Mapping Doggerland: the Mesolithic Landscapes of the Southern North Sea

 Mapping Doggerland The Mesolithic Landscapes of the Southern North Sea Edited by Vincent Gaffney, Kenneth Thomson and Simon Fitch A project funded by the Aggregates Levy Sustainability Fund and administered by English Heritage Mapping Doggerland  Mapping Doggerland In Memoriam Dr Kenneth Thomson (1966-2007) Pioneer and Explorer of the North Sea Mapping Doggerland  Mapping Doggerland Mapping Doggerland The Mesolithic Landscapes of the Southern North Sea Contents Foreword Huw Edwards (PGS) Preface Ian Oxley (English Heritage) Acknowledgements 1. Mapping Doggerland Vincent Gaffney and Kenneth Thomson 1.1 Introduction .......................................................................................................................................................... 1 1.2 The context of study ............................................................................................................................................. 1 1.3 Previous methodological approaches.................................................................................................................. 4 1.4 Towards an alternative methodology ................................................................................................................ 6 2. Coordinating Marine Survey Data Sources Mark Bunch, Vincent Gaffney and Kenneth Thomson 2.1 Introduction ........................................................................................................................................................ 11 2.2 Identification of sources: who acquires, owns or holds survey data? ........................................................... 11 2.3 Key data repositories ......................................................................................................................................... 11 2.4 Governance and surveying within territorial waters ...................................................................................... 13 2.5 Public metadata resources ................................................................................................................................. 14 2.6 Case study assessments of survey data ............................................................................................................. 16 2.6.1 Area 1: the Bristol Channel .......................................................................................................................... 16 2.6.2 Area 2: Portland ........................................................................................................................................... 21 2.6.3 Area 3: The Spurn ........................................................................................................................................ 21 2.7 Conclusions ......................................................................................................................................................... 21 3. 3D Seismic Reflection Data, Associated Technologies and the Development of the Project Methodology Kenneth Thomson and Vincent Gaffney 3.1 Introduction ........................................................................................................................................................ 23 3.2 Seismic reflection method and resolution......................................................................................................... 23 3.3 2D versus 3D seismic acquisition and interpretation ...................................................................................... 27 3.4 Interpretation strategy for the Southern North Sea........................................................................................ 28 3.5 Conclusions ......................................................................................................................................................... 30 4. Merging Technologies: The integration and visualisation of spatial data sets used in the project Simon Fitch, Vincent Gaffney and Kenneth Thomson 4.1 Introduction ........................................................................................................................................................ 33 4.2 Infrastructure ..................................................................................................................................................... 34 4.3 Software integration........................................................................................................................................... 35 4.4 Primary integration procedures........................................................................................................................ 35 4.5 Integration of volumetric information through solid modelling ................................................................... 36 4.6 Merging technologies ......................................................................................................................................... 40 i  Mapping Doggerland 4.7 Conclusions ......................................................................................................................................................... 41 5. A Geomorphological Investigation of Submerged Depositional Features within the Outer Silver Pit, Southern North Sea Simon Fitch, Vincent Gaffney and Kenneth Thomson 5.1 Introduction........................................................................................................................................................ 43 5.2 Feature descriptions .......................................................................................................................................... 44 5.3 Discussion/Feature classification ...................................................................................................................... 50 5.3.1 Temperate terrestrial geomorphological processes ...................................................................................... 51 5.3.2 Marine geomorphological processes ............................................................................................................ 51 5.3.3 Subaerial coastal landforms ......................................................................................................................... 52 5.3.4 Marine bedforms .......................................................................................................................................... 53 5.4 Environmental Interpretation .......................................................................................................................... 55 5.4.1 Sand Bank classification .............................................................................................................................. 55 5.4.2 Estuaries and Sand Bank Formation ............................................................................................................ 56 5.5 Conclusions ......................................................................................................................................................... 58 6. Salt Tectonics in the Southern North Sea: Controls on Late Pleistocene-Holocene Geomorphology Simon Holford, Kenneth Thomson and Vincent Gaffney 6.1 Introduction........................................................................................................................................................ 61 6.2 Relationships between salt structures and late Pleistocene-Holocene fluvial systems. ................................ 61 6.3 Conclusions ......................................................................................................................................................... 65 7. An Atlas of the Palaeolandscapes of the Southern North Sea Simon Fitch, Vincent Gaffney, Kenneth Thomson with Kate Briggs, Mark Bunch and Simon Holford 7.1 Introduction........................................................................................................................................................ 67 7.2 North Western Quadrant .................................................................................................................................. 72 7.2.1 Description ................................................................................................................................................... 72 7.2.2 Other Features .............................................................................................................................................. 75 7.2.2.1 Solid Geology .......................................................................................................................................... 75 7.2.2.2 Recent Geological Features ..................................................................................................................... 75 7.3 North Eastern Quadrant ................................................................................................................................... 75 7.3.1 Description ................................................................................................................................................... 75 7.3.2 Other Features .............................................................................................................................................. 81 7.3.2.1 Solid Geology .......................................................................................................................................... 81 7.3.2.2 Recent Geological Features ..................................................................................................................... 81 7.4 South East Quadrant ......................................................................................................................................... 81 7.4.1 Description ................................................................................................................................................... 81 7.4.2 Other Features .............................................................................................................................................. 82 7.4.2.1 Recent Geological Features ..................................................................................................................... 82 7.5 South Western Quadrant .................................................................................................................................. 86 7.5.1 Description ................................................................................................................................................... 86 7.5.2 Other Features .............................................................................................................................................. 89 7.5.2.1 Fluvio-Glacial features............................................................................................................................. 89 7.5.2.2 Recent Geological Features ..................................................................................................................... 89 7.6 Conclusions ......................................................................................................................................................... 89 ii  Mapping Doggerland 8. The Potential of the Organic Archive for Environmental Reconstruction: An Assessment of Selected Borehole Sediments from the Southern North Sea. David Smith, Simon Fitch, Ben Gearey, Tom Hill, Simon Holford, Andy Howard and Christina Jolliffe 8.1 Introduction ........................................................................................................................................................ 93 8.2 Potential and Rationale...................................................................................................................................... 93 8.3 Core Selection ..................................................................................................................................................... 97 8.4 Palaeoenvironmental Assessment ..................................................................................................................... 97 8.4.1 Sampling ...................................................................................................................................................... 97 8.4.2 Visual Assessment ....................................................................................................................................... 97 8.4.3 Assessment of macrofossil (insect and plant) inclusions ............................................................................. 97 8.4.4 Pollen Assessment ........................................................................................................................................ 97 8.4.5 Results .......................................................................................................................................................... 99 8.4.5.1 Vibrocore 53/02/395 (Figure 8.2) ............................................................................................................ 99 8.4.5.2 Vibrocore 54/02/215 (Figure 8.3) .......................................................................................................... 101 8.4.5.3 Vibrocore 54/02/80 (Figure 8.4) ............................................................................................................ 101 8.4.5.4 Borehole 81/50 ....................................................................................................................................... 101 8.5 Discussion .......................................................................................................................................................... 101 8.6 Conclusions ....................................................................................................................................................... 101 Appendix ........................................................................................................................................................................... 103 9. Heritage Management and the North Sea Palaeolandscapes Project Simon Fitch, Vincent Gaffney and Kenneth Thomson 9.1 Introduction ...................................................................................................................................................... 105 9.2 Future Research ............................................................................................................................................... 108 9.3 Cultural resource management procedures in the Southern North Sea ..................................................... 110 9.4 Landscape Characterisation............................................................................................................................ 110 9.5 Threat mapping ................................................................................................................................................ 116 9.6 Threat and Uncertainty Mapping ................................................................................................................... 116 9.7 Final Observations ........................................................................................................................................... 116 Bibliography ..................................................................................................................................................................... 119 Abstract............................................................................................................................................................................. 127 iii  Mapping Doggerland List of Figures Figures Figure 1.1 Hypothetical maximum extent of Doggerland ..................................................................................................... 2 Figure 1.2 Early Holocene Doggerland ................................................................................................................................. 3 Figure 1.3 Holocene shorelines ............................................................................................................................................. 5 Figure 1.4 Current extent of Southern North Sea Megasurvey 3D seismic data ................................................................... 7 Figure 2.1 BGS survey coverage on the UK continental shelf ............................................................................................ 12 Figure 2.2 Distribution of 2D and 3D seismic surveys acquired by the UKOOA ............................................................... 14 Figure 2.3 Three study areas around the English coast ........................................................................................................ 15 Figure 2.4 Area 1 ................................................................................................................................................................. 17 Figure 2.5 Area 2 ................................................................................................................................................................. 18 Figure 2.6 Area 3 ................................................................................................................................................................. 19 Figure 2.7 Intense use of space within the Southern North Sea........................................................................................... 20 Figure 3.1 Typical marine seismic reflection acquisition.. .................................................................................................. 24 Figure 3.2 Seismic resolution of a layer of varying thickness.. ........................................................................................... 24 Figure 3.3 Plots of seismic resolution as a function of burial depth and frequency............................................................. 25 Figure 3.4 A comparison between high frequency 2D seismic reflection line and low frequency 3D seismic lines ........... 26 Figure 3.5 Four possible interpretations of a channel morphology based on a coarse 2D seismic grid............................... 27 Figure 3.6 Typical 3D marine seismic reflection acquisition. ............................................................................................. 27 Figure 3.7 Poole and Christchurch bays. ............................................................................................................................. 29 Figure 3.8 A comparison of seismic images from the Dogger Bank produced using different techniques. ........................ 29 Figure 4.1 HP VISTA infrastructure diagram...................................................................................................................... 33 Figure 4.2 Inspecting data in stereo at the Visual and Spatial Technology Centre .............................................................. 34 Figure 4.3 Seismic data slice showing fluvial channel and estuary ..................................................................................... 35 Figure 4.4 3D amplitude surface within a GIS .................................................................................................................... 37 Figure 4.5 Segmentation of features of interest from a seismic volume. ............................................................................. 37 Figure 4.6 Wrapping of identified features within seismic data .......................................................................................... 38 Figure 4.7 Solid model generated by wrapping ................................................................................................................... 39 Figure 4.8 Removal of elements of the solid model within the Avizo package ................................................................... 39 Figure 4.9 Exported solid model within the GIS system ..................................................................................................... 40 Figure 4.10 GIS layers display within a fully 3 dimensional environment .......................................................................... 41 Figure 5.1 40 and 60m bathymetric contours of the Southern North Sea ............................................................................ 43 Figure 5.2 Bathymetric contours of the Outer Silver Pit area .............................................................................................. 44 Figure 5.3 Hilbert transform time slice at 0.06 seconds ...................................................................................................... 45 Figure 5.4 3-D illuminated view of Ridges A and B ........................................................................................................... 46 Figure 5.5 Dimensions of the OSP and Ridges A and B ..................................................................................................... 47 Figure 5.6 Two 2D seismic lines running through Ridge A and their location ................................................................... 48 Figure 5.7 Quaternary Geology map of the eastern end of the OSP .................................................................................... 49 Figure 5.8 RMS amplitude map of the eastern end of the OSP ........................................................................................... 50 Figure 5.9 Truncated strata on the bed of the Outer Silver Pit ............................................................................................ 52 Figure 5.10 Changes in land area with rising sea level based upon the depth to base Holocene map ................................. 54 Figure 5.11 Postulated land configuration at the time when the OSP sand banks were last active ..................................... 57 Figure 6.1 Time slices centred on prominent salt dome ...................................................................................................... 62 Figure 6.2 Collapse graben, salt swell and fluvial channel .................................................................................................. 63 Figure 6.3 BGS sparker profile 81/03/53 ............................................................................................................................. 64 Figure 7.1 An RMS timeslice covering the whole of the project study area ....................................................................... 68 Figure 7.2 Primary features identified within the Holocene landscape of the southern North Sea ...................................... 69 Figure 7.3 Pre-Holocene features recorded during mapping ............................................................................................... 70 Figure 7.4 General map of all recorded Holocene landscape features including general topographic interpretation .......... 71 Figure 7.5 The Holocene landscape and features within the northwestern quadrant ........................................................... 72 Figure 7.6 Vertical slice through salt dome exhibiting graben collapse .............................................................................. 73 Figure 7.7 Major fluvial channel deviating around an underlying salt structure ................................................................. 73 Figure 7.8 Seismic relief image of the Flamborough head disturbance ............................................................................... 74 Figure 7.9 Western end of the Outer Sliver Pit lake showing outflow channel ................................................................... 76 Figure 7.10 A seismic line across the Outer Silver Pit ........................................................................................................ 77 Figure 7.11 The Holocene landscape and features within the northeastern quadrant .......................................................... 78 Figure 7.12 Junction of rivers and coastline ........................................................................................................................ 79 iv  Mapping Doggerland Figure 7.13 A series of Tunnel Valleys crossing the Outer Silver Pit.................................................................................. 79 Figure 7.14 Location of a small structure that resemble a palaeochannel ............................................................................ 80 Figure 7.15 Modern Sandwaves directly overlying the Holocene landscape ....................................................................... 81 Figure 7.16 The Holocene landscape and features within the southeastern quadrant .......................................................... 83 Figure 7.17 Seismic line across the southeastern quadrant .................................................................................................. 84 Figure 7.18 Representative image of the "mottling" within the seismic data....................................................................... 84 Figure 7.19 Cross section over Markham's Hole (BGS line 80-01-05)................................................................................ 85 Figure 7.20 Seismic timeslice of the area interpreted as a salt marsh .................................................................................. 85 Figure 7.21 The Holocene landscape and features within the southwestern quadrant ......................................................... 87 Figure 7.22 Cross section through Well Hole (BGS Line 93-01-81) ................................................................................... 88 Figure 7.23 Cross section through the large channels in the southwestern quadrant. (BGS Line 93-01-74A) .................... 89 Figure 7.24 Image showing the complex structure of a glacial outwash plain ..................................................................... 89 Figure 8.1 Location of the boreholes in the area included in the 3D seismic survey ........................................................... 94 Figure 8.2 Location of the boreholes requested from the area and primary features identified during mapping ................. 95 Figure 8.3 Location of the boreholes inspected from the area included with the 3D seismic survey .................................. 96 Figure 8.4 Pollen diagram for Vibrocore 53/02/395 ............................................................................................................ 99 Figure 8.5 Pollen diagram for Vibrocore 54/02/215 .......................................................................................................... 100 Figure 8.6 Pollen diagram for Vibrocore 34/02/80 ............................................................................................................ 100 Figure 9.1 Major topographic or economic zones within the study area............................................................................ 107 Figure 9.2 Probable late Palaeolithic land surfaces adjacent to the Norwegian trench ...................................................... 109 Figure 9.3 Seismic data cube illustrating chronostratigraphic relationship between Holocene and earlier features .......... 109 Figure 9.4 The analytical process....................................................................................................................................... 111 Figure 9.5 Broad landscape character zones ...................................................................................................................... 113 Figure 9.6 Cross correlation of major topographic and landscape characterisation zones ................................................. 114 Figure 9.7 Potential for preservation .................................................................................................................................. 115 Figure 9.8 Red flag mapping .............................................................................................................................................. 117 Tables Table 5.1 Position, dimensions and trends of Ridges A and B ............................................................................................ 44 Table 5.2 Average dip of the flanks of Ridges A and B....................................................................................................... 46 Table 7.1 Basic quantitative data relating to identified landscape features .......................................................................... 90 Table 8.1 Summary of sedimentary samples and environmental assessment undertaken from the Southern North Sea vibrocores and boreholes.............................................................................................................................................. 98 Table 9.1 Primary landscape characterisation zones .......................................................................................................... 111 Table 9.2 Ranking of features by relative archaeological potential ................................................................................... 116 v Mapping Doggerland vi  Mapping Doggerland Foreword “The middle of the North Sea? Mammoth tusks and flint spears! Looking for Doggerland? All we need is several million dollars worth of your 3D seismic data!” I am often asked to become part of many research projects, but this phone call from Birmingham was outlining one of the most intriguing. The voice on the phone said “We will come to visit and explain what we want to try to do”. So a few weeks later, we huddled round one of our computer workstations and instead of looking deep down in the seismic data for oil, we applied the latest in petroleum exploration technology to the shallow section. To our amazement, for the first time in thousands of years, the long forgotten surface of Doggerland started to appear. Science does not get more exciting than this and it dawned on us that we were witnessing the start of a new era in marine archaeology. The project brought together a wide range of people, interests, expertise and technology. The project’s achievements, which are of international significance, are a credit to the University of Birmingham, the team and sponsoring companies for which we are proud to have received a British Archaeological Award. These achievements also stand as a lasting tribute to my friend and colleague Dr Ken Thomson, who tragically died following the project conclusion. His infectious enthusiasm for this project still brings a smile to my face as I remember his phone call to me not so very long ago. The work will go on in partnership with Birmingham supported by myself, PGS and I hope many other companies. Huw Edwards (Petroleum Geo Services) July 2007 vii Mapping Doggerland viii  Mapping Doggerland Preface We know that the seas around Britain contain an immense wealth of archaeological sites and remains, potentially without equal elsewhere in the world in terms of their number and diversity. Despite this our detailed knowledge, necessary to promote effective management, is relatively poor and more often than not based on individual sites or find-spots, lacking the opportunity to take a landscape view. Our coasts and seas are also subject to a seemingly ever-increasing rise in development pressure that represents a risk of damage or destruction to the historic environment, which in itself is unique and irreplaceable. English Heritage is the statutory advisor to the UK Government on England’s historic environment, both on land and within the English Territorial Seas and we are committed to: • helping people develop their understanding of the historic environment; • working to get the historic environment on to other people’s agenda; • enabling and promoting sustainable change to England’s historic environment; • assisting local communities to care for their historic environment; • stimulating and harnessing enthusiasm for England’s historic environment, land and sea. The 3D seismics of the Southern North Sea research programme helps us to further these aims through developing new approaches that rely on more partnerships, strategic engagement, speed and flexibility, and clarity and consistency of advice to industry, commercial awareness and customer service. Furthermore, the outcomes of the research described in this volume are of particular benefit to the aggregate extraction industry, clearly justifying our decision to support the programme through the Marine Aggregates Levy Sustainability Fund, of which we are a Distributing Body on behalf of the UK Government. The research programme is also timely in that we are involved in the intensive development of a new heritage protection regime, introduced by the UK Secretary of State for Culture, Media and Sport in November 2002, that proposes innovative changes to the types of archaeological site that could be protected by legislation to include the evidence of past occupation, or use, by humankind, at a landscape scale (i.e. area designation) regardless of whether the monument lies on land now, or has been subsequently submerged under the Territorial Seas. The University of Birmingham research also starkly reminds us that, in relation to the latter point, such remains do not in any way respect present-day administrative boundaries, and the submerged prehistory of the North Sea has value for us all. However, the vast subject area (23,000 square kilometres, yet analysed in 18 months) encompasses jurisdictions from Territorial Seas, and Continental Shelves or Controlled Waters, of many countries, with all the complications that brings in relation to legislative powers, management opportunities to further research, amenity and education, for the benefit of all. The research has been comprehensive: reviewing a range of available methodologies; appraising the geotechnical cores available for ground-truthing; unlocking previously unknown heritage management and research value from legacy commercial seismic data; developing innovative visualisation techniques; integrating marine geological interpretation; incorporating geophysical data, palaeoecological analysis and dating – all at a landscape scale. In total over 690km of palaeo-coastline was observed, together with the interpretation of 10 major estuaries, and extensive areas of salt-marsh, intertidal zone, over 1600km of fluvial systems and 24 lakes/wetlands. The implications for heritage management are also considered, acknowledging that the data generated are one of the largest samples of a well-preserved submerged Holocene landscape anywhere in Europe, indicating the potential for survival of submerged Early Mesolithic coastal sites (c. 10,000 – 8500 BP) to supplement our sparse terrestrial record. Information on adaptation to coastal change during the later Mesolithic (c. 8500 – 5500 BP) and the increasing insularity of British prehistory can also be obtained. Finally, this work is more than just of academic interest. Climate change, global warming and sea level rise are all issues in the forefront of everyone’s minds now. So as well as exploring and interpreting, in unparalleled detail, one of the most extensive, yet least known, prehistoric landscapes in Europe, this research will lend context and time depth to present day challenges for us all. Ian Oxley, Head of Maritime Archaeology, English Heritage July 2007 ix Mapping Doggerland x  Mapping Doggerland Acknowledgements The North Sea Palaeolandscapes Project was a major Martin (BP Exploration Operating Co. Ltd.), Mr Paul achievement for all who worked on it but we would like to Henni (British Geological Survey), Ms. Lia de Ruyter record here our debt to Dr Kenneth Thomson. Ken was (Geological Survey of the Netherlands), Dr Ian Selby the lecturer in Basin Dynamics at Birmingham, an ac- (Hanson Marine Aggregates), Mr Rob Ingram (Hanson knowledged expert on the interpretation and visualisation Marine Aggregates), Mr Joe Holcroft (Cemex), Dr An- of seismic data, and a Principal Investigator on the project. drew Bellamy (United Marine Dredging Ltd), Dr Ceri Tragically, he died as the project concluded on the 18th of James (British Geological Survey), Professor Mike Cowl- April 2007. To the project staff Ken was a pioneer and an ing (The Crown Estate), Dr Mike Howe (British Geologi- inspiration. To those who knew him beyond the project he cal Survey), Dr Colin Graham (British Geological Sur- remains, in our memories, a great friend and irreplaceable vey), Miss Claire Pinder (Dorset County Council), Mr colleague. Martin Foreman (Hull City Council), Mark Rae (Shell UK Limited), Arthur Credland (Hull City Council), Dr Beryl Ken, of course, would have recognised that the North Sea Lott (Lincolnshire County Council), Eileen Gillespie (Brit- Palaeolandscapes Project could not have been attempted ish Geological Survey), Rupert Hoare (WesternGeco), without the material and intellectual support of many peo- Marcia Ritthammer (BP), Mark Bennet (Lincolnshire ple and organisations. Our first debt must surely be to County Council), Dr Gordon Edge (British Wind Energy PGS Ltd who provided the data used in the original pilot Association), Sandra Lane (WesternGeco), Dr Bruno study and the current project study. The kind support pro- Marsset (IFREMER), Ian Taylor (Boskalis Group), Gra- vided by Huw Edwards deserves specific mention. The ham Singleton (CEMEX UK Marine Ltd), Gavin Douglas Aggregates Levy Sustainability Fund (managed by English (Fugro Survey Limited), Ian Wilson (Serica Energy), Dr Heritage) provided the financial support without which the Chris Pater (English Heritage), Verona Szegedi (Conoco- project could not have been carried out. Thanks to Kath Phillips UK Ltd), Neil Anderton Lynx (Information Sys- Buxton, Virginia Dellina-Musgrave, Dr Ingrid Ward, Dr tems), Dr Angela Davis (University of Wales, Bangor), Ian Oxley and Mark Dunkley at English Heritage for all John Rowley (British Geological Survey), Carol Thomson their help and support over the last two years. Following (Chevron Upstream Europe), Dr Jim Bennell (University this, we would acknowledge the work and support of all of Wales, Bangor), Mark Martin (Petroleum Geo- the members of the project management committee who Services), Ann Richards (Devon County Council), Georgia included; Dr Andrew Bellamy (BMAPA, UMD), Mr Chris Boston (Npower Renewables), Dick Lyons (GEMS-UK), Loader (PGS), Dr Virginia Dellino-Mugrave (English Julie Wallace (Centrica), Matthew Coward (CADW), Katy Heritage), Mark Dunkley (English Heritage), Mr Huw Whitaker (English Heritage), David Gurney (Norfolk Edwards (PGS), Dr Joe Holcroft (CEMEX), Dr Justin Dix County Council), Serena Cant (English Heritage), Dr (University of Southampton), Dr Nic Flemming (Univer- Penny Spikins (University of York), Zoe Crutchfield (Joint sity of Southampton), Mr Paul Hatton (Information Ser- Nature Conservation Committee), Margaret Stewart (Im- vices, University of Birmingham), Ms Valerie Scadeng perial College, University of London), Professor Greg (MCS), Mr Jason Aldridge (MCS), Dr Ingrid Ward (Eng- Tucker (CIRES, University of Colorado, Boulder), Mr lish Heritage). Steve Wallis (Dorset County Council), Mr Nick Brown (Norwest Sand and Ballast Co.), Dr Piers Larcombe Professors Geoff Bailey (York) and Martin Bell (Reading) (CEFAS), John Bingham (TCE/Royal Haskoning UK were supportive of the project throughout and assisted Ltd), Andy Barwise (Lankelma), Neil Birch (National greatly in the concluding project seminar when, respec- Wind Power), Brian Ayers (Norfolk County Council), Phil tively, they provided the introductory and summary pa- Harrison (DTI), Gwilym Hughes (CADW), Roger Jacobi pers. (British Museum), Malcom Pye (DTI – HSE), Ron Yor- ston (Tigress), Paul O'Neill (Tigress), Stephen Shorey (Ti- We would specifically like to thank HP1 (Dr Martin gress), Dierk Hebbeln (University of Bremen), Tilmann Walker, Ben Sissons and Nick Hatchard), Mercury Schwenk (University of Bremen), Hannah Cobb (Univer- Computer Systems2 (Valerie Scadeng) and Fakespace3 sity of Manchester), Rachel Baines (BGS Digital Licence (Richard Cashmore) for assistance in the launch of this manager), Joe Bulat (BGS), Paul Henni (BGS), David book. Long (BGS). In addition the following provided much needed advice At Birmingham we must record the encouragment pro- and support: vided by the Pro-Vice Chancellor, Professor Geoff Petts, and our respective heads of department, Professor Ken Mr Graham Tulloch (British Geological Survey), Mr Andy Dowden (IAA) and Professor Paul Smith (GEES). Dr Flaris (BP Exploration Operating Co. Ltd.), Ms Karen Andy Howard acted as academic reader for the publica- tion. Helen Gaffney formatted the text for publication and 1 http://welcome.hp.com/country/uk/en/welcome.html also assisted in reading the proofs. Henry Buglass worked 2 http://www.tgs.com/ assiduously to prepare the illustrations for publication and 3 http://www.fakespacesystems.com/ we thank him for the marvellous job he did. Graham Nor- xi  Mapping Doggerland rie prepared photographic illustrations for which we thank him. Paul Gaffney and Dr Niall McKeown kindly pro- vided translations for the project abstract. The staff at Birmingham Archaeology were unfailingly supportive throughout the project but we would like to thank Smon Buteux, Caroline Raynor and Alex Jones specifically. Within the VISTA division we would wish to acknowl- edge the help and support of all our colleagues including Dr Henry Chapman, Keith Challis, Mark Kincey, Steve Wilkes and Meg Watters. xii  Mapping Doggerland 1 Mapping Doggerland Vincent Gaffney and Kenneth Thomson Eventually, all things merge into one, and a river runs through it. The river was cut by the world's great flood and runs over rocks from the basement of time. On some of the rocks are timeless raindrops. Under the rocks are the words, and some of the words are theirs. I am haunted by waters. Norman Maclean (1902-90). A River Runs Through It 1.1 Introduction 1.2 The context of study The inundated prehistoric terrain of the North Sea basin This potential of the southern North Sea for geological and remains one of the most enigmatic archaeological land- archaeological research was recognised early, by Sir scapes in northwestern Europe. This region was lost to the Clement Reid, in a book on the submerged forests of the sea over a period of c. 11,000 years following the last gla- United Kingdom published in 1913. Here Reid noted, in a cial maximum and the change in relative sea levels re- remarkably perceptive paragraph that “the geologist sulted in the loss of an area larger than the United King- should be able to study ancient changes of sea-level, under dom (Coles 1998). The region therefore contains one of such favourable conditions as to leave no doubt as to the the most extensive and, presumably, best preserved prehis- reality and exact amount of these changes. The antiquary toric landscapes in Europe (Fitch et al. 2007). Moreover, should find the remains of ancient races of man, sealed up during the Mesolithic, the period primarily covered by this with his weapons and tools. Here he will be troubled by report, the area was probably an important habitat for no complications from rifled tombs, burials in older hunter-gatherer communities (Morrison 1980, 118). This graves, false inscriptions, or accidental mixture. He ought vast archaeological landscape provides Europe with an to here find also implements of wood, basketwork, or ob- immense challenge. How are we to investigate, interpret jects in leather, such as are so rarely preserved in deposits and manage the heritage of this extraordinary, but largely above the water-level.” (Reid 1913, 9). inaccessible, landscape? Following this promising start, the pioneering work of Sir This latter point is of prime importance. Although inac- Harry Godwin on moorlog (peat) deposits associated with cessible and, in most senses, invisible, the archaeology of the 1931 Colinda harpoon find from the Leman and Ower the region is as fragile as any terrestrial correlate. In terms banks, demonstrated the capacity of these extensive sub- of mineral and natural wealth the North Sea basin is a stra- merged deposits, to provide paleoenvironmental evidence tegic resource for the United Kingdom and all the coun- and proved their terrestrial origin, (Burkitt 1932, Godwin tries that surround it. Its geographical position ensures and Godwin 1933). Shortly after, Sir Graham Clarke's that this extensive region also functions as a key infra- (1936) seminal work on the “Mesolithic Settlement of structural and communications locus (Fleming 2004, 113 - Europe” acknowledged the probable settlement potential 117). The area is therefore under intensive developmental and the cultural significance of the area. It is notable, pressure from a range of threats including mineral extrac- however, that these early initiatives were not substantively tion and the direct impact of construction. Specific threats built upon. Whilst this must have largely been a conse- range from the laying of pipelines to, more recently, the quence of inaccessibility of the archaeological deposits development of wind farms, the wider issues of mineral Clement Reid (1913, 3) also, presciently, predicted that extraction and the extensive, generalised, impact of fishing “the archaeologist is inclined to say that [these deposits] and commercial trawling (Dix et al. 2004, section 1.4). belong to the province of geology, and the geologist re- The implication of such threats, in environmental terms, is marks that they are too modern to be worth his attention; probably apparent to most aware individuals and organisa- and both pass on.” The demise of active archaeological tions with an interest in the region. However, the signifi- research across the North Sea basin from the mid twentieth cance of the southern North Sea is raised in cultural terms century was paralleled by the marginalisation of the pre- when one considers that whilst the continental shelf re- sumed archaeology of the area. Whilst not denying that tains, arguably, the most comprehensive record of the Late some archaeologists were aware of the archaeological po- Quaternary and Holocene landscapes in Europe (Fitch et al tential of the region, the area was increasingly interpreted 2005), this landscape was also extensively populated by or represented as a land bridge from mainland Europe to humans and at specific periods may well have been a core Britain (Coles 1998). The largely unspoken implication habitat at a European level (Coles 1998; Flemming 2004). was that the inundated area was unimportant in cultural terms (Coles 1999, 51). In many ways it might be said that there was a spiral of indifference towards the archae- ology of the region. 