approximately triangular shape of the Selsey peninsula results from the protective presence of the Mixon reef some 2.5km seawards of Selsey Bill. This feature is composed of a relatively resistant Eocene calcareous Alveolina limestone cap rock overlying Bracklesham sands and clays. Wallace (1967, 1968, 1990 and 1996) has described a well-defined valley, up to 25m in depth and scoured by tidal currents, to the immediate south of the Mixon. The Outer and Malt Owers and The Streets are smaller bedrock reefs, but other offshore banks within 3km of the modern coastline appear to be sediment accumulations. They may be relict parts of a multistage barrier structure that was progressively segmented and submerged between 2,500 and 800 years before the present (Wallace, 1990; 1996). A remnant area of lagoonal and colluvial sediment that accumulated behind this structure survives inland of East Beach. Very fast erosion of this weak material occurred in the 50 years prior to the completion of coastal defences in 1960.

Wallace (1990; 1996) has speculated that the Mixon reef formed a part of the coastline in early Romano-British times. It may have "anchored" the contemporary position of the barrier beach mentioned above. A 17m deep sediment-infilled v-shaped gap between the Mixon and Malt Owers mark the course of the ancestral River Lavant. The latter is likely to have discharged via what is now Pagham Harbour prior to its diversion to Chichester and Fishbourne by Roman engineers in the second century A.D. Wallace (1990) also suggests that the proto-Lavant followed the line of the buried channel that runs roughly parallel to the modern East Beach coastline some 300-400m offshore. This feature has been largely infilled with late Holocene sediments, but continues to act as a local trap of mobile gravel during winter months. Some of this material is stabilised by weed growth during summer months, and may subsequently be transported by rafting to supply the Inner Owers and Kirk Arrow gravel accumulations. The strike-directed east to west valley south of the Mixon may also have been part of the course of an ancestral Lavant river.

Barrier breaching and shoreline recession associated with rising sea-level and storm events caused The Mixon to become an offshore bank, or shoal, probably at about 950-1050 AD (Wallace, 1990). It would have been emergent during mean low water, whilst the Inner Owers would, by this time, have been fully submerged. The Mixon therefore acquired its reef-like form and function from early medieval times onwards as sea-level rose further and both tide and wave-induced currents caused bedrock scour.

1.2 Hydrodynamics- Map

The tidal range is 4.9m (springs) and 2.7m (neaps) at Pagham Harbour mouth and at the entrance to Chichester Harbour, with the ebb phase shorter than the flood. The early ebb stage, gives rise to rectilinear, nearshore parallel, residual currents off the east-facing coastline. This stream moves towards the banks and reefs south of the Bill, where it is confined and movement is determined by their alignment. During the peak ebb flow, movement is north/north-eastwards. Maximum surface currents offshore the of apex of the peninsula are between 1.4ms^-1 (springs) and 0.7ms^-1 (neaps), reducing slightly at the seabed. Tidal currents adjacent to the west/south-west facing coastline flow predominantly eastwards/south-eastwards, as indicated by both float tracking and the morphology of patches of sand waves on the seabed (HR Wallingford, 1995, 1997, 2000). The protrusion of the Selsey peninsula into this net eastwards moving tidal stream creates an anticlockwise circulating gyre, (or "back-eddy") to the north-east, where residual current speeds are between 0.3 to 0.4ms^-1 at the peak of the flood stage. A smaller, clockwise moving eddy between The Streets reef and Kirk Arrow spit is set up when the ebb tidal flow is east to west (Wallace, 1990).

The offshore wave climate is dominated by waves from the south and south-west with periodic episodes of less energetic waves from the south-east. However, the shoreline wave climates are complex, as the east and west facing coastlines have contrasting orientations and western parts of Bracklesham Bay are partially sheltered by the Isle of Wight. Selsey Bill and East Beach are directly exposed to waves approaching from the south and east, but they also receive highly oblique refracted and diffracted swell waves that propagate from the south- west (HR Wallingford, 1992; 1995; 19987; 1998). West and north-west of Selsey Bill, dominant wave approach is from the south-west and wave crests are frequently parallel to the nearshore contours and shoreline. Bracklesham Bay is therefore a swash aligned shoreline, whereas Selsey Bill to Pagham Harbour is a classic drift aligned shoreline.

Wave shoaling and refraction is complicated by the presence of the submerged offshore reefs, shoals, banks, scarps and troughs. The Mixon, in particular, protects the southernmost shoreline from waves from the west and south-west, but the high incident angle of their approach is least modified immediately west of Selsey Bill. Wave climate for any one location on this coastline is a result of complex relationships between offshore to inshore transformation as a function of shoreface width and water depth; seabed relief; approach angles and interaction between wave and tidally induced currents in the breaker zone. Generally, waves steepen where tidal currents flow in opposition to dominant wind wave direction of approach. Overfalls at specific tidal states add further complications.

Given this complexity, it has been difficult to develop a quantitative wave climate. HR Wallingford (1995), using the TELURAY model, calculated a maximum annual wave height of 2.85m for the shoreline west of Selsey Bill, and 2.11m for the area 500m offshore of the east-facing coast. For mean inshore wave heights, Posford Duvivier (2001) re-ran HR Wallingford's (1995) data, using additional values derived from field measurements in 1998 and further refinements based on ENDEC model results (HR Wallingford, 1998). For annual recurrence, the maximum wave heights obtained were:

• Inner Owers: 2.8m
• South-East of Selsey Bill: 2.10m
• South-West of Selsey Bill: 2.17m
• West Wittering: 1.24m

Maximum significant wave heights are substantially greater than this, in the order of 15-20m for offshore waves off Selsey Bill (Hydraulics Research, 1974; HR Wallingford, 1992, 1993); HR Wallingford (1995) report maximum Hs (metres) to be: 4.6 (East Beach); 3.94 (Pagham Harbour entrance); 3.87 (Medmerry); 4.32 (East Wittering) and 1.44 (East Head). Inshore breaking wave heights, and incident angle of approach, directly control potential rates of longshore sediment transport, and are one basic explanation for the spatial variations summarised in Section 3.

Bracklesham Bay was one of the locations for which wave modelling excercises were undertaken as part of the DEFRA Futurecoast Project (Halcrow, 2002). An offshore wave climate was synthesised based on 1991-2000 data from the Met Office Wave Model and then transformed inshore to a prediction point in Bracklesham Bay at –4.34m O.D. Potential sensitivities to likely climate change scenarios were then tested by examining the extent to which the total and net longshore energy for each scenario varied with respect to the present situation. Results suggested that a one to two degree variation in wave climate direction could result in a 2-4% variation in longshore energy and confirmed that the Bay was significantly more sensitive to this factor than most other south coast locations, as might be expected of a swash aligned coastline.

2 Sediment Inputs

2.1 Offshore to Onshore Transport - F1 F2 F3 F4 F5 F6 References Map

Wave-transported sediment supply to the beaches of this coastline derives from several discrete sources, as detailed below. Tidal currents are not considered to be an independent mechanism of sustained onshore transport, but wave and tidal stream interaction creates complex patterns of turbulence that can entrain sediment.

