Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Targeting Archean orogenic mineralization using physical properties and integrated geophysical methods 2009

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
ubc_2009_spring_mitchinson_dianne_edith.pdf [ 9.6MB ]
ubc_2009_spring_mitchinson_dianne_edith.pdf
Metadata
JSON: 1.0052581.json
JSON-LD: 1.0052581+ld.json
RDF/XML (Pretty): 1.0052581.xml
RDF/JSON: 1.0052581+rdf.json
Turtle: 1.0052581+rdf-turtle.txt
N-Triples: 1.0052581+rdf-ntriples.txt
Citation
1.0052581.ris

Full Text

Targeting Archean Orogenic Gold Mineralization Using Physical Properties and Integrated Geophysical Methods by DIANNE EDITH MITCHTNSON B.Sc., Memorial University of Newfoundland, 2001 M.Sc., Laurentian University, 2004 A DISSERTATION SUBMITTED 1N PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Geological Sciences) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) January 2009 © Dianne Edith Mitchinson, 2009 Abstract Although Archean orogenic gold mineralization is not readily detected using geophysical methods, due to a lack of petrophysical contrast between typical low volumes of gold and hosting rocks, it is possible to use geophysics to detect other petrophysically distinct gold indicators. Geophysical inversion methods, in particular, make it possible to not only detect important gold-related rocks in the subsurface, but to map their distribution in three dimensions. The research presented examines the effectiveness of geophysical inversion as an exploration tool in the Archean orogenic gold environment through extensive physical property analysis, synthetic modeling, and inversion of various geophysical data over the Hislop gold deposit, Ontario. As understanding rock properties is imperative to interpreting geophysical data, it was necessary to establish the physical property ranges of typical host rock types, hydrothermally-altered, and mineralized rocks in this deposit setting. Felsic dikes, known to be associated with gold at Hislop, have low magnetic susceptibility and density ranges that allow them to be distinguished from mafic and ultramafic rocks. Additionally, many potentially mineralized, carbonate-altered mafic and ultramafic rocks can be isolated from their least-altered equivalents using susceptibility. Synthetic modeling showed that narrow, near-vertical felsic dikes, and sulfide- rich zones hosted by mafic and ultramafic volcanic rocks can be imaged up to 35O m in the subsurface using inversion methods. It is necessary however, to focus on small areas, to have closely spaced measurements, and small inversion cell sizes. It was demonstrated that constraining inversions through addition of basic prior geologic and physical property information, yields models with improved physical property distribution, and estimates. Applying knowledge gained from physical property, and synthetic modeling work lent confidence to interpretations of inversion results for the Hislop area. At regional scales, susceptibility and density models reveal a steep southward dip for the gold-related Porcupine-Destor Deformation Zone, and a greenstone depth of approximately 7000 m. Fe-rich mafic rocks directly hosting the Hislop deposit are 11 complexly faulted and extend to 3000 m depth. At deposit-scales, model cells with combined low susceptibilities and high chargeabilities, occurring proximal to faults, felsic intrusions, and Fe-rich mafic rocks, highlight prospective areas for further investigation. 111 Table of Contents Abstract.ii Table of Contents iv List of Tables viii List of Figures ix Acknowledgements xiv Co-authorship Statement xvi Chapter 1: Introduction 1 1.1. Combining geology and geophysical inversion for mineral exploration 1 1.1.1. Geophysics and mineral exploration 1 1.1.2. Mineral Deposit Research Unit — Geophysical Inversion Facility project 2 1.1.3. Inversion in the Archean orogenic gold environment 3 1.2. Background to geophysical inversion 5 1.3. Archean orogenic gold - geologic and geophysical background 7 1.3.1. Background on Archean orogenic gold 7 1.3.2. The Hislop gold deposit 8 1.3.3. Geophysics and gold 9 1.4. Project objectives 10 References 13 Chapter 2: Physical properties of rocks in an Archean orogenic gold environment 17 2.1. Introduction 17 2.1.1 Rationale 17 2.1.2 Objectives 18 2.2. Background 19 2.2.1 Geology and geophysics of Archean orogenic gold deposits 19 2.2.2 Geology of the study area 20 2.3. Methodology 25 2.3.1 Field and mineralogical studies 25 2.3.2 Physical property measurements 26 iv 2.4. Data and observations.29 2.4.1. Hislop deposit rock types, hydrothermal alteration, and associated mineralogy 29 2.4.2. Physical properties of the Hislop deposit 32 2.5. Interpretations 43 2.5.1. Effect of geological processes on physical properties at Hislop 43 2.6. Discussion 60 2.6.1. Exploration using physical properties 60 2.6.2. Comparison to analogous areas 67 2.7. Conclusions 75 References 77 Chapter 3: Detecting gold-related geology in Archean orogenic gold environments using geophysical inversion: a synthetic modeling study based on the Hislop gold deposit, Ontario 86 3.1. Introduction 86 3.1.1. Rationale 86 3.1.2. Objectives 87 3.2. Background 88 3.2.1. Geology of the Hislop gold deposit and relationship to other Archean orogenic gold deposits 88 3.2.2. Physical Properties of rock types and alteration zones at Hislop 90 3.2.3. General forward modeling and inversion background 95 3.3. Methods 97 3.4. Synthetic modeling results 103 3.4.1. Potential fields modeling 103 3.4.2. DC resistivity and induced polarization modeling 112 3.4.3. Improving model results with basic constraints 120 3.4.4. Other solutions for improving model results 125 3.5. Conclusions 128 References 131 v Chapter 4: 3D inversion of magnetic, gravity, DC resistivity, and induced polarization data over the Hislop gold deposit, south-central Abitibi greenstone belt 137 4.1. Introduction 137 4.1.1.Rationale 137 4.1.2. Geological background 138 4.1.3. Relationships between geophysics, physical properties, and geology 140 4.1.4. Inversion background 147 4.2. Inversion Approach 150 4.2.1. General strategy 150 4.2.2. Magnetic inversions 151 4.2.3. Gravity inversions 157 4.2.4. DC resistivity and IP inversions 158 4.2.5. Constraining magnetic inversions with reference models built in Modelbuilder 160 4.2.6. Inversion model display 162 4.3. Inversion results and analysis 165 4.3.1. Magnetic susceptibility models 165 4.3.2. Density model 175 4.3.3. Resistivity models 175 4.3.4. Chargeability models 180 4.4. Querying combined inversion results 181 4.4.1. Regional scale query (susceptibility and density) 182 4.4.2. Local scale query (susceptibility, chargeability) 184 4.4.3. Deposit scale query (susceptibility, chargeability) 186 4.5. Summary and discussion 188 References 192 Chapter 5: Summary and future work 198 5.1. Synthesis of research presented 198 5.2. Significance and contributions to the field 199 vi 5.3. Limitations of the thesis research 200 5.4. Recommendations for continued work 202 5.5. Future directions of the field of study 204 References 207 Appendix 2A — List of Abbreviations 209 Appendix 2B - Hislop Drilicore Logs, Cross-sections, and Outcrop Maps 210 Appendix 2C - Detailed and Expanded Methods 222 Appendix 2D — X-ray Diffraction Analyses 231 Appendix 2E — Physical Properties of Hislop Deposit Rocks 233 Appendix 2F — Physical Properties — Descriptive Statistics 248 Appendix 2G — correlation coefficients for Physical Properties and XRD (Rietveld) - Derived mineral abundances 251 Appendix 3A - Observed versus Prediced Data for Synthetic Inversion Models 257 Appendix 4A - Hislop 3D Magnetic, 3D Gravity, 3D DC Resistivity, and 3D IP Inversion Results 257 Appendix 4B - 2D DC Resistivity and Induced Polarization Inversion Results for Hislop 257 Appendix 4C - Observed versus Prediced Data for Hislop Inversion Models 257 vii List of Tables Table 2.1. Geophysical characteristics of Archean orogenic gold deposits 21 Table 2.2. Summary of the principal rock types found in the Hislop deposit area, and associated mineralogy 30 Table 2.3. Ranges of resistivity and chargeability for rock types similar to those occurring in the Hislop deposit area (data from Telford et al., 1990) 42 Table 2.4. Densities of the common minerals in Hislop deposit rocks (from www.mindat. org) 49 Table 2.5. Statistical data for prospective rocks at Hislop, and cut-off values used for querying physical property data 64 Table 2.6. Results from magnetic susceptibility and density queries of the Hislop physical property dataset 65 Table 3.1. Characteristics of Archean orogenic gold deposits 91 Table 3.2. Physical property values used in synthetic modeling 96 Table 3.3. Synthetic survey parameters 100 Table 3.4. Synthetic inversion parameters 100 Table 3.5. Model differences calculated between recovered and true models (the lowest model differences for each geophysical method are highlighted with bold text) 102 Table 4.1. Typical and anomalous physical property ranges for principal rock types occurring in the Hislop deposit area 142 Table 4.2. Survey parameters 155 Table 4.3. Inversion parameters 156 Table 4.4. GiFtools ModelBuilder options chosen for building Hislop reference models. 163 viii List of Figures Figure 1.1. Flow chart illustrating the role of physical properties in the inversion process. 4 Figure 2.1. Approximate location of the Hislop study area in the Abitibi greenstone belt of the Superior Province 22 Figure 2.2. Geology of the Hislop deposit area as interpreted by Power et al. (2004) from high resolution aeromagnetics 24 Figure 2.3. Cross section looking northwest through the Hislop deposit 25 Figure 2.4. Geology, alteration, magnetic susceptibility, and gold grade logs 33 Figure 2.5. Magnetic susceptibility histograms for the five main rock types found in the Hislop deposit area 34 Figure 2.6. Magnetic susceptibility histograms showing susceptibility data for a) least- altered and altered ultramafic volcanic rocks, and b) least-altered and altered mafic volcanic rocks 36 Figure 2 7. Magnetic susceptibility histograms showing susceptibility data for a) least- altered and altered intermediate dikes, b) least-altered and altered syenitic dikes, and c) least-altered and altered porphyritic rhyolite dikes 37 Figure 2.8. Density histograms for the five main rock types found in the Hislop deposit area 38 Figure 2.9. Density histograms showing density data for a) least-altered and altered ultramafic volcanic rocks, and b) least-altered and altered mafic volcanic rocks 39 Figure 2.10. Density histograms showing density data for a) least-altered and altered intermediate dikes, b) least-altered and altered syenitic dikes, and c) least-altered and altered porphyritic rhyolite dikes 40 Figure 2.11. Resistivity histograms for Hislop deposit rocks 42 Figure 2.12. Chargeability histograms for Hislop deposit rocks 43 Figure 2.13. Positive correlation between modal magnetite in Hislop rock samples (as derived from XRD analysis) and magnetic susceptibility. For calculated correlation coefficients see Appendix 2G (all rock types) 44 ix Figure 2.14. Magnetite grains (reflective grains in lower image) are destroyed within a carbonate altered zone surrounding a carbonate vein in a mafic volcanic rock from Hislop 45 Figure 2.15. Modal magnetite versus total Fe-rich carbonate abundance for all Hislop samples with measured quantities of these minerals 46 Figure 2.16. Histograms showing distribution of susceptibility for fine- and medium grained a) mafic volcanic rocks, and b) ultramafic volcanic rocks 48 Figure 2.17. Density increases for Hislop rocks with an overall increase in the abundance of Fe-rich carbonate. For calculated correlation coefficients see Appendix 2G (all rock types) 51 Figure 2.18. Measured versus calculated density for Hislop rocks 52 Figure 2.19. Porosity of ultramafic volcanic rocks at Hislop decreases with carbonate- related hydrothermal alteration 53 Figure 2.20. No relationships are indicated between porosity and density for mafic volcanic rocks at Hislop 53 Figure 2.21. Resistivity versus magnetic susceptibility 56 Figure 2.22. Resistivity versus density 57 Figure 2.23. A plot of porosity versus resistivity shows that annealing of ultramafic rocks due to precipitation of carbonate minerals during hydrothermal alteration brings about a decrease in porosity and a corresponding increase in resistivity 58 Figure 2.24. A weak positive correlation exists between pyrite abundance and chargeability 58 Figure 2.25. A negative correlation between chargeability and porosity 59 Figure 2.26. Magnetic susceptibility plotted against density for Hislop samples 60 Figure 2.27. Carbonate-alteration destroys magnetite in a) mafic and b) ultramafic volcanic rocks 62 Figure 2.28. Magnetic susceptibility histograms comparing data from Hislop rocks, and equivalent rocks from surrounding regional areas 68 Figure 2.29. Density histograms comparing data from equivalent rock types from Hislop rocks, and equivalent rocks from surrounding regional areas 69 x Figure 2.30. A comparison of magnetic susceptibility data associated with least-altered and carbonate-altered mafic rocks from the Hislop deposit, and from the greater surrounding area 70 Figure 2.31. A comparison of density data associated with least-altered and carbonate- altered mafic rocks from the Hislop deposit, and from the greater surrounding area 72 Figure 3.1. Cross-section looking northwest through the Hislop deposit 89 Figure 3.2. Plot of magnetic susceptibility versus density for major rock units at Hislop. 92 Figure 3.3. Plot of magnetic susceptibility versus density for variably altered mafic volcanic rocks, and variably altered ultramafic volcanic rocks from Hislop 93 Figure 3.4. Resistivity histograms for Hislop deposit rocks 94 Figure 3.5. Chargeability histograms for Hislop deposit rocks 94 Figure 3.6. a) 3D geological model based on the geologic setting of the Hislop gold deposit. b-e) North-facing cross-sections through 3D physical property models generated from the geologic model 99 Figure 3.7. Starting model and unconstrained magnetic inversion result for the ‘Hislop like’ magnetic susceptibility model 104 Figure 3.8. Starting models and magnetic inversion results with changes made to geometry of the target body 105 Figure 3.9. Magnetic inversion results for starting models with different physical property contrasts between the target and host rocks 106 Figure 3.10. Starting model and unconstrained gravity inversion result for the ‘Hislop like’ density contrast model 108 Figure 3.11. Gravity inversion results with changes made to geometry of the target body. 108 Figure 3.12. Inversion results with different physical property contrasts between the target and host rocks 110 Figure 3.13. Starting model and unconstrained DC resistivity inversion result (conductivity model) for the ‘Hislop-like’ resistivity model 113 Figure 3.14. DC resistivity inversion results (conductivity models) with changes made to physical property contrasts, and to the geometry of the target body 114 xi Figure 3.15. Starting model and unconstrained IP inversion result for the ‘Hislop-like’ chargeability model 116 Figure 3.16. IP inversion results with changes made to physical property contrasts and geometry of the target body 117 Figure 3.17. Inversion results for the Hislop-like susceptibility model after constraints applied 122 Figure 3.18. Inversion results for the Hislop-like conductivity model after constraints applied 124 Figure 3.19. Inversion results for the Hislop-like susceptibility model with depth weightings reduced 126 Figure 3.20. Comparison of a dipole-dipole electrode configuration and a Schlumberger configuration which resembles a Realsection array 127 Figure. 3.21. DC resistivity inversion result for resistivity data collected via a dipole- dipole survey 128 Figure 4.1. Geological map of the southwest Abitibi greenstone belt 139 Figure 4.2. Geology of the Hislop deposit area 140 Figure 4.3. Cross section looking Northwest through the Hislop deposit 141 Figure 4.4. Magnetic susceptibility plotted against density for the major rock types at Hislop 143 Figure 4.5. Magnetic susceptibility plotted against density for a) mafic and b) ultramafic volcanic rocks from the Hislop deposit area 144 Figure 4.6. Resistivity histograms for Hislop deposit rocks 146 Figure 4.7. Chargeability histograms for Hislop deposit rocks 148 Figure 4.8. Chargeability plotted against pyrite abundance for Hislop samples 148 Figure 4.9. Chargeability versus porosity for mafic rock samples from Hislop 149 Figure 4.10. Extents of magnetic data used in the deposit-, local-, and regional-scale magnetic inversions 152 Figure 4.11. Data used in regional-scale magnetic inversion 153 Figure 4.12. Data used in local-scale magnetic inversion 154 Figure 4.13. Data used in deposit-scale magnetic inversion 154 Figure 4.14. Data used in regional-scale gravity inversion 158 xii Figure 4.15. Location of DC resistivity and IP lines used for 3D DC resistivity and IP inversions. Local mine grid line numbers shown. See Figure 4.2 for geology legend.. 159 Figure 4.16. Extents of inversion model volumes, with cross-section location indicated. 164 Figure 4.17. North-south cross-section through the regional-scale unconstrained magnetic inversion result 166 Figure 4.18. Isosurface model from regional scale magnetic inversion results 167 Figure 4.19. North-south cross-sections through the local-scale a) unconstrained, and b) constrained magnetic inversion results 169 Figure 4.20. Isosurface model from local magnetic inversion results 171 Figure 4.21. North-south cross-sections through the deposit-scale a) unconstrained, and b) constrained magnetic inversion results 172 Figure 4.22. Isosurface model from deposit-scale magnetic inversion results 174 Figure 4.23. North-south cross-section through the regional-scale gravity inversion result, inverted with non-located constraints 176 Figure 4.24. Isosurface density model from regional-scale gravity inversion results... 177 Figure 4.25. North-south cross-section through the deposit-scale a) DC resistivity and b) IP inversion results 178 Figure 4.26. Isosurface models for deposit-scale a) conductivity, and b) chargeability results 179 Figure 4.27. Result for a physical property query targeting low magnetic susceptibility- low density cells within the regional-scale common earth model 183 Figure 4.28. Result for a physical property query targeting high magnetic susceptibility - high density cells within the regional-scale common earth model 184 Figure 4.29. Result for a physical property query targeting low magnetic susceptibility - high density cells within the regional-scale common earth model 185 Figure 4.30. Result for a physical property query targeting low magnetic susceptibility - high chargeability cells within the local-scale common earth model 186 Figure 4.31. Result for a physical property query targeting low magnetic susceptibility - high chargeability cells within the deposit-scale common earth model 187 xlii Acknowledgements Thanks to my supervisor Richard Tosdal for the always timely feedback and the numerous edits of my chapters, and for support during the course of the thesis research. Thanks to all the geologists and geophysicists who provided input through discussions and edits, including Claire Chamberlain, Shane Ebert, Rob Eso, Ken Hickey, Peter Lelievre, Doug Oldenburg, Nicolas Pizarro, and Victoria Sterritt. I am most especially gracious for all the geophysics help provided by Nigel Phillips and Nick Williams. Thanks for being so generous with your time, and for having so much patience. The sponsors of the MDRU-GIF project, including Geoinformatics Exploration Inc., Anglo American, Anglo Gold Ashanti, Barrick, BHP Billiton, Kennecott Exploration, Teck, Vale Inco, and Xtrata are thanked. Additional funding was provided by an NSERC postgraduate scholarship. A Hugo Dummett Mineral Discovery Fund grant from the Society of Economic Geologists provided funding for XRD and physical property analyses. Geologists and geophysicists at Geoinformatics Exploration Inc., and St. Andrew Goldfields Ltd., especially Darren Holden (Geoinformatics), and Wayne Reid (former exploration manager at St. Andrew Goldfields), are thanked for providing data, general information on the Hislop deposit and surrounding area, and use of the offices and core yard at Stock. Thanks to Brian Atkinson, Dave Truscott, and Ken Kryklywy for geological tours in and around Timmins. Elisabetta Pani is thanked for XRD analyses, and Mati Raudsepp and Sasha Wilson for help on the SEM. Kelly Russell, Steve Quane, and Krista Michol, provided guidance for density and porosity data collection. Lisa Swinnard, Lorraine Tam, and Marcia Wilson, and were all thorough and well-organized in collecting density data. Arne Toma and Kane Smith are thanked respectively, for helping me deal with various computer, and fmancial matters. xiv Thanks to my friends at UBC. You are all so smart and inspiring. And you did an amazing job of decorating the office with a stunning array of wine bottles. To my legion of former officemates, you ladies always kept it fun and funny, and kept geological conversation at tolerable levels. To Victoria and Kirsten and Amber, thanks for listening to my ramblings, and hitting the slopes and the waves hard with me between stints in the office. Thanks to my family for helping me to get back home once in a while, for their interest in my life, and of course for their constant encouragement. To Billy: thanks for helping me with core lifting and susceptibility measuring in Timmins, for cutting my rocks for me at UBC, for not (really) asking me if I am done yet, for being ok with me not having goals or plans, and for keeping everything together when I was too busy, i.e. almost all the time. Maybe you are the nicest person I will ever meet? xv Co-authorship Statement Chapters 2-4 were written as independent manuscripts that will be submitted for publication to journals focusing on applied uses of geophysics, for exploration or otherwise. Each chapter involved some input from others, in the form of instruction, discussion, editing of the work, or data collection. Those who played the largest roles in collaborating are to be recognized as co-authors on the submitted manuscripts. Their contributions are outlined below. Chapter 2: Physical properties of rocks in an Archean orogenic gold environment Authors: Dianne Mitchinson, Nigel Phillips, Elisabetta Pani, Richard Tosdal Nigel Phillips, a former research associate at the Mineral Deposit Research Unit, of the Department of Earth and Ocean Sciences helped to interpret some of the physical property data, and edited parts of the manuscript, as well as related posters and abstracts. Elisabetta Pani, researcher in the Department of Earth and Ocean Sciences, collected X ray diffraction data for the Hislop deposit suite of samples, and analyzed the data using Rietveld methods to yield mineral abundance data. My thesis supervisor Richard Tosdal contributed suggestions, and provided numerous edits of this chapter. Chapter 3: Detecting gold-related geology in Archean orogenic gold environments using geophysical inversion: a synthetic modeling study based on the Hislop gold deposit, Ontario Authors: Dianne Mitchinson, Nigel Phillips Nigel Phillips initiated the idea of completing synthetic modeling to explore the capabilities of inversion in the studied geologic setting, and provided suggestions for xvi possible variations on starting models, and on inversion parameters. He provided guidance and instruction with respect to forward and inverse modeling techniques using the University of British Columbia Geophysical Inversion Facility (UBC-GIF) inversion codes. He also edited the work. Chapter 4: 3D inversion of magnetic, gravity, DC resistivity, and induced polarization data over the Hislop gold deposit, south-central Abitibi greenstone belt Authors: Dianne Mitchinson, Nigel Phillips, Nick Williams Nigel Phillips familiarized me with inversion codes, and with the inversion modeling process in general. He helped to organize geophysical data, and provided guidance and suggestions throughout the 2D and 3D DC resistivity and induced polarization modeling. Nick Williams aided with the management and manipulation of the large datasets involved, instructed me on the use of his program ModelBuilder, and provided discussion on a number of the inversion model results. xvii Chapter 1: Introduction 1.1. COMBINING GEOLOGY AND GEOPHYSICAL INVERSION FOR MINERAL EXPLORATION 1.1.1. Geophysics and mineral exploration Geophysical techniques are used regularly to aid or supplement geologic mapping in areas where outcrop is limited. In addition to delineating surface geology with geophysics, it is possible to investigate geology at depth, where otherwise subsurface geology must be inferred from maps and structural measurements, or by drilling. Geophysics has become an especially important tool in mineral exploration. Many mineral deposit targets produce strong geophysical signatures due to high abundances of oxides and sulfides, allowing them to be distinguished from their host rocks. Geophysics is so prolific in the field of mineral exploration because of the significant amount of information it can provide for low costs (Phillips et a!., 2001). Regional geophysical data, usually magnetic and gravity data covering hundreds of kilometers of ground, is commonly available for free, or at an insignificant cost, from government geological surveys. From this data geology can be inferred, and large exploration targets spotted. With advanced stages of mineral exploration, an exploration company can have more fine-scale geophysical surveys completed for a higher cost, however, the price is minimal compared to the cost of drilling. Traditionally, geophysical data collected at the surface or from boreholes is interpreted directly after standard filtering and corrections. Estimations of sizes and shapes of features are made based on known relationships between sources and the measurement location, and through forward modeling. The relatively recent development of robust geophysical inversion methods for calculation of 3-dimensional physical property models of the subsurface allows petrophysically distinct geological features to be located in 3D space, and their geometry to be delineated at significant depths of up to 1 thousands of meters. These methods are becoming a staple in the mineral exploration industry as it is thought that most near-surface mineral deposits have been discovered, and that future resources exist at depth. 1.1.2. Mineral Deposit Research Unit — Geophysical Inversion Facility project This PhD project was completed alongside a number of others under the Mineral Deposit Research Unit — Geophysical Inversion Facility (MDRU-GIF) joint research initiative. The MDRU-GIF project was a collaborative project involving researchers and students from the University of British Columbia’s (UBC) Mineral Deposit Research Unit, and the Geophysical Inversion Facility, in addition to ten mineral exploration industry sponsors. The formal project began in 2003, and ended in the spring of 2007. The overlying objective of the MDRU-GIF project was to enhance inversion-based exploration and generate more robust 3D subsurface models through effective combination of geology, physical properties, and geophysical information. A number of more specific themes were encompassed within this principal objective including: relating physical properties to geology and geological processes (Sterritt, 2006), scaling physical property data for use at larger scales of inversion (Pizarro, 2008), and developing methods of more effectively incorporating prior geological information into geophysical inversions to yield more geologically realistic models (Phillips et al., 2007; Lelievre et al., 2008; Williams, 2008). The MDRU-GIF projects were based on data from a range of mineral deposit types including kimberlitic diamond, magmatic sulfide, orogenic gold, volcanogenic massive sulfide, and porphyry deposits, and considered different stages in exploration from regional reconnaissance to deposit delineation. This PhD project focused on the application of geophysical inversion methods to exploration in the Archean orogenic gold environment, for a range of scales of exploration. 2 1.1.3. Inversion in the Archean orogenic gold environment The Hislop deposit, a gold deposit in the south-central Abitibi greenstone belt, acted as a representative orogenic gold deposit for this work. Although the deposit is small, and was only mined for a short period, it was a good candidate for a case study deposit for this research for a number of reasons. Due to extensive exploration in the Hislop deposit area, and in nearby surrounding areas, there is a large amount of geophysical data available for use in geophysical inversions. There are numerous driliholes available for reconnaissance work on the local geology. Additionally, the area has been mapped and modeled recently (Berger, 1999 and 2002; Power et al., 2004; Reed, 2005; Mueller et al., 2006), and inversion results can be compared to known geology. Finally, it may be possible to apply concepts and results from this work to other areas, as the geology of the deposit is characteristic of other orogenic gold deposits both locally, and globally. The intent of this PhD project was to apply knowledge of orogenic gold models, of local greenstone belt geology, and of the Hislop deposit, to optimize the inversion process for this specific mineral deposit setting. The desired outcome was to generate subsurface models that are consistent with known geology in order to be able to interpret results with confidence. The project is, in essence, a multi-faceted case study, which broaches many of the themes of the MDRU-GIF project, and covers a number of stages that comprise the inversion process. PhD research encompassed understanding physical property — geology relationships, completing synthetic modeling to determine inversion imaging capabilities, and carrying out unconstrained and constrained inversions of actual geophysical data collected over the Hislop deposit. The role of physical properties in inversion is emphasized throughout this work, as they ultimately quantitatively link geology to geophysics (Fig. 1.1). Having an understanding of relationships between geology and physical properties is important for constraining geophysical inversions, determining if physical property values composing 3 model results are reasonable, and of course for interpreting geology from the recovered models. The entire process represented by the work in this thesis should be analogous to the process that an exploration company might follow if embarking on completing inversion work for a prospect, or even a more well-understood deposit where continuations of ore zones or other nearby targets are sought. Figure 1.1. Flow chart illustrating the role of physical properties in the inversion process. Physical Properties Manec SupfliIIty vs Dev, of Ibdop Rocks 0 Geophysical Modeling and Inversion 4 1.2. BACKGROUND TO GEOPHYSICAL INVERSION The UBC-GIF inversion programs are primarily used as modeling and exploration tools in this project. This thesis does not go into detail regarding the mathematics behind the inversion codes. However, in order to understand how prior geologic information can be accommodated in the inversion, and in order to appreciate features and anomalies that are manifested in the inversion results, it is important to have a general knowledge of how the codes work. Geophysical inversion can be considered the opposite process to forward modeling. Forward modeling involves generating data for a known subsurface physical property distribution. Forward modeling is sometimes used to determine the effect a specific source within the subsurface has on a measured geophysical signal. Geophysical inversion involves estimating a subsurface physical property distribution based on an observed geophysical dataset. In this case the data are known, and the location, and physical property value of the source must be calculated. To calculate a 3D subsurface model, a volume representing the earth is discretized into many model cells. A reference model or starting physical property value is assigned to the earth and physical properties within cells are perturbed over numerous iterations to attempt to fit the observed geophysical data (either collected at surface or from boreholes). The user specifies a misfit, represented by Equation 1. The misfit is essentially a measure of the difference between the observed data, and the data predicted by the recovered inversion model. Because there are far more unknowns (model cells) than there are data, there are an infinite number of possible solutions to the inversion problem. This non-uniqueness is alleviated by the addition of more information to the problem. Results can be constrained by formulating the inversion to achieve a model with particular characteristics, based on prior geological knowledge. This information is incorporated into the problem through the model objective function. 5 2 N d0bs_dPa ød(m)= [1 Equation 1. Where N is the number of geophysical data, d102s is the observed data at location I, df”’ is the predicted data at location i, and e is the standard deviation. For default inversions, the model objective function specifies that the desired model is one that is close to a given reference or background model, and is smooth in all directions. The inversion is guided toward a result honoring these specifications. The model objective function is represented in Equation 2, showing only the function controlling closeness to the reference value, and the function controlling smoothness in the x direction. These default parameters can be modified when more specific information is known about the geology. The reference value can be modified and its degree of influence on the result can be manipulated (c), and directionality can be invoked by increasing smoothing in different directions by varying amounts (c). The resulting inversion model is only acceptable if, data generated when the model is forward modeled (the predicted data) is within error of the observed data. In effect, there is no ‘best’ model, but likely a range of models that satisfy the model criteria and are geologically reasonable. øm =a f(m_mo)2dx+ax(m_mo)dx Equation 2. Where c’ is the alpha weighting determining the degree of closeness to reference model, c determines smoothing in the x direction, m is the model, and mo is the reference model. In the full equation, functions in the same form as the x-smoothing function exist for the y and z directions. 6 Detailed inversion procedures and equations are found in Li and Oldenburg (1996, 1998, and 2000). 1.3. ARCHEAN OROGENIC GOLD - GEOLOGIC AND GEOPHYSICAL BACKGROUND 1.3.1. Background on Archean orogenic gold Recent comprehensive summaries of orogenic gold deposits are given in Groves et al. (1998), Hagemann and Cassidy (2000), Goldfarb et a!. (2005), and Robert et al. (2005), and characteristics significant to the thesis are generalized here. Orogenic gold deposits are epigenetic, structurally controlled gold deposits that are hosted in orogenic belts. They are generally accepted as having formed during late stages of continental collision. Most of the discovered orogenic gold deposits in the world occur in greenstone belts situated on Archean cratons in North America, Australia, and southern Africa. Archean orogenic gold deposits typically occur proximal to large, crustal-scale faults, which are thought to represent the conduits that transported gold-bearing fluids to near-surface from depth. These deposits can occur in any host lithology, however there appears to be a common spatial relationship to felsic intrusive rocks, perhaps due to their brittle nature and ability to develop fractures, and to Fe-rich rocks, which may promote sulfidation causing gold precipitation. Hydrothermal fluids carrying gold are typically C02-rich and this is reflected in the carbonate-rich alteration mineral assemblages that accompany mineralization. Gold is most commonly hosted within or proximal to quartz carbonate veins, but may also occur in association with disseminated sulfides in spatial proximity to faults or shear zones. 7 1.3.2. The Hislop gold deposit The Hislop deposit is found in the gold and base-metal rich Abitibi greenstone belt of the Superior Province of Canada. It lies near the Porcupine Destor Deformation Zone (PDDZ), a regionally important structure with respect to gold mineralization. The general geology of the Hislop Township was mapped by Prest (1956), and more recently by Berger (1999). A geological map of the eastern Timmins area based predominantly on interpretation of high resolution aeromagnetic data, was compiled by Geoinformatics Exploration Inc. Geoinformatics also compiled an extensive database of geologic logs from drilicore derived from exploration programs run by the companies that have explored the Hislop property over the last 75 years. Berger (2002) completed an assessment on the geology and geochemistry of rocks along the eastern portion of Highway 101 (the ‘Golden Highway’), which follows the PDDZ, that included an overview of the geologic setting of mineral deposits along this corridor. The most detailed work on the Hislop deposit was completed by geologists working at St. Andrew Goldfields Ltd at the time of mining. Some underground maps were made, and petrographic and lithogeochemical work completed. At the time of the commencement of this project, the Hislop mine was closed, and most of the geologists who had worked at the mine no longer were with St. Andrew Goldfields. Much of the data on the deposit that was collected, some in digital and some in hard copy form, was scattered and difficult to compile. An internal report with significant detail on the various mineralized zones on the Hislop property was provided for this project as a reference (Roscoe and Postle, 1998). For this project, ten drillholes were re-logged, and a limited amount of geologic mapping was completed at the flooded Hislop West Area open pit, and on select outcrops in the vicinity. In general, the Hislop deposit is hosted within a series of metamorphosed mafic and ultramafic volcanic rocks. The area is structurally complex with numerous tight folds 8 and faults paralleling the regional structural trend. Gold is spatially related to a contact between a syenite dike and an ultramafic volcanic unit. Gold is refractory within disseminated pyrite, and mineralization is associated with carbonate and muscovite alteration. St. Andrew Goldfields Ltd. currently own the Hislop deposit property. The deposit is a relatively small gold deposit only mined for a few years total, producing just over 400 000 tonnes of ore, grading between 2.33 and 5.55 grams per tonne. Further gold potential has been indicated by recent drilling and sampling programs (www.standrewgoldfields.com). 1.3.3. Geophysics and gold Geophysics constitutes a useful tool in greenstone-hosted gold settings since these environments are commonly characterized by scarce outcrop. In the area between the main Timmins gold camp and the Ontario-Quebec border, where the Hislop deposit is situated, there is minimal topography. The area is heavily forested, and covered with numerous lakes. Although geophysics is heavily relied on for geologic mapping and exploration for a variety of mineral deposits in these settings, gold deposits are a notoriously elusive geophysical target. The deposits are typically low grade, and locally restricted, resulting in a poor petrophysical contrast between the target and its host rocks. Nonetheless, other geological features known to be spatially related to gold, such as host rocks, hydrothermal alteration, or sulfide mineralization, might provide petrophysically distinct targets. The geophysical methods most successfully applied for gold exploration have been DC resistivity and induced polarization (IP) methods. These methods detect conductive and chargeable sulfides commonly associated with orogenic gold. Some 9 examples of the use of these methods in gold exploration are given in Seigel et al. (1984), Johnson et al. (1989), Doyle (1990), and Halloff and Yamashita (1990) There are limited case studies using inversion in this mineral deposit environment. Some recent work includes that of Kowalczyk et al. (2002) Mira Geoscience (2005a and 2005b), and Meuller et al. (2005). It is hoped that the work completed for this PhD will contribute to an advanced understanding of the application of geophysical inversion techniques in the Archean orogenic gold setting. 1.4. PROJECT OBJECTIVES The overlying goal of this research was to optimize the geophysical inversion process to explore for gold-related rocks in the Archean orogenic gold setting. The first step in achieving this goal was to identify relationships between geology, specifically gold-related geology, and physical properties, and to delineate the key geological processes that lead to these relationships. Secondly, synthetic modeling was used ‘to determine if typical gold—related features can be regularly detected by inversion, and if inversion parameters can be modified to improve their detection. The final stage of the work involved applying prior geological and physical property knowledge to the inversion of four geophysical datasets covering the Hislop deposit area. The results of this PhD research are presented in three chapters that correlate with each of the three research stages. The thesis objectives are summarized here by chapter. Chapter 2. Initial research involved defining relationships between geology and physical properties. As mentioned, this information is critical to any geophysical or inversion work. It is obviously important with respect to interpreting results. However, it is also valuable for constraining inversions, for identifying if inversion results are sensible, and for building synthetic physical property models to test hypotheses. Magnetic susceptibility, density, resistivity, and chargeability data were collected for Hislop rock 10 samples. The goals of this initial physical property work were to outline the physical property ranges for the main rock types at Hislop, to understand any trends within physical property data, and to determine if prospective rocks could be distinguished from barren rocks based on physical properties. Additionally, to establish whether the results from this work can be applied to geophysical exploration in other areas, physical property data was compared to a large regional dataset, and data from greenstone belts in Australia. Chapter 3. Although physical property work might indicate that certain gold-related rocks have unique physical property ranges, allowing them to be distinguished from likely barren rocks, these targets may still be undetectable through inversion. This may be attributable to: geophysical survey design, data spacing, data errors, inversion discretization, inversion sensitivities, and smoothing typical in inversion results. Synthetic forward and inverse modeling tests the effectiveness of inversion to delineate desirable features in the subsurface at deposit-scales of exploration. A model based on the Hislop deposit is used, however, variations are made to the initial model and to the applied inversion parameters to explore outcomes. The research aimed to determine: whether desired targets can be imaged using inversion, whether inversion parameters could be manipulated to get a better result, which geophysical datasets yield the most useful information about the subsurface, which are best for detection of gold-related features, and what limitations exist for inversion in this setting. Chapter 4. Chapter 4 presents results from inversion of four geophysical datasets (magnetic, gravity, DC resistivity, and IP) over the Hislop deposit. Where prior physical property data is available, inversions are constrained locally and globally to generate models more consistent with known geology. The main goals of this work was to examine the subsurface geology of the Hislop deposit area, to attempt to image specific geologic units or packages of rock, to locate key geologic structures in the subsurface, and most importantly, to identify prospective areas for exploration. Geophysical datasets most 11 useful for mapping geology, and for isolating mineral exploration targets were identified. Knowledge gained from physical property work and from synthetic modeling was invoked to assess and interpret inversion results. Chapters 2 to 4 form the basis for three manuscripts to be submitted to mineral exploration-related, or applied geophysical journals. As the three chapters represent three separate deliverables, there is some overlap in information between them. 12 REFERENCES Berger, B.R., 1999, Geological investigations along Highway 101, Hislop Township: Ontario Geological Survey, Summary of Field Work and Other Activities 1999, Open File Report, 6000, p. 5-1 — 5-8. Berger, B.R., 2002, Geological synthesis of the Highway 101 area, east of Matheson, Ontario: Ontario Geological Survey, Open File Report 6091, 124 p. Doyle, H.A., 1990, Geophysical exploration for gold — a review: Geophysics, v. 55, p. 134-146. Goldfarb, R. J., Baker, T., Dube, B., Groves, D.I., Hart, C.J.R., and Gosselin, P., 2005, Distribution, character, and genesis of gold deposits in metamorphic terranes: in Hedenquist, J.W., Thompson, J.F.H., Goldfarb, R.J., and Richards, J.P., eds., 100th Anniversary Volume, Economic Geology, v. 100, p. 407-450. Groves, D.I., Goldfarb, R.J., Gebre-Mariam, M., Hagemann, S.G., and Robert, F., 1998, Orogenic gold deposits: a proposed classification in the context of their crustal distribution and relationship to other gold deposit types: Ore Geology Reviews, v. 13, p. 7-27. Hagemann, S.G., and Cassidy, K.F., 2000. Archaean orogenic lode gold deposits, in Hagemann, S.G., and Brown, P.E., eds., Gold in 2000, Society of Economic Geologists, Reviews in Economic Geology, v. 13, p. 9-68. Halloff, P.G., and Yamashita, M., 1990, The use of the IP method to locate gold-bearing sulfide mineralization, in Fink, J.B., Sternberg, B.K., McAlistar, E.O., Weiduwilt, W.G., and Ward, S.H., eds., Induced Polarization: applications and case histories, Society of Exploration Geophysicists, Tulsa, Ok., p. 227-279. 13 Johnson, I., Webster, B., Matthews, R., and McMullen, S., 1989, Time-domain spectral IP results from three gold deposits in northern Saskatchewan: CIM Bulletin, v. 82, p. 43- 49. Kowalczyk, P., Thomas, S., and Visser, 5., 2002, 3D inversion of resistivity and IP data, two case studies from mineral exploration:, SEG International Exposition and 72’ Annual Meeting extended abstract, 4 p. Lelievre, P., Oldenburg, D., and Williams, N., 2008, Constraining geophysical inversions with geologic information: Society of Exploration Geophysicists, 2008 Annual Meeting, Las Vegas, extended abstract, p. 1223-1227. Li, Y., and Oldenburg, D.W., 1996, 3D inversion of magnetic data: Geophysics v. 61, p. 394-408. Li, Y., and Oldenburg, D.W., 1998, 3D inversion of gravity data: Geophysics, v. 63, p. 109-119. Li, Y., and Oldenburg, D.W., 2000, 3D inversion of induced polarization data: Geophysics, v. 65,p.1931-45. Mira Geoscience Limited, 2005a, Detectability of mineral deposits with electrical resistivity and induced polarization methods: Ontario Geological Survey, Miscellaneous Release—Data 181. Mira Geoscience Limited, 2005b, Detectability of mineral deposits with potential field methods: Ontario Geological Survey, Miscellaneous Release — Data 177. Mueller, E.L., Reford, S.W., Dawson, D.J.W., Morrison, D.F., Pawluk, C., Grant, J., Spector, A., Rogers, D.S., and Savage, T., 2006, Acquisition, inversion and presentation of geophysical data for geoscientific profiles in the Timmins—Kirkiand Lake area: 14 Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6189 , 28 p., 15 sheets. Phillips, N., Oldenburg, D., Chen, J., Li, Y., and Routh, P., 2001, Cost effectiveness of geophysical inversions in mineral exploration: Applications at San Nicolas: The Leading Edge, v. 20, p. 1351-1360. Phillips, N., Hickey, K., Lelievre, P., Mitchinson, D., Oldenburg, D., Pizarro, N., Shekhtman, R., Sterritt, V., Tosdal, D., and Williams, N., 2007, Applied strategies for the 3D integration of exploration data: KEGS Inversion Symposium, PDAC 2007, extended abstract, 9 p. Pizarro, N., 2008, Magnetic susceptibility scaling of rocks using geostatistical analysis: an approach to geologic and geophysical model integration: Unpublished M.Sc. thesis, University of British Columbia, 177 p. Power, W. L., Byrne, D., Worth, T., Wilson, P., Kirby, L., Gleeson, P., Stapleton, P., House, M., Robertson, S., Panizza, N., Holden, D. J., Cameron, G., Stuart, R., and Archibald, N. J., 2004, Geoinformatics evaluation of the eastward extension of the Timmins Gold Camp: Geoinformatics Exploration Inc., Unpublished report for St Andrew Goldfields Ltd. Prest, V.K., 1956, Geology of the Hislop Township: Ontario Department of Mines, Annual Report, 1956, v. 65, pt. 5, 51 p. Reed, L. E., 2005, Gravity and magnetic three-dimensional (3D) modeling: Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6163, 40 p., 4 sheets. Robert, F., Poulsen, K.H., Cassidy, K.F., and Hodgson, C.J., 2005, Gold metallogeny of the Superior and Yilgam Cratons, in Hedenquist, J.W., Thompson, J.F.H., Goldfarb, R.J., 15 and Richards, J.P., eds., 100th Anniversary Volume, Economic Geology, v. 100, p. 407- 450. Roscoe and Postle Inc., 1998, Hislop Mine Property, Roscoe and Postle Associates Inc., unpublished St. Andrew Goldfields Ltd. internal report, p. 66-89. Seigel, H.O., Johnson, I., and Hennessey, J., 1984, Geophysics the leading edge: Geophysics: the Leading Edge of Exploration, v. 3, p. 32-35. Sterritt, V.A., 2006, Understanding physical property—mineralogy relationships in the context of geologic processes in the ultramafic rock-hosted mineral deposit environment: aiding interpretation of geophysical data: Unpublished M.Sc. thesis, The University of British Columbia, 172 p. Williams, N.C., 2008, Geologically-constrained UBC—GIF gravity and magnetic inversions with examples from the Agnew-Wiluna greenstone belt, Western Australia: Unpublished Ph.D. Thesis, The University of British Columbia, 479 p. 16 Chapter 2: Physical properties of rocks in an Archean orogenic gold environment1 2.1. INTRODUCTION 2.1.1 Rationale In order for geophysical inversion to be knowledgeably interpreted, it is imperative to (1) have an understanding of the rock types, alteration, and mineralization that typify the geological environment, and (2) possess an understanding of the characteristic ranges of physical properties associated with this geology. Ideally, physical property studies should be conducted on the range of representative rock types from the geological environment of interest to try and understand how, why, and on what scales, physical properties in this environment vary. It is possible to refer to published datasets for typical physical properties of rock types in a specific environment, however this information is commonly limited and the effects of hydrothermal alteration on the protolith are rarely considered. Once a clear understanding of the relationships between various physical properties and rock types, alteration, and mineralization are established, this information can be used to interpret and guide geophysical inversions. If unique relationships are present and can be statistically characterized, physical property model data generated from inversion can be queried for prospective ranges, or filtered to yield mineralogical information (Williams and Dipple, 2005). Knowledge of physical property ranges typical of a given geological environment can indicate whether an inversion has yielded realistic values. Additionally, the inversion algorithm can be manipulated to incorporate prior physical property information to drive the inversion toward a result more consistent with expected geology (Ellis and Oldenburg, 1994; Li and Oldenburg, 1996). Understanding physical property behavior, and having confidence in the data being used to constrain 1 A version of this chapter will be submitted for publication. Mitchinson, D., Phillips, N., Pani, E., and Tosdal, D., 2009, Physical properties of rocks in an Archean orogenic gold environment. 