Three-dimensional (3D) three-component (3C)shallow seismic ...

Three-dimensional (3D) three-component (3C)shallow seismicrefraction surveys across a shearzone associated with drylandsalinity at the Spicers Creek Catchment, New South Wales,Australia

Author:Nikrouz, Ramin

Publication Date:2005

DOI:https://doi.org/10.26190/unsworks/21772

License:https://creativecommons.org/licenses/by-nc-nd/3.0/au/Link to license to see what you are allowed to do with this resource.

Downloaded from http://hdl.handle.net/1959.4/20607 in https://unsworks.unsw.edu.au on 2022-07-21

Three-Dimensional (3D) Three-Component (3C) Shallow Seismic Refraction Surveys Across a Shear Zone Associated with Dryland Salinity at the Spicers

Creek Catchment, New South Wales, Australia

by

Ramin Nikrouz

B.Sc., Geology, IRAN M.Sc., Petrology, IRAN

A Thesis Submitted in Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

School of Biological, Earth and Environmental Sciences The University of New South Wales

Sydney NSW 2052 Australia

April 2005

PLEASE TYPE THE UNIVERSITY OF NEW SOUTH WALES

Thesis/Project Report Sheet Surname or Family name:

First name:

Other name/s:

Abbreviation for degree as given in the University calendar:

School:

Faculty:

Title:

Abstract 350 words maximum: (PLEASE TYPE)

Declaration relating to disposition of project report/thesis I am fully aware of the policy of the University relating to the retention and use of higher degree project reports and theses, namely that the University retains the copies submitted for examination and is free to allow them to be consulted or borrowed. Subject to the provisions of the Copyright Act 1968, the University may issue a project report or thesis in whole or in part, in photostat or microfilm or other copying medium. I also authorise the publication by University Microfilms of a 350 word abstract in Dissertation Abstracts International (applicable to doctorates only). …………………………………………………………… Signature

……………………………………..……………… Witness

……….……………………...…….… Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing to the Registrar. Requests for a longer period of restriction may be considered in exceptional circumstances if accompanied by a letter of support from the Supervisor or Head of School. Such requests must be submitted with the thesis/project report. FOR OFFICE USE ONLY

Date of completion of requirements for Award:

Registrar and Deputy Principal

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS

N:\FLORENCE\ABSTRACT

To my wife Fariba and my son Amir

CERTIFICATE OF ORIGINALITY

I hereby declare that this submission is my own work and to the best of my

knowledge it contains no materials previously published or written by another

person, nor material which to a substantial extent has been accepted for the

award of any other degree or diploma at UNSW or any other educational

institution, except where due acknowledgement is made in the thesis. Any

contribution made to the research by others, with whom I have worked at

UNSW or elsewhere, is explicitly acknowledged in the thesis.

I also declare that the intellectual content of this thesis is the product of my own

work, except to the extent that assistance from others in the project’s design

and conception or in style, presentation and linguistic expression is

acknowledged.

……………………………………………….

Ramin Nikrouz

ABSTRACT

Three-dimensional (3D) three-component (3C) shallow seismic refraction

surveys were recorded over a shear zone at two sites associated with dryland

salinity in the Spicers Creek Catchment, near Dubbo in southeastern Australia.

The seismic data were recorded with the Australian National Seismic Imaging

Resources (ANSIR) 360-trace ARAM-24 seismic system and IVI MiniVibrator.

Dryland salinity occurs extensively throughout the Spicers Creek Catchment.

The high concentration of salt in the groundwater has led to a significant decline

in agricultural productivity and a reduction in native vegetation. Furthermore,

the saline groundwater in the surface soil has caused the destruction of the clay

and soil structure and as a result, large areas of the catchment have

experienced soil erosion and extensive alteration of the landscape.

The broad objective of this study was to use seismic refraction methods to map

in detail a shear zone, which is associated with the salination. Seismic

refraction methods were selected because of their potential ability to provide

greater lateral resolution of the narrow vertical shear zone, than is currently the

norm with electrical or electromagnetic methods. This situation was confirmed

with a number of resistivity depth images generated as part of this study.

The results of the seismic refraction surveys show that the shear zone occurs

as a narrow region with low seismic velocities and increased depths of

weathering. However, there were numerous ambiguities in generating

i

consistent parameters in the refractor between adjacent receiver lines.

Furthermore, these ambiguities were not resolved with the use refraction

tomography, because even major differences with the various starting models

still generated acceptable agreement with the traveltime data. Therefore, in

order to generate consistency between the results for the various receiver lines,

it was necessary to generate detailed starting models.

A major achievement of this study is the development of a smoothing method

which significantly removes the effects of the near surface irregularities.

Termed the GRM SSM (for statics smoothing method), it essentially generates

a time-depth model of the refracting interface for which the effects of the near

surface irregularities have been minimized, by taking an average of the time-

depth profiles for a range of XY values. The GRM SSM, which takes advantage

of the unique redundancy properties of the GRM computations, was a major

factor in deriving consistent detailed starting models for refraction inversion.

Furthermore, this consistency was achieved with both S-wave components, as

well as the P-wave results.

The major geological achievement, which was made possible with the GRM

SSM, was the demonstration that there were cross cutting features associated

with the major shear zone. Therefore, it appears that saline groundwater can

discharge at the surface where increased volumes of groundwater occur at the

intersection of different sets of shears. This model provides a useful

explanation for the discontinuous occurrence of salination along the major shear

zone.

ii

ACKNOWLEDGEMENTS

First and foremost, I would like to thank God for helping me to accomplish this

project which is going to affect my academic life for some time to come. This

accomplishment would not have been possible without the help and support of

my supervisor, Derecke Palmer, who acted as a source of inspiration and

practical scientific support. I am most grateful for his keen supervision,

continuous support and encouragement during my research study. The

geophysical knowledge which I have acquired as the result of the

accomplishment of this research project is indeed the result of his insight and

expertise. I have been inspired by his endurance and perseverance.

I would also like to acknowledge my indebtedness to Paul Lennox, my co-

supervisor, who saved no effort to keep me on track in my endeavors. A

special word of thanks also goes to one of my best friends and colleagues in

Australia, Andrew Spyrow, whose friendship I value most highly. He was

especially helpful and supportive in the many computing tasks associated with

this study. I also owe a word of gratitude to David Johnston and the ANSIR

team who made it possible for me to gather what was indeed the corner stone

of my project, the data of the study.

iii

On a different note, I would like to thank the officials in the Iranian Ministry of

Science, Research and Technology and the University of Urmia of which I am a

faculty member for the financial support they provided for my family and me.

Also, I need to acknowledge my sense of gratitude and indebtedness to my

wife, Fariba and my son, Amir. They served as a source of encouragement and

support for me and were commendably patient with me although due to the

requirements of my studies, I had less time to spend with them. I am also

thankful to my parents of whose encouragement and support I am always

appreciative.

I should also take this opportunity to express my most sincere thanks to my

fellow research students at UNSW and friends, Morteza Jami and Ahmad Reza

Mokhtari, for their support and encouragement throughout the study of my

research.

Last but not the least, I sincerely thank the staff of the School of Biological,

Earth, and Environmental Sciences at UNSW, for their continued support and

encouragement. I am also grateful to many others, whose names do not

appear here. Their help is greatly appreciated.

iv

CONTENTS

Abstract-------------------------------------------------------------------------------------- i Acknowledgements---------------------------------------------------------------------- iii Contents-------------------------------------------------------------------------------------- V Illustrations---------------------------------------------------------------------------------- xi

Chapter 1 Introduction--------------------------------------------------------------------------------- 1 1.1 - Recent Advances in the Resolution of Geophysical Data-------------- -- 1 1.2 - Three-Dimentional Shallow Seismic Refraction Methods----------------- 5 1.3 - Dryland Salinity---------------------------------------------------------------------- 8 1.4 - Detailed Mapping of Shear Zones---------------------------------------------- 10 1.5 - Objectives----------------------------------------------------------------------------- 12 1.6 - Problems Solved in this Study--------------------------------------------------- 15

Chapter 2 Dryland Salinity----------------------------------------------------------------------------- 17 2.1 - Summary------------------------------------------------------------------------------- 17 2.2 - Introduction---------------------------------------------------------------------------- 18 2.3 - Extent of Dryland Salinity in New South Wales------------------------------- 19 2.4 - Causes of Dryland Salinity--------------------------------------------------------- 23 2.5 - Dryland Salinity in the Spicers Creek Catchment---------------------------- 28 2.6 - Management of Dryland Salinity-------------------------------------------------- 33

Chapter 3 Shear Zones Characteristic------------------------------------------------------------ 43 3.1 - Summary------------------------------------------------------------------------------- 43 3.2 - Introduction---------------------------------------------------------------------------- 44 3.3 - Shear Zone Types------------------------------------------------------------------- 46 3.4 - Fluid Movement in Shear Zones------------------------------------------------- 51

v

3.5 - Shear Zone Evidences---------------------------------------------------------- 54 3.5.1 - Field Evidence------------------------------------------------------------------- 54 3.5.2 - Imaging Shear Zones with Geophysical Methods---------------------- 59

Chapter 4 Three Dimension Three Component Seismic Surveys-------------------- 61 4.1 - Summary---------------------------------------------------------------------------- 61 4.2 - The Seismic Refraction Method----------------------------------------------- 62 4.3 - Seismic Velocities in the Earth------------------------------------------------ 63 4.4 - Three Dimension Seismic Refraction Methods---------------------------- 65 4.5 - Three Component Seismic Refraction Surveys---------------------------- 68 4.6 - Shear Waves Versus P Waves------------------------------------------------ 70

Chapter 5 Geology and Geophysics------------------------------------------------------------ 74 5.1 - Summary----------------------------------------------------------------------------- 74 5.2 - Introduction------------------------------------------------------------------------- 75 5.3 - Tectonic Setting-------------------------------------------------------------------- 75

5.3.1 - Lachlan Fold Belt------------------------------------------------------------ 77 5.3.2 - Gunnedah Basin------------------------------------------------------------- 82 5.3.3 - Surat Basin-------------------------------------------------------------------- 88

5.4 - Regional Geology------------------------------------------------------------------ 91 5.4.1 - Introduction-------------------------------------------------------------------- 91 5.4.2 - Climate------------------------------------------------------------------------- 92 5.4.3 - Physiography----------------------------------------------------------------- 93 5.4.4 - Dubbo Area------------------------------------------------------------------- 95 5.4.5 - Geology------------------------------------------------------------------------ 99 5.4.6 - Metamorphism--------------------------------------------------------------- 103

5.5 - Local Geology---------------------------------------------------------------------- 105 5.5.1 - Introduction------------------------------------------------------------------- 105 5.5.2 - Climate------------------------------------------------------------------------- 107 5.5.3 - Stratigraphy------------------------------------------------------------------- 107

5.5.3.1 - Palaeozoic Units------------------------------------------------------- 110

vi

5.5.3.1.1 - Okdale Formation (θco)--------------------------------------- 110 5.5.3.1.2 - Ungrouped Ordovician Intrusions (θm)--------------------- 111 5.5.3.1.3 - Gleneski Formation (Sms)------------------------------------ 112 5.5.3.1.4 - Cuga Burga Volcanics (Dgc)--------------------------------- 112 5.5.3.1.5 - Early Permian Undifferentiated (Pe)------------------------ 114 5.5.3.1.6 - Dunedoo Formation (Pd)-------------------------------------- 114

5.5.3.2 - Mesozoic Units--------------------------------------------------------- 115 5.5.3.2.1 - Boulerwood Formation (Rb)---------------------------------- 115 5.5.3.2.2 - Napperby Formation (Rp)------------------------------------- 116

5.5.3.3 - Cainozoic unit---------------------------------------------------------- 117 5.5.3.3.1 - Alluvial deposits (Qa)------------------------------------------- 117

5.5.4 - Structure----------------------------------------------------------------------- 117 5.5.4.1 - Nindethana Fault------------------------------------------------------ 121 5.5.4.2 - Narragal Thrust Sheet------------------------------------------------ 123 5.5.4.3 - Macquarie Fault-------------------------------------------------------- 123 5.5.4.4 - Neurea Fault------------------------------------------------------------ 124

5.6 - Regional and Local Geophysics----------------------------------------------- 125 5.6.1 - Introduction------------------------------------------------------------------- 125 5.6.2 - Radiometric Methods------------------------------------------------------ 127

5.6.2.1 - Theory, Application--------------------------------------------------- 127 5.6.2.2 - Interpretation Methodology----------------------------------------- 128 5.6.2.3 - Regional Interpretation----------------------------------------------- 132

5.6.3 - Aeromagnetic Methods---------------------------------------------------- 136 5.6.3.1 - Theory, Application--------------------------------------------------- 136 5.6.3.2 - Interpretation Methodology----------------------------------------- 138 5.6.3.3 - Regional Interpretation---------------------------------------------- 139

5.6.4 - Gravity Methods------------------------------------------------------------ 142 5.6.4.1 - Theory, Application-------------------------------------------------- 142 5.6.4.2 - Interpretation Methodology---------------------------------------- 144 5.6.4.3 - Regional Interpretation--------------------------------------------- 145

vii

Chapter 6 Near-Surface Effect Corrections with the GRM------------------------------- 148 6.1 - Summary----------------------------------------------------------------------------- 148 6.2 - Introduction--------------------------------------------------------------------------- 149 6.3 - Correcting for Near-Surface Effects------------------------------------------- 150 6.4 - Model 1: Irregular Refractor with Flat Topography------------------------ 152 6.5 - Model 2: Irregular Refractor with Irregular Topography----------------- 159 6.6 - Applying Near-Surface Effect Corrections to Model 2------------------- 164

Chapter 7 Methodology and data acquisition------------------------------------------------- 174 7.1 - Summary----------------------------------------------------------------------------- 174 7.2 - Introduction-------------------------------------------------------------------------- 174 7.3 - Equipment--------------------------------------------------------------------------- 175

7.3.1 - ARAM 24 Seismic Recording System---------------------------------- 176 7.3.2 - Mini Vibrator------------------------------------------------------------------ 177 7.3.3 - Geophones-------------------------------------------------------------------- 181 7.3.4 - Explosive Source------------------------------------------------------------ 182

7.4 - Field Procedure-------------------------------------------------------------------- 183 7.4.1 - Setting up the Survey Spread-------------------------------------------- 184 7.4.2 - Conducting the Survey----------------------------------------------------- 186

7.4.2.1 - Site 1---------------------------------------------------------------------- 188 7.4.2.2 - Site 2---------------------------------------------------------------------- 192

7.5 - Data quality in Site 1 and Site 2------------------------------------------------ 196 7.6 – Magnetic Data Acquisition------------------------------------------------------ 198 7.7 – Resistivity Data Acquisition----------------------------------------------------- 198

Chapter 8 Processing--------------------------------------------------------------------------------- 199 8.1 - Summary----------------------------------------------------------------------------- 199 8.2 - Introduction--------------------------------------------------------------------------- 200 8.3 - Processing of Magnetic and Resistivity Data-------------------------------- 200

8.3.1 - MagMap 2000 Software----------------------------------------------------- 201

viii

8.3.2 - Res2DINV Software---------------------------------------------------------- 201

8.4 - Processing of Seismic Refraction Data---------------------------------------- 203 8.4.1 - Microsoft Excel----------------------------------------------------------------- 204 8.4.2 - Surfer 8--------------------------------------------------------------------------- 204 8.4.3 - Seismic UN*x Program------------------------------------------------------- 205 8.4.4 - Visual-SUNT Software-------------------------------------------------------- 206 8.4.5 - Rayfract Software---------------------------------------------------------------206

8.5 - The Generalized Reciprocal Method-------------------------------------------- 207 8.5.1 - Velocity Analysis Function--------------------------------------------------- 208 8.5.2 - Time-Depth Function---------------------------------------------------------- 209

8.6 - Procedure of the Data Processing---------------------------------------------- 213 8.6.1 - Pre-processing----------------------------------------------------------------- 213 8.6.2 - Refraction Convolution Sections (RCS)--------------------------------- 214 8.6.3 - First Arrival Picking----------------------------------------------------------- 220 8.6.4 - Traveltime Graphs------------------------------------------------------------ 223 8.6.5 - The GRM Refractor Velocity Analysis Function----------------------- 228 8.6.6 - Corrections for Surface Effects-------------------------------------------- 234 8.6.7 - Seismic Velocities in the Refractor--------------------------------------- 237 8.6.8 - The GRM Time-Depth and Depth Function------------------------------240 8.6.9 - Processing of the First Arrival Amplitudes------------------------------ 244 8.6.10 - Velocity Ratios--------------------------------------------------------------- 248 8.6.11 - Three-Dimensional Images---------------------------------------------- 249 8.6.12 - Traveltime Tomography--------------------------------------------------- 249

Chapter 9 Interpretation------------------------------------------------------------------------------ 253 9.1 - Summary------------------------------------------------------------------------------ 253 9.2 - Introduction--------------------------------------------------------------------------- 255 9.3 - Magnetic Data Interpretation----------------------------------------------------- 256 9.4 - Resistivity Data Interpretation---------------------------------------------------- 257 9.5 - 3D-3C Seismic Data interpretation---------------------------------------------- 265

9.5.1 - Refractor Seismic Images at the First Site----------------------------- 265

ix

9.5.2 - Refractor Seismic Images at the Second Site------------------------- 274 9.6 - Geology and Tectonic interpretation------------------------------------------- 282 Chapter 10 Conclusions------------------------------------------------------------------------------ 285 References-------------------------------------------------------------------------------- 290 Appendices Appendix A - Linear and Non-Linear Metod--------------------------------------- 307 Appendix B - Operation Reports------------------------------------------------------ 309 Appendix C - Specification for the Equipment------------------------------------ 312 Appendix D - Station and Shot Point Coordinates------------------------------ 315 Appendix E - Geophone Coordinates----------------------------------------------- 317 Appendix F - Comment Document From Observer------------------------------ 318 Appendix G - Assessment of Data Quality----------------------------------------- 322 Appendix H - Seismic Unix Shell Scripts and C Files--------------------------- 336 Appendix I - Trace Order--------------------------------------------------------------- 348 Appendix J - Refraction Convolution Images------------------------------------- 350 Appendix K - Refractor Seismic Velocities---------------------------------------- 362 Appendix L - Time-Depth and Depth Graphs------------------------------------- 374 Appendix M - Eavepath Eikonal Traveltime Inversion-------------------------- 390 Appendix N - Traveltime Tomography Errors------------------------------------- 395 Appendix O - Refractor Images------------------------------------------------------ 396 Appendix P - GRM Tomography Images------------------------------------------ 410 Appendix Q - GRM Statics for Shallow Refraction Data----------------------- 416 Q1 – Abstract---------------------------------------------------------------- 416 Q2 – Generating Detailed Refractor Model for Inversion--------- 417 Q3 – Non-Uniqueness in Determining Detailed Refractor Velocities-------------------------------------------------------------- 422 Q4 – “Statics” Corrections for Shallow Refraction Data – Model Study -------------------------------------------------------- 427 Q5 – “Statics” Corrections for Shallow Refraction Data – Case Study----------------------------------------------------------- 431 Q6 – Conclusions----------------------------------------------------------- 434 Q7 – References------------------------------------------------------------ 437 Q8 – Figure Captions------------------------------------------------------ 441 Q9 - Figures ----------------------------------------------------------------- 445

x

ILLUSTRATIONS FIGURES Figure 1 - 1: Improvements in the resolution of geological detail have been achieved by adopting a new generation of survey specifications and presenting the resulting grid as an image. This illustration is from an area of the Ebagoola 1:250 000 sheet in northern Queensland. Top: BMR survey flown (E-W) in 1973 at a mean terrain clearance of 150 m and a line spacing of 1500 m, presented as a contour map with contour interval 10 nT. Bottom: the same area flown (E-W) by Geoterrex in 1991 with 100 m terrain clearance and 400 m line spacing (Reeves, C.V., 1992)…………………………………………………………………2 Figure 1 - 2: Results from the 3-D Mt Bulga seismic refraction survey. Seismic velocities and interpreted faults are plotted over the contours of the time-depths in milliseconds. The bold arrows indicate the directions of the higher seismic velocity (Palmer 2001b)………………………………………………………………6 Figure 1 - 3: The standard 2-D refraction depth cross section obtained across a shear zone at Mt Bulga. The shear zone (between stations 54 and 60) exhibits a decrease in the seismic velocity and an increase in the depth of weathering (Palmer, 2001b)……………………………………………………………………….6 Figure 2 - 1: Dryland salinity hazards in New South Wales (taken from Bradd, et al. 1997). The area that is within each hazard category is approximately low 75,060,000 ha; moderate 4,250,000 ha; high 470,000 ha; and very high 310,000 ha…………………………………………………………………………….21 Figure 2 - 2: Dryland salinity risk in New South Wales 2000 (From National Land & Water Resources Audit, 2000)…………………………………………….22 Figure 2 - 3: Dryland salinity risk in New South Wales 2050 (From National Land & Water Resources Audit; 2000)…………………………………………….22 Figure 2 - 4: Map of the Spicers Creek catchment with the main town, road and drainage modified from Cobbora 1:50,000 topographic map……………………28 Figure 3 - 1: (A) Fault. (B) Fault zone. (C) Shear zone (From Hobbs et al, 1976)………………………………………………………………………………….46 Figure 3 - 2: Different types of shear zones from (Ramsay and Auber, 1983)….. …………………………………………………………………………………………47 Figure 3 - 3: Idealized active shear zones of Type I and II (right) and stress / shear strain curves for each material (left). Shaded areas are current, actively straining regions, at margins (m) or in interior (i) of zones of Type I and II, respectively. Stubby arrows indicate direction of migration over time of boundaries of straining regions. Ellipses represent finite-strain gradients (From Means, 1995)…………………………………………………………………………49 Figure 3 - 4: Finite shear strain vs. time curves (solid) for the three types of shear zones, for rock at the final zone margins (m) and in the interior (i). P (dashed line) represents some process-controlling parameter (like temperature or fluid chemistry) that is assumed to change over the shearing history. In Type II zones, the marginal and interior acquire their low-strain state at about the

xi

same time and p value (heavy dots). The marginal rocks, therefore, provide a good record of former low-strain state of the interior rocks. In zones of Types I and III, the corresponding dots are farther apart in time and p, so the marginal rocks provide a less reliable record of the former low-strain state of the interior rocks. The type III diagram has been drawn for the case where the marginal rocks begin straining at the same time as the interior rocks, but more slowly. In each diagram, the right-hand dotted line is drawn at the point along the m line where its slope (the marginal strain-rate) becomes zero. The horizontal position of this line, therefore, represents the time at which rock in the shear zone margins acquired its final state of (low) strain. The left-hand line represents the time at which rock in the shear zone interior passed through this same state of (low) strain (From Means, 1995)…………………………………………………..50 Figure 3 - 5: Different types of criteria used for the determination of shear sense in shear zones (From Ramsay and Auber, 1983)……………………….58 Figure 4 - 1: Snell’s Law…………………………………………………………..62 Figure 4 - 2: Seismic refraction wave raypaths…………………………………64 Figure 4 -3: Nine-component sections recorded by P-, SH-, and SV-geophones for P-, SH-, and SV sources. Without anisotropy the sections marked ? would be blank (From Tatham and McCormack, 1991)………………………………..69 Figure 5 - 1: Tectonic units within the Lachlan Fold Belt area (Modified from: Kingham, 1998)…………………………………………………………………….76 Figure 5 - 2: Structural map of the Lachlan Fold Belt in New South Wales (Suppel and Scheibner, 1990)……………………………………………………78 Figure 5 - 3: Tectonic model for the eastern Lachlan Fold Belt (Modified from Collins and Vernon (1994))………………………………………………………..83 Figure 5 - 4: The Sydney, Gunnedah and Bowen Basin in NSW (Tadros, 1993). ………………………………………………………………………………………..85 Figure 5 - 5: Lithostratigraphy of the Gunnedah Basin (Tadros, 1993)………87

Figure 5-6: Surat Basin Stratigraphy (From: www.minerals.unsw.gov.au/geosurvey/petrol)…………………………90 Figure 5 - 7: Locality map for the Dubbo 1:250,000 geological map sheet area, showing the six component 1:100,000 map sheets and geographical features (Meakin and Morgan, 1999)………………………………………………………91 Figure 5 - 8: Mean monthly maximum and minimum temperatures, mean monthly precipitation data and the average number of wet days for Dubbo and Dunedoo, pan evaporation data for Wellington is shown at the base of Figure (Schofield, 1998)……………………………………………………………………92 Figure 5 - 9: Topography, township and surface drainage map for the Dubbo and Cobbora 1:100,000 sheets……………………………………………………94 Figure 5 - 10: Schematic structure map of the Dubbo1:250 000 sheet area (Matson, 1975)………………………………………………………………………96 Figure 5 - 11: Simplified geological framework of the Dubbo 1:250,000 map sheet area (Meakin and Morgan, 1999)………………………………………….100 Figure 5 - 12: Dubbo 1:250,000 map sheet area showing the metamorphic zones (Meakin and Morgan, 1999)……………………………………………….104 Figure 5 - 13: The road map of the Dubbo area………………………………..106

xii

Figure 5 - 14: Geological map of study area from Cobbora 1:100,000 geological map with the two sites marked. The various units are described in Table 1…108 Figure 5 - 15: Structural zones defined by Scott (1997) across the Dubbo geological map sheet………………………………………………………………118 Figure 5 - 16: Structural map of the thrust sheet on the Cowra Zone and Molong Zone of the Dubbo geological sheet (Modified from Meakin and Morgan, 1999)…………………………………………………………………………………122 Figure 5 - 17: Location of the geophysical data sets supplied by AGSO and the selected area of interest for the interpretation of this data (Schofield, 1998)..125 Figure 5 - 18: The HSV colour model – H, S, and V represent hue, saturation, and value of colour intensity, respectively. According to this model, for a colour of constant hue and saturation, if its value is decreased it darkens towards black. For a colour of constant hue and value, if the saturation is decreased it becomes whiter (Milligan and Gunn, 1977)……………………………………..131 Figure 5 – 19: Red-green-blue composite radiometric image for Dubbo area (Meakin and Morgan, 1999)……………………………………………………….134 Figure 5 - 20: Gamma-ray image of an area south of Dubbo, showing the geochemically different signatures of a number of minor intrusions in the area. Some intrusions have dominant U+Th signatures, others dominant K+U or K+Th, and yet others are high in all three elements (Jaques et al., 1997)…..135

Figure 5 - 21: TMI aeromagnetic image for the Dubbo region ( Modified from Meakin and Morgan, 1999)……………………………………………………….141 Figure 5 - 22: Point located data for the Dubbo and Ballimore regions (Schofield, 1998)……………………………………………………………………142 Figure 5 - 23: Imaged gravity data for the Dubbo region (Modified from Meakin and Morgan, 1999)…………………………………………………………………146 Figure 5 - 24: Regional features that can be identified from a comparison of the gravity and magnetic images (Schofield, 1998)…………………………………147

Figure 6 - 1: Model with a plane horizontal ground surface and a highly irregular refractor. The vertical and horizontal scales are equal (Palmer, 1980)………153 Figure 6 - 2: Traveltime curves derived from the model in Figure 6 - 1……..153 Figure 6 - 3: Coincident velocity analysis functions for XY-values from 0 to 30 m (with same reciprocal time), derived from the traveltime data in Figure 6 - 2…. ……………………………………………………………………………………….155 Figure 6 - 4: Stacked velocity analysis functions for XY-values from 0 to 30 m (with different reciprocal time), derived from the traveltime data in Figure 6-2… ……………………………………………………………………………………….155 Figure 6 - 5: Average velocity analysis graph, derived from the traveltime curves in Figure 6 - 2………………………………………………………………156 Figure 6 - 6: Linear versus non-linear refractor velocities from Figure 6 - 2. 157 Figure 6 - 7: Time-depth for XY-values from 0 to 30 m……………………….158

xiii

Figure 6 - 8: Depth section which shows particularly for the depression at 225 m and for the fault at 50 m………………………………………………………158 Figure 6 - 9: Model with irregular ground and refractor surfaces. The vertical and horizontal scales are equal (Palmer, 1980)………………………………159 Figure 6 - 10: Traveltime curves derived from the model in Figure 6 - 9….160 Figure 6 - 11: Velocity analysis functions for XY-values from 0 to 30 m derived from the traveltime data in Figure 7-10. The data for a 15 m XY-value are judged the best……………………………………………………………………161 Figure 6 - 12: Average velocity analysis graph, derived from the traveltime curves in Figure 6 - 10……………………………………………………………162 Figure 6 - 13: Refractor seismic velocity graph derived from Figure 6 – 10.163 Figure 6 - 14: Stacked time-depth for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 - 10………………………………………………163 Figure 6 - 15: Depth sections calculated from time-depth using 15-m XY-values……………………………………………………………………………….163 Figure 6 - 16: Traveltime curves derived from the model in Figure 7-10 after one correction………………………………………………………………………164 Figure 6 - 17: First raw traveltime correction……………………………………165 Figure 6 - 18: Velocity analysis functions for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 - 10 after one correction…………………166 Figure 6 - 19: Average velocity analysis graph, derived from the traveltime curves in Figure 15, after one correction…………………………………………167 Figure 6 - 20: Refractor seismic velocity graph derived from Figure 6 - 16, after one correction………………………………………………………………………168 Figure 6 - 21: Stacked time-depth for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 - 10 after 1 iteration……………………………168 Figure 6 - 22: Depth sections calculated from time-depth using 15-m XY-values after 1 iteration……………………………………………………………………..169 Figure 6 - 23: Traveltime curves derived from the model in Figure 6 – 10 after four corrections…………………………………………………………………….170 Figure 6 - 24: Fourth time-depth correction graph…………………………….171 Figure 6 - 25: Velocity analysis functions for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 - 10 after four corrections……………….171 Figure 6 - 26: Average velocity analysis graph, derived from the traveltime curves in Figure 6 - 23, after four corrections……………………………………172 Figure 6 - 27: Refractor seismic velocity graph derived from Figure 6 - 23, after four corrections…………………………………………………………………….172 Figure 6 - 28: Stacked time-depth for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 – 10 after four corrections…………………….173

xiv

Figure 6 - 29: Depth sections calculated from time-depth using 15-m XY-values after 4 iteration……………………………………………………………………173 Figure 7 - 1: P-wave versus S-wave comparison…………………………….179 Figure 7 - 2: Geophones and shot points layout at site 1……………………191 Figure 7 - 3: Geophones and shot points layout at site 2……………………195 Figure 7 - 4: P-wave shot record………………………………………………..197 Figure 8 - 1: Arrangement of the blocks used in a model together with the datum points in the pseudosection that generate by RES2DINV software…202 Figure 8 - 2: An illustration of the ray paths of the velocity analysis function. Ray paths are shown as solid lines when they are added, broken lines when they are subtracted and as broken-dashed lines when they cancel (Palmer, 2003b)………………………………………………………………………………209 Figure 8 - 3: A schematic summary of the ray paths used in the computation of the time-depth function for GRM (Palmer, 2003b)…………………………….210 Figure 8 - 4: The process of refraction migration. The offset distance (XG and YG) is the horizontal distances between the point of emergence of the ray on the refracting interface and the point of detection at the surface (top). Inversion methods which do not explicitly recognise the offset distance suffer refractor smoothing. The GRM accommodates the offset distance by calculating different values of horizontal distance between forward and reverse receivers, then selecting the XY value for which the forward and reverse rays leave the refractor at the same point and arrive at different detector positions at the surface (bottom)……………………………………………………………………212 Figure 8 - 5: Vertical shot record. The receivers on lines 1, 2, 3 and 4 are numbered 100-128, 200-228, 300-328 and 400-428 respectively. The four receiver lines are wrapped continuously in a ‘W’ shape, where the receivers for lines 2 and 4 are in the reverse order……………………………………………215 Figure 8 - 6: The shot record in Figure 1 split into individual receiver lines. The receivers on lines 1, 2, 3 and 4 are numbered 100-128, 200-228, 300-328 and 400-428 respectively. The trace order of lines 2 and 4 have been reversed..217 Figure 8 - 7: Refraction convolution section for a vertical (left) and orthogonal horizontal (right) component showing the refractor structure and amplitudes proportional to the square of the head coefficients………………………… …218 Figure 8 - 8: The enhancement of S-wave in horizontal records using records of opposite polarity. Subtraction results in the sum of the S-wave components and cancellation of P-wave components (bottom left) whilst addition results in the opposite (bottom right)…………………………………………………………….219 Figure 8 - 9: Muting and killing traces can improve the quality of first-arrival picking. The original cross line horizontal component shot record with considerable noise and dead traces (left). The same shot record after killing and muting above the first-arrivals (right)…………………………………………….220

xv

Figure 8 - 10: The picked first-arrival traveltimes overlayed on the shot records. ………………………………………………………………………………………222 Figure 8 - 11: Survey geometry with the source line offset from the receiver line. For a source point s and receiver r, when X is less than Y or when they are comparable there is a significant distortion of the traveltime curve at that point. When X is much greater than Y the distortion is negligible…………………..224 Figure 8 - 12: Traveltime graph for shot points inside the receiver spread. Since the shot points are offset from the receiver line, direct arrival traveltimes are shifted up in time and have a concave upward shape………………………..225 Figure 8 - 13: Corrected traveltime graphs for shot points inside the receiver spread. By geometric proportioning the traveltimes are changed so that the first arrival at each shot point occurs at zero time………………………………….226 Figure 8 - 14: Traveltime graph for P-wave (site1, line1) before correction..226 Figure 8 - 15: Corrected traveltime for P-wave (site1, line1) after correction.227 Figure 8 - 16: Using phantoming to combine traveltime graphs. Each traveltime curve is shifted downwards by the average separation between itself and the base traveltime graph. The average of these graphs is then taken to obtain the combined traveltime graph……………………………………………………….228 Figure 8 - 17: Forward and reverse velocity analysis graph for line1, site1, P-wave…………………………………………………………………………………229 Figure 8 - 18: The velocity analysis function calculated for different XY values and stacked. The optimum XY value can be determined by observing for which XY value the others are symmetric about, in this case a XY value of about 5 m. ………………………………………………………………………………………230 Figure 8 - 19: Forward and reverse velocity analysis graph for XY=5, P-wave, line1, site1………………………………………………………………………….231 Figure 8 - 20: Velocity analysis residuals are obtained by subtracting the average of the velocity analysis functions from the individual velocity analysis functions. Variations between the velocity analysis functions for the different XY values are generally less than 1.5 ms…………………………………………..232 Figure 8 - 21: Velocity analysis residual for P- and S-waves, site1…………233 Figure 8 - 22: Velocity analysis graph that show near surface effect……….234 Figure 8 - 23: Velocity analysis graph before correction. Notice to the line trends for V2 and V3………………………………………………………………236 Figure 8 - 24: Velocity analysis graph after correction. Notice to the line trends for V2 and V3………………………………………………………………………237 Figure 8 - 25: Lateral variations in S-wave velocities (in m/s) within the refractor…………………………………………………………………………….239 Figure 8 - 26: Traveltime of a refracted arrival at the shot point (S) (Palmer, 2003b)………………………………………………………………………………241

xvi

Figure 8 - 27: The time-depth function calculated for different XY values and stacked. The optimum XY value can be determined by observing for which XY value small irregularities in the time-depth function are most pronounced, in this case a XY value of about 5 m…………………………………………………….242 Figure 8 - 28: time-depth residuals are obtained by subtracting the average of the time-depth functions from the individual time-depth functions. Variations between the time-depth functions for the different XY values are generally less than 2 ms……………………………………………………………………………242 Figure 8 - 29: Forward and reverse distance P-wave amplitudes for shots 5 & 14 with calculating 1/X, 1/X^2 and 1/X^3. The large geometric spreading component dominates. The effects of geometric spreading have been reduced. ……………………………………………………………………………………….247 Figure 8 - 30: The product of forward and reverse amplitudes for various offset shot pair. There are gross similarities in the shape of the amplitude products for all the shot pairs, suggesting that the variations are related to variations in the seismic velocities in the refractor…………………………………………………247 Figure 8 - 31: Final sections from WET tomography. Section generated using 1D starting model with multiple layers of assumed constant vertical velocity gradients (top). Section generated using 2D starting model derived from the GRM (bottom). Despite the large differences in the sections both fit the field data to acceptable accuracy………………………………………………………252 Figure 9 - 1: 2D contour map of magnetic data at site 1………………………258 Figure 9 - 2: 2D contour map of magnetic data at site 2………………………258 Figure 9 - 3: The resistivity result at Site1-Line1……………………………….263 Figure 9 - 4: The resistivity result at Site1-line2………………………………..263 Figure 9 - 5: The resistivity result at Site1-Line3……………………………….264 Figure 9 - 6: The resistivity result at Site2………………………………………264 Figure 9 - 7: The possible P-wave undetected layer. A change from partially saturated to completely saturated sediment above the refractor will result in an increase in P-wave velocity. If this layer of completely saturated sediment is relatively thin, then it will be undetectable in the P-wave traveltime graphs, resulting in the refractor being imaged shallower than the true depth. Since S-waves do not propagate through liquids, the degree of saturation in the sediment will have no effect on its propagation, thus the S-waves will image the refractor at its true depth…………………………………………………………..267 Figure 9 - 8: Summary of the refractor velocities at site 1. The refractor can be separated into three distinct regions of different seismic velocities…………..269 Figure 9 - 9: Cross-cutting features at site1…………………………………….270 Figure 9 - 10: Amplitude product for P-wave at site1………………………….272 Figure 9 - 11: Wave-type conversion and the GRM. The forward and reverse traveltimes employed by the GRM algorithms include one set derived from SV-P-SV-waves and the other set from pure SV-waves. When this occurs, the refractor is shifted laterally and is imaged at a greater depth…………………276

xvii

Figure 9 - 12: Summary of the refractor velocities at site 2. Refractor can be separated into four distinct regions of different seismic velocities……………277 Figure 9 - 13: Cross-cutting features at site2…………………………………..280 Figure 9 - 14: Distinction among the particle-displacement vectors associated with the three fundamental modes, P, SH, and SV, which compose vector-wavefield seismic data (Hardage et al., 2003)………………………………….281 Figure 9 - 15: Simplified regional geology map of the Hill End Trough (Modified from David et al., 2003)……………………………………………………………284 TABLES Table 1- 1: Areas (hectares) with a high potential to develop dryland salinity in Australia (From: National Land and Water Resources Audit, 2001)………….9 Table 2 - 1: Summary of the local, intermediate and regional systems (Coram, 1998)…………………………………………………………………………………26 Table 2 - 2: Evaluation attributes for different dryland salinity management options (from Coram et al., 2001)…………………………………………………38 Table 2- 3: Land management recommendations for the Yass Valley (modified from Nicoll and Scown, 1993)……………………………………………………..40 Table 4 - 1: Some applications of refraction methods………………………….66 Table 5 - 1: Description of different rocks type in the Spicers Creek area (From Cobbora 100,000 geological map)………………………………………………..109 Table 5 - 2: Proposed deformation events affecting rocks across the Dubbo 1:250,000 geological sheet Scott (Morgan, 1997)………………………………119 Table 5 - 3: Survey parameters for the Gilgandra and Dubbo airborne surveys………………………………………………………………………………126 Table 5 - 4: IAEA (International Atomic Energy Agency) recommended windows for conventional 3-channel airborne gamma-ray spectrometry (IAEA, 1991).129

Table 5 - 5: Interpretation of relative concentration of K. Th, and U in rocks based on the RGB image (Morgan, 1997)……………………………………….130

40

Table 7 - 1: Trace to station relationship at site 1………………………………190 Table 7 - 2: Trace to station relationship at site 2………………………………194 Table 7 - 3: The number of shot recorded in site 1 and site 2…………………196 Table 9 - 1: Minimum, maximum and average time-depth and depth for each wave-type……………………………………………………………………………266 Table 9 - 2: Minimum, maximum and average time-depth and depth for each wave-type…………………………………………………………………..………..274

xviii

PHOTOS Photo 2 - 1: An artesian bore discharging saline groundwater. The severity of the soil erosion is evident. Note the difference between the top of the concrete collar on the casing and the existing ground surface……………………………31 Photo 2 - 2: A photo taken to the left of photo 1 (the tree and post appearing in both). The degree of soil erosion caused by saline groundwater destroying the soil structure is evident, especially along the far edge of the accumulated groundwater……………………………………………………………………….…31 Photo 2 - 3: 50 m downstream from the artesian bore in photo 1. Salt completely covers the soil around the stream, indicating the streamwater’s high salinity levels…………………………………………………………………………32 Photo 2 - 4: Upstream from photo 3 the amount of salt covering the soil is even greater. Also notice the soil erosion along the banks of the stream………..….32 Photo 7 - 1: The ARAM24 CRU………………………………………………….177 Photo 7 - 2: IVI Mini-vibrator Model T-15000…………………………………..179 Photo 7 - 3: The vibrator operating in P-wave mode. The actuator (the blue cylinder above the base plate) is positioned vertically…………………………180 Photo 7 - 4: The vibrator operating in S-wave mode. The actuator is positioned horizontally and is perpendicular to the axis of the truck………………………180 Photo 7 - 5: GS-20DM 14 Hz three component geophone. A bubble spirit level (the white circle) is used to plant the geophone in the correct orientation. Each component is transmitted through a separate cable (marked red, white and yellow)……………………………………………………………………………….181 Photo 7 - 6: A cartridge of anzomex, a high frequency explosive composed of TNT and PENT……………………………………………………………………..182 Photo 7 - 7: The Pelton Shot Pro dynamite firing system was used to remotely control detonation……………………………………………………………….….182 Photo 7 - 8: A successful detonation of a small charge……………………….187 Photo 7 - 9: four receiver lines at site 1…………………………………………189 Photo 7 - 10: One of the receiver lines ready for recording. Pin flags are first surveyed. Three receiver cables are then laid out. Receivers are planted next to the pin flags and connected to the cables. The station units are then connected to the cables………………………………………………………………………..190 Photo 7 - 11: Minivib recording at site 2…………………………………………193 Photo 7 - 12: Geophones and cable deployment at site 2…………………….193 Photo 7 - 13: A successful detonation of a small charge at site2…………….194

xix

Chapter 1

Introduction

1.1 Recent Advances in the Resolution of Geophysical Data

In the last twenty-five years, there have been significant advances in the spatial

resolution of most geophysical sets of data. Whereas geophysical data were

once commonly acquired along widely spaced profiles, it is now more usual to

obtain geophysical data along numerous closely spaced traverses. As a result,

the spatial sampling is now comparable in both horizontal directions and the

greater resolution has greatly improved the interpretation of geological features.