F1 Onshore Gravel Feed From the Kirk Arrow Spit (see introduction to sediment inputs)

Air photos indicate gravel influxes from the bank to the shore during the periods 1959-60, 1971-72, 1986-92 and 1997-99 and corresponding extensions of the bank shoreward such that they temporarily attach to the shore. This suggests that gravel is transported onshore from the inshore end of the bank by shoaling waves approaching from the south. Lewis and Duvivier (1977) concluded that this feed occurs in pulses, separated by intervening periods of erosion, but averaged 5,000m³a^-1. This estimate is based on the quantity of feed necessary to maintain beach levels over longer-term periods, in spite of output by littoral drift and beach drawdown under high-energy wave conditions. Gifford Associated Consultants (1997) propose the higher figure of approximately 10,000 m³a^-1, but this also includes an estimate of the quantity gravel that is moved eastwards of the Selsey peninsula. HR Wallingford (1995) and Posford Duvivier (2001), using both recent and historical data, estimate that 80-85% of onshore deposition is subsequently moved rapidly eastwards, whilst the remaining 15-20% either remains in place or slowly drifts westwards.

Wallace (1990) examined accretion behind groynes constructed in 1989 at Selsey Bill and estimated 15,300 m³ of gravel to have accumulated over a four month period (46,000 m³a^-1 if maintained over a year). Wallace (1990) also calculated that 5 million cubic metres of shingle have accumulated south of the entrance to Pagham Harbour since 1866, giving a mean rate of 41,677 m³a^-1, a value that corresponds well with the four month estimate from 1989. The major sources are regarded as net onshore supply from both the Kirk Arrow Spit and from the nearshore Inner Owers bank (see below), and his figures would appear to include both.

It can be concluded that strong circumstantial evidence exists indicating significant but intermittent onshore transport of gravel from the Kirk Arrow Spit. The quantitative estimates of this feed are of medium reliability because it apparently occurs as high magnitude, low frequency pulses that are not easily measured. Additional information is required on the frequency, volume and duration of typical pulses, as well as on the pattern of changes in the shape and volume of the spit itself. Better knowledge is required of how it came to expand rapidly in the late 1980s to become habitually exposed during low water spring tides, thus creating a wide inter-tidal foreshore. However, during the previous three decades, it was detached and only rarely emergent. Ultimately, over a long timescale that cannot be determined at present, the Kirk Arrow Spit represents a finite source of supply, as there is a probability that its sources of replenishment will decline over time and eventually become exhausted.

F2 Onshore Feed From the Streets and Malt Owers Reefs (see introduction to sediment inputs)

The Streets Reef and Malt Owers comprise two small denuded bedrock antiforms composed of a distinctive dark grey limestone. These interact with waves and a local tidal current gyre, causing turbulence, which can result in deposition of kelp-rafted gravel (Jolliffe and Wallace, 1973). Steep infacing slopes prevent offshore movement of mobile shingle, which tends to be driven along the strike of the beds by wave action and subsequently onto West Beach, Selsey (Harlow, 1980). The source of the kelp rafted gravel may derive from the "outer circulation" supplying East Beach (Wallace, 1990). Movement of kelp-rafted shingle, together with the pattern of deposition on the Streets, has been observed by divers (Jolliffe and Wallace, 1973; Jolliffe, 1978 and Wallace, 1990); thus, the reliability of this information is medium although quantitative details are lacking. This onshore wave driven feed has been deduced from observations of changes in beach volume at West Street, involving an estimated 1,000 m³ annually (Harlow, 1980; Wallace, 1990). These volumetric calculations are based on indirect evidence and as few details of measurements are provided it is of low reliability.

F3 Onshore Feed From the Inner Owers (see introduction to sediment inputs)

The Inner Owers are a series of mobile nearshore gravel banks, situated between East Beach, Selsey and Pagham Harbour inlet, which periodically migrate onshore. They are built onto the western margins the Pagham tidal delta and are characterised by a gentle offshore slope and a steeper inshore slope (Lewis and Duvivier, 1977). Their shape and form is determined by wave diffraction and refraction. Gravel is supplied to the local beaches in wave-driven pulses in a very similar way to Kirk Arrow Spit. The behaviour of these bars is analogous to that of swash bars that are widely associated with sand dominated estuaries in North America (Fitzgerald, 1996). The gravel supply process has been investigated by means of air photos and site observations (Lewis and Duvivier, 1977; Wallace, 1990). Based on estimations of the reasonably constant shape and volume of these gravel banks, observed during their migration phases, a total input of 10,000 m³ was calculated for the period 1970-75; a longer term average input of 3,000-5,000 m³a^-1 is quoted by Lewis and Duvivier (1977). This process has been observed directly via diver surveys, and as its contribution to beach levels is evident, this information is regarded as of medium to high reliability. Further quantitative information is currently not available, so that research and monitoring over an appropriate timescale (at least 10 years) is necessary to determine long-term supply pathways and rates.

F4 Diffuse Weed Rafted Shingle Feed (see introduction to sediment inputs)

The existence of this process was established by Jolliffe and Wallace (1973) and further confirmed by Harlow (1980). The evidence for shingle feed comprised diving observations of weed-attached shingle undergoing transport offshore and observations of beach shingle with attached holdfasts and/or kelp fronds. No further quantitative details are available.

F5 Input from the Chichester Tidal Delta (see introduction to sediment inputs)

F6 Sand Feed from the Chichester Tidal Delta to East Head (see introduction to sediment inputs)

Sand deposited on the outer Bar and East Pole Sands has the potential to be transported onshore by wave action to East Head (Webber, 1979; Harlow, 1980). Evidence for this is based upon (i) particle size variations over the tidal delta, which may involve a contribution from tidal currents during spring cycles (GeoSea Consulting, 2000); (ii) hydrographic chart evidence of up to 3m of erosion on the western margins of East Pole Sands, and 2-3m of accretion further east (ABP Research and Consultancy, 2000). This has occurred since 1923, and appears to have involved a change from a previously mobile gravel and sand surface to one, which is now stable. The latter feature was first reported by Webber (1979) and may be due to bevelling of the underlying clay substrate and its subsequent armouring by gravels.

2.2 Coastal Erosion - E1 E2 Selsey Bill East Beach, Selsey Church Norton References Map

Selsey Bill, East Beach and the coastline of Bracklesham Bay have a history of rapid erosion of low cliffs and the beach with several km being lost in historical times and a more recent maximum rate of 7.6 ma^-1, being recorded at East Beach, Selsey for the period 1932-51 (Duvivier, 1961; Millward and Robinson, 1973). The peninsula includes some spreads of raised beach deposits to the west and east of the Bill around Selsey. It is likely that these deposits were formerly much more extensive, but have been reduced greatly in extent as the headland has diminished. Erosion at such rapid rates necessitated installation of extensive and robust coast protection and defence structures, undertaken in the period 1945-61. These include groynes, revetments and seawalls which have either halted cliff erosion or reduced beach losses. The only unprotected areas where natural recession of the shoreline continues are:

(i) East Head (Photo 4 and Photo 5)
(ii) A 300 m long cliff frontage from Hillfield Road to Medmerry, Selsey (Photo 2)
(iii) Several stretches between the proximal end of Church Norton spit and the Pagham Harbour inlet.

These sites are detailed below. In addition, erosion takes place across the approximately 2000m wide shoreface between Selsey Bill and East Head. Posford Duvivier (1999), using a modification of the Brunn Rule and assuming vertical erosion, principally by waves, of 1 mma^-1 calculate an erosion yield of 19,000 m³a^-1 between Selsey and Bracklesham. This may increase slightly over the sector between East Wittering and East Head. The material released by shoreface erosion is believed to constitute fine sediments that are likely to be removed seaward in suspension.