1  Mapping Doggerland Figure 1.1 Hypothetical maximum extent of Doggerland (redrawn from Coles 1998) 2  Mapping Doggerland Figure 1.2 Early Holocene Doggerland (redrawn from Coles 1998) More recently, the significance and potential of the Coles, Dr Nic Flemming has worked unceasingly on archaeological and geomorphological record of the promoting the archaeology of the area, most notably southern North Sea has become an emerging academic through his recent edited volume on the archaeology of the interest (Flemming 2004). Within archaeology this region (Flemming 2004). The ALSF-funded assessment phenomenon can be traced directly to the 1998 review of the archaeological potential of the British continental article by Professor Bryony Coles, “Doggerland: a shelf by Dix et al. (2000) also provided a significant speculative survey” (Figure 1.1 and Figure 1.2). After context for further research around the coast of Britain. 3  Mapping Doggerland However, the fundamental consequence of these understanding of the morphology of the Holocene land- publications has actually been an increasing awareness of scape of the southern North Sea is largely based on the deficiencies of our knowledge of the North Sea in bathymetric data. This is supported by considerable ex- terms of the nature or extent of the archaeological deposits ploratory activity by the geological services of countries of the region. Such sentiments have, more recently, been bounding the sea and commercial groups seeking to ex- echoed in the contents of a series of Department of Trade ploit the area. Work by Jelgersma (1979) produced a se- and Industry regional strategic environmental assessments, ries of highly influential maps for the major changes in the or "SEA" volumes, for the mainland British marine coastline from 18,000BP to 8,300BP and, significantly, territories (Flemming 2002, 2003, 2004b, 2005; Wickham- noted the formation of an island at the Dogger Bank Jones and Dawson 2006). around 8700BP. An attempt was then made to place this landscape within a cultural context by Coles (1998) who The lack of knowledge associated with the North Sea dubbed the emergent plain “Doggerland”. This work con- Holocene surfaces was so profound that, as recently as tained hypothetical reconstructions of the coastline from 2004, Flemming noted that the inundated landscapes of the the Weichselian maximum through to 7000BP, but was Southern North Sea were essentially terra incognita. This ultimately based on the earlier study by Jelgersma. Whilst profound lack of knowledge was maintained despite the this approach has provided an overview to the area it re- results of geological studies that suggested that sediment mains true that the palaeogeography of the region re- in the area that might be associated with human occupa- mained lacking in critical detail. Researchers, including tion achieved depths of 1 to 5m thick and, locally, a Lambeck (1995), Shennan (2000), Shennan and Horton maximum thickness of 30-40m (Laraminie, 1989). The (2002) and Peltier (2004), have used isostatic rebound potential of these substantial, unexplored deposits has been models to help constrain and improve the present bathym- underscored by the significant number of human artefacts etry-based models. This has resulted in minor modifica- and mammal remains that are often trawled or dredged tions to current coastal models but the lack of detail within from the region (e.g. Van Kolfschoten T. and Van Essen the landscape (e.g. the location of fluvial systems, details 2004). It is usually assumed that such finds originated of coastline etc), and the failure to incorporate late Holo- from eroding or disturbed seabed deposits (Flemming, cene and recent sedimentation, still remain significant is- 2002; Glimmerveen et al. 2004). Flemming (2002) sug- sues (Bell et al 2006, Box 1, 14). In so far as these factors gested that richer environments for the origin and preser- have the effect of masking the true relief of the palaeo- vation of archaeological materials could include Holocene landscape it is unlikely that an adequate appreciation of fluvial valleys and the Outer Silver Pit, a vast sea inlet the human landscape can be achieved using data provided which existed to the south of the Dogger Bank from 8,000 by previous studies. - 7,500BP. In methodological terms, therefore, the investigation of These general impressions were supported by the increas- past marine environments has generally been limited by ing density of sites located around contemporary coasts available data that had serious limitations. These have that, presumably, can be extrapolated onto inundated included: coastlines beneath the North Sea. This information, clearly suggests that the lack of material associated with 1. Seabed sampling and shallow coring: These pro- deeper waters indicates an absence of evidence rather than vide high quality chronological, sedimentological evidence of absence (Fischer 2004, figure 3.3; Pedersen et and environmental data. However, data is widely al 1997). The paradox of the North Sea, therefore, is that spaced and provides a poor spatial framework although the environmental and cultural potential of the and thus limits its use in assessing the larger region remains largely unknown, it may still be correct to landscape and its archaeological significance or suggest the landscape archaeology of the region is signifi- potential. cant at a global level (Mithen 2003, 154-157). Sourcing 2. High resolution 2D seismic: Traditional shallow inundated deposits, and thereby providing an option to seismic techniques (e.g. Stright 1986; Velegrakis protect surviving archaeology, is a key, but problematic et al., 1999) have provided detailed information goal. on the architecture of sedimentary systems but as the data is generally acquired as a series of 2D 1.3 Previous methodological profiles, a weak three-dimensional framework is created due to the necessary interpolation be- approaches tween the profiles. 3. High resolution 3D seismic: These data represent If our knowledge of the archaeological deposits of the a significant advance in imaging shallow geology North Sea is so tenuous, it might be hoped that the larger (Bull et al 2005, Gutowski et al. 2005, Muller et geomorphological context of the region offers the oppor- al. 2006), but the centimetre-scale resolution of tunity to make general observations on the potential nature the data dictates that only small areas (<1km2) of preserved archaeological deposits. Unfortunately, al- can be realistically surveyed. though the North Sea has been the subject of extensive 4. High resolution bathymetry: This may provide exploration for a variety of commercial or academic rea- excellent images of the seabed topography and is sons for decades, this is probably not the case. Our current 4  Mapping Doggerland capable of providing detailed images of Late correlate to surfaces relating to earlier periods, Pleistocene and Holocene features that have a in many instance there may be a significant dif- bathymetric expression. Whilst bathymetry pro- ference (up to c. > 20 m) between them. This can vides a reasonable approximation for the land sur- lead to inaccurate representations of shoreline face for the area it can rarely consider, or attempt positions (up to 60 km difference) and past topog- to resolve, burial of features that may have oc- raphy can be markedly misinterpreted. The bed- curred during or after submersion (Cameron et rock horizon represents a minimum value that al., 1992). Consequently, the technique is unsuit- could be used in reconstruction. However, mod- able for areas including the southern North Sea ern bathymetry does not represent a maximum and the Irish Sea, where deposition has buried value as processes of erosion may have reduced most of the Quaternary and Holocene. The scale its height over time”. of this problem was clearly stated by Dix et al. (2004, 89); “although modern bathymetry can Figure 1.3 Holocene shorelines (after Jelgersma, 1979) The limitations of these methodologies are also apparent in nology and scale of archaeological application has tended the archaeological literature. There is considerable interest to restrict studies to the immediate coastal zone and to in the investigation of marine features and the identifica- relatively small, intensively surveyed areas (Mueller et al tion of marine landscapes. However, the available tech- 2006). Whilst this has been adequate for exploration of 5  Mapping Doggerland known sites (usually of the historic period, e.g. Paoletti et purposes (see Cameron et al 1992 for the BGS definition al. 2005) or micro-regional survey (Pedersen et al 1997), it of the Southern North Sea region). has largely precluded major landscape exploration. Prior to the current work, therefore, there was no plausible to- Data for the Southern North Sea is provided through a pographic or geomorphological context that could provide research agreement between the University of Birmingham a credible proxy indicator for human activity across the and Petroleum Geo-Services2 (PGS) and we are former Holocene landscapes of the North Sea. particularly indebted to Mr Huw Edwards for facilitating access to this information. PGS MegaSurveys are based 1.4 Towards an alternative on seismic data that have been released by oil companies, PGS owned seismic surveys and non-exclusive seismic methodology data made available through other geophysical contractors. Usually these data are available as 3D time migrated The impetus and opportunity to develop a methodology seismic surveys. Although quality controlled the different using 3D seismic data to deal with this challenging land- age and data acquisition methods used to collect data scape derived from doctoral research carried out at the demand that the seismics vary in quality3. University of Birmingham by Simon Fitch and under the supervision of the project’s principal investigators, Gaff- Figure 1.4 illustrates that these data exist as a significant ney and Thomson (Fitch et al. 2005). The 3D seismic continuous data source across much of the Southern North datasets acquired on the United Kingdom continental shelf Sea. Whilst the data does not currently stretch coast-to- for exploring deep geology represent a major resource for coast the total full-fold area of coverage of the Southern understanding Late Pleistocene and Holocene geology. North Sea Megasurvey is in excess of 23,000 km² and With extensive regional coverage and spatial resolutions represents more than 60 original 3D surveys belonging to of c. 12.5m such datasets provide the opportunity of map- 20 different data owners. Altogether, this data set repre- ping relatively recent geology at a regional scale and with sents the largest available data source for the exploration relative speed. Standard geophysical interpretation tech- of the palaeogeography of the Southern North Sea region niques usually used on such data to explore deeper fea- and, in archaeological terms, constitutes the largest con- tures, augmented by volume and opacity rendering, pro- tiguous archaeo-geophysical survey programme ever at- vide significant advantages in reconstructing palaeo- tempted. The work also follows the tradition of seismic geographies and allow the true 3D architecture of Late study and large-scale archaeological remote sensing pro- Pleistocene and Holocene systems to be established (see jects managed at Birmingham (Gaffney et al. 2000, Thom- Thomson and Gaffney, this volume). son 2004, Barratt et al. 2007). The original research at Birmingham coincided happily Within this context, the specific aims of the project were: with an emerging requirement to manage the archaeologi- cal heritage in the light of aggregate extraction within the • To use the existing 3D seismic datasets acquired area. Funding for a larger project was made available to on the United Kingdom continental shelf for ex- the Birmingham team through the Aggregates Levy Sus- ploring Late Quaternary and Holocene geology tainability Fund. This fund, administered by English Heri- over an area of the Southern North Sea. tage, seeks to promote best practice in planning aggregate • To provide maps of the recent geological se- extraction and to provide data to support the protection of quence at a regional scale. our marine heritage that may be impacted by such activi- ties1. This serendipitous opportunity permitted the team to • To provide detailed digital mapping of the topog- develop a methodology centred around the use of exten- raphic features of the region and to use voxel sive 3D seismic data to map Holocene features across a rendering to allow the true 3D architecture of large area of the southern North Sea. A team of three re- Late Quaternary and Holocene systems to be es- searchers was initially employed to work on this data; tablished. Kate Briggs, Simon Fitch and Dr Simon Holford. The • To compare the Holocene topographic data with papers presented in this volume present the results of this available core and borehole data to ground truth work. data and calibrate results. • To provide a model of survival potential for envi- The surfaces investigated as part of this project effectively ronmental and archaeological deposits within the represents the Holocene landscape inundated between area of the Southern North Sea to be used by the 10,000 and 7,500BP and, in archaeological terms, are aggregates industry to plan extraction and mitiga- associated with the Mesolithic period (Cameron et al. tion strategies. 1992; Jelgersma, 1979; Lambeck, 1995). Given the origin • To use data on environmental and archaeological of the data the study area was defined by the extent of potential to provide an extensive depositional available data rather than the probable historic Holocene shorelines (Figure 1.3) or notional areas defined for other 2 : http://www.pgs.com/) 1 3 http://www.english- http://www.pgs.com/business/geophysical/research/librar heritage.org.uk/server/show/nav.1315 y/mc3d/dbaFile7567.html?1=1&print=true 6  Mapping Doggerland map of the Southern North Sea for use for aggre- • To disseminate knowledge of the methodology gate developmental purposes. and outcomes of the project for the purposes of • To utilise seismic attribute analysis to map depo- supporting and developing the aggregate industry sitional systems in detail and to make calibrated and management of the mineral resource. lithological predictions that may be used in ag- gregate deposit modelling. • To provide palaeocoastline data, which may be used in the development and calibration of cur- rent sea level and palaeobathymetry models. Figure 1.4 Current extent of Southern North Sea Megasurvey 3D seismic data (source PGS http://www.pgs.com/business/products/datalibrary/nweurope/southernnorthsea/snsmegasurvey/) A full description of the technologies utilised to explore (this volume) that demonstrate the detail of specific geo- and integrate the available seismic data is provided in the morphological structures, including the nature of internal papers by Thomson and Gaffney and Fitch et al. (the atlas, features identified within the Outer Silver Pit. this volume). It is enough to note here that the signifi- cance of the first results of this work was rapidly appreci- It is true, however, that as the project proceeded there was ated, and that mapping of the area proceeded apace, as is an increasing concern within the team concerning not so demonstrated in the atlas paper by Fitch et al. This is com- much the extent of available supporting data (including 2D plemented by papers from Briggs et al. and Holford et al. seismics lines, cores etc) but the quality, or even availabil- 7  Mapping Doggerland ity, of some of this information. In particular, spot- raphic context of the region. As a consequence, the heri- checking of cores for environmental potential suggested tage agencies of countries bounding the North Sea may that the description of core data was, in some cases, mis- well have to re-assess their marine management strategies leading and that the environmental potential of some sam- in the light of this information. In this context the steps ples might have been compromised by their storage (Smith toward a historic landscape characterisation methodology, et al. this volume). Consequently, a project variation was as described in the final paper of the volume, are we be- submitted to English Heritage that supported a data audit lieve an important contribution towards the management to assess the extent of available 3D coverages, the avail- of problematic, marine landscapes. ability of other supporting datasets and their potential for research. Dr Mark Bunch was employed for these pur- Ultimately, the principal achievement of the project has poses and the results of this important work are presented been to explore and begin to interpret in unparalleled de- in summary within this volume. tail one of the most extensive, yet least known, prehistoric landscapes in Europe. Whilst our knowledge remains im- In providing this extended introduction it is worth consid- perfect the area is no longer the “terra incognita” pondered ering the wider significance of this study. Initially, it upon by Flemming less than 3 years ago (Flemming 2004). should be stressed that this volume represents our initial, Indeed, in the light of our previous lack of knowledge, the tentative steps towards providing a robust methodology for scale of the work carried out by this project is truly star- the investigation of deeply buried and inundated historic tling. The analysis of 23,000 square kilometres of seismic land surfaces. The results of the North Sea Palaeoland- data is comparable to carrying out a geophysical survey scapes Project are, we firmly believe, a major contribution over a country the size of Wales. It is a cliché to assert to our understanding of the Holocene land surfaces of the that the past is a foreign country. However, in the case of North Sea. From a methodological perspective this is of the North Sea Palaeolandscapes Project, it is hardly hyper- enormous significance. However, the Holocene land- bole to assert that, along with the outstanding contribu- scapes discussed here do not represent the total of avail- tions of Coles, Flemming, Dix and others, the project has able data for the British continental shelf. Comparable effectively begun to provide the archaeological outline of a areas of submerged, but previously habitable landscapes, previously undiscovered European realm. can also be found in the Black Sea (Ryan and Pitman 2000; Ballard et al. 2000), the Florida Gulf (Stright 1986; The final point to be made is more emotive. The loss of Faught 1988; Marks and Faught 2003; Faught 2004), the extensive late Pleistocene and Holocene landscapes, after Gulf of Arabia (Lambeck 1996) and a number of other the last glacial, represents the only previous period during regions of the world (e.g. Dortch 1997; Bailey 2004), which modern man experienced the impacts of global many of which have also been subject to extensive explo- warming at a scale predicted for the next century. The ration for mineral extraction. The work presented here is North Sea Palaeolandscapes Project provides quantitative therefore replicable elsewhere and, if implemented, the and visual evidence for the nature and significance of such results for regional research are likely to be as exciting and change. The recreation of the Mesolithic landscape and challenging as those derived for the Southern North Sea. coastline may, ultimately, be factored into improved coastal models and this is a practical and desirable out- Of course, there is room for development. This project, come. We should not forget, however, that this was a which lasted for a mere 18 months, would have benefited populated land. The loss of such extensive areas, insidious from a more substantive integration of supporting informa- and slow overall but terrifyingly fast at times, must have tion, including high resolution 2D seismic data and further been devastating for the Mesolithic populations of the core data. Unfortunately, the audit carried out as part of great northern plains. The coastlines, rivers, marshlands the study suggests that existing data will not always be and hills mapped during this project were, for thousand of available or sufficient for the purposes of refinement or years, parts of a familiar landscape to the hunter-gatherers ground truthing of results. There is, therefore, a real need of northwestern Europe. The land and its features would for dedicated, expensive ship time to provide new data to have been named; some areas might have been revered ground truth and extend the results of this study. and held personal associations or ancestral memories dear to these peoples. It is almost impossible for us now to Despite these observations, the scale of the work and the comprehend the demise of environments and ecologies fact that the landscape transcends national boundaries en- that supported communities, tribes and entire peoples. sures that, aside from primary archaeological or geomor- Whole territories may have disappeared within the mem- phological output, the implications of the results are of ory of a single generation, and the stress to the indigenous international significance in terms of heritage manage- populations is beyond our experience (Mithen 2003). The ment, at the very least. We have presumed, for nearly a memories and associations of cultures disappeared, with century, that the North Sea contained a significant ar- the landscape itself, as sea levels rose and the land re- chaeological record but it has always been a challenge to treated. manage a resource that was largely inaccessible, entirely unpredictable and, essentially, a hypothetical construct. As this project concludes, the UN Intergovernmental Panel The results presented here suggest that this record may be on Climate Change is finalising its report on the nature, traced, in part, through the reconstruction of the topog- scale and implication of global warming 8  Mapping Doggerland (http://www.ipcc.ch/). At such a time, and when climate fate of the Holocene landscapes and peoples of the North change, global warming and sea level rise are now ac- Sea may yet be interpreted, not as an academic curiosity, cepted as amongst the greatest threat to our lifestyles, the but a significant warning for our future. 9 Mapping Doggerland 10  Mapping Doggerland 2 Coordinating Marine Survey Data Sources Mark Bunch, Vincent Gaffney and Kenneth Thomson 2.1 Introduction ensure the latest navigation charts are accurate. Conserva- tion and regulatory bodies also undertake surveys to moni- The North Sea Palaeolandscapes Project (NSPP) has pri- tor environmental change arising from seabed use. marily relied upon the large Southern North Sea (SNS) 3D seismic MegaSurvey developed by Petroleum GeoServices Current government legislation promotes renewal and in- (PGS). This is a regional merge of surveys acquired by vestment in offshore industrial sectors. This is particularly the petroleum industry, oil and service companies over the notable in respect of the growth in renewable energy pro- last 20 years (Terrell et al., 2005). It offers an unparalleled jects. Offshore ‘green’ energy generation is seen by the source of data for visualising and interpreting buried fea- government as a key strategic objective in the battle to tures of the emergent Holocene landscape at a regional mitigate anthropogenic climate change. Funding systems scale providing a broad spatial framework to explore the including the Aggregates Levy Sustainability Fund offshore prehistoric landscape following the last glacia- (ALSF) have involved academia directly in combining tion. research and development with monitoring and evaluation. As stated by Thomson and Gaffney (this volume), 3D Offshore industrial surveys are specifically designed with seismic survey data is one of a diverse suite of survey data industrial application in mind. Oil companies carry out types that exists within the SNS. Different surveys pro- large-scale acoustic surveys to visualise geological struc- vide varying information about the seabed and subsurface. tures and stratigraphy at depth. Smaller-scale, shallow Survey design dictates the penetration through the subsur- geotechnical surveys are carried out on their behalf for face as well as the spatial resolution of the data set. This engineering and compliance purposes. Other industries, clearly has implications for the survey scale. Owing to the including marine aggregates producers, specifically target location of viable hydrocarbon resources and the logistics the shallow marine sediments as a resource. These groups of development, the SNS MegaSurvey is confined to a regularly conduct 2D seismic sub-bottom profiling, high region lying at least 11 miles off the English coast. There resolution, seabed bathymetric mapping using sonar and is a so-called “white band” between the coast and the direct sediment sampling by shallow core or grab. An- MegaSurvey that that may be crossed only by smaller other growing industrial application within territorial and scale surveys or 2D seismic lines. If we are to understand offshore UK waters is the development of large offshore how prehistoric societies interacted with the evolving wind farms. Aside from carrying out surveys in order to postglacial shoreline, an appropriate strategy must be de- estimate their direct environmental impact, wind farm pro- veloped to utilise surveys conducted within the white posals require engineering survey to ensure the safe instal- band. In addition, we will gain greater insight into behav- lation and anchor of turbines to the seabed. ioural patterns in the hinterland by augmenting our broad model of the emergent landscape with more focussed in- The activity of the British Geological Survey (BGS) terpretations using survey data at a finer scale. This paper uniquely occupies parts of all three sectors. The BGS is a seeks to investigate the variety of survey data available to public sector organisation that is a component of the Natu- serve these purposes and also to assess their relevance to ral Environment Research Council (NERC), a body report- the broader aims of the North Sea Palaeolandscapes Pro- ing to the Department of Trade and Industry (DTI). Sig- ject (NSPP) as identified in Gaffney and Thomson (this nificantly, it is also an academic body undertaking applied volume). research that is part-funded by the government and its own commercial activities. The BGS has undertaken extensive survey of the shallow seabed sediments in UK territorial 2.2 Identification of sources: who and offshore waters over many decades. acquires, owns or holds survey data? 2.3 Key data repositories There are three groups that require offshore survey data: Under its remit for strategic research and surveying, the governmental organisations, industry and academia. Gov- BGS has compiled the national archive of marine shelf ernmental organisations, including conservation and regu- seabed sedimentary and geological maps. These are de- latory bodies, survey for monitoring and engineering pur- rived from shallow geophysical survey and sediment sam- poses as dictated by legislation and infrastructural plan- pling. Figure 2.1 illustrates the distribution of acquired ning. For example, former public service organisations shallow geophysical survey lines and shallow sediment including British Telecom, are responsible for regularly samples around the UK according to the BGS data archive surveying seabed cable and pipeline routes to monitor their web GIS, GeoIndex, as of October 2006. It is clear that structural integrity. The British Hydrographic Office con- geophysical surveying is possible within all waters beyond tinuously maps the sea floor at a fine spatial resolution to a certain depth. Direct sediment sampling was achieved 11  Mapping Doggerland closer to shore and is particularly concentrated within en- seismic reflection survey using either a boomer, sparker or closed estuaries and embayments. pinger acoustic source (Delgado 1998, Gillespie pers. comm. 2007). Each of these is an industry name that re- A variety of geophysical survey techniques have been util- fers to the method for generating the acoustic pulse that ised as part of this process. In order to map the shape of propagates radially outward, through the water column and the sea floor all areas have been explored using an echo underlying sediment, to be reflected back at interfaces of sounder and some by more sophisticated sonar techniques. contrasting density. BGS 2D seismic survey data is par- The majority have been surveyed using a potential field ticularly suitable for mapping of detailed features as these instrument: a gravimeter or magnetometer. These give an were acquired using a high-resolution single channel re- indication of a variation in bulk density and magnetisation ceiver (hydrophone) system. The boomer generates a of seabed material laterally. However, the most useful source wavelet that tends to produce the clearest results for technique for visualising the vertical stratigraphy of sea- interpreting shallow features. bed sediments is sub-bottom seismic profiling. Most of the BGS geophysical survey lines have been covered by a Figure 2.1 BGS survey coverage on the UK continental shelf as displayed by GeoIndex, October 2006. (a) shows geophysical profile lines, (b) shows shallow sediment sample stations Sediment samples may be disposed of following catalogu- obligation to retain certain survey data in perpetuity for ing or may be retained in archive by the BGS. The store, recall by the DTI at any time. However, this obligation at Loanhead, in Edinburgh, forms part of the National does not apply to core samples or cuttings. The decision Geosciences Data Centre (NGDC), which offers an archiv- to donate these rather than dispose of them, is made at the ing and curatorial service for voluntary donations of sur- discretion of the operator. vey data, samples and field notes by both industry and amateur geoscientists. The NGDC has a section dedicated The NHDA is a joint venture between the Department for to conventional prospecting data – seismic surveys and Trade and Industry (DTI) and Common Data Access Lim- borehole logs – that is divided between onshore and off- ited (CDA), a not-for-profit subsidiary of the UK Offshore shore activities. These sections are the UK Onshore Geo- Operators Association (UKOOA), which administers data physics Library (UKOGL) and the National Hydrocarbons exchange amongst its members. The BGS manages the Data Archive (NHDA) respectively. The NHDA can act NHDA and runs the associated UK Digital Energy Atlas to relieve UKCS licensee operators of the expensive legal and Library (UK DEAL) project: a public web site aiming 12  Mapping Doggerland to promote data and information relevant to the explora- National Assembly for Wales within Welsh waters. The tion and production of hydrocarbons on the UKCS. UK regulator determines the necessity of conducting an envi- DEAL aims to provide a publicly accessible national cata- ronmental impact assessment (EIA). The environmental logue of UKCS geosciences data, their sources and own- and cultural heritage of the seabed within territorial waters ers, to facilitate data sharing between the public and/or is of concern to TCE. However, the ODPM awards li- prospective licence applicants and data owners/current censes and acts as the environmental regulator under ad- licensees. vice from conservation bodies including English Heritage (EH). Figure 2.2 shows the distribution of previously acquired UKOOA 3D seismic survey blocks. 3D seismic has been Marine aggregates comprise up to a quarter of the total the bulk survey data type used for mapping the broad fea- sand and gravel used in Britain (Bellamy, 1998). The ma- tures of the emergent landscape in the SNS. 3D surveys rine aggregates industry generates almost half TCE’s an- are ideal for reconnaissance mapping because they are nual marine income and constitutes approximately a third acquired over large areas but can still be manipulated of commercial activities in its marine portfolio (TCE, visually to resolve sufficient detail at the scales of interest 2006). Marine aggregates producers generally target shal- relevant to this project. Unfortunately, as Figure 2.2 dem- low glacial sand and gravel banks, usually within the 12- onstrates, the distribution of these data is uneven and large mile territorial limit. Wenban-Smith (2002) emphasises areas remain without coverage. The large gap off the the requirement to work with marine aggregates producers northeast coast of England is a prime example, and is in an to develop a system for identifying culturally significant area of great interest to archaeologists (Waddington 2006). deposits in order to identify significant threats prior to disturbance. In particular, he predicts the likelihood that The DTI Guidelines for the Release of Proprietary Seismic thin, well preserved sedimentary layers with high archaeo- Data UKCS 2004, and the Agreement between the Interna- logical potential directly overly Marine Aggregates Depos- tional Association of Geophysical Contractors and the DTI its (MADs) within the English Channel and SNS. Clearly, for the release of speculative seismic data 2002, govern the the work of recent ALSF projects, in collaboration with availability of survey data featured on the UK DEAL web Aggregate groups as has been the case with the NSPP, will GIS. clarify the extent that this cooperation occurs or can occur given the use of available data for cultural resource man- Petroleum industry well site surveys are an excellent agement purposes. source of high vertical resolution 2D seismic survey data. Unfortunately, these data are not ‘tied’ to the licensed oil In order to gain a licence for aggregate extraction, a ma- and gas licence block and data owners are not required to rine aggregates group must complete a two-stage applica- retain these in perpetuity. Offshore operators and the DTI tion process to TCE and the ODPM. Consultation on a continue to debate whether digital site survey data should statutory framework is currently underway. At present, be submitted by law to the NHDA. Currently, only c. this involves resource assessment and an EIA that must 1000 site survey reports have been donated to the NHDA address the potential impacts of removing seabed material voluntarily. These represent a fraction of the data acquired on coastal erosion, fishing, archaeology and marine life. offshore. Both necessarily require geophysical surveying and direct sediment sampling. 2.4 Governance and surveying within Offshore wind farms currently generate almost 16% of the territorial waters total electrical power generated by UK wind farms. Most offshore wind farms occur within the 12-mile nautical The Crown Estate (TCE) owns the seabed within the 12- limit and therefore fall under the jurisdiction of TCE for mile territorial limit and is responsible for issuing leases appraisal and licensing. Licensing for wind farm devel- and licenses for its commercial and infrastructural devel- opment begins with a request from the DTI to TCE to an- opment. These include leases of easement for seabed ca- nounce a competitive tender process for new wind farm bles and pipelines, licenses for the extraction of minerals, developments. After pre-qualification, developers must excluding oil, gas and coal (principally marine aggre- win statutory consents from two government departments, gates), and leases and licenses for the construction of off- the Offshore Renewables Consents Unit (ORCU), and the shore wind farms. Under the terms of The Marine Works Marine Consents Environment Unit (MCEU). This stage (Environmental Impact Assessment) Regulations 2007 involves submission of an EIA that includes consideration consultation document released by DEFRA in 2006, the of submerged archaeology. After securing these consents regulators of development and disposal activities within a lease or licence is granted by TCE. territorial and offshore waters of England and Wales are the bodies responsible for licensing. These are the Secre- tary of State for projects within English waters, and the 13  Mapping Doggerland Figure 2.2 Distribution of 3D seismic surveys acquired by the UKOOA 2.5 Public metadata resources environmental schemes and designations from six UK government organisations: DEFRA; English Heritage There are a number of web-based GISs detailing available (EH); NE; Environment Agency (EA); the Forestry Com- survey data across the UK coastal shelf. The MAGIC GIS mission (FC); and the Department for Communities and (http://www.magic.gov.uk) brings together information on Local Government. The MAGIC GIS features a Coastal 14  Mapping Doggerland and Marine Resource Atlas theme that presents metadata consortium of twelve partners across the UK, Ireland, the from offshore regulatory bodies including the DTI and the Netherlands, Belgium and France. The aim is to produce BGS. These can be queried elsewhere on the project web- standardised seabed habitat maps for northwest Europe. site. The MESH website hosts a basic GIS. A much larger metadata archive of all forms of marine surveying within The Mapping European Seabed Habitats (MESH – specified regions can also be queried elsewhere on the http://www.searchmesh.net/default.aspx) is an interna- website. tional marine habitat-mapping programme involving a Figure 2.3 Three study areas around the English coast The British Oceanographic Data Centre (BODC) website of the NERC MetaData Gateway, which also provides (http://www.bodc.ac.uk) hosts the European Directory of access to catalogues of data held by the other NERC data Marine Environmental Datasets (EDMED). Here, there is centres: British Atmospheric Data Centre (BADC); British information available for bathymetric, seismic and sedi- Geological Survey (BGS); NERC Earth Observation Data ment sampling (core, dredge, grab) surveys undertaken in Centre (NEODC); Antarctic Environmental Data Centre UK offshore waters by a range of research groups and in- (AEDC); Environmental Information Centre (EIC). The dustrial organisations. These fall under the “Geology – NERC MetaData Gateway is in the process of being re- Geophysics – Sedimentation” theme. The BODC is part placed by a service based on the NERC DataGrid, a web- 15  Mapping Doggerland based search facility that aims “to make the connection holes within a few tides (Murray, 1987; Newell et al., between data held in managed archives and data held by 1998). Area 1 is traversed by a number of pipelines and individual research groups so that the same tools can international telecommunications cables that continue west compare and manipulate data from both sources” to Europe and beyond the Western Approaches. These (http://ndg.badc.rl.ac.uk/). must be surveyed annually to assess their structural integ- rity. Area 1 also supports three licenses to dredge marine The DTI runs a website dedicated to the regulation of the aggregates and a single wind farm licence. Figure 2.4a oil and gas industry, and offshore development in general summarises these known threats to seabed archaeology. (http://www.og.dti.gov.uk). Users can access a repository of existing development licences and all petroleum indus- BGS survey here has been dominated by sediment sam- try survey applications (PON14) and decisions since 2001 pling, though the channel is relatively well covered by 2D (Conservation of Habitats Regulations 2001). Spread- seismic profiles shot by both the BGS and the petroleum sheets of metadata associated with these applications iden- industry (Figure 2.4b). BGS survey data is straightforward tify the applicant company, and the type, time and location to obtain by purchase order at standard rates. Unfortu- of their proposed surveying. nately, the cost of BGS data may well be prohibitive to extensive or academic exploration and this may yet prove SeaZone Solutions Limited provides a commercial source a major barrier to promoting a marine CRM agenda. Most of marine environment and coastal zone metadata to the of the petroleum industry 2D seismic survey data shown in crown, the government, industry and academia. They sell Figure 2.4 belong to BP Exploration. Petroleum compa- a range of GIS data layers relating to near-coastal land, nies are often amenable to cooperating with academic re- backshore, shoreline and near-shore environments. quests for study in the spirit of the DTI Guidelines for the Release of Proprietary Seismic Data UKCS, 2004. Indi- Archaeological artefacts recovered offshore constitute vidual data release agreements must be drafted, signed and archaeological ‘ground truthing’ of our palaeolandscape countersigned by both parties. These data are relatively models, with due acknowledgement of the frequent lack of old (early 1970s to early 1980s) and so, in the absence of an accurate provenance for such finds or indeed the accu- active exploration within the Bristol Channel, may be racy of associated positional data. There are two sources readily available. of such records. Historic Environment Records (HERs) are maintained by local authorities. The National Monu- The BGS has recently been involved in an Aggregates ments Record (NMR) is maintained by EH. Only a few Levy Sustainability Fund (ALSF) project investigating coastal HERs incorporate a marine section, and given the seabed ecology in the Outer Bristol Channel. Part of the lack of secure provenance for many artefacts this must project involved the acquisition of 2D seismic sub-bottom constitute a weakness. EH hope that all will feature a ma- profiles (Figure 2.4d). It is thought that these profile data rine aspect eventually (Roberts and Trow, 2002). The have not been interpreted with respect to seabed archae- marine HERs and NMR will form an essential tool in the ology. They would provide an ideal dataset traversing management and protection of the marine archaeological prehistoric coastlines since the Late Palaeolithic (Figure 6; resource (Robert and Trow, 2002). To this end, EH has Bell, unpublished; Tetlow, 2005). collaborated with ‘legitimate users of the sea’ to draft the Protocol for reporting finds of archaeological interest A marine aggregates license was awarded within Area 1 (2005). This protocol should ensure that the location of within the last 12 months. Recent survey data exists for artefacts dredged or caught up in fishing nets is recorded this but is still commercially sensitive. Older data exists as part of the cataloguing process. for the other two licensed areas but is often still pertinent to applications for licence extension. Access to these de- 2.6 Case study assessments of survey pends on the agreement of the aggregates producers in- volved. In the case of the Nash Bank licence, this would data involve the agreement of all three operators. The same is true of the environmental and engineering surveys under- As part of the project data audit, which forms the basis of taken prior to approval of the Scarweather Sands wind this paper, three areas adjacent to the English coastline farm. were chosen for a pilot study to assess useful survey data that have been acquired within them (Figure 2.3). These The University of Wales in Bangor (UWB) part owns a areas are geographically distinct and offer different ap- survey vessel capable of being used for 2D seismic sur- peals to offshore developers. Two are adjacent to areas veying, sonar mapping and sediment sampling. They have containing significant archaeological sites, the other lies been involved in surveying Carmarthen Bay in the north- just beyond the fringe of the area mapped by the NSPP. western part of Area 1. They hold digital data for all sur- veys conducted since the mid to late nineties. However, 2.6.1 Area 1: the Bristol Channel staff resources limit responses to requests for data. The NMR of England contains a single record of a Mesolithic The Bristol Channel has the second largest tidal range in artefact having been recovered from within Area 1 (Figure the world, 15m during spring tides. This drives a dynamic 2.4d). Dawkins and Winwood reported a number of sediment transport system capable of filling artificial scour 16  Mapping Doggerland Mesolithic blades and flakes, and Neolithic flints and Somerset coast in August 1869 (NMR Number SS 94 NE scrapers, just offshore from Minehead on the Northwest 11, Record 36754). Figure 2.4 Area1 a) Industrial threats to seabed archaeology within the Bristol Channel b) Distribution of previous BGS sampling and 2D surveys c) UKOOA surveys d) Archaeology and previous ALSF survey conducted by the BGS within Area 1. 