Figure 1-1, taken from Reeves (1992), illustrates the effects of improvements in

the spatial sampling. The upper contour map was derived from data acquired in

1973 with a fluxgate magnetometer with accuracy of ±10 nanoTeslas and a line

spacing of nominally 1.5 km. The lower grey scale image was derived from

data recorded in 1991 with a cesium vapour magnetometer with a resolution of

±1 nanoTesla and a line spacing of 400 metres. Despite the resolution ±10

nanoTeslas of the fluxgate magnetometer, it is still not possible to recognize

features such as the folds on the left hand side or the dykes on the right hand

side, in the contour map, even though they can be readily observed in the grey

scale image. The contouring process has the properties of a low pass filter, and

does not permit the retaining of the higher frequency components.

1

Figure 1 - 1: Improvements in the resolution of geological detail have been achieved by

adopting a new generation of survey specifications and presenting the resulting grid as

an image. This illustration is from an area of the Ebagoola 1:250 000 sheet in northern

Queensland. Top: BMR survey flown (E-W) in 1973 at a mean terrain clearance of 150 m

and a line spacing of 1500 m, presented as a contour map with contour interval 10 nT.

Bottom: the same area flown (E-W) by Geoterrex in 1991 with 100 m terrain clearance and

400 m line spacing (Reeves, C.V., 1992).

2

However, it is clear that the line spacing of 1.5 kilometres is too large to permit

confident correlation of subtle anomalies on adjacent profiles, and as a result,

only a low resolution result is possible. Therefore, it can be argued that despite

its format as a two-dimensional map, the contour map has essentially the same

resolution as a series of one-dimensional profiles. It is essential to employ

significantly closer line spacings, in order to achieve a genuine 2-D image.

Perhaps one of the most spectacular demonstrations of the benefits of

improved spatial resolution in both horizontal directions, has been three-

dimensional (3-D) seismic reflection methods. In the past twenty-five years,

high resolution 3-D seismic reflection methods have revolutionized the

investigation of the sedimentary basins (Weimer and Davis, 1996). The

fundamental problem facing two-dimensional (2-D) seismic methods is the fact

that most geological targets are three-dimensional. In the past, two-

dimensional methods have attempted to counter this problem by locating lines

with the strike and dips of the major features in mind, minimizing but rarely

eliminating the effect of the third dimension (Brown 1996).

French (1974) demonstrated that 3-D migration eliminates many of the lateral

correlation ambiguities which are caused by “sideswipes” and “blind structures”.

Structure maps derived from 3-D migration processing give a true and precise

picture of 3-D models, whereas the same data when processed using 2-D

migration, results in the mapped structures being distorted in shape. However,

the distortions can be reduced by increased coverage and careful interpretive

techniques (French, 1974). As a result, to obtain a precise image from seismic

3

data it is necessary to employ sampling densities and processing methods

which recognize and accommodate the three dimensions.

Furthermore, the geological targets studied in the near surface surveys often

display considerable variations in cross-line and in-line directions in terms of

depth to and seismic velocities within (Palmer, 2001b). This makes the

adoption of 3-D methods inevitable.

Perhaps one of the most compelling demonstrations of the benefits of 3-D

geophysical methods has been the study by Nestvold (1992) which

demonstrates that 3-D seismic reflection results are often significantly different

from 2-D seismic reflection results. In many cases, 2-D methods can provide

an incorrect rather than an incomplete picture of the subsurface structure. The

differences can be explained by the fact that the geological targets can vary

significantly in both horizontal directions, thereby requiring the adequate spatial

sampling obtained with 3-D methods. In other words, the incorrect subsurface

structure can often be the result of spatial aliasing. Furthermore, most seismic

reflection processing routines, especially imaging (also known as migration),

perform much better in three dimensions because they recognize and

accommodate out-of-plane signals.

4

1.2 Three-Dimensional Shallow Seismic Refraction Methods

While three-dimensional seismic refraction methods have been used in the

study of the deep earth’s crust (e.g. Miles, 1988; Bennett, 1999), they have not

been commonly applied to near surface targets. A major reason for this

situation is the fact that most shallow refraction field systems rarely consist of

more than twenty-four recording channels. As a result, the acquisition of

satisfactory volumes of data can be inconvenient.

Nevertheless, there have been several published examples which suggest that

3-D refraction methods may generate very useful results with genuine 3-D

targets. In Palmer (2001b), a 3-D refraction survey was recorded across a

mapped shear zone at Mt. Bulga in southeastern Australia. These results

showed that there were important 3-D effects associated with the nominally 2-D

shear zone (Figure 1-2). They include cross-cutting orthogonal faults that had

not been detected with an earlier 2-D profile across the shear zone and a

change in the horizontal direction of approximately ninety degrees of the

azimuth of the highest seismic velocity in one region of the survey area.

The 2-D vertical cross section shown in Figure 1-3, clearly shows a major shear

zone as a region of low seismic velocity of approximately 2000 m/s and

increased depth of weathering between stations 54 and 60. However, the

horizontal 2-D map in Figure 1-2 shows a more complex geological situation in

terms of cross cutting orthogonal faults and the rotation of the horizontal

5

Figure 1 - 2: Results from the 3-D Mt Bulga seismic refraction survey. Seismic velocities

and interpreted faults are plotted over the contours of the time-depths in milliseconds.

The bold arrows indicate the directions of the higher seismic velocity (Palmer 2001b).

Figure 1 - 3: The standard 2-D refraction depth cross section obtained across a shear

zone at Mt Bulga. The shear zone (between stations 54 and 60) exhibits a decrease in the

seismic velocity and an increase in the depth of weathering (Palmer, 2001b).direction of

6

the azimuthal anisotropy of the highest seismic velocity in one region of the

survey area.

Other 3-D shallow seismic refraction studies (e.g. Musgrave, 1970; Palmer

1986; Ebrom et al., 1992; Taner et al., 1998; Deen et al, 2000, Deen and Gohl,

2002) have demonstrated the usefulness of 3-D refraction methods in the

shallow environment.

A major benefit of high resolution geophysical data sets, is the ability to

generate pseudo-geological maps. The geological map traditionally has been a

one of the most important domains for visually summarizing geological

information. The reasons for this are that geological data have traditionally

been acquired through examination of various sites, exposures, etc on the

earth’s surface, and therefore, the geological map is a self evident summary of

these activities. Furthermore, many geological processes take place in the near

surface region, and therefore, it is frequently more convenient to conceptualize

in the horizontal domain.

On the other hand, geophysical investigations have traditionally started with a

simple traverse, and then proceed to invert the data in terms of depths and the

appropriate petrophysical properties. As a result, most geophysicists tend to

conceptualize the subsurface in terms of vertical cross sections.

Clearly, these are gross generalizations, and all competent geoscientists seek

to construct a 3-D model of the surface. However, it is worth noting that the

7

greatly increased use of airborne geophysical methods for geological mapping

was more a result of improved methods of data presentation with image

processing software, rather than more detailed methods for depth inversion.

1.3 Dryland Salinity

Dryland salinity or secondary soil salinization is a form of land degradation

which seriously affects many parts of the Australian landscape. The problem of

salinity was first identified by Walter Earnest Woods in 1924, who investigated

the deterioration of land and watercourses in the Western Australia wheat belt

in the years following clearing of native vegetation for agricultural land-use

(English, 2002). The occurrence of dryland salinity has major implications for

agricultural, industrial and domestic activities as well as for the natural

environment. This form of land degradation leads to a decline in crop and

native plant growth, increase in soil erosion, and the pollution of rivers and

groundwater

Dryland salinity is a major problem in many parts of inland Australia. According

to the National Land and Water Resources Audit (2001), over the next fifty

years, the area at risk from dryland salinity in Australia will increase from a

current 5.7 million to 17 million hectares (Coram et al. 2001). The audit states

that as a result of this three-fold increase, the area of native vegetation at risk

will expand from a current 630,000 to potentially 2,000,000 hectares, and the

length of streams at risk will be over 20,000 kilometres. The economic cost of

8

repairing the more than 200 towns expected to suffer infrastructure damage

and the 52,000 kilometres of road facing damage is substantial.

The largest areas of dryland salinity are in the agriculture zone of south-west

Western Australia. Large areas are also at risk of dryland salinity in South

Australia, Victoria and New South Wales, mainly in the Murray-Darling Basin

where ground water levels are still rising (Bradd and Gates, 1996). In New

South Wales approximately 120,000 hectares of land are estimated to be

currently affected by salinity, of which over 90 per cent occur in the five

catchments of the Murray, Murrumbidgee, Lachlan, Macquarie and Hunter

rivers. Table 1-1 is shown the areas with a high potential to developed dryland

salinity in Australia (National Land and Water Resources Audit, 2001). The

Northern Territory and the Australian Capital Territory were not included as the

dryland salinity problem was considered to be very minor.

Table 1- 1: Areas (hectares) with a high potential to develop dryland salinity in Australia

(From: National Land and Water Resources Audit, 2001).

State/Territory 2000 2050

NSW 181000 1300000

Vic 670000 3110000

Qld not assessed 3100000

SA 390000 600000

WA 4363000 8800000

Tas 54000 90000

9

The extent and severity of salinity problems are sometimes measured in terms

of the area of land affected, value of lost agricultural production and the effect

on water quality. It has been estimated that approximately 16090 km of land

in the non-arid zone of Australia is affected by irrigation salting and 324090 km

affected by natural and man-induced dryland salting (ACIL, 1983). The latter

represents approximately 35 per cent of the total area of land in Australia

affected by some form of soil degradation.

2

2

Dryland salinity occurs extensively throughout the Spicers Creek Catchment in

central eastern New South Wales, Australia. The surface discharge of saline

groundwater at several locations has led to a decline in crop and native plant

growth, increase in soil erosion, and the pollution of rivers and groundwater. At

several of these locations the saline groundwater seepage zones are

geologically induced, correlating with a shear zone inferred from geological

mapping. These zones of fractured bedrock enable the saline groundwater,

contained within deep basement aquifers, to flow upwards, where they mix with

shallow groundwater and are eventually discharged at the ground surface

causing extensive soil salinisation.

1.4 Detailed Mapping of Shear Zones

A shear zone is defined by Bates and Jackson (1980) as “a tubular zone of rock

that has been crushed and brecciated by many parallel fractures due to shear

strain”. The general characteristics of shear zones can differ in the field. The

10

largest are hundreds of kilometers long and tens of kilometers in vertical extent,

with displacements of tens to hundreds of kilometers. The smallest shear

zones observable in outcrop are typically several centimeters long and one

millimeter thick, and may have a centimeter or so of displacement. Others are

only visible in thin section (Davis and Reynolds, 1996). In the field, shear zones

are characterized by nearly vertical (75˚-90˚) foliation planes, subhorizontal (0-

15˚) lineations and in some cases, mylonitic texture (Pili et al., 1997).

In order to map shear zones, it is important to be able to visualize their key

features. Shear zones are fault-like in that they accommodate displacements.

Unlike ordinary fault surfaces, however, shear zones often do not display

physical breaks. Instead, the movement is achieved by penetrative deformation

within a zone that may be centimeters or kilometers across. Shear zones are

commonly arranged in networks or sets composed of a number of individual

shear zones. They can accommodate crustal extension and place younger,

high level rock down onto deeper, old rocks (Davis and Raynolds, 1996).

Despite their significance, the subsurface properties of shear zones are not

easy to map in detail. In many cases their existence is often inferred from

surface and sometimes underground geological mapping. Furthermore, the

precise morphology can be difficult to be determined from either surface

mapping or from boreholes. Where there is no clear evidence in the field,

geophysical methods are frequently used to detect and map the position of

shear zones.

11

Geophysical techniques, especially electrical resistivity, magnetic and

electromagnetics (EM) methods, have long proven to be successful in detecting

shear zones as regions of greatly reduced electrical resistivities or of magnetite

depletion. The success of EM and resistivity methods lies in their ability to

image these narrow zones of fractured bedrock as regions of increased

conductivity. This increase in conductivity is due to the development of

secondary porosity and permeability from fracturing, leading to an increase in

ground water saturation.

These techniques do not always map the shear zones, especially those which

lie within an extensive region of low resistivities caused by groundwater

salination. The geological noise produced by high conductive overburden

during EM and resistivity surveys can frequently obscure the conductivity

anomaly due to the shear zone in the bedrock. This condition occurred at

Spicers Creek Catchment, where the widespread occurrence and high level of

salinisation in the catchment has resulted in a highly conductive overburden and

accordingly, it was anticipated that EM and resistivity methods would not be the

most suitable for imaging the shear zone.

1.5 Objectives

Salination occurs discontinuously throughout the Spicers Creek Catchment

where saline groundwater is discharged at the surface at several locations

along a shear zone, inferred from airborne geophysical surveys and from

12

subsequent regional geological mapping by the NSW Geological Survey.

However, there has been no confirmation of the exact location of the shear

zone or its precise correlation with saline groundwater discharges at the

surface.

Although it was anticipated that the major fault, the Nindethana Fault, is an

important control in the discharge of saline groundwater, the specific geological

conditions causing the discharge were not clear. In particular, the causes of he

localized regions of discharge, generally less than 10 metres across, were not

immediately apparent. In the study area two sites (site 1 and site 2) have

chosen. At site 1, the discharge occurs where there is a minor reduction in the

topography. However, at Site 2, the discharge occurs along a gentle slope in

the topography. In addition to possible topographic effects, other causes might

include permeability barriers within the shear zone or possibly, cross cutting

structures, such as other orthogonal faults.

The major aim of the geophysical investigations was to provide detailed in-situ

mapping of the shear zone, in order to determine the most likely causes of the

discharge of saline groundwater. High resolution magnetic, electrical imaging

and 3D-3C shallow seismic refraction methods were carried out to map the

shear zones in detail.

Initially, high resolution magnetic and resistivity imaging were carried out at both

sites. While the magnetic survey did confirm the location of the shear zone as a

region of low magnetic field strength, the absence of a magnetic anomaly

13

precluded any detailed mapping of the internal features of the shear zone.

Also, resistivity imaging did not achieve satisfactory penetration through the

conductive overburden, sufficient to map the shear zone in detail.

Therefore, the seismic refraction method was employed, because shear zones

can frequently be detected as regions of increased weathering and decreased

seismic velocity. In view of the very localized expression of the discharge

zones, it was considered that the target was probably genuinely 3-D in nature.

Therefore, several closely spaced parallel lines, forming a 3-D pattern, were

employed. Finally, the use of three-component (3C) recording was employed,

to determine whether the target would have a different and possibly a more

definitive response with shear waves.

The main objectives of this study were:

1 To determine whether the shear zone occurs as a relatively narrow

region with a low seismic velocity.

2 To determine whether 3-D methods were able to provide an enhanced

image of the shear zone.

3 To determine whether the shear waves were able to provide a more

useful image of the shear zone, when compared with compressional waves.

The seismic data was recorded across two sites along strike of a major shear

zone using the Australian National Seismic Imaging Resources (ANSIR) 360-

trace system. Three component (3C) geophones were used to record shear

waves as well as compressional waves. An IVI minivibrator T-15000 was used

14

as the main source of energy for the seismic survey although some explosive

charges in shallow shot holes were used for comparison. Data were processed

using the generalized reciprocal method (GRM, Palmer 1980, 1986), Seismic

Un*x freely distributed by the Centre for Wave Phenomena at the Colorado

School of Mines and Visual-SUNT 6, generously donated by W_Geosoft.

1.6 Problems Solved in this Study

This study demonstrates that 3D – 3C seismic refraction methods are able to

provide a good estimate of the structure or shape of the refracting interface. As

discussed in a previous section, there has been no evidence to confirm the

existence of the inferred shear zone or cross-cutting features at the sites in the

Spicers Creek Catchment. This study confirms that the Nindethana fault

occurs as a region of low seismic velocities and increased depths of

weathering, and that the regions of salination correspond with the occurrence

of cross-cutting faults along the Nindethana fault.

This study demonstrates the ability to generate a reasonably detailed model of

the shear zone in the Spicers Creek Catchment in the region of the discharge

of saline groundwater. As with the Mt Bulga case history described earlier, it is

clear that the three-dimensional seismic refraction method was used able to

image the important cross-cutting features, as well as the major shear zone,

which were not as readily detectable with two-dimensional methods.

15

However, it was necessary to process the data so that the seismic signature of

the shear zone was enhanced. This required the development of methods to

include most of the field data in order to improve the statistical reliability of the

data, as well as an effective method to correct for near surface irregularities.

The latter can often result in ambiguous interpretations of the field data and

therefore can adversely effect the determination of accurate seismic velocities

in the refractor.

A new method, which is described as part of this study, takes advantages of a

unique feature of the GRM to compensate for localized variations in the near

surface seismic velocities and/or topography, termed the GRM SSM for GRM

statics smoothing method. This method significantly improved the reliability

and consistency in determining seismic velocities in the refractor.

Furthermore, this study demonstrates the importance of generating robust

starting models for inversion. In recent years, the methods of model-based

inversion have been used to the analysis of refraction data and involve the

derivation of an initial or starting model (Smith and Ferguson, 2000; Palmer

2003a) and then refining that model until the traveltimes generated with the

model, match those of the recorded field data. Wavepath eikonal traveltime

(WET) tomography was used to generate a velocity-depth section of every line

for each wave type. Starting models were generated using both one

dimensional inversion methods, namely the delta-t-V method (Gebrande and

Miller 1985), and two-dimensional inversion algorithms with the GRM.

16

Chapter 2

Dryland Salinity

2.1 Summary

Large areas throughout Australia are currently at risk of being affected by

dryland salinity. In New South Wales, approximately 125 000 hectares of land

is currently affected. Dryland salinisation develops where a source of salt

exists. This salt can be contained within the colluvium, having accumulated

there due to rainfall, aeolian processes, and the weathering of soil and rock

minerals, or it can be derived deeper in the bedrock from saline groundwater in

an unconfined aquifer.

The extent of dryland salinity in the study area, the Spicers Creek Catchment,

has severely altered the landscape, having major environmental implications.

Large areas of the catchment have experienced soil erosion resulting from the

saline groundwater in the surface soil causing the destruction of clay and soil

structure. Saline groundwater is discharged at the surface at several locations.

A significant amount of the saline groundwater continuously discharges through

shallow alluvial/colluvial aquifers at intermittent locations along an extension of

the Nindethana fault, and forms a “saline plume” contaminating the soil down

gradient from the seepage zone.

17

2.2 Introduction

Salinity is a problem, which affects about one-third of the irrigated land around

the world (Reeve and Fireman, 1967). Salinity is believed to have been a

problem with Mesopotamian agriculture about 4500 years ago (Jacobsen and

Adams, 1958). Australia is one of the few countries which experience salinity in

non-irrigation (dryland) farmland (Peck et al., 1983). Dryland salinity or

secondary soil salinisation is associated with the movement of saline

groundwater, and it is an indication of a changing water balance in the

catchment. It is associated with particular land uses that cause groundwater

levels and salt contents to rise by allowing deep percolation of rainwater

through the soil. Dryland salinity is a form of land degradation, which seriously

affects many parts of Australian landscapes.

Dryland salinity is caused by a number of earth processes and the salts

involved have multiple sources. Dryland salinity can occur where soils contain

high levels of salt because of rising groundwater levels. Usually plants and soil

organisms are killed or their productivity is severely limited on affected lands.

The occurrence of dryland salinity has major implications for agricultural,

industrial and domestic activities as well as for the natural environment. This

form of land degradation leads to a decline in crop and native plant growth,

increase in soil erosion, and the pollution of rivers and groundwater

The water quality of river systems is adversely affected by dryland salinity.

Freshwater salinity is a by-product of dryland salinity whereby salt in the surface

18

soil eventually enters river systems, leading to an increase in water salinity.

This occurs via two mechanisms. Firstly, the salt can come from overland flow

where rainfall washes salt off the affected land and into river systems at the

bottom of catchments. This is especially the case when rivers are not lined with

native vegetation which filters incoming pollutants such as salt. Secondly, the

salt can enter river systems directly from the groundwater, especially where the

rivers are low. An increase in freshwater salinity can devastate the ecology of a

river system as well as reducing water quality for domestic, industrial and

agricultural purposes.

The extent of dryland salinity in the study area, Spicers Creek Catchment, has

severely altered the landscape, having major environmental implications. The

high concentration of salt in the groundwater has led to a significant decline in

agricultural productivity and a reduction in native vegetation. Large areas of the

catchment have experienced soil erosion resulting from the saline groundwater

in the surface soil causing the destruction of clay and soil structure.

2.3 Extent of Dryland Salinity in New South Wales

Dryland salinity is a major problem in many parts of inland Australia. Bradd et

al. (1997) mapped the spatial distribution of dryland salinity hazard throughout

NSW using a prediction tool based on the correlation between the current

occurrence of dryland salinity in NSW and specific combinations of land and

physiographic attributes (Figure 2-1). The results of their work suggest that up

19

to 5 million hectares of land has the potential of being affected by dryland

salinity with 4 250 000, 470 000, and 310 000 hectares of land in the

moderate, high and very high hazard categories respectively.

In New South Wales approximately 120,000 hectares of land are estimated to

be currently affected (Bradd and Gates, 1996), while a further 5,000,000

hectares are considered to be at risk on the basis of their physical

characteristics. In New South Wales some regions are more vulnerable to

dryland salinity such as the Lachlan, Murrumbidgee and the Macquarie river

basin. Areas of major concern with high to very high dryland salinity hazard

index are located in a north-south belt just north of Canberra (Yass River

Valley), and in the south-western part of the Lachlan River catchment, east of

Wagga Wagga in the Murrumbidgee River catchment, and east of Dubbo in the

Macquarie River catchment. In the north of NSW, the potential for dryland

salinity is less (Coram, 1998). Figure 2-2 shows the dryland salinity risk in NSW

in 2001 and Figure 2-3 is predicted dryland salinity risk in NSW by 2050.

20

Figure 2 - 1: Dryland salinity hazards in New South Wales (taken from Bradd, et al. 1997).

The area that is within each hazard category is approximately low 75,060,000 ha;

moderate 4,250,000 ha; high 470,000 ha; and very high 310,000 ha.

21

Figure 2 - 2: Dryland salinity risk in New South Wales 2000 (From National Land & Water

Resources Audit, 2000).

Figure 2 - 3: Dryland salinity risk in New South Wales 2050 (From National Land & Water

Resources Audit; 2000).

22

2.4 Causes of Dryland Salinity

The general cause of dryland salinity is well known. It was described in

Australia as early as 1924 with the publication of ‘Increase in salt in soil and

streams following the destruction of native vegetation’ by Walter Wood in the

Journal of the Royal Society of Western Australia. Coram (1998) has

developed a national catchment classification system in Australia and proposed

that the broad features of common occurrences of salinisation can be described

by the fifteen hydrogeological models according to whether local, intermediate

or regional flow systems are present. The broad features of local, intermediate

and regional systems are summarized in Table 2-1. These fifteen

hydrogeological models are as below:

1) Local models:

a) Discharge over lower hydraulic conductivity structures.

b) Discharge from unweathered rock aquifers at the break of slope.

c) Discharge from weathered fractured rock aquifers at break of slope.

c) Discharge from colluvial/alluvial slopes.

d) Discharge controlled by stratigraphy (sediments/basalts).

e) Discharged from perched aquifer.

f) Discharge controlled by linear features of contrasting hydraulic

conductivity.

g) Discharge from low hydraulic conductivity aquifer.

2) Intermediate models:

23

a) Capillary discharge from unconfined aquifers in valley floors.

b) Discharge from semi-confined aquifers controlled by facies changes

and changes in aquifer geometry.

c) Discharge from unweathered fractured rock aquifer at break of slope.

d) Discharge across topographic divides controlled by large, transmissive

linear structures.

3) Regional models:

a) Regional confined and semi-confined aquifer discharge controlled by

structure.

b) Regional confined and semi-confined aquifer discharge controlled by

facies change.

c) Regional unconfined to semi-confined aquifer discharge controlled by

topography.

Coram (1998) has shown that the model 1(d) is applicable for the area around

Dubbo. In this model salinisation occurs in sedimentary sequences or basaltic

rocks because of groundwater discharging from local or perched systems where

high hydraulic conductivity layers terminate at erosional surfaces or at

faulted/unconformable contacts. Also discharge occurs from a permanent and

large saturated zone.

There are several sources of the dryland salinity in the Lachlan River catchment

in southern NSW. Acworth and Jankowski (2001) have carried out a detailed

study involving drilling, geophysics, hydrogeochemistry and groundwater

monitoring over a 10-year period at a small catchment in southeast of NSW

24

(Yass River) to investigate the sources of salt causing dryland salinity. They

demonstrated that there is no hydrogeochemical evidence to suggest

evaporative or transpirative concentration of salt in the groundwater and that

there is a likely windblown source for the salt.

Different mechanisms for controlling the development of dryland salinity have

been suggested in different parts of Australia by several authors. Knight et al.

(1989) demonstrated that there is a relationship between dryland salinisation

and saline groundwaters discharging through a fault zone. Lewis (1991)

suggested linear geological features such as dykes, fractures and shear fault

zones can be responsible for salinity.

Waltham (1997) described shear zones which were the result of drag of the salt

by the movement of the hanging wall. Dahlhaus et al. (2000) also described the

effect of discharge at a break in slope as a result of the differences in sediment

permeability. Acworth and Jankowski (2001) demonstrated that the saline

seepages originated from the fractured bedrock aquifer on the lower slopes and

valley floors, where the bedrock is close to the ground surface. Clarke et al.

(2002) illustrated the affect of geological structures on dryland salinity; and

finally the recent study by Morgan and Jankowski (2002) demonstrated that the

zone discharging salinity occurred near a fault.

25

Table 2 - 1: Summary of the local, intermediate and regional systems (Coram, 1998).

Local flow

system

Intermediate flow

system

Regional flow

system

Horizontal scale 1-3 Kilometers 5-10 kilometers >50 Kilometers

Geomorphology -Subcatchments in

higher relief on

edges of plateaus

and ranges

-Alluvial and occasionally,

glacial valley fill in foothills

and valleys

-Broad riverine

plains on

depositional basins

Geology -Fractured

metamorphics and

igneous rocks.

-Colluvial

sediments.

-Aeolian sediments

-Fractured metamorphic

and igneous rocks shallow

(<50 meters deep) alluvial

and colluvial sediments.

-Deep, interbedded

marine, alluvial and

Aeolian

sedimentary

sequences (several

hundreds of meters

deep).

Structural

features

influencing

groundwater flow

-Subsurface low

hydraulic

conductivity

features such as

bedrock highs and

dykes.

-Termination of

aquifer at erosional

surface.

-reduction in hydraulic

conductivity with distance

from sedimentary source or

aquifer weathering.

-Reductions in hydraulic

gradient or aquifer

thickness with shallowing of

the ground surface

-Subsurface low hydraulic

conductivity features such

as bedrock highs and

dykes.

-Reductions in

hydraulic

conductivity with

distance from

sedimentary source

or associated with

structural deforming

(faulting, folding).

-Reductions in

hydraulic gradient

or aquifer thickness

with shallowing of

the ground surface.

26

A more comprehensive list of salinity causes is as follows (Creelman, 2002):

Salt from the atmosphere: deflated from the sea or picked up from the land

surface, referred to as cyclical salts.

Salt from the interior of the continent: older cyclical salts that have been

accumulated in the interior of the continent, and can be transported by their

adherence to windborne dust.

Salt from anthropogenic sources: soluble products from industrial activities

such as combustion, industrial processes that use and then dispose of water.

Salt from relict seawater: this is salt that accumulates as enclosed marine

environments are evaporated. The salt is contained in the fluid inclusions within

the rocks and is partially the source of salts in basinal brines or evaporate

deposits.

Aeolian deposits and debris flow: related to geology and the hydrosphere

with atmospheric inputs.

Salt from rock weathering: altered rocks can contain over 100 times the

amount of chloride as unaltered rocks.

Salt elements exchange onto clay minerals: as seawater deposits the clays,

or as clays are deposited in seawater.

Dissolution of evaporates: within the drainage basin.

27

2.5 Dryland Salinity in the Spicers Creek Catchment

In the Spicers Creek Catchment, which is located, 50 km east of Dubbo in

central eastern New South Wales, Australia (Figure 2-4), dryland salinity occurs

extensively throughout the catchment. Saline groundwater is discharged at the

surface at several locations. A significant amount of the saline groundwater

continuously discharges through shallow alluvial/colluvial aquifers at intermittent

locations along an extension of the Nindethana fault, and forms a “saline plume”

contaminating the soil down gradient from the seepage zone. Furthermore, the

salinity is not related to the parent soil material or in situ evaporative processes.

In the area of this study Morgan and Jankowski (2002) have proposed a

seepage zone that originates in deep Palaeozoic sediments.

6420 000mN

700

000m

E

690

000m

E

6430 000mN

N BALLIMORE

680

000m

E

To DubboSpicers Creek

Talbragar River

Racec

ourseG

ully

Site 2

Site 1

Snake River

Figure 2 - 4: Map of the Spicers Creek catchment with the main town, road and drainage

modified from Cobbora 1:50,000 topographic map.

28

Previous work by Morgan and Jankowski (2004) has identified several areas in

the Spicers Creek Catchment where the discharge of saline groundwater at the

surface correlates with a north–south-trending shear zone which has been

inferred from airborne geophysical surveys and from subsequent regional

geological mapping. This zone of fractured bedrock enables the saline

groundwater, contained within deep basement aquifers, to flow upwards, mixing

with shallow groundwater within the colluvium, and eventually being discharged

at the ground surface causing extensive soil salinisation. However, there have

been no studies of the exact location of the shear zone and its correlation with

saline groundwater discharges at the surface.

The high concentration of salt in the groundwater has led to a significant decline

in agricultural productivity and a reduction in native vegetation. Large areas of

the catchment have experienced soil erosion resulting from the saline

groundwater in the surface soil causing the destruction of clay and soil

structure. The severity of soil erosion can be seen in photos 2-1 and 2-2 of an

artesian bore discharging saline groundwater. The difference between the top

of the concrete collar on the casing and the existing ground surface is

considerable. Photos 2-3 and 2-4 show the water flow approximately 50 metres

downstream from the artesian bore in photo 2-1. Salt completely covers the

majority of the stream bank, indicating the high concentrations of salt in the

stream.

At several sites, saline groundwater seepage zones are geologically induced

having formed adjacent to a north–south-trending shear zone. These zones of

29

fractured bedrock enable the saline groundwater, contained within deep

basement aquifers, to flow upwards, mixing with shallow groundwater within the

colluvium, and eventually being discharged at the ground surface causing

extensive soil salinisation.

The extent of dryland salinity in the Spicers Creek Catchment has severely

altered the landscape, having major environmental implications. The principal

land use in the Spicers Creek Catchment is agricultural. Large areas of land

are occupied by cereal cropping such as wheat, canola and legumes, whilst

sheep grazing occurs in areas of partly cleared native vegetation. The increase

in agricultural activities has resulted in the replacement of high-water use native

vegetation with low-water use crops, resulting in increased groundwater

recharge and rising water tables. This, together with the landscape’s naturally

saline environment, has led to large areas of the catchment developing dryland

salinisation. Fifty-two sites have been identified as affected (Jewel and

Associates, 2000).

30

Photo 2 - 1: An artesian bore discharging saline groundwater. The severity of the soil

erosion is evident. Note the difference between the top of the concrete collar on the

casing and the existing ground surface.

Photo 2 - 2: A photo taken to the left of photo 1 (the tree and post appearing in both). The

degree of soil erosion caused by saline groundwater destroying the soil structure is

evident, especially along the far edge of the accumulated groundwater.

31

Photo 2 - 3: 50 m downstream from the artesian bore in photo 1. Salt completely covers

the soil around the stream, indicating the streamwater’s high salinity levels.

Photo 2 - 4: Upstream from photo 3 the amount of salt covering the soil is even greater.

Also notice the soil erosion along the banks of the stream.

32

2.6 Management of Dryland Salinity

Although many management options are recognized for dryland salinity, there is

little information available on the possible effectiveness of alternative actions for

salinity control in specific area. Peck et al. (1983) demonstrated that the

following specific data and information needs for salinity management in

Australia:

1 Data on the effect of land-forming on irrigation water use.

2 Data on water use by different plants and under different management to

estimate groundwater recharge rates.

3 Methods to establish location and rate of groundwater recharge and

present or potential discharge.

4 Definition of the hydrogeological situation in which a number of

representative salt seepages have developed.

5 Application of groundwater modeling techniques to areas with

representative hydrogeology to predict the effect of management options.

6 Field studies of the effectiveness of possible treatments and

combinations of treatments for control of groundwater recharge and discharge.

7 Treatments to be considered include diversion and control of surface

waters; chemical or physical modification of soil properties; biological

manipulation of plant growth patterns; subsurface drainage; selection of crops;

and shrubs or trees.

8 Salinity models and field methodology to provide an early warning of salt

discharge before land and water are degraded.

33

9 Estimate of the time-scale for development of salinity problems and for

their control by various management actions.

10 Methods to locate the most effective sites for de-watering bores.

11 Models of river water quality which take account of the effects of land

uses on groundwater flow to the river.

12 Models of river water quality to meet more than one management

objective.

13 Studies of the long term salt distribution patterns developed in

groundwater systems associated with evaporation basins, particularly in the

absence of commercial salt production from the basin.

Coram et al. (2001) showed that dryland salinity management options can be

broadly devided into four categories. These are prevention, treatment of

causes, amelioration of symptoms, and productive uses of saline resources.

Examples of these are shown in Table 2-2.

Prevention options include the retention of native vegetation, and employing

suitable agricultural techniques in order to keep land and surface waters free

from salinity. Based on their work in the Yass River Catchment, Nicoll and

Scown (1993) developed a technique for determining dryland salinity

management strategies which is mostly preventative in nature. Dryland salinity

has long been recognized in that area, with much of the landscape being

vulnerable. Eight land management classes were identified based on the

capability of the affected land, recognizing factors such as soil type, land use,

timber cover, and terrain slope (Table 2-3).

34

Options to reduce land and surface water salinisation by treatment of the cause

involve either reducing the intake of water at the recharge zone or intercepting

the groundwater prior to infiltration into the discharge zone. The former can be

accomplished by planting communities of deep rooted trees and shallower

rooted shrubs in strategic areas of recharge. The latter is accomplished

through engineering works based on active or passive drainage.

Amelioration is achieved through a range of approaches which seek to minimize

the impact the added salt has on affected land and water resources. Alternative

farming practices such as implementing raised beds in dryland cropping,

cultivating a mixture of salt tolerant and conventional crops, and minimizing

grazing by livestock enables continued agricultural productivity whilst

diminishing land and surface water salinisation. Land and surface water

salinisation can be further diminished by managing saline discharge through the

revegetation of target areas with salt tolerant native plants and trees.

Some management options are based on not seeking to prevent or reduce

dryland salinity but rather employing alternative and productive uses for the

saline land and water resources. Aquaculture, currently one of Australia’s

fastest growing primary industries has been implemented in inland areas using

saline groundwater. The successful breeding and cultivating of seawater

species such as tiger prawns, tuna, black bream, salmon, marron and redclaw

crayfish in saline groundwater suggests that an inland aquaculture industry is

viable (Sexton, 2003). Sustainable forestry is perhaps another viable use for

35

salt affected land. Experiments in selective breeding of tree species have

produced hybrids that have both a tolerance for saline water and produce

timber. The use of such trees in sustainable forestry will enable the

management of land for water movement and salinity, whilst at the same time

maximize the productive use of the land (Sexton, 2003).

A number of options are available for the prevention and remediation of dryland

salinity in the Spicers Creek Catchment. Most preventative options are based

on managing the vegetation in areas of risk so deep drainage does not

increase. Nicoll and Scown’s (1993) land management class six (LM 6) for

saline and potentially saline country in the Yass Valley could be readily applied

in areas of the Spicers Creek Catchment. For such country, they recommend

retaining existing trees; planting salt tolerant grasses, shrubs and trees; and to

straw mulch the area for better grass germination. They also recommend that

the area be fenced in, lightly grazed for short periods only when ground cover is

established, maintain a minimum of 70 % ground cover, and not to graze

pasture below 5 cm in height.

Due to the complex hydrogeological nature of the problem, strategies for the

remediation of dryland salinity are different for every catchment. In the Spicers

Creek Catchment, the successful remediation of salinisation lies in the ability to

obtain a reasonably detailed model of the possible shear zone that is

associated with the discharge of saline groundwater. Three-dimensional three-

component seismic refraction surveys were carried out at two sites of dryland

salinity in the Spicers Creek Catchment with the objective to provide this model.

36

Variables which would be useful in determining the possible remediation

measures at the Spicers Creek Catchment include the width of the shear zone

and a measure of the degree of fracturing in terms of porosity and permeability.

The possibility of grouting as a remediation option might arise if the shear zone

is quite narrow, possibly less than 10 m. Grouting along the extent of the shear

zone and the subsequent filling of the fractured bedrock would confine the

basement aquifers. This would prevent the saline groundwater from flowing

upwards in to the colluvium.

For a wider shear zone, the most suitable remediation option might be to

intercept and drain the saline groundwater before it reaches the surface soil.

This could be done with passive drainage if the shear zone is consistently

fractured, or with active drainage if it consists of a small number of large

fractures. The construction of a passive drainage system might involve setting

porous gravels at the water table along key areas of discharge and running a

drainage line underground to a collection dam. With active drainage, a number

of bore holes could be drilled into the water table above the large fractures in

order to pump the saline groundwater into a collection dam. Once in the dam,

the water could be evaporated leaving the salt to be collect and disposed.

37

Table 2 - 2: Evaluation attributes for different dryland salinity management options (from Coram et al., 2001).

Managementobjective

Strategy Outcome sought Recommended evaluation attributes

Per cent retention of native deep-rooted vegetation in high risk areasLong-term groundw ater level trendsLong-term stream salinity and salt load trendsExtent of land salinisedLand cover changeHydrogeological and soil characteristicsClimatic characteristics

Long-term groundw ater level trendsLong-term stream salinity and salt load trendsExtent of land salinisationLand cover changeHydrogeological and soil characteristicsClimatic characteristics

Long-term groundw ater level trendsLong-term stream salinity and salt load trendsExtent of land salinisationHydrogeological and soil characteristicsClimatic characteristics

Long-term stream salinity and salt load trendsExtent of land salinisationHydrogeological and soil characteristicsClimatic characteristics

Trea

tmen

t of c

ause

Prev

enta

tion

Land and surface w aters w ill remainfree from degradation from salinity

Land and surface w ater salinisationw ill diminish

Interception and disposal of salt, and reduction of groundw ater levels in transmissin zones

Land and surface w ater salinisationw ill diminish

Land and surface w ater salinisationw ill diminish

Reduce recharge to the groundw ater system by increasing vegetation w ater use through land management practices

Interception of w ater prior to infiltration or from groundw ater upgradient of discharge zone

Manage vegetation so deep drainagedoes not increase in high risk areas

38

Table 2- 2 (continued): Evaluation attributes for different dryland salinity management options (from Coram et al., 2001).