E1 East Head (see introduction to coastal erosion)

The morphological development of the East Head spit has been fully documented by Searle (1975), May (1975), Lewis and Duvivier (1977), Harlow (1980), ABP Research and Consultancy (2000) and Baily et al. (2002) (see Section 5). These studies have reported a clockwise rotation of the spit accomplished by very rapid recession of its seaward face, at 6.8 ma^-1 during the period 1875-1896 and 2.3 ma^-1 during the period 1896-1909. This rate had slowed by 1926 and by 1963 the spit was in approximately its present position. Although East Head as a whole has retreated very little since 1963 it cannot be regarded as stable for it was breached along its neck immediately north of "The Hinge" by a storm in 1963, was overtopped in 1987 and has experienced rapid thinning since the early 1990s (Photo 5). Its apparent stability has only been achieved by extensive use of artificial structures to stimulate dune growth (Searle, 1975; Baily et al. 2002). Most authorities agree that a combination of a spring tide and a severe storm could again breach the neck of the spit resulting in further recession and possibly its permanent breaching and ultimate destruction. Several alternative breach scenarios have been modelled and evaluated (HR Wallingford, 1995, 2000; ABP Research and Consultancy, 2000). The exact cause of the current phase of erosion at The Hinge is uncertain, but involves reductions in natural sand supply from updrift longshore and nearshore sources. An additional factor may be a continuing adjustment of cross-section of the Chichester inlet mouth to the tidal prism of the harbour (ABP Research and Consultancy, 2000). It has involved lowering since 1923 of the Winner sand and gravel bank by up to three metres allowing increased wave exposure and reducing the intertidal foreshore width in front of East Head.

E2 Hillfield Road to Medmerry (see introduction to coastal erosion)

Low cliffs cut into raised beach pebble and sand deposits form a 300m long unprotected section (Photo 2) that has eroded relatively steadily at 1.09 to 1.25 ma^-1 between 1875 and 1972 (Harlow 1980; Posford Duvivier, 1997). This rate was calculated from examination of successive OS 1:2500 map editions. Earlier maps suggest that erosion was more rapid in the period 1840-1880 (Harlow, 1980). Harlow (1979, 1980) attempted a calculation of sediment yield based on an appreciation of (i) the lithology of eroding raised beach and drift sediments, (ii) calculation of cliff height variation between successive periods over nearly one hundred years and (iii) an allowance for losses via suspended transport. He suggested that up to 1,480 m³a^-1 of gravel and 3,820 m³a^-1 of fine gravel and sand could be released and potentially contribute to the upper beach and foreshore. Posford Duvivier (1997) calculate a total supply of 1,000 m³a^-1 divided equally between gravel, sand and clay. Use of historical map sources coupled with Ordnance Survey maps enabled Lewis and Duvivier (1950) to calculate mean retreat at 1.34-1.82 ma^-1 over the period 1778-1953, with a sediment yield of approximately 10,000 m³a^-1. This higher rate probably reflects the more rapid erosion in the 18th and 19th centuries along a rather longer essentially undefended coastline so that the figures calculated by Harlow (1980) and Posford Duvivier (1997) are more representative of the present-day inputs. Unpublished local authority records (Lewis and Duvivier, 1950) report that between 1930 and 1952 annual rates of erosion between Medmerry and East Beach were as high as 8m. This was the highest rate recorded for any location in England during the twentieth century.

Selsey Bill (see introduction to coast erosion)

Prior to the construction of comprehensive 'hard' sea defences between 1962 and 1969 ( Photo 1), much of the tip of the Selsey peninsula provided inputs of easily eroded sediment from wave-induced cliff and shoreface erosion. This has been the case for over 1300 years, thus accounting for over 2 km of coastline retreat since the second or third centuries AD (Ballard, 1910; Heron-Allen, 1911; White, 1934; Aldsworth, 1987; Wallace, 1990 and 1996; Castleden, 1996; Bone 1996; Thomas, 1998;). Shoreface erosion has accelerated over this period due to ongoing sea level rise and the lowering of protective off- shore rock outcrops. There is a rich, only partially explored, offshore archaeological legacy of submerged Romano-British, Saxon and early medieval landscape features, partially recorded in documentary and archival records (Heron-Allen, 1911; Wallace, 1990). Sediment yield derives not only from (former) rapid retreat of low cliffs, but abrasional scour of the complementary, expanding, shoreface platform. Hydraulics Research (1995) estimated that rapid erosion between 1850 and 1950 could have released as much as 2 million m³ of gravel from raised beach deposits. Much of this resource has now become exhausted as the headland has diminished. Lewis and Duvivier (1977) calculate an annual volume of 7,500 m³ of shingle for 1909-1962, feeding East Beach, Selsey. The quantity of sand was, and may continue to be, substantially greater, but is too fine to be retained on local beaches.

East Beach, Selsey (see introduction to coast erosion)

Posford Duvivier (2001) calculated that approximately 150m of recession of mean low water occurred between 1900 and 1950, with substantial beach drawdown and erosional loss along this south-east orientated shoreline. Losses could not be compensated by updrift littoral or nearshore transport, although some fresh supply of shingle may have derived from erosion of Raised Beach deposits periodically exposed in the back and mid shore areas. A series of successively landward relocations of the Lifeboat Station, 1909-1960, are described by Wallace (1990).

Church Norton and Pagham Harbour Inlet (see introduction to coast erosion)

Several specific sites have recorded past and recent beach erosion, considered (Gifford Associated Consultants, 1997; Posford Duvivier, 2001) to be between 4,000 and 8,000 m³a^-1. These constitute changes in the beach sediment store and are not specifically regarded as inputs. Routine nourishment of the beach fronting Church Norton spit has taken place since the early 1980s to offset this loss and thus maintain the stability and integrity of Pagham Harbour.

2.3 Beach Nourishment and Re-cycling - References Map

Medmerry

The Environment Agency and its predecessors have conducted a long-term nourishment and re-profiling scheme at the critical Medmerry barrier beach site (Environment Agency, 1998a). This structure has a long history of intermittent landward migration. Its present form dates to approximately 1600, when it was reported to have blocked a former tidal inlet channel. Subsequently, it has breached and reformed on several occasions (Bone, 1996). However, in recent years it has exhibited increasing instability, and has experienced regular cutbacks, overtopping and breaching. It has necessitated increasingly urgent nourishment and profile reconstruction in order to maintain the present defence line (Photo 6 and Photo 7). This sector of beach is in a condition of chronic disequilibrium, unable to adjust to natural gravel losses and foreshore lowering. HR Wallingford (1995) demonstrate that it is located at a focal point for wave erosion due to refraction induced by complex offshore relief. The initial nourishment was in 1976 and comprised a pilot scheme involving deposition of 14,500m³ shingle and extension of groynes to LWM. The main phase was conducted between 1976 and 1980, with the addition of 225,000m³ of gravel obtained from inland gravel pits and delivered by truck to build the berm. This was preferred to dredged marine gravels that could be deposited on the lower foreshore because Hydraulics Research (1974) were unable to confirm that this material would move onshore. The nourishment material comprised nodular flint gravels significantly coarser and more angular than the indigenous beach material. The scheme also involved insertion of 38 groynes along a 3.8 km frontage and their extension to MLWST. The artificial beach was re-profiled to a slope of 1:10, with gravel that accumulated at the most westerly downdrift groyne subsequently being recycled updrift. Further replenishment and crest elevation was completed between 1989 and 1996, involving some 90,000m³ to compensate for losses of 102,000m³ over the period 1974-1992 (HR Wallingford, 1995; Environment Agency, 1998a). Several major storm surges during the winters of 1998- 9, 2000-1, 2001-2 caused overwashing, crest lowering, beach drawdown, with a 300m breach in 1999. This has necessitated the emergency dumping (Photo 7) of over 500,000m³ of gravel (again taken from inland sources), together with profile reconstruction. By the mid-1990s the groyne field had deteriorated and was virtually redundant as a mechanism for beach sediment conservation. A modest managed re-alignment of this barrier structure has been proposed (Posford Duvivier, 2001) as the only morphodynamically sustainable strategic defence option for this site. However, even if adopted, this would not avoid the need for future re- nourishment and control, albeit on a less intense basis dependent on distance of set back relocation.

