1. Introduction
Southampton Water and the estuaries of the Test, Itchen and Hamble are drowned remnants of a major tributary of the ancestral "Solent River". This river system became fully established during the Pleistocene when fluctuating sea-level resulted in numerous phases of inundation (marine sedimentation) and emergence (fluvial down cutting). The most recent emergent phase (Devensian glacial stage) is recorded by deposition of a series of gravel terraces during stands in sea level regression (Everard, 1954; Hodson and West, 1972; Allen and Gibbard, 1993). During low sea-level, the main channel was excavated to a depth of at least -21mOD at Calshot, -14mOD at Fawley and -9mOD at Southampton Western Docks (West, 1980). The river system was inundated during the Holocene transgression and borehole data suggests that the sea had entered Southampton Water prior to deposition of the earliest dated sediments around 6500 to 6800 years B.P. (Hodson and West, 1972).
There have been several subsequent minor transgressions, regressions and periods of stationary sea-level (see unit on Quaternary History of the Solent for further details). The estuary has increased in size since the mid-Holocene through overall submergence and erosion of the north-east coast. The past 7,000 years has been characterised by sedimentation of fine material at a rate similar to local relative sea-level rise over the past 5-6000 years of 1.1 to 2.0mm per year, contributing 3-4000m³ for every 1mm of sea-level rise (Hodson and West, 1972; Long and Tooley, 1995; Long and Scaife, 1996, 2000). Dyer (1980) suggests a spatially variable accretion rate of 2-10 mm per year over the past 3-6,000 years. The Pleistocene and Holocene sediment sequence (Everard, 1956; Long and Scaife, 2000) overlies and mostly conceals underlying Eocene bedrock, but the latter are exposed in marginal cliffs on the north-east shore and where the main channel has been incised below -10mOD.
The estuary boundaries are Calshot Spit (Photo 1) in the west and the Hamble estuary (Photo 2) and Hook Spit to the east. Southampton Water is therefore a relatively narrow and enclosed meso-tidal estuary, subject to limited wave action generated in the estuary itself. Maximum significant wave heights are between 0.57 and 1.02m (extreme of 1.5m) propagated up-estuary from the Solent, but waves in excess of 0.5m occur infrequently and only when winds from the south-east are able to generate (refracted and diffracted) waves across the fetch of the eastern Solent (New Forest District Council, 2010).The west coast is relatively sheltered from wave erosion in comparison to the opposing shoreline. Thus, tidal currents dominate sediment transport, especially on western shores and in the inner estuary where fine sediments have accreted and salt marsh has provided a buffer against wave abrasion (Photo 3). The tidal regime is unusually complex controlled by a degenerate amphidromic system to the west of the Solent, and by the resonance effect of the eastwards narrowing of the English Channel, modified by the hydraulic characteristics of the Solent (Levasseur et al., 2007; Guo and Zhang, 2008). An initial flood stand lasting approximately two hours is followed by a normal flood, including high water stand of 5 hours; and a shorter ebb period of 5 hours. As a consequence, ebb currents in the mid and lower estuary are significantly more rapid than flood currents, but this asymmetry virtually disappears upstream from the Western Docks (Levasseur et al., 2007). The mean spring tide range is 4.05m and that for neaps is 2.0m. Buber et al., (2009) provide a numerically modelled frequency analysis of extreme tidal levels at the head of the estuary in the vicinity of Eling. Refer to Levasseur et al., (2007) for a more detailed account of estuary hydrodynamics, in the context of numerical modelling of its dynamic characteristics. In general terms, Southampton Water may be regarded as a low flux system, whose morphology and sediment budget are no longer in equilibrium with estuary hydrodynamics as a result of marginal land claim and other artificial modifications of the tidal prism (ABP Research and Consultancy Ltd., 1995e). Residence (flushing) time is between 5 to 10 days (according to location), with a partially-mixed water mass stratification (Dyer, 1980; Sharples, 2000; Levasseur et al., 2007).
A wide range of sediments exists in Southampton Water as a result of Holocene regressive and transgressive phases and contemporary fluvial and marine deposition. Over the past 200 years, the system as a whole has been hydrodynamically stable, but there have been substantial changes affecting the vertical and horizontal dimension of the inter-tidal area. Sediment transport and circulation is complex, with as yet limited understanding of interactions with the contiguous external Solent system.
A major new source of coastal data is from the Defra-funded National Network of Regional Coastal Monitoring Programmes. The Programmes consist of topographic beach surveys, nearshore bathymetry, aerial photography, lidar, coastal hydrodynamics (waves and tides) and terrestrial habitat mapping. Specifications for data collection are consistent for all regional programmes and the data and analysis reports are made freely available under the Open Government Licence from www.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).
The Southeast Regional Coastal Monitoring Programme measures nearshore waves using a network of Datawell Directional Waverider buoys. The nearest measurement stations to this cell are at Milford-on-Sea and Hayling Island. The buoy deployed at Milford-on-Sea is in 10mCD water depth. Between 1996 and 2012, the prevailing wave direction was southwest-by-south, with an average 10% significant wave height exceedance of 1.31m. The buoy deployed at Hayling Island is in 10mCD water depth. Between 2003 and 2012 the prevailing wave direction was south-by-west, with an average 10% significant wave height exceedance of 1.26m (CCO, 2012).
2. Sediment Inputs
2.1 Marine Sources
F1 Input of Suspended Sediments from the Solent
Entry of coarse sediments into Southampton Water from the Solent is severely limited due to lack of suitable transport mechanisms although suspended sediments can move into the estuary. The asymmetry of the tidal curve results in significantly more rapid, short-lived ebb currents compared to the flood (Webber, 1980). Bedload transport of sand and fine gravel is therefore likely to be in a net seaward direction, determined by maximum current velocity. By contrast, net suspended sediment transport is likely to be into the estuary due to the longer duration of the flood current (ABP Research and Consultancy Ltd., 2000; Dyer, 1980; Barton, 1979). Limited field evidence of this effect is available from measurements of suspended sediment concentrations, which indicate supply from the Solent. Restricted measurements in the Itchen estuary (Qadri, 1984) and Southampton Water (Dyer, 1980) suggested that concentrations increase towards potential sources.
Mean suspended sediment concentrations of 2mg/litre to 50mg/litre (mean of 35-40mg/litre) are reported by Dyer (1980) and ABP Research and Consultancy Ltd., (2000c). Concentrations tend to fall in the up-estuary direction, to average values of 5-20mg/litre between Southampton Dockhead and the Container terminal. Seasonal fluctuations are considerable, with winter peaks in excess of 100mg/litre even at the mouth of the River Test. Qadri (1984) reports very low mean concentrations, 0.6 to 2.5mg/litre in the lower Itchen in July. Sampling by Dyer was from the surface and at mid-depth over an 18 month period at three stations, with the highest concentrations recorded at mid-depth at the mouth of Southampton Water. These findings revealed that concentrations within Southampton Water were especially low compared to other regional estuaries. Gradients in concentrations indicated that the primary source of suspended sediments was the Solent and the English Channel, with especially low summer concentrations being related to reduced wave energy. Humby and Dunn (1975) reported suspended sediment concentrations of 20-200 mg/litre for other Solent estuaries, which were markedly less than concentrations of 300-2000 mg/litre and 2000 mg/litre recorded in the Thames, Mersey and other British estuaries. The unusually low concentrations characteristic of Southampton Water can be attributed to its unique tidal regime and small tidal range, but it is probable that other sediment sources, e.g. fluvial and sewage, may contribute more significantly elsewhere.
Sediments are removed from suspension in sheltered parts of the estuary, particularly the western side where accretion is facilitated by reduction of flow velocities imposed by mudflats and marshes, swards and very low wave heights and energy. The long still stand at high water assists this process.
Mineralogical analyses have revealed a clay assemblage characteristic of local Tertiary sediments (Hodson and West, 1972; Algan et al., 1994). These may have been derived from the Solent but could also have been eroded from either the north-east shoreface of Southampton Water or the catchments of the Itchen, Test and Hamble (Algan, 1993; Algan et al., 1994). Marine input of suspended sediments has not been quantified, but significant net accumulation of fine-grained sediment in the form of mudflats and marshes, has occurred since at least 6,800 BP, according to radiocarbon dating of organic horizons from borehole samples (Hodson and West, 1972, Long and Scaife, 1996; 2000).