Managementobjective

Strategy Outcome sought Recommended evaluation attributes

Managing saline discharge Long-term stream salinity and salt load trendsHydrogeological and soil characteristicsClimatic characteristics

Application of soil treatments (ameliorants) Extent of land salinisationSoil w ater characteristicsHydrogeological and soil characteristicsClimatic characteristics

Establishment of salt tolerant land cover Extent of land salinisationLand cover changeIncrease in production of biomass

Extent of land salinisationLand cover change

Optimistation of the use of non-saline resources Land cover change

Alternative use of saline land and w aterresources

Productivity from non-salt affectedland w ill increase

Productivity from salt affected landand w ater w ill increase

Prod

uctiv

e us

es o

f sal

ine

reso

urce

s

Land and surface w ater salinisationw ill diminish

Agriculture production from saline landis increased

Amel

iora

tion

of

sym

ptom

s

Productivity from salt affected landw ill increase

39

Table 2- 3: Land management recommendations for the Yass Valley (modified from Nicoll and Scown, 1993)

Land managementclass

Land use Description Recommended technique Attributes of recommendedland management technique

cropping country deep soils land capable of regular cultivationw ell structured maintain soil structure by crop rotation

protect area w ith w indbreaksavoid long, bare fallow s, avoid fallow s on f lood prone areasuse legume break crops to replace fallow smanage area to produce maximum vegetative production

rotational cropping country deep soils land capable of regular cultivationw eak structure maintain soil structure by crop rotation

prevent soil erosion by use of soil conservation earthw orksuse direct drill or reduced tillage practisesavoid long, bare fallow s, avoid fallow s on f lood prone areasestablish perennial pastures after cropping phaseprotect area w ith w indbreaks

good grazing country moderately deep soils agroforestry techniques utilise high WT and good soilsestablish w indbreaks and treelotsretain all existing timberfertilise according to soil tests, on a regular basismaintain a minimum 70% groundcover all yearspread stock w atering points to ensure even pasture grazingdo not graze pasture below 4cm in height

improved perennial pastures, potentially agroforestry as technology develops

LM 3

conservative farmingtechniques

LM 1

LM 2

minimum 4 years perennial pasture phase betw een crops

40

Table 2- 3 (continued): Land management recommendations for the Yass Valley (modified from Nicoll and Scown, 1993).

Land managementclass

Land use Description Recommended technique Attributes of recommendedland management technique

seasonal grazing country seasonally w aterlogged should not be cultivated due to the high risk of soil erosionretain w ater tolerant perennial speciesestablish perennial improved pasturesgraze w hen soils are firmmaintain even cover of vegetation, not less than 70%do not graze pasture below 4cm in heightremove stock w hen soils become w aterloggedfence out actively eroding drainage lines

poor grazing country soils highly erodible most suited to native pastures encourage re-seeding of desirable native specieslow available moisture use deep rooting perennial grasses

apply fertilisers on a regular basisavoid set-stocking natve pasturesmaintain minimum 70% ground cover all yearestablish w indbreaks and treelotsspread stock w atering points to ensure even pasture grazingdo not graze pasture below 4cm in heightaim tow ards 20% tree cover

fence out arearetain existing treesplant salt tolerant grasses, shrubs and treesimplement earthw orks as required, to stop soil erosionstraw mulch area for better germination of grasseslighlty graze for short periods onlydo not graze w hen w aterloggeddo not graze pasture below 5 cm in heightmaintain minimum 70% ground cover all year

most suited to w ater tolerant pastures. Not suitable for cultivation

LM 4

LM 5

saline and potentially saline country

saline depending onw atertable depths andsalt sources

LM 6

suited to saltland agronomy, lands to be treated for saline discharge

41

Table 2- 3 (continued): Land management recommendations for the Yass Valley (modified from Nicoll and Scown, 1993).

Land managementclass

Land use Description Recommended technique Attributes of recommendedland management technique

steep, stony shallow soils retain mature forest cover and regrow thestablish block plantings of treesexclude livestock at critical periods of tree f low eringperiodically graze to reduce fire risksmaintain minimum 70% ground cover all yearselective thinning of trees is possible on slopes

non-agricultural land very steep slopes severe soil erosion hazardsrocky areas fence area to exclude livestock and feral animalsw ater bodies encourage native animals/habitat

encourage natural forest regeneration

best suited to natural vegetation, not suited to agricultural production

LM 8

suited to green timberproduction

best protected from soil erosion w ith trees, potentially high recharge areasLM 7

42

Chapter 3

Shear Zones Characteristic

3.1 Summary

Despite their significance, shear zones are difficult to map using conventional

geological methods. Shear zones can be very difficult to detect from the

surface, and therefore the use of surface mapping to define them is frequently

unreliable. Without the use of geophysical methods, shear zones are frequently

inferred from geological mapping, with little direct evidence of their existence.

Shear zones may act either as closed or open geochemical systems with

respect to movement of fluids and elements, irrespective of the size of the zone.

Shear zones are known to perform as preferred pathways for fluid circulation.

Little is understood about the 3D geometry (the relationships between length,

width and depth) of shear zones and in particular the depth of their root zones.

In the Spicers Creek Catchment magnetic, electrical resistivity and seismic

refraction methods were used to image the inferred shear zones. It was

anticipated that in this area, seismic refraction would provide the best results

since the propagation of seismic waves would not be affected by the high level

salinisation whilst also providing possibly greater resolution.

43

3.2 Introduction

Dennis (1967) showed that the term shear zone refers to a tabular region of

pervasively faulted rock, that is, a fault zone containing a very large number of

closely spaced and anastomosing fault surfaces. A shear zone is defined by

Bates and Jackson (1980) as “a tubular zone of rock that has been crushed and

brecciated by many parallel fractures due to shear strain”. Ramsay and Auber

(1983) defined a shear zone as a planar or curviplanar zone of high deformation

which is long relative to its width (length to width ratio greater than 5:1) and

which is surrounded by rocks showing a lower state of finite strain. Shear

zones may consist of a series of interleaving, anastomosing brittle faults and

crushed rock (cataclasite) formed near the surface, or of ductile shear zones

composed of mylonitic rocks produced by faulting at great depth (Hatcher,

1995). Shear zones represent regions of weak rock as well as sources of major

water flow.

There are some differences between fault, fault zone and shear zone. A fault is

a planar discontinuity between blocks of rock that have been displaced past one

another, in a direction parallel to the discontinuity (Figure 3-1A). A fault zone is

a tabular region containing many parallel or anastomosing faults (Figure 3-1B).

A shear zone, is a zone across which blocks of rock have been displaced in a

fault-like manner, but without prominent development of visible faults (Figure 3-

1c, Hobbs, 1976). Shear zones are thus regions of localized ductile

deformation, in contrast to fault zones that are regions of localized brittle

deformation. Unlike ordinary fault surfaces, shear zones commonly do not

44

display any discrete physical break (Davis and Reynolds, 1996). Another

distinction is that the normal component of displacement, which is negligible for

fault and fault zones, but may be appreciable for shear zone. The displacement

across a shear zone can be inclined at any angle, other than 90°, to the

boundaries of the zone.

Furthermore, little is known about the 3-D geometry of shear zones, especially

the depth of their root zones (Pili et al., 1997). The root zone is considered to

be deep when the outcrop length of vertical major shear zones is from a few

hundred to a few thousand kilometers because of their narrow parallel-sided

shape.

Foliation is a very common feature within the shear zones, and it is usually

inclined at an angle to the zone’s boundaries. This foliation is commonly curved

and rotates toward parallelism to the shear zone boundaries in the central

portion of the shear zone. In practice, however, although the obliquity of

foliation is relatively common, curvature of foliation within the shear zone is rare

and in many cases, foliation occurs parallel to the shear zone boundaries. The

foliation, which may be a cleavage or schistosity, represents a profound

modification of the rock’s fabric and as such imparts a mechanical anisotropy to

the rocks within the shear zones (Ramsay and Lisle, 2000).

45

Figure 3 - 1: (A) Fault. (B) Fault zone. (C) Shear zone (From Hobbs et al, 1976).

3.3 Shear Zone Types

Ramsay (1980) has classified the shear zones as brittle, brittle-ductile and

ductile (Figure 3-2). In ductile shear zones the deformation state varies

continuously form wall to wall through the zone, brittle shear zones where the

wall are separated are separated by a discontinuity on the fracture surface, and

various intermediate types known as brittle-ductile shear zones combining these

geometric features in different proportions. Brittle, brittle-ductile or ductile shear

zones are known in various lithologies and structural levels (Ramsay, 1980)

down to upper mantle (Vissers et al., 1991).

Ramsay (1980) attributes two conditions as geometric features of ideal shear

zones. The first is shear zones generally have parallel sides and the second is

the displacement profiles along any cross section through a shear zones should

46

be identical. However, some shear zones have complex geometries and have

subparallel margins and retain a fairly consistent thickness over much of their

length.

Figure 3 - 2: Different types of shear zones from (Ramsay and Auber, 1983).

Davis and Reynolds (1996) have classified shear zones in terms of general

characteristics (continuous or discontinuous); shear zones margins (parallel,

diverging and converging); shear zone sets (parallel, anastomosing and

conjugate); deflection and offset across shear zones (dextral, sinistral, normal,

reverse); plate tectonic setting of shear zones (ocean-continental convergence,

continental collision, continental rifting, transform faulting).

Hull (1988), Mitra (1992) and Means (1995) used a three-fold classification of

shear zones as types I, II and III in terms of strain and margins (Figures 3-3 and

3-4). Type I zones thicken with time (Figure 3-3, top). This is ascribed to work-

hardening of the shearing material, such that it becomes easier to deform new

material at the zone margins than to continue deforming the interior at the same

47

rate (Figure 3-4b). Type II shear zones become narrower with time. Type II

show most actively deforming zone narrows with time.

Figure 3-3 shows idealized active shear zones of Type I and II (right) and stress

/ shear strain curves for each material (left). Type III shear zones show a non-

zero strain-rate throughout the zone during its entire history, and no significant

widening or narrowing of the most-active portion with time. These types of

shear zones show different strain rates from margin to interior.

All shear zones show a localization or concentration of deformation into a

narrow shear zone. The presence of a shear zone indicates that within a given

deforming rock mass, the distribution of strain was heterogeneous rather than

homogeneous. As a result, shear zones are characterized by spatial gradients

in the amount of strain. The amount of strain is generally highest within the

center of a shear zone, decreasing outward into the wall rocks adjacent to the

zone. If the decrease in strain away from the zone is gradual without any

distinct physical break, the shear zone is considered to be continuous. If the

decrease is more abrupt, the zone is considered to be discontinuous (Davis &

Reynolds, 1996).

48

Figure 3 - 3: Idealized active shear zones of Type I and II (right) and stress / shear strain

curves for each material (left). Shaded areas are current, actively straining regions, at

margins (m) or in interior (i) of zones of Type I and II, respectively. Stubby arrows

indicate direction of migration over time of boundaries of straining regions. Ellipses

represent finite-strain gradients (From Means, 1995).

49

Figure 3 - 4: Finite shear strain vs. time curves (solid) for the three types of shear zones,

for rock at the final zone margins (m) and in the interior (i). P (dashed line) represents

some process-controlling parameter (like temperature or fluid chemistry) that is assumed

to change over the shearing history. In Type II zones, the marginal and interior acquire

their low-strain state at about the same time and p value (heavy dots). The marginal

rocks, therefore, provide a good record of former low-strain state of the interior rocks. In

zones of Types I and III, the corresponding dots are farther apart in time and p, so the

marginal rocks provide a less reliable record of the former low-strain state of the interior

rocks. The type III diagram has been drawn for the case where the marginal rocks begin

straining at the same time as the interior rocks, but more slowly. In each diagram, the

right-hand dotted line is drawn at the point along the m line where its slope (the marginal

strain-rate) becomes zero. The horizontal position of this line, therefore, represents the

time at which rock in the shear zone margins acquired its final state of (low) strain. The

left-hand line represents the time at which rock in the shear zone interior passed through

this same state of (low) strain (From Means, 1995).

50

3.4 Fluid Movement in Shear Zones

Fault zones can act both as barriers to fluid flow and as conduits along or

across which fluid flows ( Anderson et al., 1991, Anderson et al., 1994, Hooper,

1991, Knipe, 1993, Berg and Avery, 1995, Lopez and Smith, 1995, Heneberg,

1995, Sibson, 1996, Roberts et al., 1996, Yielding et al., 1997, Knipe, 1997,).

Faults and their movements play a key-controlling element in fluid flow systems

in geological settings. The direction of fluid flow through a rock mass is

governed by the maximum hydraulic gradient (not necessarily vertical), existing

permeability anisotropy (e.g. bedding, foliation), and superimposed structural

permeability (Sibson, 1996). Furthermore, the permeability of a fault zone

depends on additional factors such as juxtaposition of permeable units across

the fault, rock properties, stress field, the amount of sand, clay, or fault gouge in

the fault zone, the presence of material with high capillary entry pressure, and

the presence or absence of an open fracture network (Smith, 1980, Allan, 1989,

Miller, 1995, Sibson, 1996, Knipe, 1997). Philips (1991) demonstrated that

fractured rocks can have much larger permeabilities, and hence hydraulic

conductivity, than that of completely homogenous rocks. As a result,

permeability may vary by many orders of magnitudes, depending upon rock

type and the pressure-temperature conditions (Braun et al., 2003).

Shear zones have long been shown as preferential fluid pathways (Beach and

Fyfe, 1972). Shear zones may act both as closed or open geochemical

systems with respect to movement of fluids and elements, irrespective of the

size of the zone (Hatcher, 1995). Shear zones are known to perform as

51

preferred pathway for fluid circulation (Beach and Fyfe, 1972). Little is

understood about the 3D geometry (the relationships between length, width and

depth) of shear zones and in particular the depth of their root zones (Pili et al.,

1997).

For an understanding of the dynamics and rates of fluid motion around and

within faults, it is necessary to understand the behavior of hydrothermal

systems. Braun et al. (2003) presented the result of simple numerical

experiments of fluid flow around and within a permeable fault embedded in a

less permeable porous medium. They showed that fluid flow is controlled by

two timescales: and , where S is the specific storage

of the porous material, l the length of the fault, and and are the

hydraulic conductivities of the porous material and the fault, respectively.

Physically, the three conditions correspond to:

Ff KSl /2=τ MF KSl /2=τ

MK FK

If ( fττ < ), fluid rapidly travels through the fault driven by the initial vertical

pressure gradient (or hydraulic head);

If ( Ff τττ << ), the fluid flow in the fault leads to pressure changes in the rock at

both extremities of the fault that result in a decrease of the pressure gradient

between the extremities of the fault;

If ( Fττ < ), fluid flow in the matrix has adapted itself to the new permeability

structure and a steady-state pressure field has developed around the fault.

The first time constant fτ , is the time required by the fluid to travel inside the

fault. fτ is therefore a function of l and . The second time constant, FK Fτ , is

52

the time required by the fluid to reach a steady-state flow and associated

pressure gradient in the matrix in the vicinity of the fault. Fτ is therefore a

function of l and . For a 1 km-long fault, MK fτ varies between 30 years and

300 kyr, depending on the value of the hydraulic conductivity, whereas Fτ varies

between 3Myr and 30 Gyr. Braun et al. (2003) showed that the length of a fault

is the most important factor which determines the value of the two time

constants, fτ and Fτ .

It has been useful to find ‘steady-state flow’ in rocks, flow at constant stress,

constant temperature, constant strain rate and constant microstructure, so that

the flow law can be expressed in a strain-independent form that allows

separation of variables, thus (Rutter, 1999):

( ) ( ) ( )ST 321 ∫∫∫Α= σε

In which ε is strain rate, σ is flow stress, T is absolute temperature and is

some measure of rate controlling microstructure, which might be grain size. It is

widely recognized that fine-grained rocks can be weaker than their coarser-

grained counterparts at high temperature (Walker, et al., 1990), if grain size

sensitive flow processes were to be activated. As Rutter (1999) discussed,

rocks in shear zones are commonly tectonically reduced in grain size, an

observation which tends to reinforce the inference that shear zones are zones

of relative mechanical weakness.

S

53

3.5 Evidence of Shear Zones

Shear zones have certain characteristic to recognize them in the field, in thin

section, on geological maps and cross sections (Davis and Reynolds, 1996),

and geophysical methods. Geophysical methods are very useful when there is

no clear evidence of shear zones in the field. Shear zones exist at all scales

and their size depends on the conditions, such as brittle, ductile, or intermediate

conditions, under which they were formed.

3.5.1 Field Evidence

The following evidence for detecting shear zones (Ramsay and Auber, 1983;

Simpson and Schmid, 1983; Passchier and Simpson, 1986; Davis and

Reynolds, 1996) might be established in outcrops, in hand specimens or in thin

sections.

1 S-C band structure (Figure 3-5A): S-C fabrics consist of foliation (S-

surfaces) and shear bands (C-surfaces). The S and C surfaces are at an angle

to one another, indicating the sense of shear. In most S-C fabrics, the S-

surfaces are a typical foliation, being defined by flattened lenticular grains and

aggregates of grains. The oblique relationship of the two structures shows a

clear indication of the displacement sense. S-C fabrics can be divided into two

general types (Lister and Snoke, 1984). Type I are the dominant S-C fabric in

most mylonitic quartzofeldspathic rocks. Type II S-C fabrics are most common

in micaceous quartzite and some mylonites. In both types of S-C fabric, the C-

54

surfaces are commonly aligned parallel to the shear zone and are associated

with the same strain gradients as any continuous shear zone.

2 Rotation of porphyroblasts or porphyroclasts (Figure 3-5B): If the

large grains instead are new metamorphic grains that grew during or after

deformation, they are called porphyroblasts. Grains interpreted to be relics from

the protolith are called porphyroclasts because they commonly represent

fragments or clasts of original phenocrysts or detrital grains. In many

metamorphic rocks it is well known that large porphyroblasts can develop;

albite, garnet, staurolite are particularly useful for the determination of shear

sense. These crystals can roll like ball bearings and by matching the inclusion

trails they contain inside the crystal ( ) with those found outside ( ), the

rotation sense can be determined. Such crystals can grow as porphyroblasts

during the shear motion when they produce a “snowball structure”.

is es

3 σ -Structure (Figure 3-5C): This structure is found around

porphyroclasts, and relates to the form of the dynamically recrystallized

pressure shadow tail in relation to the porphyroclast. In this structure the

median line of the tail does not cross the tend of the average shistosity, and the

structure appears to form because the rotation rate of the tail is higher than that

of the porphyroclast. This structure is characteristic of shear zones with low

shear strains where the recrystallization rates are higher than the rotation rates.

4 δ - Structure (Figure 3-5D): This structure, like σ -structure, is

developed around porphyroclasts, but where the recrystallized pressure shadow

55

tail is strongly rotated by the porphyroclasts. This structure shows the

extremely curved and often embayed nature of the generally narrow tail, and

where the median line crosses the general trend of the schistosity. This type of

structure is particularly characteristic of origins of high shear strains where the

recrystallization rates are lower than the rotation rates.

5 Bookshelf sliding (Figure 3-5E): When broken crystals are subjected to

shear, the individual parts are rotated in the direction of shear, a feature which

sets up a contrary shear motion between the fragments much like a collapsing

set of books. Further shearing motion can lead to separation of the individual

fragments and the development of recrystallized pressure shadow zones

between the fragments.

6 Displaced crystals (Figure 3-5F): Where crystals possessing a well

developed crystal cleavage lie with this cleavage close to the shear plane, they

often become internally detached by gliding on this surface. In highly deformed

mylonites, for example, it is not uncommon to find individual mica fragments

connected by zonal films of phyllosilicates to form free “floating mica fish”.

7 Dynamic recrystallization (Figure 3-5G): Rocks that have been very

highly deformed by crystal plastic processes frequently show an intense

banding and schistosity. Within the more deformed crystal (especially with

quartz and calcite) the bands are crossed obliquely by small grains with a

marked individual alignment.

56

8 Preferred optical orientation (Figure 3-5H): Plastic flow of crystals

leads to a variety of different types of preferred alignment patterns of the optical

axes of the crystals in the rock aggregate. Many of these complications of

interpretation of preferred orientation patterns result from the nature of the

environmental conditions during deformation.

9 Protomylonite, mylonite, and ultramylonite: Mylonites are strongly

foliated metamorphic rocks that exhibit high ductile strain and incomplete

recrystallization or recovery. Mylonite contains 50 to 90 per cent matrix.

Mylonite is typical shear zones rock. The initial stages of mylonitization produce

a weakly to moderately mylonitic rock, called a protomylonite. Protomylonite

contains less than 50 per cent fine-grained matrix. Ultramylonite contains more

than 90 per cent matrix and less than 10 per cent relict grains. Some foliation in

mylonitic rocks is the expression of thin shear zones.

57

Figure 3 - 5: Different types of criteria used for the determination of shear sense in shear

zones (From Ramsay and Auber, 1983).

58

3.5.2 Imaging Shear Zones with Geophysical Methods

Despite their significance, shear zones can be difficult to map in detail using

conventional geological methods. However, their location can often be inferred

from topographic features or geophysical methods, such as electromagnetic

techniques. Shear zones can be difficult to detect from the surface, and

therefore the use of surface mapping to define them is often unreliable. Without

the use of geophysical methods, shear zones are frequently inferred from

geological mapping, often with little evidence of their existence either directly or

tectonically. It should be emphasized that detecting and studying shear zones

are not always a straightforward procedure.

Major faults and shear zones are routinely discovered by explorationists through

geophysical methods. Nowadays, integrated use of least two geophysical

methods with geological interpretation is important to obtain accurate geological

interpretation. For example Street and Engel (1990) showed that integrated use

of geophysical methods (magnetic and seismic refraction), drilling, and sample

analysis has hound geological controls on the location of dryland salinity in

Western Australia. The success of the electromagnetic and resistivity

geophysical methods in imaging shear zones lies in their ability to image these

narrow zones as regions of high conductivity. This increase in conductivity is

due to the development of secondary porosity and permeability from fracturing

which leads to an increase in groundwater saturation. The success of the

seismic refraction method lies in its ability to image the shear zone as a narrow

region of low seismic velocity and increased depths of weathering.

59

In the Spicers Creek Catchment magnetic, electrical resistivity and seismic

refraction methods were used to image the inferred shear zones. It was

anticipated that in this area, seismic refraction would provide the best results

since the propagation of seismic waves would not be affected by the high level

salinisation whilst also providing possibly greater resolution.

60

Chapter 4

Three-Dimension Three Component

Seismic Surveys

4.1 Summary

The objective of the seismic refraction method is to determine the velocity

distribution in the sub-surface. The seismic refraction method does have a

number of advantages over the reflection method, particularly in shallow

investigations such as in geotechnical, environmental and groundwater

applications. The seismic refraction methods seek to measure the spatial

variation of petrophysical parameters by using seismic velocity.

Many geological features in the subsurface are three-dimension (3-D) in nature,

and two-dimension (2-D) seismic section is a cross section of the 3-D seismic

wave field. Having a detailed image of the subsurface usually makes the

interpretation more reliable. The aim of using 3D-3C seismic data is to find

more information about rock properties with both P- and S-waves. The most

important parameters which can be extracted from P-wave and S-wave data are

the P-wave/S-wave velocity (Vp/Vs) ratio and Poisson’s ratio.

61

4.2 The Seismic Refraction Method

The foundation of the seismic methods is the theory of elasticity. The velocities

of the propagation of the compressional and shear seismic energy are functions

of the elastic moduli or constants, and the rock densities.

The seismic refraction method is based on the principle that when a seismic

wave (P or S wave) impinges upon a boundary across which there is a contrast

in velocity, then the direction of travel of that wave changes on entry into the

new medium. The amount of change of direction is governed by the contrast in

seismic velocity across the boundary according to Snell’s Law as shown in

Figure 4-1.

Figure 4 - 1: Snell’s Law.

When a seismic signal encounters a surface separating two media having

different elastic properties, it gives rise to reflected or refracted waves. While

most refracted events have not been reflected, most reflected events have been

refracted, because refraction occurs across any velocity interface in accordance

with the Snell’s Law.

62

4.3 Seismic Velocities in the Earth

The seismic velocity is generally used as a measure of earth elasticity, since P

and S seismic velocities can be derived from the elastic constants and the

density. Due to the variability of the near surface conditions, the seismic

velocities in shallow earth materials are also highly variable. The relevant

factors include major rock forming minerals, rock formation, the nature of the

cementing material, fracturing and weathering, structural deformation, etc. In

general, seismic velocities are less in unconsolidated materials.

Furthermore, the seismic velocities of earth materials usually increase with

depth, because of decreasing weathering, and increasing water saturation,

compaction and diagenesis with depth. The velocities of seismic waves in most

igneous and metamorphic rocks depend mainly on the elastic properties of the

minerals making up the rock material itself, since these rocks have little or no

porosity. In general, igneous rocks have seismic velocities which show a much

narrower range of variation than sedimentary or metamorphic rocks. The

average velocity for igneous rocks is higher than that for other types (Dobrin

and Savit, 1988).

Because the seismic velocities generally increase with depth, seismic energy is

systematically refracted away from the vertical, until it travels in a predominantly

horizontal direction. Furthermore, the horizontally traveling energy also radiates

energy back into the overlying layers. As a result, seismic energy which is

initiated at the surfaces propagates down into the subsurface and returns to the

63

surface at some greater distance. Figure 4-2 schematically shows the basics of

the seismic refraction method.

Figure 4 - 2: Seismic refraction wave raypaths.

It might be noted that the fundamental requirement for the seismic refraction

method to be a reality is the systematic increase of seismic velocities with

depth. By contrast, there is a systematic decrease in electromagnetic velocities

with depth and as a result, refraction methods are not possible with ground

penetrating radar methods.

The seismic refraction method has a number of advantages over the reflection

method, particularly in shallow investigations such as geotechnical,

environmental and groundwater applications. The seismic refraction method is

most effective in the near-surface for depths of less than several tens of metres.

64

On land, it can be difficult to obtain good seismic reflection results from the

near-surface. By contrast, the refraction method has a unique ability to provide

detailed velocity information on the deepest refractor enabling it to detect the

location of weathered and fractured zones in bedrock.

As a result refraction methods can obtain good results with shallow targets,

whereas reflection methods are more suited to deep targets. Refraction is now

rarely employed in oil exploration because of the large scale field operation

required (Dobrin and Savit, 1988). It should be noted that the processing

methods for shallow targets with refraction methods, such as the GRM, are

suitable and probably more reliable than reflection methods when there is are

lateral changes in seismic velocities of the refractor. Table 4-1 shows some

application of seismic refraction methods.

4.4 Three Dimension Seismic Refraction Methods

Perhaps the first three-dimensional 3D seismic methods were the fan shooting

refraction methods of the 1920s and 1930s. These 3D methods, which clearly

pre-date the 3D reflection methods of the last quarter of a century, demonstrate

the fundamental maxim that 3D targets require 3D seismic methods for

efficacious exploration. The fan shooting techniques later evolved into the

broadside shooting methods of the 1950s (Richards, 1959).

65

Table 4 - 1: Some applications of refraction methods.

Seismic Refraction Applications

Authors

Exploration of alluvial deposits and sediments

(Edge and Laby, 1931, Urquhart, 1956, pakiser and Black, 1957, Hobson and Grant, 1964, Paterson, 1965, Duguid, 1968, Young and Lucas, 1988, Biehler et al., 1991, Campos-Enriquez et al., 1997)

Groundwater studies, hydrology and hydrogeological investigations

(Bonini and Hickok, 1958, Warrick and Winslow, 1960, Levshin, 1961, Wiebenga and Jasson, 1962, Lennox and Carlson, 1967a, Lennox and Carlson, 1967b, Hasselstrom, 1969, Overmeeren, 1981, Haeni, 1986, Dobrin and Savit, 1988)

Investigation of engineering sites and geotechnical factors

(Shepard, 1939, Wood, 1940, Moore, 1952, Brown and Robertshaw, 1957, Welin, 1958, Hawkins, 1961, Bartlett, 1962, Drake, 1962, Stam, 1962, Svenson and Bowering, 1963, Bigelow, 1965, Irving, 1965, Lawson et al., 1965, Mallot, 1965, Atkinson, 1970, Cummings, 1979, Hatherly and Neville, 1986, Hatherly, 1986, Kilthy et al., 1986, Laa et al., 1991)

Rock fabric and anisotropy (Bamford and Nunn, 1979, Palmer, 2000c, Palmer, 2001a, 2001b, 2001d,2003b)

66

The development of 3D refraction methods in recent times has been restricted

to deeper crustal structures, and they are as much a result of the realities of

locating receivers in accessible locations, as with the acceptance of the 3D

geometry of the subsurface targets (See page 244 Palmer 1986).

The development of 3D refraction to shallow targets has not been widespread.

Bamford and Nunn (1979) used radial profiles to determine azimuthal

anisotropy. However, no attempt was made to resolve the dual challenge of

irregular refractor parameters and azimuthal anisotropy. Dean et al (2000)

applied refraction tomography to a set of data recorded across an alluvial

channel.

Palmer (2001a and 2003a) carried out a 3D refraction survey across a large

shear zone near the massive sulphide deposit at Mt Bulga, near Orange in

southeastern Australia. The survey showed that the 3D results can display a

significantly more complicated geological picture than the standard 2D depth

cross section. Not only did the 3D results reveal significant cross cutting

features (faults?), but the azimuthal anisotropy in one section of the survey was

orthogonal to that of the major structural elements. The significantly more

detailed structural model revealed by the 3D results would have considerable

value in modelling groundwater or contaminant fluid flow, designing grouting

programs, or even designing simple engineering constructions.

67

4.5 Three Component Seismic Refraction Surveys

The aim of using three-component coordinate system is to find more information

about rock properties with both P- and S-waves. Principally, propagation

velocity of P-waves is affected by both rock incompressibility and rigidity,

whereas, S-waves is affected by rock rigidity only. P-wave velocity is a function

of three separate rock properties, the bulk modulus, shear modulus and bulk

density, whilst S-wave velocity depends upon the shear modulus and bulk

density (Domenico and Donbom, 1986). Used separately these velocities can

be ambiguous indicators of rock lithology.

The motion associated with a wave arrival is a vector quantity, and measuring

its three orthogonal components requires the use of three orthogonal

geophones. Most three-component phones use one vertical phone (P-wave)

and two horizontal phones (SV- and SH-wave). A field-recording technique

sometimes adopted in fracture studies, which successively record from

compressional wave (P source) and shear waves (SH-, and SV- sources) with

three-component geophones; this is called nine-component recording because

separate sections can be generated for each of the three-geophone outputs for

each of the three sources (Sheriff and Geldart, 1995). If the sources are pure

and the earth isotropic, three of the nine sections would be conventional P-, SH-

, and SV-sections, two would show converted waves (P to SV and SV to P), and

the other four would be blank (Figure 4-3).

68

The coordinate system for measuring the three components of a P- or S-waves

can be rotated around any axis; in particular, they can be rotated into a natural

coordinate system. Such a coordinate rotation is sometimes called Alford

rotation or polarization filtering (Sheriff and Geldart, 1995).

Three-component recording can be employed to resolve fracture problems in a

3-D survey (Lewis et al., 1991). On the other hand, if the SH- and SV- sources

and geophones are not oriented parallel and perpendicular to the natural

orientation or azimuthal anisotropy is present, then shear wave splitting will

result in energy showing on all panels.

Conventional P-wave data

Conventional SH-wave data

Conventional SV-wave data

Mode conversion P to SV

Mode conversion P to SV

?

?

?

?

SV

SH

P

P SH SV

SOURCE MOTION

Figure 4 - 3: Nine-component sections recorded by P-, SH-, and SV-geophones for P-, SH-

, and SV sources. Without anisotropy the sections marked ? would be blank (From

Tatham and McCormack, 1991).

69

4.6 Shear Waves Versus P waves

P- and S-waves are considered as body waves because these two types can

propagate through the body of an elastic solid. P-waves are also known as,

longitudinal, primary, push, dilatational, irrotational or compressional waves.

Material particles oscillate about fixed points in the direction of wave

propagation by compressional and dilatational strain, similar to a sound wave.

S-waves are also known as the transverse, secondary, rotational, or shear

waves. The seismic velocity in rocks depends on many factors, including

lithology, porosity, depth, cementation, age, pressure regim, interstitial fluids,

etc. The velocities Vp ( )α and Vs ( )β of these two waves are related to the

elastic constants and density of the medium by the equations:

2/12/12

⎟⎟⎠

⎞⎜⎜⎝

⎛=⎟⎟

⎠

⎞⎜⎜⎝

⎛ +=

ρρµλα M

2/1

⎟⎟⎠

⎞⎜⎜⎝

⎛=

ρµβ

where M is the P-wave modulus. Because the elastic constant are always

positive, α is always greater than β . The ratio Vp/Vs is defined in terms of

Poisson’s ratio and is given by:

2/1

5.01

⎟⎠⎞

⎜⎝⎛

−−

=σσ

VsVp

70

As σ decreases from 0.5 to 0, Vs/Vp increases from 0 to its maximum value,

2/1 ; thus, the velocity of the S-wave ranges from 0 to 70% of the velocity of

the P-wave (Sheriff and Geldart, 1995). Meanwhile, for fluids, µ is zero and

hence β is zero also; and as a result S-waves do not propagate in fluids.

The Poission’s ratio is one of the most important parameters which can be

extracted from P-wave and S-wave data. This parameter is independent of

density, it is a much more diagnostic lithological and pore-fluid indicators and it

can also detect the degree of rock fracturing and seismic anisotropy (Dasios et

al, 1999). In terms of rock mechanics, Poisson’s ratio is equal to the ratio of

transverse and longitudinal strain and hence can provide information on

material elasticity. The nature of interstitial fluids can be determined using the

relationships between Vp/Vs ratio and Poisson’s ratio with pore pressure and

fluid type. For example, Vp/Vs ratio is sensitive to the pore fluids found in

sedimentary rocks, with the value being much lower for gas saturation than for

liquid saturation (Tatham, 1982). Furthermore, laboratory studies have shown

that at high differential pressures, Poisson’s ratio is constant, whilst at low

differential pressures, the ratio increases for saturated rocks and decreases for

dry rocks (Carcione and Cavallini, 2002). The Vp/Vs ratio is affected by changes

in porosity. For example, a relationship between the velocity ratio and the

shape of the pore space has been shown (Toksöz et al, 1976) with flat pore

spaces resulting from fracture porosity giving a different response from rounded

pore spaces resulting from primary or secondary porosity. The anisotropy of

strata can be detected by changes in Vp/Vs ratio in different azimuthal

71

directions. This is useful in fracture detection, where velocities can be affected

differently depending on the raypath orientation with respect to these fractures.

As the velocities of S-waves are slower than that of P-waves, for S- and P-

waves with comparable frequencies and hence shorter and longer wavelengths

respectively, it might be expected that the resolution of S-wave records would

be higher than that of P-wave records. However this resolution is often not

realisable, due to the difficulty in generating high frequency S-waves, their high

attenuation, and noise from surface and compressional waves. Due to the

effect of source to ground coupling, it is more difficult to generate S-waves than

P-waves of high frequency (Helbig, 1986). In general, the quality factor of S-

waves is lower than that of P-waves, limiting their useful length of travel to a few

hundred times its wavelength. S-wave records are contaminated by a high

degree of noise with lingering reverberations from P-waves often occurring at

the same time as S-wave arrivals (Pautet et al., 2001). Also the marginally

slower velocities of surface waves mean that their separation from S-waves can

be difficult (Helbig, 1986).

The enhancement of S-waves is possible if horizontal component seismic

records are generated using sources of opposite polarity (Xia et al, 2002). S-

wave energy can be enhanced by subtracting seismic records of opposite

polarity, resulting in the cancellation of P-wave components whilst leaving the

sum of the two S-wave components. Opposite polarity records may show

amplitude and phase differences due to changes in near-surface conditions. As

a result the subtraction procedure would not lead to S-wave enhancement

72

(Dasios et al, 1999). Appropriate amplitude and phase corrections must be

applied before these enhancement operations are performed.

73

Chapter 5

Geology and Geophysics

5.1 Summary

The Dubbo 1:250,000 geology map includes sedimentary, igneous and

metamorphic rocks of Palaeozoic to Cainozoic age. This area includes the

exposed north-eastern portion of the Palaeozoic Lachlan Orogen and overlying

marginal areas of the Permo-Triassic Sydney-Gunnedah Basin and Jurasic

Surat Basin (Meakin and Morgan, 1999).

The study area (Spicers Creek Catchment) is situated on the southern margin

of the Gunnedah and Surat Basin both of which overly the western margin of

Lachlan Fold Belt. The oldest rocks in the study area are Palaeozoic in age.

The main Palaeozoic geological units are the Ordovician Okdale Formation and

the Silurian Gleneski Formation.

The study area is geologically complex due to the tectonic events it has

endured and the landscape evolution is closely related to the underlying

geology. The major fault in area is Nindethana Fault which is a major 50 km

long thrust fault that extends from the northern part of the Bathurst 1:250,000

map sheet area to the south-western corner of outcrop of the Wuuluman

74

Granite. The Nindethana Fault, for most of its length, is localized within the

Early Devonian Cunningham Formation.

5.2 Introduction

The Dubbo 1:250,000 geology map includes sedimentary, igneous and

metamorphic rocks of Palaeozoic to Cainozoic age. This area includes the

exposed north-eastern portion of the Palaeozoic Lachlan Orogen and overlying

marginal areas of the Permo-Triassic Sydney-Gunnedah Basin and Jurasic

Surat Basin (Meakin and Morgan, 1999).

In this chapter Regional and local geology and geophysics of study area will be

discussed.

5.3 Tectonic Setting

The study area is situated in the north-eastern Lachlan Fold Belt (LFB) and

overlies marginal areas of the Gunnedah Basin and Surat Basin. Figure 5-1

shows the tectonic units within the LFB area.

75

Figure 5 - 1: Tectonic units within the Lachlan Fold Belt area (Modified from: Kingham, 1998).

76

5.3.1 Lachlan Fold Belt

The Lachlan Fold Belt (LFB) is part of the eastern margin of Gondowanaland

and a major structural unit located on the eastern side of the Tasman Fold Belt

System. The LFB extends from eastern Tasmania through Victoria into New

South Wales (NSW), where it forms a large part of the state. In NSW the LFB is

bounded to the east by the younger New England Fold Belt, to the north by the

Thompson Fold Belt and to the west by the Kanmantoo Fold Belt.

The LFB is a complex orogenic belt, which according to most workers

developed from the Cambrian to the Early Carboniferous. The belt has been

subdivided into several stratotectonic units (Figure 5-2) termed anticlinorial and

synclinorial zones that reflect the geological development across the LFB

(Suppel and Scheibner, 1990).

Scheibner and Basden (1996) described the LFB as a ‘composite orogenic belt

composed of Early Cambrian to Early Devonian pre-cratonic complexes

developed over mixed ensialic (microcontinents) and ensiamatic (oceanic or

ophiolitic) basement’. Study of the stratigraphy shows that there have been four

main phases of folding as identified by stratigraphic relationships. It is probable

that tectonic activity was continuous for over 30 Ma from its commencement at

about the end of Ordovician until the Early Devonian, with significant phases in

the latest Ordovician to Early Silurian (Benambian) and Late Silurian to Early

Devonian (Bowning), followed by more discrete episodes in late Middle

77

Devonian (Tabberabberan) and Early Carboniferous time (Branagan and

Packham (2000)).

Figure 5 - 2: Structural map of the Lachlan Fold Belt in New South Wales (Suppel and Scheibner, 1990).

78

In the LFB, Cambrian greenstones are the oldest exposed rocks, which occur

as faulted-bounded slices throughout the western belt (Collins and Vernon,

1994). The Ordovician sequences are extensively exposed. Two distinct

groups occur within the Ordovician – a quartz–rich greywacke–black shale

facies and a volcanic facies (Branagan and Packham, 2000). A turbidite

current, deep-water origin has long been postulated for the Ordovician flysch.

Volcanic and volcaniclastic rocks of mainly shoshonitic composition occur in

four elongate belts in the eastern part of the Lachlan orogen (Glen et al., 1998).

Ordovician sediments in LFB were deposited across the continental margin that

stretched over 1700 Km situated about 200 Km east from the Australia craton

toward an west- dipping subduction zone (Collins and Vernon, 1994).

In the LFB, the beginning of deformation commenced during the Silurian. Early

Silurian sediments, postdating the first phase of the Benambran Orogeny, occur

in a small number of localities in the southern highlands. Deposition of

limestone, shales, dacitic volcanics and volcanic sandstones and their

deepwater equivalents, more typical of Silurian successions, commenced in the

Silurian. In the Early Silurian, several depositional troughs developed by crustal

extension. The Cowra Trough, extends from Dubbo almost to Young. It formed

between the Molong Rise and the Parkes Platform (Branagan and Packham,

2000). Deep and shallow sedimentation and deformation was associated with

widespread voluminous silicic magmatism and low-P and, locally, high-T

metamorphism during the Silurian to Early Devonian (Collins and Vernon,

1994).