Church Norton Spit

The beach fronting the southern (Church Norton) spit, protecting the entrance to Pagham Harbour, is replenished by routine artificial cycling of gravel, taken from the adjacent nearshore banks of the Pagham tidal delta, south-west of the inlet and from the Inner Owers. Since the early 1990s, this has averaged 15,000m³a^-1, but reliable figures for the previous decade are not available. Profile reconstruction is also carried out as part of this routine practice.

3. LITTORAL TRANSPORT AND BEACH DRIFT - LT1 LT2 LT3 LT4 References Map

The longshore drift system involves both gravel and sand, but a significant quantity of fine sand is probably removed in suspension directly offshore (Posford Duvivier, 1999). Tidal gyres either side of Selsey Bill created by the intrusion of the headland into the pattern of rectilinear eastwards moving residual currents may result in locally complex transport of fines by tidal streams towards (i) East Beach and (ii) Medmerry Bank (Paphitis, et al., 2000; HR Wallingford, 1995; Gifford Associated Consultants, 1997). The littoral transport system is dominated by the influence of shoaling and breaking waves. Although there are well defined net transport pathways, short-term reversals occur, especially along the swash-aligned shoreline between between Selsey Bill and East Head (HR Wallingford, 1995). Monitoring of beach levels has revealed significant fluctuations in annual drift rates since the early 1970s (Posford Duvivier, 2001; HR Wallingford, 1995, 1998).

LT1 Bracklesham Bay (Selsey to West Wittering) (see introduction to littoral transport)

All authors agree that net drift of gravel is westward from West Street, Selsey to East Head. Lewis and Duvivier (1950) and Duvivier (1961) based this on visual observations and a wind vector analysis of 7 years of records from Thorney Island. Hydraulics Research (1974) derived the same conclusion from shingle tracer experiments undertaken over a 6 month period, and Harlow (1980) and Lewis and Duvivier (1977) estimated drift on the basis of beach volume and shoreline changes since 1868. However, detailed observations and transport modelling studies based on hindcast wave climates have revealed short-term and short distance reversal of the net drift direction as a result of varying incident wave approach (Posford Duvivier, 1992, 2001). Sensitivity to wave approach direction and the capacity for drift reversals are high because Bracklesham Bay is a swash-aligned shoreline (Halcrow, 2002).

Quantitative analysis has been attempted by assuming littoral drift to be the minimum sediment volume required to explain observed beach volume trends. Using this technique Harlow (1980) calculated a drift rate (all sediment grades) of 35-40,000 m³a^-1 for the period 1965-1973; 40-50,0000 m³a^-1 for 1933-65, but only 1,000-8,000 m³a^-1 for the period 1973-77. Between 1846 and 1896 it is estimated to have been of the order of 70,000 m³a^-1. A present day mean drift volume of between 2,800 and 7,000 m³a^-1 is suggested for this unit as a whole (Posford Duvivier, 2001; ABP Research and Consultancy, 2000). The potential drift rate at West Beach, Selsey is 15,000-16,000 m³a^-1, declining to 2-6,000 m³a^-1 at Bracklesham (Photo 8), because of the reduction in wave approach angle north-westwards (HR Wallingford, 1995, 1997). These, however, are net values, so that given the frequency of drift reversal, gross values are considerably higher. It should be noted that the lower rates of drift associated with more recent decades mostly reflect the role of defence structures in reducing fresh sediment inputs and intercepting transport.

Erosion of the proximal end of East Head spit since the mid 1990s suggests that the input of sediment via littoral drift updrift of the terminal groyne is virtually zero, although a net westwards drift here of 7,000m³a^-1 is suggested by Webber (1979), revised downwards by HR Wallingford (1995) to 2,600m³a^-1. Through time, the drift rate has steadily fallen, a feature attributed both to fluctuations in the volume of onshore feed at Selsey and the effects of progressively more robust and comprehensive coastal protection structures in reducing supply from coast erosion and arresting beach drift. Lewis and Duvivier (1977), HR Wallingford (1995) and Posford Duvivier (1992, 2001) calculated a prevailing natural drift for the upper gravel beach of 2,000-5,000m³a^-1 downdrift of Medmerry, but added that this figure was only valid for the ungroyned coast; thus a negligible net drift rate of 300-500m³a^-1 was considered more realistic for the heavily groyned East Wittering to Bracklesham frontage (Photo 9) and 1,000-2,000m³a^-1 at West Beach, Selsey. In general, drift rates on this shoreline are determined by sediment availability and the condition of groynes, as deteriorating or overflowing structures can locally increase throughput for limited periods. Significant interruption to upper beach transport occurs at outfall sites, particularly Broad Rife (Photo 3), leading to immediate downdrift starvation. Thus, the downdrift benefits of the substantial gravel recharges of Medmerry beach have been surprisingly modest. Cross-shore rather than long-shore fluctuations have been dominant at Medmerry, varying between annual gains of up to 40,000m³ to annual losses of over 60,000m³ (net loss of 17,000m³a^-1 since 1974). Volumetric assessments of littoral transport are based on minimum net drift values derived from assessing inter-tidal beach changes. Throughput, which causes no discernable beach volume change, cannot be detected so that absolute drift rates could be significantly greater. Modelling studies more effectively identify the natural littoral drift potential, such rates cannot be achieved due to the effects of groynes diverting sediments into storage.

LT2 Selsey Bill to Pagham Harbour (see introduction to littoral transport)

Littoral drift of gravel is from Selsey Bill north-eastwards to the entrance to Pagham Harbour (Photo 1). The main evidence for this comprises inter-tidal beach level observations and analysis of volume changes (Lewis and Duvivier, 1955; Duvivier, 1960; Wallace, 1990; HR Wallingford, 1995; Gifford Associated Consultants, 1997; Posford Duvivier, 2001) (see Section 5). Littoral drift pathways therefore diverge in the vicinity of Selsey Bill. Detailed analysis of maps, air photos and beach level measurements and observations in groyne compartments enabled Lewis and Duvivier (1976), Harlow (1980), HR Wallingford (1995, 1997) and Posford Duvivier (2001) to locate this regionally significant drift divide between Warner Road (net westward drift) and Hillfield Road (net eastward drift). At this point, groynes and a seawall have been constructed, either side of which erosion has occurred, thus forming an artificial headland. Between Hillfield Road and Selsey Bill the eastward drift rate is low as the beach is aligned closely to the dominant wave approach (Lewis and Duvivier 1977; Posford Duvivier, 2001). Immediately north-east of Selsey Bill the potential for drift is much greater due to the sudden change in shoreline orientation to face south-east. Beach drift is also hindered in the vicinity of the Bill by the seawall and numerous groynes, but periodic offshore to onshore influxes of gravel become piled against their west sides by the prevailing drift, so that overtopping and outflanking occurs. The eastern sides of groyne compartments are usually depleted, so that the operation of any significant counter (westward) drift is unlikely. However, the precise position of the drift divide fluctuates up to 300-400m, depending on prevailing wave conditions. Thus, the Bill operates as a "one way valve" facilitating net eastward gravel transport (Lewis and Duvivier, 1977; Posford Duvivier, 2001). Estimations of rates of drift have involved both assessments of changes in beach volumes and transport modelling approaches based on hindcast wave climates. The most effective studies have sought to compare the results of the two techniques. Studies have either focussed upon accretion of the Church Norton Spit (Photo 10), or upon beach volume changes throughout the pathway as follows:

