1. Introduction
Poole Bay is a relatively shallow embayment delimited by Poole Harbour tidal inlet (Photo 1) to the southwest and Hengistbury Head/Christchurch Ledge (Photo 2) to the east. It has a very gentle and subtly sectioned evolving quasi-log spiral form, curving slightly more in the west than to the east (Harlow, 2005b; 2012a and b; Royal Haskoning, 2011.) Much of the coastline features formerly rapidly eroding soft cliffs that are now fronted and thus effectively “fossilised” by a substantial seawall/promenade built progressively eastwards, in stages, between 1878 and 1985 (principally 1907 to 1972 along the Bournemouth frontage, except for earlier short lengths of privately built seawalls - refer to Harlow, (2005b; 2012c) for details.) (Photo 3 and Photo 4). Unprotected cliffs occur along the easternmost 2km between Southbourne and Hengistbury Head (Photo 5). Beaches are predominantly dissipative and sandy, coarsening eastwards (to include a backshore gravel component east of Southbourne) in the same net direction as longshore drift along this coastline. Intensive beach management has been practiced for over a century (Lelliott, 1989, Harlow, 2001, 2005b and 2012c). The introduction of seawalls increased wave reflection, and thus the littoral drift rate, the result of which was sustained loss of beach volume; Harlow (2012e) has estimated that a volume of approximately 10 million m³ in the first decade of the twentieth century had declined to 6 and a half million m³ by 1974. Between 1915 and 1974 this issue was addressed exclusively through the building and maintenance of a groyne system to offset the effects of out flanking (lee scour) of the progressively eastward location of the terminal position of the seawall. As this was built in sections, it is now a series of segments, with offsets, an effect that is most pronounced east of Boscombe Pier (detailed in Harlow, 2012c). Details of the construction and maintenance of successive generations of initial concrete and subsequent timber groynes along the Bournemouth frontage are given in Harlow (2012b and c). Since 1974, primary reliance has been placed on periodic beach recharges using imported sand and gravel (where possible from the beneficial use of material dredged from Poole Harbour), though groynes remain critical as means of controlling accretion, drift rates and beach levels. (Harlow, 2001, 2012c and f; Cooper, 1997; Cooper, et al., 2001; May, 1990.)
The solid geology of the cliffs, and the seabed beneath Poole Bay, is composed of rocks of the Tertiary Bracklesham Group, consisting of a sequence of fine, medium and coarse sands (Bristow et al., 1991). At Hengistbury Head there are younger rocks of the Bartonian group, forming an outlier (Photo 5). The Barton Clay here is made up of a series of sands and interbedded clays, with four distinct bands of ironstone nodules. These formations dip eastwards and are cut out by a north-west to south-east trending fault. This defines the eastwards boundary of Beerpan Rocks and Christchurch Ledge, which is a seawards outcropping continuation of the resistant ironstone stratal horizons exposed in Hengistbury Head.
A substantial increase of both qualitative and quantitative knowledge of most aspects of the hydrodynamics and geomorphology of Poole Bay has been achieved in the past 55 years. This is due to the promotion of internal and commissioned research by Bournemouth and Poole Borough Councils and Poole Harbour Commissioners; strategic co-operation between them; central government support of measurement and monitoring programmes both before and after the successive renourishments of beaches; several academic research investigations, and the interest of commercial organisations. Both Poole Harbour Commissioners and the Boroughs of Poole and Bournemouth have invested in computer modelling in an attempt to better understand both wave and tidal conditions and the efficacy of control structures within the Poole Bay littoral margin. (Refer to the Bournemouth Beach Monitoring chapters and the several reports by HR Wallingford.)
A major new source of coastal data is from the DEFRA-funded National Network of Regional Coastal Monitoring Programmes. The Programmes consist of repeated topographic beach and nearshore bathymetry surveys, aerial photographic and lidar coverage, coastal hydrodynamics (waves and tides) monitoring 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.channelcoast.org. The Southeast Regional Coastal Monitoring Programme commenced in 2002. The Lead Authority is New Forest District Council, with data collection, analysis and reporting led by specialist teams at the Channel Coastal Observatory (CCO), Canterbury City Council and Adur and Worthing Councils. (See CCO Annual Survey Reports for further details). In 2012, an extensive high resolution, 100% coverage swath bathymetry dataset was collected in Poole Bay, by the Maritime and Coastguard Agency’s Civil Hydrography Programme. However, this survey did not include the area at Poole harbour entrance and approach channel, which are within Harbour Authority limits.
1.1 Coastal Evolution
Pleistocene evolution was dominated by the Solent River which flowed across the floors of Poole and Christchurch Bays and eastwards through the Solent and Spithead (Everard 1954, Bray, 2000; Velegrakis, et al., 1999). During the fluctuating sea levels of this period the Solent River and its tributaries deposited a descending sequence of gravel filled channels and terraces which mantle the predominantly sandy Eocene bedrock (Nicholls 1987; Bristow et al., 1991; Allen and Gibbard, 1993). Remnants of these fluviatile deposits occur, up to 8m in thickness, at the summit of the modern cliffline; they have also been mapped in the offshore areas of Christchurch and Poole Bays (Fitzpatrick 1987; Velegrakis, 1994). A critical factor in the creation of Poole Bay was the breaching of the Chalk ridge, which previously extended between the Needles and Handfast Point. Opinion is divided as to the date of this event. Until recently, the prevailing view favoured breaching in the early Holocene (Everard 1954, Keen 1975, Jones 1981). Interpretation of a buried channel in Poole Harbour (Devoy 1972), studies of Chalk erosion rates, offshore buried channels and the depth, relief and inclination of the plantation surface which truncates the Purbeck-Wight ridge now indicate an early to mid-Devensian breach (Wright 1982, Lacey 1985, Nicholls 1987; Velegrakis, et al., 1999; Tyhurst and Hinton, 2000; Nowell, 2000). It is thought that much of Poole Bay was opened up at this time, causing the upper Solent River (proto-Frome) river to change direction and flow south across the present area of Poole Bay (including the Chalk outcrop) as sea-levels fell during the Devensian (Wright 1982; Velegrakis, et al., 1999).
Rising sea-levels of the Holocene transgression caused rapid erosion of the soft Tertiary strata and much sediment redistribution over the floor of Poole Bay, as it assumed its present planform (Devoy, 1982). Velgrakis (1994) and Velegrakis, et al. (1999) provide a detailed stratigraphy of the sediment infill of three buried channels that are incised into the Chalk outcrop and extend further seawards. This consists of basal fluvial material, passing upwards into estuarine sediments. This sequence is a clear record of early Holocene marine invasion, creating an enclosed bay head. The simultaneous erosion of Hengistbury Head/Christchurch Ledge initiated the equally rapid, but later, erosion of Christchurch Bay (Wright 1982; Nicholls and Webber 1987a; Velegrakis, 1994; Bray, 2000; Velegrakis, et al., 1999). Further details are given in the section on Quaternary Evolution of the Solent.
Both Poole (partially) and Christchurch (fully) bays have a log spiral/zeta form planshape, but concepts of crenulate bay evolution need to recognise that Poole Bay pre-dates Christchurch Bay. Poole Bay has a quasi-stable condition fixed by the "anchoring" effect of Hengistbury Head. Accelerated erosion of this headland would induce major instability and a return to an earlier unstable configuration characterised by rapid coast erosion. (Refer to Royal Haskoning (2011) for further discussion.)
Marine erosion of Poole Bay released large quantities of sediment, some of which must now be represented by thick, discrete sediment accumulations in the western part of the bay, such as Hook Sand and in the offshore area of Studland Bay. Eastward transport into Christchurch Bay may also have occurred as the precursor to Hengistbury Head retreated and diminished (Nicholls 1985, Nicholls and Webber 1987; Halcrow, 1999). Significant quantities of sediment may also have been lost from the bay in the early stages of its erosion, although the present whereabouts of such material is conjectural and may now have been lost to the regional sediment budget.
1.2 Hydrodynamic Regime
Wave and tidal currents are the dominant sediment transport mechanisms. The wave climate varies spatially due to the sheltering effect of Handfast Point and the "Isle" of Purbeck. Halcrow (1999) and HR Wallingford (2003) assert that the plan shape of Poole Bay is adjusted to the directions of approach of swell waves. Prevailing wave direction offshore is from the south-west which also coincides with the longest fetch. Waves from this sector cannot directly enter Poole Bay, but are refracted and diffracted so as to approach from the south and south-east. The degree of shelter afforded to such waves therefore increases westwards in the lee of Handfast Point. Nearshore wave climate is characterised by transformed swell waves and waves generated by local fetches to the south and southeast. Attenuation occurs due to the presence of trough and bar topography, as well as the shallowness of Poole Bay.
The Southeast Regional Coastal Monitoring Programme has measured nearshore waves using a Datawell Directional Waverider buoy deployed 500m seawards of Boscombe Pier in 12m CD water depth. Between 2003 and 2012 the prevailing wave direction was south-by-west, with an average 10% significant wave height exceedance of 1.03m (Royal Haskoning, 2011; CCO, 2012). Harlow (2012g) provides detailed graphical presentation and statistical analysis of the Boscombe waverider record, between July 2003 and March 2011. The maximum wave height recorded was 4.5m in March 2008, an event immediately followed by the largest swell. All swell waves approached nearly normal to the shoreline, from due south (after refraction). Analysis of diurnal variations in wave height revealed the influence of reversing sea breezes, with onshore (daytime) breezes raising wave heights and the reverse during night time hours. The diurnal pattern of wave periods shows a peak around midnight, the reason for this being less clear.
Several earlier detailed wave climate studies have been undertaken in Poole Bay. A wave rider buoy was operated off Southbourne between 1974 and 1978 and recorded a maximum wave height of 8.8m, with a computed 1 in 10 year return period. The mean significant wave height (Hs) of 2m was exceeded 2% of the time, with a maximum (predicted) significant wave height of 3.9 m recorded twice during this period (Henderson and Webber, 1977; 1978; Henderson, 1979, 1980; Halcrow, 1980). In another study, using Portland wind data (1974-1984) wave hindcasting was employed to determine an offshore wave climate for Solent Beach; this indicated that Hs of 2m was exceeded 12% of the time, and Hs of 1m 39% of the time (Hydraulics Research, 1986). An appropriate wave refraction and shoaling model was then used to transform offshore conditions onshore, with accuracy improved by calibration against the measured wave buoy data. The inshore wave climate was used to determine extreme wave heights using the Weibull distribution which revealed a significant wave height of 3.9m with a one year, and 5.3m for a 20 year, return period (Hydraulics Research, 1986). A similar approach, also based on Portland wind data (1974-1990) was used to determine the offshore wave climate in the western part of Poole Bay (Hydraulics Research, 1991b). Attention was focused on an inshore point to the east of Hook Sand and it was found that waves from the southeast quadrant dominated (Hydraulics Research, 1991b). Approximately 4% of waves exceeded 2m and 1% exceeded 3m in height.
HR Wallingford (1995a; 2003) undertook a numerical modelling study of the wave climate of western Poole Bay. Swell waves were not included, but it was determined that 44% of waves generated over local fetches approach from the south-west, and 28% from the east/south-east. Hs values were rarely above 1.0m. A simplistic wave refraction analysis indicated wave energy convergence at (i) Southbourne; (ii) the central Bournemouth frontage, and (iii) Hook Sand, and other smaller sandbanks north and south of the Swash Channel. Earlier work by Henderson (1979) developed a series of computer-generated wave refraction diagrams for outer Poole Bay which concluded that longer period waves were most significantly modified by refraction, set up by Christchurch Ledge, Beerpan Rocks, Dolphin Bank and Dolphin Sand. Maximum wave focusing in this analysis was found to occur offshore Hengistbury Head (Henderson and Webber, 1979; 1980).
Brampton, et al. (1998) employed the UK Meteorological Office (UKMO) fine mesh wave model, using data for January 1991 to December 1996, to create an estimation of the wave climate for both the offshore and inshore zones of Poole Bay. In this case, swell waves were included, and it was determined that in the offshore area where water depths exceed 6m, swell waves with a one year return period have characteristic Hs values of 2.5 to 3.0m. The effect of shoaling and refraction reduced these to 1.0 to 2.0m in the inshore zone, with maximum energy focusing along the frontage of central Bournemouth.
A similar approach was used by Halcrow (1999), also based on the UKMO model but employing a 15 year continuous record of wind speed and direction for validation. Wave height and approach were determined for an offshore "node" in Poole Bay, and it was concluded that refracted waves approaching from the west and south-west affected the inshore environment of Poole Bay for 85% of the time. Although this is the prevalent wave type, dominant waves from the east and south-east affect all sectors of this coastline. For the offshore area, maximum Hs values were calculated to vary between 5.5m (1 year return period) and 7.4m (50 year recurrence) for a location off Southbourne. Mean inshore Hs values were also determined for unbroken waves, approaching from all directions. These are greatest at Southbourne, declining rapidly westwards towards Sandbanks, and reducing slightly at the Long Groyne. CH2M have completed analysis of the regional wave climate as part of the Poole Bay Coastal Defence Strategy Study 2012.