79

Most of the LFB was affected by the Late Ordovician Early Silurian Benambran

Orogeny which caused the inversion of a back arc basin resulting in a collision

thrust belt called the Wagga-Omeo Zone (Scheibner and Basden, 1996).

Extension in the Early Silurian eastern LFB was followed by the localized

Bowning-Bindi deformation in the Late Silurian-Early Devonian.

Through the Silurian and Devonian the Ordovician continental margin was

deformed, resulting in a 60% shortening of the crust and creating the 750 Km

wide LFB. The molasses deposition of extensive Late Devonian quartzose

sediments indicates the evolved, continental character of the LFB during this

time. The sediments show east-directed palaeoflow thereby implying a source

from the western LFB. Early Mid Carboniferous (Kanimblan event) deformation

was widespread, but only produced regional, gentle to open, folds and

conjugate strike-slip faults except in the eastern LFB, adjacent to the active

magmatic arc (Collins and Vernon, 1994).

In the LFB, emplacement of extensive orogenic granites occurred in the Silurian

and Devonian period (Chappell et al., 1988). This represented a major

transition in the LFB towards a major continental margin orogen and is linked

with complex sedimentary associations and explosive volcanism that occurred

on a extraordinary scale (Coney, 1992). Post orogenic Carboniferous granites

completed the development of the LFB (Kingham, 1998).

As Scheibner and Basden (1996) described, the late Middle Devonian

Tabberabberan orogeny terminated pre-cratonic development and was followed

80

by a diachronous onset of transitional tectonic volcanism and sedimentation.

The Kanimblan orogeny converted the LFB into a neocraton, with subsequent

late Carboniferous – Holocene platformal sedimentation.

In the LFB, the folding increases in intensity towards the east. Cleavage is

commonly developed in the fine-grained sediments in the east. Post-

deformation Early Carboniferous granites intruded the north-east part of the fold

belt. The largest ones are the Bathurst and Gulgong Granites; others include

the Wuuluman Granite near Wellington and small bodies south of Oberon

(Branagan and Packham, 2000).

Various models have been suggested to describe the evolution of the LFB.

Scheibner (1973) described a model involving cyclical extension and

compression related to rotation of the Australian Plate and the docking of

microcontinents and volcanic arcs. Rutland (1976) suggested a model based

on a persistent volcanic arc that remained fixed along the eastern fringe of the

Australian craton, with infilling of the foreland basin area by back arc rifting and

associated volcanism. Crook (1980) thought that the LFB developed through

accretion of up to ten fore-arc complexes that were continuously accreted to the

edge of the craton from the east. Powell (1984) illustrated two phases of post-

Cambrian subduction, the first in the Ordovician with an actively spreading

marginal sea behind the arc. This was followed by Silurian to Middle Devonian

cratonisation, dextral strike-slip faulting and silicic magmatism. The second

subduction phase commenced in the Late Devonian, along the edge of the

craton in a Chilean-style convergence producing the New England Fold Belt.

81

Collins and Vernon (1994) considered that the LFB developed as a foreland

basin between the Proterozoic craton and a westerly-dipping subduction zone

along the present day east Australian margin (Figure 5-3). Detachment of the

lithosphere from the asthenosphere started beside the western edge of the LFB

about 430 to 400 Ma and progressively moved east.

Following the delamination in the eastern LFB, it began in the western LFB, with

delamination again propagating from west to east. This phase of delamination

continued until the latest Devonian (~350 Ma). In the eastern and western LFB,

delamination of the lithosphere resulted in underplating by hot basic magma of

the asthenosphere causing higher heat flow, crustal softening, deformation,

isostatic uplift and crustal extension. Partial melting caused the emplacement

of plutons and granitic melts as well. Eastward propagation of the delamination

across both the eastern and western LFB, is indicated by the eastward

translation of crustal deformation and a general easterly younging of granitic

plutons.

5.3.2 Gunnedah Basin

The Gunnedah Basin is located between the Bowen Basin in the north and the

Sydney Basin in the south (Tadros, 1993). The Gunnedah Basin, as it is

defined in its present form, is a structural sub-basin and it developed as part of

the extensive and composite Sydney – Bowen Basin. It is about 350 km long

and up to 200 km wide, and covers an area of 50,000 km (Tadros, 1993). The 2

82

83

Figure 5 - 3: Tectonic model for the eastern Lachlan Fold Belt (Modified from Collins and Vernon (1994)).

basin is a structural trough with an elongated appearance, trending north-

northwest in northern New South Wales. It consists of three sub-basins

separated by two ridges. From west to east these have been referred to as the

Gilgandra Trough, Rocky Glen Ridge, West Gunnedah sub-basin, Boggabri

ridge and Maules Creek sub Basin (Korsch et al., 1993). The major structural

elements of the Gunnedah Basin are recognizable basement structures which

have a deep-seated origin in the upper crust and have influenced the tectonic

development of the basin throughout its history (Tadros, 1993). Figure 5-4

shows the Sydney, Gunnedah and Bowen Basin in New South Wales.

The Gunnedah Basin contains up to 1200 m sequence of marine and non-

marine Permian and Teriassic sediments and is an important Permian coal-

bearing basin (Tadros, 1993). Metavolcanics and metasediments and minor

ignimbritic volcanics of the Lachlan Fold Belt form much of the basement

underneath the western part of Gunnedah Basin. Basin extension and

subsequent subsidence initiated deposition in the Gunnedah Basin during the

early Permian, commencing with the fluvial Maules Creek Formation containing

numerous coal seams. The Late Permian marine shelf sediments of the

Porcupine Formation and the regressive Marine Watermark Formation

conformably overlie Maules Creek Formation. A second coal sequence, the

Black Jack Formation, was deposited in a deltaic and fluvial environment in the

Late Permian (Tadros, 1993).

84

Figure 5 - 4: The Sydney, Gunnedah and Bowen Basin in NSW (Tadros, 1993).

85

Triassic sediments unconformably overlie the Permian sequences. Two

depositional sequences have been identified in the Digby Formation. The lower

stage consists of a conglomerate unit deposited by a braided steam system

after uplift of the adjacent New England Fold Belt. During the late stage sandy

sediments were deposited as the elevation of the New England source area

declined, and the system changed from a braided to a meander system. A

basin-wide palaeosol horizon caps the Digby Formation. The Napperby

depositional sequence represents the upper limit of the Gunnedah Basin

sequence, with a regional unconformity existing between the Triassic and

overlying Jurassic sediments of the Surat Basin north of Liverpool Ranges

(Tadros, 1993, Kingham, 1998).

The Gunnedah Basin sequence includes a number of basic intrusions of

Mesozoic and Tertiary rocks. These are associated with massive extrusions of

the Garrawilla Volcanic complex and the Liverpool, Warrumbungle and

Nandewar Ranges. Cretaceous sediments (Bungil Formation) occur under

Tertiary basalt in the western part of the Liverpool Range (Tadros, 1993). Figure

5-5 shows the lithostratigraphy of the Gunnedah Basin from(Tadros, 1993).

In addition to coal resources, the Gunnedah Basin also has potential for the

discovery of oil and gas, especially as good petroleum source rocks are

presented at several stratigraphic horizons. Sandstones with good to excellent

reservoir potential also occur at several stratigraphic levels and many are

sealed by regionally extensive clayey units (Tadros, 1993).

86

Figure 5 - 5: Lithostratigraphy of the Gunnedah Basin (Tadros, 1993).

87

5.3.3 Surat Basin

The Jurasic-Cretaceous Surat Basin is located on the eastern part of the Great

Australian Basin (Figure 5-1). The Great Australian Basin is a composite

structural basin that includes a relict group of major Jurassic-Cretaceous

epicratonic sedimentary basin occupying about 1.7 million square kilometers of

inland eastern Australia (Cramsie and Hawke, 1984). In New South Wales the

Surat Basin has been divided into six lower category units includes the Boomi

Trough, the Lightning Ridge Shelf, the Coonamble Embayment, the Oxley

Basin, the Rawdon Ridge and the Warialda Trough (Bourke et al., 1974; Hawke

and Cramsie, 1984; Scheibner and Basden, 1996). In NSW it is also separated

from the Eromanga Basin by the Nebine Ridge, which underlies the Cunnamulla

Shelf. To the east of this is the Lightning Ridge Shelf and the Boomi Trough, to

the south of these elements is the Coonamble Embayment.

The Surat Basin is asymmetric, being wedge-shaped in cross section with the

sedimentary section gradually thickening from the western margin to the axis of

the Jurassic-Cretaceous trough, the Mimosa Syncline, and thinning more

sharply farther east to Kumbarella Ridge (Power and Devine, 1970).

The Surat Basin includes 2500 metres of mainly Jurassic clastic continental

sedimentary rocks and lower Cretaceous marine beds largely obscured by

Cainozoic alluvium. During the Jurassic period, there were 5 cycles of

terrestrial deposition. Coarse sand braided stream deposits grading up into

finer sand and silt deposits of meandering streams represent these cycles.

88

Labile sand, silt, mud and coal were laid down in swamps, lakes, deltas and

shallow seas to complete the cycle of deposition. During the mid-late Jurassic,

much of the sediments were supplied by andesitic volcanic activity (Exon,

1976). Outcropping units laid down during the Jurassic period include the

Precipice Sandstone and Evergreen Formation, the Hutton Sandstone and

Walloon Coal Measures, the Springbok Sandstone and Westbourne Formation,

the Gubberamunda Sandstone and Orallo Formation, with the final terrestrial

cycle depositing the Mooga Sandstone and Bungil Formation (Kingham, 1998).

Hawke and Cramsie (1984) described the stratigraphy of the Surat Basin in

New South Wales, as divided into four main intervals. The lowest is an Early to

Middle Jurassic, predominantly fine-grained, terrestrial sequence on the eastern

side. This is overlain by fluvial sandstone sequences of Middle Jurasic to

Neocomian age. An overlying Neocomian to early Albian essentially fine-

grained transgressive/marine/regressive sequence is overlain in the deepest,

central part of the basin by Albian transgressive/regressive sequence.

The Early Cretaceous marine transgression ended the fluvial cycles and

initiated deposition of the Rolling Downs Group. This group includes the

Doncaster and Coreena members of the Wallumbilla Formation, the Surat

Siltstone and the Griman Creek Formation (Kingham, 1998). During the Late

Cretaceous and early Tertiary, there was a deep weathering profile. Basic

volcanic activity occurred during the Oligocene-Miocene and since this time the

Surat Basin has remained relatively stable, with extensive erosion occurring

89

around the margin of Murray-Darling Basin (Exon, 1976). Figure 5-6 shows the

Surat Basin stratigraphy.

The Surat Basin hosts a large number of oil and gas fields in the adjacent state

of Queensland. Most exploration effort in the Surat Basin has focused on the

northeastern, deeper areas of the basin where it overlies the Bowen Basin.

Figure 5 - 6: Surat Basin stratigraphy (From: www.minerals.unsw.gov.au/geosurvey/petrol).

90

5.4 Regional geology

5.4.1 Introduction

The study area is located on the Dubbo 1:250, 000 sheet area (Figure 5-7).

This sheet includes six 1:100,000 maps (Dubbo (8663), Wellington (8632),

Cobbora (8733), Euchareena (8732), Gulgong (8833) and Mudgee (8832)). It is

also part of central western New South Wales and includes sedimentary,

volcanic and plutonic rocks of Early Paleozoic to Cainozoic age. The Dubbo

1:250,000 map sheet area covers the exposed north-eastern margin of the

Palaeozoic Lachlan Orogen, which is fringed by the western margin of the

Sydney Basin, the southern margin of the contiguous Gunnedah Basin and the

southern margin of the Surat Basin.

Figure 5 - 7: Locality map for the Dubbo 1:250,000 geological map sheet area, showing the six component 1:100,000 map sheets and geographical features (Meakin and Morgan, 1999).

91

5.4.2 Climate

Figure 5-8 shows the mean monthly maximum and minimum temperatures,

mean monthly precipitation data and the average number of wet days for Dubbo

and Dunedoo, and pan evaporation data for Wellington (Schofield, 1998). In

the Dubbo region rainfall is greatest in January, with an annual average of 584

mm reported on the Dubbo 1:100,000 sheet and 642 mm on the Cobbora

1:100,000 sheet. The annual distribution of rainfall is tilted toward the summer

months, but June is also a wet month. The climate is temperate with annual

mean temperatures ranging from 19 to 32 ˚C in summer, and 3 to 19˚C in winter

(Schofield, 1998).

Figure 5 - 8: Mean monthly maximum and minimum temperatures, mean monthly precipitation data and the average number of wet days for Dubbo and Dunedoo, pan evaporation data for Wellington is shown at the base of Figure (Schofield, 1998).

92

5.4.3 Physiography

The Dubbo 1:250,000 map area covers 15,600 square kilometers over parts of

the Shires of Cabonne, Wellington, Dubbo, Coolah, Mudgee, Merriwa and

Rylstone. The Western railway and several major roads, including the Mitchell

Highway, serve the area (Meakin and Morgan, 1999).

Topographically, the area falls from 740 m ASL in the east, to around 240 m

ASL near Dubbo (Figure 5-9). The topography in the Dubbo region has been

substantially reduced to a peneplain by erosion. This process mixes low-lying

lava flow relics with isolated topographic highs comprising plugs and earlier but

topographically inverted lava flows.

In the area there are three main drainage systems: Macquarie, Cudgegong and

Talbragar Rivers. The largest is the Macquarie River. The Macquarie River

flows from the western boundary of the study area to the NW through Dubbo.

The Talbragar flows in the north-west direction, through Dunedoo via Ballimore

to the Macquarie River. The two rivers link together 5 km north of Dubbo.

Talbragar River from east to west has 11 tributaries: Sandy Creek, Boomley

Creek, Spring Creek, Baragonumble Creek, Goan Creek, Spicers Creek, Back

Creek, Rocky Creek, Mitchells Creek, Jones Creek and Troy Creek. The

Gudgegong River flows in a southerly direction from the east side of Gulgong

town.

93

The Great Dividing Range extends along the eastern edge of the map area.

Surface elevation gradually decreases from as much as 1068 m ASL in the

south-east (near Harbon) to about 300 m ASL in the north-west of the map area

(Meakin and Morgan, 1999).

Figure 5 - 9: Topography, township and surface drainage map for the Dubbo and Cobbora 1:100,000 sheets.

94

5.4.4 Dubbo Area

The Dubbo 1:250,000 map sheet area covers the exposed north-eastern margin

of Lachlan Orogen. Meakin and Morgan (1999) divided the Lachlan Orogen

rocks into seven lithotectonic associations based on age, lithology and

depositional environment. These include the:

Early Ordovician to Late Ordovician generally deepwater, mafic to intermediate

volcanic and volcaniclastic rocks and intrusions;

?Early to Middle Ordovician quartz-rich turbidities;

late Early Silurian to Early Devonian shallow water to deepwater volcanic and

sedimentary sequences;

Early Devonian plutonic rocks;

Middle Devonian subaerial silicic volcanic sequences;

Late Devonian quartzose fluviatile to shallow marine sedimentary rocks; and

late Early to Late Carboniferous granite intrusions.

Figure 5-10 shows the schematic structural map of the Dubbo 1:250,000 sheet

area (Matson, 1975). During the Early Middle Ordovician, tensional forces

probably associated with rollback of the subduction zone towards the east

caused separation of a substantial part of the Girilambone flysch wedge (most

probably of Cambrian age) from the Australian Pate. The Wagga Marginal

Basin formed between the microcontinent and the major plate. In the Late

Ordovician and Early Silurian, further extension caused relatively minor splitting

between the microcontinent and the Molong Volcanic Rise such that the

95

33˚

32˚

148˚30’ 150˚

Figure 5 - 10: Schematic structure map of the Dubbo1:250 000 sheet area (Matson, 1975).

96

western segment of the volcanic rise was left attached to the Parkes Terrace

that had developed over the microcontinent. The initiation of the Cowra Trough

probably represented the depression associated with this splitting (Matson,

1975)

At the end of Ordovician, following the extension, the Australian plate advanced

eastwards causing the Benambran Orogeny. Sedimentary material filling the

Wagga Marginal Basin was deformed and metamorphosed. During Silurian

times, after this Orogeny, tensional stress developed, probably caused by

rollback (Matson, 1975). The Qidongan Orogeny occurred in this period (Crook ,

1980).

The Oidongan Orogeny during the Silurian caused further folding and uplift of

the flysch and volcanic sequences to the east, and attached to, the Molong

Volcanic Rise. Following the Oidongan Orogeny, strong tensional stresses

caused splitting of the Molong Volcanic Rise and the formation of a new

marginal sea, the Hill End Trough, with other similar features in the Lachlan

Fold Zone. After this extension in the Middle Silurian, explosive acid volcanism

commenced. The acid volcanism is indicative of the commencement of the

Bowning Tectonic Stage (Matson, 1975). In the early Devonian, closure of the

Cowra Trough marked the end of sedimentation in this area. Flysch and

volcanic rocks was deposited in the Hill End Trough during the early Devonian.

Devonian sedimentation in the Hill End Trough reached 6100 m, much greater

than the thickness of the Silurian sequence. The Silurian to Middle Devonian

97

Rise was formed, due to formation of the Sofala Anticlinorium and the Sunny

Corner Synclinorium. The formation of the Forbes Anticlinorial Zone and the

Cowra – Yass Synclinorial Zone from the former Cowra Trough was completed

during the later Kanimblan Orogeny. Tabberabberan folding and uplift

cratonized the Hill End Trough and it is now represented by the Hill End

Synclinorial zone. Locally at Mount Frome, near Mudgee, Middle Devonian

formations and the Lambie Group are conformable, but elsewhere an

unconformity has been proven below the Lambie Group (Matson, 1975).

In the Early Carboniferous the Kanimblan Orogeny caused deformation of the

Late Devonian strata, producing the Catombal and Lambie Synclinoria, and the

Lachlan Fold Belt was cratonised. During the Middle to Late Carboniferous, the

post kinematic Bathurst granites were emplaced. Prior to this emplacement,

acid volcanism, which was probably related to the granite plutonism, occurred in

the Rylstone area.

In the Permian, plate interaction and extension resulted in the formation of the

Sydney – Bowen Basin as well as New England Mobile Zone. The Hunter

Orogeny deformed the New England Fold Belt during the late Early Permian.

Following this orogeny, the sediments of the Bowen and Hawkesbury Tectonic

Stages were deposited in the Sydney – Bowen Basin a fore-deep basin

(Matson, 1975).

In the early Triassic, extension followed by depression formed the Great

Australian Basin. In the Jurassic and Cretaceous the Tasman Epi-Cratonic

Province was established, and the main subsidence was transferred from the

98

Sydney Basin to the Great Australian Basin onlapping the previous fore-deep

sediments. Also, Triassic-Jurassic and Jurassic sediments onlapped the older

rocks to the west (Matson, 1975).

5.4.5 Geology

The Dubbo 1:250,000 geology map includes sedimentary, igneous and

metamorphic rocks of Palaeozoic to Cainozoic age (Figure 5-11). The oldest

rocks in the Dubbo area belong to the Macquarie Volcanic Arc and in particular,

to the early Ordovician Mitchell Formation, which is conformably overlain by the

deepwater Hensleigh Siltstone (Meakin and Morgan, 1999).

The Early Ordovician sequence contains volcanic and volcaniclastic rocks that

range in age from the Lancefieldian (or older) to the early Bendigonian.

Because of an hiatus the Early Ordovivian strata represent a separate volcanic

event from that of the overlying Middle Ordovician Cabonne Group.

During the late Middle to Late Ordovician igneous activity generated

intermediate to mafic, medium-K to shoshonitic volcanic and shallow rocks of

the Cabonne Group, with copper-gold mineralization. A quartz-rich fan system

(Adaminby Group) developed in the south-east part of the map area during this

time. Deformation of the Lachlan Orogen occurred in the Late Ordovician to

Early Silurian. Regional extension in the Early Silurian caused the Macquarie

Volcanic Arc to split and form separate belts (Meakin and Morgan, 1999).

99

Figure 5 - 11: Simplified geological framework of the Dubbo 1:250,000 map sheet area (Meakin and Morgan, 1999).

In late Early Silurian to late Silurian, an extensive section of sedimentary and

volcanic rocks covered the Dubbo 1:250,000 map sheet area. Sedimentation

was widespread in the Hill End Trough, during the Early Devonian in the Dubbo

area, and on the adjacent Mumbil shelf to the west and the Capertee High to

the east (Mawson and Talent, 1999). The Gregra Group was deposited over 100

the Mumbil Shelf and extends into the adjacent Cowra and Hill End Troughs.

The subaerial and dominantly acid volcanic rocks of the Hyandra Creek Group

were deposited over a small area to the south of Dubbo (Meakin and Morgan,

1999). During this time, the Yeoval Batholith, consisting of I-type intrusive

igneous rocks and intermediate to mafic sills was intruded. The truncation of

folds and faults is caused by a pre-Late Devonian unconformity (Tabberabberan

event) on the Dubbo 1:250,000 map sheet.

The Carboniferous period is represented by some plutonic igneous rocks, which

range from monzodiorite to granite in composition. The Carboniferous granites

are clearly discordant with the pervasive north-south structural trends and may

indicate an extensional regime (Meakin and Morgan, 1999). The granites of the

Dubbo 1:250,000 map sheet area are largely I-type, but some phases in the

eastern part of the Gulgong plutonic suite show marginal S-type characteristics,

with the presence of common muscovite and biotite (Barron, 1997).

The Permian period is represented by reactivation of older faults and by

deposition of the Early Permian, silicic Rylstone Volcanics which host

epithermal and intrusion-related skarn mineralization. In the ?late

Carboniferous to Triassic sequence around the eastern and northern margins of

the Dubbo 1:250,000 map, early Permian sediments represent the base of the

Sydney-Gunnedah Basin. The Sydney-Gunnedah Basin is an elongated

structural basin that extends between the Lachlan Orogen to the west and the

New England Orogen to the east. Conversion of the Sydney Basin to a foreland

101

basin led to development of extensive Late Permian coal swamps (Meakin and

Morgan, 1999).

In Mesozoic age, igneous rocks extended across the Dubbo 1:250,000 map and

are commonly related to the Permian to Triassic sedimentary rocks of the

Sydney and Gunnedah Basin (Meakin and Morgan, 1999). Googodery

Trachyte is the only named Mesozoic igneous unit in this area. During this

period, fluvio-lacustrine quartzose sediments (Purlawaugh Formation, Pilliga

Sandstone) of the Surat Basin were deposited along the northern margin of the

area. They represent deposition due to erosion which commenced in the Late

Jurassic and continued into the Tertiary.

In the Cainozoic, deposits on the Dubbo 1:250,000 map are represented by

alluvial, colluvial and lacustrine sediments, basalt and duricrust. No formal

names have been applied to unconsolidated Quarternary units on the Dubbo

1:250,000 map sheet. Gravel deposition represents climatic change by high-

energy, braided river systems, which gradually diminished in energy with the

development of aridity in inland Australia. Tertiary basalt overlies and intrudes

all pre-Cretaceous rocks across the Dubbo 1:250,000 map sheet area, with

relict basalt preserved predominantly over the Mumbil Shelf and Capertee High

(Dubbo area and Liverpool Range), and also over the southern end of the Hill

End Trough (Meakin and Morgan, 1999).

102

5.4.6 Metamorphism

Figure 5-12 shows the metamorphic zones on the Dubbo 1:250,000 map sheet

area. There are four main metamorphic zones in the area (Pogson and

Watkins, 1998). The following is the four summary metamorphic zones:

M1 (blue): clay/carbonate, epidote, sericite, albite, with or without foliation.

M2 (green): prehnite/pumpellyite, normally without foliation.

M3 (purple): actinolite/biotite, with or without foliation.

MH (red): akin to the hornfels envelope surrounding an intrusion. This zone is

represented by the presence of metamorphic diopside or garnet, or other

phases such as cordierite, andalusite, or even other higher grade hornfels

assemblages.

The metamorphic map has been prepared from an assessment of the

metamorphic mineral assemblages of 1919 rocks corrected by the Geological

Survey of New South Wales from this sheet. Because of small data set, this

map is regarded as a preliminary attempt to reassess the metamorphic zones

published by (Smith, 1969). His work identified and showed the distribution of

six zones of differing metamorphic grade over the Buthurst-Dubbo region,

based on the presence or absence of key metamorphic minerals such as

carbonate, albite, chlorite, prehnite, pumpellyite, actinolite-tremolite and biotite.

Metamorphism can be seen adjacent to the intrusion on the Dubbo map. All the

Carboniferous plugs and plutons are altered to an M1 assemblage such as that

seen in the Yeoval Complex.

103

Figure 5 - 12: Dubbo 1:250,000 map sheet area showing the metamorphic zones (Meakin and Morgan, 1999).

104

5.5 Local Geology

5.5.1 Introduction

The study area is located in the Spicers Creek catchment within the Cobbora

1:100,000 map sheet area on the Dubbo 1:250,000 map (Figure 2-4). Spicers

Creek catchment is situated about 45 km from the city of Dubbo to the west, 20

km from the town of Cobbora to the east, 15 km from the town of Boomley to

the north and south 45 km to Wellington (Figure 5-13).

Cobbora region is situated in the north-eastern portion of the Lachlan Fold Belt

(LFB) and overlying marginal areas of the Gunnedah Basin and Surat Basin.

The basement rocks of the Lachlan Fold Belt have been regionally

metamorphosed and contain north-south trending faults, folds, and fractures.

The Gunnedah Basin Sequence includes a number of basic intrusions of

Mesozoic and Tertiary rocks. These are associated with massive extrusions of

Garrawilla Nolanic complex and the Liverpool, Warrumbungle and Nandewar

Ranges (Tadros, 1993). The Surat Basin contains 2500 metres of mainly

Jurrasic continental sedimentary rocks and lower Cretaceous marine beds

largely obscured by Cainozoic alluvium (Exon, 1976).

The area of study, is geologically complex due to the tectonic events it has

endured and the landscape evolution is closely related to the underlying

geology (Morgan and Jankowski, 2002). The Palaeozoic rocks range in age

from Ordovician through to Devonian. The depositional environments ranges

from marine to sub-marine (Schofield and Jankowski, 2000). Lithologies 105

include submarine felsic, rhyolitic tuffs, tuffaceous sedimentary rocks and latitic

lavas and intrusives (Meakin and Morgan, 1999). Permian to Triassic

sediments contain marine and fluvial sediments with some volcanics. Jurassic-

Cretaceous sediments consist of terrestrial-sourced shales, coals, and well-

sorted sands. Depositional environments include floodplains, meandering

streams and high energy braided streams. These sedimentary cover rocks dip

in a northwesterly direction and Tertiary basalts intrude the sediments as sills,

plugs, dykes, and relic basalt flows. Quaternary alluvial and collovial sediments

blanket drainage lines and low relief areas (Morgan and Jankowski, 2002).

Figure 5 - 13: The road map of the Dubbo area.

106

5.5.2 Climate

In the Spicers Creek catchment the highest temperatures are January with

maximum mean of 33˚C and the lowest temperatures are in July with a

minimum mean of 1.8˚C. The mean average rainfall for the catchment was

measured as 563 mm. Summer is the wettest season receiving 31% of the

annual rainfall (Jewel, 2000, Morgan and Jankowski, 2002).

5.5.3 Stratigraphy

The rock units that are represented in the Cobbora area are categorized into

those relating to basement and those that constitute the cover rocks. The

Ordovician to Late Devonian, mainly marine basement rocks were deformed

during compressive Middle Devonian and Middle Carboniferous orogenic

episodes, whilst the mainly terrestrial Late Carboniferous to Late Jurassic cover

rocks were deposited in an extensional tectonic regime.

Figure 5-14 shows the geological map around the area are scale of 1:100,000.

Table 5-1 is a summary of the different rock types found in the area.

Throughout the following sections the characteristics of the geological units

around the study area are documented (Meakin and Morgan, 1999).

107

Figure 5 - 14: Geological map of study area from Cobbora 1:100,000 geological map with the two sites marked. The various units are described in Table 1.

108

Table 5 - 1: Description of different rocks type in the Spicers Creek area (From Cobbora 100,000 geological map).

Symbol from

fig 2-14

Rock

Formation

/Group

Age

Composition

Qa

--- Quaternary Alluvial silt, clay and sand, variable humic content,

sporadic pebble to cobble-sized unconsolidated

conglomeratic lenses.

Rp Napperby

Formation

Triassic Siltstone thinly interbedded with fine-medium grained,

lithic-quartz sandstone, minor conglomerate;

coarsening-up sequences.

Rb Boulderwood

Formation

Triassic Thick-bedded quartz-lithic conglomerate, coarse

pebbly quartz-lithic sandstone and siltstone.

Pd

Dunedoo

Formation

Permian

Pebbly quartz-lithic sandstone to conglomerate,

breccia, white claystone, thin coal seams.

Pe

---- Permian Carbonaceous siltstone, quartz lithic sandstone,

conglomerate and coal lenses rare varves.

Dgc Cuga Burga

Volcanics

Devonian Latitic, crystal-lithic sandstone, breccia, siltstone, tuff,

latite and lesser andesite, basalt, minor

allochthonous limestone, quartzose sandstone.

Sms Gleneski

Formation

Silurian Rhyolitic to latitic lava, intrusive and tuff and

volcaniclastic sandstone.

θm Ungrouped

Ordovician

Intrusions

Late

Ordovician

Monzonite and diorite to more felsic syenite and

granite.

θco Oakdale

Formation

Ordovician Basalt, basaltic andesite, latite lava and intrusions,

volcaniclastic breccia, conglomerate, sandstone and

siltstone, minor allochthonous limestone.

109

5.5.3.1 Palaeozoic units

The Palaeozoic rocks range in age from Ordovician through to Devonian. The

early Ordovician sequence represents the oldest strata on the Dubbo 1:250,000

map sheet area and the study area.

5.5.3.1.1 Oakdale Formation (θco)

The Oakdale Formation is one of the most widespread units of the Cabonne

Group. The Formation takes it name from the Oakdale property, at the

southern end of the Oakdale Anticline (Meakin and Morgan, 1999). It forms a

broad, structurally complex, largely fault–bounded belt about 120 km in strike

length which extends from 35 km north of Wellington to 20 km south-east of

Orange.

Constituent units contain a number of allochthonous, and possibly rarely

autochthonous, limestone bodies, as well as limestone breccia and calcareous

sandstone. Primary volcanic rocks of the Oakdale Formation include

predominantly shoshonite, basalt to basaltic andesite and less commonly latite

lavas and high level intrusions. Volcanoclastic breccia and conglomerate are

common in the Oakdale Formation, particularly in the area to the north of

Wellington. The units are interbedded with sandstone turbidite packages and

primary volcanic layers.

110

The units are massive to coarsely bedded, clast- to matrix-supported, poorly

sorted and range from 3 m to over 20 m in thickness. This formation is

unconformably overlain in the main belt by late Early to Late Silurian Mumbil

Shelf strata, represented south of Wellington by the Dripstone Formation, and

north of Wellington by the Narragal limestone and Gleneski Formation of the

Mumbil Group.

In the Oakdale Formation, graptolite and brachiopod fauna have been

described by a number of researchers from the siltstone units (Meakin and

Morgan, 1999).

5.5.3.1.2 Ungrouped Ordovician Intrusions (θm)

Many small intrusive bodies of presumed Ordovician age occur within the

Oakdale Formation. None of these units have been isotopically dated. The

intrusions range from monzonite and diorite to more felsic syenite and granite in

composition and the outcrops range in size from less than 100 metres up to 900

metres in diameter.

Metamorphism by these intrusions is generally limited to weak to moderate

development of chlorite-sericite-epidote alteration of plagioclase and

ferromagnesian minerals, indicating a lower greenshist grade (Meakin and

Morgan, 1999).

111

5.5.3.1.3 Gleneski Formation (Sms)

The formation takes it name from the Gleneski property north of the Macquarie

River (Meakin and Morgan, 1999). The Gleneski Formation is comprised

predominantly of felsic to rhyolitic tuff and tuffaceous sedimentary rocks, with

lesser rhyolitic to felsic to latitic lava and intrusions, shale, and minor limestone.

Massive felsite, quartz felsite, quartz latite and rhyolite lava and intrusions

locally are intercalated with the volcanoclastic rocks of the Gleneski Formation.

These lithologies are interbedded on a variety of scales. Rare limestone and

calcareous sandstone are interbedded with volcanoclastic rocks of the Gleneski

Formation.

The massive and graded beds, erosional bases and local cross bedding in

clastic rocks of the Gleneski Formation suggest that they were deposited

subaqueously by turbidity currents. The magma source for the Gleneski

Formation was acid, but both primary volcanic and pyroclastic rocks, as well as

volcanic and lithic fragments in tuffaceous units, are deficient in quartz

phenocrysts (Meakin and Morgan, 1999).

5.5.3.1.4 Cuga Burga Volcanics (Dgc)

The name is derived from Cuga Burga Creek, which cuts through the formation

to the south-east of the Oakdale Anticline (Meakin and Morgan, 1999). The

Cuga Burga Volcanics contains no autochthonous fossils and age constraints

for the formation are provided mainly by conodont data for overlying and 112

underlying formations. It consists mainly of shoshonitic latitic volcanic and

volcanoclastic rocks. A thick sequence of pillow lavas is well-exposed in the

representative section of this units. A low radiometric character characterizes

the Cuga Burga Volcanics, although locally, the formation has a strong

potassium (pink) response.

The belt of Cuga Burga Volcanics separating the Yeoval Complex from the

Yennora Granite in the western area shows a complex intrusive history. Many

dykes are present, varying in composition from microdiorite and graphic

microgranite to dacite and rhyolite to basalt. The Cuga Burga Volcanics

contains no authochthonous fossils and age constraints for the formation are

provided mainly by conodont data for the overlying and underlying formations

(Meakin and Morgan, 1999).

The Cuga Burga Volcanics is characterized by a low radiometric character (dark

signature on radiometric images), although locally, such as in the fault-bounded

belt north of The Gap, the formation has a strong potassium (pink) response.

The formation has an overall anomalous magnetic high signature on total

magnetic intensity images. Igneous rocks of the formation have a high magnetic

susceptibility while volcaniclastic rocks are more variable (Meakin and Morgan,

1999).

113

5.5.3.1.5 Early Permian Undifferentiated (Pe)

This unit is largely exposed in a series of north-west trending lobes which

extend northwards under the Late Permian to Jurassic sedimentary rocks of the

Gunnedah Basin and the overlying Great Australian Basin. It contains

carbonaceous siltstone, quartz lithic sandstone, conglomerate coal lenses and

rare varves. This unit is somewhat better exposed in the Spicers Creek. In the

Spicers Creek the basal Early Permian sequence consists of poorly sorted,

sedimentary polymictic breccia and conglomerate which contains subangular to

rounded clasts of sandstone, phyllite and quartz which range up to 40 cm in

size.

The undifferentiated Early Permian, a possible distal equivalent to the Rylstone

Volcanics, correlates with the Late Carboniferous to earliest Permian, thick non-

marine fluvio-glacial conglomerates of the Talaterang Group in the southern

Sydney Basin, and the fluvio-lacustrine sediments of the Seaham Formation

(Meakin and Morgan, 1999).

5.5.3.1.6 Dunedoo Formation (Pd)

The Dunedoo Formation forms a relatively thin unit in the area of study. It

unconformably overlies the undifferentiated Early Permian (Pe) sedimentary

rocks and is disconformably overlain by Early Triassic sedimentary rocks. The

Dunedoo Formation is only moderately exposed at scattered localities, mainly in

114

creeks or occasionally on steeper slopes immediately below the more resistant

overlying Boulderwood Formation.

The formation consists of breccia, conglomerate and fine – to coarse- grained

sandstone and pebbly sandstone of quartzose to quartz-lithic and

quartzofeldspathic composition with a dominantly kaolinitic matrix. The lower

section of this formation is generally only poorly exposed. Surface exposures of

the Dunedoo Formation range up to 60 m on the Cobbora 1:100,000 map sheet

area. Common fining-up sequences in the Dunedoo Formation reflect fluvial

deposition (Meakin and Morgan, 1999).

5.5.3.2 Mesozoic units

Igneous rocks of Mesozoic age are scattered across Dubbo 1:250 000 map

sheet area, and are commonly associated with Permian to Triassic sedimentary

rocks of the Sydney and Gunnedah Basins. The Mesozoic rocks in the Spicers

Creek area are limited to the Boulerwood Formation and the Napperby

Formation. They cover of a large part of the north of area.

5.5.3.2.1 Boulerwood Formation (Rb)

The name is derived from Boulderwood homestead (Meakin and Morgan,

1999). It is typically exposed at the margins of slightly elevated plateaus

capped by the early Jurrasic Purlawaugh Formation. This formation ovelies the

115

Dunedoo Formation with probably disconformity or slight angular unconformity

and is conformably overlain by the Napperby Formation.

The Boulderwood Formation consists of very coarse-grained pebbly to

conglomeratic, lithic-quartz sandstone and grit, together with subordinate flaggy,

planar laminated, white, medium-grained quartz sandstone, claystone and

siltstone. Clasts are subangular to subrounded and predominantly quartz, with

minor fine silicic volcanic fragments and other siliceous metasedimentary

basement lithologies. The formation was deposited by a fluvial system.

Channel floor, channel bar and point bar sequences are common (Meakin and

Morgan, 1999).

5.5.3.2.2 Napperby Formation (Rp)

The name is derived from calcareous sandstone and laminated mudstone

exposed on Napperby station, near the Pine Vala well on the Coonabarabran-

Mullaley road (Meakin and Morgan, 1999). It conformably overlies the

Boulderwood Formation.

The formation consists of white, fine – to medium-grained, moderately to poorly

sorted lithic-quartz and quartzose, often flaggy, ferruginous sandstone with a

white clayey matrix thinly interbedded or interlaminated with grey siltstone and

minor conglomerate lenses. The Napperby Formation represents lacustrine-

delta deposition (Meakin and Morgan, 1999).

116

5.5.3.3 Cainozoic Unit

In the area of study the Cainozoic deposits are represented by alluvial deposits.

5.5.3.3.1 Alluvial Deposits (Qa)

No formal names have been applied to unconsolidated Quarternary units.

Alluvial deposits are geomorphological units. They can be regarded as falling

into three distinct morpho-stratigraphic types: actively depositing Holocene

zones; inactive Holocene/Pleistocene zones; and relict/elevated Pleistocene

zones. Generalized lithologies for the alluvium consist of silt and clay, with a

variable humic content.

5.5.4 Structure

The Spicers Creek region is situated in the east of the Dubbo 1:250,000 map

sheet and in the west of the Cobbora 1:100,000 sheet. It includes four of the

structural entities of the LFB; including the Cowra Trough, Molong High, Hill End

Trough and Capertee High (see Figure 5-10). Scott (1997) re-defined the

structural entities of the northeastern LFB into the four zones including the

Cowra-Molong Zone, Nindethana Zone, Hill End Zone and Capertee Zone

(Figure 5-15). He proposed there were three deformation events affecting rocks

across the map sheets. These are summarized in Table 5-2.

117

Site 2

Site 1Dubbo

Figure 5 - 15: Structural zones defined by Scott (1997) across the Dubbo geological map sheet.

D1 deformation in the late Middle Devonian was related to N-S compression

that produced east-trending folds and thrusts. The Ordovician and Silurian

strata were uplifted during this deformation because of thrusting them over early

Devonian sediments. This was followed by erosion as the elevated topography

was flattened. D2 deformation commenced in the early Carboniferous through

E-W shortening. It was a major deformation with an early compressional phase

which produced the dominant axial cleavage forming deformation. This

118

deformation varies in intensity style between the four structural zones as

described below.

Table 5 - 2: Proposed deformation events affecting rocks across the Dubbo 1:250,000 geological sheet Scott (Morgan, 1997).

Deformation Age Form Direction of Compression

D1 late Middle Devonian Folding and thrusting N-S compression

D2(major deformation) Early Carboniferous Cleavage, folding and E-W compression thrusting

D3 Early Carboniferous Kinking N-S compression

In the Cowra-Molong zone, the D2 deformation is represented by west-dipping

thrusts and east-dipping normal faults, with N-NE trending folds with near

vertical to west-dipping axial surfaces. The presence of small-scale synthetic

and antithetic strike-slip faults was probably associated with major strike-slip

faulting along the peripheries of the zone during the Early Carboniferous.

D2 deformation in Hill End Zone is not intense. Compression varied in

orientation from ENE-WSW in the north of the zone to ESE-WNW in the

southern part of the zone. This deformation is characterized by continuous,

shallowly plunging, open to tight folds with four-way dip closure. Folds have a

well-developed, axial plane cleavage.

119

In Nindethana Zone, D2 deformation is represented by E-NE dipping thrusts

and normal faults that cause repetition of the eastward-younging strata.