Church Norton Spit Growth (see introduction to littoral transport)

The approach involves measurement of accretion immediately south of Pagham Harbour entrance and attribution of all material accumulating to littoral drift from Selsey and East Beach. Wallace (1990) determined a mean drift rate of 41,500m³a^-1 over the period 1866- 1989, with a maximum of 76,000m³a^-1 in 1962. Lewis and Duvivier (1977) calculated the marginally higher potential rate of 50,000m³a^-1 for the period 1875-1909. Modelling using the LITPACK numerical model (Gifford Associated Consultants, 1997) suggests drift of 71,000m³a^-1 for all sediment grades Both studies neglect possible direct inputs to beaches flanking Pagham Harbour inlet from sources such as offshore gravel banks, and so probably overestimate actual drift rates. Their approaches also fail to allow for variations in supply resulting from both periodic upgrading of groynes and recharge operations. HR Wallingford (1995) calculated a drift rate of 33,000m³a^-1 for East Beach, reducing to 8,000m³a^-1 when adjusted for assumed groyne efficiency in their model studies. For Selsey Bill, their equivalent figures are 13,700 and 5,500m³a^-1. Barcock and Collins (1991) have re-calculated the prevailing drift rate between East Beach and Pagham Harbour entrance to be between 24,000 and 42,000m³a^-1. This is based data on the frequency distribution of wave heights and approach directions and considers sediment exchanges with Pagham tidal delta. Using HR Wallingford's (1995) DRCALC model, updated by later wave climate information, Posford Duvivier (2001) propose a potential drift rate of 32,000m³a^-1. Actual rates are considered to be 25-30% of the above volumes because of the role of groynes.

Beach Volume Changes from Selsey Bill to Church Norton(see introduction to littoral transport)

The approach involves estimation of drift from a consideration of all beach volume changes between Selsey Bill and Church Norton. Using this approach Wallace (1990) recorded movement of 15,300 m³ in 4 months at Selsey Bill, and Lewis and Duvivier (1977) deduced an average input close to Selsey Bill of 5,000-6,000m³a^-1 (1959-75) composed mainly of pulses of gravel onshore from the Kirk Arrow Spit. Posford Duvivier (2001) estimated a drift rate of 13,700m³a^-1 at Selsey, reducing to 5,500m³a^-1 once groyne performance is factored in. Drift potential was found to increase north-eastwards on East Beach, where it is between 15,000 and 25,000m³a^-1. Much of this beach frontage is managed, so that actual drift depends on groyne performance and sediment availability. HR Wallingford (1995, 1997) state that a mean quantity of 5,000m³a^-1 is supplied from the Kirk Arrow Spit and is augmented by a further 5,000m³a^-1 from the Inner Owers, giving a total drift potential of 25-30,000m³a^-1. This figure is based on the fact that on formerly unprotected stretches erosion of up to 10,000m³a^-1 occurred to maintain drift rates; it corresponds closely with the estimated gravel input from erosion behind East Beach before coast protection. The sediment supply and transport system had previously achieved an equilibrium so that no net beach erosion occurred between East Beach and Church Norton. Terminal scour and thus some downdrift starvation has been a significant past problem at the north-east end of this protected frontage. HR Wallingford (1995) use a modelling approach informed by measured beach level changes to calculate drift to be approximately 32,000m³a^-1 along the Church Norton spit, although the rate is less than 17,000m³a^-1 for the upper gravel beach alone.

Overall, there is some variation in rates of processes along this complex frontage and numerous alternative estimates of drift have been produced. There is some consensus that potential drift in recent decades of all sediments approximates to around 30,000m³a^-1 with actual drift being in the range 7,000 to 11,000m³a^-1 being controlled by beach management practices. The mean rates reported above make assumptions on the magnitude and frequency of pulses of supply and the overtopping of groynes. Despite this, the overall pattern and volume of drift has been established at medium to high reliability with moderate to good correspondence between modelled and observed processes.

The effect of sustained long-term unidirectional north-east drift along this sector, over at least the last millennia, has been to deliver much sediment to build the southern spit that helps to define the entrance to Pagham Harbour (Photo 10). It has a history of fluctuation, thus indicating temporal variation in littoral and offshore drift supply (see Section 5). It is currently in a phase of depletion, but potential erosion losses are offset by deliberate cycling of gravel from co-extensive nearshore bars, at a rate of approximately 11,000m³a^-1 since the late 1980s. The northern (Pagham) spit is the product of a local 'counter' drift, resulting from a transport divide some 2000m north-east of the harbour entrance. This itself is the outcome of interaction between tidal currents generated by the inlet and complex wave refraction over the Pagham tidal delta (Geodata Institute, 1994). With a maximum SW drift throughput of 5,000m³a^-1 (Barcock and Collins, 1991; Collins, et al., 1995), this northern spit has had less capacity for growth and change than its southern counterpart.

Offshore tidal current transport of sand, inferred from sample surveys of bedforms and numerical modelling (Barcock and Collins, 1991) is considered to be towards the south-west, or west, in water depths of less than 15m.

LT3 Reversed Littoral Drift of Sand in Bracklesham Bay (see introduction to littoral transport)

Experiments employing fluorescent sand and shingle tracers at Medmerry (Hydraulics Research, 1974) have indicated that sand transport may be reversed on the lower foreshore, seaward of groynes, due to strong eastward residual tidal currents. Any reversal was, however, considered to be local to the study site because the westward tidal current increases in velocity west from Medmerry so inducing net sand transport westward to East Wittering; a transport divide may therefore exist in the vicinity of Bracklesham. The tracer experiments were only conducted over a 6 month period so it is possible they were unrepresentative of typical conditions. The experiments offer the only direct information on this component of sediment transport, although HR Wallingford (1995) state that an anticlockwise circulating tidal eddy exists outside the entrance to Chichester Harbour that transports sand from the foreshore between East and West Wittering onto East Pole Sands. Other studies fail to acknowledge such a net drift reversal and Lewis and Duvivier (1977) argued that all sand transport on the lower foreshore was, on balance, westward (i.e. in the same direction as the shingle of the upper foreshore). Their evidence was sand distribution observed within groyne compartments, analysis of residual offshore currents, and the lack of sand at Selsey contrasted with its abundance at East Pole Sands and East Head. Modelling by HR Wallingford (1995) computed the transport rate for sand to be some 25,000m³a-1from east to west

LT4 Drift Divergence North of Pagham Harbour Entrance (see introduction to littoral transport)

A local reversal of drift occurs at the eastern boundary of this frontage for the net direction of drift immediately east of Pagham Harbour inlet is considered to be westwards over a 500- 700m long frontage at an approximate rate of 5,000m³a^-1. The Pagham Harbour ebb tidal delta and wide, accreting foreshore sets up complex local wave refraction and provides protection against the dominant south-westerly waves enabling a very local dominance of south-easterly waves (Barcock and Collins, 1991; Gifford Associated Consultants, 1997; Posford Duvivier, 2001). A drift divide marked by a zone of persistent beach erosion is therefore identified in the vicinity of Pagham Beach estate (Wallace, 1990; Posford Duvivier, 2001).

