Sediments, Morphology and Sedimentary Processes on Continental Shelves E-Book

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: International Association Of Sedimentologists Series
• Sprache: Englisch

Seitenzahl: 972

Veröffentlichungsjahr: 2012

Beschreibung

The application of multibeam and sediment transport measurement technologies and the adoption of multi-faceted research methodologies have greatly advanced our understanding of the sedimentary processes on continental shelves in the last decade.

This book uniquely blends  cutting-edge research and state-of-the art review articles that take stock of new advances in multibeam mapping and sediment transport technologies, spatial analysis and modelling, and the applications of these advances to the understanding of shelf sediments, morphodynamics, and sedimentary processes. Case studies are also presented to illustrate the utilization of seabed property and process knowledge in habitat mapping and ocean management

With its mix of papers focusing on technological advances, integration of shelf morphology and processes, and the application of these advances to coastal and ocean management, this Special Publication volume will serve as a milestone reference for professional marine scientists and as advanced text for students in marine geology, sedimentology and oceanography.

This book is part of the International Association of Sedimentologists (IAS) Special Publications.

The Special Publications from the IAS are a set of thematic volumes edited by specialists on subjects of central interest to sedimentologists. Papers are reviewed and printed to the same high standards as those published in the journal Sedimentology and several of these volumes have become standard works of reference.

Leseprobe

Contents

Cover

Other publications of the International Association of Sedimentologists

Special Publications

Reprint Series

Title Page

Copyright

Preface

Part 1: Sediments and morphology in shelf and coastal systems

Chapter 1: Optimal Use of Multibeam Technology in the Study of Shelf Morphodynamics

Introduction

Bathymetry

Backscatter

Demonstration of Change Detection Capability and Limitations

Conclusions

Acknowledgements

References

Chapter 2: Palaeogeographic Reconstruction of Hecate Strait British Columbia: Changing Sea Levels and Sedimentary Processes Reshape a Glaciated Shelf

Introduction

Setting

Data Collection

Results

Discussion

Conclusions

Acknowledgements

References

Chapter 3: Changes in Submarine Channel Morphology and Slope Sedimentation Patterns From Repeat Multibeam Surveys in the Fraser River delta, western Canada

Introduction

Setting

Methods

Results

Discussion

Conclusions

Acknowledgements

References

Chapter 4: Recent Sedimentary Processes in the Cap de Creus Canyon Head and Adjacent Continental Shelf, NE Spain: Evidence from Multibeam Bathymetry, Sub-Bottom Profiles and Coring

Introduction

Methodology

Results

Discussion

Conclusions

Acknowledgements

References

Chapter 5: Geology Metrics for Predicting Shoreline Change Using Seabed and Sub-Bottom Observations from the Surf Zone and Nearshore

Introduction

Study Area

Methodology

Results

Discussion

Conclusions

Acknowledgements

References

Chapter 6: Re-Examination of Sand Ridges on the Middle and Outer New Jersey Shelf Based on Combined Analysis of Multibeam Bathymetry and Backscatter, Seafloor Grab Samples and Chirp Seismic Data

Introduction

Background Information on Sand Ridge Evolution

Data

Results

Discussion

Conclusions

Acknowledgements

References

Chapter 7: Sedimentary Facies of Shoreface-Connected Sand Ridges off the East Frisian Barrier-Island Coast, Southern North Sea: Climatic Controls and Preservation Potential

Introduction

Physical Setting

Methods

Results

Discussion

Conclusions

Acknowledgements

References

Part 2: Sediment transport processes, sedimentation and modelling

Chapter 8: Recent Advances in Understanding Continental Shelf Sediment Transport

Introduction

Bottom Boundary Layer Hydrodynamics and Sediment Suspension

The Complexity of Shelf Sedimentation: Future Research Challenges

Acknowledgements

References

Chapter 9: Recent Advances in Instrumentation Used to Study Sediment Transport

introduction

Measuring the Fluid Flow

Measuring Bedload

Measuring Suspended Sediment Concentration

Measuring bed Morphology

Measuring Particle Size and Settling Velocity

Putting it all Together

Future Prospects

References

Chapter 10: Seabed Disturbance and Bedform Distribution and Mobility on the Storm-Dominated Sable Island Bank, Scotian Shelf

Introduction

Study Region and Methods

Results

Discussion

Conclusions

Acknowledgements

References

Chapter 11: Temporal Variability, Migration Rates and Preservation Potential of Subaqueous Dune Fields Generated in the Agulhas Current on the Southeast African Continental Shelf

Introduction

Physical Setting

Methods

Results

Discussion and Conclusions

Acknowledgements

References

Chapter 12: Measurement of Bedload Transport in a Coastal Sea Using Repeat Swath Bathymetry Surveys: Assessing Bedload Formulae Using Sand Dune Migration

Introduction

Study Area

Methods

Results

Discussion

Conclusion and Suggestions for Further Work

Appendix

Acknowledgements

Nomenclature

References

Chapter 13: Analyzing Bedforms Mapped Using Multibeam Sonar to Determine Regional Bedload Sediment Transport Patterns in the San Francisco Bay Coastal System

Introduction

Study Area

Previous Work

Methods

Results and Discussion

Conclusions

Acknowledgements

References

Chapter 14: Sediment Transport on Continental Shelves: Storm bed Formation and Preservation in Heterogeneous Sediments

Introduction

Shelf Sedimentation During Storms

Modern Storm Bed Deposition

Relating Modern Storm Beds to the Rock Record

Summary and Conclusions

Acknowledgements

References

Chapter 15: Tidal Influence on the Transport of Suspended Matter in the Southwestern yellow Sea at 6 ka

Introduction

Study Area

Methods

Results

Discussion

Summary and Conclusion

Acknowledgements

References

Chapter 16: Origin, Transport Processes and Distribution Pattern of Modern Sediments in the Yellow Sea

Introduction

General Aspects of the Yellow Sea

Materials and Methods

Results

Discussion

Conclusions

Acknowledgements

References

Chapter 17: Seasonal and Spatial Variation in Suspended Sediment Characteristics off the Changjiang Estuary

Introduction

Materials and Methods

Results

Discussion

Conclusions

Acknowledgements

References

Chapter 18: Factors Controlling Downward Fluxes of Particulate Matter in Glacier-Contact and Non-Glacier Contact Settings in a Subpolar Fjord (Billefjorden, Svalbard)

Introduction

Study Area

Materials and Methods

Results

Discussion

Conclusions

Acknowledgements

References

Part 3: Application and management

Chapter 19: On Seabed Disturbance, Marine Ecological Succession and Applications for Environmental Management: A Physical Sedimentological Perspective

Introduction

Disturbance and Ecological Succession

Models of Continental Shelf Current-Related Disturbance Regimes

Conclusions

Acknowledgements

References

Chapter 20: Benthic Habitat Mapping from Seabed Acoustic Surveys: Do Implicit Assumptions Hold?