2.2 Fluvial Input
FL1 Test; FL2 Itchen; FL3 Hamble
Three major rivers flow into Southampton Water, the Test, Itchen and Hamble, draining a combined catchment area of a little over 1500km². Part of the discharge of these rivers is derived from Chalk groundwater which results in fairly stable hydrographs (Webber, 1980). Southern Water Authority and Environment Agency discharge data has been analysed revealing mean flows of 5.37 (range of 3-13) m³ per second for the Itchen (Wright and Barnard, 1964; Webber, 1980; Qadri, 1984; Rendel Geotechnics and University of Portsmouth, 1996); 11.02 (range of 6-25) m³ per second for the Test (Wright and Barnard, 1964; Webber, 1980; Rendel Geotechnics and University of Portsmouth, 1996) and mean flow of 0.47 (range of 0.6 to 1.3) m³ per second for the Hamble (Wright and Barnard, 1964; ABP Research and Consultancy Ltd., 2000c). Hydes (2000) gives an estimate of 1.54x10ˉ⁶m³ per year of freshwater discharge into Southampton Water. Both discharge and velocity characteristics are not conducive to bedload transport as extreme values have not been recorded.
Samples of suspended sediments were collected at four sites on the Itchen estuary between Itchen Bridge (estuary mouth) and Woodmill Bridge (near the tidal limit). Measurements covered spring and neap tides and surface, mid-depth and bottom levels and revealed that suspended sediment concentrations were greatest at the mouth (1.5 - 2.5mg per litre) and declined upstream (0.6mg per litre). It has been suggested that seawater entering Southampton Water on the flood current is much richer in suspended sediments than the fluvial discharge (Qadri, 1984). Rendel Geotechnics and the University of Portsmouth (1996); Velegrakis et al. (1999) calculate that the Itchen and Test rivers contribute between 2,000 to 4,300 tonnes per year of suspended load to the upper reaches of Southampton Water. This is less than 5% of potential supply, most of which is diverted into storage in the extensive flood plains of both rivers. The Hamble River is unlikely to supply in excess of 300-350 tonnes per year. Webber (1980) calculated that fluvial discharge into Southampton comprised only 1.3% of the neap tidal prism. Information relating to bedload inputs have not been monitored, but supply may be restricted to occasional peak discharges. Figures of estimated potential bedload supply of 3,668 tonnes per year (Itchen) and 11,324 tonnes per year (Test) are an order of magnitude above actual delivery to Southampton Water owing to numerous upstream sources of storage (Rendel Geotechnics and University of Portsmouth, 1996). Bray (2000) proposes a total potential fluvial sediment input of a maximum of 17,000m³ per year, but this includes bedload materials that would be stored in the lower river channel reaches and upper estuaries without being supplied directly to Southampton Water.
2.3 Coastal Erosion
Three distinct areas undergo erosion: the shoreline between the Hamble and Itchen rivers; the saltmarsh margins of the south-west shore between Calshot and Eling; and intertidal flats in several areas of the estuary. Potential erosion losses are prevented along some 26km of estuary-side frontage by defences maintained by a mix of statutory authorities, commercial and private ownerships (Halcrow, 1998; ABPMer, 2006).
Analysis of Coastal Monitoring Programme 2003 and 2012 lidar, aerial photography and topographic baseline survey data, combined with other datasets, academic research and historical studies has enabled sediment budgets, transport rates and directions to be identified or verified.
The coast between the Hamble and Itchen rivers comprises low cliffs up to 9m height cut in sandy gravel and medium to coarse gravel sediments of the Pleistocene (Solent River) (Photo 4). At Hamble Common, a cliff line is cut into alluvium directly overlying bedrock. Historically, erosion was probably continuous, but over the past 100 years there has been piecemeal construction of defences and significant lengths are now protected to differing standards (Halcrow Group, 2006) (Photo 5). At present, unprotected cliffs are actively eroding through basal undercutting and slab failure (ABPMer, 2006). Wave heights are characteristically low due to swell attenuation and limited internal fetch, but a maximum height of 1.1m was recorded during a storm in 2001. Weathering makes a small additional contribution, but there is no evidence of seepage erosion.
A slow rate of cliff retreat at Weston Point and between Netley Castle and Netley Hard (Hydraulics Research, 1987; Posford Duvivier, 1994; 1997; New Forest District Council, 1993) is supported from Coastal Monitoring Programme data. Although the extent of vegetation cover on cliff top and slopes limits quantification for the entire cliff face, analysis for the slopes at the eastern section of the Country Park indicates rates of erosion and volumes are consistent and low, less than 1,000m³ per year. South of Royal Victoria Park, the cliff line is vegetated and partly wooded, thus suppressing potential erosion. Map comparisons over the period 1870-1965 have revealed retreat at a rate of 0.1 to 0.5m per year, (Posford Duvivier, 1994; 1997; 1999) with some coastal advance by land claim at Hamble Point (Hooke and Riley, 1987). Highest recession rates (0.5m per year) appear to be at Hamble Common (Photo 6) and between Netley Abbey and the Royal Victoria Country Park. Site observations in 1988 by Southampton City Council revealed a low (2-3m) gravel and clay cliff 250m long, which was at that time retreating rapidly due to basal erosion attributable to sea wall deterioration. Approximately 20-25% of eroded material, released by cliff erosion, constituting approximately 50-100m³ per year is sand and gravel that is retained on local beaches, with the remainder about 400m³ per year presumed to be removed as suspended load (Posford Duvivier, 1997). Although the yield is very low it appears to have been sufficient in the past to supply the bulk of sand and gravel on the adjacent narrow beaches (Hydraulics Research, 1987). Total supply rate from cliff degradation is influenced by either intermittent slope destabilisation due to the disturbance of vegetation by basal erosion and/or shallow slides. It is also spatially differentiated by protection structures, which vary in their effectiveness.
Saltmarsh
Although most areas of mudflat and saltmarsh have eroded considerably during the 20th Century, narrow margins and increasingly fragmented areas of inter-tidal mudflat and saltmarsh are still present between Hythe and Calshot on the northwest shore of Southampton Water, and within the River Hamble. E2 arrows have been added to represent input of fine sediment (silt and clay) from saltmarsh erosion. Retreat of Spartina marsh and consequent release of fine sediments has been widely reported for the western shore of Southampton Water - see Photo 7 and Photo 8 (Goodman et al., 1959; Hydraulics Research, 1987; ABP Research and Consultancy Ltd., 2000a, 2000c; Halcrow, 1998; Johnson 1996, 2000; Baily and Pearson, 2007; Cope et al., 2008) and within the Hamble estuary (Goodman et al., 1959; Nature Conservancy Council, 1984c; Hooke and Riley, 1987; Gray et al., 1993; Baily and Pearson, 2007; Cope et al., 2008; Bray, 2010). This erosion commonly takes two forms; (a) frontal retreat resulting in formation of cliffs 0.5m - 1.5m high separating marshes and mudflats; and, (b) internal dissection and fragmentation of the marsh by erosion of channel margins and development of erosion "pans" (Goodman et. al., 1959; Baily and Pearson, 2007), although this is less developed here than in other parts of the Solent (ABP Research and Consultancy Ltd., 2000c; Williams, 2006). One important, and arguably dominant, cause appears to be "dieback" of Spartina anglica around inter-creek 'pans' which reduces the stability of accumulated saltmarsh sediments, making them susceptible to wave and/or tidal abrasion. Further discussion of this phenomenon is given in section 5.4. Sediments eroded from saltmarshes must be regarded as a largely or partly re-circulating input released from previous storage.
Spartina now appears unable to naturally colonise previously unvegetated mud flats, or to re-occupy "die back" areas. Field trials conducted by Dicks and Levett (1989); Dicks and Iball (2005) only achieved regeneration by transplanting techniques. Sediments released by erosion are therefore unlikely to be re-incorporated into the small areas of expanding inter-tidal marshes elsewhere, although they could be deposited upon existing marsh surfaces. Evidence of erosion extends back to the 1930s when the 'die-back' phenomenon was initiated (Goodman et al., 1959; CEGB, 1988; Gray and Benham, 1990; Gray et al., 1991). Ordnance Survey map comparisons covering the period 1870-1965 are not an effective means of documenting longer-term saltmarsh erosion because it is uncertain how accurately the marsh margins were mapped at the beginning of this period (Hooke and Riley, 1987). Some of the most reliable map information relates to the Marchwood-Eling segment, where significant marsh growth was recorded for the period 1910-1932 followed by frontal erosion at 1.7m per year over the period 1932-1965 (Hooke and Riley 1987). A similar rate of erosion was recorded for the edge of the saltmarsh between Ashlett and Calshot over the period 1910-1965, but in other areas map evidence is inconclusive and only indicates increased saltmarsh dissection (Hooke and Riley 1987). Air photo analysis covering the period 1962-1985 revealed saltmarsh erosion at up to 4m per year between Calshot and Fawley (CEGB 1988). This suggests that erosion may have accelerated in recent decades, but analysis only covered a limited spatial area.