HR Wallingford (1999; 2003) report the results of an independent analysis of the offshore wave climate using the HINDWAVE numerical prediction and TELURAY wave refraction modules. These integrate the effects of refraction and shoaling. Mean Hs values are computed for several approach directions, wave periods and return frequencies. For locations offshore Poole Head, Bournemouth Pier and Southbourne, mean Hs for annual recurrence varies between 2.16m (155°) and 1.01m (245°); 2.46m (155°) and 2.80m (245°); and 2.26 (155°) and 3.61 (245°) respectively. Similar orders of difference are computed for a 1 in 10 year recurrence, with the highest value being 4.88m for waves approaching Southbourne from 245°.
These studies demonstrate the effect of increasing shelter in the west of Poole Bay as waves with a dominant south to south-west approach in this area are less severe. By contrast, conditions along the eastern sector are more energetic, with waves having a dominant south, south-east and south-south-east approach. Wave rays indicate that energy is focused on Hengistbury Head. These differences affect the directions and rates of longshore sediment transport and the nature of beach material distribution (Sections 3 and 5).
Tidal range is low, approximately 2.0m during spring cycles and 1.0m during neaps. A weak "double high water" component occurs, mean range increasing westwards. These characteristics concentrate wave action into a relatively narrow height range. Bed stresses created by tidal currents alone are not usually sufficient to entrain sand, but they act in conjunction with waves to transport material both onshore and offshore. The most rapid currents are in (i) the extreme western part of the bay, where the peak ebb flow at Poole Harbour mouth and in the East Looe Channel approaches 2.5ms-¹, and (ii) offshore the Long Groyne at Hengistbury Head and in the area of Christchurch Ledge. Tidal flow in the west is deflected south by prevailing south-easterly currents and generates currents up to 1.0ms-¹ off Handfast Point (BP 1991; Riley, et al., 1994). Tidal flow has been studied in the western part of the bay by mathematical models (Hydraulics Research, 1986, 1988, 1991b; HR Wallingford, 1995a; 2003). These have been validated against data obtained by Ocean Surface Current Radar (OSCR) and direct depth-averaged tidal current measurements from three locations (BP 1991, Hydraulics Research, 1991b). Peak near and offshore velocities are highest in deeper water, but are generally less than 0.6ms-¹, with values of 0.15-0.25ms-¹ in the vicinity of Bournemouth and Boscombe piers (Riley, et al., 1994; HR Wallingford, 1995a). These figures relate to spring tides. Ebb flow is predominantly westward off Hengistbury Head and south or south westward further west. Flood flow is north or north-east in the west, becoming eastward in direction towards Hengistbury Head. Southward and south-westward residual flow occurs in the west out of Poole Bay (Hydraulics Research, 1991b; Brampton, et al., 1998). There is also an eastwards residual flow towards Christchurch Ledge, with velocities progressively higher with distance offshore (Brampton, et al., 1998; Osborne, 1991; HR Wallingford, 2003). Current metering of surface and sub-surface residual currents in the offshore zone has indicated that speeds decrease from the surface towards the seabed, with an imposed anticlockwise rotation (Osborne, 1991). In the offshore zone, tidal streams flow approximately parallel to the shoreline.
Brampton, et al. (1998) used the TELEMAC two-dimensional numerical model to simulate tidal current vectors and velocities. They demonstrated pathways of sand transport, for median grain sizes of 200þm and 100þm, consistent with direction of residual tidal currents in the offshore zone. For inner Poole Bay, it was concluded that tidal current velocities alone are below sediment entrainment thresholds, and that stresses imparted by shoaling and breaking waves are a necessary auxiliary to tidal currents to effect significant sand transport. Waves become the dominant force in the inshore zone, as demonstrated by the SANDFLOW numerical model. Halcrow (1999) confirmed this general pattern except in the entrance channel to Poole Harbour, where ebb tidal current velocities in the Swash Channel - particularly along its eastern boundary - are higher than elsewhere in western Poole Bay. However, even here, it is wave and tide generated stresses working together that account for sand mobility (Hydraulics Research, 1986; Brampton, et al., 1998). Transport rates have been observed to increase substantially under the action of storm waves.
Harlow (2004) provides a summary analysis of data obtained from the Class 1 tide gauge installed at Bournemouth Pier in June 1996. The extreme high tidal level recorded was 2.6m CD, whilst the extreme low was 0.3m CD. Mean sea-level rose 9.3mm per year, 1996 to 2003, reflecting an increase in the height of maximum surges during this period. Pirazzoli et al. (2006) state that the maximum recorded surge, 1996 t0 2002, was 100cm CD, whilst the maximum combined astronomical tide and surge was 280cm CD. Royal Haskoning (2011) provide data on the ten highest and lowest (i.e. positive and negative) surges recorded at Bournemouth Pier, 1996-2007.
2. Sediment Inputs
2.1 Marine (off to Onshore) Input
Two categories of marine input are recognised, comprising: (i) input of fresh sediment external to the Poole Bay system; and (ii) sediment feed to beaches and Poole Harbour entrance from existing nearshore/offshore stores within the Poole Bay system.
This feature can be interpreted as a part of the ebb tidal delta of Poole Harbour entrance composed of sediments from Poole Bay, and it functions as a significant sediment store, supplying sediment to shore-normal and alongshore transport pathways in the nearshore zones adjacent to Sandbanks and Studland peninsulas. Over its northern area, in the East Looe Channel, tidal currents are capable of moving fine sand. Part of the crest of Hook Sand lies above -1mOD causing refracted waves to break. It is suggested that this causes: (i) sand to be driven onshore (north-westwards) from the crest (HR Wallingford, 2000; 2003), and (ii) some sediment to move offshore in the shallows of Poole Head and then southwards along the east side of the bank (Hydraulics Research, 1986; 1991a).
However, refraction and shoaling models based on a 10-year hindcast offshore wave climate indicated a high potential for net southward transport of both suspended and bedload sediments in excess of 25,000m³ per year. Sand supplied by this pathway may periodically partially infill the Swash Channel and/or be transported further south to Poole Bar, although there is no clear evidence for this (Brampton, et al., 1998; HR Wallingford, 2003).
Examination of the pattern of sediment accretion against groynes first constructed at Sandbanks between 1896 and 1906 to prevent potential breaching of the peninsula revealed a possible onshore feed from Hook Sand (Robinson, 1955). Evidence of this process is provided by examination of historic maps and charts covering the period 1785-1953. These showed a tendency for offshore bars to form between Poole Head and Sandbanks and it is envisaged that their onshore migration could supply sand to the beach (Poole Harbour Commissioners, 1995-6; Royal Haskoning, 2011). Anecdotal evidence is available for onshore feed of sand from Hook Sands. The Sandbanks frontage adjacent to Hook Sands has never been replenished but beach accretion occurred during the period of groyne construction.
Post-1991 beach erosion induced by a sequence of south-easterly storms has tended to support HR Wallingford’s (1991b) analysis that indicated an excess of transport potential along the Sandbanks peninsula. Subsequent studies by HR Wallingford (1994, 1995a and b and 2000) confirmed the occurrence of net eastward drift over recent decades that might have been expected to have completely depleted the beaches and partly eroded the peninsula. That such severe erosion did not occur led HR Wallingford (2000) to infer that the beaches must have been at least partly sustained by onshore transport of sand from the direction of Hook Sand. Based upon the results of these studies, rock groynes and a rock revetment were constructed in 1995 (Photo 1) and 2001 (Photo 7) to control potential erosion of the beach. In addition, 1200m² of vegetated dunes were created to provide additional defence against potential beach overtopping.
2.2 Fluvial Input
No rivers of any significance discharge directly into Poole Bay from this sector of coastline. The river Bourne has discharged via a culvert since the late nineteenth century. The several deeply-dissected chines that interrupt the cliffline west of Bournemouth Pier do occasionally discharge run-off and sediment following prolonged or intense rainfall. Harlow (2001) quotes the example of Middle Chine, which discharged approximately 500m³ of sediment onto the beach in 1991 following 75mm of rainfall in 3 hours. Boscombe Chine has also discharged several inputs in excess of 3-500m³ over recent decades.
2.3 Coast Erosion
» E1
The coast between Poole Harbour entrance and Hengistbury Head was previously subject to continuous erosion throughout the Holocene, resulting in development of steep cliffs between 4m and 36m in height and thus supply of sand and gravel to the beach. Harlow (2012a) estimates that over the past 9,000 years, the rate of recession was approximately 1.0m per year, yielding each year some 125,000m³ per year of sediment. This situation was altered from the 1890s onwards, starting with construction of coast protection structures at Sandbanks. Schemes involving further protection by seawalls and groynes followed. The last section of sea wall along the Poole BC shoreline was completed in 1985 at Canford Cliffs. Groyne rebuilding and other controlling measures continues to date. The Bournemouth frontage from Poole Head to Solent Road was progressively protected by a seawall promenade between 1907 and 1972 (Lacey, 1985; Lelliott, 1989; Harlow, 2012b.) These measures have involved protecting the toe (Photo 3), local cutback, slope grading and stabilisation to 35° of the previously freely eroding cliffs (Photo 3), as well as drainage, including sand drains, (Photo 8) and planting. (Harlow (2012a) does not consider the last measure to be an effective erosion control.) Thus the supply of sediment to adjacent beaches was progressively reduced as protection spread eastward along the frontage (Wigmore, 1951; Bournemouth Borough Council, 1991c; Harlow, 2012a.) Map comparisons over the period 1867-1933 indicated cliff retreat at 0.15-0.40m per year between Poole Head and Bournemouth Pier, and 0.20-0.50m per year between Bournemouth and Southbourne before exclusion of marine erosion (Lacey 1985).
Cliff retreat rates were combined with details of cliff height and sediment composition to determine past rates of cliff sediment supply. The analysis for 1867-1933 calculated a total supply of 115,000m³ per year, of which 91,000m³ per year was sufficiently coarse to remain on the beach. Similar analysis for 1933-1967, taking into account the extension of the protected frontage during this period revealed quantities of 77,000m³ per year and 66,000m³ per year respectively (Lacey, 1985). Gao and Collins (1994) undertook an independent analysis of cliff erosion yield prior to protection, and concluded that yield was 136,000m³ per year, of which 46% (approximately 100,000m³) was stable on adjacent beaches. Contemporary supply is restricted to the 2.5km length of eroding cliffs between Solent Road and Hengistbury Head and is estimated at between 4,000 and 8,000m³ per year (Lacey, 1985; Royal Haskoning, 2011) of which 1,400m³ per year is gravel (Bray, 1993; Harlow, 2001). Coast protection has thus had a progressive and dramatic effect in reducing sediment supply from coast erosion and this is widely cited as a major cause of the marked reduction in both drift rates and volumes and beach levels recorded over the past 80 years (Lelliott, 1989; Harlow and Cooper, 1994; Posford Duvivier, 1998, 1999; Harlow, 2001, 2012a, b).
Despite cliff stabilisation by slope regrading and planting of grasses, shrubs etc., sub-aerial processes of weathering and mass movement continue to operate along the Bournemouth frontage (Bournemouth Borough Council, 1991a, b and c; Harlow, 2012a). These include wind erosion, surface wash and gulleying, the latter promoted by groundwater seepage where there are clay and silt horizons between sandstones. Harlow (2001) estimates an ongoing rate of 1cm per year of cliff top retreat for the frontage between Branksome Chine and Southbourne, giving a yield of 15m³ per year, although this negligible volume of material does not contribute to the foreshore as it is either retained on slope or landward behind the promenade retaining wall. All of this accumulates at the junction between the cliff foot and the promenade, and is periodically removed. The cliffs immediately east and west of Alum Chine have not been re-modelled, and are subject to gulleying and periodic small scale failures induced by critical pore pressures in gravels and sands overlying clays (Halcrow, 1999). Another site of past failure occurs east of the Toft 'Zig-Zag' path.
Two morphological units are recognised along this coast. Low cliffs composed of Valley Gravels with overlying blown sand rise eastward from Solent Road to attain 4.5mOD height at Double Dykes (Photo 9). Predominantly sandy cliffs at Hengistbury Head are composed of Barton Clay, Hengistbury Beds and overlying Boscombe Sands capped by Plateau Gravels (Photo 5). These cliffs rise to 36mOD and contain ironstone nodules and thin pebble bands (May, 1971; Lacey, 1985; Bray, 1993; Halcrow, 1999; Daley and Balson, 2000). The upper cliff (75°) is distinct from the lower cliff facet (45-55°), and both decline in gradient towards the headland tip. Recession has been studied using a variety of sources and timescales including OS maps (1840-1960) and Bournemouth Borough Council surveys (1923-1986). These surveys reveal cliff-top erosion at the following rates for the period 1840-1986: 0.75 to 1m per year at Solent Road; 1.12 to 1.5m per year at Double Dykes and between 0.50 and 3.5m per year at Hengistbury Head (Wigmore, 1951; May, 1966 and 1971; Hydraulics Research, 1986; Parker and Thompson, 1988; Turner, 1998; Royal Haskoning, 2011). These are mean values, as losses of up to 10m have occurred as a result of single failures, with negligible change during both preceding and subsequent decades. Cliff foot retreat rates were everywhere significantly lower, e.g. an average of 0.15m per year (yielding 5,000m³ per year) along the sector 500m east of Double Dykes, 1980-2003 (Halcrow, 2004.) Only approximately 10% of sediment delivered here by erosion is stable on the beach, and therefore available to longshore drift (Harlow, 2012b); the remainder is presumed to be lost offshore.
