Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Targeting Archean orogenic mineralization using physical properties and integrated geophysical methods Mitchinson, Dianne Edith 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 ]
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
Original Record: 1.0052581 +original-record.json
Full Text
1.0052581.txt
Citation
1.0052581.ris

Full Text

Targeting Archean Orogenic Gold Mineralization UsingPhysical Properties and Integrated GeophysicalMethodsbyDIANNE EDITH MITCHTNSONB.Sc., Memorial University of Newfoundland, 2001M.Sc., Laurentian University, 2004A DISSERTATION SUBMITTED 1N PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Geological Sciences)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)January 2009© Dianne Edith Mitchinson, 2009AbstractAlthough Archean orogenic gold mineralization is not readily detected usinggeophysical methods, due to a lack of petrophysical contrast between typical lowvolumes of gold and hosting rocks, it is possible to use geophysics to detect otherpetrophysically distinct gold indicators. Geophysical inversion methods, in particular,make it possible to not only detect important gold-related rocks in the subsurface, but tomap their distribution in three dimensions. The research presented examines theeffectiveness of geophysical inversion as an exploration tool in the Archean orogenicgold environment through extensive physical property analysis, synthetic modeling, andinversion of various geophysical data over the Hislop gold deposit, Ontario.As understanding rock properties is imperative to interpreting geophysical data, itwas necessary to establish the physical property ranges of typical host rock types,hydrothermally-altered, and mineralized rocks in this deposit setting. Felsic dikes, knownto be associated with gold at Hislop, have low magnetic susceptibility and density rangesthat allow them to be distinguished from mafic and ultramafic rocks. Additionally, manypotentially mineralized, carbonate-altered mafic and ultramafic rocks can be isolatedfrom 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 inthe 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 demonstratedthat constraining inversions through addition of basic prior geologic and physicalproperty information, yields models with improved physical property distribution, andestimates. Applying knowledge gained from physical property, and synthetic modelingwork lent confidence to interpretations of inversion results for the Hislop area. Atregional scales, susceptibility and density models reveal a steep southward dip for thegold-related Porcupine-Destor Deformation Zone, and a greenstone depth ofapproximately 7000 m. Fe-rich mafic rocks directly hosting the Hislop deposit are11complexly faulted and extend to 3000 m depth. At deposit-scales, model cells withcombined low susceptibilities and high chargeabilities, occurring proximal to faults,felsic intrusions, and Fe-rich mafic rocks, highlight prospective areas for furtherinvestigation.111Table of ContentsAbstract.iiTable of ContentsivList of Tables viiiList of FiguresixAcknowledgements xivCo-authorship StatementxviChapter 1: Introduction11.1. Combining geology and geophysical inversion for mineral exploration11.1.1. Geophysics and mineral exploration11.1.2. Mineral Deposit Research Unit — Geophysical Inversion Facilityproject 21.1.3. Inversion in the Archean orogenic gold environment31.2. Background to geophysical inversion51.3. Archean orogenic gold - geologic and geophysical background71.3.1. Background on Archean orogenic gold71.3.2. The Hislop gold deposit81.3.3. Geophysics and gold91.4. Project objectives10References13Chapter 2: Physical properties of rocks in an Archean orogenic gold environment172.1. Introduction172.1.1 Rationale172.1.2 Objectives182.2. Background192.2.1 Geology and geophysics of Archean orogenic golddeposits 192.2.2 Geology of the study area202.3. Methodology252.3.1 Field and mineralogical studies252.3.2 Physical property measurements26iv2.4. Data and observations.292.4.1. Hislop deposit rock types, hydrothermal alteration, and associated mineralogy 292.4.2. Physical properties of the Hislop deposit 322.5. Interpretations 432.5.1. Effect of geological processes on physical properties at Hislop 432.6. Discussion 602.6.1. Exploration using physical properties 602.6.2. Comparison to analogous areas 672.7. Conclusions 75References 77Chapter 3: Detecting gold-related geology in Archean orogenic gold environmentsusing geophysical inversion: a synthetic modeling study based on the Hislop golddeposit, Ontario 863.1. Introduction 863.1.1. Rationale 863.1.2. Objectives 873.2. Background 883.2.1. Geology of the Hislop gold deposit and relationship to other Archean orogenicgold deposits 883.2.2. Physical Properties of rock types and alteration zones at Hislop 903.2.3. General forward modeling and inversion background 953.3. Methods 973.4. Synthetic modeling results 1033.4.1. Potential fields modeling 1033.4.2. DC resistivity and induced polarization modeling 1123.4.3. Improving model results with basic constraints 1203.4.4. Other solutions for improving model results 1253.5. Conclusions 128References 131vChapter 4: 3D inversion of magnetic, gravity, DC resistivity, and inducedpolarization data over the Hislop gold deposit, south-central Abitibi greenstone belt1374.1. Introduction 1374.1.1.Rationale 1374.1.2. Geological background 1384.1.3. Relationships between geophysics, physical properties, and geology 1404.1.4. Inversion background 1474.2. Inversion Approach 1504.2.1. General strategy 1504.2.2. Magnetic inversions 1514.2.3. Gravity inversions 1574.2.4. DC resistivity and IP inversions 1584.2.5. Constraining magnetic inversions with reference models built in Modelbuilder1604.2.6. Inversion model display 1624.3. Inversion results and analysis 1654.3.1. Magnetic susceptibility models 1654.3.2. Density model 1754.3.3. Resistivity models 1754.3.4. Chargeability models 1804.4. Querying combined inversion results 1814.4.1. Regional scale query (susceptibility and density) 1824.4.2. Local scale query (susceptibility, chargeability) 1844.4.3. Deposit scale query (susceptibility, chargeability) 1864.5. Summary and discussion 188References 192Chapter 5: Summary and future work 1985.1. Synthesis of research presented 1985.2. Significance and contributions to the field 199vi5.3. Limitations of the thesis research 2005.4. Recommendations for continued work 2025.5. Future directions of the field of study 204References 207Appendix 2A — List of Abbreviations 209Appendix 2B - Hislop Drilicore Logs, Cross-sections, and Outcrop Maps 210Appendix 2C - Detailed and Expanded Methods 222Appendix 2D — X-ray Diffraction Analyses 231Appendix 2E — Physical Properties of Hislop Deposit Rocks 233Appendix 2F — Physical Properties — Descriptive Statistics 248Appendix 2G — correlation coefficients for Physical Properties and XRD (Rietveld) -Derived mineral abundances 251Appendix 3A - Observed versus Prediced Data for Synthetic Inversion Models 257Appendix 4A - Hislop 3D Magnetic, 3D Gravity, 3D DC Resistivity, and 3D IP InversionResults 257Appendix 4B - 2D DC Resistivity and Induced Polarization Inversion Results for Hislop257Appendix 4C - Observed versus Prediced Data for Hislop Inversion Models 257viiList of TablesTable 2.1. Geophysical characteristics of Archean orogenic gold deposits21Table 2.2. Summary of the principal rock types found in the Hislop deposit area,andassociated mineralogy 30Table 2.3. Ranges of resistivity and chargeability for rock types similar to those occurringin the Hislop deposit area (data from Telford et al., 1990)42Table 2.4. Densities of the common minerals in Hislop deposit rocks (from www.mindat.org)49Table 2.5. Statistical data for prospective rocks at Hislop, and cut-off values usedforquerying physical property data64Table 2.6. Results from magnetic susceptibility and density queries of the Hislopphysicalproperty dataset 65Table 3.1. Characteristics of Archean orogenic gold deposits91Table 3.2. Physical property values used in synthetic modeling 96Table 3.3. Synthetic survey parameters100Table 3.4. Synthetic inversion parameters100Table 3.5. Model differences calculated between recovered and true models(the lowestmodel differences for each geophysical method are highlighted with bold text)102Table 4.1. Typical and anomalous physical property ranges for principal rock typesoccurring in the Hislop deposit area142Table 4.2. Survey parameters155Table 4.3. Inversion parameters156Table 4.4. GiFtools ModelBuilder options chosen for building Hislop reference models.163viiiList of FiguresFigure 1.1. Flow chart illustrating the role of physical properties in the inversion process.4Figure 2.1. Approximate location of the Hislop study area in the Abitibi greenstone beltof the Superior Province 22Figure 2.2. Geology of the Hislop deposit area as interpreted by Power et al. (2004) fromhigh resolution aeromagnetics 24Figure 2.3. Cross section looking northwest through the Hislop deposit 25Figure 2.4. Geology, alteration, magnetic susceptibility, and gold grade logs 33Figure 2.5. Magnetic susceptibility histograms for the five main rock types found in theHislop deposit area 34Figure 2.6. Magnetic susceptibility histograms showing susceptibility data for a) least-altered and altered ultramafic volcanic rocks, and b) least-altered and altered maficvolcanic rocks 36Figure 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 37Figure 2.8. Density histograms for the five main rock types found in the Hislop depositarea 38Figure 2.9. Density histograms showing density data for a) least-altered and alteredultramafic volcanic rocks, and b) least-altered and altered mafic volcanic rocks 39Figure 2.10. Density histograms showing density data for a) least-altered and alteredintermediate dikes, b) least-altered and altered syenitic dikes, and c) least-altered andaltered porphyritic rhyolite dikes 40Figure 2.11. Resistivity histograms for Hislop deposit rocks 42Figure 2.12. Chargeability histograms for Hislop deposit rocks 43Figure 2.13. Positive correlation between modal magnetite in Hislop rock samples (asderived from XRD analysis) and magnetic susceptibility. For calculated correlationcoefficients see Appendix 2G (all rock types) 44ixFigure 2.14. Magnetite grains (reflective grains in lower image) are destroyed within acarbonate altered zone surrounding a carbonate vein in a mafic volcanic rock fromHislop 45Figure 2.15. Modal magnetite versus total Fe-rich carbonate abundance for all Hislopsamples with measured quantities of these minerals 46Figure 2.16. Histograms showing distribution of susceptibility for fine- and mediumgrained a) mafic volcanic rocks, and b) ultramafic volcanic rocks 48Figure 2.17. Density increases for Hislop rocks with an overall increase in the abundanceof Fe-rich carbonate. For calculated correlation coefficients see Appendix 2G (all rocktypes) 51Figure 2.18. Measured versus calculated density for Hislop rocks 52Figure 2.19. Porosity of ultramafic volcanic rocks at Hislop decreases with carbonate-related hydrothermal alteration 53Figure 2.20. No relationships are indicated between porosity and density for maficvolcanic rocks at Hislop 53Figure 2.21. Resistivity versus magnetic susceptibility 56Figure 2.22. Resistivity versus density 57Figure 2.23. A plot of porosity versus resistivity shows that annealing of ultramafic rocksdue to precipitation of carbonate minerals during hydrothermal alteration brings about adecrease in porosity and a corresponding increase in resistivity 58Figure 2.24. A weak positive correlation exists between pyrite abundance andchargeability 58Figure 2.25. A negative correlation between chargeability and porosity 59Figure 2.26. Magnetic susceptibility plotted against density for Hislop samples 60Figure 2.27. Carbonate-alteration destroys magnetite in a) mafic and b) ultramaficvolcanic rocks 62Figure 2.28. Magnetic susceptibility histograms comparing data from Hislop rocks, andequivalent rocks from surrounding regional areas 68Figure 2.29. Density histograms comparing data from equivalent rock types from Hisloprocks, and equivalent rocks from surrounding regional areas 69xFigure 2.30. A comparison of magnetic susceptibility data associated with least-alteredand carbonate-altered mafic rocks from the Hislop deposit, and from the greatersurrounding area70Figure 2.31. A comparison of density data associated with least-altered and carbonate-altered mafic rocks from the Hislop deposit, and from the greater surroundingarea 72Figure 3.1. Cross-section looking northwest through the Hislop deposit89Figure 3.2. Plot of magnetic susceptibility versus density for major rock unitsat Hislop.92Figure 3.3. Plot of magnetic susceptibility versus density for variably alteredmaficvolcanic rocks, and variably altered ultramafic volcanic rocks from Hislop93Figure 3.4. Resistivity histograms for Hislop deposit rocks94Figure 3.5. Chargeability histograms for Hislop deposit rocks94Figure 3.6. a) 3D geological model based on the geologic setting of the Hislop golddeposit. b-e) North-facing cross-sections through 3D physical propertymodels generatedfrom the geologic model99Figure 3.7. Starting model and unconstrained magnetic inversion result for the‘Hisloplike’ magnetic susceptibility model104Figure 3.8. Starting models and magnetic inversion results with changes madetogeometry of the target body105Figure 3.9. Magnetic inversion results for starting models with different physicalpropertycontrasts between the target and host rocks106Figure 3.10. Starting model and unconstrained gravity inversion result for the ‘Hisloplike’ density contrast model108Figure 3.11. Gravity inversion results with changes made to geometryof the target body.108Figure 3.12. Inversion results with different physical property contrasts betweenthetarget and host rocks110Figure 3.13. Starting model and unconstrained DC resistivity inversion result(conductivity model) for the ‘Hislop-like’ resistivity model113Figure 3.14. DC resistivity inversion results (conductivity models) with changesmade tophysical property contrasts, and to the geometry of the target body114xiFigure 3.15. Starting model and unconstrained IP inversion result for the ‘Hislop-like’chargeability model 116Figure 3.16. IP inversion results with changes made to physical property contrasts andgeometry of the target body 117Figure 3.17. Inversion results for the Hislop-like susceptibility model after constraintsapplied 122Figure 3.18. Inversion results for the Hislop-like conductivity model after constraintsapplied 124Figure 3.19. Inversion results for the Hislop-like susceptibility model with depthweightings reduced 126Figure 3.20. Comparison of a dipole-dipole electrode configuration and a Schlumbergerconfiguration which resembles a Realsection array 127Figure. 3.21. DC resistivity inversion result for resistivity data collected via a dipole-dipole survey 128Figure 4.1. Geological map of the southwest Abitibi greenstone belt 139Figure 4.2. Geology of the Hislop deposit area 140Figure 4.3. Cross section looking Northwest through the Hislop deposit 141Figure 4.4. Magnetic susceptibility plotted against density for the major rock types atHislop 143Figure 4.5. Magnetic susceptibility plotted against density for a) mafic and b) ultramaficvolcanic rocks from the Hislop deposit area 144Figure 4.6. Resistivity histograms for Hislop deposit rocks 146Figure 4.7. Chargeability histograms for Hislop deposit rocks 148Figure 4.8. Chargeability plotted against pyrite abundance for Hislop samples 148Figure 4.9. Chargeability versus porosity for mafic rock samples from Hislop 149Figure 4.10. Extents of magnetic data used in the deposit-, local-, and regional-scalemagnetic inversions 152Figure 4.11. Data used in regional-scale magnetic inversion 153Figure 4.12. Data used in local-scale magnetic inversion 154Figure 4.13. Data used in deposit-scale magnetic inversion 154Figure 4.14. Data used in regional-scale gravity inversion 158xiiFigure 4.15. Location of DC resistivity and IP lines used for 3D DC resistivity and IPinversions. Local mine grid line numbers shown. See Figure 4.2 for geology legend.. 159Figure 4.16. Extents of inversion model volumes, with cross-section location indicated.164Figure 4.17. North-south cross-section through the regional-scale unconstrained magneticinversion result 166Figure 4.18. Isosurface model from regional scale magnetic inversion results 167Figure 4.19. North-south cross-sections through the local-scale a) unconstrained, and b)constrained magnetic inversion results 169Figure 4.20. Isosurface model from local magnetic inversion results 171Figure 4.21. North-south cross-sections through the deposit-scale a) unconstrained, and b)constrained magnetic inversion results 172Figure 4.22. Isosurface model from deposit-scale magnetic inversion results 174Figure 4.23. North-south cross-section through the regional-scale gravity inversion result,inverted with non-located constraints 176Figure 4.24. Isosurface density model from regional-scale gravity inversion results... 177Figure 4.25. North-south cross-section through the deposit-scale a) DC resistivity and b)IP inversion results 178Figure 4.26. Isosurface models for deposit-scale a) conductivity, and b) chargeabilityresults 179Figure 4.27. Result for a physical property query targeting low magnetic susceptibility-low density cells within the regional-scale common earth model 183Figure 4.28. Result for a physical property query targeting high magnetic susceptibility -high density cells within the regional-scale common earth model 184Figure 4.29. Result for a physical property query targeting low magnetic susceptibility -high density cells within the regional-scale common earth model 185Figure 4.30. Result for a physical property query targeting low magnetic susceptibility -high chargeability cells within the local-scale common earth model 186Figure 4.31. Result for a physical property query targeting low magnetic susceptibility -high chargeability cells within the deposit-scale common earth model 187xliiAcknowledgementsThanks to my supervisor Richard Tosdalfor the always timely feedback and thenumerous edits of my chapters,and for support during the courseof the thesis research.Thanks to all the geologists andgeophysicists who provided input throughdiscussionsand edits, including Claire Chamberlain,Shane Ebert, Rob Eso, Ken Hickey, PeterLelievre, Doug Oldenburg, NicolasPizarro, and Victoria Sterritt. Iam most especiallygracious for all the geophysicshelp provided by Nigel Phillips and NickWilliams.Thanks for being so generous with your time,and for having so much patience.The sponsors of the MDRU-GIF project, includingGeoinformatics ExplorationInc., Anglo American, Anglo Gold Ashanti,Barrick, BHP Billiton, KennecottExploration, Teck, Vale Inco, and Xtrataare thanked. Additional funding was providedby an NSERC postgraduate scholarship. AHugo Dummett Mineral Discovery Fund grantfrom the Society of Economic Geologistsprovided funding for XRDand physicalproperty analyses.Geologists and geophysicists at GeoinformaticsExploration Inc., and St. AndrewGoldfields Ltd., especially DarrenHolden (Geoinformatics), and WayneReid (formerexploration manager at St. AndrewGoldfields), are thanked for providing data,generalinformation on the Hislop deposit andsurrounding area, and use of theoffices and coreyard at Stock. Thanks to Brian Atkinson,Dave Truscott, and Ken Kryklywy forgeological tours in and around Timmins.Elisabetta Pani is thanked for XRD analyses,and Mati Raudsepp and SashaWilson for help on the SEM.Kelly Russell, Steve Quane, and KristaMichol, providedguidance for density and porositydata collection. Lisa Swinnard, LorraineTam, andMarcia Wilson, and were all thorough andwell-organized in collecting densitydata. ArneToma and Kane Smith are thankedrespectively, for helping me deal withvariouscomputer, and fmancial matters.xivThanks to my friends at UBC. You are all so smartand inspiring. And you did anamazing job of decorating the office with a stunningarray of wine bottles. To my legionof former officemates, you ladies always kept itfun and funny, and kept geologicalconversation at tolerable levels. To Victoria andKirsten and Amber, thanks for listeningto my ramblings, and hitting the slopes and thewaves hard with me between stints in theoffice.Thanks to my family for helping me to get back home once ina while, for theirinterest in my life, and of course for their constant encouragement.To Billy: thanks forhelping me with core lifting and susceptibility measuringin Timmins, for cutting myrocks for me at UBC, for not (really)asking me if I am done yet, for being ok with menot having goals or plans, and forkeeping everything together when I was too busy,i.e.almost all the time. Maybe you are the nicestperson I will ever meet?xvCo-authorship StatementChapters 2-4 were written as independent manuscripts that will be submitted forpublication to journals focusing on applied uses of geophysics, for explorationorotherwise. Each chapter involved some input from others, in the form of instruction,discussion, editing of the work, or data collection. Those who played the largestroles incollaborating are to be recognized as co-authors on thesubmitted manuscripts. Theircontributions are outlined below.Chapter 2: Physical properties of rocks in an Archean orogenic gold environmentAuthors: Dianne Mitchinson, Nigel Phillips, Elisabetta Pani, Richard TosdalNigel Phillips, a former research associate at the Mineral Deposit Research Unit,of the Department of Earth and Ocean Sciences helped to interpret someof the physicalproperty data, and edited parts of the manuscript, as wellas related posters and abstracts.Elisabetta Pani, researcher in the Department of Earth and Ocean Sciences,collected Xray diffraction data for the Hislop deposit suite of samples, and analyzed thedata usingRietveld methods to yield mineral abundance data. My thesis supervisorRichard Tosdalcontributed suggestions, and provided numerous edits of this chapter.Chapter 3: Detecting gold-related geology in Archean orogenic gold environmentsusing geophysical inversion: a synthetic modeling studybased on the Hislop golddeposit, OntarioAuthors: Dianne Mitchinson, Nigel PhillipsNigel Phillips initiated the idea of completing synthetic modelingto explore thecapabilities of inversion in the studied geologic setting, and providedsuggestions forxvipossible variations on starting models, and on inversionparameters. He providedguidance and instruction with respect to forward andinverse modeling techniques usingthe University of British Columbia GeophysicalInversion Facility (UBC-GIF) inversioncodes. He also edited the work.Chapter 4: 3D inversion of magnetic, gravity, DC resistivity,and inducedpolarization data over the Hislop gold deposit, south-central Abitibigreenstone beltAuthors: Dianne Mitchinson, NigelPhillips, Nick WilliamsNigel Phillips familiarized me with inversion codes, andwith the inversionmodeling process in general. He helped to organizegeophysical data, and providedguidance and suggestions throughout the 2Dand 3D DC resistivity and inducedpolarization modeling. Nick Williams aided withthe management and manipulation ofthe large datasets involved, instructed me on theuse of his program ModelBuilder, andprovided discussion on a number of the inversionmodel results.xviiChapter 1: Introduction1.1. COMBINING GEOLOGY AND GEOPHYSICAL INVERSION FORMINERAL EXPLORATION1.1.1. Geophysics and mineral explorationGeophysical techniques are used regularly to aid or supplement geologic mappingin areas where outcrop is limited. In addition to delineating surface geology withgeophysics, it is possible to investigate geology at depth, where otherwise subsurfacegeology must be inferred from maps and structural measurements, or by drilling.Geophysics has become an especially important tool in mineral exploration. Manymineral deposit targets produce strong geophysical signatures due to high abundances ofoxides and sulfides, allowing them to be distinguished from their host rocks. Geophysicsis so prolific in the field of mineral exploration because of the significant amount ofinformation it can provide for low costs (Phillips et a!., 2001). Regional geophysical data,usually magnetic and gravity data covering hundreds of kilometers of ground, iscommonly available for free, or at an insignificant cost, from government geologicalsurveys. From this data geology can be inferred, and large exploration targets spotted.With advanced stages of mineral exploration, an exploration company can have morefine-scale geophysical surveys completed for a higher cost, however, the price is minimalcompared to the cost of drilling.Traditionally, geophysical data collected at the surface or from boreholes isinterpreted directly after standard filtering and corrections. Estimations of sizes andshapes of features are made based on known relationships between sources and themeasurement location, and through forward modeling. The relatively recent developmentof robust geophysical inversion methods for calculation of 3-dimensional physicalproperty models of the subsurface allows petrophysically distinct geological features tobe located in 3D space, and their geometry to be delineated at significant depths of up to1thousands of meters. These methods are becoming a staple in the mineral explorationindustry 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 projectThis PhD project was completed alongside a number of others under the MineralDeposit Research Unit — Geophysical Inversion Facility (MDRU-GIF) joint researchinitiative. The MDRU-GIF project was a collaborative project involving researchers andstudents from the University of British Columbia’s (UBC) Mineral Deposit ResearchUnit, and the Geophysical Inversion Facility, in addition to ten mineral explorationindustry 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-basedexploration and generate more robust 3D subsurface models through effectivecombination of geology, physical properties, and geophysical information. A number ofmore specific themes were encompassed within this principal objective including:relating physical properties to geology and geological processes (Sterritt, 2006), scalingphysical property data for use at larger scales of inversion (Pizarro, 2008), anddeveloping methods of more effectively incorporating prior geological information intogeophysical 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 froma range of mineral deposit types including kimberlitic diamond, magmatic sulfide,orogenic gold, volcanogenic massive sulfide, and porphyry deposits, and considereddifferent stages in exploration from regional reconnaissance to deposit delineation. ThisPhD project focused on the application of geophysical inversion methods to explorationin the Archean orogenic gold environment, for a range of scales of exploration.21.1.3. Inversion in the Archean orogenic gold environmentThe Hislop deposit, a gold deposit in the south-centralAbitibi greenstone belt,acted as a representative orogenic gold deposit forthis work. Although the deposit issmall, and was only mined for a short period, it was a good candidatefor a case studydeposit for this research for a number of reasons. Due to extensiveexploration in theHislop deposit area, and in nearby surroundingareas, there is a large amount ofgeophysical data