Introduction

Conclusions

Acknowledgements

References

List of Reviewers

Index

Other publications of the International Association of Sedimentologists

Special Publications

43 Quaternary Carbonate and Evaporite Sedimentary Facies and Their Ancient Analogues

      A Tribute to Douglas James Shearman

      Edited by Christopher G. St. C. Kendall and Abdulrahman S. Alsharhan

      2010, 494 pages, 263 illustrations

42 Carbonate Systems During the Olicocene-Miocene Climatic Transition

      Edited by Maria Mutti, Werner E. Piller and Christian Betzler

      2010, 304 pages, 154 illustrations

41 Perspectives in Carbonate Geology

      A Tribute to the Career of Robert Nathan Ginsburg

      Edited by Peter K. Swart, Gregor P. Eberli and Judith A. McKenzie

      2009, 387 pages, 230 illustrations

40 Analogue and Numerical Modelling of Sedimentary Systems

      From Understanding to Prediction

      Edited by P. de Boer, G. Postma, K. van der Zwan, P. Burgess and P. Kukla

      2008, 336 pages, 172 illustrations

39 Glacial Sedimentary Processes and Products

      Edited by M.J. Hambrey, P. Christoffersen, N.F. Glasser and B. Hubbard

      2007, 416 pages, 181 illustrations

38 Sedimentary Processes, Environments and Basins

      A Tribute to Peter Friend

      Edited by G. Nichols, E. Williams and C. Paola

      2007, 648 pages, 329 illustrations

37 Continental Margin Sedimentation

      From Sediment Transport to Sequence Stratigraphy

      Edited by C.A. Nittrouer, J.A. Austin, M.E. Field, J.H. Kravitz, J.P.M. Syvitski and P.L. Wiberg

      2007, 549 pages, 178 illustrations

36 Braided Rivers

      Process, Deposits, Ecology and Management

      Edited by G.H. Sambrook Smith, J.L. Best, C.S. Bristow and G.E. Petts

      2006, 390 pages, 197 illustrations

35 Fluvial Sedimentology VII

      Edited by M.D. Blum, S.B. Marriott and S.F. Leclair

      2005, 589 pages, 319 illustrations

34 Clay Mineral Cements in Sandstones

      Edited by R.H. Worden and S. Morad

      2003, 512 pages, 246 illustrations

33 Precambrian Sedimentary Environments

      A Modern Approach to Ancient Depositional Systems

      Edited by W. Altermann and P.L. Corcoran

      2002, 464 pages, 194 illustrations

32 Flood and Megaflood Processes and Deposits

      Recent and Ancient Examples

      Edited by I.P. Martini, V.R. Baker and G. Garzón

      2002, 320 pages, 281 illustrations

31 Particulate Gravity Currents

      Edited by W.D. McCaffrey, B.C. Kneller and J. Peakall

      2001, 320 pages, 222 illustrations

30 Volcaniclastic Sedimentation in Lacustrine Settings

      Edited by J.D.L. White and N.R. Riggs

      2001, 312 pages, 155 illustrations

29 Quartz Cementation in Sandstones

      Edited by R.H. Worden and S. Morad

      2000, 352 pages, 231 illustrations

28 Fluvial Sedimentology VI

      Edited by N.D. Smith and J. Rogers

      1999, 328 pages, 280 illustrations

27 Palaeoweathering, Palaeosurfaces and Related Continental Deposits

      Edited by M. Thiry and R. Simon Coinon

      1999, 408 pages, 238 illustrations

26 Carbonate Cementation in Sandstones

      Edited by S. Morad

      1998, 576 pages, 297 illustrations

25 Reefs and Carbonate Platforms in the Pacific and Indian Oceans

      Edited by G.F. Camoin and P.J. Davies

      1998, 336 pages, 170 illustrations

24 Tidal Signatures in Modern and Ancient Sediments

      Edited by B.W. Flemming and A. Bartholomä

      1995, 368 pages, 259 illustrations

23 Carbonate Mud-mounds

      Their Origin and Evolution

      Edited by C.L.V. Monty, D.W.J. Bosence, P.H. Bridges and B.R. Pratt

      1995, 543 pages, 330 illustrations

Reprint Series

4 Sandstone Diagenesis: Recent and Ancient

      Edited by S.D. Burley and R.H. Worden

      2003, 648 pages, 223 illustrations

3 Deep-water Turbidite Systems

      Edited by D.A.V. Stow

      1992, 479 pages, 278 illustrations

2 Calcretes

      Edited by V.P. Wright and M.E. Tucker

      1991, 360 pages, 190 illustrations

This edition first published 2012 © 2012 by International Association of Sedimentologists

Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell's publishing program has been merged with Wiley's global Scientific, Technical and Medical business to form Wiley-Blackwell.

Registered office: John Wiley & Sons, Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial offices: 9600 Garsington Road, Oxford, OX4 2DQ, UK

The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

111 River Street, Hoboken, NJ 07030-5774, USA

For details of our global editorial offices, for customer services and for information abouthow to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell.

The right of the author to be identified as the author of this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought.

Library of Congress Cataloging-in-Publication Data

Sediments, morphology, and sedimentary processes on continental shelves: advances in technologies, research, and applications / edited by Michael Li, Christopher R. Sherwood, Philip R. Hill.

p. cm. – (Special publication / International Association of Sedimentologists; no. 44)

Includes bibliographical references and index.

ISBN 978-1-4443-5082-1 (hardcover: alk. paper) 1. Sedimentation and deposition. 2. Sedimentology. 3. Sediment transport. 4. Continental margins. I. Li, Michael. II. Sherwood, Christopher R. III. Hill, Philip R. IV. International

Association of Sedimentologists.

QE571.S423 2012

551.3′03–dc23

2011040562

Preface

Sediments, Morphology and Sedimentary Processes on Continental Shelves: Advances in Technologies, Research, and Applications

The application of multibeam and sediment transport measurement technologies and the adoption of integrated techniques in the approach to research have greatly advanced our understanding of sediments, morphology and sedimentary processes on continental shelves in the last decade. This book focuses on the applications of new multibeam mapping and sediment transport measurement technologies, the integration of morphology and processes and the utilization of shelf seabed property and process knowledge in coastal and ocean management.

The volume grew out of a technical session “Sediments and Sedimentary Processes on Continental Shelves” organized by the editors at the 17th International Sedimentological Congress held in Fukuoka, Japan, from 27 August to 1 September, 2006. Thirteen of the twenty contributions were originally presented in that technical session. The other seven papers were solicited to cover the latest advances in multibeam mapping technology, advances and applications of sediment transport measurement techniques, and the application of sediment property and process knowledge in habitat mapping and ocean management. The articles in this book were contributed by authors from eight countries and cover a range of topics. The aim of the book is to take stock of the impact that new advances in technology, spatial analysis and modeling have brought to the understanding of shelf sedimentology. With the mix of primary research and review papers, the boo will serve as a milestone for the world's shelf sedimentology and ocean management communities.