A major contributory factor to former rapid erosion at Hengistbury Head was the foreshore and nearshore mining of ironstone boulders over the period 1848-1870 (Paris, 1954; Tyhurst, 1985a, Hydraulics Research, 1986; Bray, 1993; Turner, 1998; Royal Haskoning, 2011). This involved direct removal of material providing cliff toe protection, and also facilitated an increased volume of littoral drift; both factors causing accelerated beach depletion. Since about 1880, rates of cliff toe erosion have declined at Hengistbury due to cessation of ironstone mining, construction of the Long Groyne in 1938 (Photo 10) and World War II anti-invasion works. These measures all promoted beach accretion and stability at, and west of the tip of the headland (Photo 2), which has continued up to the present (Hydraulics Research, 1986; Bray, 1993; Turner, 1998). Halcrow (2004) concluded from analysis of ortho-rectified air photo cover that there had been no measurable cliff retreat, 1978-2003, along the sector 500m west of the Long Groyne. However, the beneficial (protective) effects of the Long Groyne have not extended as far west as Double Dykes or Solent Beach, and the low cliffs here suffered continuing erosion at up to 1m per year over the period 1933-67 (Lacey 1985) and 0.4-1.1m per year over the period 1976-1986 (Hydraulics Research, 1986). This situation is directly related to beach sediment starvation resulting from progressive eastward construction of groynes and seawalls along the updrift coastline. These measures both prevent sediment input from cliff erosion and intercept some eastward drift of sediment.
Contemporary cliff top recession at Hengistbury Head is measured at 0.1m per year, mostly comprising losses from gulleying, small mudflows, rockfalls and physical weathering (Photo 5). Analysis of Coastal Monitoring Programme 2006-11 lidar and aerial photography datasets indicates continued cliff erosion eastward of Double Dykes, with accumulations of in the order of 75m³ of cliff-derived sands and gravels retained as a talus store overlying the upper beach; approximately 10% of the eroded material. This indicative rate of significantly less than 1,000m³ per year is a reduction from the estimated 2004 rate of 3-10,000m³ per year. Parker and Thompson (1988) estimate that sub-aerial weathering removes some 780m³ per year from the headland cliffline, west to Double Dykes.
Rate of cliff erosion, used in conjunction with cliff height and lithology along the unprotected cliffline (1360m in length) enabled calculation of cliff sediment input (Lacey, 1985). This analysis indicated an input (>0.08mm diameter) of 26,000m³ per year for 1867-1933, 14,000m³ per year for 1933-67 and 4,000m³ per year for 1984. Supply has declined due to slower erosion of Hengistbury Head due to increasing toe protection afforded by the beach accreting against the Long Groyne, protection of Solent Beach by groynes and beach replenishment (see Section 2.4) as well as declining relief landward as the cliffs at Double Dykes have retreated. Cliff foot surveys indicated erosion of 0.9m per year over the period 1974-82 at Solent Beach/Double Dykes, sufficient to supply 750m³ per year of gravel (Lacey, 1985). Royal Haskoning (2011) estimate that there is a potential sediment supply of 8,000m³ per year from this section. Wright (1986, 1982) asserted that gravel on the upper beach between Solent Beach and the Long Groyne was largely derived from local cliff erosion. Naturally occurring and managed dunes close to the Long Groyne, which form an element of local cliff conservation management (Turner, 1998; Bray and Hooke, 1998), partially obscure coarse sediment forming the upper beach.
2.4 Beach Replenishment
» N1, N2 · N3
Within the Borough of Poole Council’s frontage the completed beach replenishments schemes between Sandbanks and Branksome Dene Chine include:
- In 2003 using dredged material from the Poole Swashway, 105,000m³ of sand was placed between Sandbanks and Shore Road.
- In 2005 to 2006 using beneficial dredged material from Poole Harbour Commissioners’ capital dredging to deepen the main shipping channel into Poole Harbour, Poole, Bournemouth and Swanage beaches were replenished. 450,000m³ of sand was placed on Poole’s beaches between Sandbanks and Branksome Dene Chine, some of which initially fed the far western sector of Bournemouth beach. Further information on scheme details can be found at www.poolebay.net/PhaseI/poole.htm
N1, N2, Sandbanks to Southbourne
Four phases of nourishment (locally known as Beach Improvement Schemes (BIS) have been undertaken by Bournemouth Borough Council, between Alum Chine and Southbourne as follows:
- BIS1 - a pilot project in 1970 involving 84,000m³ of dredged sand from Pot Bank was placed at MLW 200m offshore west of Bournemouth Pier. Some was pumped ashore; whilst monitoring revealed that much of the remainder migrated onshore naturally in succeeding months (Wilmington, 1982; Lacey, 1985; Harlow and Cooper, 1994; Cooper and Harlow, 1996 and 1998; Cooper, 1997 and 1998).
- BIS2 - a major scheme involving 1.4 million m³ marine dredged sand along an 8.5km frontage over the period July 1974 - July 1975. 654,000m³ was pumped onto the inter-tidal beach and 750,000m³ was placed in nearshore dumpsites: much of this latter quantity subsequently migrated over 5 years onto the foreshore. The borrow source was from 8km offshore Boscombe Pier, involving sand similar in median grain size and grain size distribution to the indigenous beach material. High initial losses were recorded, probably due to overfilling of groyne bays (Harlow, 2012c.)
- BIS3 - a further two part nourishment scheme involving: (a) 990,000m³ sand, and some gravel, dredged from the Swash Channel approach to Poole Harbour in 1988, pumped onto the beach and allowed to form its own profile. This occurred in three phases between Alum Chine to Bournemouth Pier and Boscombe Pier to Southbourne (Turner, 1994; Cooper, 2004). 45% of material was placed below low water mark during the first phase, with all of the rest pumped onto the inter-tidal beach, to a crest level of 2mAOD. (b) 143,000m³ of marine dredged gravel placed on Solent Beach, 1988/9 (see below) (Photo 6).
- BIS4 – a five-part nourishment scheme commencing in 2006 involving two major sand nourishments using sediment dredged from Poole Harbour entrance and Aggregate Licence Area 451 (south-east of the Isle of Wight) respectively, and followed by three annual top-ups in 2008, 2009 and 2010. Total volume of sand introduced was 1,627,000m³. The recharged beach profile crest level was 2m AOD, allowing for the development of a more “natural” profile than was achieved with BIS3 (Harlow, 2012c).
Intensive monitoring of the Alum Chine - Hengistbury Head shoreline has been undertaken since July 1974, using topographically surveyed beach profiles, hydrographic survey out to 450m seawards of mean low water and systematic sediment sampling. Detailed evaluations of the accuracy, limitations and possible margins of error of the techniques used are given by Hodder (1986), Cooper (1997) and Harlow (2001, 2012 a to g). The first post-nourishment survey in 1976 revealed an intertidal gain of 745,000m³, although only 650,000m³ was actually pumped onto the beach. This gain was due to sand transported independently onshore by wave action during or shortly after the nourishment operation (Webber, 1980; Lacey, 1985; Harlow and Cooper, 1996; Harlow, 2001). Subsequent surveys showed substantial volume loss from the intertidal zone, which was attributed in part to over-supply as well as to rapid eastward littoral drift (the initial nourished beach overtopped pre-existing groynes), transport of finer materials to the nearshore zone and profile adjustment. It has been suggested that the 1 in 10 slope of the face of the nourished beach was not compatible with the wave climate in the eastern part of Poole Bay (Newman, 1978; Lacey, 1985), although the slightly smaller size of imported (compared to indigenous) sediment may also have been a factor (Hodder 1986). Intertidal losses were 287,000m³ per year in 1976 and 1977, but diminished within two years of nourishment to a mean of 26,100m³ per year over the next 9 years (95,000m³ per year between 1979 and 1981) suggesting eventual attainment of an equilibrium between beach profile, sediment size and wave climate (Lacey,1985; Harlow and Cooper, 1996; Cooper, 1997). A factor, which could also have contributed to volume loss, is compaction of the beach subsequent to nourishment. This does not involve actual loss of material but is impossible to calculate retrospectively without packing density measurements from the freshly nourished beach (Harlow, 2001; Cooper, 1997).
A different view of post-nourishment beach behaviour was obtained by detailed volumetric calculations for the sediment prism extending between mean low water and a distance 300m offshore (Hodder 1986). This analysis revealed that total volume was 1 million m³ greater in October 1978 than in July 1974, and 350,000m³ greater than in July 1975. It is therefore inferred that substantial offshore and longshore losses of 270,000m³ per year subsequent to nourishment were offset by wave driven onshore feed of 620,000m³ from the nearshore dump-sites. Losses from the intertidal zone were therefore not necessarily losses to the beach system, for much sediment was retained in the nearshore zone, contributing up to a 1m rise in seabed levels and thus facilitating dissipation of wave energy (Hodder, 1986; Gao and Collins, 1994; Harlow, 2001). Spatial analysis of these volumetric trends indicated that nourishment was most effective between Bournemouth and Boscombe Piers, where imported sediment was of a slightly larger grain size and a high proportion remained within the intertidal and nearshore zones. Indigenous beach sediment size increases eastward (Lacey 1985) and thus imported sediment became progressively smaller than the "background" in this direction. Losses from the nourished frontage were shown to benefit downdrift beaches, as the Solent Road - Hengistbury Head segment had accreted to 100,000m³ above its July 1974 volume by 1979 (Hodder 1986).
Overall beach volumes along the nourished segment remained at 550,000m³ above pre-nourishment values by November 1982 (0-300m offshore), with an annual loss of 95,000m³ in 1979, 1980 and 1981 following exhaustion of inputs from nearshore supply sites. Intertidal volumes had returned to near 1974 values at most locations by the mid-1980s (Hodder, 1986; May, 1990).
The reliability of the volumetric analysis was evaluated by Hodder (1986) and it was found that some inaccuracy resulted from survey error, spacing error (failure of profile interval to adequately represent intervening beach segments) and seasonal profile variations. Overall confidence of volumetric information for the 0-300m zone was estimated at +/- 131,000m³ for the nourished beach, as a whole. Volumetric accuracy could be increased by undertaking representative measurements of packing density immediately after nourishment so as to calibrate volumetric analysis for subsequent compaction effects.
BIS3: post-completion beach behaviour
Routine monitoring involved continued measurement of the same 50 cross-sectional profiles for beach and offshore survey and sediment sampling as for the earlier nourishment. Analysis of beach volume change employed a "rolling average" of five successive surveys to minimise the impact of survey errors (Harlow and Cooper, 1996; Harlow, 2001; Cooper, 1997). Results revealed that 400,000m³ was lost between 1990 and 1993 (Harlow and Cooper, 1994). In 1994, loss of recharge material was 92,000m³, and by 2000 very little of the material added 12 years previously was still retained (Harlow, 2001). The overall behaviour of the replenished beach was comparable to that following the earlier recharge. Initial rapid loss was due in part to over-filling of groyne bays, which therefore under-performed immediately following re-nourishment. Cooper (1997) states that the highest rates of post-nourishment decline were on the frontage between Boscombe Pier and the end of the promenade at Southbourne. This coincides with a (then) larger groyne spacing to length ratio in comparison to updrift beaches, but also correlates with the steepest beach profiles and coarser background sediments. This is understood as a response to the progressive west to east increase in exposure to wave energy that causes an acceleration of drift and possibly offshore winnowing of the finer grain sizes. Steepening also appeared to be positively linked to grain size. Harlow (2001; 2012d) notes that monitoring revealed gradual change of grain size distribution over time, with the pattern of eastwards coarsening becoming more marked. He states that a grain size of 1.8 phi is critical, with smaller particles more likely to move seawards to form nearshore bars.
Cooper (1997, 1998); Harlow and Cooper (1994; 1996); Harlow (2000; 2001; 2012a, b and d ), Cooper and Harlow (1998) have undertaken critical comparative analysis of both sand recharge schemes, and have observed the following shared behaviour trends:
- Initially rapid profile adjustment and volume losses, especially fine material lost by suspension, followed by 'peak' volumes up to four years following nourishment as material placed offshore or nearshore is moved into the inter-tidal zone.
- Thereafter, several successive years of relatively low volume losses, close to what might be considered the "background" erosion rate defined by the deficit of input into the littoral drift transport system.
Cooper (1998) applied two different empirical numerical models to the evidence for volumetric loss, both of which assume post peak reduction to be at an exponential rate. Data for volume decline following the 1988/89 recharge provided a good fit for the Verhagen model, as the latter includes a linear component that accommodates "background" erosion. Extrapolation of the rate of net volume decline was found to intercept the critical baseline defining a perceived need for further replenishment in 2003. This was derived by multiplying the active width of the beach "envelope" (450m) by the frontage length and an assumed rate of sea level rise (6mm per year). Harlow (2001), however, remarks that as beach levels had returned to their pre-nourishment (1988/9) levels by 2000, the calculation of critical baseline should also include an exponent of wave climate. Storm frequency might be the most appropriate.