available for use in geophysical inversions. Thereare numerousdriliholes available for reconnaissance work on thelocal geology. Additionally, the areahas been mapped and modeled recently (Berger, 1999and 2002; Power et al., 2004;Reed, 2005; Mueller et al., 2006), and inversion resultscan be compared to knowngeology. Finally, it may be possible to apply concepts and results fromthis work to otherareas, as the geology of the deposit is characteristic of other orogenic golddeposits bothlocally, and globally.The intent of this PhD project was to apply knowledge oforogenic gold models,of local greenstone belt geology, and of the Hislopdeposit, to optimize the inversionprocess for this specific mineral deposit setting. Thedesired outcome was to generatesubsurface models that are consistent with known geology in orderto be able to interpretresults with confidence. The project is, in essence, a multi-facetedcase study, whichbroaches many of the themes of the MDRU-GIF project, and coversa number of stagesthat comprise the inversion process. PhD research encompassed understandingphysicalproperty — geology relationships, completing synthetic modeling todetermine inversionimaging capabilities, and carrying out unconstrained and constrainedinversions of actualgeophysical data collected over the Hislop deposit.The role of physical properties in inversion isemphasized throughout this work,as they ultimately quantitatively link geology to geophysics (Fig.1.1). Having anunderstanding of relationships between geology and physical propertiesis important forconstraining geophysical inversions, determining if physical propertyvalues composing3model results are reasonable, and of course for interpreting geology from the recoveredmodels.The entire process represented by the work in this thesis should be analogous tothe process that an exploration company might follow if embarking on completinginversion work for a prospect, or even a more well-understood deposit wherecontinuations 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 PropertiesManec SupfliIIty vs Dev, of Ibdop Rocks0GeophysicalModeling andInversion41.2. BACKGROUND TO GEOPHYSICAL INVERSIONThe UBC-GIF inversion programs are primarily used as modeling and explorationtools in this project. This thesis does not go into detail regarding the mathematics behindthe inversion codes. However, in order to understand how prior geologic information canbe accommodated in the inversion, and in order to appreciate features and anomalies thatare manifested in the inversion results, it is important to have a general knowledge ofhow the codes work.Geophysical inversion can be considered the opposite process to forwardmodeling. Forward modeling involves generating data for a known subsurface physicalproperty distribution. Forward modeling is sometimes used to determine the effect aspecific source within the subsurface has on a measured geophysical signal. Geophysicalinversion involves estimating a subsurface physical property distribution based on anobserved geophysical dataset. In this case the data are known, and the location, andphysical property value of the source must be calculated.To calculate a 3D subsurface model, a volume representing the earth is discretizedinto many model cells. A reference model or starting physical property value is assignedto the earth and physical properties within cells are perturbed over numerous iterations toattempt to fit the observed geophysical data (either collected at surface or fromboreholes). The user specifies a misfit, represented by Equation 1. The misfit isessentially a measure of the difference between the observed data, and the data predictedby 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 inversionproblem. This non-uniqueness is alleviated by the addition of more information to theproblem. Results can be constrained by formulating the inversion to achieve a model withparticular characteristics, based on prior geological knowledge. This information isincorporated into the problem through the model objective function.52N d0bs_dPaød(m)=[1Equation 1. Where N is the number of geophysical data,d102sis the observed data atlocation 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 desiredmodel is one that is close to a given reference or background model, and is smooth in alldirections. The inversion is guided toward a result honoring these specifications. Themodel objective function is represented in Equation 2, showing only the functioncontrolling closeness to the reference value, and the function controlling smoothness inthe x direction. These default parameters can be modified when more specificinformation is known about the geology. The reference value can be modified and itsdegree of influence on the result can be manipulated (c), and directionality can beinvoked by increasing smoothing in different directions by varying amounts (c). Theresulting inversion model is only acceptable if, data generated when the model is forwardmodeled (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 themodel criteria and are geologically reasonable.øm=af(m_mo)2dx+ax(m_mo)dxEquation 2. Where c’ is the alpha weighting determining the degree of closeness toreference model, c determines smoothing in the x direction, m is the model, and mo isthe reference model. In the full equation, functions in the same form as the x-smoothingfunction exist for the y and z directions.6Detailed inversion procedures and equations are found in Li andOldenburg(1996, 1998, and 2000).1.3. ARCHEAN OROGENIC GOLD - GEOLOGIC ANDGEOPHYSICALBACKGROUND1.3.1. Background on Archean orogenic goldRecent comprehensive summaries of orogenic golddeposits are given in Groveset al. (1998), Hagemann and Cassidy (2000), Goldfarbet a!. (2005), and Robert et al.(2005), and characteristics significant to the thesis are generalizedhere. Orogenic golddeposits are epigenetic, structurally controlled gold deposits that arehosted in orogenicbelts. They are generally accepted as having formedduring late stages of continentalcollision. Most of the discovered orogenic gold depositsin the world occur in greenstonebelts situated on Archean cratons in North America, Australia, andsouthern Africa.Archean orogenic gold deposits typically occur proximal to large,crustal-scalefaults, which are thought to represent the conduitsthat transported gold-bearing fluids tonear-surface from depth. These deposits can occur in any host lithology,however thereappears to be a common spatial relationship to felsic intrusive rocks,perhaps due to theirbrittle nature and ability to develop fractures, andto Fe-rich rocks, which may promotesulfidation causing gold precipitation. Hydrothermal fluids carryinggold are typicallyC02-rich and this is reflected in the carbonate-rich alterationmineral assemblages thataccompany mineralization. Gold is most commonlyhosted within or proximal to quartzcarbonate veins, but may also occur in association with disseminatedsulfides in spatialproximity to faults or shear zones.71.3.2. The Hislop gold depositThe Hislop deposit is found in the gold and base-metalrich Abitibi greenstonebelt of the Superior Province of Canada. It lies near the PorcupineDestor DeformationZone (PDDZ), a regionally important structure withrespect to gold mineralization.The general geology of the HislopTownship was mapped by Prest (1956), andmore recently by Berger (1999). A geological map ofthe eastern Timmins area basedpredominantly on interpretation of high resolution aeromagneticdata, was compiled byGeoinformatics Exploration Inc. Geoinformatics also compiled anextensive database ofgeologic logs from drilicore derived from exploration programs runby the companiesthat have explored the Hislop property over the last75 years. Berger (2002) completedan assessment on the geology and geochemistry of rocksalong the eastern portion ofHighway 101 (the ‘Golden Highway’), which follows the PDDZ,that included anoverview of the geologic setting of mineral deposits alongthis corridor. The mostdetailed work on the Hislop deposit was completed by geologistsworking at St. AndrewGoldfields Ltd at the time of mining. Someunderground maps were made, andpetrographic and lithogeochemical work completed. At the time of the commencementofthis project, the Hislop mine was closed, and most of the geologistswho had worked atthe mine no longer were with St. Andrew Goldfields. Much of the dataon the deposit thatwas collected, some in digital and some in hard copy form, was scatteredand difficult tocompile. An internal report with significant detail on the various mineralizedzones onthe Hislop property was provided for thisproject as a reference (Roscoe and Postle,1998).For this project, ten drillholes were re-logged, and a limitedamount of geologicmapping was completed at the flooded Hislop West Area open pit, and on select outcropsin the vicinity.In general, the Hislop deposit is hosted within a seriesof metamorphosed maficand ultramafic volcanic rocks. The area is structurally complex withnumerous tight folds8and faults paralleling the regional structural trend. Gold is spatially related to a contactbetween a syenite dike and an ultramafic volcanic unit. Gold is refractory withindisseminated pyrite, and mineralization is associated with carbonate and muscovitealteration.St. Andrew Goldfields Ltd. currently own the Hislop deposit property. Thedeposit is a relatively small gold deposit only mined for a few years total, producing justover 400 000 tonnes of ore, grading between 2.33 and 5.55 grams per tonne. Further goldpotential has been indicated by recent drilling and sampling programs(www.standrewgoldfields.com).1.3.3. Geophysics and goldGeophysics constitutes a useful tool in greenstone-hosted gold settings since theseenvironments are commonly characterized by scarce outcrop. In the area between themain Timmins gold camp and the Ontario-Quebec border, where the Hislop deposit issituated, there is minimal topography. The area is heavily forested, and covered withnumerous lakes.Although geophysics is heavily relied on for geologic mapping and explorationfor a variety of mineral deposits in these settings, gold deposits are a notoriously elusivegeophysical target. The deposits are typically low grade, and locally restricted, resultingin a poor petrophysical contrast between the target and its host rocks. Nonetheless, othergeological features known to be spatially related to gold, such as host rocks,hydrothermal alteration, or sulfide mineralization, might provide petrophysically distincttargets.The geophysical methods most successfully applied for gold exploration havebeen DC resistivity and induced polarization (IP) methods. These methods detectconductive and chargeable sulfides commonly associated with orogenic gold. Some9examples 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 and2005b), and Meuller et al. (2005). It is hoped that the work completed for this PhD willcontribute to an advanced understanding of the application of geophysical inversiontechniques in the Archean orogenic gold setting.1.4. PROJECT OBJECTIVESThe overlying goal of this research was to optimize the geophysical inversionprocess to explore for gold-related rocks in the Archean orogenic gold setting. The firststep in achieving this goal was to identify relationships between geology, specificallygold-related geology, and physical properties, and to delineate the key geologicalprocesses that lead to these relationships. Secondly, synthetic modeling was used ‘todetermine if typical gold—related features can be regularly detected by inversion, and ifinversion parameters can be modified to improve their detection. The final stage of thework involved applying prior geological and physical property knowledge to theinversion of four geophysical datasets covering the Hislop deposit area. The results ofthis PhD research are presented in three chapters that correlate with each of the threeresearch stages. The thesis objectives are summarized here by chapter.Chapter 2.Initial research involved defining relationships between geology and physicalproperties. As mentioned, this information is critical to any geophysical or inversionwork. It is obviously important with respect to interpreting results. However, it is alsovaluable for constraining inversions, for identifying if inversion results are sensible, andfor building synthetic physical property models to test hypotheses. Magneticsusceptibility, density, resistivity, and chargeability data were collected for Hislop rock10samples. The goals of this initial physical property work were to outline the physicalproperty ranges for the main rock types at Hislop, to understand any trends withinphysical property data, and to determine if prospective rocks could be distinguished frombarren rocks based on physical properties. Additionally, to establish whether the resultsfrom this work can be applied to geophysical exploration in other areas, physical propertydata was compared to a large regional dataset, and data from greenstone belts inAustralia.Chapter 3.Although physical property work might indicate that certain gold-related rockshave unique physical property ranges, allowing them to be distinguished from likelybarren rocks, these targets may still be undetectable through inversion. This may beattributable to: geophysical survey design, data spacing, data errors, inversiondiscretization, inversion sensitivities, and smoothing typical in inversion results.Synthetic forward and inverse modeling tests the effectiveness of inversion to delineatedesirable features in the subsurface at deposit-scales of exploration. A model based on theHislop deposit is used, however, variations are made to the initial model and to theapplied inversion parameters to explore outcomes. The research aimed to determine:whether desired targets can be imaged using inversion, whether inversion parameterscould be manipulated to get a better result, which geophysical datasets yield the mostuseful information about the subsurface, which are best for detection of gold-relatedfeatures, 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 propertydata is available, inversions are constrained locally and globally to generate models moreconsistent with known geology. The main goals of this work was to examine thesubsurface geology of the Hislop deposit area, to attempt to image specific geologic unitsor packages of rock, to locate key geologic structures in the subsurface, and mostimportantly, to identify prospective areas for exploration. Geophysical datasets most11useful for mapping geology, and for isolating mineral exploration targets were identified.Knowledge gained from physical property work and from synthetic modeling wasinvoked to assess and interpret inversion results.Chapters 2 to 4 form the basis for three manuscripts to be submitted to mineralexploration-related, or applied geophysical journals. As the three chapters represent threeseparate deliverables, there is some overlap in information between them.12REFERENCESBerger, B.R., 1999, Geological investigations along Highway 101, Hislop Township:Ontario Geological Survey, Summary of Field Work and Other Activities 1999, OpenFile 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, 124p.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: inHedenquist, J.W., Thompson, J.F.H., Goldfarb, R.J., and Richards, J.P., eds.,100thAnniversary 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 crustaldistribution 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, inHagemann, 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-bearingsulfide 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 ofExploration Geophysicists, Tulsa, Ok.,p.227-279.13Johnson, I., Webster, B., Matthews, R., and McMullen, S., 1989, Time-domain spectralIP 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, 4p.Lelievre, P., Oldenburg, D., and Williams, N., 2008, Constraining geophysical inversionswith 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.1931-1945.Mira Geoscience Limited, 2005a, Detectability of mineral deposits with electricalresistivity and induced polarization methods: Ontario Geological Survey, MiscellaneousRelease—Data 181.Mira Geoscience Limited, 2005b, Detectability of mineral deposits with potential fieldmethods: 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 presentationof geophysical data for geoscientific profiles in the Timmins—Kirkiand Lake area:14Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6189 , 28p.,15sheets.Phillips, N., Oldenburg, D., Chen, J., Li, Y., and Routh, P., 2001, Cost effectiveness ofgeophysical inversions in mineral exploration: Applications at San Nicolas: The LeadingEdge, 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 the3D integration of exploration data: KEGS Inversion Symposium, PDAC 2007, extendedabstract, 9p.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, 177p.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., andArchibald, N. J., 2004, Geoinformatics evaluation of the eastward extension of theTimmins Gold Camp: Geoinformatics Exploration Inc., Unpublished report for StAndrew Goldfields Ltd.Prest, V.K., 1956, Geology of the Hislop Township: Ontario Department of Mines,Annual Report, 1956, v. 65, pt. 5, 51p.Reed, L. E., 2005, Gravity and magnetic three-dimensional (3D) modeling: DiscoverAbitibi Initiative, Ontario Geological Survey, Open File Report 6163, 40p.,4 sheets.Robert, F., Poulsen, K.H., Cassidy, K.F., and Hodgson, C.J., 2005, Gold metallogeny ofthe Superior and Yilgam Cratons, in Hedenquist, J.W., Thompson, J.F.H., Goldfarb, R.J.,15and Richards, J.P., eds.,100thAnniversary 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 thecontext of geologic processes in the ultramafic rock-hosted mineral deposit environment:aiding interpretation of geophysical data: Unpublished M.Sc. thesis, The University ofBritish Columbia, 172p.Williams, N.C., 2008, Geologically-constrained UBC—GIF gravity and magneticinversions with examples from the Agnew-Wiluna greenstone belt, Western Australia:Unpublished Ph.D. Thesis, The University of British Columbia, 479p.16Chapter 2: Physical properties of rocks in an Archean orogenicgold environment12.1. INTRODUCTION2.1.1 RationaleIn order for geophysical inversion to be knowledgeably interpreted, it isimperative to (1) have an understanding of the rock types, alteration, and mineralizationthat typify the geological environment, and (2) possess an understanding of thecharacteristic ranges of physical properties associated with this geology. Ideally, physicalproperty studies should be conducted on the range of representative rock types from thegeological 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 datasetsfor typical physical properties of rock types in a specific environment, however thisinformation is commonly limited and the effects of hydrothermal alteration on theprotolith are rarely considered.Once a clear understanding of the relationships between various physicalproperties and rock types, alteration, and mineralization are established, this informationcan be used to interpret and guide geophysical inversions. If unique relationships arepresent and can be statistically characterized, physical property model data generatedfrom inversion can be queried for prospective ranges, or filtered to yield mineralogicalinformation (Williams and Dipple, 2005). Knowledge of physical property ranges typicalof a given geological environment can indicate whether an inversion has yielded realisticvalues. Additionally, the inversion algorithm can be manipulated to incorporate priorphysical property information to drive the inversion toward a result more consistent withexpected geology (Ellis and Oldenburg, 1994; Li and Oldenburg, 1996). Understandingphysical property behavior, and having confidence in the data being used to constrain1A version of this chapter will be submitted for publication. Mitchinson, D., Phillips, N., Pani, E., andTosdal, D., 2009, Physical properties of rocks in an Archean orogenic gold environment.17inversions is critical; changinginversion parameters, or using referencemodels toconstrain inversions, canchange a model significantly (Phillips,2002; Williams, 2006).A physical property study ofthe Hislop deposit aims toprovide a detailedinvestigation into physical propertyrelationships within an Archeanorogenic golddeposit environment. Physicalproperties considered are magneticsusceptibility, density,resistivity, and chargeability.An important goal of these studiesis to identify the physicalproperty datasets, alone, orin combination, which are most effectivein detecting Archeanorogenic gold-relatedmineralization or proxies to mineralization.2.1.2 ObjectivesThe objectives of this research areto:1. Review the key characteristicsof Archean orogenic goldenvironments, and thegeophysical methods commonlyemployed in exploration for them;2. Characterize the principalhost rocks, alteration characteristics, andstyles of goldmineralization at Hislop;3. Document relationshipsbetween physical properties androcks at Hislop throughpetrographic work and mineralanalyses;4. Explain the controls onphysical property variations;5. Outline magneticsusceptibility, density, resistivity, andchargeability ranges thatspecifically characterize the hostrocks, alteration mineral assemblages,andmineralization at Hislop;6. Define the most usefulphysical properties for targetingpotentially mineralizedrocks at Hislop;7. Assess whether physicalproperty values are representative ofArchean orogenicgold settings elsewhere.182.2. BACKGROUND2.2.1 Geology and geophysics of Archean orogenic gold depositsGeological characteristics of Archean orogenic gold depositsOrogenic gold deposits are epigenetic, structurally-controlled gold deposits hostedin metamorphosed orogenic belts (Groves et al., 1998). This work focuses specifically onthe physical property analysis of rocks associated with orogenic gold deposits hostedwithin an Archean age greenstone belt setting. Although Archean orogenic gold depositsare 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 betransported 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 asstockwork mineralization (Roberts, 1988; Hodgson, 1993; Hagemann and Cassidy, 2000;Goldfarb et al., 2005). The Archean gold deposits considered in this study do not includeArchean-age placer, or banded iron formation (BIF)-hosted gold deposits.Archean orogenic gold deposits have for many years been an important source ofgold 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 explorationfor these types of deposits, and an initiative to improve exploration methods for theirdiscovery.19Geophysical characteristics of orogenic gold depositsGold 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 inArchean orogenic gold deposits, in contrast to massive-style mineralization represented involcanogenic massive sulfide deposits or nickel sulfide deposits, which form largergeophysical targets in distinct contrast to host rocks. Defining alternative targets, orindicators, with known relationships to gold, and sufficiently distinct physical propertycharacteristics, is required to fully utilize geophysical tools (Seigel et al., 1984; Doyle,1990).Geophysical methods used to target gold-associated structures, host rocks, andalteration zones include magnetics, gravity, electrical methods (DC resistivity andinduced polarization), and electromagnetic methods. Table 2.1 lists various geologicalfeatures commonly related to gold mineralization, and examples of the geophysicalmethods that are most effective in targeting them. Ideally some combination oftechniques can be employed to target a variety of gold-related features at a particularlocality, 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 areaRegional geological settingThe Superior Province of the Canadian Shield is the largest Archean craton onearth. It is composed of a number of northeast-trending, amalgamated volcano-plutonic,granitic-gneissic, and sedimentary terranes (Card and Ciesielski, 1986). Boundaries of theterranes, or subprovinces, are structural or metamorphic zones that juxtapose contrasting20geological and geophysicalterranes (Card andCiesielski, 1986; Card,1990; Williams etal., 1991). The studyarea for this projectis located within thesouth-central Abitibisubprovince, or greenstonebelt (Fig. 2.1).Table 2.1. Geophysicalcharacteristics of Archeanorogenic gold depositsFeature ScaleGeophysical characterMethods of detection1 Greenstone RegionalOverall low, but ‘rough’ magneticAirborne magneticsterranes 1000 kmscharacterGranitoids commonlylower density than Airbornegravitygreenstone2 Large scaleRegional to Low magneticsignature attributed toAirborne/ground magneticsfaults districtoxidation/alterationl00kms E . .. .High/low resistivity zones dependanton DC resistivitydegree of annealing3 Lithological DistrictVarious depending on physicalproperties Various dependingon rock type ofmarker units 10 kmsof rock type of interestinterest4 Hydrothermal LocalMagnetic lows resulting fromdestruction Airborne/groundmagneticsalteration 10 mof magnetite; less commonlymagnetichighs, due to influx of Fe-richfluidsHigh resistivity if silicificationDC resistivity5 MineralizationLocal Disseminated sulfideassociation with DC resistivityand Induced10 m gold - conductiveand chargeable PolarizationMagnetic pyrrhotiteMagnetics if pyrrhotite is mainFe-sulfide associated with goldI) Grant, 1985; Isles etal., 1989; Doyle, 1990; Williamset al., 1991; Gunn and Dentith,1997; 2) Henkeland 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) Holsnerand Schneer, 1961;Grant, 1985; Harron et al.,1987; Doyle, 1990; Williams,1994; Lapointe et al.,1986; Johnson et a!., 1989;Doyle 1990; 5) Johnson etal., 1989; Seigel et al., 1984,Doyle, 1990; Hallof and Yamashita,1990; Dockeryetal., 1984.Many of the golddeposits in Abitibi greenstonebelt gold camps,like theTimmins-Porcupine, andKirkland Lake camps,are spatially related toprominent, largescale crustal structures,including the east-westtrending Porcupine-DestorDeformationZone and Larder-Lake-CadillacDeformation Zone (Colvineet al., 1988;Kerrich, 1989;21Hodgson and Hamilton, 1990; Jackson and Fyon, 1991). Most gold deposits are notlocalized by these larger “first order” faults, but by secondary or tertiary splays (Kerrich,1989; Robert, 1990; Hodgson, 1993; McCuaig and Kerrich, 1998; Hagemann andCassidy, 2000).Figure 2.1. Approximate location of the Hislop study area in the Abitibi greenstone beltof the Superior Province. Modified after Card and Ciesielski (1986).Hislop deposit geology and gold settingThe Hislop deposit area is underlain mainly by interlayered mafic and ultramaficvolcanic rocks (Fig. 2.2). The volcanic rocks are complexly folded and are presentlyaligned northwest-southeast. They are intruded by coarse-grained syenites, fine-grainedquartz-feldspar phyric rhyolite dikes, and dacitic to andesitic dikes, usually alongnorthwest-southeast trending faults (Prest, 1956; Berger, 1999; Power et al., 2004).ioLocation ofStudy AreaSuperior province22Gold is localized near thenortheast and southwest contactsof an elongate,approximately 30 m -100 mwide, northwest-trending syenite (Cooper,1948; Prest, 1956;Roscoe and Postle, 1998;Berger, 1999 and 2002), as depictedin the cross-section in Fig.2.3. The majority of goldat Hislop is associated with disseminatedpyrite within, what isrecorded in mine and geologicalsurvey documents