17 inversions is critical; changing inversion parameters, or using reference models to constrain inversions, can change a model significantly (Phillips, 2002; Williams, 2006). A physical property study of the Hislop deposit aims to provide a detailed investigation into physical property relationships within an Archean orogenic gold deposit environment. Physical properties considered are magnetic susceptibility, density, resistivity, and chargeability. An important goal of these studies is to identify the physical property datasets, alone, or in combination, which are most effective in detecting Archean orogenic gold-related mineralization or proxies to mineralization. 2.1.2 Objectives The objectives of this research are to: 1. Review the key characteristics of Archean orogenic gold environments, and the geophysical methods commonly employed in exploration for them; 2. Characterize the principal host rocks, alteration characteristics, and styles of gold mineralization at Hislop; 3. Document relationships between physical properties and rocks at Hislop through petrographic work and mineral analyses; 4. Explain the controls on physical property variations; 5. Outline magnetic susceptibility, density, resistivity, and chargeability ranges that specifically characterize the host rocks, alteration mineral assemblages, and mineralization at Hislop; 6. Define the most useful physical properties for targeting potentially mineralized rocks at Hislop; 7. Assess whether physical property values are representative of Archean orogenic gold settings elsewhere. 18 2.2. BACKGROUND 2.2.1 Geology and geophysics of Archean orogenic gold deposits Geological characteristics of Archean orogenic gold deposits Orogenic gold deposits are epigenetic, structurally-controlled gold deposits hosted in metamorphosed orogenic belts (Groves et al., 1998). This work focuses specifically on the physical property analysis of rocks associated with orogenic gold deposits hosted within an Archean age greenstone belt setting. Although Archean orogenic gold deposits are not restricted to one particular rock type, spatial relationships to felsic intrusive rocks, and to Fe-rich mafic rocks are common (Hodgson and Troop, 1988; Hodgson, 1990; Groves and Foster, 1991; Goldfarb et al., 2005; Robert et al., 2005). Gold is thought to be transported in C02-rich fluids (Bohike, 1989; Ridley and Diamond, 2000) and as such, mineralization is usually associated with carbonate-rich alteration mineral assemblages (Fyon and Crockett, 1983; Kishhida and Kerrich, 1987; Meuller and Groves, 1991; McCuaig and Kerrich, 1998). Gold occurs most commonly within quartz- and carbonate- filled vein systems, and occurs less frequently as disseminated replacement zones, or as stockwork mineralization (Roberts, 1988; Hodgson, 1993; Hagemann and Cassidy, 2000; Goldfarb et al., 2005). The Archean gold deposits considered in this study do not include Archean-age placer, or banded iron formation (BIF)-hosted gold deposits. Archean orogenic gold deposits have for many years been an important source of gold in Australia, Africa, India, and North America (Goldfarb et al., 2005; Robert et al., 2005). With a rise in gold prices in recent years, there has been a revival in exploration for these types of deposits, and an initiative to improve exploration methods for their discovery. 19 Geophysical characteristics of orogenic gold deposits Gold is notoriously difficult to detect using geophysics (Seigel et al., 1984; Doyle, 1990). Although gold itself is dense and conductive (19.3 g/cm3, and 5 x S/rn, respectively; Doyle, 1990), it is usually only present in relatively small quantities in Archean orogenic gold deposits, in contrast to massive-style mineralization represented in volcanogenic massive sulfide deposits or nickel sulfide deposits, which form larger geophysical targets in distinct contrast to host rocks. Defining alternative targets, or indicators, with known relationships to gold, and sufficiently distinct physical property characteristics, is required to fully utilize geophysical tools (Seigel et al., 1984; Doyle, 1990). Geophysical methods used to target gold-associated structures, host rocks, and alteration zones include magnetics, gravity, electrical methods (DC resistivity and induced polarization), and electromagnetic methods. Table 2.1 lists various geological features commonly related to gold mineralization, and examples of the geophysical methods that are most effective in targeting them. Ideally some combination of techniques can be employed to target a variety of gold-related features at a particular locality, in order to prioritize areas of interest. Magnetics and induced polarization (IP) are historically the most useful methods in delineating lithologies, structures, alteration, and sulfide distribution related to gold mineralization. 2.2.2 Geology of the study area Regional geological setting The Superior Province of the Canadian Shield is the largest Archean craton on earth. It is composed of a number of northeast-trending, amalgamated volcano-plutonic, granitic-gneissic, and sedimentary terranes (Card and Ciesielski, 1986). Boundaries of the terranes, or subprovinces, are structural or metamorphic zones that juxtapose contrasting 20 geological and geophysical terranes (Card and Ciesielski, 1986; Card, 1990; Williams et al., 1991). The study area for this project is located within the south-central Abitibi subprovince, or greenstone belt (Fig. 2.1). Table 2.1. Geophysical characteristics of Archean orogenic gold deposits Feature Scale Geophysical character Methods of detection 1 Greenstone Regional Overall low, but ‘rough’ magnetic Airborne magnetics terranes 1000 kms character Granitoids commonly lower density than Airborne gravity greenstone 2 Large scale Regional to Low magnetic signature attributed to Airborne/ground magnetics faults district oxidation/alteration l00kms E . . . .High/low resistivity zones dependant on DC resistivity degree of annealing 3 Lithological District Various depending on physical properties Various depending on rock type of marker units 10 kms of rock type of interest interest 4 Hydrothermal Local Magnetic lows resulting from destruction Airborne/ground magnetics alteration 10 m of magnetite; less commonly magnetic highs, due to influx of Fe-rich fluids High resistivity if silicification DC resistivity 5 Mineralization Local Disseminated sulfide association with DC resistivity and Induced 10 m gold - conductive and chargeable Polarization Magnetic pyrrhotite Magnetics if pyrrhotite is main Fe- sulfide associated with gold I) Grant, 1985; Isles et al., 1989; Doyle, 1990; Williams et al., 1991; Gunn and Dentith, 1997; 2) Henkel and Guzman, 1977; Boyd, 1984; Grant, 1985; Doyle, 1990; Coggon, 1984; 3) Doyle, 1990; Groves eta!., 1984; Gunn and Dentith, 1997; Boyd, 1984; Grant, 1985; Flood eta!., 1982; 4) Holsner and Schneer, 1961; Grant, 1985; Harron et al., 1987; Doyle, 1990; Williams, 1994; Lapointe et al., 1986; Johnson et a!., 1989; Doyle 1990; 5) Johnson et al., 1989; Seigel et al., 1984, Doyle, 1990; Hallof and Yamashita, 1990; Dockery etal., 1984. Many of the gold deposits in Abitibi greenstone belt gold camps, like the Timmins-Porcupine, and Kirkland Lake camps, are spatially related to prominent, large scale crustal structures, including the east-west trending Porcupine-Destor Deformation Zone and Larder-Lake-Cadillac Deformation Zone (Colvine et al., 1988; Kerrich, 1989; 21 Hodgson and Hamilton, 1990; Jackson and Fyon, 1991). Most gold deposits are not localized by these larger “first order” faults, but by secondary or tertiary splays (Kerrich, 1989; Robert, 1990; Hodgson, 1993; McCuaig and Kerrich, 1998; Hagemann and Cassidy, 2000). Figure 2.1. Approximate location of the Hislop study area in the Abitibi greenstone belt of the Superior Province. Modified after Card and Ciesielski (1986). Hislop deposit geology and gold setting The Hislop deposit area is underlain mainly by interlayered mafic and ultramafic volcanic rocks (Fig. 2.2). The volcanic rocks are complexly folded and are presently aligned northwest-southeast. They are intruded by coarse-grained syenites, fine-grained quartz-feldspar phyric rhyolite dikes, and dacitic to andesitic dikes, usually along northwest-southeast trending faults (Prest, 1956; Berger, 1999; Power et al., 2004). io Location of Study Area Superior province 22 Gold is localized near the northeast and southwest contacts of an elongate, approximately 30 m -100 m wide, northwest-trending syenite (Cooper, 1948; Prest, 1956; Roscoe and Postle, 1998; Berger, 1999 and 2002), as depicted in the cross-section in Fig. 2.3. The majority of gold at Hislop is associated with disseminated pyrite within, what is recorded in mine and geological survey documents as, “carbonate-breccia”, south of the syenite (Cooper, 1948; Prest, 1956; Roscoe and Postle, 1998). The carbonate breccia is predominantly a strongly carbonate-altered brecciated equivalent of an ultramafic unit at Hislop. Gold also occurs to a lesser extent within quartz veinlets, stockworks and fractures in mafic volcanic flows north of the syenite (Roscoe and Postle, 1998). Generally, there is little gold within the syenite, with the exception of weak mineralization occurring within a zone approximately 3 m from the southern contact with carbonate breccia. (Cooper, 1948). High gold grades at Hislop are also associated with rhyolite porphyries, which are found as narrow, discontinuous intrusive bodies in mafic and ultramafic units south of the syenite (Fig. 2.2). A number of northeast-trending, sinistral separation cross-faults offset the syenite and bounding mafic and ultramafic flows in places (Cooper, 1948; Prest, 1956; Power et al., 2004). Gold-bearing zones widen, and gold grade commonly increases where these cross faults intersect mineralization along the syenite (Roscoe and Postle, 1998) Two principal mineralized zones, the Shaft Area and the West Area (Fig. 2.2), were mined by St. Andrew Goldfields, Ltd., at Hislop over three separate intervals between 1990 and 2006. In 1990 and 1991, 215 990 tonnes of ore grading 5.55 g/t were mined, between 1999 and 2000, 185 100 tonnes or ore grading 3.4 g/t were mined, and recently in 2006, 10147 tonnes of ore grading 2.33 g/t were mined (www.standrewgoldfields.com). 23 I I LDO — late diorite/dolerite _____ SSG - greywacke I SLO — mudstone - siltstone j Soc — sediment, undivided IFD/IFO — felsic intrusive dyke! felsic intrusive undivided I I 100 — intrusive, undivided ______ ISO — syenite intrusive, undivided I VFO — felsic volcanic, rhyolite, rhyodacite r, I VLJO — ultramafic volcanic, undivided 2 VMF — magnetic mafic volcanic I VMO — mafic volcanic, basalt, aridesite Figure 2.2. Geology of the Hislop deposit area as interpreted by Power et al. (2004) from high resolution aeromagnetics. Locations of two mined areas on the Hislop property (West Area open pit; Shaft Area underground) are outlined in red. Also shown are 10 drill holes (one overlapping) logged for this study. The cross-section shown in Figure 2.3 is based on core logging of three drill holes that were drilled in the West Area. 24 DDH H9601 DDH Ext 280, GK 280, and H9605 Figure 2.3. Cross section looking northwest through the Hislop deposit, showing locations of carbonate-dominated alteration and gold mineralization. Cross section interpreted from drill core logged from the West Area of the Hislop property. 2.3. METHODOLOGY 2.3.1 Field and mineralogical studies Ten drill holes from the 1996 and 1997 St. Andrew Goldfields Ltd. drill programs were re-logged for this study. Geology, alteration and structure were recorded down-hole (Appendix 2B). Surface geology at the West Pit was mapped. Small areas, approximately Multi-lithic Volcanic Breccia Lamprophyric Dike Intermediate Dike Porphyritic Rhyolite Dike Syenite Intrusive Mafic Volcanic Rock Ultramafic Volcanic Rock Fault Drill trace 25 10 m by 10 m were mapped in detail for outcrop scale studies of magnetic susceptibility (Appendix 2B). Petrographic, and mineralogical studies (scanning electron microscope and X-ray diffraction studies) allow for characterization of host rocks and alteration mineral assemblages within the Hislop deposit area. This work also constrains geological processes that control physical property variations. The presence, abundance, and composition of minerals, such as magnetite, pyrite, and carbonate, which have particularly significant influences on physical properties, were documented. Whereas petrographic and SEM work defines the minerals present in the various samples, quantitative XRD work using Rietveld refinement methods, described by Raudsepp and Pani (2003) and outlined in Appendix 2C, contributes relative mineral proportions for 37 samples at Hislop. This quantitative information is useful for comparisons to physical property data, and for calculations of density data. 2.3.2 Physical property measurements Magnetic susceptibility Magnetic susceptibility data from Hislop was recorded using a hand-held magnetic susceptibility meter, the Exploranium KT-9 Kappameter. Susceptibilities are reported in i0 SI Units. Magnetic susceptibility readings were taken every 5 m along drill core for all drill holes re-logged for this project. Measurements were made on all samples collected from drill core and from outcrop. Magnetic susceptibility readings were taken at 10 different points over each sample, and the average value was used in analyses of this data. Magnetic susceptibility readings were also collected systematically over six roughly 10 m2 grids over mapped outcrops to understand controls on susceptibility at the surface at outcrop scale. Typically 2-5 readings were taken at each site and the average was used. In total magnetic susceptibility was determined for 432 samples. Greater than 1000 additional readings were collected from drill core and 26 outcrop. Corrections applied to susceptibility measurements to account for core diameter, and split core intervals are outlined in Appendix 2C. The magnetic susceptibility dataset represents the largest physical property dataset from the Hislop physical property study. Density Density measurements were made for 414 drill core and hand samples from Hislop using the buoyancy or hydrostatic method and calculations outlined by Johnson and Olhoeft (1984). To calculate grain density: Pg = pw*Wi/(W1W2) where is grain density, W1 is the mass of the oven-dried sample in air, and W2 is the mass of the sample submerged in water. To obtain the mass of the sample in water, the sample is placed on a tray which is suspended from a weighing scale positioned above a small tank of water. The scale is tared with the tray hanging suspended in the water bath, and the sample is added to the submerged tray. Pw is the density of the water, which is assumed to be 1 g/cm3. Density is reported in g/cm3. Additional density measurements were made using an alternate method, the geometric method, to confirm data accuracy, and results are presented in Appendix 2C. For grain density calculations, porosity is not considered. However, for later interpretations of some trends in the Hislop physical property data, it was of interest to calculate porosity. Porosity is calculated from dry and saturated rock masses. Isolated porosity (inaccessible to air or water) is not accounted for by this method. The equation used is (Cas and Wright, 1987): = 100*(W3 - W1)/(W3- W2) 27 where 4) is porosity, W1 is the mass of the oven-dried sample in air, W2 is the mass of the sample submerged in water, and W3 is the mass of the water-saturated sample in air (Cas and Wright, 1987). Porosity is reported as %. Resistivity and chargeability Resistivity and chargeability data for 67 representative drill core and hand samples were measured by Zonge Engineering and Research Organization, Inc. Resistivity and chargeability measurements are collected simultaneously after samples have been moisture-saturated. They are made in time-domain. A current is established between opposite ends of the samples using a constant current transmitter, which conducts currents as low as 100 nA. Resistivity is calculated based on the length, and cross-sectional area of the sample, the amplitude of the current, and the change in potential recorded across the sample. Resistivity is reported in Ohm-rn. Conductivity can be calculated from resistivity by taking the inverse value. Conductivities are expressed in S/rn. The chargeability of a sample is based on the rate of decay of the voltage after the applied current is turned off. For the Hislop samples, it was determined using an 8 second period, measured during the 0.45 — 1.1 seconds window after the current is turned off. The resultant value is the average chargeability value of a sample based on 16 cycles. Chargeability is reported in milliseconds (ms). Large scale permeability, not necessarily exhibited in the drill core or hand sample, may control measurements made in the field. As such, measurements of resistivity taken on drill core or hand samples are commonly higher than measurements made in-situ (www.zonge.com/LabIP.html). This must be considered if sample-scale resistivity measurements are to be used to constrain geophysical inversions. Chargeability data collected from core or hand samples are thought to be sufficiently representative of larger scale measurements (www.zonge.com/LabIP.html). 28 2.4. DATA AND OBSERVATIONS 2.4.1. Hislop deposit rock types, hydrothermal alteration, and associated mineralogy Rock Types All rocks at Hislop have been metamorphosed to greenschist facies, however, for simplification purposes, the prefix meta- is herein ignored. The rock protoliths are recognizable based on textures and characteristic metamorphic mineral assemblages, and the protolith name is used hereafter. The five principal rock types at Hislop are, ultramafic volcanic rocks (predominantly komatiites), mafic volcanic rocks (tholeiitic basalts), intermediate (andesitic-basaltic) dikes, syenite intrusions, and feldspar (+1- quartz) porphyritic rhyolite dikes. Other rock types occurring less frequently in this area, which will not be discussed in detail, include mafic intrusions, lamprophyric dikes, and multi-lithic volcanic breccia units. Descriptions and typical mineralogy of the main Hislop rock types as determined through petrographic, scanning electron microscope, and X-ray diffraction work, are given in Table 2.2, and detailed results of XRD mineral abundance analyses are found in Appendix 2D. Hydrothermal alteration The most common hydrothermal alteration mineral assemblage at Hislop is a carbonate + muscovite rich assemblage that occurs predominantly in intermediate dikes, and mafic and ultramafic volcanic rocks. This is manifested as siderite or ankerite (grouped, and simplified herein as Fe-carbonate) + muscovite alteration in mafic volcanic rocks and intermediate dikes, and as either Fe/Mg-carbonate (ankerite to dolomite) + muscovite alteration, or magnesite (Mg-carbonate) + fuchsite (Cr-muscovite) alteration in ultramafic volcanic rocks. Carbonate + muscovite alteration was noted in drill core and outcrop to occur near faults and contacts, and in proximity to syenite and rhyolite 29 Table 2.2. Summary of the principal rock types found in the Hislop deposit area, and associated mineralogy. mag, +1- ph, +1- sp B. Fe-carbonate + muscovite: cb (dolo), qtz, ms, chi, +1-a b, +1- mic C. Magnesite + fi,chsite: mgs/dolo, qtz, ms (flich), chi, +/-il, +1- cr -sp ab, chi, aug, act, cal, qtz, ep, ms, clzo, uivo, mag B. Fe-carbonate: ab, mic, ank/dolo/sid, ms, ±1- py, +1- chi Least- altered Ultramafic volcanic rocks Hand Sample Description Dark black to brown (oxidized) in color; soft; commonly sheared; fine to medium grained textures; qtz +7- cb veins; relict spinifex textures adjacent to margins of some flows; relict cumulate textures represent internal parts of the flow. Dark grey-green to dark purple color; massive or pillowed (+l_ variolites) flows up to lOOm thick; fine to medium grained rock; pillows are well-preserved with qtz + cb filled amygdules increasing near margins; thin chl +1- ep altered selvedges. chl, dol, qtz, +1- tIc, +1- hbl, +1- mag Dark grey to mauve in color; massive, homogeneous, and fine-grained; hbl fsp phyric, hbl-phyric, or aphyric; phenocrysts approximately 1mm in length and euhedral; groundmass is very fine-grained and composed mainly of microlitic fsp. Pink to mauve in color; massive; very coarse-grained; composed of mesoperthitic potassium fsp and ab; fsp crystals up to 3 cm; rare interstitial mafic minerals. ab, cb, ma, qtz, .4- chl, +/-mag, +1- pyMineralogy Alteration Alteration mineralogy Massive; generally fsp (+7. qtz) porphyritic; aphanitic grey to pink groundmass composed of very fine grained fsp and qtz; fsp phenocrysts comprise 3% — 80% of the rock; qtz phenocrysts make up 2 - 5% of the rock; minor mafic minerals. ab, kspar, dol/ank/cal, py, +1- qtz, +7- ma, +1- chl sb, qtz, mic, chl, +1- dol ms, ab, sid, dol, qlz, chi, B. Fe-carbonate + albile (+7-) quartz: ab, dolo/ank, qtz, mic, chl, mag 4. Talc + chlorite: chl, tIc, dolo, cal, A. Fe-carbonate + muscovite - ank, A. Fe-carbonate + muscovite: cb, A. Muscovite: ab, mic, ms, ank/dol, A. Muscovite: ab, qtz, ms, cal, +7- py ms, ab, qtz, py B. Fe-carbonate: ab, qtz, ank, ms, +7-B. Fe/Mg-carbonate: ab, cb, ma, py 4-I- py, ab = albite; act = actinolite; ank = ankerite; aug = augite; cal = calcite; cb = carbonate; chl = chlorite; clzo = clinozoisite; cr-si = Cr-apinel; dol = dolomite; ep = epidote; fu = fuchsite (Cr-muscovite); hbl = hornblende; il = ilmenite; mag = magnetite; mgn = magriesite; mic = microcline; isis muscovite; ph = phiogopite; py = pyrite; qtz = quartz; ser = sericite; sp = serpentine; sid = siderite; tIc = talc; ulv ulvospinel; (Fe iron; Mg magnesium). sole: mineralogy based on I samplefor ach ofA and B intrusions. Most of the mined Hislop ore came from an Fe-carbonate-altered ultramafic breccia, as such, carbonate-related alteration is considered an important vector to mineralization. Carbonate + muscovite alteration is commonly mapped as a distal, pervasive alteration surrounding Archean orogenic gold deposits, and this is the case for many gold deposits elsewhere in the Abitibi greenstone belt (Fyon and Crockett, 1983; Hodgson, 1990). Fe-carbonate + albite alteration occurs at Hislop over narrow intervals within mafic volcanic rocks in drill core near some of the known high grade gold zones. Albite rich alteration assemblages occur proximal to gold at other gold deposits in the Abitibi, including the Holloway deposit, near the Ontario-Quebec border (Ropchan et al., 2002) and the Kerr-Addison deposit in the Kirkland Lake district (Kishida and Kerrich, 1987). Muscovite alteration is the predominant alteration affecting Hislop syenite intrusives and porphyritic rhyolite dikes. Fe-carbonate alteration affects these rocks to a lesser extent. The overall lack of Ca, Mg, and Fe in felsic rocks at Hislop hinders the formation of carbonate minerals when exposed to CO2 rich fluids, as discussed in Roberts (1988). Most prospective rock types and alteration at Hislop From Archean orogenic gold deposit models (Roberts, 1988; Groves and Foster, 1991; Hodgson, 1993; Groves et al., 1998; Hagemann and Cassidy, 2000; Goldfarb et al., 2005), and from previous work done on nearby gold deposits (Moore, 1936; Prest, 1956; Troop, 1986; Berger 1999; Berger, 2002), and at Hislop (Roscoe and Postle, 1998; Power et al., 2004), it is possible to outline the prospective rocks at Hislop. Rock types known to have a close spatial relationship to gold at Hislop include Fe-rich volcanic rocks, syenite intrusive rocks, and porphyritc rhyolite dikes. Carbonate dominated hydrothermal alteration is frequently associated with gold in these deposits, and is known to be related to gold at Hislop. 31 2.4.2. Physical properties of the Hislop deposit All physical property measurements made on Hislop deposit samples, including magnetic susceptibility, density, resistivity, chargeability, and porosity measurements, are compiled in Appendix 2E. Descriptive statistics, and correlation coefficients for physical properties and mineral abundances can be found in Appendices 2F and 2G, respectively. Magnetic susceptibility Magnetic susceptibility logs Selected geology and susceptibility logs from various parts of the Hislop property illustrate the behavior of magnetic susceptibility associated with characteristic rock types and alteration styles at Hislop (Fig. 2.4). Most, but not all, unaltered or weakly altered (where alteration minerals are restricted to veins) mafic and ultramafic volcanic rocks are high susceptibility. Susceptibilities are also high where ultramafic rocks are characterized by talc-chlorite metamorphic mineral assemblages (depicted by the dark green color in Column 2 of the drill logs in Fig. 2.4). There is a regular drop in magnetic susceptibility where Fe-carbonate + albite alteration, Fe-carbonate + muscovite alteration, or magnesite + fuchsite alteration has been superimposed on mafic and ultramafic volcanic rocks. Less pervasive, weakly fracture-focused alteration, such as an Fe-rich dolomite alteration that lends a pink color to some intermediate dikes and mafic volcanic rocks, do not appear to have a consistent effect on magnetic susceptibility values. There are seemingly no obvious patterns between altered syenite intrusives and porphyritic rhyolite dikes and magnetic susceptibility. Other rock types generally occur as very narrow units, and susceptibility readings for these rocks are sporadic. 32 Lithology Multi-lithic volcanic breccia Lamprophyric dike Intermediate-mafic dike Porphyritic rhyolite dike Syonite intrusive Mafic volcanic rock Ultramafic volcanic rock Alteration Weak to moderate Fe-cb + ms Strong Fe-cb + ma Fe-cb + ab Tic-chl metamorphic assemblage Mg-Gb (magnesite) + ms (fuchsite) Chlorite j Sericite Fe-cb + ab (intermediate dikes) Ms/ser (syenite and rhyolite dikes) Fe-cb (syenite and rhyolite dikes) Pink Fe-rich dol veins Epidote veins Hematite - pervasive Hematite along fractures Magnetite2 Figure 2.4. Geology, alteration, magnetic susceptibility, and gold grade logs for four Hislop drill holes logged for this study. The most consistent susceptibility trends include: low susceptibility of felsic intrusive rocks, high susceptibility of talc-chlorite assemblage ultramafic volcanic rocks, and some mafic volcanic rocks, and low susceptibility of carbonate-altered ultramafic and mafic volcanic rocks. For explanations of abbreviations in legend see bottom of Table 2.2. _________ Au abundances between 0.15- 1 ppm Au abundances between 1 - 5 ppm Au abundances > 5 ppm Column 1: Lithology Column 2: Alteration Column 3: Magnetic Susceptibility (x103 SI Units) *not all intervals sampled 33 Magnetic susceptibility data - all rock samples Magnetic susceptibility data collected from drill core and hand samples are summarized in a series of histograms (Fig. 2.5). A wide range of susceptibilities, spanning 2 and 3 magnitudes, characterize the main rock types at Hislop. The histograms show a steady decrease in magnetic susceptibility values from ultramafic to felsic rocks. Mafic and ultramafic rocks have distinct bimodal magnetic susceptibility distributions. Extended ranges of susceptibility for intermediate and felsic rocks may be attributed to a small number of outliers. 40 30 20 10 8 6 4 2 8 6 4 2 0.01 0.1 1 10 100 Magnetic Susceptibility (x iO SI Units) Figure 2.5. Magnetic susceptibility histograms for the five main rock types found in the Hislop deposit area. Mean values are given for general comparison, however the mean may not be an appropriate descriptor for populations with bimodal distributions. 10 S 12 8 4 Mean 12 O7 - ntermedtedes _JlJItfhmFL fl .n HTh Mean Syenite intrusives - : 1.19 - I - 1000 34 Magnetic susceptibility data - least altered and altered rock samples As was indicated in the magnetic susceptibility logs, some of the variation within ultramafic and mafic rock data may be attributed to effects of alteration. When magnetic susceptibility data for ultramafic and mafic rocks at Hislop is subdivided into least- altered, and carbonate-altered populations, it is apparent that the carbonate-altered populations have lower overall magnetic susceptibilities (Fig. 2.6). Intermediate dikes display a slight decrease in average magnetic susceptibility with Fe-carbonate, and Fe- carbonate + muscovite alteration (Fig. 2.7). Syenites and porphyritic rhyolite dikes, exhibit generally restricted ranges of magnetic susceptibility (Fig. 2.7). Alteration results in a minimal decrease in susceptibility for these intrusive rocks. Density Density data - all rock samples From density histograms (Fig. 2.8), there is a decrease in density from ultramafic to felsic rocks. Narrow ranges in density characterize syenite intrusive rocks and porphyritic rhyolite dikes. Ultramafic and mafic rock densities span a larger range than density values for intermediate and felsic rocks Density data - least altered and altered rock samples Alteration of ultramafic rocks correlates with a slight increase in average density relative to least-altered ultramafic rocks. There is a minor decrease in average density for Fe-carbonate + albite altered mafic volcanic rocks (Fig. 2.9). Intermediate dikes undergo a marginal density increase with Fe-carbonate + muscovite alteration (Fig. 2.10). There are no significant changes in densities between unaltered and altered equivalents of syenites and rhyolite dikes at Hislop - data peaks are generally consistent between the subpopulations (Fig. 2.10). 35 42 4 2 a) Variably altered ultramafic volcanic rocks 0.01 0.1 1 10 100 Magnetic Susceptibility (x i0 SI Units) 15 5 30 20 10 6 4 2 b) Variably altered mafic volcanic rocks Figure 2.6. Magnetic susceptibility histograms showing susceptibility data for a) least- altered and altered ultramafic volcanic rocks, and b) least-altered and altered mafic volcanic rocks. 8 4 8 6 4 Mean 6.03 Least-altered dolomite+chlorite lage Mean 5.731 Fe/Mg-carbonate+rnuscovite I Illilili Mean 0.75 Magnesite+fuchsite 1000 [Mean 40.O9 Least-altered chlorite+albiteflasmbla{y Mean 6.86] Fe-carbonate+muscovite IMean 1.71*[i ri Fe+carbonate+albite 0.01 0.1 1 10 100 1000 Magnetic Susceptibility (x iO SI Units) 36 42 6 3 6 4 2 b) Syenite intrusives 0,01 0.1 1 10 100 1000 Magnetic Susceptibility (x 10 SI Units) 4 2 a 2 f-1 Least-Itered rhyolite dikes Mean O.35[ Mucovite fl Mean 0.48J Fe/Mg-carbonate. ri U 0.01 0.1 1 10 100 1000 Magnetic Susceptibility (x 10 SI Units> Figure 2 7. Magnetic susceptibility histograms showing susceptibility data for a) least- altered and altered intermediate dikes, b) least-altered and altered syenitic dikes, and c) least-altered and altered porphyritic rhyolite dikes. a) Intermediate dikes --1 intermed. dikes .LHFP Mean 0.74 Fe-carbonate rL +muscovite Mean 10.95 Fe/Mg-carbonate. I I 0.01 0.1 1 10 100 Magnetic Susceptibility CX 10 SI Units) 1000 [Mean 2 87j Lest-aIteed I syenite intrusives 1 4 2 6 4 2 4 2 I . Muscovite. Mean 0.36 Fe/Mg-carbonate Mean 0.23 c) Porphyritic rhyolite dikes 4 2 37 20 15 10 5 Figure 2.8. Density histograms for the five main rock types found in the Hislop deposit area. Jitramafic volcanic rocks J iiai Intermediate dikes 25 15 5 8 4 10 5 10 5 IilTii Porphyritic rhyolite dikes Mean 2.70 , II i—I1F1—f—[_i-i fl ri 2.4 2.5 2.6 2.7 2.8 2.9 Density (glcm3) 3.0 3.1 3.2 38 20 15 10 5 20 15 10 5 Density (g/cm’) 2.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2 Density (g/cm3) Figure 2.9. Density histograms showing density data for a) least-altered and altered ultramafic volcanic rocks, and b) least-altered and altered mafic volcanic rocks. a) Variably altered ultramafic volcanic rocks 4 2 20 15 10 5 6 4 2 6 4 2 Least-aIterei dolornite+ Mean 2.85 -chlorite assemblage - TaIc+chlorite 4 Fe/Mg-carbonate÷ Mean 2.87J muscovite - - I I I Magnesite+fuchsite [Mean 2.92 2.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2 39 3 2 3 2 6 4 2 2.4 6 4 2 6 4 2 4 3 2 c) Porphyritic rhyolite dikes Least-altered - rf rhyolite dikes Mean 2.69 I — I Muscovite - i r—fL Mean 2.68 Fe/Mg-carbonate Mean 2.73 — 2.4 2.6 2.8 3.0 3.2 Density (gIcm) Figure 2.10. Density histograms showing density data for a) least-altered and altered intermediate dikes, b) least-altered and altered syenitic dikes, and c) least-altered and altered porphyritic rhyolite dikes. 40 a) Intermediate dikes Mean 2.83 Last-lterd dikes n 2 85 Fe-carbonate Mean 2.6 2.8 3.0 3.2 Density (gIcm) b) Syenite intrusives Leat-aIkred’ eniteintw:s . Muscovite I Mean 2.72 Fe/Mg-carbonate LMn. 2.4 2.6 2.8 3.0 3.2 Density (gIcm) 8 6 4 2 6 4 2 4 3 2 Resistivity Resistivity data - all rock samples Ranges of resistivity for Hislop rocks are large and overlap one another significantly (Fig. 2.11). However, they are roughly comparable to published data for similar rock types (Tab. 2.3). From resistivity histograms, it is evident that ultramafic rocks have the lowest resistivities of the five main rock types at Hislop. Mafic volcanic rocks, intermediate dikes, syenite intrusives, and porphyritic rhyolite dikes have similar average resistivities. There was insufficient sample numbers to evaluate effects of hydrothermal alteration on the various rock types Chargeability Chargeability data - all rock samples Ranges of chargeability values for the various Hislop rock types generally overlap one another, with some outliers (Fig. 2.12). Hislop chargeability data falls into the chargeability ranges considered to be characteristic of these rock types (Tab. 2.3), although many of these published chargeability ranges largely overlap. As with resistivity data, there were too few samples to compare the effects of alteration on chargeability values for the five rock types. 41 Figure 2.11. Resistivity histograms for Hislop deposit rocks. Data indicates lower overall resistivities for ultramafic volcanic rocks from Hislop. Table 2.3. Ranges of resistivity and chargeability for rock types similar to those occurring in the Hislop deposit area (data from Telford et al., 1990). Rock Type Resistivity (Ohm-rn) feldspar porphyry 4 x 1 3 (wet) porphyry (various) 60 - i04 syenite 102 - 106 andesite 1.7 x 1 2 (dry) basalt 10- 1.3 x i07 (dry) peridotite 6.5 x (dry) calcarious/mica schists 20 - Rock Type Chargeability (ms) schists 5-20 precambrian volcanics 8-20 dense volcanic rocks 100-500 granites, granodiorites 10-50 2-8 % sulfides 500-1000 8-20% sulfides 1000-2000 20% sulfides 2000-3000 1000 1( Resistivity (Ohm-rn) 42 10 8 6 4 2 6 4 2 Ultramafic volcanic rocks Maflc volcanic rocks - Intermediate dikes - LMean 29.24 Syenite intrusives Mean 14.71t Porphyritic rhyolite dikes _______ I Mean 10.47 10 100 Chargeability (milliseconds) Figure 2.12. Chargeability histograms for Hislop deposit rocks. Chargeability ranges for the individual rock types overlap and are not unique. 2.5. INTERPRETATIONS 2.5.1. Effect of geological processes on physical properties at Hislop Magnetic susceptibility From petrographic, SEM, and XRD work, it was established that magnetite is the only significant magnetic mineral in the Hislop deposit rocks. The trend of decreasing magnetic susceptibility from ultramafic to felsic rocks observed at Hislop, reflects decreasing magnetite abundance. A plot of modal magnetite, as derived from XRD and Rietveld analyses, plotted against magnetic susceptibility (Fig. 2.13) shows a positive correlation between these data, supporting this interpretation. I IMean 6.91[ 4 2 6 4 2 6 4 2 1000 43 9 Rock Types 8 Ultramafic rocks Maflc rocks X intermediate dikes 6 A Syenite intrusives ORhyo porphyries 2 1 I I 0 20 40 60 80 100 Magnetic Susceptibility (x104 SI Units) Figure 2.13. Positive correlation between modal magnetite in Hislop rock samples (as derived from XRD analysis) and magnetic susceptibility. For calculated correlation coefficients see Appendix 2G (all rock types). Large susceptibility ranges for the major rock types at Hislop are not atypical and result from the broad range of mineralogy that can be encompassed under a given rock classification; classification schemes do not normally take into account oxide and sulfide accessory minerals, the minerals primarily controlling susceptibility (Clark, 1997). Bimodal populations are common in magnetic susceptibility data and are interpreted to represent distinct populations whereby Fe has partitioned mainly into paramagnetic minerals (weakly magnetic phases including silicates and carbonates) or into ferromagnetic minerals (strongly magnetic minerals, such as magnetite and pyrrhotite). A change in magma composition, or in oxidation state, may cause a rock to fall into one population or another (Clark, 1997). Magnetite in mafic volcanic rocks, intermediate dikes, and felsic rocks at Hislop is primary igneous magnetite. Magnetite is not typically a primary igneous mineral in komatiitic rocks as chromite is the principal spinel that forms (Clark, 1997). Magnetite forms in ultramafic rocks usually as a product of serpentinization of olivine during early, retrograde metamorphism (Bucher and Frey, 2002). 44 In mafic rocks, the decrease in magnetic susceptibility with Fe-carbonate + muscovite, and Fe-carbonate + albite alteration is predominantly attributed to the conversion of magnetite in mafic rocks to Fe-carbonate upon exposure to C02-rich hydrothermal fluids (Roberts, 1988). Figure 2.14 shows this process occurring at Hislop, adjacent to a Fe-carbonate-filled fracture in a mafic volcanic rock. A negative correlation between modal magnetite and total Fe-rich carbonate abundance (ankerite + siderite + dolomite) further corroborates this relationship at Hislop (Fig. 2.15). Figure 2.14. Magnetite grains (reflective grains in lower image) are destroyed within a carbonate altered zone surrounding a carbonate vein in a mafic volcanic rock from Hislop. Plane polarized and reflected light photomicrographs. 45 a, 0 Dolomite + Ankerite + Siderite (Wt. %) Figure 2.15. Modal magnetite versus total Fe-rich carbonate abundance for all Hislop samples with measured quantities of these minerals. Bubble size represents relative magnetic susceptibility. A decrease in susceptibility correlates with a decrease in modal magnetite and an increase in Fe-rich carbonate abundance. Magnetite in ultramafic rocks formed during serpentinization is thought to be similarly affected by carbonate alteration. Clark (1997) explains that upon interaction with C02-rich fluids, magnetite is first redistributed in ultramafic rocks, and then is destroyed. Conversion of magnetite to Fe-rich carbonate was not directly observed in ultramafic rocks during petrographic or mineralogical work on Hislop rocks, however, the lack of magnetite in carbonate-altered ultramafic rocks compared to least-altered equivalents, is assumed to be due to alteration-related magnetite destruction. Some intermediate dikes, porphyritic rhyolite dikes, and syenite intrusive rocks have low abundances of primary igneous magnetite, thus typically magnetite-destructive alteration does not affect magnetic susceptibility significantly (Fig. 2.10). Hydrothermal alteration processes affecting mafic and ultramafic volcanic rocks do not explain all of the measured variation in magnetic susceptibility, as is indicated by additional heterogeneity in susceptibility readings from recorded unaltered intervals in drill core (Fig. 2.4). Variations in magnetic susceptibility in the absence of obvious 46 hydrothermal alteration could be related to a range of factors. Based on the rock types and mineralogy at Hislop, the most likely factors causing variable susceptibility in generally unaltered rocks at Hislop include an uneven distribution of primary or secondary magnetite, grain size, and irregular oxidation of magnetite to form hematite. An uneven primary distribution of magnetite in mafic rocks, and uneven secondary distributions of magnetite in ultramafic rocks may explain non-alteration related magnetic susceptibility variations in these rocks. Some small scale variations must be expected, as rocks are not likely to be perfectly homogeneous in their modal mineralogy. Formation of magnetite in a mafic volcanic rock is dependant on many factors including the magma composition, the degree of differentiation, and the temperature and pressure conditions under which the rock is formed or metamorphosed (Clark, 1997). For ultramafic rocks, the formation of magnetite from olivine during serpentinization may be influenced by location of fluid pathways in the rock. Small magnetite grain sizes are usually more susceptible than larger grain sizes as they do not easily retain remnanant magnetism (Clark, 1997). To examine the role of visible grain size in non-alteration related variations in magnetic susceptibility, least- altered fine-grained and medium-grained samples are plotted separately. Magnetite grain size here is assumed to be consistent with the overall grain size of the samples. The resulting histograms (Fig. 2.16) illustrate that fine-grained, and medium-grained mafic and ultramafic rocks have similar ranges and distributions of magnetic susceptibility, and similar average susceptibilities. Thus, variations in magnetic susceptibility data for these rocks are not likely to be strongly controlled by grain size. In some mafic and ultramafic rock samples, hematite rims magnetite grains indicating some oxidation of these rocks has occurred. A consistent pattern related to a particular alteration event, or having specific lithological or structural control, was not recognized during petrographic or mineralogical (SEM and XRD) analyses. Irregular oxidation of magnetite to hematite in mafic flows however, could contribute to decreases in magnetic susceptibility unrelated to hydrothermal alteration in mafic rock samples. 47 a) Mafic volcanic rocks 10 5 4 3 2 5 4 3 2 b) Ultramafic volcanic rocks I I I Fine-grained - Mean = 1.06 x103 SI Units Figure 2.16. Histograms showing distribution of susceptibility for fine- and medium grained a) mafic volcanic rocks, and b) ultramafic volcanic rocks. Similar distributions between fine- and medium-grained subsets indicates that grain size is not a major control on susceptibility at Hislop. Fine-grained 40 -Mean 0.77 xl o SI Units 30 — 20 — Medium-grained — Mean = 1.09 x103 SI Units 0.01 0.1 1 10 100 1000 Magnetic Susceptibility (x iO SI Units) Th[1 4 3 2 1F1 Medium-grained - Mean 0.75 x103 SI Units mimi ri 0.01 ________ Ii 0.1 1 10 100 1000 Magnetic Susceptibility (x iO SI Units) 48 Density Mineralogy and porosity are considered to be the main controls on density at Hislop. Both mineralogy and porosity are affected by geological processes including igneous fractionation! differentiation, metamorphism, and hydrothermal alteration. Mineralogy plays a significant role in determining rock densities. Igneous and volcanic rock densities generally decrease with increasing Si02 content (Johnson and Olhoeft, 1984; Telford et al., 1990), reflecting an increase in the abundance of low density felsic minerals, and a corresponding decrease in the abundance of higher density Fe- and Mg-rich mafic minerals. This is consistent for Hislop samples. From Table 2.4, it is apparent that minerals that typically characterize ultramafic and mafic rocks at Hislop are higher in density on average than those that characterize felsic rocks. Table 2.4. Densities of the common minerals in Hislop deposit rocks (from www.mindat. org). ________________________ Mineral Density (glcm3) Quartz 2.62 Microcline 2.56 Albite 2.62 Actinolite 3.04 Epidote 3.45 Augite 3.4 Chlorite 2.65 Muscovite 2.82 Calcite 2.71 Ankerite 3.05 Siderite 5 Dolomite 2.84 Magnesite 3 Talc 2.75 Serpentine 2.53 Pyrite 5.01 Magnetite 5.15 Hematite 5.3 49 The modal mineralogy of syenites and porphyritic rhyolite dikes brings about their narrow density ranges. They are dominated by a small number of similarly dense minerals, specifically quartz and feldspar. Densities of ultramafic and mafic rocks span a larger range of densities than those making up intermediate and felsic rocks which is a result of their more complex and varied mineralogy (refer to Tab. 2.2). A slight increase in the average measured density of ultramafic volcanic rocks corresponds with Fe/Mg-carbonate + muscovite, and magnesite + fuchsite alteration. This relationship can be attributed to changes in mineralogy accompanying alteration. Based on published mineral densities (Tab. 2.4, mineral densities from www.mindat.org), a change in the bulk mineralogy of an ultramafic rock containing predominantly chlorite, plus carbonate, talc, quartz, and magnetite, to a rock composed of abundant Fe-rich and Mg-rich carbonate, plus muscovite, and quartz, should theoretically result in a denser rock. Carbonate minerals are expected to have a significant influence on rock density. On average, they are denser than those silicate minerals that dominate the mineralogy of igneous and volcanic rocks, Carbonates containing Fe would be especially influential, having densities as high as 5 g/cm3 (e.g. siderite, Tab. 2.4). Figure 2.17 illustrates the correlation between increasing density values with increasing Fe-rich carbonate abundance in Fe-carbonate bearing Hislop deposit samples. Minor variations in the density of mafic volcanic rocks from Hislop may be similarly attributable to alteration. The lower average density values for Fe-carbonate + albite altered samples, as compared to least-altered and Fe-carbonate + muscovite samples, is considered to be related to bulk mineralogy (Fig. 2.9). The increased relative abundances of low-density albite in rocks with Fe-carbonate and albite-dominated alteration assemblages has likely lowered the density. Some changes in density for subpopulations of altered intermediate dikes (Fig. 2.10) are difficult to interpret due to irregular data populations that might have come about through oversimplified sample groupings. There is little change in density between the variably altered syenites and porphyritic rhyolites (Fig. 2.10). It is assumed that for 50 these rock types, bulk mineralogy changes do not add or subtract significant dense minerals, and thus alteration has little influence on the density of these rocks. 50 Rock Types C Ultramafic rocks X - 40 • Mafic rocks X Intermediate dikes + A Syenite intrusives ,. 0 Rhyolite porphyries 3U I- ..P . x 20 A x 10 00 AA A = 0.56 2.60 2.70 2.80 2.90 3.00 Density (glcm3) Figure 2.17. Density increases for Hislop rocks with an overall increase in the abundance of Fe-rich carbonate. For calculated correlation coefficients see Appendix 2G (all rock types). Calculating densities from modal mineralogy as determined from XRD analysis and published mineral density data helps to determine what the densities of the rock should theoretically be, if the density is controlled solely by mineralogy. When compared to measured densities, discrepancies will indicate that there are factors aside from bulk mineralogy affecting the rock. Density is calculated simply by using volume concentrations of minerals (C) and their grain densities (p) as given in Johnson and Olhoeft (1984): p = Ci* P i + C2* P2+ C3* P 3... C* p 51 A lack of strong correlation between some of the measured and calculated densities for ultramafic and mafic rocks (Fig. 2.18), suggests that there may be other controls on density. Two possible explanations for the incongruity include not accounting for porosity in samples, and limitations in mineral identification using XRD methods. 3.40 3.20 3.00 2.80 2.60 2.40 2.40 2.60 2.80 3.00 3.20 3.40 Measured density (glcm3) Figure 2.18. Measured versus calculated density for Hislop rocks. Discrepancies between the density values obtained from the two methods for ultramafic and mafic volcanic rocks could indicate that bulk mineralogy does not solely control density. Density is known to decrease with increasing porosity (Telford et al., 1990; Johnson and Olhoeft, 1984), and porosity is thought to be a factor in some of the density variations at Hislop. To test the possible influence of porosity on the density of mafic and ultramafic rocks, a suite of samples in varying states of alteration were measured for porosity using the method described in section 3.2 Figure 2.19 shows that there is an overall negative correlation between density and porosity for ultramafic rocks at Hislop. Talc-chlorite assemblage rocks are most porous and least dense in accordance with their typically strong foliation. Strongly carbonate-altered samples have lower porosities and higher densities. Figure 2.20 indicates no obvious relationship between density and porosity for mafic volcanic rocks. = 0.69 (excluding outlying intermediate dike sample) 52 Unaltered to Altered Ultramafic Rocks DLst, altd. (Dol-chi) •TIc+cN ultramafic 3 DFeCb+ms afld. utiramafic DFe/MgCb+fu altd ultraniafic •1 r2=0.28 0 I I I 2.70 2.75 2.80 2.85 2.90 2.95 3.00 3.05 Density (glcm3) Figure 2.19. Porosity of ultramafic volcanic rocks at Hislop decreases with carbonate- related hydrothermal alteration, due to annealing of this commonly sheared rock. This brings about a corresponding increase in density. Abbreviations: Lst. altd. least altered; dol+chl dolomite + chlorite; tlc+chl = talc + chlorite; FeCb+ms = Fe-carbonate + muscovite; Fe/MgCb+fu = Fe(Mg)-carbonate + fuchsite. For calculated correlation coefficients see Appendix 2G (ultramafic rocks). Unaltered to Altered — Mafic Rocks — Lst. altd. (Chi-ab) FeCt,+ms altd. niafic 3 oFeb+ab altd. mafic > 2.70 2.75 2.80 2.85 2.90 2.95 3.00 3.05 Density (glcm3) Figure 2.20. No relationships are indicated between porosity and density for mafic volcanic rocks at Hislop. Abbreviations: Lst. altd. = least altered; chl+ab = chlorite + albite; FeCb+ms = Fe-carbonate + muscovite; FeCb+ab Fe-carbonate + albite. For calculated correlation coefficients see Appendix 2G (mafic rocks). 