4. SEDIMENT OUTPUTS

4.1 Estuarine Outputs - EO1 EO2 References Map

EO1 Chichester Harbour Entrance (see introduction to estuarine outputs)

Eastward shoreline drift from Hayling Island and westward drift from West Wittering and East Head converge at the harbour entrance (Photo 11). Sand and gravels entering the Chichester tidal channel are flushed offshore by the strong ebb residual tidal current and deposited at varying distances from the entrance depending upon sediment size, wave conditions and water depth. This is confined by the orientation and shape of bar topography (HR Wallingford, 2000) and sediment trend analysis (Geosea Consulting, 2000). Gravel can be transported a maximum of 2km offshore and sand a maximum of 3.5km offshore (Webber, 1979). The result of this offshore flushing of sediments has been the accumulation of some 25 million cubic metres of sediment within a major ebb tidal delta (Webber, 1979). Sediment sampling by Harlow (1980) and GeoSea Consulting (2000) revealed a series of sedimentary zones and potential transport pathways, suggesting that wave action can mobilise sediments on the tidal delta and drive them back shoreward towards Eastoke, Hayling and West Wittering. The net result of these processes appears to be an anticlockwise circulation in the east of Chichester entrance as reported by ABP Research and Consultancy (2000). Part of this circulation is depicted in Photo 11.

The volume of sediment transported and deposited offshore by tidal currents has not been calculated, but fresh supply to the tidal delta could be estimated from littoral drift inputs at the entrance. Drift inputs to the tidal channel were undoubtedly greater in the past (over 70,000m³a^-1), but have substantially declined over the past 100 years as coast protection has intercepted and reduced drift along the shoreline of Bracklesham Bay (Harlow, 1980; Posford Duvivier, 2001). Not only have inputs to the delta reduced, but losses due to dredging of Chichester bar have increased since 1973 (see Section 4.2). In fact, analyses of bathymetric data have suggested that the ebb tidal delta suffered a net loss of 1.4 million cubic metres from 1974 to 2000 (Posford Duvivier, 2001).

The rotation and recession of East Head during the nineteenth century (Section 5.3) caused the formerly deep entrance channel close to the east shore of Hayling Island widen eastwards along a north to south axis, thus reducing the Winner Bank. This may continue a long-term trend, as the absence of drift deposits beneath Sandy Point (Hayling Island) suggests that the harbour channel was further west at least two or three millennia Before the Present.

Analysis of the hydraulic regime at the entrance channel, via hydrodynamic modelling (ABP Research and Consultancy, 2000; HR Wallingford, 1998) has shown that there has been a rather variable pattern of channel narrowing and deepening; widening and shallowing since the mid-nineteenth century. The channel initially decreased in size, as the Winner bank accreted, 1887-1923; but overall, given the rotation and retreat of East Head and subsequent lowering of the Winner, the cross-sectional area of the channel has increased over the last 150 years. This suggests that it is adjusting towards a new equilibrium condition, but is below its optimum cross-sectional area given the tidal prism of Chichester Harbour. It stimulates the suggestion (ABP Research and Consultancy, 2000) that the harbour mouth has adjusted, or is adjusting, to a change from a wave-dominated littoral transport fed sediment budget to one which is controlled more strongly by tidal currents.

EO2 Pagham Harbour Entrance (see introduction to estuarine outputs)

Ebb tidal currents are moderate and their influence does not extend very far seaward compared to those generated at the much larger Chichester Harbour. Wave action is modified by local refraction induced by complex bathymetry, but HR Wallingford (1993) calculated a mean significant wave height of 1m, and a maximum of 4.5m, at Pagham Harbour Entrance. Wave induced currents oppose seaward transport and tend to drive material back landward where ebb tidal currents are weak. A consequence is that the ebb tidal delta is located close to the inlet and is relatively small. Sediment therefore has a short residence time within the delta and is liable to being driven back ashore within swash bars to the west and east of the inlet.

Net seaward discharge that generates accretion of the ebb tidal delta at Pagham entrance is in the order of 16,000m³a^-1 representing the balance estimated between landwards (flood tide and wave-driven) input of 18,000m³a^-1, and seawards removal of 34,000m³a^-1 (Geodata Institute, 1994).

Flushing processes and the ebb delta sediment budget have undoubtedly changed over the past four centuries due to reduction of the tidal prism of the harbour as a consequence of land claim. The earliest record of this process is in 1580, with the first major flood ebmbankment constructed in 1637 (Brown, 1981; Cavis-Brown, 1910; Graves, 1981; Environment Agency, 1998). Approximately 1.26km² were reclaimed between 1672 and 1809 and the remainder was reclaimed in 1877. Breaching in 1910 led to rapid inundation of the present harbour area of 2.83km². It can be postulated that the ebb tidal would have reduced in size and area as the tidal prism was reduced and it is likely that much sediment was driven ashore from 1876 to 1910 when there was no active inlet.

4.2 Dredging

Dredging across Chichester Bar, on the ebb tidal delta, is carried out routinely to maintain a channel for navigation; some of the material is used for beach renourishment on south-east Hayling Island. Dredging for aggregates, on a small scale on The Winner, was carried out between the early nineteenth century and about 1920. Dredging for navigation access, however, did not become significant until 1973.

A total of 600,00m³ of gravel was thus removed over the period 1974-1982 (Harlow, 1985) and between 1988-1996 dredging was permitted up to an annual limit of 20,000 tonnes (12,500m³a^-1). Analysis of hydrographic surveys indicated that water depths increased over the dredged area in the 1970s (Webber, 1979). Despite this, it was difficult to attribute this change solely to dredging because the Chichester tidal delta is a large sediment reservoir (25 million m³) characterised by major natural internal fluctuations and sediment redistribution (ABP Research and Consultancy, 2000). From bathymetric information (Posford Duvivier, 2001) it can be concluded that the Chichester tidal delta has more recently lost material due to reduction of littoral supply from Bracklesham Bay in combination with outputs by dredging. For the period 1974 to 2000 this can be estimated at 1.4 million cubic metres. This quantity is quite significant in comparison to the total estimated volume of the delta, particularly when it is considered that the bulk of output comprises gravel from the inner bar whilst much of the estimated stored volume comprises sand on the outer bar. Continued dredging and onshore feed might therefore be expected to deplete reserves, such that natural onshore feed could reduce in the near future (or may already have done so).

5. SEDIMENT STORES - References Map

Sediments are stored along this shoreline within its beaches, spits and within ebb tidal deltas associated with harbour inlets.