as, “carbonate-breccia”,south of thesyenite (Cooper, 1948; Prest,1956; Roscoe and Postle, 1998).The carbonate breccia ispredominantly a strongly carbonate-alteredbrecciated equivalent of anultramafic unit atHislop. Gold also occurs to alesser extent within quartz veinlets,stockworks andfractures in mafic volcanicflows north of the syenite (Roscoeand Postle, 1998).Generally, there is little goldwithin the syenite, with the exceptionof weakmineralization occurring within a zoneapproximately 3 m from the southerncontact withcarbonate breccia. (Cooper,1948). High gold grades at Hislopare also associated withrhyolite porphyries, which are foundas narrow, discontinuousintrusive bodies in maficand ultramafic units south of thesyenite (Fig. 2.2).A number of northeast-trending,sinistral separation cross-faultsoffset the syeniteand bounding mafic andultramafic flows in places (Cooper,1948; Prest, 1956; Power etal., 2004). Gold-bearing zoneswiden, and gold grade commonlyincreases where thesecross faults intersect mineralizationalong the syenite (Roscoe and Postle,1998)Two principal mineralized zones,the Shaft Area and the West Area(Fig. 2.2),were mined by St. AndrewGoldfields, Ltd., at Hislop overthree separate intervalsbetween 1990 and 2006.In 1990 and 1991, 215 990 tonnes ofore grading 5.55 g/t weremined, between 1999 and 2000,185 100 tonnes or ore grading3.4 g/t were mined, andrecently in 2006, 10147 tonnesof ore grading 2.33 g/twere mined(www.standrewgoldfields.com).23I I LDO — late diorite/dolerite_____SSG - greywackeI SLO — mudstone - siltstonejSoc — sediment, undividedIFD/IFO — felsic intrusive dyke!felsic intrusive undividedI I 100 — intrusive, undivided______ISO — syenite intrusive, undivided_____IVFO — felsic volcanic, rhyolite, rhyodaciter, I VLJO — ultramafic volcanic, undivided2VMF — magnetic mafic volcanicI VMO — mafic volcanic, basalt, aridesiteFigure 2.2. Geology of the Hislop deposit area as interpreted by Power et al. (2004) fromhigh 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 10drill holes (one overlapping) logged for this study. The cross-section shown in Figure 2.3is based on core logging of three drill holes that were drilled in the West Area.24DDH H9601 DDH Ext 280, GK 280, and H9605Figure 2.3. Cross section looking northwest through the Hislop deposit, showinglocations of carbonate-dominated alteration and gold mineralization. Cross sectioninterpreted from drill core logged from the West Area of the Hislop property.2.3. METHODOLOGY2.3.1 Field and mineralogical studiesTen drill holes from the 1996 and 1997 St. Andrew Goldfields Ltd. drill programswere 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, approximatelyMulti-lithic Volcanic BrecciaLamprophyric DikeIntermediate DikePorphyritic Rhyolite DikeSyenite IntrusiveMafic Volcanic RockUltramafic Volcanic RockFaultDrill trace2510 m by 10 m were mappedin detail for outcrop scale studies ofmagnetic susceptibility(Appendix 2B).Petrographic, and mineralogicalstudies (scanning electron microscopeand X-raydiffraction studies) allow forcharacterization of host rocks andalteration mineralassemblages within theHislop deposit area. This work alsoconstrains geologicalprocesses that control physicalproperty variations. The presence,abundance, andcomposition of minerals,such as magnetite, pyrite, and carbonate,which haveparticularly significant influenceson physical properties, weredocumented. Whereaspetrographic and SEMwork defines the minerals presentin the various samples,quantitative XRD work usingRietveld refinement methods,described by Raudsepp andPani (2003) and outlined in Appendix2C, contributes relative mineralproportions for 37samples at Hislop. This quantitativeinformation is useful for comparisonsto physicalproperty data, and for calculationsof density data.2.3.2 Physical property measurementsMagnetic susceptibilityMagnetic susceptibilitydata from Hislop was recordedusing a hand-heldmagnetic susceptibility meter,the Exploranium KT-9 Kappameter.Susceptibilities arereported in i0 SI Units. Magneticsusceptibility readingswere taken every 5 m alongdrill core for all drill holesre-logged for this project.Measurements were made onallsamples collected from drillcore and from outcrop. Magneticsusceptibility readingswere taken at 10 differentpoints over each sample, andthe average value was used inanalyses of this data. Magneticsusceptibility readings were alsocollected systematicallyover six roughly 10 m2grids over mapped outcropsto understand controls onsusceptibility at the surface atoutcrop scale. Typically 2-5 readingswere taken at eachsite and the average was used.In total magnetic susceptibilitywas determined for 432samples. Greater than 1000additional readings were collectedfrom drill core and26outcrop. Corrections applied to susceptibility measurements to account for core diameter,and split core intervals are outlined in Appendix 2C. The magnetic susceptibility datasetrepresents the largest physical property dataset from the Hislop physical property study.DensityDensity measurements were made for 414 drill core and hand samples fromHislop using the buoyancy or hydrostatic method and calculations outlined by Johnsonand 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 themass of the sample submerged in water. To obtain the mass of the sample in water, thesample is placed on a tray which is suspended from a weighing scale positioned above asmall 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.Pwis the density of the water, which isassumed to be 1 g/cm3.Density is reported in g/cm3.Additional density measurementswere 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 laterinterpretations of some trends in the Hislop physical property data, it was of interest tocalculate porosity. Porosity is calculated from dry and saturated rock masses. Isolatedporosity (inaccessible to air or water) is not accounted for by this method. The equationused is (Cas and Wright, 1987):= 100*(W3- W1)/(W3- W2)27where4)is porosity, W1 is the mass of the oven-dried sample in air, W2 is the mass of thesample submerged in water, and W3 is the mass of the water-saturated sample in air (Casand Wright, 1987). Porosity is reported as %.Resistivity and chargeabilityResistivity and chargeability data for 67 representative drill core and handsamples were measured by Zonge Engineering and Research Organization, Inc.Resistivity and chargeability measurements are collected simultaneously after sampleshave been moisture-saturated. They are made in time-domain. A current is establishedbetween opposite ends of the samples using a constant current transmitter, whichconducts currents as low as 100 nA. Resistivity is calculated based on the length, andcross-sectional area of the sample, the amplitude of the current, and the change inpotential recorded across the sample. Resistivity is reported in Ohm-rn. Conductivity canbe calculated from resistivity by taking the inverse value. Conductivities are expressed inS/rn. The chargeability of a sample is based on the rate of decay of the voltage after theapplied current is turned off. For the Hislop samples, it was determined using an 8 secondperiod, 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 handsample, may control measurements made in the field. As such, measurements ofresistivity taken on drill core or hand samples are commonly higher than measurementsmade in-situ (www.zonge.com/LabIP.html). This must be considered if sample-scaleresistivity measurements are to be used to constrain geophysical inversions. Chargeabilitydata collected from core or hand samples are thought to be sufficiently representative oflarger scale measurements (www.zonge.com/LabIP.html).282.4. DATA AND OBSERVATIONS2.4.1. Hislop deposit rock types, hydrothermal alteration, and associated mineralogyRock TypesAll rocks at Hislop have been metamorphosed to greenschist facies, however, forsimplification purposes, the prefix meta- is herein ignored. The rock protoliths arerecognizable based on textures and characteristic metamorphic mineral assemblages, andthe protolith name is used hereafter. The five principal rock types at Hislop are,ultramafic volcanic rocks (predominantly komatiites), mafic volcanic rocks (tholeiiticbasalts), 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, andmulti-lithic volcanic breccia units. Descriptions and typical mineralogy of the mainHislop rock types as determined through petrographic, scanning electron microscope, andX-ray diffraction work, are given in Table 2.2, and detailed results of XRD mineralabundance analyses are found in Appendix 2D.Hydrothermal alterationThe most common hydrothermal alteration mineral assemblage at Hislop is acarbonate+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 volcanicrocks and intermediate dikes, and as either Fe/Mg-carbonate (ankerite to dolomite) +muscovite alteration, or magnesite (Mg-carbonate)+fuchsite (Cr-muscovite) alteration inultramafic volcanic rocks. Carbonate + muscovite alteration was noted in drill core andoutcrop to occur near faults and contacts, and in proximity to syenite and rhyolite29Table 2.2. Summary of the principal rock types found in the Hislop deposit area, and associated mineralogy.mag, +1- ph, +1- spB. Fe-carbonate + muscovite: cb(dolo), qtz, ms, chi, +1-a b, +1- micC. Magnesite + fi,chsite: mgs/dolo,qtz, ms (flich), chi, +/-il, +1- cr -spab, chi, aug, act, cal, qtz, ep, ms, clzo,uivo, magB. Fe-carbonate: ab, mic,ank/dolo/sid, ms, ±1- py, +1- chiLeast-alteredUltramafic volcanic rocksHandSampleDescriptionDark black to brown (oxidized) incolor; soft; commonly sheared; fine tomedium grained textures; qtz +7- cbveins; relict spinifex textures adjacentto margins of some flows; relictcumulate textures represent internalparts of the flow.Dark grey-green to dark purple color;massive or pillowed (+l_ variolites)flows up to lOOm thick; fine tomedium grained rock; pillows arewell-preserved with qtz + cb filledamygdules increasing near margins;thin chl +1- ep altered selvedges.chl, dol, qtz, +1- tIc, +1- hbl, +1- magDark grey to mauve in color; massive,homogeneous, and fine-grained; hblfsp phyric, hbl-phyric, or aphyric;phenocrysts approximately 1mm inlength and euhedral; groundmass isvery fine-grained and composedmainly of microlitic fsp.Pink to mauve in color; massive; verycoarse-grained; composed ofmesoperthitic potassium fsp and ab;fsp crystals up to 3 cm; rareinterstitial mafic minerals.ab, cb, ma, qtz, .4- chl, +/-mag, +1- pyMineralogyAlterationAlterationmineralogyMassive; generally fsp (+7. qtz)porphyritic; aphanitic grey to pinkgroundmass composed of very finegrained fsp and qtz; fsp phenocrystscomprise 3% — 80% of the rock; qtzphenocrysts make up 2 - 5% of therock; minor mafic minerals.ab, kspar, dol/ank/cal, py, +1- qtz, +7-ma, +1- chlsb, qtz, mic, chl, +1- dolms, ab, sid, dol, qlz, chi,B. Fe-carbonate + albile (+7-)quartz: ab, dolo/ank, qtz, mic, chl,mag4. 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- pyms, ab, qtz, pyB. Fe-carbonate: ab, qtz, ank, ms, +7-B. Fe/Mg-carbonate: ab, cb, ma, py4-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; ulvulvospinel; (Fe iron; Mg magnesium).sole: mineralogy based on I sampleforach ofA and Bintrusions. Most of the mined Hislop ore came from an Fe-carbonate-altered ultramaficbreccia, as such, carbonate-related alteration is considered an important vector tomineralization. Carbonate + muscovite alteration is commonly mapped as a distal,pervasive alteration surrounding Archean orogenic gold deposits, and this is the case formany gold deposits elsewhere in the Abitibi greenstone belt (Fyon and Crockett, 1983;Hodgson, 1990).Fe-carbonate+ albite alteration occurs at Hislop over narrow intervals withinmafic volcanic rocks in drill core near some of the known high grade gold zones. Albiterich 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 syeniteintrusives and porphyritic rhyolite dikes. Fe-carbonate alteration affects these rocks to alesser extent. The overall lack of Ca, Mg, and Fe in felsic rocks at Hislop hinders theformation of carbonate minerals when exposed to CO2 rich fluids, as discussed in Roberts(1988).Most prospective rock types and alteration at HislopFrom 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; Poweret al., 2004), it is possible to outline the prospective rocks at Hislop. Rock types known tohave a close spatial relationship to gold at Hislop include Fe-rich volcanic rocks, syeniteintrusive rocks, and porphyritc rhyolite dikes. Carbonate dominated hydrothermalalteration is frequently associated with gold in these deposits, and is known to be relatedto gold at Hislop.312.4.2. Physical properties of the Hislop depositAll physical property measurements made on Hislop deposit samples, includingmagnetic susceptibility, density, resistivity, chargeability, and porosity measurements, arecompiled in Appendix 2E. Descriptive statistics, and correlation coefficients for physicalproperties and mineral abundances can be found in Appendices 2F and 2G, respectively.Magnetic susceptibilityMagnetic susceptibility logsSelected geology and susceptibility logs from various parts of the Hislop propertyillustrate the behavior of magnetic susceptibility associated with characteristic rock typesand 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 arehigh susceptibility. Susceptibilities are also high where ultramafic rocks are characterizedby talc-chlorite metamorphic mineral assemblages (depicted by the dark green color inColumn 2 of the drill logs in Fig. 2.4).There is a regular drop in magnetic susceptibility where Fe-carbonate + albitealteration, Fe-carbonate + muscovite alteration, or magnesite+fuchsite alteration hasbeen superimposed on mafic and ultramafic volcanic rocks. Less pervasive, weaklyfracture-focused alteration, such as an Fe-rich dolomite alteration that lends a pink colorto some intermediate dikes and mafic volcanic rocks, do not appear to have a consistenteffect on magnetic susceptibility values.There are seemingly no obvious patterns between altered syenite intrusives andporphyritic rhyolite dikes and magnetic susceptibility. Other rock types generally occuras very narrow units, and susceptibility readings for these rocks are sporadic.32LithologyMulti-lithic volcanic brecciaLamprophyric dikeIntermediate-mafic dikePorphyritic rhyolite dikeSyonite intrusiveMafic volcanic rockUltramafic volcanic rockAlterationWeak to moderate Fe-cb + msStrong Fe-cb+maFe-cb + abTic-chl metamorphic assemblageMg-Gb (magnesite) + ms (fuchsite)ChloritejSericiteFe-cb + ab(intermediate dikes)Ms/ser (syenite and rhyolite dikes)Fe-cb (syenite and rhyolite dikes)Pink Fe-rich dol veinsEpidote veinsHematite - pervasiveHematite along fracturesMagnetite2Figure 2.4. Geology, alteration,magnetic susceptibility, and goldgrade logs for four Hislop drillholes logged for this study. Themost consistent susceptibilitytrends include: low susceptibilityof felsic intrusive rocks, highsusceptibility of talc-chloriteassemblage ultramafic volcanicrocks, and some mafic volcanicrocks, and low susceptibility ofcarbonate-altered ultramafic andmafic volcanic rocks. Forexplanations of abbreviations inlegend see bottom of Table 2.2._________Au abundances between 0.15- 1 ppmAu abundances between 1 - 5 ppmAu abundances > 5 ppmColumn 1: LithologyColumn 2: AlterationColumn 3: Magnetic Susceptibility (x103 SI Units)*notall intervals sampled33Magnetic susceptibility data - all rock samplesMagnetic susceptibility data collected from drillcore and hand samples aresummarized in a series of histograms (Fig. 2.5).A wide range of susceptibilities,spanning 2 and 3 magnitudes, characterize the main rocktypes at Hislop. The histogramsshow a steady decrease in magnetic susceptibility valuesfrom ultramafic to felsic rocks.Mafic and ultramafic rocks have distinct bimodalmagnetic susceptibilitydistributions. Extended ranges of susceptibility forintermediate and felsic rocks may beattributed to a small number of outliers.40302010864286420.01 0.1 1 10 100Magnetic Susceptibility (x iO SI Units)Figure 2.5. Magnetic susceptibility histograms for thefive main rock types found in theHislop deposit area. Mean values are given for general comparison,however the meanmay not be an appropriate descriptor for populations withbimodal distributions.10S1284Mean 12 O7- ntermedtedes_JlJItfhmFL fl .n HThMean Syenite intrusives -:1.19- I -100034Magnetic susceptibility data - least altered and altered rock samplesAs was indicated in the magnetic susceptibility logs, some of the variation withinultramafic and mafic rock data may be attributed to effects of alteration. When magneticsusceptibility data for ultramafic and mafic rocks at Hislop is subdivided into least-altered, and carbonate-altered populations, it is apparent that the carbonate-alteredpopulations have lower overall magnetic susceptibilities (Fig. 2.6). Intermediate dikesdisplay 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 resultsin a minimal decrease in susceptibility for these intrusive rocks.DensityDensity data - all rock samplesFrom density histograms (Fig. 2.8), there is a decrease in density from ultramaficto felsic rocks. Narrow ranges in density characterize syenite intrusive rocks andporphyritic rhyolite dikes. Ultramafic and mafic rock densities span a larger range thandensity values for intermediate and felsic rocksDensity data - least altered and altered rock samplesAlteration of ultramafic rocks correlates with a slight increase in average densityrelative to least-altered ultramafic rocks. There is a minor decrease in average density forFe-carbonate + albite altered mafic volcanic rocks (Fig. 2.9). Intermediate dikes undergoa marginal density increase with Fe-carbonate+ muscovite alteration (Fig. 2.10). Thereare no significant changes in densities between unaltered and altered equivalents ofsyenites and rhyolite dikes at Hislop - data peaks are generally consistent between thesubpopulations (Fig. 2.10).354242a) Variably altered ultramafic volcanic rocks0.01 0.1 1 10 100Magnetic Susceptibility (x i0 SI Units)155302010642b) Variably altered mafic volcanic rocksFigure 2.6. Magnetic susceptibility histograms showing susceptibility data for a) least-altered and altered ultramafic volcanic rocks, and b) least-altered and altered maficvolcanic rocks.84864Mean 6.03 Least-altered dolomite+chloritelageMean5.731Fe/Mg-carbonate+rnuscoviteI IllililiMean 0.75Magnesite+fuchsite1000[Mean 40.O9Least-altered chlorite+albiteflasmbla{yMean 6.86] Fe-carbonate+muscoviteIMean1.71*[iriFe+carbonate+albite0.01 0.1 1 10 100 1000Magnetic Susceptibility (x iO SI Units)364263642b) Syenite intrusives0,01 0.1 1 10 100 1000Magnetic Susceptibility (x 10 SI Units)42a2f-1Least-Iteredrhyolite dikesMean O.35[MucoviteflMean 0.48JFe/Mg-carbonate.ri U0.01 0.1 1 10 100 1000Magnetic 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--1intermed. dikes.LHFPMean 0.74 Fe-carbonaterL+muscoviteMean 10.95Fe/Mg-carbonate.I I0.01 0.1 1 10 100Magnetic Susceptibility CX 10 SI Units)1000[Mean 2 87jLest-aIteedI syenite intrusives14264242I. Muscovite.Mean 0.36Fe/Mg-carbonateMean 0.23c) Porphyritic rhyolite dikes42372015105Figure 2.8. Density histograms for the five main rock types found in the Hislop depositarea.Jitramafic volcanic rocksJiiaiIntermediate dikes2515584105105IilTiiPorphyritic rhyolite dikesMean 2.70, II i—I1F1—f—[_i-i fl ri2.4 2.5 2.6 2.7 2.8 2.9Density (glcm3)3.0 3.1 3.23820151052015105Density (g/cm’)2.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2Density (g/cm3)Figure 2.9. Density histograms showing density data fora) least-altered and alteredultramafic volcanic rocks, and b) least-altered and altered mafic volcanic rocks.a) Variably altered ultramafic volcanic rocks422015105642642Least-aIterei dolornite+Mean 2.85-chlorite assemblage -TaIc+chlorite4Fe/Mg-carbonate÷Mean 2.87Jmuscovite -- I I IMagnesite+fuchsite[Mean 2.922.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.23932326422.4642642432c) Porphyritic rhyolite dikesLeast-altered -rfrhyolite dikesMean 2.69I—IMuscovite -i r—fLMean 2.68Fe/Mg-carbonateMean 2.73—2.4 2.6 2.8 3.0 3.2Density (gIcm)Figure 2.10. Density histograms showing density data for a) least-altered and alteredintermediate dikes, b) least-altered and altered syenitic dikes, and c) least-altered andaltered porphyritic rhyolite dikes.40a) Intermediate dikesMean 2.83 Last-lterddikesn 2 85Fe-carbonateMean2.6 2.8 3.0 3.2Density (gIcm)b) Syenite intrusivesLeat-aIkred’eniteintw:s. MuscoviteIMean 2.72Fe/Mg-carbonateLMn.2.4 2.6 2.8 3.0 3.2Density (gIcm)8642642432ResistivityResistivity data - all rock samplesRanges of resistivity for Hislop rocks are large and overlap one anothersignificantly (Fig. 2.11). However, they are roughly comparable to published data forsimilar rock types (Tab. 2.3). From resistivity histograms, it is evident that ultramaficrocks have the lowest resistivities of the five main rock types at Hislop. Mafic volcanicrocks, intermediate dikes, syenite intrusives, and porphyritic rhyolite dikes have similaraverage resistivities. There was insufficient sample numbers to evaluate effects ofhydrothermal alteration on the various rock typesChargeabilityChargeability data - all rock samplesRanges of chargeability values for the various Hislop rock types generally overlapone another, with some outliers (Fig. 2.12). Hislop chargeability data falls into thechargeability ranges considered to be characteristic of these rock types (Tab. 2.3),although many of these published chargeability ranges largely overlap. As withresistivity data, there were too few samples to compare the effects of alteration onchargeability values for the five rock types.41Figure 2.11. Resistivity histograms for Hislop deposit rocks. Data indicates lower overallresistivities for ultramafic volcanic rocks from Hislop.Table 2.3. Ranges of resistivity and chargeability for rock types similar to those occurringin the Hislop deposit area (data from Telford et al., 1990).Rock Type Resistivity (Ohm-rn)feldspar porphyry 4 x 13(wet)porphyry (various) 60 - i04syenite 102- 106andesite 1.7 x 12(dry)basalt 10- 1.3 x i07 (dry)peridotite 6.5 x (dry)calcarious/mica schists20 -Rock Type Chargeability (ms)schists 5-20precambrian volcanics 8-20dense volcanic rocks 100-500granites, granodiorites 10-502-8 % sulfides 500-10008-20% sulfides 1000-200020% sulfides 2000-30001000 1(Resistivity (Ohm-rn)42108642642Ultramafic volcanic rocksMaflc volcanic rocks -Intermediate dikes -LMean 29.24Syenite intrusivesMean 14.71tPorphyritic rhyolite dikes_______IMean 10.4710 100Chargeability (milliseconds)Figure 2.12. Chargeability histograms for Hislop deposit rocks. Chargeability ranges forthe individual rock types overlap and are not unique.2.5. INTERPRETATIONS2.5.1. Effect of geological processes on physical properties at HislopMagnetic susceptibilityFrom petrographic, SEM, and XRD work, it was established that magnetite is theonly significant magnetic mineral in the Hislop deposit rocks. The trend of decreasingmagnetic susceptibility from ultramafic to felsic rocks observed at Hislop, reflectsdecreasing magnetite abundance. A plot of modal magnetite, as derived from XRD andRietveld analyses, plotted against magnetic susceptibility (Fig. 2.13) shows a positivecorrelation between these data, supporting this interpretation.IIMean 6.91[426426421000439Rock Types8 Ultramafic rocksMaflc rocksX intermediate dikes6 ASyenite intrusivesORhyo porphyries21I I0 20 40 60 80 100Magnetic Susceptibility (x104 SI Units)Figure 2.13. Positive correlation between modal magnetite in Hislop rock samples (asderived from XRD analysis) and magnetic susceptibility. For calculated correlationcoefficients see Appendix 2G (all rock types).Large susceptibility ranges for the major rock types at Hislop are not atypical andresult from the broad range of mineralogy that can be encompassed under a given rockclassification; classification schemes do not normally take into account oxide and sulfideaccessory minerals, the minerals primarily controlling susceptibility (Clark, 1997).Bimodal populations are common in magnetic susceptibility data and are interpreted torepresent distinct populations whereby Fe has partitioned mainly into paramagneticminerals (weakly magnetic phases including silicates and carbonates) or intoferromagnetic minerals (strongly magnetic minerals, such as magnetite and pyrrhotite). Achange in magma composition, or in oxidation state, may cause a rock to fall into onepopulation or another (Clark, 1997).Magnetite in mafic volcanic rocks, intermediate dikes, and felsic rocks at Hislopis primary igneous magnetite. Magnetite is not typically a primary igneous mineral inkomatiitic rocks as chromite is the principal spinel that forms (Clark, 1997). Magnetiteforms in ultramafic rocks usually as a product of serpentinization of olivine during early,retrograde metamorphism (Bucher and Frey, 2002).44In mafic rocks, the decrease in magnetic susceptibility with Fe-carbonate +muscovite, and Fe-carbonate+albite alteration is predominantly attributed to theconversion of magnetite in mafic rocks to Fe-carbonate upon exposure to C02-richhydrothermal 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 correlationbetween 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 acarbonate altered zone surrounding a carbonate vein in a mafic volcanic rock fromHislop. Plane polarized and reflected light photomicrographs.45a,0Dolomite + Ankerite + Siderite (Wt. %)Figure 2.15. Modal magnetite versus total Fe-rich carbonate abundance for all Hislopsamples with measured quantities of these minerals. Bubble size represents relativemagnetic susceptibility. A decrease in susceptibility correlates with a decrease in modalmagnetite and an increase in Fe-rich carbonate abundance.Magnetite in ultramafic rocks formed during serpentinization is thought to besimilarly affected by carbonate alteration. Clark (1997) explains that upon interactionwith C02-rich fluids, magnetite is first redistributed in ultramafic rocks, and then isdestroyed. Conversion of magnetite to Fe-rich carbonate was not directly observed inultramafic rocks during petrographic or mineralogical work on Hislop rocks, however,the lack of magnetite in carbonate-altered ultramafic rocks compared to least-alteredequivalents, is assumed to be due to alteration-related magnetite destruction.Some intermediate dikes, porphyritic rhyolite dikes, and syenite intrusive rockshave low abundances of primary igneous magnetite, thus typically magnetite-destructivealteration does not affect magnetic susceptibility significantly (Fig. 2.10).Hydrothermal alteration processes affecting mafic and ultramafic volcanic rocksdo not explain all of the measured variation in magnetic susceptibility, as is indicated byadditional heterogeneity in susceptibility readings from recorded unaltered intervals indrill core (Fig. 2.4). Variations in magnetic susceptibility in the absence of obvious46hydrothermal alteration couldbe related to a range of factors. Based onthe rock typesand mineralogy at Hislop, themost likely factors causing variable susceptibilityingenerally unaltered rocks at Hislop includean uneven distribution of primaryorsecondary magnetite, grain size, and irregularoxidation of magnetite to formhematite.An uneven primary distributionof magnetite in mafic rocks,and unevensecondary distributions of magnetite inultramafic rocks may explain non-alterationrelated magnetic susceptibilityvariations in these rocks. Some small scalevariations mustbe expected, as rocks are notlikely to be perfectly homogeneousin their modalmineralogy. Formation of magnetite in amafic volcanic rock is dependant onmanyfactors including the magmacomposition, the degree ofdifferentiation, and thetemperature and pressure conditionsunder which the rock is formed ormetamorphosed(Clark, 1997). For ultramafic rocks, theformation of magnetite from olivine duringserpentinization may be influenced bylocation of fluid pathways in therock.Small magnetite grain sizes areusually more susceptible than larger grainsizes asthey do not easily retain remnanantmagnetism (Clark, 1997). To examinethe role ofvisible grain size in non-alterationrelated variations in magnetic susceptibility,least-altered fine-grained andmedium-grained samples are plotted separately.Magnetite grainsize here is assumed to be consistentwith the overall grain size of the samples.Theresulting histograms (Fig. 2.16)illustrate that fine-grained, and medium-grainedmaficand ultramafic rocks have similarranges and distributions of magneticsusceptibility, andsimilar average susceptibilities. Thus,variations in magnetic susceptibilitydata for theserocks are not likely to be stronglycontrolled by grain size.In some mafic and ultramaficrock samples, hematite rimsmagnetite grainsindicating some oxidation of theserocks has occurred. A consistent pattern relatedto aparticular alteration event, or having specificlithological or structural control, wasnotrecognized during petrographic or mineralogical(SEM and XRD) analyses. Irregularoxidation of magnetite to hematite in maficflows however, could contributeto decreasesin magnetic susceptibility unrelatedto hydrothermal alteration in maficrock samples.47a) Mafic volcanic rocks1054325432b) Ultramafic volcanic rocksI I IFine-grained- Mean = 1.06 x103 SI UnitsFigure 2.16. Histograms showing distribution of susceptibility for fine- and mediumgrained a) mafic volcanic rocks, and b) ultramafic volcanic rocks. Similar distributionsbetween fine- and medium-grained subsets indicates that grain size is not a major controlon susceptibility at Hislop.Fine-grained40 -Mean0.77 xl o SI Units30 —20 —Medium-grained— Mean = 1.09 x103 SI Units0.01 0.1 1 10 100 1000Magnetic Susceptibility (x iO SI Units)Th[14321F1Medium-grained- Mean 0.75 x103 SI Unitsmimiri0.01________Ii0.1 1 10 100 1000Magnetic Susceptibility (x iO SI Units)48DensityMineralogy and porosity are considered to be the main controls on density atHislop. Both mineralogy and porosity are affected by geological processes includingigneous fractionation! differentiation, metamorphism, and hydrothermal alteration.Mineralogy plays a significant role in determining rock densities. Igneous andvolcanic rock densities generally decrease with increasing Si02 content (Johnson andOlhoeft, 1984; Telford et al., 1990), reflecting an increase in the abundance of lowdensity felsic minerals, and a corresponding decrease in the abundance of higher densityFe- and Mg-rich mafic minerals. This is consistent for Hislop samples. From Table 2.4, itis apparent that minerals that typically characterize ultramafic and mafic rocks at Hislopare 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).