53 Discrepencies between measured and calculated density data could also be attributed to generalization of modal mineralogy during Rietveld analysis. For some minerals that exist as a solid solution, such as dolomite and ankerite, a proper name is not assigned for intermediate compositions. The density values for these end members differ significantly, and the resulting calculated density would be affected accordingly if one end member mineral classification was chosen over the other. For intermediate and felsic intrusive rocks, calculated and measured density values match closely, indicating primarily mineralogical control on density. Changes in density between unaltered and altered versions of these rock types thus must be explained by relative additions or subtractions of more and less dense minerals. Resistivity Least-altered metamorphosed volcanic and igneous rocks at Hislop have resistivity ranges similar to published ranges for equivalent rocks types (Tab. 2.3). Published resistivity ranges for most minerals are very large and not as specific as density values for given minerals. This makes it difficult to assess the combined resistivity affects of minerals making up a rock. This being said, the role of mineralogy on resistivity at Hislop is thought to be minimal. The majority of minerals making up Hislop rocks are poor to intermediate conductors, or resistors (> 1 Ohm-rn; Telford et al., 1990). Most sulfides, and some oxides, are known to be good conductors (low resistivity, <1 Ohm-rn), and there is a small percentage of these minerals in Hislop samples. Variations in resistivity at Hislop are interpreted to be primarily controlled by rock texture and porosity. Resistivity is known to drop considerably with increasing water content of rocks (Telford et aL, 1990), thus to be related to the porosity of a rock (Halloff, 1992). As such, the low average resistivity of ultramafic rocks compared to the other Hislop rock types is interpreted to be a result of the relatively high porosities of 54 talc-chlorite assemblage ultramafic rocks, the most common ultramafic rock subpopulation sampled during this study. Resistivity is plotted against magnetic susceptibility and density (Fig. 2.21 and Fig. 2.22), two properties shown to vary with alteration in ultramafic and mafic volcanic rocks at Hislop. In Figure 2.21a, the ultramafic samples are the only samples to outline a trend between resistivity and magnetic susceptibility. With ultramafic samples colored to represent their dominant alteration assemblages, it is obvious that the trend is related to alteration. This variation in resistivity is interpreted to be related specifically to alteration effects on porosity. Figure 2.23 demonstrates that a decrease in porosity of ultramafic rocks with carbonate alteration causes the rock to become more resistive. Thus, altered ultramafic samples are resistive and, as was indicated previously, are characterized by low magnetic susceptibilities due to magnetite destruction. When resistivity is compared with density (Fig, 2.22), again a weak correlation emerges only for ultramafic samples. When colored based on alteration, the relationship of increasing resistivities and densities with carbonate alteration is apparent for the majority of the samples, and is explained by a decrease in porosity for altered rocks. Chargeability The main control on the chargeability of Hislop rocks is thought to be the presence of disseminated sulfides. Disseminated sulfides in rocks are readily chargeable where subjected to an induced current, due to the chargeable nature of the metallic grains coupled with the large surface area provided by a disseminated texture (Telford et al., 1990). Other known controls on chargeability include presence of clay minerals and graphite, both of which are absent from Hislop rocks. A positive relationship between pyrite abundance based on XRD analyses, and chargeability for syenites and porphyritic rhyolites is indicated in Figure 2.24. However, there is not a similarly convincing relationship indicated for other Hislop rock types. 55 1000000 — ci. . 100000 110000 ____ 1000 RockTypes 15 C Uftramafic rocks Mafic rocks I VU x Intermediate dikes A Syenite intrusives 10 Rhyolite porphyries _________________________________________________ 0.01 0.1 1 10 100 1000 Magnetic Susceptibility (x103 SI Units) 1000000 1 b. — 100000 10:0: Unaltered to Altered Ultramafic Rocks 100 DLst. altd. (Dol-chi) •TIc+chl ultramafic — DFeCb+msaltd. utiramafic i — 0.53 DFe/MgCb+fu altd. ultramafic (ultramafic rocks) 10 I I 0.01 0.1 1 10 100 1000 Magnetic Susceptibility (x103 SI Units) Figure 2.21. Resistivity versus magnetic susceptibility, a) Ultramafic volcanic rock samples indicate a trend between these physical properties, whereas variations in resistivity and magnetic susceptibilty are more irregular for other rock types. b) When data points are colored to represent the various ultramafic alteration assemblages, it is apparent that the relationship between restivity and susceptibility is controlled in part by carbonate alteration. For abbreviations, see Fig. 2.19. For calculated correlation coefficients see Appendix 2G (ultramafic rocks). 56 1000000 — a. 100000 E 10000 xA A xn 1000 Rock Types C UItramaflc rocks Maflc rocks 1 00 x Intermediate dikes A Syeriite intrusives 10 ORhyolite porphyries I 2.60 2.65 2.70 2.75 2.80 2.85 2.90 2.95 3.00 Density (glcm3) 1000000 - b. 100000 10000 t A4< A Unaltered to Altered Ultramafic Rocks QLst. altd. (Dol-chi) I ‘“‘ •TIc+chl ultramafic — DFeCb±ms altd. utiramafic — 10 DFe/MgCb+fualtd. ultramafic I I (ultromafic rocks) 2.60 2.65 2.70 2.75 2.80 2.85 2.90 2.95 3.00 Density (glcm3) Figure 2.22. Resistivity versus density. a) As with resistivity versus magnetic susceptibility, trends in data when plotted based on rock type are not obvious. b) Subdividing ultramafic rocks based on alteration assemblage reveals that increasing resistivities and densities can be to some extent attributed to carbonate alteration. For abbreviations, see Fig. 2.19. For calculated correlation coefficients see Appendix 2G (ultramafic rocks). 57 Unaltered to Altered Ultramaflc Rocks QLst. altd. (Dol-chl) •TTC+chl ultramafic 3 DFeCb+ms altd. DFe/MgCb+fu altd. ultramafic 0 —----—- --- 0_i • n E1 LU r2=O.70 0 I 10 100 1000 10000 100000 Resistivity (Ohm-rn) Figure 2.23. A plot of porosity versus resistivity shows that annealing of ultramafic rocks due to precipitation of carbonate minerals during hydrothermal alteration brings about a decrease in porosity and a corresponding increase in resistivity. For abbreviations, see Figure 2.19. For calculated correlation coefficients see Appendix 2G (ultramafic rocks). 10 Rocktypes O Ultramaflc rocks Mafic rocks X 0 Xlnterrnediate dikes A ASyenite intwsives - o Rhyolite porphyries A 2 AA A 0 o 0 r2=O.53 (felsic rocks, excluding rhyolite dike outlier) 0 1.00 10.00 100.00 Chargeability (ms) Figure 2.24. A weak positive correlation exists between pyrite abundance and chargeability, however the trend is mainly controlled by porphyritic rhyolite dike and syenite samples. There is no evidence of a consistent relationship between chargeability and pyrite abundance for intermediate to ultramafic volcanic rocks. For calculated correlation coefficients see Appendix 2G (felsic rocks). 58 This indicates that there may be variables affecting chargeability other than, or in addition to, sulfide abundance. Sulfide grain size and texture, and the relationship between sulfide grains in the rock, are all potential factors that can influence the rock’s chargeability (Pelton et a!., 1978). As chargeability should increase with increased surface area of sulfide minerals, chargeability values may depend on whether sulfides are disseminated, concentrated in a stockwork system, or controlled by fractures or veins. Variable porosity may affect the chargeability of mafic volcanic rocks. Chargeability can decrease with porosity; increased fluid pathway volume can be more conducive to electrolytic conduction, prohibiting polarization. For example, chargeabilities may be higher for a crystalline igneous rock containing disseminated sulfides, than for a more porous sedimentary rock containing sulfides, (Telford et al., 1990). Although the dataset is small (few samples have both chargeability data and porosity), there is a weak relationship between porosity and chargeability for mafic rocks at Hislop (Fig. 2.25). 10 IeMafi r2=0.24 0 1.00 10.00 100.00 1000.00 Chargeability (ms) Figure 2.25. A negative correlation between chargeability and porosity in this plot indicates that increases in porosities of mafic volcanic rocks at Hislop may hinder the ability for metallic minerals to become charged. For calculated correlation coefficients see Appendix 2G (mafic rocks). 59 2.6. DISCUSSION 2.6.1. Exploration using physical properties Physical properties most useful for isolating prospective rocks at the Hislop deposit The most useful physical properties for delineating prospective rocks at Hislop from those more likely to be barren are magnetic susceptibility and density. Magnetic susceptibility and density are equally capable of discerning prospective syenite intrusive rocks and porphyritic rhyolite dikes at Hislop from intermediate, mafic, and ultramafic rocks (Fig. 2.26). These physical properties however, do not distinguish between hydrothermally altered and least-altered felsic rocks, as mineralogical changes in these rocks related to alteration processes do not add or remove any significant quantities of dense or magnetic minerals. 3.20 3.10 —. 3.00 2.90 2.80 2.70 0 2.60 2.50 2.40 0.01 0.1 1 10 100 1000 Magnetic Susceptibility (x103 SI Units) Figure 2.26. Magnetic susceptibility plotted against density for Hislop samples. Syenite intrusives and porphyritic rhyolite dikes have distinctly low density and magnetic susceptibility ranges allowing them to be distinguished from intermediate, mafic, and ultramafic rocks at Hislop. Rock Types Ultramsfic rocks Mafic rocks X Intermediate dikes ASyenite intrusives o Rhyolite porphyries c 0 4 A) x 0A 60 Magnetic susceptibility and density ranges for intermediate, mafic and ultramafic rocks are large, and generally overlap. These rock types cannot be independently distinguished from one another based on these two physical properties. That being said, potentially prospective carbonate-altered intermediate, mafic, and ultramafic rocks have typically low magnetic susceptibilities; carbonate-altered rocks almost exclusively occur in the lower susceptibility ranges for these rocks (Figs. 2.27a and 2.27b). Thus, when this low range is isolated, the majority of carbonate-altered rocks are targeted. Unfortunately, due to variability in magnetite abundance and distribution in intermediate to ultramafic rocks, and irregular hematization of magnetite, there are relatively unaltered, low- susceptibility rocks at Hislop. Carbonate-altered rocks at Hislop cannot be exclusively delineated from a physical property dataset as a result of this overlap. Nonetheless, targeting low susceptibility rocks would be effective in delineating many prospective carbonate-altered rocks from high susceptibility rocks more likely to be barren of mineralization. Density provides an additional measure of alteration of ultramafic volcanic rocks only. If ultramafic rocks were isolated, density values could be used to delineate the higher density magnesite + fuchsite rocks from other ultramafic rocks, specifically those with lower density talc + chlorite assemblages. Resistivity may be useful in distinguishing ultramafic rocks from other rocks in the Hislop physical property dataset, however this is likely of no significance with respect to mineralization, as these rock types are not uniquely mineralized. If dealing solely with ultramafic rocks however, higher resistivity values may be indicative of carbonate altered, low-porosity ultramafic volcanic rocks. Chargeability values do not distinguish between rock types at Hislop. Although there may be a relationship between pyrite abundance and chargeability for felsic rocks, there are likely other influences on the chargeability of rocks at Hislop, like the texture of sulfides, or that of the host rock itself. 61 3.20 a.3.10 _. 3.00 C., 2.90 P 0, >2.80 (0, 0 ____________ 2.60 Unaltered to AlteredMafic Rocks ILst. altd. (Chi-ab)2.50 IIOFeCb+ms alid. mafic OFeCb+ab altd. maf’ic 2.40 0.1 1 10 100 1000 Magnetic Susceptibility (x103 SI Units) 3.20 -___________________ 3.10 b. r’ 3.00 D 290 0, 2.80 —B B ____ 2.70 Unaltered toAltered Ultramafic Rocks 2.60 QLst. altd. (Dol-chi) •TIc+chl ultramafic 2.50 FeCb+ms altd. utiramafic Fe/MgCb+fu altd. ultramafic 2.40 I 0.1 1 10 100 1000 Magnetic Susceptibility (xl O SI Units) Figure 2.27. Carbonate-alteration destroys magnetite in a) mafic and b) ultramafic volcanic rocks, causing magnetic susceptibility to drop. Density values increase slightly for altered ultramafic rocks. For abbreviations in a) and b) see Figs. 2.20 and 2.19, respectively. 62 Prospective physical property ranges Magnetic susceptibility and density constitute the two most well understood physical properties at Hislop. They were determined to be the most useful of the four physical properties studied in delineating some of the prospective rocks at Hislop. Table 2.5 summarizes the prospective ranges for magnetic susceptibility and density for Hislop. These ranges were established using the statistical analysis program SPSS Statistics, and anomalously high and low values (extreme cases occurring beyond 3x the interquartile range of values) were eliminated to yield a tighter, more representative, range of values for each of the rock types. These prospective cut-off values are used to query the Hislop physical property database for the purposes of determining the effectiveness of these cut-offs to distinguish between possible gold-related rocks and rocks likely to be barren. The dataset was queried first to isolate prospective felsic rocks (susceptibility <0.42 x i0 SI Units, density <2.8 g/cm3), and then queried to identify altered intermediate to ultramafic rocks (susceptibility <5.96 x i0 SI Units, density between 2.67 g/cm3 and 2.97 glcm3) from the remaining data. The results can be assessed in two ways: (1) No. of targeted rock types recalled / total known to occur in database; (2) No. of targeted rock types recalled / total recalled in query where the ‘targeted rock types’ refer to felsic rocks, or hydrothermally altered rocks. Table 2.6 compiles the results from this query. Out of 70 total felsic samples in the database, 55 were recalled, falling within the statistically significant susceptibility and density ranges for these rocks, yielding a 79% success rate. However, this query yielded 73 samples in total, out of which 18 were not felsic intrusive rocks, thus mislabeling 25% of the results. 63 Table 2.5. Statistical data for prospective rocks at Hislop, and cut-off values used for querying physical property data. Rock Type Magnetic Susceptibility (1O- SI) Density (glcm3) Cut-off values for querying data No. Mean Median Range No. Mean Median Range Rock Type Mag. Sus. Density Unaltered ultramafic 8 6.03 1.49 0.57-12.5 8 2.85 2.86 2.82-2.89 (dolomite-chlorite assemblage) Unaltered ultramafic (talc- 46 24.12 14.61 0.44-84.4 46 2.85 2.84 2.79-2.94 chlorite assemblage) Fe-carbonate-muscovite 16 5.73 1.01 0.41-5.96 15 2.87 2.85 2.80-2.91 altered ultramafic Carbonate-altered 0.41-5.96 2.80-2.96 ultramafic Magnesite-fuchsite altered 9 0.75 0.62 0.49-0.95 9 2.91 2.92 2.85-2.96 ultramafic Unaltered mafic 107 40.09 21.50 0.35-141 101 2.87 2.87 2.70-3.08 Fe-carbonate-muscovite 75 6.86 0.60 0-2.19 71 2.85 2.86 2.78-2.97 altered mafic Carbonate-altered 0.2-2.19 2.76-2.97 mafic Fe-carbonate-albite altered 14 1.71 0.58 0.28-1.27 13 2.82 2.81 2.76-2.86 mafic Unaltered intermediate 11 41.11 18.40 0.24-135.29 11 2.83 2.81 2.72-2.95 intrusive Carbonate altered 22 10.95 0.58 0.13-3.8 22 2.79 2.78 2.67-2.95 intermediate intrusive Carbonate-altered intermediate 0.13-3.8 2.67-2.95 Carbonate-muscovite 10 0.75 0.69 0.32-1.55 10 2.85 2.86 2.76-2.94 intrusive altered intermediate intrusive Syenite 34 1.19 0.20 0.07-0.42 32 2.70 2.69 2.64-2.74 . .Felsic intrusives 0.05-0.42 2.57-2.80Rhvolite norDhvrv 37 1.59 0.16 0.05-0.41 35 2.69 2.67 2.57-2.80 Table 2.6. Results from magnetic susceptibility and density queries of the Hislop physical property dataset. Target rock Total target rock Total recalled No. target rock % target rock % target rock samples samples in samples from samples recalled samples out of out of total recalled database query using query known amount in samples (2) database (1) Felsic intrusive rocks 70 73 55 79 75 Generally-altered 188 221 142 76 64 intermediate, mafic, and ultramafic rocks Carbonate-altered 146 221 112 77 51 intermediate, mafic and ultramafic rocks Of 188 variably altered intermediate, mafic, and ultramafic samples (this includes some obscure alteration types not thoroughly reported on in this work, in addition to carbonate altered rocks), 142 altered samples were recalled by the query, yielding a 76% success rate. Seventy-nine out of the 221 total recalled samples were relatively unaltered samples, thus 36% of the resulting sample set were misclassified. Out of 146 total dominantly carbonate-altered samples in the dataset 112 were recalled by the same query, giving a 77% success rate in detecting these samples from the dataset. The mafic and ultramafic samples misidentified as being altered, upon examination, are largely unaltered low susceptibility mafic volcanic rocks that overlap the physical property ranges of carbonate-altered mafic volcanic rocks. Although some unprospective, low susceptibility rocks would inevitably be targeted, many of the barren rocks are eliminated from consideration. Results of such queries would not provide definitive targets for exploration, but could act as important mineral vectoring criteria for consideration in association with any other geological, geophysical, geochemical, or mineralogical data available from the area. 65 Physical properties and 3D geophysical inversion modeling It is anticipated that physical property cut-off values similar to those used to target prospective samples from the Hislop physical properties database would be equally successful when applied to 3D physical property models generated from geophysical inversions in the Hislop area. However, the number of rock types at Hislop, and their structurally complicated relationships to one another, would make direct referencing to specific rock types and alteration assemblages based on physical property data difficult. At larger scales of modeling low magnetic susceptibility values may be effective in isolating felsic rocks and strongly carbonate-altered intennediate, mafic and ultramafic rocks. Density information would help further confirm identification of felsic rocks, isolating them from other magnetic susceptibility lows. With perhaps more localized inversion modeling, smaller scale variations in physical properties, like for example, subtle changes in mafic and ultramafic units related to the presence of felsic intrusions or of carbonate-alteration zones, could become apparent in regions that appear to be more homogeneous at a larger scale. The use of physical property data to highlight mineralization, or prospective geology and alteration, would generally occur at a later stage in exploration when an acceptable inversion model has been established for a property or deposit. Prior to this stage, physical property data can play an important role in guiding geophysical inversions. Knowledge of characteristic physical property values of rock types from the area of exploration, and of any relationships between physical properties and mineralization, can be input into the inversion to constrain it, which can significantly improve the inversion result (e.g. Williams, 2006). 66 2.6.2. Comparison to analogous areas Comparison to regional variations in physical property data An important goal of this work is to compile a dataset of typical physical properties expected to occur within a representative Archean orogenic gold environment, for future use in guiding and interpreting inversions both at Hislop and in similar mineral deposit environments. Before this data is used, however, it is important to determine whether the physical property values and ranges from Hislop represent those typically found in this environment. A regional physical property study covering Matheson and Kirkland Lake areas to the west and south of the Hislop deposit area, respectively, was completed for a large sample set of over 1000 samples (Ontario Geological Survey, 2001). Magnetic susceptibility, density, and resistivity were measured. Comparing the magnetic susceptibility and density data from the OGS study to the Hislop data helps to define the local extent to which these physical properties vary in this part of the Abitibi greenstone belt. This dataset was assessed and rock types considered to be equivalent to the primary rock types at Hislop were compiled and subdivided. Histograms comparing OGS physical property data to Hislop data are presented in Figure 2.28 and Figure 2.29. Magnetic susceptibility Mafic and ultramafic rocks from the two studies have similar magnetic susceptibility distributions. Fe-carbonate-altered mafic volcanic rocks from the Matheson and Kirkland Lake areas have similar data distributions as Fe-carbonate altered rocks from the Hislop area (Fig. 2.30). It was not possible to compare any other altered rock data from the OGS dataset to similar altered rocks from Hislop as no other samples in the OGS dataset were subdivided based on alteration assemblages. 67 12 9 6 3 12 9 6 3 12 8 4 4 3 2 8 6 4 2 10 8 6 4 2 8 6 4 2 Ultramafic Volcanic Rocks Syenite Intrusives 0.1 1 10 100 Magnetic Susceptibility (x 10 SI Units) Figure 2.28. Magnetic susceptibility histograms comparing data equivalent rocks from surrounding regional areas. from Hislop rocks, and 68 Hislop study FL jn - Mafic Volcanic Rocks Hislop study35 25 15 S 60 40 20 Intermediae Intrusive Hislop study - I study [illU 6 4 2 0.01 1000 20 15 10 5 10 8 6 4 2 25 15 5 30 20 10 10 8 6 4 2 4 3 2 12 9 6 3 8 6 4 2 12 9 6 3 8 6 4 2 Ultramafic Volcanic Rocks 2.4 2.5 2.6 2.7 2.8 2.9 Density (g/cm’) Figure 2.29. Density histograms comparing data from equivalent rock types from Hislop rocks, and equivalent rocks from surrounding regional areas. 69 Mafic Volcanic Rocks Intermediate Intrusives Hiskp study OGStudy fh Syenite Intruives Hislop study GSstudy Feldspar Porphyries HisIopstudy I g OGS tudy r 3.0 3.1 3.2 Least-altered rocks 40 30 20 10 Hislop maflc ., - volcanic rocks I1 - JL I W -OGSmafic - voIcanicro_fl, 01 0.1 1 10 100 1000 Magnetic Susceptibility (x io SI Units) 0.1 1 10 100 Magnetic Susceptibility (x iO SI Units) Figure 2.30. A comparison of magnetic susceptibility data associated with least-altered and carbonate-altered mafic rocks from the Hislop deposit, and from the greater surrounding area. Susceptibilities for local and regional intermediate intrusive samples overlap, however, there are very few regional samples overall. A comparison of syenite and felsic intrusive magnetic susceptibility data from the two datasets illustrates that regionally, there is more variation in magnetic susceptibility of these rock types than what is represented in the Hislop area, with three to four populations distinguishable. A comparison between Hislop and OGS data indicates outliers in Hislop syenite and felsic intrusive data may fall into the higher susceptibility ranges observed for similar rocks in the OGS dataset. The large regional range in susceptibilities may make it difficult to discriminate higher susceptibility, magnetite-rich syenites and felsic intrusives from mafic and ultramafic rocks at the regional scale. If it could be determined that low- susceptibility syenites are more commonly associated with mineralization, then the overlap would not cause a problem for physical property based exploration, and may 70 20 15 10 5 120 80 40 0 Carbonate-altered rocks Hislop maiic volcanic rocks -_ OGS mafic Volcanic Rofl[.....Fl 10 5 0.01 1000 actually allow for the identification, based on susceptibility, of prospective syenites from a larger syenite database. Density Density distributions for Hislop ultramafic rocks and the regional scale ultramafic rocks are similar, however there is a gap in data in the OGS dataset between 2.85 and 2.90 g/cm3. This may be attributable to the smaller size of the OGS ultramafic rock dataset, which has about half the number of samples of the Hislop dataset. The variety of ultramafic rocks that occur in the region may not be represented in the OGS sample set. Matheson and Kirkland Lake intermediate intrusive rocks have similar ranges of density values compared to Hislop intermediate dikes, however, the Matheson and Kirkland Lake samples, are much fewer in number. Mafic volcanic rock, syenite intrusive, and felsic intrusive data from these studies are not as comparable to one another. There are a greater number of high density mafic rocks regionally than at Hislop, suggesting there are high density regional scale mafic rocks that are not represented at Hislop. Regionally, syenites have slightly lower densities, and felsic intrusive rocks have slightly higher densities than the equivalent Hislop rocks. Magnetic susceptibility data for regional syenites and feldspar porphyries indicated that there are multiple populations that exist for these rock types that were not recognized or sampled at Hislop. The different subpopulations of these rocks at the regional scale likely differ in mineral composition, which would explain the inconsistencies between Hislop and OGS sample densities. Where separated into least-altered and altered rock populations (Fig. 2.31), ranges of density for least-altered mafic rocks are generally equivalent for the local and regional datasets, with the exception of the previously mentioned high density population in the unaltered regional mafic rock dataset. Densities of carbonate-altered mafic rock suites from the individual studies also overlap, however data populations in each sample set do 71 not match, with an anomalous high density population in the OGS dataset between 3.0 - 3.2 g/cm3. This high density population corresponds with higher density least-altered mafic volcanic samples in the OGS dataset. Perhaps these samples were mislabeled, or incompletely labeled originally and actually represent a population of anomalously high density carbonate-altered mafic rocks not encountered at Hislop. This however would infer that there is an alteration process which yields higher densities in mafic volcanic rocks, which was not observed during Hislop physical property studies Least-altered rocks 20 - Hislop mafic volcanic rocks 10 - 5 426 272.82.93.03.1 3.2 Density (g/cm3) Carbonate-altered rocks Hislop mafic 20 -volcanic rocks - 10 - - OGS mafic -volcanic rocks - 10 5— - ______ r1111 2.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2 Density (g/cm3) Figure 2.31. A comparison of density data associated with least-altered and carbonate- altered mafic rocks from the Hislop deposit, and from the greater surrounding area. Due to differences in instruments (the OGS susceptibility data were collected using a Bartington MS-2 susceptibility meter) and techniques used to collect the physical property data, some discrepancies would be expected between the two datasets. Additional differences may arise from misplacement of samples from the OGS dataset into incorrect rock categories for comparison to the Hislop sample suite. As there were no detailed descriptions for the OGS samples, it was not possible to be entirely confident in 72 assigning the samples to the proper congruent categories. Finally, lack of correspondence between some populations may be the result of rock types not being sampled with equal frequency at the local and regional scales. In summary, regional magnetic susceptibility data is more representative of local Hislop rocks than density data, especially with respect to intermediate, mafic, and ultramafic rocks. Least-altered mafic and ultramafic rocks from the local and regional datasets have similar means and ranges of magnetic susceptibility, as do altered rocks. This means that physical property queries used in this study should be capable of delineating a significant proportion of carbonate-altered rocks from suites of intermediate to ultramafic volcanic rocks throughout the larger area. Density values are not as consistent between the different scale studies, and are less useful in targeting particular rock types or alteration assemblages at the regional scale. However, since regional syenite densities are always as low as Hislop syenite densities, or lower, these important rock types may be distinguishable at the regional scale using appropriate physical property cut-offs. Effect of metamorphism on physical property data. Although greenschist facies rocks are the typical hosts for Archean orogenic gold deposits, these deposits also occur, albeit to a lesser extent, in amphibole or even higher grade rocks (Meuller and Groves 1991; Groves, 1993; Hagemann and Cassidy, 2000). Varying metamorphic grade can have a significant effect on physical property behavior, and must be considered prior to interpretation of sample-based, or geophysical inversion- derived physical property data. A comparison to physical property data from studies of the Weebo/Wildara and Southern Cross greenstone belts in the Yilgarn Craton, Australia, reveals similarities and differences in physical property data from similar geological environments of varying metamorphic grade (Boume et a!, 1993). Metamorphism can invoke changes in mineralogy or texture that can significantly influence the physical property value of a rock. An excellent example is the formation of magnetite during serpentinization. Bourne et al. (1993) have shown that densities and magnetic 73 susceptibilities are higher overall for amphibolite facies ultramafic and mafic rocks than for greenschist facies ultramafic and mafic rocks. They explain that increases in density of mafic rocks of amphibole facies grade is due to the destruction of low density plagioclase (2.61-2.77 g/cm3) to form hornblende (3.02-3.45 g/cm3) from actinolite/tremolite. Ultramafic rocks of higher metamorphic grade have increased densities relative to less metamorphosed ultramafic rocks which is related to the replacement of serpentine and talc (2.7 g/cm3), by olivine (3.3 g/cm3). Magnetic susceptibility increases with metamorphic grade in both mafic and ultramafic rocks due to increased magnetite content by volume in amphibolite grade rocks, and increased magnetite grain sizes which increases low-field magnetic susceptibility. The increase in susceptibility with metamorphic grade in ultramafic rocks is not consistent with the results of Clark et al. (1992) from the Agnew-Wiluna belt of the Yilgarn Block. A decrease in susceptibilities of ultramafic rocks with metamorphic grade in the Agnew Wiluna belt may indicate that hydrothermal alteration played a larger role in destroying magnetite that was formed during serpentinization. Since rock composition influences the products of hydrothermal alteration, for variably metamorphosed rocks there will be different alteration mineral products (Meuller and Groves, 1991; McCuaig and Kerrich, 1998). These variations in alteration mineral assemblages are generally consistent between gold deposits in rocks of the same metamorphic grade. Thus, as long as there are no other significant physical property- altering variables at work competing with mineralogical controls, some predictions can be made regarding the physical property characteristics of hydrothermally altered zones in metamorphosed rocks. An example of a mineralogical change related to increased temperatures and pressures of hydrothermal alteration-related sulfide precipitation that would have a particularly strong effect on physical property behavior, is formation of pyrrhotite instead of pyrite as the main gold-related sulfide (McCuaig and Kerrich, 1998; Hagemann and Cassidy, 2000). This is a high susceptibility mineral in its monoclinic form. The presence of monoclinic pyrrhotite would increase the susceptibility of mineralized areas, and could provide an important vector to gold mineralization. 74 Effects of alteration of metamorphic mineral assemblages must always be considered. Roberts (1988) explains how amphibolite facies rocks are known to be hydrothermally altered in the Archean orogenic gold setting to mineral assemblages reminiscent of a retrograde metamorphic assemblage (with chlorite, quartz and carbonate), or to an alteration assemblage similar to the amphibolite facies mineral assemblage (with biotite, garnet, anthophyllite, cummingtonite, cordierite, gedrite, and, staurolite). These changes to mineralogy will likely affect physical properties, such as magnetic susceptibility and density, which are known to be strongly controlled by mineralogy. 2.7. CONCLUSIONS Mineralogical and textural modifications within and between the different rock suites explain many of the physical property variations at Hislop. These are related to the range of geological processes, including igneous differentiation/fractionation, metamorphism, and hydrothermal alteration, that have affected the rocks throughout their history. The physical properties most useful for detecting prospective rocks at Hislop are magnetic susceptibility and density. This study illustrates predictable relationships between low susceptibility values and prospective felsic intrusive rocks and carbonate- altered mafic and ultramafic rocks in the immediate Hislop deposit area. Low density values will help confirm the presence of felsic rocks. The magnetic susceptibility and density cut-off values used to query the Hislop dataset in this study are considered useful for targeting prospective rocks in the Hislop area within physical property datasets generated from drill core measurements. The same cut-offs could be used for locating prospective areas within a 3D physical property model generated from geophysical inversions. Due to overlap between less prospective mafic and ultramafic volcanic rocks with low modal magnetite and prospective, carbonate altered rocks, any low susceptibility targets would have to be considered alongside other exploration criteria. The cut-off values could be used to filter physical property data as a first pass method of eliminating areas most likely to be barren. 75 In addition to using physical properties as a means to delineate prospective rocks in the Archean orogenic gold deposit environment, mean values, ranges, and standard deviations of physical property data for the different rock types, and altered subpopulations can be used to constrain geophysical inversions. Although there is some indication of relationships between hydrothermally altered rocks and resistivity there are not enough electrical property data to confidently use these relationships to identify prospective rocks. Far more magnetic susceptibility and density data were collected and analyzed during the course of this study than resistivity and chargeability data, and as such, there is increased confidence in interpreting magnetic susceptibility and density data, and 3D susceptibility and density inversion models. By comparison to a more regional scale physical property dataset, the Hislop physical property dataset is generally representative of rocks in this part of the Abitibi greenstone belt, with the exception of there being a greater variability in compositions of felsic intrusives at the regional scale. The Hislop dataset may be less representative of rocks in a similar mineral deposit environment at a different metamorphic grade. Obtaining prior information about an exploration site, conducting reconnaissance in the area of interest, and collecting representative rock samples would enable a geologist to determine if metamorphic grade, and any overprinting hydrothermal alteration might affect typical physical property ranges characteristic of the Archean orogenic gold deposit environment. Some knowledge of physical properties, be it expected values based on known rock types, or mineral assemblages, will vastly improve the interpretation of physical property models reulting from geophysical inversion. 76 REFERENCES Berger, B.R., 1999, Geological investigations along Highway 101, Hislop Township: Ontario Geological Survey, Summary of Field Work and Other Activities 1999, Open File Report, 6000, p. 5-1 — 5-8. Berger, B.R., 2002, Geological synthesis of the Highway 101 area, east of Matheson, Ontario: Ontario Geological Survey, Open File Report 6091, 124 p. Bohlke, J.K., 1989, Comparison of metasomatic reactions between a common CO2-rich vein fluid and diverse wall rocks: intensive variables, mass transfers, and Au mineralization at Alleghany, California: Economic Geology, v.84, p. 29 1-327. Bourne, B.T., Dentith, M.C., Trench, A., and Ridley, J., 1993, Physical property variations within Archean granite-greenstone terrane of the Yilgarn craton, Western Australia: the influence of metamorphic grade: Exploration Geophysics, v. 24, p. 367- 374. Boyd, D., 1984, Aeromagnetic surveys and gold, in Doyle, H.A., ed., Geophysical Exploration for Precambrian Gold Deposits, Geology Department and University Extension, University of Western Australia, Publication 10, p. 81-96. Bucher, K., and Frey, M., 2002, Petrogenesis of metamorphic rocks, seventh edition, Springer-Verlag Berlin Heidelberg, New York, 431 p. Card, K.D., 1990, A review of the Superior Province of the Canadian Shield, a product of Archean accretion: Precambrian Research, v. 48, p. 99-156. Card, K. D., and Ciesielski, A., 1986, DNAG Subdivisions of the Superior Province of the Canadian Shield: Geoscience Canada, v. 13, p. 5-13. 77 Cas, R.A.F., and Wright, J.V., 1987, Volcanic successions, modem and ancient: a geological approach to processes, products, and successions: Allen & Unwin, London, United Kingdom, 528 p. Clark, D.A., 1997, Magnetic petrophysics and magnetic petrology: aids to geological interpretation of magnetic surveys: Journal of Australian Geology and Geophysics, v. 17, p. 83-103. Clark, D.A., French, D.H., Lackie, M.A., and Schmidt, P.W., 1992, Magnetic petrology: application of integrated rock magnetic and petrological techniques to geological interpretation of magnetic surveys: Exploration Geophysics, v. 23, p. 65-68. Coggon, J.H., 1984, Geophysics as an aid to gold exploration in the Yilgarn block, in Doyle, H.A., ed., Geophysical Exploration for Precambrian Gold Deposits, Geology Department and University Extension, University of Western Australia, Publication 10, p. 113-13 8. Colvine, A.C., Fyon, J.A., Heather, K.B., Marmont, S., Smith, P.M., and Troop, D.G., 1988, Archaean lode gold deposits in Ontario: Ontario Geological Survey Miscellaneous Paper 139, 136 p. Cooper, M.A., 1948, Report on Kelore Mines Ltd., Ramore, Ontario, unpublished, 22 p. Dockery, B.A., 1984, Geophysical case history of the Lady Susan gold prospect, in Doyle, H.A., ed., Geophysical Exploration for Precambrian Gold Deposits, Geology Department and University Extension, University of Western Australia, Publication 10, p. 165-182. Doyle, H.A., 1990, Geophysical exploration for gold — a review: Geophysics, v. 55, p. 134-146. 78 Ellis, R.G., and Oldenburg, D.W., 1994, Applied geophysical inversion: Geophysical Journal International, v. 116, p.5-11. Fowler, C.M.R., Stead, D., Pandit, B.I., Janser, P.W., Nisbet, E.G., and Nover, G., 2005, A database of physical properties of rocks from the Trans-Hudson Orogen, Canada, Canadian Journal of Earth Sciences, v. 42, p. 555-572. Fyon, J.A., and Crockett, J.H., 1983, Gold exploration in the Timmins District using field and lithogeochemical characteristics of carbonate alteration zones, in Hodder, R.W., and Petruk, W., eds, Geology of Canadian Gold Deposits, Canadian Institute of Mining and Metalurgy, Special Volume 24, p. 113-129. Goldfarb, R. J., Baker, T., Dube, B., Groves, D.I., Hart, C.J.R., and Gosselin, P., 2005, Distribution, character, and genesis of gold deposits in metamorphic terranes, in Hedenquist, J.W., Thompson, J.F.H., Goldfarb, R.J., and Richards, J.P., eds., 100th Anniversary Volume, Economic Geology, v. 100, p. 407-450. Grant, F.S., 1985, Aeromagnetics, geology, and ore environments, II. Magnetite and ore environments: Geoexploration, v. 23, p. 335-362. Groves, D.I., 1993, The crustal continuum model for late-Archaean lode-gold deposits of the Yilgarn Block, Western Australia: Mineralium Deposita, v. 28, p. 366-374. Groves, D.I., and Foster, R.P., 1991, Archaean lode gold deposits, in Foster, R.P., ed., Gold Metallogeny and Exploration, Blackie and Son, Glasgow, United Kingdom, p. 63- 103. Groves, D.I., Ho, S.E., and Houstoun, S.M., 1984, The nature of Archean gold deposits in Western Australia with particular emphasis on parameters relevant to geophysical exploration, in Doyle, H.A., ed., Geophysical Exploration for Precambrian Gold Deposits, Geology Department and University Extension, University of Western Australia, Publication 10, p. 1-63. 79 Groves, D.I., Goldfarb, R.J., Gebre-Mariam, M., Hagemann, S.G., and Robert, F., 1998, Orogenic gold deposits: a proposed classification in the context of their crustal distribution and relationship to other gold deposit types: Ore Geology Reviews, 13, p. 7- 27. Gunn, P.J., and Dentith, M.C., 1997, Magnetic response associated with mineral deposits: Australian Geological Survey, Journal of Australian Geology and Geophysics Organisation, v.17,p.145-58. Hagemann, S.G., & Cassidy, K.F., 2000, Archaean orogenic lode gold deposits, in Hagemann, S.G., and Brown, P.E., eds., Gold in 2000, Society of Economic Geologists, Reviews in Economic Geology, v. 13, p. 9-68. Hallof, P.G., 1992, Resistivity and induced polarization, the spectral induced polarization method, in Van Blaricom, R., compiler, Practical Geophysics II for the Exploration Geologist, Northwest Mining Association, USA, p. 39-138. Hallof, P.G., and Yamashita, M., 1990, The use of the IP method to locate gold-bearing sulfide mineralization, in Fink, J.B., Sternberg, B.K., McAlistar, E.O., Weiduwilt, W.G., and Ward, S.H., eds., Induced Polarization: applications and case histories, Society of Exploration Geophysicists, Tulsa, Ok., p. 227-279. Harron, G.A., Middleton, R.S., Durham, R.B., and Philipp, A., 1987, Geochemical and geophysical gold exploration in the Timmins area, Ontario: a case history: Canadian Mining and Metallurgical Bulletin, v. 82, p. 52-57. Henkel, H., and Guzman, M., 1977, Magnetic features of fracture zones: Geoexploration, v. 13,p. 173-181. Hodgson, C.J., 1990, An overview of the geological characteristics of gold deposits in the Abitibi subprovince, in S.E. Ho, F. Robert, and D.I. Groves, compilers, Gold and Base Metal Mineralization in the Abitibi Subprovince, Canada, with Emphasis on the Quebec 80 Segment, Short Course Notes, Geology Department (key center) and University Extension, the University of Western Australia, publication No. 24, p. 63-100. Hodgson, C.J., 1993, Mesothermal lode gold deposits, in Kirkham, R.V., Sinclair, W.D., Thorpe, R.I., and Duke, J.M., eds., Mineral Deposit Modeling, Geologic Association of Canada, Special Paper 40, p. 635-678. Hodgson, C.J., and Hamilton, J.V., 1990, Gold mineralization in the Abitibi greenstone belt: end-stage results of Archean collisional tectonics, in Keays, R.R., Ramsey, W.R.H., and Groves, D.I., eds., The Geology of Gold Deposits: the Perspective in 1988, Economic Geology Monograph 6, p. 86-100. Hodgson, C.J., and Troop, D.G., 1988, A new computer-aided methodology for area selection in gold exploration: a case study from the Abitibi greenstone belt: Economic Geology, v. 83, p. 952-977. Hood, P., Irvine, J., and Hansen, J., 1982, The application of the aeromagnetic gradiometer survey technique to gold exploration in the Val d’Or mining camp, Quebec: Canadian Mining Journal, v.103, p. 21-39. Hoisner, W.T., and Schneer, C.J., 1961, Hydrothermal magnetite: Geological Society of America Bulletin, v. 72, p. 369-3 85. Isles, D.J., Harman, P.G., and Cunneen, J.P., 1989, The contribution of high resolution aeromagnetics to Archean gold exploration in the Kalgoorlie Region, Western Australia, in Keays, R.R., Ramsey, W.R.H., and Groves, D.I., eds., The Geology of Gold Deposits: the Perspective in 1988, Economic Geology Monograph 6, p. 3 89-397. Jackson, S. L., and Fyon, A. J., 1991, The western Abitibi subprovince in Ontario, in P.C. Thurston, H.R. Williams, R.H. Sutcliffe and G.M. Stott, eds., Geology of Ontario, Ontario Geological Survey, Special Volume 4, Part 1, p. 405-482. 81 Johnson, I., Webster, B., Matthews, R., and McMullen, 5., 1989, Time-domain spectral IP results from three gold deposits in northern Saskatchewan: CIM Bulletin, v. 82, p. 43- 49. Johnson, G.R., and Olhoeft G.R., 1984, Density of rocks and minerals, in Carmichael, R.S., ed., Handbook of physical properties of rocks; Volume III, CRC Press, Florida, p. 1-38. Kerrich, R., 1989, Geodynamic setting and hydraulic regimes: shear zone hosted mesothermal gold deposits, in Bursnall, J.T., ed., Mineralization and Shear Zones, Geological Association of Canada, Short Course Notes 6, p. 89-128. Kishida, A., and Kerrich, R., 1987, Hydrothermal alteration zoning and gold concentration at the Kerr-Addison Archean lode gold deposit, Kirkland Lake, Ontario: Economic Geology, v. 82, p. 649-690. Lapointe, P., Morris, W.A., and Harding, K.L., 1986, Interpretation of magnetic susceptibility: a new approach to geophysical evaluation of the degree of rock alteration: Canadian Journal of Earth Sciences, v. 23, p. 393-40 1. Li, Y., and Oldenburg, D.W., 1996, 3-D inversion of magnetic data: Geophysics v. 61, p. 394-408. McCuaig, T.C., and Kerrich, R., 1998, P-T-t-deformation-fluid characteristics of lode gold deposits: evidence from alteration systematics: Ore Geology Reviews, v. 12, p. 381- 453. Meuller, A.G., and Groves, D.I., 1991, The classification of Western Australian greenstone-hosted gold deposits according to wall-rock alteration mineral assemblages: Ore Geology Reviews, v. 6, p. 291-33 1. 82 Moore, E.S., 1936, Geology and ore deposits of the Ramore area: Ontario Department of Mines, Annual Report, 1936, v. 45, Pt. 6, 37 P. Ontario Geological Survey, 2001, Physical rock property data from the physical rock property study in the Timmins and Kirkland Lake Areas: Ontario Geological Survey, Miscellaneous Release — Data 91. Pelton, W.H., Ward, S.H., Hallof, P.G., Sill, W.R., and Nelson, P.H., 1978, Mineral discrimination and removal of inductive coupling with multifrequency IP, Geophysics, v.43, p. 588-609. Phillips, N.D., 2002, Geophysical inversion in an integrated exploration program: examples from the San Nicolas deposit: Unpublished M.Sc. thesis, University of British Columbia, 237 p. Power, W. L., Byrne, D., Worth, T., Wilson, P., Kirby, L., Gleeson, P., Stapleton, P., House, M., Robertson, S., Panizza, N., Holden, D. J., Cameron, G., Stuart, R., Archibald, N. 3., 2004, Geoinformatics evaluation of the eastward extension of the Timmins Gold Camp: Geoinformatics Exploration Inc., Unpublished report for St Andrew Goldfields Ltd. Prest, V.K., 1956, Geology of the Hislop Township: Ontario Department of Mines, Annual Report, 1956, v. 65, pt. 5, 5lp. Raudsepp, M., and Pani, E., 2003, Application of Rietveld analysis to environmental mineralogy, in Jambor, J.L., Blowes, D.W., and Ritchie, A.I.M., eds., Mineralogical Association of Canada, Short Course Series, Volume 31, Chapter 8,p.165-80. Ridley, J.R., and Diamond, L.W., Fluid chemistry of orogenic lode gold deposits and implications for genetic models, in Hagemann, S.G., and Brown, P.E., eds., Gold in 2000, Society of Economic Geologists, Reviews in Economic Geology, 13, p. 141-162. 83 Robert, F., 1990, Structural setting and control of gold-quartz veins of the Val D’Or area, southeastern Abitibi subprovince, in S.E. Ho, F. Robert, D.L and Groves, compilers, Gold and Base Metal Mineralization in the Abitibi Subprovince, Canada, with Emphasis on the Quebec Segment, Short Course Notes, Geology Department (key center) and University Extension, the University of Western Australia, publication No. 24, p. 167-212. Robert, F., Poulsen, K.H., Cassidy, K.F., and Hodgson, C.J., 2005, Gold metallogeny of the Superior and Yilgarn Cratons, in Hedenquist, J.W., Thompson, J.F.H., Goldfarb, R.J., and Richards, J.P., eds., 100th Anniversary Volume, Economic Geology, v. 100, p. 407- 450. Roberts, R.G., 1988, Archean lode gold deposits, in Roberts, R.G., and Sheahan, P.A., eds., Ore Deposit Models, Geoscience Canada, Reprint Series 3, p. 1-19. Ropchan, J.R., Luinstra, B., Fowler, A.D., Benn, K., Ayer, J., Dahn, R., Labine, R., and Amelin, Y., 2002, Economic Geology, v. 97, p. 29 1-309. Roscoe and Postle, 1998, Hislop Mine Property, Roscoe and Postle Associates Inc., St. Andrew Goldfields Ltd. internal report, unpublished, p. 66-89. Seigel, H.O., Johnson, I., and Hennessey, J., 1984, Geophysics the leading edge: Geophysics: the Leading Edge of Exploration, v. 3, p. 32-35. Telford, W.M., Geldart, L.P., and Sheriff, R.E., 1990, Applied Geophysics, Second Edition, Cambridge University Press, 770 p. Troop, D.G., 1986, Multiple orebody types and vein morphologies, Ross Mine, District of Cochrane: Ontario Geological Survey, Summary of Field Work and Other Activities, v. 132, p. 413-420. Williams, H., Stott, G.M., Thurston, P.C., Sutcliffe, R. H., Bennett, G., Easton, R.M., and Armstrong, D.K., 1991, Tectonic evolution of Ontario: summary and synthesis, in P.C. 84 Thurston, H.R. Williams, R.H. Sutcliffe and G.M. Stott, eds., Geology of Ontario, Ontario Geological Survey, Special Volume 4, Part 2, p. 1255—1332. Williams, N.C., 2006, Applying UBC-GIF potential fields inversions in greenfields or brownfields exploration: Australian Earth Sciences Convention, 2006, Melbourne, Australia, 10 p. Williams, N. and Dipple, G., 2005, Identifying sulfide mineralization from physical property measurements and its application to mineral exploration inversions, Geological Society of America, 2005 Salt Lake City Annual Meeting, Abstracts with Programs, v. 37, p. 23. Williams, P.K., 1994, Relationship between magnetic anomalism and epigenetic gold mineralization in the Victory-Defiance area, Western Australia, in, Dentith, M.C., Frankcombe, K.F., Ho, S.E.., Shephard, J.M., Groves, D.I., and Trench, A., eds., Geophysical signatures of Western Australian Mineral Deposits, Geology and Geophysics (Key Center) and UWA Extension, The University of Western Australia, Publication 26, Australian Society of Exploration Geophysicists, Special Publication 7, p. 283-296. www.mindat.org, Mindat.org mineral and locality database. www.standrewgoldfields.com, website for St. Andrew Goldfields, Ltd. www.zonge.com/LabIP.html, website for Zonge Engineering and Research Organization, IP and resistivity measurements. 