A photogrammetric study of saltmarsh change between 1971 and 2001 revealed continuing major losses in saltmarsh area, primarily as a result of the die-back and loss of Spartina marsh (Colenutt, 1999, 2001; Bray and Cottle 2003; Baily et al. 2002, Baily and Pearson, 2007). Furthermore, Colenutt (2001) estimated that saltmarsh erosion losses were likely to continue in the future. On this basis, it is clear that fine sediments are likely to continue to be released into the estuary as marsh retreat proceeds.
The sediment store of fine-grained sediment currently "locked up" as saltmarsh and mudflat is being relatively rapidly transferred to suspended load throughput and loss from the Solent system. The implication is that the current hydrodynamic regime strongly favours store depletion, and that mudflats and outer marshes are in disequilibrium with forcing factors. As the causes promoting erosion are unlikely to alter in the foreseeable future, this situation will continue. Indeed, as relative sea-level rise accelerates, it will probably intensify in coming decades.
The following table presents saltmarsh extent loss for Southampton Water and the River Hamble, mapped from historical aerial photography interpretation, calibrated by map analysis and ground control, using a sequence of sorties between 1946 and 2000 to 2002 (Baily and Pearson, 2007; Cope et al., 2007; Cope et al., 2008). The findings are presented in the Solent Dynamic Coast Project (Cope et al., 2008) which adds to the datasets mapped by Baily et al., (2002) and Baily and Pearson (2007) as part of the Solent Coastal Habitat Management Plan (Bray and Cottle, 2003).
Table 1 presents the earliest (Year 1) to most recent (Year 2) saltmarsh extent mapping for Southampton Water and the River Hamble. The % saltmarsh loss per year can be used to compare the differences in loss with other units such as the West Solent or Portsmouth, Langstone and Chichester Harbours. Variability in % saltmarsh loss per year is attributed to historical land reclamation, exposure to wave attack, elevation of the marsh, the presence of salt pan formation, sea level rise leading to coastal squeeze, Spartina dieback and dredging (Baily and Pearson, 2007; Cope et al., 2008). 0.8% of the 1.2% saltmarsh loss per year in Southampton Water is attributed to historical reclamation. Cope et al. (2008) consider that edge erosion has been the principal process driving saltmarsh decline, with the highest rates of loss occurring between 1971 and 1984. Some acceleration in contraction has occurred since the early to mid-1990s. There has been some variation in the rates and spatial patterns of degeneration and extinction in different parts of the western shoreline of Southampton Water. For example, at Calshot saltmarsh cover declined by 73%, 1940- 2001 (Cope et al., 2008), a rate of 1.2% per year; the latter was 2.45% per year, 1991 to 2006 and increasing to 4.3% per year in the latter part of this period, 2001-2007 (Williams, 2006; Bray, 2010). Almost a third of the total loss at Calshot, at least up to 2001, was due to land claim associated with berthings and approach channels for the Fawley oil refinery terminal and power station. At Eling and Marchwood, 36ha of saltmarsh in 1971 reduced to 18.7ha in 2001, i.e. a 49% loss (Baily and Pearson, 2007). However, this is a site where there has been a significant loss of mudflats and marsh as a result of dredging to give improved access to the adjacent container port. Edge retreat at Hythe occurred at 2.9m per year, 1971-2001, whereas it was 4.3m per year at Calshot during the same period (Baily and Pearson, 2007). It is likely that anthropogenic factors account for these contrasts, but there are, as yet, no certain explanations. Decline is not universal, however. New Forest District Council (2010), for instance, record expansion as well as northwards migration of saltmarsh (26%, 1996-2008) in the lower Test estuary, north of the Redbridge Causeway.
Inter-tidal Mudflat Erosion
Map comparisons have revealed retreat of Low Water Mark at a significantly more rapid rate than High Water Mark, so that the intertidal zone has narrowed and steepened This occurred during the period 1870-1965 and was recorded for all parts of Southampton Water except the reclaimed docks between the Itchen mouth and Redbridge (Hooke and Riley 1987). Although the low gradient of the intertidal zone makes specification of Low Water Mark susceptible to error due to sea level datums being changed over successive map editions, the reality of net erosion was confirmed by the distribution of eroding mudflat margins, evident from field inspection (Hooke and Riley 1987). Intertidal narrowing was quantified by recession rates derived from map comparisons at the following locations: Calshot-Ashlet (1.7m per year 1870 to 1965), Fawley Refinery (0.6m per year 1868 to 1966); Weston Point (up to 3.2m per year 1870 to 1965); and Itchen mouth to Hamble Point (3.2m per year 1870 to 1965). Intervening areas showed similar trends but at generally lesser rates (Hooke and Riley 1987, Halcrow, 1998). Comparable rates, are reported by Oranjewoud (1988, 1992) and ABP Research and Consultancy Ltd., (2000c) and are reviewed by Halcrow (1998). Short-term erosion rates as high as 8m per year may have prevailed in the 1940s, but the fastest recession rate for recent decades is approximately 4m per year, between Calshot and Fawley, 1962 to 1985 (ABP Research and Consultancy Ltd., 1995c).
Causes of marsh and mudflat erosion are uncertain but several, probably inter-related, possibilities have been suggested. These include an absolute shortage of sediment (Geodata Unit, 1987), on-going relative sea-level rise; land claim, coast defences, channel dredging and the failure of some intertidal marshes and mudflats to store sediment released by co-adjacent erosion. Barton (1979) reported that some deeply dredged navigation channels were as close as 200m to the leading edges of inter-tidal mudflats, e.g. Marchwood-Eling (Photo 8). Here, inter-tidal slopes were frequently much steeper (9-13 degrees) than those more remote from dredged channels and where cut into Pleistocene gravels. This was attributable, at least in part, to failure of oversteepened dredged channel margins. Some localised losses have specific causes, such as oil spillage on Fawley marshes causing marsh plant mortality (Oil Pollution Research Unit, 1994); there has been a positive effort to encourage regeneration at this location, thus compensating for erosion losses. Waves created by shipping movements (48,430 vessel movements in 1998) may make a small input to the erosion of mudflat margins in the mid and upper estuary. ABP Research and Consultancy Ltd., (2000a) estimate that they contribute about 2% of energy received at Netley Shore, but some 10% at Hythe. Most of the larger vessel movements are likely at or close to high tide when waves generated are likely to impact on saltmarshes rather than be dissipated by the lower foreshore. Chellow et al. (2010) examined evidence for inter-tidal mudflat erosion caused by wave reflection from backshore defences along the Netley shoreline, detailed in Section 5.3 of the following text.
Research using monitored sites at Hythe Flats measured suspended and bedload transport pathways due to reversing tidal currents (varying between 0.02 and 0.2m per second) (da Quaresma, 2004; da Quaresma et al., 2007). Highest concentrations of suspended sediments occurred during the young flood and later ebb stages, with net movement out of the site. By contrast, net bedload transport proved to be landwards, as indicated by tracking marked surficial shells. Shell transport rates were up to 0.7m per tide over the mudflat surface during highest wave energy conditions. Mudflat lowering, assisted by shell abrasion, was thus inferred, and a relationship between bed erosion and retreat of the saltmarsh cliff edge observed. It was apparent that mudflats were more stable during periods of highest rates of cliff recession, and vice-versa. This can be ascribed to inter-tidal profile flattening caused by higher wave energy, releasing sediment and subsequently raising flat elevation.
Inter-tidal and erosional losses have also taken place in the estuaries of the main tributaries of Southampton Water; however, only the River Hamble has been examined in any detail. Volumetric analysis of bathymetric charts, 1965-1993 (ABP Research and Consultancy Ltd., 1994a), shows an overall trend for erosional loss downstream from the M27, railway and Bursledon road bridges, but a slight positive balance upstream. Most of the losses and gains were related to changes in the position and depth of the main channel. However, Cundy and Croudace (1995) reported consistent sub-tidal mudflat erosion adjacent to vertically accreting saltmarshes, but with no correlation between accumulation rates and site elevations. It is uncertain if changes in the sediment budget are primarily the result of channel dredging and marina construction. A net volume loss of subtidal sediment of 156,000m³ is computed for the previous ten years based on an average rate of erosional retreat of mean low water mark of 4.7mm per year. Refer to unit on River Hamble to Portsmouth Harbour Entrance for additional detail.