17  Mapping Doggerland Figure 2.5 Area 2. Distribution of BGS and UKOOA surveys 18  Mapping Doggerland Figure 2.6 Area 3 a) Industrial threats to archaeology b) Distribution of BGS sediment and 2D surveys c) UKOOA 2 and 3D surveys. 19  Mapping Doggerland Figure 2.7 Intense use of space within the Southern North Sea 20  Mapping Doggerland 2.6.2 Area 2: Portland These are part of a group of four surveys fringing the PGS MegaSurvey and belong to WesternGeco. Three are ‘un- Area 2 has not proven attractive to the petroleum or ma- released speculative’ survey data that would be expensive rine aggregates industries. In 2001 UNESCO inscribed to obtain. The fourth is a ‘released proprietary’ survey the coastline alongside Area 2 as part of a World Heritage originally owned by BP Exploration. This should be rela- Site – the British Mesozoic Coast. Area 2 is traversed tively straightforward to obtain for academic research. only by 2D geophysics survey lines acquired by the BGS and the petroleum industry. It also contains a few BGS Area 3 contains four zones licensed for shallow marine shallow sediment core/borehole sample stations and a sin- aggregates dredging, each licensed to a different marine gle exploratory well bore. Figure 2.5 summarises these aggregates producer. All underwent the same licence ap- survey data. plication process despite the fact that one zone, licence 440, lies beyond the 12-mile territorial limit. Technical The petroleum industry 2D seismic survey data within environmental statements for these zones are not available Area 2 belongs to many companies and was shot between in electronic form. Any other data that has been retained the early 1970s and the early 1990s. Development did not would prove useful in future licence renewal applications follow, although the Maersk Oil North Sea UK Ltd sank and are likely to be subject to the constraints of commer- well 97/12-1 in late 1995, following extensive surveying cial sensitivity. including side scan sonar high resolution 2D sub-bottom seismic profiling and vibrocoring, carried out by Kerr The Greater Wash area has become a major focus for the McGee Oil (UK). second generation of offshore wind farm projects. One of two “super” farms currently under consideration by the Three coastal sites are where prehistoric peat and wildlife DTI and DEFRA, Triton Knoll, falls within Area 3. This have been recorded adjacent to Area 2. However, the Eng- and the other smaller project, Race Bank, are currently at a lish NMR and Dorset HER contain no record of Meso- late stage of the DTI consents process. Survey data is lithic finds recovered offshore. likely to remain commercially sensitive until 2010. 2.6.3 Area 3: The Spurn As area 3 is adjacent to the NSPP study area it is useful to consider a larger area in respect of marine infrastructure. Area 3 is situated north west of the Wash and is of particu- This information is provided in Figure 2.7. Study of the lar importance in that it impinges on the NSPP study area. data reveals the intensity of industrial use of the area. This has been an important area for oil and gas explora- Query of the NMR of England, the HERs of Lincolnshire tion, fishing and shipping and is traversed by gas and and Norfolk and the literature revealed there have not been chemical pipelines. Recently, it has become a focal point any Mesolithic artefacts recovered and catalogued within for marine aggregates production and offshore wind farm Area 3. However, a wider, regional view of these com- development. Area 3 occupies part of the “white band” bined data suggests that the SNS is a relative hotbed for between the current extent of 3D seismic data and the offshore finds by UK standards. However, Figure 2.7 also shore (Terrell et al., 2005). However, the area has been suggests a general absence of finds within the region intensively investigated by other forms of survey primarily where there is a high concentration of petroleum industry aimed at characterising the shallow seabed sediments. infrastructure. This may well be the result of exclusion zones around rigs and wellheads that prohibit deep fishing Figure 2.6b and c illustrate the distribution of BGS and and dredging: activities often responsible for disrupting UKOOA survey data within Area 3. A request, dictated prehistoric archaeological deposits (Wenban-Smith, 2002). by data availability, was made for a strategic selection of Such deposits may therefore have actually been preserved seismic lines and shallow sediment core records from the by the presence of this infrastructure. On the other hand, BGS. The BGS was able to supply these as a selection of the absence of a unified database of finds supported by image files of scanned paper seismic lines and geological sediment or geomorphological data, as provided by the reports. Several items from the original enquiry were un- NSPP, is highlighted when one considers that the route of available. Geological and survey reports for the gas pipe- the new Langeled gas pipeline from Norway passed across line featured in Figure 2.6a were available from the current part of the project study area and, possibly, through a site owner. where moorlog (fossilised peat) was recorded (Kooijmans, 1971, Figure 2.6). The petroleum industry 2D seismic surveys belong to many different companies. They were shot between the 2.7 Conclusions mid 1960s and the early 1990s. Continuing industrial re- generation of this area lends these data considerable ongo- The requirement for a greater understanding between off- ing commercial value and, consequently, some may be less shore developers and marine conservationists is becoming easy to obtain. more acute as the UK makes greater use of its marine re- sources in ever more diverse ways. Central to the effec- Half of Area 3 has been investigated by petroleum indus- tive management of the seabed is collaboration between try 3D seismic surveys (Figure 2.6c and Figure 2.7). developers and researchers that aims to mitigate our im- 21  Mapping Doggerland pact on the environment. The archaeological resource The final point to be made is central to the use of existing contained within the Holocene sediments of the seabed resources. It is acknowledged that geographic distribution remains poorly understood. This is primarily due to the of acquired survey data has been led by the location of practical limitations of carrying out archaeological investi- natural resources and location-specific legislation. How- gations. However, the NSPP has revealed dramatic in- ever, the physical location of resources and related data- sights into the nature of the emergent prehistoric landscape sets is equally disparate. Many survey datasets are distrib- around our present-day coast and these insights were uted amongst the archives of the BGS and individual off- gained using existing exploration survey data. The latter shore operators. Some data are held within academia but point is an important one and is emphasised by the extent these are often bound by the strictures of specific legal and variety of available information revealed as part of agreements linked to a specific activity. Recent govern- this data audit. It should now be clear that much substan- ment policy is opening up the flow of data between indus- tive information on the nature of the marine resource could try, academia and the public. However, it remains for re- be recovered through the use of the existing data. In some searchers to prove the usefulness of their work for industry cases the strategic collection of new data may be required, and regulators and to assist in defining the threat that in- where gaps have been identified or on those occasions dustrial activities pose to the marine environment. In so where further information is required for specific archaeo- doing, researchers can help industry meet the demands of logical purposes. an increasingly rigorous regulatory framework. 22  Mapping Doggerland 3 3D Seismic Reflection Data, Associated Technologies and the Development of the Project Methodology. Kenneth Thomson and Vincent Gaffney 3.1 Introduction seismic reflection data types and how these considerations influenced the project methodology. A variety of methods and datasets were potentially avail- able for use in the project. However, as the choice of 3.2 Seismic reflection method and methods, and hence data types, controls the volume and resolution quality of the results an optimal approach needed to be developed. A crucial consideration was the need to mini- Seismic reflection surveying involves the transmission of mise the time involved in the analysis, and hence the ex- acoustic energy into the subsurface and recording the en- pense of the project, whilst at the same time maximising ergy reflected from acoustic impendence contrasts. The the spatial coverage and detail. Consequently, the avail- reflections produced at acoustic impendence contrasts are able technologies, the costs involved in acquiring new data predominantly the product of changes in lithology with the and the possibilities for using existing data needed to be impendence contrast, or reflection coefficient (the ratio of evaluated when planning the project. amplitude of the reflected wave to the incident wave, or how much energy is reflected.), given by the equation: The Southern North Sea (SNS) contains an extensive col- lection of data including seabed samples, shallow core, R = (ρ2V2 - ρ1V1)/ (ρ2V2 + ρ1V1) bathymetry data and seismic reflection profiles collected for the investigation of near seabed features or deeper hy- Where: R = reflection coefficient drocarbon exploration. This suggested that the project ρ1 = density of medium 1 could potentially achieve its aims by exploiting existing ρ2 = density of medium 2 data with future acquisition of bespoke datasets being con- V1 = velocity of medium 1 sidered on the basis of the project's results. The existing V2 = velocity of medium 2 datasets from the SNS were acquired for a variety of pur- poses and consequently have differing strengths and With appropriate processing this allows the production of weaknesses. Seabed samples and shallow coring can pro- pseudo-depth sections of the subsurface structure with the vide chronological, sedimentological and environmental vertical axis being two-way travel time to the reflector. data. However, such data provides a poor spatial frame- work. Although the basics of this technique are common, the details vary for a range of applications including the inves- Although high resolution bathymetry data can provide tigation of deep crustal structure (Klemperer and Hobbs, excellent images of the seabed topography, and hence de- 1991), hydrocarbon exploration (Bally, 1987) and near tailed images of early Holocene features that have a seabed sediment structure (e.g. Salomonsen and Jensen, bathymetric expression, many of the important geomor- 1994; Velegrakis and Dix, 1999; Praeg, 2003 and Bulat, phological features are, at least, partially buried in the 2005). These diverse applications dictate different acqui- SNS. Consequently, there was a need for regionally ex- sition parameters that in turn determine the resolution and tensive datasets that have the capability to image below depth of penetration of the survey as well as the costs in- the seabed. The only existing data within the SNS that volved in acquiring the data. Consequently, the relative could meet these requirements were seismic reflection merits of a range of available seismic reflection data types datasets. Use of such surveys could permit the generation needs to be assessed when considering the investigation of of regional maps for buried Early Holocene landscape fea- submerged, and partially buried Holocene features. tures. These datasets would then provide the framework into which data from shallow boreholes, seabed samples Standard marine acquisition involves towing an energy and bathymetry could be integrated. source and a cable (streamer) containing pressure sensitive receivers to record the reflections from the underlying Marine seismic acquisition is undertaken for a variety of strata (Figure 3.1). In single fold data, only one reflection purposes, with varying data densities, coverage, depths of is received from any point in the subsurface. However, penetration and resolution. Consequently, there was a many seismic profiles are multi-fold. In this case several choice between differing seismic reflection data types, shot-receiver pairs are of the correct geometry to collect each being acquired for specific purposes that may not acoustic energy reflected from the same point. These re- have been compatible with the project requirements. This flections can then be summed in order to increase the sig- paper will discuss the critical parameters for differing nal-to-noise ratio of the seismic profile. 23  Mapping Doggerland Figure 3.1 Typical marine seismic reflection acquisition. The vessel travels through the water and regularly fires the seismic source. The sound wave travels through the water column (and underlying sediments) and is partially reflected at acoustic impendence contrasts. Receivers within the long towed cable astern of the vessel detect the reflected wave, which is then transmitted to the vessel and recorded. The characteristics of the seismic energy source are dampening effect of the top few hundred metres of intimately linked to the required resolution and depth of overburden is relatively small and consequently seismic penetration. The vertical resolution of seismic reflection sources with frequencies in excess of 100Hz can be data requires a minimum separation between two employed. In contrast, 2D and 3D seismic data acquired interfaces that will give rise to two separate reflections. At for hydrocarbon exploration need to image to depths of separations of less than ¼λ the reflections from the two several kilometres and consequently employ sources with acoustic impedance contrasts constructively interfere with frequencies of less than 100Hz. This, combined with the maximum amplitude occurring at ¼λ, known as the increasing velocity with depth, results in a significantly tuning thickness (Figure 3.2). However, it is not until ½λ higher vertical resolution for 2D seismic data specifically that the two reflections are separable (Figure 3.2). acquired for the investigation of shallow geology (<1km) Consequently, the vertical resolution of seismic reflection compared to standard 2D or 3D seismic data required for data can be defined as either the minimum resolvable (¼λ; hydrocarbon exploration. This is demonstrated in Figure Figure 3.3) or the minimum separable (½λ; Figure 3.3). 3.4 where a high frequency 2D seismic line specifically designed to image Holocene and Pleistocene features is As the vertical resolution is dependent upon the compared with a line from exactly the same position wavelength it is therefore dependent on the velocity of the extracted from a 3D seismic dataset acquired for medium and the frequency of the seismic source/reflected hydrocarbon exploration. The high-resolution line (Figure wave. Ideally, a high frequency source (>100Hz) would 3.4a) shows a channel and its complex infill pattern. In be used in all circumstances. However, as the geology contrast, the low-resolution 3D seismic line (Figure 3.4b) progressively dampens high frequency seismic signals is unable to image the channel. with depth, the seismic source needs to be chosen with consideration to the required depth of penetration. The Figure 3.2 Seismic resolution of a layer of varying thickness. The blue lines show the top and base of a layer. Using a 400Hz Ricker wavelet, and assuming a sediment velocity of 1600ms-1, the reflections from the top and base can be shown to constructively interfere at thicknesses less than ¼λ (1 metre) with the maximum amplitude at ¼λ. Furthermore, the top and the base of the layer do not perfectly align with peak and troughs. At thicknesses of ½λ (2 metres) and greater the top and the base of the layer are separable with peak and troughs aligning perfectly with the position of the top and base respectively. 24  Mapping Doggerland Figure 3.3 Plots of seismic resolution as a function of burial depth and frequency. (a) The minimum resolvable vertical resolution. (b) The minimum separable vertical resolution. (c) The horizontal resolution of unmigrated seismic data. (d) The assumed velocity-depth structure. The lateral resolution of seismic reflection data is higher lateral resolution compared to lower frequency dependent on the fact that seismic energy travels data collected for hydrocarbon exploration (Figure 3.3). through the subsurface and encounters the reflecting surfaces over discrete areas. The energy travels as wave However, the procedure of migrating seismic data, fronts and the region on the reflector where the seismic which ensures reflected energy is correctly positioned energy is reflected constructively is known as the within the subsurface, considerably enhances resolution. Fresnel Zone (Sherrif, 1977). Lateral resolution is Consequently, for migrated data, lateral resolution determined by the radius of the Fresnel Zone, which depends on trace spacing, the length of the migration itself depends on the wavelength of the acoustic pulse operator, time/depth of the reflector and the bandwidth and the depth of the reflector. Thus, in non-migrated of the data. If completely successful then the lateral seismic data, lateral resolution is dependent on the resolution of the high frequency seismic section shown frequency of the seismic source, the interval velocity in Figure 3.4a would be approximately 12m compared and on the travel time to the reflector. As with the to 50m for the low-resolution 3D seismic line (Figure vertical resolution, this implies that high frequency 3.4b). seismic reflection data will provide a significantly 25  Mapping Doggerland Figure 3.4 A comparison between (a) high frequency 2D seismic reflection line and (b) low frequency 3D seismic line from the same location. Note that higher frequencies yield greater vertical detail. 26  Mapping Doggerland 3.3 2D versus 3D seismic acquisition This method of acquisition has two main disadvantages. Firstly, the reflected seismic energy is assumed to have and interpretation originated from a point directly beneath the profile even though it could have originated from a point laterally offset Traditional seismic reflection data is generally referred from the profile. This aliasing means that the location of a to as 2D as it is acquired as a series of discrete vertical feature cannot be accurately constrained, as the spacing be- profiles using a single streamer towed behind the vessel. tween lines is too wide to correct this error. Secondly, the This acquisition pattern results in the collection of sev- spacing between lines is sufficiently wide that it can be diffi- eral profiles with the spacing between profiles being cult to map the position of a morphological feature across the several orders of magnitude greater than the trace spac- region of interest. For example, Figure 3.5 demonstrates ing (i.e. the horizontal sampling interval along the pro- how wide line spacing can lead to several equally valid in- file). terpretations. Figure 3.5 (a-d) Four possible interpretations of a channel morphology based on a coarse 2D seismic grid. Each interpretation is equally valid. (e-h) Schematic illustrations of how each of the interpretations shown in a-d would appear on a timeslice from a laterally continuous, binned 3D seismic volume. This demonstrates that 3D seismic data has the potential to distinguish between the possible alternatives. Figure 3.6 Typical 3D marine seismic reflection acquisition. The vessel travels through the water and regularly fires the seismic source. The sound wave travels through the water column (and underlying sediments) and is partially reflected at acoustic impendence contrasts. Receivers within the towed cables astern of the vessel detect the reflected wave, which is then transmitted to the vessel and recorded. In contrast to Figure 3.1, 3D acquisition involved multiple towed cables with each cable being capable of producing a seismic profile. 27  Mapping Doggerland In contrast, 3D reflection seismic data involves the towing directions and estimates of three-dimensional seismic co- of multiple streamers (Figure 3.6), which allows the rapid herence are obtained. Small regions within the seismic collection of multiple closely spaced lines. The survey volume containing stratigraphic anomalies such as chan- configuration provides significant advantages as it gener- nels have a different seismic character compared to the ally involves multifold collection, which is containing corresponding regions of neighbouring traces. This results several reflections from the same subsurface location, and in a sharp discontinuity in local trace-to-trace coherence allows reflected energy to be correctly positioned in space. and allows the rapid identification of stratigraphic features Thus it eliminates the potential positioning errors associ- (Figure 3.8; Bahorich and Farmer, 1995). ated with 2D seismic data. This is achieved through the conversion into a binned dataset with, in the case of data Another advance in 3D seismic interpretation has been the acquired for hydrocarbon exploration, a bin spacing of development of opacity rendering techniques (Kidd 1999). 12.5m x 12.5m x 4 milliseconds, or multiples thereof. The technique converts conventional 3D seismic data into a voxel volume. Each voxel contains information from the Each bin is then populated by those reflections that origi- original portion of the 3D seismic volume that it occupies nated from within the bin space. Once binned the data together with an additional user-defined variable that con- format provides additional benefits. Firstly a laterally con- trols its opacity. The opacity of individual voxels can then tinuous binned data volume means that a geomorphologi- be varied as a function of their seismic amplitude (or any cal feature can be mapped from bin to bin, removing the other seismic attribute), allowing the user to examine only potential errors involved in the interpretation of 2D data those voxels that fall within the particular amplitude (or (Figure 3.5 and Figure 3.7). However, 3D seismic is also attribute) range of interest. By using appropriate opacity more versatile than 2D data as it can be interrogated in a filters it is possible to image the depositional systems such number of ways. Instead of relying on vertical profiles, as buried fluvial channels. This exploits seismic charac- the volume can be sliced in any direction. Of particular teristics, which are in part lithologically dependent, and importance to the investigation of relatively shallow, and different from the surrounding materials, thus permitting flat, Holocene features is the ability to produce a horizon- the surrounding rock to be made transparent whilst pre- tal slice (timeslice) through the data as this can, in many serving all but the smallest channels as opaque features cases, be interpreted as a geological map showing a range (Figure 3.8; Fitch et al., 2005). of sedimentary features and facies patterns (Figure 3.8; Fitch et al., 2005). 3.4 Interpretation strategy for the The interpretation of 3D seismic data has improved sig- Southern North Sea nificantly in recent years due to the development of a range of new techniques originally designed to improve The above discussion demonstrates that the ideal dataset geological interpretation associated with hydrocarbon ex- for the investigation of submerged Holocene/Mesolithic ploration and production. Once a stratigraphic marker of landscapes within the region would be high resolution interest has been identified, it can be mapped across the (>100Hz) 3D seismic data with appropriate borehole con- 3D seismic volume to produce a horizon. This can then be trol. Such a dataset would provide high (metre or less) examined using a variety of attributes (e.g. depth, seismic vertical and lateral resolution and a laterally continuous amplitude, dip, azimuth) with each attribute having the coverage, thus removing the need to interpret the location potential to reveal different characteristics of the feature of of a feature between data points. It would also support the interest (Posamentier, 2005). However, simply applying application of the latest advances in seismic attribute artificial illumination from a number of directions can analysis and visualisation to aid interpretation Unfortu- prove highly effective at identifying subtle geomor- nately, high-resolution 3D seismic surveying equipment, phological features. such as the 3D CHIRP system developed by the National Oceanography Centre, Southampton (Gutowski, 2005), is Seismic attributes can also play a crucial role in the inter- a recent development and consequently such data is ex- pretation of 3D seismic data. Extracting RMS amplitude tremely rare. Furthermore, such systems currently utilise (root-mean-square) over a two-way time interval, defined small vessels that are not suitable for deployment beyond by two mapped horizons or two timeslices, is commonly the immediate coastal waters. However, it is the high employed to differentiate zones of different seismic ampli- resolution of the system that is the most significant handi- tude within a seismic volume (e.g. Den Hartog Jager et al., cap. High-resolution seismic acquisition involves much 1993). As seismic amplitude is a function of density slower surveying rates and thus higher costs. For exam- and/or velocity contrasts it is often closely related to the ple, the 3D CHIRP system described by Gutowski (2005) depositional facies (Figure 3.8). Further enhancement of is capable of surveying approximately 0.02km2 per day. the seismic slices, both vertical and horizontal as well as This contrasts with the lower resolution 3D seismic data mapped horizons, can be achieved through the generation acquired for hydrocarbon exploration which, although it of a coherence (or semblance) seismic volume. The co- involves large, expensive and custom-built vessels, can herence cube (Bahorich and Farmer, 1995) calculates lo- survey 40km2 a day at a cost of c. $5000 per square kilo- calised waveform similarity in both inline and crossline metre (Bacon et al., 2003). 28  Mapping Doggerland Figure 3.7 (a) Geological map of Poole and Christchurch bays with the artificially illuminated plan view of the seabed reflector. (b) Artificially illuminated plan view of the seabed reflector. (c) Artificially illuminated plan view of the seabed reflector with palaeo-channels as mapped using 3D seismic. (d) Artificially illuminated plan view of the seabed reflector with palaeo-channels as mapped using 2D seismic data (Velegrakis et al. 1999). Note the poor correlation between the 2D interpretation and the actual channel locations. Figure 3.8 A comparison of seismic images from the Dogger Bank produced using different techniques. (a) Simple seismic amplitude timeslice revealing some Holocene depositional features (channels). (b) RMS amplitude extraction for the top 200ms of the data. (c) Opacity rendering of the same seismic volume as (b). Note that both the RMS extraction and the manipulation of the opacity allow some improvement in channel definition. (d) Seismic coherence (semblance) timeslice at the same level as (a). Note that coherence provides some additional detail of the channels. 29  Mapping Doggerland Although high-resolution 3D seismic data is expensive to opportunity to rapidly develop a regional framework into acquire and not readily available it is interesting to con- which other, higher resolution datasets, could subsequently sider the advantages of using existing 3D seismic data be integrated. The approach was: acquired for hydrocarbon exploration. The frequency spectra for the top 200ms (the likely depth/two-way time 1. To map regionally significant reflectors using the range of interest) of the dataset used in this study has regional 3D seismic dataset. 98.7% of the frequency content in the 3-72Hz range with a 2. To interpret these surfaces using artificial illumi- mean frequency of 14.7Hz. Consequently, a mean fre- nation and horizon attributes such as amplitude quency of 14.7Hz provides a vertical resolution of 27m and dip to identify morphological features and the although the higher frequency components suggest that a developmental chronology. vertical resolution of 10m or less may be possible. The 3. To generate seismic attributes for the regional 3D limit of horizontal resolution for unmigrated seismic data, seismic dataset. the Fresnel Zone (Sherrif, 1977), for the mean frequency 4. To sequentially timeslice these attribute volumes of 14.7Hz would provide a Freznel Zone width of 66m. (e.g. amplitude, coherence, RMS amplitude) and However, this can be considered an extremely conserva- to employ opacity rendering techniques to iden- tive estimate as the higher frequency components suggest tify morphological features and the developmen- lateral resolutions of around 30m may be possible (Emery tal chronology. and Myers, 1996). In addition, as migration of the seismic 5. Integrate the above to develop a first order geo- data considerably enhances lateral resolution with the limit morphic model. being dependent upon trace spacing, length of the migra- 6. Use existing high resolution 2D seismic data and tion operator and the bandwidth of the data. Conse- shallow borehole data to refine the geomorphic quently, the lateral resolution of the top 200ms of the mi- model, resolve interpretational and chronological grated dataset used in this study may actually approach the ambiguities, and to provide palaeoenvironmental line spacing of 50m. data. Although these parameters may suggest that the commer- This strategy provided several advantages. Firstly, it op- cial 3D seismic datasets acquired for hydrocarbon explora- timised the use of existing data thus reducing the cost of tion are not suitable for the exploration of Holo- the project. More importantly, the speed of the project cene/Mesolithic landscapes in the Southern North Sea the was significantly increased as the 3D seismic permitted the reverse is actually true. For example, Figure 3.7 is an arti- rapid development of a regional model and the identifica- ficially illuminated map of the seabed reflector from Poole tion of key localities for detailed work. These could be and Christchurch bays in the English Channel and mapped identified on the basis of the need for clarification or a using commercial 3D seismic data from over the Wytch recognition of their environmental importance. Farm oilfield. The map shows a number of north-south trending channels that do not possess a current bathymetric 3.5 Conclusions expression. Instead, Velegrakis et al. (1999) demon- strated, using 2D seismic data, that these channels are The key requirement for understanding the Mesolithic completely infilled and are sub-seabed features. Conse- archaeological potential of the Southern North Sea is the quently, Figure 3.7 demonstrates that the vertical resolu- development of a detailed regional landscape model. tion of approximately 10m results in a mean response from Given the vast area under consideration the traditional both the seabed and the acoustic impedance contrasts from archaeological approach would suggest a requirement for several metres below it (c.f. Bulat, 2005). This implies the acquisition of a large bespoke geophysical survey of that mapping the seabed, or any other near seabed reflec- the area, with appropriate stratigraphic, sedimentological tor, can provide maps containing information from several and environmental controls from cores. The expense, and metres of Holocene strata. Furthermore, a timeslice can logistical complexity, involved in such a survey would be also be considered to provide information from a strati- prohibitive. Consequently, there existed a need to develop graphic interval several metres thick. Given a bin spacing a regionally extensive and detailed landscape model utilis- of 50m, and an areal coverage of >20,000km3, this sug- ing existing data. Fundamental to developing such a gests that timeslicing and mapping regionally significant model would be access to regionally extensive seismic reflectors has the potential to provide an extensive recon- reflection datasets. However, given the range of uses for naissance tool for the investigation of submerged, Holo- which data are acquired, the acquisition parameters vary cene landscapes (Figure 3.8). considerably and hence have varying degrees of applica- bility to the study of Early Holocene landscapes. Tradi- The above considerations therefore suggest that an alterna- tional 2D high frequency, high resolution seismic reflec- tive interpretation strategy could be employed for the in- tion data have been the preferred dataset for the investiga- vestigation of Holocene/Mesolithic landscapes beneath the tion of Holocene geology and archaeology. Unfortunately, Southern North Sea. This approach was completely de- such datasets are limited in their use as regional mapping pendent on the donation of >23,000km3 of commercial 3D tools as they are prone to spatial aliasing errors and require seismic data by PGS Reservoir. This would provide an extrapolation of geomorphic features between relatively 30  Mapping Doggerland widely spaced data points. Conversely, 3D seismic data analysis and seismic visualisation to be applied. These acquired for hydrocarbon exploration have significantly advantages permit the rapid development of a regional lower resolution but provide a relatively complete spatial geomorphic model into which higher resolution seismic coverage. This diminishes the aliasing issues and also reflection data, bathymetry, seabed samples and shallow allows a range of techniques such as timeslicing, attribute core can be integrated. 31 Mapping Doggerland 32  Mapping Doggerland 4 Merging Technologies: Integration and Visualisation of Spatial Data Simon Fitch, Vincent Gaffney and Kenneth Thomson 4.1 Introduction sets used by the project and to provide access to high-end stereo-projection systems to visualise and to quality check The primary goal of the North Sea Palaeolandscapes Pro- the data as it was processed and interpreted. Given the ject (NSPP) was to explore the Late Pleistocene and Holo- nature of the project and its significant technical demands cene landscapes of the Southern North Sea through the use some description of the technical context of the project is of c. 23,000 km2 of contiguous 3D seismic data provided required. Technical management of the project ran through by PGS UK (www.PGS.com). Processing this large vol- the HP Visual and Spatial Technology Centre (HP ume of data required considerable investment in both VISTA). HP VISTA is a division of Birmingham Archae- hardware and software. In line with most archaeological ology and is situated within the Institute of Archaeology projects utilising large amounts of digital, remote sensed and Antiquity at the University of Birmingham. The labo- data, there was a requirement to utilise specialist softwares ratory was developed, exceptionally, for the visualisation for the processing of primary seismic datasets and standard of large archaeological data sets and the facilities provided geographic information systems to manage, manipulate are, currently, unique within British archaeology. A sche- and display supporting data and interpretative layers matic of the Centre's specialist infrastructure, as utilised by (Gaffney and Gater, 2003 Chapter 5; Chapman 2006). the NSPP, is provided in Figure 4.1. What is, perhaps, less usual within archaeology has been the requirement to provide access to high bandwidth net- works and storage to cope with the large volumetric data- Figure 4.1 HP VISTA infrastructure diagram 33  Mapping Doggerland 4.2 Infrastructure level of performance, image quality and resolution inde- pendently, so a large model can be rapidly manipulated at The primary engines for the majority of analysis were a lower quality/higher performance setting and then a par- three HP xw8200 workstations with 64bit dual processors, ticular region of interest examined at a high quality/lower 16Gb of RAM, 2Tb of local storage and high-end graphics performance setting, without having to interrupt the visu- cards. Whilst the majority of primary processing was car- alisation. The SV7 supported projection across a 4.27m ried out on these machines, specialist softwares used to by 1.8m dual channel, rear projection Fakespace Power- carry out attribute analysis on the seismic data, for exam- Wall. This provided a geometrically accurate stereoscopic ple, were resident on machines elsewhere on the campus. display using active stereo glasses (Figure 4.2). This Project data was held centrally on a secure 5Tb NAS stor- combination of analytical power and sophistication of dis- age system resident within the University Information play, rare within archaeology, was demanded by the nature Services' machine room. The need to move data between of the datasets under investigation and is comparable with these machines and to the central storage was supported by specialist facilities dealing directly with petroleum geol- gigabit-enabled cabling between most project computers, ogy and remote sensing data sets. To put this in context, a and assisted by the availability of a dedicated fibre optic paper published 10 years ago and also concerned with the link between the NAS and the main display centre in the issues of integrating GIS and other spatial technologies HP VISTA centre. The extent and complexity of the within archaeological projects at Birmingham, records that volumetric data, and the utilisation of solid modelling and the entire archaeological computing group was then ser- opacity rendering techniques, required that the data should viced by a single server with 64 megabytes of RAM and 4 be viewed stereoscopically and also that displays were gigabytes of mass storage (Buteux et al. 1997, Gaffney large enough to provide the opportunity of group viewing. and Gaffney 2000a). The NSPP's computing require- This was provided within the HP VISTA centre via an HP ments, and its demand for exponentially increased storage SV7 scaleable visualisation system. This is a multi-pipe capacity, presumably indicates the scale of change that can system providing high-end, high quality visualisation. A be anticipated in future, comparable digital landscape pro- particular benefit of the system is the ability to set the jects. Figure 4.2 Inspecting data at the Visual and Spatial Technology Centre (HP VISTA) 34  Mapping Doggerland 4.3 Software integration based geophysical surveys (e.g. magnetometry and resis- tivity) as well as remotely sensed imagery is also reasona- Having acquired the infrastructure needed to implement bly well established (Gaffney et al. 2000). The planar na- the project it is appropriate to consider the requirement for ture of these traditional land based geophysical surveys software integration that arose during the course of the facilitates their easy integration into a traditional GIS's study. The increasing application of geographic informa- map style interface. However, the representation of true tion systems to manage, manipulate and display spatially three dimensional volume data, including 3D seismics, is a referenced data sets within archaeology has been a major challenge to systems that are, essentially, 2D (Kvamme trend over the past decade (Chapman 2006, Connolly and 2006, Watters 2006). Lake 2006). Not surprisingly, the majority of the NSPP's interpretative and supporting spatial data has been held Whilst the representation of a third dimension is possible within an ARCGIS database made available to all team within certain GIS viewers, such applications are not ideal members. A similar trend is discernable within other for the purposes of representing volumetric geophysical comparable groups concerned with hydrocarbon explora- data, as these do not allow for the representation of voxels. tion. The ability of GIS technology to handle a variety of Such systems are therefore unable to adequately display a spatial data, in conjunction with its analytical capacity, volumetric representation of a cube of seismic 3D data. make GIS an invaluable tool for petroleum exploration Within the NSPP there was, therefore, always a require- (Gaddy 2003). Further, the ability of a GIS to visualise ment to explore the means by which volume data could be and manage data throughout a project's life has also integrated with standard GIS map layers in a manner that proved invaluable to the petroleum industry (c.f. Lawley retained the integrity of the original volumetric data. and Booth, 2004). 