This work thus indicates that each of the two major sand replenishments BIS2 and BIS3 had an effective 'life' of 12-14 years. "Borrow" material accounted for approximately 30% of net beach volume in 1997 (Cooper, 1997) A well designed and efficient groyne field has proved essential to retaining recharge materials for an extended period (Cooper, 1997; Harlow, 2001), thus demonstrating that nourishment alone is not sufficient when there is steady or periodically rapid littoral drift. Harlow and Cooper (1994; 1996) have asserted that differences in the particle size distribution of the borrow material used in each replenishment have not proved critical to beach stability, although comments by Hodder (1986) relating to the effects of compaction on initial volume losses are valid.
Harlow (2012c) estimates that the rate of volume loss from BIS4 has been approximately 70,000m³ per year. Potential loss of sand by deflation (between 10 to 20,000m³ per year) is recovered by recycling where it is deposited on the promenade.
Further details on beach morphodynamics are given in Section 5. Analysis of profile and volume changes post 1974 indicates the strong influence of beach renourishment.
N3 1988/89 Nourishment of Solent Beach
A gravel nourishment scheme involving 148,000m³ of dredged material was undertaken on Solent Beach in 1988/9 (Photo 6) preceded by construction of three rock groynes and a gabion revetment at Double Dykes in 1986/87 (Harlow, 1989; Cooper, 1997; Cooper and Harlow 1994; 1998; Harlow and Cooper 1996). A major reason for maintaining beach levels here is the potential threat of a breach at Double Dykes under storm surge conditions that could establish a new tidal inlet to Christchurch Harbour (Bray, 1993; Hydraulics Research, 1986; Parker and Thompson, 1988; H R Wallingford, 1993; Turner 1998; Royal Haskoning, 2011.) Concern was expressed at the time of recharge that beach retention may be low east of Double Dykes where the beach remains ungroyned (Lelliott, 1989). This was confirmed by 1995, when analyses of beach volumes indicated that most of the material added via recharge had been lost, much of it within the first year (Bournemouth Borough Council, 1995; Harlow, 2001). Proposals to build a sequence of five rock groynes between Double Dykes and a location 400m west of the Long Groyne (HR Wallingford, 1993; Bray, 1993), as a measure to conserve beach sediment, have not yet been implemented. Offshore loss of fines is a characteristic response of newly nourished beaches and is probably a factor promoting loss of volume at this location.
3. Littoral Drift
Introduction
Analysis of Coastal Monitoring Programme lidar, aerial photography and baseline topographic data confirms net eastwards to northeastwards littoral drift within the bay, although frequent reversals in response to changes in prevailing wind direction are an important feature especially in the eastern sector of Poole Bay (Harlow, 2012f). The sheltering (wave refracting and diffracting) effects of the Isle of Purbeck and Hook Sand, together with strong tidal currents generated at Poole Harbour entrance have resulted in a complex transport regime in the west of the bay that is not yet understood fully.
To better understand processes occurring throughout the bay, a systematic biannual measurement programme of 33 beach profiles (from 1974) and monitoring of drift inferred from beach levels against groynes (from 1993) was undertaken by Bournemouth Borough Council [BBC] up to 2002. (From 2003, monitoring of 90 cross-sectional beach profiles has been undertaken by the CCO Regional Coastal Monitoring Survey - refer to subsequent text). This dataset, which incorporates over 80,000 observations from 1565 surveys extending up to August 2011 has been analysed to determine major trends and patterns in littoral drift and beach accretion/erosion occurring between Poole Head and Solent Beach (Harlow 1995, 2001, 2005a and 2012f). The main points are:
- Between 1993 and 2011, 59.6% of all measurements revealed net easterly drift, whilst 31.5% indicated westerly movement; the remaining 10% were indeterminate. (refer also to HR Wallingford (2003) for a simulation of potential nearshore and offshore sediment transport directions; this is based on HRW’s PISCES model, which integrates wave propogation and tidal streams). The longest period of uninterrupted easterly drift was 239 days and of westerly drift 49 days. Locations of maximum drift rates in both directions were identified - for easterly drift these are opposite Flaghead Chine, Alum Chine and Fisherman’s Walk (Southbourne beach). Harlow (2005c) provided a preliminary estimation of resultant drift directions based on measurements of the directions of wave approach from the Boscombe waverider record, July 2003 to May 2005. 57% of littoral transport was eastwards, and 41% westwards, but as Harlow acknowledges direction alone can only provide an approximation of operative drift.
- Assessment of net drift directions has also been corroborated by measurements of changes in the area of inter-groyne beach profiles between successive surveys, rather than simply depending upon records of up and downdrift beach levels against groynes.
- Drift reversals are initially evident on the beach foreshore above mean low water, as this area is subject to wave action for a longer period during each tidal cycle. Sustained reversals (i.e. exceeding several tidal cycles) correlate significantly with changes in wind direction, quantified as a "Cumulative Wind Forcing Factor". Changes in incident breaking wave heights, steepness and angle of approach were not systematically recorded and have not been quantitatively related to near simultaneous drift directions.
- Discontinuities of transport are introduced by the rock and timber groynes, but also by Bournemouth and Boscombe piers. The piles of these latter structures act as offshore breakwaters, thus reducing breaking wave energy and the rate of longshore transport immediately downdrift; this effect is apparent from the presence of beach salients, or nesses. Subtle spatial variations in quantities of littoral movement are also due to slight changes in the alignment of the seawall and esplanade due to its "staged" construction. A small salient coinciding with the former site of Southbourne pier (demolished in 1907) also significantly influences the local rate and direction of longshore drift. Without the presence of these artificial transport discontinuities, littoral drift rates would be significantly higher than those now operative, perhaps approaching a potential 200,000m³ per year (Harlow, 2012b.)
- Survey frequency was sufficient to identify even very transient drift convergence and divergence zones, established by alongshore variations in net drift direction due to spatial variation of wave conditions in response to changes of incident winds.
- Littoral drift involves both the intertidal beach face and the nearshore zone, but the latter has not been directly monitored. In discussing the morphodynamics of beaches in western Poole Bay, Cooper (1997) reported inconclusive research that net eastwards drift was accompanied by net westwards movement in the nearshore zone.
- Drift rates cannot be determined directly from changes in inter-groyne beach volumes. This is because of the influences of spatially variable groyne design and geometry on sediment trapping and by-passing, as well as antecedent beach gradients. Harlow (2000) calculated a potential gross drift rate for the whole Bournemouth frontage of 124,000m³ per year, but control structures, lack of sediment supply and other factors substantially reduce this in practice. Cooper, Hooke and Bray (2001) proposed that the actual rate is closer to 20,000m³ per year, with rates increasing eastwards in response to average wave energy.
- Drift rates are at a maximum on the main shore parallel bar, fall to a minimum in the adjacent shorewards trough and increase to a subsidiary maximum close to Mean High Water before declining to zero at the limit of wave run-up. They increase significantly when storm waves operate (HR Wallingford, 2003).
- This detailed record is an excellent database for assessing and thus improving groyne performance- refer to Harlow (2012b, f) for detailed discussion of attributes such as material composition, permeability, dimensions and storage capacity, spacing, orientation etc. based on specific experience.
The morphological form of the Sandbanks peninsula would appear to be indicative of east to west littoral drift of sand, although studies over the past decade suggest that a complex transport situation exists. In particular there is some uncertainty over the net drift direction and a tendency for frequent drift reversals is identified.
Analysis of Coastal Monitoring Programme lidar, aerial photography and baseline topographic data confirms a net northeastwards littoral drift of approximately 3,000m³ per year, although drift rates and direction are variable over short timescales. Observations of accumulation of sediment against the Shore Road outfall indicates sub-tidal drift east to west. Onshore and offshore sediment transport occurs between the beach face and inter-tidal foreshore to the sub-tidal limit of the beach profile, rather than fully offshore. The 2004 arrows, based on modelling outputs, indicated more than 20,000m³ per year was transported both northeastwards and southwestwards.
The timber groynes at Sandbanks that were built in 1996 to supplement or replace earlier structures were designed as training walls, as the tide flowing in the East Looe Channel was destabilising the channel. Anecdotal information from the 1996 groyne construction indicated net beach accretion with beach width increasing approximately 15-20m. Since construction the beach has remained stable implying an onshore supply of sand from Hook Sands.
The majority of the nearshore zone of Poole Bay is characterised by a shallow, gently sloping, featureless seabed. Within the bay the nearshore homogenous fine to medium sand, muddy sands and gravel sediment is generally sufficiently thick to mask the underlying geology, although between Canford Cliffs and Bournemouth Pier, and ranging between 400-800m offshore, there are isolated rock outcrops, and some 2.5km offshore a larger collection of outcrops called the Poole Rocks. The lack of bedforms or sediment accumulations related to these outcrops provides no evidence of a net sub-tidal transport direction.
Littoral drift modelling revealed a drift reversal west of Sandbanks car park, with net westward drift continuing at 20,000m³ per year towards the Haven Hotel (Hydraulics Research, 1991b). Southwestward drift fails to accumulate in the vicinity of the Haven Hotel so it was envisaged that flood tidal currents at Poole Harbour Entrance entrain sediment arriving and transport it into the main channel. The flood dominant East Looe channel flows across the subtidal beach toe so that it could contribute significantly to transport and ultimately control beach development at this location. Indeed, when littoral drift was modelled at different levels across the beach profile results revealed greatest transport rates below mean low water where tidal currents have higher velocities (Hydraulics Research, 1991b; Halcrow, 2004).
HR Wallingford (2000) identified several distinct phases between 1886 and 1994 during which drift alternated between net easterly and westerly directions. West to east net movement has prevailed since 1952, though with several short-term periods of reversal. This variability is considered to result primarily from fluctuations in incident wave climate. During periods of uninterrupted unidirectional drift, throughput rates may be up to 100,000m³ per year. During transition phases of reversal, net rates are lower than 5,000m³ per year, in either west or east directions. The potential for rapid transport is also indicated by rapid beach losses that occurred following groyne removal in 1990/91 and significant accretion that occurred within the embayments of five rock groynes constructed in 1994/95 (Photo 1) and 2001 (Photo 7) (HR Wallingford, 2000). This recent experience recalls earlier fluctuations of alternating beach depletion and recovery, dating back more than a century (Royal Haskoning, 2011.)
Both HR Wallingford (2000) and Brampton, et al. (1998) observe that potential drift west of the westernmost groyne (Photo 1) is towards Poole Harbour entrance, but that little material is actually retained on the beach at this location.
Harlow (2009) indicated that a littoral drift reversal occurred in the past, but perhaps due to anthropogenic changes at the coast or on Hook Sands, it was not sustainable and has ceased to operate.
One difficulty that would appear to arise here is an explanation for the growth and development of the Sandbanks peninsula. It is mostly readily understood as a conventional spit that has extended southwestwards in the direction of dominant littoral drift, with its proximal end near Poole Head. In doing so, it has, together with the northwards growth of South Haven spit, narrowed the entrance channel to Poole Harbour. However, arguably both spits may have a partial origin as former transgressive barrier beach structures.
Analysis of Coastal Monitoring Programme lidar, aerial photography and baseline topographic data provides no evidence for drift reversal, with a net northeastward transport; between Sandbanks and Poole Head a drift of 1-3,000m³ per year is confirmed, which may coincide with potential reversals in drift, depending on prevailing wave conditions. This represents a change in direction and rate from the 2004 arrows based on modelled outputs, which indicated more than 20,000m³ per year in both northeastward and southwestward directions.
In 2012, an extensive high resolution, 100% coverage swath bathymetry dataset was collected in Poole Bay, by the Maritime and Coastguard Agency’s Civil Hydrography Programme. However the shoreward limit was defined by the 2mCD contour and did not cover the shore-parallel bar that operates between MLW and 400m offshore, Hook Sands, the Looe Channel and the Swash Channel.
The orientation of the Sandbanks spit, favoured historic westward drift therefore Hydraulics Research (1991b) hypothesised that eastward drift is restricted to the upper beach and that net drift across the lower profile is westward.
Hydraulics Research (1991b) modelling suggested a southwest drift from Branksome Chine to Poole Head fed by beach erosion or onshore transport of approximately 50,000m³ per year. However Robson (2003) confirmed that there is no evidence for either process operating at such magnitude; field observations, supported by recent Coastal Monitoring Programme data, have suggested that net eastward drift increases up to 3-10,000m³ per year between Poole Head and Branksome Chine. Rates are likely to increase progressively eastwards across this sector.
Analysis of Coastal Monitoring Programme lidar, aerial photography and baseline topographic data, in combination with observations of beach morphology over the past decade, provides no evidence for drift reversal, with a net eastward transport direction confirmed, with a net drift in the order of 10-20,000m³ per year between Canford cliffs and Branksome Chine, increasing to more than 20,000m³ per year eastward to Durley Chine. The effectiveness of groynes in restricting drift on this sandy beach is limited due to a marked propensity for drift in the nearshore zone seaward of the toe of the longest groynes.