Chart comparisons (ABP Research and Consultancy Ltd., 2000c) indicate that the planform of most sub-tidal areas within Southampton Water have been relatively stable over the past 100-200 years, but some fluctuations that may have occurred during this period cannot easily be detected from this evidence. There is no specific evidence to support the view that erosion of inter-tidal areas has been balanced by gains in the sub-tidal environment. However, most authorities propose that suspended load sediment derived from marshes and mudflat erosion is subject to re-circulation at various timescales, within the estuary. The fate of such material, however, is not understood fully.
Using a largely theoretical approach based on the assumption of a rate of 1mm per year vertical erosion, Posford Duvivier (1999) calculate that: (i) approximately 4,500m³ per year to 13,500m³ per year of fine sediments are removed from the inter and sub-tidal shoreface between Hythe and Fawley; (ii) a loss of 900-2,600m³ per year at Fawley, where there are protection measures and attempts to stimulate Spartina recolonisation; and (iii) loss of 3-9,000m³ per year between Fawley and the marshes protected by Calshot Spit. Spatial variations of erosion losses and shoreface width would appear to correlate closely. These losses are in addition to those involving marsh retreat on the mid to upper foreshore.
3. Littoral Transport
Littoral drift towards the entrance of Southampton Water from both the west and east Solent is indicated by the presence of Calshot and Hook Spits respectively. These features may operate as partial transport boundaries to gravel and coarse sand, with some of the material stored and some transported offshore into the Solent by strong ebb currents in the main tidal channels and/or by wave action. Boreholes reveal several gravel horizons beneath the saltmarshes in the lee of Calshot Spit, suggesting cyclic or episodic erosion and accretion, although stratigraphic and historic records indicate that the position of the Spit has been stable for several millennia (Hodson and West, 1972). A relative lack of mobile gravel and coarse sand deposits within Southampton Water suggests that entry of these materials is very limited or currently non-operative (Dyer, 1980).
Analysis of Coastal Monitoring Programme lidar, aerial photography and topographic baseline data indicates that discernible sediment transport is only evident for the beach and foreshore between the mouths of the Hamble and Itchen rivers where narrow upper beaches of gravel and coarse sand are developed (Photo 9 and Photo 10). A possible transient weak drift divide between Royal Victoria Country Park and Hamble Common is indicated with north-westward and south-eastward drift evident, although rates are consistent and less than 1,000m³ per year.
Webber (1980) considered that bedload and suspended load longshore transport were in opposite directions along this entire frontage, the latter moving northwards. The temporal stability of this dual, divergent transport system has not been assessed and no quantitative details are available. Maximum significant wave heights generated by the longest available fetch of 5140m are between 0.57 and 1.02m, although in reality waves from this narrow window are infrequent and prevailing waves are significantly lower. It is thought that significant drift can only be generated when waves combine with peak tidal stream velocities (ABP Research and Consultancy Ltd., 2000c; Posford Duvivier, 1999). There is no evidence for longshore transport along the western shore of Southampton Water, where maximum theoretical mean significant wave heights are between 0.53 and 0.94m (extreme of 1.5m) at the Hythe shore, from a maximum south-easterly fetch of 4150m (ABP Research and Consultancy Ltd., 2000c). Again, these waves are infrequently generated in reality and the shoreline is sheltered from prevailing south-westerly waves.
4. Sediment Outputs
4.1 Estuarine Transport
EO1 Ebb tidal south-eastward out of the estuary
Ebb tidal currents in the estuary downstream of the Docks are of greater velocity than flood currents due to a shorter period of flow, due to the long high water stand. It therefore follows that sediments subject to bedload transport should undergo net output from the estuary, widely used as an explanation for relatively slow siltation in the main channels (Robinson, 1963; ABP Research and Consultancy Ltd., 1993, 2000c). The scouring process of the ebb tidal current has not been examined in detail and only limited information exists as to grades and volumes of sediment output (ABP Research and Consultancy Ltd., 2000a and c). Despite this, several studies have measured current velocities, either in connection with estuary hydrodynamics or for estimation of suspended sediment concentrations and siltation rates. Currents are generally strongest in the main channel between Calshot and Hythe. Peak surface ebb velocities of 1.02m per second and of 1.01m per second have been recorded off Fawley and Hythe respectively (Flood, 1981; Dyer, 1970) and Hamble (Gifford and Partners, 1989). A peak ebb velocity of 0.52m per second was recorded just above the bed of the main channel by Flood (1981). Seabed stresses imparted by ebb currents are approximately four times those of the corresponding flood flows. Current speeds decrease upstream, with maximum surface ebb velocities of 0.79m per second in the Hamble (Qadri, 1984), and less than 0.55m per second off the Royal Pier, Southampton (ABP Research and Consultancy Ltd., Ltd, 1989; 2000b). Velocities are fastest in the deepest parts of the lower estuary channel but as the water mass moves upstream on the flood tide, water levels become higher. This induces downstream flow from the shallow inter-tidal areas marginal to the principal channels. Thus there are two vectors of water flow operating simultaneously, which are most pronounced at the mid-flood stand. Motion is slowed by the frictional contact in shallow water with mudflat surfaces. A critical factor is that there is slack water in the area between opposing flows, thus promoting potential sedimentation by inhibiting re-erosion by ebb currents of previously deposited sediment (McLaren, 1987; Price and Townend, 2000).
Flume studies by Wright and Leonard (1959) simulated a variety of flow conditions in the main channel and determined a series of critical velocities for gravel transport (>13mm diameter size). These were established both across and along a variety of slopes to recreate conditions over dredged channels. The results indicated lower values when currents were normal to slope contours and when slopes were steep; the lowest critical velocity was 0.91m per second on a 32 degree slope. Although these laboratory studies simplified actual field conditions, the lowest critical velocity was almost twice as great as the maximum bottom velocity measured in the field (0.52m per second, Flood, 1981). Whilst ebb-flood current asymmetry results in some selective seawards movement of coarser sediment grades, it appears unlikely that any significant gravel output can occur by ebb currents in any part of the main channel even on the less stable sloping margins of dredged channels. However, there is considered to be some potential for output of fine gravel under exceptional conditions. Although the Brambles Bank, in the central Solent, represents in part the ebb tidal delta of Southampton Water it is unlikely to actually receive much contemporary material flushed out of Southampton Water. Instead, it is either a relic feature, or is has formed from sediment movements within the Solent that are influenced by ebb tidal flows emerging from Southampton Water. The intertidal and subtidal portions of Hamble Spit, approximately 1000m south of Hamble Common could be considered as being derived from the ebb tide delta of the Hamble, composed of coarse silt, sand and fine gravel.
Sediment transport in the main estuary channel is clearly indicated by linear erosion furrows aligned parallel to ebb current flow (see map). These bedforms are developed in predominantly fine sediments, suggesting that only these grades are transported in this area (Dyer,1970; Flood, 1981; Ziedler, 1990). Furrows reported by Flood (1981) represent the erosion of some 25,000m³ of fine-grained sediment, but over an indeterminate time scale. However, any net output of suspended load clay and silt by tidal currents is at least balanced by input (section 2.1).
Sediment transport in Southampton Water is thus limited compared to many other English estuaries, due to shelter from wave action, restricted tidal range and a long period of standing water. Tidal current velocities are modest and combined tide-wave interaction is limited by low wave heights imposed by very restricted wave fetch. The net direction of tidal transport differs for suspended sediments compared to bedload sediments, suspended sediment transport being inward and bedload transport outward.
Further evidence of net seaward transport along this pathway has been provided by a series of sandwaves identified by side-scan sonar traverses of the main channel between Fawley and Hythe (Flood, 1981; Alluvial Mining Co. Ltd., 1994). These indicate net seaward transport, with sampling revealing a composition of muddy coarse sands and fine gravels. Current measurements indicate that peak ebb velocities decline upstream to Southampton Docks, so the maximum sediment size transported must also decline in this direction.
In the inner estuary (Dockhead to Redbridge) tidal current asymmetry rapidly reduces, and current velocities are also less than in lower Southampton Water. The Docks are therefore a nodal point for both tidal hydraulics and sediment movement.