The book is divided into three sections. Section I, “Sediments and Morphology in Shelf and Coastal Systems” opens with an article by Hughes-Clarke that reviews the application of state-of-the-art multibeam technology to shelf sedimentology research. Resolution and accuracy in both bathymetry and backscatter data are explored and examples illustrate the optimal use of multibeam technology. The next three papers demonstrate the integration of multibeam bathymetry mapping with seismic surveying and coring. Barrie & Conway use these techniques to understand how changing sea levels and sedimentary processes reshaped the glaciated shelf in Hecate Strait, British Columbia. Hill demonstrates how repeated multibeam surveys can be used to interpret significant changes in submarine channel morphology and slope sedimentation patterns in the Fraser River delta, western Canada. Garcia-Garcia et al. interpret sedimentary processes in the Cap de Creus canyon head and adjacent continental shelf, NE Spain.

Two papers apply these integrated techniques to the study of sand ridges on continental shelves. Goff and Duncan integrate multibeam bathymetry and backscatter data in an analysis of sand ridge response to the present-day hydrologic regime of the middle and outer New Jersey shelf. Son et al. investigate the sedimentary facies and preservation potential of shoreface-connected sand ridges in the southern North Sea through analyses of multibeam bathymetry, seabed samples and internal sedimentary structures derived from box-cores. The last paper in this section, by McNinch and Miselis, explores the links between shoreline changes, morphology and sediment distribution in the surf zone off the coast of North Carolina based on almost a decade of repeated field observations that included interferometric swath bathymetry, chirp sub-bottom profiles, sediment vibracores and radar data.

Eleven papers are included in Section II, “Sediment transport processes, sedimentation and modelling”. The first two contributions review advances in instrumentation and understanding in the research of sediment transport processes on continental shelves. Wright reviews highlights of shelf sediment transport research and identifies new directions, with an emphasis on gravity-driven transport, within negatively-buoyant hyperpycnal layers, as an important mechanism of across-shelf sediment transport. Williams focuses on recent technological innovations in sediment transport instrumentation and examines, by example, how these novel technologies contribute to simultaneous measurements of fluid flow, bed morphology and sediment transport at unprecedented spatial and temporal resolution in the laboratory and in the field. He also predicts how developments in instrumentation over the coming decade may enable more accurate forecasting of sediment transport processes.

The following papers present recent findings on seabed forcing, bedform distribution and migration, the implications of these to bedload transport calculations and regional sediment transport patterns and sedimentary strata formation and preservation under a range of current and storm conditions. Regional bathymetric maps, sidescan sonar and multibeam bathymetric surveys, sediment samples and model predictions of seabed disturbance are integrated by Li et al. to characterize the distribution, metrics and mobility of bedforms on the storm-dominated Sable Island Bank, Scotian Shelf. Flemming and Bartholomä examine a serial sidescan sonar data set spanning nearly 16 years to shed light on the temporal variability, migration rates and preservation potential of subaqueous dune fields beneath the Agulhas Current on the southeast African continental shelf. Duffy & Hughes-Clarke use repeat multibeam bathymetry surveys and velocity profile measurements to quantify long-term sediment transport and evaluate the performance of several formulations for predicting bedload transport rate over sand dunes in the Bay of Fundy. Barnard et al. present a comprehensive analysis of more than 3000 bedforms measured from high-resolution multibeam survey data and demonstrate how this highly detailed, quantitative information can be used to both determine regional sediment transport patterns and assist management of the sediment resource in the San Francisco Bay coastal system. Keen et al. use observations and numerical model predictions to examine the source, transport processes and deposition of modern storm beds. By comparison with historical data, they evaluate the recurrence frequency of modern storm beds and the preservation potential of such beds within a storm-dominated shelf sequence.

The last four articles in Section II discuss the origin, characteristics, seasonal and spatial variations, transport and sedimentation of suspended particulate matter. The understanding of cohesive sediment dynamics and deposition presents its own set of challenges that can be addressed by new technologies, integration of different spatial data sets and numerical modelling. Shi et al. use various sediment characteristics, suspended particulate matter concentration and trend analysis to establish a depositional process model for the Yellow Sea. Uehara & Saito use numerical modeling to examine the tidal influence on the transport of suspended matter in the southwestern Yellow Sea in the mid-Holocene. Wang et al. demonstrate the use of in-situ particle size and concentration measurements to understand seasonal changes in suspended particle dynamics off the Changjiang estuary. Szczuciski & Zajczkowski integrate hydrology and sediment trap data to investigate particulate matter fluxes and controlling factors in a sub-polar fjord in the European Arctic.

Section III reviews the application of shelf sedimentology research to habitat mapping and ocean management. Harris explores the inter-relationships between seabed disturbance, marine ecological succession and ecosystem-based management, demonstrating the need for an understanding of sedimentology in the design of marine protected areas (MPAs). The volume concludes with a conceptual article by Kostylev that examines the validity of the basic assumptions in using acoustic seabed surveys to produce benthic habitat maps. He calls for a better understanding of the processes linking benthic communities to seabed geology so that remotely-sensed geological information from, for example, multibeam surveys can be used more confidently in benthic habitat mapping.

This Special Publication volume would not have been possible without the contribution and support of a large group of people. First and foremost, the guest editors would like to acknowledge the support and cooperation from all the authors of the papers included in this book. We are grateful to the Special Publication editors Drs Ian Jarvis, Stella Bignold and Thomas Stevens for their support and advice during the editing process of this book. We acknowledge our respective institutions for allowing us to take on this endeavour in addition to our normal organizational responsibilities. Special thanks go to Nina Parry for her very able co-ordination of all correspondence during the manuscript review process. Finally, each paper in this book was evaluated by at least two reviewers and their comments and critiques have greatly improved all the papers in this volume. We deeply appreciate the effort of these reviewers who are listed at the end of this book.

The completion of this Special Publication volume is made possible by the support of our families, particularly our wives Ping, Patricia and Jennifer. We are grateful for their understanding and encouragement.

Michael Z. Li Christopher R. Sherwood Philip R. Hill

Part 1

Sediments and morphology in shelf and coastal systems

Chapter 1

Optimal Use of Multibeam Technology in the Study of Shelf Morphodynamics

John E. Hughes Clarke

Department of Geodesy and Geomatics Engineering, University of New Brunswick, Canada (E-mail: [email protected])

Abstract

Many of the recent advances in our understanding of sedimentary processes on the continental shelf have come about as a result of the use of multibeam sonar systems. These systems provide wide area coverage of seafloor variations in bathymetry and backscatter at typical horizontal resolutions as small as 2% of the water depth. The narrowest beam systems now provide backscatter data at resolutions approaching towed sidescan sonar while simultaneously providing co-registered, equivalent-resolution topography.

Even more valuable than the static view of the seabed is an ability, through resurvey, to monitor temporal variations in the seabed. By adding the time dimension, insights can be provided into the sedimentary processes rather than just the resulting sediment distribution. To achieve this, however, requires particular attention to be placed on the limitations of these survey systems, which affect repeatable accuracy. To assess the total achievable accuracy one needs to account for all the integrated components of the survey system.