In 2012, an extensive high resolution, 100% coverage swath bathymetry dataset was collected in Poole Bay, by the Maritime and Coastguard Agency’s Civil Hydrography Programme. However the shoreward limit was defined by the 2mCD contour and did not cover the shore-parallel bar that operates between MLW and 400m offshore. The majority of the nearshore zone of Poole Bay is characterised by a shallow, gently sloping, featureless seabed. Within the bay the nearshore homogenous fine to medium sand, muddy sands and gravel sediment is generally sufficiently thick to mask the underlying geology and paleao-channels that drained the chines, although between Canford Cliffs and Bournemouth Pier, and ranging between 400-800m offshore, there are isolated rock outcrops. The lack of bedforms, sediment accumulations connected to rock outcrops and the longevity of discernible paleaochannels and sub-tidal drainage features may suggest minimal transport of sand and gravel grade material.
Wave refraction and shoaling was analysed in Poole Bay, based on a hindcast offshore wave climate derived from wind data for 1977. Inshore wave conditions were calibrated against those recorded at the Southbourne wave rider site and corrected inshore wave climates were employed to calculate longshore wave energy flux (Henderson, 1979; Henderson and Webber, 1979). These calculations revealed a potential for the divergence of littoral drift pathways in the vicinity of Durley Chine. The role of groynes in impeding littoral drift also needs to be calculated, but detailed observations and measurements between 1993 and 2011 of accretion against groynes failed to identify evidence of a sustained drift divide at this location (Harlow, 2001; 2012f), although it may have a brief presence during short periods of drift reversal. The 2004 map showing this as a permanent divide has therefore been corrected.
Analysis of Coastal Monitoring Programme lidar, aerial photography and baseline topographic data, in combination with observations of sand distribution within groyne compartments, confirms a net eastward transport direction, with a net drift in the order of more than 20,000m³ per year between Durley Chine and Southbourne (Photo 4). The effectiveness of groynes in restricting drift on this sandy beach is limited due to a marked propensity for drift in the nearshore zone seaward of the toe of the longest groynes.
In 2012, an extensive high resolution, 100% coverage swath bathymetry dataset was collected in Poole Bay by the Maritime and Coastguard Agency’s Civil Hydrography Programme. However the shoreward limit was defined by the 2mCD contour and did not cover the shore-parallel bar that operates between MLW and 400m offshore. The majority of the deeper nearshore zone of Poole Bay is characterised by a shallow, gently sloping, featureless seabed. Within the bay the nearshore homogenous fine to medium sand, muddy sands and gravel sediment is generally sufficiently thick to mask the underlying geology and paleao-channels that drained the chines. The lack of bedforms and the longevity of discernible paleao-channels and sub-tidal drainage features may suggest minimal transport of sand and gravel grade material.
Different authors have determined different rates of beach volume loss according to the timescale of analysis (e.g. the effect of beach levels determining wave refection by the seawall promenade) and various assumptions on transport conditions. Loss of 65,000m³ per year was calculated by Newman (1978), 91,000m³ per year by Lacey (1985) and 40,000m³ per year, over 9 years, by Hodder (1986). These values were then used to calibrate sediment transport equations and littoral drift was hindcast based on the wave refraction analysis of Henderson (1979). The estimates of Hodder (1986) are regarded as the most reliable, for reasons given in the preceding section. Cooper, Hooke and Bray (2001) quote a mean drift rate of 35,000m³ per year following the effect of the second replenishment programme. All authors agree that net drift is eastward and that rates increase from Bournemouth Pier to Southbourne due to increasing wave energy in this direction (Henderson, 1979). Harlow (2001) observes, from a detailed programme of drift monitoring, that Bournemouth and Boscombe piers locally reduce longshore transport, and promote some immediate updrift accretion. Local variations in longshore transport rates are induced by subtle variations in seawall alignment.
Fluorescent sand tracer experiments were undertaken over a limited period between Bournemouth and Boscombe piers (Kent 1988, Whitehouse, et al., 1997). Wave and longshore current measurements were employed for empirical derivation of littoral drift rates and results compared favourably with those obtained by Henderson (1979) and Lacey (1985). Calculations of drift rates based on tracer results were significantly lower by comparison; this was attributed to failure to adequately model the effect of tidal range and onshore-offshore transport, as well as a low level of tracer recovery. It was concluded that longshore transport could not be reliably predicted using existing transport equations because these fail to account properly for all factors influencing movement. Another short-term study measured suspended sediment concentration in breaking waves at a site west of Boscombe Pier (Taylor 1990). Although of too limited scope for conclusive results, the study showed a relationship between sediment concentration with elevation above sea-bed and breaker type. Measurements of this type are ultimately necessary to develop more refined sediment transport models which overcome the problems of excessive simplification noted by Kent (1988)
Analysis of Coastal Monitoring Programme lidar, aerial photography and baseline topographic data confirms a net eastward transport direction, with a net drift in the order of more than 20,000m³ per year between Southbourne and Hengistbury Head.
In 2012, an extensive high resolution, 100% coverage swath bathymetry dataset was collected in Poole Bay, by the Maritime and Coastguard Agency’s Civil Hydrography Programme. However the shoreward limit was defined by the 2mCD contour and did not cover the shore-parallel bar that operates between MLW and 400m offshore. The majority of the nearshore zone of Poole Bay is characterised by a shallow, gently sloping, featureless seabed. Within the bay the nearshore homogenous fine to medium sand, muddy sands and gravel sediment is generally sufficiently thick to mask the underlying geology, although the western flank of the northwest-southeast oriented rock outcrop of Christchurch Ledge is clearly evident southwest of the Long Groyne. Further offshore, the seabed has a thin veneer of sand in summer, but may be bare of sediment most winters, suggesting erosion. (Harlow (2012b) offers a tentative calculation of annual loss.) Sediment accretion to the west of an outfall feature, between Southbourne and Double Dykes indicates eastward sub-tidal transport.
For the calculation of drift rates along this sector, it is particularly important to correctly calibrate sediment transport equations because the beach face is composed of a higher proportion of flint gravel, which exhibits different hydraulic behaviour to sand and is less readily transported. Henderson (1979); Lacey (1985) and Hodder (1986) derived gross drift rates within the range 53,000m - 70,000m³ per year at Southbourne, and indicated a slightly increased drift rate of gravel and sand between Double Dykes and Hengistbury Head of 63,000 to 90,000m³ per year. Hydraulics Research (1986) suggest a lower rate, at 45,000m³ per year, based on a modelling approach. Most of the transport estimated involves movement of sand rather than gravel in the nearshore zone. Differences in the estimates are attributed to use of different wave climate data and differences in calibration of sediment transport equations to represent the mixed grain sizes of the beach.
Valuable information was derived from three separate tracer experiments employing aluminium pebbles similar to indigenous clasts in terms of size-range, distribution and shape (Wright, Cross and Webber, 1978 and 1980; Wright, 1982). Tracer recoveries were quite high (28-61%), thus enabling determination of the width, depth and longshore transport of coarse beach sediment. The maximum distance of alongshore displacement was greatest to the west (167m), but the average results indicated 58m movement to the west and 87m to the east. Drift volumes calculated from this information were related to three sets of incident wave conditions (each with characteristic breaking wave heights) to calibrate transport equations, and these were then used to estimate annual drift rates. Gravel transport rates of 11,000m³ per year and 28,000m³ per year were derived from the two main experiments, but transport was only accurately monitored over short periods that were not fully representative of annual conditions. The drift estimates of Henderson (1979), Lacey (1985) and Hodder (1986) are also likely to be excessively high because the presence of gravel on a sandy beach causes decreased sand transport (Wright, 1982) and presence of sand on a dominantly gravel beach causes enhanced gravel transport. This suggests that it may be necessary to calculate gravel and sand transport separately and employ appropriate correction factors depending on beach sediment size distributions. A further problem with overall drift estimates derived for this segment is that the calibration of sand transport was derived using a groyned nourished beach, whilst that of gravel was from a natural beach. This is a cause of inaccurate prediction of total sediment transport, because sand and gravel calibrations were combined arithmetically and errors with one element affect the whole equation.
The limitations apparent in existing transport predictions led HR Wallingford (1986) to undertake further studies based on mathematical modelling. A hindcast offshore wave climate based on Portland wind data (1974-84) was validated against inshore wave measurements from the Southbourne wave rider buoy. Refraction and shoaling analysis was employed to determine shoreline longshore wave energy flux, from which an eastward drift of 45,000m³ per year was calculated. This estimate however was based on the assumption that the Long Groyne was not present. As this major structure causes progressive eastward change in beach orientation, it is possible that actual drift rates would decrease in this direction due to the more stable beach plan shape produced by accretion against the groyne. Analysis of accretion rates immediately updrift of the Long Groyne reveal accumulation equivalent to eastward drift of 600-900m³ per year (Wright 1982). However, this is an underestimate because the groyne is only a partial barrier to transport and subject to overtopping (Photo 10) and outflanking, especially by sand (Hydraulics Research, 1986; Harlow, 2001; Harlow, 2012c). The net eastwards drift of 45,000m³ per year is more representative of conditions at Double Dykes where the Long Groyne has no updrift effect. Construction of rock groynes on Solent Beach intercepted drift and a multi-line beach plan model was employed to investigate these changes (Hydraulics Research, 1986). The model calculated the distribution of littoral drift across the profile and showed that drift on the upper gravel beach was small (less than 9,000m³ per year) compared to that further down profile and in the nearshore zone. The groynes therefore intercepted only a small proportion of total drift, mostly the gravel fraction. By analogy this situation may also exist at the Long Groyne. Cooper (1997) concluded, from analyses of beach behaviour either side of the Long Groyne, between 1987 and 1993, that sand from the 1988 replenishment bypassed within 3 to 4 years of its emplacement updrift.
Analysis of Coastal Monitoring Programme lidar, aerial photography and baseline topographic data confirms a net eastward transport direction with less than 1,000m³ per year of sand accumulation at the Long Groyne. Calculations of foreshore levels above MLW suggest a net drift in the order of more than 20,000m³ per year of fine sand by-passes the rock reinforced concrete Long Groyne into the nearshore zone of Christchurch Bay and along Mudeford Spit, probably in the order of 35,000m³ per year. Bypassing may be assisted by flood tide generated eddies (Harlow, 2012c). A significant volume of material may be contained in the subtidal extension of the foreshore including the mobile bar system (which varies in position ranging from MLW to approximately 400m offshore), but it is has not been possible to quantify these volumes.
In 2012, an extensive high resolution, 100% coverage swath bathymetry dataset was collected in Poole Bay, by the Maritime and Coastguard Agency’s Civil Hydrography Programme. However the shoreward limit was defined by the 2mCD contour and did not cover the shore-parallel bar that operates between MLW and 400m offshore. The majority of the nearshore zone of Poole Bay is characterised by a shallow, gently sloping, featureless seabed. Within the bay the nearshore homogenous fine to medium sand, muddy sands and gravel sediment is generally sufficiently thick to mask the underlying geology, except on the northwest-southeast oriented Christchurch Ledge, which extends continuously approximately 6km offshore from Hengistbury Head towards the Dolphin Bank. This is the dominant seabed feature at the eastern end of the bay.
Sand, and some gravel, accretion against the south-facing sides of a set of rock groynes on Mudeford Spit indicates northward transport (Tyhurst 1985). Accumulation of sediment against new rock groynes built in 1986/87 by Bournemouth Borough Council was so rapid that a planned renourishment was not carried out (Harlow, 2012c). In the absence of knowledge of the specific source, it is generally held (but unproven) that sediments accumulating on the spit are at least partially derived from bypassing of the Long Groyne. Visual observations suggest that overtopping of the structure by sediment is probable (Photo 10). Physical modelling of the Long Groyne and its surrounding area (Hydraulics Research, 1986) indicated that coarse sediment could overtop the groyne during storms and sand could outflank it. Although subject to scaling effects, the physical modelling results corroborate findings of earlier drifter research, which indicated a strong trend for eastward transport around the Long Groyne (Watson, 1975, Tyhurst, 1976, Turner, 1990). Thus a significant proportion of the predicted 45,000m³ per year arriving at this point (Hydraulics Research, 1990) may outflank it. It is also considered likely that the rapid filling of the groyne bays along Mudeford Spit soon after their construction indicates an offshore to onshore sediment supply pathway- refer to discussion in unit on Christchurch Bay.
4. Sediment Outputs
4.1 Offshore Transport
In 2012, an extensive high resolution, 100% coverage swath bathymetry dataset was collected in Poole Bay, by the Maritime and Coastguard Agency’s Civil Hydrography Programme. Within the bay the nearshore homogenous fine to medium sand, muddy sands and more patchy gravel sediment is generally sufficiently thick to mask the underlying geology, except on the northwest-southeast oriented Christchurch Ledge, which extends continuously approximately 6km offshore from Hengistbury Head towards the Dolphin Bank. This submarine ridge, which varies in width between 400-1200m is the dominant seabed feature at the eastern end of the bay, due to the presence of several resistant bands of ironstone concretions (Velegrakis, 1994; Brampton, et al., 1998). Its crest is at approximately -8m CD.
