The Thames Estuary lies towards the southern edge of the London Basin bounded by upland areas to the south (North Downs) and north (Chiltern Hills) composed of Cretaceous Chalk. It was not until the Late Cretaceous (around 65 million years ago) that a major rise in sea level across Europe led to a significant deepening of the sea and without the influence of sediments brought in from nearby landmasses a very pure marine limestone was deposited in the warm sea; this is the Chalk, which may constitute up to 98% calcium carbonate. The Chalk forms the sub-crop of sections of the middle part of the Inner Thames (Erith downstream to Tilbury and parts of Woolwich and Gallions Reaches) (Sumbler, 1996). Elsewhere the Chalk is covered by Tertiary muds and sands (Balson and Dâ€™Olier, 1989; British Geological Survey, 1997).
A fall in sea level allowed the emergence of large areas of land and a considerable thickness of Chalk was eroded away. However, around 60 million years ago, sea level rose again and a shallow sea invaded the area depositing a series of Tertiary muds and sands reflecting changes in sea level and the transgression and regression of this sea (Sumbler, 1996). The oldest Tertiary sediments beneath the Thames Estuary belong to the Thanet Sand Formation and Lambeth Group. The bulk of the Thanet Sand consists of shallow marine silty sand with the main outcrops in south-east London (e.g. Howland, 1991) and north Kent. The Lambeth Group comprises sands (the Upnor Formation) deposited in a shallow sea and the overlying Woolwich Formation, comprising a varied assortment of sediments including clays and sands deposited in brackish, estuarine or coastal lagoon environments.
Following a rise in sea level, shallow marine conditions were again established in the Thames area, and the Harwich Formation was deposited, made up of several distinct units of mud and sand. Sea level continued to rise during the Eocene (55-35 million years ago) leading to the deposition of the thick bluish-brown London Clay which is the most widespread and best known of the Tertiary deposits of the London Basin and underlies much of Greater London and the Thames.
The Tertiary units are overlain by a complex suite of sediments deposited during the glacial and interglacial phases of the Quaternary, including those of the Holocene (last 10,000 years). Between the Anglian glaciation and the Devensian glaciation (the last Ice Age) the River Thames and its tributaries became established in their modern valleys and formed wide expanses of river terrace sands and gravels (Bridgland, 1994). These are mainly remnants of floodplains, representing phases in the gradual downcutting of the river during the Pleistocene; the highest terrace being the oldest and the lowest the youngest. This gently terraced landform is now almost completely obscured by urban development. The last major phase of terrace formation was during the Devensian glaciation when the River Thames was graded to a level at least 25 m below present sea level. The Late Devensian River Thames appears to have followed a braided course, crossing a wide floodplain until the early Holocene when it gradually developed into a single channel river (Wilkinson and Sidell, 2000). The deposits are now covered by estuarine alluvium, deposited as sea level rose during the Holocene interglacial (10,000 years ago to present).
Following the melting of the ice sheet at the end of the Devensian glaciation there has been a significant rise in sea level. The Thames Estuary was flooded around 8000 years ago and complex sequences of marine/brackish sediments intercalated with freshwater peats were deposited on the youngest terrace sands and gravels (Devoy, 1977, 1979, 2000; Marsland, 1986). The Holocene sediments cover the floodplain approximating to the area that has been flooded by high water spring tides, including that presently protected by flood defences; they occur on both sides of the estuary and occupy an overall width of 3-10 km (Royal Haskoning, 2004).
The width of the Thames Estuary floodplain deposits is partially controlled by the position and strength of the Cretaceous and Tertiary sub-crops. The most significant change occurs at Tilbury where relatively soft Tertiary deposits downstream are replaced by relatively hard Chalk upstream, resulting in greater confinement of the river upstream. As a consequence the width of the floodplain deposits narrows rapidly from 10 km in the Coryton area to 3 km at Tilbury-Gravesend. The thickness of the deposits increases downstream, reaching a maximum of about 35 m at the eastern end of Canvey Island (Marsland, 1986).