85 Chapter 3: Detecting gold-related geology in Archean orogenic gold environments using geophysical inversion: a synthetic modeling study based on the Hislop gold deposit, Ontario2 3.1. INTRODUCTION 3.1.1. Rationale Three dimensional geophysical inversion modeling, involving the estimation of physical property distributions within the earth’s subsurface from observed geophysical data, is used widely as a tool to explore for a range of mineral deposit types. Yet, as different deposit types are characterized by unique combinations of rock types, mineralogy, structure, and morphology, each deposit type may not be equally well imaged by inversion. Expectations regarding the detectability and delineation of orebodies and related rocks in a given mineral deposit setting can be generated through synthetic forward and inverse modeling prior to actual geophysical inversion work. This study employs synthetic modeling to test the capabilities of geophysical inversion as an exploration tool in the Archean orogenic gold environment. In contrast to its more extensive use in imaging higher tonnage and higher grade deposits like volcanogenic massive sulfide, magmatic sulfide, and porphyry deposits (Oldenburg et al., 1997; Phillips, 2002; Farquharson et al., 2008), geophysical inversion is not as commonly used to explore for, or map, Archean orogenic gold deposits. As a result, there are fewer case histories successfully demonstrating its application, and thus there is less familiarity with the range of outcomes that can accompany inversion of different geophysical datasets over these deposits. Prior to any geophysical work, it is important to identify the geological and physical property characteristics of typical exploration targets. Gold mineralized rocks 2A version of this chapter will be submitted for publication. Mitchinson, D., and Phillips, N, 2009, Detecting gold-related geology in Archean orogenic gold environments using geophysical inversion: a synthetic modeling study based on the Hislop gold deposit, Ontario 86 are not easily detectable using geophysics due to low tomlages and grades (Doyle, 1990). Alternate exploration targets must therefore be sought. In the case of many orogenic gold deposits, geological features spatially related to gold include faults, felsic intrusive rocks, and hydrothermal alteration zones. These are commonly narrow, near vertical features that extend to depth, and are hosted within deformed and steeply-dipping stratigraphy. Physical property studies completed on rocks associated with the Hislop gold deposit, an orogenic gold deposit located east of the world renowned Timmins-Porcupine gold camp in Ontario, indicate that some known gold-related features have distinct physical property ranges that may allow them to be distinguished from likely barren host rocks (see Chapter 2). Synthetic forward and inverse modeling completed on simple 3D models based on the Hislop deposit tests whether these petrophysically distinct gold- related rocks can be detected using inversion methods. Synthetic modeling investigates whether the physical property contrasts are sufficiently strong, and if sizes, shapes, and locations, of gold-related geological features are such that they can be detected within a discretized earth model at a 1 km scale of investigation. Tests are devised to explore the effects of the addition of geological and physical property constraints. Results of the modeling reveal whether realistic physical property values can be recovered, thus lending confidence to the interpretation and querying of the recovered physical property model. Results furthermore highlight possible limitations of inversion at this scale. It indicates maximum depths of investigation, and can identify features not caused by a known source, but rather are artifacts or byproducts of the inversion algorithm. 3.1.2. Objectives Synthetic modeling work aims to answer a series of questions related to how well inversion is able to image prospective geologic features expected in the Archean orogenic gold setting. Specific questions include: 87 1. Can a feature of interest be imaged using unconstrained inversion at a <1 km scale of investigation? What range of geometry, and physical property contrasts, can we expect to image within this mineral deposit setting. 2. How well does the inversion reproduce the true model? What are the significant differences between the recovered and true models? What are the causes of discrepancies? 3. Can the model result be improved with addition of basic prior geological knowledge, and what types of constraining information are most effective in improving the model? What differences between the true and recovered models persist? 4. Which geophysical datasets are most beneficial to invert for orogenic gold exploration? What information can each data type provide to help better understand the geology of the subsurface? 3.2. BACKGROUND 3.2.1. Geology of the Hislop gold deposit and relationship to other Archean orogenic gold deposits The synthetic models presented herein are based on a simplified version of the geology of the Hislop gold deposit, and on average physical property-values determined for the range of significant Hislop rock types (Chapter 2). The Hislop gold deposit is located approximately 13 km southeast of Matheson, Ontario, in the gold and base metal-rich Abitibi greenstone belt of the Superior Province (Fig. 2.1, Chapter 2). The Hislop deposit area is underlain mainly by mafic and ultramafic 88 volcanic rocks that have been deformed into near-vertical structural panels (Prest, 1956; Berger, 1999; Power et al., 2004) Gold at Hislop (Fig. 3.1) is spatially related to an elongate, northwest-trending 30 m to 100 m wide, syenite dike occurring between a mafic, and an ultramafic volcanic unit (Cooper, 1948; Prest, 1956; Roscoe and Postle, 1998; Berger, 1999). The majority of gold occurs with disseminated pyrite within a strongly Fe-carbonate-altered, brecciated equivalent of the ultramafic unit lying adjacent to the southwest margin of the syenite (Cooper, 1948; Prest, 1956; Roscoe and Postle, 1998). Lesser gold occurs within quartz veinlets, stockworks and fractures in mafic volcanic flows north of the syenite, as well as in association with nearby porphyritic rhyolite dikes striking parallel to stratigraphy. Figure 3.1. Cross-section looking northwest through the Hislop deposit, showing areas of carbonate-dominated alteration and gold mineralization. Cross-section interpreted from drill core logged from the Hislop property. DDH H9601 DDH Ext 280, GK 280, and H9605 Multi-lithic Volcanic Breccia Lamprophyric Dike Intermediate Dike Porphyritic Rhyolite Dike Syenite Intrusive [] Mafic Volcanic Rock Ultramafic Volcanic Rock Fault — .‘ Drill trace 89 Geology and alteration mineral assemblages at Hislop are common to many greenschist facies-hosted Archean orogenic gold deposits globally. Orogenic gold deposits are epigenetic, and structurally-controlled. They can be hosted in any rock type, although Fe-rich mafic and felsic intrusive rocks are commonly in spatial proximity. Fe- carbonate + muscovite dominate gold-related hydrothermal alteration mineral assemblages, which usually extend only short distances (centimeter to meter scale) orthogonal to mineralized veins and structures. Gold occurs predominantly adjacent to, or within quartz-carbonate veins, or directly within host rocks associated with disseminated sulfides. A summary of characteristics defining Archean orogenic gold deposits is given in Table 3.1. Because of the shared characteristics between the Hislop deposit and other orogenic gold deposits, results from this study may be useful in guiding inversion work, and interpreting inversion results for other Archean orogenic gold deposits in the Abitibi greenstone belt, and globally. 3.2.2. Physical Properties of rock types and alteration zones at Hislop Petrophysical contrasts between likely mineralized, and unmineralized rocks are necessary to yield a geophysical target, and as such they must be identified and understood. One difficulty in targeting Archean orogenic gold deposits using geophysics is that, although gold itself is a conductive and dense mineral, it is usually low grade and thus does not contrast significantly enough from the host rocks to be directly detected by geophysical methods (Doyle, 1990). This means that other petrophysically distinct vectors to gold are required. At Hislop, petrophysically distinct target rocks include syenite and rhyolite dikes, carbonate-altered mafic and ultramafic volcanic rocks, and sulfide-rich zones. Results from a physical property study on the Hislop deposit (see Chapter 2) show that gold-related syenites and porphyritic rhyolite dikes in the Hislop area have low susceptibility and density ranges distinguishing them from higher susceptibility mafic and ultramafic volcanic rocks (Fig. 3.2). Magnetic susceptibility further separates most low 90 Table 3.1. Characteristics of Archean orogenic gold deposits. Age Tectonic setting Structural Host rocksllithological Hydrothermal alterationI Mineralization associationlcontrols on associations geochemical signature mineralization Examples: Form in extensional, Spatially associated with Can form in any rock Carbonate alteration Usually hosted in 2710-2670 Ma compressional, and large scale crustal type, however, Fe-rich muscovite/sericite alteration, throughgoing, quartz (Abitibi); transtensional structures; mainly mafic and ultramafic silicification, and albitization; carbonate veins, less 2630 Ma environments during controlled by second and volcanic supracrustal addition of CaO, CC2, Fe203 commonly as (Yilgarn deformational third order faults that rocks, sedimentary rocks (carbonate alteration), Si02 disseminated replacement Craton); processes at occur as splays off of the (fluviatile sequences), and 1<20, Ba, and Na20 zones, or as stockwork 2670 Ma convergent plate main fault zone; steeply felsic intrusives, are (muscovite alteration, veins. (Midlands margns. reverse to oblique, brittle common hosts in the silicification, albitization). greenstone to ductile shears zones Abitibi; gold-related faults belt, commonly occur at Zimbabwe contacts between contrast Craton) Darbyshire et al., 1997; Fyon and Crockett, 1983; Groves et al., 1995; Groves et al., 1998; Hagemann and Cassidy, 2000; Hodgson, 1989; Hodgson, 1990; Hodgson, 1993; Hodgson and Hamilton 1990; Hodgson and MacGeehan, 1982; Hodgson and Troop, 1988; Kent et al., 1996; Kerrich, 1989; Kerrich and Cassidy, 1994; Kerrich and Wyman, 1990; Kishida and Kerrich, 1987; McCuaig and Kerrich, 1998; Meuller and Groves, 1991; Robert, 1990; Robert, 2001; Roberts, 1988; Sibson et al., 1988; Weinberg et al., 2004. susceptibility carbonate-altered mafic and ultramafic rocks from high susceptibility, least altered precursors (Fig. 3.3). 3.20 ————- — 3.10 3.00 - 2.90 ________ 2.80 2.70 2.60 2.50 2.40 0.01 Electric properties, resistivity and chargeability, do not uniquely distinguish prospective rocks at Hislop (Figs. 3.4 and 3.5). There is a large overlap in resistivity values for the rock types studied at Hislop, with the only distinct resistivity range related to sheared talc-chlorite rich ultramafic rocks. These rocks, although not considered prospective, exhibit a fabric which results in lower resistivities (or higher conductivities) than other rocks in the area. Although not explicitly documented in the Hislop physical property study, it is expected that sulfide-rich areas would be conductive. Hence, conductivity ranges for sulfide-rich rocks in the synthetic models are derived from other sources documenting electric properties of rocks (Telford et al., 1990; Connell et al., 2000). As with resistivity, specific chargeability ranges do not characterize the individual Rock Types Ultramafic rocks Maflc rocks X Intermediate dikes A Syenite intrusives 0 Rhyolite porphyriesE U) 0 >b x e 0.1 I 10 100 Magnetic Susceptibility (x103 SI Units) 1000 Figure 3.2. Plot of magnetic susceptibility versus density for major rock units at Hislop. Syenite intrusives and porphyritic rhyolite dikes have distinctly low density and magnetic susceptibility ranges, allowing them to be distinguished from intermediate, mafic, and ultramafic rocks at Hislop. 92 3.20 E 0, 0 I 3.10 3.00 2.90 2.80 2.70 2.60 2.50 2.40 0.1 3.20 3.10 3.00 2.90 2.80 2.70 2.60 2.50 2.40 1 10 100 Magnetic Susceptibility (x103 SI Units) b. 1000 EJa; *Ip,” Unaltered to Altered Ultramafic Rocks QLst. altd. (Dol-chi) •Tlc+chi ultramafic DFeCb+ms altd. utlramafic flFe/MgCb+fu altd. ultramafic 0.1 1 10 100 1000 Magnetic Susceptibility (x103 SI Units) Figure 3.3. Plot of magnetic susceptibility versus density for variably altered mafic volcanic rocks, and variably altered ultramafic volcanic rocks from Hislop. Carbonate- rich alteration (pink and yellow diamonds, and pale green and yellow squares) destroys magnetite in mafic and ultramafic volcanic rocks, causing magnetic susceptibility to drop. Abbreviations in legends: Lst. altd. = least altered assemblage; Chl+ab = chlorite+albite assemblage; FeCb+ms = Fe-carbonate+muscovite; FeCb+ab = Fe carbonate+albite; Dol+chl = dolomite+chlorite assemblage; Tlc+chl = talc+chlorite; Fe/MgCb+fu = Fe/Mg-carbonate+fuchsite (chrome-muscovite). 93 6 IJitramafic volcanic rocks Mean 2816 4 - 2 _ Mafic volcanic rocks Mean 284321 4 - 2 - -Intermediate dikes rMean 9759 ri-1- 6 4 2 8 6 4 2 teintrusiveean6760i Porphyritic rhyolite dikes 11534 E 10 100 1000 10000 100000 1000000 Resistivity (Ohm-rn) Figure 3.4. Resistivity histograms for Hislop deposit rocks. Data indicates lower overall resistivities for ultramafic volcanic rocks from Hislop. 10 - Ultramaflc volcanic rocks - 8— — 6— — 4— _______— 2 - . Mean 6.91L 6 - Mafic volcanic rocks - 2 — lMean31.32r Intermediate dikes 6 Syenite intrusives - 4— - 2 - Mean 141 6 Porphyritic rhyolite dikes _____ I 1 10 100 1000Chargeability (milliseconds) Figure 3.5. Chargeability histograms for Hislop deposit rocks. Chargeability ranges for the individual rock types overlap and are not unique. 94 prospective rock types at Hislop, and in general, all rocks have low background values of chargeability consistent with published values (Chapter 2). However, values are expected to be high where the rock contains anomalous sulfide abundances (Telford et a!., 1990). In the synthetic models, sulfide zones are given values corresponding with the highest chargeabilities found at Hislop, and values published in Telford et al. (1990). Chargeability values referenced from the Hislop physical property study were divided by 1000 to yield values that correspond to the unitless 0-1 chargeability values that are output from induced polarization inversions. Physical property values used in the synthetic models are given in Table 3.2. It is important to note that gravity inversions produce density contrast models, thus, to build the starting models, density contrasts were used. For this study, the density contrast of each rock type is the difference between the rock’s density, and the average density value for all the major rock types (2.81 g/cm3). Density values are presented as both densities and density contrasts in Table 3.2. Note also that conductivities are used in the starting models for DC resistivity work, and that conductivity models are the product of DC resistivity inversions. Conductivity can be converted to resistivity by taking the reciprocal. 3.2.3. General forward modeling and inversion background This research employs forward modeling and inversion codes from the University of British Columbia Geophysical Inversion Facility (UBC-GIF). This section provides a brief overview of these applications. Further details are found in Li and Oldenburg (1996, 1998, and 2000). Forward modeling is essentially a tool for hypothesis testing. Subsurface physical property models are devised, a geophysical survey is simulated over the top of the model, and data are collected. The value of the data collected at each survey point are related to the location of the source in the subsurface, its physical properties, and the strength of the 95 inducing field in the cases of magnetic, electromagnetic, DC resistivity, and induced polarization methods. Table 3.2. Physical property values used in synthetic modeling. Density (glcrn3)ISusceptibility Conductivity ChargeabilityDensity contrastRock Type (S/rn) (ms)(SI Units) (glcm3) Syenite/rhyolite dike 0.00025 2.71-0.11 1.80E-04 0.016 Maficvolcanicrock 0.032 2.88/+0.07 1.50E-04 0.016 Ultramafic volcanic 0.0096 2.85/+0.04 2.27E-03 0.016 rock Carbonate altered ultramafic/mafic volcanic rock (for corn parison) Moderately sulfide-rich I .40E-02 0.16 rock 3.OOE-02 0.3Sulfide-rich rock Chapter 1 - Chapter 1; Sulfide AnomalousFrom Chapter 1 - rich rock values chargeabilities(Hislop deposit From Chapter 1 from Telford et al., from highestSource physical property 1990, and Connell chargeability study) et al., 2000 samples from Hislop study Anomalous low and high values are highlighted by blue and red borders, respectively 2.82/+0.01 3.20E-04 0.016 Geophysical inversion involves calculation of the subsurface physical property distribution from collected geophysical data. Subsurface physical properties are calculated based on the known physical relationships between sources and measurement locations at the Earth’s surface. Unlike forward modeling, the solution is non-unique. There are far more unknowns than there are data, and thus there are an infinite number of solutions. To reduce the number of possible solutions, a model objective function is defined. For default inversions the model objective function specifies that the model is required to be close to a background value, referred to as the reference model, and has to be smoothly varying in all directions. With increased knowledge of geology or physical 96 properties, the degree of closeness to the reference model can be manipulated, and the smoothing in the x, y, and z directions can be increased or reduced, by adjusting weightings within the model objective function. In addition to the model objective function, a data misfit is defined. The data calculated by forward modeling the inversion result must be sufficiently close to the observed data. The misfit and model objective function, respectively, are written: m =a $(m_mo)2dx+ax(m_mo)dx (d +aI—(m—m0)Idydy ) N2 +a..I—(m—m0)Idz dz .1) where N is the number of geophysical data, d0 is the observed data at location i, df’’ is the predicted data at location i, and e, is the standard deviation. ci is the alpha weighting determining the degree of closeness to the reference model, a, and determines smoothing in the x, y, and z directions, respectively, m is the model, and m0 is the reference model. 3.3. METHODS The 3D ‘Hislop-like’ geologic model shown in Figure 3.6a was converted to the four initial physical property models (Figs. 3.6b-3.6d), based on values in Table 3.2. Physical property models were created using the University of British Columbia 8i 97 Geophysical Inversion Facility’s (UBC-GIF) Meshtools3D program. Relative dimensions and scales of geologic features in these models are similar to those at Hislop. Unconstrained, geophysical inversions were completed using synthetic magnetic, gravity, DC resistivity, and induced polarization (IP) data generated from forward modeling magnetic susceptibility, density, conductivity, and chargeability models, respectively. Table 3.3 summarizes synthetic survey parameters used, and Table 3.4 summarizes inversion parameters. Observed data ‘collected’ over the starting physical property model, and predicted data generated from forward modeling the inversion result, are compared after each inversion to detennine if results are acceptable. Observed and predicted data, and achieved misfit values are found in Appendix 3A. Magnetic and gravity inversions are investigated first. Based on physical property work, this data is expected to be useful in distinguishing prospective low susceptibility and low density syenite dikes. The synthetic susceptibility model (Fig. 3 .6b) depicts a narrow vertical low susceptibility dike located between higher susceptibility mafic and ultramafic rocks. The density model (Fig. 3.6c) consists of a low density dike within higher density mafic and ultramafic rocks. For susceptibility analysis, the dike might also act to represent a low-susceptibility, strongly carbonate altered zone, along a fault between two higher susceptibility units. DC resistivity and IP methods are investigated for their ability to locate conductive and chargeable sulfides in the subsurface. These synthetic starting models have six sulfide-rich zones extending vertically to depth at the ultramafic rock-syenite dike contact (Figs. 3.6d and 3.6e). Additionally the conductivity models contain a talc+chlorite-rich ultramafic schist, incorporated to determine the effect of its unique range of conductivity values (Fig. 3 .6d). DC resistivity and IP data were collected using a Realsection electrode array, the configuration used in the collection of actual DC resistivity and IP data over the Hislop deposit in 1996 for exploration purposes. This type of electrode array employs widely spaced transmitter electrodes placed at a distance outboard of closely spaced receiver electrodes to collect data easily and quickly over 98 a. 3D geologic model .... Gold-related disseminated / i?rsulfldes / Carbonate+muscovite _j alteration of ultramafic rocks Syenite Intrusive fl Mafic Volcanic Rock Ultramafic Volcanic Rock Figure 3.6. a) 3D geological model based on the geologic setting of the Hislop gold deposit. b-e) North-facing cross-sections through 3D physical property models generated from the geologic model: b) magnetic susceptibility model, c) density model, d) conductivity model, e) chargeability model. Susceptibility and density modeling tests detectability of the syenite dike (the alteration zone is not considered here - syenite detection is focused on). Resistivity and chargeability modeling tests detectability of sulfide-rich zones, and low resistivity talc-chlorite dominated ultramafic rocks (carbonate alteration zone is included here). 300 b) Magnetic susceptibility Syenite dike c) Density 300 200 100 0 -100 Ultramafic volcanic rock V 2Dm ramafic canic rock Mafic volcanic rock Syenite C - •1 Mafic volcarrock b. 553800 554100 554400 553800 o os - 100417 200 50203 0.025 100 0,0107 0 -100 300 I 0000 200 0.00251 000202 000150 100 0.707 00 000054 0 Se-005 -100 ) Charceabilild) Conductivity 300 200- 100- 0- -100. 553800 Ultramfic sulfide zone volcan1 rock m Mafic volcanic Carhunate- rock altered ultramafic ______ d. P 0.20.158 0,117 0.070 0.0 000 -0.0 07 3 0 -0.05 I 016 0.136 0112 0.0882 0.0643 00404 00165 554100 554400 99 Table 3.3. Synthetic survey parameters. Model Data area (UTM) Lines Line Station Height # Data Data Other information spacing Spacing errors Magnetics x: 553800- 554400 E-W 50 m 10 m 320 m 793 5%; Inclination:75°; y: 5373000 - 5373600 5% DecIination:-12; floor Strength: 57478 nT Gravity x: 553800- 554400 E-W 50 m 10 m 320 m 793 0.01 y: 5373000 - 5373600 mGal floor DC x: 552800- 555400 E-W 100 m 20 m 300 m 2485 5% Realsection survey - Resistivity y: 5373000- 5373600 (ground) 5 Tx spacings: 1000 m, 1500 m, 2000 m, 2400 m, 3200 m IP x: 552800- 555400 E-W 100 m 20 m 300 m 2485 15% Realsection survey - y: 5373000 - 5373600 (ground) 5 Tx spacings: 1000 m, 1500 m, 2000 m, 2400 m, 3200 m Table 3.4. Synthetic inversion parameters. Inversion # Data Inversion core extents # Core Core cell # Padding Other (UTM) cells size cells Magnetic 793 x:553800-554400 144000 10m3 8000 y: 5373000 - 5373600 z: 300- (-)100 Density 793 x:553800-554400 144000 10m3 8000 y: 5373000 - 5373600 z: 300- (-)100 DC 2485 x: 553800 - 554400 18000 20 m3 30976 Near-surface cell Resistivity y: 5373000 - 5373600 weightings applied to z: 300- (-)100 reduce electrode noise lP 2485 x: 553800 - 554400 18000 20 m3 30976 Near-surface cell y: 5373000 - 5373600 weightings applied to z: 300 - (-)100 reduce electrode noise 100 large areas. Surface weighting files were used in the DC resistivity and IP inversion calculations in an attempt to subdue the tendency for high conductivity and chargeability values to accumulate at electrode locations (DCIP3D user manual, Version 2.1). After testing the four initial ‘Hislop-like’ models from Figure 3.6, the geometry of prospective features, and physical property contrasts between target and host rocks, are manipulated to explore the range of results. Inconsistencies between all true models and recovered models are identified and the cause of these discrepancies is assessed. For select cases, attempts are made to further minimize differences between true and recovered models by applying basic constraints based on prior physical property knowledge. These ‘non-located’ constraints (Phillips et al., 2007) are globally applied by adjusting the model objective function, the inversion function defining the type of model desired (e.g. a smooth model, a model close to a reference model; Li and Oldenburg, 1996). For this work, the difference between recovered and true models was calculated to compare the closeness of the recovered model to the true model, and to determine if a resulting model has improved with constraints applied. The sum of the differences in physical property values between each pair of equivalent cells from two identically sized models is calculated: Model difference = I m — m where N = the number of data, m = the physical property value of the th cell in the true model, and m’ = the physical property value of the equivalent cell in the recovered inversion model. This value gives a global relative measure of difference between the models. Since physical properties related to the different geophysical methods have different characteristic numerical ranges, model difference values might be much smaller for the results of one method versus another. Thus, calculated values can only be 101 compared between models generated by the same geophysical method. Model differences for all results are given in Table 3.5. Table 3.5. Model differences calculated between recovered and true models (the lowest model differences for each geophysical method are highlighted with bold text). Model Model Difference Magnetic susceptibility 20 m syenite between mafic and ultramafic rocks 1781.7 60 m syenite between mafic and ultramafic rocks 1656.9 60 rn syenite, buried 1818.8 20 rn syenite in mafic volcanic rocks 2396.4 20 m syenite in ultramafic volcanic rocks 721.5 constrained, reference model 0.03 SI Units 1232.4 constrained, upper bounds 0.035 SI Units 906.0 constrained, alpha y and z increased (100) 1644.3 constrained, alpha y and z increased, bounds 0.035 904.5 depth weighting decreased (13 and z0 decreased by 1/4) 705.4 depth weighting decreased, and upper bounds set at 0.035 701.8 Density 20 m syenite between mafic and ultrarnafic rocks 4561.7 60 rn syenite between mafic and ultrarnafic rocks 4702.0 60 rn syenite, buried 5355.3 20 rn syenite in mafic volcanic rocks 5613.3 20 m syenite in ultramafic volcanic rocks 3441.0 Conductivity 40 m sulfide-rich zones near ultramafic - syenite contact 158.4 40 m sulfide-rich zones - higher conductivity (0.03 S/m) 161.0 40 m sulfide-rich zones - laterally extensive zone 312.1 40 m sulfide-rich zones - one anomalous zone, no sheared ultrarnafic 160.8 40 m sulfide-rich zones - constrained, reference model 0.001 S/rn 173.3 40 m sulfide-rich zones - constrained, alpha y and z increased relative 154.4 to x 40 m sulfide-rich zones - constrained, reference model 0.001, alpha y 132.6 and z increased relative to x Dipole-Dipole survey 198.8 Chargeabi I ity 40 m sulfide-rich zones near ultramafic - syenite contact 1632.8 40 m sulfide-rich zones - higher chargeability (0.3 rns) 1742.4 40 rn sulfide-rich zones - laterally extensive zone 3362.9 102 3.4. SYNTHETIC MODELING RESULTS 3.4.1. Potential fields modeling Magnetic susceptibility models Hislop-like model: 20 m syenite dike hosted between mafic and ultramafic volcanic units The contact between the mafic volcanic unit in the east and the central syenite dike is well-resolved to depth by magnetic inversion (Fig. 3.7). In contrast, the contact between the ultramafic volcanic rock in the west and the syenite is essentially unresolved, and as such, the low susceptibility syenite dike is not imaged. Susceptibility values greater than 0.05 SI Units are attained within the area known to be occupied by the high susceptibility mafic volcanic unit. These susceptibilities are over-estimated compared to the known susceptibility of 0.032 SI Units for these rocks. Susceptibilities are underestimated where the ultramafic volcanic unit is present, assuming values close to 0 SI Units compared to true susceptibilities of 0.0096 SI Units. Near the surface, there appears to a low susceptibility ‘overburden’, where surface cell susceptibility values drop to 0 SI Units. Varying geometry Two geometrical variations on the previous model were tested. These new models encompass a syenite dike of greater width, and a buried syenite dike. A 60 m wide syenite dike is resolved near the surface in its correct location, down to about 150 m (Fig. 3.8a). The contact between the syenite and both mafic volcanic and ultramafic volcanic units are detected. With depth however, the geologic contacts are no longer well constrained. The buried syenite dike inversion result was comparable to the result for the 20 m dike model with the mafic volcanic unit-syenite dike contact being well-imaged to depth (Fig. 3.8b). Again, the ultramafic rock-syenite dike contact is poorly detected and 103 the syenite is not fully resolved. For the geometrically varied models, problems with over- and underestimation of susceptibility persist. Figure 3.7. Starting model and unconstrained magnetic inversion result for the ‘Hislop like’ magnetic susceptibility model. Results are shown at the same susceptibility scale as the starting model. The contact between the mafic volcanic unit and the syenite is detected to depth, whereas the contact between the syenite and the ultramafic unit is undetected. Susceptibility values in association with the mafic volcanic rock unit in the recovered susceptibility model are overestimated (>0.05 SI Units, compared to -M.03 SI Units in true model). Varying physicalproperty contrasts Two additional models are tested to explore the effect of varying the susceptibility contrast between the target rocks — the syenite dike — and the host rocks. The 20 m syenite dike is first modeled within a mafic volcanic host, and then within an ultramafic volcanic host rock. Geophysical inversion over a syenite dike hosted in high susceptibility mafic volcanic rocks successfully locates a narrow vertical low susceptibility zone near surface, and down to approximately 350 m depth (Fig. 3.9a). As in previous results, the central low susceptibility zone smoothes outward with depth in the model. The low susceptibility values within surface cells persist, and susceptibility values assigned to areas 300 05 4 - 00417 200 00323 - 0025 00 57 000 33 - 0 SI Units Starting model Recovered susceptibility model I 001 0.0417 0 .03 2 3 0.525 0.07 67 0 .0 05 33 SI Units 100 -100• 553800 554100 554400 553800 554100 554400 104 corresponding with the location of mafic volcanic rocks are overestimated, especially at depth. The host rock to the syenite dike is next changed from relatively high susceptibility mafic volcanic rock, to a relatively moderate susceptibility ultramafic volcanic rock to investigate resulting inversions. Results indicate the presence of a low susceptibility zone down to 250 m (Fig. 3.9b). In general, although there is a higher contrast between mafic volcanic rocks and syenite, the inversion results for the syenite dike hosted within ultramafic rocks has recovered values more consistent with the true model susceptibility values (see model difference values in Table 3.5). Figure 3.8. Starting models and magnetic inversion results with changes made to geometry of the target body. a) Result for the 60 m syenite hosted by mafic and ultramafic rocks. The location of the syenite is well-imaged near-surface. b) Result for the buried 60 m syenite. The syenite dike is undetected, and the result similar to the initial ‘Hislop-like’ susceptibility model. I°- 0.0417 0.0033 - 0.025I: olSI Units 554400 I 00457 00323 0.025 0.0161 0,30633 SI Units 554400553800 554100 105 300 Starting model Recovered susceptibility model a. 0.05 005 200 - - I 0417 10 0411 0.0333 00333 100 — 0.025 0.025 0.0161 00161 0.00633 0.00533 0 7 SI Units SI Units -100 553800 554100 554400 553800 554100 554400 553800 554100 554400 I 0050 0411 0.0 3 33 0.025 0 0167 0,0 08 33 SI Units 553800 554100 554400 I 1,05 0. 0411 0 0333 0 026 0 0167 0.00833 SI Units Figure 3.9. Magnetic inversion results for starting models with different physical property contrasts between the target and host rocks. a) Result for the 20 m syenite dike hosted within high susceptibility mafic volcanic rocks. The narrow syenite is detected to depth. Susceptibility values for mafic volcanic rocks are overestimated. b) Result for the 20 m syenite dike hosted in moderately susceptible ultramafic volcanic rocks. Again the syenite is detected to depth. The susceptibility scale is kept the same for all models for comparison — however the features in the recovered model in Figure 3.9b would be better visualized with the scale set to have a lower maximum value. 106 Density models Hislop-like model. 20 m syenite dike hosted between mafic and ulframafic volcanic units The starting Hislop-like density model differs from the susceptibility model in that the ultramafic and mafic volcanic host rocks have similarly high densities, and contrast nearly equally with the low density 20 m syenite, compared to the variable susceptibility contrast between the syenite and rock units on either side in susceptibility models. The gravity inversion result reveals the low density syenite and indicates its extent down to about 200 m depth (Fig. 3.10). The body terminates beyond this as the model becomes smooth. The slightly smaller density contrast between the ultramafic volcanic rock and the syenite, compared to the mafic volcanic rock and the syenite, is apparent in the marginally weaker detection of the western contact of the syenite. In general density values are well-estimated throughout the central region of the model. However, as with magnetic inversion results, there is some overestimation at depth with estimated density contrasts for mafic volcanic rocks of approximately 0.117 g/cm3 versus known values of 0.07 g/cm3. Varying geometry The eastern contact between the syenite dike and adjacent mafic volcanic rock is better resolved to depth where the syenite width is increased to 60 m, with the central low density zone now being imaged to approximately 300 m (Fig. 3.lla). Where the 20 m syenite is buried deeper (Fig 3.11 b), the low density body is essentially unresolved by gravity inversion. It is apparent that there is some decrease in density from east to west across the model, but distinct geological units are not obvious. Despite this, density values are still relatively well-estimated throughout the central part of the model. 107 Starting model Figure 3.10. Starting model and unconstrained gravity inversion result for the ‘Hislop like’ density contrast model. The contact between the mafic volcanic unit and the syenite is better located than the contact between the ultramafic unit and the syenite. Density contrasts are reasonably estimated throughout the central parts of the recovered model. Figure 3.11. Gravity inversion results with changes made to geometry of the target body. a) Result for the 60 m syenite hosted by mafic and ultramafic rocks. The syenite is well- imaged to depth. b) Result for the buried 60 m syenite. The syenite dike is essentially undetected, however an overall change in density from east to west is detected by the inversion, with the contact between mafic and ultramafic rocks being detected near- surface. 553800 554100 554400 I020,158 0,117 0,075 0.0323 -000833 .0.05 gfcm 553800 554100 554400 I020188 0111 0018 0.0333 -000833 -005 gicrn3 Starting model300 200 100 0 -100 200 300 I020.758 0,711 ‘0.318 0.8333 -000833 0 -005 gicm 100 553800 554100 554400 I0,20158 0,7 1 0,015 00303 ‘0 00832 -000 g/cm -100- 553800 554100 554400 108 Varyingphysicalproperty contrasts As the two host rocks in the Hislop-like model are characterized by similar density values, the results of the single-host inversions (Figs. 3.1 2a and 3.1 2b) are not dramatically different from the two-host results. Where either mafic volcanic rocks or ultramafic volcanic rocks are the sole host for a syenite dike, the inversion is capable of imaging the low density syenite to depths ranging from 150-200 m. These depths are slightly less than the depths resolved in the susceptibility inversion results for the single- host scenario. The density model becomes smoother with depth. Discussion of potential fields inversion modeling results Magnetics The presence of a 20 m syenite dike hosted between an ultramafic, and a mafic volcanic unit is not obvious in synthetic magnetic inversion results. The inversion only detects an overall gradient here between the lower and higher susceptibility areas. The narrow dike is more successfully imaged between the two different hosts when it has a slightly greater width, or when it is hosted by a single high susceptibility rock type. The most significant differences between the recovered and true magnetic susceptibility models include 1) smoothing across known contacts, especially across contacts where there is a low susceptibility contrast (such as the contact between the syenite dike and the ultramafic volcanic rock), 2) smoothing with depth, and 3) incorrect estimation of susceptibilities through the model, which generally yields higher maximum susceptibility values than in the true model. The third item encompasses the issue of low susceptibility values being incorrectly assumed near surface. 109 I02o so 0117 O 075 o oooo .0 008)3 -005 gIcm Figure 3.12. Inversion results with different physical property contrasts between the target and host rocks. a) Result for the 20 m syenite dike hosted within higher density mafic volcanic rocks. The narrow syenite is detected down to about 200 m depth. b) Result for the 20 m syenite dike hosted in ultramafic volcanic rocks. There is a marginally deeper detection of the syenite. 300 200 100- 0 300 200 100 0- 553800 554100 554400 I ° 0158 all, 0076 00333 .0 .005 gfcm -100. 553800 300 200 100 0- 554100 554400 200I SO 0.158 0117 00’s 0.0 33 3 -o c-ooo .005 glcm1 4(50.6 100 300 I020 50 0117 0075 00333 .0008: .005 g/cm 0 553800 554100 554400 553800 554100 554400 Most of the smoothing within the model is a byproduct of the inversion algorithm and choice of model norm. The model objective function for linear potential fields inversions is written such that a smooth, simple model result is calculated. This is facilitated using a L2 norm calculation which minimizes structure over the volume (Li and Oldenburg, 1996). A smooth model, in the case where there is little prior information about the subsurface, would be desired over a more structured and complicated model. Smoothing at depth is related to the hr3 decay of the magnetic signature with depth in the subsurface, and the inversion cannot resolve features deeper than the sources that have influenced the magnetic data collected. 110 Poor susceptibility estimation in near-surface cells may be related to depth weightings. A default depth weighting is written into all potential field inversion calculations (Li and Oldenburg, 1996 and 1998). As there is no inherent depth resolution in potential fields data, when an unweighted inversion is carried out, all susceptibility will occur at the surface as it is the simplest solution explaining the observed data (Li and Oldenburg, 1996 and 1998). Although depth weighting is necessary to offset this, the surface cells appear to be less sensitive, and tend to assume reference model values, which for default inversions is 0 SI Units. The generally overestimated susceptibility values for the model as a whole may also be explained by the depth weighting. To compensate for the lack of susceptibility at the surface, it is necessary for the inversion to place high susceptibility values at depth and increase their overall magnitude, in order to fit the observed magnetic data. Gravity The essentially equal contrast between higher density mafic volcanic rocks and ultramafic volcanic rocks, and the lower density syenite, allows for consistent detection of the narrow syenite dike, unless it is buried. Gravity inversions follow similar calculations as magnetic inversions, with the main difference being only the forward model solution - the physical relationships between subsurface sources and data observations. Thus explanations for discrepancies between true and recovered density models are similar to those for magnetic susceptibility models. As with magnetic inversions, the significant discrepancies are related to smoothing and depth resolution. Smoothing can be related to the choice of model norm used in the inversion algorithm, and decreasing resolution with depth is explained by the known hr2 decay of the gravity signal with depth. Contacts are not resolved as deep as they are with magnetic inversions. Relative density contrasts versus magnetic susceptibility contrasts might cause the depth detection to be inconsistent between the magnetic and gravity inversion results (i.e. the contrast between low 111 susceptibility syenite and high susceptibility mafic volcanic rocks is greater than the density contrast between the two rocks). The tendency of the near-surface cells to assume values near 0 g/cm3, the reference model value, likely necessitates having overestimated densities at depth. Combining magnetic and gravity inversion results would better detect the syenite where it is not resolved in the two-host model by magnetic inversions alone. The density result would better locate the ultramafic-syenite contact, and the magnetic inversion result would contribute by providing better depth information. 3.4.2. DC resistivity and induced polarization modeling Synthetic conductivity and chargeability models are likely less representative of true subsurface geology than susceptibility and density models. Rock type and associated mineralogy strongly influence magnetic susceptibility and density values (Chapter 2) and thus using geology to create starting physical property models is easily justified. Although rock type can play a role in determining conductivity and chargeability values and distribution, these physical properties are more strongly controlled by rock texture, permeability, and the presence of fluids (Telford et al., 1990). Thus it must be kept in mind that in nature, more complicated conductivity and chargeability distributions likely exist than can be represented by the synthetic models. Resistivity Models Hislop-like model: 40 m wide sulfide rich zones near ultramafic rock-syenite dike contact Six high conductivity zones (only three visible on cross-section) within significantly lower conductivity host rocks, but proximal to a moderately high conductivity, sheared ultramafic volcanic unit were not resolved through DC resistivity 112 inversion (Fig. 3.13). The moderately high conductivity ultramafic rocks, however, were imaged, but only to a depth of approximately 200 m. The recovered conductivity anomaly associated with the ultramafic unit extends faintly toward the general location of the high conductivity sulfide-rich zones. Overall, recovered conductivity values for low conductivity areas are close to true values, however conductivities associated with the sheared ultramafic unit, and obviously those associated with the anomalous conductivity zone, are underestimated. Figure 3.13. Starting model and unconstrained DC resistivity inversion result (conductivity model) for the ‘Hislop-like’ conductivity model. The sulfide-rich high conductivity zones are undetected. The moderately conductive sheared talc-chlorite rich ultramafic rock is detected near-surface, and resolved only to about 200 m depth. Varyingphysicalproperty contrasts and geometry Doubling the conductivity for the discrete sulfide zone model so as to represent a more sulfide-rich rock (Table 3.2), does not improve the recovery of the sulfide rich zones, as they remain undetected (Fig. 3.1 4a). Since raising the conductivity values of the sulfide zones was ineffective, the high conductivity zone was modified to be a continuous zone extending from north to south and vertically to depth to test whether a more extensive feature might lead to a better recovery (Fig 3.14b). Although the zone is now persistent and reaches the surface, no longer consisting of discrete small zones at depth, it remains undetected by inversion of DC resistivity data. As with the discrete sulfide zone model, a conductivity high is detected in association with the sheared ultramafic unit, and 300 200 0.014 S/rn I 0003 0 00251 0,00202 000153 1001 000100 0000042 50.005 S/rn 300. Startina mode’ 0003 200 - 0,00251 0.00 202 100- 0,00103 0.001 03 0 - 0.000542 50-005 S/rn -100- 553800 554100 554400 0 -lOG. 553800 554100 554400 113 30O 300 0.002 1000251 2001 0,00202 0.00152 0.00103 0.000542 51-005 S/rn 1001 300 Starting model 0 003 200 0,00202 100- 000152 0 00102 0 .0 00 54 2 50-000 S/rn -100 553800 554100 554400 Starting model 300 200 100 -100 553800 554100 554400 0 553800 554100 554400 0- -i00i 553800 554100 0 554400 I 0.00010 0 000101 0 500172 0.000153 0. 0001 54 0 000145 0000135 S/rn Figure 3.14. DC resistivity inversion results (conductivity models) with changes made to physical property contrasts, and to the geometry of the target body. a) Conductivities are doubled for the sulfide-rich zones, but the targets remain undetected. b) The target is made to be continuous. There is a weak indication of the conductive zone. c) All host rocks are given the same low conductivity background value. Comparing the result at the same scale, (i), the target is essentially undetected. Adjusting the scale (ii) reveals a conductive body, but with highly underestimated associated conductivities. I 0000 0 0 00251 000202 0 001 000103 3 0 000542 50-005 S/rn 200 I 0.003 0.0025 0 002 0.5015 0.0 01 0.0 005 S/rn 200- 100- 1 553800 I 0.003 0,00 20 1 0.0 0202 000102 0.001 03 0 00004 50-005 S/rn 554100 554400 553800 554100 554400 0.03 S/rn b. C. 0.03 S/rn 114 this zone extends laterally out toward the continuous sulfide zone, marginally more so than in the previous results. The moderately conductive ultramafic unit was removed, and all other units, aside from the sulfide-rich zone, were assigned a low conductivity value representing mafic volcanic rocks, to test a single continuous high conductivity zone in a low conductivity background (Fig. 3.l4c). Comparing the results to the true model at the same conductivity scales, demonstrates there is essentially a complete lack of resolution of the target feature. When the scale is adjusted to encompass the true recovered maximum and minimum values, the feature is revealed. The recovered body is correctly located near the surface, but extends only about 150 m depth. The values coinciding with the high conductivity zone however, are very low and near background values, and would not be considered of an anomalous nature. Chargeability Models Hislop-like model: 40 m wide sulfide rich zones near ultramafic-syenite contact Six sulfide-rich zones (only three visible in the cross-section) are assigned anomalous chargeability values. The anomalous values (Table 3.2) are chosen based on the highest chargeabilities measured from the Hislop deposit chargeabilty studies (see Chapter 2), and are divided by 1000 to correspond with IP inversion outputs. They are modeled within a low chargeability background. The zones are detected as a single, small anomalous area near the surface, which coincides with the top of the upper sulfide-rich zone (Fig. 3.15). The chargeability values estimated by the inversion are low compared to those of the true model. 115 300 Recovered charreability model Figure 3.15. Starting model and unconstrained IP inversion result for the ‘Hislop-like’ chargeability model. The sulfide-rich high chargeability zones are only detected near surface down to about 125 m. The chargeability values estimated by the inversion for the chargeable zones are much lower than known values. Varying geometry andphysicalproperties The chargeability of the sulfide zones is doubled from the initial model to test their subsequent resolution (Fig. 3.1 6a). The inversion result is essentially identical to the previous model result, but with marginally higher chargeability values coinciding with known sulfides. The upper chargeability zone is detected, but chargeability values are underestimated. Where the sulfide zone is made a continuous feature with doubled chargeability values, it is well-located in the subsurface by the inversion, extending down to approximately 300 m depth (Fig. 3.16b). The representative chargeability values are underestimated for the sulfide zone. There is some excess structure occurring near the surface and at depth, which does not occur in the true model. 