5.1 Beach Morphology and Sedimentology (see introduction to sediment stores)

Along much of this coast the upper beaches are steep and composed of flint gravel, whilst the lower foreshore has a relatively shallow slope and is composed of medium and fine sand (Hydraulics Research, 1974; Lewis and Duvivier, 1977; Harlow, 1980; Posford Duvivier, 2001). Patchy gravel frequently overlies foreshore sand, which in turn normally conceals the sandy clays of the Eocene Bracklesham Series. The main beach morphodynamic and sedimentary features are distinctive along the several discrete sectors of this coastline, namely:

(i) East Head: The seaward side is composed of a wide gently sloping sandy foreshore (ABP Research and Consultancy, 2000), which narrows abruptly at the point of distal curvature. Here, a steep convex shingle beach has become a more established feature in recent years, and may represent the re-exposure of the original shingle platform on which the spit has been developed. A very narrow shingle backshore beach at The Hinge has been eliminated by recent erosion, but has been retained immediately updrift by groynes.

(ii) Bracklesham Bay: Beach crest levels are typically maintained at around 5.4m above Ordnance Datum (AOD). An exception to this is along and immediately west of the Medmerry frontage (Photo 6) where maximum levels exceed 6m AOD due to artificial profile reconstruction or "beach scraping" (Harlow, 1980; Hydraulics Research, 1974; Posford Duvivier, 2001). Beach width decreases south-eastwards whilst the slope of both the upper and lower foreshore increases north-westwards. Beach sediment sampling between 1km west of Selsey Bill and 0.5km east of Bracklesham indicated that backshore pebbles were generally larger than those on the foreshore and at Bracklesham (Photo 8) mean size was larger than at Selsey (Cole, 1980). No statistically significant sorting trends were detected across the profiles other than for size. Although the measurements and analyses were carried out carefully, the spatial extent of sampling was limited and the temporal variability of the beach sediments was not considered in this study. Harlow (1980) sampled beach sediments from 124 sites between Gilkicker Point (Gosport) and Selsey Bill. Surface sediment samples were collected after a storm affecting the shoreline between East Wittering and Selsey in an attempt to describe the form to which the beach is tending in response to dominant waves. He noted that: (a) gravels on both the upper storm beach and the mixed midbeach increase in size from east to west. (b) Medium sand on the midbeach showed limited coarsening westward from Selsey Bill. (c) Fine to medium sand on the lower foreshore showed no sorting trend. It should be noted that the sedimentology of this beach would have been affected considerably by the numerous beach replenishments and beach scraping undertaken for coastal defence

(iii) West Beach, Selsey and Selsey Bill: The beach is generally narrower and steeper than the Bracklesham Bay frontage and is backed by hard defences. It is usually fairly empty of gravel in the vicinity of Hillfield Road (the littoral drift divide), but gravel accumulations increase both eastward and westward from this point (Lewis and Duvivier, 1977). Sand is only of significance at the lower beach fronting Selsey Bill, though it is not a constant feature (HR Wallingford, 1997).

(iv)East Beach, Selsey: The beach has been frequently described as dominantly gravel, and steeper and narrower than West Beach (Duvivier, 1961; Hydraulics Research, 1974; Posford Duvivier, 2001). Immediately east of Selsey Bill, Lewis and Duvivier (1977) reported that the beach was composed of loose gravel sloping steeply down to a gravely pavement near or below LWMST. Further north-east, a variable proportion of sand was an impersistent feature of the lower foreshore. In this vicinity there were exposures of clay and other in-situ strata after winter storms, which tended to be concealed in summer by renewed beach accretion. Mean grain size is reported to increase rapidly northwards for the sandy-gravels of the foreshore, suggesting selective longshore transport, but is fairly constant for the upper beach (Gifford Associated Consultants, 1997). Quantitative analysis of beach sediments has not been undertaken, although Duvivier (1964) noted that Selsey Beach consisted predominantly of rounded flint clasts. Attrition tests were carried out by placing weighed samples in a revolving steel drum and re-weighing and resieving the residue at intervals. This technique established that attrition rate was inversely related to pebble size.

(v) Church Norton and the Pagham Entrance Spits: The beach at Church Norton is composed mostly of flint shingle (Lewis and Duvivier, 1977; Wallace, 1990; Posford Duvivier, 2001). Barcock and Collins (1991) report that Pagham Beach (the north eastern shingle spit) consists of a steep upper beach, a shallower mid-beach, and an extensive low gradient foreshore. This beach is mostly flint gravel, but grades to coarse sand and granules on the lower foreshore. The series of closely-spaced ridges that make up much of the Church Norton and spit beaches may result from the bifurcation of the ebb channel immediately seaward of the harbour entrance. The abandoned channel then fills with gravel, forming a linear bank that subsequently migrates on shore. Repetition of this process creates the pattern of multiple ridges (Barcock and Collins, 1991). Beach morphology along this sector has been considerably modified by nourishment and profile reconstruction.

5.2 Beach Volumes (see introduction to sediment stores)

Volumetric information has been collected by a variety of sources and covers most of the area:

(i) East Head. The total volume of material (sand and gravel) comprising the inter-tidal beaches of East Head has been calculated for a variety of dates using MHW and MLW on successive OS maps (Harlow, 1980). Its volume was 499,400m³ in 1846, increased to 894,000m³ by 1965, and remained stable up to 1975. The total volume of sand comprising the entire spit and dune system exceeds 2.2 million m³ (ABP Research and Consultancy, 2000), but there are no recent calculations of beach volume alone. The volume of gravel on East Head beach was calculated independently by Webber (1978) and Jarosz (1979). Webber measured 7 profiles and estimated volume at 22,000m³, while Jarosz measured 3 profiles and estimated volume at 21,183m³. Direct measurement of sediment thickness (depth) was not employed in either of these investigations thus the volumes quoted are approximate. The similarity of estimates by Jarosz and Webber suggests the quoted volumes may be representative, but the study by Harlow (1980) indicated that inter-tidal sediment volume can vary greatly at East Head according to incident winds and waves. The foreshore of the Winner in front of East Head has eroded and lowered by up to 3m since 1923 (ABP Research and Consultancy, 2000).

(ii) West Wittering to West Beach, Selsey. A series of profiles were measured at monthly intervals along this frontage by Hydraulics Research (1974). Beach volume was calculated above 0.16m OD, a level equivalent to the toe of Medmerry Beach. Calculated beach volumes revealed seasonal variations, with significantly larger volumes in summer. Mean quantities (i.e. inter-annual values) were:

West Wittering: 135,000m³
East Wittering: 167,000m³
Medmerry: 450,000m³
West Beach, Selsey: 30,000m³


Distinction was not made between gravel and sand, for the two were frequently intermixed on the foreshore and their relative proportions at depth were unknown. The temporal representativeness of these volumes was moderate to low because profiles were only measured during a 9 month period, insufficient to include exposure to the full range of wave energy states. Severe winter storms were observed to remove all foreshore sediment and expose the underlying substrate to wave abrasion (Lewis and Duvivier, 1977). However, much of this material was recovered during the following summer so that the operation of seasonal "cut" and "fill" cycles is likely to be an inherent feature of this sector of coastline.

Analysis of beach levels, 1973-1995 (HR Wallingford, 1995 and Gifford Associated Consultants, 1997) using Environment Agency ABMS profiles revealed that the upper gravel beach was narrow at Bracklesham (Photo 8), and did not appear to derive much benefit from up-drift replenishment. Further north-west it was more substantial, although there were site-specific variations in width and height reflecting groyne trapping efficiency. The lower foreshore appeared to have maintained its form, but levels dropped progressively throughout the above period, especially at West Wittering, Bracklesham and Medmerry. High rates of volume loss and profile flattening along the Medmerry frontage, associated with storm waves, have been offset by intensive renourishment and reprofiling (see Section 2.3). This has been especially marked in the late 1990s (Posford Duvivier, 2001).