In this paper, the contributions of the various sources of systematic bathymetric and backscatter error within a typical shelf multibeam survey are described. To optimize the bathymetric data, strategies for dealing with imperfections in tidal models and knowledge of the sound speed structure are described. In order to improve the backscatter data, strategies for predicting the combined effect of beam pattern residuals and the seabed angular response are detailed.

To illustrate a typical result, a pair of overlapping surveys employing widely differing source sensor resolution and accuracy is combined to try to predict the relative importance of active and relict shelf morphodynamic processes.

Keywords: Multibeam, multi-sensor integration, calibration, backscatter reduction.

Introduction

Routine application of multibeam sonar bathymetry and backscatter has revolutionized our understanding of continental shelf morphodynamics. The ability to view a near-continuous topographic surface together with variations in seabed backscatter strength provides an overview analogous to aerial photography, resulting in a vastly improved ability to interpret the seafloor sedimentary processes (Hughes Clarke et al., 1996).

One of the most immediate results of this new technology has been the recognition, for the first time, of the continuity and juxtaposition of long wavelength features such as drowned beach ridges and reefs (e.g. Gardner et al., 2005), or moraine complexes (e.g. Todd et al., 1999). But the real challenge to maximizing the usefulness of this data will lie in the finer details revealed. The detail is in the shorter wavelength morphology that lies close to the limits of resolution of these systems.

After the first pass interpretation of the current state of the shelves, future research will be increasingly focused on monitoring their temporal evolution. The first view provides a snapshot. That snapshot allows inferences to be made about likely sedimentary processes. However, proof of the activity of those processes awaits repetitive surveying. Proof that the seabed has changed requires confidence in the absolute accuracy of both the bathymetric and backscatter output of the integrated sonar system.

Obvious change, such as new slide scars (Brucker et al., 2007), overprinted iceberg scours (Sonnichsen et al., 2005), freshly emplaced debris flows (Kammerer et al., 1998) or significantly-migrated bedform positions (Duffy and Hughes Clarke, 2005) can be discerned from imperfect data. However, more subtle transitions, such as accretion of thin sand sheets, deflation of near shore sand bodies, deepening of pockmarks or migration of ripples requires a level of absolute accuracy that lie at the limit of many of the integrated systems.

This paper explores the resolution and accuracy capabilities in both bathymetry and backscatter that is realistically available from currently state-of-the-art multibeam sonar systems. Practical examples are provided, illustrating the advantages and limitations of this sort of data for shelf morphodynamic research.

Bathymetry

Resolution

The power of a multibeam system lies in its ability to resolve sedimentary structures at wavelengths small enough to infer the processes active. Many of the sediment transport mechanisms can be inferred from the short wavelength relief. Most notably, bedforms, such as transverse dunes or ripples and longitudinal ribbons provide a clear indication of active sediment transport. Similarly, erosional scour and pockmarks are indicative of modern or relict sedimentary processes. However, such features, which have spatial scales of decimetres to a few tens of metres, often lie at the limit of the spatial resolution of the system. In the case of surface hull-mounted sonars, the resolution decays roughly linearly with depth. However, the question needs to be asked: does the disappearance of a specific short wavelength morphology with depth indicate a change in sedimentary environment, or merely a defocusing of the instrument over increasing range?

Sedimentologists wishing to conduct multibeam surveys may not have the luxury of choice of system due to logistical or financial constraints. When interpreting the available data, however, it is important to establish the achievable resolution of the utilized specific sonar system. To this end, there are a number of components that need to be considered, including:

Beam Width, Spacing and Detection Algorithm

Sonar systems are routinely quoted with beam width dimensions. Such dimensions need to be specified in two directions (Fig. 1A), along track (controlled by the transmit beam width) and across track (controlled by the receive beam width) as they may differ (Miller et al., 1997).

Fig. 1. (A) Representation of an oblique narrow beam footprint on a typical seabed terrain. (B) Variation of size and spacing of a series of multibeam profiles, comparing and contrasting the bottom detection solution (represented by stars alternating black and white from beam to beam) spacing offered by equi-angle, equi-distant and high density beam forming (multiple solutions per beam). (C) The method of describing the across-track relief within a single beam footprint by looking at the evolution of differential phase over the across-track beam dimension. For this method, the centre of the beam corresponds to the point at which the differential phase is zero. For conventional detection, this point is the only one located (by regression through the phase slope), whereas for high definition, multiple points on the phase curve (corresponding to multiple angles with respect to the beam centre) are identified.

In order to appreciate the potential of the beam footprint, its solid angle needs to be projected to the seabed over the range of depths and angles used. It is readily apparent that the minimum, resolvable dimension is strongly linked to the size of this footprint (Fig. 2). Resolution needs to be described separately for along and across track.

Fig. 2. Comparison of the backscatter and bathymetric imaging resolution of two generations of common multibeam sonar systems. Wrecks and boulder fields in 25–40 m of water. Note the increased definition in both bathymetry and backscatter imaging of the boulder targets. Note also the pronounced drop in resolution for the outermost beams in the case of the equi-angle beam spacing utilized by the EM1000. Beam widths are given for both scanners.

For an amplitude detection (deMoustier, 1993), the resolvable dimension cannot be smaller than this footprint as the echo is integrated over that dimension. Few sonars today, however, still use amplitude detection outside the near nadir or near specular region. Phase detection using a split aperture (deMoustier, 1993), in which the elevation angle within the beam footprint is defined by phase rather than peak intensity, is almost universally used. In this manner, discrimination across track can be achieved based on phase (Fig. 1C). For the long, lower grazing angle echoes, phase (and thus feature definition) can be discerned at a scale significantly finer than the beam footprint dimension (Hughes Clarke et al., 1998). For most sonars this is achieved by having beam spacings across track that are tighter than the beam footprint dimension. The most common example of this is the “Equi-Distant” beam spacing (EDBS) mode (Fig. 1B) increasingly offered. For conventional phase detection, each beam still has only one depth solution (what is termed the “zero phase crossing, solution 0 in Fig. 1C), but it is based on just the phase slope in the central part of the beam.

Figure 2 (left; EM1000 images) illustrates the resolution achieved using equi-angular beamspacing when the EDBS philosophy is not employed. As can be seen, the definition of the boulders degrades notably as one moves to the outer part of the swath. The compromise in EDBS is that, for a finite number of beams, the beam spacing in the near nadir region has to be compromised to accommodate the extra solutions at lower grazing angles (see beam spacing in Fig. 1B). For example for the EM1002, which has 111 beams over a 150° sector, in equi-angular mode (EABS) the near nadir beams are spaced at 1.35°, whereas in EDBS they are spaced at 3.84° (resulting in lower nadir resolution and wider than the 2° beam width, resulting in corrupted backscatter data).

Most recently, the limitation of EDBS has been removed through the use of “high definition” beam forming (Kongsberg, 2005) in which, for phase detection, multiple points on the phase slope are used within a single beam footprint (Fig. 1C, solutions −1, +1 and +2). The physical beam spacing is actually equi-angular, but more depth solutions than beams are generated by subdividing the lower grazing angle beams. This gets around the compromise in conventional EDBS as optimal beam spacing for amplitude detection is retained.