________________________MineralDensity (glcm3)Quartz2.62Microcline2.56Albite2.62Actinolite3.04Epidote3.45Augite3.4Chlorite 2.65Muscovite 2.82Calcite 2.71Ankerite 3.05Siderite 5Dolomite 2.84Magnesite 3Talc 2.75Serpentine 2.53Pyrite 5.01Magnetite 5.15Hematite 5.349The modal mineralogy of syenites and porphyritic rhyolite dikes brings abouttheir narrow density ranges. They are dominated by a small number of similarly denseminerals, specifically quartz and feldspar. Densities of ultramafic and mafic rocks span alarger range of densities than those making up intermediate and felsic rocks which is aresult of their more complex and varied mineralogy (refer to Tab. 2.2).A slight increase in the average measured density of ultramafic volcanic rockscorresponds with Fe/Mg-carbonate+muscovite, and magnesite + fuchsite alteration. Thisrelationship can be attributed to changes in mineralogy accompanying alteration. Basedon published mineral densities (Tab. 2.4, mineral densities from www.mindat.org), achange 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 andMg-rich carbonate, plus muscovite, and quartz, should theoretically result in a denserrock. Carbonate minerals are expected to have a significant influence on rock density. Onaverage, they are denser than those silicate minerals that dominate the mineralogy ofigneous 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 thecorrelation between increasing density values with increasing Fe-rich carbonateabundance in Fe-carbonate bearing Hislop deposit samples.Minor variations in the density of mafic volcanic rocks from Hislop may besimilarly attributable to alteration. The lower average density values for Fe-carbonate +albite altered samples, as compared to least-altered and Fe-carbonate + muscovitesamples, is considered to be related to bulk mineralogy (Fig. 2.9). The increased relativeabundances of low-density albite in rocks with Fe-carbonate and albite-dominatedalteration 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 comeabout through oversimplified sample groupings. There is little change in density betweenthe variably altered syenites and porphyritic rhyolites (Fig. 2.10). It is assumed that for50these rock types, bulk mineralogy changes do not add or subtract significant denseminerals, and thus alteration has little influence on the density of these rocks.50Rock TypesC Ultramafic rocksX- 40• Mafic rocksX Intermediate dikes+A Syenite intrusives,. 0 Rhyolite porphyries3UI- ..P. x20Ax1000AAA= 0.562.60 2.70 2.80 2.90 3.00Density (glcm3)Figure 2.17. Density increases for Hislop rocks with an overall increase in the abundanceof Fe-rich carbonate. For calculated correlation coefficients see Appendix 2G (all rocktypes).Calculating densities from modal mineralogy as determined from XRD analysisand published mineral density data helps to determine what the densities of the rockshould theoretically be, if the density is controlled solely by mineralogy. When comparedto measured densities, discrepancies will indicate that there are factors aside from bulkmineralogy affecting the rock. Density is calculated simply by using volumeconcentrations of minerals (C) and their grain densities (p) as given in Johnson andOlhoeft (1984):p= Ci*Pi+C2*P2+C3*P3...C*p51A lack of strong correlation between some of the measured and calculateddensities for ultramafic and mafic rocks (Fig. 2.18), suggests that there may be othercontrols on density. Two possible explanations for the incongruity include not accountingfor porosity in samples, and limitations in mineral identification using XRD methods.3.403.203.002.802.602.402.40 2.60 2.80 3.00 3.20 3.40Measured density (glcm3)Figure 2.18. Measured versus calculated density for Hislop rocks. Discrepancies betweenthe density values obtained from the two methods for ultramafic and mafic volcanic rockscould 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 densityvariations at Hislop. To test the possible influence of porosity on the density of mafic andultramafic rocks, a suite of samples in varying states of alteration were measured forporosity using the method described in section 3.2 Figure 2.19 shows that there is anoverall negative correlation between density and porosity for ultramafic rocks at Hislop.Talc-chlorite assemblage rocks are most porous and least dense in accordance with theirtypically strong foliation. Strongly carbonate-altered samples have lower porosities andhigher densities. Figure 2.20 indicates no obvious relationship between density andporosity for mafic volcanic rocks.= 0.69(excluding outlyingintermediate dike sample)52Unaltered to AlteredUltramafic RocksDLst, altd. (Dol-chi)•TIc+cN ultramafic3DFeCb+ms afld. utiramaficDFe/MgCb+fu altd ultraniafic•1r2=0.280 I I I2.70 2.75 2.80 2.85 2.90 2.95 3.00 3.05Density (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. Thisbrings 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 correlationcoefficients see Appendix 2G (ultramafic rocks).Unaltered to Altered —Mafic Rocks —Lst. altd. (Chi-ab)FeCt,+ms altd. niafic3 oFeb+ab altd. mafic>2.70 2.75 2.80 2.85 2.90 2.95 3.00 3.05Density (glcm3)Figure 2.20. No relationships are indicated between porosity and density for maficvolcanic rocks at Hislop. Abbreviations: Lst. altd. = least altered; chl+ab = chlorite+albite; FeCb+ms = Fe-carbonate + muscovite; FeCb+ab Fe-carbonate+ albite. Forcalculated correlation coefficients see Appendix 2G (mafic rocks).53Discrepencies between measuredand calculated density data could also beattributed to generalization of modalmineralogy during Rietveld analysis. For someminerals that exist as a solid solution, suchas dolomite and ankerite, a proper nameis notassigned for intermediate compositions.The density values for these end members differsignificantly, and the resulting calculated densitywould be affected accordingly if oneend member mineral classification was chosen overthe other.For intermediate and felsic intrusive rocks, calculatedand measured densityvalues match closely, indicating primarilymineralogical control on density. Changesindensity between unaltered and altered versions of theserock types thus must be explainedby relative additions or subtractions of more and lessdense minerals.ResistivityLeast-altered metamorphosed volcanic and igneousrocks at Hislop haveresistivity ranges similar to published ranges forequivalent rocks types (Tab. 2.3).Published resistivity ranges for most mineralsare very large and not as specific as densityvalues for given minerals. This makes itdifficult to assess the combined resistivityaffectsof minerals making up a rock. This beingsaid, the role of mineralogy on resistivity atHislop is thought to be minimal. The majorityof minerals making up Hislop rocks arepoor to intermediate conductors, or resistors(>1 Ohm-rn; Telford et al., 1990). Mostsulfides, and some oxides, are known tobe good conductors (low resistivity, <1 Ohm-rn),and there is a small percentage of these mineralsin Hislop samples.Variations in resistivity at Hislop are interpretedto be primarily controlled byrock texture and porosity. Resistivity isknown to drop considerably with increasingwater content of rocks (Telford etaL, 1990), thus to be related to the porosity of a rock(Halloff, 1992). As such, the low averageresistivity of ultramafic rocks compared to theother Hislop rock types is interpreted tobe a result of the relatively high porosities of54talc-chlorite assemblage ultramafic rocks, the most common ultramafic rocksubpopulation sampled during this study.Resistivity is plotted against magnetic susceptibility and density (Fig. 2.21 andFig. 2.22), two properties shown to vary with alteration in ultramafic and mafic volcanicrocks at Hislop. In Figure 2.21a, the ultramafic samples are the only samples to outline atrend between resistivity and magnetic susceptibility. With ultramafic samples colored torepresent their dominant alteration assemblages, it is obvious that the trend is related toalteration. This variation in resistivity is interpreted to be related specifically to alterationeffects on porosity. Figure 2.23 demonstrates that a decrease in porosity of ultramaficrocks with carbonate alteration causes the rock to become more resistive. Thus, alteredultramafic samples are resistive and, as was indicated previously, are characterized bylow magnetic susceptibilities due to magnetite destruction. When resistivity is comparedwith density (Fig, 2.22), again a weak correlation emerges only for ultramafic samples.When colored based on alteration, the relationship of increasing resistivities and densitieswith carbonate alteration is apparent for the majority of the samples, and is explained bya decrease in porosity for altered rocks.ChargeabilityThe main control on the chargeability of Hislop rocks is thought to be thepresence of disseminated sulfides. Disseminated sulfides in rocks are readily chargeablewhere subjected to an induced current, due to the chargeable nature of the metallic grainscoupled with the large surface area provided by a disseminated texture (Telford et al.,1990). Other known controls on chargeability include presence of clay minerals andgraphite, both of which are absent from Hislop rocks.A positive relationship between pyrite abundance based on XRD analyses, andchargeability 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.551000000 —ci.. 100000110000____1000RockTypes15C Uftramafic rocksMafic rocksI VUx Intermediate dikesA Syenite intrusives10Rhyolite porphyries_________________________________________________0.01 0.1 1 10 1001000Magnetic Susceptibility (x103SI Units)10000001b.— 10000010:0:Unaltered to AlteredUltramafic Rocks100DLst. altd. (Dol-chi)•TIc+chl ultramafic—DFeCb+msaltd. utiramafici— 0.53DFe/MgCb+fu altd. ultramafic (ultramafic rocks)10 II0.01 0.11 10 1001000Magnetic Susceptibility (x103 SIUnits)Figure 2.21. Resistivity versusmagnetic susceptibility,a) Ultramafic volcanic rocksamples indicate a trendbetween these physical properties,whereas variations inresistivity and magnetic susceptibiltyare more irregular for other rocktypes. b) Whendata points are colored to representthe various ultramafic alterationassemblages, it isapparent that the relationshipbetween restivity and susceptibilityis controlled in part bycarbonate alteration. For abbreviations,see Fig. 2.19. For calculatedcorrelationcoefficients see Appendix2G (ultramafic rocks).561000000 —a.100000E10000xAAxn1000Rock TypesCUItramaflc rocksMaflc rocks1 00x Intermediate dikesA Syeriite intrusives10ORhyolite porphyries I2.60 2.65 2.70 2.75 2.80 2.85 2.90 2.95 3.00Density (glcm3)1000000 -b.10000010000tA4<AUnaltered to AlteredUltramafic RocksQLst. altd. (Dol-chi)I‘“‘ •TIc+chl ultramafic —DFeCb±ms altd. utiramafic —10DFe/MgCb+fualtd. ultramafic II (ultromafic rocks)2.60 2.65 2.70 2.75 2.80 2.85 2.90 2.95 3.00Density (glcm3)Figure 2.22. Resistivity versus density. a) As with resistivity versus magneticsusceptibility, trends in data when plotted based on rock type are not obvious. b)Subdividing ultramafic rocks based on alteration assemblage reveals that increasingresistivities and densities can be to some extent attributed to carbonate alteration. Forabbreviations, see Fig. 2.19. For calculated correlation coefficients see Appendix 2G(ultramafic rocks).57Unaltered to AlteredUltramaflc RocksQLst. altd. (Dol-chl)•TTC+chl ultramafic3 DFeCb+ms altd.DFe/MgCb+fu altd. ultramafic0—----—- ---0_i•n E1LU r2=O.700 I10 100 1000 10000 100000Resistivity (Ohm-rn)Figure 2.23. A plot of porosity versus resistivity shows that annealing of ultramafic rocksdue to precipitation of carbonate minerals during hydrothermal alteration brings about adecrease in porosity and a corresponding increase in resistivity. For abbreviations, seeFigure 2.19. For calculated correlation coefficients see Appendix 2G (ultramafic rocks).10RocktypesO Ultramaflc rocksMafic rocksX0Xlnterrnediate dikesAASyenite intwsives -o Rhyolite porphyriesA2AAA0o 0r2=O.53(felsic rocks, excludingrhyolite dike outlier)01.00 10.00 100.00Chargeability (ms)Figure 2.24. A weak positive correlation exists between pyrite abundance andchargeability, however the trend is mainly controlled by porphyritic rhyolite dike andsyenite samples. There is no evidence of a consistent relationship between chargeabilityand pyrite abundance for intermediate to ultramafic volcanic rocks. For calculatedcorrelation coefficients see Appendix 2G (felsic rocks).58This indicates that there may be variables affecting chargeability otherthan, or inaddition to, sulfide abundance. Sulfide grain size andtexture, and the relationshipbetween sulfide grains in the rock, are all potential factors that can influencethe rock’schargeability (Pelton et a!., 1978). As chargeability should increasewith increasedsurface area of sulfide minerals, chargeability values may depend on whethersulfides aredisseminated, concentrated in a stockwork system, or controlled by fractures or veins.Variable porosity may affect the chargeability of mafic volcanicrocks.Chargeability can decrease with porosity; increased fluid pathway volumecan be moreconducive to electrolytic conduction, prohibiting polarization. Forexample,chargeabilities may be higher for a crystalline igneous rock containingdisseminatedsulfides, than for a more porous sedimentary rock containing sulfides,(Telford et al.,1990). Although the dataset is small (few samples haveboth chargeability data andporosity), there is a weak relationship between porosity and chargeabilityfor mafic rocksat Hislop (Fig. 2.25).10IeMafir2=0.2401.00 10.00 100.001000.00Chargeability (ms)Figure 2.25. A negative correlation between chargeabilityand porosity in this plotindicates that increases in porosities of mafic volcanic rocks at Hislopmay hinder theability for metallic minerals to become charged. For calculated correlationcoefficientssee Appendix 2G (mafic rocks).592.6. DISCUSSION2.6.1. Exploration using physical propertiesPhysical properties most useful for isolating prospective rocks at the Hislop depositThe most useful physical properties for delineating prospective rocks at Hislopfrom those more likely to be barren are magnetic susceptibility and density. Magneticsusceptibility and density are equally capable of discerning prospective syenite intrusiverocks and porphyritic rhyolite dikes at Hislop from intermediate, mafic, and ultramaficrocks (Fig. 2.26). These physical properties however, do not distinguish betweenhydrothermally altered and least-altered felsic rocks, as mineralogical changes in theserocks related to alteration processes do not add or remove any significant quantities ofdense or magnetic minerals.3.203.10—. 3.002.902.802.7002.602.502.400.01 0.1 1 10 100 1000Magnetic Susceptibility (x103SI Units)Figure 2.26. Magnetic susceptibility plotted against density for Hislop samples. Syeniteintrusives and porphyritic rhyolite dikes have distinctly low density and magneticsusceptibility ranges allowing them to be distinguished from intermediate, mafic, andultramafic rocks at Hislop.Rock TypesUltramsfic rocksMafic rocksX Intermediate dikesASyenite intrusiveso Rhyolite porphyriesc04A)x0A60Magnetic susceptibility and density ranges for intermediate, mafic and ultramaficrocks are large, and generally overlap. These rock types cannot be independentlydistinguished from one another based on these two physical properties. That being said,potentially prospective carbonate-altered intermediate, mafic, and ultramafic rocks havetypically low magnetic susceptibilities; carbonate-altered rocks almost exclusively occurin the lower susceptibility ranges for these rocks (Figs. 2.27a and 2.27b). Thus, when thislow range is isolated, the majority of carbonate-altered rocks are targeted. Unfortunately,due to variability in magnetite abundance and distribution in intermediate to ultramaficrocks, and irregular hematization of magnetite, there are relatively unaltered, low-susceptibility rocks at Hislop. Carbonate-altered rocks at Hislop cannot be exclusivelydelineated from a physical property dataset as a result of this overlap. Nonetheless,targeting low susceptibility rocks would be effective in delineating many prospectivecarbonate-altered rocks from high susceptibility rocks more likely to be barren ofmineralization.Density provides an additional measure of alteration of ultramafic volcanic rocksonly. If ultramafic rocks were isolated, density values could be used to delineate thehigher density magnesite + fuchsite rocks from other ultramafic rocks, specifically thosewith lower density talc + chlorite assemblages.Resistivity may be useful in distinguishing ultramafic rocks from other rocks inthe Hislop physical property dataset, however this is likely of no significance with respectto mineralization, as these rock types are not uniquely mineralized. If dealing solely withultramafic rocks however, higher resistivity values may be indicative of carbonatealtered, low-porosity ultramafic volcanic rocks. Chargeability values do not distinguishbetween rock types at Hislop. Although there may be a relationship between pyriteabundance and chargeability for felsic rocks, there are likely other influences on thechargeability of rocks at Hislop, like the texture of sulfides, or that of the host rock itself.613.20a.3.10_. 3.00C.,2.90P0,>2.80(0,0____________2.60Unaltered to AlteredMafic RocksILst. altd. (Chi-ab)2.50 IIOFeCb+ms alid. maficOFeCb+ab altd. maf’ic2.400.1 1 10 100 1000Magnetic Susceptibility (x103SI Units)3.20 -___________________3.10b.r’3.00 D2900,2.80—BB___________2.70Unaltered toAlteredUltramafic Rocks2.60QLst. altd. (Dol-chi)•TIc+chl ultramafic2.50FeCb+ms altd. utiramaficFe/MgCb+fu altd. ultramafic2.40 I0.1 1 10 100 1000Magnetic Susceptibility (xl O SI Units)Figure 2.27. Carbonate-alteration destroys magnetite in a) mafic and b) ultramaficvolcanic rocks, causing magnetic susceptibility to drop. Density values increase slightlyfor altered ultramafic rocks. For abbreviations in a) and b) see Figs. 2.20 and 2.19,respectively.62Prospective physical property rangesMagnetic susceptibility and density constitute the two most well understoodphysical properties at Hislop. They were determined to be the most useful of the fourphysical properties studied in delineating some of the prospective rocks at Hislop. Table2.5 summarizes the prospective ranges for magnetic susceptibility and density for Hislop.These ranges were established using the statistical analysis program SPSS Statistics, andanomalously high and low values (extreme cases occurring beyond 3x the interquartilerange of values) were eliminated to yield a tighter, more representative, range of valuesfor each of the rock types.These prospective cut-off values are used to query the Hislop physical propertydatabase for the purposes of determining the effectiveness of these cut-offs to distinguishbetween possible gold-related rocks and rocks likely to be barren. The dataset wasqueried 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)fromthe 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 querywhere 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 thedatabase, 55 were recalled, falling within the statistically significant susceptibility anddensity ranges for these rocks, yielding a 79% success rate. However, this query yielded73 samples in total, out of which 18 were not felsic intrusive rocks, thus mislabeling 25%of the results.63Table 2.5. Statistical data for prospective rocks at Hislop, and cut-off values used for querying physical property data.Rock TypeMagnetic Susceptibility (1O- SI) Density (glcm3)Cut-off values for querying dataNo. Mean Median Range No. Mean Median Range Rock Type Mag. Sus. DensityUnaltered ultramafic 8 6.03 1.49 0.57-12.5 8 2.85 2.86 2.82-2.89(dolomite-chloriteassemblage)Unaltered ultramafic (talc- 46 24.12 14.61 0.44-84.4 46 2.85 2.84 2.79-2.94chlorite assemblage)Fe-carbonate-muscovite 16 5.73 1.01 0.41-5.96 15 2.87 2.85 2.80-2.91altered ultramaficCarbonate-altered0.41-5.96 2.80-2.96ultramaficMagnesite-fuchsite altered 9 0.75 0.62 0.49-0.95 9 2.91 2.92 2.85-2.96ultramaficUnaltered mafic 107 40.09 21.50 0.35-141 101 2.87 2.87 2.70-3.08Fe-carbonate-muscovite 75 6.86 0.60 0-2.19 71 2.85 2.86 2.78-2.97altered maficCarbonate-altered0.2-2.19 2.76-2.97maficFe-carbonate-albite altered 14 1.71 0.58 0.28-1.27 13 2.82 2.81 2.76-2.86maficUnaltered intermediate 11 41.11 18.40 0.24-135.29 11 2.83 2.81 2.72-2.95intrusiveCarbonate altered 22 10.95 0.58 0.13-3.8 22 2.79 2.78 2.67-2.95intermediate intrusiveCarbonate-alteredintermediate 0.13-3.8 2.67-2.95Carbonate-muscovite 10 0.75 0.69 0.32-1.55 10 2.85 2.86 2.76-2.94intrusivealtered intermediateintrusiveSyenite 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 2.57-2.80Table 2.6. Results from magnetic susceptibility and density queries of the Hislop physicalproperty dataset.Target rock Total target rock Total recalled No. target rock % target rock % target rock samplessamples in samples from samples recalled samples out of out of total recalleddatabase query using query known amount in samples (2)database (1)Felsic intrusive rocks 70 73 55 79 75Generally-altered 188 221 142 76 64intermediate, mafic, andultramafic rocksCarbonate-altered 146 221 112 77 51intermediate, mafic andultramafic rocksOf 188 variably altered intermediate, mafic, and ultramafic samples (this includessome obscure alteration types not thoroughly reported on in this work, in addition tocarbonate 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 unalteredsamples, thus 36% of the resulting sample set were misclassified. Out of 146 totaldominantly 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, uponexamination, are largely unaltered low susceptibility mafic volcanic rocks that overlapthe physical property ranges of carbonate-altered mafic volcanic rocks.Although some unprospective, low susceptibility rocks would inevitably betargeted, many of the barren rocks are eliminated from consideration. Results of suchqueries would not provide definitive targets for exploration, but could act as importantmineral vectoring criteria for consideration in association with any other geological,geophysical, geochemical, or mineralogical data available from the area.65Physical properties and 3D geophysical inversion modelingIt is anticipated that physical property cut-off values similar to those used to targetprospective samples from the Hislop physical properties database would be equallysuccessful when applied to 3D physical property models generated from geophysicalinversions in the Hislop area. However, the number of rock types at Hislop, and theirstructurally complicated relationships to one another, would make direct referencing tospecific rock types and alteration assemblages based on physical property data difficult.At larger scales of modeling low magnetic susceptibility values may be effective inisolating felsic rocks and strongly carbonate-altered intennediate, mafic and ultramaficrocks. Density information would help further confirm identification of felsic rocks,isolating them from other magnetic susceptibility lows. With perhaps more localizedinversion modeling, smaller scale variations in physical properties, like for example,subtle changes in mafic and ultramafic units related to the presence of felsic intrusions orof carbonate-alteration zones, could become apparent in regions that appear to be morehomogeneous at a larger scale.The use of physical property data to highlight mineralization, or prospectivegeology and alteration, would generally occur at a later stage in exploration when anacceptable inversion model has been established for a property or deposit. Prior to thisstage, physical property data can play an important role in guiding geophysicalinversions. Knowledge of characteristic physical property values of rock types from thearea of exploration, and of any relationships between physical properties andmineralization, can be input into the inversion to constrain it, which can significantlyimprove the inversion result (e.g. Williams, 2006).662.6.2. Comparison to analogous areasComparison to regional variations in physical property dataAn important goal of this work is to compile a dataset of typical physicalproperties expected to occur within a representative Archean orogenic gold environment,for