4.2 Land Claim
Significant areas of Southampton Water have been reclaimed beginning in 1836 (Eastern Docks, Southampton) (refer to maps in Cope et al., 2008) and extending up to the late 1980s (Dibden Bay). Land claim at the docks at Southampton is described by Barton (1979) and Long and Scaife (2000), and the western shore at Fawley, Hythe and Marchwood by Coughlan (1979), Hooke and Riley (1987) Hydraulics Research (1987), ABP Research and Consultancy Ltd. (1995b; 2000c) and Williams (2006). The last two sources give a comprehensive documentation, with total land claim amounting to over 500 hectares (160ha, 1927-34; 187ha, 1935-1978). 125ha were reclaimed at Fawley, 1928-1963, most of it in 1950-51 to extend the refinery site using dredge spoil from the Calshot approach channel. Around 200ha were involved in the 1970s to 1980s reclamation of Dibden Bay. Land claim immobilises large volumes of estuary sediment and must therefore be regarded as an output. It also may have other effects, which have been studied using numerical models, calibrated by field measurements, for the evaluation of proposed reclamation schemes (ABP Research and Consultancy Ltd., 2000b; 2000c). Two main effects are recognised: firstly, land claims protruding into the estuary tend to constrict tidal flow and locally increase current velocities and sediment transport potential (Gifford and Partners, 1989). Secondly, reclamation reduces the estuary tidal prism so that current velocities are reduced in the main channels, lowering the potential for sediment input and output, especially bedload output.
It should be noted that the cumulative effects of the historical reclamations have not been evaluated specifically in terms of the intertidal sediments that have been impounded, or the reductions in tidal prism that have resulted.
4.3 Dredging for Navigation
Both capital and maintenance dredging of the main channel, dock approaches and berths have been undertaken in Southampton Water for over two centuries. Four main sediment types are removed from the estuary by this process, and taken to licensed offshore dumping sites:
- Thin veneers of recently deposited clays and silts including re-settled sediments distributed by previous dredging operations, and occasional slumped and degraded sediments derived from oversteepened channel slopes (ABP Research and Consultancy Ltd., 2000a).
- Mid to Late Holocene clays and silts deposited over the past 5000 - 6000 years.
- Pleistocene gravels deposited as river terraces during the Devensian establishment of the 'Solent River' system.
- Tertiary (Eocene) sands, silts and clays forming the underlying substrate.
Major capital dredging operations have all involved deepening and widening of the main access channel between Fawley and Southampton Docks to provide a minimum channel depth of 7.4m below present chart datum in 1889, -9.3m in 1907; -10.2m in 1931 and -12.6m in 1996-97 to the Container Port (ABP Research and Consultancy Ltd., 1993, 1995c and d; Elderfield, 1999; Greenwell, 2000). New approach channels and berths were created at the Western Docks in 1931-36; Fawley, 1960-62 and the Container Terminal between 1967 and 1977.
A total of over 11 million m³ of sediment was removed from Southampton Water during these operations, including 3 million m³ in 1951 and 6.6 million m³ in 1996-7 (Webber 1980). Between 2002 and 2007 dredging of the main approach channel removed 1,245,000 tonnes of mixed calibre sediment. Detailed descriptions are provided by Anon (1951), Anon (1963), Barton (1979), Webber (1980), Long and Scaife (2000), as well as the sources quoted above. Routine maintenance dredging of berths and channels is also conducted and mean estimates of material removed are in the order of 90,000m³ per year (Robinson 1963) to 100,000m³ per year (Webber 1980). 978,000 tonnes were removed from the Itchen berths and channels, and 470,000 tonnes from the Fawley oil refinery terminal, between 2002 and 2008. All of these volumes are relatively small compared to dredging undertaken in many other British estuaries, so it may be concluded that natural siltation rates are low, particularly in the main channel. This is in spite of the fact that channel dredging can truncate, and therefore steepen, adjacent sub-tidal profile gradients leading to downslope movements of materials into dredged channels. These modest siltation rates can be attributed to natural sediment scour and transport by the ebb currents (Robinson, 1963; Greenwell, 2000), although it may also relate to limited sediment availability as indicated by low suspended sediment concentrations recorded throughout the estuary. Siltation is also relatively slow in sheltered dock berths after dredging (Barton, 1979), indicating a low rate of input from suspended sediments, once immediate site disturbance effects have ceased.
A review of the impact of dredging in the Hamble River estuary, 1990-97 (Universities of Newcastle and Portsmouth, 2000) calculated removal of a minimum of 2,900m³ over this period. Most of this was undertaken as maintenance dredging of the various marinas and around jetties, berths and pontoons. A further 2,200m³ appears to have been removed as part of the regular programme of maintenance of the main navigation channel, 1990-2000.
The impact of marina construction, over the period 1975 to the early 1990s, in the Hamble estuary has been examined for its effects on tidal flow velocities (ABP Research and Consultancy Ltd., 1994b). Overall, increases of velocity are apparent for sites adjacent to the main channel, thereby increasing rates of erosion. However, eddy circulations within marinas have also been initiated, resulting in high rates of accretion over co-adjacent areas. This is confirmed by Cundy and Croudace (1995), who report accretion as high as 0.2m per year immediately following marina construction.
5. Sediment Stores
5.1 Beaches
Sand and gravel beaches are only present between the Hamble and Itchen rivers mouths because much of the low energy northern and western margins of the estuary are either reclaimed and protected by embankments, or are fronted by saltmarsh and mudflats. The north-eastern coast is most exposed to wave action and is fringed by beaches of varying width and stability, composed of materials from the adjacent eroding cliffs (Hydraulics Research, 1987; Posford Duvivier, 1994)
At Weston Point, a low bank of sand and gravel (reported as unstable) has been artificially replenished (Southampton City Council, 1988). Further east, a stable upper gravel beach backs a small embayment at Weston Park (Photo 10). East of Weston Park, both the foreshore and beach narrow; the latter is intermittently backed by sea walls of various construction types and dates - see Photo 5 (New Forest District Council, 1993; Posford Duvivier, 1994). At Victoria Park, the narrow gravel upper beach has been characterised by falling levels causing undermining and the erosion of sea walls (Photo 9). The foreshore widens towards Hamble Point where land was reclaimed by dumping of rubble in the 1920s and 30s (Hydraulics Research, 1987; New Forest District Council, 1993; Posford Duvivier, 1996).
The lower foreshore is generally composed of silt and clay with overlying patches of gravel, cobbles and occasional boulder-sized clasts. These coarser sediments were probably derived from reworking of Pleistocene terrace gravels (Hodson and West, 1972) and became stranded as a 'lag' deposit by coastal recession, rising sea level and selective removal of fine-grained sediment. Their weed covered state suggests that current velocities are insufficient for transport of this coarse sediment.
Information on these beaches is derived from limited site observations and little quantitative data is available, other than map evidence of historical and recent narrowing of the lower intertidal foreshore. Details of short and long-term beach morphodynamics and sedimentology remain uncertain.
5.2 Sand and Gravel Banks
No significant mobile deposits composed of these sediment types have been described for Southampton Water. Gravel and sand are present in some intertidal areas but is immobile and derives almost certainly from reworking of relic Pleistocene terraces. Small concentrations of coarse sand and gravel-sized clasts, with occasional boulders, are present on the eastern foreshore of the lower Test and at Weston Point.
Some shell banks are present near Hythe and Marchwood Power Station, although it has been stated that shells are not frequent within the estuary (Dyer, 1980). Chenier accumulations, consisting of mixed shell, fine gravel and coarse sand, are present at several eroding saltmarsh margins, such as Hythe (da Quaresma, 2004; da Quaresma et al., 2007) (Photo 7). Their origin is uncertain, but their shelly materials may have been winnowed out of eroding mudflats and saltmarshes prior to sorting and deposition by tidal currents and wave action. It is the view of da Quaresma et al. (2007) that shell material initially derives from the main estuary channels, is incorporated into mudflat deposits and subsequently revealed and then concentrated on by wave action. Cheniers on saltmarshes are the product of wave washover across marsh edges during storms, and may be subject to some later retreat when extreme wave heights are attained.
The closest major accumulations of coarse sand and gravel are located out of the estuary at Brambles Bank and Calshot Spit respectively. The latter has been stable in configuration over a long period of time, but there is some debateable historical evidence for periods of both extension and retreat, indeed a breach may have occurred in the early seventeenth century (ABP Research Consultancy Ltd., 2000c). These erosive phases could have released some quantities of coarse material into the estuary, but its subsequent pathway is uncertain although some gravel spreads have been located buried by fine sediments immediately behind the spit. At present, there is a substantial store of gravel extending as a lobe shaped inter and sub tidal projection aligned with the main axis of the spit. Evidence suggests that it is currently immobile. Under prevailing hydrodynamic conditions, the input of gravel at the mouth of Southampton Water is thought to be unlikely.