Sediment provenance is uncertain, but the occurrence of coarse materials seawards of the modern shoreface is probably mostly relict, derived from both terrestrial and marine deposits dating from the mid or late Quaternary (Brampton, et al., 1998). In the nearshore zone, and seawards to the central areas of Poole Bay, predominantly sandy sediments are likely to mostly derive from past cliff and shoreface erosion. Seismic studies (Velegrakis, 1994; Brampton, et al., 1998) have revealed that seabed sediments rest on an erosion surface cut across previously varied relief. This was the result of valley incision during the Devensian stages of low sea-level, when the present day area of Poole Bay was exposed to sub-aerial denudation. This is indicated by buried palaeochannels, infilled with a transgressive suite of Holocene fluvial, estuarine and marine sediments (Velegrakis, et al., 1999).
Submarine sandbanks have developed at Hook Sand seaward of the mouth of Poole Harbour and at Dolphin Sands. The latter may function as a sink for both Christchurch and Poole Bays, although some of the sand arriving there is in transit, ultimately moving into the deeper water of the central English Channel. All other sandbanks are stores, as revealed by their dynamic behaviour. Brampton, et al., (1998) report that Dolphin Sand is up to 8m above the level of the surrounding seafloor, and is an average of 6.5 km in length and between 1.0 and 2.0km in width. This represents an accumulation of over 25 million m³, with clear evidence of a basic stability of form from the end of the nineteenth century. Given these dimensions, it is difficult to argue that all of this material is mobile, even though hydrographic chart comparisons for 1980 to 1990 indicate that the northern flank of Dolphin Sands eroded at much the same rate as its southern flank gained material.
Analysis of Coastal Monitoring Programme and swath bathymetry data collected by the Maritime and Coastguard Agency’s Civil Hydrography Programme, provides no conclusive evidence to support a net eastward transport or circulation offshore. Further research is proposed to determine whether wave driven offshore to onshore transport is an active process within the outer reaches of the bay.
Sea-bed drifter experiments indicated marked eastward transport in areas seaward of the -10mOD contour and east of Bournemouth (Watson 1975, Tyhurst 1976, Turner 1990). These studies have shown a tendency for movement from Poole to Christchurch Bay within the offshore zone, but are contrary to transport directions indicated by bedforms (Dyer 1970, Fitzpatrick 1987) and modelling (Hydraulics Research, 1991; 1990a; Brampton, et al., 1998). However, they are similar to those determined for surface currents by OSCR (BP 1991; Osborne, 1991). It is therefore possible that drifter results are an unreliable guide to bedload sediment transport. Confusion may derive from profile asymmetry of smaller bedforms, subject to reversal by tidal currents. It should be emphasised that almost all survey data has been obtained under calm, summer conditions, and therefore is not representative of potentially very complex transport conditions under high energy waves (Hydraulics Research, 1990a; Brampton, et al., 1998; HR Wallingford, 2003).
This postulated seabed transport pathway is opposed by the findings of drifter experiments (Watson, 1987, Tyhurst, 1976, Turner, 1990). These studies show a net eastward transport potential over all areas of the sea bed east of a line drawn south of Bournemouth pier. The reliability of these studies is questionable because it is uncertain what drifters actually measure and it has been shown that they are not necessarily reliable indicators of residual seabed currents and seabed mobility (Collins and Barrie, 1979). Additionally, recoveries were between 38 and 75% and there was uncertainty as to the accuracy of some information returned and the exact routes followed by drifters. This postulated pathway cannot be verified until the results of further studies become available.
Analysis of Coastal Monitoring Programme and swath bathymetry data collected by the Maritime and Coastguard Agency’s Civil Hydrography Programme indicates a range of bedforms, including gravel and sand waves and sand ribbons. However, profile asymmetry provides no conclusive evidence to support a net southwestward transport or circulation offshore, as simulated by HR Wallingford (2003).
Sediment type throughout much of this area is moderately well sorted medium/fine sand, which forms waves moving across a stable gravel basement. Analysis of the profile asymmetry of these bedforms indicated transport away from Dolphin Sand to the southwest, with transport veering more directly southward further offshore. Typical Eocene substrate-derived heavy minerals are present in the north of the area, e.g. epidote, glauconite, apatite and also in smaller quantities south of the eroded Purbeck-Needles ridge, despite their relative absence from the in-situ sea-bed outcrops. This implies a southward supply from Poole Bay, but as no evidence of corresponding northward return feed is available, it is tentatively concluded that this pathway comprises a net output from the Poole Bay system. Much material may have been supplied from Christchurch Bay via Dolphin Sand and possibly only passes through Poole Bay en route to a final sediment sink in the central English Channel (Halcrow, 1999; Brampton, et al., 1998; Cooper, Hooke and Bray, 2001).
4.2 Estuarine Sediment Transport
EO1 Poole Harbour Entrance, the Swash Channel and Hook Sand
Ebb tidal currents are shorter-lived, but more rapid (up to 2.5ms-¹) than corresponding flood currents at Poole Harbour entrance (BP, 1991; Hydraulics Research, 1986, 1988, 1990b, 1991), thus the potential exists for seaward transport along the Swash Channel (Appleton, 1994; Halcrow, 1999; HR Wallingford, 2003). This feature is 3.2km in length, 150m in width and is flanked by linear bars that function as temporary sediment stores that provide input to onshore directed transport of sand. Its maximum depth is at the harbour mouth, due to channel definition by natural tidal flushing and maintenance by dredging. Tidal modelling based on bathymetry recorded on a 400m resolution grid was used to determine sand transport using empirical calculations (Hydraulics Research, 1986; BP Exploration, 1991). These studies showed increasing transport potential southward along the Swash Channel and towards Handfast Point. Hydrodynamic analysis also showed that the localised ebb tidal "jet" from Poole Harbour entrance was deflected southward by interaction with dominant though comparatively weak south-westerly tidal flow within Poole Bay, further confirmed by large asymmetric bedforms (Brampton, et al., 1998; HR Wallingford, 2003). The zone of maximum interaction was at Poole Bar, where sediment transport of 20,000m³ per year across the bar was predicted. This conclusion that net transport is seawards of the harbour entrance was supported by limited sediment sampling which showed progressive grain size reduction along this pathway, from a dominance of coarse gravel at the harbour entrance to fine sand and silt at Poole Bar (Halcrow, 1999; HR Wallingford, 2003). Deepening of the Swash Channel in 1988/89 and 2005/006 due to dredging may have been sufficient to reduce sediment supply to Hook Sand as well as the littoral transport pathway in Studland Bay.
A more detailed mathematical tidal current model was developed on a 44m resolution grid, validated against float tracking, and coupled with an empirical sandflow model (Hydraulics Research, 1986). To provide a more realistic transport prediction, wave effects were also included. This latter information was obtained from a hindcast offshore wave climate (Portland wind data 1974-84) adjusted for refraction and shoaling effects due to Hook Sand. Bed conditions were specified according to available information and three contrasting wave/tide conditions were tested. The Swash Channel itself was not examined, as little material was present on the bed, but its margins were modelled in detail. During calm conditions (70% of the time), the mean transport rate was limited, and was from north-east to south-west across the bar at an average of 15,000m³ per year. Under the influence of "typical" waves (occurring 30% of the time) the gross transport rate increased four-fold and net annual movement of 29,000m³ was estimated. Storm waves (once per year) caused a ten-fold increase in transport, but net movement was only 2,000m³ per year due to the low frequency of this condition. Net north-east to south-west residual transport of 46,000m³ per year of suspended sediment was estimated from Hook Sand southward to Poole Bar. This was most intense under the combination of high energy wave action and spring tides. These calculations are of medium reliability because site conditions are complex and modelling necessitated simplification. For example, wave-current interactions are not fully understood and bedload transport was not adequately modelled. In addition, long-term trends were not established because only a limited number of conditions could be tested. However, the basic observations of this study were confirmed by Brampton, et al. (1998) using the SANDFLOW 2-D sand transport model, calibrated for tidal current velocities. This study and HR Wallingford (2003) concluded that only waves and tidal currents operating together could account for the predicted volumes of predominantly suspended sediment moved.
A comprehensive sediment transport study of the area immediately seawards of the harbour entrance was undertaken to determine the potential effects of constructing a proposed artificial island north-east of Hook Sand (Hydraulics Research, 1991b; BP et al., 1991). This involved modelling the existing regime, against which to compare predicted changes that might be caused by the island. Sand transport was determined from mathematical wave and tide models validated against appropriate field data. Sensitivity tests were conducted to evaluate the relative importance of key variables, e.g. wave approach direction, and a series of base conditions established. Transport was studied during calm, "typical" and storm conditions and it was confirmed that most rapid transport occurred with high energy wave conditions and spring tidal currents working together. Potential, but perhaps temporary, deposition zones were identified in the harbour entrance, in the Swash Channel, near the crest of Hook Sand and on Poole Bar. Erosion was predicted on both the seaward face and landward flank of Hook Sand. In places, it is naturally armoured by coarse sand of lower mobility, which could protect the bank and inhibit erosion. Effects of factors such as these are difficult to model and their simplification can cause over-estimation of transport rates. Detailed analysis of the conditions in the Swash Channel revealed that transport was strongly dependent upon combined wave and tidal current action with 50-100 times more movement under “typical” waves and 500-1000 times more during the operation of storm waves. Deposition/erosion profiles along the Swash Channel revealed accretion on the northern margin of the bar during calm conditions, accretion further seaward with "typical" waves, and erosion of the seaward face of the bar under storm wave action.
Due to short-term variability indicated by the sensitivity tests, these results are for a limited number of conditions. They cannot be extrapolated as long-term trends without more detailed knowledge of the interactions of variables such as wave direction, wave breaking and tidal phase. Accurate long-term study is therefore required to examine small systematic short-term changes which might be cumulative and cause major long-term trends.
Hook Sand
This feature is included here, as it can be interpreted as a part of the ebb tidal delta of Poole Harbour entrance composed of sediments from Poole Bay. Over its northern area, in the East Looe Channel, tidal currents are capable of moving fine sand. Part of the crest of Hook Sand lies above -1mOD causing refracted waves to break. It is suggested that this causes: (i) sand to be driven onshore (north-westwards) from the crest (HR Wallingford, 2000; 2003), and (ii) some sediment to move offshore in the shallows of Poole Head and then southwards along the east side of the bank (Hydraulics Research, 1986; 1991a). Refraction and shoaling models based on a 10-year hindcast offshore wave climate indicated a high potential for net southward transport of both suspended and bedload sediments in excess of 25,000m³ per year. Sand supplied by this pathway may periodically partially infill the Swash Channel and/ or be transported further south to Poole Bar, although there is no clear evidence for this (Brampton, et al., 1998). Hydrographic chart comparisons covering the period 1785-1990 revealed that the dimensions of Swash Channel and Hook Sand were subject to periodic fluctuation, but overall were relatively stable in position and planform. A small net reduction in area of Hook Sands over this period was detected, and of course there were probable seasonal, annual and inter-annual variations that could not be resolved from the cartographic evidence. This implies a long-term equilibrium between sediment supply and loss (Velegrakis, 1994; Hydraulics Research, 1986, 1991c; HR Wallingford, 1994; Halcrow, 1999) despite considerable sediment throughput, thus suggesting that Hook Sand is not the ultimate sediment sink for this transport pathway. It does however function as a significant sediment store, supplying sediment to shore-normal and alongshore transport pathways in the nearshore zones adjacent to Sandbanks and Studland peninsulas. Reliability of transport-direction is high due to corroboration by model studies (Hydraulics Research, 1988, 1991b; HR Wallingford, 2003), but quantitative information on transported volumes is of low to medium reliability because the independent contribution of tidal currents remains uncertain.
There is no unambiguous evidence of wave driven sediment from offshore within the bay and the speculative 2004 arrows indicating such transport have been removed.
4.3 Aeolian Transport and Deposition
A1 Losses to cliff face and cliff access routes
Sand is entrained from the beach by the wind, blown along the coast and up the cliff face (Harlow, 2013). Detachment of sand grains by wind erosion is a minor process but makes a contribution to the accumulation of cliff foot debris. Observations (Harlow, 2012a), discussions and analysis of the Regional Coastal Monitoring Programme data indicates a net loss of fine-grained cohesionless sand landward from the foreshore to the cliff face, or restrained by sediment traps installed to minimise transfer to the public promenade. Harlow (2012a) estimates that on a bay-wide basis, approximately 10-20,000m³ per year of sand may be lost from the beach surface to the cliff face and landwards of the promenade. Cliff access routes, such as Boscombe and Double Dykes, literally act as direct pathways of sediment from foreshore to the cliff top, where sand accumulates and is subsequently removed from the footpath or highway through Council maintenance operations (Harlow, 2012a).