4.4 Primary integration procedures Despite this promising situation, the size of the project database and the incorporation of volumetric data sources Whilst proprietary tools do exist to display the products of proved problematic during the course of the research. It is the analysis of seismic data within a GIS, this has gener- notable that the application of GIS systems within the pe- ally been achieved through the rectification of a flat image troleum industry, in contrast to archaeology, has run in from which interpretation can be undertaken. With respect parallel with the development of 3D modelling systems to seismic surveys, integration of a cube of data can be and remote sensing packages that link interpretive and achieved thought the export of serial planar, timeslices. geophysical data (Gaddy 2003, 1). The requirements of This facilitates interpretation alongside GIS layers in a the petroleum industry have produced highly flexible in- traditional fashion (Goodman and Nishimura, 2000). terpretation packages and softwares, such as Tigress and However, for the NSPP it would not be desirable to intro- SMT Kingdom, developed to image vast, complex geo- duce all possible time slices into the GIS. Indeed, to dis- physical datasets and to facilitate the mapping, manage- play the number of seismic attributes used in this project ment and planning of petroleum data within an easily would result in tens of thousands of slices and associated visualised environment. Industry requirements are there- data layers. This would add unnecessary complexity to the fore very similar to those of the NSPP (see Thomson and database and vastly increase the amount of data to be ma- Gaffney, this volume), and the utilisation of these devel- nipulated and stored. Indeed, management of such a vol- oped and reliable packages was highly desirable in the ume of data would probably be beyond the present capac- context of the analysis carried out as part of this project. ity of most GISs and, probably, incomprehensible to a human operator (Kvamme 2006). Within marine archaeology, the use of GISs to assist in archaeological management is fundamental (Groom and Consequently, in the first instance a selection of slices Oxley 2001, 56). Distributional analysis of marine and from the most commonly used data attribute amplitudes associated resources, or analysis of absence of evidence, were selected to provide an overview of the dataset. This permits targeted use of resources in curatorial terms and selection, however, does leave open the possibility that provides a greater insight into the structure of the marine significant information may be missed between slices. To database (Groom and Oxley 2001; Allen and Gardiner compensate for this the NSPP utilised RMS amplitude 2000, Fitch et al. 2005, 194). The utilisation of geophysi- slices (root mean squared, see Thomson and Gaffney, this cal information in this role within a GIS is also appreciated volume) to facilitate the display information from the by the archaeological computing community (e.g. Buteux missing volume within a standard 2D image. The resultant et al. 2000, Gaffney et al. 2000b). This has proved invalu- slice therefore shows areas of anomalous seismic ampli- able when monitoring landscapes that contain poorly un- tudes within the selected volume, and can be useful in im- derstood archaeological resources (Chapman et al. 2001). aging channels (see Figure 4.3). The resultant output is a In projects, such as the NSPP, where archaeological sur- planar slice, admirably suited for the integration into a vey and interpretation is severely limited by the prevailing standard GIS system, and satisfies some of the require- physical environment, the ability to combine data sources ments of Kvamme (2006) for displaying this data type. including geophysics, physical samples and findspots to Furthermore, the method of generation means that it is provide a proxy environment for interpretation is crucial. highly accurate in terms of spatial location, and is only The use of GIS to explicitly integrate traditional land limited by the properties of the original seismic survey. 35  Mapping Doggerland Thus such an approach is compatible with standard proc- prospection (see Watters (2006) for a comparable process essing and display methodologies used in archaeological used to process GPR data). Figure 4.3 Seismic data slice illustrating a fluvial channel and estuary 4.5 Integration of volumetric ogy groups and to use the sophisticated display technolo- gies available at Birmingham to implement analyses per- information through solid mitting solid modelling and full 3D and stereo visualisa- modelling tion to maximise information extraction. Whilst RMS slicing allows for the display of more of the 3D surface modelling has been utilised within archaeo- information contained within the volume of seismic data, logical geophysics for some time (e.g. Neubauer and Eder- its generation as a planar slice results in the loss of integ- Hinterleitner, 1997). Yet, fundamentally, these still repre- rity of complex structures because the three dimensional sent only a single layer and are, at best, 2.5 dimensional in component of any anomaly is not adequately represented nature. They are not, therefore, suitable for the representa- (Figure 4.4). The potential to lose significant data is there- tion of the volumetric 3D seismic data or for the explora- fore very real. From the outset of the project it was de- tion of an internal structure within complex volume fea- cided to follow common practise amongst petroleum geol- tures. 36  Mapping Doggerland Figure 4.4 3D amplitude surface within a GIS. The green lines are bathymetric contours, and allow for the visual correla- tion between the anomalies and seabed topography. Figure 4.5 Segmentation of features of interest from a seismic volume. 37  Mapping Doggerland Mercury Software’s ‘Avizo’ visualisation package was to define the voxels contained within each feature. Al- utilised within the NSPP to provide a visual insight into though this can be an automated process noise, at survey the volumetric components of anomalies observed within boundaries and within the top of the data column pre- the 3D seismic data (Figure 4.5). This was supplemented vented such a simple implementation. Consequently, it by purchase of the Avizo “Very Large Data Pack” which was necessary to utilise a semi-automatic process of iso- supports datasets that may be hundreds of gigabytes in surfacing, utilising user set boundaries and thresholds to size. Representation of anomalies from industry standard constrain the process of automatic voxel selection. With SEG-Y data is possible utilising the recent features in the required information extracted from the seismic data- Avizo software developed to facilitate geological visuali- set, a solid model was built for each segmented feature by sation services for the oil and gas industry. Following generating a three dimensional frame for each feature import, Avizo possesses a suite of tools, which make it an model (Figure 4.7). Once constructed, it is possible to effective environment to explore, analyse and display disassemble a feature through slicing and gain further in- many types of remotely sensed data. The potential to spe- sights into the three-dimensional structure (see Figure 4.8). cifically extract information associated with voxels within Further information can also be gained through volumetric a series of solid models using this software has recently analysis within the solid modelling package. For example been demonstrated by Watters using ground penetrating the determination of an average voltex value within a vol- radar data (2006). ume according to assigned values or attributes may have particular significance. The calculation of channel volume To extract the required volumetric models, the original is an obvious output from such a model. However, whilst seismic data was directly segmented utilising picking valuable in their own right such modelling packages are techniques that are commonly employed in the oil and gas constrained, in analytical terms, as they lack many of the sector (Figure 4.5). A series of user determined lists are spatial procedures available to the majority of GISs. generated following segmentation, which contain features of interest. Fully automated selection tools can be utilised Figure 4.6 Wrapping of identified features within seismic data 38  Mapping Doggerland Figure 4.7 Solid model generated by wrapping Figure 4.8 Removal of elements of the solid model within the Avizo package from Figure 4.7 permits visualisation of the internal structure of the system 39  Mapping Doggerland As the volume forming the solid model has real world at- vironment into a GIS permits a representation of the shape tributes in all three dimensions it becomes possible to gen- and size of the anomalies, it still fails to represent the at- erate CAD models that can be imported into a GIS for tributes of the anomaly and the volume of the original sur- display and spatial analysis. This mode of representation vey. Although it is possible to transfer attribute data to is more suited to GIS as it is composed of polygonal ele- GIS data layers, solid modelling remains a superior ments, which, as they wrap the entire anomaly, can visual- method to represent 3D anomalies contained within volu- ise and facilitate the user’s appreciation of the actual vol- metric data and can be used in preference to standard pla- ume and size of anomalies. The results of part of such a nar slices. process can be seen in Figure 4.9. However, although export of these models from a fully three-dimensional en- Figure 4.9 Exported solid model within the GIS system. The feature can be seen to be composed of a contour model and a series of shapes representing the geophysical anomaly at the given time slice. 4.6 Merging technologies from that data (Figure 4.10). The ability to display associ- ated GIS information, and the ability to produce GIS inter- Commercial software groups have appreciated the increas- pretation layers directly within the package helps reduce ing importance of GIS data and the requirement for inte- spatial positioning errors, which may occur during tradi- gration with specialist softwares and data. 3D seismic in- tional planar slice integration. The direct generation of terpretation packages, including Tigress and SMT's King- GIS-compatible interpretative layers within specialist dom, have the ability to create and display datasets derived packages also improves quality of the interpretation from, or exported to, a separate GIS. Consequently, GIS through the availability of advanced attribute and opacity layers can be integrated within a fully 3D environment rending techniques not available within a standard GIS which provides the capacity to display the voxel volume of (see Thomson and Gaffney, this volume). the dataset and also vector or polygon features derived 40  Mapping Doggerland Figure 4.10 GIS Layers display within a fully 3 Dimensional environment - A cube of seismic data (displayed as a voxel volume) and a 3D surface & 3D GIS polygons (interpretation layer) 4.7 Conclusions information derived from volumetric geophysical survey and GIS. It is apparent that the use of solid modelling packages to display and analyse seismic volumetric data can assist in Despite this, it is accepted that the integration of all data integrating complex 3D data within environments that sources utilised by archaeological, or related, projects is utilise standard GISs for data management. However, it probably not feasible and, perhaps, may not even be required must still be acknowledged that the process of exporting at this point in time. However, what is incontrovertible is data divorces the solid model from the original volumet- that the complexity of our analyses demands that we are able ric datasets and, potentially, important attribute data or to transfer the rich spatial data that we utilise between tech- specialist attribute derivatives. Ultimately, this must nologies that permit us to visualise them in an appropriate limit the analytical potential of derived data. However, and increasingly sophisticated manner. Indeed, this may integration of volume data within a GIS provides en- ultimately be the most significant point. Our ability to visu- hanced opportunities for spatial analysis plus integration alise data is increasingly a primary driver and is linked di- with other, supporting spatial datasets and this cannot be rectly into novel interpretative positions. When considering achieved adequately within any proprietary seismic 3D data sets specifically, it is our capacity to facilitate visu- processing package. Until the technologies merge fur- alisation, through linkage between diverse softwares, which ther, and this is a real trend in software development, allow us to view data in a variety of new and exciting man- linking disparate data types through alternative tech- ners. Within the larger context of available spatial data nologies, including solid modelling softwares, offer the sources it is this process that will progress our interpretation best way forward for the integration and visualisation of of novel data sets and, ultimately, our understanding of past landscapes. 41 Mapping Doggerland 42  Mapping Doggerland 5 A Geomorphological Investigation of Submerged Depositional Features within the Outer Silver Pit, Southern North Sea Kate Briggs, Kenneth Thomson and Vincent Gaffney 5.1 Introduction esses (e.g. Valentin, 1957; Robinson, 1968; Balson and Jeffery, 1991; Praeg, 2003) or catastrophic drainage events A distinct east-west trending bathymetric deep, the Outer in an ice marginal environment (Wingfield, 1990). Alter- Silver Pit (OSP), lies at approximately 54˚N 2˚E on the natively, Donovan (1965) postulated that strong tidal cur- bed of the North Sea (Figure 5.1). This deep is the largest rents in the SNS during the early Holocene marine trans- of a series of offshore depressions in the Southern North gression were responsible for eroding such deeps. Sea (SNS) that are thought to have formed during Quater- nary glaciations, either as the product of subglacial proc- Figure 5.1 40 and 60m bathymetric contours of the Southern North Sea. The major bathymetric depressions are labelled. 43  Mapping Doggerland Whilst there has been considerable research into the for- most prominent of these being two elongate ridges that mation of the OSP, the localised geomorphology within also have bathymetric expression (Figure 5.2). These fea- the depression has been largely ignored despite its poten- tures will form the focus of this paper. It is the aim of this tial to provide further insights into the processes that led to study to classify the elongate ridge features, and extract the formation of the depression. The availability of 3D any information about the environments and conditions in seismic data to this study provided an ideal dataset for the which they were formed, through close examination of investigation of the morphological and stratigraphical their morphology and locality, and by detailed comparison characteristics of the OSP. This chapter will demonstrate to modern bedforms with similar morphology. that 3D seismic data can be used to identify a number of distinctive geomorphological features within the OSP, the Figure 5.2 Bathymetric contours (40, 50, 60, 70 and 80m below sea level) of the Outer Silver Pit area; depicted in white are the elongate ridge features 5.2 Feature descriptions 50m below sea level at their crests (Figure 5.3). Ridge A is elongate and tapers towards the south east but is Figure 5.3 and Figure 5.4 show the two elongate ridges rounded towards the north west end. It is discretely lo- within the OSP (Figure 5.2). Further details of the loca- cated within the OSP, entirely disconnected from the tions, dimensions and trends of the ridges are contained in banks of the depression. Although tapered at its south- Table 5.1 and Figure 5.2. The ridges are situated at the western end, Ridge B broadens to the northeast until it eastern limit of the OSP (Figure 5.2) and stratigraphically connects to the bank of the OSP at its eastern extremity. they lie at, or very close to, the seabed, approximately 40- Table 5.1 Position, dimensions and trends of Ridges A and B Ridge A Ridge B Length c.18km c.16.5km Width c.2km c.3km Height c.30m c.20m Trend ESE/WNW WNW/ENE 44  Mapping Doggerland Figure 5.3 Hilbert transform time slice at 0.06 seconds (top) depicting the two ridges. Two arbitrary lines are annotated which pass through Ridges A (X-X) and B (Y-Y). 45  Mapping Doggerland Figure 5.4 3-D illuminated view of Ridges A and B, with a depression referred to in the text highlighted in red. The ridges are vertically exaggerated. Table 5.2 Average dip of the flanks of Ridges A and B Ridge A Ridge B Shallow Slope Steep Slope Shallow Slope Steep Slope Major Axis 0.07˚ 0.1˚ 0.09˚ Minor Axis 0.035˚ 0.84˚ 0.035˚ 1.13˚ In cross section, Ridge A displays an asymmetry along 1.13˚ on the steeper slope of Ridge B. Given that the trends its major axis and both ridges are asymmetric along of the ridges are offset the steeper slope of Ridge A faces their minor axes (Figure 5.3 and Figure 5.4); the asym- south east whereas the steep flank of Ridge B faces south metry of the minor axis of Ridge A decreases towards west (Figure 5.3 and Figure 5.4). Cross sectional views of its southeasterly end. The steeper slopes of the ridges the features (Figure 5.3) also reveal that the crests of the are slightly concave whereas the shallower slopes are ridges are relatively flat. predominantly convex (Figure 5.3). The dip of the ridges’ flanks are relatively shallow (Table 5.1 and Fig- ure 5.5), reaching an approximate maximum of only 46  Mapping Doggerland Figure 5.5 Dimensions of the OSP and Ridges A and B 47  Mapping Doggerland Figure 5.6 Two 2D seismic lines running through Ridge A and their location. Marked in red are the internal reflectors and at the base, the undulating surface upon which the ridge lies 48  Mapping Doggerland Figure 5.7 Quaternary geology map of the eastern end of the OSP overlain with contours of the depth to the base of the Holocene sediments (black) and the bathymetry of Ridges A and B (dashed red). (Adapted from Larminie, 1989a and 1989b) Two available 2D seismic lines (Figure 5.6), which pass northeasterly direction and is approximately 7.5km wide through Ridge A, provide further details not apparent in with a shallow northern flank. The second branch, to the the lower resolution 3D data. Firstly, the surface of south of Ridge B, trends in a southwesterly direction and is ridge is smooth on both the steep and shallow slopes. approximately 5.5km wide with relatively steep banks on Secondly, the ridge lies on a surface that undulates rela- both sides. It is not possible to constrain the length of the tively sharply, truncates the strata below and is seen either branch as they continue beyond the confines of the throughout most of the OSP, where a relatively thin data. Also, to the immediate south of Ridge B lie two lesser veneer of sediments covers it. Finally, the 2D data depressions measuring approximately 2km and 0.75km in shows that the internal structure of Ridge A is composed width (Figure 5.4 and Figure 5.7) that emerge from the raised of a series of dipping internal reflectors (foresets). ground to the east of the OSP. The Quaternary geology map of the area reveals that these small depressions correspond On a broader scale, the ridges are located at the conflu- with outcrops of the Botney Cut Formation (Figure 5.7), a ence of several features. At the eastern end of the OSP, system of partially or completely infilled subglacial valleys and to the sides of Ridge B, the depression forks into dated to the late Weichselian (Larminie, 1989b). two lesser branches (Figure 5.2, Figure 5.5 and Figure 5.7). The branch to the north of Ridge B trends in an 49  Mapping Doggerland The geology map of the Holocene and seabed sediments 300µm) that have been derived from Pleistocene glacial and (Figure 5.7) reveals many significant properties of the periglacial deposits (Larminie, 1989a). These sediments are ridges. The contours of the depth to the base of the overlain by sands at the seabed and in relation to the sur- Holocene sediments in the Silverwell area shows an rounding areas this is a larger sediment size than present at absence of the ridges in the topography, thus it can be the seabed of the OSP depression but of the same size or concluded that the ridges are of Holocene age. The smaller than the seabed sediments on the nearby higher Holocene sediments of the ridges consist of the Tershel- ground. Furthermore, Figure 5.8, an RMS amplitude map of lingerbank Member (of the Nieuw Zeeland Gronden the ridge area, depicts the sand ridges as having a high am- Formation), which is defined as open marine sediments plitude signal commonly associated with sandy sediments. and are slightly muddy sands (mean grain size 120- Figure 5.8 An RMS amplitude map from 0.004 to 0.05 seconds (TWT) of the eastern end of the OSP. N.B. the lighter col- ours depict high amplitude response areas 5.3 Discussion and Feature classifi- mately 120m (Shackleton, 1987) over the past c.20,000 years. In the British Isles a large proportion of the deglacia- cation tion had occurred by 13,000 BP (Lowe and Walker, 1997) uncovering vast tracts of the present North Sea Basin as a Deterioration of ice masses following the last glacial land surface that was gradually inundated during the Holo- maximum has led to a global sea level rise of approxi- cene. Shennan et al., (2000) reconstructed the pattern and 50  Mapping Doggerland timing of inundation in the North Sea basin during the which they currently exist. However, without going into Holocene using reliable indicators of past sea levels further details of the morphology and formation of such bars obtained from sediment core analysis and geophysical it is possible to discount, with some confidence, Ridges A models that integrated ice sheet reconstructions, the and B as being fluvial bedforms from the very early Holo- Earth's rheology, eustasy and glacio- and hydro- cene. This is because significant evidence exists that is sug- isostasy. It was predicted that the OSP had become a gestive of at least one major erosional event in the OSP dur- shallow estuary by 9,000 BP, the Dogger Bank had be- ing the early Holocene. come isolated from mainland Europe at high tide by 8,000 BP and by 6,000 BP the Dogger Bank had be- The Quaternary geology map of the OSP area reveals that, come completely submerged. Thus, since it is known aside from the occasional patches, there is a general absence from the Holocene geological maps and the bathymetry of late Pleistocene/Late Glacial Maximum sediments in the that Ridges A and B were formed at some time during depression, which makes it somewhat distinctive from the the Holocene (10,000 BP to present), it is suggested that continuous cover in the surrounding area. Although the ab- they are the product of either terrestrial or marine proc- sence of these sediments cannot be taken as sole evidence of esses. their erosion, it is a notable spatial correlation and reasoning would suggest that it is highly improbable that the ice sheet 5.3.1 Terrestrial geomorphological processes that covered the OSP area during the Late Glacial Maximum (Carr et al., 2006) would not have deposited material either During the early Holocene, in the period prior to marine directly, or in the form of proglacial outwash, in such a se- inundation (10,000 BP to 9,000 BP), a number of terres- lective manner in this region1. Evidence of erosion in the trial landscape processes common to temperate envi- OSP during the early Holocene is reported by Larminie ronments are likely to have sculpted the land surface of (1989a). This indicates that where the Holocene sediments the SNS. There are two such classes of processes that overlie the Botney Cut Formation (a Late Glacial Maximum could have led to the reworking and deposition of such tunnel valley infill), they are separated by an erosional con- volumes of sediment retained in Ridges A and B, tact. This is supported by 2D lines crossing the OSP. namely mass movement and fluvial processes. Mass movement, the transfer of material down a slope under In Figure 5.9 a 2D line, situated to the north of Ridges A and the influence of gravity (Summerfield, 1991), is capable B, shows the base Holocene reflector truncates the strata of generating significant deposits of fine (colluvium) to below (the Botney Cut Formation). The sediments recorded large grained (talus) sediment. Inherently, sediments as overlying the erosional surface in the OSP are described transported by mass movement are deposited at the base as fully marine (Larminie, 1989a); consequently, it is possi- of the slope from which they originated. Thus, the ble that marine processes led to the erosion of late Pleisto- situation of Ridge A at 4-5km distance from the proxi- cene sediments and any early Holocene terrestrial sediments mal slope eliminates the possibility of it being deposited that may once have been present in the OSP. It is also likely by a gravitational process. Although Ridge B is at- that the strong erosional processes capable of removing these tached to a slope of the OSP depression, it too can be sediments would also have reworked Ridges A and B had disregarded as a mass movement deposit as the feature they been present. As the 2D lines in Figure 5.6 show, the is approximately the same height as the adjacent slope ridges overly the truncated surface and thus it is likely that and is thus by no means a mass of material deposited at they are a product of marine processes that operated subse- the base of it. quent to the recorded erosional event. It is more reasonable to assume that fluvial processes 5.3.2 Marine geomorphological processes contributed to the shaping of the landscape in the OSP during the very early Holocene. Runoff would have The entrainment and subsequent transportation of the large been routed to, and exploited, any local depressions. volumes of sand sized sediments that comprise Ridges A and Thus it is a distinct possibility that Ridges A and B were B would have required relatively strong currents over a sus- formed in a fluvial environment. Fluvial bedforms, spe- tained period of time. At present the OSP is a low energy cifically bars, can achieve significant dimensions, de- environment thought to be sheltered from the high tidal cur- veloping lengths that are comparable to the width of the rent velocities, conveyed from the north, by the shallow channel in which they form (Knighton, 1998). They can Dogger Bank (Eisma, 1975); it is also understood that the take on a variety of shapes and occur in a range of con- OSP is void from the influence of surface waves that are not ditions (Knighton, 1998). The point of branching in the OSP channel at its eastern end (in the vicinity of the two ridges - see feature descriptions), may have hosted a 1 The Holocene map also shows that the early Holocene, brackish confluence of two rivers under fluvial conditions; sev- marine sediments of the Elbow Formation, that exist in the SNS and eral observations exist of bars that have formed at con- in the locality of the OSP are also absent from the depression. fluences in modern fluvial systems (e.g. Melis et al., Because the OSP is a closed depression it is possible that there 1994; Rhoads and Kentworthy, 1995; De Serres et al., were no Holocene intertidal deposits; when sea level rose 1999). It is therefore probable that features similar to sufficiently to cross the threshold of the depression water would Ridges A and B would have formed in the position in have flooded the pit thus the inundation would have been relatively sudden and not a gradual incursion. 51  Mapping Doggerland considered to penetrate beyond 30m in depth in the SNS (McCave, 1971). Figure 5.9 Truncated strata on the bed of the Outer Silver Pit (location depicted in the insert). The red horizon is the Holocene Base reflector (Hebbeln and Meggers, 1999); the blue horizons show the strata that have been truncated Such low hydraulic energy has led to sedimentation on records to further constrain the precise marine influenced the bed of the OSP2 of fine sands and silts (Veenstra, environment the ridges were formed in, i.e. estuarine, strait 1965; Larminie, 1989a) which are locally laminated3 or shallow clastic shelf environment. Therefore, in order to (Larminie, 1989a). Indeed, the late Holocene sedimen- gain such palaeoenvironmental insight the morphology of the tary record of the OSP is similar to that of the present features must be examined in respect to modern features of (Larminie, 1989a) suggesting that the depositional envi- similar form. ronment was also similarly low energy. Furthermore, Shennan et al., (2000) propose that by 6,000BP the The distinctive elongate morphologies of the ridges are simi- coastline at the margins of the North Sea was compara- lar to a number of landforms that develop subaerially along ble to that at present, and therefore it is feasible to sug- coasts and on the bed of continental shelves or tidal inlets. gest that the currents and water depths around the OSP were similar. Consequently, it is likely that Ridges A 5.3.3 Subaerial coastal landforms and B were formed in the early Holocene when the wa- ter depth in the OSP was relatively shallow and the cur- Of the abundance of subaerial coastal landforms that exist rents relatively strong. However, insufficient evidence globally, Spits and Barrier Islands are identifiable as being can be derived from the available sediment and sea level depositional elongate ridges. Spits protrude into estuary/bay mouths or the sea whilst attached to the mainland at one end 2 (Masselink and Hughes, 2003). They generally have one or As mentioned in the previous section the sediments on the more landward pointing recurves at their distal ends, al- top of Ridges A and B are of a larger grain size than those on though they can be entirely linear (Bird, 2000). Spits form the bed of the OSP but similar to those at equivalent depths when there is a sudden break in the direction of the coastline below sea level. This is suggestive of the action of currents and/ or wave action on the tops of the ridges. but sediment transport and deposition continues along the 3 The laminations in the OSP form as a consequence of the original pathway (Haslett, 2000). They are built up above yearly generation and deterioration of a thermocline at 20-30m high tide level and lagoons and marshes develop on the shel- depth that is able to develop as a consequence of the minimal tered landward side. impact of tidal action (Eisma, 1975). 52  Mapping Doggerland There are several aspects of Ridges A and B and their environment that exists between barrier islands and the context which suggest that it is unlikely that they are mainland today. spits. Although Ridge B protrudes from 'land' at a change in direction of that land body, Ridge A is posi- 5.3.4 Marine bedforms tioned independently at a point where there is no direc- tional change in the 'coast' and thus suggests that Ridge The possibility that Ridges A and B are either submerged A was not formed by processes of continued longshore Spits or Barrier Islands is rejected and alternatively a suite of drift of sediment along a pathway not contiguous with marine bedforms are considered. There are three main elon- the coastline. If these features were spits, it would be gate sandy bedforms that are commonly identified in marine likely that the amplitude signal from the 'sheltered' area or tidally influenced environments. These are sand ribbons, on the landward side would be noticeably lower as a sand waves and sand banks/ridges. Sand ribbons are longi- result of the fine sediments that accumulate in low en- tudinal bedforms that develop parallel or sub-parallel to the ergy environments of lagoons or marshes. dominant tidal flow current (Cameron et al., 1992). They occur in areas with relatively high surface current velocities, Figure 5.8 shows that there is no notable difference in often in excess of 100cm/s (Johnson and Baldwin, 1996). the amplitude signal between the land adjacent to the Although sand ribbons vary greatly in size they are generally ridges on either side, suggesting the absence of a lagoon less than 15km in length, 200m wide and up to 1m thick or marsh. However, it must be noted that it is possible (Kenyon, 1970), however, as a result of their diminutive that these sediments were eroded during the Holocene thickness they tend not to be identifiable on seismic images. transgression. The longshore drift processes that are responsible for the formation of spits are likely to be Sand waves, small/medium subaqueous dunes or megarip- minimal within an estuary or strait environment where ples as they are sometimes referred to, are flow transverse the flow dynamics are likely to be dominated by the features, i.e. their crests lie approximately perpendicular to flood and ebb currents of the tide. It may have been the direction of the main current (Cameron et al., 1992; possible for a spit to form at a point during transgression Blondeaux, 2001). They are greater in height than sand rib- if the easternmost land body in the OSP had been an bons, generally falling between 1.5m and 10m in thickness open coastline. Figure 5.10 however, shows that it was (Ashley, 1990; Johnson and Baldwin, 1996) and they gener- not and that it was inundated at an early point relative to ally occur in areas where current velocities exceed 65cm/s other local land bodies. Finally, although in certain cir- (Johnson and Baldwin, 1996). Sand waves can be both cumstances it is possible that spits can be linear, Ridges symmetrical and asymmetrical in form (Blondeaux, 2001) A and B lack the recurved end that is frequently charac- with maximum slope angles of 10-12˚ (Cameron et al., teristic of spits. 1992). Barrier Islands are linear, shore parallel sand bodies Ridges A and B are both significantly larger (Table 5.1) than that, like spits, extend above sea-level (Masselink and the reported dimensions of sand ribbons or sand waves and Hughes, 2003). They vary greatly in size and can be up thus it is concluded that they are neither feature. Conversely, to 100m high, 100s of metres wide and 1000s of metres sand banks/ ridges are much larger features commonly long (Haslett, 2000). Further similarity to spits extends measuring up to 80km long, 1-3km wide and 10-50m high to the low energy lagoons that are often situated be- (Johnson and Baldwin, 1996; Dyer and Huntley, 1999) and tween the barrier and the mainland (Masselink and are of similar dimensions to Ridges A and B. Furthermore, in Hughes, 2003). Again, there are several features of agreement with the morphology of Ridges A and B, Johnson Ridges A and B and their locality that suggests these and Baldwin (1996) describe sand banks/ ridges as linear features are not barrier islands. Firstly, neither of the bedforms that are asymmetrical in cross section and com- ridges lies parallel to a potential shoreline. Secondly, at posed of medium to fine sands. a time when these features would still be in part above sea level, they would have been situated in an inlet (es- The morphological evidence is strongly suggestive that tuary/strait) environment and not on the open coast Ridges A and B are indeed sand banks/ridges in terms of where Barrier Islands are observed to form (Figure both the similarity which exists between them and also the 5.10). Finally, as previously stated there is no sugges- elimination of a range of other possible features based on the tion in the amplitude data of the presence of fine la- available data. However, this study is lacking in the provi- goonal sediments in a possible lee area of either ridge, sion of data that may provide validation for such claims. thus suggesting the (possible) absence of the sheltered 53  Mapping Doggerland Figure 5.10 Changes in land area with rising sea level based upon the depth to base Holocene map (Larminie, 1989b). Ridges A and B are depicted in red 54  Mapping Doggerland For example, higher resolution seismic data and sedi- that operate to form analogous landforms in the present day. ment cores may be able to provide greater detail of the Thus it is the purpose of the following section to utilise the internal structure of the features and thus reveal the available data in an attempt to reconstruct the processes that presence/absence of certain sedimentary structures char- operated in the OSP when the sand banks were last active, acteristic of sand banks/ ridges4. Also access to high- i.e. the Early Holocene6. Due to the rapid and significant density sediment analysis would provide information of extent of sea level rise in the North Sea during the early the existence/absence of any sediment grain size grada- Holocene, marine conditions within the OSP would have tions that occur across the features5. However, the body undergone significant transformations e.g. from estuarine, to of evidence available is considered to be sufficient to strait, to open sea. Therefore the suite of processes which classify, albeit tentatively, the features in the OSP as initially led to the formation of the sand banks may have sand banks/ridges. Thus the following inferences based been different to those that ultimately shaped and maintained on this classification, are made with a degree of uncer- them prior to their becoming moribund. tainty. 5.4 Environmental Interpretation It is possible to make the distinction between sand ridges and sand banks based upon the criteria that ridges 5.4.1 Sand Bank classification have a length/width ratio that exceeds 40 (Amos and King, 1984). The length/width ratios of Ridges A and B Sand banks develop in a number of marine settings where are just 9 and 5.5 respectively and so are herein referred there is an abundant supply of sand and currents strong to as 'sand banks'. enough to transport the material. Dyer and Huntley (1999) produced a classification system of the various types of sand Sand banks may exist in either an active or moribund banks resulting in their following subdivision: state. Active sand banks are present in areas where the tidal currents are relatively strong (>50cm/s). Their i) offshore banks, crests are shallow in the water and are generally quite ii) estuary mouth banks (including those formed sharp, except when they approach sea level at which on tidal deltas and those formed in wide estuary point they display flattened tops. The steep slopes of mouths where there is not a delta) and active ridges are relatively steeply inclined at c.6˚ and iii) headland associated banks (separating those they are flanked with sand waves. Conversely, mori- banks formed around stable and recessional bund sand banks are situated in relatively deep water headlands). where currents have diminished to <50cm/s and are thus insufficient to transport sand on the seabed (Johnson and Offshore sand banks are described by Blondeaux (2001) as Baldwin, 1996; Dyer and Huntley, 1999). They have rhythmical features that generally have a crest spacing of a rounded profiles, shallow slopes of <1˚ and an absence few kilometres. They form with their long axes orientated at of sand waves on their flanks (Johnson and Baldwin, a small oblique angle to the peak tidal flow direction, where 1996; Dyer and Huntley, 1999). The surrounding sea there is a convergence in bed load transport paths (Dyer and floor is covered by sandy or muddy sediments as op- Huntley, 1999). The crests of the banks are often at or only a posed to larger gravel sized sediments found in the vi- few metres below the sea surface at low tide. Although there cinity of active banks (Johnson and Baldwin, 1996; is more than one 'ridge' within the OSP, sets of offshore sand Dyer and Huntley, 1999). As previously stated it is banks/ridges that have been identified in the SNS commonly known that the currents in the OSP are at present very occur in groups of ten or more (Cameron et al., 1992). Fur- weak, the crests of the sand banks are flattened and at a thermore at a point when the crests of the banks would have depth of approximately 40-50m below sea level and the been at or close to sea level at low tide (50-40m below pre- steep slopes have mean angles of only 0.84˚ and 1.13˚. sent) the OSP would have been an enclosed seaway (i.e. an The 2D seismic data depicts the surfaces of the sand estuary or a strait)7 not an offshore environment. Thus it is banks as smooth and thus void of sand waves. Thus the concluded that the sand banks within the OSP cannot be sandbanks in the OSP are interpreted as being moribund classified as 'offshore banks’. on the basis of the geomorphological observations. It is also possible to conclude that Ridges A and B are not The moribund sand banks within the OSP are thus relict headland associated banks. The crests of the sand banks in features formed in conditions that no longer exist at this the Outer Silver Pit are at approximately the same depth be- location. Such features can provide insight into past low sea level (c.50m) as the surrounding 'land/potential environments through the understanding of processes headland'. Thus at the point when the water depth in the 4 6 The internal structure of linear sandbanks within estuaries is As previously noted, by 6,000 BP the North Sea sea levels, and likely to vary according to the precise controls upon its consequently the conditions in the OSP, were similar to those at formation and maintenance. However, one may expect to present. 7 observe such structures as clay drapes, rippled beds, 'ebb Assuming a tidal range of less than 10 metres based on the current foreset packages' (Fenies et al., 1999). (~8m) and Holocene tidal ranges (<8m) of the Humber estuary on 5 Larger particle sizes are likely to be observed on the flank the east coast of England (Shennan and Horton, 2002). Data for where the dominant currents operated. Holocene tidal ranges of the OSP are not currently available. 