200 lOOi 0 100I I Oil 0136 0,112 O 0602 O 0643 O 0484 00160 116 300 Figure 3.16. IP inversion results with changes made to physical property contrasts and geometry of the target body. a) Chargeabilities are doubled for the sulfide-rich zones. The resolved feature remains restricted to the near surface, but has slightly higher chargeabilities than the previous model result. b) The high chargeability target is made to be continuous. The target is well-located to almost 300 m depth, but has underestimated chargeabilities. Some additional structure in the model is found in near-surface cells and at depth. Discussion of DC resistivity and IP inversion modeling results DC Resistivity The most significant discrepancy between true and recovered conductivity models is the lack of resolution of the central ‘sulfide-rich’, high conductivity zones. Even when the starting model contains only one persistent conductive feature, the recovered values are too low to be considered anomalous. This suggests that the feature may be too narrow, or the contrast between host and target rocks too weak. The fact that the western 200 100 1001 553800 I 0.16 0.136 0.112 00842 00643 0.0404 00165 554100 554400 553800 554100 554400 300 Recovered charleabilit model 0.16 016 jO,136 200 10.136 0112 0.112 00862 100 00882 00643 0.0643 00404 0 00404 0.01 65 00166 -1001 553800 554100 554400 553800 554100 554400 117 ultramafic volcanic unit is resolved, a unit characterized by lower conductivities than the sulfide-rich body suggests that a lack of contrast can not be the explanation. The width of the feature is more likely to hinder its detection. In 3D forward modeling processes, to calculate voltages at a point, conductivities of cells within a 3D mesh must be averaged with those of neighboring cells over the cell volume (Dey and Morrison, 1979). In the case of the Hislop model, there is only one cell boundary that divides the anomalously conductive sulfide zone (the unit is 40 m thick and the cells 20 m wide), and the high value is essentially retained only in the averaging over this one particular boundary. At the dike’s contacts, the high conductivities are averaged with the lower conductivities of the host rock, bringing the values down quickly to background conductivities. The aberration in conductivity is thus only weakly represented in the data to be inverted. Making cell sizes smaller, may give the feature a better chance to be accounted for, however, increasing the number of model cells causes the inversion problem to be exceedingly large, especially when employing data from Realsection arrays, where a large mesh is required. The resolved moderately conductive ultramafic unit is only imaged to about 200 m depth, highlighting the second major discrepancy between true and recovered models - the lack of resolution at depth. This can mainly be attributed to the specific electrode array used. A Realsection array is similar to gradient or Schlumberger arrays in that transmitter electrodes are spaced at large distances outboard of the receiver electrodes (Telford et al., 1990). The poor resolution associated with wide electrode spacings is discussed by Hallof and Yamashita (1990). With increasing distance from the transmitter, the current weakens. Therefore a distant source that is intersected by the current will produce a weak signal, and will likely only be detected when the receiver electrode is sufficiently close, and in this case receiver electrodes are on surface. The final discrepancy between true and resolved models is the underestimation of conductivity values for known higher conductivity areas. This is interpreted to be primarily associated with smoothing and the subsequent dispersal of conductivity over larger volumes within the mesh. Based on the model objective function defined, the 118 inversion solution favors the assignment of low conductivities over many cells, rather than high conductivities within more compact volumes. Inducedpolarization The chargeable zones are imaged to various degress in each of the IP inversion model results, with depth detection improving with increased chargeability contrast between the sulfides and the host rocks. However, the recovered chargeability values corresponding to the sulfide-rich features in each case are underestimated. As with conductivity model results, the small width of the feature likely limits its detection. However, in contrast to the forward modeling of conductivity models, an averaging of chargeabilities between neighboring cells is not used in the IP forward model solution. This means the chargeability values are not as diminished during this step of the synthetic modeling process, which explains the slightly better resolution of the small chargeability anomalies. The poor resolution at depth known for Realsection surveys further reduces the possibility of resolving the sulfide-rich zones. As with recovered conductivity models, smoothing resulting from the model norm contributes to the dispersion of chargeabilities over a larger volume. The smoothing of the conductivity anomaly over more cells than what are known to contain anomalous values, means that each cell requires less chargeability overall in order to explain the observed data. The irregularly dispersed high chargeability values at surface thought to be related to the increased sensitivities at electrodes may also partly explain why chargeability values are lower than those from equivalent areas within true models — surface cells might already be taking up some of the required chargeabilty needed to explain the observed data, resulting in lower values elsewhere. 119 3.4.3. Improving model results with basic constraints Inversion model results can be improved by constraining the inversion with additional geologic and physical property information (Phillips, 2002; Williams, 2006). This information can be inferred from the exploration deposit model, from published data, or from direct reconnaissance work. Prior information can be incorporated into the inversion calculation through basic manipulation of the model objective function to produce a model consistent with known geology, physical properties, and geometry (Li and Oldenburg, 1996 and 1998). As with all inversion results, the result must still fit the data within the specified misfit. In this section, basic constraints are applied to inversions to try and reduce the discrepancies between recovered and true models evident from unconstrained inversions. The constraints are tested only for the Hislop-like resistivity and magnetic susceptibility models. The unconstrained inversions for these models did not fully delineate the target rocks, and as such, these two cases constitute good candidates for testing the possibility of model improvement with constraints. For the globally constrained inversions, physical property values known to be representative of the subsurface geology are used to recover more accurate physical property values, and knowledge of general structural orientations is applied to encourage smoothing of features in the desired directions. With significant amounts of prior physical property knowledge in the form of physical property measurements or geological 3D models, thorough and complex constraints can be applied, however, in this work simple solutions to modifying the models are explored. The first constraint tested is use of a reference model, which in this case is a single physical property value that is considered representative of expected values. The inversion is required to yield a model close to this reference model, while satisfying the remaining terms of the inversion algorithm. The second constraint tested is setting physical property bounds on the inversion results. The default setting for UBC-GIF inversions usually allows a large range of values to be assumed by the model cells. The bounds can be adjusted to yield results within the range of known or expected values. The 120 third constraint tested is a geometrical constraint and is chosen based on known geological directionality, or preferred orientations. The table of model difference values (Table 3.5) can be referred to here to evaluate the quantitative improvements in model estimation in accordance with the various constraints applied. A decrease in the calculated model difference reflects smaller differences between true and recovered physical properties. Magnetic constraints: 20 m syenite dike between ultramafic and mafic volcanic units Reference model A constant reference model value of 0.03 SI Units is used, representing expected high susceptibilities of mafic and ultramafic rocks. As in unconstrained results, only the contact between the contrasting mafic volcanic rock and syenite is detected (Fig. 3.1 7a). The susceptibility contrast between the ultramafic volcanic unit and the syenite continues to be elusive. Near-surface cells show again the tendency to acquire reference model values, in this case values around 0.03 SI Units (previously 0 SI Units reflecting the default reference model). The calculated model difference is an improvement in overall recovery of the true susceptibility values compared to unconstrained results. This improved susceptibility estimate is explained by the new reference model value (> 0 SI Units) being assumed by the surface cells, resulting in the necessary lowering of susceptibility at depth, where susceptibilities in relation to the mafic volcanic unit were initially highly overestimated to compensate for low surface values. Geological contacts are slightly better located. 121 300 pre!erence model 300 Constrained using bounds Figure 3.17. Inversion results for the Hislop-like susceptibility model after constraints applied, a) Inversion result with reference model set to 0.03 SI Units; b) Result with bounds set from 0 to 0.03 5 SI Units; c) Result after ctz and cty increased relative to cLx; d) Result with upper bounds set to 0.035 SI Units, and alpha cLz and cty increased. Bounds Bounding the model using a upper susceptibility bound of 0.03 5 SI Units, a value slightly higher than the known susceptibility of mafic volcanic rocks here, and more suitable than the default upper bound of 1 SI Unit, yields a more qualitatively and quantitatively accurate model (Fig. 3.17b). The cap on the susceptibility values keeps susceptibility from being significantly overestimated. Low susceptibilities near-surface, thought to be caused in previous models by depth weighting, are minimized with appropriate bounds. An overall lowering of the susceptibilities in the vicinity of the mafic volcanic rocks means the high susceptibility zones are not pushed as deep to effectively reproduce the observed data. The mafic volcanic unit/syenite dike contact is well-located, I 200I 06 00617 60633 0026 60167 0,00633 SI Units 100 I -100 553800 554100 554400 0 300 Constrained using alpha weightings 0025 60107 000033 SI Units 554400 I 0.05 0.0417 0.0333 0025 0. 01 67 7 00033 SI Units 553800 554100 554400 553800 554100 554400 122 extending to a depth of about 400 m, however the presence of the syenite dike, is still not obvious. Alpha weightings Use of alpha (cL) weightings to achieve smoothing along the z and y axes reflecting known structural orientations, results in sharper contacts within the model, but causes unnecessary vertical exaggeration (Fig. 3.1 7c). Model difference calculations (Tab. 3.5) indicate that susceptibility value estimation has not improved with respect to the unconstrained result. Without putting any restrictions on susceptibility values, susceptibilities are still overestimated. Combined bounds and directional weighting By combining alpha weighting in y and z directions with use of more appropriate upper bounds values, a well-estimated model results, with slightly sharper contacts than when bounds alone are constrained (Fig. 3.17d). Model difference values show this result is not necessarily an improvement on setting upper bounds exclusively. Although the gold-related feature, the syenite dike, is not better imaged, the physical property model values are better estimated, and thus geological interpretations of the model will improve. Resistivity constraints: 40 m wide sulfide rich zones near ultramafic rock-syenite dike contact. Reference values The reference value for the conductivity inversion was set to 0.00 1 S/rn, a value lying approximately between the higher conductivity ultramafic and mafic volcanic rocks in order to improve the overall conductivity estimations within the model. The sheared ultrarnafic rocks are resolved to a slightly greater depth than in the unconstrained DC resistivity result (Fig. 3.1 8a). Poorly estimated values in this result now appear to be related to the less sensitive, deeper model cells’ tendency toward higher reference model values of 0.001, a value higher than those cells at the same depth in the true models. The 123 effect is shallower cells have more underestimated values than previously, not requiring high conductivities since higher conductivities exist in the deeper cells. Changing the reference model is good practice as seeing where reference values take over at depth within the model allows for determination of maximum depth of investigation, which can be estimated as the depth where a range of models consistently revert to reference model values (Oldenburg and Li, 1999). Figure 3.18. Inversion results for the Hislop-like conductivity model after constraints applied, a) Inversion result with reference model set to 0.001 S/rn; b) Result with az and ay increased relative to ax; c) Result with reference model set to 0.001 S/rn, and alpha cLz and ay increased relative to ax. Alpha weightings Increasing alpha weightings in the z and y directions relative to the x direction to increase smoothing parallel to known structures and contacts causes conductivities 0003 0.00201 0,00202 0.00153 0.00103 0000542 50-005 S/rn onstrained with alpha weights and reference model 0.003 000255 000252 000553 0.001 03 0000542 50-005 S/rn 124 related to sheared ultramafic rocks to extend marginally deeper (Fig. 3.1 8b). The greater volume encompassed by the conductivity anomaly means the values are lower overall, in comparison to initial inversion results, to explain the observed data. Combined reference model and directional weighting Combining alpha weighting with appropriate reference value assignment yields a good, geologically reasonable result with the conductivity anomaly extended to depth as in the true model (Fig. 3.1 8c). Model difference values for this result are lower than for all previous DC resistivity inversion results. The use of the reference model value of 0.00 1 S/m keeps conductivity values high, unlike the application of alpha weightings alone. Again, constraining the model using prior physical property and geological knowledge does not improve imaging of the gold target, but yields a more accurate physical property model, which will in turn, lead to improved geological interpretations. 3.4.4. Other solutions for improving model results Experimentation with additional modifications to inversion and survey parameters were attempted to try to improve the model results where they have not been improved by constraints. Magnetic susceptibility model improvements — adjusting depth weightings One of the causes of discrepancy between true and recovered magnetic susceptibility models is the applied depth weightings. Depth weighting written into the potential fields inversion codes (Li and Oldenburg, 1996 and 1998) are necessary to offset the natural decay of the magnetic and gravity signal, and for distributing physical properties to depth. However depth weighting appears to lead to low sensitivities at the surface in susceptibility models, and subsequent overestimation of susceptibility at depth. To alleviate the problem of surface cells assuming reference model values, the default depth weighting was reduced by decreasing values of f3 and z0 (arbitrarily by a 125 quarter of their default values which were 3 and 20.92, respectively), the variables within the depth weighting function controlling the offset of the natural magnetic decay (Li and Oldenburg, 1996). This should allow more sensitivity within the upper cells in the model. The model results are improved over results with the default weighting used. Model difference values drop with a decrease in the weighting (Fig. 3.19). This change emphasizes that a significant portion of the disagreement between true and recovered models stems from the poor estimation of susceptibility near surface. It is recommended that for magnetic inversions at this scale, the depth weighting be manipulated for comparison to other unconstrained and constrained model results. For larger scale models, there might be problems associated with this manipulation of the default depth weighting, in terms of loss of information at depth. Figure 3.19. Inversion results for the Hislop-like susceptibility model with depth weightings reduced. The susceptibility model is better estimated in the near-surface cells, and susceptibility values are more accurate overall. Smoothing increases with depth. Resistivity model improvements - survey design DC resistivity and IP surveys completed using Realsection or Schlumberger arrays tend to have better depth detection than arrays with more closely spaced transmitters, but less spatial resolution overall due to the weakening of the current over 300 Magnetic result with reduced depth we 200’ 100 0 1005 10.0417 0.0333 0.025 0.0167 000833 SI Units 554400 -100W 553800 554100 126 the large distances covered. Hallof and Yamashita (1990) discuss the importance of using closely spaced electrodes for detection of small sulfide-rich zones. The use of a different array configuration where transmitter electrodes are not distant with respect to receiver electrodes might enhance spatial resolution of the conductivity anomalies in the Hislop model. A dipole-dipole array, compared to a Schiumberger array in Figure 3.20, with a = 40 m, and n = 1-10, was used to test this hypothesis. The result seems to be an improvement on the Realsection inversion result with better estimated conductivity values, and imaging of the upper parts of the conductive sulfide-rich zones (Fig 3.21). Although a full investigation of the effectiveness of different electrode arrays is beyond the scope of this research, this example shows that the chosen survey design can determine whether a feature will be detected in the geophysical data and in inversion results. The possibilities should be well-researched in advance of exploration with consideration of the types of information required and the characteristic sizes and depths of targets. Dipole - Dipole C2 Cl P1 P2 .E—a.( na Sch I urn berger Cl P1 P2 C2 •( na )•( a—>.#-——na )• Notes: Cl current source &ectrodes (transmitters) P1 potential electrodes (receivers) a = &ectrode separation; n = an integer Figure 3.20. Comparison of a dipole-dipole electrode configuration and a Schiumberger configuration which resembles a Realsection array. Dipole-dipole surveys employ closely spaced current and potential electrodes. For the Schiumberger array, current electrodes are distal to potential electrodes (figure modified from Inversion for Applied Geophysics resource package, UBC-GIF). 127 Figure. 3.21. DC resistivity inversion result for resistivity data collected via a dipole- dipole survey. Depth resolution has not increased compared to Realsection results, but there is better spatial resolution and the sulfide-rich zone is detected in addition to the sheared ultramafic unit. 3.5. CONCLUSIONS Synthetic modeling is important to conduct prior to inversion work. It will reveal whether or not a feature of particular shape, size, and of certain contrast with the host rocks can be resolved using inversion methods. In doing this, limitations of inversion are also revealed, establishing where caution in interpreting results is necessary (e.g. where physical properties recovered may be inaccurate, or if there are artifacts that are byproducts of the inversion), and determining when confidence can be placed in the interpretation or querying of recovered models. Synthetic modeling also tests the effects of constraints, and determines if it is possible to improve the model through their application. Synthetic modeling was completed in this study to determine whether geological features, specifically rocks related to gold mineralization and characteristic of Archean orogenic gold deposits, are detectable in the subsurface. From previous geological and DC Resistivity result from dipole-dipole survey 300. I C 003 0.00251 000202 000153 0.00103 0 000642 5e.005 S/rn 554400553800 554100 128 physical property studies it was determined that prospective features typical of this mineral deposit setting, which are also petrophysically distinct from host rocks include, near-vertical, or steeply-dipping faults, felsic dikes, carbonate alteration zones, and sulfide rich zones. These features are modeled within ultramafic and mafic volcanic rocks, common hosts to orogenic gold mineralization. The scale of the study is reminiscent of deposit-scale exploration, and inversions are done on a 1 km by 1 km by 600 m mesh. Synthetic magnetic, gravity, DC resistivity, and IP data were modeled. In general, this work shows there are significant enough contrasts between gold- related targets and host rocks in this environment for geophysics, and geophysical inversion, to be useful exploration tools. Magnetic and gravity inversions were successful in resolving low susceptibility and low density gold-related syenite dikes to depths around 200-3 50 m. The narrow, 20 m dike is most poorly imaged by magnetic inversion where hosted between a mafic and an ultramafic unit. Where there are two hosts of different susceptibility, yet both of higher susceptibility than the dike, the signature of the dike is lost in the overall gradient from low to high susceptibility areas. A combination of magnetic and gravity inversion results would best detect the syenite dike, with the density resolving the ultramafic-syenite contact, and magnetics resolving features slightly deeper. The main differences between recovered and true magnetic and gravity models result from smoothing due to the choice of inversion model norm, and due to the natural decay of the geophysical signal with depth. Depth weights also cause discrepancies between true and recovered models, which can bring about high estimates of susceptibility or density at depth, leading to poor distributions of physical properties throughout the model. Resistivity modeling using a Realsection electrode array does not detect narrow anomalous conductive zones related to sulfide-rich rocks, unless the zones are quite anomalous and laterally continuous, and in this case, their associated conductivity values are underestimated. DC Resistivity inversions are, however, effective at modeling larger conductive geological units, but only to shallow depths within the subsurface. Induced polarization inversions detect chargeable zones, especially where they are extensive and 129 continuous. However, as with conductivity models, the result is only reliable near surface, and anomalous values are underestimated. The lack of resolution at depth accompanying DC resistivity and IP survey methods cause the poor depth resolution in inversion results. The underestimation of anomalous physical property values is due to dispersal of anomalous values over larger areas as a result of smoothing brought on by the inversion model objective function. Constraining inversion results acts most importantly in limiting physical property ranges and producing a better distribution of physical properties throughout the model. While use of global constraints does not improve the resolution of targets that were not previously detected, they yield qualitatively and quantitatively better results. Applying constraints permits assessment of the range of possible results. Features that persist between various model results are required to exist to fit the observed data and satisfy the requirements of the model objective function, and are likely real. For all geophysical methods, different results would be expected at larger scales of exploration and modeling. At larger data spacings, there is greater depth of resolution, but inversion models would display less detail since larger model cells are required to keep inversion computation times to a minimum. Synthetic modeling, and subsequent inversion of true geophysical data, requires knowledge about relationships between physical properties and expected rock types. This is best achieved by having a reconnaissance knowledge of the geology being investigated. Using downhole susceptibility information, or surface sample data, it is recommended that typical physical property ranges be determined for the important rock types and altered equivalents. Sourcing published data is an option where data collection has not been carried out. Physical property studies in various geological settings are becoming more commonplace, and physical property data is being compiled currently on a more regular basis. This information is becoming increasingly valuable with the drive in recent years toward using geophysical methods to explore for deposits in the deeper subsurface. 130 REFERENCES Berger, B.R., 1999, Geological investigations along Highway 101, Hislop Township: Ontario Geological Survey, Summary of Field Work and Other Activities 1999, Open File Report, 6000, p. 5-1 — 5-8. Connell, S., Scromeda, N., Katsube, T.J., and Mwenifumbo, J., 2000, Electrical resistivity characteristics of mineralization and unmineralized rocks from the Giant and Con mine areas: Geological Survey of Canada, Yellowknife, NWT, Current Research 2000-E9. Cooper, M.A., 1948, Report on Kelore Mines Ltd., Ramore, Ontario, unpublished, 22 p. Darbyshire, D.P.F., Pitfield, P.E.J., and Campbell, S.D.G., 1996, Late Archean and Early Proterozoic gold-tungsten mineralization in the Zimbabwe Archean craton: Rb-Sr and Sm-Nd isotope constraints: Geology, v. 24, p. 19—22. DCIP3D user manual, Version 2.1, University of British Columbia Geophysical Inversion Facility. Dey, A., and Morrison, H.F., 1979, Resistivity modeling for arbitrarily shaped three dimensional structures: Geophysics, v. 44, p. 753-780 Doyle, H.A., 1990, Geophysical exploration for gold — a review: Geophysics, v. 55, p. 134-146. Farquharson, C.G, Ash, M.R., and Miller, H.G, 2008, Geologically constrained gravity inversion for the Voisey’s Bay ovoid deposit: The Leading Edge, v. 27, p. 64-69. Fyon, J.A., and Crockett, J.H., 1983, Gold exploration in the Timmins District using field and lithogeochemical characteristics of carbonate alteration zones, in Hodder, R.W., and 131 Petruk, W., eds, Geology of Canadian Gold Deposits, Canadian Institute of Mining and Metalurgy, Special Volume 24, p. 113-129. Groves, D.I., Ridley, J.R., Bloem, E.M.J., Gebre-Mariam, M., Hagemann, S.G., Hronsky, J.M.A., Knight, J.T., McNaughton, N.J., Ojala, J., Vielreicher, R.M., McCuaig, T.C., and Holyland, P.W., 1995, Lode gold deposits of the Yilgarn block: products of late Archean crustal-scale overpressured hydrothermal systems, in Coward, M.P., and Ries, A.C., eds., Early Precambrian Processes, Geological Society Special Publication, no. 95, p. 155-172. Groves, D.I., Goldfarb, R.J., Gebre-Mariam, M., Hagemann, S.G., and Robert, F., 1998, Orogenic gold deposits: a proposed classification in the context of their crustal distribution and relationship to other gold deposit types: Ore Geology Reviews, v. 13, p. 7-27. Hagemann, S.G., and Cassidy, K.F., 2000, Archaean orogenic lode gold deposits, in Hagemann, S.G., and Brown, P.E., eds., Gold in 2000, Society of Economic Geologists, Reviews in Economic Geology, v. 13, p. 9-68. Hallof, P.G., and Yamashita, M., 1990, The use of the IP method to locate gold-bearing sulfide mineralization, in Fink, J.B., Sternberg, B.K., McAlistar, E.O., Weiduwilt, W.G., and Ward, S.H., eds., Induced Polarization: applications and case histories, Society of Exploration Geophysicists, Tulsa, Ok., p. 227-279. Hodgson, C.J., 1989, The structure of shear-related, vein-type gold deposits: a review: Ore Geology Reviews, v. 4, 231-273 Hodgson, C.J., 1990, An overview of the geological characteristics of gold deposits in the Abitibi subprovince, in Ho, S.E., Robert, F., and Groves, DI, compilers, Gold and Base Metal Mineralization in the Abitibi Subprovince, Canada, with Emphasis on the Quebec Segment, Short Course Notes, Geology Department (key center) and University Extension, the University of Western Australia, publication No. 24, p. 63-100. 132 Hodgson, C.J., 1993, Mesothermal lode gold deposits, in Kirkham, R.V., Sinclair, W.D., Thorpe, R.I., and Duke, J.M., eds., Mineral Deposit Modeling, Geologic Association of Canada, Special Paper 40, p. 63 5-678. Hodgson, C.J., and Hamilton, J.V., 1990, Gold mineralization in the Abitibi greenstone belt: end-stage results of Archean collisional tectonics, in Keays, R.R., Ramsey, W.R.H., and Groves, D.I., eds., The Geology of Gold Deposits: the Perspective in 1988, Economic Geology Monograph 6, p. 86-100. Hodgson, C.J., and MacGeehan, P.J., 1982, Geological characteristics of gold deposits of the Superior Province of the Canadian Shield: Canadian Institute of Mining and Metallurgy, Special Volume 24, p. 211-228. Hodgson, C.J., and Troop, D.G., 1988, A new computer-aided methodology for area selection in gold exploration: a case study from the Abitibi greenstone belt: Economic Geology, v. 83, p. 952-977. Kent, A.J.R., Cassidy, K.F., and Fanning, CM., 1996, Archean gold mineralization synchronous with the final stages of cratonization, Yilgarn Craton, Western Australia: Geology, v. 24, p. 879-882. Kerrich, R., 1989, Geodynamic setting and hydraulic regimes: shear zone hosted mesothermal gold deposits, in Bursnall, J.T., ed., Mineralization and Shear Zones, Geological Association of Canada, Short Course Notes 6, p. 89-128. Kerrich, R., and Cassidy, K.F., 1994, Temporal relationships of lode gold mineralization and accretion, magmatism, metamorphism, and deformation — Archean to present: a review: Ore Geology Reviews, v. 9, p. 263-3 10. Kerrich, R., and Wyman, D., 1990, The geodynamic setting of mesothermal gold deposits: an association with accretionary tectonic regimes: Geology, v. 18, p. 882-885. 133 Kishida, A., and Kerrich, R., 1987, Hydrothermal alteration zoning and gold concentration at the Kerr-Addison Archean lode gold deposit, Kirkland Lake, Ontario: Economic Geology, v. 82, p. 649-690. Li, Y., and Oldenburg, D.W., 1996, 3-D inversion of magnetic data: Geophysics v. 61, p. 394-408. Li, Y., and Oldenburg, D.W., 1998, 3D inversion of gravity data: Geophysics, v. 63, p.109-19. Li, Y., and Oldenburg, D.W., 2000, 3D inversion of induced polarization data: Geophysics, v. 65,p.1931-45. McCuaig, T.C., Kerrich, R., 1998, P-T-t-deformation-fluid characteristics of lode gold deposits: evidence from alteration systematics: Ore Geology Reviews, v. 12, p. 381-453. Meuller, A.G., and Groves, D.I., 1991, The classification of Western Australian greenstone-hosted gold deposits according to wall-rock alteration mineral assemblages: Ore Geology Reviews, v. 6, p. 291-33 1. Oldenburg, D.W., and Li, Y., 1999, Estimating depth of investigation in dc resistivity and IP surveys: Geophysics, v. 64, p. 403-4 16. Oldenburg, D.W., Li, Y., and Ellis, R.G., 1997, Inversion of geophysical data over a copper gold porphyry deposit: A case history for Mt. Milligan: Geophysics, v. 62, p. 1419-143 1. Phillips, N.D., 2002, Geophysical inversion in an integrated exploration program: examples from the San Nicolas deposit: Unpublished M.Sc. thesis, University of British Columbia, 237 p. 134 Phillips, N., Hickey, K., Lelievre, P., Mitchinson, D., Oldenburg, D., Pizarro, N., Shekhtman, R., Sterritt, V., Tosdal, D., and Williams, N., 2007, Applied strategies for the 3D integration of exploration data: KEGS Inversion Symposium, PDAC 2007, extended abstract, 9 p. Power, W. L., Byrne, D., Worth, T., Wilson, P., Kirby, L., Gleeson, P., Stapleton, P., House, M., Robertson, S., Panizza, N., Holden, D. J., Cameron, G., Stuart, R., and Archibald, N. J., 2004, Geoinformatics evaluation of the eastward extension of the Timmins Gold Camp: Geoinformatics Exploration Inc., Unpublished report for St Andrew Goldfields Ltd. Prest, V.K., 1956, Geology of the Hislop Township: Ontario Department of Mines, Annual Report, 1956, v. 65, pt. 5, 51 p. Robert, F., 1990, Structural setting and control of gold-quartz veins of the Val D’Or area, southeastern Abitibi subprovince, in Ho, S.E., Robert, F., and Groves, D.I., compilers, Gold and Base Metal Mineralization in the Abitibi Subprovince, Canada, with Emphasis on the Quebec Segment, Short Course Notes, Geology Department (key center) and University Extension, the University of Western Australia, publication No. 24, p. 167- 212. Robert, F., 2001, Syenite-associated disseminated gold deposits of the Abitibi greenstone belt, Canada: Mineralium Deposita, v. 36, p. 503-5 16. Roberts, R.G., 1988, Archean lode gold deposits, in Roberts, R.G., and Sheahan, P.A., eds., Ore Deposit Models, Geoscience Canada, Reprint Series 3, p. 1-19. Roscoe and Postle, 1998, Hislop Mine Property, Roscoe and Postle Associates Inc., St. Andrew Goldfields Ltd. internal report, unpublished, p. 66-89. 135 Sibson, R.H., Robert, F., and Poulsen, K.H., 1988, High-angle reverse faults, fluid pressure cycling, and mesothermal gold-quartz deposits: Geology, v. 16, p. 55 1-555. Telford, W.M., Geldart, L.P., and Sheriff, R.E., 1990, Applied Geophysics, Second Edition, Cambridge University Press, 770 p. Weinberg, R.F., Hodkiewicz, P.F., and Groves, D.I., 2004, What controls gold distribution in Archean terranes: Geology, v. 32, p. 545-548. Williams, N.C., 2006, Applying UBC-GIF potential fields inversions in greenfields or brownfields exploration: Australian Earth Sciences Convention, 2006, Melbourne, Australia, 10 p. 136 Chapter 4: 3D inversion of magnetic, gravity, DC resistivity, and induced polarization data over the Hislop gold deposit, south-central Abitibi greenstone belt3 4.1. INTRODUCTION 4.1.1. Rationale Magnetic, gravity, DC resistivity, and induced polarization (IP) data were inverted to investigate subsurface geology within a section of the south-central Abitibi greenstone belt hosting the Hislop gold deposit. A large amount of historic drilling has been completed, but much of it is shallow, and concentrated on specific mineral exploration properties. The irregular drilling coverage, and an overall lack of outcrop in the Hislop deposit area suggests that geophysical inversion, the calculation of subsurface distributions of physical properties from geophysical data, could be an extremely useful tool for understanding subsurface geology and establishing mineral exploration targets in this part of the gold-rich Abitibi greenstone belt. In contrast to its more extensive application for delineation of massive sulfide- style deposits (Oldenburg et al., 1997; Phillips, 2002; Farquharson et al., 2008), geophysical inversion has not been as commonly applied for similar purposes in the orogenic gold environment. The reason for this is that orogenic gold deposits, like the Hislop gold deposit, are characterized by small, discontinuous, and low grade orebodies that do not have a strong petrophysical contrast with typical host rocks. Geological units, hydrothermal alteration zones, and structures that are known to be related to gold, however, can provide larger scale exploration targets. There are only a few examples of case studies employing inversion methods to detect gold-related rocks in Archean orogenic gold settings (UBC-GIF Inversion for Applied Geophysics CD-ROM, 2000- A version of this chapter will be submitted for publication. Mitchinson, D., Phillips, N., and Williams, N., 2009, 3D inversion of magnetic, gravity, DC resistivity, and induced polarization data over the Hislop gold deposit, south-central Abitibi greenstone belt. 137 2006; Kowalczyk et al., 2002; Mira Geoscience Ltd., 2005a and 2005b; Mueller et al., 2006). The usefulness of these methods as an exploration tool in this mineral deposit environment may therefore not yet be fully appreciated. The large amount of geophysical data available, and a thorough background understanding of the relationships between geology and physical properties for the Hislop deposit area (Chapter 2), creates an opportunity to conduct a comprehensive study of the types of information that might be acquired by inverting a suite of geophysical data at a range of scales in this mineral deposit setting. 4.1.2. Geological background The area investigated in this study is located in the south-central Abitibi greenstone belt (Fig. 4.1), east of the Timmins-Porcupine gold camp, which is known for it’s world-class Archean orogenic gold deposits (Hollinger-Mclntyre and Dome deposits), and in general for its large number of gold and base metal deposits and occurrences. The study area (Fig. 4.2) is underlain by northwest-southeast trending ultramafic to mafic volcanic rock sequences, with lesser felsic volcanic units (Prest, 1957; Berger, 1999; Power et al., 2004; Roscoe and Postle, 1998). The volcanic sequences are intruded by variably sized, fine to coarse-grained felsic and intermediate intrusives and dikes. A major crustal-scale fault zone, the Porcupine-Destor Deformation Zone (PDDZ), runs northwest-southeast through the area, parallel to the general regional trend. Gold deposits in this part of the Abitibi greenstone belt have a close spatial relationship with the PDDZ (Jackson and Fyon, 1991; Berger, 2002). It is interpreted to have acted as a conduit through which C02-rich and gold-bearing fluids ascended upward through the crust to sites of eventual gold deposition (Kerrich, 1989). Sedimentary rocks from the Porcupine and Timiskaming assemblages, lie north of the PDDZ, likely having originally accumulated in a structurally controlled basin during fault development late in the formation of the greenstone belt (Ayer et al., 2002). 138 Local and deposit scale geophysical inversions completed for this study focus on the Hislop gold deposit. The Hislop deposit is a structurally controlled Archean orogenic gold deposit. Gold occurs with disseminated pyrite and is distributed within host rocks in proximity to a fault that occurs between a coarse-grained syenite dike, and a metamorphosed ultramafic volcanic unit (Fig. 4.3). Lesser mineralization occurs within small veins and vein stockworks in magnetite-bearing Fe-rich tholeiitic basalts north of the syenite dike. Gold is accompanied by Fe-rich carbonate, and muscovite alteration (Prest, 1956; Roscoe and Postle, 1998; Berger, 2002). The Hislop deposit area was explored by numerous exploration groups from the early 1900’s onward, and there are many existing driliholes and associated logs providing geological information for this property. The Hislop deposit was mined during three separate efforts between 1990 to 2006, by St. Andrew Goldfields, Ltd., from underground workings (Shaft Area on map in Fig. 4.2) and a small open pit (West Area). Over 400,000 ..- Major faults Proterozoic rocks Archean sedimentary r: ] Archean granitoid rocks • rchean volcanic rocks Archean mafic intrusive rocks Figure 4.1. Geological map of the southwest Abitibi greenstone belt (modified after Poulsen et al., 2000). The Hislop deposit study area is shown with respect to the Timmins-Porcupine gold camp. White circles represent gold deposits and black circles represent world class gold deposits (>100 t). PDDZ = Porcupine Destor Deformation Zone, LLCDZ = Larder Lake—Cadillac Deformation Zone. 139 tonnes of ore was mined over this period, grading between 2.33 g/t and 5.55 g/t (www.standrewgoldfields.com). The mine is currently not in operation. SSG - greywacke S00 — sediment, undivided IFD!IFO — felsic intrusive dyke! felsic intrusive undivided 100 — intrusive, undivided Figure 4.2. Geology of the Hislop deposit area as interpreted by Power et a!. (2004) from high resolution aeromagnetics, and previous mapping in the Abitibi greenstone belt. Locations of two mined areas on the Hislop property (West Area open pit; Shaft Area underground) are outlined in red. Also shown are 10 drill holes (one overlapping) logged for this study. The cross-section shown in Figure 4.3 is based on core logging of four drill holes that were drilled in the West Area. 4.1.3. Relationships between geophysics, physical properties, and geology Although gold itself is a dense and conductive mineral, its characteristic low grades in orogenic gold deposits make its direct detection using geophysical methods _____ LDO — late diorite/dolerite I I SLO — mudstone - siltstone ISO — syenite intrusive, undivided VFO — felsic volcanic, rhyolite, rhyodacite VUO — ultramafic volcanic, undivided > VMF — magnetic mafic volcanic VMO — mafic volcanic, basalt, andesite 140 difficult (Seigel et al., 1984; Doyle, 1990). Known vectors to gold such as hosting structures, lithology, or related hydrothermal alteration mineral assemblages, however, may still be targeted remotely. Rock property studies on the Hislop deposit revealed the existence of consistent relationships between certain physical properties, and potentially mineralized rocks (Chapter 2). A summary of the results of these physical property studies are presented in the following sections. Table 4.1 summarizes the physical property ranges for each rock type, and indicates anomalous ranges unique to some of the prospective rocks in the Hislop area. Figure 4.3. Cross section looking Northwest through the Hislop deposit, showing locations of carbonate-dominated alteration and gold mineralization. Cross section interpreted from drill core logged from the West Area of the Hislop property (see Figure 4.2). DDH H9601 DDH Ext 280, GK 280, and H9605 Multi-lithic Volcanic Breccia Lamprophyric Dike LII Intermediate Dike Porphyritic Rhyolite Dike Syenite Intrusive Mafic Volcanic Rock Ultramafic Volcanic Rock Fault . Drill trace 0 141 Table 4.1. Typical and anomalous physical property ranges for principal rock types occurring in the Hislop deposit area. Unaltered ultramafic (dolomite-chlorite assemblage) Unaltered ultramafic (talc- chlorite assemblage) Fe-carbonate-muscovite altered ultramafic Magnesite-fuchsite altered ultramafic rocks) 111-8546 Range (all ultramafic rocks) 2.90-15.967 Unaltered mafic Fe-carbonate-muscovite altered mafic Fe-carbonate-albite altered mafic 0.35-141 2.70-3.08 0-2.19 2.78-2.97 0.28-1.27 2.76-2.86 (all mafic rocks) 541-58754 (all mafic rocks) 2.07-46.5 Carbonate- 0.2-2.19 2.76-2.97 altered mafic Unaltered intermediate intrusive Carbonate altered intermediate intrusive Carbonate-muscovite altered intermediate intrusive 0.24-135.29 2.72-2.95 0.13-3.8 2.67-2.95 0.32-1.55 2.76-2.94 (all intermediate rocks) 2314-22613 (all intermediate rocks) 9.45-15.4 Carbonate- altered intermediate intrusive 0.13-3.8 2.67-2.95 L’J Syenite Rhyolite porphyry 0.07-0.42 2.64-2.74 0.05-0.41 2.57-2.80 (all syenites) 2631-9400 (all rhyolite dikes) 8976-11525 (all syenites) 6-15.67 Felsic (all rhyolite dikes) intrusives 2-20.4 0.05-0.42 2.57-2.80 Mag. Sus. Density Resistivity Chargeability Cut-off values for querying data Rock Type (x103 SI) (glcm3) (Ohm-rn) (ms) Range Range Range I Res. I Chg. 0.57-12.5 2.82-2.89 (all ultramafic 0.44-84.4 2.79-2.94 0.41-5.96 2.80-2.91 0.49-0.95 2.85-2.96 Iock Type I Mae. Sus. I Density Ultramafic rocks 111-85 46 Carbonate- altered 0.41-5.96 2.80-2.96 ultramafic Sulfide-rich rocks Based on anomalies in inversion results Anomalous sulfides <1540 >120 Magnetic susceptibility Syenite and porphyritic rhyolite dikes have low susceptibility ranges distinct from most intermediate to ultramafic rocks at Hislop (Fig. 4.4). Their susceptibility values range from 0.05 — 0.42 x103 SI Units. Mafic and ultramafic volcanic rocks, and intermediate intrusive rocks have bimodal susceptibility populations (Fig. 4.4). This distribution indicates there are two distinct populations that make up the data. The high susceptibility population (> -10 x103 SI Units) is predominantly composed of least-altered Fe-rich tholeiitic basalts and ultramafic volcanic rocks (mainly talc-chlorite schists). Physical property studies revealed that the lower susceptibility population is partly composed of carbonate-altered intermediate, mafic and ultramafic rocks. These altered rocks (Figs. 4.5a and 4.5b) have susceptibility ranges from 0.13 — 5.96 x103 SI Units. Thus, for exploration targeting purposes, any rocks with susceptibilities above 5-10 x103 SI Units, where the break in data in Figures 4.4 and 4.5 occurs, can be excluded as less prospective. 3.20 3.10 3.00 2.90 ___________ 2.80 2.70 2.60 2.50 2.40 — 0.01 0.1 10 100 1000 Rock Types 9 Ultramafic rocks Mafic rocks X Intermediate dikes A Syenite intrusives 0 Rhyolite porphyriesEC.) 4-, C’, a) ø>bx 1 Magnetic Susceptibility (xl0 SI Units) Figure 4.4. Magnetic susceptibility plotted against density for the major rock types at Hislop. Syenite intrusives and porphyritic rhyolite dikes have distinctly low density and magnetic susceptibility ranges allowing them to be distinguished from intermediate, mafic, and ultramafic rocks at Hislop. 143 3.20 3.10 - 3.00 2.90 2.80 2.70 C) 2.60 2.50 2.40 1 10 100 Magnetic Susceptibility (x1O3 SI Units) b. 10000.1 3.20 3.10 3.00 2.90 2.80 2.70 2.60 2.50 2.40 I mfc.LIf1; *“!‘Unaltered to AlteredUltramafic Rocks DLst. altd. (Dol-chi) •Tlc+chl ultramafic QFeCb+ms altd. utiramafic DFe/MgCb+fu altd. ultrarnafic 0.1 1 10 100 1000 Magnetic Susceptibility (xl O SI Units) Figure 4.5. Magnetic susceptibility plotted against density for a) mafic and b) ultramafic volcanic rocks from the Hislop deposit area. Carbonate-alteration destroys magnetite in mafic and ultramafic volcanic rocks, causing magnetic susceptibility to drop. Density values increase slightly for altered ultramafic rocks. Abbreviations in legends: Lst. altd. = least altered assemblage; Chl+ab = chlorite+albite assemblage; FeCb+ms = Fe carbonate+muscovite; FeCb+ab = Fe-carbonate+albite; Dol+chl dolomite+chlorite assemblage; Tlc+chl = talc+chlorite; Fe/MgCb+fu = Fe/Mg-carbonate+fuchsite (chrome muscovite). 144 Low susceptibilities, however, do not uniquely identify felsic intrusive rocks and carbonate-altered samples. Fe-poor tholeiitic basalts (not differentiated from Fe-rich basalts on plots) have low susceptibilities that overlap with the susceptibility range of prospective carbonate altered rocks. This means that targeting low susceptibility areas will not exclusively target prospective rocks, and if possible, other criteria should be used to further discriminate the different rocks types that exist within the low susceptibility range. Density Density studies indicate that syenite and porphyritic rhyolite dikes have low densities compared to other rock types in and around the Hislop deposit area with ranges from 2.57-2.80 g/cm3 (Fig. 4.4, and Tab. 4.1). All other rock types and their altered equivalents have higher density ranges generally greater than 2.75 g/cm3. Density data may thus be used to further distinguish between low susceptibility felsic intrusive rocks, and low-susceptibility carbonate-altered rocks or Fe-poor tholeiitic basalts, where felsic rocks would have low susceptibilities and low densities, and carbonate-altered rocks and Fe-poor tholeiites would have low susceptibilities and higher densities. Although density ranges for least-altered and altered mafic and ultramafic rocks generally overlap, a trend of increasing density in altered ultramafic rocks with carbonate alteration was indicated (Fig. 4.5b, and Chapter 2). This suggests that where ultramafic rocks are knowi to dominate within an area, it may be possible to identify carbonate- altered rocks using density in addition to susceptibility. Resistivity and chargeability Resistivity values measured in the lab are not consistently representative of larger scale measurements since there can be large scale features in the rock controlling resistivity that are not present at the hand sample scale (www.zonge.com!LabIP.html). For interpreting resistivity data and relationships to geology, sample measurements are best compared to one another on a relative scale. From Hislop physical property studies, resistivity was determined to be partly controlled by rock texture, specifically porosity and schistosity. Low resistivity (or high conductivity) values associated with sheared and 145 porous ultramafic volcanic rocks may distinguish them from other Hislop rock types, which otherwise have higher, overlapping ranges of resistivity (Fig. 4.6). A pattern of increasing resistivity with carbonate-alteration occurs in ultramafic rocks. The increased resistivity ranges related to carbonate-altered ultramafic rocks, however, begin to overlap with the resistivity ranges for other rock types. 4 3 2 6 4 2 8 6 Although most sulfides are known to be conductive (Telford et al., 1990), there were no significant correlations observed between pyrite abundances derived from XRI) (Rietveld) analyses and resistivity during physical property work (Chapter 2, Appendix 2G). Compared to resistivity measurements, chargeability measurements made in the lab are less inconsistent with larger scale measurements, and can thus be trusted to better represent in-situ chargeability. Chargeability values do not distinguish between different Ultramafic volcanic rocks6 4 2 4 2 Mafic volcanic rocks Intermediate dikes lear! 2816 ___ I ___ Mean 28432 n 9759 Porphyritic rhyolite dikes Mean 11534 10 100 1000 10000 100000 1000000 Resistivity (Ohm-rn) Figure 4.6. Resistivity histograms for Hislop deposit rocks. Data indicates lower overall resistivities for ultramafic volcanic rocks from Hislop. 146 rock types at Hislop as chargeability ranges essentially overlap for the suite of samples (Fig. 4.7). Although it is widely known that chargeability is strongly controlled by the presence of sulfide minerals (Telford et al., 1990), physical property studies at Hislop indicate only a weak trend between pyrite abundance and chargeability, and only for felsic rocks (Fig. 4.8). Chargeability studies at Hislop also suggest that porosity may decrease chargeabilities, complicating relationships between this physical property and mafic volcanic rocks (Fig. 4.9, Chapter 2). Despite the lack of direct correlation between sulfides and higher chargeabilities for Hislop drillhole and surface samples, induced polarization has been used successfully to target sulfides in previous exploration efforts in similar geological settings (Johnson et al., 1989; Bate et al., 1989; Hallof and Yamashita, 1990). 4.1.4. Inversion background Geophysical inversion methodology is regularly used throughout industry, government, and academia, to investigate the Earth’s subsurface geology and explore for mineral deposits (Oldenburg et al., 1998). Geophysical inversion can be considered the opposite of geophysical forward modeling processes. Whereas forward modeling involves calculation of a geophysical response from a known, or hypothetical subsurface physical property model, geophysical inversion involves a calculation of the subsurface arrangement of physical properties, based on surface measurements, that is capable of causing an observed dataset. 