(iii) Selsey Bill. HR Wallingford (1995), using ABMS extending data back to 1973, identifed a steady lowering of foreshore levels, particularly west of Hillfield Road. At this point, and north-west to West Beach, there is no direct feed from the Kirk Arrow Spit. The narrow depleted strip of upper gravel beach is also a result of wave reflection from backing seawalls. Further east, towards and at Selsey Bill, the shoreline is less exposed to waves from a wide approach sector. This is apparent from the comparatively wide mixed shingle and sand beach, though substantial drawdown occurs when there are incident waves from the south-east. The foreshore close to the Bill (Photo 1) has tended to resist the trend towards depletion, showing some modest accretion in recent years probably due to receipt of influxes of gravel from the Kirk Arrow Spit since 1997 (Posford Duvivier, 2001). HR Wallingford (1995) calculated an overall upper beach gravel loss of 1,000 m³ for 1973 to 1992 prior to the recent gravel influxes. Hydraulics Research (1974) calculated that the beach at Selsey Bill (Hillfield Road) stored some 65,000m³ of sediment; this is a mean value that accounted for seasonal fluctuation.

(iv) East Beach, Selsey. Lewis and Duvivier (1977) note that beach volume is variable spatially and temporally, but estimate that 50,000-55,000m³ of material is permanently retained on the beach. This calculation was undertaken using aerial photographs (1972- 75) and beach morphology observations. Analysis is complicated by the fact that there are significant differences in beach levels north and south of the lifeboat station. Gifford Associated Consultants (1997) calculated a net loss of 40,000m³a^-1, 1972-1992, for the entire sector between Selsey Bill and Pagham Harbour entrance, based on application of the LITPAK numerical model. For East Beach alone, a loss of 6,000m³a^{- 1} was derived from analysis of ABMS beach profiles for the same period (HR Wallingford, 1995). Taking into account the complexities of loss and gain on this beach, especially nearshore and inshore exchanges and the retention of gravel by closely-spaced groynes, annual depletion over this period is considered to be close to 4,000m³a^-1. This, however, is substantially less than the rate of loss between 1900 and 1950, when recession at about 3ma^-1 occurred prior to the construction of defences. During this period, underlying Raised Beach deposits were intermittently exposed, thus adding to sediment supply and helping to offset losses (Posford Duvivier, 2001). Under present hold the line policies this source is no longer available.

A number of specific locations on East Beach, Selsey, involving both upper and lower beaches, have exhibited net depletion and lowering over the most recent 30 years (HR Wallingford, 1995). However, profiles at Inner Owers and Church Norton have been characterised since the early 1900s by net periodic accretion at rates of up to 4,800m³a^-1 (100,000m³ between 1970 and 1994), probably due to sequential "welding" of onshore moving bars (Barcock and Collins, 1991; HR Wallingford, 1995). Depletion, here, and on the Pagham spit, was apparent between 1910 and 1940 and by the mid-1980s; the latter was losing material at a minimum rate of 2,000m³a^-1, necessitating gravel recycling and re-profiling in an attempt to resist beach losses. These trends are a continuation of those measured by Lewis and Duvivier (1977), using aerial photography, for the period 1967-1975. However, small scale spatial variability of net accretion/depletion patterns are apparent, and appear to be most directly related to groyne re-construction, the inshore movement of gravel banks and bars and spatial variation in exposure to wave energy. Terminal scour at the easternmost groyne was evident before the latter was buried following replenishment.

The erosion losses from East Beach need to be placed in an historical context. Excepting locations subject to irregular nourishment, LWM has retreated at a rate of between 0.3 and 3.0 ma^-1 since 1875 (Gifford Associated Consultants, 1997). As HWM is held static by seawalls and embankments along much of this frontage, this has led to a long-term trend of profile steepening. Foreshore width has diminished by up to 650m over the past 125 years at the most rapidly receding locations. The fastest rates of retreat at specific points occurred within a few years of the completion of seawalls, which was undertaken in piecemeal fashion between 1910 and 1969.

Taking the sector from Selsey Bill to Pagham Harbour entrance as a whole, beach budgets are negative, with groyne-assisted accretion and gravel recycling along the sector north of East Beach only partly counteracting overall losses of beach volume further south. Details of site-specific gains and losses, 1973-1992, are given in HR Wallingford (1995).

5.3 Stores: spits and estuarine sediments (see introduction to sediment stores)

The principal stores are the spits at either side of the entrance to Pagham Harbour; East Head Spit, the ebb tidal delta offshore of Pagham Harbour, and the estuarine sediments within Pagham Harbour.

Pagham Harbour Spits

Pagham Harbour (2.83km²) is a product of Holocene sea-level submergence of the former mouth of the river Lavant prior to its diversion in Roman times (Wallace, 1990). However, its present extent is the result of storm surge inundation on 6th December 1910 following complete enclosure and land claim in 1876 (Environment Agency, 1998b). Sitlation of the original larger estuary and progressive, piecemeal, reclamation took place earlier, the latter commencing in the seventeenth century (Graves, 1981; Brown, 1981).

The convergent gravel spits that define the Pagham Harbour entrance channel have behaved in a highly dynamic fashion over at least the past seven centuries (Robinson, 1955; Robinson and Williams, 1983; Barcock and Collins, 1991; Geodata Institute, 1994; Gifford Associated Consultants, 1997; Environment Agency, 1998b; Posford Duvivier, 2001). The earliest reasonably reliable map evidence (1587) suggests that the southern (Church Norton) spit had a configuration similar to the present, possibly in response to one or more breaches dating back to 1340-1410. Between 1672 and 1724, it extended some 90 m northeastwards, with a rapid acceleration in this extension of almost 900 m between 1774 and 1885. Episodes of breaching interrupted the spit extension in 1820, 1829 and 1840. As the spit extended and thinned during this 100 year period, there may not have been a significant supplementary supply from inshore sources (i.e. migratory bars associated with the tidal delta). The main source of sediment feed was probably delivered by littoral drift along the shoreline from the south west. Rapid shore erosion occurring around the Selsey peninsula at this time would have provided a local source of sediment. Over this same period, the northern (Pagham) spit, the product of 'counter drift' determined by a transport divide north-east of the harbour entrance, experienced net erosion and recession. The entrance channel cut into the low clay cliff on the northern side, resulting in 170m of coastline retreat at this point between 1780 and 1840.

Natural change ceased in 1876, when both spits were partially stabilised and the inlet channel closed to effect the final land claim of Pagham Harbour in 1877. Although this helped to create almost 120m of foreshore progradation, a major storm in December 1910 breached the Church Norton spit, creating a 160m wide channel near Church Norton and flooding the newly reclaimed area. Subsequent events have been documented by the sources mentioned above, notably the Environment Agency (1998b), as follows. Further north-eastwards growth of the Church Norton spit narrowed the entrance and deflected it some 700-800m to the NE towards Pagham. The deflected entrance began to threaten bungalows built on the Pagham spit in the 1920s, so a new narrow entrance, close to the 1910 breach, was established artificially in 1937. The pre-1937 entrance gradually closed becoming marked by a low gravel berm. A further breach over a wide front between the 1937 and pre 1937 inlets occurred shortly after 1955 and by 1958 the entrance was some 700m wide. Rapid drift thereafter led to a further extension of the Church Norton spit across the inlet narrowing it to 250m by 1961. Following continued narrowing a stabilisation of the entrance ch