Case studies of a submerged landscape: the Bristol Channel, UK
The recent significant increase in building work in the coastal zone (eg. new nuclear plants and port extension/construction) and offshore (eg. offshore windfarms and gravel extraction) provides archaeologists with an opportunity to gather data from a range of sources, and across larger areas than has often previously been possible. As such we can now create fully-integrated, well dated (through radiocarbon dating of material gained from cores), palaeogeographic reconstructions of significant sections of the continental shelf.
This is because much of the data acquired for the engineering and planning components of such projects is the same as that used to survey for archaeological sites and landscapes; as such we can share data to better understand change through time and the archaeological record. Through a combination of extant regional scale data (tens to hundreds of kilometres) and site-scale studies (tens of metres through to a few kilometres: boreholes and vibrocores), it is possible to adopt an integrated approach to such reconstructions, creating a “total” terrestrial palaeo-landscape view. This case study is an example of one such project, focused on recent infrastructure developments at Hinkley Point, on the southern margin of the Bristol Channel, in the south-west of the UK (Image 1).
This area of the Bristol Channel (Image 1 and 2) has been subject to an integrated terrestrial (lidar, aerial photography, coring and archaeological survey) and marine geophysical (derived UKHO bathymetry, swath bathymetry, side scan, boomer sub-bottom data), geotechnical (boreholes and vibrocores), and archaeological surveys to provide a seamless Holocene palaeo-landscape reconstruction for the whole of the Bristol Channel area. The Mesolithic (10000 – 4000 lBC) archaeological evidence from the marginal (currently terrestrial) areas of the Bristol Channel and Severn Estuary is internationally recognised as of significance, giving a rare glimpse into prehistoric life. It includes human and animal footprints, wood, flint and stone artefacts, along with associated submerged forests and intertidal peat sequences (see M.Bell, 2007, Prehistoric Coastal Communities: The Mesolithic in western Britain or the website of the Severn Estuary Levels Research Committee for details). However, our understanding of the offshore sections of this region has been limited.
Image 1 shows the major east-west flow of the meandering “palaeo-Severn” channel system (the old course of the river now submerged beneath the sea), which can be traced from Sharpness several kilometres up the Severn Estuary to the mouth of the Bristol Channel. This river system has a number of features including sharp and erosive channel edges, which suggest at least two phases of river adjustment to sea-level change and prominent streamlined bedrock islands, typical of very high river flow rates. These probably formed in response to at least two periods of lower sea level, which are typically associated with periods of major ice sheet growth. The presence of the margins of the last Glacial Maximum ice sheet close to the south Wales coast at these times, could have been responsible for the high river discharges, and sufficient erosive power to cut these bedrock channels.
Image 2: UKHO bathymetric data showing a) colour coded bathymetry; b) Hillshade representations; c) Slope layer draped over the hillshade; d) Aspect layer. Copyright: British Crown Copyright and SeaZone Solutions Ltd. All rights reserved. Product Licence 032007.016.
Image 2 shows the ability to enhance data interpretation of the bathymetry by using a combination of hillshade (equivalent to casting a light over a land surface and mapping the shadows and strong reflections to give a pseudo-3D relief), slope (the maximum gradient between each depth point) and aspect (the direction of that maximum gradient). All of these techniques are standard in both free and purchased GIS systems and are a common way to manipulate such data.
There is a distinct difference in the catchment characteristics to the north and south of the main “palaeo-Severn”. To the north, discrete north-south flowing, relatively deep (> 8 m), dendritic tributary systems are found flowing across a submerged platform (thus cutting across the geological stratigraphy) toward the main channel. By comparison, there are limited expressions of discrete palaeochannels in the south and, where they do exist, such as offshore Hinkley, they are very shallow and appear to conform to irregularities in the pre-existing bedrock surface.
The presence of deep valleys in the adjacent terrestrial Somerset Levels suggests that volumes of water need to discharge across this landscape, but this could be achieved through broad, poorly-constrained flooding surfaces rather than discrete channels. Such contrasts in river form could have significant effects on the human interaction with these landscapes during the Upper Palaeolithic and Early Mesolithic, as well as the nature of the preserved archaeological record. It is interesting to consider that the Upper Palaeolithic occupants (occupation dated to c. 14750 calBP) of the cave systems of the Cheddar Gorge, 35 km to the east would have had an almost uninterrupted view of this vast cascading river system.
The nature of this landscape would also control, significantly, the style of marine inundation (flooding as sea level rose). Naturally, sea level rise would start with the deep, main palaeochannel, becoming a narrow tidal river, followed by a rapid inundation of the main low lying bedrock platform in between. Data taken from buried sections of this bedrock platform have identified extensive freshwater peats overlain by estuarine sediments, which provide a date for this marine inundation of c. 7900 calBC. Further, although not calibrated by core data, a large seabed exposure of potential peat deposits, can be seen in the swath bathymetry. Such exposures, even on the seabed, could be sources of large volumes of material for further palaeoenvironmental investigations at this critical part of the sea-level record and further identification of human activity.
This work has been supported by EDF Energy and Historic England.
© University of Southampton, 2017