Deformation in the west of zone is higher than in the east part. Scott (Morgan,

1997) suggested that the zone was initially subjected to E-W compression

which produced north-trending faults. Subsequent ENE-WSW compression

resulted in dextral shear along these fault creating NW trending folds. In this

zone northwest trending folds have near vertical, east-dipping axial surfaces,

and are oblique to the north-trending faults.

D2 deformation in Capertee zone is characterised by west-dipping thrusts which

repeat the west-dipping stratigraphy. The direction of compression was ENE-

WSW, resulting in some dextral strike-slip faulting. Folds are open to tight with

an N-NE axial surface direction and have a well developed axial surface

cleavage.

D3 Early Carboniferous deformation involved N-S compression and

development of large scale easterly trending kink axes. There was no major

deformation after D3, but lineated faulting and cleavage development in some

parts the Wuuluman Granite and the Gulgong Granite. In the north of the map

sheet the sedimentary basin reflects the reactivation of the NNW trending faults,

with thick fault bound Permian outliers showing the significance post-Permian

faulting.

In the area of the study within the Molong Zone there are many faults and

thrusts striking north-south (Figure 5-16). These include the Nindethana Fault,

120

Narragal thrust, Neurea Fault and Macquarie fault. Throughout the following

sections the characteristic of the faults near the study area are documented.

5.5.4.1 Nindethana Fault

The Nindethana Fault is a major 50 km long thrust fault that extends from the

northern part of the Bathurst 1:250,000 map sheet area to the south-western

corner of outcrop of the Wuuluman Granite. The Nindethana Fault, for most of

its length, is localized within the Early Devonian Cunningham Formation.

However, west of Lake Burrendong, rocks east of the fault represented the Late

Silurian Mumbil Group and the overlying Early Devonian Cuga Burga Volcanics

in the cores and limbs of meridional anticlines (Meakin and Morgan, 1999).

Direct exposure of the Nindethana Fault occurs around Galwadgere (GR

692200 - 6383750) where it represents a 100 m wide shear zone, with strong

pyrite lineation. The dip of the fault is estimated to be 45-50˚ to the east

(Meakin and Morgan, 1999). The fault loses displacement northwards, so that

immediately south-west of the Wuuluman Granite, the footwall consists of Cuga

Burga Volcanics rather than the Cunningham Formation (Meakin and Morgan,

1999). The fault terminates to the north against a west-north-west trending

cross fault that forms part of a west-north-west corridor.

121

Figure 5 - 16: Structural map of the thrust sheet on the Cowra Zone and Molong Zone of the Dubbo geological sheet (Modified from Meakin and Morgan, 1999).

122

5.5.4.2 Narragal Thrust Sheet

The Narragal Thrust Sheet represents a combination of regional folds and

thrusts, and thus shows a structural transitional zone between the Molong Zone

to the west which is dominated by thrusts, and the Hill End Zone to the east that

is dominated by folds (Scott, 1997). This thrust sheet is bounded to the west

(below) by the east-dipping Narragal fault that marks the roof of Neurea Thrust

Sheet and to the east (end below) by the Nindethana Fault that also dips east.

The Narragal Thrust Sheet thus lies between two parallel thrusts (Meakin and

Morgan, 1999).

Cleavage in the Narragal Thrust Sheet dips steeply to the east and west and

swings in strike from NNW in the southern third of the map sheet area to NNE in

northern two third where it is approximately parallel in strike to fold axial traces.

The general obliquity with the meridonal bounding faults suggests a component

of left-lateral movement on them during deformation (Meakin and Morgan,

1999).

5.5.4.3 Macquarie Fault

The Macquarie Fault has a NNW trend and east dip. Its southern extent is

uncertain and it may disappear within the Garra Formation or link into a WNW

cross fault that cuts off the northern end of the Neurea Fault. This fault

juxtaposes Ordovician rocks in the hanging wall against the Late Devonian

Catombal Group in the Mount Arthur Syncline in the foot wall. The fault is

interpreted as one of a series of east-dipping thrusts with lesser displacement

123

and different geometry. The displacement history of these faults is uncertain.

They could represent either normal faults with varying displacement from north

to south or strike-slip fault (Meakin and Morgan, 1999).

5.5.4.4 Neurea Fault

The Neurea fault belongs to Neurea Thrust Sheet. The Neurea Thrust sheet is

bounded to the west by the Neurea Fault, which is interpreted to be an east-

dipping, oblique thrust. It is also bound to the east by the Narragal Fault,

interpreted to be an east-dipping thrust that juxtaposes Late Silurian Mumbil

Group and younger rocks in the Narragal Thrust sheet over Late Ordovician

Oakdale Formation (Meakin and Morgan, 1999).

In the Neurea Thrust Sheet some fold plunges are steep. These steep plunges

and the presence of strike-slip fibres/striae (plunging gently north) on steeply

east-dipping fault surface in the volcanics are interpreted in terms of strike-slip

movement on faults superimposed on vertical (east-side-up), dip-slip, thrust

movement.

124

5.6 Regional and Local Geophysics

5.6.1 Introduction

The Australian Geological Survey Organisation (AGSO) has acquired regional

geophysical data consisting of airborne radiometric and magnetic data and land

based gravity data. The geophysical data was provided as ERMapper grids,

one for both the Dubbo and Cobbora 1:100,000 sheets and a third grid for the

Gilgandra 1:250,000 sheet (Figure 5-17). The gravity data was provided as

Bouguer anomaly data corrected to a density of 2.67 tonnes/m.

Figure 5 - 17: Location of the geophysical data sets supplied by AGSO and the selected area of interest for the interpretation of this data (Schofield, 1998).

125

The radiometric and magnetic data for the Gilgandra Map Sheet was acquired

in 1979. Its resolution is significantly lower than that of the data from the Dubbo

Map Sheet which was flown as part of the National Geoscience Mapping

Accord (NGMA) in 1991. The survey parameters are listed in Table 5-3.

Table 5 - 3: Survey parameters for the Gilgandra and Dubbo airborne surveys.

Gilgandra Survey Dubbo Survey

Operator New South Wales Department New South Wales Department

of Mineral Resources of Mineral Resources

Date 1979 1991

Line spacing 1,500m 400m

Altitude 150m 100m

Sampling interval 60m (magnetic) 7m (magnetic)

60m (radiometric) 67m (radiometric)

Line direction E-W E-W

Instruments ٠ MF57 Fluxgate ٠ Geometrics G833 helium

magnetometer vapour magnetometer

٠Four-channel gamma ray ٠Four-channel gamma ray

spectrometer (crystal spectrometer (crystal

volume 16.8 litres) volume 33.6 litres)

126

5.6.2 Radiometric Methods

5.6.2.1 Theory, Application

Elements whose atomic nuclei contain the same number of protons but different

numbers of neutrons are called isotopes. They are forms of the same element

with different atomic weights. Certain nuclei disintegrate spontaneously

emitting α particles (helium nuclei He) and β particles (electrons and positrons).

This is the phenomenon of radioactivity. These emissions alter the nuclear

charge, α by –2, β (positron) by –1 and β (electron) by +1. As a result the

disintegrating nucleus is transformed into a nucleus of another element.

The natural decay of radioactive elements produces a variety of types of

radiation (alpha, beta, gamma) at specific energy levels. Only gamma-ray

radiation has sufficient energy to be useful for geological mapping or

exploration. There are a large number of radioactive minerals, but the more

common are potassium (K), uranium (U), and thorium (Th). The gamma

radiation is emitted when these elements decay gradually through time (see

Telford et al., 1976), for a description of the individual decay process and

series).

Radiometric surveying is employed in the search for radioactive deposits and

also for deposits associated with radioactive elements such as zirconium and

titanium (Kearey and Brooks, 2001). Radiometric surveys used in geological

mapping as different rock types can be recognized from their distinctive

signature (Pires and Harthhill, 1989). It is an important source of information for

127

soil, regolith and geomorphological studies, including their mineralogy, texture,

chemistry and style of weathering (Wilford et al., 1997).

As Jaques et al. (1997) demonstrated gamma-ray spectrometry is an important

tool for distinguishing igneous rock bodies because of composition and degree

of differentiation, as U, Th and K. These elements are critical elements involved

in magmatic differentiation processes.

5.6.2.2 Interpretation Methodology

Airborne radiometric methods measure gamma radiation emitted from the

Earth’s surface and it is recorded using a gamma-ray spectrometer. The most

diagnostically valuable of these for geological mapping correspond to potassium

(1.46 MeV), bismuth-214 (1.765MeV) from the uranium-238 decay series, and

thallium-232 (2.614 MeV) from the thorium-232 decay series (Horsfall, 1997).

When K, U and Th decay they create gamma radiation of differing energy, K

producing the lowest and Th the highest (Table 5-4).

The sensor detects the energy levels of the gamma spectrum, and the relative

contribution of each band is recorded as a separate channel along with a total

signal channel (Horsfall, 1997). Modern airborne gamma-ray data acquisition

for regional mapping consists of a multichannel spectrometer be able to

measure at least 256 channels of data in the energy range 0-3 MeV (Minty,

1997).

128

Table 5 - 4: IAEA (International Atomic Energy Agency) recommended windows for conventional 3-channel airborne gamma-ray spectrometry (IAEA, 1991).

Element analyzed Isotope used Gamma ray energy Energy window

Potassium K 1.46 MeV 1.370-1.570 MeV

Uranium Bi 1.76 MeV 1.660-1.860 MeV

Thorium Tl 2.61 MeV 2.410-2.810 MeV

The three bands of K, U and Th can be interpreted individually or combined as

a red-green-blue (RGB) composite image (Milligan and Gunn, 1977). The mix

of the three bands as the three primary colours produces a variety of colours,

which help in the interpretation of the relative concentrations of K (red), Th

(green) and U (blue) in the rock (Table 5-5). In creating images, the range of

each of the three colours is divided into 256 levels. For example, in RGB

images, the brightness of a particular colour equates to its relative abundance –

a dark red colour indicates a high level of potassium, with little or no thorium or

uranium. Furthermore, another factor for interpreting is using a cylindric

coordinate system in terms of colour variation perceived by the human eye as

the hue, saturation, and value model (HSV; Milligan and Gunn (1977)).

In the HSV model, hue indicates the combinations of primary colours (RGB),

value is the intensity (or energy of colour), and saturation is the relative lack of

white in the colour (Milligan and Gunn, 1977). Figure 5-18 shows the HSV

colour model that hue is measured around the vertical axis from 0˚ (red) to 360˚,

saturation varies from 0 on the vertical axis to 1 on the triangular surfaces of the

129

hexagonal, and value varies from 0 (black) through shades of grey along the

central axis to 1 (white) at the top. ( Milligan and Gunn, 1977).

Table 5 - 5: Interpretation of relative concentration of K. Th, and U in rocks based on the RGB image (Morgan, 1997).

K Th U Resultant colour

High Low Low Red

Low High Low Green

Low Low High Blue

High High Low Yellow

High Low High Purple

Low High High Light Blue

High High High White

An interpretation of gamma-ray data needs to incorporate many variables.

Besides the geometry and physical property contrasts of the radioactive source,

environmental effects can influence the interpretation of data. Environmental

factors include air temperature and pressure, soil moisture, non-radioactive

overburden, rainfall, vegetation and precipitation (Minty, 1997). Barren

overburden can reduce the radiation output from the Earth’s surface because of

its high density. In some areas, dense vegetation may shield the source of

radiation which would be equivalent to 50 m of air. The trunks of trees in dense

130

forests also reduce the effect of radiation from the ground. Changing

temperature and pressure can lead to a change in air density by up to 30 per

cent. Rainfall and soil moisture decrease the effect of radiation from the surface

as well (Minty, 1997)

Figure 5 - 18: The HSV colour model – H, S, and V represent hue, saturation, and value of colour intensity, respectively. According to this model, for a colour of constant hue and saturation, if its value is decreased it darkens towards black. For a colour of constant hue and value, if the saturation is decreased it becomes whiter (Milligan and Gunn, 1977).

131

5.6.2.3 Regional Interpretation

Figure 5-19 shows the RGB ternary radiometric image for the Dubbo area. It

shows that over the Dubbo 1:250,000 map sheet area the Ordovician volcanic

rocks are the generally low radioactivity, with slightly higher levels of potassium

parting a characteristic dark red to red appearance in RGB images. The

Silurian sedimentary rocks of the Cowra Trough show commonly of low to

moderate radioactivity, with thorium radiation dominant. This gives a

characteristic dark green anomaly in the RGB radiometric image. The

sedimentary rocks of the Hill End Trough represents the low to moderate levels

of radiation, with elevated thorium radiation. The Silurian sedimentary rocks of

the Capertee High appear in the RGB images as areas of dark green, indicating

low levels of radiation with slightly elevated levels of thorium radiation (Meakin

and Morgan, 1999).

The Devonian sedimentary rocks produce little radioactivity, resulting in dark

areas on the radiometric RGB image. For example, the Devonian Cuga Burga

Volcanics produces extremely low levels of radioactivity, giving a characteristic

black anomaly on RGB images. The Carboniferous intrusions over the Dubbo

area (such as the Gulgong Granite) exhibit moderate to high potassium

radioactivity with red colour in the RGB image. The northern outcrop of the

Gulgong Granite (East of Dubbo) shows high levels of potassium and thorium

radiation, giving a characteristic yellow colour in the RGB images. Permian to

Mesozoic sedimentary and Tertiary volcanic rocks generally emit low to

moderate levels of radioactivity.

132

In the north-eastern part of Dubbo 1:250,000 map sheet, the basalts may be

mapped from the radiometric images but in the north-western part, the generally

lower radioactivity levels emitted by the Great Australian Basin (Surat Basin)

sedimentary rocks makes it more difficult to distinguish between the

sedimentary rocks and the basalts (Meakin and Morgan, 1999).

As Jaques et al. (1997) showed there are some associated intrusive bodies in

the south of Dubbo ( probably Mesozoic age) with the development of alteration

and rare-metal mineralisation such as Zr, Hf, Nb, Ta, light REE and Y. The

gamma-ray image shows that each intrusion has a different geochemical

signature. Some have dominant U+Th signatures, others dominant K+U or

K+Th, and some are high in all three elements (Figure 5-20).

133

Figu

re 5

-19:

Red

-gre

en-b

lue

com

posi

te ra

diom

etric

imag

e fo

r Dub

bo a

rea

(Mea

kin

and

Mor

gan,

199

9).

134

Figure 5 - 20: Gamma-ray image of an area south of Dubbo, showing the geochemically different signatures of a number of minor intrusions in the area. Some intrusions have dominant U+Th signatures, others dominant K+U or K+Th, and yet others are high in all three elements (Jaques et al., 1997).

135

5.6.3 Aeromagnetic Methods

5.6.3.1 Theory, Application

Magnetic surveys can be carried out in the air from helicopters or fixed wing

aircraft, because large areas can be covered in great detail and relatively

inexpensively. The operation is known as airborne magnetic surveying or

aeromagnetic surveying. Magnetic survey can also be carried out on the ground

for more detail. The instrument used to measure the magnetic field is called a

magnetometer.

Airborne magnetic surveys are used extensively to map regional geology in

areas of poor outcrop. The usual procedure is to record data continuously

along equally spaced parallel flight lines covering the survey area. In Australia,

airborne surveys are flown routinely at line spacings of 400-500 m and in areas

where greater detail is required, flight-line spacing is reduced to 200 m. For

regional mapping the flight line direction is usually oriented orthogonal to the

predominant strike of the geology. For more specific applications, such as

mineral exploration targets, the flight-line separation and direction will be

selected to maximise the magnetic signature (Horsfall, 1997).

Ferromagnetic minerals in the earth’s crust are the source of anomalous

magnetic field. The major magnetic minerals are magnetite, titanmagnetite,

titanhematite, maghemite, pyrrhotite and native iron. These minerals can show

magnetic anomalies, because of their abnormally large magnetic susceptibilities

or because they have high remnant magnetizations. Of the magnetic minerals

136

that are found in the nature, magnetite is the most abundant (Grant, 1984).

Igneous rocks have a higher content of magnetic minerals than sediments and

can be mapped in sedimentary basins from magnetic data (Gunn, 1997a).

The contrasting proportions of these minerals in different crustal rocks produce

the magnetic anomalies that are the targets of magnetometer surveying.

Airborne magnetic surveys measure the total magnetic field that includes

induced magnetism and remnant magnetism. Any rock containing magnetic

minerals may possess both induced and remanent magnetizations (Gunn and

Dentith, 1997).

When a material is placed in a magnetic field it may acquire a magnetization in

the direction of the field which is lost when the material is removed from the

field. This phenomenon is known as induced magnetization or magnetic

polarization. The inherited magnetization remaining after removal of the applied

field is known as remanent, or permanent, magnetization (Kearey and Brooks,

2001).

The magnetic field decreases approximately as the inverse of the square of the

distance from the magnetic source. Therefore to record small variations in the

field, surveys must be flown close to the ground. The survey attitude is usually

in the range 80-60 m. Surveys with 200 m line spacing are typically flown 80 m

above ground while surveys with 100 m line spacing may be flown at 60 m

above ground (Horsfall, 1997).

137

5.6.3.2 Interpretation Methodology

Rock magnetism results from the magnetic minerals they contain. Magnetic

minerals also belong to iron-titanium-oxygen solid solution series from

magnetite (FeO) to ulvospinel (FeTiO) or the iron-sulphur group pyrrhotite

(FeS). Magnetite can occur as either primary magmatic minerals or as

metamorphic minerals. Basic igneous rocks are usually highly magnetic due to

their relatively high magnetite content. In general magnetite concentrations

decrease with increasing acidity in magmatic rocks (Kearey and Brooks, 2001).

Magnetic anomalies range in amplitude from a few tens of nanoteslas over

deep metamorphic basement to several hundred nanoteslas over basic

intrusions and may reach an amplitude of several thousand nanoteslas over

magnetite ores. The magnetic field for typical magnetic bodies in the

subsurface decays rapidly away from its source (Kearey and Brooks, 2001). As

a result shallow bodies are subjected by high wavenumber components,

creating narrower relatively low amplitude anomalies, whereas deep bodies

produce broader, smoother anomalies (Schofield, 1998).

Ferromagnetic rocks or sediments show positive magnetic susceptibility

whereas diamagnetic rocks or sediments show negative magnetic susceptibility.

For example pure salt comprising hatite, gypsum and anhydrite, is diamagnetic

and it has a negative magnetic susceptibility and as a result represent a

negative magnetic contrast relative to surrounding sediments (Gunn, 1997b).

138

The aim of magnetic interpretation is to obtain geological information from

magnetic survey data (Gunn, 1997b). The results of magnetic survey are

usually presented as a map, or an image, of the magnetic anomalies. The

magnetic anomaly is the product of the susceptibility of the rock and the size of

the body or rock. A colour total magnetic intensity (TMI) image provides a

useful method for interpreting geological units and their structure. Boundaries

between different geological units are determined on the basis of their magnetic

response. Common causes of magnetic anomalies include dykes, faulted,

folded, magnetite ore bodies and metamorphic basement rocks (Kearey and

Brooks, 2001).

There are some environmental factors that can influence magnetic anomalies

include temperature and weathering. It is unusual to find magnetite in

unmetamorphosed sedimentary rocks except in very minor amounts. Magnetite

is unstable in the low-temperature, highly oxidizing environment of chemical

weathering and sedimentation. Four factors effect the chemical weathering and

sedimentation of iron covsist of chemical weathering, leaching, transportation

and compaction and early diagenesis (Grant, 1984). Figure 5-21 presented the

total magnetic intensity (TMI) image for the Dubbo region.

5.6.3.3 Regional Interpretation

Sedimentary rocks over the Dubbo area show low to medium magnetic

susceptibility, but contain indications of internal north-south structure. In the

western part of Dubbo, the Silurian sedimentary rocks in the Cowra Trough 139

generally show low magnetic susceptibility. The magnetic susceptibility of the

Silurian volcanic rocks in this area is commonly low (range 10x10SI to

25x10SI). Silurian rocks on the Capertee High represent low magnetic

susceptibility (<20x10SI) that is reflected in low-amplitude magnetic anomalies

(Meakin and Morgan, 1999).

Devonian sedimentary rocks crop out over large portions of the Dubbo

1:250,000 map sheet area. They have low magnetic susceptibility (<20x10SI)

in relation to four structural zones including Cowra Zone (Cowra Trough),

Molong Zone, Hill End Zone and Capertee Zone (Meakin and Morgan, 1999).

The Carboniferous intrusions (e.g. Wuuluman Granite and Gulgong Granite) are

generally in the low to moderate range for magnetic susceptibility (between

20x10SI to 2000x10SI). The magnetic susceptibility values measured for the

Permian and Mesozoic sedimentary rocks of the Sydney Basin, Gunneda Basin

and Surat Basin are generally very low (<10x10SI). With low contrasts in

magnetic susceptibility, they do not produce significant magnetic anomalies

(Meakin and Morgan, 1999). However, the magnetic susceptibility values for

the Tertiary volcanic rocks are in the medium range (300x10SI to 1500x10SI).

Flows In the north-eastern part of Dubbo and in the west of Dubbo overlie

sedimentary rocks and older volcanic materials respectively, which have

moderate to high magnetic susceptibility values. Therefore, it is difficult to

distinguish the anomaly pattern of the Tertiary basalts from those of the

underlying rocks (Meakin and Morgan, 1999).

140

Survey area

Figure 5 - 21: TMI aeromagnetic image for the Dubbo region ( Modified from Meakin and

Morgan, 1999).

141

5.6.4 Gravity Methods

5.6.4.1 Theory, Application

Australian Geological Survey Organisation (AGSO) carried out numerous

traverses throughout the Dubbo and Ballimore region. The stations are on

average about 4 km apart. Figure 5-22 shows that the distribution of the

stations is not regular and accordingly parts of the area have dense coverage

while in others parts coverage is quite sparse. The gravity data are based on

the IGSN71 gravity datum and the ISOGAL84 scale.

Figure 5 - 22 Point located data for the Dubbo and Ballimore regions (Schofield, 1998).

142

Gravity surveying measures variations in the Earth’s gravitational field produced

by differences of density between subsurface rocks. The theory that explains

gravitational methods is based on Newton’s law. This law is expressing the

force of mutual attraction between two particles in terms of their masses and

separation. The force of attraction F between two masses m and m, that

dimensions are small with respect to the distance r between them, is given by

221

rmGmF =

where G is the Gravitational Constant.

The variation in gravity depends on lateral changes in the density of earth

materials in the vicinity of the measuring point (Dobrin and Savit, 1988). Gravity

method is a natural field method, which measure a bulk petrophysical property

such as density. Density depends on both composition and porosity. In

sedimentary rocks, porosity is the main cause of density variations and density

tends to increase with depth because of compaction, and with age due to

cementation. Igneous and metamorphic rocks have small porosity, and

composition is the main cause of density variation. In igneous rocks density

increase as acidity decreases.

Like radioactivity and magnetic methods, gravity is a natural source method in

which local variations in density of rocks near the surface cause small changes

in the main gravity field. Some applications of gravity surveying are in regional

geological studies, hydrocarbon exploration, geodesy, monitoring volcanoes

and archaeogeophysics (Kearey and Brooks, 2001). Gravity studies have

143

shown that the graben portions of the rift system correspond to narrow gravity

lows caused by low density sediments along the rift (Gunn, 1977a,Swain,

1992).

Gravity surveying can be particularly useful in the initial exploration of water

covered shelf areas, if there is no geological information (Dobrin and Savit,

1988). However gravity appears to have more limited applicability in mineral

exploration because of the good contrasts between the densities of base metal

sulphides and the host rocks. Some ores, like chromites, have a high density

compared with the material that surrounds them and gravity methods can be

used directly by detailed gravity surveys.

5.6.4.2 Interpretation Methodology

The geological interpretation of gravity data is more difficult and needs more

corrections (such as Bouguer, elevation, free air, hight correction, tide and drift

corrections) rather than magnetic methods. It is known that the gravity on the

Earth’s surface depends on five factors, latitude, elevation, topography of the

surrounding terrain, earth tides and variation in density in the subsurface.

Because of the variations in the density in the subsurface, the interpretation of

potential field anomalies for gravity data is inherently ambiguous (Kearey and

Brooks, 2001). One of the most difficult problems in gravity interpretation is the

separation of residual anomalies from the regional Bouguer gravity anomaly

(Mickus et al., 1991).

144

Gravity anomalies can be either positive or negative, depending on whether the

body is more or less dense than its surroundings. The magnitude and shape of

a gravity anomaly can indicate the depth, shape and size of the causative body.

Gravity anomalies are smaller and much smoother than magnetics because

density variations are relatively small and regular compared to changes in

magnetic susceptibility.

5.6.4.3 Regional Interpretation

Figure 5-23 shows the gravity image for the Dubbo region. Over the Dubbo

area, a major gravity high, show in red, runs from the south-western corner to

the central northern boundary. The gravity anomaly image is of low magnitude,

but shows broads of internal north-south structure. This structure is consistent

with aeromagnetic data described previously. The western gravity ridge marks

the position of the Ordovician volcanic rocks. On the center of map, the gravity

low indicates the position of the Hill End Trough.

Figure 5-24 shows the regional features that can be identified from a

comparison of the gravity and magnetic image for Dubbo region (Schofield,

1998).

145

148 30 00E 149 00 00E 149 30 00E

32 0

0 00

S32

30

00S

31 4

5 00

S

149 00 00E

32 0

0 00

S

Survey area

Figure 5 - 23: Imaged gravity data for the Dubbo region (Modified from Meakin and

Morgan, 1999).

146

Figure 5 - 24: Regional features that can be identified from a comparison of the gravity and magnetic images (Schofield, 1998).

147

Chapter 6

Near-Surface Effect Corrections with the GRM

6.1 Summary

The determination of seismic velocities in the unweathered layer with shallow

seismic refraction methods is often an ill posed problem. Very small variations

in the parameters used to derive the seismic velocities in the refractor can result

in significant variations in the computed velocities. In many cases, where small

receiver spacings are employed to define very shallow targets, such as is

commonly the case with geotechnical investigations, the variations in the

computed time parameters are often as large as the errors in the traveltime data

(Palmer et al, 2005).

The standard methods for correcting for surface irregularities are usually not

practical where the very near surface irregularities are of limited lateral extent.

In such circumstances, the GRM smoothing statics method (SSM) is a simple

and robust approach, which can facilitate more accurate estimates of refractor

velocities (Palmer et al., 2005).

The Near-Surface Effect Corrections with the GRM generates a smoothing

“statics” correction by subtracting an average of the time-depths computed with

a range of XY values from the time-depths computed with a zero XY value

148

(where the XY value is the separation between the receivers used to compute

the time-depth). The time-depths to the deeper target refractors do not vary

greatly with varying XY values, and therefore an average is much the same as

the optimum value. However, the time-depths for the very near surface

irregularities migrate laterally with increasing XY values and they are

substantially reduced with the averaging process.

As a result, the time-depth profile averaged over a range of XY values is

effectively corrected for the near surface irregularities. Also, the time-depths

computed with a zero XY value are the sum of both the near surface effects and

the time-depths to the target refractor. Therefore, their subtraction generates

an approximate “statics” correction, which in turn, is subtracted from the

traveltimes.

6.2 Introduction

Variations in the topography and thickness of the near-surface soil layers can

have a major affect on traveltimes, because the very low seismic velocities,

often as low as 250 m/s, can result in large increases in traveltimes. These

layers can lead to significant problems with the detailed inversion of refraction

data, commonly as artefacts in the refractor velocities if no attempt is made to

recognize and remove their effects. As such, it is usually necessary to correct

for variations in the near surface, if a detailed model of the refractor is required

as a starting models for tomographic inversion.

149

With the traveltime data recorded at Spicers Creek, it was found that the

consistency between adjacent lines, which were only 10 m apart, was not

always immediately obvious, even though the traveltime data generated

excellent refraction tomograms. However, when the near surface effects were

satisfactorily accommodated, the consistency between adjacent lines was

improved considerably.

The near-surface corrections are adjustments to the original field seismic

refraction data, in order to remove the effects of varying topography or surface

velocities. As part of this study, a new method for correcting near-surface

effects based on the generalized reciprocal method (GRM, Palmer, 1980, 1986)

has been developed and has been applied to synthetic models as well as field

data. In general, the method, whilst being simple to implement, provides a

significantly more reliable measures of detailed refractor parameters.

6.3 Correcting for Near-Surface Effects

Most methods for correcting for near-surface effects attempt to model the near

surface irregularities and to calculate the traveltimes through these

irregularities. The corrections are usually computed with the first arrival

refraction data, if there is adequate coverage of these irregularities. The

standard methods of correcting for any near surface variations, using

traveltimes obtained with intermediate shot-to-receiver spacings is described in

Dobrin (1983, p. 334–336) and Palmer (1986, p.152–155). If these corrections

150

are not made, the determination of refractor velocities can be subject to

considerable ambiguity, especially where the detailed resolution of narrow

features, such as shear zones, is concerned.

However, where the irregularities do not extend over more than a few receiver

intervals, these deterministic methods cannot be employed, because there is

usually insufficient reversed traveltime data to define them accurately. In such

situations, an alternative approach is required.

The new method for correcting near-surface effects, which was developed as

part of this study, takes advantage of a unique feature of the GRM in which a

variable refraction migration distance, known as the XY distance, is employed.

Since the GRM time-depths computed with a range of XY values generally do

not exhibit large differences, the average of the time-depths for a range of XY

values is largely similar to that computed with the optimum XY value. However,

anomalies caused by surface irregularities propagate laterally throughout the

traveltimes and time-depths (Palmer, 1986, p.107-111). Therefore, when the

time-depths are averaged for a range of XY values, the time-depth anomalies

and hence the near-surface effects are averaged and reduced.

The average of these time-depths is then subtracted from the time-depths

computed with a zero XY. The result is generally a good estimate of the time-

depth anomalies caused by the near-surface variations. This difference is

subtracted from the traveltimes, thereby affecting a good correction for the near-

surface variations. This process can then be repeated on the updated

151

traveltimes iteratively until there is a significant improvement in the computed

seismic velocities in and the depth to the refractor.

The use of the GRM to remove near-surface effects from seismic refraction data

is shown in the following sections. The method is applied to various synthetic

models taken from Palmer (1980, pages 17-30) to illustrate its theoretical

performance over a range of near-surface variations. In some cases, the near

surface variations vary smoothly, while in other cases, the variations are large

and discontinuous. The first model consists of an irregular refractor with plane

topography whilst the second model introduces an irregular topography. The

method described has also been applied to the data acquired during the Spicers

Creek seismic refraction survey.

6.4 Model 1: Irregular Refractor with Flat Topography

The model shown in Figure 6-1 has a plane horizontal ground surface and a

highly irregular refractor with dips of approximately 18 degrees on the sloping

surfaces. The seismic velocity within the refractor changes laterally from 4000

to 5000 m/s at station 175, whilst the seismic velocity of the layers above the

refractor is constant at 1500 m/s. The first arrival traveltimes, shown in Figure

6-2, were obtained by wavefront construction. This method was used since it

conveniently accommodates dipping refractors, interfering head waves, and

diffractions (Palmer, 1974). Since the model has a plane horizontal topography,

there are no effects from variations in the near surface, thus no corrections

152

were necessary. The inversion of the traveltime data for model 1 using the

GRM produce is shown in Figures 6-2 to 6-8.

20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340

20

0

-40

-20

20

0

-20

-40

1500 m/s

4000 m/s 5000 m/s

Figure 6 - 1: Model with a plane horizontal ground surface and a highly irregular refractor. The vertical and horizontal scales are equal (Palmer, 1980).

First Arrival Times Graph

0

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Tim

e (m

s) Forward

Reverse

Figure 6 - 2: Traveltime curves derived from the model in Figure 6 - 1.

153

The velocity analysis functions for XY-values ranging from 0 to 30 are shown in

Figure 6-3. They have been stacked in order to clearly show the minor

differences between each function (Figure 6-4). Detailed study of the velocity

analysis functions indicates that the optimum XY-value is between 15 and 20 m,

thus an optimum XY-value of 20 m was chosen. For the XY-value of 20, the

velocity analysis data fall very close to two straight lines. Whilst for other XY-

values, the points are scattered about theoretically correct straight lines. For

the zero XY value, the real lateral change in the refractor velocity is obscured by

numerous fictitious velocity changes, which are related to the irregular refracting

interface.

The seismic velocities within the refractor were determined using a standard

“linear method” (see Appendix A). Figure 6-5 shows the average velocity

analysis function, which was computed by averaging the functions for a range of

XY values which were symmetrical about the optimum XY value of 20 m, i.e.

10m<XY<30m. A number of straight lines were fitted onto this function, with the

seismic velocities obtained by taking the reciprocal of the gradients of the lines.

In Figure 6-5, two straight lines were fitted onto the average velocity analysis

function indicating one lateral change of the seismic velocity within the refractor

at station 180. The gradients of these lines give seismic velocities of

approximately 4000 and 5000 m/s and they are within 1 percent of the correct

values.

154

Stacked Velocity Analysis Graph

0

20

40

60

80

100

120

0 100 200 300

Distance (m)

Stac

ked

Velo

city

Ana

lysi

s (m

s)XY=0

XY=5

XY=10

XY=15

XY=20

XY=25

XY=30

Figure 6 - 3: Coincident velocity analysis functions for XY-values from 0 to 30 m (with same reciprocal time), derived from the traveltime data in Figure 6 - 2.

Stacked Velocity Analysis Graph

0

20

40

60

80

100

120

0 100 200 300Distance (m)

Stac

ked

Velo

city

Ana

lysi

s (m

s) XY=0

XY=5

XY=10

XY=15

XY=20

XY=25

XY=30

Figure 6 - 4: Stacked velocity analysis functions for XY-values from 0 to 30 m (with different reciprocal time), derived from the traveltime data in Figure 6-2.

155

Average Velocity Analysis Graph

20

40

60

80

100

120

0 100 200 300

Distance (m)

Ave

rage

d V

eloc

ity A

naly

sis

(ms)

4000 m/s

5000 m/s

Figure 6 - 5: Average velocity analysis graph, derived from the traveltime curves in Figure 6 - 2.

In order to verify the seismic velocities obtained by the “linear method”, a “non-

linear method” was also used (see Appendix A). This method uses a series of

numerical differentiation formulas derived by differentiating Lagrangian

interpolation formulas (Beyer, 1975). The refractor velocities were calculated by

substituting the values of the average velocity analysis function into the

numerical differentiation formulas, to give the derivative at each point along the

function and then taking the reciprocal to give the refractor velocity at each of

these points. In Figure 6-6, the refractor seismic velocities obtained using both

the linear and non-linear methods are shown.

156

Refractor Seismic Velocities Graph

2000

3000

4000

5000

6000

7000

0 50 100 150 200 250 300 350

Distance (m)

Ref

ract

or S

eism

ic V

eloc

ity (m

/s)

3 point5 point7 pointlinear

Figure 6 - 6: Linear versus non-linear refractor velocities from Figure 6 - 2.

The time-depth functions for XY-values ranging from 0 to 30 are shown in

Figure 6-7. As with the velocity analysis function, they have been stacked in

order to more clearly show the minor differences between each XY value. All

the XY-values give essentially the same time-depths except where there are

changes in dip. At these points, use of the non-optimum XY values can

produce significant smoothing.

The resolution of the irregularities in the refractor is maximised with the use of

the optimum XY-value which in this example is 20 m. Depth section is shown

in Figure 6-8 which is particularly evident for the depression at 225 m and for

the fault at 50 m.

157

Stacked Time-depth Graph

0

5

10

15

20

25

30

35

0 50 100 150 200 250 300 350

Distance (m)St

acke

d Ti

me-

dept

h (m

s)

XY=0XY=5XY=10XY=15XY=20XY=25XY=30

Figure 6 - 7: Time-depth for XY-values from 0 to 30 m.

0

10

20

30

40

0 25 50 75 100 125 150 175 200 225 250 275 300 325 350

Distance (m)

Dep

th (m

)

Figure 6 - 8: Depth section which shows particularly for the depression at 225 m and for the fault at 50 m.

158

6.5 Model 2: Irregular Refractor with Irregular Topography

The second model, shown in Figure 6-9, has the same very irregular refractor

interface with the lateral change in the seismic velocity as the first model, but

introduces an irregular topography. The first arrival traveltimes are shown in

Figure 6-10. A comparison with the traveltime graphs for the first model in

Figure 6-2 shows that the effects of near-surface variations are evident. Rapid

changes in topography, such as at 120 and 250 m, are shown as time-shifts in

both the forward and reverse traveltime graphs.

20 4060 80

100 120140

160180

200220 240

260 280300 320 340

20

0

-40

-20

20

0

-20

-40

1500 m/s

4000 m/s 5000 m/s

Rapid changes in topography

Figure 6 - 9: Model with irregular ground and refractor surfaces. The vertical and horizontal scales are equal (Palmer, 1980).

159

First Arrival Times Graph

0

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Tim

e (m

s)

ForwardReverse

Near surface effects

Figure 6 - 10: Traveltime curves derived from the model in Figure 6 - 9.

The stacked velocity analysis functions for XY-values ranging from zero to 30 m

are shown in Figure 6-11. Whilst none of the velocity analysis functions show

two distinct straight-line segments, an XY-value of 15 m was determined to be

the optimum. As with the first model, the refractor seismic velocities for the

second model were calculated using both the linear and non-linear methods. In

Figure 6-12, five straight lines were fitted on to the average velocity analysis

function, giving refractor seismic velocities of 4000, 3800, 8000, 3600, and 5000

m/s. It is evident that the velocity analysis has been adversely affected by the

variations in the near surface, particularly between 190 and 215 where the

velocity reaches an implausible 8000 m/s. Furthermore, the velocities are not

resolved with the non-linear method, shown in Figure 6-13, which returns an

erratic set of velocities.

160

Figure 6-14 shows the stacked time-depth functions for XY-values ranging from

0 to 30. The time-depths for all XY-values are highly irregular with the effects of

near-surface variations evident. This is particularly the case at 120 and 250 m

where the time-depths shift by 4 ms for the XY-value of zero and propagate

laterally across the time-depths for successive XY-values. Depth sections

calculated with XY-value of 15 m is shown in Figure 6-15. The computed

refractor surface, although following the general shape of the refractor in the

original model, is severely distorted. Without applying corrections for near-

surface effects, the inversion of the seismic refraction data has not generated

an accurate image of the subsurface.

Stacked Velocity Analysis Graph

0

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Stac

ked

Velo

city

Ana

lysi

s (m

s)

XY=0XY=5XY=10XY=15XY=20XY=25XY=30

Figure 6 - 11: Velocity analysis functions for XY-values from 0 to 30 m derived from the traveltime data in Figure 7-10. The data for a 15 m XY-value are judged the best.

161

Average Velocity Analysis Graph

V1=4000 m/s

V2=3800 m/s

V3=8000 m/s

V4=3600 m/s

V5=5000 m/s

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Ave

rage

Vel

ocity

Ana

lysi

s (m

s)

V1V2V3V4V5

Figure 6 - 12: Average velocity analysis graph, derived from the traveltime curves in Figure 6 - 10.

Refractor Seismic Velocities Graph

2000

3000

4000

5000

6000

7000

8000

9000

0 50 100 150 200 250 300 350

Distance (m)

Ref

ract

or S

eism

ic V

eloc

ity (m

/s)

3 point5 point7 pointlinear

Figure 6 - 13: Refractor seismic velocity graph derived from Figure 6 - 10.

162

Stacked Time-depth Graph

0

5

10

15

20

25

30

35

0 50 100 150 200 250 300 350

Distance (m)

Stac

ked

Tim

e-de

pth

(ms)

XY=0XY=5XY=10XY=15XY=20XY=25XY=30

Figure 6 - 14: Stacked time-depth for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 - 10.

0

10

20

30

40

0 25 50 75 100 125 150 175 200 225 250 275 300 325 350

Distance (m)

Dep

th (m

)

Figure 6 - 15: Depth sections calculated from time-depth using 15-m XY-values.

163

6.6 Applying Near-Surface Effect Corrections to Model 2

In order to obtain an accurate image of the subsurface, corrections for near-

surface effects should be made to the traveltime data. Figure 6-16 shows the

updated traveltimes for the second model after one iteration of the GRM method

for correcting near-surface effects. A comparison with the raw traveltimes

suggests a significant reduction of the near-surface effects, particularly at 120

and 250 m where the time-shifts have been removed. The variable nature of

the near-surface effects is illustrated in Figure 6-17, which gives the amount by

which the raw traveltimes were corrected.

First Arrival Times Graph, Correction 1

0

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Tim

e (m

s)

ForwardReverse

Figure 6 - 16: Traveltime curves derived from the model in Figure 7-10 after one correction.

164

Correction - 1st Iteration

-2

-1

0

1

2

0 50 100 150 200 250 300 350

Distance (m)

Tim

e (m

s)

Figure 6 - 17: First raw traveltime correction.

Figures 6-18 and 6-19 show the stacked velocity analysis and the average

velocity analysis graphs obtained from the updated traveltimes respectively.