However tight the beam spacing in the across track dimension, in the along track direction, the beam dimension and its spacing will still limit resolution. Thus narrower transmit beam widths are to be favoured. For a given transmit beam width and depth, the fore-aft dimension of the footprint grows with obliquity. Thus for geological purposes, resolution will generally decay away from the nadir region. Again it is important that this limitation be noted when interpreting the distribution of features close to the limit of resolution such as ripples or boulders.

There is a wide variety of multibeam sonars available, but the ones most commonly used on the continental shelf are those in the 100 kHz range. The EM1000, operating at 95 kHz with a beam width of 2.4° × 3.3°, first appeared in 1992 and has been used extensively in continental shelf surveys worldwide. Large tracks of the US conterminous continental shelf have been covered with this sonar (Gardner et al., 2005, Valentine, 2005, Butman et al., 2006). The RESON 8111 (100 kHz, 1.5° × 1.5°) appeared in 1996 and has been used commercially for similar scale continental shelf mapping (Wilson et al., 2005, Intelmann et al., 2006). The EM1000 was superseded by the very similar but higher resolution EM1002 (2.0° × 2.0°) in 1998, but many were still used until 2005. The EM1002 has been employed on a regional scale for continental shelf geological mapping by Canadian agencies (Pickrill & Todd, 2002, Conway et al., 2004). The EM710 (Fig. 2, right hand side images) represents one example of the next generation of sonar systems that are replacing the 1000/1002 series with beamwidths now as narrow as 0.5° × 1.0°, and for the first time include yaw stabilization. The practical examples here compare and contrast the EM1000 and EM710 sonars.

Roll, Pitch and Yaw Stabilization

In order to optimize the resolution, the sounding density along track should be as high and as even as possible. Ping rate for single ping systems is controlled by the two way travel time (TWTT) to the outermost beams. The wider the angular sector, the lower the ping rate. Thus for a given speed, resolution will decay with sector width resulting in a competition between lateral coverage and resolution. This is starting to be solved with the recent use of multiple swaths per ping cycle system. This is now offered (but only delivered in July 2008) by a number of manufacturers and promises to improve this limitation.

Irrespective of the along track vessel movement between pings, the outermost beams may be displaced more or less depending on the vessel rotations and the form of stabilization (Fig. 3). Roll stabilization is essential if the full swath is to be used, but does not affect the along track density. Pitch stabilization is more important in deeper water. But the biggest issue in continental shelf depths is yaw. In a cross-sea, vessel heading is hard to maintain, and as the water depth becomes shallower the helmsman is forced to take stronger corrective action to maintain minimal survey line offset. The requirement for yaw stabilization depends on the inter-ping yaw shift and the transmit beam width. As narrow transmit beams are being used to increase resolution, the requirement for yaw stabilization is increasing.

Fig. 3. The strategy and result of active roll, pitch, and yaw stabilization. Note particularly the improvement in even sounding density achieved by using the multi-sector strategy. This strongly impacts on the ability to maintain resolution for all regions ensonified.

To achieve yaw stabilization requires the use of multiple sectors (Fig. 3). For a single sector system, the full swath is illuminated using a single broad transmit beam that can only utilize a single steering angle, which must be chosen as a compromise whereby both sides of the swath are aligned as best as possible. For the case of multiple sectors, a succession of individual transmissions is generated, closely spaced in time (separated in time only by the length of each pulse). Each sector/transmission addresses only a specific subset of the total swath and can thus have a unique steering angle that best aligns that subset of the swath. In this manner the compromise inherent in single sector systems can be avoided, allowing yaw stabilization that requires, as a minimum, opposite-sense steering angles for each side. Without yaw stabilization, there will be zones of lower sounding density (on the outside of shallow corners; Fig. 3) where the target resolution, and thus geological interpretation, is compromised.

Accuracy

Achievable resolution is no guarantee of absolute survey accuracy at that level. Any survey consists of a series of systematically offset corridors of data, normally called swaths. The combination of multiple swaths requires a common reference datum. Absolute accuracy limits will corrupt the data in two ways: (1) when blending the overlap, the view of the seabed in the region of overlap will be defocused; and (2) when comparing the swath with data collected at other times, only scales of seabed change larger than the combination of the achievable accuracies of both surveys will be discernable.

While manufacturers' brochures tend to emphasize the sonar-relative range and angle accuracy, these usually input an uncorrelated random noise in the sounding data rather than a systematic bias. For operations where repeat surveys are required for sedimentary change assessment, it will be the systematic errors that are more important as they will generate biases that can be confused with true sediment accretion or deflation. There are a number of components in addition to the sonar range and angle measurement that contribute to the achievable degree of accuracy.

Positioning Systems – Horizontal

Positioning requirements are normally quite different for horizontal and vertical. The Global Positioning System is now used universally for the horizontal component of marine surface surveys. The achievable accuracy depends on the type of GPS chosen. Stand alone versions (non-differential) will allow 10–15 m accuracy, sufficient for deep-sea operations (where the resolution is below this), but not for shelf investigations where some form of differential GPS will be required.

Differential corrections from a coastal (usually Coastguard) service will provide sub-2 m horizontal accuracy, adequate for outer continental shelf surveys. To obtain better accuracy than this would require an interpolated correction service, such as Fugro OmniSTAR (Visser, 2007) or C&C CNav (Chance et al., 2003), often referred to as Globally-corrected GPS (GcGPS). Such services provide decimetre level horizontally, meeting practically all the needs of shelf and inshore surveys.

To get to a centimetric level positioning requires a local base station and “Kinematic GPS” (USACE, 2002). This is not practically needed for horizontal positioning but, as outlined below does provide the necessary level of vertical positioning to account fully for tides and squat.

Assuming that the horizontal accuracy of the positioning system meets the needs of the seabed change detection requirement, one still has to ensure proper integration of that position. The most common issue is one of time delays between sonar and positioning sensor clocks. Delays will result in systematic, along survey-line displacement of the swaths of data. This will generate apparent migration of seabed features that could be confused with real change. Detection of such offsets is normally quite easy by comparing the displacement of linear targets such as bedrock outcrop ridges or sand wave crests within a single survey.

Angular Measurements – Accuracy and Alignment

All sonar relative ranges and bearing need to be adjusted for array orientation at transmit and receive operations. Generally the stated angular accuracies (<0.05°) of the high-end GPS-integrated inertial motion sensors are more than adequate for the accuracy levels needed for operations. However, it is not the instrument accuracy that most concern us, but rather the integration of sensor data. Proper integration requires knowledge of sonar to motion sensor alignment and timing calibration.

Misalignment or mistiming of sensors relative to each other can create both static biases (for example a roll bias) and dynamic residuals (so called wobbles). For a full review of the sources of dynamic motion residuals, the reader is referred to Hughes Clarke (2003). From the point of view of the sedimentologist, the effects have two end-member results. Firstly the static biases impede the ability to measure change, and secondly the dynamic motion residuals can be confused with, or obscure, real seabed terrain.