55  Mapping Doggerland OSP was sufficient to submerge the sandbanks, the ad- at the time the sand banks were last active. For the jacent ‘land’ would also have been submerged and thus same reasons as stated above in point 1, this may could not have been a headland. However, it is possible not have been so. The topographic data in the depth that at some time in their development the ridges were a to the base of the Holocene map may have pre- lesser height in relation to the 'land' but subsequently sented a better representation of this land surface8. gained height during inundation and landform evolution. However, because the sand banks are Holocene fea- Nonetheless, regardless of the stage of inundation, at a tures they are not depicted. In using the bathymetric time when these sandbanks could have been active, the data it is argued that a more realistic comparison of OSP would have been a seaway, not an open coastline the sand banks' elevation in relation to that of the with significant longshore processes to generate head- surrounding land is acquired as late Holocene depo- land banks. Furthermore, several components remain of sition is compensated for; this is of course making the sandbanks' morphology that suggests they are not the further assumption that the sedimentation rate headland-associated banks. Headland associated banks was spatially continuous. occur in zones of littoral sediment transport conver- gence where there is an acute change in the direction of 3. The tidal range is assumed to be 10m, with low tide the coastline (Dyer and Huntley, 1999). Sandbanks being at, what is today, 50m below sea level. The form on one or both sides of the headland and are sepa- level of low tide is based on the further assumption rated from the headland by a deep narrow channel; this that the crest of the active sand banks became ex- precludes Ridge B from being a headland-associated posed at low tide. This is a somewhat robust as- ridge as it protrudes from the cusp of the 'land' to which sumption based on observations of present day sand it is attached. Headland associated banks are generally banks in estuaries (e.g. Wright, et al., 1975). How- only a few kilometres in length and thus are much ever, in constraining low tide to 50m, i.e. the height shorter than the sand banks in the OSP (Table 5.1). The of the crest, all of the above errors outlined in point sand banks in the OSP also do not portray the pear 1 are subsumed. The assumption of a 10m tidal shaped form commonly associated with headland banks range is an unavoidable gross assumption, as there (Dyer and Huntley, 1999). is no data available for Holocene tidal ranges/prisms within the OSP. Nonetheless, the as- On the premise that, when last active, the sand banks sumption is founded in the knowledge of present within the OSP would have been close to the water sur- day tidal ranges of the nearby east coast of England face or exposed at low tide it is possible to infer, based and the current understanding of the effects of such on the submerged topographical data, that the sand a land configuration on tidal prisms. The exact banks were probably formed in a strait or estuary envi- range of 10m was taken for the purpose of simplic- ronment. Figure 5.11 is a generalised depiction of the ity in the production of Figure 5.11 as the contour possible land surface, intertidal areas and inundated spacing of the bathymetry is 10m. This assumption areas using bathymetric data and based on a series of is very limiting for this study, as the land surround- assumptions that, at present, cannot be further con- ing the OSP is very low relief and therefore a small strained: difference in the tidal range could produce very dif- ferent marine conditions from high to low tide. 1. It is assumed that the crests of the sand banks within the OSP were at, what is now, 50m be- 5.4.2 Estuaries and Sand Bank Formation low sea level when they were last active. However, this depth is derived from the Estuaries are varied and complex environments; the unique bathymetric data which has a 10m contour in- suite of processes that operate in a given estuary are a func- terval and so theoretically the crest of the bank tion of several individual factors (e.g. tides, waves, fluvial could lie at a lesser depth, up to c.41m below input and morphology), which are temporally dynamic over present sea level. Also, the height of the sand the short- and long-term. In order to gain insight and aid bank may have changed from when it was last interpretation of estuaries and their many facets, several clas- active as a result of post inundation deposition. sification schemes of various estuarine attributes have been Larminie (1989a) suggests a generalised range proposed. Such classifications are also valuable in the inter- of only 1-5m of Holocene sedimentation in the pretation of past estuarine conditions based on relict features, locality of the sand banks. It is not thought that thus potentially aiding an interpretation of past conditions erosion would have had a significant impact within the OSP founded on the available information of the upon the elevation of the sand banks since be- moribund banks and their surroundings. coming moribund, because for them to be moribund the local currents operating must be so weak as to not transport sand sized material. 2. It is assumed that the topographic expression of 8 Although, problems exist with quantifying and identifying the the seabed surface represented in the bathymet- spatial and temporal distribution of erosion, and subsequently ric data approximates to that of the land surface inferring the effects the change in topography may have had on hydraulic processes relative to the stages of sandbank formation. 56  Mapping Doggerland Figure 5.11 Postulated land configuration at the time when the OSP sand banks were last active. The contours are the bathymetric contours from Larminie (1989a). The map is based on a number of assumptions highlighted in the text Hayes (1975) proposed that estuaries could be grouped In macro-tidal/tidally dominated9 estuaries independent and according to tidal range. He applied Davies' (1964) mutually evasive flood- and ebb-dominated channels are scheme of tidal classification that identified three widely observed (e.g. Robinson, 1960; Price, 1963; Ludwick, classes of tidal range, micro-tidal (0-2 m), meso-tidal (2- 1975; Wright et al., 1975). These mutually evasive channels 4m) and macro-tidal (>4m). Based on extensive obser- develop as a result of the tidal asymmetry that occurs in such vations of shorelines from around the world, Hayes estuaries. Tidal asymmetry refers to the asymmetry in mag- (1964) concluded that distinctions could be made be- nitude, velocity and duration between the flood and ebb tides tween tidal ranges and their associated suite of deposi- in a given estuary (Masselink and Hughes, 2003). The tional landforms. Most notably he observed that linear asymmetry is generated as a result of differences in the mag- sand banks were widely associated with macrotidal es- nitude of influence exerted by friction upon the flow of each tuaries. The currents in macro-tidal estuaries are strong tide. In flow contained within the channel a lesser propor- and capable of transporting the relatively coarse grain tion of the volume of the flood tide (i.e. the crest of the tidal sediments that form sand banks. 9 As a consequence of large tidal ranges, the driving morphological processes in such estuaries are tidally dominated. 57  Mapping Doggerland wave) is in contact with the channel surface (and there- banks are commonly orientated obliquely to the direction of fore friction) than that of the shallow ebb tide. Thus the peak tidal flow (Kenyon et al., 1981). velocity of the channel flow of the flood tide is greater than that in the ebb. Given that the discharge volume Only a restricted volume of quantitative studies on sand through the channel of the flood tide is similar to that of banks in estuaries exists. Such quantitative study would the ebb tide, it follows that the duration of the ebb tide is broaden and improve the depth of understanding of the proc- greater in order to compensate for the reduction in ve- esses operating to build and maintain the banks. This there- locity. fore limits the level of environmental interpretation that can be derived from the sand banks within the OSP. Neverthe- In reality the tides in many estuaries are not fully con- less, it is possible to derive some basic understanding of the tained within the main channel along its entire length or processes in the OSP during the early Holocene based on the throughout the full tidal cycle and, frequently, broad sand banks. zones of intertidal land flank the channel. Friedrichs and Aubrey (1988) examined the relationship between The presence of the sand banks suggests that the OSP was channel shape and tidal asymmetry. They suggested macrotidal and therefore tidally dominated. The tidal cur- that where large intertidal areas become submerged dur- rents would have been strong and capable of transporting the ing the flood tide, the proportion of the water volume sand sized sediments in the sand banks. Tidal asymmetry is exposed to surface friction effects, and indeed the sever- likely to have led to the formation of ebb- and flood- ity of friction10, is increased and thus the efficiency of dominated channels between which the sand banks may have the conveyance of the tide up-estuary is reduced. Con- formed. It is possible that 'Ridge A' was formed in such a versely, during the ebb tide, when water levels are way, as it lies independently in the 'main channel' and trends lower, flow is conveyed within the channel where there in such a manner relative to the surrounding topography, that is a lesser frictional effect. Consequently, the ebb tide it is plausible to suggest it would have formed at an oblique has a greater velocity than the flood tide in areas with angle relative to the direction of the dominant current. How- significant tracts of intertidal land, leading to an overall ever, the positioning of 'Ridge B', attached to a mass of land ebb-dominance. In areas of ebb-dominance, ebb- and at the confluence of the branch in the OSP depression, dominated currents/channels exist within the main may suggest an alternative mode of formation. The conver- channel and the flood dominant currents prevail over the gence or divergence of water moving downstream or up- shallow, intertidal areas. To the contrary, in areas of stream along the branches of the OSP most probably resulted flood dominance, it is the flood currents that dominate in certain hydraulic processes that would have led to a loss of in the main channel and the ebb currents that are preva- fluid energy and the subsequent deposition of the sand bank lent along the shallow margins of the channel. It is not sediments. Furthermore, the two smaller depressions di- uncommon for estuaries to exhibit portions of both rectly to the south of Ridge B (see feature descriptions) may flood- and ebb-dominance as a result of the changing have accommodated fluvial and/or tidal channels that are morphology along their lengths. Ebb-dominance more also likely to have influenced the hydrodynamics controlling frequently occurs towards the head of estuaries where the processes forming/maintaining the sandbanks. However, there are larger intertidal zones and flood-dominance it is beyond the scope of this study to examine in further de- mostly occurs towards the mouth. At a given point tail the precise fluid mechanics, which may have operated in these regions will interdigitate (Harris, 1988) and, as this area, which resulted in the deposition of 'Ridge B'. It is previously mentioned, the flood and ebb channels tend however, important to suggest that sedimentological analysis to be mutually evasive. and 2D seismic data depicting the internal structure of Ridge B has the potential to provide clarification of the precise It is between the mutually evasive ebb and flood chan- processes that operated to form the sandbanks. nels within estuaries, where bed loads converge and sand banks occur (e.g. Wright et al., 1975; Harris, 1988; 5.5 Conclusions Harris et al. 1992; Dyer and Huntley, 1999). The sand waves that exist on the sides of sand banks are com- The two elongate ridge features identified in the OSP in the monly observed to be ebb or flood orientated on oppo- 3D seismic data are interpreted as moribund sand banks that site sides (e.g. Wright et al., 1975; Harris, 1988). Harris formed in as estuarine environment during the early Holo- (1988) suggests that this is indicative of a 'circulatory cene marine transgression between approximately 10,000 pattern around the sandbank crest'. In most cases the and 6,000 BP. From this interpretation and knowledge of tidal flow in one direction will dominate; it is this that modern analogues, it is inferred that the tidal range of the causes the sand bank cross-sectional asymmetry (Ken- OSP was probably macro-tidal and the tidal currents con- yon et al., 1981). In estuaries sand banks migrate away veyed in the estuary were relatively strong. The identifica- from their steep slopes, as it is the steep slope that is tion of a strongly undulating truncation surface in the 2D actively eroded (Dyer and Huntley, 1999). Also sand seismic data suggestive of a major erosional event is suppor- tive of the theories of Donovan (1965; 1975) which suggest 10 that strong marine currents were, at least, in part responsible The roughness coefficient (measure of friction) is generally in the formation of the OSP depression. greater over intertidal areas than within channel as a result of the vegetation that colonises the periodically emergent land. 58  Mapping Doggerland It is of paramount importance to highlight that this could provide the opportunity for dating and therefore con- study and its findings are fundamentally restricted by straint of the timing when the ridges were last active. Fi- the limitations of both the available data and the level nally, current limits in quantitative understanding of the of current understanding of modern, analogous sys- formation and dynamics of sand banks in estuaries, as pre- tems. In terms of the available data, limitations arising viously mentioned, render the prediction of the processes from the relatively low resolution of the 3D seismic that led to the formation of the sandbanks in the OSP prob- images were, in part, overcome by use of supplemen- lematic. As Knighton (1998) suggests 'ancient deposits and tary 2D seismic data. However, the data quality re- bed forms are a key element in palaeohydraulic reconstruc- mained insufficient to examine/identify very fine scale tion'; however, 'the reliability of such reconstructions de- sedimentary structures, for example clay drapes and pends on an adequate understanding of the formative proc- rippled beds, which have the potential to provide vali- esses operating within the present-day environment'. dation of the geomorphological classification of the sand banks. Superior resolution 2D line data (e.g. high Despite the limitations and restrictions faced by this study it frequency sonar source) through the sand banks may has highlighted the suitability of 3D seismic data for the have allowed for examination of such structures. How- identification and broad interpretation of submerged, large ever, this data was not available. A lack of physical scale, geomorphological features and their surroundings. evidence in the way of sedimentary data from cores Furthermore, this study has reflected the vital and signifi- was unavailable for this study, indeed such data could cant role that relict landforms play in the reconstruction of have provided the means for ground truthing and vali- past environments and specifically palaeohydraulic proc- dation of the interpretations made by providing, for ex- esses, whilst fervently asserting the importance of the reli- ample, information of grain size distribution and de- ability and wealth of data in producing reconstructions that tailed internal structure. Furthermore, sediment cores are robust and of scientific value. 59 Mapping Doggerland 60  Mapping Doggerland 6 Salt tectonics in the Southern North Sea: controls on late Pleistocene-Holocene geomorphology. Simon Holford, Ken Thomson and Vincent Gaffney 6.1 Introduction 6.2 Relationships between salt structures and late Pleistocene- During the Upper Permian (c. 260 to 251 million years BP) over 1000 m of marine evaporites (the Zechstein Holocene fluvial systems. Supergroup) accumulated in the North Sea (Cameron et al. 1992). Their subsequent burial promoted mobility on By comparison with the adjacent onshore landscapes of geological timescales resulting in thickness variations East Anglia and Continental Europe (Belgium, Denmark) from less than 50 m, in regions of salt withdrawal, to more which are characterised by similar subsurface geology, the than 2500 m in some of the major salt diapirs (Cameron et late Pleistocene-Holocene landscape of the SNS is likely al. 1992). Today, the deformed salt deposits encompass a to have been defined by a low-relief, relatively flat land wide range of structural morphologies (Jenyon 1986) and, surface. Once Holocene modifications due to sediment in places, the crests of salt structures are within 100 m of accretion (e.g. sandbanks; Stride et al. 1982) or erosion the seabed (Cameron et al. 1992). This proximity to the (e.g. by tidal scouring) are factored into consideration, the present-day depositional surface suggests that the uplift present-day bathymetry of the SNS reveals a relatively flat and penetration of the overburden by Upper Permian surface with water depths of only 20 to 40 m in the study evaporites may well have influenced the topography and area (Cameron et al. 1992). It is difficult to reconcile the hence depositional systems in this region in the recent present-day bathymetry to any distinct structural control geological past. by salt tectonics, but it is likely that the subaerial late Pleistocene-early Holocene landscape may well have been Measured uplift rates of emergent and immediately influenced by near-surface salt bodies e.g. in the form of subsurface salt diapirs range from 2 to 7 mm yr-1 relative topographic highs above salt cored anticlines. (Bruthans et al. 2006). This invariably leads to the There is evidence to support this hypothesis from 3D deformation of the overlying rock layers and hence exerts seismic data from the north of this study area. Figure 6.1 an important control on synkinematic sedimentation presents a series of timeslices (amplitude, Hilbert patterns, and hence geomorphic processes. Topographic transform and phase, 0.1 seconds) from the south of the relief produced when salt approaches the land surface in project area. Near the centre of the timeslice a series of continental settings can vary from 45 m (Al Salif, Yemen; broadly concentric reflectors, diagnostic of a salt diapir Davison et al. 1996) to up to 1500 m (Zagros Mountains, (Stewart 1999) are clearly visible (Figure 6.1b). An Iran; Talbot and Alavi 1996), whilst in offshore settings arbitrary seismic line through the 3D seismic volume e.g. the Mississippi Delta; topographic relief varies (Figure 6.1e) confirms the structure is a diapir and hence between 100 and 240 m (Jackson et al. 1994). likely to have been expressed by a relative topographic high on the late Pleistocene-Holocene land surface prior to However, a more appropriate analogy for the influence of early Holocene marine transgression. There is some salt tectonics on landscape evolution in the SNS during the faulting associated with the southern flank of the salt late Pleistocene-Holocene is that of the Five Islands, south diapir (Figure 6.1b). To the south west of the salt structure central Louisiana. Located on a low-relief landscape near an approximately west north west - east south east the western boundary of the Mississippi River delta plain, trending sinuous feature is identified from the seismic time the Five Islands comprise five salt domes aligned in an slices (Figure 6.1a-d). This feature is interpreted as a approximately north west-south east trend, which have fluvial channel. To the southwest of the presumed fluvial pierced and uplifted overlying late Pleistocene meander channel there appears to be another, more elongate salt belt deposits (Autin 2002). The domes are all nearly structure. The fluvial channel can therefore be interpreted circular in plan, surrounded by lowland Pleistocene and/or as occupying a relative topographic low, or even a shallow Holocene delta plain marshes, and attain maximum valley, within the late Pleistocene-Holocene land surface, elevations which range from c. 23 m asl on Jefferson between two relative topographic highs cored by actively Island to c. 52 m asl on Weeks Island (Autin 2002). The upwelling salt. The topographic low may result from the geomorphic impact of the salt domes is most clearly withdrawal of salt in the subsurface, thereby causing the exemplified by Avery Island, where proximal fluvial overburden and land surface to downwarp. Several channels, although modified by engineering in places, apparent meander loops can be identified within the fluvial show an overall sub-concentric pattern, encircling the salt channel, and one meander appears to coincide with one of dome and following the topography closely. the radial faults associated with the salt diapir, suggesting a further tectonic control on channel geometry and evolution. 61  Mapping Doggerland Figure 6.1 (a) Amplitude time slice (0.1 seconds) centred on prominent salt dome with a characteristic concentric reflector pattern. (b) Amplitude time slice (0.1 seconds) with interpretation. A fluvial channel that flows sub-parallel to the salt dome is identified, and it is suggested that salt-related faults have also influenced the direction and geometry of the fluvial channel. (c) Uninterpreted arbitrary seismic line A-A’, confirming that the concentric feature identified in (a-d) is a salt dome Figure 6.2 contains an amplitude timeslice (0.076 seconds) package of mostly sub parallel Pleistocene reflections, the and seismic section. These show an apparently north thickness of which varies laterally. In several places along westerly draining network of channels, which flow the line the reflections within the Pleistocene succession towards the eastern end of the OSP. The first-order terminate against packages of chaotic and complex tributary to this drainage network overlies the axis of a reflections, which most likely represent infilled tunnel salt-cored anticline (Figure 6.2b). A seismic profile valleys. The base of the Pleistocene succession is marked through the channel shows that the channel body directly by an approximately horizontal unconformable surface, overlies a possible collapse graben above the salt swell beneath which is the steeply dipping, truncated, pre- (Figure 6.2c and d). This observation suggests that the Pleistocene succession. Here, the pre-Pleistocene formation of the collapse graben led to the development of succession has been folded above an upwelling body of a topographic depression in the land surface that was salt into an asymmetric anticline, the hinge of which has exploited by the fluvial channel. been truncated. Near the truncated hinge of the fold the Pleistocene succession shows a marked thinning. This is More conclusive evidence for the direct control of collapse indicative of growth of the fold during Pleistocene times graben on late Pleistocene-Holocene fluvial systems is (Figure 6.3b). At the hinge itself, the seafloor has a presented in Figure 6.3. This figure contains part of an striking horst-graben-horst type morphology (Figure 6.3c east -west trending shallow seismic (sparker) profile, and d). This suggests that the overburden above the salt- 81/03/53, located in the western part of the OSP. Sparker cored fold has begun to collapse. The fact that the seafloor profiles provide higher resolution data compared to itself shows such a spectacular collapse, graben-style, offshore 2D and 3D seismic reflection surveys. suggests that this overburden deformation has occurred Immediately below the seafloor reflector is a thin package within the very recent past, and may indeed be continuing of Holocene reflections. These are draped over a thicker through to the present day. 62  Mapping Doggerland Figure 6.2 (a) Amplitude time slice (0.076 seconds). (b) As (a) but with interpretation; late Pleistocene-Holocene fluvial channels signified by yellow arrows, Holocene sand waves shown in purple, with the green annotation signifying a Pleistocene tunnel valley. (c) Seismic section. (d) Part of seismic section (c) with interpretation 63  Mapping Doggerland Figure 6.3 (a) Part of BGS sparker profile 81/03/53 that trends east-west through the OSP. (b) Sparker profile 81/03/53 with interpretation. Seafloor reflector and seafloor multiples shown in red. Tunnel valleys shaded in green, Pleistocene sediments in purple and pre-Pleistocene sediments in yellow. Reflections within the Pleistocene succession are picked in dark purple. Within the pre-Pleistocene succession discernable reflections are marked by dark blue picks, whilst diffractions (which mark the radial scattering of incident seismic energy) are shown in light blue. (c) Close up of boxed region indicated in (b). (d) Close up of collapse graben with interpretation, showing channel development within collapse graben. 64  Mapping Doggerland These observations provide convincing evidence for structure locally influenced late Pleistocene-Holocene neotectonic activity driven by salt diapirism. Examination drainage patterns, with rivers either flowing around the of the Pleistocene succession within the collapse graben relative topographic highs above near surface salt domes, reveals a series of characteristic seismic facies units, with or flowing along the topographic depressions created by sequences of sub-horizontal reflectors downlapping onto collapse structures in the immediate overburden above inclined reflections; these are interpreted as channel fill actively upwelling salt structures. A shallow seismic deposits. It seems apparent therefore that the collapse profile from the OSP (81/03/53) provides spectacular graben identified here was active during late Pleistocene evidence that recent salt movement leading to collapse times, and moreover, that several fluvial channels were graben formation has not only controlled the location of flowing along the axis of the graben. late Pleistocene-Holocene fluvial systems, but also directly controls the morphology of the seafloor. It is clear 6.3 Conclusions therefore that halokinetic activity, which dominates the structural fabric of the SNS basin, has also played a Salt tectonics is responsible for some of the late critical role in the late Pleistocene-Holocene geomorphic Pleistocene-Holocene geomorphology of the SNS. 2D and evolution of this region. 3D seismic reflection datasets provide evidence that salt 65 Mapping Doggerland 66  Mapping Doggerland 7 An Atlas of the Palaeolandscapes of the Southern North Sea Simon Fitch, Vincent Gaffney, Kenneth Thomson with Kate Briggs, Mark Bunch and Simon Holford 7.1 Introduction The results of this process for the whole study area are presented in Figure 7.1. For presentation purposes the The preceding papers in this volume have provided the detail of the results have been presented as four quadrants essential background to the results of the NSPP mapping reflecting the underlying BGS geological maps. When programme. Here, we need only be concerned with a de- considering these maps, it should be stressed that the pri- scription of the analytical processes that led to the identifi- mary intent of the atlas is to provide information on fea- cation of features within the available seismic data and a tures that had a clear physical expression upon the Holo- description and interpretation of these features. A provi- cene landscapes under review, in many cases earlier fea- sional interpretation of the results will be presented in the tures could be identified. final chapter. However, these were not consistently mapped, except on The analysis presented here utilised the top 0.5s of the those occasions when earlier features were reflected within SNS Mega Survey provided by PGS for use within the the later Holocene landscape. For instance, deeply in- NSPP. The Mega Survey comprises a series of surveys cised, earlier structures can be demonstrated to form ba- that have been pre-balanced, stacked and migrated into a sins for lacustrine features and to constrain later channel single contiguous data set. The intrinsic spatial qualities development. Despite this, older features which were of the data permit the inclusion of supporting data sets to mapped, but which had no relation to the Holocene land- assist interpretation (see Thomson and Gaffney and Fitch scape, are presented in Figure 7.3 for the sake of com- et al, this volume). For this purpose, high-resolution pleteness. boomer surveys acquired for the examination of the Qua- ternary sediments of the North Sea were sourced from the A further point should be made in relation to topographic British Geological Survey (BGS) to assist in the investiga- variation. Aside from features with clear structural integ- tion. In addition, other high-resolution seismic informa- rity, it was also possible to derive general topographic ex- tion and traditional data sources including core log infor- pression for the wider landscape from the seismic data. mation were also integrated to provide a framework within This was recorded where observed, and this provides the which archaeological landscape interpretation could be basis for the general topographic maps for each atlas quad- attempted. rant and for the final overall map (Figure 7.4). It was not possible, given the time constraints of the project and the Interpretation of the very large data set used in the project, extent of the data, to identify extensive surfaces (even and the limited period available for study, necessitated the where possible) and more detailed information on surface division of data between team members. However, to en- topography. This information must, therefore, be used sure consistency, interpretative procedures were standard- with care. ised across the team. This process was facilitated by the use of a consistent colour scheme to represent specific Although the results are presented here as a series of paper features and the results from each area were appraised by maps, this is not an adequate reflection of the richness of each member of the project team (Figure 7.2). the data or the utility of the information as a digital model. Recently, the team at Birmingham has been investigating When a final area interpretation had been achieved, a sec- the potential of quadrant Internet mapping and, specifi- ond evaluation of the interpretation was carried out as a cally, Google Earth format to distribute digital mapping to group exercise. Following this, the interpretation was in- a wider audience (Barratt et al 2007). Aside from archiv- tegrated within the project GIS and assessed against sup- ing with the Archaeology Data Service 1, further informa- porting data. Geophysical timeslice information was also tion on this and interim data releases may be found on the integrated via a series of pseudo-timeslices generated for Project's web site2. specific time intervals. Once created, attribute information for each seismic slice was extracted and applied to the pseudo slice. The resultant information was then extracted as a series of ASCII files containing the necessary X, Y, Z, and attribute information. This data was then used to pro- duce an image of the geophysical timeslice. An RMS Slice was utilised for this purpose, where it was necessary to display an image representing the features within a series 1 http://ads.ahds.ac.uk/ of timeslices. 2 http://www.iaa.bham.ac.uk/research/fieldwork_research_ themes/projects/North_Sea_Palaeolandscapes/index.htm. 67  Mapping Doggerland Figure 7.1 An RMS timeslice covering the whole of the project study area 68  Mapping Doggerland Figure 7.2 Primary features identified within the Holocene landscape of the southern North Sea 69  Mapping Doggerland Figure 7.3 Pre-Holocene features recorded during mapping 70  Mapping Doggerland Figure 7.4 General map of all recorded Holocene landscape features including general topographic interpretation 71  Mapping Doggerland Figure 7.5 The Holocene landscape and features within the northwestern quadrant 7.2 North Western Quadrant identification of a fluvial feature that appears to run di- rectly across the feature. 7.2.1 Description A third structure, (Figure 7.5 C), is located at the mouth of The landscape of the northwestern quadrant displays the OSP depression. This structure is surrounded by rela- strong geological influence. This in part is due to the thin tively flat land. However, this graben collapse structure Pleistocene sediment cover within the area, which allows forms a slight, but distinct, depression in this area. The underlying geological relief to influence the overlying graben collapse is surrounded by slight rises formed by Holocene landscape (Lumsden 1986a). In addition to this, upstanding geology. This area is also likely to have con- four active salt domes in this region are associated with tained a marshy depression during the Mesolithic period, graben collapse features and these form dominant struc- with the upstanding lip forming a slight, but visible, rise in tures within the landscape. The first of these circular the ground level. The location of this depression close to structures (Figure 7.5 A) appears as a depression in the the edge of the OSP suggests that during inundation, this south of this quadrant. It features an upstanding lip on the structure may have bounded a channel joining the OSP to west of this structure, and was formed by an outcrop of the wider marine environment. solid geology that remained upstanding after the collapse of the salt dome graben. In the Holocene, this structure The fourth, and final, structure is smaller than the others. would have created low hills of only a few metres in However this crestal collapse differs in being surrounded height. These would have partially surrounded the main by a clear ring of upstanding geology (Figure 7.5 D). The graben collapse, which would have created a lower area, outer ring is disturbed in places by geological faulting as- possibly containing a marshy area. sociated with the main graben collapse. The upstanding nature of this ring is clearly demonstrated by an adjacent The second structure (Figure 7.5 B) appears in close prox- fluvial feature that is clearly channelled around the struc- imity to the previous feature. The expression of this struc- ture (Figure 7.7). This circular structure, however, pos- ture is very slight, and it is possible that this had little visi- sesses several interesting properties. ble impact on the landscape. This is also suggested by the 72  Mapping Doggerland Figure 7.6 Vertical slice through salt dome exhibiting graben collapse Figure 7.7 The major fluvial channel (blue) can be clearly seen to deviate to respect the topographic rise formed during the Holocene by the underlying salt structure 73  Mapping Doggerland It may be that the basin is surrounded by solid geology observations of recent salt movements in the North Sea which, given the right conditions, may have retained wa- (Holford et al. this volume). The large Dogger Bank ter. If this is so, two possibilities are suggested. The first earthquake recorded in 1931 indicates that the area is still is that this structure retained water and formed a lake that active (BGS 2007). was surrounded by low hills. If so this would have formed a highly attractive environment for hunter-gatherers. An- The underlying geology is also evident in other areas of other option is that it may have contained a general marshy the quadrant. To the far west, several prominent features or wetland area, if the faulting and/or geological perme- may also be observed (Figures 7.5E and 7.8). One is di- ability prevented significant water build up. Unfortu- rectly correlated with the Flamborough Head disturbance, nately, neither scenario can be advanced with certainty. which appears on BGS mapping as being directly exposed Indeed, as the water table was raised prior to inundation a on the seabed. The flanks of this feature are covered by a marshy area may well have become a lake over time. very thin veneer of Pleistocene material. This would have represented a significant Holocene landscape feature and The fact that these salt structures formed enough of a to- appeared as a dominating, but low, ridge extending from pographic expression to produce regional highs and lows the present coastline out into the North Sea. Given the strongly suggests that at least some salt structures within thin sediments within this region it is unlikely that exten- this region were active during the Late Pleistocene and sive archaeological sediments are preserved in the area. Early Holocene. This evidence is not inconsistent with Figure 7.8 Seismic relief image of the Flamborough Head disturbance, note the visible positive relief it imparts to the Holocene landscape 74  Mapping Doggerland Further south, the Holocene landscape would have risen as the offshore regions (<-10m) are therefore likely to be it drew nearer to the present shoreline. Within this area a incised enhancing their survival and detection. series of small, fragmentary and truncated fluvial features can be observed. Their patchy nature is almost certainly a 7.2.2.2 Recent Geological Features reflection of post inundation erosion (Flemming 2002, Cameron 1992). The poor resolution of these features may A series of linear features can be discerned in the centre of also be due to noise caused by the shallow water column the study area, near the outflow of the OSP Lake. Their in this area. position in the upper sections of the seismic section and their structure revealed that these features were reflections The dominant feature in the northeast of this quadrant is of large sand waves on the seabed surface that are of re- the western end of the OSP (Figure 7.5 F and Figure 7.9). cent origin (Lumsden 1986b). A distinct fluvial feature can be seen running west north west from the OSP towards the topographic depression in 7.3 North Eastern Quadrant between Flamborough Head and the Dogger Bank. This feature may be a channel flowing from, or even feeding, 7.3.1 Description the lake that must have filled the OSP during the early Holocene. This channel was certainly active prior to the The northeastern quadrant of the survey area has provided inundation of this region c. 9.5 Ka BP (Shennan 2000). A one of the most complete pictures of the emergent similar channel was advocated by Coles (1998) in her map landscape of the SNS. This area was mapped as part of a of the region, although at the time no evidence was avail- pilot study prior to this project, and the results of that work able to support such a proposal. In any case, the presence have been confirmed and enhanced during this larger of such a channel suggests that the OSP contained a sig- exercise (Fitch et. al. 2005). The level nature of this nificant freshwater body prior to marine inundation that landscape largely reflects the presence of deep Late would have been very attractive to hunter gathers. Pleistocene sediments within the region. The area possesses a topographic high over the area of the Dogger With respect of the OSP itself, it is clear that the bulk of Bank that gently descends into the lower lying plain this structure has suffered from Early Holocene marine surrounding the OSP. One minor topographic high can be erosion (Figure 7.10). Large scour marks are clearly visi- observed in the southwest of this quadrant, located ble within the depression, and the seismic data clearly approximately over the Outer Well Bank (Figure 7.11 B, shows truncation of deposits. This observation suggests Laraminie 1989). This topographic high is related to a that any lacustrine deposits that remain within this feature facies change within the Late Pleistocene deposits, and is are likely to be intermittent. Coastlines at the edges of the due to a change in depositional environment during the Outer Silver Pit are characterised by a strong response and Late Pleistocene (Laramine 1989). However the dominant appear as distinct boundaries. Tidal scour marks are also topographic feature in the area remains the OSP, which visible and reflect differences in tidal flow. Although no forms a significant depression in the south of the quadrant clear dating evidence is available, isostatic models suggest (Figure 7.11 C). that this coastline was active at around 9,500 BP (Shennan 2000). Within this quadrant the predominant trend of all the fluvial systems is to the southeast. These drain the area of 7.2.2 Other Features the Dogger Bank, to the north, and converge on the OSP, (Figure 7.11 D, A). All of the Holocene fluvial features 7.2.2.1 Solid Geology can be seen to be incised into the underlying Late Pleistocene Dogger Bank Formation. This relationship The underlying bedrock, along with associated faults, is demonstrates that these features are likely to have been seen clearly within the upper timeslices of the data. Al- latest Pleistocene or Early Holocene in date. The majority though there are clear indications that the solid geology of of these features are highly developed sinuous systems this region is near the surface, not all of this need have a with a high stream order. The geographic location of these topographic expression. Indeed the BGS maps for this systems, in relation to the early Holocene topography, region record that late Pleistocene cover in this region is suggests that they were sub-aerially exposed for a longer very thin over much of the area (Laraminie 1989). This period than most of the survey area and that the systems allows the strongly reflective bedrock to swamp the signal are better developed as a consequence. A number of of all but the largest of Holocene features. Thus it is likely abandoned meanders can be observed, in association with that this area may contain features that could not be identi- a developed main central channel, and these join the coast fied during mapping. This observation must also be tem- via a well-formed estuary. The latter feature can just be pered by the observation that it is possible that active ero- seen through heavy striping within the dataset. sion may have effectively truncated the Holocene deposits and features within in this area. The remaining features in  Mapping Doggerland Figure 7.9 Western end of the OSP lake showing outflow channel and associated topographic highs and lows 76  Mapping Doggerland Figure 7.10 A seismic line across the OSP. Pronouced scouring (red line) occured during the early Holocene marine transgression of the area . The surface is overlain by later early Holocene marine sands and muds. (Data provided by the University of Bremen) 77  Mapping Doggerland The abandonment of channels is presumably a response to gatherer groups within the landscape and provided a wide changes in the fluvial regime. In this case, the most likely variety of hunting and gathering opportunities. cause is a response to regional sea-level rise. The final landscape features in this quadrant that require description As the fluvial systems progress down to the OSP, clear are a series of bulbous basins that occur next to a coastlines are observed framing the OSP itself (between B prominent fluvial system referred to as the "Shotton and D (Figure 7.11). These are characterised by a strong River” (Figure 7.11 E, Fitch et. al. 2005). Initially response and are often accompanied on the seaward side recorded during the pilot project these basins have been by tidal scour marks. A number of rivers meet the interpreted as wetlands or lakes. The correlation between coastline in this area, widen and form small estuaries these features and an underlying tunnel valley within the (Figure 7.12). The coast (or lakeside) is clearly defined, seismic dataset suggests that the earlier depression may located between the -40 metre and -50 metre contours and have been filled with impermeable glacial fill material coincide with the contemporary outline topography of the leading to lake or marsh formation. Again, such wetland OSP. systems could have been gathering points for hunter Figure 7.11 The Holocene landscape and features within the northeastern quadrant If we utilise the sea level curves provided by Jelgersma Unfortunately, the profusion of sea level curves for this (1979), Shennan (2000) and Peltier (2004) then an region, combined with the difficulties of relating such approximate age between 9,500 BP and 8,500 BP is models to the real world (Bell et al. 2006), demands that suggested. an accurate age for this shoreline awaits the recovery of suitable, dateable material. 78  Mapping Doggerland Figure 7.12 The junction of rivers and coastline can clearly be seen in the inset seismic image. The river channels can be observed to widen and form small estuaries Figure 7.13 A series of tunnel valleys (green outline) crossing the OSP and observed to underlie the Holocene landscape (coastline marked in orange)  Mapping Doggerland Figure 7.14 The location of one of the small structures that resembles a palaeochannel. The origin of these features remains to be determined 80  Mapping Doggerland Figure 7.15 Modern Sandwaves directly overlying the Holocene landscape However, this area would have provided a diverse and productive environment for the occupants of the It is likely that these formed small extensions of tunnel landscape. The OSP during this period would have hosted valleys located within this area; although it is equally pos- a variety of intertidal and estuarine environments within sible that these are remnants of a more recent palaeoland- which food would have been relatively abundant. scape. Unfortunately the origin of these features could not be determined utilising the available data, and the mode of Two large prominent ridges can be observed in the formation and age of these structures remains undeter- southern central section of the quadrant (near C in Figure mined. 