5.3 Mudflats
The majority of intertidal areas are mudflats, up to 1000m in width and covering some 1.1x107m² (1100ha), composed of stores of organic-rich silts, silty clays and clays. There is a dense, but apparently stable, network of dissecting creeks. The deposits are substantially of marine origin, moved into Southampton Water by suspended transport, although some organic (peat) and fluviatile sediments have been revealed at depth by boreholes and excavations at the Western Docks, Southampton (Everard, 1956; Hodson and West, 1972; Barton, 1979); Dibden Bay (ABP Research and Consultancy Ltd., 2000c) and elsewhere. Similar deposits also flank the margins of the main channels and are described in section 5.5. Overall, mudflats have experienced significant erosion since approximately 1940 as evidenced by retreat of Mean Low Water mark. (Hooke and Riley, 1987; Williams, 2006; Baily and Pearson, 2007). However, the spatial pattern of loss is complex, and in some areas, such as Hythe, net accretion has been recorded in recent decades (ABP Research and Consultancy Ltd., 2000c; Williams, 2006). This, combined with general erosion of upper saltmarsh, has had the effect of flattening of inter-tidal profiles (Gray et al., 1993; Johnson, 1998). See E2 for further details. It has been suggested (ABP Research and Consultancy Ltd., 2000a; Price and Townend, 2000) that intertidal mudflats are currently moving towards a more stable morphodynamic condition following the "perturbation" effect of Spartina growth and then decline on adjacent saltmarshes. This is confirmed by Chellow et al., (2010), who applied a numerical model, calibrated by ten years of preceding annual surveys, to several adjacent profiles orthogonal to the Netley shoreline. An exception was observed in the case of a profile backed by a seawall, where erosion loss from beach drawdown and flat abrasion was ascribed to wave reflection. Friend et al., (2005) demonstrated, from an experimental plot at Hythe, that there is a complex relationship between the physical and biogeochemical properties of affecting the stability of cohesive mudflat sediments, with a distinct diurnal rhythm. During the night there is a reduction (approximately 30%) in surface stability, at least partly due to the “armouring” presence of a diatom film during daytime hours, which is absent at night when diatoms migrate. This effect might be accentuated by more energetic waves, so could result in significant mudflat bevelling during these conditions. Much of the fine-grained sediment eroded from marshes and inter-tidal mudflats may be gained by sub-tidal areas (redistribution) or flushed into the Solent and lost to the system.
5.4 Spartina Saltmarsh
Significant areas of Spartina (Cordgrass) marsh are present along the western shore (mostly S. anglica, but notably including small areas of S. maritima), but their extent is significantly less than formerly due to land claim and erosion resulting from "die-back" of S. anglica. For Southampton Water as a whole, 449ha of saltmarsh in 1946 had reduced to 191ha in 1996, representing a cumulative loss of 59% (ABP Research and Consultancy Ltd., 2000c). Bray and Cottle (2003) quote a decline from 258ha in 1971 to 165 ha in 2001 (excluding Test marshes and the Hamble estuary) based on previous photogrammetric survey (see section E2 for additional estimates of loss calculated by more recent research).
S. anglica is a fertile hybrid of the native S. maritima and the accidentally introduced American Smooth Cordgrass, S. alterniflora. The latter was first observed in the Itchen estuary in 1816, but a sterile hybrid, S. x. townsendii did not appear - at Hythe - until about 1870 (S. x. townsendii can only spread vegetatively). S. anglica colonised large areas between approximately 1890 and 1920 due to its reproductive vigour, spreading quickly into other estuaries of the Solent and beyond. Its decline appears to date from the 1920s, due to a combination of edge erosion and pan "dieback" (Williams, 2006; Baily and Pearson, 2007; Cope et al., 2008). There are probably a number of causes working together, but anaerobic soil conditions due to waterlogging and genetic change are arguably the more convincing. Decline and loss continues to the present, although S. anglica remains healthy at some locations, e.g. between Calshot and Fawley. Native saltmarsh mortality may have been in slow progress at least a century before the appearance of S. anglica, e.g. in the lower Test estuary. In this case, some recovery is apparent, with mixed saltmarsh now invading reed swamp.
The colonising swards of cordgrass interrupted water flow, (reduced turbulence at or near bed level) and increased sedimentation, so that significant quantities of suspended sediments were trapped and stored, thus raising marsh elevation (Goodman et al., 1959; Raybould et al., 2000; Neumeier and Amos, 2006). Initial reports of degeneration and dieback of S. anglica date from the mid-1920s (CEGB, 1988; Gray et al., 1990; Raybould et al., 2000; Cope et al., 2008). As root systems of dead plants decomposed, previously bound sediments around pan edges and marsh margins became mobilised and they also became susceptible to lateral erosion by tidal scour and by wave attack. The marsh edge and channel (creek) margins receded, with further release of sediments to lower foreshore and sub-tidal mudflats. Two key processes are thus recognised: (i) frontal erosion of marsh margins (with abrasion and vegetation degradation by chenier washover in areas of shell accumulation), and (ii) fragmentation of saltmarsh blocks. Recolonisation of mudflats has been negligible and by the mid/late 1950s significant recession and cliffing of the marshes was evident (Goodman et al., 1959; Gray and Benham, 1990; Baily and Pearson, 2007; Cope et al., 2008) at rates of up to 2.5m per year. The "dieback" process has also occurred in the Hamble River, but has been relatively less rapid and extensive, at least on the east bank (Goodman et al., 1959; Gray et al., 1993; Johnson, 1998; Bray, 2010). Rates of erosion of the front edge of the saltmarshes declined in some areas in the 1980s and 1990s, for example to an average of 0.9m per year in the Hythe to Calshot area (ABP Research and Consultancy Ltd., 2000c). Refer to Section E2 for further details of the rates and extent of saltmarsh contraction determined by the Solent Dynamic Coast Project (Cope et al., 2008) and Baily and Pearson, 2007.
Although reasons for loss are uncertain, (refer to section E2) "dieback" areas have generally been restricted to marshes where the substratum is particularly soft, fine-grained and waterlogged. It has been suggested that sediment input may be insufficient to maintain marginal sedimentation at a rate comparable to relative sea level rise at some sites, thus leading to waterlogging of Spartina roots. (Geodata Unit, 1987; Gray et al., 1993; Tsuzaki, 2004, 2010). A range of other causes has also been examined, (Gray & Benham, 1990; Raybould et al., 2000; Gray et al., 1991), whose precise applicability to Southampton Water is uncertain. Limited areas have undergone dieback due to contamination by refinery effluents 1km upstream and downstream of the outfall at Fawley. Improvements to effluent quality beginning in 1971 had resulted in some regrowth of degenerate marsh by 1981, although no natural recolonisation of bare mudflats has yet been recorded (Dicks and Levett, 1989; Oil Pollution Research Unit, 1994; May, 2000; Dicks and Iball, 2005) except for small areas of experimental planting. This may be due to the historical burden of hydrocarbon contaminants trapped in the mudflat sediments, so recovery by Spartina may not be possible until an overlying layer of uncontaminated sediments has been deposited. Vertical accretion of 3.5-8mm per year (Hythe) and 2-8mm per year (Hamble estuary) for recent decades, based on 137Cs, 60Co, 210Pb and other dating methods (Lewes, 1997; Cundy et al., 1997) indicates that upper saltmarshes may trap some of the fine sediment removed from mudflats and lower elevation marshes subject to longer periods of immersion. These rates are considered to be in response to contemporary sea-level rise (Cundy and Croudace, 1996). Spartina marshes protect sea walls, embankments and low natural cliffs from wave attack, so their recession could ultimately affect the backshore boundary. Depletion of Spartina marshes also releases significant quantities of fine sediment which may be deposited in navigation channels, and is periodically removed from Southampton Water by maintenance dredging. There is likely to be a complex link between saltmarsh and mudflat erosion, with the latter potentially being partly the result of a sediment deficit created by the dredging of the main estuary channel. Wave erosion is only an effective erosional process when high water occasionally coincides with easterly and southeasterly winds. Saltmarsh retreat is not universal within Southampton Water, but exceptions are due to special local factors. For example, mean high water at the south west end of Hamble Common has extended seawards some 150m since the early 1930s, following extinction of foreshore saltmarsh between 1910 and 1925. This is apparently due to waste dumping, and has attracted some new colonisation (Halcrow, 1998; Bray, 2010).