Water Column Sound Speed Structure

An integral component of an accurate depth measurement is the proper accounting of sound wave propagation and refraction in the water column (Beaudoin et al., 2004). This depends on an adequate knowledge of the sound speed structure in the ocean. Failure to account for this properly will result in either a dynamic residual (Hughes Clarke, 2003) or a systematic, convex or concave across track bias (Fig. 4), the magnitude of which depends on the unmonitored changes in the water column. The water column is changing continually in time and space and thus the magnitude and sign of the error will reflect the time and/or distance that has passed since the last sound speed measurement (Hughes Clarke et al., 2000). Figure 4 illustrates a typical summer continental shelf oceanographic section, illustrating the rapid change in refraction conditions from fresh water stratified, to a tidal mixed area to a thermally stratified ocean within distances of tens of kilometres.

Fig. 4. Variability in the sound speed field for a typical summer-time continental shelf. Example dogleg transect illustrated from A to B to C across the Bay of Fundy. Note transition from a fresh-water stratified environment, to a thermally stratified environment, punctuated by zones of increased mixing, separated by abrupt tidal fronts, spaced at times by only a few kilometers along track. Such changes, which alter the location of the velocline, lower left, result in systematic across track biases in the resulting bathymetric data, lower right. Data were derived from a continuously operating underway profiler at 12 knots, with sample spacing of about 1 km.

In order to minimize the significance of sound speed errors, a variety of strategies may be developed including continuously monitoring the sound speed (Cartwright & Hughes Clarke, 2002), reducing the angular sector of the swath, and reviewing archived information about likely water mass variability and then designing the survey to take that into account.

The illustrated profile (Fig. 4) required updated sound speed structure information at approximately half hour intervals to maintain the full ± 65° swath within IHO order 1 specifications (International Hydrographic Organization, 1998). Such a profiling frequency is not practical unless underway profiling strategies are available. Prior knowledge of this oceanographic variability would allow the prudent user to break up the survey into regions of common watermass type.

In all cases, it should be appreciated that reducing the angular sector is the most reliable way of minimizing these errors. This results in a low rate of coverage, but will improve the data density (as the required maximum two-way travel time is reduced) and thus increase the resolution. A practical example is presented (Fig. 5) showing inter-survey bathymetric surface differences for Squamish Delta in Howe Sound, British Columbia. The upper delta is accreting 1 m yr−1 on average. The delta has been the subject of investigation with multibeam since 2004 (Brucker et al., 2007). Maintaining sufficient sound speed data to survey represents a challenge due to the presence and variability of a freshwater plume emanating from the mouth of the Squamish River that is modulated over a tidal cycle.

Fig. 5. Example of the effect of water column uncertainty on estimates of seafloor change. (A) difference map between two surveys (EM1002) in 50–200 m of water. (B) bathymetry of the Squamish Delta, both surveys using sparse sound speed profiles, obtained several kilometres away. (C) difference map between two surveys (EM3002). For both surveys, the greyscale used is the same between −3 m and +3 m. Differences greater than this are thresholded to black or white.

The first difference map (Fig. 5A) shows the apparent changes based on two regional surveys that did not undertake extra sound speed measurements close to the river mouth. While it is immediately apparent that gross change has occurred in the delta foreset channel and on the proximal lobe to the SW, there is a conspicuous pattern of striping developed over the difference map on the rest of the delta surface that does not obviously correlate with likely depositional or erosional processes. These are a result of refraction residuals (Fig. 4D) in each survey. Note that the residuals are actually only 0.5–1.5 m in 100–200 m of water which is well within International Hydrographic Organization standards, which are typically ± 1.5% of water depth (International Hydrographic Organization, 1998).

By contrast, the second difference map (Fig. 5B) was obtained using much more frequent sound speed profiles collected in the local area throughout the survey. As can be seen, the striping is nearly absent (the two surveys were run orthogonally to each other and thus the contribution of each survey should be apparent). Only using these methods can one start to assess the scale of the over-bank sedimentation that contributes to the long term growth of the delta front.

Tidal Reduction, Measurement and Models

However good all the other integrated components of the depth measurement are, ultimately the depth must be referenced to a stable vertical datum. Tidal reduction has always been a necessary step. For traditional coastal hydrographic surveys the standard has been to install a local gauge.

This approach is valid for regions in the local area that share the same phase and amplitude of the tide. However, as one moves along restricted coastal areas or onto the open continental shelf, knowledge is required of the propagation of the tidal wave. This is often expressed in terms of a co-tidal chart, where regions are defined in which tides at a reference station need to be scaled and delayed to be valid in adjacent regions (Admiralty, 1969). For coastal areas, this knowledge is based on historic location of tide gauges at adjacent locations along the coast. However, as one moves out onto the continental shelf, adequate tidal data are lacking. For an open coastline the propagation of the tidal wave from the edge of the continental shelf was previously poorly known. However, recent modelling, based on analysis of Topex Poseidon sea surface elevation data (e.g. Dupont et al., 2002) has resulted in the development of hydrodynamic models that predict the propagation of the wave across outer-continental shelf regions. In this manner, a dynamic tidal solution may be calculated along the track of the survey vessel, appropriate for the location and time of the vessel at every point.

In Fig. 6, an example on one such model for the Bay of Fundy is presented. The model is based on the WebTide (Department of Fisheries and Oceans, 2005) hydrodynamic model, which is available for the entire Canadian continental shelf. The resolution is variable and uses a finite element triangulated network (Fig. 6A). The Bay of Fundy is a region in which the amplitude of the tide (Fig. 6B) more than doubles as one moves up the bay and the phase is successively delayed (Fig. 6C) particularly at constricted regions (Greenberg, 1979).

Fig. 6. (A) Resolution of the hydrodynamic model for the Bay of Fundy; (B) the resulting distribution of tidal amplitude; (C) the phase of the constituents (M2 illustrated here, but available for all of M2, S2, N2, O1, K1); (D) in order to take advantage of GPS heighting, the addition of an ellipsoid-geoid separation model is required. The profile represents the correction necessary to shift measurements from the ellipsoid to the geoid over a 20-day period of operation (the ship tracks are illustrated in D).

Of most concern are constricted regions where the tidal wave is impeded and the phase contours (Fig. 6C) are tightly spaced. In these regions, a local gauge can become invalid within a few kilometres. The model illustrated here has been adopted as the prime reference for reprocessing of multibeam surveys from 1992 to 2007 in the region. This approach was chosen over conventional tide gauges due to the problems associated with the necessity of maintaining multiple gauges and accurately defining the tide in the central bay. Disadvantages of tidal models are twofold:

1. Their accuracy offshore is hard to assess. It is generally only tested against point stations on the coast;

2. They cannot predict non-tidal sea level signatures due to, for example, atmospheric pressure variations, or wind-driven sea surface run up.

In the event of there being an unmodelled amplitude or phase error in the applied tidal profile, it will not be immediately apparent if sequential survey lines are just a few tens of minutes apart. This is because the magnitude and sign of the residual error will change only with periods similar to the tidal forcing. But if a pair of lines is run with a time gap between them of several hours (more strictly with a significant change in tidal phase) then the sign and magnitude of the tidal residual error is unlikely to be the same.