147 a- Cu 0 10 8 6 4 2 6 4 2 4 2 10.00 Chargeability (ms) Figure 4.8. Chargeability plotted against pyrite abundance for Hislop samples. A weak positive correlation exists between pyrite abundance and chargeability, however the trend is mainly controlled by porphyritic rhyolite dike and syenite samples. There is no evidence of a consistent relationship between chargeability and pyrite abundance for intermediate to ultramafic volcanic rocks. Ultramafic volcanic rocks [n 6.9LL - Mafic volcanic rocks :Mean 31.32r Intermediate dikes - rrl ean4 6 - Syenite intrusives - 6 - Porphyritic rhyolite dikes_ 2 _______ Mean 10.47 1 10 100 1000Chargeability (milliseconds) Figure 4.7. Chargeability histograms for Hislop deposit rocks. Chargeability ranges for the individual rock types overlap and are not unique. 10 1 0 Rock Types 0 Ultramaflc rocks •Mafic rocks x Intermediate dikes A ASyenite intrusives o Rhyolite porphyries A A A 0 0 0 1.00 100.00 148 10 Mafic rocks (j)l 0 0 0 0 0 I 1.00 10.00 100.00 1000.00 Chargeability (ms) Figure 4.9. Chargeability versus porosity for mafic rock samples from Hislop. A negative correlation between chargeability and porosity in this plot indicates that increases in porosities of mafic volcanic rocks at Hislop may hinder the ability for metallic minerals to become charged. One of the limitations of inverting geophysical data is that the solution is non unique. Due to the fact that there are a greater number of unknowns (i.e. cells in the discretized model volume), than there are data, the problem is underdetermined. There are many distributions of physical properties that can cause the same observed data set. To alleviate this non-uniqueness, a model objective function, or model ‘goal’, is defined, so that the model outcome is consistent with expected geology. Additionally, a specific misfit must also be achieved between observed data and predicted data calculated from the recovered model. The inversion process is an iterative process. The model will be re computed numerous times in an attempt to minimize differences between the predicted and observed data sets, and to satisfy the terms of the model objective function. Detailed inversion calculations can be found in Li and Oldenburg’s (1996, and 1998) papers on 3D gravity and magnetic inversions. Where geology is better understood, and/or where physical property data is available (published data, downhole data, drill core, or outcrop measurements) inversions can be more thoroughly constrained. Physical property data, or reference models are incorporated into the descretized volume of interest. The model then has to be estimated 149 such that the incorporated data is honored. Physical property bounds can be specified to limit the range of values that are allowed to be taken up by the model cells. Finally, smoothing of physical properties in the x, y, or z directions using weightings written into the inversion algorithm (alpha weightings, cii, c’ and ct) can invoke geological directionality. Constraining inversions should result in more accurate models, and better estimated physical properties throughout the model (Phillips, 2002; and Williams, 2006; also Chapter 3). It also further reduces non-uniqueness. Generating multiple inversion models with varying constraints will result in improved interpretations of the models - consistently occurring features between model results can be assigned higher confidence. 4.2. INVERSION APPROACH 4.2.1. General strategy Unconstrained inversions of airborne magnetic data, airborne gravity data, and DC resistivity and IP data, were completed over the Hislop deposit area. As there exists a significant amount of magnetic susceptibility data for Hislop deposit area rocks, and there are well-established relationships between magnetic susceptibility and geology, magnetic inversions are also inverted with constraints incorporated via reference models built using William’s (2008) GiFtools ModelBuilder software. Constraining data include downhole susceptibility measurements, and surface sample susceptibility measurements. Additionally, cells within the model mesh where the physical property values can be estimated based on interpreted geology are also constrained. Further details on ModelBuilder applications are given in section 2.5. Both unconstrained and constrained magnetic inversion results are presented herein. Gravity, DC resistivity, and IP inversions are constrained using only ‘non-located’ constraints, as described by Phillips et al., (1997), which are applied globally to the model. Non-located constraints, such as global reference models, and bounds on physical property ranges, were used successfully to improve inversion results in synthetic modeling studies (Chapter 3). Only constrained gravity, DC 150 resistivity, and IP inversion results are presented here, although all results are included in Appendix 4A. Inversion results are interpreted with respect to mapped and interpreted surface geology, and geology logged from drill core. Recovered models are interrogated through querying based on relationships between geology (lithology, alteration, mineralization), and physical properties (magnetic susceptibility, density, resistivity, and chargeability), identified during physical property studies on the Hislop gold deposit (Chapter 2). 4.2.2. Magnetic inversions A high resolution airborne magnetic survey completed in 2002 covers an area of roughly 58 km by 20 km in the eastern Abitibi greenstone belt. Lines were flown north- south, at spacings of 50 m, and data was collected along lines at 7-10 m intervals. The magnetic data extents are shown in Figure 4.10, and the magnetic data are given in Figures 4.11 to 4.13. The datasets to be used in the inversion must include estimated standard deviations. For the Hislop magnetic datasets, the assigned standard deviation was 2-5%, and a floor value of 2-5% of the maximum measured field strength (in nT) was added, such that very low values do not have unrealistically low errors. All survey parameter details are compiled in Table 4.2. Magnetic inversions are completed at three scales, referred to in this study as ‘regional scale’ (20 km x 18 km), ‘local scale’ (6 km x 4 1cm), and ‘deposit scale’ (2 km x 1.5 km). Cells making up the core volume of interest in the regional, local, and deposit scale models are 200 m2, 50 m2, and 25 m2, respectively. Inversion parameters are detailed in Table 4.3. Local and deposit scale inversions are completed from magnetic data that has had larger scale magnetic signatures removed using regional removal methods described by Li and Oldenburg (1998). There is not sufficient data coverage to remove any larger scale geophysical signatures for the Hislop 20 km x 18 km regional scale inversions. Surface data extents and inversion volumes (Tabs. 4.2 and 4.3), were chosen based on maximum coverage required to explain any subsurface features that might occur within the core volume of interest. Padding cells were added along the perimeters of the inversion 151 Figure 4.10. Extents of magnetic data used in the deposit-, local-, and regional-scale magnetic inversions (red outlines), of gravity data (blue outline) used in the regional-scale gravity inversion, and of DC Resistivity and IP data used in corresponding deposit and local scale inversions (yellow outline). Geological map from Power et al., 2004. See Figure 4.2 for geology legend. IRegional Scale F I 71Z [l E L 16.5 km 152 I 4961 3928 5371507_ C t 567753_ 5356493 -1238 nT 547158 550678 554198 557719 561239 564759 568280 Easting (m) Figure 4.11. Data used in regional-scale magnetic inversion. Local-scale magnetic dataset outlined. Refer to Figure 4.10 for corresponding geology. 5382767_ Observed Magnetic Data 12355data, 1= 75, D-12 5379014_ 5375260_ 5364000 - 5360246 - 2895 - 1862 828.5 -204.6 153 Figure 4.12. Data used in local-scale magnetic inversion. Deposit-scale magnetic dataset outlined. Refer to Figure 4.10 for corresponding geology. Figure 4.13. Data used in deposit-scale magnetic inversion. Refer to Figure 4.10 for corresponding geology. Observed Magnetic Data 27976data, 1=75, D=-12 5376170 5806 4634 5374274 - 3461 572378 - 2288 570482 - 1115 5368586 -57.7 5366691_ - -1231 I nT 547067 I I I 548932 550798 552663 554528 556393 558259 Eastinq Cm) Observed Magnetic Data 92O9data, 175, D-12 5373140_ 2708 - 2126 5372440 1543 571740 2 0) 571040 - 378.4 5370340 - -204 5369640 -786.5 I I I I I nT 550665 551325 551984 552644 553303 553963 554622 Easting (m) 154 Table 4.2. Survey parameters. Magnetics x: 547067 - 568259 (local) y: 5366691 - 5376170 Gravity x: 545003 - 561995 y: 5365060 - 5377899 x: 550100 - 555300 y: 5370100 - 5372400 IP (local) x: 550100 - 555300 y: 5370100 - 5372400 IP (deposit) x: 551100-554300 y: 5370400 - 5372400 N-S margins margins draped 27982 2%; (lOOm) (100) 50m floor centre centre l4OnT (—50 m) (—50) N-S 50 m 25 m draped 9209 2%; 50 m floor 5OnT E-W 500 m 250 m constant 1850 0.01 468 m mGal floor SW-NE lOOm 20m ground 6545 10% max. Voltage SW-NE lOOm 20m ground 4576 10% max. Voltage SW-NE 100 m 20 m ground 6545 10% max. Voltage SW-NE 100 m 20 m ground 4576 10% max. Voltage lnclination:75°; Declination:-12°; Strength: 57478 nT 2002 lnclination:75°; Declination:-12°; Strength: 57478 nT 2002 lnclination:75°; Declination:-12°; Strength: 57478 nT 2003 1996 Realsection survey - 5 Tx spacings: 1000 m, 1500 m, 2000 m, 2400 m, 3200 m (26 lines) 1996 Realsection survey - 5Tx spacings: 1000 m, 1500 m, 2000 m, 2400 m, 3200 m (20 lines) 1996 Realsection survey - 5Tx spacings: 1000 m, 1500 m, 2000 m, 2400 m, 3200 m (26 lines) 1996 Realsection survey - 5Tx spacings: 1000 m, 1500 m, 2000 m, 2400 m, 3200 m (20 lines) Model Data area (UTM) Lines Line Station Height # Data Data Year Other information spacing Spacing errors Magnetics x: 547158 - 568280 N-S 200 m 200 m draped 12361 5%; 2002 (regional) y: 5356493 - 5382767 50 m 300nT Magnetics (deposit) floor x: 550665 - 554622 y: 5369640 - 5373140 DC Resistivity (local) DC Resistivity (deposit) x: 551100-554300 y: 5370400 - 5372400 Table 4.3. Inversion parameters. Inversion # Data 12361 Inversion core extents (UTM) x: 541700 - 563500 y: 5381350 - 5361550 z: 500 - (-)4700 Magnetics 27982 x: 550050 - 555150 (local) y: 5369780 - 5373080 z: 450 - (-)2150 Magnetics 9209 x:551630-553630 (deposit) y: 5370640 - 5372140 z: 450 - (-)550 Gravity 1850 x: 547500 - 559500 y: 5367400 - 5375400 z: 500 - (-)1500 IP (local) 6545 x: 550700 - 554700 y: 5370100 - 5372400 z: 400 - (-)800 IP (deposit) 4576 x: 551500 - 553900 y: 5370400 - 5372400 z: 400 - (-)200 Topography used in all models unconstrained model: default alphas; constrained model: using reference model built in GIFtools (Tab. 4.4); L=200 9013/ unconstrained model: default a values; 9142 constrained model: using reference model built in GlFtools (Tab. 4.4); L100 1892/ unconstrained: L=500, L=400 1856 constrained: L750, L=600 6529/ Near-surface cell weightings applied; 6455 unconstrained: default alphas; constrained: L200; L=100; reference value 0.00015 S/rn 4670/ Near-surface cell weightings applied; 4486 unconstrained: default alphas; constrained: L=1 00; L=50; reference value = 0.00015 S/rn 6369/ Near-surface cell weightings applied; 6445 unconstrained: default alphas; constrained: L200; L= 100; reference value = 0.031 ms 4412/ Near-surface cell weightings applied; 4519 unconstrained: default alphas; constrained: L1 00; L=50; reference value = 0.031 rns Magnetics (regional) # Core Core cell # Padding Achieved Other cells size cells misfit 280566 200rn3 178101 12321 unconstrained:a=0.0001,a=16; 165312 centre 50 rn3 margins 100 m3 116928 29190 (unconstr.) I 28452 (constr.) DC Resistivity (local) DC Resistivity (deposit) 6545 x: 550700 - 554700 y: 5370100 - 5372400 z: 400 - (-)800 4576 x: 551500 - 553900 y: 5370400 - 5372400 z: 400 - (-)200 192000 25m3 236064 15360 250 rn3 56460 (xy); 200 rn2 (z) 51520 50m3 159680 184320 25rn3 249856 51520 50rn3 159680 184320 25m3 249856 cJ Relationship of a (alpha weight) to L (length scale): (L)2 = (a/a)2;similar for L, L volumes to a distance that is required to explain any features that might occur in the dataset but not directly within the volume of interest. Topographical information was used in all inversions. Topography data was downloaded from the Shuttle Radar Topography Mission (SRTM) online database. Data was collected at approximately 90 m spacings. Located constraints applied to magnetic inversions are discussed in Section 2.5. 4.2.3. Gravity inversions Airborne gravity surveys over the northeastern, northwestern, and southern Timmins areas were completed in 2003 as part of the Discover Abitibi Initiative (Ontario Geological Survey, 2004). For the eastern Timmins survey (Fig. 4.10 and Fig. 4.14), lines are east-west trending, and 500 m apart. Data spacing is 120 m (Tab. 4.2). Two north- south trending tie lines occur 5 km apart. The sparse data spacing meant that only a regional scale inversion could be performed with cell sizes of 250 m to correspond with an intermediate spacing between the 500 m lines and the 120 m data spacing (Tab. 4.3). Regional removal was not performed on the gravity data, as the gravity dataset does not extend far enough beyond the chosen regional scale area to effectively remove a regional signature. A good correlation between gravity data and mapped geology indicates that a larger regional trend does not contribute strongly to the dataset, and the lack of a regional removal should not be detrimental to the inversion result. Gravity data was assigned standard deviations of 0.01 mGal for Hislop gravity inversion work. Regional gravity inversions were constrained with non-located constraints. Bounds were used to restrict densities in the inversion result to within the range expected for the rocks in the study area, and inversion smoothing weightings were increased. 157 Figure 4.14. Data used in regional-scale gravity inversion. Refer to Figure 4.10 for corresponding geology. 4.2.4. DC resistivity and IP inversions A combined DC Resistivity and induced polarization (IP) survey was completed in 1996 (Roscoe and Postle, 1998). Thirty-nine lines of Realsection data were collected (Fig. 4.15). Lines were spaced 100 m apart, transmitter electrodes were spaced from 1000 m up to 3200 m along lines, and receiver electrode spacing was 20 m. Measurements were made in time domain for both DC resistivity and IP surveys. The required data format for inversion of DC resistivity data is the potential in Volts normalized by the current (DCIP3D User Manual, version 2.1). 5377899 Observed Gravity Data 1850 data 5375331 - 15.53 p72763 - C 7O195_ 111.35 - 7.176 5367627_ 2.999 5365060_ -1.177 I I I 545003 547835 550667 553499 556331 Easting (m) -5.354 559163 561995 -9.531 mGaI 158 Figure 4.15. Location of DC resistivity and IP lines used for 3D DC resistivity and IP inversions. Local mine grid line numbers shown. See Figure 4.2 for geology legend. Induced polarization effects are caused by the build up of charge at physical interfaces within a medium. Chargeability is measured by assessing the decay of voltage over time when the electrical current is shut off (Telford et al., 1990). Data over the Hislop deposit was collected over 10 different time windows. Measured data is in mV/V and a total apparent chargeability is calculated for this work as the sum of the voltages recorded for time windows 2 to 8, multiplied by 0.8. The value is divided by 1000 to get data into the form V/V, the correct units for IP inversion calculations. Standard deviations on DC resistivity data and IP data are assigned at 10%, with a floor of 10% of the maximum voltage to avoid small errors on low data values (Tab. 4.2). When DC resistivity and IP data are displayed as pseudosections, the depth is usually arbitrarily assigned for visualization purposes based on n-spacings, or the distance between the transmitters in this case (Telford et al., 1990). Pseudosections are simply a method of displaying the data, and the z-scale does not represent depth. The positions and shapes of features also do not likely reflect the true geology. Inverting DC I E 0 0 CD C) 7100 m 159 resistivity and IP data is thus very useful as it can indicate the correct location of features, and can resolve the true shape of features within the subsurface. 2D DC resistivity and IP inversion models were initially completed separately for each of the 39 survey lines prior to running 3D inversions. This helped to determine appropriate errors for 3D inversions, and to examine depth of investigation (Oldenburg and Li, 1999). 2D results were also compared to cross-sections through 3D inversion results, and consistencies between the two indicated robust, consistent modeling (Appendix 4B). A 4 km x 2.3 km area, and a 2.5 km x 2 km area immediately surrounding the Hislop deposit were focused on for the 3D DC resistivity and IP inversions. The core volume for the ‘local’ scale inversions was discretized into 50 m2 cells, and the core volume for the ‘deposit’ scale models was discretized into 25 m2 cells (Tab. 4.3). Non-located constraints were used to refine local scale and deposit-scale DC resistivity and IP inversions. A global reference model of 0.00015 S/m was applied to DC resistivity inversions, and a reference model of 0.031 ms was applied to IP inversions. This acts to guide results toward reasonable values consistent with prior physical property information. and ct, were increased relative to ct, to impart known geological fabrics. 4.2.5. Constraining magnetic inversions with reference models built in Modelbuilder Magnetic inversions at the local and deposit scales were constrained using geological and physical property data collected during Hislop physical property studies (Chapter 2). A test version of the UBC GIFtools ModelBuilder program (Williams, 2008) was used to compile the geological and physical property data into a reference model. To build susceptibility reference models, susceptibility data from the Hislop physical property study, as well as from a regional physical property study focused on 160 geology west and south of the Hislop deposit area (Ontario Geological Survey, 2001), were used. Although the OGS regional study examined geology outside the extents of the Hislop study area, geological units are generally continuous across the greenstone belt, and relationships between physical properties and geology are expected to be consistent (Chapter 2). Physical properties measurements made down hole, and on surface samples or outcrop, can be input into the reference model with the appropriate associated drillhole collar and survey information, and with XYZ locations. These measurements, along with their linked geological and alteration information, form the basis of a physical properties database for the reference model being created. Any other empirical geological information, including geological maps, outcrop maps, downhole geology logs, and 3D geological volumes, can be painted onto the model cells and subsequently translated into physical properties by way of the program looking up the average physical property value calculated from previous property measurements for the rock type identified in the model cell. Thus, it is possible to populate an entire layer of cells with physical properties based on a geological map that covers the area, or to populate all cells intersected by a drillhole without actual measurements made on the core. Since potential fields geophysics is sensitive to near-surface sources it is important to be sure that constraints used near- surface are reliable. Synthetic modeling testing inversion results for a Hislop-like geological setting (Chapter 3) showed that poor susceptibility estimation for near-surface cells can affect the distribution of susceptibility throughout the whole model. Thus, while geological maps are available covering the area surrounding the Hislop deposit, only outcrop observations were used to populate near-surface cells. Geological contacts and rocks types from outcrop maps were considered more reliable than larger scale regionally interpreted geology maps. It is not uncommon to have more than one physical property data existing within one cell. For the Hislop deposit inversions, cells are 25 to 50 m. Susceptibility measurements were collected every 5 m on Hislop drillcore, adding 5 to 10 measurements to a single cell as a result. Where cells have more than one type of data 161 (e.g. downhole physical property measurements, plus data assigned based on geological mapping or logging) a single representative value must be chosen. The ModelBuilder program presents a number of options for choosing the representative value, depending on the types and amount of data available. For the Hislop susceptibility reference models, this value is the average of: a) the mean of actual physical property measurements, and b) the mean of measurements assigned based on geological observations, with both information sources considered equally reliable. Smallness weights are assigned to the constrained cells. These weights relay to the inversion the degree of reliability of its assigned physical properties. If a high smallness weight is specified, the inversion will attempt to achieve values close to the cell’s reference value. If properties within a cell are expected to be consistent over a surrounding volume, a ‘buffer’ can be designed around the cell. The information within the central cell is extrapolated to the cells of the buffer. Buffer cells might be assigned a low smallness weight, having a lower reliability than cells containing measured data. Refer to Table 4.4 for all constrained model parameters chosen for magnetic inversions. The resulting susceptibility reference models constrain 10% of the local scale susceptibility model, and 8.4% of the deposit scale susceptibility models. 4.2.6. Inversion model display Figure 4.16 outlines the surface extents of each of the model results to be discussed herein. Inversion results are displayed as cross-sections through the recovered 3D model for comparison to overlying mapped and interpreted geology. The location of the cross-section, indicated in Figure 4.16, is consistent between the displayed results, and represents a north-south slice through the model directly beneath the Hislop deposit. Isosurface models from each inversion result highlight the 3D distributions of anomalous material in the subsurface, and are interpreted based on previously noted correlations with geology. It is difficult to show the full 3D distribution of physical properties in a single 3D representation. To appreciate the shapes and depths of anomalous areas throughout 162 Table 4.4. GiFtools ModelBuilder options chosen for building Hislop reference models. Local scale Deposit scale GiFtools parameters magnetic magnetic inversion inversion Default parameters Lowest possible measurement (lower values rejected) Highest possible measurement (higher values rejected) Reference property value (where no constraining data exists in a cell) Smallness weight (reliability weight - defines desired degree of closeness to reference model values) Property lower bound (default upper bound where no data) Property upper bound (default upper bound where no data) 0 3000x10 SI 0 1 (low) 0 I 000xl 0 SI 0 3000x10 SI 0 1 (low) 0 1000x103Sl Source data Downhole property measurements Surface sample measurements Drillholes with geological observations and property measurements Drillholes with geological observations Weights and bounds Bounds assigned to a cell are controlled by the contained property data, and are defined by data within the confidence interval of: Representative block size % of block requied to be filled before bounds allowed to be applied Relative smallness weight (reliability weight) for surface measurements Relative smallness weight for drilling measurements Relative smallness weight for drilling geology logs Relative smallness weight for outcrop geology map Buffers Smooth interpolation across cells Maximum buffer distance for surface measurements Maximum buffer distance for drilling measurements Maximum buffer distance for drilling geology Maximum buffer distance for outcrop map Strike Dip Pitch % model constrained 10 10 1934 903 99.7% 99.7% Models built Reference model & smallness weights Lower & upper bounds model Smoothness weights 1034 113 1034 58 25 m 75% 25 m 75% 100 100 100 50 50 200 m 200 m 200 m 200 m 100 50 50 lOOm lOOm lOOm lOOm 115 115 90 90 0 0 10% 8.4% 163 the model volume, the models should be viewed with a 3D viewer such as UBC-GIF’s MeshTools3D, or with a Gocad viewer. The models and their associated meshes are included, along with a MeshTools3D model viewer, as an appendix on a CD accompanying the thesis (Appendix 4A). Instructions on how to use the viewer can be found on the UBC-GIF website, http://www.eos.ubc.ca!ubcgif/, under “Software manuals”. Observed versus predicted data for all inversion results are plotted in Appendix 4C (on CD). E sity regional volume Figure 4.16. Extents of inversion model volumes, with cross-section location indicated. 164 4.3. INVERSION RESULTS AND ANALYSIS 4.3.1. Magnetic susceptibility models Regional scale 18 km x 20 km model from unconstrained magnetic inversion High susceptibilities (>10 x 1 0 SI Units) in the regional magnetic model are associated with the dark green units on the Hislop area geologic map, which correspond to Fe-rich tholeiitic basalts (Fig. 4.17). The faulting and folding of a series of Fe-rich mafic rocks near the center of the map area, seems to thicken this rock package causing the significant central high susceptibility anomaly. The central faulted package of Fe-rich basalt units that dominate the northern part of the Hislop deposit stratigraphy appear to bottom-out at a depth of -3000 m. The anomaly conveys a steep dip to the southwest. A low susceptibility zone south of the central Fe-rich basalt package represents volcanic stratigraphy dominated by Fe-poor tholeiitic basalts and felsic volcanic rocks. It is not possible to distinguish between these two low susceptibility rock types in the susceptibility model result. High susceptibilities correlating with a series of Fe-rich volcanic flows persist through the southern region of the model, extending to depths of around 7000 m. A strong contrast occurs between the central high susceptibility zone, and low susceptibility rocks in the north, which is interpreted to be the manifestation of the location of the Porcupine Destor Deformation Zone. The belt scale PDDZ, mapped at the surface to occur along the southern margin of an Fe-poor basalt unit south of the Porcupine and Timiskaming assemblage sedimentary rocks, is indicated to dip about 45° - 60° southward beneath the interlayered mafic and ultramafic volcanic strata. This structure may be truncating mafic and ultramafic rock packages at depth. The very low susceptibility volume north of the interpreted fault likely represents the sedimentary rock sequences of the Porcupine and Timiskaming assemblages, or a combination of sedimentary rocks and Fe-poor mafic volcanic sequences. 165 The isosurface model in Figure 4.18 shows the regional subsurface distribution of Fe-rich mafic and ultramafic rocks recovered by the magnetic inversion. Figure 4.17. North-south cross-section through the regional-scale unconstrained magnetic inversion result, with overlying geologic map of the greater Hislop deposit area. For geological legend see Figure 4.2. Inset shows extent of model volume and cross-section location. Figure 4.16 can also be referred to. 166 Regional scale magnetic susceptibility model Isosurface cut-off: 29 x iO SI Units Figure 4.18. Isosurface model from regional scale magnetic inversion results. Ultramafic intrusion Low susceptibility areas dominated by felsic volcanics and sediments 1OO N Rotation of magnetic mafic and ultramafic stratigraphy across Ross and Hislop faults 167 Local scale 4 km x 6 km model This inversion essentially zooms in on the structure of the central high susceptibility body seen in the regional magnetic inversion result. Unconstrained inversion The highest susceptibilities are related to the central Fe-rich tholeiitic basalts (Fig. 4.1 9a). The local scale inversion result suggests a 25OO m depth for this basalt package, slightly shallower than the depth indicated in the regional result. This may be related to inversion-related smoothing over smaller distances in accordance with smaller cell sizes. The local scale model indicates the main susceptibility body is more structured than suggested in the regional model. The central susceptibility bodies extending in segments to depth gives the appearance of having once been one coherent unit, that was later dissected by near-vertical faults. A vertical low susceptibility zone in the south projects upward to correlate with a fault interpreted at the surface (the Ross Fault, Fig. 4.2). Magnetite in the rocks adjacent to these faults may have been destroyed as a result of structurally controlled C02-rich hydrothermal fluid circulation. As in the regional results, the Fe-rich basalts appear to dip generally southward. Other, more narrow Fe-rich mafic and ultramafic units are associated with shallow high susceptibility bodies. A high susceptibility body to the north is likely related to a mapped ultramafic unit. The associated susceptibilities of this northern body are consistent with those of talc-chlorite rich ultramafic rocks from the Hislop physical property studies, being somewhat lower than susceptibilities characteristic of Fe-rich basalts (Chapter 2). Low susceptibilities are associated with Fe-poor basalts, rhyolite units, felsic intrusives, and faulted areas. The extremely low susceptibility area north of the PDDZ and at depth is presumably reflecting thick packages of sedimentary rocks, which 168 Fi gu re 4. 19 . N or th -s ou th cr o ss -s ec tio ns th ro ug h th e lo ca l-s ca le a) u n co n st ra in ed , a n d b) co n st ra in ed m ag ne tic in ve rs io n re su lts ,w ith o v er ly in g ge ol og ic m ap s. Th e cr o ss se ct io n sp an s th e m ai n o re zo n e at H isl op . F or ge ol og ic al m ap le ge nd se e Fi gu re 4. 2. In se ts ho w s ex te nt o fm o de l v o lu m e an d cr o ss -s ec tio n lo ca tio n. Fi gu re 4. 16 ca n al so be re fe rre d to . U nc on st ra in ed lo ca ls c a le m a gn et ic su sc ep tib ili ty C on st ra in ed lo ca l s c a le m a gn et ic su sc ep tib ili ty M ag ne tic Su sc ep tib ili ty (xl o SI U ni ts) 10 0 15 0 were mapped north of the PDDZ on the geological map (Fig. 4.2), or Fe-poor mafic volcanic rocks, also mapped in the northern areas. The inferred Porcupine Destor Deformation Zone, mapped just north of the northern ultramafic units, separates high and low susceptibility regions. Its dip is slightly shallower in this result than in the regional model result. Constrained inversion The constrained local scale results exhibit some noticeable differences from unconstrained results (Fig. 4.1 9b). The core high susceptibility body is clearly separated from a smaller susceptibility anomaly closer to the surface. The ultramafic body north of the high susceptibility Fe-rich basalt units is more clearly disconnected from the basalts, and now has a more wedge-like appearance. Constraining the result also pushes high susceptibility bodies to a greater depth, steepening the dip angle of the inferred PDDZ, making it more consistent with the - 600 angle suggested in the regional model results. The steep dip angle of the central mafic volcanic rock package, and additional features not seen in the cross-section are illustrated in the isosurface model in Figure 4.20. Deposit scale 1.5 km x 2 km model This inversion focuses on the core portion of the central high susceptibility basalts interpreted from the local scale magnetic inversions. The goal is to attempt to uncover more fine scale structure, and to locate narrow low susceptibility syenite and rhyolite dikes, and alteration zones known to be spatially related to gold mineralization. Unconstrained inversion From the recovered model (Fig. 4.21a), a sharp gradient is obvious between the mapped Fe-rich basalt units, and the gold-related syenite. From synthetic modeling work (Chapter 3), an ultramafic rock-syenite dike contact was not detectable at these scales of inversion, and the equivalent contact is not obvious here. The ultramafic rocks south of the central syenite dike are 170 apparently low susceptibility, which, from physical property studies, could indicate their alteration to a carbonate-rich assemblage. Rhyolite dikes mapped to intrude the ultramafic unit could also be lowering the susceptibility here. Low susceptibilities are spatially related, in Figure 4.21, to Fe-poor mafic units, ultramafic units, and faulted rocks. Figure 4.20. Isosurface model from local magnetic inversion results. 5372880 5372155 West 5370705 Tightly folded magnetic mafic and ultramafic units Normal fault displacing Fe-rich mafic units to depth / Felsic intrusion 5369980 550050 Local scale magnetic susceptibility model Isosurface cut-off: 101 x iO SI Units //4: / 5S150 171 Unconstrained deposit scale magnetic susceptibility N — i15O Magnetic Susceptibility (x10 SI Units) 50 100 150 I I I 200 Figure 4.21. North-south cross-sections through the deposit-scale a) unconstrained, and b) constrained magnetic inversion results, with overlying geologic maps. The cross section spans the main ore zone at Hislop. For geological legend see Figure 4.2. Inset shows extent of model volume and cross-section location. a) Constrained deposit scale magnetic susceptibility N 172 The depth of the anomaly (-l000 m), and dip angle of the bottom of the high susceptibility body is similar to the local inversion outcome. But, despite cell sizes being smaller in the deposit-scale model (25 m3), there is little more resolution gained. In fact, a separation in the anomaly apparent in the local result does not occur in the unconstrained deposit model. The apparent lower resolution at the deposit scale might be explained by the default ci weightings or length scales (Mag3D User Manual, version 3.0, 2005). Length scales are applied in inversion work to manipulate smoothing in given directions according to cell size and prior geologic information. The default length scales used in magnetic inversions corresponds to cells sizes of 50 m (as was used in the local-scale inversions). Since the length scales were not reduced to correspond to smaller sizes in the unconstrained deposit-scale model, smoothing in the x, y, and z directions may be excessive. No obvious narrow low susceptibility zones characteristic of felsic dikes or altered rocks are distinguished within the susceptibility anomaly. The 50 m x 25 m spacing of the magnetic data used for this inversion limits the resolution of features smaller than this. In addition, as discussed in Chapter 3, smoothing inherent in the inversion result brought about by the choice of a model norm that gives priority to smooth results, would further obscure small-scale low susceptibility zones. Constrained inversion Constrained deposit-scale results indicate there is more complex structure within the high susceptibility zones related to the Fe-rich basalts (Fig. 4.21b). Although the irregular shape of the low susceptibility area within the central high susceptibility area is obviously an artifact of the location of the drillhole, and buffer zones, used to constrain the inversion, rendering the result not particularly geologically realistic, it indicates the existence of more fine scale structure within the central susceptibility anomaly. The internal low susceptibility zones were determined from drill core assessment to be related to the presence of syenites, Fe-poor basalts, and carbonate-altered rocks. Hislop drill core logs presented in Chapter 2 indicate that significant changes in rock type and alteration mineral assemblages, and thus susceptibility, can occur at the centimeter scale. The cell sizes in the inversion limits the ability to resolve these fine scale fluctuations. Nonetheless, constraints can offset some of the smoothing that occurs in the 173 inversion result and highlight some of these small scale features. The presence of the low susceptibility zones forces susceptibility to be redistributed within the model, and its magnitude to increase in the upper portion of the anomaly. As with the local-scale constrained inversion results in the previous section, a high susceptibility zone to the north interpreted to be related to ultramafic rocks, now appears to be more detached from the central high susceptibility Fe-rich basalts. An isosurface model for the deposit-scale constrained magnetic inversion is shown in Figure 4.22. 5372240 West N 5371 81 5 Distribution of Fe-rich mafic volcanic rocks in the Hislop deposit area - where highs drop away from 5371390 surface, may be representative of felsic intrusions, alteration or faults 5370540 SS1 530 I I 1 5370965 Deposit scale magnetic susceptibility model Isosurface cut-off: 85 x iO SI Units Part of • .,.ultramafk unit dl 52630 Figure 4.22. Isosurface model from deposit-scale magnetic inversion results. 174 4.3.2. Density model Gravity data spacing over the greater Hislop deposit area is large (500 m x 120 m), as such only a regional scale inversion was carried out. The corresponding large cell sizes (250 m) means only large scale features representing larger volumes dominated by low density felsic or sedimentary rocks versus high density mafic and ultramafic rocks are resolved. Fe-rich and Fe-poor mafic, and ultramafic volcanic rocks that underlie the central portion of the mapped area, cause a high density zone to dominate the core of the inversion volume (Fig. 4.23). To the north, there exists low density material likely related to sedimentary rocks of the Porcupine and Timiskaming assemblages. Low magnetic susceptibilities in the same location confirm the dominance of sedimentary rocks at depth. The boundary between the central high density area and the northern low density area here is not the same boundary that separates high and low susceptibility rocks in the regional magnetic results (see Fig. 4.17). This represents the contact between dense Fe-poor tholeiitic basalts north of the PDDZ, and the adjacent sedimentary assemblages. This may explain the difference in the apparent dip of the geologic units composing the central Hislop area between the magnetic and gravity inversion results. Low density areas in the southern regions of the model correlate with a package of rhyolitic volcanic rocks (pale yellow unit in inset of Fig. 4.23) and sedimentary rocks that extend southeastward out of the section. The 3D distribution of mafic and ultramafic rocks versus felsic and sedimentary rocks can be observed from the isosurface model Figure 4.24. 4.3.3. Resistivity models DC resistivity and IP inversion modeling was completed at two scales (Tabs. 4.2 and 4.3), however, since results are similar, only figures corresponding to the deposit-scale results are shown. 175 Figure 4.23. North-south cross-section through the regional-scale gravity inversion result, inverted with non-located constraints. The geologic map of the greater Hislop deposit area overlies the model. For geological legend see Figure 4.2. Inset shows extent of the model volume and cross-section location. Regional scale density with global constraints Ii (r\r. _•— Density (glcm3) 2.7 2.75 2.8 2.85 2.9 2.95 I I I I __ 176 5375150 Figure 4.24. Isosurface density model from regional-scale gravity inversion results. From physical property studies, high conductivities (low resistivities) were found to be associated with metamorphosed ultramafic rocks (talc-chlorite schists), whereas other rock types were less conductive. Higher conductivities in the DC resistivity inversion result, as expected, are associated with ultramafic rocks in the northern and central parts of the model (Fig. 4.25a). High conductivities however, may be instead, or additionally, correlated with Hislop deposit sulfides near the center of the map, or with interpreted faults. Two significant anomalies not represented in the cross-section, but seen in the isosurface model (Fig. 4.26a), correlate spatially with felsic intrusive rocks. Since felsic rocks are normally N 5371 400 Low density areas are dominated by r” sedimentary unfts.. / • .and felsic volcanic rocks and intrusives.. 1/ / Regional scale gravity model Isosurface cut-off: 2.83 g/cm3 177 Figure 4.25. North-south cross-section through the deposit-scale a) DC resistivity and b) IP inversion results, both inverted with non-located constraints. Hislop area geologic map overlies models. For geological legend see Figure 4.2. Inset shows extent of the model volume and cross section location. 178 b) Deposit scale chargeability model Isosurface cut-off: 0.031 Figure 4.26. Isosurface models for deposit-scale a) conductivity, and b) chargeability results. N Area dominated by ultramafic rocks j Conductivity anoma_,,,_..1 weakly associated with syenite dike South a) Deposit scale conductivity model Isosurface cut-off: 0.00064 S/rn north Area where a NW-SE fault, west of the Ross fault, intersects the N. Arrow fault. N Correlating anomalies near felsic intrusive! syenite dike South Chargeability anomaly follows syenite dike 554300 179 resistive, these anomalies may reflect the presence of sulfides. Alternatively, high conductivities may be associated with a conductive overburden, a feature indicated in inversion models generated for this area by Mueller et al. (2006). Outside of ultramafic rock-dominated areas and faulted areas, where geology is dominated by mafic and felsic rocks, lower conductivities (higher resistivities) occur. Depth of investigation tests (Oldenburg and Li, 1999) were conducted for selected lines during 2D inversion work on the Hislop DC resistivity and IP datasets (Appendix 4B), and indicate that subsurface features are generally not resolvable below about 400- 600 m depth. 4.3.4. Chargeability models A distinct high chargeability zone, likely to represent the presence of sulfides, occurs in the subsurface beneath the mapped syenite dike (slightly obscured in figure) at Hislop, and extends to depth, dipping slightly to the southwest (Fig. 4.25b). This anomaly disperses horizontally at depth. The horizontal displacement does not correspond to any known features and is similar to artifacts in synthetic inversion models for IP data in Chapter 3. Additionally, from depth of investigation tests, it is suspected that the model is essentially unreliable at these depths. An additional small anomaly occurs just north of the central chargeability anomaly in proximity to some interpreted faults, and overlapping with a conductivity high. When the 3D model is viewed, the central chargeability anomaly extends a few hundred meters to the northwest and to the southeast following the syenite dike, before it detaches from the surface and moves to depth (Fig. 4.26b). Northwest of the West Area open pit, a significant anomaly correlates with the northern Arrow Fault (refer to Fig. 4.2). In the isosurface model, three chargeability highs correlate with conductivity highs. Two of the anomalies are in proximity to felsic rocks, and the third occurs in the 180 northeast in mafic rocks north of the PDDZ. The correlation between the two physical properties could indicate sulfide-rich rocks. 4.4. QUERYING COMBINED INVERSION RESULTS Data from Hislop 3D inversion models were combined using Gocad 3D GIS software with Mira Links add-ons, and the resulting ‘common earth models’ queried in an attempt to define spatial extents of rock units, and potentially prospective areas for exploration targeting. It is important to note that the local and deposit-scale magnetic susceptibility models used are from constrained inversions, whereas all other inversion results making up the common earth models are unconstrained. This process involved projecting properties from the different inversion results onto one discretized mesh. For Hislop, three common earth models were created, a regional scale model where susceptibility and density were projected onto a mesh with 200 m cells, a local scale model, where susceptibility and chargeability from local scale inversions were projected onto a mesh with 50 m cells, and a deposit scale model, with susceptibility and chargeability data held in 25 m cells. During data projection, each ‘client’ cell in the common earth model grid takes on the value of the closest ‘server’ (inversion) cell center. Physical property cut-offs used to query the common earth model were determined using descriptive statistics calculated during Hislop physical property studies. In essence, susceptibility and density are queried at the regional scale with expectations of modeling lithological units, or significant packages of rocks, and susceptibility and chargeability are used at the local and deposit scales to find sulfide-bearing felsic intrusives, and carbonate-altered rocks. Conductivity values do not uniquely define prospective rock types, or hydrothermal alteration (Tab. 4.2), and is thus not used in the queries. High conductivites can indicate the presence of faults that act as important 181 structural traps for gold mineralization, or sulfide-rich rocks, but high conductivities can also be related to least-altered and likely unmineralized ultramafic volcanic rocks. The cut-off values used for common earth model queries are based on physical property ranges characterizing rock types and alteration at Hislop (Tab. 4.2). The query results are presented in plan-view in the corresponding figures, with a transparent geologcal map overtop. 4.4.1. Regional scale query (susceptibility and density) Three different queries were applied to the regional combined susceptibility- density model in an attempt to target the three populations of rocks indicated in the Hislop susceptibility versus density plot (Fig. 4.4): 1. low susceptibility-low density felsic rocks (Fig. 4.27), 2. high susceptibility-high density least-altered mafic and ultramafic rocks (Fig. 4.28), and 3. low susceptibility-high density carbonate-altered or Fe-poor mafic rocks (Fig. 4.29). By targeting low susceptibility (<3 x iü SI Units) and low density (<2.75 g/cm3) areas of the model at the regional scale, two felsic intrusives in the south, and sedimentary rocks mainly associated with the Porcupine and Timiskaming assemblages in the northeast map area are isolated (Fig. 4.27). Felsic intrusive bodies near the center of the mapped area overlying this model are not detected by this query. This might relate to the large cell sizes used and the overwhelming of smaller lower susceptibility and density zones by the more abundant susceptible and dense mafic and volcanic units, an effect noticed in synthetic modeling results (Chapter 3). A query targeting high susceptibility (>5 x iO SI Units), and high density (>2.8 g/cm3)cells in the regional scale common earth model targets areas dominated by Fe-rich basalts and ultramafic volcanic rocks in the central and southern parts of the map area 182 (Fig. 4.28). High susceptibilities and densities stop abruptly at the mapped location of the PDDZ. Figure 4.27. Result for a physical property query targeting low magnetic susceptibility- low density cells within the regional-scale common earth model. Anomalous zones extend to greater than 2000 m depth. Plan view with transparent geology. Geological legend in Figure 4.2. A low susceptibility (<3 x SI Units) and high density (>2.8 g/cm3) query identifies areas dominated by Fe-poor basalts, or possibly areas of carbonate-altered Fe- rich basalts or ultramafic rocks (Fig. 4.29). Cells highlighted by this query, underlying mapped Fe-rich basalts and ultramafic rocks, may warrant further inspection as the low susceptibilities here could indicate carbonate alteration of these normally high susceptibility rocks. A northern zone of low susceptibility-high density cells extends from a sequence of mafic rocks just north of the PDDZ, into the mapped Porcupine assemblage sedimentary rocks. Sedimentary rocks elsewhere have typically low densities, and this anomaly might indicate that the contact is interpreted incorrectly. Showsallcellsthat meetquerycriteria —ç romsu ace •%%‘i%%%IlIIIIdownward • a a • I I I I I I I I I I I I I I I I I I øøø#ø.Fø•• a • • a I I I I I I I I I I I I I I I III 140 • • • P P I I I I I I I I I I I I I Illi • •••.p..•. a • 1 . a • • i a i a a a • a • - a a - • S S I a S S 544 I — S S S S S S S S S S • • eaa-. a L- a a a a a a a a a a seSaseeS a.... — see — a — — a a a . • a a a — — a s —— • • — 5 esa a a a 5 a.—...— Regional scale low susceptibility - low density query / 11111111 ulla_IlIl. Ia III III •• a.. -. 11111111 1 km 183 Figure 4.28. Result for a physical property query targeting high magnetic susceptibility - high density cells within the regional-scale common earth model. Anomalous zones extend to greater than 2000 m depth. Plan view with transparent geology. Geological legend in Figure 4.2. 4.4.2. Local scale query (susceptibility, chargeability) Cells in the local scale common earth model containing low susceptibilities (<3 x i03 SI Units) combined with high chargeabilities (>0.12) were targeted to identify potentially prospective felsic intrusive rocks, carbonate altered zones, and sulfide-rich areas. This query result highlighted a number of areas focused near the Hislop deposit. These zones are concentrated along geological contacts (marginal to the faulted Fe-rich basalt unit), and especially where the contacts are faulted, or where two or more faults intersect. Along the southern central syenite dike contact the highlighted zones are 184 %%%%‘ a - a I • a aaaaa •‘a — Regional scale high susceptibility - high density query • S S S S S S SøØ -a...... • alaaa a a..a..a. — aaa.a.a.a • a a a. a a a. •aaa.%aaaa.aa - .a.aaaeai..aaIa • a aa..a * a a a a a a a a a ,- - a a S • S • a a a a a a a a . a a a • a a • a a a a ,.-: — S S — — — S — — — • • • a aSa.a...a. a. a a.. S a. a a ..weeS — • aaaaa_a S — a .aaaaasa a anaaa a — — — a a a a a a a a a a a a a a a a — a — a a aaaas.. a...a. a a a a a a a — a.aaaaaaaaa — —— _—as.sa. a a • a a a a.•s Sa.j a..rna .a a a a a a a a a a a a a a. a a aaaaaaaaaa a — a a a •aaaa a a — —h. • •.... a. a a . i a a a a a a a • . a a a a. aaa a ...4J1sInn.r iLPØØØØS•• a III a a m,vIi1$.