future use in guiding and interpreting inversions both at Hislop and in similar mineraldeposit environments. Before this data is used, however, it is important to determinewhether the physical property values and ranges from Hislop represent those typicallyfound in this environment.A regional physical property study covering Matheson and Kirkland Lake areas tothe west and south of the Hislop deposit area, respectively, was completed for a largesample set of over 1000 samples (Ontario Geological Survey, 2001). Magneticsusceptibility, density, and resistivity were measured. Comparing the magneticsusceptibility and density data from the OGS study to the Hislop data helps to define thelocal extent to which these physical properties vary in this part of the Abitibi greenstonebelt. This dataset was assessed and rock types considered to be equivalent to the primaryrock types at Hislop were compiled and subdivided. Histograms comparing OGS physicalproperty data to Hislop data are presented in Figure 2.28 and Figure 2.29.Magnetic susceptibilityMafic and ultramafic rocks from the two studies have similar magneticsusceptibility distributions. Fe-carbonate-altered mafic volcanic rocks from the Mathesonand Kirkland Lake areas have similar data distributions as Fe-carbonate altered rocksfrom the Hislop area (Fig. 2.30). It was not possible to compare any other altered rockdata from the OGS dataset to similar altered rocks from Hislop as no other samples in theOGS dataset were subdivided based on alteration assemblages.671296312963128443286421086428642Ultramafic Volcanic RocksSyenite Intrusives0.1 1 10 100Magnetic Susceptibility (x 10 SI Units)Figure 2.28. Magnetic susceptibility histograms comparing dataequivalent rocks from surrounding regional areas.from Hislop rocks, and68Hislop studyFL jn -Mafic Volcanic RocksHislop study352515S604020Intermediae IntrusiveHislop study -Istudy[illU6420.01 1000201510510864225155302010108642432129638642129638642Ultramafic Volcanic Rocks2.4 2.5 2.6 2.7 2.8 2.9Density (g/cm’)Figure 2.29. Density histograms comparing data from equivalent rock typesfrom Hisloprocks, and equivalent rocks from surrounding regionalareas.69Mafic Volcanic RocksIntermediate IntrusivesHiskp studyOGStudyfhSyenite IntruivesHislop studyGSstudyFeldspar PorphyriesHisIopstudyIgOGS tudyr3.0 3.1 3.2Least-altered rocks40302010Hislop maflc ., -volcanic rocks I1- JLI W-OGSmafic -voIcanicro_fl,01 0.1 1 10 100 1000Magnetic Susceptibility (x io SI Units)0.1 1 10 100Magnetic Susceptibility (x iO SI Units)Figure 2.30. A comparison of magnetic susceptibility data associated with least-alteredand carbonate-altered mafic rocks from the Hislop deposit, and from the greatersurrounding area.Susceptibilities for local and regional intermediate intrusive samples overlap,however, there are very few regional samples overall. A comparison of syenite and felsicintrusive magnetic susceptibility data from the two datasets illustrates that regionally,there is more variation in magnetic susceptibility of these rock types than what isrepresented in the Hislop area, with three to four populations distinguishable. Acomparison between Hislop and OGS data indicates outliers in Hislop syenite and felsicintrusive data may fall into the higher susceptibility ranges observed for similar rocks inthe OGS dataset. The large regional range in susceptibilities may make it difficult todiscriminate higher susceptibility, magnetite-rich syenites and felsic intrusives frommafic and ultramafic rocks at the regional scale. If it could be determined that low-susceptibility syenites are more commonly associated with mineralization, then theoverlap would not cause a problem for physical property based exploration, and may70201510512080400Carbonate-altered rocksHislop maiicvolcanic rocks-_OGS maficVolcanicRofl[.....Fl1050.01 1000actually allow for the identification, based on susceptibility, of prospective syenites froma larger syenite database.DensityDensity distributions for Hislop ultramafic rocks and the regional scale ultramaficrocks are similar, however there is a gap in data in the OGS dataset between 2.85 and2.90 g/cm3.This may be attributable to the smaller size of the OGS ultramafic rockdataset, which has about half the number of samples of the Hislop dataset. The variety ofultramafic 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 densityvalues compared to Hislop intermediate dikes, however, the Matheson and Kirkland Lakesamples, are much fewer in number.Mafic volcanic rock, syenite intrusive, and felsic intrusive data from these studiesare not as comparable to one another. There are a greater number of high density maficrocks regionally than at Hislop, suggesting there are high density regional scale maficrocks that are not represented at Hislop.Regionally, syenites have slightly lower densities, and felsic intrusive rocks haveslightly higher densities than the equivalent Hislop rocks. Magnetic susceptibility data forregional syenites and feldspar porphyries indicated that there are multiple populationsthat exist for these rock types that were not recognized or sampled at Hislop. Thedifferent subpopulations of these rocks at the regional scale likely differ in mineralcomposition, which would explain the inconsistencies between Hislop and OGS sampledensities.Where separated into least-altered and altered rock populations (Fig. 2.31), rangesof density for least-altered mafic rocks are generally equivalent for the local and regionaldatasets, with the exception of the previously mentioned high density population in theunaltered regional mafic rock dataset. Densities of carbonate-altered mafic rock suitesfrom the individual studies also overlap, however data populations in each sample set do71not 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-alteredmafic volcanic samples in the OGS dataset. Perhaps these samples were mislabeled, orincompletely labeled originally and actually represent a population of anomalously highdensity carbonate-altered mafic rocks not encountered at Hislop. This however wouldinfer that there is an alteration process which yields higher densities in mafic volcanicrocks, which was not observed during Hislop physical property studiesLeast-altered rocks20- Hislop maficvolcanic rocks10 -5426 272.82.93.03.1 3.2Density (g/cm3)Carbonate-altered rocksHislop mafic20 -volcanic rocks -10 - -OGS mafic-volcanic rocks -105— -______r11112.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2Density (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 collectedusing a Bartington MS-2 susceptibility meter) and techniques used to collect the physicalproperty data, some discrepancies would be expected between the two datasets.Additional differences may arise from misplacement of samples from the OGS datasetinto incorrect rock categories for comparison to the Hislop sample suite. As there were nodetailed descriptions for the OGS samples, it was not possible to be entirely confident in72assigning the samples to the proper congruent categories. Finally, lack of correspondencebetween some populations may be the result of rock types not being sampled with equalfrequency at the local and regional scales.In summary, regional magnetic susceptibility data is more representative of localHislop rocks than density data, especially with respect to intermediate, mafic, andultramafic rocks. Least-altered mafic and ultramafic rocks from the local and regionaldatasets 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 ofdelineating a significant proportion of carbonate-altered rocks from suites of intermediateto ultramafic volcanic rocks throughout the larger area. Density values are not asconsistent between the different scale studies, and are less useful in targeting particularrock types or alteration assemblages at the regional scale. However, since regionalsyenite densities are always as low as Hislop syenite densities, or lower, these importantrock types may be distinguishable at the regional scale using appropriate physicalproperty cut-offs.Effect of metamorphism on physical property data.Although greenschist facies rocks are the typical hosts for Archean orogenic golddeposits, these deposits also occur, albeit to a lesser extent, in amphibole or even highergrade 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 ofthe Weebo/Wildara and Southern Cross greenstone belts in the Yilgarn Craton, Australia,reveals similarities and differences in physical property data from similar geologicalenvironments of varying metamorphic grade (Boume et a!, 1993). Metamorphism caninvoke changes in mineralogy or texture that can significantly influence the physicalproperty value of a rock. An excellent example is the formation of magnetite duringserpentinization. Bourne et al. (1993) have shown that densities and magnetic73susceptibilities are higher overallfor amphibolite facies ultramaficand mafic rocks thanfor greenschist facies ultramaficand mafic rocks. Theyexplain that increasesin densityof mafic rocks of amphibolefacies grade is due tothe destruction of lowdensityplagioclase (2.61-2.77g/cm3) to form hornblende(3.02-3.45 g/cm3)fromactinolite/tremolite. Ultramaficrocks of higher metamorphicgrade have increaseddensities relative toless metamorphosed ultramaficrocks which is related tothereplacement of serpentineand talc (2.7 g/cm3),by olivine (3.3 g/cm3). Magneticsusceptibility increases withmetamorphic grade inboth mafic and ultramaficrocks dueto increased magnetitecontent by volume in amphibolitegrade rocks, and increasedmagnetite grain sizes whichincreases low-field magneticsusceptibility. The increaseinsusceptibility with metamorphicgrade in ultramafic rocksis not consistent withtheresults of Clark et al.(1992) from the Agnew-Wilunabelt of the Yilgarn Block.Adecrease in susceptibilitiesof ultramafic rocks with metamorphicgrade in the AgnewWiluna belt may indicatethat hydrothermal alterationplayed a larger rolein destroyingmagnetite that wasformed during serpentinization.Since rock composition influencesthe products of hydrothermalalteration, forvariably metamorphosedrocks there will bedifferent alteration mineralproducts(Meuller and Groves,1991; McCuaig and Kerrich,1998). These variationsin alterationmineral assemblagesare generally consistent betweengold deposits in rocks ofthe samemetamorphic grade.Thus, as long as there are noother significant physicalproperty-altering variables atwork competing with mineralogicalcontrols, some predictionscanbe made regarding the physicalproperty characteristicsof hydrothermally alteredzonesin metamorphosed rocks. Anexample of a mineralogical changerelated to increasedtemperatures and pressuresof hydrothermal alteration-relatedsulfide precipitation thatwould have a particularlystrong effect on physicalproperty behavior, isformation ofpyrrhotite instead ofpyrite as the main gold-relatedsulfide (McCuaig and Kerrich,1998;Hagemann and Cassidy,2000). This is ahigh susceptibility mineral inits monoclinicform. The presenceof monoclinic pyrrhotitewould increase the susceptibilityofmineralized areas, and couldprovide an importantvector to gold mineralization.74Effects of alteration of metamorphic mineral assemblages must always beconsidered. Roberts (1988) explains how amphibolite facies rocks are known to behydrothermally altered in the Archean orogenic gold setting to mineral assemblagesreminiscent of a retrograde metamorphic assemblage (with chlorite, quartz andcarbonate), or to an alteration assemblage similar to the amphibolite facies mineralassemblage (with biotite, garnet, anthophyllite, cummingtonite, cordierite, gedrite, and,staurolite). These changes to mineralogy will likely affect physical properties, such asmagnetic susceptibility and density, which are known to be strongly controlled bymineralogy.2.7. CONCLUSIONSMineralogical and textural modifications within and between the different rocksuites explain many of the physical property variations at Hislop. These are related to therange of geological processes, including igneous differentiation/fractionation,metamorphism, and hydrothermal alteration, that have affected the rocks throughout theirhistory. The physical properties most useful for detecting prospective rocks at Hislop aremagnetic susceptibility and density. This study illustrates predictable relationshipsbetween low susceptibility values and prospective felsic intrusive rocks and carbonate-altered mafic and ultramafic rocks in the immediate Hislop deposit area. Low densityvalues will help confirm the presence of felsic rocks.The magnetic susceptibility and density cut-off values used to query the Hislopdataset in this study are considered useful for targeting prospective rocks in the Hisloparea within physical property datasets generated from drill core measurements. The samecut-offs could be used for locating prospective areas within a 3D physical property modelgenerated from geophysical inversions. Due to overlap between less prospective maficand ultramafic volcanic rocks with low modal magnetite and prospective, carbonatealtered rocks, any low susceptibility targets would have to be considered alongside otherexploration criteria. The cut-off values could be used to filter physical property data as afirst pass method of eliminating areas most likely to be barren.75In addition to using physical properties as a means to delineate prospective rocksin the Archean orogenic gold deposit environment, mean values, ranges, and standarddeviations of physical property data for the different rock types, and alteredsubpopulations can be used to constrain geophysical inversions.Although there is some indication of relationships between hydrothermally alteredrocks and resistivity there are not enough electrical property data to confidently use theserelationships to identify prospective rocks. Far more magnetic susceptibility and densitydata were collected and analyzed during the course of this study than resistivity andchargeability data, and as such, there is increased confidence in interpreting magneticsusceptibility and density data, and 3D susceptibility and density inversion models.By comparison to a more regional scale physical property dataset, the Hislopphysical property dataset is generally representative of rocks in this part of the Abitibigreenstone belt, with the exception of there being a greater variability in compositions offelsic intrusives at the regional scale. The Hislop dataset may be less representative ofrocks in a similar mineral deposit environment at a different metamorphic grade.Obtaining prior information about an exploration site, conducting reconnaissancein the area of interest, and collecting representative rock samples would enable ageologist to determine if metamorphic grade, and any overprinting hydrothermalalteration might affect typical physical property ranges characteristic of the Archeanorogenic gold deposit environment. Some knowledge of physical properties, be itexpected values based on known rock types, or mineral assemblages, will vastly improvethe interpretation of physical property models reulting from geophysical inversion.76REFERENCESBerger, B.R., 1999, Geological investigations along Highway 101, Hislop Township:Ontario Geological Survey, Summary of Field Work and Other Activities 1999, OpenFile 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, 124p.Bohlke, J.K., 1989, Comparison of metasomatic reactions between a common CO2-richvein fluid and diverse wall rocks: intensive variables, mass transfers, and Aumineralization at Alleghany, California: Economic Geology, v.84,p.29 1-327.Bourne, B.T., Dentith, M.C., Trench, A., and Ridley, J., 1993, Physical propertyvariations within Archean granite-greenstone terrane of the Yilgarn craton, WesternAustralia: the influence of metamorphic grade: Exploration Geophysics, v. 24,p.367-374.Boyd, D., 1984, Aeromagnetic surveys and gold, in Doyle, H.A., ed., GeophysicalExploration for Precambrian Gold Deposits, Geology Department and UniversityExtension, 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, 431p.Card, K.D., 1990, A review of the Superior Province of the Canadian Shield, a product ofArchean accretion: Precambrian Research, v. 48,p.99-156.Card, K. D., and Ciesielski, A., 1986, DNAG Subdivisions of the Superior Province ofthe Canadian Shield: Geoscience Canada, v. 13,p.5-13.77Cas, R.A.F., and Wright, J.V., 1987, Volcanic successions, modem and ancient: ageological approach to processes, products, and successions: Allen & Unwin, London,United Kingdom, 528p.Clark, D.A., 1997, Magnetic petrophysics and magnetic petrology: aids to geologicalinterpretation 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 geologicalinterpretation 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, inDoyle, H.A., ed., Geophysical Exploration for Precambrian Gold Deposits, GeologyDepartment 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 MiscellaneousPaper 139, 136p.Cooper, M.A., 1948, Report on Kelore Mines Ltd., Ramore, Ontario, unpublished, 22p.Dockery, B.A., 1984, Geophysical case history of the Lady Susan gold prospect, inDoyle, H.A., ed., Geophysical Exploration for Precambrian Gold Deposits, GeologyDepartment 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.134-146.78Ellis, R.G., and Oldenburg, D.W., 1994, Applied geophysical inversion: GeophysicalJournal 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 fieldand lithogeochemical characteristics of carbonate alteration zones, in Hodder, R.W., andPetruk, W., eds, Geology of Canadian Gold Deposits, Canadian Institute of Mining andMetalurgy, 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, inHedenquist, J.W., Thompson, J.F.H., Goldfarb, R.J., and Richards, J.P., eds.,100thAnniversary Volume, Economic Geology, v. 100,p.407-450.Grant, F.S., 1985, Aeromagnetics, geology, and ore environments, II. Magnetite and oreenvironments: Geoexploration, v. 23,p.335-362.Groves, D.I., 1993, The crustal continuum model for late-Archaean lode-gold deposits ofthe 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 inWestern Australia with particular emphasis on parameters relevant to geophysicalexploration, in Doyle, H.A., ed., Geophysical Exploration for Precambrian GoldDeposits, Geology Department and University Extension, University of WesternAustralia, Publication 10,p.1-63.79Groves, 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 crustaldistribution 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 GeophysicsOrganisation, v.17,p.145-158.Hagemann, S.G., & Cassidy, K.F., 2000, Archaean orogenic lode gold deposits, inHagemann, 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 polarizationmethod, in Van Blaricom, R., compiler, Practical Geophysics II for the ExplorationGeologist, Northwest Mining Association, USA,p.39-138.Hallof, P.G., and Yamashita, M., 1990, The use of the IP method to locate gold-bearingsulfide 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 ofExploration Geophysicists, Tulsa, Ok.,p.227-279.Harron, G.A., Middleton, R.S., Durham, R.B., and Philipp, A., 1987, Geochemical andgeophysical gold exploration in the Timmins area, Ontario: a case history: CanadianMining 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 theAbitibi subprovince, in S.E. Ho, F. Robert, and D.I. Groves, compilers, Gold and BaseMetal Mineralization in the Abitibi Subprovince, Canada, with Emphasis on the Quebec80Segment, Short Course Notes, Geology Department (key center) and UniversityExtension, 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 ofCanada, Special Paper 40,p.635-678.Hodgson, C.J., and Hamilton, J.V., 1990, Gold mineralization in the Abitibi greenstonebelt: 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, EconomicGeology Monograph 6,p.86-100.Hodgson, C.J., and Troop, D.G., 1988, A new computer-aided methodology for areaselection in gold exploration: a case study from the Abitibi greenstone belt: EconomicGeology, v. 83,p.952-977.Hood, P., Irvine, J., and Hansen, J., 1982, The application of the aeromagneticgradiometer 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 ofAmerica Bulletin, v. 72,p.369-3 85.Isles, D.J., Harman, P.G., and Cunneen, J.P., 1989, The contribution of high resolutionaeromagnetics 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.405-482.81Johnson, I., Webster, B., Matthews, R., and McMullen, 5., 1989, Time-domain spectralIP 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 hostedmesothermal 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 goldconcentration 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 magneticsusceptibility: 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 lodegold 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 Australiangreenstone-hosted gold deposits according to wall-rock alteration mineral assemblages:Ore Geology Reviews, v. 6,p.291-33 1.82Moore, E.S., 1936, Geology and ore deposits of the Ramore area: Ontario Department ofMines, Annual Report, 1936, v. 45, Pt. 6, 37P.Ontario Geological Survey, 2001, Physical rock property data from the physical rockproperty 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, Mineraldiscrimination 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 BritishColumbia, 237p.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 GoldCamp: Geoinformatics Exploration Inc., Unpublished report for St Andrew GoldfieldsLtd.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 environmentalmineralogy, in Jambor, J.L., Blowes, D.W., and Ritchie, A.I.M., eds., MineralogicalAssociation of Canada, Short Course Series, Volume 31, Chapter 8,p.165-180.Ridley, J.R., and Diamond, L.W., Fluid chemistry of orogenic lode gold deposits andimplications 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.141-162.83Robert, F., 1990, Structural setting and control of gold-quartz veins of the Val D’Orarea,southeastern Abitibi subprovince, in S.E. Ho, F. Robert, D.L and Groves, compilers,Goldand Base Metal Mineralization in the Abitibi Subprovince, Canada, with Emphasison theQuebec Segment, Short Course Notes, Geology Department (key center) and UniversityExtension, 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 metallogenyofthe Superior and Yilgarn Cratons, in Hedenquist, J.W., Thompson, J.F.H.,Goldfarb, R.J.,and Richards, J.P., eds.,100thAnniversary 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., andAmelin, Y., 2002, Economic Geology, v. 97,p.29 1-309.Roscoe and Postle, 1998, Hislop Mine Property, Roscoe and Postle AssociatesInc., St.Andrew Goldfields Ltd. internal report, unpublished,p.66-89.Seigel, H.O., Johnson, I., and Hennessey, J., 1984, Geophysics the leadingedge:Geophysics: the Leading Edge of Exploration, v. 3,p.32-35.Telford, W.M., Geldart, L.P., and Sheriff, R.E., 1990, AppliedGeophysics, SecondEdition, Cambridge University Press, 770p.Troop, D.G., 1986, Multiple orebody types and vein morphologies, RossMine, District ofCochrane: Ontario Geological Survey, Summary of Field Work and OtherActivities, v.132,p.413-420.Williams, H., Stott, G.M., Thurston, P.C., Sutcliffe, R. H., Bennett,G., Easton, R.M., andArmstrong, D.K., 1991, Tectonic evolution of Ontario: summary and synthesis,in P.C.84Thurston, 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 orbrownfields exploration: Australian Earth Sciences Convention, 2006, Melbourne,Australia, 10p.Williams, N. and Dipple, G., 2005, Identifying sulfide mineralization from physicalproperty measurements and its application to mineral exploration inversions, GeologicalSociety 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 goldmineralization 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 andGeophysics (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.85Chapter 3: Detecting gold-related geology in Archean orogenicgold environments using geophysical inversion: a syntheticmodeling study based on the Hislop gold deposit, Ontario23.1. INTRODUCTION3.1.1. RationaleThree dimensional geophysical inversion modeling, involving the estimation ofphysical property distributions within the earth’s subsurface from observed geophysicaldata, is used widely as a tool to explore for a range of mineral deposit types. Yet, asdifferent deposit types are characterized by unique combinations of rock types,mineralogy, structure, and morphology, each deposit type may not be equally wellimaged by inversion. Expectations regarding the detectability and delineation oforebodies and related rocks in a given mineral deposit setting can be generated throughsynthetic forward and inverse modeling prior to actual geophysical inversion work.This study employs synthetic modeling to test the capabilities of geophysicalinversion as an exploration tool in the Archean orogenic gold environment. In contrast toits more extensive use in imaging higher tonnage and higher grade deposits likevolcanogenic massive sulfide, magmatic sulfide, and porphyry deposits (Oldenburg et al.,1997; Phillips, 2002; Farquharson et al., 2008), geophysical inversion is not as commonlyused to explore for, or map, Archean orogenic gold deposits. As a result, there are fewercase histories successfully demonstrating its application, and thus there is less familiaritywith the range of outcomes that can accompany inversion of different geophysicaldatasets over these deposits.Prior to any geophysical work, it is important to identify the geological andphysical property characteristics of typical exploration targets. Gold mineralized rocks2A 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: asynthetic modeling study based on the Hislop gold deposit, Ontario86are 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 golddeposits, geological features spatially related to gold include faults, felsic intrusive rocks,and hydrothermal alteration zones. These are commonly narrow, near vertical featuresthat extend to depth, and are hosted within deformed and steeply-dipping stratigraphy.Physical property studies completed on rocks associated with the Hislop golddeposit, an orogenic gold deposit located east of the world renowned Timmins-Porcupinegold camp in Ontario, indicate that some known gold-related features have distinctphysical property ranges that may allow them to be distinguished from likely barren hostrocks (see Chapter 2). Synthetic forward and inverse modeling completed on simple 3Dmodels based on the Hislop deposit tests whether these petrophysically distinct gold-related rocks can be detected using inversion methods. Synthetic modeling investigateswhether the physical property contrasts are sufficiently strong, and if sizes, shapes, andlocations, of gold-related geological features are such that they can be detected within adiscretized earth model at a 1 km scale of investigation. Tests are devised to explore theeffects of the addition of geological and physical property constraints. Results of themodeling reveal whether realistic physical property values can be recovered, thus lendingconfidence to the interpretation and querying of the recovered physical property model.Results furthermore highlight possible limitations of inversion at this scale. It indicatesmaximum depths of investigation, and can identify features not caused by a knownsource, but rather are artifacts or byproducts of the inversion algorithm.3.1.2. ObjectivesSynthetic modeling work aims to answer a series of questions related to how wellinversion is able to image prospective geologic features expected in the Archean orogenicgold setting. Specific questions include:871. Can a feature of interest be imaged using unconstrained inversion at a <1 km scaleof investigation? What range of geometry, and physical property contrasts, can weexpect to image within this mineral deposit setting.2. How well does the inversion reproduce the true model? What are the significantdifferences between the recovered and true models? What are the causes ofdiscrepancies?3. Can the model result be improved with addition of basic prior geologicalknowledge, and what types of constraining information are most effective inimproving the model? What differences between the true and recovered modelspersist?4. Which geophysical datasets are most beneficial to invert for orogenic goldexploration? What information can each data type provide to help betterunderstand the geology of the subsurface?3.2. BACKGROUND3.2.1. Geology of the Hislop gold deposit and relationship to other Archean orogenicgold depositsThe synthetic models presented herein are based on a simplified version of thegeology of the Hislop gold deposit, and on average physical property-values determinedfor 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 ultramafic88volcanic 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 30m 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 ofgold occurs with disseminated pyrite within a strongly Fe-carbonate-altered, brecciatedequivalent 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 quartzveinlets, stockworks and fractures in mafic volcanic flows north of the syenite, as well asin association with nearby porphyritic rhyolite dikes striking parallel to stratigraphy.Figure 3.1. Cross-section looking northwest through the Hislop deposit, showing areas ofcarbonate-dominated alteration and gold mineralization. Cross-section interpreted fromdrill core logged from the Hislop property.DDH H9601 DDH Ext 280, GK 280, and H9605Multi-lithic Volcanic BrecciaLamprophyric DikeIntermediate DikePorphyritic Rhyolite DikeSyenite Intrusive[]Mafic Volcanic RockUltramafic Volcanic RockFault—.