The result of the correction is to smooth the velocity analysis graphs. Four

straight lines were fitted on to the average velocity analysis function giving

seismic velocities of 4000, 3700, 4750, and 4800 m/s. These velocities are a

significant improvement than those obtained with the raw traveltimes. The

seismic velocities obtained after one iteration of the correction method are

within 7 percent of the correct values, whilst the velocity high of 8800 m/s from

the raw traveltimes has been removed. However there are still some

uncertainties in the validity of the velocities. This can be seen in the velocities

obtained by the “non-linear” method, Figure 6-20, where there are distinct highs

at 200 m and 270 m.

In Figure 6-21, the time-depths derived from the updated traveltimes, are

markedly different to those derived from the raw traveltimes. A considerable

degree of smoothing has occurred, whilst the pronounced time-shifts at 120 and

250 m have been removed. Figure 6-22 shows the depth section computed

165

with the optimum XY-value of 15. A comparison with this depth section and that

computed with the raw traveltimes indicates only slight differences between

them. The major differences occur between 200 and 280 m, whilst the rest

remains largely the same. Although the first iteration of the GRM method for

correcting near-surface effects has resulted in a significant improvement in the

refractor seismic velocities, the depth section is still severely distorted and

markedly different to the correct depth section. Perhaps further iterations of the

correction method will resolve this.

Stacked Velocity Analysis Graph, Correction 1

0

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Stac

ked

Velo

city

Ana

lysi

s (m

s)

XY=0XY=5XY=10XY=15XY=20XY=25XY=30

Figure 6 - 18: Velocity analysis functions for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 - 10 after one correction.

166

Average Velocity Analysis Graph, Correction 1

V2=3700 m/s

V3=4750 m/s

V4=4800 m/s

V1=4000 m/s

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Ave

rage

Vel

ocity

Ana

lysi

s (m

s)

V1V2V3V4

Figure 6 - 19: Average velocity analysis graph, derived from the traveltime curves in Figure 15, after one correction.

Refractor Seismic Velocities Graph, Correction 1

2000

3000

4000

5000

6000

7000

8000

0 50 100 150 200 250 300 350

Distance (m)

Ref

ract

or S

eism

ic V

eloc

ity (m

/s)

3 point5 point7 pointlinear

Figure 6 - 20: Refractor seismic velocity graph derived from Figure 6 - 16, after one correction.

167

Stacked Time-depth Graph, Correction 1

0

5

10

15

20

25

30

35

0 50 100 150 200 250 300 350

Distance (m)

Stac

ked

Tim

e-de

pth

(ms)

XY=0XY=5XY=10XY=15XY=20XY=25XY=30

Figure 6 - 21: Stacked time-depth for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 - 10 after 1 iteration.

0

10

20

30

40

0 25 50 75 100 125 150 175 200 225 250 275 300 325 350

Distance (m)

Dep

th (m

)

Figure 6 - 22: Depth sections calculated from time-depth using 15-m XY-values after 1 iteration.

168

The updated traveltimes after four iterations of the GRM method for correcting

near-surface effects are shown in Figure 6-23. After this many iterations the

difference between the updated traveltimes and that of the previous iteration are

small, as shown in Figure 6-24, suggesting that further iterations would result in

minimal improvement. The velocity analysis graphs, shown in Figures 6-25 and

6-26, provide very accurate seismic velocities for the refractor. Two straight

lines have been fitted to the average velocity analysis function and give values

of 3914 and 4866 m/s, which are within 3 percent of the correct values.

Furthermore, these results are verified by their close correlation with the

velocities obtained using the “non-linear” method, Figure 6-27.

The time-depths obtained after four iterations of the GRM method for correcting

near-surface effects, Figure 6-28, are significantly different to the time-depths

derived from the raw traveltimes. As well as smoothing and removing the time-

shifts, after four iterations, the time-depths are quite similar to those derived

from the first model. Figure 6-29 shows the depth section computed with the

optimum XY-value of 15. There is a significant improvement in this depth

section when compared with that derived from the raw traveltimes or the first

updated traveltimes. Also, there is a close similarity between it and the depth

section derived from the first model, especially from 180 to 350 m, suggesting a

successful removal of the near-surface effects. However, the near-surface

effects have not been completely accommodated for. This is evident at 80 and

125 m where the refractor depth is still distorted. Furthermore, further iterations

would probably not completely resolve these distortions.

169

The GRM correction method as applied to synthetic models shows that in

general, it is reliable and consistent in removing the effects of variations in the

near-surface. The method is particularly effective with seismic velocities,

returning accurate results after one iteration, whilst significantly improving the

refractor shape after four. Perhaps its major advantages are that it is simple

and quick to implement, it requires no knowledge of the near-surface velocities

or other petrophysical parameters, and most importantly can be easily included

into the standard GRM inversion process.

First Arrival Times Graph, Correction 4

0

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Tim

e (m

s) ForwardReverse

Figure 6 - 23: Traveltime curves derived from the model in Figure 6 – 10 after four corrections.

170

Correction - 4th Iteration

-2

-1

0

1

2

0 50 100 150 200 250 300 350

Distance (m)

Tim

e (m

s)

Figure 6 - 24: Fourth time-depth correction graph.

Stacked Velocity Analysis Graph, Correction 4

0

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Stac

ked

Velo

city

Ana

lysi

s (m

s)

XY=0XY=5XY=10XY=15XY=20XY=25XY=30

Figure 6 - 25: Velocity analysis functions for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 - 10 after four corrections.

171

Average Velocity Analysis Graph, Correction 4

V2=5000 m/s

V1=4000 m/s

20

40

60

80

100

120

0 50 100 150 200 250 300 350

Distance (m)

Ave

rage

Vel

ocity

Ana

lysi

s (m

s)

V1

V2

Figure 6 - 26: Average velocity analysis graph, derived from the traveltime curves in Figure 6 - 23, after four corrections.

Refractor Seismic Velocities Graph, Correction 4

2000

3000

4000

5000

6000

7000

0 50 100 150 200 250 300 350

Distance (m)

Ref

ract

or S

eism

ic V

eloc

ity (m

/s)

3 point5 point7 pointlinear

Figure 6 - 27: Refractor seismic velocity graph derived from Figure 6 - 23, after four corrections.

172

Stacked Time-depth Graph, Correction 4

0

5

10

15

20

25

30

35

0 50 100 150 200 250 300 350

Distance (m)

Stac

ked

Tim

e-de

pth

(ms) XY=0

XY=5XY=10XY=15XY=20XY=25XY=30

Figure 6 - 28: Stacked time-depth for XY-values from 0 to 30 m derived from the traveltime data in Figure 6 – 10 after four corrections.

0

10

20

30

40

0 25 50 75 100 125 150 175 200 225 250 275 300 325 350

Distance (m)

Dep

th (m

)

Figure 6 - 29: Depth sections calculated from time-depth using 15-m XY-values after 4 iteration.

173

Chapter 7

Methodology and Data Acquisition

7.1 Summary

Three-dimensional (3D) three-component (3C) shallow seismic refraction

surveys were recorded over a shear zone at two sites associated with dryland

salinity in the study area. The seismic data were recorded with the Australian

National Seismic Imaging Resources (ANSIR) 360-trace ARAM-24 seismic

system and IVI MiniVibrator. Survey geometry consisted of four parallel

receiver lines 10 m apart, set orthogonal to the inferred shear zone. Each line

consisted of 29 three-component geophones at 5 m intervals. 532 shots were

recorded during the two surveys resulting in around 1.5GB of data.

7.2 Introduction

The fieldwork was carried out at two sites in the Spicers Creek Catchment from

the 6th to the 12th of April 2003, using 360 trace ARAM24 seismic system and

IVI Minivib seismic energy source operated by Australian National Seismic

Imaging Resource (ANSIR). ANSIR is operated jointly by the Australian

Geological Survey Organisation (AGSO) and the Australian National University.

174

The Minivib can be conFigured in the field to generate P- and S-waves (see

Appendix B for an Operations Reports).

The field crew consisted of personnel from Geoscience Australia, including the

party manager, vibrator operator and geophone engineer, the observer from

Trace Energy Service, and field hands from UNSW.

The survey at site 2 was conducted first. The majority of the first day was spent

setting out the geophone spread and testing the system. Recording started on

the afternoon of the first day and continued for another two and a half days.

Recording at site 2 was completed on the morning of the fourth day. The

spread was retrieved and then re-deployed at site 1. Recording at site 1

commenced in the afternoon of the fourth day and continued till the end of the

sixth day, after which the spread was retrieved.

7.3 Equipment

The equipment listed below were used for recording data during the field

procedure

1) ARAM24 seismic recording system consisting of 360 channels.

2) IVI Minivib T-15000

3) 120 three component geophones GS-20DM (natural frequency, 14 HZ).

4) Dynamites (Dyno Nobel HDP 150).

175

5) Dynamite firing system (Shot Pro).

6) Cables for 5 m station separation.

7) Remote Acquisition Module (RAM) and batteries.

8) Plastic pin markers.

9) GPS (Gamin’s GPS 12 personal navigator).

10) Auger for digging shallow hole.

Appendix C shows the detail of specifications for the equipment.

7.3.1 ARAM24 Seismic Recording System

The seismic recorder was a 360-channel ARAM24 system manufactured by

Geo-X systems. The ARAM24 is a telemetry seismic recording system, which

is capable of 3D acquisition. It consists of a Central Recording Unit (CRU) from

which the system operated. The CRU is capable of data QC, vibroseis

correlation. The system was mounted in a purpose-built truck. The truck

served as mobile processing station. It contained the equipment shown in

photo 7-1 that includes from left to right: printer, ARAM24 processor, and

monitor for displaying data.

176

Photo 7 - 1: The ARAM24 CRU.

7.3.2 Mini-Vibrator

The seismic energy source used in the most of the survey was the IVI Minivib

Model T-15000 manufactured by Industrial Vehicles International. The Mini-

vibrator is designed for small scale, high-resolution studies and it is pictured

below in photo 7-2. The Minivibe can be used in the normal P-wave

configuration and the S-wave configuration. The Minivib has an output of 6,000

pounds (2721.1 kg), and additional weights can be added to increase the hold

down weight to 17,000 pounds (7709.8 kg).

177

Photo 7 - 2: IVI Mini-vibrator Model T-15000.

The Minivib is a high fidelity, controllable, vibratory seismic source developed

for geophysical surveys. It is capable of generating a frequency-modulated

signal in the range of 10 to 550 Hz, which is almost six octaves. Among its

features are the following:

1. The Minivib is an environmentally “soft” source, which puts energy into

the ground over an extended period of time.

2. The computer-based controller allows the Minivibe to generate linear,

non-linear, and segmented sweeps.

3. The source signal can be optimised on site to suit local conditions and

the survey objectives.

178

The Minivib has the capability to rotate the vibrator mechanism from the vertical

to the horizontal, so that shear waves, as well as compressional waves can be

generated. Figure 7-1 shows the actuator in the two positions for the

generating of P- and S-waves. The vibrator in P-wave mode is shown in photo

7-3 with the actuator positioned vertically, whilst in photo 7-4 it is in S-wave

mode with the actuator positioned horizontally. This design also allows for the

rotation and positioning of the horizontally mounted actuator at any angle

relative to the line.

The ‘minivib’ P-waveconfiguration.

The ‘minivib’ S-wave configuration.

Figure 7 - 1: P-wave versus S-wave comparison.

179

Photo 7 - 3: The vibrator operating in P-wave mode. The actuator (the blue cylinder above

the base plate) is positioned vertically.

Photo 7 - 4: The vibrator operating in S-wave mode. The actuator is positioned

horizontally and is perpendicular to the axis of the truck.

180

7.3.3 Geophones

The receivers were three-component (3C) geophones and used GS-20DM 14

Hz elements in a PC-3D case. Each component was transmitted through a

separate cable. A spirit bubble level and an arrow on the case were used to

plant them in the correct orientation. Overall, 116 geophones (3C) were used

during field procedure. Photo 7-5 shows a receiver with the bubble spirit level

(the white circle) and the three cables (marked red, white and yellow).

Photo 7 - 5: GS-20DM 14 Hz three component geophone. A bubble spirit level (the white

circle) is used to plant the geophone in the correct orientation. Each component is

transmitted through a separate cable (marked red, white and yellow).

181

7.3.4. Explosive Source

The explosive source was ANZOMEX, which consists of penta-erythritol

tetranitrate (PETN) and trinitrotoluene (TNT). Each cartridge weighed 100 g

(photo 7-6). This explosive is generally used as a booster for ANFO mixtures

and as such, has a very rapid detonation producing a high frequency signal.

Detonation was controlled remotely using the Pelton Shot Pro Dynamite Firing

System (photo 7-7).

Photo 7 - 6: A cartridge of anzomex, a high frequency explosive composed of TNT and

PENT.

Photo 7 - 7: The Pelton Shot Pro dynamite firing system was used to remotely control

detonation.

182

7.4 Field Procedure

Survey geometry consisted of four parallel receiver lines 10 m apart, set

orthogonal to the inferred shear zone. Each line consisted of 29 three-

component geophones at 5 m intervals (Figure 7-2 and 7-3). This results in a

total of 116 receiver stations, and in turn 348 recording channels over an area

of 140 by 30 m. Several lines of source points were shot both cross-line and in-

line to the receivers using the mini-vibrator, with source points at 20 m intervals,

and dynamite at larger intervals. Plastic red and white numbered pin markers

were positioned at every station and hand held GPS used to determine the

coordinates (Appendix D and Appendix E). Appendix F shows comments

document from observer.

The weather over the seven days was typical for the area at that time of the

year. Temperatures were cold in the morning, and increased during the day to

a maximum in the mid to high 20s. Winds were slight, with gusts occurring up

occasionally, especially during the last two days. The weather was fine for most

of the survey, with overcast weather on the last two days. Light rain fell on two

days, and although there was some degradation in data quality, recording

continued, except for a short period of several hours when the rain was much

heavier. Heavier down pours resulted in half a day of non-operational time.

Firstly, site 2 and then site 1 were recorded.

183

7.4.1 Setting Up the Survey Spread

The first step in the field procedure was to determine the shear zone’s location

and trend at the survey site. This was determined from previous magnetic and

resistivity surveys, its surface expression, such as saline groundwater discharge

and salt scalding, from the presence of salt resilient vegetation which were

planted in order to combat the dryland salinity, from the location of piezometers

which were used during previous studies, and from the 1:100 000 geological

map of the area.

Once the general location and trend of the shear zone was determined the

receiver lines were then set out. A measuring tape was used to measure 140 m

over the and along a bearing orthogonal to the inferred shear zone. Pin flags

were then placed at 5 m intervals along the tape. The pin flags were kept in line

with each other by forward and backward siting. The adjacent lines were offset

orthogonally from the first line by 10 m intervals, with pin flags being placed at 5

m intervals. The second, third and fourth lines were then set out in similar

manner to the first line. The first Receiver station was 100 and if, positioned on

the lower-right corner of the rectangle, then the last station, number 187, was

on the upper-left.

The source points were located with a 300 m length of rope with markings at 5

m intervals. The rope was laid along a number of different lines parallel,

perpendicular and oblique to the receiver lines with pin flags being placed at 20

m intervals.

184

The accuracy in setting out the receiver and source lines was generally quite

good. This can be seen in the GPS readings, which were taken at both ends

and at the centre of each receiver line and at every source point. For site 1,

from the GPS readings the distance between the receiver lines ranged from 9.8

to 10.6 m, whilst the distance between the source points was 20 +/- 1.8 m. Site

2 had less precise measurements with 10 +/- 1.8 m between the receiver lines,

and 20 +/- 2.6 m between the source points.

After the receiver lines were set out, receiver cables were laid out using the

geophone truck. Three cables were used, one for each component, going up

the first receiver line, then down the second, back up the third, and down the

fourth. These cables were then connected to the recording vehicle.

Since the seismic recorder only had 3D capabilities for 4 lines, it was decided to

use individual line inputs for each component, and to concatenate the four lines

on the ground into a single time.

The three component geophones were planted at 5m intervals along the

receiver lines and connected to the three cables. Horizontal planting of the

geophones was readily achieved with the aid of a bubble spirit level on the top

of each unit and correct orientation of the horizontal components was achieved

by ensuring that was orthogonal to the receiver lines. These two steps ensured

that the directions of each of the components were the same for every

geophone. When connecting the geophone outputs to the cables, care was

185

taken to ensure that the correct component output was connected to the correct

cable.

Three RAMs and battery sets, one for each component, were required for every

eight geophones, four in front and four behind. The RAMs filtered, amplified,

digitised and formatted then sent the signals from the geophones to the

recording truck. Each amplifier was powered by a battery pack.

After setting up the recording spread, all the components were checked from

the recording truck. This process involved testing the response of each

geophone element to ensure that they were all connected and functioning

correctly, checking battery power levels, and testing the amplifiers.

7.4.2 Conducting the Survey

The recording procedure involved recording all the source points for the first

survey with the vibrator in P-wave mode then converting to S-wave mode, for

which all the source points were re-shot parallel to, anti parallel to, and in both

perpendicular directions to the receiver lines. This process was reversed for the

second site with S-waves then P-waves recorded. The dynamite shots were

recorded whilst the mini-vibrator was being converted from P-wave to S-wave

mode, a process which took approximately to four hours. Photo 7-8 shows the

successful detonation of a small charge.

186

Photo 7 - 8: A successful detonation of a small charge.

Recording the source points with the mini-vibrator was an efficient operation. At

each source point the mini-vibrator made four linear sweeps from 10 to 200 Hz

over 8s. The sampling rate was 1 ms, with samples taken over 2 s. The

vibrator deployed its pad at the first source point, swept, retracted its pad then

moved forward 20 m to the next source point within a few minutes. This was

repeated until the mini-vibrator reached the end of the source line, where it then

moved to the beginning of the next source line. The observer’s log (Appendix

F) shows the time taken for the Mini-vibrator to shoot each source line.

Due to its efficient operation, records were taken with the mini-vibrator

sweeping in all four orthogonal horizontal directions. A considerable volume of

redundant data was acquired, thereby permitting selection of the best quality

187

shot records for processing. Over 540 shot records were obtained during the

two surveys resulting in approximately 1.5 GB of data.

7.4.2.1 Site 1

Site 1 was located in a ploughed paddock at grid coordinates 691500mE,

6433400mN (Map Grid of Australia, 149:02:36 East, 32:13:27 South) with the

receiver lines running at a bearing of 311 degrees. Lack of vegetation and

smooth topography resulted in precise survey setup and good accessibility for

the vibrator truck. The paddock sloped uphill to the north east of the survey.

This slope was gently and was not considered to have had an adverse effect on

acquisition. A dry erosional gully ran east west to the northwest of the receiver

lines, and it necessitated careful negotiation by the vibrator truck.

Due to the ploughing, a thin layer of loose soil, up to 5 cm thick was present

above a more compacted soil. Geophone ground coupling was not affected by

this, since the geophone spike was long enough to penetrate into the more

compacted soil especially after the removal of the surrounding loose soil.

Nevertheless, there were still some local variations in the near surface soil

properties, such as the moisture content, which affected the seismic velocities.

Those variations resulted in the development of a smoothing method, in order to

minimise their effect on the determination of the seismic velocities in the

refractor.

188

Also, the coupling between the vibrator pad and the ground was not affected.

Preparation of the deep shot holes was generally not easy as the compacted

soil became hard and dry at a depth of 40 cm making drilling with the manual

auger difficult. Rain during the last 2 days of the survey made the loose soil

greasy, which had a nuisance affect on the vibrator truck’s mobility.

Receiving station 100 was positioned on the lower-right corner of the rectangle

(station 187 on the upper-left), the long axis of the rectangle being parallel to

the drainage. Photo 7-9 shows four receiver lines at site 1. Photo 7-10 shows

one of the receiver lines ready for recording. Figure 7-2 shows the geophones

and shotpoints layout. There were three LIC’s (Line Interface Cards) at this

site, which led to table 7-1 configurations. Recording length and sample interval

were 1s, 0.5 ms respectively.

Photo 7 - 9: four receiver lines at site 1.

189

Photo 7 - 10: One of the receiver lines ready for recording. Pin flags are first surveyed.

Three receiver cables are then laid out. Receivers are planted next to the pin flags and

connected to the cables. The station units are then connected to the cables.

Table 7 - 1: Trace to station relationship at site 1.

Trace number Station number Component

1 100 Vertical

2 101 Vertical

etc. etc.

116 215 Vertical

117 100 Horizontal 1

118 101 Horizontal 1

etc. etc.

232 215 Horizontal 1

233 100 Horizontal 2

234 101 Horizontal 2

etc. etc.

348 215 Horizontal 2

190

Figure 7 - 2: Geophones and shot points layout at site 1.

191

7.4.2.2 Site 2

Site 2 was recorded first. Site2 was located in grid coordinates 691000mE,

6434550mN (Map Grid of Australia, 149:01:47 East, 32:12:39 South) with the

receiver lines at a bearing of 206 degrees. The site was on a disused paddock

consisting of a number of rows of eucalypt saplings which had been planted to

promote lowering of the water table. The density and height of the vegetation

resulted in minor inconveniences in setting out geophone spread, especially

where the lines of saplings ran through the receiver lines. At many source

points the vegetation had to be removed from under the vibrator pad since it

would have adversely affected the coupling with the ground. The vibrator truck

mobility was not affected since the vegetation was generally consisted of small

sampling (Photo 7-11 and 7-12). Figure 7-3 shows the geophones and shot

points layout at site2.

The soil was hard and compacted making penetration of the geophone spike

difficult but nevertheless, still providing good geophone coupling with the

ground. Augering into the compacted soils was difficult, resulting in shallow

dynamite shot holes. Photo 7-13 shows the successful detonation of a small

charge at site2.

Receiver station 100 was positioned on the lower-right corner of the rectangle

with (station 187 on the upper-left), the long axis of the rectangle being at right

angles to the drainage. Also, there were only two useable LIC’s (Line Interface

Cards) at this site so the end of the second cable was plugged into the start of

192

the third cable. This led to the configuration shown in Table 7-2. Recording

length and sample interval were 1s, 0.5 ms respectively.

Photo 7 - 11: Minivib recording at site 2.

Photo 7 - 12: Geophones and cable deployment at site 2.

193

Photo 7 - 13: A successful detonation of a small charge at site2.

Table 7 - 2: Trace to station relationship at site 2.

Trace number Station number on

records

Station number on

ground

Component

1 100 100 Vertical

2 101 101 Vertical

etc. etc.

116 215 215 Vertical

117 100 100 Horizontal 1

118 101 101 Horizontal 1

etc. etc.

232 215 215 Horizontal 1

233 216 Not used

234 217 Not used

235 218 Not used

236 219 Not used

237 220 100 Horizontal 2

238 221 101 Horizontal 2

etc. etc.

352 335 215 Horizontal 2

194

Figure 7 - 3: Geophones and shot points layout at site 2.

195

7.5 Data Quality in Site1 and Site2

532 shots were recorded during the two surveys resulting in around 1.5GB of

data. Table 7-3 shows the number of shot recorded in site 1 and site 2.

Table 7 - 3: The number of shot recorded in site 1 and site 2.

P-Wave S-Wave Dynamite Total

Site 1 50 216 9 275

Site 2 83 171 3 257

Total 133 387 12 532

The first site has 354 traces per record (except for FFID 90-94, which have 356

traces i.e., two extra auxiliary traces as CHAN 355 and 356) as below:

CHAN 1-116 Station 100-215

CHAN 117-232 Station 100-215

CHAN 232-236 Dead

CHAN 237-354 Station 100-215

CHAN 353-354 Auxiliary

The second site has 350 traces per record, i.e. 348 data traces and 2 extra

auxiliary traces as CHAN 351&352.

The record quality achieved using the mini-vibrator varied significantly between

source points. The surface conditions on which the vibrator operated seemed

to have a significant influence on record quality with the source to receiver 196

distance playing a lesser role. Rain had adverse affects on the vertical

recording component, with the light rain that fell during the second last day

resulting in an increase in the signal to noise ratio. Strong winds induced a

swaying movement of the pin flags and vegetation, which had adverse affects

on the horizontal recording components.

Appendix G shows the assessment of data quality that including the rating from

1 to 5 in site 1 and site 2. The quality increases from 1 (bad record) to 5 (the

best record). In site 1 at about 6.5 per cent and in site 2 at about 17.7 per cent

of data were not useful to processing. Overall, 17.7 per cent of 532 shot

records do not have good quality to use. Figure 7-4 shows an example of good

shot record.

Figure 7 - 4: P-wave shot record.

197

7.6 Magnetic Data Acquisition

The magnetic survey was acquired at the Spicers Creek Catchment using a G-

858 magnetometer. At the first site the survey consisted of 18 lines, 5 m apart

and 130 m in length and at the second site consisted of 40 lines, 10 m apart

and 100 m in length. Station spacing was approximately 0.25 m along each

line.

7.7 Resistivity Data Acquisition

At the Spicers Creek Catchment, the resistivity survey was acquired using an

ABEM Terrameter. Using a Wenner configuration, resistivity measurements

were acquired along three different lines (line 1 = 240m; line 2 = 260m; line 3 =

180m) at the first site and along one line (420 m) at the second site (refer to

chapter 8).

198

Chapter 8

Processing

8.1 Summary

Processing of the Spicers Creek Catchment survey was done using a number

of software packages for magnetic (Magmap 2000), resistivity (Res2DINV) and

seismic refraction data. The main stage was the processing of seismic

refraction data by using the Seismic Un*x program, developed by the Centre for

Wave Phenomena at the Colorado School of Mines, Visual-SUNT and Rayfract

softwares.

In particular these software packages were used to generate three-dimensional

images, perform the generalized reciprocal method, calculate velocity ratios of

P- and S-waves, the picking of first-arrival traveltimes, display the shot records,

processing on the full shot records and wave eikonal traveltime (WET)

tomography images.

The major challenges of the processing of the seismic refraction data were to

reduce the very large volume of data, primarily through averaging processes,

and to generate detailed starting models for tomographic inversion. Near

surface irregularities were a major cause of inconsistencies between adjacent

199

lines. The development of the GRM SSM (statics smoothing method) is a major

achievement of this study.

8.2 Introduction

Processing of the geophysical data was done in two stages, using a number of

software packages. The first stage involved the processing of magnetic and

resistivity data collected from 2 to 7 April 2002, using MagMap 2000 and

RES2DINV software. The second stage involved the processing of the seismic

refraction data collected from 6 April 2003 to 13 April 2003, with the generalized

reciprocal method (GRM; Palmer, 1980, 1986). Processing of these data was

done using Microsoft Excel, Surfer 8, Seismic Un*x, Visual_SUNT, and

RAYFRACT software.

8.3 Processing of Magnetic and Resistivity Data

Magnetic and electrical resistivity surveys were carried out to determine the

general location of the probable shear zones. These are the most common

geophysical methods for imaging shear zones in the near surface and they are

widely used in many geotechnical and environmental investigations. The

magnetic data was processed using MagMap 2000 and the resistivity data

using RES2DINV.

200

8.3.1 MagMap 2000 Software

High resolution magnetic data were acquired at two sites at Spicers Creek using

a Geometrics G858 magnetometer. MagMap 2000 (Geometrics software,

version 4.26) is a post acquisition processing software used to edit and correct

magnetic data from several models of Geometrics magnetometers. MagMap

2000 was used to initially download the data from the magnetometer. It was

then employed to perform various editing functions, include moving lines and

positions, changing the direction of lines, rotating lines, and deleting parts of

lines or entire lines. The MagMap 2000 software also enables the performing of

diurnal corrections, plotting the acquired magnetic field, and filtering for spikes

and dropouts. The results of processing are shown in Chapter 9.

8.3.2 RES2DINV Software

Electrical resistivity imaging surveys were carried out along four lines at the two

sites at Spicers Creek with an ABEM Terrameter. The data were processed

with RES2DINV. The two-dimensional (2-D) starting model consists of a

number of rectangular blocks loosely tied to the distribution of the data points in

the pseudosection (shown in Figure 8-1).

201

Figure 8 - 1: Arrangement of the blocks used in a model together with the datum points in

the pseudosection that generate by RES2DINV software.

The distribution and size of the blocks are generated so that the number of

blocks does not exceed the number of data points whilst the depth to the bottom

row of blocks is set equal to the maximum electrode spacing which is

approximately equivalent to the depth of investigation (Edwards, 1977). A 2-D

apparent resistivity model of the subsurface is generated using a forward

modeling subroutine on the block starting model and the resistivity data

obtained from electrical imaging surveys (Griffiths and Barker, 1993).

Using a non-linear least-squares inversion routine, true resistivity-depth section

were generated from the apparent sections (de Groot-Hedlin and Constable,

202

1990), the results of which are shown in Chapter 9. The inversion routine used

by the program is a smoothness-constrained least-squares optimization

technique (de Groot-Hedlin and Constable, 1990, Sasaki, 1992) and is based

on the following equation:

(J T J + µF)d = J g T

where: F=f f + f f XTX Z

TX

f = horizontal flatness filter X

f = vertical flatness filter Z

J= matrix of partial derivatives

µ= damping factor

d= model perturbation vector

g= discrepancy vector.

8.4 Processing of Seismic Refraction Data

The processing of the seismic refraction data was performed with a number of

software packages. The processing of the seismic data took more than a year

to carry out. Due to the considerable volume of data, an initial interpretation

was carried out in order to gain an initial understanding of the problems. It then

became apparent that the expressions of the shear zone were subtle, and

careful re-processing would be required. The processing, particularly the

picking of first-arrival traveltimes was done up to four times using different

methods in order to obtain accurate values.

203

8.4.1 Microsoft Excel

Microsoft Excel is powerful spreadsheet software used to efficiently evaluate

and revise data, calculate and compare numbers, and create charts and

graphs. This software was used to process the first-arrival traveltimes and their

amplitudes. In particular it was used to determine the degree of geometric

spreading; calculate amplitude products; review and combine traveltime curves;

perform the GRM analyses including the generation of velocity analysis, time-

depth and depth graphs, and the calculation of P- and S-wave velocities and

their ratios.

8.4.2 Surfer 8

Surfer by Golden Software is a grid-based graphics program. Surfer

interpolates irregularly spaced XYZ data into a regularly spaced grid using a

number of gridding algorithms. The grid is then used to produce different types

of maps including contour, vector, wireframe, image, shaded relief, and surface

maps. Maps can be displayed and enhanced in Surfer by the addition of

boundary information, posting data points, combining several maps, adding

drawing, and annotating with text. Surfer was used to generate three-

dimensional images of the refractor velocity and depth, and velocity-depth

sections.

204

The main statistical gridding algorithm used was kriging. This method is one of

the more flexible methods and is useful for gridding almost any type of data set,

producing visually appealing maps from irregularly spaced data. Triangulation

with linear interpolation was also used. This algorithm, based on optimal

Delaunay Triangulation creates triangles by drawing lines between data points.

The original points are connected so that none of the triangle’s edges are

intersected by others. Each triangle defines a plane over the grid nodes lying

within the triangle, with the tilt and elevation of the triangle determined by the

three original data points defining the triangle.

8.4.3 Seismic Un*x Program

The Seismic Un*x (SU) package is a seismic processing and research

environment developed by the Center for Wave Phenomena, Colorado School

of Mines and is distributed free with full source code. This package operates in

the UNIX or LINUX environments. For convenience, the LINUX environment

was created with the freeware package CYGWIN which operations within the

Microsoft Windows environment (Stockwell et al., 2000). Seismic Un*x was

used to prepare the seismic data before processing, pick first-arrival traveltimes

and amplitudes, display the shot records, and perform processing on the full

shot records.

205

8.4.4 Visual_SUNT Software

The picking of the first-arrival traveltimes and amplitudes from more complex

shot records were done using Visual_SUNT software developed by W_Geosoft.

This software reads seismic data in SU format. After picking the first-arrival

traveltimes and amplitudes with Visual_SUNT, the data was transferred to Excel

for further processing.

8.4.5 RAYFRACT Software

Traveltime tomography was performed using RAYFRACT by Intelligent

Resources, a Windows 32-bit software package. The tomographic algorithm is

based on wavepath eikonal traveltime inversion (Schuster and Quintus-Bosz,

1993) and is suited for processing of seismic profiles and interpretation of both

P- and S-waves. RAYFRACT performs the Wavefront method as an optimized

version of the GRM, with laterally varying and automatically estimated receiver

separation i.e. XY distance.

Wavepath Eikonal Traveltime (WET) tomography was used to generate a

velocity-depth section of every line for each wave type from the GRM initial

models. WET tomography accommodates multiple signal paths contributing to

each first break and is a computationally efficient and geophysically robust

method (for a brief discussion on WET theory see Appendix M).

206

8.5 The Generalized Reciprocal Method

The aim of this section is to give a general overview of generalized reciprocal

method (GRM) of seismic refraction inversion as described by Palmer (1980,

1981, 1986, 1990 and 2003b). Only those parts of the GRM that have been

used for the processing of the Spicers Creek refraction data are described.

The GRM is a seismic inversion method that uses the refracted arrival

traveltime data for both forward and reverse shots, and the reciprocal time. The

GRM is able to provide good estimates of the seismic velocities within the

refractor, and the structure or shape of the refracting interface.

The major difference between the GRM and other seismic refraction inversion

methods is that the former applies a variable refraction migration which can be

often be useful in resolving undetected layers, variable velocity media and

anisotropy under favourable conditions. With the GRM, a range of offset

distances is computed, usually from zero to a value in excess of the estimated

true value. The optimum value, for which the refractor velocity analysis function

displays a minimum of artifacts caused by the irregular refractor interface, is

selected by inspection of the graphs.

The GRM consists of two algorithms: the velocity analysis function from which

the refractor velocity is derived, and the time-depth function which is a measure

of the depth to the refractor in units of time. The two functions are calculated for

different values of the horizontal separation between forward and reverse

207

receivers (XY values), with the optimum value being determined from inspection

of the resulting functions.

8.5.1 Velocity Analysis Function

The GRM refractor velocity analysis function, t , is defined by: V

( )FRRXFYV tttt +−= 2/1 (1)

This equation is applied to pairs of reversed traveltimes at the receivers X and Y

which are located between the forward and reverse shot points F and R from

the same refractor. The computed value is referred to G which is the midpoint

between X and Y. The raypaths used to form this equation are shown

schematically in Figure 8-2.

For moderately dipping refractor, the traveltimes from Q to Y and from P to X

are approximately equal and hence cancel. Therefore, tV reduces to the

average of the traveltimes from F to P and from F to Q, or approximately from F

to R. This approximation is more reasonable when the XY separation is such

that P and Q are very close together. Since t is computed for successive

receiver locations, the increments represent the additional time required to

travel the additional receiver spacing in the refractor, and hence provides a

method for determining the seismic velocity within the refractor.

V

208

Figure 8 - 2: An illustration of the ray paths of the velocity analysis function. Ray paths

are shown as solid lines when they are added, broken lines when they are subtracted and

as broken-dashed lines when they cancel (Palmer, 2003b).

8.5.2 Time-Depth Function

Following the determination of the refractor velocity, the next step of the GRM

involves the computation of the time-depth function. The time-depth function t

is defined by:

G

( )( )nFRRXFYG VXYtttt /2/1 +−+= (2)

where

Vn = Seismic velocity in the refractor

As with the velocity analysis function, the time-depth is computed for pairs of

reversed traveltimes at the receivers X and Y, which are located between the

shot points F and R from the same refractor. The computed value of t is G

209

referred to G which is the midpoint between X and Y. The ray paths used to

form this equation are shown schematically in Figure 8-3. The ray paths that

remain are from the refractor to the receivers, namely from P to X and from Q to

Y, minus the normal projection of those ray paths on the refractor at the

refractor velocity.

Figure 8 - 3: A schematic summary of the ray paths used in the computation of the time-

depth function for GRM (Palmer, 2003b).

The XY/V term can also be accommodated by approximating it with the

difference between the velocity analysis function at the X and Y locations. This

approach accommodates all velocity variations, including the artifacts related to

non-optimum XY values, and as a result can emphasize the artifacts related to

non-optimum values.

n

The optimum value is taken as that XY separation for which the velocity

analysis function shows a minimum of lateral inhomogeneities and the time-

210

depths show a minimum of smoothing of the refractor topography for data which

have been corrected for near surface effects. At the optimum XY value, both

the forward and reverse rays are assumed to have been critically refracted at or

very near the same point on the refractor.

The optimum XY value for the analysis of refractor velocity is determined as

being that value which gives a time-depth function most closely approximating a

series of straight lines, that is, there is a minimum of artefacts generated by an

irregular refractor. Where there are lateral variations in the refractor velocity,

the optimum value of XY may differ with distance. For the selection of the

optimum XY value for the time-depth function, the function which exhibits most

detail should be chosen. In general, the optimum XY value is usually the same

for both the velocity analysis and time-depth functions (Reynolds, 1997).

However, Palmer (2003b) showed the determination of the correct XY value is

not simple task, because the time differences for most studies show that the

time differences computed with the algorithm for the various XY values are

generally very small, and frequently less than a millisecond for common near

surface targets.

The GRM explicitly recognizes and accommodates for the offset distance, by

selecting from inspection, the XY value for which the forward and reverse rays

leave the refractor at approximately the same point, rather than arriving at the

same receiver (Figure 8-4). Using this, the GRM can also compute an average

velocity of all layers above the target refractor with the equation:

211

V=nG

n

VtXYXYV

2

2

+ (3)

_

The GRM is unique in being able to determine both average and individual layer

velocities. Averaging the refractor velocity analysis function for a range of XY

values centered on the optimum, can minimize artifacts related to lateral

variations in depth to and seismic velocity within the refractor.

Figure 8 - 4: The process of refraction migration. The offset distance (XG and YG) is the

horizontal distances between the point of emergence of the ray on the refracting

interface and the point of detection at the surface (top). Inversion methods which do not

explicitly recognise the offset distance suffer refractor smoothing. The GRM

accommodates the offset distance by calculating different values of horizontal distance

between forward and reverse receivers, then selecting the XY value for which the forward

and reverse rays leave the refractor at the same point and arrive at different detector

positions at the surface (bottom).

212

8.6 Procedure of the Data Processing

8.6.1 Pre-processing

Before the seismic data could be processed, a number of steps had to be

performed. Firstly the data had to be converted from SEG-Y file format to SU

file format, in order to be read by Seismic Un*x. SEGY data format consists of

three parts. The first part is a 3200 byte EBCDIC card image header which

contains 40 cards images of text data describing the tape. The second part is a

400 byte binary header containing information about the contents of the tape

reel. The third portion of the SEGY format consists of the actual seismic traces.

Each trace has a 240-byte trace header (Stockwell et al., 2000).

The SU data format is based on the trace portion of the SEGY format. The

primary difference between the SEGY traces and SU traces is that the data

portion of the SU format are floats, written in the native binary format. Next,

since SEG-Y files may use the unassigned portion of the trace headers which

SU also uses, it was necessary to zero out those fields meaningful to SU. Also

of the 352 traces in each record, four were auxiliary and were removed.

The next step in the preparation of the data was to generate the individual shot

records for each component. This was done using the SU shell script ‘shots’

(which is shown in Appendix H together with the other shell scripts and C

programs used). The 29 receivers on each line are numbered 100-128, 200-

228, 300-328 and 400-428 for lines 1, 2, 3 and 4 respectively. The resultant

shot records have the four receiver lines wrapped continuously in a ‘W’ shape,

213

where the receivers for lines 2 and 4 are in reverse order. The time interval was

from 0 to 0.2 s for the vertical component and 0 to 0.3 s for the horizontal

components. Figure 8-5 shows a shot record.

Once each shot record component was generated, the geophone locations in

their headers were set using the SU shell script ‘setgx’. Care was taken to

ensure that the geophone numbers were correctly assigned to the order of the

trace field number (see Appendix I). After the geophone locations were set the

seismic data was ready for processing.

8.6.2 Refraction Convolution Sections (RCS)

The refraction convolution sections (RCS) is a method for imaging shallow

refraction data which employs full trace processing by the convolution of

forward and reverse shot records (Palmer, 2001c). The benefits of using this

method are (Palmer 2001b):

The S/N ratio shows less variation.

Improves the resolution of lateral variations in both amplitudes and time-depth.

Results in more detailed geological models of the subsurface through the

convenient use of amplitudes and traveltimes.

Increased resolution of ambiguities in refractor models and azimuthal

anisotropy.

214

0

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

Tim

e (s

econ

ds)

3.364 3.366 3.368 3.370 3.372x104Trace Number

Record 497 P

Line 1 Line 2 Line 3 Line 4

428400 328300200128 228100

Figure 8 - 5: Vertical shot record. The receivers on lines 1, 2, 3 and 4 are numbered 100-

128, 200-228, 300-328 and 400-428 respectively. The four receiver lines are wrapped

continuously in a ‘W’ shape, where the receivers for lines 2 and 4 are in the reverse

order.