%j a I II a. a 00’ a a PWIØWFYIO a.. -: a. I lila. at a a. at a a a laa a at %% %aaa 1km aaaa.ia.aIaaa..aaaa%% %%% Figure 4.29. Result for a physical property query targeting low magnetic susceptibility - high density cells within the regional-scale common earth model. Anomalous zones extend up to 1500-2500 m depth. Plan view with transparent geology. Geological legend in Figure 4.2. coincident with some high gold abundances (Fig. 4.30). Returned anomalies also coincide with higher gold concentrations near the northern Arrow Fault, and where the northern Arrow Fault intersects the north-south trending fault west of the Ross Fault. An anomaly northwest of the mapped PDDZ is also marginal to drillholes with anomalous gold. One, low susceptibility, high chargeability zone occurring just east of the Hislop deposit, near a felsic intrusion, is in an area of minimal to no drilling. Some high gold values occur in association with a drilled area southwest of the Hislop deposit, near the Hislop Fault. Only about half of this drilled area is contained within the common earth model. The query did not identify prospective rocks here. Referring back to the regional susceptibility-density queries however, this area correlates with low susceptibilities. It is possible that the gold here is not associated with sulfides Regional scale low susceptibility - high density query a. a .d’ø a a a a• .ø• sm—a • a. • . .. a.%%’• • • a a a a a • • . a a a . . • a. • • • • • • • . ,at sat...... .•...aa...•.•• .5 S... •S•I a.....p,... .• S. SS •••S• •.••.•.a... .. S.. .t....t.I....u...aa..a.. ...a S • S S I I I I I I a • a a a a a a . I I • S S a — S S S S • I • • a • a • . • a a S a •. . . — • — S 5 a a a a — S S S • S • • S • S • S 5 5 5 5 &es 5 — — - •SSSS •SJ•s a S • •• • _____ — a — nSSS S • tSSaSflaaSSSS.5. __S..• S S • a •.a.SSSF - a a a a.aaaatt •., S. •...a.att.. as Ia••.S•S • s-,. 5••.IS .5 Isa... Seem . •li III.. a_s_se ...ø S... - 505555. • ___ . ISP 55 asS.... 5505..a.I a •IIIS ..S.S... .55Sa. : • 4 • a.aaa.aa.a.55155 • ..aaaa.... 555 - • .lla%aaaa ... . • • • . a . • . • • •• V. %_ •: .0.1 1kmr’. 185 and thus chargeability values are not anomalous at this location. This area is currently being explored by Stroud Resources (www.stroudresourcesltd.com), and the endeavor is confusingly called the Hislop Project. Figure 4.30. Result for a physical property query targeting low magnetic susceptibility - high chargeability cells within the local-scale common earth model. Anomalous zones extend up to 3 00-600 m depth. Anomalous downhole gold assays are indicated. Plan view with transparent geology. Geological legend in Figure 4.2. 4.4.3. Deposit scale query (susceptibility, chargeability) Querying the deposit scale common earth model using the low susceptibility and high chargeability criteria, returns several zones with the desired characteristics, whose locations are generally consistent with local scale query results (Fig. 4.31). There is slightly more detail compared with the local scale results, with some additional small regions highlighted, and others eliminated. Again, most prospective areas are spatially 1 km ..fl..flfl... S5% ‘! N” ‘N, •. . “‘. :::::::: , N. Arrow Fault . ::::::::.. \ ‘ “• .. .• •:.•H 4 S.ArrOW_ Local scale low susceptibility - highchargeability query Gold (ppm) 4 6 8 10 12 14 16 186 associated with areas of complex faulting, and are also proximal to felsic intrusives and dikes (some narrower dikes south of the main syenite dike are obscured by the anomalies and plotted gold assays, but can be see more clearly in Fig. 4.2). As with the local-scale results, there are areas of high gold concentrations not detected by the query that may represent mineralization not accompanied by disseminated sulfides. Figure 4.31. Result for a physical property query targeting low magnetic susceptibility - high chargeability cells within the deposit-scale common earth model. Anomalous zones extend up to 60-500 m depth. Anomalous downhole gold assays are indicated. Plan view with transparent geology. Geological legend in Figure 4.2. The query results are not expected to detect all subsurface areas meeting the criteria. Constrained deposit scale magnetic inversion results indicate that there are small scale low susceptibility zones that can be masked by smoothing of high susceptibility 1 km . . 4 ... flu a.... •:..:::.n :::::. 1iF:EE...jJEr a. “ ... Deposit scale low susceptibility - high chärgeability query S. ArroW9it Gold (ppm) 2 4 6 8 10 12 14 16 I I I I I 187 values in the inversion result. There is likely more detail in high susceptibility rocks that cannot be resolved using inversion, or querying techniques at this scale. 4.5. SUMMARY AND DISCUSSION Geophysical inversion of a series of geophysical datasets over the Hislop gold deposit in the south-central Abitibi greenstone belt, was carried out at a range of scales of investigation. Results show that inversion is a useful tool for detecting specific lithologic packages, and altered and mineralized rock, and for interrogating their 3D subsurface distribution. Regional scale inversion results highlight large scale structures, and lithological boundaries. Magnetic results show packages of susceptible rocks, dominated by Fe-rich tholeiitic basalts, extending to depth in the crust up to about 7 km. This depth is consistent with published depths for crustal rocks above granitic basement rocks in the Abitibi greenstone belt (Reed et al., 2005). Magnetic inversion traces the crustal scale Porcupine Destor Deformation Zone, a regionally important gold-related structure, into the subsurface from its interpreted location at the surface, and indicates a southward dip (about 45° - 60°) as it undercuts and possibly truncates the overlying packages of volcanic rock. This southward dip is consistent with results for recent seismic work, and magnetic and gravity inversions completed in the Currie Township west of the Hislop deposit (Reed, 2005; Reed et al., 2005). The dip of the PDDZ is interpreted to vary along its trace, however, changing from 45°- 65° in the Hislop area to steeper angles closer to the Ontario-Quebec border (Berger, 2002). At the regional scale, major lithologic units and domains are mapped using combined magnetic and gravity inversion results. Querying the combined results revealed three petrophysically distinct lithological packages, and importantly, allowed felsic and sedimentary rocks (low susceptibility-low density) to be distinguished from Fe-poor tholeiites and potentially carbonate-altered mafic and ultramafic rocks (low 188 susceptibility-high density). Exploration at the regional scale might focus generally on the range of low susceptibility regions, which are expected to contain dominantly felsic rocks, and carbonate-altered rocks. Associated lithogeochemical studies testing various alteration indices (Davies et a!., 1990; Eilu et al, 1995; Piche and Jebrak, 2003) could be helpful in further distinguishing least-altered Fe-poor mafic rocks from carbonate-altered rocks. Regional scale inversions may be more appropriate for mapping geology in areas of poor outcrop than for generating targets directly. Although high susceptibility and high density rocks are likely to reflect mainly least-altered mafic and ultramafic rocks, it is possible that smaller zones of prospective low susceptibility-low density rocks within these larger rock packages are not being detected. At the local and deposit scales of magnetic inversion, more detail is resolved within the subsurface. Distributions of Fe-rich basalt, versus Fe-poor basalt and felsic rocks are better defined, and locations and orientations of near-vertical faults dissecting a central package of high susceptibility Fe-rich basaltic rocks are discernible. There is not a significant increase in detail visible in 25 m3 cell deposit scale inversion over the 50 m3 cell local scale inversion. Features smaller than 25 m, which might constitute narrow, mineralization-related felsic dikes and alteration zones, are simply not detectable at these scales, with magnetic data spacing limited to 50 m x 25 m, and typical inversion-related smoothing occurring. More detail was indicated however, when constrained magnetic inversions are completed at the deposit scale, and it becomes apparent that there are small scale heterogeneities in the physical property distributions that are not being detected by unconstrained inversions. Synthetic modeling studies showed that narrow low susceptibility and low density zones can be imaged down to at least a few hundred meters at deposit-scales of exploration, when data spacing is 50 m x 10 m and cells are 10 m3 (Chapter 3). It should be noted however, that with smaller cell sizes and smaller data spacing, the computation times for inversion are increased significantly. From physical property studies (Chapter 2), and synthetic modeling work (Chapter 3), it is clear that density is a useful physical property for targeting felsic dikes, and helpful in distinguishing low susceptibility mafic rocks from felsic rocks. 189 Unfortunately gravity data available for this work was extremely sparse and could not be used to model the subsurface at the local or deposit scales. Having both high resolution magnetic and gravity data available for deposit-scale exploration would be very useful for geological mapping, and detecting gold-related rock types at smaller scales of investigation. DC resistivity inversions image some ultramafic volcanic units in the Hislop area, as was predicted by the consistently low resistivities (high conductivities) of these commonly sheared rocks indicated from physical property studies on hand samples (Chapter 2). Through inversion work, it was shown that high conductivities are also correlated spatially with faults. Whether this is related to fluid content and porosity which increases conductivity (Telford et al., 1990) or to the presence of sulfides was not determined. Conductivity was not used in common earth model queries. Conductivity does not consistently detect prospective rocks in the Hislop area. High conductivities may indicate faults which do not necessarily host mineralization, or conductive talc-chlorite schists which are not typically mineralized. Induced polarization inversions locate chargeability anomalies that are interpreted to be due to the presence of sulfides. These anomalies align with the immediate location of the Hislop deposit proximal to the central northwest-southeast trending syenite dike. Again, by combining the results of different inversions, the most information is gained. At the local and deposit scales, queries of combined susceptibility and chargeability data were used to try and locate sulfide-rich felsic dikes and carbonate- altered rocks. Query results highlighted zones focused along the mainly faulted contacts between Fe-rich basalts and other rocks, and near cross-cutting faults. These areas highlighted by the queries in the area of the Hislop deposit are geologically ideal gold targets, with faults providing conduits and structural traps for hydrothermal fluids, and nearby Fe-rich rocks that promote sulfidation processes leading to gold precipitation (Mikucki, 1998). Some areas where high gold grades were intersected during drilling were targeted by the queries, confirming prospectivity. 190 Although gold mineralization hosted in greenstone facies rocks does not usually have a strong geophysical signature due to its typically low grades, it is still possible to remotely target Archean orogenic gold deposits using alternative exploration vectors such as hosting lithology, and alteration mineral assemblages. Geophysical inversion not only allows detection of prospective gold-related rocks but can indicate the spatial extent of these rocks in the subsurface. Geophysical based mapping of geology, and exploration target generation is so valuable in Archean greenstone terranes since they are often characterized by low percentages of outcrop. The key to getting the most from inversions is by understanding relationships between physical properties in the geological environment or mineral deposit setting of interest. With this prior knowledge inversions can be constrained to yield results consistent with geological observations, inversions will be more confidently interpreted overall, and queries can be developed to target model cells with the desired combination of physical property characteristics representative of mineralized rocks. 191 REFERENCES Ayer, J., Amelin, Y., Corfu, F., Kamo, S., Ketchum, J., Kwok, K, and Trowell, N., 2002, Evolution of the southern Abitibi greenstone belt based on U-Pb geochronology: autochthonous volcanic construction followed by plutonism, regional deformation and sedimentation: Precambrian Research, v. 115, p. 63-95. Bate, S.J., Thorsen, K.R., and Jones, D., 1989, The Casa Berardi area; an exploration case history: Proceedings of Exploration ‘87, Ontario Geological Survey, Special Volume 3, p. 855-870. Berger, B.R., 1999, Geological investigations along Highway 101, Hislop Township: Ontario Geological Survey, Summary of Field Work and Other Activities 1999, Open File Report, 6000, p. 5-1 — 5-8. Berger, B.R., 2002, Geological synthesis of the Highway 101 area, east of Matheson, Ontario: Ontario Geological Survey, Open File Report 6091, 124 p. Davies, J.F., Whitehead, R.E.S., Huang, J., and Nawaratne, 5., 1990, A comparison of progressive hydrothermal carbonate alteration in Archean metabasalts and metaperidotites: Mineralium Deposita, v.25, p. 65—72. DCIP3D User Manual, version 2.1, A program library for forward modelling and inversion of DC resistivity and induced polarization data over 3D structures: UBC Geophysical Inversion Facility, University of British Columbia. Doyle, H.A., 1990, Geophysical exploration for gold — a review: Geophysics, v. 55, p. 134-146. Eilu, P., Groves, D.I., Mikucki, E.J., McNaughton, N.J., and Ridley, J.R., 1998, Alteration indices and pathfinder elements in wallrock alteration zones around Archaean 192 lode-gold deposits, in Pasava, J., Kribek, B., and Zak, K., eds., Mineral Deposits; from their Origin to their Environmental Impacts, Third Biennial SGA Meeting, Prague, Czech Republic, p. 113-116. Farquharson, C.G, Ash, M.R., and Miller, H.G, 2008, Geologically constrained gravity inversion for the Voisey’s Bay ovoid deposit: The Leading Edge, v. 27, p. 64-69. Hallof, P.G., and Yamashita, M., 1990, The use of the IP method to locate gold-bearing sulfide mineralization, in Fink, J.B., Sternberg, B.K., McAlistar, E.O., Weiduwilt, W.G., and Ward, S.H., eds., Induced Polarization: applications and case histories, Society of Exploration Geophysicists, Tulsa, Ok., p. 227-279. Jackson, S. L., and Fyon, A. J., 1991, The western Abitibi subprovince in Ontario, in P.C. Thurston, H.R. Williams, R.H. Sutcliffe and G.M. Stott, eds., Geology of Ontario, Ontario Geological Survey, Special Volume 4, Part 1, p. 405-482. Johnson, I., Webster, B., Matthews, R., and McMullen, S., 1989, Time-domain spectral IP results from three gold deposits in northern Saskatchewan: CIM Bulletin, v. 82, p. 43- 49. Kerrich, R., 1989, Geodynamic setting and hydraulic regimes: shear zone hosted mesothermal gold deposits, in Bursnall, J.T., ed., Mineralization and Shear Zones, Geological Association of Canada, Short Course Notes 6, p. 89-128. Kowalczyk, P., Thomas, S., and Visser, S., 2002, 3D inversion of resistivity and IP data, two case studies from mineral exploration: SEG extended abstracts, SEG International Exposition and 72nd Annual Meeting, 4 p. Li, Y., and Oldenburg, D.W., 1996, 3-D inversion of magnetic data: Geophysics v. 61, p. 394-408. 193 Li, Y., and Oldenburg, D.W., 1998, 3D inversion of gravity data: Geophysics, v. 63, p.109-19. Mag3D User Manual, version 4.0, 2005, A program library for forward modelling and inversion of magnetic data over 3D structures: UBC-Geophysical Inversion Facility, University of British Columbia. Mueller, E.L., Reford, S.W., Dawson, D.J.W., Morrison, D.F., Pawluk, C., Grant, J., Spector, A., Rogers, D.S., and Savage, T., 2006, Acquisition, inversion and presentation of geophysical data for geoscientific profiles in the Timmins—Kirkiand Lake area: Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6189 , 28 p., 15 sheets. Mikucki, E.J., 1998, Hydrothermal transport and depositional processes in Archean lode- gold systems: A review: Ore Geology Reviews, v. 13, p. 307—321. Mira Geoscience Limited, 2005a, Detectability of mineral deposits with electrical resistivity and induced polarization methods: Ontario Geological Survey, Miscellaneous Release—Data 181. Mira Geoscience Limited, 2005b, Detectability of mineral deposits with potential field methods: Ontario Geological Survey, Miscellaneous Release — Data 177. Oldenburg, D.W., and Li, Y.,1999, Estimating depth of investigation in dc resistivity and IP surveys: Geophysics, v. 64, p. 403-416. Oldenburg, D.W., Li, Y., Farquharson, C.G., Kowalczyk, P., Aravanis, T., King, A., Zhang, P., and Watts, A., 1998, Applications of geophysical inversions in mineral exploration problems: The Leading Edge, v. 17, p. 461 - 465. 194 Oldenburg, D.W., Li, Y., and Ellis, R.G., 1997, Inversion of geophysical data over a copper gold porphyry deposit: A case history for Mt. Milligan: Geophysics, v. 62, p. 1419-143 1. Ontario Geological Survey 2001, Physical rock property data from the Physical Rock Property Study in the Timmins and Kirkland Lake Areas: Ontario Geological Survey, Miscellaneous Release — Data 91. Ontario Geological Survey, 2004, Ontario airborne geophysical surveys, gravity data, northeast, northwest, and south Timmins areas: Ontario Geological Survey, Geophysical Data Set 1051. Phillips, N.D., 2002, Geophysical inversion in an integrated exploration program: examples from the San Nicolas deposit: Unpublished M.Sc. Thesis, University of British Columbia, 237 p. Piche, M., and Jebrak, M., 2003, Normative minerals and alteration indices developed for mineral exploration: Journal of Geochemical Exploration, v. 82, p. 59-77. Poulsen, K.H., Robert, F., and Dube, B., 2000, Geological classification of Canadian gold deposits: Bulletin - Geological Survey of Canada, Report 540, 106 p. Power, W. L., Byrne, D., Worth, T., Wilson, P., Kirby, L., Gleeson, P., Stapleton, P., House, M., Robertson, S., Panizza, N., Holden, D. J., Cameron, G., Stuart, R., and Archibald, N. J., 2004, Geoinformatics evaluation of the eastward extension of the Timmins Gold Camp: Geoinformatics Exploration Inc., Unpublished report for St Andrew Goldfields Ltd. Prest, V.K., 1956, Geology of the Hislop Township: Ontario Department of Mines, Annual Report, 1956, v. 65, 51 p. 195 Reed, L. E., Snyder, D. B., and Salisbury, M. H., 2005, Two-dimensional (2D) reflection seismic surveying in the Timmins-Kirkiand Lake area, northern Ontario; acquisition, processing, interpretation: Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6169, 68 p., 10 plates. Reed, L. E., 2005, Gravity and magnetic three-dimensional (3D) modeling: Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6163, 40 p., 4 sheets. Roscoe and Postle, 1998, Hislop Mine Property, Roscoe and Postle Associates Inc., St. Andrew Goldfields Ltd. internal report, unpublished, p. 66-89. Seigel, H.O., Johnson, I., and Hennessey, J., 1984, Geophysics the leading edge: Geophysics: the Leading Edge of Exploration, v. 3, p. 32-35. Telford, W.M., Geldart, L.P., and Sheriff, R.E., 1990, Applied Geophysics, Second Edition: Cambridge University Press, 770 p. UBC-GIF Inversion for Applied Geophysics CD-ROM, 2000-2006, a teaching and learning CD-ROM, by Oldenburg, D.W., and Jones, F.M.: University of British Columbia, Geophysical Inversion Facility. Williams, N.C., 2006, Applying UBC-GIF potential fields inversions in greenfields or brownfields exploration: Australian Earth Sciences Convention, 2006, Melbourne, Australia, extended abstract, 10 p. Williams, N.C., 2008, Geologically-constrained UBC—GIF gravity and magnetic inversions with examples from the Agnew-Wiluna greenstone belt, Western Australia: Unpublished Ph.D. Thesis, The University of British Columbia, 479 p. www.standrewgoldfields.com, website for St. Andrew Goldfields, Ltd. 196 www.stroudresourcesltd.com, website for Stroud Resources Ltd. www.zonge.comlLablP.html, website for Zonge Engineering and Research Organization, IP and resistivity measurements. 197 Chapter 5: Summary and future work 5.1. SYNTHESIS OF RESEARCH PRESENTED The goal of this research was to apply an understanding of the characteristic geology and physical properties of a typical Archean orogenic gold deposit to geophysical inversion for improved mapping of geology and delineation of gold-related rocks in the subsurface. The Hislop deposit of eastern Ontario was used as an example of this deposit type in the case study. Key relationships between geology and alteration, and physical properties were established for Hislop, and ranges of physical properties representing more prospective geology were identified. Physical property information was eventually used to improve inversion results through their incorporation as inversion constraints. Synthetic modeling revealed the sizes and depths, and necessary physical property contrasts required to image petrophysically distinct gold-related features in the subsurface. It allowed depths of investigation to be roughly determined, and allowed certain inversion artifacts to be identified. Preliminary physical property, and synthetic forward and inverse modeling work contributed strongly to how eventual inversion models for the Hislop deposit were interpreted. At larger scales of investigation, magnetic and density data can be used for mapping geology, and for determining regional-scale exploration targets based on the distribution of the geologic units and structures modeled. Results are especially useful in the parts of the Abitibi greenstone belt that were modeled during this study, as outcrop percentages are low. At deposit-scales of investigation, induced polarization (IP) inversion methods were effective in detecting sulfide-rich rocks, with chargeability anomalies correlating well with known mineralization. DC resistivity inversion results were not easily interpretable due to the variable behavior of conductivity. Some correlations between chargeability and conductivity anomalies in select areas surrounding 198 the Hislop deposit, however, may suggest the presence of potentially gold-bearing, sulfide-rich rocks. Magnetic inversions at the local and deposit scales identified a complex distribution of faults characterized by low susceptibilities possibly brought on by magnetite-destructive carbonate alteration. Smaller scale features, such as the gold- related syenite dike at Hislop, were obscured by smoothing of higher susceptibilities within the inversion volume. Synthetic modeling work has indicated that more detail can be derived from inversion, permitting better resolution of the narrow features that characterize typical Archean orogenic gold deposits. But to attain this detail, it is necessary to focus on a small area, collect closely-spaced data, and to use small inversion cell sizes. The density data available for the Hislop area was very widely spaced. From physical property studies and inversion investigations, it is expected that smaller scale density data in combination with closely spaced magnetic data would be effective in establishing geological contacts and rejecting least-prospective rocks at the deposit scale. 5.2. SIGNIFICANCE AND CONTRIBUTIONS TO THE FIELD There is limited published information detailing geophysical inversion modeling efforts in Archean orogenic gold environments. The work presented in this thesis provides a comprehensive case study focused on the application of inversion methods for orogenic gold exploration, and may act as a reference point for others embarking on using inversion to explore in this deposit setting. As previously discussed, an understanding of physical properties lays the groundwork for applying geophysics or geophysical inversion as exploration tools. The extensive physical property work completed constituted a major component of this thesis and is an important contribution to geophysics-based exploration. A significant amount of physical property data was generated for Hislop deposit rocks, and physical property ranges for typical host rock types and for prospective rocks were delineated. This data may eventually be contributed to a regional or national physical property databases, 199 enhancing the sources on which to draw for geophysics-based exploration in similar areas where little sampling or physical property reconnaissance has been done. Synthetic modeling of a typical gold deposit provided insight into the features that will, and will not be imaged for a given survey design and mesh discretization. It also allowed application of various basic constraints to be tested to assess their influence on recovered models. This compilation might provide some guidance for geophysical survey, or inversion design, in a similar setting. Inversion of the range of geophysical data available over Hislop, at a range of scales made for a unique case study with significant breadth. Querying combined physical property models was shown to be a valuable application of inversion results. It was demonstrated that the combination of magnetic susceptibility and density models were useful for distinguishing sedimentary and felsic rocks from Fe-rich mafic and ultramafic rocks, and Fe-poor mafic and ultramafic rocks, and for outlining their 3D subsurface distributions at the regional scale. The queries used constitute important mapping tools in areas of poor outcrop in this part of the Abitibi greenstone belt. At smaller scales, prospective areas can be distinguished by combining chargeability and susceptibility results, as was indicated by correlation between known mineralization, and high chargeability-low susceptibility anomalies. Physical property studies (Chapter 2) indicated similarities in local and regional scale physical property ranges and distributions. As such, these queries could be applied to other inversion results regionally. 5.3. LIMITATIONS OF THE THESIS RESEARCH Due to their ease of collection, it was possible to amass a large number of magnetic susceptibility and density measurements for Hislop samples. An equivalent number of measurements for resistivity and chargeability were not generated, as equipment was not available to make the measurements in-house. Measurements had to be completed at the physical properties laboratory at Zonge Engineering and Research 200 Organization, Inc., resulting in a limited dataset. This meant the rock types in their least- altered and variably altered states are not well-represented. It was not possible to make a thorough assessment of the effects on alteration on these two electrical properties, and in addition, to have confidence in relationships that were indicated between these properties, and sulfide abundance or porosity. All modeling possibilities were not considered, and synthetic modeling and inversion work could both be expanded on. For example, synthetic modeling was only carried out at one representative scale of exploration, the deposit-scale, with data collected only on a 50 m x 10 m grid. With anticipation of completing regional inversions it would be beneficial to model the deposit at a more regional scale. Only select variations on the geological setting of the modeled gold deposit were considered during synthetic modeling studies, and constraints only demonstrated for a subset of these scenarios. There are obviously many different scenarios that can be tested, but it would take considerable time to assess them all. Similar expansions on work could be applied to inversion of actual data over Hislop. Different combinations of constraining information could be applied to each of the models to explore the full range of possible outcomes. Throughout the course of inversion studies, it was indicated that inversion results can also vary dramatically when geophysical data are scaled differently, and when errors are changed. These parameters might also be investigated more extensively through additional inversions. With the array of possible modifications, it is feasible that there is a better model to be generated in each case. An additional limitation of the research relates to the previous comments. One of the interesting challenges encountered in completing this project was dealing with the rapid rate at which inversion concepts and methods are developing. At times, a series of models would be completed only to discover that there was a newer version of the inversion code available! This is a relatively new field, and the Geophysical Inversion Facility at UBC are at the forefront of it. The UBC-GIF has developed robust inversion codes that are used worldwide, and the codes are constantly being updated to adapt to the modeling needs of exploration and environmental communities. This means that more 201 effective codes, or programs with increased functionality, are becoming available on a regular basis, and that the models presented herein might be improved on with application of newer software. 5.4. RECOMMENDATIONS FOR CONTINUED WORK There were several ideas proposed during the course of this research that were not followed up on. Some of the ideas worthy of further investigation are listed here, along with additional suggestions. More resistivity and chargeability data is needed to better define relationships between geology and physical properties. Since IP methods are so effective in delineating sulfides at Hislop, and have been shown to be effective in detecting mineralization for other gold deposits, more chargeability data would be useful. It would be beneficial to do a more in depth analysis of relationships between chargeability to sulfides types, sulfide textures and abundances, as well as attempt to define a relationship between gold and chargeability. As chargeability data was collected at multiple time windows during IP work both in the field, and in the laboratory, there is potentially more information to be gained. To calculate chargeability for this thesis, the value representing the voltage decay over these time windows was chosen to be 80% of the sum of voltages over eight of the time windows. This choice of representative value is somewhat arbitrary, and there exist other standard measurements in the industry. The consistency of measurement methods for a suite of data is of more importance than choice of calculation. By assessing the entire decay curve, or looking at voltages from individual time windows, instead of calculating a representative value, relationships between chargeability and mineralization not previously identified may be revealed. 202 It may be constructive to automate the synthetic modeling process. Constructing the range of synthetic models and testing them was time-consuming, and only select scenarios were represented. Such a program could automatically vary geometry and physical property contrasts of a target feature for given survey parameters and inversion cell sizes, and assesses model difference values (Chapter 4) to determine conditions where the difference between true and recovered models are low. This may be an effective way to know more accurately and efficiently when a feature is too small or too deep, or has too low of a contrast from host rocks, to be imaged. From a data management standpoint, another program might be devised to help manipulate the typically large datasets to be used in inversions. Some basic unofficial programs exist, but a formal one could be made. The program should be able to cut a specific range of data from a dataset that covers a larger area, and decimate data to get spacing to correlate with inversion cell sizes, perhaps allowing more dense data at the core and sparse data in outer regions. A formal program that reorganizes DC resistivity and chargeability into an inversion-friendly format would also save time. Regarding the inversion models, some may be rerun to test application of various combinations of constraints to get a more thorough idea of the range of results possible. There was limited testing of constraints for density, DC resistivity, and IP inversions, although prior information exists to expand on this. Additional constraints can be added to magnetic inversions. An example would be the use of the entire regional geology map, rather than just outcrop geology, to populate all surface cells with reference physical properties, or the construction of 3D domains based on large packages of similar rock, that can be assigned appropriate background reference values. Initially it was proposed that a 3D geologic model of the Hislop deposit would be created for use as a reference model for inversions, and for general comparison to inversion results and incorporation into common-earth models. The model was initiated, based on cross-sections drawn from select drillholes, however it was not completed. The process of building a 3D geological model requires significant time, reasonable 203 experience in GIS modeling, and a thorough understanding of geology. There was simply not enough information collected during this study to build anything but a very simple model that is extensively interpreted. There is, however, potential for a 3D model to be built for the Hislop deposit in the future, as there is a wealth of information from the many drillholes that were logged in this area, and now there are geophysical models which can help with geological interpretations at depth. The geological model must be completed with contribution from geologists that are well-familiarized with the geology and structure of the deposit The Hislop common-earth model can be further developed with the addition of a 3D geological model, and with the contribution of other existing data. Data from a large scale 3D model of the area created in the Fracsis GIS program by Geoinformatics Exploration Inc., including fault and geological contact surfaces, can be converted to forms usable in Gocad. A large quantity of drilling information, along with gold assays, and geochemical information collected by numerous workers throughout Hislop’s exploration history can be incorporated into the model for the purposes of mapping and target generation. At the start of this project lithogeochemical data was obtained with the anticipation that there may be relationships existing between this data and physical property data which would allow chemistry to be used to predict physical properties. Unfortunately, no statistically relevant trends emerged. Although geochemical data do not appear to be useful as a direct proxy for mineralogy or physical properties at Hislop, the collected lithogeochemical data might be beneficial to include in common earth models of Hislop. Anomalous abundances of elements reflecting carbonate, muscovite (or sericite), and albite-dominated alteration, such as C02, K, and Na, would act as a additional exploration criteria for querying along with geophysical inversion results. 5.5. FUTURE DIRECTIONS OF THE FIELD OF STUDY The field of geophysical inversion-based exploration is young. The inversion codes developed at UBC are constantly being updated and refined in order to allow more 204 flexibility with respect to incorporating geological information. They will continue to develop as they are being used to a greater extent in practice. The GIFtools ModelBuilder program of Williams (2008) is still in development. This program, or at least this type of program, will become a standard in the field that allows all prior geological and physical property knowledge to be input into inversions as constraints. The influence of the input data on the model is determined by the user based on the confidence the user has in the data. Additional programs to help input geological information into inversions, or make the results consistent with expected geology, are in progress. Diagonal dips and structural trends outside of north, south, east, and west directions can now be input using codes being developed by Lelievre et al. (2008). This is would be of use for inversions in any geologic setting, however could be especially useful in Archean greenstone terranes where there is commonly a strong structural fabric that should be relayed in the inversion. Smoothing inherent in inversions causes physical property values to grade between low and high anomalies. This may not be considered representative of the true geological or physical property situation. Phillips et al., (2007) initially introduced a method that restricts ranges of physical properties allowed to be taken up by model cells. This would be a useful tool where geology is simple, with only a few rock types present, and specific physical property ranges are expected. Lelievre et al. (2008) demonstrate how this application can be used. This technique might be constructively applied to inversions in greenstone belts where smoothing in inversion results can obscure important contacts between petrophysically distinct mafic and felsic units. The importance of physical property data collection is being increasingly recognized, especially in light of the need to use geophysics to explore for deeper mineral deposits. Large scale, publicly accessible physical property databases will become more common in the future, allowing geoscientists to cull physical property information from specific geographic areas, geologic regimes, and deposit types, to fortify geophysical work. A large data collection effort initiated by the Ontario Geological Survey (2001) in 205 the central Abitibi greenstone belt was mentioned in Chapter 2. A national physical property database is currently being compiled by the Geological Survey of Canada and Mira Geoscience Ltd. (Parsons and McGaughey, 2007). The best inversion results are generated when geologists and geophysicists collaborate on the problem. Geologists and geophysicists need to combine efforts to research or investigate physical properties in a given environment prior to inversion. Geologists can play a larger role in geophysical investigations, and will benefit the exploration effort by doing so. Geologists can provide insight when surveys are being designed, and can aid the inversion process by contributing prior geologic information including dominant structural fabrics, typical stratigraphic thicknesses, proportions and volumes of rock types or alteration present, and shapes and sizes of typical orebodies. Significant geologic information can be incorporated into inversions, by directly manipulating basic input parameters or with a complex reference model building program like that of Williams (2008). Recent collaborations between geologists and geophysicists for the greater understanding of a geological region took place during the Discover Abitibi Project. Greenstone architecture and mineral deposit settings were investigated indepth using a combination of geology and geophysics (Ayer et a!., 2005, Reed, 2005, Reed et al., 2005, Mueller et al., 2006). Similar types of collaborations are likely in the future. In order to have the greater community of geoscientists appreciate the benefits of collaboration between the two disciplines, case studies need to be presented in more general forums or as short courses that will attract members from both fields. 206 REFERENCES Ayer, J.A., Thurston, P.C., Bateman, R., Dube., B., Gibson., H.L., Hamilton, M.A., Hathway, B., Hocker., S.M., Houle, M.G., Hudak, G., Ispolatov, V.0., Lafrance, B., Lesher, C.M., MacDonald, P.J., Peloquin, A.S., Piercey, S.J., Reed., L.E., and Thompson, P.H., 2005, Overview of results from the Greenstone Architecture Project: Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6154, 146 p., 3 sheets. Lelievre, P., Oldenburg, D., and Williams, N., 2008, Constraining geophysical inversions with geologic information: Society of Exploration Geophysicists, 2008 Annual Meeting, Las Vegas, extended abstract, p. 1223-1227. Mueller, E.L., Reford, S.W., Dawson, D.J.W., Morrison, D.F., Pawluk, C., Grant, J., Spector, A., Rogers, D.S., and Savage, T., 2006, Acquisition, inversion and presentation of geophysical data for geoscientific profiles in the Timmins—Kirkland Lake area: Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6189 , 28 p., 15 sheets. Ontario Geological Survey 2001, Physical rock property data from the Physical Rock Property Study in the Timmins and Kirkland Lake Areas: Ontario Geological Survey, Miscellaneous Release — Data 91. Parsons, S., and McGaughey, J., 2007, Rock property database system: Proceedings of Exploration ‘07, Toronto, Ontario, p. 933-938. Phillips, N., Hickey, K., Lelievre, P., Mitchinson, D., Oldenburg, D., Pizarro, N., Shekhtman, R., Sterritt, V., Tosdal, D., and Williams, N., 2007, Applied strategies for the 3D integration of exploration data: KEGS Inversion Symposium, PDAC 2007, extended abstract, 9 p. 207 Reed, L. E., 2005, Gravity and magnetic three-dimensional (3D) modeling: Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6163, 40 P., 4 sheets. Reed, L. E., Snyder, D. B., and Salisbury, M. H., 2005, Two-dimensional (2D) reflection seismic surveying in the Timmins-Kirkland Lake area, northern Ontario; acquisition, processing, interpretation: Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6169, 68 p., 10 plates. Williams, N.C., 2008, Geologically-constrained UBC—GIF gravity and magnetic inversions with examples from the Agnew-Wiluna greenstone belt, Western Australia: Unpublished Ph.D. Thesis, The University of British Columbia, 479 p. 208 APPENDIX 2A - LIST OF ABBREVIATIONS Rock Type Minerals IF feisic intrusive ab albite IFP feldspar-phyric rhyolite dike act actinolite IQFP quartz-feldspar-phyric rhyolite dike al alunite II intermediate dike an anatase lix brecciated intermediate dike ank ankerite IM mafic dike ap(hy) hydroxyl apatite KMXmag brecciated K-fsp vein in magnetic mafic volcanic rock au augite L lamprophyric dike bt biotite ML)( multi-lithic breccia cal calcite QMX brecciated quartz vein in mafic volcanic rock clz clinozoisite QUX brecciated quartz vein in ultramafic volcanic rock dc clinochlore S syenite dike chi chlorite Seds sedimentary rocks dol dolomite T volcanic tuff ep epidote VM mafic volcanic rock Fe-cb Fe-carbonate VMX brecciated mafic volcanic rock Mg-cb Mg-carbonate VMmag magnetic mafic volcanic rock Fecb ankerite+dolomite+siderite VMXmag brecciated magnetic mafic volcanic rock hem hematite VMP pillowed mafic volcanic rock hbl hornblende VMPX brecciated pillowed mafic volcanic rock ksp potassium feldspar VU ultramafic volcanic rock mag magnetite VUX brecciated ultramafic volcanic rock mns magnesite mc(int) microcline (intermediate) mc(or) microcline(ordered) Alteration ms muscovite B carbonate+muscovite alteration (bleached) ms(Mg) muscovite (magnesium) B+P carbonate+muscovite+albite alteration mus(tot) total muscovite C chlorite or orthoclase CB chlorite+carbonate+sericite par pargasite CH chlorite+hematite pnt paragonite Cs chlorite+sericite per peridlase F carbonate+fuchsite alteration ph phlogopite FC Fe-carbonate alteration p1 plagioclase FC+H Fe-carbonate+hematite alteration py pyrite FC+H+S Fe-carbonate+hematite÷sericite alteration qtz quartz FC+Q Fe(Mg?)-carbonate+quartz alteration rut rutile FC+5 Fe(Mg?)-carbonate-’-muscovite/sericite alteration ser sericite H hematite alteration sid siderite S muscovite/sericite alteration sm smithsonite S+Q sericite+quartz alteration sp sphalerite T talc-chlorite metamorphic assemblage tc/tlc talc U generally unaltered wt witherite 209 APPENDIX 2B - HISLOP DRILLCORE LOGS, CROSS-SECTIONS, AND OUTCROP MAPS LDO — late diorite/dolerite SSG - greywacke SLO — mudstone - siltstone S00 — sediment, undivided IFD/lFO — felsic intrusive dykel felsic intrusive undivided 100 — intrusive, undivided I I ISO — syenite intrusive, undivided ______ VFO — felsic volcanic, rhyolite, rhyodacite I I VUO — ultramafic volcanic, undivided > VMF — magnetic mafic volcanic I I VMO — mafic volcanic, basalt, andesite Figure 2B.l. Hislop deposit area geology map (modified from Power et a!., 2004) showing locations of mined areas (West Area, and Shaft Area), ten driliholes logged for this study, five geologic cross-sections compiled from drill logs (1-5), and 2 outcrops (A and B). /1 Modified from Power et al., 2004 N SLO 500m 4 •H9711 I H9Th7H2 AnuwFauft LDO H9606 oP7V’ H9605 N 1i97ó8 (N)4EXT20 VMO\/ VMO oi GK 204 210 Hislop drill log legend. Alteration IJ Weak to moderate Fe-cb + ms L_i Strong Fe-cb + ms Fe-cb + ab Tic-chl metamorphic assemblage Mg-Cb (magnesite) + ms (fuchsite) Chlorite E Sericite Fe-cb + ab (intermediate dikes) Ms/ser (syenite and rhyolite dikes) Fe-cb (syenite and rhyolite dikes) Pink Fe-rich dol veins Epidote veins Hematite - pervasive Hematite along fractures Magnetite? n II H H II II Lithology Multi-lithic volcanic breccia Lamprophyric dike lritermediate-mafic dike Porphyritic rhyolite dike Syenite intrusive Mafic volcanic rock Ultramafic volcanic rockU Example of column layout: 750 B 750 _ r 775 775 800 1 2 80O Column 1: Lithology Column 2: Alteration Column 3: Magnetic Susceptibility (x103 SI Units) Column 4*: ________ Au grades between 0.15- 1 ppm Au grades between 1 - 5 ppm Au grades > 5 ppm *not all intervals sampled Fault Abbreviations: see first page of appendices 211 Loo for H9601 25 25 50 00 75 -75 100 -100 125 125 150 150 170 ç 175 200 200 225 225 250 250 275- 275 300 -- - 300 325_— 325 350 -350 ____/ -- 375 375 400 - 400 425 425 455 - 450 475 : 475 500 - 500 525 525 550 - 550 575” -575 - 4 600 ____ ____ - -605 625 625 0 Lo!p[ H9602 0 I 1 - 2oj.JL -2: 75 -100 125 -150 175 r - 200 -225 -250 275 -, 300 [I 325 H 350 400 425 450 b 475 13 500 700 725 750 800 > -i 0 32 — 650 675 700 725 750 775 800 212 LogforH96O60 25 55 75 100 125 150 175 250 225 280 275 350 / 626 375 / / 650 400 425 700 450 725- 475 / 750 500 1 775 525 1 850 650 1 575 1 0 Lo for H9605 254V 25 50 50 /A 100____ 100 125 - 125 150- -150 175 -/ 175 200 200 225 225 250 250 275 275 305 300 0 153 0 25 50• 75- 100 - 125 150 175 200 225 250 275 300 325 350 400: 425 450 - 525 575 - 600 - 625 :650 675 700 725 750 775 :555 0 149 213 o Log for H9707 Lo9 for H9708 25 —25 25 -25 50 L50 50 .% ‘- 0 75 -75 75- — ) 75 100 —100 100 100 125 —125 125 125 150 —150 150 150 175 _ =175 170 175 r 200 200 200 - 200 225 —225 225 225 250 — 250 250 ; 250 275 -- 275 — 275 ‘- 275 300 300 300 ) 300 325 --325 - — 325 325 350 :350 350 .- 350 375 -375 S 400 - - 400 - 400 400 425 — 425 425 = 425 625 - 625 F _ :: ::: > 550 50 = 575 75 600 —600 600- 214 250 275 300 325 350 375 400 - for EXT 280 250 -275 -300 325 375 400 0 49 - 0- 25- 50- 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450 475 500 525 550 Log for H9711 / 0 169 -0 :25 •50 75 100 125 150 -175 :225 :250 -275 300 325 300 375 -400 -425 450 -475 525 550 215 Legend applies to all following cross-sections Multi-lithic Volcanic Breccia Lamprophyric Dike Intermediate Dike Porphyritic Rhyolite Dike Syenite Intrusive J Fe-poor Mafic Volcanic Rock Fe-rich Mafic Volcanic Rock Ultramafic Volcanic Rock . Fault .. — . Drill trace 216 k) Cross-section 2. s-- I r 100 mRl. Cross-section 3. DDH Ext 280, GK 280, and H9605 218 4. L’J c. KI 5Omj cCross-SE sw-I 200 mRL 100 mRL DDH H9708 DDH H9707 / -200 mRi l.1øJ C Outcrop map A. 47to643 (4) 167 to347 (1) 35to167 (12) • 4to 35 (7) • Oto 4(30) Magnetic susceptibility Jd3-3-Skintt3 — Carbonate altered mafic volcanic rock Syenite Darker colors represent outcrop + Outcrop map B. • Overburden Unaltered variolitic mafic volcanic •107b0 173 (7) 5Oto 107 (24) Bleached and oxidized variolitic mafic volcanic ot 50 (14) • 8to 20 (16) • Oto 8 (33) Magnetic susceptibility (xl 0-3 SI units) 221 APPENDIX 2C - DETAILED AND EXPANDED METHODS XRD Analysis - Reitveld analysis The standardless Rietveld refinement method was used to determine mineral abundances for Hislop samples. Samples were prepared and run, and data was analyzed, by Elizabetta Pani at the University of British Columbia. X-ray diffraction (XRD) analyses are first run on powdered bulk rock samples. The sample must first be ground such that particle sizes are <10pm to avoid inaccurate diffraction peak intensities and preferred orientation of grains. Samples are ground in ethanol using a McCrone Micronising Mill with corundum elements. The sample is placed into a back-loading mount. The top of the mount is fit with a textured glass to minimize preferred orientation on the surface of the sample, and is removed before analysis. A modified razor blade may be used additionally to create texture in the top of the sample. Standard X-ray diffraction patterns are collected for samples using a Siemens D5000 diffractometer. X-ray diffraction data are collected in increments of 0.04°, from 3° to 700 29. The counting time is 2 seconds/step, and CuKa radiation is used. The diffractometer used includes a diffracted-beam monochromator, 10 divergence and anti-scatter slits, a 0.6 mm receiving slit, and an incident-beam Soller slit which was removed. A long-fine-focus Cu X-ray tube is used and operated at 40kV and 4OmA, with a take-off angle of 6°. The mineral phases are determined using conventional search-match procedures. The XRD data are analyzed by Rietveld refinement using the program Topas 2.0 (Bruker AXS 2000). For this method, information regarding the crystal structure of all detected phases is used to calculate a diffraction pattern for each phase present. These patterns are summed and then fitted to the collected diffraction pattern using a least squares refinement. Numerous parameters are considered in the refinement including a series of global parameters (e.g. background, radiation wavelength, correction for the monochromator crystal), and mineral phase-dependant parameters (e.g. atomic coordinates, size and shape of the unit cell, site-occupancy). Using Rietveld methods, the relative masses for each phase can be calculated by considering the scaling factor 222 determined when observed and calculated data were being fit, the number of formula units per unit cell, the mass of the formula unit, and the volume of the unit cell. Detailed methods are found in Raudsepp and Pani (2003). Magnetic susceptibility corrections Core diameter corrections An Exploranium KT-9 Kappameter was used to collect magnetic susceptibility data from 3.6 cm diameter drill core. Since the meter can only be set to take measurements from drill core with a diameter that is a whole number, the KT-9 was set to take readings for 4 cm drill core, the closest whole number to the diameter of the drill core being used. Through some experimentation, it was determined how to correct the magnetic susceptibility values for 3.6 cm from the 4 cm diameter field data. Susceptibility readings were taken along a few samples of whole core at particular data points (16 different points, Fig. 2C.1) with the meter set to various diameter settings. After this, the data was plotted to try and find a relationship between the magnetic susceptibility values and changes in diameter. First, diameter was plotted against magnetic susceptibility for select sample points to determine a relationship (for an example of this see Fig. 2C.2; for the full experimental dataset, see the spreadsheet labeled Appendix 2C on accompanying CD). It was noted that this relationship changes with variations in magnetic susceptibility between the different point locations measured; the relationship takes the form y = mx, but a and m are different depending on the susceptibility at a given point. Coefficients a and m calculated from select sample points on four different samples were plotted against susceptibility measured on the 4 cm diameter setting (Figs. 2C.3 and 2C.4). From this it was determined that there were linear relationships between m and magnetic susceptibility and a and magnetic susceptibility. It was concluded that readings at 4 cm can be plugged into the equations m = 3.2295x - 33.133 and a = -0.0004x - 0.6506 (where x = magnetic susceptibility at 4 cm) to get m and a, and then m, a and the diameter we want to correct to (3.6 cm) can be input back into the 223 equation y = mx, where x is the new magnetic susceptibility (at 3.6 cm). This correction was applied to all of the whole core data collected in the field. Split core corrections Some of the drill core was split lengthwise for sampling purposes, and only one half of these particular intervals was available to test. The KT-9 meter setting was kept on 4 cm diameter for these intervals. It was noticed that there existed discrepancies between the magnetic susceptibilities of whole and split core of similar rock types and it was necessary to correct for this. Similar experiments to the core diameter tests were completed, and susceptibility readings taken along whole core at particular data points (at various diameters), after which the core was split lengthwise using a rock saw and readings taken along the split pieces at the same designated data points. The data collected at each point was compared between the different diameters (this can be seen in spreadsheet 4 in Appendix 2C on the accompanying CD) with the anticipation that the change in values between whole and split core was simply proportional. For each sample the average ratio between whole core and split core values is consistent between various diameter settings. However, from one sample to the next (samples range from felsic rocks which have the lowest magnetic susceptibilities, to ultramafic rocks with higher magnetic susceptibilities) the ratio changes slightly (ranging from about 0.83 to 0.89). Low magnetic susceptibility samples do not consistently change by a different ratio than do high magnetic susceptibility samples when split. From spreadsheet 4, an attempt was made to determine if there was a relationship between susceptibility and this ratio. There appears to be a weak trend that indicates that a higher ratio can be used to correct for low magnetic susceptibility samples while a lower ratio can be used to correct for higher magnetic susceptibility samples (Fig. 2C.5). But the trend is not good, and appears to break down at low magnetic susceptibilities where there appears to be a broad range in “correction factors”. The average ratio was considered to be 0.85, and thus a correction factor of 1.15 was applied to split core samples to get approximate whole core susceptibilities. 