Following laboratory experiments, Tsuzaki (2004; 2010) found anaerobic soil conditions with impeded drainage to be the most likely cause of the dwarf growth forms and lack of re-colonisation of pans and mudflats by Spartina anglica on the south coast of England. The thesis concludes that the ultimate demise of the Spartina anglica marshes of the south coast of England is the result of frontal and creek erosion of the mature marsh and the failure of Spartina anglica to establish itself on the newly exposed sediments of the foreshore.
5.5 Channel Sediments
The main channels of Southampton Water have been dredged to their present depths, so sediments of the bed and margins are those of the underlying geological succession with additional cover by a veneer of recently deposited fine sediments (Barton, 1979; ABP Research and Consultancy Ltd., 1993; 2000b). In upper parts of the estuary where the buried Pleistocene channel lies closer to the surface e.g. Western Docks, navigation channels have been cut through its associated gravels -up to 5 m thick- and into the sandy clays of the Bracklesham Formation (Barton, 1979). Between Fawley and Calshot the buried channel is deeper and the navigation channel is generally floored by Pleistocene gravel deposits covered by a relatively thin sequence of Holocene silts and clays (Hodson and West, 1972). North of Fawley, the latter are rarely in excess of 5m in thickness (Barton, 1979), but towards Calshot, they increase in thickness to 13m and form the major slopes flanking the navigation channel (Long and Scaife, 2000). Pleistocene deposits chiefly comprise sub-angular flints and brown sandy clays (Barton 1979) that probably form relatively stable slopes. This is apparent from the evidence of flume experiments in which the majority of gravels could not be moved by simulation of the current velocities measured in the channel (Wright and Leonard, 1959). Sidescan sonar and echo-sounding revealed linear erosional furrows (with widths of up to 5m and depths of 1m continuous for nearly 4km (Flood, 1981) from which sediments were sampled by divers. An average size distribution yielded 1% sand, 54% silt and 45% clay. Slightly coarser sediments, including shell layers, were found within the furrows.
A similar range of deposits are present in the Itchen estuary, where silts and clays are the main channel deposits together with some superficial patches of fine gravel, sand and cobbles upstream to Woodbridge (Qadri, 1984). The provenance of these coarser sediments is not known.
Estuarine deposits reach a maximum thickness of 13m at Fawley and decrease upstream to 6m at the Container Docks (Hodson and West, 1972; Barton, 1979). This information is reliable as it was based on numerous boreholes and foundation excavations.
5.6 Siltation Rates
Although the overall sediment transport system in Southampton Water has been relatively poorly studied, research has been conducted into siltation rates due to its impact on navigation access to the Port of Southampton and at Fawley.
This work confirms that net accretion rates in dredged channels in Southampton Water are low because relatively little maintenance dredging is required compared to most UK estuary-head ports (Robinson, 1963; Barton, 1979; CEGB, 1988; ABP Research and Consultancy Ltd., 1995f and 2000c). Comparison of Southampton Port Authority hydrographic surveys indicated no measurable changes in bed levels over most of Southampton Water between 1970-1985 (CEGB, 1988 and 1990-95 (ABP Research and Consultancy Ltd., 1995f). However, a sedimentation rate of 0.3m per year was measured at a berth at the Container Docks, and similar rates were recorded in the Bury swinging ground (Photo 8) soon after (Barton, 1979) and at Hythe and Ocean Village marinas within 6-9 months of their initial dredging. Some of this may be attributed to resettlement of sediment disturbed by dredging operations. Some sedimentation also possibly occurs by slumping of the banks of navigation channels. In both cases, accretion is attributable to disturbance, thus natural "background" siltation rates are low (Barton, 1979; ABP Research and Consultancy Ltd., 2000b; 2000c).
A detailed study of echo-sounding traces recorded over the period 1939-1968 revealed a significant area of sedimentation on the west side of the main channel between Fawley and Hythe (Flood, 1981). The major accumulation area was subject to sedimentation at approximately 0.11m per year with a maximum of 0.20m per year in a zone 2-3km long and 100m wide, previously dredged in 1926. A second, delta-shaped accumulation area was identified opposite Eastland Creek and was subject to sedimentation at 0.12m per year (Flood, 1981). This has been confirmed by more recent surveys (ABP Research and Consultancy Ltd., 2000a), albeit at lower conjectured rates. As part of this area of sedimentation is opposite the cooling water outfalls from Fawley, it was suggested that hydrocarbons from the effluent became adsorbed on suspended sediment particles causing local intensification of flocculation and deposition (Knap, 1978). Overall, contemporary siltation is more rapid to the west of the main channel, (e.g. Marchwood Military Port) whilst that to the east is extremely slow and was measured at 0.02 - 0.04m per year (Flood, 1981). This may be due to the fact that channel deepening causes artificial steepening of sub-tidal profiles, thus siltation represents an attempt to restore equilibrium.
ABP Research and Consultancy Ltd., (1995e; 2000b) have undertaken numerical modelling studies to evaluate the impact on suspended sediment transport in the upper estuary of a proposal to construct a container terminal at Dibden Bay. Under present circumstances accretion during the period of slack water and erosion on the succeeding ebb phase are approximately balanced, with a tendency towards net erosion on spring tides. Further towards the estuary head, net deposition of suspended load is apparent, estimated at 137 to 161,000m³ per year (ABP Research and Consultancy Ltd., 2000c). Net accretion is probably less between Weston and Netley because of relatively greater exposure to sediment entrainment and removal by wave or combined tidal/wave action.Further information may be found at http://www.scopac.org.uk/sediment-sinks.html.
6. Summary and Sediment Budget
Within this section key sediment transfer processes are summarised.
Sediment Inputs
Marine sources are likely to be the most significant, however, analysis of tidal conditions indicates that net input comprises suspended rather than bedload sediments. Littoral drift is a negligible input because most material transported towards Southampton Water is retained on Calshot and Hook spits or entrained and flushed back into the Solent. Fluvial input is considered small because river discharge is limited, low in suspended sediment concentrations by comparison to tidal exchange and bedload sediments are likely to be stored within floodplains. Erosion of cliffs between the Itchen and Hamble river mouths supply gravel to local beaches, but the quantity is small due to slow retreat and low cliff height. A significant trend for erosion of Spartina marsh margins and narrowing of the intertidal zone has been identified. Whilst this process undoubtedly releases significant quantities of predominantly fine sediments, it does not comprise much input of new sediments but, rather, re-distribution of the existing estuarine sediment store. Although a considerable thickness of sediments is stored within the estuary, only those sediments at the surface and particularly at the estuary margins are susceptible to erosion transport. Sediments immobile under the prevailing hydraulic region can be released into the dynamic estuarine system only by dredging of the navigation channels. Dredged material may re-settle locally or be transported and contribute to siltation further afield.
Sediment Outputs
Natural output is possible by tidal flow, whilst human activities result in output by reclamation (impoundment of mudflat and saltmarsh sediments) and dredging. Ebb current tidal streams are significantly more rapid than corresponding flood currents, so net output occurs by near bedload transport. This is mostly coarse silt and sand because tidal current velocities are insufficient to transport coarser sediments. This sediment transport pathway is confirmed by the presence of linear furrows and sandwaves indicating bedload transport out of the estuary although its magnitude is difficult to discern. Land claim and dredging involves direct sediment loss from the estuary and also affects the sediment budget indirectly by interfering with the hydraulic regime of the estuary. Both practices have resulted in localised variation in tidal velocities, but with net reduction of the tidal prism and enlargement of the submerged profile of the channel. These changes suggest possible reduction of output by bedload transport with potential reduction in marine inputs due to the reduced tidal exchange. Historically, sediment inputs have exceeded outputs with consequent sedimentation over the past 6500 years. It is possible that this balance has been upset by recent reclamation and dredging so that these practices may be contributory factors in the widely reported erosion of saltmarsh and intertidal areas. Dredged channels reach to within 200m of the edge of the foreshore zone in places. Mudflat erosion may therefore be partly ascribed to sediment "demand" induced by channel deepening over more than 150 years. This, in turn, can be transferred to saltmarshes in an attempt to restore sediment budget equilibriu.
Sediment Stores
Beaches are extremely small and limited to the shore between the Hamble and Itchen rivers. Sedimentological and volumetric information is not available so beach behaviour is uncertain. The main sediments within Southampton Water are estuarine silts and clays, which form the mudflats of the intertidal zones and the flanks of the channels. The main navigation channel has been significantly deepened by dredging and parts are now floored by Pleistocene gravels. These gravels are more resistant to scour and can maintain higher slope angles than the estuarine silts and clays. Siltation following episodes of dredging results in deposition of thin layers of silts and clays over the gravels. Mobile sand and gravel banks have not been recorded in Southampton Water, although shell banks were recorded at Marchwood. It must be concluded that sedimentation over the past 6500 years has mostly comprised silts and clays in Southampton Water. Coarser materials, by contrast, are relatively scarce amongst the surface sediments. Formerly extensive areas of Spartina marsh are now much reduced by land claim and erosion involving net estuarine sediment loss and release respectively. Protection afforded by Spartina marshes to defences of low-lying land is therefore declining, and is unlikely to improve because natural recolonisation has mostly ceased.