‘ Drill trace89Geology and alteration mineral assemblages at Hislop are common to manygreenschist facies-hosted Archean orogenic gold deposits globally. Orogenic golddeposits 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 mineralassemblages, which usually extend only short distances (centimeter to meter scale)orthogonal to mineralized veins and structures. Gold occurs predominantly adjacent to, orwithin quartz-carbonate veins, or directly within host rocks associated with disseminatedsulfides. A summary of characteristics defining Archean orogenic gold deposits is givenin Table 3.1. Because of the shared characteristics between the Hislop deposit and otherorogenic 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 Abitibigreenstone belt, and globally.3.2.2. Physical Properties of rock types and alteration zones at HislopPetrophysical contrasts between likely mineralized, and unmineralized rocks arenecessary to yield a geophysical target, and as such they must be identified andunderstood. One difficulty in targeting Archean orogenic gold deposits using geophysicsis that, although gold itself is a conductive and dense mineral, it is usually low grade andthus does not contrast significantly enough from the host rocks to be directly detected bygeophysical methods (Doyle, 1990). This means that other petrophysically distinctvectors to gold are required. At Hislop, petrophysically distinct target rocks includesyenite and rhyolite dikes, carbonate-altered mafic and ultramafic volcanic rocks, andsulfide-rich zones.Results from a physical property study on the Hislop deposit (see Chapter 2) showthat gold-related syenites and porphyritic rhyolite dikes in the Hislop area have lowsusceptibility and density ranges distinguishing them from higher susceptibility mafic andultramafic volcanic rocks (Fig. 3.2). Magnetic susceptibility further separates most low90Table 3.1. Characteristics of Archean orogenic golddeposits.Age Tectonic setting Structural Host rocksllithologicalHydrothermal alterationI Mineralizationassociationlcontrols on associations geochemical signaturemineralizationExamples: Form in extensional, Spatially associatedwith Can form in any rock Carbonate alteration Usually hostedin2710-2670 Macompressional, and large scale crustal type, however, Fe-richmuscovite/sericite alteration, throughgoing, quartz(Abitibi);transtensional structures; mainly mafic and ultramaficsilicification, and albitization; carbonate veins, less2630 Maenvironments during controlled by second and volcanic supracrustaladdition of CaO, CC2,Fe203 commonly as(Yilgarndeformational third order faults that rocks, sedimentary rocks(carbonate alteration), Si02, disseminated replacementCraton);processes at occur as splays off of the (fluviatile sequences), and1<20, Ba, and Na20zones, or as stockwork2670 Maconvergent plate main fault zone; steeply felsic intrusives, are(muscovite alteration,veins.(Midlandsmargns. reverse to oblique, brittle common hosts in thesilicification, albitization).greenstoneto ductile shears zones Abitibi; gold-related faultsbelt,commonly occur atZimbabwecontacts between contrastCraton)Darbyshire et al., 1997; Fyon and Crockett, 1983; Groves etal., 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 andCassidy, 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, leastaltered precursors (Fig. 3.3).3.20 ————- —3.103.00 -2.90________2.802.702.602.502.400.01Electric properties, resistivity and chargeability, do not uniquely distinguishprospective rocks at Hislop (Figs. 3.4 and 3.5). There is a large overlap in resistivityvalues for the rock types studied at Hislop, with the only distinct resistivity range relatedto sheared talc-chlorite rich ultramafic rocks. These rocks, although not consideredprospective, exhibit a fabric which results in lower resistivities (or higher conductivities)than other rocks in the area. Although not explicitly documented in the Hislop physicalproperty 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 othersources documenting electric properties of rocks (Telford et al., 1990; Connell et al.,2000). As with resistivity, specific chargeability ranges do not characterize the individualRock TypesUltramafic rocksMaflc rocksX Intermediate dikesA Syenite intrusives0 Rhyolite porphyriesEU)0>b xe0.1 I 10 100Magnetic Susceptibility (x103 SI Units)1000Figure 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 magneticsusceptibility ranges, allowing them to be distinguished from intermediate, mafic, andultramafic rocks at Hislop.923.20E0,0I3.103.002.902.802.702.602.502.400.13.203.103.002.902.802.702.602.502.401 10 100Magnetic Susceptibility (x103SI Units)b.1000EJa;*Ip,”Unaltered to AlteredUltramafic RocksQLst. altd. (Dol-chi)•Tlc+chi ultramaficDFeCb+ms altd. utlramaficflFe/MgCb+fu altd. ultramafic0.1 1 10 100 1000Magnetic Susceptibility (x103 SI Units)Figure 3.3. Plot of magnetic susceptibility versus density for variably altered maficvolcanic rocks, and variably altered ultramafic volcanic rocks from Hislop. Carbonate-rich alteration (pink and yellow diamonds, and pale green and yellow squares) destroysmagnetite in mafic and ultramafic volcanic rocks, causing magnetic susceptibility todrop. Abbreviations in legends: Lst. altd. = least altered assemblage; Chl+ab =chlorite+albite assemblage; FeCb+ms = Fe-carbonate+muscovite; FeCb+ab = Fecarbonate+albite; Dol+chl = dolomite+chlorite assemblage; Tlc+chl = talc+chlorite;Fe/MgCb+fu = Fe/Mg-carbonate+fuchsite (chrome-muscovite).936IJitramafic volcanic rocksMean 28164 -2_Mafic volcanic rocksMean 2843214 -2 --Intermediate dikes rMean 9759ri-1-6428642teintrusiveean6760iPorphyritic rhyolite dikes11534E10 1001000 10000 100000 1000000Resistivity (Ohm-rn)Figure 3.4. Resistivity histograms for Hislop deposit rocks. Data indicates lower overallresistivities for ultramafic volcanic rocks from Hislop.10- Ultramaflc volcanic rocks -8— —6— —4—_______—2 - . Mean6.91L6- Mafic volcanic rocks -2— lMean31.32rIntermediate dikes6Syenite intrusives -4— -2- Mean 1416Porphyritic rhyolite dikes_____I110 100 1000Chargeability (milliseconds)Figure 3.5. Chargeability histograms for Hislop deposit rocks. Chargeability ranges forthe individual rock types overlap and are not unique.94prospective rock types at Hislop, and in general, all rocks have low background values ofchargeability consistent with published values (Chapter 2). However, values are expectedto 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 highestchargeabilities found at Hislop, and values published in Telford et al. (1990).Chargeability values referenced from the Hislop physical property study were divided by1000 to yield values that correspond to the unitless 0-1 chargeability values that areoutput from induced polarization inversions.Physical property values used in the synthetic models are given in Table 3.2. It isimportant to note that gravity inversions produce density contrast models, thus, to buildthe starting models, density contrasts were used. For this study, the density contrast ofeach rock type is the difference between the rock’s density, and the average density valuefor all the major rock types (2.81 g/cm3). Density values are presented as both densitiesand density contrasts in Table 3.2. Note also that conductivities are used in the startingmodels for DC resistivity work, and that conductivity models are the product of DCresistivity inversions. Conductivity can be converted to resistivity by taking thereciprocal.3.2.3. General forward modeling and inversion backgroundThis research employs forward modeling and inversion codes from the Universityof British Columbia Geophysical Inversion Facility (UBC-GIF). This section provides abrief 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 physicalproperty 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 tothe location of the source in the subsurface, its physical properties, and the strength of the95inducing field in the cases of magnetic, electromagnetic, DC resistivity,and inducedpolarization methods.Table 3.2. Physical property values used in syntheticmodeling.Density (glcrn3)ISusceptibility ConductivityChargeabilityDensity contrastRock Type(S/rn) (ms)(SI Units)(glcm3)Syenite/rhyolite dike 0.00025 2.71-0.11 1.80E-040.016Maficvolcanicrock 0.032 2.88/+0.07 1.50E-040.016Ultramafic volcanic0.0096 2.85/+0.04 2.27E-03 0.016rockCarbonate alteredultramafic/maficvolcanic rock (forcorn parison)Moderately sulfide-richI .40E-02 0.16rock3.OOE-02 0.3Sulfide-rich rockChapter 1 -Chapter 1; Sulfide AnomalousFrom Chapter 1- rich rock values chargeabilities(Hislop depositFrom Chapter 1 from Telford et al., from highestSourcephysical property1990, and Connell chargeabilitystudy)et al., 2000 samples fromHislop studyAnomalous low and high values are highlighted by blue and redborders, respectively2.82/+0.01 3.20E-04 0.016Geophysical inversion involves calculationof the subsurface physical propertydistribution from collected geophysical data. Subsurfacephysical properties arecalculated based on the known physical relationships between sourcesand measurementlocations 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 arean infinite number ofsolutions. To reduce the number of possible solutions,a model objective function isdefined. For default inversions the model objectivefunction specifies that the model isrequired to be close to a background value, referred to as the referencemodel, and has tobe smoothly varying in all directions. With increased knowledgeof geology or physical96properties, the degree of closeness to the reference model can be manipulated, and thesmoothing in the x, y, and z directions can be increased or reduced, by adjustingweightings within the model objective function. In addition to the model objectivefunction, a data misfit is defined. The data calculated by forward modeling the inversionresult must be sufficiently close to the observed data. The misfit and model objectivefunction, respectively, are written:m=a$(m_mo)2dx+ax(m_mo)dx(d+aI—(m—m0)Idydy)N2+a..I—(m—m0)Idzdz.1)where N is the number of geophysical data,d0is the observed data at location i, df’’ isthe predicted data at location i, and e, is the standard deviation. ci is the alpha weightingdetermining the degree of closeness to the reference model, a, and determinessmoothing in the x, y, and z directions, respectively, m is the model, and m0 is thereference model.3.3. METHODSThe 3D ‘Hislop-like’ geologic model shown in Figure 3.6a was converted to thefour 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 Columbia8i97Geophysical Inversion Facility’s (UBC-GIF) Meshtools3D program. Relative dimensionsand 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 modelingmagnetic susceptibility, density, conductivity, and chargeability models, respectively.Table 3.3 summarizes synthetic survey parameters used, and Table 3.4 summarizesinversion parameters. Observed data ‘collected’ over the starting physical propertymodel, and predicted data generated from forward modeling the inversion result, arecompared after each inversion to detennine if results are acceptable. Observed andpredicted data, and achieved misfit values are found in Appendix 3A.Magnetic and gravity inversions are investigated first. Based on physical propertywork, this data is expected to be useful in distinguishing prospective low susceptibilityand low density syenite dikes. The synthetic susceptibility model (Fig. 3 .6b) depicts anarrow vertical low susceptibility dike located between higher susceptibility mafic andultramafic rocks. The density model (Fig. 3.6c) consists of a low density dike withinhigher density mafic and ultramafic rocks. For susceptibility analysis, the dike might alsoact to represent a low-susceptibility, strongly carbonate altered zone, along a faultbetween two higher susceptibility units.DC resistivity and IP methods are investigated for their ability to locateconductive and chargeable sulfides in the subsurface. These synthetic starting modelshave six sulfide-rich zones extending vertically to depth at the ultramafic rock-syenitedike contact (Figs. 3.6d and 3.6e). Additionally the conductivity models contain atalc+chlorite-rich ultramafic schist, incorporated to determine the effect of its uniquerange of conductivity values (Fig. 3 .6d). DC resistivity and IP data were collected using aRealsection electrode array, the configuration used in the collection of actual DCresistivity and IP data over the Hislop deposit in 1996 for exploration purposes. This typeof electrode array employs widely spaced transmitter electrodes placed at a distanceoutboard of closely spaced receiver electrodes to collect data easily and quickly over98a. 3D geologic model.... Gold-related disseminated/i?rsulfldes/ Carbonate+muscovite_j alteration of ultramafic rocksSyenite IntrusiveflMafic Volcanic RockUltramafic Volcanic RockFigure 3.6. a) 3D geological model based on the geologic setting of the Hislop golddeposit. b-e) North-facing cross-sections through 3D physical property models generatedfrom the geologic model: b) magnetic susceptibility model, c) density model, d)conductivity model, e) chargeability model. Susceptibility and density modeling testsdetectability of the syenite dike (the alteration zone is not considered here - syenitedetection is focused on). Resistivity and chargeability modeling tests detectability ofsulfide-rich zones, and low resistivity talc-chlorite dominated ultramafic rocks (carbonatealteration zone is included here).300b) Magnetic susceptibilitySyenite dikec) Density3002001000-100Ultramaficvolcanic rockV2Dmramaficcanic rockMaficvolcanicrockSyenite C -•1Maficvolcarrockb.553800 554100 554400553800o os- 100417200502030.0251000,01070-100300I00002000.002510002020001501000.707 000000540Se-005-100)Charceabilild) Conductivity300200-100-0--100.553800Ultramficsulfide zonevolcan1orockmMaficvolcanicCarhunate- rockaltered ultramafic______d.P0.20.1580,1170.0700.0 000-0.0 07 3 0-0.05I0160.13601120.08820.06430040400165554100 55440099Table 3.3. Synthetic survey parameters.Model Data area (UTM) Lines Line Station Height # Data Data Other informationspacing Spacing errorsMagnetics x: 553800- 554400 E-W 50 m 10 m 320 m 793 5%; Inclination:75°;y: 5373000 - 5373600 5% DecIination:-12;floor Strength: 57478 nTGravity x: 553800- 554400 E-W 50 m 10 m 320 m 793 0.01y: 5373000 - 5373600 mGalfloorDC x: 552800- 555400 E-W 100 m 20 m 300 m 2485 5% Realsection survey -Resistivity y: 5373000- 5373600 (ground) 5 Tx spacings: 1000m, 1500 m, 2000 m,2400 m, 3200 mIP x: 552800- 555400 E-W 100 m 20 m 300 m 2485 15% Realsection survey -y: 5373000 - 5373600 (ground) 5 Tx spacings: 1000m, 1500 m, 2000 m,2400 m, 3200 mTable 3.4. Synthetic inversion parameters.Inversion # Data Inversion core extents # Core Core cell # Padding Other(UTM) cells size cellsMagnetic 793 x:553800-554400 14400010m38000y: 5373000 - 5373600z: 300- (-)100Density 793 x:553800-554400 14400010m38000y: 5373000 - 5373600z: 300- (-)100DC 2485 x: 553800 - 554400 1800020 m330976 Near-surface cellResistivity y: 5373000 - 5373600 weightings applied toz: 300- (-)100 reduce electrode noiselP 2485 x: 553800 - 554400 1800020 m330976 Near-surface celly: 5373000 - 5373600 weightings applied toz: 300 - (-)100 reduce electrode noise100large areas. Surface weighting files were used in the DC resistivity and IP inversioncalculations in an attempt to subdue the tendency for high conductivity and chargeabilityvalues 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 ofprospective features, and physical property contrasts between target and host rocks, aremanipulated to explore the range of results. Inconsistencies between all true models andrecovered models are identified and the cause of these discrepancies is assessed. Forselect cases, attempts are made to further minimize differences between true andrecovered models by applying basic constraints based on prior physical propertyknowledge. These ‘non-located’ constraints (Phillips et al., 2007) are globally applied byadjusting the model objective function, the inversion function defining the type of modeldesired (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 tocompare the closeness of the recovered model to the true model, and to determine if aresulting model has improved with constraints applied. The sum of the differences inphysical property values between each pair of equivalent cells from two identically sizedmodels is calculated:Model difference=Im — mwhere N = the number of data, m = the physical property value of thethcell in the truemodel, and m’ = the physical property value of the equivalent cell in the recoveredinversion model. This value gives a global relative measure of difference between themodels. Since physical properties related to the different geophysical methods havedifferent characteristic numerical ranges, model difference values might be much smallerfor the results of one method versus another. Thus, calculated values can only be101compared between models generated by the same geophysical method. Model differencesfor all results are given in Table 3.5.Table 3.5. Model differences calculated between recovered and true models (the lowestmodel differences for each geophysical method are highlighted with bold text).Model Model DifferenceMagnetic susceptibility20 m syenite between mafic and ultramafic rocks 1781.760 m syenite between mafic and ultramafic rocks 1656.960 rn syenite, buried 1818.820 rn syenite in mafic volcanic rocks 2396.420 m syenite in ultramafic volcanic rocks 721.5constrained, reference model 0.03 SI Units 1232.4constrained, upper bounds 0.035 SI Units 906.0constrained, alpha y and z increased (100) 1644.3constrained, alpha y and z increased, bounds 0.035 904.5depth weighting decreased(13and z0 decreased by 1/4) 705.4depth weighting decreased, and upper bounds set at 0.035 701.8Density20 m syenite between mafic and ultrarnafic rocks 4561.760 rn syenite between mafic and ultrarnafic rocks 4702.060 rn syenite, buried 5355.320 rn syenite in mafic volcanic rocks 5613.320 m syenite in ultramafic volcanic rocks 3441.0Conductivity40 m sulfide-rich zones near ultramafic - syenite contact 158.440 m sulfide-rich zones - higher conductivity (0.03 S/m) 161.040 m sulfide-rich zones - laterally extensive zone 312.140 m sulfide-rich zones - one anomalous zone, no sheared ultrarnafic 160.840 m sulfide-rich zones - constrained, reference model 0.001 S/rn 173.340 m sulfide-rich zones - constrained, alpha y and z increased relative 154.4to x40 m sulfide-rich zones - constrained, reference model 0.001, alpha y 132.6and z increased relative to xDipole-Dipole survey 198.8Chargeabi I ity40 m sulfide-rich zones near ultramafic - syenite contact 1632.840 m sulfide-rich zones - higher chargeability (0.3 rns) 1742.440 rn sulfide-rich zones - laterally extensive zone 3362.91023.4. SYNTHETIC MODELING RESULTS3.4.1. Potential fields modelingMagnetic susceptibility modelsHislop-like model: 20 m syenite dike hosted between mafic and ultramafic volcanic unitsThe contact between the mafic volcanic unit in the east and the central syenitedike is well-resolved to depth by magnetic inversion (Fig. 3.7). In contrast, the contactbetween 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 valuesgreater than 0.05 SI Units are attained within the area known to be occupied by the highsusceptibility mafic volcanic unit. These susceptibilities are over-estimated compared tothe known susceptibility of 0.032 SI Units for these rocks. Susceptibilities areunderestimated where the ultramafic volcanic unit is present, assuming values close to 0SI Units compared to true susceptibilities of 0.0096 SI Units. Near the surface, thereappears to a low susceptibility ‘overburden’, where surface cell susceptibility values dropto 0 SI Units.Varying geometryTwo geometrical variations on the previous model were tested. These new modelsencompass a syenite dike of greater width, and a buried syenite dike. A 60 m widesyenite 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 volcanicunits are detected. With depth however, the geologic contacts are no longer wellconstrained. The buried syenite dike inversion result was comparable to the result for the20 m dike model with the mafic volcanic unit-syenite dike contact being well-imaged todepth (Fig. 3.8b). Again, the ultramafic rock-syenite dike contact is poorly detected and103the syenite is not fully resolved. For the geometrically varied models, problems withover- and underestimation of susceptibility persist.Figure 3.7. Starting model and unconstrained magnetic inversion result for the ‘Hisloplike’ magnetic susceptibility model. Results are shown at the same susceptibility scale asthe starting model. The contact between the mafic volcanic unit and the syenite isdetected to depth, whereas the contact between the syenite and the ultramafic unit isundetected. Susceptibility values in association with the mafic volcanic rock unit in therecovered susceptibility model are overestimated (>0.05 SI Units, compared to -M.03 SIUnits in true model).Varying physicalproperty contrastsTwo additional models are tested to explore the effect of varying the susceptibilitycontrast between the target rocks — the syenite dike — and the host rocks. The 20 msyenite dike is first modeled within a mafic volcanic host, and then within an ultramaficvolcanic host rock.Geophysical inversion over a syenite dike hosted in high susceptibility maficvolcanic 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 centrallow susceptibility zone smoothes outward with depth in the model. The low susceptibilityvalues within surface cells persist, and susceptibility values assigned to areas300054 - 00417 20000323- 002500 57000 33- 0SI UnitsStarting model Recovered susceptibility modelI0010.04170 .03 2 30.5250.07 670 .0 05 33SI Units100-100•553800 554100 554400 553800554100 554400104corresponding with the location of mafic volcanic rocks are overestimated, especially atdepth.The host rock to the syenite dike is next changed from relatively highsusceptibility mafic volcanic rock, to a relatively moderate susceptibility ultramaficvolcanic rock to investigate resulting inversions. Results indicate the presence of a lowsusceptibility zone down to 250 m (Fig. 3.9b). In general, although there is a highercontrast between mafic volcanic rocks and syenite, the inversion results for the syenitedike hosted within ultramafic rocks has recovered values more consistent with the truemodel susceptibility values (see model difference values in Table 3.5).Figure 3.8. Starting models and magnetic inversion results with changes made togeometry of the target body. a) Result for the 60 m syenite hosted by mafic andultramafic rocks. The location of the syenite is well-imaged near-surface. b) Result forthe buried 60 m syenite. The syenite dike is undetected, and the result similar to the initial‘Hislop-like’ susceptibility model.I°- 0.04170.0033- 0.025I:olSI Units554400I00457003230.0250.01610,30633SI Units554400553800 554100105300Starting model Recovered susceptibility modela.0.05 005200 --I0417 10 04110.033300333100 —0.025 0.0250.0161 001610.00633 0.005330 7SI Units SI Units-100553800 554100 554400 553800 554100 554400553800 554100 554400I0050 04110.0 3 330.0250 01670,0 08 33SI Units553800 554100 554400I1,050. 04110 03330 0260 01670.00833SI UnitsFigure 3.9. Magnetic inversion results for starting models with different physical propertycontrasts between the target and host rocks. a) Result for the 20 m syenite dike hostedwithin 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 msyenite dike hosted in moderately susceptible ultramafic volcanic rocks. Again thesyenite is detected to depth. The susceptibility scale is kept the same for all models forcomparison — however the features in the recovered model in Figure 3.9b would be bettervisualized with the scale set to have a lower maximum value.106Density modelsHislop-like model. 