Figure 7 illustrates the analysis of a repeat survey on the continental shelf in which the inter-survey differences are clearly dominated by the tidal signature. Two types of residual are seen. Gradually changing magnitude and sign of the difference across the survey progression indicate either a phase or amplitude error in the tide. Abrupt steps in the sign of the difference along a shiptrack indicate that the survey has been broken for an unspecified period. Only the second type of error will show up in the short wavelength morphology as an abrupt inter-line step. The first type of error, results in only a few centimetres difference in the error between adjacent lines (even though both are actually wrong). The example in Fig. 7 is of two surveys, one day apart and using an identical platform just using orthogonal survey line orientations. As most of the error sources cancelled out, the inter-survey bias was minimal (1 cm), but the tidal errors are seen to be the dominant signature even though the tide gauge was only 10 km away. Both Fig. 7 and Fig. 5A illustrate that it is important to know the survey line orientation when examining surface difference maps. Any apparent lineation that is parallel to one of the two survey line orientations should be treated with suspicion. If a certain sedimentary process that has a preferred grain is suspected, the survey lines should be oriented so that any systematic biases would not be confused with the natural process of interest.

Fig. 7. Map of inter-survey differences for two EM1002 survey performed 24 hours apart. The region is an extremely low relief sand seafloor in the English Channel. Two rock outcrops can be seen on either side of the map. They appear as a disturbance in the difference map due to slight timing and bottom tracking uncertainty. There was no real change in the seafloor between the surveys, yet the difference map illustrates that the inter-survey errors are not random. Rather they are dominated by imperfections in the tidal model. Such patterns need to be understood before any real sedimentary change can be interpreted.

As the magnitude and sign of the tidal error only change over periods of hours, a strategy of avoiding long survey lines that only generate overlap after several hours should be adopted, thereby minimizing interline errors. The preferred sequence would be to break up large areas into several regions with line lengths no more than about an hour. Note that the error is still present but is not manifested as abrupt line to line mismatches. This provides a much clearer view of the geomorphology. Steps will still be generated at survey region boundaries. Similarly survey strategies that involve “race track” strategies, where alternate lines are run with fill in lines at other phases of the tides are to be avoided. This is often implemented for vessels that have a large turning radius compared with the line spacing.

Note that this strategy of breaking up large areas into smaller sub-regions is actually complementary to the aim of minimizing water mass variability as the data collected are within a similar water mass, and sound speed sampling strategies can be designed around a shifting box survey region. An increasingly-used alternative to the tidal measurement and modelled interpolation is to use a GPS-derived ellipsoid height. Conventional differential GPS heights are in the ± 5 m vertical accuracy range and thus of no value. Kinematic GPS offers the best solution but are limited by separation of platform and base station (generally to less than 20 km, USACE, 2002).

An emerging approach involves the GcGPS services such as C-Nav or OmniSTAR which offer a vertical accuracy of several decimetres. With smoothing, this provides an adequate result for tidal correction in continental shelf waters where one is beyond the practical range of kinematic GPS and the tidal propagation models are uncertain. The major problem with these services is reliability (Hughes Clarke et al., 2005). The vertical profile needs to be filtered and edited to account for discontinuities and interruptions.

Before any ellipsoid height model can be used, the separation between that surface and the desired reference vertical datum (usually either Chart Datum or Mean Sea Level) needs to be established. For a small area (less than a few kilometres) a single shift can often be applied, but typical geoid-ellipsoid surface slopes are in the 3 to 10 cm km−1 range and thus for continental shelf areas, one needs to have a model of the geographic variation in the separation. Figure 6D illustrates the EGM96 (Lemoine et al., 1998) ellipsoid to geoid surface separation used for the Bay of Fundy. The superimposed ship tracks run several hundred kilometres up and down the bay and thus require a continuously varying separation to be applied to the data (profile inset in Fig. 6D). In this manner repeat surveys may be conducted and referenced to a stable datum (the ellipsoid) wherein one can start to estimate sedimentary change at a vertical scale of a few decimetres.

Backscatter

Increasingly, spatial variations in the seabed backscatter strength are being used as an additional tool to aid in interpretation of shelf sedimentary processes. In order to use this effectively, a proper understanding of both the physical controls on seabed scattering and the effect of sonar radiometric and geometric imaging is required.

Physical Controls on Seabed Scattering

Seabed backscatter strength is driven by the seabed's physical properties (Jackson et al., 1986) and thus is potentially a useful indicator of sedimentary environment. A direct correlation between acoustic backscatter strength and a simple quantity such as grain size has been inferred (e.g. Borgeld et al., 1999) but in general remains elusive because spatial variations in backscatter may reflect changes in one or all of the following:

Impedance contrast of the seabed/seawater interface (controlled by the bulk density and sound speed in the sediment);Interfacial roughness of that sediment water boundary;Volume heterogeneity – changes in the patchiness and contrast in the very shallow subsurface impedance;Changing grazing angle (Fig. 8).

Fig. 8. The effect of grazing angle on multibeam geometry and typical angular response curves. Cartoons illustrating the changing role of the differing physical scattering processes for the three main parts of the angular response curve are shown (Vol. Scat. = Volume Scattering, Critical = angle beyond which no sound is refracted into the seabed).

Even at a fixed grazing angle, it can thus be ambiguous as to whether a change viewed is resulting from a change in impedance, roughness, or volume heterogeneity. Hamilton & Bachman (1982) demonstrated that, for terrigenous sediments, the impedance is strongly correlated with grain size. It would be convenient if this were the principal control on backscatter strength but, for a given grain size, the interface roughness is linked to other factors such as sorting or rippling or the presence of shell hash. For fine-grained sediment (where there is significant penetration into the sediment), the volume heterogeneity is controlled strongly by bioturbation and/or the presence or absence of buried shell debris or glacial dropstones.

Distinguishing outcrop or cobbles from fine-grained unconsolidated sediments is not an issue as the backscatter strength contrast between gravel and mud is unambiguous for all grazing angles (Fig. 8B). For the case of most temperate continental shelves, however, the variations of interest often range only from muddy sands to sandy muds. Under these conditions, the simple grain-size correlation can be obscured by other factors such as sorting, rippling, bioclastic debris and bioturbation.

Grazing Angle Effects

Even for a given set of sediment physical properties, the backscatter strength will vary with grazing angle (Fig. 8B). A typical swath will image from vertical incidence (90° grazing) to grazing angles usually as low as 25° (Fig. 8A). Thus, a measure of bottom backscatter strength will vary across the swath, providing at first glance, a misleading picture of the sediment distribution. For practical mapping purposes, the geological interpreter wishes to view an image that reflects regional sediment variations without having to continuously be aware of the imaging geometry. To achieve that, a compensation algorithm needs to be established that effectively “flattens” the angular response curves (Fig. 8B). To do this, of course, requires a priori knowledge of the shape of that curve.