5. Sediment Stores and Sinks; Beach Morphodynamics
The frontage is divided into the following three zones representing differences in transport regime and style of behaviour:
North Haven Point, Sandbanks, to Branksome Chine
HR Wallingford (2000) presents a detailed summary of the major phases of beach behaviour from the late nineteenth century through to 1994. Between 1886 and 1900, accretion and erosion were experienced at different locations, in the absence of any substantial defence or control structures. Groynes were introduced in 1896 in response to local beach sediment losses but were not comprehensive until 1925. Over this period, there was steady accretion in groyne bays over the central section, but net erosion at either flank. For the next 30 years, accretion was confined to the groynes in the western sector, with modest erosion affecting the eastern/north-eastern part of the beach. This suggested either net east to west drift, or a tendency for sand to move from offshore and built up the beach behind Hook Sand. After the mid-1950s, beach stability prevailed over the eastern sector but there was some erosion to the west, indicating a prevailing west to east littoral transport pathway. Between approximately 1910 and 1980, some 60 m of seaward advance occurred over central and eastern beaches; most of the original 1890s groynes were removed in the 1950s. Starting in the late 1980s and accelerating over the next 10 years, erosion became dominant over the entire beach frontage; some 40 m of retreat (1.8m per year) was experienced between 1991 and 1998 on central Sandbanks beach. Refer to Halcrow (2004) for numerical modelling of beach response under 1:1 to 1:100 year storm conditions.
This abrupt change from accretion/stability to erosion affecting the central and eastern sectors of this beach, and a significantly increased erosion rate west to Haven Hotel, has been the subject of research and remedial action (HR Wallingford, 1994; 1995a, 2000). Accretion during much of the twentieth century has been ascribed to a non-uniform net onshore feed of sand from Hook Sand (Robinson, 1955). Under higher energy conditions, refracted (and refracting) waves entrain sand from the crest of this offshore bank. HR Wallingford (1994) Brampton et al. (1998) and Halcrow (1999) report some landward movement of Hook Sand between 1990 and 1996, which tends to confirm this process. However, direct onshore transport is modified by the presence of the flood dominant East Looe channel, between Hook Sand and the beaches of the Sandbanks peninsula. Its origin is not fully understood, but it is likely that tidal currents, flowing towards the Poole Harbour entrance, transport fine sand. Some of this is likely to be supplied by wave erosion of Hook Sand, and it is probable that waves also contribute to this shore-parallel pathway (HR Wallingford, 2000; 2003). Further complications are introduced by waves that move across East Looe Channel during high water, and by wave modification by both flood and ebb currents moving into and out of the harbour mouth. All of these mechanisms could create a net offshore feed to Hook Sand, which therefore functions as a dynamic store. Hydrographic chart analysis, from 1800, reveals that both Hook Sand and East Looe Channel (Poole Harbour Commissioners, 1995-6)) experience constant changes in the relative positions of their boundaries (Gao and Collins, 1994; HR Wallingford, 1994 and 1995a and b; Halcrow, 1999). Periods of beach foreshore erosion and steepening correlate closely - though not exactly - with the inshore migration of East Looe Channel.
Although rapid beach drawdown along the Sandbanks shoreline can be correlated with the incidence of a series of southeasterly storm waves between 1991 and 1994, further explanation may be linked to beach management practices. The previous phases of accretion and stability had encouraged the removal of groynes, first installed in the early twentieth century, as they deteriorated and appeared to be redundant. Rock armouring of the seawall was carried out, to reduce wave reflection, and consequent enhanced offshore transport, in 1991/2. Rock groynes were built along the southern sector of the coastline, together with a limited recharge of 40,000m³, in 1995, which were successful in retaining sand (Photo 1). Simultaneously, small backshore dunes were created, which were subsequently stabilised by vegetation planting. The primary purpose of these groynes was to act as training walls to deflect the East Looe Channel offshore, or stop it moving inshore as it was steepening the beach and making it unstable. However, these groynes appear to have promoted scour to the northeast, with a loss of beach width of 0.8m per year, 1995-1999 (HR Wallingford, 2000). This northern sector of the Sandbanks beach system did not benefit from the additional effect of the new groynes in "deflecting" tidal currents away from the foreshore. (HR Wallingford, 1995b, 2000). The effects experienced indicate that dominant littoral drift was eastwards, although short periods of reversal occurred (Harlow, 2001). In 2001, an additional four rock groynes were constructed along the eroding section between Sandbanks car park and Poole Head (Photo 7). Whilst these control measures along the western end of the beach were successful in preventing further losses, depletion continued along the central and eastern sectors. This issue was addressed in 2003 by a sand replenishment of 105,000m³, creating a beach of an average 50m width along a 1,020m frontage. Further smaller beach replenishments along the Poole Frontage have taken place in 2005/2006 from Sandbanks to Branksome Dene Chine.
Branksome Chine to Southbourne
Between 1974 and 2002, Bournemouth Borough Council undertook biannual surveys of a set of 33 beach profiles, at roughly 200m intervals. This has been continued since 2003 by the Regional Beach Monitoring Survey (CCO), thus providing a continuous record of beach morphology and volumetric changes. Data has been obtained from integrated topographic surveys to mean low water and hydrographic surveys to a distance seawards from origin of 450m (occasionally 600m) or where water depth increases to more than 10m. At around 450m there is a distinct break of slope, which can be taken as the closure point of the active beach “envelope”. Landwards the beach gradient is 1 in 50, seawards it is near-horizontal at 1 in 400. Although there has been some reservations concerning hydrographic survey accuracy, this is nonetheless an exceptionally long as well as a high quality monitoring record. Details of survey methodology, together with a critique, are set out in detail in Harlow (2012d). Surveys have been conducted during spring and autumn of each year, in an effort to "even out" seasonal fluctuations of profile form, area and volume. A comprehensive graphical record and statistical analysis (including explanatory comments and some interpretation) is given by Harlow (2001, 2005a; 2012d), with further analysis of earlier parts of this data set by Gao and Collins (1994) and Cooper (1997). Summaries are also given in Brampton, et al. (1998) and Halcrow (1999). The following paragraphs draw heavily upon data synthesis and critical interpretation by Harlow (1994; 2001, 2005a, 2005b and 2012d).
Net beach area and volume is a function of exposure to wave energy, (and over shorter-term periods littoral drift rates and directions), characterised overall by a progressive reduction in width from west to east. Greatest variability, as revealed by a measure of standard deviation, occurs at Southbourne; it is least where the beach is relatively more sheltered from incident waves, notably between the two main piers and in the west of this frontage. Beach gradients are characteristically low, but again increase eastwards. This is in response not only to wave energy, but also to a downdrift (eastwards) coarsening of mean particle size. Where there are local variations in beach orientation resulting from the influence of past or present defence structures, beach volumes also fluctuate. Without recharge, a combination of limited natural sediment supply and the presence of a wave reflecting steep, backing seawall results in "squeezing" of the intertidal beach and a consequent steady reduction of volume and steepening of gradient (Harlow, 2005e and 2012d). Estimated rates of loss of beach fill following BIS2, 3 and 4 are detailed in Harlow, (2012d), (see also previous section on Beach Replenishment) averaging 17,000m³ per year for the beach envelope seawards to 100m and 43,000m³ per year (a loss of elevation of 3.7mm per year) out to 450m. These figures may be underestimates, with combined loss closer to 70,000m³ per year. Volume reduction is in part a response to substrate erosion, which forms a concave profile between -3.2 to -4.8m below the “average” beach elevation (Harlow, 2012d). Substrate levelling occurs when the beach sediment prism is mobilised and temporarily removed during high energy events. Gao and Collins (1994) give consideration to beach stability or equilibrium conditions along this frontage, concluding that recovery from "perturbations", primarily storm and surge events, is relatively rapid.
Although there is abundant evidence for seasonal variations in beach height, area and profile form, the timing of surveys - together with analytical methods designed to "smooth" fluctuations - does not clearly reveal this (Harlow, 2012b). However, maximum area is normally characteristic of late summer and early autumn. High-energy storms can result in as much as 1 m of beach lowering in a few hours (Henderson and Webber, 1977) when it is thought that sands are transported seawards to form dynamic, migratory and fluctuating nearshore bar and trough. Typically, this material is returned to the beach during calmer intervals with an associated decay of nearshore bars. These processes of cyclic, short term "cut" and "fill" have not been studied directly, and although apparent from the record of profile oscillations (H R Wallingford, 2006) they have not been fully assessed in detail in Poole Bay.
Overall convexity of profile form out to 450 m is apparent from much of the survey record (Gao and Collins, 1994). A major reason for this is the presence of the near shore sand bar, below low water, that have characteristic heights of 1.5 m and form parallel to the shoreline. These features are present in water depths of up to 10m for approximately 80% of the hydrographic survey record and are best developed during winter, especially following storms. This type of inshore morphology is relatively unusual on the south coast of England, as it only occurs where tidal range is low. Gao and Collins (1994) propose that the bars are related to breaking waves (i.e. "break-point" bars), but the fact that they occur in progressively greater water depths, and are more persistent in an eastwards direction throws doubt on this conclusion. An alternative explanation, one that accommodates the fact that they are not permanent features along the entire frontage, is that bar topography is the result of the reflection of long period swell waves from the beach face (Harlow, 2001, 2005a). No clear associations between changes in bar and trough morphology on the one hand, and inter-tidal beach erosion and accretion on the other, has been determined. The PSD of the bar sand is significantly finer than the PSD at MLW & MHW (Harlow, 2012e).
Beach elevation has not been specifically analysed, but forms an inherent element of volume change, discussed below. It is clear that both past and present groynes have had a critical effect in this respect (Harlow, 1995; 2001; 2012b and d).
Analysis of beach and inshore sediment sampling in Bournemouth, carried out in conjunction with many of the beach surveys, indicate a high degree of both longshore and cross-shore sorting. Coarsening towards the east and around Mean High Water is evident. A value of 1.8 phi is a critical size, with finer material likely to move offshore; some of this is deposited on the shore-parallel bar. The beach replenishment of 1974/5 did not significantly alter overall particle size distribution, as 'borrow' material was close to that of the indigenous beach. However, in the five or six years following the recharge scheme of 1988/9, the less well-sorted sands obtained from dredging the Swash channel approach to Poole Harbour resulted in an overall finer distribution (Cooper, 1997). Harlow (2012e) provides detailed graphical data and analytical comments on the spatial and temporal variation of particle size distribution revealed by over 1,600 sediment samples obtained repeatedly at MHW, MLW and 200m offshore (400m in the case of Solent Beach) along established survey profiles of Bournemouth beach, 1974 to 2011. Sand is the dominant fraction, changing with depth to a sandy-gravel basement. The beach became finer in texture following BIS2, but coarser as a result of the recharge of BIS3. Consistent with the dominant direction of littoral drift, beach composition coarsens progressively eastwards. Refer also to the analyses of the Channel Coast Observatory (2007-11), which provides continuity with data previously processed by Bournemouth Borough Council.
There has been considerable analysis of net beach volume changes, 1974 to 1997 (Harlow, 2001; Cooper, 1997 and summarised in Harlow and Cooper, 1996). This indicates up to 500,000m³ of gross volume fluctuation between successive surveys, but the use of a 5 survey "rolling average" (to smooth out shorter-term irregularities) indicates variation between 6 million m³ in 1974 and 9 million m³ in 1990. There were fluctuations in between due to the effects of sand nourishment in 1975 and 1988/9 and subsequent decay due to selective longshore and offshore transport. Cooper (1997) and Harlow (2001; 2012b, d, e) have undertaken detailed analysis of the volumetric changes induced by replenishments of different sectors of the Bournemouth beach system. The major variables are location with respect to dump sites; antecedent profile form and beach area; beach orientation; exposure to wave energy; the nature of local control structures and particle size distribution of introduced material. As with many renourishment projects, rapid losses or movements were experienced within two years of emplacements, but thereafter stabilised at levels characteristic of "background" erosion. For example, volume losses resulting from the 1988/9 recharge were 255,000m³ per year over the succeeding 3 years, but thereafter reduced to 44,000m³ per year. Groyne design and spacing was often critical to beach sediment retention, demonstrating its importance to the viability of management by renourishment. The return to pre-nourishment volumes was fastest along the eastward part of this frontage, between Boscombe Pier and the end of the continuous seawall (Cooper, 1997). Beach levels along this sector also fell rapidly following ad hoc replenishment in 2000, necessitating the addition of rock armour in front of the seawall. A location at Fisherman’s Walk was used to deploy scour monitors in front of the seawall, May to June 2005 (HR Wallingford, 2006; DEFRA, 2007). Though a short-term study, this revealed the relationships between wave height and water level changes in relation to beach level fluctuations as well as liquefaction potential of the seabed subject to various wave loadings. The latter increases with saturation, and may induce sustained reduction of level. Gao and Collins (1994) estimate that the annual loss of beach sediment for the entire Bournemouth beach system, 1970-1990, was close to 100,000m³.
Further evidence of the negative budget prevailing over recent decades, (most recently confirmed for the period 2008-12) and probably since at least the early twentieth century following the progressive extension eastwards of defences (Bray, et al., 1996; Cooper, Hooke and Bray, 2002), is given by Harlow (2001). He uses the Bournemouth Borough Council profile database to calculate recession of both High and Low Water Marks between 1974 and 1997. Between 1974 and 1987, MHW receded by 1.36m per year over successive autumn surveys, and 0.3m per year between 1991 and 1997. The reduced rate during the later period reflects the short period elapsing since re-nourishment in 1988/89.