Devoy (1977, 1979) proposed two Holocene relative sea-level curves from the estuary, one for Tilbury and one from sites to the west of Tilbury. Although the curves from both areas followed the same trend, the Tilbury curve plotted c. 1.5 to 3 m below the west of Tilbury curve. Various reasons have been put forward for this anomaly, including the possibility of differential subsidence on an east-west axis (Devoy, 1979). However, a re-interpretation of the data (Haggart, 1995; Long, 1995), removed the need for eastward trending subsidence.
The most recent model proposed for the Holocene evolution of the Thames Estuary (Long, 2000; Long et al., 2000) describes sedimentation within a three-stage sequence based on estuarine development:
- Stage 1 - The early Holocene rapid rise in relative sea level and flooding of the estuary between 8000 and c. 6000 years ago (Wilkinson and Sidell, 2000) leading to the widespread deposition of the silt and clay.
- Stage 2 – A major expansion of peat-forming communities between c. 6000 and 3500 years ago: Beginning in the lower estuary, the initial formation replaced estuarine mudflat and saltmarsh sedimentation. Further west in London the rising water table allowed peats to form on top of Devensian terrace sands and gravels. Peat accumulation had a significant impact on the geometry of the estuary, reducing the spatial extent of intertidal environments. At Cross Ness, the intertidal area narrowed by 4 km. It is likely that the reason for initiation of peat formation at this time is a reduction in the rate of relative sea-level rise between c. 6000 and 4000 years ago. In the Thames Estuary the slow down in sea level rise would have encouraged the expansion of saltmarsh and then freshwater communities across areas of former intertidal mudflat.
- Stage 3 - Between 4000 and 3000 years ago the peats of the lower estuary were inundated with later inundation of middle and upper estuary areas (2500 years ago at Silvertown, Wilkinson et al., 2000). By c. 2500 to 2000 years ago almost all of the once extensive peat forming communities throughout the estuary downstream of Woolwich had been replaced by intertidal conditions. Hence, the tidal Thames expanded and was once again flanked by extensive mudflats and saltmarshes that continued to develop, with only occasional still-stand phases until c. 150 years ago when much of the previously intertidal area was land-claimed for docks and associated installations.
Historic relative Sea-Level Change
The most recent relative sea-level curve (Wilkinson and Sidell, 2000) shows that there is a general rise of sea level through time, with an initial rapid rise of 3.5 mm per year, slowing down around c. 6000 years ago to 0.7 mm per year. This is supported by a wider analysis of land-level and sea-level change around Britain (Shennan and Horton, 2002), which calculated a late Holocene (last 4000 years) relative sea-level rise of 0.74 mm per year for the Thames, 0.85 mm per year for Essex and 0.67 mm per year for Kent. This can be compared with those for 20th century sea-level changes published by Woodworth et al. (1999) using tide gauges. They showed relative sea-level rises of 1.22+/-0.24 mm per year at Southend-on-Sea, 1.58+/-0.91 mm per year at Tilbury and 2.14+/-0.15 mm per year at Sheerness. Overall, these figures suggest an additional rate of relative sea level change in the 20th century of around 1 mm per year, as compared to the Late Holocene. This is in general agreement with the view that global sea levels have increased by 100-200 mm over the last century.
Shennan and Horton (2002) suggest, however, that some deficiencies may be inherent in the 20th century dataset. These include the unequal distribution of measurements and the considerable amount of interannual (typically decadal) variability present in all tide gauge records. Littlewood and Crossman (2003) also questioned the degree of accuracy of the tide gauge data based on concerns that they may not have remained at the same level relative to Ordnance Datum throughout their period of deployment. They indicated that the gauges were levelled to Ordnance Datum over 40 years ago, and since that time differential ground subsidence may have caused their perceived level to be different to their actual level. Monitoring using GPS at the tide gauge locations at Richmond, Tower Pier, Silvertown, Erith, Tilbury and Southend-on-Sea has shown that between March 1997 and July 1999, the movement of ground levels at these locations was statistically insignificant.