7.11). These elongate ridge features, have already been discussed in an earlier paper in this volume and were 7.3.2.2 Recent Geological Features interpreted as moribund sand banks that formed in as estuarine environment during the early Holocene marine No features of recent geological origin were observed in transgression (see Briggs et al. this volume; Shennan the seismic data within this region, however 2D BGS 2000). From this it is inferred that the tidal range of the seismic lines do show that minor sand structures are OSP was probably macro-tidal and the tidal currents located in this area (Figure 7.15). conveyed in the estuary were relatively strong. 7.3.2 Other Features 7.4 South East Quadrant 7.3.2.1 Solid Geology 7.4.1 Description A series of tunnel valleys can be clearly be observed underlying the Holocene landscape and crossing the OSP Few major topographic features were identified within the within this quadrant (Figure 7.13). These features do not area (Figure 7.16) and the majority of the surface area appear to have any topographic expression, but are visible within the quadrant was determined to be extremely flat due to the absence of the Later Pleistocene deposits in this (Figure 7.17). Although the significant data striping pre- depression. These are directly related to features observed sent in the southeast of this quadrant hindered interpreta- on BGS mapping ascribed to the Middle Pleistocene tion (Figure 7.18), data quality is reasonable elsewhere, Swarte Bank Formation (Laramine 1989). Due to time and little significant landscape variation is discernable. and focus constraints only those features that were The southern margin of the OSP is visible in the northern significant to the interpretation of the Holocene landscape most part of the quadrant (Figure 7.16 B), along with asso- were digitised. ciated intertidal features. A significant depression, which retains a bathymetric expression today, can be observed to There are, however, a number of small features that the north. However, the majority of the fluvial features resemble a series of palaeochannels. These are located within this region run across a large and relatively flat within the Outer Silver Pit, incised into Lower to Middle plain. The seismic signal in this area generates a "mot- Pleistocene deposits, and are covered by recent sediment tled" appearance, the origin of which is uncertain (Figure (Figure 7.14). 7.18). A few minor fluvial channels can be observed 81  Mapping Doggerland within this mottled zone (Figure 7.16 A). Although de- Boulders Bank Formation, can be observed to terminate at tailed interpretation within this area is problematic the the Early Holocene coastline. These relationships suggest area, although undistinguished in topographic terms, may that this feature dates to the Late Pleistocene or Early contain significant archaeological potential. The BGS Holocene, and strongly suggests that this feature may have mapping for this region, for example, records an extensive been a drainage channel for this depression. The depres- coverage of the archaeologically important early Holocene sion may have contained a lake during the early Holocene. Elbow Formation over the plain. This interpretation is supported by the BGS mapping for the area that records the presence of Late Weichselian to Where fluvial features can be defined within this quadrant Holocene glacio-lacustrine deposits throughout this de- they can be divided into two groups. The first group is pression (Brown 1986). located to the northwest of the quadrant and flow from the southwest to the northeast (Figure 7.16 C). These features Further to the north, in an area bound by Markham's Hole, run directly across Holocene floodplains but have little the Botney Cut and the OSP, are a series of channels trav- observed sinuosity. The project was unable to resolve the ersing a low-lying area (Figure 7.16 F). The area itself is channels located within these floodplains. However, given clearly defined but does not possess any specific character- the topography, it is likely that the channels have a higher istic other than the presence of these channels. This might degree of sinuosity than is visible within the data. Adja- suggest a delta system (Figure 7.20) however there are no cent to the coast, a clearly defined and developed estuary structures visible in the seismic data to support this inter- can be seen at the termination of the fluvial channels (near pretation. Its relatively flat prospect and position, adjacent Figure 7.16 B). The character of these estuaries suggests to one of the OSP marine inlets, suggests that the area that their formation has been controlled by a series of in- might have been a salt marsh for part of its history at least. undations during a period of rising sea level. The channels appear to show two distinct formation phases. A primary, slightly larger, channel may be older, The second group of features is located in the southeast of whilst a series of smaller channels may form a network the quadrant and trend southwest - southeast. They may over and around this structure at a slightly later date. Once flow towards a location, suggested by Coles (1998), to again, this sequence may be the consequence of rising sea contain a deep-water channel (Figure 7.16 A, D). Unfor- levels. tunately, the noise within this region is enough to hinder the resolution of these features. In one area, near that The OSP coastline is still well pronounced in this area, studied by Praeg (1997), a clear and well-developed sinu- although the basin appears to widen as it nears the area of ous channel system can be discerned. This suggests that if the postulated salt marsh. The coastline is characterised the issue of noise could be overcome, possibly following by a strong seismic response although tidal scour marks, access to newer surveys, the route of these channels could which are clearly visible elsewhere, are not clearly defined be resolved. in this area. This might reflect a difference in tidal flow at this point. A deeply incised inlet can be seen in the inter- The most significant landscape feature within the quadrant tidal area, adjacent to the postulated salt marsh. The origin is the large depression that forms Markham's Hole (Figure of this feature is a partially filled glacial tunnel valley, 7.16 E and Figure 7.19). Located in the northeast of the which was inundated during the marine transgression at quadrant, this large, partially in-filled valley retains a around 9,500 BP. bathymetric expression. The seismic data actually reveals that this feature is much deeper than the bathymetry sug- 7.4.2 Other Features gests (Figure 7.19). 7.4.2.1 Recent Geological Features BGS cross-sections for the area, along with 2D seismic data made available to the project, suggests the existence In the west of the study area are a series of very small lin- of significant deposits within this structure. These depos- ear features. Their position, high in the seismic column its can be directly related to the Late Pleistocene Botney and their structure revealed that these features are reflec- Cut Formation and are directly overlain by sediments of tions of large sand waves on the seabed surface and are of recent origin. A channel system attached to the end of this recent origin. tunnel valley, which is incised into the Late Pleistocene 82  Mapping Doggerland Figure 7.16 The Holocene landscape and features within the southeastern quadrant 83  Mapping Doggerland Figure 7.17 Seismic line across the southeastern quadrant. Little landscape variation can be discerned in this area Error! Figure 7.18 A representative image of "mottling" within the seismic data. Whilst minor channels can be discerned, the clarity of the interpretation is hindered. 84  Mapping Doggerland Figure 7.19 Cross section (BGS line 80-01-05) over Markham's Hole. This clearly shows the deep incision of this feature (location marked by red line on inset). The inset timeslice of this feature shows the location of the outflow channel (marked orange) Figure 7.20 Seismic timeslice image of the palaeochannels within an area interpreted as a salt marsh 85  Mapping Doggerland 7.5 South Western Quadrant Geological beds pushed to near-surface positions by the crests of deep underlying salt structures form these slopes. 7.5.1 Description The result of this process would have been a landscape comprising a broad, low valley bounded on either side by The landscape which has been observed within the seismic gentle slopes. Given that the majority of fluvial features data for this quadrant reflects the variable influence of the observed within the area appear to drain into this valley underlying geology upon the observed siesmic signal. (Figure 7.21 C, D), it is suggested that this area formed a Areas in which the geological influence is weak are gener- low lying wetland plain, occupied by fluvial systems. ally associated with the presence of deep Late Pleistocene Within the valley, one large upstanding feature can be ob- deposits. However, the overall character of this landscape served (Figure 7.21 E). This indicates the presence of an is as a relatively gentle plain sloping from the modern underlying salt dome that has gently raised the geological coastline onto a lower plain in the north and west of the beds within this area. This would have formed a low hill quadrant (Figure 7.21). within a relatively flat area. Such a hill might represent an important locale for hunter-gatherers. Aside from potential Beginning in the east of the quadrant, the results suggest a for settlement, game could have been observed from this relatively flat lying terrain, with a number of large, inter- vantage point as it migrated up the valley (e.g. Fischer spersed depressions. These depressions represent unfilled 2004, 34). tunnel valleys that may have contained lakes due to their pre-existing low lying topography and basin-like nature. Slightly to the north is a salt dome (right of Figure 7.21 D) This is supported in part by the presence of lacustrine de- where the break of slope coincides, again, with underlying posits relating to the Late Plesitocene and Early Holocene geological formations. A rise is located within the north within sediments recovered from some of these depres- of this feature and relates to the peak of the underlying salt sions (Cook 1991). However later erosional events makes dome. This feature is considered as indicative of a rela- the calculation of the full lacustrine extent problematic tively raised area in comparison to the valley. This would (Cook 1991). The smaller of these basins probably corre- have formed a distinctive topographic high within the con- sponds to the Well Hole (Figure 7.21 A). However, the temporary landscape. shape of this feature only partially corresponds to the fea- ture observable within the 3D seismic data. A cross sec- The large depression associated with the Inner Silver Pit is tion through this feature was provided by an additional located to the southwest of the hill mentioned above high-resolution 2D digital line (93-01-81) provided by the (Figure 7.21 F). The seismic data is, unfortunately, heav- BGS (Figure 7.22). This 2D line reveals a partly eroded ily striped in this area. Despite this, the broad outline of deposit, which directly overlies the erosion surface but is the feature can be determined and available geological beneath modern sediments. Similar results are recorded mapping for the area suggests that the depression is heav- on the BGS maps for the area, which describe the deposits ily incised and that all sediment has been removed from within this feature as pertaining to the Botney Cut Forma- within the bulk of the feature. Whilst BGS mapping indi- tion. This deposit dates from the latest Pleistocene/earliest cates an absence of Botney Cut Formation deposits in the Holocene and is associated with a glacio-lacustrine origin. area it is possible that patches of sediments may remain Given that no obvious outflow is observed, this tunnel within the Inner Silver Pit. Whilst it is impossible to de- valley presumably formed a lake within the surrounding termine if this depression formed a lake or wetland during flat lying landscape. the early Holocene, the feature is so pronounced today that it is extremely unlikely that it had no topographic expres- The other major depression within this plain corresponds sion during the early Holocene. On that basis, it seems to the Sole Pit (Figure 7.21 B). The available high- reasonable to suggest that the feature would have con- resolution 2D seismic data reveals that the base of this pit tained a lacustrine environment of some sort. A series of has also suffered significant erosion, removing any depos- palaeochannels can be seen to emerge from this feature its of Late Pleistocene/Early Holocene age. Minor depos- (between F and C on Figure 7.21). Unfortunately, noise its of this material do occur on the flanks of this feature. It within the data renders it impossible to determine if these is suggested that this feature would have formed a similar features are related to the Holocene or earlier landscapes. lacustrine feature to 7.21 A. Holocene erosion within this feature ensures that confirmation of such an interpretation The southwestern quadrant is associated with three princi- is likely to be problematic. pal groups of fluvial systems. The first, located in the northwestern area (Figure 7.21 G), may be equated to Further to the west of the quadrant, significant landscape those described in the northwest quadrant. They comprise features reflect the influence of underlying solid geology a series of small, truncated channels that appear to be gen- (Figure 7.21 C). This is indicated by the clarity of solid erally poorly preserved. Although some larger sections geological structures within the seismic data and their in- provide a clearer response their fragmentary nature is al- fluence upon the early Holocene landscape. The seismic most certainly a product of erosion. The second group data indicates a slight, but significant, depression located includes those features that are a continuation of channel in the southwestern corner of the quadrant. This is systems observed in the southeastern quadrant (Figure bounded on either side by low but significant scarp slopes. 7.21 H). These also drain towards the northeast, and a few 86  Mapping Doggerland abandoned meander bends are associated with these chan- these riverine features, they may well have provided an nels. They appear large and well developed and a cross important route through the landscape, and to and from the section through one indicates a substantial sedimentary coast (Barton and Roberts 2004, 352). profile (Figure 7.23). Given the size and significance of Figure 7.21 The Holocene landscape and features within the southwestern quadrant 87  Mapping Doggerland Figure 7.22 Cross section through Well Hole (BGS Line 93-01-81. Note the preservation of the Botney Cut Formation (base marked red on interpreted section) 88  Mapping Doggerland Figure 7.23 Cross section through the large channels in the southwestern quadrant. (BGS Line 93-01-74A) The third group of features is located in the far northeast- observed beneath the Holocene land surface (Figure 7.24). ern corner of the quadrant, adjacent to the coastline Although this has little observed impact upon the overly- (Figure 7.21 I). These features are similar to those ob- ing Holocene landscape, the feature may have archaeo- served in the northeastern quadrant and display dendritic logical significance as it may contain deposits relating to tributaries with well-developed floodplains. the earlier Palaeolithic occupation of this landscape. This group displays evidence for change in sea level indi- 7.5.2.2 Recent Geological Features cated by the abandonment of part of the channel and the overlying formation of an associated estuary or tidal flat. Four main sets of sand ridges were observed in the west of The estuary has a clear bathymetric expression within the the study area. These sand ridges cluster as a series of BGS Digbath250 dataset, which suggests that any archaeo- parallel lines and are reflected in the upper sections of the logical deposits may be near the surface. This area could seismic column. These are interpreted as large seabed, well represent a location with significant potential for fu- sand waves of recent origin. ture palaeoenvironmental sampling. 7.6 Conclusions 7.5.2 Other Features The mapping and recording of such an extensive archaeo- 7.5.2.1 Fluvio-Glacial features logical landscape has been a daunting task. The primary statistics relating to the extent of features recorded during A series of channel structures were observed to lie beneath the project is provided in Table 7.1. Holocene features (Figure 7.23). In the west of the quad- rant, and near the location of the later depression at Figure 7.21 B, a large fluvio-glacial outwash plain may also be 89  Mapping Doggerland Table 7.1 Basic quantitative data relating to identified landscape features Coastline Length Observed 691 km Marine Area Observed 1791 km2 Lakes/Wetlands Observed 24 Salt Marsh Area 309 km2 Intertidal Zone area observed 293 km2 Major Estuaries Observed 10 Total Fluvial Stream Length 1612 km Fluvial Related Features Observed 305 Number of Stream Segments 719 Total Area covered by Fluvial Features 526 km Mean Strahler Order 1.52 Mean Shrever Order 3.64 Average Angle of Stream Join 68 degrees In considering the results of the NSPP mapping exercise it be considered to resolve specific issues or enhance inter- should be stressed that the interpretation provided here is pretation. historic in several senses. The 3D seismic data was ac- quired over a period of 20 years at least. The PGS Mega In conclusion, it seems reasonable to stress that the quanti- survey does not, therefore, represent a snap shot of any tative data and associated mapping presented here repre- single period of time and the information presented here sents a quantum shift in respect of our knowledge of the may not represent the full effect of modern changes to the Holocene landscapes of the North Sea. The study has preserved elements of the landscape. However, the effects revealed a hunter gatherer landscape that is, currently, of burial and erosion across such a vast landscape are without parallel in Europe and, moreover, may prove to be likely to be relatively small given the time period under an optimum area of settlement during the Early Holocene. consideration. As such, the general thrust of this chapter Interpretation of this landscape in terms of habitation po- remains valid and the data can, with some caution, be used tential and sediment survival will undoubtedly affect our in support of larger archaeological synthesis and for man- understanding of the regional archaeology of all the coun- agement purposes. However, it is clear that the results of tries bounding the survey area. Whilst consideration of the this mapping will require future ground truthing. Coring is wider archaeological significance of the data will be pre- required to provide samples for palaeoenvironmental study sented in the final paper of this volume, the results pre- and dating, whilst higher resolution survey might usefully sented here represent a major achievement and reflect the combined efforts of all involved in the project. 90  Mapping Doggerland Figure 7.24 An image showing the complex structure of a glacial outwash plain. A cross section through this feature (Line A-A') shows that this feature is located deep beneath the seabed (marked on in green) 91 Mapping Doggerland 92  Mapping Doggerland 8 The Potential of the Organic Archive for Environmental Reconstruction: An Assessment of Selected Borehole Sediments from the Southern North Sea. David N. Smith, Tom C. B. Hill, Ben R. Gearey, Simon Fitch, Simon Holford, Andy J. Howard and Christina Jolliffe 8.1 Introduction 8.2 Potential and Rationale Prior to the inundation of the region, during the eustatic In the Holocene, river valleys and associated floodplain sea level rise of the early Holocene, the landscape of the wetlands of the scale recorded in the 3D seismic survey North Sea basin would have presented early human settlers data are widely recognised as key locations for the preser- with a range of ecosystems, resources for food and shelter, vation of environmental and biological archives (Brown as well as barriers that restricted their movement. 1997; Howard and Macklin 1999). Classic ‘landward’ Therefore, the application of appropriate environmental examples of this are recorded by a range of studies from archaeological methodologies, that help to elucidate the Great Ouse (Dawson 2000), the Severn (Brown 1983), signals of both natural and anthropogenic landscape the Thames (Sidell et al. 2000) and the Trent (Knight and change, within a well constrained chronostratigraphic Howard 2005). These sedimentary archives are usually framework, will be integral to the development of any associated with abandoned (in-filled) river channels, wet- archaeological research framework for the North Sea (Bell lands and associated sand and gravel splays, with some and Walker 2005). spanning considerable periods of time (Brayshay and Din- nin 1999; Smith et al. 2006). However, despite relatively The identification of peat deposits in the present intertidal abundant evidence for preservation, the majority of land- zone fringing the North Sea Basin (e.g. Horton et al. 1999) ward archives are associated with middle to late Holocene and further offshore (e.g. Shennan et al. 2000) palaeoenvironments (e.g. Greenwood and Smith 2005; demonstrates the potential for organic preservation and Howard 2005) and reworking of the floodplain may be a associated palaeoenvironmental reconstruction using a key factor in the under-representation of earlier Holocene range of proxy indicators. The analysis and interpretation organic records. In terms of the southern North Sea, it is of 3D siesmic survey data as part of this project has probable that the channel deposits identified during the identified a range of natural sedimentary traps capable of analysis and interpretation of the 3D seismic survey data containing further environmental remains, such as also contain valuable archives of this type, and further- palaeochannels and floodplain wetlands; however, these more, have the potential to fill this missing gap within the need to be ‘ground truthed’ in order to prove their early Holocene. environmental potential. Such an approach will allow previous interpretations of landscape evolution and Within the perimarine zone, the potential of intertidal archaeological potential of the region to be assessed (e.g. sedimentary archives for environmental reconstruction, Coles 1998; Flemming 2004; Shennan et al. 2000; including elucidating the nature of sea level change has Shennan and Horton 2002) as well as aid the development been clearly demonstrated by recent studies (e.g. Bell et al. of new models. 2000); therefore, the potential of now submerged intertidal areas is no less. There is also clear evidence that many Logistically, obtaining environmental samples from early and middle Holocene ‘peats’ from modern intertidal beneath the southern North Sea presents many obstacales. areas have suffered erosion during periods of later Holo- For example, delimiting the position of suitable sediment cene sea level change (Bell et al. 2000), which has clear traps and the significant expense of commissioning marine implications for the quality of the archive. Furthermore, sampling are two of the most obvious issues. However, along areas of the northern British coastline there is grow- preliminary enquiries indicated that the British Geological ing evidence that extreme events such as the ‘tsunami’ Survey (BGS) has an archive of Late Pleistocene- associated with the Storegga landslide may have resulted Holocene material, which they have recovered from shal- in the erosion of environmental archives (Boomer et al. low boreholes and vibrocores. Therefore, the aim of this 2007). Though the effects of the Storegga landslide on the part of the project was to assess whether any of the contemporary coastal margins of the central North Sea are existing sedimentary material stored at BGS offices not clear, it is unlikely that late Holocene and modern ero- (National Geoscience Data Centre, Keyworth; Murchison sion will have affected the buried landscapes recorded in House, Edinburgh) could be used to provide meaningful the seismic survey. Effectively, this means that the central enviromental evidence from natural features identified in North Sea may actually have the potential to contain some the 3D siesmic survey. the best-preserved Early Holocene palaeoenvironmental records for Northern Europe. 93  Mapping Doggerland Figure 8.1 Location of the boreholes in the area included in the 3D seismic survey 94  Mapping Doggerland Figure 8.2 Location of the boreholes requested from the area and primary features identified during mapping 95  Mapping Doggerland Figure 8.3 Location of the boreholes inspected from the area included in the 3D seismic survey 96  Mapping Doggerland 8.3 Core Selection 8.4 Palaeoenvironmental Assessment Two groups of site-investigation material are held by the BGS in low-temperature storage facilities. The first com- 8.4.1 Sampling prises a series of shallow borehole cores drilled from the 1960s onwards as part of a program of systematic mapping Sampling of the remaining five borehole cores occurred on of the offshore geology of the United Kingdom Continen- the 15th June 2006. After discussions with curatorial staff tal Shelf (UKCS). In total, twenty-nine shallow boreholes at the BGS, it was agreed that sampling quotients should were drilled within the UKCS quadrants that cover the be limited to a maximum of four samples per 1 metre sec- area of the 3D seismic survey (Figure 8.1). In addition, tion of core and the size taken should be limited to 50g of vibrocore samples were acquired over this extended time material. In total, fifty samples were obtained from three period to aid in the drafting of 1:250,000 seabed sediment vibrocores (six samples from 53/02/395, nineteen samples maps. from 54/02/215 and twenty-three from 54/02/80) and a single sample and a wood fragment from one shallow Initially, a sample of nineteen shallow cores and vibro- borehole (81/50). The borehole details, sample numbers cores were requested from the BGS (Figure 8.2). The se- and sample depths are summarised in Table 8.1. lection of these was based on two factors: (1) their proxim- ity to landform features (i.e. sedimentary traps) that had 8.4.2 Visual Assessment been identified during analysis and interpretation of the 3D seismic data; (2) indications for the presence of ‘peat’ in Samples from vibrocore 53/02/395 (samples NSP01-06) these borehole records. were predominantly medium grained, light yellow-brown sands with a relative abundance of disarticulated shell 0f the (nineteen) cores requested, six were not available fragments. Samples from vibrocore 54/02/215 (samples and four existed only as paper archives. As a result, only NSP07-25) were fine to medium grained, light yellow- four shallow boreholes (79/01, 81/49, 82/21, 89/05) and brown sands, which appear to become slightly darker yel- four vibrocores (53/02/395, 54/00/205, 54/02/80 and low-brown with depth, possibly in response to an increase 54/02/215) from the area under study were available for in humic organic content. There also appeared to be a examination (Figure 8.3). Given the low numbers of cores gradual reduction in abundance of disarticulated shell available, it was therefore decided to also request shallow fragments with depth. Samples from vibrocore 54/02/80 core 81/50, drilled to the south of the area of the 3D were fine-grained, light yellow-brown sands with occa- seismic survey (because the field descriptions suggested sional disarticulated sand fragments (although in lower that it had organic inclusions). Between the 21st and 24th abundance than found in the previous two vibrocores). of March 2006, one of the authors (Simon Holford) recorded the nine selected cores at Edinburgh, describing 8.4.3 Assessment of macrofossil (insect and plant) their sedimentary characteristics, condition and potential inclusions for further environmental analyses. Previous research had identified the ‘Elbow Formation’; a unit of fine-grained Initially, it was intended to use the assessment procedures muddy sands and interbedded clay of early Holocene date, for insect and plant remains outlined by Kenward et al. which is mappable in the eastern and southern part of the (1986) and to assess the degree of preservation following project survey area (Cameron et al. 1992; Ward et al. Kenward and Hall (2006). All forty-nine samples of mate- 2006). the key to further work, therefore, was to identify rial from the boreholes were rapidly scanned under a low- the presence of organic (peat) sediments that might be of power binocular microscope and no organic macrofossils Holocene age. were observed. The only exception to this was a small fragment of poorly preserved wood from borehole 81/50 In terms of quality of storage and general degree of pre- (NSP49). Given that the wood came from unconsolidated servation, the core samples were found to be in a good sands, its context was poorly constrained and the pollen condition given their age (the shallow boreholes are named assemblage from the sample was poorly preserved it was after the year and order in which they were drilled i.e. decided not to date this fragment. 79/01 was the first borehole drilled in 1979). However, some surfaces of the cores were covered in mould (sec- 8.4.4 Pollen Assessment tions 89/05 and 54/02/215) and all were largely desiccated. Both of these conditions inhibit the preservation of organic In total, sixteen samples were selected for the assessment material and mould growth certainly limits any potential to of plant microfossils (i.e. pollen grains, moss fragments date materials using radiocarbon techniques. Abundant and plant spores). This selection was based on the conclu- shell fragments of a mollusc (possibly Spisula subtrunca- sion that the sediment samples from the basal parts of the ta) were recovered in many of the cores. However, these boreholes and vibrocores afforded the greatest environ- fragments were considered unsuitable for detailed further mental potential, especially since the upper parts of the study, such as stable isotope analysis or amino acid race- cores might be reworked. In addition, the slightly darker mization dating techniques. A full description of the cores yellow-brown nature of some of the deeper sediments was examined and their condition is given in the Appendix. also considered to be an indicator of the potential presence 97  Mapping Doggerland of organic remains, which in turn, increases the likelihood of plant macrofossil preservation. Table 8.1 Summary of sedimentary samples and environmental assessment undertaken from the Southern North Sea vibrocores and boreholes Beetle Plant macrofossil Borehole/core name Sample number Depth Pollen Analysis Analysis Analysis 53/02/395 NSP01 6.68 m * * * 53/02/395 NSP02 7.13 m * * * 53/02/395 NSP03 8.24 m * * * 53/02/395 NSP04 9.13 m * * * 53/02/395 NSP05 10.23 m * * * 53/02/395 NSP06 11.25 m * * * 54/02/215 NSP07 0.2 m * * 54/02/215 NSP08 0.4 m * * 54/02/215 NSP09 0.6 m * * 54/02/215 NSP10 1.2 m * * 54/02/215 NSP11 1.5 m * * 54/02/215 NSP12 1.8 m * * 54/02/215 NSP13 2.2 m * * 54/02/215 NSP14 2.5 m * * 54/02/215 NSP15 2.8 m * * 54/02/215 NSP16 3.2 m * * 54/02/215 NSP17 3.5 m * * 54/02/215 NSP18 3.8 m * * 54/02/215 NSP19 4.2 m * * * 54/02/215 NSP20 4.4 m * * 54/02/215 NSP21 4.6 m * * * 54/02/215 NSP22 4.8 m * * 54/02/215 NSP23 5.2 m * * * 54/02/215 NSP24 5.4 m * * 54/02/215 NSP25 5.6 m * * * 54/02/80 NSP26 0.4 m * * 54/02/80 NSP27 0.6 m * * 54/02/80 NSP28 0.8 m * * 54/02/80 NSP29 1.2 m * * 54/02/80 NSP30 1.4 m * * 54/02/80 NSP31 1.6 m * * 54/02/80 NSP32 1.8 m * * 54/02/80 NSP33 2.2 m * * 54/02/80 NSP34 2.4 m * * 54/02/80 NSP35 2.6 m * * 54/02/80 NSP36 2.8 m * * 54/02/80 NSP37 3.2 m * * 54/02/80 NSP38 3.4 m * * 54/02/80 NSP39 3.6 m * * 54/02/80 NSP40 3.8 m * * * 54/02/80 NSP41 4.2 m * * 54/02/80 NSP42 4.4 m * * * 54/02/80 NSP43 4.6 m * * 54/02/80 NSP44 4.8 m * * * 54/02/80 NSP45 5.2 m * * 54/02/80 NSP46 5.4 m * * * 54/02/80 NSP47 5.6 m * * 54/02/80 NSP48 5.8 m * * * 81/50 NSP50 11.9 m * * * 98  Mapping Doggerland Pollen Diagram for Vibracore 53/02/395 TREES SHRUBS HERBS SPORES . et nd )i te le id m no e ae H a ce iac e ic od ea er iu o ed a e Pt p o d a ( m ea um Er nop ll ac e n ea ta m le e a c yll h e hy yp m l y id m yr a u e ol ue s us ac Po o p h M ce C rn u To g n Po p s s iu Q in u La era C op lP lu la Co u s rc er s s u o ha l ix ry ry tu nu nu bu ax ct i er lm Ca Be Sa Sp Lithology Fr Pt Al Pi Vi U 6.40 6.80 7.20 7.60 8.00 8.40 Depth (metres) 8.80 9.20 9.60 10.00 10.40 10.80 11.20 20 20 40 20 20 20 20 20 50 100 150 % TLP No. of grains counted Figure 8.4 Pollen diagram for Vibrocore 53/02/395 Preparation for pollen assessment followed the standard tions since these were universally low. None of the pollen techniques including KOH digestion and acetylation spectra recovered reaches the recommended assessment (Moore et al. 1991). Due to the high inorganic content of count of 125 let alone the 350 recommended for full all samples, a very small amount of residue remained once analysis. As a result there is no attempt here to use this pollen preparation had been completed. Each sample was data to reconstruct prevailing vegetation or to reconstruct mounted on a 24x40mm coverslip and glass slide for landscape. counting. The pollen assessments were carried out on a Meiji MX5000 microscope at x400 magnification. 8.4.5.1 Vibrocore 53/02/395 (Figure 8.4) To ensure a statistically valid environmental assessment is Six samples (6.68, 7.13, 8.24, 9.13, 10.23 and 11.25m achieved, pollen analysis normally requires a minimum of along the core) were assessed. The low counts of pollen 125 total land pollen grains (TLP) excluding aquatics and recorded preclude detailed interpretation. Pollen occur- spores to be counted for each sample. However, pollen rence was low although increased with depth. Trees and occurrence was very low in all samples, resulting in the shrubs are well represented (80%+ TLP), initially domi- minimum 125 TLP not being achieved in any single sam- nated by Pinus (pine) and Quercus (oak) with small ple. A complete microscope slide was therefore traversed amounts of Alnus (alder) and Betula (birch). Corylus (ha- for each sample in order to assess species abundance, di- zel) increases in abundance vertically through the se- versity and conditions of preservation. Pollen nomencla- quence. Ulmus (elm) and Salix (willow) are recorded at ture follows Moore et al. (1991). trace values. Herbs are recorded in the form of Poaceae (wild grasses), Cyperaceae (sedges) and Chenopodiaceae 8.4.5 Results (fat hen), whilst Pteridium (bracken) and Sphagnum (bog moss) spores are also present, the former increasing down The results are presented as pollen diagrams produced the profile after 9.13m. using the computer program TILIA (Grimm 1991). The diagrams display percentage values rather than concentra- 99  Depth (metres) Depth (m etres) 5.80 5.60 5.40 5.20 5.00 4.80 4.60 4.40 4.20 4.00 6.00 5.80 5.60 5.40 5.20 5.00 4.80 4.60 4.40 4.20 4.00 3.80 3.60 Lithology Lithology Al nu s 20 Al % TLP nu s 20 % TLP Be Pollen Diagram for Vibracore 54/02/215 tu Pollen Diagram for Vibracore 54/01/80 Be F a la tu gu la Fr s 20 ax Pi P i in u nu nu s s s 20 20 Q 40 ue Fr rc 20 ax u s Q i nu ue s Ti rc lia 20 Figure 8.6 Pollen diagram for Vibrocore 34/02/80 us U Figure 8.5 Pollen diagram for Vibrocore 54/02/215 Ti lia lm TREES U S a us l ix lm C C us TREES or or ylu 20 yl s 20 us Vi bu 100 Sa A r rn u li te m Vi x C m is bu yp i a er 20 C rn u ac ar m y C ea he e Mapping Doggerland C op y p hy E r no p SHRUBS er lla ic od a 20 ac c L a c e ia c e a ea c t ae e a e e e C P o uc e SHRUBS he a ae E r no p P l ce a i c od an e a M ta g P l ce i ac yr o an a e e Po tag ae S a iop h l an ac o l x yl ce Pt a n e a P t ifr ag l um ol at er a a e r e ce HERBS ol P o op s P t op s at l y i da e r id a P t p od (m HERBS S p id i u a ( m e r iu o ha m o S p id iu m no l To gn n ol h m et ta um et e) e) T o ag n in lP ta um 20 in de ol lP t. le de 50 n t ol . le n 40 100 60 SPORES SPORES No. of grains counted No. of grains counted 150  Mapping Doggerland 8.4.5.2 Vibrocore 54/02/215 (Figure 8.5) sampling. However, it is probably more likely that it represents sediment reworking, with previously deposited In total, four samples were assessed for pollen from pollen grains being re-eroded and subsequently deposited vibrocore 54/02/215. All samples were taken from within the sedimentary archive. towards the base of the core profile; 4.20m, 4.60m, 5.20m and 5.60m (Figure 8.5). The impression is of mixed dense The overall vegetation reflected is generally similar for all woodland with oak, pine and hazel into which alder seems three sequences, although some differences are observed, to have subsequently expanded. Tree and shrub pollen such as slightly higher values for grasses and sedges in (85%) again dominate the spectra, initially with pine, oak, vibrocore 53/02/395. The dominance of tree and shrub hazel and birch, with lower values for Tilia (lime), Ulmus taxa including pine, birch, oak and alder suggests a (elm) Fraxinus (ash), Ericaceae (heather family) and Salix. Holocene timeframe. Without an adequate radiometric Alnus increases above 4.60m with Fagus (beech) also dating framework, it is impossible to provide a precise age appearing in the uppermost sample. Clearance and but the pollen spectra do suggest an early to middle openings in the woodland are suggested by trace values for Holocene date. For example, lime is absent from these Poaceae (grasses) and Cyperaceae (sedges), along with cores but is relatively common after the mid-Holocene Plantago lanceolata (ribwort plantain), Chenopodiaceae, (Huntley and Birks 1983). Elm, which is recorded in some Lactuceae (dandelions etc.) and Artemisia (mugwort). of the samples, is usually present at higher values in the Spores in the form of Pteridium, Sphagnum and Pteropsida early part of the Holocene (eg. Huntley and Birks 1983). (fern) are present across the spectra. Both the scale and However, given the considerable chance that causes of these clearances cannot be speculated upon. taphonomic/depositional bias, local edaphic factors or problems of pollen production and representation could be 8.4.5.3 Vibrocore 54/02/80 (Figure 8.6) patterning these very small spectra this provisional age estimate should be viewed with caution. Five samples were assessed for pollen from vibrocore 54/02/80. All samples were taken from towards the base The presence of grasses and sedges does suggest the pres- of the core profile; 3.80m, 4.40m, 4.80m, 5.40m, 5.80m ence of open habitats within these woodlands. Whilst (Figure 8.6). Pollen concentrations were again generally these openings in the canopy may reflect areas of wetter low; however despite this, pollen preservation was soil, well known indicators of ‘clearance’ such as ribwort reasonable for most of the samples. Tree and shrub pollen plantain and dandelions, are encountered in vibrocore dominate the spectra (90%+TLP), with Alnus, Betula, 54/02/215 and vibrocore 54/02/80. However, given the Pinus, Corylus and Quercus. Trace values are recorded constraints of the data set, it is perhaps unwise to make for Tilia, Ulmus and the shrubs Salix and Viburnum. Few detailed comment regarding the possible role of human herbaceous taxa are present, although low values for communities in the creation or exploitation of any such Poaceae and Cyperaceae increase slightly towards the top open areas. of the sequence. The only other herbs present are Plantago lanceolata, Chenopodiaceae and Caryophyllaceae (pink family). Sphagnum and Pteropsida 8.6 Conclusions spores are found in low quantities towards the base of the core. There is no doubt that deposits associated with the land- scape features identified in the southern North Sea have 8.4.5.4 Borehole 81/50 significant potential to provide proxy records of past vege- tation, climate and land use history, which can be placed At a depth of 11.90m in borehole 81/50, a wooden frag- within a secure chronostratigraphic framework. Unfortu- ment was observed within the sedimentary profile. This nately, this environmental assessment suggests that the sample was therefore subsequently also assessed for pol- existing late Pleistocene-Holocene sedimentary archives len. Pollen abundance was extremely low within the sam- retained by the BGS have limited value to address these ple, with only eight pollen grains counted. As a result, research questions. Despite being well curated, they do pollen concentrations were too low to permit any tentative not appear to contain the type of material needed for this conclusions to be drawn. study. This is not surprising since these cores were taken for distinctly different research purposes to those of this 8.5 Discussion current survey. Equally, they were drilled before detailed palaeogeographic maps of the resolution provided by this The generally low occurrence and apparent concentration project were available and hence do not appear to have of pollen in all the sequences discussed preclude detailed been extracted from the most promising palaeoenviron- interpretation. All of the spectra recovered commonly mental localities. contained grains that showed evidence of fragmentation, distortion and occasionally corrosion. Species identifica- In order to provide more substantial palaeoenvironmental tion was commonly hindered as a result of the poor pollen understanding of the environmental archive within the preservation. It is possible that this poor preservation North Sea, further ‘targeted’ sampling from the seabed is represents the degradation of material in storage after necessary. The locations of any future borehole drilling 101  Mapping Doggerland and sediment sampling programme should be pre- well-developed radiometric dating framework and imme- determined by reference to the features identified within diate palaeoenvironmental assessments should be under- the 3D seismic atlas. Promising, major landform features taken prior to storage in order to prevent sample loss for initial targeted sampling include the ‘Shotton River’ through decomposition and/or desiccation and to stabilise and associated wetlands and salt marshes. Any further any environmental evidence for future analysis. environmental analysis should be undertaken within a 102  Mapping Doggerland Appendix: Sediment and preservation descriptions of the Holocene / Late Pleistocene sections of the cores examined in this survey Shallow boreholes from this borehole were found to be in excellent condition (with some local exceptions where the core material was 79/01 comparatively desiccated and fragmented. Between 0 and 6m, the inspected sediments comprised mostly dark olive- This borehole was located at the southern extreme of the grey, fine-grained shelly sands and gravels. Below 6m PGS southern North Sea 3D seismic survey. Unfortunately there is a transition to interbedded fine sands and silty the drilling of this borehole was dogged with technical clays, with grey scattered shells. Organic materials were problems (completion reports, Murchison House) resulting not observed at any levels within the examined section of in limited recovery of late Pleistocene-Holocene deposits. core, even within the more clay-rich units. The inspected sediments were held in sample jars and comprised mostly brown, shelly sand. It was therefore 89/05 concluded that material from this borehole had limited potential. This borehole is located several km to the west of 82/21, and hence was chosen for examination for similar reasons. 