Analysis of the outer part of the beach between 100m and 450m of origin showed some volumetric increase following renourishment in 1975 and 1988 even though this part of the profile was not directly recharged. However, a major loss of volume in 1975, and a sustained gain between 1992 and 1995 that could not be linked to renourishment were also revealed. One implication is that large scale but short-term losses and gains in the nearshore area may occur independent of inter-tidal beach volume changes. Significant net onshore and offshore, but unsteady, transport pathways therefore would appear to operate.
Solent Beach to the Long Groyne, Hengistbury Head
This length of beach is characterised by a narrow upper gravel backshore and wide sandy foreshore with only relatively small-scale seasonal fluctuations of profile. Littoral drift is predominantly west to east, evident from updrift accumulation against the Long Groyne. Between 1938, its year of construction, and 1980, Lacey (1985) calculated that the position of Mean High Water advanced seawards by over 70m (1.8m per year). For Mean Low Water, advance over the same period was close to 100m. Sand dune accumulation occurs across the backshore close to the Long Groyne, assisted since the late 1980s by the trapping of saltating sand by gabion cages and fencing. This is due to the drying out of beach sand as the tide falls. Harlow (2001) estimates net dune sand accumulation of approximately 1,000m³ per year.
The morphodynamics of Solent Beach, and the co-adjacent beach between Double Dykes and the Long Groyne has been complicated since the mid-1970s by three artificial influences. The first is the effect of updrift sand replenishments (BIS2 in 1974/5, BIS3 in 1988/9, and BIS4 in 2006/10) which were transferred to this beach, after a time lag of two to three years, by eastwards littoral drift (Cooper, 1997; Harlow, 2012d). The second is a direct gravel recharge of Solent Beach in 1988, involving close to 150,000m³. The last is the construction of three rock groynes at Double Dykes in 1987, with additional armouring in 1991 (Photo 9). These have been relatively successful in accumulating sediment, especially sand and fine gravel (HR Wallingford, 1993; Halcrow, 1999). Inter-groyne beach area has increased as a result.
In comparison to updrift beaches, width (except adjacent to the Long Groyne) is relatively narrow due to greater exposure to wave energy and a net offshore sediment transport pathway estimated at 25,000m³ per year (Hydraulics Research, 1986). This is the result of deflection of littoral drift by Christchurch Ledge and Beerpan Rocks. Harlow (2001) calculates beach volume, 1974 to 1998, to be a mean of 1,578,000m³, but with peaks in 1977 and 1992 resulting from earlier updrift replenishment schemes. The natural tendency is for erosion, calculated by Gao and Collins (1994) to be 250m³ per metre length for the period 1974 to 1987; Harlow (2001) calculates that a loss of 200m³ per metre prevailed between 1979 and 1984 across a regularly re-surveyed profile approximately 800 m west of the Long Groyne. It was assumed that this period post-dated any benefit from the updrift beach nourishment in 1974/5. Cooper (1997) noted that net erosion east of Double Dykes was apparent by 1993, giving only some 4 years of temporary accretion following the BIS3 in 1988/89. Throughout the period 1974 to 1989 (Hydraulics Research, 1986; Cooper, 1997; Harlow, 2001), the beach 100-150 m west of the Long Groyne steadily accreted but erosion at an average rate of 100m³ per metre set in after 1993.
Since the early 1990s, there have been proposals to extend the sequence of rock groynes at Double Dykes further downdrift (eastwards) to help conserve beach sediment (HR Wallingford 1993; Bray, 1993). These have yet to be implemented, although current strategic management policy (Royal Haskoning, 2011) is to maintain this beach at a critical width that will allow only modest basal cliff erosion. Further information may be found at http://www.scopac.org.uk/sediment-sinks.html.
6. Summary of Sediment Transport Pathways
- Poole Bay has been formed by the inundation and erosion of soft Tertiary strata within the former basin of the Frome/Piddle River over at least the past 8,000 years. It is envisaged that marine erosion would have released vast quantities of predominantly sandy sediments as the embayment was formed.
- The Bay operates as a partially enclosed sediment circulation system that exports some sediment eastward to Christchurch Bay and also south and south-westward to offshore.
- Its sediment budget would have been sustained by cliff erosion, but following almost complete protection at the shoreline over the 20th Century it appears to have become completely closed to natural fresh sediment inputs (except for some continuing cliff erosion at Hengistbury Head). Four major episodes of beach replenishment have sustained the beaches and littoral drift pathways over the past five decades, although each have been characterised by some rapid initial losses and persistent longer term decline of volume.
- Relatively little sediment is stored within the beaches or seabed deposits of the bay. Exceptions are the large accumulations of Hook Sand and Studland Bay in the west.
- A well-established net eastwards drift operates throughout most of the bay and transports littoral sediments towards Hengistbury Head. The volume of drift has diminished over the 20th Century due to the effects of coastal protection although there have been notable short-term accelerations associated with the beach replenishment episodes. Drift is more complex along the Sandbanks peninsula, where reversals of net direction may occur and the littoral transport regime is affected by the presence of the East Looe tidal channel close inshore and the refraction of incoming waves over Hook Sand.
- A south-westward sand transport pathway operates across the offshore bed of Poole Bay. The final destination for this material remains uncertain, but it is probably lost offshore.
- An extremely complex transport regime operates in the vicinity of Hook Sand and Poole Harbour entrance. Hook Sand may be fed from offshore sources, although much of its volume could be inherited from earlier intervals in the erosion of Poole Bay. It is also fed by material moving into Poole Harbour Entrance that is flushed seawards by dominant ebb tidal currents, although those fluxes appear quite weak. Waves drive material onshore from the crest of the bank to the Sandbanks Peninsula where it may either drift eastward along the beach or move west to become entrained by tidal currents at Poole Harbour entrance and be flushed seawards back towards Hook Sand.
- Large pulses of sediment may move from Hook Sand south-westward into Studland Bay. This process is significant for it is the only feasible source for the rapid accretion of the South Haven peninsula. However, some 50,000 to 100,000m³ per year has accumulated during specific periods over the past 500 years (and accretion is continuing), involving a quantity greater than the present volume of Hook Sand. It could be that creation of the foundations of South Haven peninsula was either a "one-off" redistribution of stored material derived from the late-Holocene erosion of Poole Bay or, (partially related) the product of quasi-barrier landwards translation of sand in response to Holocene sea-level rise. The historical sediment dynamics of Poole Bay provide substantial scope for further research.
- The sediment dynamics of Hook Sand is of great importance for it would appear to have become an essentially fossil deposit with limited sources of fresh supply. If this is correct, the continued flux of sediment pulses into Studland Bay could deplete the bank, lowering its crest and exposing the vulnerable Sandbanks peninsula to increased wave energy. It should be noted that chart comparisons of the bank covering 1785 to 2005 have indicated relative stability.
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.channelcoast.org.
The Southeast Regional Coastal Monitoring Programme commenced in 2002. The Lead Authority is New Forest District Council, with data collection, analysis and reporting led by specialist teams at the Channel Coastal Observatory (CCO), Canterbury City Council and Adur and Worthing Councils. 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.
Detailed monitoring of sediment losses from past episodes of beach replenishment have been used by a variety of researchers as a basis to determine rates of drift. However definitive results could not be produced due to the non-typical initial behaviour immediately following placement of fill and an inability to reliably separate losses occurring due to net offshore transport from those resulting from drift.
A variety of modelling studies have also been undertaken, but these have tended to produce a range of transport estimates and on occasions predicted drift direction did not correspond with observations of sediment accretion within groyne embayments. Problems facing modelling involve selection of appropriate wave climates, the complex conditions occurring in the western parts of the bay, the extensive groyne fields that intercept drift and a tendency for cross shore and transport to dominate over longshore.
Two main opportunities for future research can be identified:
- Further work at Solent Beach and Hengistbury Head, where accretion against the Long Groyne may provide a means of validating drift on the upper beach. For this to be effective would require improved estimations of sediment bypassing of the Long Groyne by overtopping and outflanking. The work of Wright (1982) would provide a foundation for updating and extension of studies.
- It is recommended that modelling should attempt to better define the efficiency of the groynes in intercepting drift and to study the manner in which drift responds as sediment availability varies e.g. following a replenishment. Refer to discussion in Harlow (2012b) of the interconnected issues of the alignment of the existing defence line, the historic orientation and spacing of groynes and how these should be addressed by future management to achieve the reduction or elimination of beach morphological and sediment transport discontinuities.
8. Knowledge Limitation and Monitoring Requirements
The Southeast Regional Coastal Monitoring Programme commenced in 2002. Analysis of the 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.
Most of the shoreline of this sector has benefited from what is probably the longest established and most comprehensive and detailed beach monitoring programme carried out in the United Kingdom. Starting in 1974, Bournemouth Borough Council (Regional Beach Monitoring Survey since 2003) have undertaken biannual (spring and autumn) beach profile surveys for over 37 profiles. Topographical survey has been employed for the intertidal beach down to Mean Low Water with hydrographical survey used to extend this to 450m seawards. In addition, particle size analysis has been routinely undertaken for samples from 8 sectional profiles. From 1993 to 2012, detailed monitoring of beach levels adjacent to groynes, and computation of drift direction, has been routinely and systematically recorded along the Bournemouth frontage. Full details of this work, together with analytical conclusions, are given in Harlow (2001; 2005b; 2012a to g).
Other regular monitoring activities includes (i) frequent bathymetric surveys of the Swash channel and Poole Harbour entrance channel by Poole Harbour commissioners, and (ii) regular intertidal beach profile resurveys of the Sandbanks peninsula, by Borough of Poole Council. Other continuous data sets include tidal levels, at Bournemouth Pier, and wind speed and direction, together with a directional wave rider buoy offshore of Boscombe.
Notwithstanding results from the Southeast Regional Coastal Monitoring Programme, and the summarised information collated in the Hurst Spit to Durlston Head SMP2 (Royal Haskoning, 2010b), recommendations for future research and monitoring that might be required to inform management are as follows:
- Existing surveys, commencing in 1992, are limited to the inter-tidal beach width. There are complex, and as yet inadequately understood, transfers of sediment between Hook Sand and the adjacent shoreline. Repetitive hydrographic surveys, out to a distance of approximately 1 km, would give a clear view of the mobility of the seabed of this dynamic zone, especially the stability of the East Looe Channel and the crest of Hook Sands. This work could be complemented by sediment samples and detailed mapping of bedforms.
- Some more frequent beach profiling is needed to determine the typical seasonal changes of the beaches and their responses to storms. Due to a well-established tendency for cross shore sediment exchanges associated with the growth and decay of the nearshore bar the sub-tidal portions of the active profile would also need to be surveyed.
- To understand beach profile changes it is important to have knowledge of the beach sedimentology (grain size and sorting). Sediment size and sorting can alter significantly along this frontage due to beach management, especially following major recharge operations. Ideally, previously established sampling programmes need to be continued to determine longer-term variability within intervals between beach replenishments. 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.
- The dynamic behaviour of Hook Sand is a crucial element of the sediment budget of western Poole Bay. Changes in planshape since the mid nineteenth century, revealed by comparison of successive hydrographic charts, indicate fluctuations in volume and also alternating phases of erosion and accretion on both its eastern and western flanks. Sediment transport pathways to and from this sandbank are also uncertain. Detailed and more frequent bathymetric surveys of the whole of the bank would give better quantitative knowledge of the magnitude and frequency of phases of net erosion and deposition. Ideally, they should be combined with some seabed sediment sampling. The latter would provide valuable information on the potential for onshore sediment transport through the compilation of large-scale maps of sediment distribution, and analyses of particle size and sorting to derive bed transport vectors as was undertaken for Chichester Harbour entrance by Geosea Consulting Ltd (1999). It would be useful also to map bedforms throughout as they could provide additional evidence of directions of sediment transport. It might, in particular, throw light on the important question of the stability of the bank and its capacity to feed sediments ashore. A second application of the bathymetry would be to enable reliable wave transformations and determination of breaking wave climates along the shoreline. These, in turn, should enable improved modelling of sediment transport.
- Linked to the foregoing point, the details of sediment transport in the nearshore and offshore zones of all parts of Poole Bay are subject to considerable uncertainty. Further work is needed to confirm the bedload movement of sand and give improved insights into the pathways of transport that move sediment offshore from beaches of the eastern sector of this coastline. Is it removed westwards, or does it move out of this sub-cell through removal to Dolphin Sand? Is it possible for sand released at Hook Sand to move initially towards the adjacent shoreline, but thereafter to move offshore and south/southwestwards to contribute to stores such as those in Studland Bay and at Poole Bar? There are probably several alternative pathways, depending on grain size in relation to stresses set up by both wave motion and tidal currents.
- Additional work is needed to establish the extent that Hengistbury Long Groyne functions to intercept drift and to produce reliable estimates of quantities of bypassing that provide inputs to the Christchurch Bay unit.
- Potential of beach recycling from offshore stores e.g. Dolphin Sands and Hook Sands.