Sediment Transport
Sediment transport is predominantly by tidal currents because wave generation is limited by the narrow and sheltered nature of the estuary. Limited littoral drift is reported on beaches between the rivers Hamble and Itchen, with a drift divergence between Victoria Park and Hamble Common. Bedload and suspended load transport along the axis of the main channel are in opposite directions because of more rapid but shorter duration ebb currents. Suspended sediments such as fine silts and clays undergo net transport up the estuary and into various creeks, channels and saltmarshes, although concentrations within the water column are considerably lower than recorded at many comparable British estuaries. Bedload sediments (mostly coarse silts, sand and fine gravels) are transported down the estuary and result in the formation of bedforms such as gravel sand waves and linear furrows. Channels are relatively stable with low natural siltation and stable bedforms. Thus, it is generally the view that sediment transport rates are low within the estuary. This conclusion appears accurate for the present, where channels are well aligned to peak tidal flows, but numerical model studies indicate that new channels cut across the natural flow direction could experience siltation at up to 84,000m³ per year. Sediment transport and subsequent siltation can also increase in the short-term due to disturbance caused by dredging of navigation channels. Widening of channels close to intertidal areas can also lead to drawdown, narrowing of intertidal areas and steepening of foreshore profiles. Erosion of fine-grained sediments (mudflats and saltmarshes) may be effective when maximum fetch coincides with the long high water stand. The impact of ship-induced bow waves will also be greater during high water, but has not been quantified fully. Sedimentation in Southampton Water is generally restricted to the western side of the main channel where rates of 0.12m per year are characteristic for limited areas. Overall vertical sedimentation rates over the past 6500 years are estimated at 2mm -10mm per year and 1mm - 2mm per year. It is suggested that sedimentation has kept pace with sea-level rise over recent millennia. More rapid 20th Century rates of 4-8mm per year have been recorded from the Hythe and Hamble saltmarshes.
Sediment Budget
Despite several areas of uncertainty, together with lack of representative and reliable quantitative data for most elements, ABP Research and Consultancy Ltd., (2000c) attempted a provisional 20th century estuary-wide budget. Its elements were critically evaluated by Bray (2000) who has proposed the following based upon the ABP work:
The cliff inputs are very high compared to estimates of 500m³ per year presented in section 2.3 and river input refers to potential rather than actual transport.
Some elements are not quantified e.g. sediment stored within the water column, impoundment losses resulting from reclamation. Furthermore, it could be argued that capital dredging alters the sediment storage capacity of the estuary. Note especially that the marine input is inferred as the amount required for balancing the budget and assumes that the budget actually is balanced. It has yet to be measured or checked independently.
The change in storage elements are summed to give a net value that refers primarily to sediments yielded by saltmarsh and intertidal erosion that would appear not to become re-deposited in the estuary (subtidal gains are small by comparison). It suggests that significant losses have occurred. Posford Duvivier (1999) calculated a rather lower erosion rate of between 8,500 and 25,000m³ per year from the shoreface between Calshot and Hythe, but this only covers some 30% of the total estuary shoreline.
Despite the considerable uncertainty, and the wide range of estimated values, it is apparent that the total estuary budget is in deficit given sustained dredging and an apparent failure to retain all of the products of saltmarsh and mudflat erosion. It is partly in this context that the impoundment of intertidal sediments resulting from land claim proposals (e.g. Dibden Bay) must be viewed. Despite these variations, the cumulative effect of land claim within the estuary as a whole has been to reduce the tidal prism. This has had the direct effect of reducing output via bedload transport and potentially reducing input of suspended sediments.
More detailed budget assessments have been prepared for distinct portions of the estuary. For example, the area upstream of the Dockhead is an area of net deposition, with an estimated 40% coming from erosion of inter-tidal areas and 60% from marine sources; most but perhaps not all of this gain is removed via berthside and channel dredging (ABP Research and Consultancy Ltd., 2000a). The mouth of the Itchen appears to have a steady balance between losses and gains, whilst an erosion loss of 780m³ per year is calculated for the lower Hamble estuary (ABP Research and Consultancy Ltd., 2000a).
A clear implication is that marine input is essential to maintain the existing budget although its actual magnitude is not known. Dredging represents a major loss and some of the sediments eroded from saltmarshes and mudflats would also appear to be lost, although it is suggested that further studies and refinements of estimates are required. These initial findings would appear to be contrary to the long established trend for net sedimentation that has prevailed over the past 6000 years.
Examination of the Southampton Water estuary regime reveals that its cross section area at the mouth (and also at intervals further upstream) is larger than the equilibrium value that might be expected for its tidal prism (ABP Research and Consultancy Ltd., 2000c; Bray and Cottle, 2003). The non-equilibrium regime appears to have been inherited from the Test/Itchen valley that was inundated by rising sea-levels some 6,000 years ago. Sediment transport and deposition should normally act to reduce the inlet cross section towards its equilibrium value. However, infilling appears to have operated slowly within most Solent estuaries due to a limited sediment supply and shelter from wave action at their entrances.
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. Although at present a 12 year time series of data has been collected, 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, although the lack of significant wave energy and the poor development of beaches means that shorelines in this unit are not suited for definitive studies of drift.
8. Knowledge Limitations and Monitoring Requirements
Notwithstanding results from the Southeast Regional Coastal Monitoring Programme, and the summarised information and assessments from the North Solent SMP2 (New Forest District Council, 2010), recommendations for future research and monitoring that might be required to inform management include:
- Previous investigations have not regarded sediment distribution and availability as a critical issue, so it is difficult to assess the overall sediment budget. Processes transporting sediment have not been previously perceived as contributing inputs and/or outputs and consequently have received only cursory examination. Comprehensive and quantitative budgetary analysis is not therefore possible at present, and much of the existing knowledge base requires verification by further research to enhance understanding of complex inter-relations between hydrodynamics and sedimentation processes.
- Outputs and inputs at the estuary mouth are supported by strong direct (e.g. tidal current tracking and analysis) or indirect (e.g. bedforms) information, but little integrative quantitative information is available. Details are lacking of both suspended sediment input and bedload output. By far the most important transport mechanism is tidal flow, a process, which can be modelled using appropriate sediment, transport equations and verified by field measurements. A useful first step would be to collect measurements of suspended load concentrations within the water column over a representative range of tidal conditions. This should assist analysis of suspended sediment transport using tidal models and possibly lead to more definitive estimates of the magnitude of marine input.
- Bedform surveys should be coupled with systematic sediment sampling, so that distribution maps can be compiled. Examination of particle size distribution and sorting patterns can then be used to infer transport directions and the grades of sediment being transported as was undertaken for Chichester Harbour entrance by Geosea Consulting Ltd (1999).
- The major natural process causing concern is narrowing of the intertidal zone and associated saltmarsh erosion. This should continue to be monitored, with saltmarsh perimeters and mudflat levels carefully measured, and volumes of eroded/deposited sediment determined by map comparison of different parts of the estuary. Field monitoring of upper saltmarsh surfaces is needed, to determine the rates of vertical accretion throughout the estuary. This would provide improved estimates of sediment storage that could contribute to sediment budget calculations.
- Siltation rates have been measured in the main channels and newly dredged berths immediately following operations, but these are unlikely to be representative of normal conditions in the estuary as a whole because disturbance by dredging causes enhanced localised siltation. Hydrographic chart comparisons for a limited area between Fawley and Hythe, for example, revealed spatially and temporally variable siltation patterns. Analyses of this type could be extended to cover all appropriate parts of the estuary and thus obtain an overall view. Such information would provide a valuable input into the compilation of a sediment budget of the estuary, and perhaps indicate more clearly the fate of sediments released by mudflat and saltmarsh erosion. Constructing a more soundly-based sediment budget for Southampton Water is a major medium to longer term research priority.
- Studies are needed into the cumulative effects of historical reclamation and dredging upon the estuary. Improved knowledge of past estuary responses to these factors should provide insights into future sensitivity. It is recommended that studies should focus upon reconstructing historical sediment budgets and tidal prism analyses in order to determine the effects of these processes.