20 m syenite dike hosted between mafic and ulframafic volcanic unitsThe starting Hislop-like density model differs from the susceptibility model inthat the ultramafic and mafic volcanic host rocks have similarly high densities, andcontrast nearly equally with the low density 20 m syenite, compared to the variablesusceptibility contrast between the syenite and rock units on either side in susceptibilitymodels. The gravity inversion result reveals the low density syenite and indicates itsextent down to about 200 m depth (Fig. 3.10). The body terminates beyond this as themodel becomes smooth. The slightly smaller density contrast between the ultramaficvolcanic rock and the syenite, compared to the mafic volcanic rock and the syenite, isapparent in the marginally weaker detection of the western contact of the syenite. Ingeneral density values are well-estimated throughout the central region of the model.However, as with magnetic inversion results, there is some overestimation at depth withestimated density contrasts for mafic volcanic rocks of approximately 0.117 g/cm3versusknown values of 0.07 g/cm3.Varying geometryThe eastern contact between the syenite dike and adjacent mafic volcanic rock isbetter resolved to depth where the syenite width is increased to 60 m, with the central lowdensity 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 isessentially unresolved by gravity inversion. It is apparent that there is some decrease indensity 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 partof the model.107Starting modelFigure 3.10. Starting model and unconstrained gravity inversion result for the ‘Hisloplike’ density contrast model. The contact between the mafic volcanic unit and the syeniteis better located than the contact between the ultramafic unit and the syenite. Densitycontrasts 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 essentiallyundetected, however an overall change in density from east to west is detected by theinversion, with the contact between mafic and ultramafic rocks being detected near-surface.553800 554100 554400I020,1580,1170,0750.0323-000833.0.05gfcm553800 554100 554400I020188011100180.0333-000833-005gicrn3Starting model3002001000-100200300I020.7580,711‘0.3180.8333-0008330-005gicm100553800 554100 554400I0,201580,7 10,01500303‘0 00832-000g/cm-100-553800 554100 554400108Varyingphysicalproperty contrastsAs the two host rocks in the Hislop-like model are characterized by similardensity values, the results of the single-host inversions (Figs. 3.1 2a and 3.1 2b) are notdramatically different from the two-host results. Where either mafic volcanic rocks orultramafic volcanic rocks are the sole host for a syenite dike, the inversion is capable ofimaging the low density syenite to depths ranging from 150-200 m. These depths areslightly 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 resultsMagneticsThe presence of a 20 m syenite dike hosted between an ultramafic, and a maficvolcanic unit is not obvious in synthetic magnetic inversion results. The inversion onlydetects an overall gradient here between the lower and higher susceptibility areas. Thenarrow dike is more successfully imaged between the two different hosts when it has aslightly greater width, or when it is hosted by a single high susceptibility rock type.The most significant differences between the recovered and true magneticsusceptibility models include 1) smoothing across known contacts, especially acrosscontacts where there is a low susceptibility contrast (such as the contact between thesyenite dike and the ultramafic volcanic rock), 2) smoothing with depth, and 3) incorrectestimation of susceptibilities through the model, which generally yields higher maximumsusceptibility values than in the true model. The third item encompasses the issue of lowsusceptibility values being incorrectly assumed near surface.109I02o so0117O 075o oooo.0 008)3-005gIcmFigure 3.12. Inversion results with different physical property contrasts between thetarget and host rocks. a) Result for the 20 m syenite dike hosted within higher densitymafic 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 amarginally deeper detection of the syenite.300200100-03002001000-553800 554100 554400I°0158all,007600333.0.005gfcm-100.5538003002001000-554100 554400200ISO0.158011700’s0.0 33 3-o c-ooo.005glcm14(50.6100300I020 500117007500333.0008:.005g/cm0553800 554100 554400 553800 554100 554400Most of the smoothing within the model is a byproduct of the inversion algorithmand choice of model norm. The model objective function for linear potential fieldsinversions is written such that a smooth, simple model result is calculated. This isfacilitated using a L2 norm calculation which minimizes structure over the volume (Liand Oldenburg, 1996). A smooth model, in the case where there is little prior informationabout 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 thesubsurface, and the inversion cannot resolve features deeper than the sources that haveinfluenced the magnetic data collected.110Poor susceptibility estimation in near-surface cells may be related to depthweightings. A default depth weighting is written into all potential field inversioncalculations (Li and Oldenburg, 1996 and 1998). As there is no inherent depth resolutionin potential fields data, when an unweighted inversion is carried out, all susceptibilitywill occur at the surface as it is the simplest solution explaining the observed data (Li andOldenburg, 1996 and 1998). Although depth weighting is necessary to offset this, thesurface cells appear to be less sensitive, and tend to assume reference model values,which for default inversions is 0 SI Units. The generally overestimated susceptibilityvalues for the model as a whole may also be explained by the depth weighting. Tocompensate for the lack of susceptibility at the surface, it is necessary for the inversion toplace high susceptibility values at depth and increase their overall magnitude, in order tofit the observed magnetic data.GravityThe essentially equal contrast between higher density mafic volcanic rocks andultramafic volcanic rocks, and the lower density syenite, allows for consistent detectionof the narrow syenite dike, unless it is buried.Gravity inversions follow similar calculations as magnetic inversions, with themain difference being only the forward model solution - the physical relationshipsbetween subsurface sources and data observations. Thus explanations for discrepanciesbetween true and recovered density models are similar to those for magneticsusceptibility models. As with magnetic inversions, the significant discrepancies arerelated to smoothing and depth resolution. Smoothing can be related to the choice ofmodel norm used in the inversion algorithm, and decreasing resolution with depth isexplained by the known hr2 decay of the gravity signal with depth. Contacts are notresolved as deep as they are with magnetic inversions. Relative density contrasts versusmagnetic susceptibility contrasts might cause the depth detection to be inconsistentbetween the magnetic and gravity inversion results (i.e. the contrast between low111susceptibility syenite and high susceptibility mafic volcanic rocks is greater than thedensity contrast between the two rocks). The tendency of the near-surface cells to assumevalues near 0 g/cm3,the reference model value, likely necessitates having overestimateddensities at depth.Combining magnetic and gravity inversion results would better detect the syenitewhere it is not resolved in the two-host model by magnetic inversions alone. The densityresult would better locate the ultramafic-syenite contact, and the magnetic inversionresult would contribute by providing better depth information.3.4.2. DC resistivity and induced polarization modelingSynthetic conductivity and chargeability models are likely less representative oftrue subsurface geology than susceptibility and density models. Rock type and associatedmineralogy strongly influence magnetic susceptibility and density values (Chapter 2) andthus using geology to create starting physical property models is easily justified.Although rock type can play a role in determining conductivity and chargeability valuesand 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 inmind that in nature, more complicated conductivity and chargeability distributions likelyexist than can be represented by the synthetic models.Resistivity ModelsHislop-like model: 40 m wide sulfide rich zones near ultramafic rock-syenite dike contactSix high conductivity zones (only three visible on cross-section) withinsignificantly lower conductivity host rocks, but proximal to a moderately highconductivity, sheared ultramafic volcanic unit were not resolved through DC resistivity112inversion (Fig. 3.13). The moderately high conductivity ultramafic rocks, however, wereimaged, but only to a depth of approximately 200 m. The recovered conductivityanomaly associated with the ultramafic unit extends faintly toward the general location ofthe high conductivity sulfide-rich zones. Overall, recovered conductivity values for lowconductivity areas are close to true values, however conductivities associated with thesheared ultramafic unit, and obviously those associated with the anomalous conductivityzone, are underestimated.Figure 3.13. Starting model and unconstrained DC resistivity inversion result(conductivity model) for the ‘Hislop-like’ conductivity model. The sulfide-rich highconductivity zones are undetected. The moderately conductive sheared talc-chlorite richultramafic rock is detected near-surface, and resolved only to about 200 m depth.Varyingphysicalproperty contrasts and geometryDoubling the conductivity for the discrete sulfide zone model so as to represent amore sulfide-rich rock (Table 3.2), does not improve the recovery of the sulfide richzones, as they remain undetected (Fig. 3.1 4a). Since raising the conductivity values of thesulfide zones was ineffective, the high conductivity zone was modified to be a continuouszone extending from north to south and vertically to depth to test whether a moreextensive feature might lead to a better recovery (Fig 3.14b). Although the zone is nowpersistent and reaches the surface, no longer consisting of discrete small zones at depth, itremains undetected by inversion of DC resistivity data. As with the discrete sulfide zonemodel, a conductivity high is detected in association with the sheared ultramafic unit, and3002000.014 S/rnI00030 002510,00202000153 1001000100000004250.005S/rn300.Startina mode’0003200-0,002510.00 202100-0,001030.001 030- 0.00054250-005S/rn-100-553800 554100 5544000-lOG.553800 554100 55440011330O3000.002100025120010,002020.001520.001030.00054251-005S/rn1001300Starting model0 0032000,00202100-0001520 001020 .0 00 54 250-000S/rn-100553800 554100 554400Starting model300200100-100553800 554100 5544000553800 554100 5544000--i00i553800 5541000554400I0.000100 0001010 5001720.0001530. 0001 540 0001450000135S/rnFigure 3.14. DC resistivity inversion results (conductivity models) with changes made tophysical property contrasts, and to the geometry of the target body. a) Conductivities aredoubled for the sulfide-rich zones, but the targets remain undetected. b) The target is made to becontinuous. There is a weak indication of the conductive zone. c) All host rocks are given thesame low conductivity background value. Comparing the result at the same scale, (i), the targetis essentially undetected. Adjusting the scale (ii) reveals a conductive body, but with highlyunderestimated associated conductivities.I000000 002510002020 00100010330 00054250-005S/rn 200I0.0030.00250 0020.50150.0 010.0 005S/rn200-100-1553800I0.0030,00 20 10.0 02020001020.001 030 0000450-005S/rn554100 554400 553800 554100 5544000.03 S/rnb.C.0.03 S/rn114this zone extends laterally out toward the continuous sulfide zone, marginallymore sothan in the previous results.The moderately conductive ultramafic unit was removed, and all other units, asidefrom the sulfide-rich zone, were assigned a low conductivity value representingmaficvolcanic rocks, to test a single continuous high conductivity zone in a lowconductivitybackground (Fig. 3.l4c). Comparing the results to the true modelat the sameconductivity scales, demonstrates there is essentially a complete lack of resolutionof thetarget feature. When the scale is adjusted to encompass the true recovered maximumandminimum values, the feature is revealed. The recovered body is correctly locatednear thesurface, but extends only about 150 m depth. The values coinciding with the highconductivity zone however, are very low and near background values, and would notbeconsidered of an anomalous nature.Chargeability ModelsHislop-like model: 40 m wide sulfide rich zones near ultramafic-syenite contactSix sulfide-rich zones (only three visible in the cross-section) are assignedanomalous chargeability values. The anomalous values (Table3.2) are chosen based onthe highest chargeabilities measured from the Hislop deposit chargeabilty studies(seeChapter 2), and are divided by 1000 to correspond with IP inversion outputs. Theyaremodeled within a low chargeability background. The zones are detected as a single, smallanomalous area near the surface, which coincides with the top of the uppersulfide-richzone (Fig. 3.15). The chargeability values estimated by the inversion are low comparedtothose of the true model.115300Recovered charreability modelFigure 3.15. Starting model and unconstrained IP inversion result for the ‘Hislop-like’chargeability model. The sulfide-rich high chargeability zones are only detected nearsurface down to about 125 m. The chargeability values estimated by the inversion for thechargeable zones are much lower than known values.Varying geometry andphysicalpropertiesThe chargeability of the sulfide zones is doubled from the initial model to testtheir subsequent resolution (Fig. 3.1 6a). The inversion result is essentially identical to theprevious model result, but with marginally higher chargeability values coinciding withknown sulfides. The upper chargeability zone is detected, but chargeability values areunderestimated.Where the sulfide zone is made a continuous feature with doubled chargeabilityvalues, it is well-located in the subsurface by the inversion, extending down toapproximately 300 m depth (Fig. 3.16b). The representative chargeability values areunderestimated for the sulfide zone. There is some excess structure occurring near thesurface and at depth, which does not occur in the true model.200lOOi0100IIOil01360,112O 0602O 0643O 048400160116300Figure 3.16. IP inversion results with changes made to physical property contrasts andgeometry of the target body. a) Chargeabilities are doubled for the sulfide-rich zones. Theresolved feature remains restricted to the near surface, but has slightly higherchargeabilities than the previous model result. b) The high chargeability target is made tobe continuous. The target is well-located to almost 300 m depth, but has underestimatedchargeabilities. Some additional structure in the model is found in near-surface cells andat depth.Discussion of DC resistivity and IP inversion modeling resultsDC ResistivityThe most significant discrepancy between true and recovered conductivity modelsis the lack of resolution of the central ‘sulfide-rich’, high conductivity zones. Even whenthe starting model contains only one persistent conductive feature, the recovered valuesare too low to be considered anomalous. This suggests that the feature may be toonarrow, or the contrast between host and target rocks too weak. The fact that the western2001001001553800I0.160.1360.11200842006430.040400165554100 554400 553800 554100 554400300Recovered charleabilit model0.16 016jO,13620010.1360112 0.112008621000088200643 0.0643004040004040.01 65 00166-1001553800 554100 554400 553800 554100554400117ultramafic volcanic unit is resolved, a unit characterized by lower conductivities than thesulfide-rich body suggests that a lack of contrast can not be the explanation. The width ofthe feature is more likely to hinder its detection. In 3D forward modeling processes, tocalculate voltages at a point, conductivities of cells within a 3D mesh must be averagedwith those of neighboring cells over the cell volume (Dey and Morrison, 1979). In thecase of the Hislop model, there is only one cell boundary that divides the anomalouslyconductive sulfide zone (the unit is 40 m thick and the cells 20 m wide), and the highvalue is essentially retained only in the averaging over this one particular boundary. Atthe dike’s contacts, the high conductivities are averaged with the lower conductivities ofthe host rock, bringing the values down quickly to background conductivities. Theaberration 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 beexceedingly large, especially when employing data from Realsection arrays, where alarge mesh is required.The resolved moderately conductive ultramafic unit is only imaged to about 200m 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 electrodearray used. A Realsection array is similar to gradient or Schlumberger arrays in thattransmitter electrodes are spaced at large distances outboard of the receiver electrodes(Telford et al., 1990). The poor resolution associated with wide electrode spacings isdiscussed by Hallof and Yamashita (1990). With increasing distance from the transmitter,the current weakens. Therefore a distant source that is intersected by the current willproduce a weak signal, and will likely only be detected when the receiver electrode issufficiently close, and in this case receiver electrodes are on surface.The final discrepancy between true and resolved models is the underestimation ofconductivity values for known higher conductivity areas. This is interpreted to beprimarily associated with smoothing and the subsequent dispersal of conductivity overlarger volumes within the mesh. Based on the model objective function defined, the118inversion solution favors the assignment of low conductivities over many cells, ratherthan high conductivities within more compact volumes.InducedpolarizationThe chargeable zones are imaged to various degress in each of the IP inversionmodel results, with depth detection improving with increased chargeability contrastbetween the sulfides and the host rocks. However, the recovered chargeability valuescorresponding to the sulfide-rich features in each case are underestimated. As withconductivity model results, the small width of the feature likely limits its detection.However, in contrast to the forward modeling of conductivity models, an averaging ofchargeabilities 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 syntheticmodeling process, which explains the slightly better resolution of the small chargeabilityanomalies. The poor resolution at depth known for Realsection surveys further reducesthe possibility of resolving the sulfide-rich zones.As with recovered conductivity models, smoothing resulting from the model normcontributes to the dispersion of chargeabilities over a larger volume. The smoothing ofthe conductivity anomaly over more cells than what are known to contain anomalousvalues, means that each cell requires less chargeability overall in order to explain theobserved data.The irregularly dispersed high chargeability values at surface thought to be relatedto the increased sensitivities at electrodes may also partly explain why chargeabilityvalues are lower than those from equivalent areas within true models — surface cellsmight already be taking up some of the required chargeabilty needed to explain theobserved data, resulting in lower values elsewhere.1193.4.3. Improving model results with basic constraintsInversion model results can be improved by constraining the inversion withadditional geologic and physical property information (Phillips, 2002; Williams, 2006).This information can be inferred from the exploration deposit model, from publisheddata, or from direct reconnaissance work. Prior information can be incorporated into theinversion calculation through basic manipulation of the model objective function toproduce a model consistent with known geology, physical properties, and geometry (Liand Oldenburg, 1996 and 1998). As with all inversion results, the result must still fit thedata within the specified misfit.In this section, basic constraints are applied to inversions to try and reduce thediscrepancies between recovered and true models evident from unconstrained inversions.The constraints are tested only for the Hislop-like resistivity and magnetic susceptibilitymodels. The unconstrained inversions for these models did not fully delineate the targetrocks, and as such, these two cases constitute good candidates for testing the possibilityof model improvement with constraints. For the globally constrained inversions, physicalproperty values known to be representative of the subsurface geology are used to recovermore accurate physical property values, and knowledge of general structural orientationsis applied to encourage smoothing of features in the desired directions. With significantamounts of prior physical property knowledge in the form of physical propertymeasurements or geological 3D models, thorough and complex constraints can beapplied, 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 asingle physical property value that is considered representative of expected values. Theinversion is required to yield a model close to this reference model, while satisfying theremaining terms of the inversion algorithm. The second constraint tested is settingphysical property bounds on the inversion results. The default setting for UBC-GIFinversions usually allows a large range of values to be assumed by the model cells. Thebounds can be adjusted to yield results within the range of known or expected values. The120third constraint tested is a geometrical constraint and is chosen based on knowngeological directionality, or preferred orientations.The table of model difference values (Table 3.5) can be referred to here toevaluate the quantitative improvements in model estimation in accordance with thevarious constraints applied. A decrease in the calculated model difference reflects smallerdifferences between true and recovered physical properties.Magnetic constraints: 20 m syenite dike between ultramafic and mafic volcanic unitsReference modelA constant reference model value of 0.03 SI Units is used, representing expectedhigh susceptibilities of mafic and ultramafic rocks. As in unconstrained results, only thecontact 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 continuesto be elusive. Near-surface cells show again the tendency to acquire reference modelvalues, in this case values around 0.03 SI Units (previously 0 SI Units reflecting thedefault reference model). The calculated model difference is an improvement in overallrecovery of the true susceptibility values compared to unconstrained results. Thisimproved susceptibility estimate is explained by the new reference model value(>0 SIUnits) being assumed by the surface cells, resulting in the necessary lowering ofsusceptibility at depth, where susceptibilities in relation to the mafic volcanic unit wereinitially highly overestimated to compensate for low surface values. Geological contactsare slightly better located.121300pre!erence model300Constrained using boundsFigure 3.17. Inversion results for the Hislop-like susceptibility model after constraintsapplied, a) Inversion result with reference model set to 0.03 SI Units; b) Result withbounds 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.BoundsBounding the model using a upper susceptibility bound of 0.03 5 SI Units, a valueslightly higher than the known susceptibility of mafic volcanic rocks here, and moresuitable than the default upper bound of 1 SI Unit, yields a more qualitatively andquantitatively accurate model (Fig. 3.17b). The cap on the susceptibility values keepssusceptibility from being significantly overestimated. Low susceptibilities near-surface,thought to be caused in previous models by depth weighting, are minimized withappropriate bounds. An overall lowering of the susceptibilities in the vicinity of the maficvolcanic rocks means the high susceptibility zones are not pushed as deep to effectivelyreproduce the observed data. The mafic volcanic unit/syenite dike contact is well-located,I200I0600617606330026601670,00633SI Units100I-100553800 554100 5544000300Constrained using alpha weightings002560107000033SI Units554400I0.050.04170.033300250. 01 677 00033SI Units553800 554100 554400 553800 554100 554400122extending to a depth of about 400 m, however the presence of the syenite dike, is still notobvious.Alpha weightingsUse of alpha (cL) weightings to achieve smoothing along the z and y axesreflecting known structural orientations, results in sharper contacts within the model, butcauses unnecessary vertical exaggeration (Fig. 3.1 7c). Model difference calculations(Tab. 3.5) indicate that susceptibility value estimation has not improved with respect tothe unconstrained result. Without putting any restrictions on susceptibility values,susceptibilities are still overestimated.Combined bounds and directional weightingBy combining alpha weighting in y and z directions with use of more appropriateupper bounds values, a well-estimated model results, with slightly sharper contacts thanwhen bounds alone are constrained (Fig. 3.17d). Model difference values show this resultis not necessarily an improvement on setting upper bounds exclusively. Although thegold-related feature, the syenite dike, is not better imaged, the physical property modelvalues are better estimated, and thus geological interpretations of the model will improve.Resistivity constraints: 40 m wide sulfide rich zones near ultramafic rock-syenite dikecontact.Reference valuesThe reference value for the conductivity inversion was set to 0.00 1 S/rn, a valuelying approximately between the higher conductivity ultramafic and mafic volcanic rocksin order to improve the overall conductivity estimations within the model. The shearedultrarnafic rocks are resolved to a slightly greater depth than in the unconstrained DCresistivity result (Fig. 3.1 8a). Poorly estimated values in this result now appear to berelated to the less sensitive, deeper model cells’ tendency toward higher reference modelvalues of 0.001, a value higher than those cells at the same depth in the true models. The123effect is shallower cells have more underestimated values than previously, not requiringhigh conductivities since higher conductivities exist in the deeper cells. Changing thereference model is good practice as seeing where reference values take over at depthwithin the model allows for determination of maximum depth of investigation, which canbe estimated as the depth where a range of models consistently revert to reference modelvalues (Oldenburg and Li, 1999).Figure 3.18. Inversion results for the Hislop-like conductivity model after constraintsapplied, a) Inversion result with reference model set to 0.001 S/rn; b) Result with az anday increased relative to ax; c) Result with reference model set to 0.001 S/rn, and alphacLz and ay increased relative to ax.Alpha weightingsIncreasing alpha weightings in the z and y directions relative to the x direction toincrease smoothing parallel to known structures and contacts causes conductivities00030.002010,002020.001530.00103000054250-005S/rnonstrained with alpha weights and reference model0.0030002550002520005530.001 03000054250-005S/rn124related to sheared ultramafic rocks to extend marginally deeper (Fig. 3.1 8b). The greatervolume encompassed by the conductivity anomaly means the values are lower overall, incomparison to initial inversion results, to explain the observed data.Combined reference model and directional weightingCombining alpha weighting with appropriate reference value assignment yields agood, geologically reasonable result with the conductivity anomaly extended to depth asin the true model (Fig. 3.1 8c). Model difference values for this result are lower than forall previous DC resistivity inversion results. The use of the reference model value of0.00 1 S/m keeps conductivity values high, unlike the application of alpha weightingsalone. Again, constraining the model using prior physical property and geologicalknowledge does not improve imaging of the gold target, but yields a more accuratephysical property model, which will in turn, lead to improved geological interpretations.3.4.4. Other solutions for improving model resultsExperimentation with additional modifications to inversion and survey parameterswere attempted to try to improve the model results where they have not been improvedby constraints.Magnetic susceptibility model improvements — adjusting depth weightingsOne of the causes of discrepancy between true and recovered magneticsusceptibility models is the applied depth weightings. Depth weighting written into thepotential fields inversion codes (Li and Oldenburg, 1996 and 1998) are necessary tooffset the natural decay of the magnetic and gravity signal, and for distributing physicalproperties to depth. However depth weighting appears to lead to low sensitivities at thesurface in susceptibility models, and subsequent overestimation of susceptibility at depth.To alleviate the problem of surface cells assuming reference model values, thedefault depth weighting was reduced by decreasing values of f3 and z0 (arbitrarily by a125quarter of their default values which were 3 and 20.92, respectively), the variables withinthe depth weighting function controlling the offset of the natural magnetic decay (Li andOldenburg, 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. Modeldifference values drop with a decrease in the weighting (Fig. 3.19). This changeemphasizes that a significant portion of the disagreement between true and recoveredmodels stems from the poor estimation of susceptibility near surface. It is recommendedthat for magnetic inversions at this scale, the depth weighting be manipulated forcomparison to other unconstrained and constrained model results. For larger scalemodels, there might be problems associated with this manipulation of the default depthweighting, in terms of loss of information at depth.Figure 3.19. Inversion results for the Hislop-like susceptibility model with depthweightings 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 designDC resistivity and IP surveys completed using Realsection or Schlumbergerarrays tend to have better depth detection than arrays with more closely spacedtransmitters, but less spatial resolution overall due to the weakening of the current over300Magnetic result with reduced depth we200’1000100510.04170.03330.0250.0167000833SI Units554400-100W553800 554100126the large distances covered. Hallof and Yamashita (1990) discuss the importance of usingclosely spaced electrodes for detection of small sulfide-rich zones. The use of a differentarray configuration where transmitter electrodes are not distant with respect to receiverelectrodes might enhance spatial resolution of the conductivity anomalies in the Hislopmodel. 