215

It was necessary to split each shot record into the four individual receiver lines

and to reverse the trace order of the second and fourth lines. These two stages

were done using the SU shell script ‘linebreaker’ (see Appendix H). Figure 8-6

shows the result of this operation on the shot record in Figure 8-5. The

individual lines have been separated where the receivers on lines 1, 2, 3 and 4

are numbered 100-128, 200-228, 300-328 and 400-428 respectively. After

splitting the shot records into the four individual lines it was possible to generate

refraction convolution sections (see Appendix J).

The refraction convolution sections were generated by using the SU shell script

‘convolve’ which uses the SU program ‘surefcon’ to convolve pairs of forward

and reverse shot records. Figure 8-7 shows the results in the structure and

amplitudes of the refractor being displayed in a time cross-section.

Shot records were generated for horizontal sources of opposite polarity

enabling the enhancement of the S-wave energy. This operation was

performed by subtracting seismic records of opposite polarity, resulting in the

cancellation of P-wave components whilst leaving the sum of two S-wave

components (Helbig, 1986). Figure 8-8 shows the S-wave enhancement by

using the SU shell script ‘sumdiff’ (see Appendix H). These S-wave enhanced

waves are especially evident in the orthogonal components which generally had

poor convolution sections before this process.

216

0

0.05

0.10

0.15

Tim

e (s

econ

ds)

100 105 110 115 120 125Geophone

Record 497 P

0

0.05

0.10

0.15

Tim

e (s

econ

ds)

200 205 210 215 220 225Geophone

Record 497 P

0

0.05

0.10

0.15

Tim

e (s

econ

ds)

300 305 310 315 320 325Geophone

Record 497 P

0

0.05

0.10

0.15

Tim

e (s

econ

ds)

400 405 410 415 420 425Geophone

Record 497 P

Figure 8 - 6: The shot record in Figure 1 split into individual receiver lines. The receivers

on lines 1, 2, 3 and 4 are numbered 100-128, 200-228, 300-328 and 400-428 respectively.

The trace order of lines 2 and 4 have been reversed.

217

0

0.05

0.10

0.15

Tim

e (

seconds)

400 405 410 415 420 425Geophone

RCS, Records 497 and 510 P

0

0.05

0.10

0.15

0.20

0.25

0.30

Tim

e (

seconds)

300 305 310 315 320 325Geophone

RCS, Records 415 and 432 Sh2

Figure 8 - 7: Refraction convolution section for a vertical (left) and orthogonal horizontal

(right) component showing the refractor structure and amplitudes proportional to the

square of the head coefficients.

218

0

0.05

0.10

0.15

0.20

0.25

0.30

Tim

e (s

econ

ds)

100 105 110 115 120 125Geophone

Record 323 Sh1

0

0.05

0.10

0.15

0.20

0.25

0.30

Tim

e (s

econ

ds)

100 105 110 115 120 125Geophone

Record 272 Sh1

0

0.05

0.10

0.15

0.20

0.25

0.30

Tim

e (m

s)

100 105 110 115 120 125Geophone

Record 323 minus 272 h1

0

0.05

0.10

0.15

0.20

0.25

0.30

Tim

e (m

s)100 105 110 115 120 125

Geophone

Record 323 plus 272 h1

Figure 8 - 8: The enhancement of S-wave in horizontal records using records of opposite

polarity. Subtraction results in the sum of the S-wave components and cancellation of P-

wave components (bottom left) whilst addition results in the opposite (bottom right).

219

8.6.3 First-Arrival Picking

Before picking first arrivaltimes, two procedures which muted and killed traces

were performed to improve of the first-arrival picks. By using muting, traces can

be zeroed out either before or after the first-arrivals, whilst killing enables the

removal of entire traces. Muting was done by the shell script ‘mute’ and killing

was done by the SU command ‘sukill’. Figure 8-9 shows the result of killing and

muting traces for one example shot record.

0

0.05

0.10

0.15

0.20

0.25

0.30

Tim

e (

seco

nds)

3.388 3.390 3.392 3.394 3.396 3.398x104Trace Number

Record 96 h2

0

0.05

0.10

0.15

0.20

0.25

0.30

Tim

e (

seco

nds)

3.388 3.390 3.392 3.394 3.396 3.398x104Trace Number

Record 96 h2 muted

Figure 8 - 9: Muting and killing traces can improve the quality of first-arrival picking. The

original cross line horizontal component shot record with considerable noise and dead

traces (left). The same shot record after killing and muting above the first-arrivals (right).

220

After muting and killing, first arrivals from the shot records were picked to obtain

traveltimes by using a series of shell scripts in SU. The shell script ‘pics’ was

employed to display the individual shot records. The shell script ‘amps’ was

used to interpolate the maximum positive and negative amplitudes and their

times over the record within a specific time window, which was 20 ms, starting

at those points generated by ‘pics’ shell script. The quality of the interpolating

procedure in picking first-arrivals was checked using ‘plotAmps’ which displays

the shot records with first-arrivals overlayed.

Figures 8-10 show the result of the ‘plotAmps’. They show the varying quality of

the picking procedure in determining first-arrival traveltimes. In the record on

the top, the maximum negative first-arrivals (in white), form a smooth line, whilst

the maximum positive first-arrivals (in black) form an irregular line. In the record

on the bottom both the maximum negative and positive first-arrivals form fairly

smooth lines, although some irregularities do occur, such as between trace

numbers 5880 and 5900, and at 5920.

Due to poor record quality, picking first-arrivaltimes using Seismic Un*x was

often difficult (especially for many S-wave records). Instead Visual_SUNT was

used to pick the first-arrival times for those shot records with a complex set of

head-waves. With Visual_SUNT traces can be viewed individually, together

with every time increment in the sampling rate and their amplitudes. Once a

whole shot record was picked by this way, which was a lengthy process, the

quality of the picking could be checked using ‘plotAmps’ shell script. In some

shot records, both Seismic Un*x and Visual_SUNT were used.

221

The majority of the first-arrival processing was performed on the shot records

from inline shots. Also shot records from offset shots, that is shot points outside

the receiver spread were used more often since it ensured that these arrivals

were refracted and not direct first-arrivals.

Figure 8 - 10: The picked first-arrival traveltimes overlayed on the shot records.

222

8.6.4 Traveltime Graphs

The first arrival traveltimes for each wave type (P-, SH-, SV-waves) were plotted

against distance to produce traveltime graphs. These graphs were then

checked to ensure that the minimum time of each in-spread traveltime graph

coincides with the shot point. The intercept times at each in-spread shot point

were also determined from the traveltime graphs using parallelism (Sjogren,

1980).

The traveltime graphs were distorted since the source lines were offset from the

receiver lines (Figure 8-11), due to the impracticality of operating the mini-vibe

next to the receivers. The distorting effect was most prominent for the receivers

near the source points, where the distance X along the line and Y the lateral

offset from the line are comparable, especially when X was less than Y. As a

result the traveltimes of direct arrivals were distorted significantly, such as in

Figure 6-12 where the traveltime graphs around the source points are shifted up

in time by over 20 ms and have a broad concave upward shape. For the

refracted arrivals the receivers are far from the source points, with Y being

much less than X, thus the distorting effect was less prominent on the

traveltimes of these arrivals.

The traveltime curves were corrected for the distortion effect using geometric

proportioning. Each traveltime was multiplied by the proportioning factor X/L,

where L is the direct and X is the inline distance between the source and

receiver for that traveltime (Figure 8-11). The factor X/L has a value of 0 when

223

the receiver and source are at the same inline point, that is X=0 and L=Y, and

approaches a limit of 1 when the receiver is far from the source, that is X>>Y

and L→X. The effect of this factor was thus greatest for the direct arrivals and

diminished progressively on the further refracted arrivals. In Figure 8-13 the

traveltime graph of Figure 8-12 has been corrected. The traveltime graphs at

the source points now occur at zero time and they do not have the broad

concave shape. Also, the refracted arrivals remain largely unaffected.

The traveltime graphs were corrected for each wave type (P-, SH-, SV-waves)

totalling 24 traveltime graphs. These corrected traveltimes were then used for

processing with the GRM (Figures 8-14 and 8-15).

Figure 8 - 11: Survey geometry with the source line offset from the receiver line. For a

source point s and receiver r, when X is less than Y or when they are comparable there is

a significant distortion of the traveltime curve at that point. When X is much greater than

Y the distortion is negligible.

224

In order to reduce the amount of time required to perform the GRM on all the

pairs of offset traveltime graphs, phantoming was used to combine the forward

traveltime graphs and to combine the reverse traveltime graphs. The

advantage of this process was that the GRM was applied only once for each

component-line combination by using the averaged forward and reverse

traveltime graphs.

0

10

20

30

40

50

60

70

80

0 20 40 60 80 100 120 140distance (m)

trave

ltime

(ms)

S6

S7

S8

S9

S10

S11

S12

S13

Figure 8 - 12: Traveltime graph for shot points inside the receiver spread. Since the shot

points are offset from the receiver line, direct arrival traveltimes are shifted up in time

and have a concave upward shape.

225

0

10

20

30

40

50

60

70

80

0 20 40 60 80 100 120 140distance (m)

trave

ltime

(ms)

S6

S7

S8

S9

S10

S11

S12

S13

Figure 8 - 13: Corrected traveltime graphs for shot points inside the receiver spread. By

geometric proportioning the traveltimes are changed so that the first arrival at each shot

point occurs at zero time.

Traveltime Graph- Site1, Pwave Line1 (S1-S18)

0

20

40

60

80

100

120

0 25 50 75 100

125

Distance (m)

Tim

e (m

s)

tminS1tminS2tminS3tminS4tminS5tminS6tminS7tminS8tminS9tminS10tminS11tminS12tminS13tminS14tminS15tminS16tminS17tminS18

Figure 8 - 14: Traveltime graph for P-wave (site1, line1) before correction.

226

Corrected Traveltime(P) -Site1, Line 1

0

10

20

30

40

50

60

70

80

90

100

0 20 40 60 80 100 120 140

Distance (m)

Tim

e (m

s)

S1

S2

S3

S4

S5

S6

S7

S8

S9

S10

S11

S12

S13

S14

S15

S16

S17

S18

Figure 8 - 15: Corrected traveltime for P-wave (site1, line1) after correction.

Phantoming involves shifting each traveltime graph by the average separation

between itself and the nominated base traveltime graph, then taking the

average of all these graphs. This is shown in Figure 8-16 where the traveltime

graph of shot 1 has been shifted downwards by the average separation

between itself and shot 3. Similarly the traveltime graph of shot 2 has been

shifted downwards by the average separation between itself and shot 3. The

shifted traveltime graphs for these shots (1 and 2) and that of shot 3 are then

averaged to give the combined traveltime graph. The reciprocal times for each

pair of combined offset traveltime graphs were determined using the standard

reciprocal time formula which uses an intermediate reciprocal time (Palmer,

2003b).

227

40

50

60

70

80

90

0 20 40 60 80 100 120 140distance (m)

trave

ltime

(ms)

shot 1

shot 2

shot 3

combined

Traveltime curves

are shifted

Figure 8 - 16: Using phantoming to combine traveltime graphs. Each traveltime curve is

shifted downwards by the average separation between itself and the base traveltime

graph. The average of these graphs is then taken to obtain the combined traveltime

graph.

8.6.5 The GRM Refractor Velocity Analysis Function

The GRM refractor velocity analysis function, computed with shots 1 and 18 for

the P-wave arrivals for line 1 at site 1, using equation 1, is shown in Figure 8-

17. XY values from –10 m to 20 m were used for the following means.

The graphs in Figure 8-17 show that the values computed with various XY

values are different between 0 m and 60 m. Although the range of the

variations is only, a few milliseconds at most, it is still possible to fit a wide

228

range of seismic velocities to these computations. These velocities are

artefacts, related to the change in depth to the refracting interface, rather than

true lateral changes in the seismic velocities in the refractor.

Velocity Analysis (P), Site1, Line1, S1-S18

25

35

45

55

65

75

0 20 40 60 80 100 120 140Distance (m)

Tim

e (m

s)

XY=-10mXY=-5 m

XY=0 m

XY=5 m

XY=10 m

XY=15 m

XY=20 m

Figure 8 - 17: Forward and reverse velocity analysis graph for line1, site1, P-wave.

Figure 8-18 shows the application of equation 1 to the same data, but with

different reciprocal times for each set of XY values, in order to separate the

computation for clarity. It is clear that the graphs are symmetrical about the

values computed with XY=5m. In order to remove the effects of the artefacts

generated with non-optimum XY values, the complete set of computations has

been averaged.

229

20

25

30

35

40

45

50

55

60

65

70

75

0 20 40 60 80 100 120 140

Distance (m)

Stac

ked

Velo

city

Ana

lysi

s (m

s)

XY=-10 m

XY=-5 m

XY=0 m

XY=5 m

XY=10 m

XY=15 m

XY=20 m

Figure 8 - 18: The velocity analysis function calculated for different XY values and

stacked. The optimum XY value can be determined by observing for which XY value the

others are symmetric about, in this case a XY value of about 5 m.

The averaging operation is most effective, where the range of XY values is

centred on the optimum value, which in this case is 5 m. Furthermore, the

averaging operation is also more effective, where there is a reasonable number

of values is moderate to large. Therefore, the velocity analysis function was

evaluated over the range -10m<XY<20m. Although there is probably very

limited geophysical significance in the negative XY values, nevertheless, their

use generates a more useful average. Figure 8-19 compares the average with

the velocity analysis values computed with a 5 m value. The agreement is quite

close.

230

Velocity Analysis Graph (P-wave), site1, Line1, XY=5, S1-S18

2530354045505560657075

0 20 40 60 80 100 120 140

Distance (m)

Tim

e (m

s)

XY=5 m

Averages

Figure 8 - 19: Forward and reverse velocity analysis graph for XY=5, P-wave, line1, site1.

Residuals were obtained by subtracting the average velocity analysis function

from the velocity analysis functions calculated with the different XY values. The

residuals were quite small, mostly less than 1 ms indicating the relatively small

variations in the different velocity analysis functions. In Figure 8-20 they are

generally less than 1 ms and are less than 0.25 ms between 65 and 95 m. The

residuals are spatially correlated and symmetry is apparent with the values

derived from XY values greater than the optimum show the opposite trend to

those derived from XY values less than the optimum. This symmetry confirms

the selection of an optimum XY value of between 2.5 and 5 m.

Residual velocity analysis functions for P- and S-waves are shown in Figure 8-

21. There is a good symmetry for in residuals between 60 m and 110 m.

231

Although they show reasonable changes in the refractor velocity analysis

function, they can very systematically and the optimum XY can be considered to

be about 5 m.

-1.50

-1.00

-0.50

0.00

0.50

1.00

1.50

0 20 40 60 80 100 120 140

Distance (m)

Velo

city

Ana

lysi

s R

esid

uals

(ms)

XY=-5 mXY=0 mXY=5 mXY=10 mXY=15 m

Figure 8 - 20: Velocity analysis residuals are obtained by subtracting the average of the

velocity analysis functions from the individual velocity analysis functions. Variations

between the velocity analysis functions for the different XY values are generally less than

1.5 ms.

232

Residuals Velocity Analysis(P-wave) -Site1, Line 1, S1-S18

-1.5

-1

-0.5

0

0.5

1

1.5

0 20 40 60 80 100 120 140

Distance (m)

Tim

e di

ffere

nce

(ms)

XY=-10 m

XY=-5 m

XY=0 m

XY=5 m

XY=10 m

XY=15 m

XY=20 m

Residuals Velocity (H2N) - Site1, Line 1, S1-S18

-3

-2

-1

0

1

2

3

4

5

0 20 40 60 80 100 120 140

Distance (m)

Tim

e di

ffere

nce

(ms)

XY=-10 m

XY=-5 m

XY=0 m

XY=5 m

XY=10 m

XY=15 m

XY=20 m

ResidualsVelocity (H1W)- Site1, Line 1, S1-S18

-5

-4

-3

-2

-1

0

1

2

3

4

5

6

0 20 40 60 80 100 120 140

Distance (m)

Tim

e di

ffere

nce

(ms)

XY=-10 m

XY=-5 m

XY=0 m

XY=5 m

XY=10 m

XY=15 m

XY=20 m

Figure 8 - 21: Velocity analysis residual for P- and S-waves, site1.

233

8.6.6 Corrections for Surface Effects

The averaging of the refractor velocity analysis function over the range of XY

values which were symmetrical about the mean, in this case –10m < XY < 20m,

was quite successful in removing artefacts which were caused by changes in

the depth to the main refractor. However, there were still some inconsistencies

in determining the seismic velocities in the refractor, between adjacent lines and

they appeared to be related to surface irregularities (Figure 8-22; Palmer, 1986,

p.107-111).

Velocity Analysis Graph (P-wave) -Site1, Line 1, S1-S18

25

30

35

40

45

50

55

60

65

70

75

0 20 40 60 80 100 120 140

Distance (m)

Tim

e (m

s)

XY=0 m

XY=5 m

XY=10 m

Averages

near surface effect

Figure 8 - 22: Velocity analysis graph that show near surface effect.

234

The definition of any surface irregularities is normally straightforward with the

standard 2-D methods of profiling, in which explosives sources are used. It

involves using traveltimes from sources close to the receivers. However, with

3D methods, it can be impractical to record shot points along every recording

line. Furthermore, the offset between the receiver line and the source line

which is an integral part of surface sources, together with the lack of useful

arrivals for receivers close to the source, results in a lack of suitable traveltimes

to make the corrections. Accordingly, it was necessary to develop an

alternative approach which was cognizant of the realities of 3D data acquisition

with surface seismic sources.

The new method developed as part of this study, takes advantage of a unique

feature of the GRM. The time-depths computed with a range of XY values

generally do not exhibit large differences. Therefore, an average of the time-

depth for a range of XY values, is usually similar to that computed with the

optimum XY value.

However, the anomalies caused by surface irregularities propagate laterally

throughout the time-depths (Palmer, 1986, p.107-111). When the time-depths

are averaged for a range of XY values, the time depth anomalies are averaged

and reduced. Therefore, if the average of the time-depth for a range of XY

values is subtracted from the time-depth for zero XY, the result is a good

estimate of the time-depth anomalies caused by the surface sources. This

difference is then subtracted from the traveltime graphs, thereby affecting a

good correction for the surface irregularities. In general, this method

235

significantly improved the reliability and consistency in determining seismic

velocities in the refractor. Figures 8-23 and 8-24 show the velocities on the

velocity analysis graph before and after correcting near surface effects

respectively.

Average Velocity (P-wave), Before Correcting, Site1, Line1,S1-S18

V1=5515 m/s

V2=3932 m/s

V3=3681 m/s

V4=3055 m/s

0

10

20

30

40

50

60

70

80

0 20 40 60 80 100 120 140

Distance (m)

Tim

e (m

s)

V1V2V3V4

Figure 8 - 23: Velocity analysis graph before correction. Notice to the line trends for V2

and V3.

236

Average Velocity Analysis (P-wave), After correcting, Site1, Line1 , S1-S18

V1=5417 m/s

V2=3700 m/s

V3=3091 m/s

0

10

20

30

40

50

60

70

80

0 20 40 60 80 100 120 140

Distance (m)

Tim

e(m

s) V1

V2

V3

Figure 8 - 24: Velocity analysis graph after correction. Notice to the line trends for V2 and

V3.

8.6.7 Seismic Velocities in the Refractor

Appendix K shows refractor velocities for all the average velocity graphs. The

determination of the seismic velocities in the refractor was a four-stage process,

which aimed to use all of the data, and to drive values which were free from

artefacts. In the first stage, the traveltimes from different source points were

phantomed then averaged. This process generated traveltimes which were

largely free of picking random errors and which therefore, were quite accurate.

237

In the second stage, the time-depth algorithm of the GRM was used to generate

corrections for any near-surface irregularities. These corrections minimised the

likelihood of generating artefacts related to the near-surface irregularities. In

the third stage, the velocity analysis algorithm of the GRM was evaluated then

averaged over a range of XY values. This averaging process effectively

minimised the likelihood of generating artefacts related to variations in the depth

to the refractor.

In the final stage, the seismic velocities in the refractor were obtained by linear

regression. However, the selection of the lateral intervals over which the

analysis were to be performed, was made by visual inspection of the refractor

velocity analysis function after the first three stages had been applied. This

stage was potentially the most subjective, as is shown by Figure 8-25, in which

there is a large number of possible lateral variations in the refractor S-wave

velocity. However, the application of the first three stages resulted in graphs of

the computed refractor velocity functions which consisted of a relatively small

number of lateral variations with quite distinct changes. As a result, the

confidence in determining the number of lateral changes, and the seismic

velocities is considered to be quite high.

It must be restated that the determination of accurate seismic velocities in the

refractor is an ill posed problem because very small variations in the refractor

velocity analysis function can result in significant variations in the computed

velocities. The four-stage process described above provides an extremely

238

useful approach for minimizing the generation of artefacts from a number of

sources.

55

65

75

85

95

105

115

125

135

0 20 40 60 80 100 120 1distance (m)

aver

age v

elocit

y ana

lysis

(ms)

40

1271

1864

1468

19103270

2737

Figure 8 - 25: Lateral variations in S-wave velocities (in m/s) within the refractor.

The final step was to compute the residuals by subtracting the average velocity

analysis function from the velocity analysis functions calculated with the

different XY value. This calculation was performed for four lines (in-line) at site

1 and site 2. The results show that the residual analyses for all in-line lines

were quite small, mostly less than 1 ms, thereby indicating the relatively small

variation in the different velocity analysis function.

Figure 8-20 shows that the residuals are generally less than 1 ms and are less

than 0.25 ms between 65 and 95 m. The residuals are symmetrical about the

optimum XY, which is seen to be between 3.5 and 5 m.

239

8.6.8 The GRM Time-depth and Depth Function

The GRM time-depth function was calculated for a range of XY values from –10

to 20 in 5 m increments by using equation (2) as described in section 6-4-2.

The time-depth was then compared with the half intercept time, the two being

mathematically equivalent. The intercept time is the parameter obtained by

extrapolating the segment of the traveltime graph with arrivals from the

refractor, back to the shot point. Whilst the two measures had a similar trend in

the majority of cases, the time-depth was often displaced vertically to a greater

time. This was attributed to errors in the estimation of the reciprocal time.

These errors were corrected by adjusting the time-depths to better reflect the

half intercept times which are computed independently of the reciprocal time.

As Figure 8-26 shows, if only the traveltime at the shot point is required, that is

the intercept time, then it can be computed with equation, from the difference

between the traveltime to the shot point S from an offset shot point P, minus the

average of the differences between the two sets of refracted arrivals (Palmer,

1986, p.48-51)

T = T -(T -T GS ) (4) S PS PG

The time-depth functions were stacked to determine how the shape of the

functions varied with the different XY values. All the XY values show essentially

the same time-depths in respect to general trends and absolute value.

However the resolution of small irregularities were maximised with the optimum

240

XY value. This can be seen in Figure 8-27, in particularly at 20, 55 and 120 m,

where the irregularities are most pronounced for an XY value of 5. The small

variations in the time-depths were further verified with the residuals which were

computed in the same way as the velocity analysis residuals and were usually

less than 1.5 ms, an example of which is shown in Figure 8-28.

Figure 8 - 26: Traveltime of a refracted arrival at the shot point (S) (Palmer, 2003b).

After calculating the time-depth, the absolute depth needs to be calculated. For

this reason, the seismic velocities above the refractor and the depth conversion

factor (DCF) must be determined. The average seismic velocity above the

target refractor was calculated using equation (4). The depth conversion factor

(DCF) is the expression with units of velocity which relates time-depth to layer

thickness (5).

DCFtZ G=1 (5)

241

0

5

10

15

20

25

30

0 10 20 30 40 50 60 70 80 90 100 110 120 130 140

Distance (m)

Stac

ked

time-

dept

h (m

s)

XY=-10

XY=-5

XY=0

XY=5

XY=10

XY=15

XY=20

Figure 8 - 27: The time-depth function calculated for different XY values and stacked. The

optimum XY value can be determined by observing for which XY value small

irregularities in the time-depth function are most pronounced, in this case a XY value of

about 5 m.

-1.5

-1

-0.5

0

0.5

1

1.5

0 20 40 60 80 100 120 140

Distance (m)

Tim

e-de

oth

resi

dual

(ms)

XY=-10 mXY=-5 mXY=0 mXY=5 mXY=10 mXY=15 mXY=20 m

Figure 8 - 28: time-depth residuals are obtained by subtracting the average of the time-

depth functions from the individual time-depth functions. Variations between the time-

depth functions for the different XY values are generally less than 2 ms.

242

The DCF’s exact form requires all of the angles of refraction to be known. This

can be difficult when the refractor is irregular and dip angles change

unpredictably between successive receiver locations. Instead horizontal layer

approximations of the DCF are used. The horizontal layer approximation of the

two layer DCF was evaluated at each point along the profile using the refractor

velocity and average above refractor velocity (6).

21

21

VV

VVDCFn

n

−= (6)

The depth to the refractor was calculated by multiplying the time-depth function

of the optimum XY value with the depth conversion factor at each point (6). In

order to avoid large discontinuities in the depth profile caused by lateral

changes in the DCF, the DCF was smoothed around these changes.

For all traveltimes over all the lines for both sites, the GRM velocity analysis and

time-depth algorithm were computed to give detailed image of the refractor.

The final results in regard to absolute depth for both sites show consistent

results for P- and S-waves. In general, the first site exhibits a deeper refractor

(between 13-17 m) than at the second site (between 9-12 m). Appendix L

shows all time-depth and depth graphs for the first and second site.

243

8.6.9 Processing of First Arrival Amplitudes

Processing of the first-arrival amplitudes are useful in seismic refraction studies

because they can indicate the gross velocity model and facilitate the convenient

recognition of amplitude variations related to changes in the seismic velocities

in the refractor (Palmer, 2001e). As Palmer (2001b) concluded the higher the

contrast in the refractor wave speed the lower the head wave amplitude and

vice versa. The expression for calculating the amplitude of the head wave for a

thick refractor with a plane horizontal interface is showed by Heelan (1953) and

Werth (1967).

Amplitude=KF(t) / (rL 3 ) 2/1

Where K is the head coefficient, which is a function of the contrasts in seismic

velocities and/or densities between the upper layer and the refractor. F(t) is the

displacement potential of the incident pulse, r is the source to receiver distance,

and L is the distance the wave has traveled within the refractor.

The geometric spreading is defined as the component of the variation in the

amplitude of the refracted arrivals which can be attributed solely to the shot-to-

receiver distance (Palmer, 2003a):

Geometric spreading component = 1/((rL ) ) 3 2/1

244

For most shallow seismic refraction surveys, the geometric spreading

component dominates the head wave amplitude, and thus must be removed in

order to recognize amplitude variations related to changes in refractor.

The expression for K given by (Werth, 1967) is:

K= 22212

222/12

222

21

])21()21([)1()]2()21([2

γχλρχλργγγγρλγλρχγ

mmmm

+++−+−−++

where:

21 / PP VV=γ

21 / ρρρ =

)/()/( 21

22

21

21 PSPS VVVVm −= ρ

2/12 )1( γχ −=

2/1221

211 )/( γλ −= SP VV

2/1222

212 )( γλ −+= SP VV

V = compressional wavespeed in upper media 1P

V = shear wavespeed in upper media 1S

1ρ = density in upper medium, and similarly for the lower medium 2.

For strong contrasts in wavespeeds, ie for γ ⎯→⎯ 0,

K α 2211 / pp VV ρργρ =

The evaluation of K for a selected set of elastic parameters shows that the

amplitude decreases as the contrast in the seismic velocity between two media

increases (Cerveny and Ravindra, 1971, O'Brien, 1967).

245

In order to recognize any amplitude variations related to the head coefficient, it

was necessary to reduce the effect of geometrical spreading. This was done by

computing the amplitude product for each forward and reverse shot pair

(Palmer, 2003b). The product of the forward and reverse shot amplitudes

largely compensates for the effects of geometric spreading and dipping

refractors (Palmer, 2003b, Palmer, 2001b). The amplitude product was

calculated only for the P-waves since the S-waves did not show good

consistency.

The effectiveness of this operation can be seen in Figure 8-29. The Figure

shows the large amount of geometric spreading from a shot pair 20 m from the

receiver spread, with the amplitudes decaying much faster than the reciprocal of

the distance squared. The product of this shot pair as well as a number of other

offset shot pairs is shown in Figure 8-30. The removal of the geometric

spreading component is most effective between 20 and 100 m. Also there are

gross similarities in the shape of the amplitude products for all the shot pairs.

This suggests that the variations are related to the head coefficient and hence

to variations in the seismic velocities in the refractor.

246

Maximum Amplitude Shots 5 & 14 - Line 1

0

10

20

30

40

50

60

70

80

90

100

20 40 60 80 100 120 140 160D ist ance ( m)

1/x (Forw ardDistance)1/X^2 (Forw ardDistance)1/X^3 (Forw ardDistance)1/X (ReverseDistance)1/X^2 (ReverseDistance)1/X^3 (ReverseDistance)Forw ard AmplitudeProductReverse AmplitudeProduct

Figure 8 - 29: Forward and reverse distance P-wave amplitudes for shots 5 & 14 with

calculating 1/X, 1/X^2 and 1/X^3. The large geometric spreading component dominates.

The effects of geometric spreading have been reduced.

Amplitude Products

0

20

40

60

80

100

0 20 40 60 80 100 120 140distance

scale

d amp

litude

prod

uce (

%)

Shots 5 &14

Shots4&15

Shots1&18

Shots2&17

Shots 3&16

Figure 8 - 30: The product of forward and reverse amplitudes for various offset shot pair.

There are gross similarities in the shape of the amplitude products for all the shot pairs,

suggesting that the variations are related to variations in the seismic velocities in the

refractor.

247

8.6.10 Velocity Ratios

Shear wave (S-wave) data can be readily combined with the interpretation of

conventional seismic compressional-wave (P-wave) data. After determining the

seismic velocities for both P- and S-waves, it was possible to calculate the

velocity ratio Vp/Vs or Poisson’s ratio along each profile. Although Poisson’s

ratio (σ) is not a constant of proportionality between stress and strain, it can be

expressed in terms of elastic constants, and is often used in the same context

as the elastic constants (Tatham, 1985).

The ratio Vp/Vs is defined in terms of Poisson’s ratio (σ) and is given by

(Reynolds, 1997):

2/1

2/11

⎟⎠⎞

⎜⎝⎛

−−

=σ

σVsVp

where:

2/13/4

⎟⎟⎠

⎞⎜⎜⎝

⎛ +=

ρµkVp

( ) 2/1/ ρµ=Vs

k=Bulk modulus

µ = Shear (rigidity) modulus

ρ = density

Note that µ =0 for a fluid, as fluids cannot support shear-waves, and the

maximum value of Poisson’s ratio is 0.5; σ % 0.05 for very hard rocks, %0.45 for

loose, unconsolidated sediments, average % 0.25.

248

8.6.11 Three-Dimensional Images

Three-dimensional images were generated by using the results from the

processing of the first-arrival traveltimes and amplitudes across the four lines at

the first and second site. Image maps in Surfur (8) were used to generate

three-dimensional images.

8.6.12 Traveltime Tomography

Wavepath eikonal traveltime (WET) tomography was used to generate a

velocity-depth section of every line for each wave type. WET tomography

accommodates multiple signal paths contributing to each first break and is a

computationally efficient and geophysically robust method. Wavepath eikonal

traveltime inversion is a high frequency traveltime tomographic method. It is a

computationally efficient method, being an order of magnitude faster than wave-

equation traveltime inversion (yet comparable in effectiveness) since only

solutions to the eikonal equation are involved. It models multiple signal

propagation paths contributing to one first-arrival and as such is superior to

conventional ray-tracing tomography which is limited to the modelling of just one

ray per first-arrival (for a brief discussion on WET theory see Appendix M).

The first step was to generate starting model by using both one and two

dimensional inversion methods. The delta-t-V method was used to generate 1D

starting models (Gebrande and Miller, 1985). This method is a pseudo-

249

tomographic method that yields one-dimensional velocity profiles for each

common midpoint (CMP). This turning ray inversion method delivers

continuous depth vs. velocity profiles for all profile stations. The first step of the

WET inversion algorithm is to propose an initial slowness model, which may be

generated assuming constant vertical velocity gradients such as with 1D tau-p

inversion (see Appendix M), or assuming discrete velocity changes such as with

the generalised reciprocal method (Palmer, 2003a). The depth sections and

seismic velocities obtained with the GRM (as discussed previously) were used

to generate 2-D starting models.

The starting models were then refined with WET tomography until ideally the

modelled traveltimes matched the field traveltimes. A perfect match between

the traveltimes however was never realised, with the mean unsigned error

around 2 ms (Appendix N). This was because after a certain number of

iterations in the WET tomography algorithm, the amount of change between the

updated velocity fields of consecutive iterations decreased, that is the rate of

convergence between the modelled and field traveltimes decreased. At this

point the WET algorithm was terminated.

The differences between the final models generated using the 1-D and 2-D

starting models were significant. Furthermore the errors between them and the

field data were both comparable and within the bounds of acceptable accuracy.

This non-uniqueness is illustrated in Figure 6-31, the top and bottom sections

being derived from 1-D and 2-D starting models respectively. Both fit the field

data to acceptable accuracy. However the differences between the two are

250

large. A refracting interface is easily identifiable in the 2-D derived section at a

depth of 15 m. This is not the case with the 1-D derived section, with a

refracting interface tentatively being placed at a depth of 20 m along the 1000

m/s contour.

The second major difference between the sections is the depths to which they

are imaged. The 1-D derived section has been imaged to a depth of 80 m. The

distances between the offset source points and the receivers were of the same

order, and as such this depth seems unrealistically large. The 2-D derived

section has been imaged to a depth of 20 m which is more plausible for

refracted waves where the minimum source to receiver distances are greater

than four times the depth of the target.

251

0 20 40 60 80 100 120 140-80

-70

-60

-50

-40

-30

-20

-10

0

0 20 40 60 80 100 120 140-30

-20

-10

0

Figure 8 - 31: Final sections from WET tomography. Section generated using 1D starting

model with multiple layers of assumed constant vertical velocity gradients (top). Section

generated using 2D starting model derived from the GRM (bottom). Despite the large

differences in the sections both fit the field data to acceptable accuracy.

252

Chapter 9

Interpretation

9.1 Summary

Three-dimensional three-component shallow seismic refraction surveys at two

sites of dryland salinity in the Spicers Creek Catchment New South Wales have

verified that the shear zone exists as a relatively narrow region of decreased

seismic velocity and increased depths of weathering. Refraction tomography

indicated that the shear zone’s seismic expression continued below the

refracting interface and into the bedrock.

Three-dimensional surveying provided an improvement in the structural imaging

capabilities of the refraction method. Despite showing gross trends along the

orthogonal two-dimensional profiles, variations in the in-line direction suggest

that shear zone has considerable lateral variations parallel to their strike. The

validity of these in-line variations are emphasised by their consistency in the

results of all the wave types. Furthermore, these variations are not predictable

on the basis of orthogonal two-dimensional profiles.

A detailed analysis of the refractor seismic velocities and amplitudes at two

sites, show a number of linear features parallel to and cross-cutting the shear

253

zone (Nikrouz and Palmer, 2004). These lineaments, on basis of the Dubbo

geological map, have been interpreted as a series of recent faults which cut the

older shear zone. The discharge of saline groundwater appears to occur along

the intersection of these faults with the shear zone.

At the first site the results show the seismic geophysical anomaly of half a shear

zone. The velocity-depth sections show that the shear zone’s seismic

expression extends below the refracting interface into the bedrock. All the

wave-types have similar spatial distribution in both time-depth and depth

images and have been interpreted as the boundary between bedrock and

overlying sediment. Changes in refractor seismic velocities occur in both the

cross-line and in-line directions. In the in-line direction, the velocities for all

wave-types show a general decreases with the refractor being separated into

three main regions of different velocities. Furthermore, the seismic velocities in

the refractor in the cross-line direction suggest the presence of a number of

cross cutting and parallel lineaments which have been interpreted to be a series

of faults.

At the second site, the results show the seismic response of a shear zone as a

narrow region of low seismic velocity and increased depth of weathering. The

P-wave/S-wave velocity ratios, which is quite sensitive to pore fluid saturant and

lithology, and Poisson’s ratios (σ ) show distinctive high values over the region.

This has been attributed to an increase in secondary porosity and permeability

from fracturing, leading to an increase in groundwater saturation. At site2, as at

site1, an interpretation of the refractor images of the seismic velocities and

254

comparison with amplitudes show a number of lineaments parallel to and cross-

cutting the shear zone. These lineaments have been interpreted as a series of

recent faults that cut the older shear zone. Furthermore, these cross-cutting

features are consistent with the tectonic interpretation in the south part of the

study area. The discharge of saline groundwater occurs at the intersection of

these faults with the shear zone.

The shear zone’s existence at both sites confirms that there is a relationship

between the geological structure and the high levels of salinisation at the

discharge zones. Groundwater continuously discharges through shallow

alluvial and collovial aquifers along the faults. This forms a “saline plume”

which contaminates the soil down gradient from the seepage zone.

Furthermore, hydrogeology studies by Morgan and Jankowski (2004) have

supported this model of the salination process.

9.2 Introduction

Seismic exploration consists of three main stages: data acquisition, processing,

and interpretation (Yilmaz, 1988). The data acquisition and processing have

been discussed in the previous chapters. In this chapter the interpretation of

data is discussed. The interpretation is based on geophysical methods, which

include magnetic, resistivity and 3D-3C seismic refraction method, and

geological method which includes the interpretation of 1:100 000 Cobbora

geological map.

255

In the Spicers Creek Catchment magnetic, electrical resistivity and seismic

refraction methods were used to image the inferred shear zones. It was

anticipated that in this area, seismic refraction would provide the best results

since the propagation of seismic waves would not be affected by the high level

salinisation whilst also providing possibly greater resolution.

In general, the geophysical and geological interpretation show a consistent

result in terms of the being of shear zone and cross-cutting features at two sites

in the Spicers Creek Catchment.

9.3 Magnetic Data Interpretation

Figures 9-1 and 9-2 show the magnetic field in nanoteslas as a 2D contour plot

for the two sites. The magnetic results show reductions of approximately 2100

nanoteslas at site1 and 600 nanoteslas at site2 over the inferred zones of the

shear zones. The low magnetic values are inferred to be caused by the

enhanced oxidation of magnetite, the magnetite mineral, to hematite, goethite or

maghemite, resulting from fracturing of the rock associated with the shear zone.

Magnetite is unstable in the typically low temperature and highly oxidizing

environment of chemical weathering and sedimentation. Groundwater

containing CO causes the ferrous iron minerals to eventually break down with

Fe being slightly soluble in water containing CO . Meanwhile, the oxygen

fugacity sets the upper limit at which magnetite can exist as a stable compound.

2

+22

256

It is given by the fO2 vs T relationship that governs the oxidation of magnetite to

hematite (Grant, 1985):

)(32/1)(2 32243 hematiteOFeOmagnetiteOFe ⇔+

At the Spicers Creek Catchment, the magnetics show lows over the inferred

location of the shear zones. These magnetic lows are consistent with the

weathering of magnetite by the saline groundwater. Furthermore, the local

magnetic data is consistent with the airborne magnetic data obtained by AGSO

over the Dubbo 1:250,000 map (as discussed in chapter 5).

9.4 Resistivity Data Interpretation

Electrical resistivity survey methods have been employed for many decades to

image the near surface where they can provide useful information on geological

structure, lithologies and subsurface water resources. Resistivity methods are

widely employed in hydrogeological investigations (Kearey and Brooks, 2001),

because earth resistivity can be readily related to porosity and water quality.

2D resistivity imaging is a geophysical method which aims to generate a cross

section of the electrical properties of the subsurface by passing an electric

current along many different paths and measuring the associated potential

differences. This can be accomplished by connecting a linear array of

257

Figure 9 - 1: 2D contour map of magnetic data at site 1.

Inferred shear zone

Inferred shear zone

Figure 9 - 2: 2D contour map of magnetic data at site 2.

258

grounded electrodes through a multicore cable to a computer controlled

switching module and a resistivity meter. However, with the electrical resistivity

surveys at the Spicers Creek, the data were acquired with a traditional field

system using two current and two potential electrodes. A large number of

apparent resistivity measurements were made using differing combinations of

four electrodes along the array in a Wenner configuration. The apparent

resistivity measurements were used to plot pseudosections, which are

qualitatively related to the distribution of true resistivity. The pseudosections

are inverted using either a manual interactive approach or an automatic iterative

method to improve the detail and correctly scale the apparent resistivity image

giving a true resistivity-depth section (Griffiths and Barker, 1993).