224 Figure 2C. 1. Data points along a whole piece of drill core. Magnetic susceptibility readings were taken at each of 16 data points at various diameter settings (2.54, 3, 4, and 5 cm). 225 H-97-07 456m 350 300 - .. — -0.7249 - ________________ E — -0.7101y — 452.02x —Power (d5) 100 ---- — —Power (d7) y = 451 .85x°7298 —Power (d9) 50 •-. 0 0 1 2 3 4 5 6 diameter Figure 2C.2. Readings for five sample points (data points dl, d3, d5, d7, d9) for sample H9707-456 with meter set at 2.52, 3, 4, and 5 cm diameters. For a particular sample, there is not a distinct relationship between susceptibility and diameter, i.e., there is not simply one equation. The equation (y = mxj changes with variations in magnetic susceptibility between the different points tested. 226 Figure 2C.3. Relationship between susceptibility measured on the 4 cm diameter setting, and m from equation 1, for select points from four different samples. The different samples are obvious as the four clusters of data occurring within narrow susceptibility ranges. E • Series I I — Linear (Series I) mag sus 227 0Figure 2C.4. Relationship between susceptibility measured on the 4 cm diameter setting, and a from equation 1, for select points from four different samples. The different samples are obvious as the four clusters of data occurring within narrow susceptibility ranges. 40100 200 300 400 -0.1 y= -0.0004x- 0.6506 • Seriesi — Linear (Series 1) -0.9 mag sus 228 0.87 0 0.86 . 0.85 0 0.84 sus (of split core) Figure 2C.5. Possible relationship between the susceptibility of split core and the ratio between split core and whole core values. Relationship seems to break down at low susceptibilities. 0.90 0.89 0.88 1 ‘1 1 ‘1 I y = -9E-05x + 0.8583 • Seriesi —Linear 0.83 0.82 0.81 0 100 200 300 400 500 229 Comparison of hydrostatic and geometric calculations of density Density measurements for Hislop samples were made using the hydrostatic method described in Chapter 2, section 4.2.2. To test if this method was generating reliable density measurements, a geometric method was applied to select samples for comparison. Density values are determined by dividing the mass of the sample by its measured volume. The volume of the sample was determined by measuring the length and diameter of a piece of whole drill core using a Mitutoyo caliper (l2in!300mm). The drill core was first cut as evenly as possible on each end. Multiple measurements were made of the diameter and length to account for slight irregularities, and the average value was used to calculate the samples volume. Figure 2C.6 shows a plot of density measurements made by the geometric method versus density measurements made using the hydrostatic method. A strong correlation indicates values are accurate. Figure 2C.6. Density data calculated using the hydrostatic method versus the geometric method. U 0) (C C C) •0 U E 0 C) 0 2.95 2.9 2.85 2.8 2.75 2.75 2.80 2.85 2.90 2.95 Hydrostatic density (glcm3) 230 APPENDIX 2D - X-RAY DIFFRACTION ANALYSES _____ dol ep Fecb hem hbl ksp maci mns 20.7 18.7 0.5 14.9 22.4 2.1 4.0 28.0 18.7 1.1 46.8 33.2 1.0 0.9 0.4 3.7 5.1 3.2 Sample Rock Altn act al an ank an(hv’ au bt cal clz jjfl H9604-57 VMP U 37.1 0.8 14.7 17.3 4THD8 VMP U 27.1 30.7 10.7 22.0 4THDIO4 VMP U 38.4 4.2 23.4 15.5 12.2 H9601-439 VM B 44.0 5.2 H9601-122 VM U 49.6 12.5 H9604-28 VM U 27.6 1.0 0.5 11.1 7.2 30.6 H9604-122 VM B 3.1 0.5 14.9 H9605-111 VMmag U 8.0 2.0 H9707-373 VMmag U 34.9 29.6 8.3 1.9 0.9 10.8 4THDII7 VMmaq U 41.7 0.6 2.1 7.1 7.1 2.1 4THDIOO VMmag U 36.4 15.6 7.7 13.8 4THDII6 VMmag B 34.3 16.1 16.1 4THDII4 VMmag U 16.2 19.1 21.7 15.6 3THDI5 VMmag U 35.6 13.8 21.4 6.8 3.2 7.9 H9711-281 VU B 2.6 16.7 22.9 22.9 H9606-179 VU F 2.0 35.0 19.0 19.0 H9601-200 VU T 4.0 3.3 11.7 42.1 6.5 6.5 H9601-252 VU U 36.6 23.0 23.0 H9604-379 VU T 10.4 49.7 5.2 5.2 H9601-396.5VU B 1.2 5.4 20.6 20.6 3THDI VU B 17.3 4.0 6.1 20.5 20.5 H9601-410 S U 59.3 0.8 2.2 2.2 H9605-176 S FC 63.0 0.4 4THDII5 S FC+S 37.2 0.5 0.5 H9601-417 S B H9605-210 S FC+S 59.4 8.1 0.6 8.1 16.2 3THD6 S U 48.5 1.0 1.0 H9601-406 S S 58.6 5.2 5.2 H9601-298 IQFP U 64.8 0.4 1.2 1.2 H9604-214 IFP 5 54.0 3.2 2.2 3.2 6.4 H9601-322 IQFP U 62.8 5.7 5.7 H9601-302 IQFP U 59.8 0.8 0.8 H9707-137 IM U 32.2 1.7 32.8 11.6 11.6 H9606-66 II F 2.5 21.1 44.2 H9606-230 II FC 33.0 20.2 24.3 H9606-173 II B+P 45.8 14.2 14.2 H9604-444 L U 14.3 20.7 25.4 25.6 6.0 8.2 2.5 1.8 4.6 2.9 44.2 0.6 31.2 3.6 3.1 1.5 0.6 Sample Rock H9604-57 VMP 4THD8 VMP 4THDIO4 VMP H9601-439 VM B H9601-122 VM U H9604-28 VM U H9604-122 VM B H9605-111 VMmag U H9707-373 VMmag U 4THDII7 VMmag U 4THDIOO VMmag U 4THDII6 VMmag B 4THDII4 VMmag U 3THDI5 VMmag U H9711-281 VU B H9606-179 VU F H9601 -200 VU T H9601-252 VU U H9604-379 VU T H9601 -396.5 VU B 3THDI VU H9601-410 S H9605-176 S 4THDII5 S H9601-417 S H9605-210 S 3THD6 S H9601-406 S H9601-298 IQFP H9604-214 IFP H9601 -322 IQFP U H9601-302 IQFP H9707-137 IM H9606-66 II H9606-230 II H9606-173 II H9604-444 L Altn U U U U U U F FC B+P 9.9 4.9 1.0 14.4 3.3 2.7 3.1 22.8 4.8 32.9 18.4 17.3 23.7 38.0 35.0 6.7 1.8 19.1 9.3 33.5 1.4 32.9 17.3 38.0 13.1 48.0 8.3 11.1 23.3 15.4 5.8 28.6 17.7 15.3 6.2 qtz 23.6 3.8 2.0 0.9 11.1 1.0 7.2 11.2 1.3 19.4 34.0 1.7 17.0 0.5 19.7 10.8 2.4 34.0 1.6 7.9 27.3 15.1 0.5 30.5 29.2 17.3 25.6 27.1 1.4 1.2 1.3 3.5 0.9 0.6 43.0 1.1 7.9 2.9 1.2 1.2 1.7 11.2 0.3 26.2 0.3 25.0 1.0 30.4 0.5 30.6 1.1 6.5 21.0 9.3 4.6 7.4 2.9 15.1 _______ par pgt per ph(1M) p1 py rut sid sm sp tc(1A) wI 1.3 1.8 mc(int)mc(ord) ms(2M1) ms(1M,Mg) mus(tot) or 3.5 3.5 4.7 4.7 4.2 6.7 1.8 2.6 0.9 3.1 33.5 24.0 22.8 4.8 3.1 B U FC FC+S B FC÷S U S U S 0.3 1.1 12.6 8.4 25.6 17.6 19.5 38.3 6.7 8.4 2.3 12.7 6.9 0.4 5.8 0.4 23.1 4.1 0.3 1.6 14.2 21.4 APPENDIX 2E - PHYSICAL PROPERTIES OF HISLOP DEPOSIT ROCKS Sample No. HolelD From To Rock type Gr. Altn notes MS Den Chrg Res Por size (x104SI) (glcm3) (ms) (Ohm-rn) (%) FC 2.793THDIO hndsmp IM equigranular 4.15 3THDII hndsmp VU sheared 24.80 3THD12 hndsmp VUX B “carbonate breccia” 0.96 3THD13 hndsmp Vmmag? U 17.20 3THD14 hndsmp VMmag? U 50.40 3THD15 hndsmp VMmag? U 88.80 3THDI6 hndsmp VMmag? U 82.30 3THDI7 hndsmp VMmag? B+P north of main mine syenite 12.40 3THDI8 hndsmp S FC+S 0.27 3THDI9 hndsmp Vmmag U 94.70 3THDIA hndsmp VU B breccia; Fe-carbonate and qtz matrix, 11.20 disseminated py 3THD2 hndsmp S FC+S 0.16 3THD2O hndsmp Vmmag? U 41.60 3THD21 hndsmp VMP U pillow tops to NE (060), 10cm to 0.5 m, 15.60 some disseminated py 3THD21 hndsmp VMP disseminated py 15.60 3THD3 hndsmp IM FC equigranular; disseminated py and cpy 0.86 3THD4 hndsmp L massive 12.40 3THD5 hndsmp IM FC syenite “veins”, cut by Fe-carbonate veins 6.21 2.73 3THD6 hndsmp S U 0.17 2.64 3THD7 hndsmp VMmag? U syenite “veins”; disseminated py and cpy 203.00 3.07 3THD9 hndsmp IF FC carb and disseminated py/cpy fills fractures 0.41 2.67 4THDIO hndsmp VM f U 0.97 2.92 4THD100 hndsmp VMmag U 1.14 3.00 4THD1OI hndsmp VMP f U pillows 20cm to Im; chlorite selvedges; tops 0.55 2.94 to 075 4THDIO2 hndsmp 3.06 4THD1O3 hndsmp VM 2.89 4THD1O4 hndsmp VMP U epidote in selvedges 2.98 4THD1O5 hndsmp VMP f U epidote in selvedges 2.97 4THDIO6 hndsmp VU F Royal Oak pit? 0.62 2.95 4THDIO7 hndsmp VM B Royal Oak pit? 0.37 2.88 4THDIO8 hndsmp VMP U Royal Oak pit? 0.52 2.82 3.01 3.01 2.94 2.98 20.87 2548.9 2.87 2.79 11.17 22052 2.78 9.83 4757.6 2.67 10.57 5852.6 2.89 2.77 13.40 6507.1 15.17 5034.1 6.63 10687 0.28 8.50 33638 0.23 IM U U 33.00 69.50 12.80 alteration along fractures size (x1O3SI) hndsmp Seds U Porcupine sediments Sample No. HolelD From To Rock type Gr. Altn notes MS Chrg Res Por (ms) (Ohm-rn) (%) 9.73 3478.5 0.54 Den (cm) 4THDIO9 0.21 2.73 4THD11O hndsmp Seds U Porcupine sediments 0.28 2.76 4THD111 hndsmp IM c U latedike? 35.10 3.06 36.60 16472 4THD112 hndsmp IM c U latedike? 26.20 3.04 4THD1I3 hndsmp VMX U 0.43 2.84 4THDI14 hndsmp Vmmag U 74.10 2.92 37.73 58754 0.13 4THD115 hndsmp S c FC+S 0.20 2.69 6.27 3343.8 4THDII6 hndsmp VMMag B 0.15 2.81 7.57 2130.9 4THD1I7 hndsmp VMMag U 54.30 2.78 28.18 3075 4THD2 hndsmp VMP c U 0.68 2.91 4THD3 hndsmp VU B similar to rock from Royal Oak pit; Fe-carb 1.44 2.88 alteration along fractures 4THD4 hndsmp II U 0.24 2.80 4THD5 hndsmp VM U disseminated py 0.35 2.76 4THD6 hndsmp VM B similar to rock from Royal Oak pit; Fe-carb 0.38 2.81 4THD7 hndsmp VMP U variolitic, flow-banded pillow basalt; 0.72 2.90 coalescing varioles 4THD8 hndsmp VMP U pillows 30 cm; epidote in selvedges 0.88 2.99 3.38 42488 0.57 4THD9 hndsmp VM U epidote in veins 32.40 2.92 Ext280-258 EXT 280 258.25 258.45 VU f T sheared, fractured; qtzlFe-carb vein 21.06 2.84 Ext280-274 EXT 280 274.25 274.5 VU f T rare veins 0.55 2.86 Ext280-275 EXT 280 275 275.2 VU f T in-situ fragmented and sheared 27.76 2.86 7.80 283.3 1.49 Ext280-287 EXT 280 287.45 287.65 VU f T massive to sheared, abundant qtz 26.90 2.87 amygdules, <1mm Ext280-306 EXT 280 306.65 306.9 IFP f S 20% fsp phenocrysts average <1 to 1mm 0.16 2.66 Ext280-308 EXT28O 308 308.2 IFP f S phenocrystsdiffuse 0.12 2.65 Ext280-319 EXT 280 319.3 319.55 IFP f U unaltered phenocrysts clearly visible 0.16 2.64 Ext280-341 EXT28O 341.2 341.45 IFP f B 0.16 2.61 Ext280-353 EXT 280 353.3 353.55 IFP f B 0.13 2.66 Ext280-358 EXT 280 358.2 358.4 II f U sharp contacts 0.39 2.86 Ext280-365 EXT 280 365.2 365.4 VMmag m B 1.28 2.67 Ext280-365.5 EXT 280 365.45 365.7 IFP f S 0.12 2.88 Ext280-372 EXT 280 372.3 372.55 VMmag m B 0.55 2.89 Ext280-375 EXT 280 375.65 375.85 II f H 3.29 2.84L’J Ext280-388 EXT 280 387.75 388 Ext280-401 EXT 280 401.45 401.7 Ext280-402 EXT 280 402.1 402.3 VUX 2mm to 4cm Ext280-406 EXT 280 406 406.25 II f Ext280-407 EXT 280 407.4 407.6 II f Ext280-410 EXT 280 410.45 410.7 II f Ext280-411 EXT28O 411.35 411.55 II f GK204-101 GK204 101.50 101.8 S c GK204-109 GK204 108.97 109.27 S c GK204-116 GK204 116.13 116.43 S c GK204-128 GK204 127.71 128.02 VMmag f GK204-129 GK204 129.08 129.39 MLX GK204-168 GK204 168.71 168.86 VU f GK204-25 GK204 24.99 25.3 VMmag m GK204-60 GK204 59.74 60.05 II m GK204-74 GK204 73.80 74.1 IF c GK204-76 GK204 75.90 76.2 IF c FC+S H hornblend phenocrysts, 1-3mm, 5-10% S FC+S C FC+S FC+S B breccia B+P mylonite? T U equigranular U S FC+Q 0.14 2.68 3.30 2.79 0.31 2.76 0.13 2.69 0.07 2.68 0.14 2.70 0.18 2.74 0.41 2.86 0.53 2.94 14.12 2.81 182.35 3.00 67.50 2.79 0.30 2.67 0.24 2.67 GK204-84 GK204 84.58 84.89 VMmag m GK204-93 GK204 92.96 93.27 vMmag m GK204-97 GK204 97.08 97.38 S c H9601-107 H9601 106.85 107.05 II m H9601-122 H9601 121.95 122.2 VM m H9601-126 H9601 126.5 126.8 VU m H9601-132 H9601 132.4 132.6 VU m H9601-133 H9601 133.55 133.75 IM m H9601-135 H9601 135.6 135.8 IM m H9601-150 H9601 150.4 150.6 vu m H9601-156 H9601 156.4 156.6 vu m H9601-164 H9601 164 164.2 vu m H9601-172 H9601 171.8 172 IM c H9601-186 H9601 186.55 186.75 IQFP f H9601-200 H9601 200.65 200.85 VU m H9601-210 H9601 210.45 210.6 IQFP f C B FC+S sulfides U U u spinifex T U U T spinifex T U FC u crowded porphyry T spinifex 67.20 2.83 2.06 2.85 0.14 2.84 0.70 2.94 18.70 2.86 158.17 5400 0.63 0.71 2.83 1.56 2.81 3.20 2.79 53.60 2.75 21.60 2.85 52.60 2.85 12.50 2.87 11.20 2.77 2.45 2.67 20.40 5969 4.42 2.82 20.40 1422.4 1.13 Sample No. HolelD From To Rock type Gr. AItn notes MS Den Chrg Res Por size (x103 SI) (qlcm3) (ms) (Ohm-m) (%) VU f T sheared; qtz and Fe-carb veins 2.77 2.84 f H 0.49 2.85 T sheared; qtz and hem(?) altered fragments, 1.18 2.85 B crowded porphyry 0.18 2.75 H9601-229 H9601 228.8 229.05 IQFP H9601-24 H9601 24.7 24.9 VU H9601-241 H9601 241.25 241.45 IQFP H9601-252 H9601 252.6 252.8 VU H9601-258 H9601 258.4 258.6 MLX H9601-266 H9601 266.05 266.25 IQFP H9601-286 H9601 286.4 286.6 QUX H9601-294 H9601 294.6 294.85 VUX H9601-298 H9601 298.2 298.4 IQFP H9601-302 H9601 302.1 302.2 IQFP H9601-308 H9601 308.7 308.9 QUX H9601-314 H9601 314.6 314.8 VU H9601-32 H9601 32.75 32.95 VU H9601-320 H9601 320.6 320.8 IQFP H9601-322 H9601 322 322.2 IQFP H9601-324 H9601 323.9 324.1 VUX equigranular crowded porphyry m U equigranular B chaotic, multi-lithic breccia f FC weakly qtz and fsp porphyritic B ultramafic breccia, quartz clasts? f T brecciated, Fe-carb matrix f U 1.23 ppm Au from core box qtz fragment breccia; 2.69 ppm Au 5.16 ppm Au from core box 3.15 ppm Au from core box qtz veins 0.57 2.89 0.16 2.70 2.24 2.84 5.25 1062.5 0.72 16.30 2.80 0.28 2.73 0.65 2.85 4.71 2.86 0.06 2.64 2.23 7856 0.05 2.63 2.53 8976 1.93 2.85 13.76 2.87 0.61 2.85 0.06 2.64 2.90 9580 0.07 2.65 4.20 23970 H9601-331 H9601 331.4 331.6 VU H9601-338 H9601 338.1 338.3 II H9601-361 H9601 361 361.2 VU H9601-371 H9601 371.2 371.5 VU H9601-383 H9601 383 383.15 II H9601-393 H9601 392.85 393.05 VU f H9601-396 H9601 395.95 396.15 VM f H9601-396.5 H9601 396.6 396.8 VU f H9601-405 H9601 405 405.2 S H9601-406 H9601 405.9 406.1 S H9601-410 H9601 410.4 410.6 S H9601-417 H9601 417.3 417.5 S H9601-419 H9601 419 419.2 VMmag f H9601-422 H9601 422.3 422.5 VMmag f H9601-43 H9601 43.3 43.55 VUX H9601-436 H9601 435.9 436.1 VMmag f H9601-439 H9601 439.1 439.4 VM f m T equigranular f FC m T m U spinifex f FC hbl and fsp phenocrysts T B+P fractured B fractured FC-S 2.13 ppm Au from core box FC-S 5.89 ppm Au from core box U B B+P B B U B strong mag sus contrast between bleached 0.48 2.79 0.26 2.67 21.41 2.87 0.74 2.87 4.33 698.39 0.51 107.00 2.86 0.44 2.82 0.28 2.77 0.78 2.92 2.31 2.86 27.40 7872.4 0.18 2.73 0.16 2.70 25.80 9400.2 0.12 2.74 15.67 5703.6 0.46 2.82 2.12 2.93 0.41 2.91 60.70 2.86 ?? 33.23 49812 Sample No. HolelD From To Rock type Gr. AItn notes MS Den Chrg Res Por size (x103 SI) (gicm3) (ma) (Ohm-rn) (%) f B crowded porphyry 0.20 2.73 m U f U f U B f T m U f U f U T sheared, brecciated to massive; Fe-carb and 11.07 2.84 and unbleached areas H9601 -474 H9601 H9601-491 H9601 H9601-496 H9601 H9601 -507 H9601 H9601-516 H9601 H9601-529 H9601 H9601-541 H9601 H9601-55 H9601 H9601 -551 H9601 H9601 -571 H9601 H9601 -581 H9601 H9601 -600 H9601 H9601 -614 H9601 H9601-79 H9601 H9601-95 H9601 H9602-103 9602 H9602-119 9602 H9602-141 9602 H9602-143 9602 H9602-162 9602 H9602-188 9602 H9602-203 9602 H9602-217 9602 H9602-234 9602 H9602-242 9602 H9602-247 9602 H9602-254 9602 H9602-272 9602 H9602-292 9602 H9602-295 9602 H9602-303 9602 H9602-320 9602 H9602-321 9602 474.35 474.55 VMmag f 491.45 491.65 VMmag f B 496.3 496.5 VMXmag m B+P 507.4 507.6 VMPX m B+P 516.6 516.8 VMmag m B 529 529.2 VMmag m U 541.6 541.8 VMmag m B 54.95 55.15 VMP 551.55 551.7 VM 614 614.2 II f U 78.75 79 VMP f U 95.4 95.6 VMP f U 103.05 103.25 VMP f U 119.7 119.9 VM m U 141.15 141.35 VMmag m FC sulfides 143.2 143.4 VMmag m U 161.8 162 IM c U 188.6 188.9 VMP f U 203.25 203.45 VMmag m U 217.2 217.45 VU f T 234.6 234.8 VMmag f U 241.95 242.2 IFP f S 247.75 247.95 vMxmag f B 254.2 254.4 S c C+H sulfides 271.85 272.05 VU T 292.15 292.4 VU 295.7 295.9 vu 303.25 303.45 vu 320.1 320.35 vu 321.7 321.9 IM f C 0.56 2.85 0.56 2.86 0.51 2.81 0.55 2.83 1.54 2.79 0.48 2.80 0.66 2.77 0.76 2.88 1.06 2.75 13.90 2.93 0.41 2.72 15.40 16160 0.57 2.79 0.72 2.83 1.74 2.96 3.95 2.90 27.80 2.86 83.70 2.84 37.88 2.97 23.20 2.94 0.65 2.93 25.80 2.97 52.10 2.85 14.59 2.71 132.94 2.87 0.16 2.71 52.70 2.87 1.01 2.94 0.52 2.80 0.73 2.97 0.54 2.98 0.54 2.81 Sample No. HolelD From To Rock type Gr. AItn notes MS Den Chrg Res Por size (x103 SI) (glcm3) (ms) (Ohm-rn) (%) B 0.51 2.82 571.2 571.4 VMXmag 581.25 581.55 II 599.9 600.1 vu f U mafic phenos m B+P pink rhodachrosite or dolomite carbonate veining? B f U m T L’J m T C T f T T sulfides spinifex H9602-326 9602 H9602-343 9602 H9602-364 9602 H9602-450 9602 H9602-47 9602 H9602-476 9602 H9602-479 9602 H9602-498 9602 H9602-56 9602 H9602-73 9602 H9604-105 H9604 H9604-108 H9604 H9604-113 H9604 H9604-117 H9604 H9604-122 H9604 H9604-126 H9604 H9604-14 H9604 H9604-150 H9604 H9604-173 H9604 H9604-182 H9604 H9604-190 H9604 H9604-201 H9604 H9604-204 H9604 H9604-205 H9604 H9604-214 H9604 H9604-229 H9604 H9604-232 H9604 H9604-260 H9604 H9604-279 H9604 H9604-28 H9604 H9604-292 H9604 H9604-303 H9604 H9604-312 H9604 H9604-327 H9604 326.4 326.6 VU 343.6 343.8 IM 363.8 364 vu 450.25 450.45 VU 47.1 47.35 VM 476.35 476.55 IFP 479.45 479.65 vu 498.4 498.6 VMP 56 56.25 VM 73.75 73.95 VMP 104.95 105.15 VM 108.6 108.8 VM 113.15 113.3 VM 117.6 117.8 VM 122.7 122.9 vM 126.3 126.45 VM 14.4 14.6 VM 150.6 150.85 VM 173.15 173.25 VU 181.8 182 VM 189.9 190.1 VM 201.05 201.25 vu 204.2 204.35 IFP 205.1 205.3 IFP 214.2 214.4 IFP 229.4 229.65 IFP 232.45 232.65 IFP 260.7 260.9 IFP 279.3 279.5 vu 28.25 28.45 VM 291.85 292.05 IFP 303.5 303.7 IFP 312.3 312.5 vu 327 327.2 S U hbl phenocrysts m T T m LI epidote f B C T f u u f U f B B f B varioles? f B varioles? m U f B f F f U f B f T strongly sheared f S f U f S f S f S f U f T f Li f U f S+Q C U 85.90 2.79 81.10 2.76 80.10 2.80 6.99 2.87 0.08 2.65 0.52 2.77 0.55 2.87 1.63 2.92 69.90 2.92 0.44 2.83 0.51 2.84 3.67 2.82 6.10 2.95 0.89 2.95 39.40 2.87 0.60 2.88 0.49 2.88 0.57 2.80 0.60 2.88 18.35 2.87 0.08 2.57 0.13 0.09 2.64 0.14 0.18 2.67 71.80 2.82 6.54 2.88 0.21 2.71 2.18 2.61 109.00 2.87 0.55 2.65 4.30 405.27 N/A >70000 0.25 Sample No. HolelD From To Rock type Gr. Altn notes MS Den Chrg Res Por size (x1O3 SI) (glcm3) (ms) (Ohm-rn) (%) f C 0.55 2.79 f f f B B sulfide-filled amygdules? varioles? m icrofragmental 2.97 3994.3 0.40 fragmental, angular fragments 0.67 2.93 5.73 18578 0.63 2.07 541.24 1.54 10.70 11357 0.13 2.68 10.70 14759 sheared dark black mineral (?) fills amygdules m T mottled texture, patches of black minerals Go Sample No. HolelD From To Rock type Gr. Altn notes MS Den Chrg Res Por size (x104 SI) (glcm3) (ms) (Ohm-rn) (%) c U 0.75 2.84H9604-339 H9604 338.9 339.1 IF H9604-357 H9604 357.1 357.3 VU f T H9604-379 H9604 379 379.25 VU f T H9604-405 H9604 405.7 405.9 II f U H9604-407 H9604 407.65 407.85 L U H9604-414 H9604 414.45 414.65 II m U H9604-425 H9604 425.5 425.7 S c FC H9604-428 H9604 428.3 428.5 S c U H9604-433 H9604 432.95 433.15 S c S H9604-442 H9604 441.85 442.05 S c S H9604-444 H9604 444.3 444.5 L U H9604-447 H9604 447 447.25 S c FC H9604-475 H9604 475.65 475.85 S c S H9604-504 H9604 504.5 504.7 S c S+Q H9604-512 H9604 512 512.2 II FC H9604-517 H9604 517.55 517.75 VMmag f B H9604-532 H9604 531.8 532 S c S H9604-545 H9604 545.5 545.7 VMmag f U H9604-555 H9604 555.05 555.25 VMmag f U H9604-57 H9604 57.7 57.9 VMP f U H9604-574 H9604 574.2 574.4 II m U H9604-585 H9604 585.45 585.65 VMmag f B H9604-595 H9604 595.6 595.8 VMmag f B H9604-603 H9604 603 603.25 VMmag f B H9604-609 H9604 609.6 609.8 VMmag f B+P H9604-615 H9604 615.45 615.65 VMmag f H H9604-625 H9604 625.25 625.45 VMmag m U H9604-635 H9604 635.5 635.75 VMmag m U H9604-673 H9604 672.75 673 VMmag m U H9604-702 H9604 702.7 702.95 VMmag f U H9604-716 H9604 716.7 716.9 VMmag f C+B H9604-719 H9604 719 719.2 VMmag f B H9604-727 H9604 727.6 727.8 IFP f H9604-738 H9604 738.6 738.8 II f FC some brecciation 84.40 2.83 3.37 197.89 2.09 some brecciation 41.20 2.74 4.18 111.16 3.09 hbl and fsp phenocrysts 69.41 2.81 49.30 2.87 67.65 2.95 0.25 2.67 0.20 2.67 0.20 2.70 0.15 2.68 0.93 2.69 15.58 19664 0.26 2.59 intensely altered 0.13 2.67 0.18 2.71 15.67 7245.2 1.94 2.69 98.82 2.89 0.42 2.69 106.00 2.78 mafic phenocrysts 140.00 amygs increase at pillow margin 0.59 2.79 15.98 155090 0.26 135.29 2.89 115.57 2313.6 148.24 2.85 0.72 1.94 19.41 2.92 12.43 14346 3.91 2.82 5.24 2.79 21.50 2.85 141.53 49590 0.22 soft green mineral filling fractures, with pink 0.78 3.03 carbonate core pink mineral in vein (dolomite?) 0.61 46.20 2.77 0.58 2.88 0.48 2.83 FC sparce phenocrysts 1.08 2.78 46.20 10657 L’) 0.47 2.77 H9605-142 H9605 142.5 142.7 VMmag c B+P H9605-144 H9605 144.7 144.9 VMXmag f H9605-145 H 9605-150 H 9605-157 H 9605-162 H9605 145.35 145.55 S H9605 150.5 150.75 S H9605 157.3 157.5 S H9605 161.8 162 II 0.24 2.73 0.14 2.64 0.15 2.65 14.43 2630.6 0.51 2.83 10.13 4354.6 Sample No. HolelD From To Rock type Gr. AItn notes MS Den Chrg Res Por size (x103 SI) (glcm3) (ms) (Ohm-rn) (%) 53.18 2.90H9604-745 H9604-757 H9604-763 H9604-772 H9604-776 H9604-784 H9604-792 H9604-90 H9604-96 H9605-1 11 H 9605-128 H 9605-135 H9605-1 39 H9604 745.3 745.5 VMmag f U H9604 757.5 757.75 VMmag f B+P H9604 763.25 763.45 II c B H9604 772.2 772.4 VU m T H9604 776 776.2 II c B H9604 784.1 784.3 VU m T H9604 792.1 792.3 II m FC H9604 90.3 90.5 VM m B H9604 96.2 96.4 II FC H9605 111.5 111.7 vMmag c U H9605 128.1 128.3 VMmag c B H9605 135.7 135.9 II c B+P H9605 139.65 139.9 II f 0.59 2.81 1.55 2.80 11.29 2.89 0.32 2.87 38.00 2.82 57.70 2.77 0.52 2.78 0.35 2.84 “syenite” (K-spa-rich) veins 115.00 2.85 syenite” (K-spa-rich) veins 2.00 2.79 0.28 2.78 FC f-gr inrusive, purple color, possibly assd 0.18 2.67 w/nearby syenite 0.71 2.80 B bleached clasts, average 1cm; bleached and 0.62 2.97 chaotic matrix c FC+S c FC+S c FC+S f B mafic phenocrysts, <1 to 2mm, 3%, slightly elongate H9605-167 H9605 167 167.2 S c FC+S 0.14 2.68 H9605-176 H9605 176.4 176.6 S c FC 0.21 2.68 7.65 4150.6 H9605-180 H9605 180.2 180.4 S c FC 0.24 2.66 H9605-20 H9605 19.8 20 vMmag f B 14.40 2.80 H9605-210 H9605 210.7 210.9 S c FC+S 0.22 2.69 10.28 14889 H9605-217 H9605 217.3 217.5 5 c FC+S 0.21 2.68 H9605-224 H9605 224.15 224.35 VU f B sheared, qtz and Fe-carb veining 0.48 2.80 6.97 5820.4 0.32 H9605-238 H9605 237.85 238.05 vu f B sheared, qtz and Fe-carb veining 5.96 2.85 H9605-264 H9605 263.85 264.05 VU f T massive 21.18 2.87 H9605-266 H9605 266.5 266.75 VU f T in-situ brecciation, angular clasts, 2cm 22.00 2.83 4.93 366.1 1.08 H9605-27 H9605 27.7 27.95 VMmag f U 34.70 2.76 H9605-272 H9605 272.2 272.4 Il f U mafic (hbl?) phenocrysts, average 1mm, 91.18 2.85 10%, elongateL’J H9605-278 H9605 278 278.2 VM H9605-28 H9605 28.4 28.6 VMmag f H9605-301 H9605 301.3 301.5 VU f H9605-58 H9605-66 H9605-73 H9605 58.4 58.65 VMmag m H9605 66.65 66.85 IM C H9605 72.8 73 VMXmag U B U T F F F mylonite? B+P FC F T F FC B B F C C B F multi-lithic mylonite? FC FC FC+H B+P T FC+H 0.54 0.31 0.93 0.99 0.15 0.95 0.61 3.85 180.00 1.18 2.90 6.45 2077.9 0.40 2.78 13.60 8868.9 2.80 2.85 2.76 2.93 2.95 2.67 2.88 2.88 Sample No. HolelD From To Rock type Gr. Altn notes MS Den Chrg Res Por size (xlO3S ) (glcm3) (rns) (Ohm-rn) (%) f U 0.96 2.90 U 51.10 2.87 T in-situ brecciation, sheared, fragments 11.88 2.82 stretched U FC+S B+P in-situ brecciation, altered fragments average 0.5-1cm 327.00 3.04 0.82 2.87 0.43 2.82 29.40 2.76 0.14 2.68 0.59 2.81 37.00 2.85 0.32 2.76 0.68 2.95 0.62 2.92 0.16 2.77 9.45 4243 0.59 2.92 0.54 2.85 9.88 5135.4 0.31 0.51 2.83 H9605-8 H9605 8.25 8.45 VMmag f H9605-99 H9605 99.45 99.65 IF C H9606-104 H9606 104.4 104.6 L H9606-119 H9606 118.85 119.05 VU m H9606-147 H9606 147.1 147.3 II f H9606-152 H9606 151.75 152 II f H9606-154 H9606 154.4 154.6 VUX m H9606-173 H9606 173.2 173.45 II f H9606-174 H9606 174.5 174.7 II f H9606-179 H9606 179 179.25 VU m H9606-206 H9606 206.15 206.4 VU m H9606-22 H9606 22.35 22.6 VUX m H9606-230 H9606 230.25 230.45 II f H9606-232 H9606 231.85 232.05 II f H9606-233 H9606 233.5 233.7 II f H9606-247 H9606 247.7 247.9 II f H9606-270 H9606 270 270.2 VU H9606-276 H9606 276.4 276.6 II H9606-278 H9606 278 278.2 II H9606-295 H9606 295.5 295.7 VUX H9606-305 H9606 305 305.2 II m H9606-323 H9606 323.5 323.7 II H9606-327 H9606 327.75 327.95 VMmag f H9606-338 H9606 338.05 338.25 VMmag f H9606-357 H9606 357.3 357.5 II c H9606-370 H9606 370.05 370.3 VMmag m H9606-40 H9606 40.4 40.6 VUX m fragmental, mylonite? 1.00 2.91 8.97 8545.9 0.34 0.61 2.86 0.64 2.90 0.45 2.84 1.24 2.93 5.98 2.83T sheared, fragmenta’ H9606-411 H9606 411.05 411.3 VMmag m H9606-42 H9606 41.85 42.05 II m H9606-421 H9606 421.55 421.75 MLX H9606-424 H9606 424.3 424.5 IIX H9606-430 H9606 429.8 430 II m H9606-452 H9606 452.3 452.5 VMmag m H9606-459 H9606 459.75 459.95 VU m H9606-465 H9606 465.75 465.95 vu m H9606-471 H9606 471.5 471.7 VU m H9606-496 H9606 496.35 496.55 MLX H9606-501 H9606 500.9 501.1 vu m H9606-537 H9606 537.05 537.25 II H9606-538 H9606 538 538.2 II f H9606-539 H9606 539.5 539.7 vu m H9606-569 H9606 569 569.25 vMmag f H9606-59 H9606 59.5 59.7 vu m H9606-604 H9606 604.4 604.6 VMmag f H9606-619 H9606 618.9 619.1 VMP f H9606-628 H9606 628.25 628.45 II m H9606-629 H9606 629 629.25 II H9606-630 H9606 630.15 630.35 II H9606-631 H9606 630.85 631.05 VM f H9606-649 H9606 649.3 649.55 II m H9606-66 H9606 65.8 66 II f H9606-660 H9606 660.2 660.4 VMP f H9606-675 H9606 675 675.2 VMP f H9606-675.5 H9606 675.7 675.9 II H9606-706 H9606 706.7 706.95 VMP f H9606-708 H9606 708.15 708.4 QMX H9606-713 H9606 713.55 713.75 VM f H9606-718 H9606 718.15 718.4 VMP f H9606-721 H9606 721.7 721.9 VMP f H9606-725 H9606 725.25 725.5 VMP m H9606-745 H9606 745.1 745.35 VMP m H C+S multi-lithic breccia; large variation in sus B buff and purple-colored clasts B FC+H C massive U massive B B U Bi-P FC B sheared U F massive U U FC fsp phenocrysts FC FC B U F U qtz amygdules near pillow margins B B B B B B B B B weak fabric 39.53 2.78 14.82 2.85 0.74 2.94 0.75 2.88 124.71 2.89 24.00 2.82 27.65 2.82 36.82 2.85 7.59 2.84 3.24 2.86 15.97 3752.2 0.70 4.95 2.92 0.25 2.73 1.06 2.83 25.80 2.61 0.54 2.96 3.37 1009.04 0.50 53.20 2.83 0.63 2.78 0.74 2.83 0.45 2.77 0.68 2.83 2.48 2.83 18.40 2.75 0.61 2.92 0.60 2.79 0.59 2.81 0.42 2.76 0.67 2.88 0.35 2.87 0.55 2.89 0.54 2.88 0.40 0.39 2.80 0.48 2.90 0.55 2.90 0.33 Sample No. HolelD From To Rock type Gr. AItn notes MS Den Chrg Res Por size (x103 SI) (glcm3) (ms) (Ohm-rn) (%) FC+H 4.85 2.90 11.30 22613 L’) H9606-753 - H9606 H9606-757 H9606 H9606-784 H9606 H9606-79 H9606 H9606-89 H9606 H9606-90 H9606 H9608-102 9708 H9608-118 9708 H9608-129 9708 H9608-145 9708 H9608-176 9708 H9608-190 9708 H9608-192 9708 H9608-198 9708 H9608-209 9708 H9608-217 9708 H9608-217 9708 H9608-224 9708 H9608-24 9708 H9608-258 9708 H9608-259 9708 H9608-27 9708 H9608-272 9708 H9608-306 9708 H9608-327 9708 H9608-346 9708 H9608-361 9708 H9608-37 9708 H9608-378 9708 H9608-39 9708 H9608-405 9708 H9608-419 9708 H9608-422 9708 H9608-444 9708 H9608-463 9708 752.8 753.05 II 757.1 757.3 QMX 784.1 784.35 VMP 79.25 79.45 II 89.2 89.45 VU 90.1 90.3 II 102.55 102.8 IF 117.85 118.05 VM 129.45 129.65 VM 145 145.2 VM 176 176.2 VMP 190 190.2 VMP 192.45 192.65 IM 198.2 198.4 IM 209.4 209.6 T 217 217.3 MLX 217.3 217.4 MLX 224.65 224.85 VMP 24.3 24.5 VMP 258.4 258.6 VM 258.8 259 VM 27.55 27.75 T 272 272.2 VMPX 306.4 306.6 VM 327.35 327.55 VM 346.15 346.4 VM 361.2 361.4 VMPX 37.15 37.35 VM 378.5 378.75 VMP 38.8 39 VM 405 405.2 VMP 419.45 419.65 VMP 422.65 422.85 VMP 444.6 444.8 VMP 463.3 463.5 VMP f B f B f U f C FC B FC B u f U f C B u u C m C m u U f U U f B B f C f B f B f B f u 0.46 2.86 0.60 2.80 0.40 2.82 25.18 2.84 0.80 2.82 2.47 2.85 0.71 2.80 0.51 284 0.65 2.84 0.67 2.79 31.53 2.79 67.53 2.80 1.00 2.89 0.65 2.84 0.61 2.81 1.52 2.81 2.54 2.83 70.50 2.91 0.45 2.85 1.57 2.91 18.50 2.97 0.72 2.83 12.50 2.85 14.80 2.88 0.62 0.60 2.88 34.50 2.95 0.68 2.92 14.10 2.93 0.91 2.81 0.54 2.84 0.58 2.83 0.62 2.93 0.53 2.77 Sample No. HolelD From To Rock type Gr. Altn notes MS Den Chrg Res Por size (x104 SI) (glcm3) (ms) (Ohm-rn) (%) m FC 0.55 2.91 f U F T f T C B f B B qtz-carbonate breccia; sheared spinifex in-situ brecciation purple mineral (a carbonate?) in vein epidote and cal veins unaltd pillow basalt; cal veinsI’J H9608-508 9708 H9608-522 9708 H9608-537 9708 H9608-554 9708 H9608-562 9708 H9608-573 9708 H9608-600 9708 H9608-639 9708 H9608-650 9708 H9608-657 9708 H9608-674 9708 H9608-68 9708 H9608-712 9708 H9608-85 9708 H9608-98 9708 H9707-102 H9707 H9707-111 H9707 h9707-122 H9707 H9707-137 H9707 H9707-141 H9707 H9707-144 H9707 H9707-160 H9707 H9707-162 H9707 H9707-171 H9707 508.25 508.45 VMP 522.2 522.45 VMP 537.6 537.8 VMP 554 554.2 VMP 562.3 562.5 VMP 573.3 573.55 VMP 600.1 600.35 VMP 639 639.2 VMP 649.9 650.1 VMP 657.3 657.5 VM 673.9 674.1 VM 68.5 68.7 VMP 712.7 712.9 VM 85.25 85.45 VM 97.9 98.1 VM 102.5 102.7 VU 111.2 111.4 IF 121.85 122.1 IF 137.3 137.5 IM 141.6 141.8 VU 144.35 144.55 IF 160.75 160.95 II 162.65 162.85 VU 171.2 171.4 MLX f B in-situ brecciation f B f B f f f f F C U m U C U m T m FC f f chlorite amygdules chlorite amygdules chlorite amygdules chlorite amygdules 0.41 2.86 0.41 2.86 0.34 2.84 0.45 2.87 0.42 2.90 0.53 2.84 0.38 2.86 0.46 2.89 0.47 2.86 0.38 2.90 44.30 2.89 0.54 2.90 4.03 2.86 1.29 2.79 1.66 2.90 0.40 2.80 0.16 2.66 to sub-angular H9707-179 H9707 H9707-188 H9707 H9707-194 H9707 H9707-204 H9707 H9707-219 H9707 H9707-241 H9707 H9707-265 H9707 H9707-273 H9707 H9707-301 H9707 H9707-307 H9707 179.7 179.9 VMmag f U 188.15 188.4 VMXmag f C 194.65 194.85 II f B 204.7 204.9 VMmag f U 219.7 219.9 VMP f U 241.1 241.3 VMmag m U 265.1 265.3 VMmag m U 273.2 273.4 VMmag m U 301.05 301.25 VMmag m U 307.05 307.3 VMmag m U 1-2 cm fragments, clast-supported amygdules at margins syenite (Kfsp) veins 2.21 2.87 147.06 2.99 0.60 2.88 57.10 2.62 0.77 2.80 35.80 2.47 2.87 2.92 42.70 2.94 57.10 3.00 119.00 2.87 Sample No. HolelD From To Rock type Gr. AItn notes MS Den Chrg Res Por size (x103 SI) (qlcm3) (ms) (Ohm-rn) (%) f C 0.56 2.87 B B B B f B m B m B f C m U f U m U sheared; zoned fragments? 49.53 2.83 8.87 758.82 0.76 slightly sheared; abund quartz amygdules 10.60 2.83 2.90 2241.6 0.88 30.24 2.76 FC abundant py along fracures 18.59 2.95 T qtz fragments from shearing of veins 10.22 2.81 FC+H felsic and mafic clasts, < 1cm; sub-rounded 14.59 2.75 H9707-31 7 H 9707-351 H9707-36 H9707-373 H9707-38 H9707-388 H9707-403 H9707-420 H9707-456 H9707-478 H9707-488 H9707-494 H9707-495 H9707-512 H9707-51 5 H9707-522 H9707-540 H9707-554 H9707-56 H9707-575 H9707-599 H9707-603 H9707-614 H9707-625 H9707-627 H9707-637 H9707-655 H9707-67 H9707-677 H9707-685 H9707-690 H9707-84 H9707-96 H9707 603 603.2 VMmag H9707 614.35 614.55 VMmag f H9707 625.25 625.45 VMmag H9707 627.1 627.3 VMmag H9707 637.3 637.5 VMmag f H9707 655.75 655.95 VMmag m H9707 67.15 67.35 II m H9707 677.4 677.6 VMmag m H9707 684.8 685 VMmag m H9707 690.3 690.5 VMmag m H9707 84.4 84.6 II f H9707 96.65 96.85 IF c F U F U U U U B U U B B FC+S U U U T U 2.26 2.89 0.69 2.86 97.53 32.12 105.65 12.59 62.90 1.72 0.42 107.00 83.50 131.00 0.69 0.14 2.93 2.91 3.08 3.14 2.99 2.78 2.84 2.87 3.02 2.87 2.68 Sample No. HolelD From To Rock type Gr. Aftn notes MS Den Chrg Res Por size (x104 SI) (q!cm3) (ms) (Ohm-rn) (%) U 60.50 3.00H9707 317.15 317.4 VMmag m H9707 351.15 351.4 VMmag m H9707 36.5 36.7 VU H9707 373.65 373.9 VMmag m H9707 38.25 38.5 II f H9707 388.75 388.95 S H9707 403.7 403.9 VMmag m H9707 419.8 420 VMmag m H9707 456.05 456.25 VMmag m H9707 478.25 478.45 VMmag m H9707 488.3 488.5 VMmag m H9707 494.1 494.3 VMmag m H9707 495.5 495.7 VMXmag H9707 511.8 512 VMmag m H9707 515.2 515.4 VMXmag H9707 522.45 522.65 VMmag m H9707 540.5 540.7 VMmag m H9707 553.8 554 VMmag m H9707 56.7 56.9 VU f H9707 575.5 575.7 VMmag m H9707 599.65 599.85 KMXmag U syenite (Kfsp) veins 130.00 2.88 sheared 0.81 2.94 77.80 2.96 46.50 30556 0.26 0.51 2.91 30.82 2.69 128.00 2.95 48.30 2.92 141.00 2.97 2.19 2.58 88.24 2.89 1% cpy interstitial strongly bleached fragments <1cm, some qtz clasts abundant qtz/Fe-carb veins sheared U Kspar rich clasts; <2cm frags, may have been a vein U in-situ brecciation U U in-situ brecciation <1cm, perlitic; varioles? B U U T U U U FC+S minor qtz amygdules (<1mm) U 2.68 84.20 26.90 9.43 19.40 15.10 0.55 46.82 2.85 2.82 2.84 2.83 2.80 2.84 2.86 2.87 I’—) H9711-106 H9711 106.25 106.45 VM H9711-116 H9711 115.95 116.15 VM H9711-134 H9711 133.9 134.2 VM H9711-149 H9711 148.85 149.05 VMX H9711-149.5 H9711 149.4 149.6 VMX H9711-162 H9711 162 162.2 VM H9711-178 H9711 178 178.2 VM H9711-192 H9711 192.55 192.75 VM H9711-211 H9711 211.6 211.8 VM H9711-247 H9711 247.3 247.5 VMX H9711-250 H9711 250 250.2 VU H9711-252 H9711 252.15 252.35 S H9711-265 H9711 265.45 265.65 VU H9711-281 H9711 281.5 281.7 VU H9711-296 H9711 296.65 296.85 VMmag f H9711-313 H9711 313.5 313.7 VM f H9711-320 H9711 320.6 320.8 VM f H9711-344 H9711 344.2 344.4 VMmag f H9711-368 H9711 368.5 368.8 VM f H9711-378 H9711 378.5 378.7 VM f H9711-401 H9711 401.6 401.8 VM f H9711-430 H9711 430.4 430.6 VMmag f B÷P varioles; bleached clasts B varioles B B m U m U mafic phenocrysts B+P B sheared FH 0.56 0.53 2.83 0.68 2.74 0.44 0.61 0.61 1.07 2.80 0.67 2.81 0.53 2.76 0.76 2.89 0.46 2.72 0.75 2.87 B beige “clay-looking mineral - leucoxene (I.e. 0.69 was ilmenite) FC+H +s B B bleached varioles U B varioles B÷P B soft green material in veins FC+H 1.25 2.87 0.85 2.88 24.82 2.88 2.95 2.89 1.27 2.93 1.29 2.91 11.06 2.93 N/A >70000 H9711-439 H9711 439.1 439.35 VMmag f H9711-446 H9711 445.95 446.15 VMmag f H9711-466 H9711 465.8 466 VMmag f H9711-470 H9711 470.05 470.25 S f H9711-475 H9711 475.75 475.95 VU f H9711-496 H9711 496.4 496.6 VU H9711-521 H9711 521.65 521.85 IFP f H9711-53 H9711 53.05 53.25 VM H9711-551 H9711 551.65 551.85 VU H9711-65 H9711 65.5 65.7 VM B varioles U U FC B sheared T U phenocrysts, 1-2 mm, 5-10% m U T 14.35 2.91 44.59 2.82 26.47 2.82 0.80 2.84 3.05 2.85 38.60 3.08 0.22 2.68 5.66 53.80 2.79 N/A >70000 Sample No. HolelD From To Rock type Gr. Altn notes MS Den Chrg Res Por size (x104 SI) (gicm3) (ms) (Ohm-m) (%) m B÷P 0.49 m FC+S f B f f m m f m m B 2.87 5.34 5164 0.49 32.00 2.94 m B 1.19 Sample No. HolelD From To Rock type Gr. Altn notes MS Den Chrg Res Por size (x104 SI) (g!cm3) (ma) (Ohm-rn) (%)H9711-99 H9711 99.5 99.7 VM m FC+H 0.86 2.72HGP-site I hndsmp VM m 9.00 7453.7 0.50 APPENDIX 2F - PHYSICAL PROPERTIES - DESCRIPTIVE STATISTICS Magnetic Susceptibility SUSCEPTIBILITY (x1O3 SI Units) No. Mean Std. Dev. Log mean Median Range Ultramafic volcanic rocks - all 82 15.63 22.82 4.29 3.83 0.41-109 Least altered (dol÷chl) 8 6.03 9.61 2.15 1.49 0.57-27.65 Talc-chlorite assemblage 46 24.12 26.57 9.6 14.61 0.44-109 Fe-carbonate+muscovite altered 16 2.73 10.41 1.86 1.01 0.41-36.82 Magnesite+fuchsite altered 9 0.75 0.37 0.7 0.62 0.49-1.66 Mafic volcanic rocks - all 218 26.11 45.18 3.86 1.42 0.15-327 Leastaltered 107 40.09 52.94 9.78 21.5 0.35-327 Fe-carbonate+muscovite altered 75 6.86 25.24 1.03 0.6 0.15-148.24 Fe-carbonate+albite altered 14 1.71 3.21 0.84 0.58 0.28-12.4 Intermediate dikes-all 59 12.07 29.09 1.21 0.61 0.13-135.29 Least altered 11 41.11 47.24 7.02 18.4 0.24-135.29 Fe-carbonate+muscovite altered 10 0.75 0.35 0.68 0.69 0.32-1.55 Fe/Mg carbonate altered 22 10.95 25.91 1.26 0.58 0.13-107 Syeniteintrusives-aII 34 1.19 5.25 0.25 0.2 0.07-30.82 Least altered 12 2.87 8.8 0.42 0.27 0.16-30.82 Muscovitealtered 11 0.36 0.65 0.21 0.15 0.14-2.31 Fe/Mg carbonate altered 4 0.23 0.021 0.23 0.23 0.21-0.25 Porphyritic rhyolite dikes-all 37 1.59 5.41 0.26 0.16 0.05-30.24 Least altered 15 0.35 0.61 0.445 0.16 0.05-2.45 Muscovite altered 13 0.48 0.82 5.71 0.14 0.08-2.47 Fe/Mg carbonate altered 7 2.39 5.39 2.19 0.28 0.12-14.59 00 Density DENSITY (glcm3) No. Mean Std. Dev. Median Range Ultramafic volcanic rocks-all 81 2.86 0.057 2.85 2.74-3.08 Least altered (dol+chl) 8 2.85 0.023 2.86 2.82-2.89 Talc-chlorite assemblage 46 2.85 0.061 2.84 2.74-3.08 Fe-carbonate÷muscovite altered 15 2.87 0.054 2.85 2.78-3.01 Magnesite+fuchsite altered 9 2.91 0.035 2.92 2.85-2.96 Mafic volcanic rocks - all 205 2.86 0.1 2.87 1.94-3.14 Least altered 101 2.87 0.094 2.87 2.47-3.08 Fe-carbonate+muscovite altered 71 2.85 0.13 2.86 1.94-3.14 Fe-carbonate+albite altered 13 2.81 0.051 2.81 2.74-2.93 Intermediate dikes - all 59 2.82 0.078 2.83 2.67-2.95 Least altered 11 2.83 0.077 2.81 2.72-2.95 Fe-carbonate+muscovite altered 10 2.85 0.054 2.86 2.76-2.94 Fe/Mg carbonate altered 22 2.79 0.088 2.78 2.67-2.95 Syenite intrusives - all 32 2.7 0.057 2.69 2.59-2.86 Least altered 10 2.7 0.057 2.69 2.64-2.84 Muscovite altered 11 2.72 0.072 2.69 2.64-2.86 Fe/Mg carbonate altered 4 2.67 0.0096 2.68 2.66-2.68 Porphyritic rhyolite dikes - all 36 2.61 0.45 2.67 2.63-2.88 Least altered 15 2.51 0.7 2.67 2.63-2.84 Muscovite altered 12 2.68 0.074 2.67 2.57-2.85 Fe/Mg carbonate altered 7 2.73 0.082 2.71 2.65-2.88 249 Resistivity RESISTIVITY (Ohm-rn) Mean Std. Dev. Log mean Median 2815.99 2451.99 1574.58 2159.75 1127.75 1376.22 573.95 385.69 4451.55 1714.49 3952.65 4757.60 28431.94 36820.16 13004.05 16462.00 17910.02 17881.25 8344.70 16462.00 Range 111.16-8545.9 111.16-3994.3 1009.04-3217.4 541.24-155090 541.24-49812 Chareabi1itv CHARGEABILITY (ms) Std. Dev. Median 4.46 5.46 5.89 4.24 2.45 5.68 44.97 13.80 11.04 9.37 Range 2.9-20.4 2.9-20.4 3.37-9.88 2.07-1 58.17 2.07-33.23 No. 20 8 7 18 6 Ultramafic volcanic rocks - all Talc-chlorite assemblage Carbonate+muscovite altered Mafic volcanic rocks - all Carbonate+muscovite or albite altered Intermediate dikes - all Syenite intrusives - all Porphyritic rhyolite dikes - all 6 9758.85 10 6759.51 11 11534.00 8025.69 3528.59 5359.31 7193.33 6027.61 10523.10 6611.75 6179.85 11136.00 2313.6-22613 2630.6-14889 4576-23970 No. Ultramafic volcanic rocks - all 20 Talc-chlorite assemblage 8 Carbonate+muscovite altered 7 Mafic volcanic rocks - all 18 Carbonate+muscovite or albite 6 altered Intermediate dikes - all 6 Syenite intrusives - all 10 Porphyntic yoIite dikes-all 11 Mean 6.91 6.36 6.63 31.32 12.03 29.24 14.71 10.47 42.35 7.04 13.19 12.45 13.92 4.20 9.45-115.57 6.27-27.4 2.2-46.2 250 APPENDIX 2G - CORRELATION COEFFICIENTS FOR PHYSICAL PROPERTIES AND XRD (RIETVELD) - DERIVED MINERAL ABUNDANCES Calculated using the statistical analysis software SPSS Statistics. Correlation coefficients are calculated for all physical properties. Not all minerals are considered, only ones occurring most commonly. Spearman’ s correlation coefficient calculations are used as they are appropriate wheredata do satisfy normality assumptions. ALL ROCK TYPES (PAGE 1) LOG_MA I IMAGSUS GSUS I DEN Cl-IRG RES LOG RES POR AB J CAL CLCSpearmaon rho MAGSUS Correlation Coefficient 1000 1 000-i 444j .278 -.212 .2*4 175 -.5131 178 - 188Sig. (2-tailed) 000 GOt) .034 ] - 69 07 363 003 601 442N 365 385 369 I 5 58 58 29 32 t 19LOG_MAGSUS Correlation Coefficient i.ooor 1.000 - 427T-212 -214 175 - 513H 176 188Sig. (2-tailed) 000 000 034 109 107 363 003 601 442N 386 395 369 58 59 50 28 32 11 19DEN Correlation Coefficient 444W 444 1000 076 036 - 038 3$1 646 405 - 030Sig. (2-tailed) 013Cr 000 - 568 789 778 041 000 26 8133N ‘9 365 370 59 58 58 28 32 1 19CHRG Correlation Coefficient . 78 27s .076 000 352 354i 519j -115 j 030 243Sig. (2-tailed) **34 034 .569 006 005 I 006 I 560 I 834 348N 58 55 58 80 80 80 27 26 10 17RES Correlation Coefficient -212 -212 -036 1005 1.000j - 8t3 .021 .146 -287Sig. (2-tailed) 109 109 789 1 00d. 000 000 9*4 688 300N 58 58 58 Ø4 60 I 60 27 28 10 17LOG_RES Correlahon Coefficient -214 -214 -038 1 000 1 000 815 020 168 - 264Sig. (2-tailed) 107 107 778 000 000 919 643 3i37N 58 50 58 60 J_,ffl,._60 27 28 10 1?POR Correlation Coefficient 175 175 —aBr ..815 1000 -29t I -26t 569Sig. (2-tailed) 363 363 04t . - - - 000 I 366 68 042N 29 29 99 - - 27 -- 27 27 30 12 8 13AB Correlation Coefficient : ‘- - 64fl - 115 02 020 -291 1.000 - 74T -395Sig. (2-tailed> 900 560 914 919 358 014 j 104N 42 21j 32 28 28 28 2 33 10 L 18CAL Correlation Coefficient 178 178 405 -030 146 168 -261 042* 1000 57*Sig. (2-tailed) 601 601 2 6 934 688 643 618 014 I .135N 11 II II tO to to 6 56 1 5CLC Correlabon Coefficient -188 -198 -630 -243 -267 264 569 -395 57 1.000Sig. (2-tailed) 442 .442 903 348 300 307 842 104 139N 19 19 15 17 17 17 3 10 8 20DOL Correlation Coefficient 417* 417 578 95 -067 -013 -060 922 -378 -099Gig. (2-tailed> 634 fli34 ,)062 385 974 954 888 001 403 748N 26 26- l’ 26 22 22 22 8 25 7 13CB_TOT Correlation Coefficient 160 80 I .6.469* - 68 063 065 087 - 6’0 301 -037Sig. (2-tailed) 332 332 625 403 753 745 800 SOC 369 889N 31 3 31 27 27 27 ii I 30 1 7HEM Correlation Coefficient .. .7W 256’ 072 I 850 048 048 -350 450 1 000’ 056Sig. (2-tailed> --1ffl 036 878 I 007 910 910 6 4 310 .913N ,-3-,,, 71 71 8 8 8 6 7 3 6MAO Correlation Coefficient - 857’ 957 518 083 -083 -083 -.800 -190 00 07Si9 (2-tailed) 407 007 188 83 831 831 200 651 .873 867N 8 0 8 9 9 9 4 8 5 8MC_INT Correlation Coefficient -456 -456 - ‘84’ 051 - 19 - 191 -371 i 975 - 725 -367Sig. (2-tailed> 088 088 864 5 3 513 468 - - .$- 193 332N 15 15 45- 14 14 14 6 - 1$- 6 9MS Correlation CoefficIent -515 -515 -492 -.297 -442 - 462 -300 -000 -l 090 -429Sig (2-tailed) 128 28 148 409 200 179 624 987 1.000 337N 10 10 10 9 10 10 5 10 2 7MUS_TOT Correlation Coefocient -527 -527 -592 -297 -442 -462 -305 -006 - 009 -429Sig. (2-tailed) 096 096 055 465 200 179 624 987 I 000 337N Ii II Ii 10 10 10 5 10 2 7PY Correlation Coefficient 381 38 .494’ 380 -242 264 -347 -789 600Sig. (2-tailed> 132 132 044- 159 385 341 73 112 .208N 17 17 1 18 15 15 1 7 5 6QTZ Correlation Coefficient -235 -235 -163 -255 -081 -077 308 167 008 032Sig. (2-tailed) 211 211 389 158 693 708 330 386 983 900N 30 30 30 26 26 26 2 29 8 8FECB_TOT Correlation Coefficient - 411’ 43V 886 .259 251 248 000 - c72 -450 -242Sig (2-tailed) -i,$26 82$- - - 000 233 248 255 1 500 - 000 310 426N - 7 27 - 27 23 23 23 8 25 7 13Correlation in oigoificnnt at the .51 level (2-tailed) - • . . - . . Relationships addressed in Chapter 2 -Correlation is significant at the .05 level (2-tailed) 251 C) °‘ ‘2 I Z C o C )Z C o C )Z C o C )Z C o o Z C o C )Z C O C ) o 8C ) a C ) o C ) . 0 - t C) Co 0 - — 0 -1 J V V 8 0 0 ni 0) z . C ) C ) 0 C) C) > g Z C o C )Z C oC )Z C oC )Z C oC )Z O O C ) a C ) 8 C ) c , C ) C ) t t 3 ;u C) 0 m z m 0 Co j: Z C o C )Z C o C )Z C o C )Z C o C o 8 o C Cl) 0 0 C) Co -1 I 0 -C C) C o . — — — V 0 0 0 O 0 rC 0 )0 0 Co 0 00 o b - o - b b - - n C o . g . • . C V 0) g0 ) g0 ) V m C) — — V , V . • V V - — - — - — — — — — - — - _ _ _ .1 • ‘0! - ‘ 0 0 ’ V •V - - — — V CD !1 r1 “ - V — V — — — V V : ’ 0 ’ 0 I. V IV I ‘ 0 0 0 , (0 ‘0 ‘0 ‘0 0 (0 0 N 0 (0 0 (N (0 (0 (N (0 00 (N (0 0 0 (N V (N V (0 0) (0 0 (0 ‘0 ‘0 0( 0 W 0 0 0 f l 0 ’ 0 ( 0 . - 0 _ 0 _ N 0 0 f l 0 - 0 (0 0 0 -(- 0 0 ‘ 0 ’ ‘ 0 ‘ 0 0 0 )0 0 (0 (0 0 - ‘ 0 ( .0 0 0 ((((0 0 0 (0 ‘ ‘ 0 CD0 ((08 V ’ 0 ( 0 ’ 0 ( f l 0 ( (0 (0 0 0 - . .0 ( 0 (0 0 (0 0 ‘ 0 0 (0 0 0 (0 0 - 0 ( ( V ( 0 - 0 0 0 0 0 0 0 ‘ 0 0 ((((0 W ( — — — — — ‘0(D ( 0 ( 0 ( 0 ( 0 ( ( ( g 0 0 g 0g o o — V V V V V - V V CD0 — ‘ 0 ’ 0 g g ’ 0 . 0 o , o V V V • V V V • V V V V V V V V •V V - V — V — V V V V V rC)CD V 0 0 ’ 0 _ ’ 0 0 ’ 0 0 1 ‘ 0 ’ 0 0 O C ) b O f l ( a V 0 - . _ 0) 0 ‘0 0 0 CD (0 (0 0) (0 (0 0 0 0 0 ( 0 - ( ( 0 ( 0 ( . ( 0 ( 0 ( V . CD V (0 CD V (0 0 00 0 , CD 0 0 (0 0) 0 0 ( 0 ((((0 (0 O ( (0 0 ) ‘0 ‘0 ‘0 (0 * 0 0 ‘ 0 0 - . 0 — — — < V CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD CD 8 8 8 c 8 8 c 3 ( C C C : — — 0 , : 0DOD 0 — I E0 0 O¶1 CzC 253 O C C 1 0 0 t I 0 0 0 0 > - u r- r 0 0 r g I Z 0 ‘ - u I (0 d 0 m I (0 - I 00 C) C z ø O z ( 0 o z ø O z ø o z w o z ø o z ø o z w O z ø o z w o z o ,o z w o z m o z o ,o z o o z o ,o z ø o z 0 7 0 z G ,o z w n z c n o e e 2 e2 e e C C C o e n C o C o C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) C) (0 , . . C) - 2 — - - - - - - - - - - - - - - - - )• 0 S g 0 C) - - — — - - — — , (0 , 0 — - - - j - - - , - I-, a O i p - z 0 0 C H H rn x 3 0 3 0 8 0 3 0 3 0 3 0 8 0 3 0 3 0 8 0 3 0 3 0 3 0 8 0 3 0 3 0 3 0 3 0 3 0 3 0 S o a a a a a a a a a a a a a a a a a a a a 0 . - - , - 0 0 - - - - - j — — j ; - : - : 0 . a E1 0 C, C C I’ J FELSIC ROCKS - SYENITE INTRUSIVES AND RIJYOLITE DIKES FECB_TOT CorrelaSon Coefficient 087 518. (2-laded) .801) N 11 .999 .)99 .6W .6W -22) -.247 .306 .366[ -.438 .000 001 005 .005 .373 324 282 298 177 70 67 15 18 18 18 It 10 ii -.643 565 .575 -.770 600 119 054 .051 015 4 7 1) 12 8 opoarrranc cc aco.o CorrelatiOn COOt00eñL one 519. (2-laded) N 70 LOG_MA 1 1 1MAGOUS GSUS DON CI-IRG LOG CoG RES LOG RES AS DOL I CB TOT I MC INT MS I MUS TOT PY LOG Ph’ OTZ FECB TOT .730 262 .097 .600 LOG_MAGSUS Correlason Coefficient .999 1,000 ,384 .642 .64r -216 -.2)9 -.356 .366 -.438 .738 -.400 -.643 .865 .575 -.770 .087Sig.(2-toded) .000 . .001 .004 .054 .395 .339 .262 298 .177 262 .600 119 056 .651 .015 800N 70 70 67 18 18 18 18 It 10 Ii 4 4 7 12 12 9 IIDEN Conretafion Coefficient ::::. 6 1.000 .655 .660 -.096 . 715 -.462 .494 -.069 250 .400 .667 SaP .576 -.73l .402So (2tol d) 8 601 003 003 698 649 153 147 040 742 800 102 048 051 025 220N 67 67 18 18 18 18 II IC ft 4 4 7 12 12 9 IiCHRG Correlas n Co ffictent 655 I 000 I OOT 095 109 600 ISO 317 1 000 41)0 143 568 508 676 2675.9 (21 led) 9 903 705 667 088 651 .951 600 787 074 074 084 488N 15 18 18 18 18 9 8 9 4 4 9 10 10 7 9LOG_CHG C I I C Ill I 655 1 000 I 000 099 169 600 190 377 000 400 143 588 505 679 267Sg (21 I d) 8 053 708 667 088 651 406 600 787 074 074 594 489N 95 18 16 15 16 9 9 9 4 4 6 16 IC 7 9RES Correla9on Coefficient -223 -.216 -.098 -.095 -.095 1.005 .99T .233 .033W .850 -200 -.400 -.600 .081 .091 .107 .833Sig. (2-tailed) 373 390 .698 .708 .708 0115 .546 .013 .004 .800 . .600 208 803 .803 .819 .005N 19 18 18 10 18 18 16 9 8 9 4 4 6 10 10 7 9LOG_RES Correlation Coetficlenl -247 -.239 -.115 -.109 -.109 .999 7.000 210 814W .879 .200 - 405 -.600 .036 .036 .162 .820Sig. (2-tailed) .324 .339 .649 .667 .667 .000 974 014 .002 .000 .600 .208 920 .920 .728 .006N 18 18 18 18 19 58 19 9 8 9 4 4 6 tO tO 7 9AB Cornela8onCoefficient -.356 -.356 - 462 -.600 -.600 .233 218 1.000 -.015 -.036 -l.000 -1.OOffi -.486 -.355 -.328 .452 -.282Sig. (2-tailed) 262 262 .753 .088 .068 .546 .574 . .960 .975 .000 .060 .329 .285 .325 .260 .401N II II II 9 9 9 - - 9 11 10 11 4 3 6 11 11 9 IlDOL Correla9on Coefficient .366 .366 .494 .190 .190 .937 . .8l4 -.018 1000 .652 000 -1000 -.700 624 .620 -.476 .964Sig. (2-teiled) .298 .298 .147 .657 .651 ./2t1) 14. .960 . 069 1 000 1000 .188 .054 .056 233 .000N 10 10 15 8 6 4- 8 10 10 tO 4 2 0 tO 10 8 10CB_TOT C I Ii C ffi nt 438 438 069 317 317 $4 8Th 039 612 1 600 865 500 371 027 005 571 736Sg (2toI d( 177 177 040 406 406 94 658 915 060 200 667 468 937 989 139 950N II II 11 9 9 9 8 11 10 1 4 3 6 II II 8 ItMC_INT Conolason Coefficient .739 .739 259 1.500 7.000 -.209 - 200 -1 SOT .000 -.800 1 000 . 1 000 .800 .500 -.506 .000Sig. (2-tailed) .262 .262 .742 . . .000 .800 .000 1.060 200 . . . .200 .200 .667 1.000N 4 4 4 4 4 4 4 4 4 4 4 0 2 4 4 3 4MS Correlation Coefficient -400 -.400 .400 -400 -.400 -.400 -400 -1 000 -1.000 .500 . 1.000 800 . 400 -400 1 000 500Sig. (2-tailed) .600 .600 .600 .600 600 .600 .660 .000 7.050 .667 . . .200 .605 .603 . .667N 4 4 4 4 4 4 4 3 2 3 0 4 4 44 2 3MUS_TOT Cnrrela8oc Coefficient -.643 -.643 667 .143 .143 -.600 -.600 -.486 -.700 -.371 1.000W .800 I 000 -.534 -.536 .400 . 371Org. (2-ta,ler .719 .119 .102 .787 .787 .208 .208 .329 .198 .468 . .200 . .215 .219 .600 .490N 7 7 7 6 6 6 6 6 5 6 2 4 7 7 7 4 6Ph’ Correlation Coefficient .565 .565 - SOP .588 .588 .091 .036 -.355 .624 .027 .800 -.400 -.536 1,600 .998 -.633 .482Sig. (2-toiled) .056 .056 048 .074 .074 .803 .920 .285 .054 .937 .200 .600 .215 . .000 .061 133N 12 12 12 10 10 tO 10 11 10 ii 4 4 7 12 12 9 11LOG_PY Correlation Coefficient .575 .575 .575 .588 .586 .691 .036 -.328 .620 .005 .850 -.450 -.536 .... - 913131 7.000 -,669 .478Sg (21 I d( 051 051 051 974 074 803 920 325 056 989 260 600 215 1)60 049 137N 12 12 12 In 10 In 10 II 10 If 4 4 7 - -12 12 0 IlQTZ Cornela8onCoeffinient -.77-3 - - -.Zl -.679 -.679 .107 .162 .452 -.476 .571 -.500 1.000 .400 -.633 -.669 1.000 -.476Sig. (2-toiled) 075 -- 015 -- - - 035 .094 .094 .819 728 .260 .233 .139 .667 .8130 .567 049 - .233N 9 .5 .9 7 7 7 7 8 8 8 3 2 4 9 9 9 8.087 .800 11 Carnela9on is significant at the .01 level (2-tailed). -. Ccrnelason is significant at the 05 level (2-tailed). Relotrocships addressed in Chapter 2 .402 .267 267 . .. o.ao sco -282 eon ton.- 000 .500 -371 .492 .478 -.476 1.000.220 .488 408 - -- 066 401 0011 515 t 000 .667 .468 .133 .137 .23315 9 9 .9’J-i’ri 13 II fll3 77 4 3 6 75 II 5 11 UI ON Appendices on accompanying CD: APPENDIX 3A - OBSERVED VERSUS PREDICED DATA FOR SYNTHETIC INVERSION MODELS APPENDIX 4A - HISLOP 3D MAGNETIC, 3D GRAVITY, 3D DC RESISTIVITY, AND 3D IP INVERSION RESULTS APPENDIX 4B - 2D DC RESISTIVITY AND INDUCED POLARIZATION INVERSION RESULTS FOR HISLOP APPENDIX 4C - OBSERVED VERSUS PREDICED DATA FOR IIISLOP INVERSION MODELS 257

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
Russia 14 0
United States 6 21
Canada 6 5
Finland 5 0
India 4 0
Germany 3 9
France 2 0
Japan 1 0
China 1 29
City Views Downloads
Unknown 25 18
Mumbai 4 0
Calgary 3 1
Washington 2 0
Ashburn 2 0
Hangzhou 1 0
Wilmington 1 0
Kapuskasing 1 0
Montreal 1 0
Tokyo 1 0
Vancouver 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items