81/49 The completion log stated that this borehole encountered Holocene sediments (mostly muddy sands) down to depths This borehole is located in the western part of the PGS of c. 14 m., below which it penetrated sands of Middle megasurvey. A BGS completion report for this borehole Pleistocene and earlier date. The first few metres of the stated that between 0 and 11.3m it contained deposits of Holocene succession, containing homogenous, soft muddy Pleistocene-recent age. Examination of the available core sand (but no observable organic matter) was in excellent material over this depth range revealed it to be generally condition. Between 1.5 and 10.0m however, much of the very fine-grained, soft sand-mud deposits, dark grey- core material was covered by variable amounts of mould brown in colour, often interbedded with fine clay laminae. raising the possibility that any organic material present, Towards the base of 'Pleistocene-recent' succession the could have been contaminated. Although the presence of sediments are more consolidated, desiccated and richer in numerous wood fragments within the Middle Pleistocene sand, lacking the clay laminae observed at shallower le- succession had been noted in the completion log, it was vels. The entire 'Pleistocene-recent succession' is lacking apparent upon inspection that this borehole had been ex- in shell fragments or organic matter. At around 11m there tensively sampled for micropalaeontological analyses at an is a clear transition to dark grey, indurated mudstones of earlier date leaving few horizons containing organic mat- Jurassic age. In summary, no material of potential pa- ter. laeoenvironmental use was found to be present in this bo- rehole. Vibrocores 81/50 54/02/80 This borehole is located outside the extent of available This vibrocore was an obvious candidate for examination seismic coverage, 48 km east of Lowestoft. However, a given its proximity to the large NW/SE trending early Ho- BGS report on both onshore and offshore boreholes ac- locene fluvial system recorded by seismic data (the so quired during 1981 stated that this borehole contained a called Shotton River channel). This vibrocore recovered horizon of wood fragments at shallow levels (<15 m) with- 5.8m of sedimentary deposits, most probably all Holocene in the Pleistocene-recent succession. Between depths of 0 in age. The sediments are in a good condition and com- and 10m, yellow-brown, medium-grained unconsolidated prise mostly fine-grained well-sorted sands, grey-brown in sands, often with abundant shell fragments, characterize colour with some bands of shell fragments, and slightly the retrieved sediments. In places the core is in quite poor muddy layers that may contain datable organic material. condition probably resulting from the unconsolidated ma- terial shifting. In places the core has also clearly been ex- 53/02/395 tensively sampled. Inspection of the core at the depth of 11.82 to 12.0m found only a few small wood fragments. This core was obtained near the southern margin of the The apparent paucity of wood fragments is probably a Outer Silver Pit. 5.98m. of core were recovered as grab function of previous sampling efforts. samples, and 5.90m of continuous core were recovered below this. Lithologically, the sediments are very similar 82/21 to vibrocore 54/02/80; yellow-grey, fine-grained biotur- bated sands with fragmented mud laminae and scattered This borehole was selected for inspection due to its loca- shell debris. Some dark horizons, possibly containing or- tion within the Outer Silver Pit. Overall, the core samples ganic material were identified, but no definitive wood 103  Mapping Doggerland fragments or ‘peat’ horizons were observed. The core cores were in a largely excellent condition, again some samples were generally in good condition, but some sec- sections were covered with mould. tions were covered with mould. 54/00/205 54/02/215 This vibrocore was drilled in the extreme east of the study This core was drilled on the opposite side of the Outer area. Whereas most of the vibrocores were soft and damp, Silver Pit. As well as a grab sample, 6m of continuous the sediments from 54/00/205 were completely dry and core were available for inspection. Interestingly, the ac- desiccated. Overall, the sediments are best described as companying BGS report described the vibrocore as having olive-brown fine-grained sands. The shallower units often penetrated "probably into the top of a small palaeovalley contain abundant shell fragments and pebbles, although in this area". The vibrocore samples are best described as the quantities of both decrease with depth. Again, no ma- soft, olive-coloured muddy sands, well sorted and fine terial deemed suitable for palaeoenvironmental analyses grained. No organic material was observed. Although the was found during the examination of these deposits. 104  Mapping Doggerland 9 The Archaeology of the North Sea Palaeolandscapes Simon Fitch, Vincent Gaffney and Kenneth Thomson 9.1 Introduction that the emergent landscape of the SNS should possess seasonally visited base camps during the Early Mesolithic. The map data generated as part of this project represents However, information from Scandinavian suggests a con- one of the largest samples of a, potentially, well preserved trasting lifestyle utilising only resources within a maritime early Holocene landscape surviving in Europe and it is zone (Indrelid 1978:169-70, Nygaard 1990:232), and it is essential that some consideration of the archaeological possible that the contemporary occupants of the "Dogger- context of the mapped remains is presented here. The land" coastline might have followed a similar lifestyle. If European cultural period associated with this landscape is this comparison were correct it would contrast with con- the Mesolithic which lasts between c. 10,000 BP and c. ventional models of Mesolithic movement for England at 5,500 BP, dependent on geographic position. Tremendous least (Darvill 1995, figure 20; Smith 1992). Indeed, whilst environmental change forms the backdrop to cultural it must be acknowledged that previous models have rarely events throughout this period. Sea level rise, associated had access to data from the original coastlines, recent dis- with climate change, resulted in the loss of more than coveries at sites including Howick in Northumberland 30,000 km2 of habitable landscape across the southern (Waddington et al. 2003), suggest that we might expect North Sea basin during the Mesolithic alone, and the inun- significantly more complexity and diversity in economic dation of this immense area has essentially left us with a and social practise than previously imagined or currently ‘black hole’ in the archaeological record for northwestern experienced. Any enhancement of our knowledge derived Europe as a whole. This situation is made worse by the from the submerged landscape of the North Sea is there- fact that finds from the region only rarely possess an accu- fore likely to provide information that will significantly rate provenance or context (Koojimans 1971; Verhart refine our appreciation of the Early Mesolithic within the 2004). larger region. Whilst the Early Mesolithic (10,000 to 8,500BP) record The Later Mesolithic in Britain (8,500 to 5,500BP) has from the North Sea region is, essentially, a blank, the ter- often been interpreted as a time of economic change and restrial record does provide some insight into what might increasing divergence from cultural developments in be expected within the area of the Southern North Sea it- Europe (Jacobi 1973; Wymer 1991). Jacobi (1976) for self (Jacobi 1973; Wymer 1991). The early stages of the example, concluded that such discrepancies were related to English Mesolithic are best represented in England by a the submergence of parts of the North Sea and the in- small number of sites including Star Carr, Thatcham, creased difficulty of maintaining connections between Broxbourne and Horsham (Clark 1972, Healy et al. 1992, Europe and Britain. Certainly, the effect of the final inun- Warren et al. 1934, Jacobi 1978). These sites do show dation of the North Sea emergent landscape during the some variation in culture indicative of differing social Later Mesolithic would have been significant to the many groupings (Reyner 1998). In the past there has been a groups who must have lived on, or adjacent to, the North trend to group British sites with those of the Maglemosian Sea plain. As the historic landscape was gradually lost to of Denmark, and frequently to see parallels with the "Du- the sea the area would have fragmented into islands. vensee" culture (Clark 1975). However, there are difficul- Whilst some of these isolated areas, at least, would have ties with such comparison and they add little to our under- continued to be populated as marine transgression pro- standing of the archaeology of the North Sea region as it gressed, habitation of this region would have become in- stands. creasingly tenuous and migration from the region must have occurred (Coles 1999). The consequences for the In general terms, all of these early sites demonstrate utili- groups who moved, or for those who lived in the areas into sation of a range of resources, primarily focused upon which they migrated, are largely unknown and only rarely game animals and plants. The Early Mesolithic in England considered. Whilst some consideration has been given to does not yet record substantive evidence of the use of ma- the issue of migration to areas including Norway and Scot- rine resources. However, given the emerging knowledge land, for the earlier period of inundation (Nummedal 1924, of coastal change it is likely that the majority of the areas Bjerk, 1995; Fuglestvedt 2003, Warren 2005, 37), the sig- that might record such economic practices are actually nificance of population movement during the final periods submerged. Evidence from Scandinavia, where substantial of flooding has hardly ever received consideration (Coles areas of early coastline survive, suggests that these re- 1999, 54). sources would not have been ignored (Norqvist 1995). Conventionally, the Early Mesolithic has been seen as The isolation of Britain that is assumed to have derived period where populations moved between base camps on from these changes is often stressed in the literature. The the coastline to inland camps to forage (Clark 1972, Smith absence of formal burial in the British archaeological re- 1992, Fischer et al. 2004). This interpretation suggests cord, for instance, is notable and suggests a cultural differ- ence. It may be that there were separate customs regard- 105  Mapping Doggerland ing burial in Britain but it is equally possible that formal of the new mapping for assessing the character of the ar- burial sites do exist and that these may have been located chaeological record and, in particular, the potential sur- near the coast, in areas which have now been lost to the vival of palaeoenvironmental data. Following this one can sea (Barton and Roberts 2004, Chatterton 2005, 108). assess how current management options may be changed However, as Funnell (1995) and Coles (1998) observe, or adapted to use the new data and, finally, it will be nec- Britain did not become an island until c. 5,500BP and the essary to discuss the potential of the data to plan research actual effect of the North Sea as a barrier to cultural con- strategies that may begin to answer some of the research tacts must be open to question. The use of major river questions outlined in the previous section. systems for communication seems uncontroversial (Rob- erts 1987, Reyneir 1998), and it is not inconceivable that Clearly the resolution of the data produced by the NSPP the shallow marine areas of the North Sea could have been does not permit a fine-grained assessment of the archae- traversed and contact with European groups maintained ology of the area. However, many of the features identi- (e.g. Coles 1998, 76). One might even suggest that the fied through the analysis have the potential to achieve ar- potential for communication by boat, via shallows, might chaeological significance (Table 9.1). Paramount amongst have actually enhanced the potential for contact rather than these features is the OSP. This basin dominates the acted as a barrier. Consequently, whilst the overall picture mapped landscape both in extent but also the manner in provided by the available evidence for the Late Mesolithic which so many other features are linked to or drain into it. within Britain suggests a mosaic of localised groupings we It also represents, of course, a major economic resource: should be cautious when assuming that this reflects en- whether considered a lake or a marine outlet. Surrounded forced isolation (Morrison 1980). by nearly 700 kilometres of coastline, or lakeshore, merg- ing with 10 major estuaries and a salt march covering Another traditional characteristic associated with the tran- more than 300 square kilometres, the OSP must have acted sition from the Early to Late Mesolithic, and often as- as a prime economic resource for human groups across a sumed to be a consequence of the change in sea level, is massive area. Waterfowl, fish and other animals must the assertion that there is an increasing focus upon coastal have been abundant in this area, as would reeds or other resources (Rowley Conwy 1983). This shift has been in- vegetational resources that hunter gatherer groups might terpreted as a response to higher population levels or mo- require. In its later incarnation as a marine estuary the bility caused by sea level rise (Mithen 1999). However, OSP also provided a significant point of access to the ma- this period is characterised by an increasing visibility of rine resources missing from much of the English terrestrial activity in landscapes that had previously been under- archaeological record. Presumably this is an area where represented in the archaeological record, e.g. estuaries we could seek evidence for intense utilisation of marine which had lain beyond the contemporary coastal margins. resources and any differing social and settlement struc- These unexplored areas provided a diverse range of re- tures that might result from access to such resources. sources that were unlikely to be ignored during any period Away from this imposing area of water, the twenty four of human occupation (Allen 1997, Clarke 1978). Once lakes or wetlands and the 1,600 kilometres of rivers or again, we should be cautious about the extent, or signifi- streams recorded by project staff would also have provided cance, of such apparent change (Milner 2004). The cur- similar opportunities for hunter gatherer groups, whilst rent picture may well be the result of increased visibility of also directing paths and tracks through the landscape. coastal resource utilisation rather than substantive eco- These features achieve further significance as volumetric nomic change. The provision of information that permits and sedimentary analysis suggests that many of these areas adequate comparison or assessment of development during are most likely to provide palaeoenvironmental evidence. this period is again predicated on the availability of repre- It may be that we will never be able to explore settlement sentative data for the period overall. On that basis, the associated with these features but proxy evidence for set- potential for the North Sea to provide critical information tlement and land use gained through a programme of di- for such an assessment seems clear. rected coring planned on the basis of the results presented here is a real possibility. The probable significance of the results derived from the NSPP mapping of the Holocene surfaces should be clear The character of this area cannot, of course, be represented from the previous discussion. It is essentially true that our simply as a sequence of river and lakes. It was a land- current interpretative position for the Mesolithic of the scape in the fullest sense and we must not sidestep the maritime regions of north western Europe stand largely as responsibility of treating it as such. Figure 9.1 provides a a consequence of the lack of information from the North general interpretation of the mapped landscape data made Sea. The ability to identify significant landscape features available through the project. Here it is important to stress within the area of the North Sea, or information that can that whilst the character of the area studied was, essen- support directed exploration or future data gathering, is tially, a plain it was never a featureless plain. Mesolithic therefore a considerable opportunity for research. It would communities of the North Sea would have been sensitive also be a serious challenge for the heritage communities to the economic and social significance of this subtle vari- within all the countries that bound the North Sea basin ability. (Maarleveld and Peeters 2004). The nature of this chal- lenge can be assessed initially by considering the potential 106  Mapping Doggerland Figure 9.1 Major topographic or economic zones within the study area Although the current work was only able to provide a course, attractive for human groups and there is at least a broad outline of topography away from major incised chance that faults in these areas might also expose other features, this remains an achievement in the light of how useful resources including lithic sources. little we knew previously. We can guess that the large, low valley in the west of the study area would have been Contemporary populations, of course, would have under- attractive for a variety of reasons (figure 9.1). The low stood the great North Sea plain in a much more intense hill, tentatively identified within this valley, suggests and personal manner. Small groups would have been in- opportunities for settlement or even for hunting stands. timate with features of the landscape that we cannot detect Features associated with salt domes are also particularly using current technologies. The emotive relationship be- interesting in landscape terms. In some cases upswelling tween individuals and their surrounding landscape can domes would have formed low hills but, where there is hardly be understood through a study that can only evidence for graben collapse, the centre of these domes vaguely discern the trend of the land or map its grosser may have contained wetlands or lakes. Such areas are, of features. It is, however, the best we possess and on that 107  Mapping Doggerland basis we must consider what implications may be drawn in area (Boomer et al. 2007; Coles 1998). Equally signifi- respect of further research. cantly, similar studies are limited by the extent of available seismic data. There is a major gap in the availability of 9.2 Future Research 3D seismic data in the marine areas associated with North- ern England (Bunch et al. this volume). There is also an Despite the limitations outlined above it is incontrovertible attenuation of response to 3D seismic survey in shallow that the data presented as part of this project has demon- waters. Consequently, there exists a "white band" which strated the potential of marine, remote sensed data for the surrounds the modern coast and within which our knowl- exploration of the inundated Holocene land surfaces of the edge of the palaeotopography and, by inference, the ar- North Sea. In comparison to the situation described by chaeology of the area, is severely limited. Our ability to Flemming, a mere 3 years ago, the North Sea is no longer tie together the data from the Southern North Sea with terra incognita (Flemming 2004). It is, of course, ac- terrestrial archaeology in a seamless manner, although knowledged that the current product still represents a lim- desirable, is therefore limited. In the deeper marine areas ited interpretation and could be substantially refined by the there will be a reliance upon 2D seismics to fill this gap, integration of further data sets, including high-resolution with a concomitant loss of the extensive data associated 2D seismic surveys. However, the resolution and detail of with 3D data sets. In shallow waters traditional methods the derived landscape can provide a substantive basis for of marine prospection may be employed to effect (diving, further prospection or exploration of this unique land- high resolution seismic survey etc.): although the extent of scape. In particular the need, outlined in the paper by such data is a limiting factor when considering heritage Smith et al. (this volume), for further coring to support requirements to manage the resource. There is, therefore, palaeoenvironmental research is substantially supported by an urgent need to collect new data sets to fill these gaps or our ability to identify and map deposits with enhanced to investigate methods to integrate other data sets in a environmental potential with some confidence. We may more imaginative and productive manner. now be freed, to some extent, from our current reliance on serendipitous finds with poor contextual value. Future There is another point to be made in relation to the remain- work should now be directed towards detailed pa- ing archaeological potential of the North Sea. There is laeoecological studies of the type commonly carried out in increasing evidence that we should expect, at the least, terrestrial landscapes around the North Sea basin (Peeters low-level occupation in areas north of the study area dur- 2006). ing the Later Palaeolithic. Wickham-Jones and Dawson (2006, 19) suggest that the melting of the Devensian ice We should also note the potential of the data for support- sheet north of Scotland would have been rapidly followed ing novel and exciting behavioural models with real ar- by marine inundation and that the areas to the north and chaeological potential (Ch'ng et al 2004). The Holocene west would not have been available for occupation. How- landscapes of the North Sea were never an abstract con- ever, whilst excluding consideration of even earlier peri- cept. This land was both habitable and inhabited, and the ods, the spatial extent of surviving late Palaeolithic land landscape data we possess, or can now acquire, offers us surfaces that have a potential for preserving traces of hu- the opportunity to explore archaeological predictive mod- man occupation is actually bounded by the Norwegian elling that can, in turn, be used to refine our concepts of Trough and encompasses the Viking Bergen Hills. land use and enhance the potential of directed, invasive exploration to answer archaeological questions. Other Not surprisingly, traces of occupation in these areas are research programmes have already begun to generate ap- few and the context of a single worked lithic, recovered plied models and some, including the "Danish Fishing from a vibrocore at a depth of 143 metres off the Viking Model", are reported to be very successful (Fischer 1995, Bank, remains uncertain (Long et al. 1986). However, 375). Most of these have used localised bathymetry as a Wickham-Jones and Dawson conclude that current ab- topographic proxy but this is inevitably less successful in sence of evidence for Later Palaeolithic habitation in Scot- deeper water where burial of the landscape has occurred land should not be regarded as evidence of absence and (Fisher 1995, 377). The utilisation of information from that "the submerged landscape of the Scottish shelf is thus seismic data should help improve modelling strategies and the most likely location for the preservation of traces of the exploration of predictive models using the North Sea early settlement" (2006, 34). The significance of 3D seis- seismic data is part of research currently being carried out mic data sets from northern waters can be demonstrated at Birmingham (Fitch et al 2007). and the result of analysis of one small area, not far from the Viking Bank find indicates that new insights are possi- Despite the apparent success of the NSPP it should not be ble for early landscapes in the deeper waters to the north presumed that the current work represents a final product (Figure 9.2). either in spatial or chronological terms. This area studied here does not represent the whole, or even the available, Such observations are also pertinent to this study. Whilst extent of land surfaces that could be investigated. The not consistently mapped as part of this project, results in- shoreline of the great North Sea Holocene plain would dicate that Later Palaeolithic surfaces are also amenable to have extended north along the current shoreline of north- study within the Southern North Sea (Figure 9.3 and atlas). ern England and further to the east of the present study One must assume that the potentially, well preserved Later 108  Mapping Doggerland Palaeolithic deposits that underlie the current study area, for research and heritage management. and which stretch far to the north, must rate as priorities Figure 9.2 Sample seismic data illustrating probable Later Palaeolithic land surfaces adjacent to the Norwegian Trench. The Viking Bank flint illustration is from Long et al. (1986) Figure 9.3 Seismic data cube illustrating chronostratigraphic relationship between Holocene and earlier features 109  Mapping Doggerland 9.3 Cultural resource management an international group with explicit interests in policy is- sues related to the prehistory of the North Sea and its asso- procedures in the Southern North ciated coastline as a whole. In support of such initiatives, Sea the overt inclusion of marine issues in the emerging Pa- laeolithic Research Framework1 may also prove significant The primary qualities of the North Sea archaeological re- over the longer term. source, in management terms, are its general inaccessibil- ity and an uncertainty concerning the nature, or even loca- Within this larger context, it is clear that the extent and tion of any archaeological remains. This contrasts sharply detail of information provided through this project for the with intertidal or shallow marine zones where there is usu- Holocene land surfaces of the North Sea is currently ally some opportunity to physically record, known sites, to unique and tasks heritage agencies with providing an ap- analyse their distributions and therefore to provide some propriate management response. Within England, at least, degree of protection or management. The depth of depos- there are some basic guidelines to guide action, Robert and its, or water column, overlying the presumed North Sea Trow's (2002) publication “Taking to the Water: English landscape has generally ensured that the presence of ar- Heritage's Initial Policy for the management of Maritime chaeological deposits could only be inferred on the basis Archaeology in England” sets out the general principle of contemporary correlates from terrestrial or shallow wa- that the marine resource “and terrestrial archaeological ter contexts (Flemming 2004). Paradoxically, whilst there remains provide a seamless physical and intellectual con- is a general assumption that the depth of water and overly- tinuum” and that maritime heritage should “enjoy parity of ing deposits might mask substantial, preserved archaeo- esteem and treatment with their terrestrial counterparts” logical deposits, the archaeological material trawled from (Robert and Trow 2002, 4). Of particular importance the area, which is generally our only guide to the distribu- within this document is the requirement to consider the tion of deposits, presumably suggests continuing damage marine environment as an historic landscape. Given the to relict deposits. Unfortunately, whilst acting as an im- scale of the North Sea study area, this is clearly a premise portant proxy for direct examination, this material also has from which we can begin to provide a management re- a low locational or interpretative value. sponse (Oxley and O'Regan 2001). Management of such a resource, essentially indefinable or With this in mind, and given the available data provided without adequate positional information, is an unenviable through the project, there are three heritage products that challenge but one that cannot be avoided (Roberts and one might anticipate from this study: Trow 2002; Oxley and O'Regan 2001). There are a num- ber of legal or treaty obligations that govern regional and • A general characterisation and interpretation of national responsibilities for marine heritage. Wickham the available data in landscape terms Jones and Dawson (2007, 7-14) and Flemming (2005, 3- • An assessment for the likely potential of the 10) provide substantive reviews of national legislation and available data in respect of archaeological re- international obligations that apply or impact upon British search maritime territory (Wenban-Smith, 2002). What need be • An assessment of the reliability of the interpreta- stressed here is that whilst English Heritage’s direct re- tion and its value for mitigation mapping sponsibility for marine heritage carries only to the 12 mile limit around the coast, many government agencies retain a The general process by which such work could be carried wider interest in the marine heritage (Oxley nd). More- out as part of the project is provided in Figure 9.4. over, the nation, through treaty or international obligation, is often required to consider marine heritage issues across 9.4 Landscape Characterisation territorial waters. It is also true, following the extensive recent activity related to marine archaeology (frequently For the past 16 years landscape heritage management related to ALSF funding), that agency interests in the within much of the United Kingdom has been concerned wider issues of the North Sea are becoming more explicit with historic landscape characterisation (HLC). HLC, in (Oxley nd). This much is clear in published reviews of the its current form, derives from the recognition that there is potential of the marine resource (e.g. Dix et al. 2004), or a requirement to provide a comprehensive characterisation projects with applied methodological value (Bates et al. of entire landscapes as a management tool (Aldred and nd). Fairclough 2002, Fairclough and Rippon 2002). Previous heritage strategies were undermined by an overemphasis A number of ALSF projects have also studied specific on known archaeological sites that, unintentionally, tended heritage management issues including the provision of to privilege isolated areas to the detriment of of the wider codes of practice for reporting marine finds and investigat- landscape. ing the application of historic landscape characterisation programmes to marine seascapes (Wessex Archaeology 2003; 2005). The establishment of the North Sea Prehis- tory Newsletter by Dr Peter Murphy (English Heritage) is also worth stressing here as this indicates the emergence of 1 http://www.iceage.org.uk/Framework.html 110  Mapping Doggerland Figure 9.4 The analytical process In contrast, HLC sought to appreciate the overall impor- with other project staff indicated that there were signifi- tance of the landscape itself. No specific part of any land- cant difficulties in applying HLC in areas where early fea- scape is deemed as more valuable than another within an tures were obscured by later sediments (Baker et al. 2007 HLC programme, and all aspects of a landscape are avail- figure 8.1). In contrast, the scale at which the North Sea able for classification and therefore management. Land- project has operated, and the extensive topographic data scape is treated as a contemporary and dynamic entity in- generated, has a clear HLC application. corporating past activity as one contributing factor to the final classification. There are, however, also significant differences faced when attempting to apply HLC to the NSPP data. In the All HLC projects seek to assess every significant land par- first instance the available mapping is not consistent across cel of a study area, assigning it to the type which best the whole of the study area. Seismic response is variable represents its predominant landscape character, as far as and the response less good in the southern and western this is determinable. Once detailed HLC types have been sectors of the study area where the water column is rela- defined, simplified and interpreted cultural landscape tively shallow. The resolution of the mapped data from zones can be derived. This approach enables general pat- the North Sea also falls well below that expected by terres- terns of the landscape to be discerned and provides a basis trial HLC projects, which commonly use 1:1250-1:10,000 for further classification or research if required. The pri- map scales according to the context of the project (Aldred mary output from a traditional HLC programme is there- and Fairclough 2002, 26). However, the poor resolution of fore a broad-brush interpretation of landscape character the data may be offset by the nature of the landscape under that supports the management of change across an entire study. The North Sea data effectively represents a par- landscape by providing a continuous assessment of the tially mapped Mesolithic landscape in topographic terms, whole area (Fairclough 2006). and the notional resolution of the data (Thomson and Gaffney, this volume), supports mapping of generalised Heritage practitioners have not ignored the potential value economic/landscape units which may, ultimately, reflect of HLC within the context of maritime archaeology and a broad land use patterns within a Mesolithic economy. number of initiatives, including English Heritage's historic Whilst unencumbered by later cultural development the seascapes programme, have attempted to implement the landscape’s post-depositional taphonomy should also be broad concepts of historic landscape characterisation in a considered part of the landscape's character. In this sense, number of marine environments (Wessex Archaeology the data seems, a priori, to possess the potential for a suc- 2005; Hill et al. 2001). At the time of writing, few of cessful HLC implementation. these projects had reached fruition, but informal discussion 111  Mapping Doggerland Table 9.1 Primary landscape characterisation zones Class Area (Km2) Logical ID Description General Areas extension/ Speculative (Km2) 1 1872.33 261.63 Early Mesolithic Similar to the Severn Estuary, base The Outer Silver Pit Seaway shows tidal scour marks and presence of abandoned bedforms. 2 2872.34 0 Dominated by geology Area is typified by thin deposits Offshore Lincolnshire with fluvial systems and near surface solid geology, which illustrate modern erosion, a few fragmentary fluvial systems present. 3 3154.09 932.17 Dominated by geology Area of very strong solid geology Spurn, Sole Pit Region and with some fluvial response with large salt structures. Eastermost Rough systems It is likely that these formed regional features within the landscape. 4 412.55 157.45 Inlet area - partial The inlet of the Outer Silver Pit. Outer Silver Pit scour The area only shows partial scouring, and possesses a channel that may have drained the lake which once existed in the Outer Silver Pit. 6 4390.82 124.84 Landscape influenced An area where the fluvial channels Offshore Norfolk, South by underlying glacial appear to be influenced by Central Dogger Bank, Outer deposit underlying glacial deposits. Well bank 7 3760.9 0 Area of smaller This area contains many smaller Swarte Bank, Indefatigable Holocene channels channels, only visible in part, but Bank appear to have extended across the whole of this area. 8 823.52 0 Area surrounding large Drainage in this region appears to South Botney Cut, South East lacustrine feature have been dominated by the Outer Silver Pit Markhams Hole Lacustrine System. 9 2762.3 65.54 Low lying areas with Lower lying areas of the Southern Doggerbank, South soft coastline Doggerbank region that possess West Patch and South West extensive fluvial systems which Spit extend into soft coastline areas. 10 1626.13 0 Area of reuse of Later This area is dominated by two South Eastern Outer Silver Pit, Pleistocene features major fluvial systems which appear Well Hole to be utilising existing Late Pleistocene courses. 11 146.37 0 Lacustrine feature Area that would have formed Sole Pit, Silver Pit, Well Hole lakes/wetlands during the Early and Markhams Hole Mesolithic. 12 1033.29 0 Doggerbank fluvial This area holds the clearest and Dogger bank, Eastermost systems area best preserved of the fluvial Shoal systems in the region, the data in these regions suggest that the landscape is well preserved. 13 535.23 0 Areas with clear These areas clearly show evidence Botney Cut Region, Cleaver indication of marine of being altered by marine Bank, South Rough, transgression (salt incursion. Eastermost Shoal marsh) 14 293.23 42.67 Early Holocene The Early Mesolithic coastline. Outer Silver Pit, North East coastline Doggerbank 112  Mapping Doggerland Figure 9.5 Broad landscape character zones 113  Mapping Doggerland Figure 9.6 Cross correlation of major topographic and landscape characterisation zones On that basis, a strategy was designed to provide a basic However, provision of the broad topographic variation of HLC product using the available data. Initially, the land- the landscape, picked from the Holocene land surfaces, scape was classified into a series of broad areas based permitted refinement of this image by cross-tabulating the upon their depositional history, and major historic land- primary topographic variation of the Mesolithic landscape scape features. At the outset an attempt was made to with the primary landscape zones defined from the initial automate this process by treating the data as a hyperspec- characterisation phase. A total of 80 separate land classes tral image. This followed a procedure designed to catego- were generated through this process and the result is rise poorly mapped landscapes explored by the team in the shown in Figure 9.6. This data is interesting as it probably context of research at Fort Hood, in Central Texas (Barratt represents the best general zonation, in terms of potential et al 2007; White and Ray 2000). This approach, however, Holocene land use, currently achievable using the avail- failed at an early stage due to the mosaic nature of the able data and, to the extent that it may correlate with broad seismic images. The landscape characterisation was there- economic activity, may carry considerable potential to act fore performed manually using the available mapping, and as the basis for more detailed behavioural modelling. was primarily guided by geomorphological and hydrologi- cal characteristics to provide broad landscape zones. On It should be acknowledged that the data presented here this basis, the entire area was classified into the areas de- does not currently represent a full HLC product, as it does tailed in Table 9.1, and the description added to the poly- not truly incorporate contemporary landscape features. In gon layer as an attribute to provide graphical display fact a further reclassification of the primary HLC image (Figure 9.5). The dividing lines for many of the landscape was generated incorporating a zoned bathymetry layer. zones observed coincided broadly with known watersheds This produced an excessively complex image that, al- between observed fluvial features. though potentially of use for management, is not presented here and is retained in archive. 114  Mapping Doggerland Figure 9.7 Potential for preservation 115  Mapping Doggerland Table 9.2 Ranking of features by relative archaeological potential River potential Lakes Marsh/Wetlands Coastlines Deadzones/Background 0 = Absent 0 = Absent 0 = Absent 0 = Absent Scoured = 0 1 = Low 4 = present 4 = present 4 = Present Landscape present = 1 4 = high (Ranking based on a modified Strahler stream ordering) 9.5 Threat mapping tion or development plans modified in the light of the po- tential of such areas to contain undiscovered features. Whilst the HLC data are significant in their own right they are probably not, in themselves, an adequate basis for a When considering these issues it seems reasonable to sug- larger management strategy for the southern North Sea. In gest that the further we are from identifiable structures the particular, this requires an assessment of the potential of more likely it is that other factors may be preventing dis- the area under study for preservation of archaeological covery. Following such an argument, a separate map was materials. This may be provided by the multiplication of prepared representing threat and uncertainty as a single two normalised data sets for the interpreted archaeological measure linked to horizontal distance from feature and potential of identified landscape features and the depth of accessibility (ranked according to the overlying depth of overlying sediments derived from published sources in- sediment and water column). These three factors were cluding BGS mapping. The relative values for landscape normalised and summed to provide a single value and the feature potential, prior to normalisation, are provided in results are presented in Figure 9.8. This map provides a Table 9.2. Following this process, areas with a lack of continuous assessment across the study area in which areas known features and an absence of significant sedimenta- of high threat and low uncertainty (shallow water column tion score low, whilst probable scoured areas, including or sediments proximate to identified features) grade into the Outer Silver Pit, produce a value of 0. Areas with areas with low threat and high uncertainty (greater water probable archaeological potential and with significant column or sediments at an increasing distance from known overlying deposits score high. The mapped data is pro- features). vided in Figure 9.7. This procedure results in a simple but highly effective form of "red flag" mapping that can be usefully compared Not surprisingly, Figure 9.7 emphasises areas that are to Figure 9.2, which primarily reflects probable archaeo- likely to be of prime archaeological interest, most notably logical potential. Setting aside areas which may be lacustrine environments, marsh areas and coasts. How- scoured (the Outer and Inner Silver Pits), this measure ever, the areas around the large river systems to the north highlights significant areas in the southern and western of the Outer Silver Pit are emphasised overall; a conse- parts of the study area as zones which might contain fea- quence of the association of a dense network of major tures, which are not amenable to current mapping tech- channels and protective sediments. The apparent potential nologies but which may be more prone to development associated with the large sand bank systems in the south threat. east of the project area may be misleading as this reflects depth of sediment associated with highly mobile features. 9.7 Final Observations 9.6 Threat and Uncertainty Mapping The North Sea Palaeolandscapes Project mapped more than 23,000 square kilometres of Holocene land surfaces The data provided in Figure 9.7 is useful in assessing the and presented these for publication in slightly over 18 overall significance of features identified through the months. The product of this work is one of the largest seismic analysis. However, such mapping does not pro- analyses of remotely sensed data ever attempted for ar- vide substantial guidance in areas within which features chaeological purposes and this has brought to the attention have not been identified. Hence the extensive areas that of the archaeological and heritage communities one of the are suggested as having a relatively low potential may be most extensive and best preserved prehistoric landscapes misleading. Earlier papers in this volume have noted that in Europe at least. The methodologies demonstrated here our ability to identify Holocene structures is limited in a have wide application in similar landscapes elsewhere, number of areas, notably those associated with a shallow when appropriate remote sensed data is available. How- water column. Consequently, a primary concern, after ever, whilst such research is technically appealing, we direct identification of Holocene features, must be to iden- should not lose sight of the fact that its fundamental im- tify areas that may contain features and might also be un- portance is its ability to inform research into the Early der threat. Such zones may be chosen for further prospec- Holocene communities of northwestern Europe. 