The conductivity* of geological materials exhibits one of the largest ranges of all

physical properties (Reynolds, 1997). In a normal crust, from the surface to up

to 15 km depth, bulk conductivities are largely controlled by aqueous electrolytic

conduction (Ward, 1990). The major factors affecting the bulk electric

conductivity of soil or rock are (McNeill, 1990):

1. Porosity and permeability

2. Degree of saturation (fraction of pore space/fractures filled with moisture)

3. Presence of clays with moderate to high cation exchange capacity (CEC)

4. Conductivity of interstitial fluid

* Resistivity is the inverse of conductivity and has units of ohm-metres (Ωm). The following will refer to conductivity, which has units of siemens/metre (S/m). 259

Geological processes such as dissolution, faulting, shearing, and weathering

usually increase porosity and fluid permeability. Thus the result of such process

on a rock formation will be an increase in conductivity.

The fluid saturation of a rock formation is a direct influence on its conductivity.

A large proportion of pore spaces, fractures and other gaps filled with fluid in a

rock mass will enhance aqueous electrolytic conduction. In general, the

conductivity of a rock increases as the percent fluid saturation increases.

The presence of clay adds an additional component to conductivity, with clay

particles acting as separate conducting paths to the electrolyte path. Due to the

large cation-exchange capacity of clay, the cations of the diffuse layer are free

to move under the influence of an applied electric field. This leads to an

increase in the density of charge carriers and hence an increase in conductivity.

The salinity of interstitial fluid is perhaps the greatest factor determining the

conductivity of a rock. The ions that enable the flow of electric current in a fluid

result from the dissociation of salts, which occurs when salts are dissolved in

water. Since the quantity of charge that can be carried by an ion is finite, the

more ions available in a solution, the greater the total charge, and hence the

higher the conductivity. In general, a rock which contains water within its pores

will have a greater conductivity when the salinity of the water is high than when

it is low.

260

The resistivity of geological materials exhibits one of the largest ranges of all

physical properties, from m for native silver to m for pure

sulfur (Reynolds, 1997). In sedimentary rocks, the resistivity of the interstitial

fluid is probably more important than that the host rock. Archie (1942)

developed an empirical formula for the effective resistivity of a rock formation in

terms of the resistivity and volume of the pore water present as below:

Ω× −8106.1 Ω1610

wcb fa ρφρ −−=

Where φ is the porosity, the fraction of pores containing water of resistivity f

wρ and , and are empirical constants (a b c 5.0 5.2≤≤ a , , and

). Korvin (1982) has proposed a theoretical basis to account for Archie’s

Law. Salin groundwater may have a resistivity as low as 0.05 Ωm and some

groundwater can have resistivities in excess of 1000

5.23.1 ≤≤ m

2≈n

Ωm.

The 2D resistivity imaging results are shown in Figures 9-3 to 9-6. For each

Figure the top pseudosection is the measured apparent resistivity, whilst the

middle is the apparent resistivity starting model calculated by the inversion

software. The bottom is the resistivity-depth section resulting after three

iterations of the inversion program.

Evidence of the shear zone is not apparent in the resistivity-depth sections.

The sections do not show narrow vertical zones of low resistivity as would be

expected from the presence of the shear zone. Rather, they all show an

approximately layered subsurface with increasing resistivity with depth. The

very near surface does however show a degree of lateral variation in resistivity

261

such as at the second site (Figure 9-6) where there are very erratic resistivity

variations in the top 8 metres of the section. These near surface variations

represent geological noise which is prevalent where the overburden has

irregular resistivity or thickness.

A large resistivity contrast exists at line 2 of the first site (Figure 9-4) where

there is a change from 7 to 17 Ωm at an increasing depth along the line. This

contrast corresponds to the boundary between the bedrock and colluvium. The

depth to bedrock is not evident in the other sections despite all being imaged to

a depth of 20 m. This is possibly due to the inversion algorithm which, although

correcting the depth variations in resistivity, images sharp boundaries as

gradational. As a result resistivity contrasts appear less than true (Barker,

1997).

At the Spicers Creek Catchment, it was anticipated that resistivity methods

would not be the most suitable for imaging the shear zone. The widespread

occurrence and high level of salinisation in the catchment has resulted in a

highly conductive overburden. The geological noise produced by this

conductive overburden during resistivity surveys would obscure the conductivity

anomaly due to shear zone in the bedrock. In addition the resistivity image

sections show low resistivity zones in the area of the inferred shear zones.

However, the regions of low conductivity were quite extensive laterally and this

together with the limited penetration did not facilitate the precise delineation of

the shear zones.

262

Figure 9 - 3: The resistivity result at Site1-Line1.

Figure 9 - 4: The resistivity result at Site1-line2.

263

Figure 9 - 5: The resistivity result at Site1-Line3.

Figure 9 - 6: The resistivity result at Site2.

264

9.5 3D-3C Seismic Data Interpretation

Seismic data interpretation for the first and second sites is based on the plans

of the various refractor properties (shown in Appendix O) and velocity-depth

sections obtained from wavepath eikonal traveltime (WET) inversion of a two

dimensional starting model derived from the generalized reciprocal method

(GRM) (shown in Appendix P). Appendix O shows the different properties of

the refractor including time-depth, depth, seismic velocity, seismic velocity ratio,

Poisson’s ratio, amplitude product, and amplitude product ratio. Appendix P

includes GRM refractor seismic velocity images for both the P-wave and S-

waves for each line.

9.5.1 Refractor Seismic Images at the First Site

On the basis of the velocities, the refracting interface at the first site has been

interpreted as the boundary between the bedrock and the overlying sediment.

Refractor time-depth maps and depth maps for each wave-type are shown in

Appendix O. The time-depth and depth maps show a similar spatial distribution

and the depth to the refractor for all wave-types (P-, SV- and SH) and steadily

decrease in the increasing cross-line direction. The spatial distribution is most

prominent for SV- and SH- waves which have similar depth maps where the

refractor increases from a depth of 13 to 17 m. Table 9-1 shows the minimum,

maximum depth and time-depth for each wave-type.

265

In general, the refractor depths as imaged by the SH- and SV- waves across

the whole image are comparable. From cross-line 0 to 60, the refractor depths

as imaged by the P-waves and S-waves are similar. However, after cross-line

60, the rate of increase in depth to the S-waves refractor is greater than the P-

wave. By the end of profile, the S-wave images the refractor 3 m deeper than

the P-wave. This is especially evident on the time-depth and depth profiles for

the individual lines as show in Appendix L chapter 8.

Table 9 - 1: Minimum, maximum and average time-depth and depth for each wave-type.

Time-depth (ms) Depth (m)

Wave-

type

Minimum Maximum Minimum Maximum

P-wave -17 -22 12.5 15.5

SH- wave -37 -52 13.5 18.5

SV- wave -33 -55 11 19

The difference in depths that the P-waves and S-waves image the refractor can

be explained by the likely occurrence of an undetected layer consisting of

saturated sediments. The sediment overlying the bedrock changes from being

damp or partially saturated to completely saturated at a point along the bedrock

interface (shown in Figure 9-7). This increase in saturation represents a

genuine increase in P-wave velocity, from 1150 m/s to a likely 1750 m/s.

266

Bedrock

Partially saturated sediment

Saturated sediment

Vp1Vs1

Vp2 > Vp1Vs2 = Vs1

Vp3 > Vp2Vs3 > Vs2

Figure 9 - 7: The possible P-wave undetected layer. A change from partially saturated to

completely saturated sediment above the refractor will result in an increase in P-wave

velocity. If this layer of completely saturated sediment is relatively thin, then it will be

undetectable in the P-wave traveltime graphs, resulting in the refractor being imaged

shallower than the true depth. Since S-waves do not propagate through liquids, the

degree of saturation in the sediment will have no effect on its propagation, thus the S-

waves will image the refractor at its true depth.

However, the saturated layer is not thick enough to generate refracted P-wave

first arrivals, and therefore is not recognisable in the traveltime graphs. The

result of this undetected layer is to image the P-wave refractor shallower than

its true depth. Furthermore, in this case the problem of the undetected layer is

not overcome by the use of an average seismic velocity above the refractor

derived with the GRM. This is because the saturated sediment is not present

along the whole profile and also progressively changes thickness.

267

Since S-waves do not propagate through liquids, the degree of saturation in the

sediments has no influence on the velocity of these waves. Thus the presence

of the completely saturated layer along part of the profile will have no effect on

the S-wave image of the refractor. The layer of saturated sediment is present

where the refractor is imaged at differing depths by the P-waves and S-waves,

and is absent where the refractor is imaged at the same depth by both waves.

The refractor images for all wave-types are shown in Appendix O. Changes in

refractor seismic velocities (Figure O3) occur in both the cross-line and the in-

line directions. In the in-line direction, the velocities for all wave-types show a

general decrease with the refractor being separated into three main regions of

different velocities (Figure 9-8). The P-wave velocities are between 4700-5300

m/s between 0 and 30 m, around 4000 m/s between 30m and 120m and 3000

m/s between 120m and 140m. The constant in seismic velocities between the

regions is not as large for the S-waves, with both the SV- and SH-waves having

velocities greater than 2500 m/s in the first region, 1700 for the second, and

around 1400 m/s for the third region. The consistent zoning of refractor

velocities for all wave-types is compelling evidence for genuine lateral changes

within the bedrock. At the first site, the existence of half of a shear zone is

evident from the first 40 m in the in-line direction (line AB on Figure 9-8). Line

AB is coincident with a major fault in the Spicers Creek Catchment as is

discussed in the geology section.

268

B

AB

A B

A

Figure 9 - 8: Summary of the refractor velocities at site 1. The refractor can be separated

into three distinct regions of different seismic velocities.

An analysis of the seismic velocities in the refractor in the cross-line direction

suggests the presence of a number of cross cutting and parallel lineaments

which have been interpreted to be a series of faults (Figure 9-9). There is a

consistency in the patterns of the seismic velocities for all wave-types, which

suggests the existence of faults. These features can be seen such as the

velocity high at cross-line 35m and in-line 10m, which is also exist in all wave-

types; the velocity low at cross-line 40m and in-line 20m-30m. The similarities

269

in the spatial features and lineaments for all wave-types suggest that the

seismic velocities are true representations of the refractor, with minimal

artefacts from the inversion algorithm, or errors in the first arrival traveltimes.

V~4.9

V~5.3

V~3

V~3.5

V~4

V~2.2

V~2.5

V~2.7

Cross-line

In-li

ne

Cross-line

In-li

ne

Cross-line

In-li

ne

Figure 9 - 9: Cross-cutting features at site1.

270

Variations in P-wave/S-wave velocity (Vp/Vs) ratio and Poisson’s ratio are

associated with the distribution of crack or pore geometry and the subsequent

degree of fluid saturation, rather than with a change in the elastic constants of

the minerals comprising the matrix (Dohr, 1985). Increases in these ratios are

generally an indication of increasing porosity and permeability, and fluid

saturation. At the first site, the Vp/Vs ratio and Poisson’s ratio show a gross

increase in the cross-line direction. Vp/Vs ratio increases from around 1.4 to

over 2.2 (Appendix O, Figure O4). Poisson’s ratio ranges from -0.5 to 0.4

(Appendix O, Figure O5), which is within the theoretical limits of -1 for a solid of

very high rigidity and 0.5 for a fluid. The increase in Vp/Vs and Poisson’s ratio

can be attributed to an increase in fracturing and fluid saturation in the bedrock.

Overlaid on the gross variations in Vp/Vs ratio and Poisson’s ratio are features

related to changes in the SV- and SH-wave velocity. The S-wave velocity highs

at cross-line 5m, in-line 10m is represented by lows in Vp/Vs and Poisson’s

ratios, whilst the S-wave velocity low at cross-line 120 m to 140 m from in-line 0

to 20 m is represented by highs. The differences between the SV- and SH-

wave velocities are small as indicated by the SV-wave/SH-wave velocity ratio

which has an average of 1 and ranges from 0.75 to 1.25.

The amplitudes products are shown in Appendix O Figure O6. The amplitude

products are variable for P-waves and S-waves. However, there are similarities

with the lineaments on the amplitude products and the lineaments interpreted

from the seismic velocities, both showing a number of parallel and cross-cutting

features. Perhaps one features of interest is the wide zone of low amplitude

271

between cross-line 60m and 120 m for P-waves and SV and meanwhile cross-

line 60 m to 100 m for SH.

Figure 9-10 is the wireframe map of amplitude product with the geometric effect

correction being applied. The map can be divided into three regions. A region

of high amplitude product exists between cross-line between 0 to 60 m, a region

of low amplitude product exists between cross-line between 60m to 100 m, and

a region of high amplitude product exists between cross-line between 100 m to

140 m. The presence of these regions suggests that variations in amplitudes

are associated with changes in the refractor. A considerable degree of noise is

also evident on the amplitude products and is related to lateral variations in the

near surface layers. The low amplitude between cross-line 0 to 20 m and high

amplitude between cross-line 100 m to 120 m are possibly due to near surface

effects in the study area.

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

0.022

0.024

Amplitude Product (S1*S18) , P-wave, Site 1

Amplitude Product

Cross-line

In-line

Figure 9 - 10: Amplitude product for P-wave at site1.

272

As with the amplitude products, the amplitude product ratios show little

consistency between wave-types. However, a zone between cross-line 60 m to

100 m is similar to the amplitude low mentioned above. The cross-cutting

features are not obvious in amplitude product maps possibly due to near

surface effects.

Figures P1 to P3 (Appendix P) are velocity-depth sections obtained from

wavepath eikonal traveltime (WET) inversion of a two dimensional starting

model derived from the generalized reciprocal method (GRM) for all wave-

types.

The refractor depths for all wave-types are between 11 m to 13 m and are

similar to the depth sections as shown in Appendix L chapter 8. The velocities

for all wave-types show a general decrease in the increasing cross-line. The

refractor can be separated into three main regions of different seismic

velocities. The first region has P-wave velocities of between 5200 – 3600 m/s,

and around 2800 m/s and 2000 m/s for the second and third regions

respectively. For SH waves the first region has velocities of between 3600-

2200 m/s, and around 1600 and 800 m/s for the second and third regions

respectively. The distinction between regions is less evident for the SV waves

but lateral changes are obvious.

In general, the images show the lateral changes in refractor in term of velocity

for all type-waves. The presence of half a shear zone can be deduced from

these GRM tomography images. The existence of lateral changes is due to the

273

different geological strata and the presence of the fault at study area. Cross

cutting features are not obvious on these sections.

9.5.2 Refractor Seismic Images at the Second Site

Figures O8 and O9 (Appendix O) are the refractor time-depth and depth images

at the second site. All the wave-types have similar spatial distribution in both

images and have been interpreted as the boundary between bedrock and

overlying sediment. There is an increase in refractor depth in the middle of the

profile between cross-line 40 m to 100 m. The P-waves and SH-waves image

the refractor at a maximum depth of around 10.5 m between cross-line 40 m

and 90 m. The shallowest depths are at the ends of profile around cross-line 20

m and cross-line 120 m. The SV-waves image the refractor at a maximum

depth around 14 m between 30 m and 60 m, and are the shallowest at the ends

of profile around cross-line 15 m and 130 m. Table 9-2 shows the minimum,

maximum depth and time-depth for each wave-type.

Table 9 - 2: Minimum, maximum and average time-depth and depth for each wave-type.

Time-depth (ms) Depth (m)

Wave-

type

Minimum Maximum Minimum Maximum

P-wave -5 -12 4.5 10.5

SH- wave -10 -22 5.5 10.5

SV- wave -12 -26 6 14

274

The SH-wave depth image shows that the refractor has been shifted laterally by

20 m in the decreasing cross-line direction and is 3 m deeper than the other

wave-types. The lateral shift in the refractor depth is probably due to SV-P-SV

wave-type conversion (Figure 9-11) and the inconsistent use of the first-arrival

traveltimes in the GRM algorithms. SV-P-SV wave-type conversion occurs

when an SV-wave generated at the surface, propagates into the subsurface at a

velocity Vs. At the refracting interface, SV-wave is critically refracted and is

converted to a P-wave and travels along the refractor at a greater velocity Vp.

The head waves generated by this wave are then converted back to SV-waves

and travel at a velocity Vs to the surface. Thus waves which are thought to be

pure SV-waves are in fact SV-P-SV-waves. A lateral shift of the refractor may

result if the forward and reverse traveltimes used by the GRM algorithms has

one set derived from SV-P-SV-waves and the other set from pure SV-waves

(shown in Figure 9-11). Furthermore, the use of the faster SV-P-SV-wave

velocity in the depth conversion factor will result in the refractor being imaged at

a greater depth.

Unlike the first site, there is no considerable difference in the refractor’s depth

as imaged by the P-waves and SH-waves. This suggests that the layer of

partially saturated or completely saturated sediment is absent. The surface

velocities at the second site are greater than at the first, suggesting the absence

of the partially saturated layer.

275

A C D Bsurface

SVSVSV

SV refracting

PP

Figure 9 - 11: Wave-type conversion and the GRM. The forward and reverse traveltimes

employed by the GRM algorithms include one set derived from SV-P-SV-waves and the

other set from pure SV-waves. When this occurs, the refractor is shifted laterally and is

imaged at a greater depth.

The velocities for all wave-types are shown in Figure O10 Appendix O. As with

the first site, the change of velocities in the refractor occurs in both directions.

In the cross-line direction the refractor can be separated into four regions of

different seismic velocities, with a region of low seismic velocity occurring

between regions of high seismic velocities (Figure 9-12).

276

5

2.8

Figure 9 - 12: Summary of the refractor velocities at site 2. Refractor can be separated

into four distinct regions of different seismic velocities.

For all wave-types the boundary between the region of low and high seismic

velocity is quite distinct. The first region is between cross-line 0 to 20 m, the

second between 20 m to 50 m, the third between 50 m to 110 m, and the fourth

277

between 110 m to 140 m. The first region has P-wave velocities around 3000

m/s, and around 5000 m/s and 3800 m/s and 5400 m/s for the second and third

and fourth regions respectively. The seismic velocities for the SH-wave are

around 2000 m/s, 2300 m/s and 1500 m/s and 2500 m/s for the first, second,

third and fourth regions respectively. The seismic velocities for the SV-wave

are around 3500 m/s, 2800 m/s, 2000 m/s and 3000 m/s for the first, second,

third, and fourth regions respectively.

The velocity contrast at third region and the other three regions are smaller for

the P-waves than for S-waves. This is possibly due to the degree of fluid

saturation in the bedrock. The P-wave velocity contrast between unfractured

rock and fluid-saturated fractured rock is smaller than between unfractured rock

and dry fractured rock. Since they do not propagate through liquids, the velocity

of S-waves in fractured rock is the same no matter what the degree of fluid

saturation. The high degree of fluid saturation in the bedrock has led to a

decrease in the velocity contrast of the P-waves whilst not affecting the velocity

contrast of the S-waves.

The consistent zoning of refractor velocities for all wave-types is compelling

evidence for genuine lateral lithological changes within the bedrock especially

for the third mentioned region. These seismic velocity models together with the

depth images suggest a narrow shear zone at the third region (cross-line 50 m

to 110 m) characterized by low seismic velocities and increased depths of

weathering.

278

As with the first site, in the inline direction a detailed analysis of the seismic

velocities in the refractor suggests the presence of a number of cross cutting

and parallel lineaments which have been interpreted to be a series of faults

(Figure 9-13). There are a number of spatial similarities between all wave-types

velocities between cross-line 0 to 50 m, in-line 0 to 30 m, and between cross-

line 110 m to 140 m, in-line 0 to 30 m which seem to be related to the faults.

While The cross cutting features can be seen on both sides on the images for

all wave types, the SH-wave shows these features the best. This maybe due to

the direction of propagation SH-waves throughout the refractor. Figure 9-14

shows the direction of propagation for all wave-types.

There is a region of increased P-wave/SH-wave velocity ratio and associated

Poisson’s ratio between cross-lines 40 m and 110 m (Figure O11, O12,

Appendix O). The increase is considerable with a Vs/Vp ratio of 2.25 and

Poisson’s ratio of 0.35, contrasted with Vs/Vp ratio of 1.75 and Poisson’s ratio

of 0.3 in the surrounding regions. This increase correlates with the region of

low refractor velocity and is possibly related to an increase in porosity and

permeability, and a subsequent increase in fluid saturation.

The lateral variation in the SV-wave velocity can be resolved by examining the

velocity ratios. A P-wave/SV-wave velocity ratio high of greater than 2.25

occurs between cross-lines 20 m and 40 m, whilst the SV-wave/SH-wave

velocity ratio high of greater than 1.75 occurs between cross-lines 90 m and

110 m.

279

V~3.5

V~2.4

V~4.8

V~5.8

V~4.8

V~5.6

2.8

V~3.8 V~3.3

V~2.6 V~3.2

V~3.3

V~2.6

V~2.5

V~2

V~2.2

V~2.4

V~3

V~2

Cross-line

In-line

Cross-line

In-line

In-line

Cross-line

Figure 9 - 13: Cross-cutting features at site2.

280

Figure 9 - 14: Distinction among the particle-displacement vectors associated with the

three fundamental modes, P, SH, and SV, which compose vector-wavefield seismic data

(Hardage et al., 2003).

Figure O13 (Appendix O) shows the amplitude products for all wave-types.

There is a general increase in the increasing cross-line direction. Like the first

site there is no correlation between the amplitudes and the seismic velocities in

the refractor. However, there are similarities with a number of cross-cutting and

parallel faults in the seismic velocity profiles. The amplitude images show the

lateral changes in the cross-line related to the faults in the bedrock.

Variations in amplitudes related to changes in the bedrock may be resolved

using the P-wave/S-wave ratios of the amplitude product (O14). Features such

281

as the high between cross-line 40 m and 100 m in the P-wave/SV-wave ratio,

and between 50 m and 110 m in the P-wave/SH-wave ratio, are clearly

detectable. This region of high P-wave amplitude and low S-wave amplitude

may be due a high degree of fracturing causing the preferential attenuation of

S-waves with respect to P-waves as S-waves are more directly affected by

changes in rigidity modulus and density.

Figures P4 to P6 (Appendix P) show the GRM velocity-depth sections for all

wave-types. The low velocity zone can be seen for most of images including P-

waves (line1 and line4), SH-waves (line1, line3, line4) and SV-waves (line4).

There is no obvious low velocity zone for some images because of the limited

penetration of some of the wave-types into the refractor. The low velocity zones

support a region of increased porosity and permeability, and the subsequent

increases in the degree of groundwater saturation, that correlates with the

presence of the shear zone at the second site.

9.6 Geology and Tectonic Interpretation

As discussed in the geology chapter, the study area is situated in the mid-

Silurian to Early Devonian Hill End Trough in the eastern Lachlan Fold Belt

(LFB) and overlies marginal areas of the Gunnedah Basin and Surat Basin.

The Hill End Trough is bounded to the south by the Carboniferous Buthurst

Batholith and the Lachlan Transverse Zone (Glen and Walshe, 1999) and to the

east and west by Ordovician volcanics that are overlain by thin Silurian-Early

282

Devonian rocks and intruded in part by Carboniferous granites (David et al.,

2003). Figure 9-15 shows the simplified regional geology map of the Hill End

Trough (David et al., 2003).

The geology and tectonic interpretation shows a detailed model of the cross-

cutting features at the first and second site. They confirm that there is a good

relationship between the interpretation of seismic data and that of the geology

at the Spicers Creek Catchment. The detection of lineaments and cross-cutting

features through the interpretation of the seismic velocities and amplitude

products was made possible via the use of three dimensional seismic refraction

methods.

283

Figure 9 - 15: Simplified regional geology map of the Hill End Trough (Modified from

David et al., 2003).

284

Chapter 10

Conclusions

Three-dimensional (3D) three-component (3C) shallow seismic refraction

surveys were recorded over a shear zone at two sites associated with dryland

salinity in the Spicers Creek Catchment, near Dubbo in southeastern Australia.

The seismic data were recorded with the Australian National Seismic Imaging

Resources (ANSIR) 360-trace ARAM-24 seismic system and IVI MiniVibrator.

Dryland salinity occurs extensively throughout the Spicers Creek Catchment.

The high concentration of salt in the groundwater has led to a significant decline

in agricultural productivity and a reduction in native vegetation. Furthermore,

the saline groundwater in the surface soil has caused the destruction of the clay

and soil structure and as a result, large areas of the catchment have

experienced soil erosion and extensive alteration of the landscape.

The broad objective of this study was to use seismic refraction methods to map

in detail a shear zone, which is associated with the salination. Although the

shear zone had been recognized in airborne geophysical surveys and

subsequent regional geological mapping as the Nindethana Fault, the exact

geological factors which produced the discontinuous occurrences of saline

groundwater at the surface, were not known. Detailed ground magnetic surveys

carried out as part of this study, confirmed the location of the shear zone, but

285

they did not provide further explanation of the causes of salination. Therefore,

the detailed objectives were to map the extent of the shear zone as well as

other properties, such as the rock fabric.

Seismic refraction methods were selected because of their potential ability to

provide greater lateral resolution of the narrow vertical shear zone, than is

currently the norm with electrical or electromagnetic methods. This situation

was confirmed with a number of resistivity depth images generated as part of

this study. Furthermore, it was anticipated that the high electrical conductivities

of the near surface areas would have no significant impact on the depth of

investigation of the seismic energy. Finally it was also anticipated that the head

wave amplitude might provide a more detailed measure of the seismic velocities

as a function of azimuth and therefore of fracture porosity.

The 3D 3C survey was set out as four parallel lines 10 m apart with twenty-nine

3C geophones, 5 m apart, in each line. The source lines were located parallel

and orthogonal to the lines of geophones and passing through the geometric

centre of the receiver spread. Source points were nominally 10 m apart and

both P- and S-waves were generated.

The results of the seismic refraction surveys show that the shear zone occurs

as a narrow region with low seismic velocities and increased depths of

weathering. However, there were numerous ambiguities in generating

consistent parameters in the refractor between adjacent receiver lines.

Furthermore, these ambiguities were not resolved with the use refraction

286

tomography, because even moderate differences with the various starting

models still generated acceptable agreement with the traveltime data.

Therefore, in order to generate consistency between the results for the various

receiver lines, it was necessary to generate detailed starting models.

It was found that near surface irregularities of a very limited lateral extent were

the major sources of uncertainty in deriving detailed starting models. In the

Spicers Creek Catchment, the variations in the traveltimes caused by the near

surface irregularities, were frequently of a comparable magnitude to those

caused by the changes in depths and seismic velocities in the refractor. As a

result, the derivation of detailed parameters of the refractor using the

generalized reciprocal method (GRM) was subject to significant uncertainty

unless they are accommodated with the GRM SSM.

A major achievement of this study is the development of a smoothing method

which significantly removes the effects of the near surface irregularities.

Termed the GRM SSM (for statics smoothing method), it essentially generates

a time-depth model of the refracting interface for which the effects of the near

surface irregularities have been minimized, by taking an average of the time-

depth profiles for a range of XY values. Although the time-depths to the target

refractor show very little variation with changing XY distances, the traveltime

anomalies for the near surface irregularities migrate laterally through the time-

depths. Therefore, the average time-depth profile is largely free of the near

surface anomalies. When this averaged time-depth profile is subtracted from

that for zero XY, the result is essentially a set of corrections for the near surface

287

irregularities. These corrections are then subtracted from the field traveltimes,

and the GRM computations are then repeated with the corrected traveltimes.

The GRM SSM, which takes advantage of the unique redundancy properties of

the GRM computations, was a major factor in deriving consistent detailed

starting models for refraction inversion. Furthermore, this consistency was

achieved with both S-wave components, as well as the P-wave results.

The major geological achievement, which was made possible with the GRM

SSM, was the demonstration that there were cross cutting features associated

with the major shear zone. Therefore, it appears that saline groundwater can

discharge at the surface where increased volumes of groundwater occur at the

intersection of different sets of shears. This model provides a useful

explanation for the discontinuous occurrence of salination along the major shear

zone.

As is usually the case, increased numbers of recording channels would have

been beneficial. In particular, the four recording lines separated by 10 m did not

provide as high resolution in the cross-line direction as in the in-line direction.

Therefore, where cross cutting features are considered likely to occur and to be

significant, than re-deploying the recording spread in the orthogonal direction

provides an acceptable alternative to employing field systems with significantly

larger recording capacity.

The measurement of S-wave velocities provided a measure of Poisson’s ratio,

which is a useful elastic parameter. Although a detailed analysis of the head

288

wave amplitudes did not generate useful results, nevertheless, the fact that the

head wave amplitude is a function of the densities, as well as the seismic

velocities, suggests that the joint inversion of seismic refraction traveltimes and

head wave amplitudes, with detailed gravity profiles, should facilitate the

determination of both seismic velocity and density models, and in turn, the

derivation of elastic constants. It is likely that the detailed geotechnical

characterization of sites with 3D 3C seismic refraction methods, together with

detailed gravity profiles, could provide useful quantitative models for the

analysis of groundwater flow in fractured rock masses, as well as for the

traditional engineering construction site applications.

289

REFERENCES ACIL, Australia Pty Ltd, 1983, Causes, Extent and Effect of Salinity in Victoria:

A Report Prepared for the Salinity Committee of the Victorian Parliament, Melbourne, pp. 175.

Acworth, R. I. and Jankowski, J., 2001, Salt source for dryland salinity-evidence

from an upland catchment on the Southern Tablelands Of New South Wales: Aust. J.Soil.Res., 39, pp. 39-59.

Allan, U. S., 1989, Model for hydrocarbon migration and entrapment within

faulted structures: AAPG Bulletin, 73, pp. 803-811. Anderson, R. N., Flemings, P., Losh, S., Austin, J. and Woodhams, R., 1994,

Gulf of Mexico growth fault drilled, seen as oil, gas migration pathway: Oil and Gas J., 12, pp. 97-102.

Anderson, R. N., He, W., Hobart, M. A., Wilkinson, C. R. and Nelson, H. R.,

1991, Active fluid flow in the Eugene Island area, offshore Louisiana: The Leading Edge, 10, pp. 12-17.

Archie, G. E., 1942, The electrical resistivity log as an aid to determining some

reservoir characteristics: A.I.M.E., 146, pp. 389-406. Atkinson, T., 1970, Ground preparation by ripping in open pit mining: Mining

Magazine, 122, pp. 458-469. Bamford, D. and Nunn, K. R., 1979, In situ seismic measurements of crack

anisotropy in the carboniferous limestone of northwest England: Geophys. Prosp., 27, pp. 322-338.

Barker, R. D., 1997, Electrical imaging and its applications in engineering

investigations, In: McCann, D. M., Eddleston, M., Fenning, P.J., Reeves, G.M.(Ed): Modern geophysics in engineering geology, The Geological Society, London, pp. 37-43.

Barron, L. M., 1997, Igneous petrology overview of the Dubbo 1:250,000

geological sheet: Geological Survey of New South Wales, Report GS1997/080.

Bartlett, A. H., 1962, Refraction seismic survey in civil engineering: Australian

Road Research Board 1st Biennial Conference. Bates, R. L. and Jackson, J. A., 1980, Glossary of Geology: American

Geological Institute, U.S.A. Beach, A. and Fyfe, W. S., 1972, Fluid transport and shear zones at Scourie,

Sutherland: evidence for overthrusting?: Contrib.Mineral.Petrol., 36, pp. 175-180.

290

Bennett, G., 1999, 3-D seismic refraction for deep exploration targets: The Leading Edge, 18, pp. 186-191.

Berg, R. R. and Avery, A. H., 1995, Sealing properties of tertiary growth faults,

Texas Gulf Coast: AAPG Bulletin, 79, pp. 375-393. Beyer, W. H., 1975, Handbook of mathematical science, 5th edition: CRC

Press, Florida. Biehler, S., Ferguson, J., Baldridge, W. S., Jiracek, G. R., Aldren, J. L.,

Martinez, M., Fernandez, R., Romo, J., Gilpin, B., Braile, L. W., Hersey, D. R., Luyendyk, B. P. and Aiken, C. L., 1991, A geophysical model of the Espanol Basin, Rio Grande rift, New Mexico: Geophysics, 56, pp. 340-353.

Bigelow, N., 1965, Seismograph operations by marine state highway

commission: Highway Research Record No.81, Highway Research Board, Washington, pp. 9-15.

Bonini, E. W. and Hickok, E. A., 1958, Seismic refraction method in ground-

water exploration: AIME Trans., 211, pp. 485-488. Bourke, D. J., Hawke, J. M. and Scheibner, E., 1974, Structural subdivision of

the Great Australian Basin in New South Wales: Geological Survey of New South Wales, Quarterly Notes, 16, pp. 10-16.

Bradd, J. M. and Gates, G., 1996, Dryland salinity in New South Wales. A state

perspective: Department of Land and Water Conservation, Technical Report No. TS95.113.

Bradd, J. M., Milne-Home, W. A. and Gates, G., 1997, Overview of factors

leading to dryland salinity and its potential hazard in New South Wales, Australia: Hydrogeology Journal, 5, pp. 51-67.

Branagan, D. F. and Packham, G. H., 2000, Field geology of New South Wales:

Department of Mineral Resources New South Wales, Sydney, Australia. Braun, J., Munroe, S. and Cox, F., 2003, Transient fluid flow in and around a

fault: Geofluids, 3, pp. 81-87. Brown, A. R., 1996, Interpretation of three-dimensional seismic data: AAPG

Memoir, 42. Brown, P. D. and Robertshaw, J., 1957, Determining the thickness of

unconsolidated deposits overlying shallow mine working by seismic methods in methods and case histories in mining geophysics: Sixth Commonwealth Mining and Metallurgical Congress, Montreal.

291

Campos-Enriquez, J., Delgado-Redriguez, O., Chavez-Segura, R., Gomez-Contreras, P., Flores-Marquez, E. L. and Birch, F. S., 1997, The subsurface structure of the Chalco sub-basin (Mexico City) inferred from geophysical data: Geophysics, 62, pp. 23-35.

Carcione, J. M. and Cavallini, F., 2002, Poisson's ratio at high pore pressure.:

Geophysical Prospecting, 50, pp. 97-106. Cerveny, V. and Ravindra, R., 1971, Theory of seismic head waves: University

of Toronto Press. Cerveny, V., Molotkov, I.A., Psencik, I., 1977, Ray Method in Seismology:

Praha, Universita, Karlova. Chappell, B. W., White, A. J. R. and Hine, R., 1988, Granite provinces and

basement terranes in the Lachlan Fold Belt, Southeaster Australia: Australian Journal of Earth Sciences, 35, pp. 505-521.

Clarke, C. J., George, R. L., Bell, R. W. and Hatton, T. J., 2002, Dryland salinity

in south-western Australia: its origins, remedies and future research directions: Australian Journal of Soil Research, 40, pp. 93-113.

Collins, W. J. and Vernon, R. H., 1994, A rift-drift-delamination model of

continental evolution: Paleaozoic tectonic development of eastern Australia: Tectonophysics, 235, pp. 249-275.

Coney, P. J., 1992, The Lachlan belt of eastern Australia and Circum-Pacific

tectonic evolution: Tectonophysics, 214, pp. 1-25. Coram, J., 1998, National classification of catchments for land and river salinity

control: RIRDC Publication, No: 98/78. Coram, J., Dyson, P. and Evans, R., 2001, An evaluation framework for dryland

salinity: Department of Agriculture, Fisheries and Forestry Australia, Australian Government Publishing Service, Canberra.

Cramsie, J. N. and Hawke, J. M., 1984, Introduction In:Hawke, J. M. and

Cramsie, J. N.(Eds) Contribution to the geology of the Great Australian Basin in New South Wales: Geological survey of New South Wales, Department of Mineral Resources, Bulletin No.31, pp. 1-22.

Creelman, R., 2002, Reviewing the Science of Salinity: First Annual Symposium

on Salinity Assessment and Management, University of Sydney. Crook, K. A. W., 1980, Fore-arc evolution in the Tasman Geosyncline: the origin

of the southern Australia continental crust: Journal of the Geological Society of Australia, 27, pp. 215-232.

292

Cummings, D., 1979, Determination of depths an irregular interface in shallow refraction surveys using a pocket calculator: Geophysics, 44, pp. 1987-1998.

Dahlhaus, P. G., MacEwan, R. L., Nathan, L. and Morand, J., 2000, Salinity on

the southeastern Dundas Tableland, Victoria: Australian Journal of Earth Sciences, 47, pp. 3-11.

Dasios, A., McCann, C., Astin, T. R., McCann, D. M. and Fenning, P., 1999,

Seismic imaging of the shallow subsurface: shear-wave case histories: Geophysical Prospecting, 47, pp. 565-591.

David, V., 2003, Cross section four, Combo to Guntawang in Hill End Trough,

CD Geoscience Database: Geological Survey of New South Wales. David, V., Glen, R. A. and Spencer, R., 2003, Three-dimensional extension of

the Hill End Trough based on the modeling of the regional gravity data. In: Vassallo, J. J. G., R.A.(Ed), Evolution of the Hill End Trough: Hill End Trough CD Geoscience Data Package. Open File GS Report 2003/291, Geological Survey of New South Wales, Sydney, Australia, pp. 43-48.

Davis, H. D. and Reynolds, S. J., 1996, Structural geology of rocks and regions:

John Wiley & sons Inc., U.S.A. de Groot-Hedlin, C. and Constable, S., 1990, Occam's inversion to generate

smooth, two-dimensional models form magnetotelluric data: Geophysics, 55, pp. 1613-1624.

Deen, T. and Gohl, K., 2002, 3-D tomographic seismic inversion of a

paleochannel system in central New South Wales, Australia: Geophysics, 67, pp. 1364-1371.

Deen, T. J., Gohl, K., Leslie, C., Papp, E. and Wake-Dyster, K. D., 2000,

Seismic refraction inversion over a palaeochannel in the Lachlan fold belt, central NSW: Exploration Geophysics, 31, pp. 389-393.

Dennis, J. G., 1967, International tectonic dictionary: Am. Assoc. Pet. Geol.

Tusla, pp. 196. Dobrin, M. B., 1983, Introduction to geophysical prospecting: McGraw-Hill.Inc,

Japan. Dobrin, M. B. and Savit, C. H., 1988, Introduction to geophysical prospecting:

McGraw Hill. Dohr, G., 1985, Seismic shear waves: Geophysical Press, London. 293

Domenico, S. N. and Danbon, S. H., 1986, Shear-wave technology in petroleum exploration - past, current, and future In: Dunbom, S. H. and Domenico, S. N.(Eds) Shear wave exploration, Society of Exploration Geophysicists, pp. 3-18.

Drake, C. L., 1962, Geophysics, "Geophysics" and engineering: Geophysics,

27, pp. 193-197. Duguid, J. O., 1968, Refracted determination of water table depth and alluvium

thickness: Geophysics, 33, pp. 481-488. Ebrom, D., McDonald, J. and Tatham, B., 1992, Sideswipe in a 3-D, 3-C model

survey: The Leading Edge, 11, pp. 45-51. Edge, A. G. and Laby, T. H., 1931, The principles and practice of geophysical

prospecting: Cambridge University Press, London. Edwards, L. S., 1977, A modified pseudosection for resistivity and induced

polarization: Geophysics, 42, pp. 1020-1036. English, P., 2002, Overview of salinity research at CSIRO land and water: First

Annual Symposium on Salinity Assessment and Management, University of Sydney.

Exon, N. F., 1976, Geology of the Surat Basin: Australian Government

Publishing Service, Canberra, French, S. W., 1974, Two-dimensional and three-dimensional migration of

model-experiment reflection profiles: Geophysics, 39, pp. 265-277. Gebrande, H. and Miller, H., 1985, Refraction seismology In: Bender, F.(Ed),

Applied geoscience, methods of applied geophysics and mathematical procedures in geosinces, F.Enke Publishing House, Stuttgart, pp. 226-260.

Glen, R. A. and Walshe, J. L., 1999, Cross Structure in the Lachlan Orogen

Australia: the Lachlan Transverse Zone example: Australian Journal of Earth Sciences, 46, pp. 641-658.

Glen, R. A., Walshe, J. L., Barron, L. M. and Watkins, J. J., 1998, Ordovician

convergent-margin volcanism and tectonism in the Lachlan sector of east Gondwana: Geology, 26, pp. 751-754.

Glen, R. A. and Watkins, J. J., 1999, Implications of Middle Devonian

deformation of the eastern part of the Hill End Trough, Lachlan Orogen, New South Wales: Australian Journal of Earth Sciences, 46, pp. 35-52.

Grant, F. S., 1985, Aeromagnetics, geology and ore environments, I.Magnetite

in igneous, sedimentary and metamorphic rocks, An overview: Geoexploration, 23, pp. 303-333.

294

Griffiths, D. H. and Barker, R. D., 1993, Two-dimensional resistivity imaging and modelling in areas of complex geology: Journal of Applied Geophysics, 29, pp. 211-226.

Gunn, P. J., 1997a, Regional magnetic and gravity responses of extensional

sedimentary basins: AGSO Journal of Australian Geology and Geophysics, 17, pp. 115-131.

Gunn, P. J., 1997b, Application of aeromagnetic surveys to sedimentary basin