The curve shape however, is highly variable between differing sediment types with strong specular peaks, of varying width and differing rates of roll-off with low grazing angle (Fig. 8B). Thus there is a need to locally adjust the compensation algorithm to reflect the local angular response (AR) curve. However, this is potentially a circular argument, as the AR curve needs to be derived from the seafloor and thus one needs to assume that the sediment type is constant from side to side in a single (or series of adjacent) swath. For continental shelf depths (50–200 m) this translates into an assumption of spatial sediment invariance over a distance of 200 to 800 m. Without this assumption one risks interpreting across track sediment changes as unusual AR curves which could then be compensated incorrectly.

Sonar Radiometric and Geometric Influence on the Received Scattering Intensity

All of the above discussion assumes that one has a calibrated measure of the bottom backscatter strength (BS). To achieve this requires a complete knowledge of the sonar system settings.

Source Level and Receiver Gain Settings

Depending on the sonar system, the source level and receiver gains may or may not have already been compensated for. For Kongsberg systems, the receiver gains are automatically set to adjust for source level, spherical spreading, attenuation and pulse length and seabed backscatter variations (but assuming a locally flat seafloor) (Hammerstad, 2000). The only compensation necessary for these systems is slight adjustments for exact pulse length used, beam pattern residuals (see later section) and true seafloor slope (see later section). In contrast, the RESON family of sonars maintain a fixed receiver gain ramp but log all the radiometric parameters including source level, pulse length and fixed gain steps. Before data can be used for geological interpretation, all the calculations need to be applied in post processing (e.g. Beaudoin et al., 2002).

The most fundamental measure is the source level of the sonar. Few multibeam systems are precisely calibrated and thus an absolute level cannot be relied upon. The usual proxy is that, for the duration of a deployment, the source level is a constant. Such an assumption will break down if a survey consists of multiple deployments with changing sonar hardware. Overlapping coverage between surveys performed with different hardware settings may be the only way to maintain a stable relative calibration (Hughes Clarke et al., 2008). Thus when trying to quantitatively assess whether a change in backscatter imagery between two surveys is real, the user must attempt to grossly shift the data to match in regions where it is believed that the seabed sedimentary environment is unaffected. Even if the data in a certain region are fixed, one needs to account for the effect of changing seasonal oceanography, which is expressed in the seawater attenuation coefficient.

Seawater Attenuation

The received intensity is a function of the attenuation taking place in the seawater. This attenuation is dependent on the frequency of choice and varies significantly with temperature and salinity (Francois & Garrison, 1982a, 1982b). It is up to the user to apply the appropriate value. One of the main variations reflects the changes in fresh water influence as one moves within the coastal zone. For example at given salinities the received intensity (all at 10 m depth, 10 °C, 100 kHz):

33 ppt salinity (typical coastal ocean): 32 dB km−1;27 ppt salinity (typical distal river plume or fjord basin): 27 dB km−1;15 ppt salinity (typical brackish estuary): 16 dB km−1.

Fortunately river plumes are normally restricted to the upper few metres of the water column and thus the depth-averaged attenuation coefficient is less affected. But within fjord basins, separated by sills, the bulk change in the salinity from basin to basin, if not accounted for, will alter the apparent backscatter strength of the basin floor. For example in 100 m water depth using a 60° beam (400 m round trip) this corresponds to a 2 dB change for a change from 27 to 33 ppt. Thus, unless compensated for (not standard in most post-processing software), one cannot discern whether there is a change in the sediment type up the fjord, or merely a change in the water mass.

Perhaps more misleading is the fact that such a bias is depth dependent. For a single basin, the image will appear consistent, but, with the wrong attenuation coefficient, the interpreter may infer a depth-correlated change in sediment type. Using the same example (33 v. 27 ppt), the same material will appear 4 dB weaker from the beach to 200 m depth and 8 dB weaker at 400 m depth. Compared with that, the BS variation between fine sand and mud is only 2–6 dB. Many sedimentary environments are depth-dependent as they depend on surface wave activity or current strength and thus the user is easily led into believing depth-related apparent sediment variations.

Another effect is the seasonality of the water temperature (all at 10 m depth, 33 ppt S – 100 kHz):

5 °C – 27 dB km−1;10 °C – 32 dB km−1;15 °C – 35 dB km−1.

Thus if a regional survey starts in the spring (5 °C), but continues, or is compared with one in the late summer (15 °C), a 100 m depth solution using a 60° beam, (a 400 m round trip) will exhibit a 3.2 dB difference.

Pulse Length Changes

Except when operating in the shallowest range of depth, most sonar systems are operating at full power the upper level of which is normally restricted by cavitation issues. As the water gets deeper, the received signal strength will drop, resulting in a loss of signal to noise. To circumvent this, one needs to increase the pulse length. Doing so for the same source level increases the instantaneously ensonified area resulting in both a stronger signal and, for narrow band signals, a lower range resolution.

This has three effects on the geological interpretation:

1. Unless compensated for, the seabed intensity will appear to change. Even for those systems that do so, the compensation is never perfect;

2. If the interpreter is relying on the pixel speckle characteristics to discern different sediment type, the speckle pattern will coarsen with longer pulses;

3. If the interpreter is looking to resolve small features, the longer pulse will be defocused, making some short wavelength features such as ripples or cobble fields disappear.

To compensate for effect 1, a measure of the pulse-length needs to be maintained with the data and, based on analysis of the shift at changes, a bulk and/or range and angle-dependent offset needs to be applied.

The second effect can be most damaging to some of the automated textural classification software on the market (e.g. Milvang et al., 1993; Preston et al., 2001) both of which in part rely on the “Pace” features (Pace & Dyer, 1979; Pace & Gao, 1988). At this time, this approach cannot take into account pulse length changes and thus automated classification is limited to regions where a single pulse length is used.

For the third effect, the loss in resolution is generally less of an issue because, at the point at which the pulse length needs to be shifted, the data with the shorter pulse length are compromised in any case by the lower signal to noise levels.

Beam Patterns – Single and Multiple Sectors

Both the transmitter and each of the individual receiver beam patterns have intensity/sensitivity variations with elevation angle. The combined effect of these two beam patterns will generate variations in received intensity across the swath that might be confused with seabed sediment changes.

The simplest configuration is a single-headed, single sector, multibeam in which the entire swath is illuminated by one transmitter. In this case, the transmit beam patterns are generally simple, varying only slowly with angle (e.g. RESON 8111, Beaudoin et al., 2002). A notable exception is the original EM1000 which used a barrel array for the transmit and thus variations in the intensity from the staves within the barrel could produce complex transmit beam patterns. In both cases, the pattern is fixed with respect to the array. For the EM1000, because each receiver channel, which is roll-stabilized, uses a separate amplifier, any inter-amplifier differences will show up as an apparent vertically-referenced beam pattern residual.

An additional complication is found in multi sector systems. Several systems: EM12, EM1002, EM300, EM120, EM710 use multiple sectors. This is done to provide advantages in multiple suppression, improved pitch and yaw stabilization (Fig. 3