116  Mapping Doggerland Figure 9.8 Red flag mapping. This figure combines threat and uncertainty data based on distance to feature and depth of overlying sediment. The lack of sediment cover and direct association with identified features with archaeological potential rate as high threats with little uncertainty. Deep overlying deposits lying further from recorded features rank as low threat areas but with significant levels of uncertainty Reynier (2005, 1) recently described research into the comparison, these can never truly be used as a proxy for early Mesolithic as currently "listless", perhaps due to the settlement more than 120km away, in the centre of the difficulties presented by the archaeological record. In part Great North Sea plain. this may be a consequence of our lack of knowledge of the prehistoric archaeology of the North Sea. Currently, the The results of the NSPP should be taken as a wake-up call. Holocene archaeology of the region is infrequently con- The landscapes mapped here represent areas that would sidered within the literature and the lack of available evi- have been prime habitable zones linking and, perhaps, dence is tacitly presented as an absence of evidence. Con- explaining much of the archaeological variation we see sequently, the area appears to occupy a proximal role in around the North Sea basin. In contrast, the present terres- the literature and our interpretative position. It remains trial archaeological record, which approximates the sum of true, however, that only a few terrestrial sites are actually our knowledge, in Britain at least, may better be repre- available to support our current interpretative models for sented as areas that were peripheral locations for the the earlier Holocene (Milner and Woodman 2004, 5), and Mesolithic occupants of the North Sea basin (Morrison fewer have provided adequate environmental evidence for 1980). The data provided by the NSPP mapping pro- this period (Whitehouse and Smith 2004). This is a par- gramme provides significant support for a radical shift in lous position and we should be assured that the apparent our interpretative position for the Mesolithic in north density of sites that have been identified or explored in western Europe. Previously unimaginable, the Holocene Europe, most notably in Denmark, will not actually fill landscapes revealed here allow us to discriminate between this gap (Fischer 2004). Few of these sites are located environmental zones, characterise areas of archaeological further than 5km from the coast and, whilst useful for potential and, possibly, provide the opportunity to explore 117  Mapping Doggerland the southern North Sea with the likelihood of archaeologi- than 7,000 years ago. Whilst it is easy to be overwhelmed cal success. In doing so we can anticipate the exploration by the sheer scale of the surviving historic landscape, the of an entirely new European country whose study may re- heritage of the North Sea is fragile. Our previous lack of invigorate research into the Mesolithic and later Palaeo- knowledge has permitted unsympathetic development and lithic occupation of northwestern Europe. poorly managed exploitation. Having rediscovered "Dog- gerland" the United Kingdom, and all the countries bound- In promoting academic change the project can, so far, be ing the North Sea, must assume the responsibility associ- judged a success. However, the exploration and mapping ated with our shared heritage. How we respond to such a of uncharted rivers and hills also brings other responsibili- unprecedented challenge will be judged by future genera- ties. The archaeology of the North Sea has remained es- tions. sentially untouched since the land was lost to the sea more 118  Mapping Doggerland References Aldred, O., and Fairclough, G. 2003. Historic Landscape Characterisation: taking stock of the Balson, P. S., and Jeffery, D. H. 1991. The glacial sequence method. English Heritage and Somerset County of the southern North Sea. In J.Ehlers, P. L. Gibbard, and J. Council, London. Rose, (eds.) Glacial Deposits in Great Britain and Ireland. Rotterdam: Balkena, 245-253. Allen, J. R. L. 1997. Subfossil mammalian tracks (Flandrian) in the Severn Estuary, S.W. Britain: Barratt, G., Dingwall, L., Gaffney, V., Fitch, S., Huckerby, mechanics of formation, preservation and distribution. C., and Maguire, T. 2007. Heritage management at Fort Philosophical Transactions of the Royal Society B352 Hood, Texas: Experiments in Historic Landscape Charac- (513): 481-518. terisation. Archaeopress. Allen, M. J., and Gardiner, J. 2000. Our Changing Barton, N., and Roberts, A. 2004. The Mesolithic Period in Coast: a survey of the intertidal archaeology of England: Current perspectives and new research. In: A. Langstone Harbour, Hampshire. CBA Research Report Saville (ed.) Mesolithic Scotland and its Neighbours. Society 124. of Antiquaries of Scotland. Amos, C. L. and King, E. L. 1984. Bedforms of the Bates, R., Dean, M., Lawrence, M., Robertson, P., and Canadian Eastern seaboard: a comparison with global Tempera, F. 2005. Innovative Approaches to Archaeological occurrences. Marine Geology. 57: 167-208. Site surveying and Evaluation. ALSF Project 3837. http://www.coastms.co.uk/Conferences/Outputs%20and%20 Ashley, G. M. 1990. Classification of large-scale Reports/ALSF%202005/ALSF%202005%20Bates.pdf subaqueous bedforms: a new look at an old problem. Journal of Sedimentary Petrology. 60: 160-172. Bell, M., Caseldine, A., and Neumann, H. 2000. Prehistoric Intertidal Archaeology in the Welsh Severn Estuary (CBA Autin, W. J. 2002. Landscape evolution of the Five Research Report 120). London: Council for British Islands of south Louisiana: scientific policy and salt Archaeology. dome utilization and management. Geomorphology. 47: 227-244. Bell, M., and Walker, M.J.C. 2005. Late Quaternary Environmental Change (second edition). Harlow: Pearson. Bacon, M., Simms, R., and Redshaw, T. 2003. 3-D Seismic Interpretation. Cambridge University Press. Bell, T., O' Sullivan, A., and Quinn, R. 2006. Discovering ancient landscapes under the sea. Archaeology Ireland 20 Bahorich, M., and Farmer, S. 1995. 3-D seismic dis- (2): 12-17. continuity for faults and stratigraphic features: The co- herence cube. The Leading Edge 14: 1053-1058. Bellamy, A. G. 1998. The UK marine sand and gravel dredging industry: an application of Quaternary geology. In: Bailey, G. 2004. The wider significance of submerged J. - P. Latham, (ed.) Advances in Aggregates and archaeological sites and their relevance to world prehis- Armourstone Evaluation. Geological Society, London, tory. In Flemming N.C (ed.) Submarine Prehistoric Engineering Special Publications, 13: 33-45. Archaeology of the North Sea, CBA Research Report 141. BGS. 2007. - Seismicity and Earthquake Hazard in the UK - located at Baker, M.V., Tapper, B, Johns, C., and Herring P. http://www.earthquakes.bgs.ac.uk/hazard/Hazard_UK.htm - 2007. England’s Historic Seascapes, Scarborough to visited 14/02/2007. Hartlepool and Adjacent Marine Types. Historic Seascape Characterisation. Historic Environment Bird, E. 2000. Coastal Geomorphology: An Introduction. Service, Environment and Heritage, Cornwall County John Wiley and Sons, 322. Council Report No: 2007R021. Bjerk, H. B. 1995. The North Sea Continent and the pioneer Ballard, R. D., Colemand, D. F., and Rosenburg G. settlement of Norway. In. A. Fischer (ed.) Man and Sea in 2000. Further Evidence of Abrupt Holocene Drowning the Mesolithic. Oxbow. 131- 144. of the Black Sea Shelf. Marine Geology. 170: 253-61. Blondeaux, P. 2001. Mechanics of coastal forms. Annual Bally, A. W. 1987. Atlas of seismic stratigraphy Vol- Review of Fluid Mechanics. 33: 339-370. ume 1. AAPG studies in Geology: 27. American Asso- ciation of Petroleum Geologists, Tulsa, Oklahoma, Boomer, I., Waddington, C., Stevenson, T., and Hamilton, D. U.S.A. 2007. Holocene coastal change and geoarchaeology at 119  Mapping Doggerland Howick, Northumberland, UK. The Holocene 17 (1): 89-104. Ch'ng, E., Stone, R. J., and Arvantis, T. N. 2004. The Shotton River and Mesolithic Dwellings: Recreating the Past Brayshay, B. A. and Dinnin, M. 1999. Integrated from Geo-Seismic Data Sources. The 5th International palaeoecological evidence for biodiversity at the Symposium on Virtual Reality, Archaeology and Cultural floodplain-forest margin. Journal of Biogeography 26 Heritage, VAST (2004), in cooperation with ACM (1): 115-131. SIGRAPH and The Eurographics Association, 7-10 December, Brussels, Belgium. 125-133. Brown, A. G. 1983. Floodplain deposits and accelerated sedimentation in the lower Severn basin in K. J. Chapman, H. P. 2006. Landscape Archaeology and GIS. Gregory (ed) Background to Palaeohydrology. London: Tempus Publishing Ltd. 60. Wiley. 375-397. Chapman, H. P., Fletcher, W. G. and Thomas, G. 2001. Brown, A. 1997. Alluvial Archaeology. Cambridge: Quantifying the effects of erosion on archaeology of inter- Cambridge University Press. tidal environments: a new approach and its implications for their management. Conservation and Management of Ar- Brown, M. 1986. Indefatigable Sheet (53N 02E): chaeological Sites 4: 233-240. Quaternary Geology. HMSO, London Chatterton, R. 2005. Ritual. In. C. Conneller and G. Bruthans, J., Filippi, M., Geršl, M., Zare, M., Melková, Warren. Mesolithic Britain and Ireland: New Approaches. J., Pazdur, A., and Bosák, P. 2006. Holocene marine Tempus. terraces on two salt diapirs in the Persian Gulf, Iran: age, depositional history and uplift rates. Journal of Clarke, G. 1936. The Mesolithic Settlement of Europe. Quaternary Science. Cambridge University Press. Bulat, J. 2005. Some considerations on the interpreta- Clark, J. G. D. 1972. Star Carr: a Case Study in tion of seabed images based on commercial 3D seismic Bioarchaeology. Reading (MA): Addison-Wesley in the Faroe-Shetland Channel. Basin Research 17: 21- (Addison Module in Anthropology 10). 42. Clark, J. G. D. 1975. The Earlier Stone Age Settlement of Bull, J. M., Gutowski, M., Dix, J. K., Henstock, T. J., Scandinavia. Cambridge University Press. Hogarth, P., Leighton, G., and White, P. R. 2005. Ma- rine Geophysical Researches 26: 157-169. Clarke, D. 1978. Mesolithic Europe: The Economic Basis. Duckworth. Burkitt, M. C. 1932. A Maglemose Harpoon dredged up recently from the North Sea. Man 32 (May, 1932): Coles, B. J. 1998. Doggerland: a speculative survey. Pro- 118. ceedings of the Prehistoric Society 64: 45-81. Buteux, S., Gaffney, V., van Leusen, M., and White, R. Coles, B. J. 1999. Doggerland's loss and the Neolithic. In: 1997. Establishing GIS services at the Birmingham B. Coles, J. Coles and M. Schon Jorgensen (eds.) Bog Bod- University Field Archaeology Unit: Problems and ies, Sacred Sites and Wetland Archaeology. WARP Occa- potential. In A. Gotarelli (ed) Sistemi Informativi e Reti sional Paper. 12: 51-57 Geografiche in Archeologia: GIS-Internet. Firenze, 59- 77. Coles, B. J. 1998. Doggerland: a speculative survey. Proceedings of the Prehistoric Society. 64: 45-81. Buteux, S., Gaffney, V., White, R., and van Leusen, M. 2000. Wroxeter Hinterland Project and geophysical Connolly, J., and Lake, M. 2006. Geographical Information survey at Wroxeter. Archaeological Prospection 7: 69- Systems in Archaeology. Cambridge Manuals in Archae- 80. ology, Cambridge University Press. Cameron, T. D. J., Crosby, A., Balson, P. S., Jeffery, D. Cook P.J. 1991. Spurn Sheet (53N 00E): Quaternary H., Lott, G. K., Bulat, J., and Harrison, D. J. 1992. Geology. HMSO, London United Kingdom offshore regional report: the geology of the Southern North Sea. HMSO for the British Geo- Darvill, T. C. 1995. Prehistoric Britain. Routledge logical Survey. Davies, J. L. 1964. A morphogenetic approach to world Carr, S. J., Holmes, R., Van Der Meer, J. J. M., and shorelines. Zeitschrift fur Geomorphologie. 8: 127-142. Rose, J. 2006. The Last Glacial Maximum in the North Sea Basin: micromorphological evidence of extensive Davison, I., Bosence, D.W., Alsop, G.I. & Alawi, M. 1996. glaciation. Journal of Quaternary Science. 21: 131-153. Deformation and sedimentation around active Miocene salt 120  Mapping Doggerland diapirs on the Tihama Plain, northwest Yemen. In: G. I. Archaeological Resource Management in the UK, Second Alsop, D. J. Blundell, and I. Davison, (eds.) Salt Edition. Sutton. 250-270. Tectonics. Geological Society, London, Special Publications. 100: 23-39. Fairclough G and Rippon S. 2002. Europe's Cultural Heritage: archaeologists and the management of change. Dawson M. 2000 (ed). Prehistoric, Roman and post- Europae Archaeoloiae Consilium. Brussels. Roman landscapes of the Great Ouse Valley (Council for British Archaeology Research Report 119) York: Faught, M. K. 1988. Inundated Sites in the Apalachee Bay Council for British Archaeology. area of the Eastern Gulf of Mexico. Florida Anthropologist. 41: 185-190. De Serres, B., Roy, A. G., Biron, P. M., and Best, J. L. 1999. Three-dimensional structure of flow at a conflu- Faught, M. K. 2004. The Underwater Archaeology of Pa- ence of river channels with discordant beds. Geomor- laeolandscapes, Apalachee Bay, Florida. American Antiq- phology. 26: 313-335. uity, 69 (2): 275-289. Delgado, J. 1998. Encyclopedia Of Underwater And Fenies, H., de Resseguier, A., and Tastet, J.-P. 1999. Inter- Maritime Archaeology. Yale University Press. tidal clay-drape couplets (Gironde estuary, France). Sedi- mentology. 46: 1-15. Den Hartog Jager, D., Giles, M. R., and Griffiths, G. R. 1993. Evolution of Palaeogene submarine fans of the Fischer, A. 1995. An Entrance to the Mesolithic world North Sea in space and time. In: J. R. Parker, (ed) Pe- beneath the Ocean. Status of ten years' work on the Danish troleum geology of Northwest Europe: Proceedings of sea floor. In A. Fischer (ed) Man and Sea in the Mesolithic: the 4th Conference. Geological Society of London, Bath, Coastal settlement above and below present sea level, 59-72. Proceedings of the International Symposium, Kalundborg, Denmark 1993. Oxbow. Dix, J., Quinn, R., and Westley, K. 2004. A Re- assessment of the Archaeological Potential of Continen- Fischer, A. 2004. Submerged Stone Age: Danish examples tal Shelves. Located at: and North Sea potential. In: N. C. Flemming (ed.) Subma- http://www.arch.soton.ac.uk/Research/Aggregates/shelv rine Prehistoric Archaeology of the North Sea. CBA Re- e-report.htm search Report 141: 21-36. Donovan, D. T. 1965. Holocene origin of erosion hol- Fitch, S., Thomson, K., and Gaffney, V. L. 2005. Late lows in the North Sea. Challenger Society Annual Re- Pleistocene and Holocene depositional systems and the pa- port. 3 (17): 43-44. laeogeography of the Dogger Bank, North Sea. Quaternary Research 64: 185-196. Donovan, D. T. 1975. The Geology and origin of the Silver Pit and other closed basins in the North Sea. Fitch, S., Gaffney, V. L., and Thomson, K. 2007. In Sight Proceedings of the Yorkshire Geological Society. 39 of Doggerland: From speculative survey to landscape explo- (2): 267-293. ration. Internet Archaeology 22. (http://intarch.ac.uk). Dortch, C. 1997. Prehistory Down Under: archaeologi- Flemming, N. C. 2002. The scope of Strategic Environ- cal investigations of submerged Aboriginal sites at Lake mental Assessment of North Sea areas SEA3 and SEA2 in Jasper, Western Australia. Antiquity 71 (271): 116-123. regard to prehistoric archaeological remains. Department of Trade and Industry. Report TR_014. Dyer, K. R., and Huntley, D. A. 1999. The origin, clas- sification and modelling of sand banks and ridges. Con- Flemming, N. C. 2003. The scope of Strategic Environ- tinental Shelf Research. 19: 1285-1330. mental Assessment of Continental Shelf Area SEA4 in re- gard to prehistoric archaeological remains. Department of Eisma, D. 1975. Holocene sedimentation in the Outer Trade and Industry. Silver Pit Area (Southern North Sea). Marine Science Communications. 1: 407-426. Flemming N. C. 2004. Submarine Prehistoric Archae- ology of the North Sea. CBA Research Report 141. Emery, D. and Myers, K.J. 1996. Sequence Stratigra- phy. Blackwell Science. Flemming, N. C. 2004b. The scope of Strategic Environ- mental Assessment of North Sea areas SEA5 in regard to English Heritage. (nd). (http://www.english- prehistoric archaeological remains. Department of Trade heritage.org.uk/server/show/nav.1315) and Industry. Fairclough, G. J. 2006. From assessment to Flemming, N. C. 2005. The scope of Strategic Environ- characterisation. In J. Hunter, and I. Ralston, (eds.) mental Assessment of Irish Sea Area SEA6 in regard to 121  Mapping Doggerland prehistoric archaeological remains. Department of Valley, in D. N. Smith, M. B. Brickley and W. Smith (eds). Trade and Industry Report. Fertile Ground: Papers in Honour of Professor Susan Limbrey (AEA Symposia no. 22). Oxford: Oxbow Books 53- Flemming N. C. 2006. Area SEA6 with regard to 67. prehistoric and early historic archaeological remains. Department of Trade and Industry. Grimm, E. 1991. TILIA and TILIA.GRAPH. Illinois State (http://www.offshore- Museum. sea.org.uk/site/scripts/category_info.php?categoryID=3 7) Groom, D., and Oxley, I. 2001. Maritime Fife: managing Fife's underwater heritage: a feasibility study for a mari- Friedrichs, C. T., and Aubrey, D. G. 1988. Non-linear time archaeological GIS. In D. W. Wheatley, G. P. Earl tidal distortion in shallow well mixed estuaries: a syn- and S. J. Poppy (eds.) Contemporary Themes in Archaeo- thesis. Estuarine and Coastal Shelf Science. 27: 521- logical Computing. Oxbow Books. 546. Gutowski, M. 2005. 3D high-resolution sub-bottom profil- Fuglestveldt, I. 2003. Encultring the Landscape beyond ing - 3D Chirp. Hydro International. Doggerland. In L. Larsson, H. Kindgren, K. Knutsson, D. Leoffler, and A. Åkerlund (eds.). Mesolithic on the Gutowski, M., Bull, J., Dix, J., Henstock, T., Hogarth, P., Move: Papers presented at the Sixth International Leighton, T., and White, P. 2005. True 3D high resolution Conference on the Mesolithic in Europe, Stockholm sub-bottom profiling: 3D Chirp system design and data ex- 2000. Oxbow. 103-106. amples. Geophysical Research Abstracts, 7: No.06616, EGU. Funnel, B. M. 1995. Global sea-level and the (pen-) insularity of late Cenozoic Britain. In R. C. Preece (ed.) Harris, P. T. 1988. Large-scale bedforms as indicators of Island Britain: a Quaternary Perspective, 3 -13. mutually evasive sand transport and the sequential infilling London: Geological Society. of wide-mouthed estuaries. Sedimentary Geology. 57: 273- 298. Gaddy, D. E. 2003. Introduction to GIS for the Petro- leum Industry. Penwell publishers. Harris, P. T., Pattiaratchi, C. B., Cole, A. R., and Keene, J. B. 1992. Evolution of subtidal sandbanks in Moreton Bay, Gaffney, C., and Gaffney, V. 2000a. Non-invasive eastern Australia. Marine Geology. 103: 225-247. Investigations at Wroxeter at the end of the 20th Cen- tury. Archaeological Prospection 7(2). Haslett, S. K. 2000. Coastal Systems. Routledge. 218. Gaffney, C., Gater, J., Linford P., Gaffney V., and Hayes, M. O. 1975. Morphology of sand accumulation in White R. 2000b. Large-scale systematic fluxgate gra- estuaries: an introduction to the symposium. In L. E. Cronin dio-metre at the Roman city of Wroxeter. Archaeologi- (ed). Estuarine Research. Volume II: Geology and Engineer- cal Prospection 7: 81-99. ing. Academic Press, New York. 3-22. Gaffney, C., and Gater, J. 2003. Revealing the Buried Healy, F., Heaton, M., and Lobb, S. J. 1992. Excavation of Past: Geophysics for Archaeologists. Tempus Pub- a Mesolithic site at Thatcham, Berkshire. Proceedings of the lishers. Prehistoric Society. 58: 41-76. Glimmerveen J., Mol D., Post K., Reumer J. W. F., van Hill, M., Brigg, J., Minto, P., Bagnal, D., Foley, K., and der Plicht, H., de Vos, J., van Geel, B., van Reenen, and Williams, A. 2001. Guide to Best Practice in Seascape Pals, J. P. 2004. The North Sea project: the first palae- Assessment. Countryside Council for Wales. Brady ontological, palynological and archaeological results. Shipman Martin, University College Dublin. In: N. C. Flemming 2004 (ed.). Submarine Prehistoric http://www.ccw.gov.uk/Images_Client/Reports/ACF1676.pd Archaeology of the North Sea. CBA Research Report f. 141: 21-36. Horton, B. P., Innes, J. B., Shennan, I., Gehrels, W. R., Godwin H and Godwin M. E. 1933. British Lloyd, J. M., McArthur, J. J. and Rutherland, M. M. 1999. Maglemose harpoon sites. Antiquity VII (25): 36-48. The Northumberland Coast in D. R. Bridgland, B. P. Horton and J. B. Innes (eds) The Quaternary of North-East England. Goodman, D., and Nishimura, Y. 2000. Ground- Field Guide. London: Quaternary Research Association. penetrating Radar Survey at Wroxeter. Archaeological 147-165. Prospection, 7(2): 101-107. Howard, A. J. 2005. The contribution of geoarchaeology to Greenwood, M., and Smith, D. N. 2005. A survey of understanding the environmental history and archaeological Coleoptera from sedimentary deposits from the Trent 122  Mapping Doggerland resources of the Trent Valley, U.K. Geoarchaeology 20, Archeologische Rapporten 32). Amersfoort: Rijksdienst 93-107. voor het Oudheidkundig Bodemonderzoek. 183-198. Howard, A. J., and Macklin, M. G. 1999. A generic Kenyon, N. H. 1970. Sand ribbons of European tidal seas. geomorphological approach to archaeological Marine Geology. 9: 25-39. interpretation and prospection in British River valleys: A guide for archaeologists’ investigation of Holocene Kenyon, N. H., Belderson, R. H., Stride, A. H., and Johnson, Landscapes. Antiquity 73, 527-41. M. A. 1981. Offshore tidal sandbanks as indicators of net sand transport and as potential deposits. International Asso- Huntley, B., and Birks, H. J. B. 1983. An atlas of past ciation of Sedimentologists, Special Publication. 5: 257-268. and present pollen maps of Europe, 0-13,000 years BP. Cambridge University Press. Kidd, G. D. 1999. Fundamentals of 3D seismic volume visualization. The Leading Edge 18: 702-709. Indrelid, S. 1978. Mesolithic economy and settlement patterns in Norway. In P. A. Mellars (ed.) The early Klemperer, S., and Hobbs, R. 1991. The BIRPS atlas: deep postglacial settlement of Northern Europe: An seismic reflection profiles around the British Isles. Cam- ecological perspective. Duckworth. 147-76. bridge University Press. Jackson, M. P. A., and Talbot, C. J. 1994. Advances in Knight, D., and Howard. A. J. 2005. Trent Valley Land- Salt Tectonics. In: P. L. Hancock, (ed.) Continental scapes. Heritage, Marketing and Publications Ltd. Deformation. Pergamon Press, 159-179. Knighton, D. 1998. Fluvial Forms and Processes: A New Jacobi, R. M. 1973. Aspects of the Mesolithic Age in Perspective. Arnold. 383. Great Britain. In S. K. Kozlowski (ed.) The Mesolithic in Europe. Warsaw University Press. 237-266. Kooijmans, L. P. 1971. Mesolithic bone and antler implements from the North Sea and from the Netherlands. Jacobi, R. 1976. Britain inside and outside Mesolithic Berichtien van de Rijksdienst voor het Oudheidkunde Europe Proceedings of the Prehistoric Society 42, 67- Bodemonderzoek, 20-21: 27-73. 84. Kvamme, K. 2006. Integrating multidimensional geo- Jacobi, R.M. 1978. The Mesolithic of Sussex. In. P.L physical data. Archaeological Prospection 13: 57-72. Drewett (ed.) Archaeology in Sussex to AD 1500. Council for British Archaeology Research Report, 29, Lambeck, K. 1995. Predicted Shoreline from Rebound 15-22. Models. Journal of the Geological Society 152: 437-448. Jelgersma, S. 1979. Sea-level changes in the North Sea Lambeck K. 1996. Shoreline reconstructions for the Persian basin. In: E. Oele, R. T. E. Schuttenhelm, and A. J. Gulf since the last glacial maximum. Earth and Planetary Wiggers, (eds) The Quaternary history of the North Sea. Science Letters 142: 43-57. Acta Universitatis Upsaliensis: Symposium Universita- tis Upsaliensis Annum Quingentesimum Celebrantis: 2: Larminie, F. G. 1988. Silverwell Sheet (54N 02E): Sea Bed 233-248. Sediments and Holocene Geology. HMSO, London. Jenyon, M. K. 1986. Salt Tectonics. Elsevier, London. Laraminie, F. G. 1989a. Silverwell Sheet (54N 02E): Seabed Sediments and Holocene Geology. London, HMSO. Johnson, H. D., and Baldwin, C. T. 1996. Shallow Clastic Seas. In H. G. Reading (ed.) Sedimentary Envi- Larminie, F. G. 1989b. Silverwell Sheet (54N 02E): ronments: Processes, Facies and Stratigraphy. Black- Quaternary Geology. HMSO, London. well Science. 232-280. Lawley, R., and Booth, S. 2004. Skimming the Surface. Kenward, H. K., Engleman, C., Robertson, A., and Geoscientist 14: 4-6. Large, F. 1986. Rapid scanning of urban archaeological deposits for insect remains. Circaea 3, Long, D., Wickham-Jones, C. R., and Ruckley, N. A. 1986. 163-172. A flint artefact from the northern North Sea. In D. Roe (ed.) Studies in the Upper Palaeolithic of Britain and North Kenward, H. K., and Hall, A. R. 2006. Easily decayed Western Europe. BAR International Series. 296: 55-62. organic remains in urban archaeological deposits: value, threats, research directions and conservation, in O. Lowe, J. J., and Walker, M. J. 1997. Reconstructing Brinkkemper, J. Deeben, J. van Doesburg, D. Hallewas, Quaternary Environments. 2nd Edition. Prentice Hall. E. M. Theunissen, and A. D. Verlinde, (eds), Vakken in vlakken: Archeologische kennis in lagen. (Nederlandse 123  Mapping Doggerland Ludwick, J. C. 1975. Tidal currents, sediment transport (Seamap-3D case study). Geophysical Research Abstracts 8. and sand banks in Chesapeake Bay entrance. In, L. E. Cronin, (ed.) Estuarine Research. Volume II: Geology Murray, J. W. 1987. Biogenic indicators of suspended and Engineering. Academic Press, New York. 365-380. sediment transport in marginal marine environments: quantitative examples from SW Britain. Journal of the Lumsden, G. I. 1986a. California Sheet (54N 0E): Geological Society, London, 144: 127-133. Quaternary Geology. HMSO, London. Neubauer W., and Eder-Hinterleitner, A. 1997. 3D- Lumsden, G. I. 1986b. California Sheet (54N 0E): Interpretation of Postprocessed Archaeological Magnetic Seabed Sediments. HMSO, London. Prospection Data. Archaeological Prospection 4: 191-205. Maarleveld, T. J. and Peeters, H. 2004. Can we Newell, R. C., Seiderer, L. J., and Hitchcock, D. R. 1998. manage? In N. C. Flemming (ed). Submarine The impact of dredging works in coastal waters: a review of Prehistoric Archaeology of the North Sea. CBA the sensitivity to disturbance and subsequent recovery of research Report. 141: 102-111. biological resources on the seabed. Oceanography and Marine Biology: an Annual Review, 36: 127-178. Marks, B. S., and Faught, M. K. 2003. Another Prehis- toric Site submerged on the continental shelf of North- Nordqvist, B. 1995. Coastal Adaptations in the Mesolithic: west Florida. Current Research in the Pleistocene 20: A study of coastal sites with organic remains from the Boreal (8JE1577) 49-51. and Atlantic periods in Western Sweden. GOTARC series B, 13. Gothenburg University. Erbrink and Tacoma. Masselink, G., and Hughes, M. G. 2003. Introduction to Coastal Processes and Geomorphology. Hodder Nummedal, A. 1924. Om flintpladsene. Norsk geologisk Arnold. Tidsskrift. Oslo. 7: 89-141. McCave, I. N. 1971. Wave effectiveness at the sea bed Nygaard, S. 1990. Mesolithic Western Norway. In P. M. and its relationship to bed-forms and deposition of mud. Vermeersch and P. Van Peer (eds.) Contributions to the Journal of Sedimentary Petrology. 41: 89-96. Mesolithic in Europe. Leuven University Press. 227-237. Melis, T. S., Webb, R. H., Griffiths, P. G., and Wise, T. Oxley, I. nd. Developments in marine Archaeological W. 1994. Magnitude and frequency data for historic Resource management Relevant to the Revision of the debris flows in Grand Canyon National Park and Eastern Counties Research Framework. Vicinity, Arizona. US Geological Survey Water (http://www.eaareports.demon.co.uk/FW_Oxley.pdf) Resources Investigations Report. 94-4214. Oxley I and O'Regan D. 2001. The Marine Archaeological Milner, N., and Woodman, P. 2004. Looking into the Resource. Institute Of Field Archaeologists Paper No. 4. canon's mouth: Mesolithic studies in the 21st century. In N. Milner and P. Woodman (eds.) Mesolithic Studies (www.archaeologists.net/modules/icontent/inPages/docs/pub at the beginning of the 21st Century. Oxbow. s/maritime_resource.pdf -) Milner, N. 2004. Seasonal Consumption Practices in Pedersen, L., Fischer, A., Aaby, B. (eds). 1997. The Danish the Mesolithic: Economic, Environmental, Social or Storebaelt since the Ice Age. A/S Storebaelt Fixed Link. Ritual? In N. Milner and P. Woodman (eds.) Mesolithic Studies at the beginning of the 21st Century. Oxbow. Peeters, J. H. M. 2006. Hoge Vaart-A27 in Context: Towards a model of Mesolithic-Neolithic land use dynamics Mithen, S. 1999. Hunter-gathers of the Mesolithic. In as a framework for archaeological heritage management. J. Hunter and I. Ralston (eds.) The Archaeology of Rijkedienst voor arccgeologie cultuurlanschap en Britain. Routledge. 35-58. monumenten. Amersfoort. Mithen, S. 2003. After the Ice. Routledge. Peltier, W. R. 2004. Global Glacial Isostasy and the Surface of the Ice-Age Earth: The ICE-5G (VM2) Model and Moore, P. D., Webb, J. A. and Collinson, M. E 1991. GRACE. Ann. Rev. Earth and Planet. Sci., 32: 111-149. Pollen Analysis (Second Edition). London: Blackwell. Paoletti, V., Secomandi, M., Priomallo, M., Giordano, F., Morrison, A. 1980. Early Man in Britain and Ireland. Fedi, M., and Repolla, A. 2005. Magnetic Survey at the Croom Helm. Submerged Archaeological Site of Baia, Naples, Southern Italy. Archaeological Prospection 12: 51-59. Mueller, C., Luebke, H., Woelz, S., Jokisch, T., Wendt, G., and Rabbel, W. 2006. Marine 3-D seismic investi- Posamentier, H. W. 2005. Application of 3D seismic visu- gation of a late Ertebolle settlement in Wismar Bay alisation techniques for seismic stratigraphy, seismic geo- 124  Mapping Doggerland morphology and depositional systems analysis: exam- Rowley-Conwy, P. 1983. Sedentary hunters: the Ertebolle. ples from fluvial to deep-marine depositional environ- In G. N. Bailey (ed.) Hunter-Gather Economy in Prehistory: ments. In: A. G. Dore and B. A. Vining, (eds.) North- a European Perspective. Cambridge University Press. 111- West Europe and Global Perspectives: Proceedings of 126. the 6th Petroleum Geology Conference. Geological Society of London 1565-1576. Ryan, W., and Pitman, M. 2000. Noah's Flood: The New Scientific Discoveries About the Event That Changed His- Praeg D., 1997. Buried fluvial channels: 3D-seismic tory. Schuster. geomorphology. In: T. A. Davies, T. Bell, A. K. Cooper, H. Josenhans, L. Polyak, A. Solheim, M. S. Salomonsen, I., and Jensen, K. A. 1994. Quaternery ero- Stoker, J. A. Stravers, (eds.), Glaciated Continental sional surfaces in the Danish North Sea. Boreas 23: 244- Margins: An Atlas of Acoustic Images. Chapman and 253. Hall. 162-163. Shackleton, N. J. 1987. Oxygen isotopes, ice volumes and Praeg, D. 2003. Seismic imaging of mid-Pleistocene sea level. Quaternary Science Reviews. 6: 183-190. tunnel-valleys in the North Sea Basin - high resolution from low frequencies. Journal of Applied Geophysics Shennan, I., Lambeck, K., Flather, R., Horton, B., McArthur, 53: 273-298. J., Innes, J., Lloyd, J., Rutherford, M., and Wingfield, R. 2000. Modelling western North Sea palaeogeographies and Price, W. A. 1963. Patterns of flow and channelling in tidal changes during the Holocene in I. Shennan and J. An- tidal inlets. Journal of Sedimentary Petrology. 33: 279- drews (eds.) Holocene Land-Ocean Interaction and Envi- 290. ronmental Change around the North Sea. Geological Soci- ety, London, Special Publication 166: 299-319. Reid, C. 1913. Submerged Forests. Oxford. Shennan, I., and Horton, B. 2002. Holocene land and sea- Reynier, M. J. 1988. Early Mesolithic settlement in level changes in Great Britain. Journal of Quaternary Sci- England and Wales: some preliminary observations. In ence 17 (5-6): 511-526. N. Ashton, F. Healy and P. Pettitt (eds.) Stone Age Archaeology: Essays in Honour of John Wymer. Sherrif, R. E. 1977. Limitations on resolution of seismic Oxbow (Oxbow Monograph 102/Lithic Studies Society reflections and geologic detail deriveable from them. In: C. Occasional Paper 6). 174-184. E. Payton, (ed.) Seismic Stratigraphy - Applications to hy- drocarbon Exploration. Memoir of the American Associa- Reynier, M. J. 2005. Early Mesolithic Britain: Origins, tion of Petroleum Geologists, Tulsa. development and directions. BAR British Series 393. Archaeopress. Sidell, J., Williams, K., Scaife, R., and Cameron, N. 2000. The Holocene Evolution of the London Thames (MoLAS Rhoads, B. L., and Kentworthy, S. T. 1995. Flow Monograph 5). London: Museum of London Archaeology structure at an asymmetrical stream confluence. Service. Geomorphology. 11: 273-293. Smith, C. 1992. Late Stone Age Hunters of the British Isles. Roberts, A. J. 1987. Late Mesolithic occupation of the Routlege Publishers. Cornish coast at Gwithian: preliminary results. In P. Orwley-Conwy, M. Zvelebil and H. P. Blankholm (eds.) Smith, D. N., Roseff, R., Bevan, L., Brown, A. G., Butler, S, Mesolithic North West Europe: Recent Trends. Hughes, G., and Monckton A. 2005. Archaeological and Department of Archaeology and Prehistory, University Environmental Investigations of a Late Glacial and Holocene of Sheffield. 131-8. river valley sequence on the River Soar, at Croft, Leicestershire. The Holocene 15: 353-377. Roberts, P., and Trow, S. 2002. Taking to the Water: English Heritage’s Initial Policy for The Management Stewart, S. A. 1999. Seismic interpretation of circular of Maritime Archaeology in England. Unpublished geological structures. Petroleum Geoscience. 5: 273-285. policy statement, English Heritage, (http://www.english- heritage.org.uk/server/show/nav.8381). Stride, A.H., Belderson, R.H., Kenyon, N.H. & Johnson, M.A. 1982. Offshore tidal deposits: sand sheet and sand bank Robinson, A. H. W. 1960. Ebb-flood channel systems facies. In: Stride, A.H. (ed) Offshore tidal sands: processes in sandy bays and estuaries. Geography. 45: 183-199. and deposits. London, Chapman and Hall, 95-125. Robinson, A. H. W. 1968. The submerged glacial Stright, M. J. 1986. Evaluation of archaeological site poten- landscape off the Lincolnshire coast. Transactions of tial on the Gulf of Mexico continental shelf using high- the Institute of British Geographers. 44: 119-132. resolution seismic data. Geophysics 51 (3): 605-622. 125  Mapping Doggerland Summerfield, M. A. 1991. Global Geomorphology. Southern North Sea. Environmental Archaeology 11: 207- Longman. 218. Talbot, C.J. & Alavi, M. 1996. The past of a future Warren, G. 2005. Mesolithic Lives in Scotland. Tempus syntaxis across the Zagros. In: Alsop, G.I., Blundell, Publishers. D.J. & Davison, I. (eds) Salt Tectonics. Geological Society, London, Special Publications, 100, 89-109. Warren, S. H., Clark, J. G. D., Godwin, M. E., Godwin, H., MacFadyen, W. A. 1934. An Early Mesolithic Site at Terrell, N., Edwards, H., Scoffield, P., and Martin, M., Broxbourne Sealed Under Boreal Peat. The Journal of the 2005. PGS Faroe Shetland Basin MegaSurvey – the Royal Anthropological Institute of Great Britain and Ireland Key to New Discoveries in a Maturing Frontier Area? LXIV: 101-128. In: H. Ziska, (ed.) Faroe Islands Exploration Conference: Proceedings of the 1st Conference, Annales Watters, M. S. 2006. Geovisualization: an example from the Societatis Scientiarum Færoensis (Faroese Society of Catholme Ceremonial Complex. Archaeological Prospec- Sciences and Humanities). Supplementum 43, tion. 13: 282-290. Tórshavn. 176-182. Wenban-Smith, 2002. Palaeolithic and Mesolithic Tetlow, E. A. 2005. The palaeoentomology of the salt Archaeology on the Sea Bed: Marine Aggregate Dredging marches and coastal woodlands of the Severn estuary. and the Historic Environment. Unpublished report for PhD Thesis, Institute of Archaeology and Antiquity, BMAPA and RCHME, Wessex Archaeology, University of Birmingham, UK. (http://www.bmapa.org/media.htm). Thomson, K. 2004. Overburden deformation associ- Wessex Archaeology. 2003. Artefacts from the Sea. ated with halokinesis in the Southern North Sea: impli- (http://www.wessexarch.co.uk/projects/marine/alsf/artefacts_ cations for the origin of the Silver Pit Crater. Visual sea/artefacts_sea.html). Geosciences 9: 1-9. Wessex Archaeology. 2005. England's Historic Seascapes. Valentin, H. 1957. Glazialmorphologische Historic Environment Characterisation in England's untersuchungen in Ostengland: ein beitrag sum problem Intertidal and Marine Zone. (http://www.english- der letzen vereisung im Nordseeraum. Abhandlungen heritage.org.uk/server/show/nav.8684). des Geographischen Instituts der Freien Universität Berlin. 4: 1-86. Wickham-Jones, C. R., and Dawson, S. 2006. The scope of Strategic Environmental Assessment of North Sea areas Van Kolfschoten, T., and Van Essen, H. 2004. Palaeo- SEA7 with regard to prehistoric and early historic ar- zoological heritage from the bottom of the North Sea. chaeological remains. Department of Trade and Industry. In N. C Flemming (ed.) Submarine Prehistoric Archae- (http://www.offshore- ology of the North Sea, CBA Research Report 141. sea.org.uk/site/scripts/category_info.php?categoryID=37) Veenstra, H. J. 1965. Geology of the Dogger Bank Whitehouse, N. J., and Smith, D. N. 2004. “Islands” in area, North Sea. Marine Geology. 3: 245-262. Holocene forests: implications for forest openness, landscape clearance and “culture-steppe” species. Environmental Velegrakis, A. F., Dix, J. K. and Collins, M. B. 1999. Archaeology. 9(2): 203-212. Late Quaternary evolution of the upper reaches of the Solent River, Southern England, based upon marine Wingfield, R. T. R. 1990. The origin of major incisions geophysical evidence. Journal of the Geological Soci- within the Pleistocene deposits of the North Sea. Marine ety, London. 156: 73-87. Geology. 91: 31-52. Verhart, L. B. M. 2004. The implications of prehistoric Wright, L. D., Coleman, J. M., and Thom, B. G. 1975. finds on and off the Dutch coast. In N. C. Flemming Sediment transport and deposition in a macrotidal river (ed.) Submarine Prehistoric Archaeology of the North channel: Ord River, Western Australia. In, L. E. Cronin, Sea. CBA research Report. 141: 57-64. (ed.) Estuarine Research. Volume II: Geology and Engineering. Academic Press, New York. 309-321. Waddington, C., Bayliss, A., Bailey, G., Boomer, I., Milner, N., Pedersen K., and Shiel R. 2003. A Wymer, J. 1991. Mesolithic Britain. Princes Risborough: Mesolithic coastal site at Howick, Northumberland. Shire Publications. Antiquity. 77 (295). Ward, I., Larcombe, P., and Lillie, M. 2006. The dating of Doggerland – post glacial geochronology of the 126  Mapping Doggerland Mapping Doggerland: The prehistoric landscapes of the North Sea basin are amongst the most enigmatic archaeological landscapes in northwestern Europe. Whilst the region contains one of the most extensive and, probably, best preserved hunter-gatherer landscapes in Europe, global warming resulted in the loss of a vast area of habitable land over a period of c.11,000 years. The challenge to investigate, interpret and manage the heritage of this extraordinary, but largely inaccessible landscape is enormous. 3D seismic datasets, acquired to explore deep geology, present a major opportunity to explore this landscape at a regional scale. The North Sea Palaeolandscapes Project utilised c. 23,000 km2 of 3D seismic data to provide detailed digital mapping of Late Pleistocene and Holocene topographic features across the area. This publication provides an assessment of the archaeological potential of the interpreted landscape and related environmental sources. It outlines associated archaeological issues, a methodology for implementing historic landscape characterisation within the area, as well as an assessment of data sources for further exploration of the British coastal shelf. The results of this study will be of interest to archaeologists, geomorphologists and cultural resource managers working in analogous environments, whilst the methodology outlined may be applied to similar landscapes with comparable supporting data. La Cartographie de Doggerland Les paysages préhistoriques du bassin de la Mer du Nord sont parmi les plus énigmatiques de l’Europe du Nord-Ouest. Tout en admettant que la région possède un des plus vastes paysages, et selon toute probabilité, le paysage le mieux conservé des chasseurs-cuilleurs en Europe, le réchauffement de la planète a entraîné la perte d’une région énorme de terrain habitable à travers une période d’environ 11,000 ans. Le défi énorme est d’étudier, d’interpréter, et de gérer l’héritage de ce paysage qui est à la fois extraordinaire et, pour la plupart, inaccessible. Des données sismiques à trois dimensions qui ont été acquises pour explorer la géologie profonde, nous offrent une opportunité importante d’explorer ce paysage sur un plan régional. Le Projet des Paléo-paysages de la Mer du Nord a utilisé environ 23,000 km2 de données sismiques en relief afin de créer une carte digitale détailée des caracteristiques topographiques de la fin de l’ère Pleistocene et Holocene de la région. Cette publication fournit une évaluation des possibilités archéologiques du paysage interprété et des sources environnementales. Elle met en relief des questions archéologiques, une méthodologie pour réaliser la caractéristaion des paysages historiques dans ce domaine, et une évaluation des sources des données pour explorer d’avantage l’écueil anglais. Les résultats de cette étude intéresseront des archéologistes, des géomorphologistes, et des directeurs de ressources culturelles qui travaillent dans des environnements pareils, et, on pourra utiliser la méthodologie soulignée avec d’autres paysages qui sont comparables et qui ont des données similaires. Die kartographische Aufnahme von ‘Doggerland’ Die prähistorischen Landschaften unter der Nordsee gehören zu den rätselhaftesten archäologischen Landstrichen des nordwestlichen Europas. Obgleich die Region eine der weiträumigsten und, wahrscheinlich, eine der best erhaltensten Landschaften prähistorischer Jäger und Sammler in Europa umfaßt, hat die Erwärmung der Erdatmosphäre über einen Zeitraum von ungefähr elftausend Jahren zum Verlust einer ausgedehnten Fläche bewohnbaren Landes geführt. Die Herausforderung, diese außergewöhnliche aber größtenteils unerreichbare Landschaft zu untersuchen, zu interpretieren und ihre Hinterlassenschaft zu bewahren, ist enorm. Seismische 3D Datenbestände, die zur geologischen Erforschung tieferer Schichten angelegt wurden, bieten uns eine einzigartige Gelegenheit, diese Landschaft auf regionaler Ebene zu erkunden. Das Nordsee Paläo-Landschafts Projekt hat seismische 3D Daten auf einer Fläche von ungefähr 23 000 km2 genutzt um eine detaillierte Karte der topographischen Merkmale dieses Gebietes während des späten Pleistozäns und des Holozäns herzustellen. Dieses Buch bietet eine Beurteilung des archäologischen Potentials dieser Landschaft und eine Einschätzung der verwandten Quellen zur Umwelt. Es umreißt die damit verbundenen archäologischen Fragen, skizziert eine Methodik für die Erfassung des Charakters historischer Landschaften in diesem Gebiet und beurteilt andere Arten von Materialien für die weitere Erforschung des Meeresbodens vor der britischen Küste. Die Ergebnisse dieser Studie sind für alle Archäologen, Geomorphologen und Leiter kultureller Ressourcen, die auf diesem und ähnlichem Gebiet tätig sind, von Interesse, während die hier vorgeschlagene Methodik auch auf andere Landschaften dieser Art, für die vergleichbare Daten zur Verfügung stehen, angewandt werden kann. 127 Mapping Doggerland 128