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 animprovement on the Realsection inversion result with better estimated conductivityvalues, 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 beyondthe scope of this research, this example shows that the chosen survey design candetermine whether a feature will be detected in the geophysical data and in inversionresults. The possibilities should be well-researched in advance of exploration withconsideration of the types of information required and the characteristic sizes and depthsof targets.Dipole - DipoleC2 Cl P1 P2.E—a.(naSch I urn bergerCl P1 P2 C2•( na )•( a—>.#-——na )•Notes:Clcurrent source &ectrodes (transmitters)P1potential electrodes (receivers)a = &ectrode separation; n = an integerFigure 3.20. Comparison of a dipole-dipole electrode configuration and a Schiumbergerconfiguration which resembles a Realsection array. Dipole-dipole surveys employ closelyspaced current and potential electrodes. For the Schiumberger array, current electrodesare distal to potential electrodes (figure modified from Inversion for Applied Geophysicsresource package, UBC-GIF).127Figure. 3.21. DC resistivity inversion result for resistivity data collected via a dipole-dipole survey. Depth resolution has not increased compared to Realsection results, butthere is better spatial resolution and the sulfide-rich zone is detected in addition to thesheared ultramafic unit.3.5. CONCLUSIONSSynthetic modeling is important to conduct prior to inversion work. It will revealwhether or not a feature of particular shape, size, and of certain contrast with the hostrocks can be resolved using inversion methods. In doing this, limitations of inversion arealso revealed, establishing where caution in interpreting results is necessary (e.g. wherephysical properties recovered may be inaccurate, or if there are artifacts that arebyproducts of the inversion), and determining when confidence can be placed in theinterpretation or querying of recovered models. Synthetic modeling also tests the effectsof constraints, and determines if it is possible to improve the model through theirapplication.Synthetic modeling was completed in this study to determine whether geologicalfeatures, specifically rocks related to gold mineralization and characteristic of Archeanorogenic gold deposits, are detectable in the subsurface. From previous geological andDC Resistivity result from dipole-dipole survey300.IC 0030.002510002020001530.001030 0006425e.005S/rn554400553800 554100128physical property studies it was determined that prospective features typical of thismineral deposit setting, which are also petrophysically distinct from host rocks include,near-vertical, or steeply-dipping faults, felsic dikes, carbonate alteration zones, andsulfide rich zones. These features are modeled within ultramafic and mafic volcanicrocks, common hosts to orogenic gold mineralization. The scale of the study isreminiscent of deposit-scale exploration, and inversions are done on a 1 km by 1 km by600 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 geophysicalinversion, to be useful exploration tools. Magnetic and gravity inversions were successfulin resolving low susceptibility and low density gold-related syenite dikes to depthsaround 200-3 50 m. The narrow, 20 m dike is most poorly imaged by magnetic inversionwhere hosted between a mafic and an ultramafic unit. Where there are two hosts ofdifferent susceptibility, yet both of higher susceptibility than the dike, the signature of thedike is lost in the overall gradient from low to high susceptibility areas. A combination ofmagnetic and gravity inversion results would best detect the syenite dike, with the densityresolving the ultramafic-syenite contact, and magnetics resolving features slightly deeper.The main differences between recovered and true magnetic and gravity models resultfrom smoothing due to the choice of inversion model norm, and due to the natural decayof the geophysical signal with depth. Depth weights also cause discrepancies betweentrue and recovered models, which can bring about high estimates of susceptibility ordensity at depth, leading to poor distributions of physical properties throughout themodel.Resistivity modeling using a Realsection electrode array does not detect narrowanomalous conductive zones related to sulfide-rich rocks, unless the zones are quiteanomalous and laterally continuous, and in this case, their associated conductivity valuesare underestimated. DC Resistivity inversions are, however, effective at modeling largerconductive geological units, but only to shallow depths within the subsurface. Inducedpolarization inversions detect chargeable zones, especially where they are extensive and129continuous. However, as with conductivity models, the result is only reliable nearsurface, and anomalous values are underestimated. The lack of resolution at depthaccompanying DC resistivity and IP survey methods cause the poor depth resolution ininversion results. The underestimation of anomalous physical property values is due todispersal of anomalous values over larger areas as a result of smoothing brought on bythe inversion model objective function.Constraining inversion results acts most importantly in limiting physical propertyranges 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 notpreviously detected, they yield qualitatively and quantitatively better results. Applyingconstraints permits assessment of the range of possible results. Features that persistbetween various model results are required to exist to fit the observed data and satisfy therequirements of the model objective function, and are likely real.For all geophysical methods, different results would be expected at larger scalesof 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 tokeep inversion computation times to a minimum.Synthetic modeling, and subsequent inversion of true geophysical data, requiresknowledge about relationships between physical properties and expected rock types. Thisis best achieved by having a reconnaissance knowledge of the geology beinginvestigated. Using downhole susceptibility information, or surface sample data, it isrecommended that typical physical property ranges be determined for the important rocktypes and altered equivalents. Sourcing published data is an option where data collectionhas not been carried out. Physical property studies in various geological settings arebecoming more commonplace, and physical property data is being compiled currently ona more regular basis. This information is becoming increasingly valuable with the drivein recent years toward using geophysical methods to explore for deposits in the deepersubsurface.130REFERENCESBerger, B.R., 1999, Geological investigations along Highway 101, Hislop Township:Ontario Geological Survey, Summary of Field Work and Other Activities 1999, OpenFile Report, 6000,p.5-1 — 5-8.Connell, S., Scromeda, N., Katsube, T.J., and Mwenifumbo, J., 2000, Electricalresistivity characteristics of mineralization and unmineralized rocks from the Giant andCon mine areas: Geological Survey of Canada, Yellowknife, NWT, Current Research2000-E9.Cooper, M.A., 1948, Report on Kelore Mines Ltd., Ramore, Ontario, unpublished, 22p.Darbyshire, D.P.F., Pitfield, P.E.J., and Campbell, S.D.G., 1996, Late Archean and EarlyProterozoic gold-tungsten mineralization in the Zimbabwe Archean craton: Rb-Sr andSm-Nd isotope constraints: Geology, v. 24, p. 19—22.DCIP3D user manual, Version 2.1, University of British Columbia GeophysicalInversion Facility.Dey, A., and Morrison, H.F., 1979, Resistivity modeling for arbitrarily shaped threedimensional structures: Geophysics, v. 44,p.753-780Doyle, 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 gravityinversion 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 fieldand lithogeochemical characteristics of carbonate alteration zones, in Hodder, R.W., and131Petruk, W., eds, Geology of Canadian Gold Deposits, Canadian Institute of Mining andMetalurgy, 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., andHolyland, P.W., 1995, Lode gold deposits of the Yilgarn block: products of late Archeancrustal-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 crustaldistribution 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, inHagemann, 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-bearingsulfide 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 ofExploration 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, 231-273Hodgson, C.J., 1990, An overview of the geological characteristics of gold deposits in theAbitibi subprovince, in Ho, S.E., Robert, F., and Groves, DI, compilers, Gold and BaseMetal Mineralization in the Abitibi Subprovince, Canada, with Emphasis on the QuebecSegment, Short Course Notes, Geology Department (key center) and UniversityExtension, the University of Western Australia, publication No. 24,p. 63-100.132Hodgson, 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 ofCanada, Special Paper 40,p.63 5-678.Hodgson, C.J., and Hamilton, J.V., 1990, Gold mineralization in the Abitibi greenstonebelt: 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, EconomicGeology Monograph 6,p.86-100.Hodgson, C.J., and MacGeehan, P.J., 1982, Geological characteristics of gold deposits ofthe Superior Province of the Canadian Shield: Canadian Institute of Mining andMetallurgy, Special Volume 24,p.211-228.Hodgson, C.J., and Troop, D.G., 1988, A new computer-aided methodology for areaselection in gold exploration: a case study from the Abitibi greenstone belt: EconomicGeology, v. 83,p.952-977.Kent, A.J.R., Cassidy, K.F., and Fanning, CM., 1996, Archean gold mineralizationsynchronous 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 hostedmesothermal 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 mineralizationand accretion, magmatism, metamorphism, and deformation — Archean to present: areview: Ore Geology Reviews, v. 9,p.263-3 10.Kerrich, R., and Wyman, D., 1990, The geodynamic setting of mesothermal golddeposits: an association with accretionary tectonic regimes: Geology, v. 18,p.882-885.133Kishida, A., and Kerrich, R., 1987, Hydrothermal alteration zoning and goldconcentration 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.109-119.Li, Y., and Oldenburg, D.W., 2000, 3D inversion of induced polarization data:Geophysics, v. 65,p.1931-1945.McCuaig, T.C., Kerrich, R., 1998, P-T-t-deformation-fluid characteristics of lode golddeposits: evidence from alteration systematics: Ore Geology Reviews, v. 12,p.381-453.Meuller, A.G., and Groves, D.I., 1991, The classification of Western Australiangreenstone-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 andIP surveys: Geophysics, v. 64,p.403-4 16.Oldenburg, D.W., Li, Y., and Ellis, R.G., 1997, Inversion of geophysical data over acopper 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 BritishColumbia, 237p.134Phillips, N., Hickey, K., Lelievre, P., Mitchinson, D., Oldenburg, D., Pizarro, N.,Shekhtman, R., Sterritt, V., Tosdal, D., and Williams, N., 2007, Applied strategies for the3D integration of exploration data: KEGS Inversion Symposium, PDAC 2007, extendedabstract, 9p.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., andArchibald, N. J., 2004, Geoinformatics evaluation of the eastward extension of theTimmins Gold Camp: Geoinformatics Exploration Inc., Unpublished report for StAndrew 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 Emphasison the Quebec Segment, Short Course Notes, Geology Department (key center) andUniversity Extension, the University of Western Australia, publication No. 24,p.167-212.Robert, F., 2001, Syenite-associated disseminated gold deposits of the Abitibi greenstonebelt, 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. 66-89.135Sibson, R.H., Robert, F., and Poulsen, K.H., 1988, High-angle reverse faults, fluidpressure 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, SecondEdition, Cambridge University Press, 770p.Weinberg, R.F., Hodkiewicz, P.F., and Groves, D.I., 2004, What controls golddistribution in Archean terranes: Geology, v. 32,p.545-548.Williams, N.C., 2006, Applying UBC-GIF potential fields inversions in greenfields orbrownfields exploration: Australian Earth Sciences Convention, 2006, Melbourne,Australia, 10p.136Chapter 4: 3D inversion of magnetic,gravity, DC resistivity,and induced polarization dataover the Hislop gold deposit,south-central Abitibi greenstonebelt34.1. INTRODUCTION4.1.1. RationaleMagnetic, gravity, DC resistivity, andinduced polarization (IP) data wereinverted to investigate subsurfacegeology within a section of the south-central Abitibigreenstone belt hosting the Hislopgold deposit. A large amount of historic drillinghasbeen completed, but much of it is shallow,and concentrated on specific mineralexploration properties. The irregular drillingcoverage, and an overall lackof outcrop inthe Hislop deposit area suggests that geophysicalinversion, the calculation of subsurfacedistributions of physical properties fromgeophysical data, could be an extremelyusefultool for understanding subsurfacegeology and establishing mineral explorationtargets inthis part of the gold-rich Abitibi greenstonebelt.In contrast to its more extensive applicationfor delineation of massive sulfide-style deposits (Oldenburg et al.,1997; Phillips, 2002; Farquharson etal., 2008),geophysical inversion has not beenas commonly applied for similar purposes intheorogenic gold environment. The reasonfor this is that orogenic gold deposits, liketheHislop gold deposit, are characterizedby small, discontinuous, and low gradeorebodiesthat do not have a strong petrophysical contrastwith typical host rocks. Geological units,hydrothermal alteration zones, and structuresthat are known to be relatedto gold,however, can provide largerscale exploration targets. There are onlya few examples ofcase studies employing inversion methodsto detect gold-related rocks in Archeanorogenic gold settings (UBC-GIF Inversionfor Applied Geophysics CD-ROM, 2000-A version of this chapter will be submittedfor publication. Mitchinson, D., Phillips, N., and Williams,N.,2009, 3D inversion of magnetic, gravity,DC resistivity, and induced polarizationdata over the Hislop golddeposit, south-central Abitibi greenstone belt.1372006; Kowalczyk et al., 2002; Mira Geoscience Ltd.,2005a and 2005b; Mueller et al.,2006). The usefulness of these methodsas an exploration tool in this mineraldepositenvironment may therefore not yet befully appreciated.The large amount of geophysical data available,and a thorough backgroundunderstanding of the relationships betweengeology and physical properties for the Hislopdeposit area (Chapter 2), creates anopportunity to conduct a comprehensivestudy of thetypes of information that might be acquiredby inverting a suite of geophysical dataat arange of scales in this mineral depositsetting.4.1.2. Geological backgroundThe area investigated in this study is locatedin the south-central Abitibigreenstone belt (Fig. 4.1), east of the Timmins-Porcupinegold camp, which is known forit’s world-class Archean orogenicgold deposits (Hollinger-Mclntyre andDomedeposits), and in general for its large number ofgold and base metal deposits andoccurrences. The study area (Fig. 4.2) is underlainby northwest-southeast trendingultramafic to mafic volcanic rocksequences, with lesser felsic volcanicunits (Prest,1957; Berger, 1999; Power etal., 2004; Roscoe and Postle, 1998). Thevolcanicsequences are intruded by variably sized,fine to coarse-grained felsic and intermediateintrusives and dikes. A major crustal-scalefault zone, the Porcupine-DestorDeformationZone (PDDZ), runs northwest-southeastthrough the area, parallel to the generalregionaltrend. Gold deposits in this part of the Abitibigreenstone belt have a closespatialrelationship with the PDDZ (Jackson andFyon, 1991; Berger, 2002). It is interpretedtohave acted as a conduit through whichC02-rich and gold-bearing fluids ascendedupwardthrough the crust to sites of eventual gold deposition(Kerrich, 1989). Sedimentary rocksfrom the Porcupine and Timiskamingassemblages, lie north of the PDDZ, likelyhavingoriginally accumulated in a structurally controlledbasin during fault development lateinthe formation of the greenstone belt (Ayeret al., 2002).138Local and deposit scale geophysical inversions completed for this study focus onthe Hislop gold deposit. The Hislop deposit is a structurally controlled Archean orogenicgold deposit. Gold occurs with disseminated pyrite and is distributed within host rocks inproximity to a fault that occurs between a coarse-grained syenite dike, and ametamorphosed ultramafic volcanic unit (Fig. 4.3). Lesser mineralization occurs withinsmall veins and vein stockworks in magnetite-bearing Fe-rich tholeiitic basalts north ofthe 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 theearly 1900’s onward, and there are many existing driliholes and associated logs providinggeological information for this property. The Hislop deposit was mined during threeseparate efforts between 1990 to 2006, by St. Andrew Goldfields, Ltd., from undergroundworkings (Shaft Area on map in Fig. 4.2) and a small open pit (West Area). Over 400,000..- Major faultsProterozoic rocksArchean sedimentary r:]Archean granitoid rocks•rchean volcanic rocksArchean mafic intrusiverocksFigure 4.1. Geological map of the southwest Abitibi greenstone belt (modified afterPoulsen et al., 2000). The Hislop deposit study area is shown with respect to theTimmins-Porcupine gold camp. White circles represent gold deposits and black circlesrepresent world class gold deposits (>100 t). PDDZ = Porcupine Destor DeformationZone, LLCDZ = Larder Lake—Cadillac Deformation Zone.139tonnes 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 - greywackeS00 — sediment, undividedIFD!IFO — felsic intrusive dyke!felsic intrusive undivided100 — intrusive, undividedFigure 4.2. Geology of the Hislop deposit area as interpreted by Poweret a!. (2004) fromhigh resolution aeromagnetics, and previous mapping in the Abitibi greenstone belt.Locations of two mined areas on the Hislop property (West Area open pit; ShaftAreaunderground) are outlined in red. Also shown are 10 drill holes (one overlapping) loggedfor this study. The cross-section shown in Figure 4.3 is based on core logging of four drillholes that were drilled in the West Area.4.1.3. Relationships between geophysics, physical properties, and geologyAlthough gold itself is a dense and conductive mineral, its characteristic lowgrades in orogenic gold deposits make its direct detection using geophysical methods_____LDO — late diorite/doleriteI ISLO — mudstone - siltstoneISO — syenite intrusive, undividedVFO — felsic volcanic, rhyolite, rhyodaciteVUO — ultramafic volcanic, undivided>VMF — magnetic mafic volcanicVMO — mafic volcanic, basalt, andesite140difficult (Seigel et al., 1984; Doyle, 1990). Known vectors to gold such as hostingstructures, lithology, or related hydrothermal alteration mineral assemblages, however,may still be targeted remotely. Rock property studies on the Hislop deposit revealed theexistence of consistent relationships between certain physical properties, and potentiallymineralized rocks (Chapter 2). A summary of the results of these physical propertystudies are presented in the following sections. Table 4.1 summarizes the physicalproperty ranges for each rock type, and indicates anomalous ranges unique to some of theprospective rocks in the Hislop area.Figure 4.3. Cross section looking Northwest through the Hislop deposit, showinglocations of carbonate-dominated alteration and gold mineralization. Cross sectioninterpreted from drill core logged from the West Area of the Hislop property (see Figure4.2).DDH H9601 DDH Ext 280, GK 280, and H9605Multi-lithic Volcanic BrecciaLamprophyric DikeLIIIntermediate DikePorphyritic Rhyolite DikeSyenite IntrusiveMafic Volcanic RockUltramafic Volcanic RockFault. Drill trace0141Table 4.1. Typical and anomalous physical property ranges for principal rock types occurring inthe Hislop deposit area.Unaltered ultramafic(dolomite-chloriteassemblage)Unaltered ultramafic (talc-chlorite assemblage)Fe-carbonate-muscovitealtered ultramaficMagnesite-fuchsite alteredultramaficrocks)111-8546Range(all ultramaficrocks)2.90-15.967Unaltered maficFe-carbonate-muscovitealtered maficFe-carbonate-albite alteredmafic0.35-141 2.70-3.080-2.19 2.78-2.970.28-1.27 2.76-2.86(all mafic rocks)541-58754(all mafic rocks)2.07-46.5Carbonate-0.2-2.19 2.76-2.97altered maficUnaltered intermediateintrusiveCarbonate alteredintermediate intrusiveCarbonate-muscovite alteredintermediate intrusive0.24-135.29 2.72-2.950.13-3.8 2.67-2.950.32-1.55 2.76-2.94(all intermediaterocks)2314-22613(all intermediaterocks)9.45-15.4Carbonate-alteredintermediateintrusive0.13-3.8 2.67-2.95L’JSyeniteRhyolite porphyry0.07-0.42 2.64-2.740.05-0.41 2.57-2.80(all syenites)2631-9400(all rhyolite dikes)8976-11525(all syenites)6-15.67Felsic(all rhyolite dikes) intrusives2-20.40.05-0.42 2.57-2.80Mag. Sus. Density Resistivity Chargeability Cut-off values for queryingdataRock Type(x103 SI) (glcm3)(Ohm-rn) (ms)Range Range RangeIRes.IChg.0.57-12.5 2.82-2.89 (all ultramafic0.44-84.4 2.79-2.940.41-5.96 2.80-2.910.49-0.95 2.85-2.96Iock Type I Mae. Sus. I DensityUltramaficrocks111-85 46Carbonate-altered 0.41-5.96 2.80-2.96ultramaficSulfide-rich rocks Based on anomalies in inversion results Anomalous sulfides <1540 >120Magnetic susceptibilitySyenite and porphyritic rhyolite dikes have low susceptibility ranges distinct frommost intermediate to ultramafic rocks at Hislop (Fig. 4.4). Their susceptibility values rangefrom 0.05 — 0.42 x103 SI Units. Mafic and ultramafic volcanic rocks, and intermediateintrusive rocks have bimodal susceptibility populations (Fig. 4.4). This distributionindicates there are two distinct populations that make up the data. The high susceptibilitypopulation (> -10 x103 SI Units) is predominantly composed of least-altered Fe-richtholeiitic basalts and ultramafic volcanic rocks (mainly talc-chlorite schists). Physicalproperty studies revealed that the lower susceptibility population is partly composed ofcarbonate-altered intermediate, mafic and ultramafic rocks. These altered rocks (Figs. 4.5aand 4.5b) have susceptibility ranges from 0.13 — 5.96 x103 SI Units. Thus, for explorationtargeting purposes, any rocks with susceptibilities above 5-10 x103 SI Units, where thebreak in data in Figures 4.4 and 4.5 occurs, can be excluded as less prospective.3.203.103.002.90___________2.802.702.602.502.40 —0.01 0.1 10 100 1000Rock Types9 Ultramafic rocksMafic rocksX Intermediate dikesA Syenite intrusives0 Rhyolite porphyriesEC.)4-,C’,a) ø>bx1Magnetic Susceptibility (xl0 SI Units)Figure 4.4. Magnetic susceptibility plotted against density for the major rock types atHislop. Syenite intrusives and porphyritic rhyolite dikes have distinctly low density andmagnetic susceptibility ranges allowing them to be distinguished from intermediate, mafic,and ultramafic rocks at Hislop.1433.203.10- 3.002.902.802.70C)2.602.502.401 10 100Magnetic Susceptibility (x1O3SI Units)b.10000.13.203.103.002.902.802.702.602.502.40Imfc.LIf1;*“!‘Unaltered to AlteredUltramafic RocksDLst. altd. (Dol-chi)•Tlc+chl ultramaficQFeCb+ms altd. utiramaficDFe/MgCb+fu altd. ultrarnafic0.1 1 10 100 1000Magnetic Susceptibility (xl O SI Units)Figure 4.5. Magnetic susceptibility plotted against density for a) mafic and b) ultramaficvolcanic rocks from the Hislop deposit area. Carbonate-alteration destroys magnetite inmafic and ultramafic volcanic rocks, causing magnetic susceptibility to drop. Densityvalues increase slightly for altered ultramafic rocks. Abbreviations in legends: Lst. altd. =least altered assemblage; Chl+ab = chlorite+albite assemblage; FeCb+ms = Fecarbonate+muscovite; FeCb+ab = Fe-carbonate+albite; Dol+chl dolomite+chloriteassemblage; Tlc+chl = talc+chlorite; Fe/MgCb+fu = Fe/Mg-carbonate+fuchsite (chromemuscovite).144Low susceptibilities, however, do not uniquely identify felsic intrusive rocks andcarbonate-altered samples. Fe-poor tholeiitic basalts (not differentiated from Fe-richbasalts on plots) have low susceptibilities that overlap with the susceptibility range ofprospective carbonate altered rocks. This means that targeting low susceptibility areas willnot exclusively target prospective rocks, and if possible, other criteria should be used tofurther discriminate the different rocks types that exist within the low susceptibility range.DensityDensity studies indicate that syenite and porphyritic rhyolite dikes have lowdensities compared to other rock types in and around the Hislop deposit area with rangesfrom 2.57-2.80 g/cm3 (Fig. 4.4, and Tab. 4.1). All other rock types and their alteredequivalents have higher density ranges generally greater than 2.75 g/cm3.Density datamay thus be used to further distinguish between low susceptibility felsic intrusive rocks,and low-susceptibility carbonate-altered rocks or Fe-poor tholeiitic basalts, where felsicrocks would have low susceptibilities and low densities, and carbonate-altered rocks andFe-poor tholeiites would have low susceptibilities and higher densities.Although density ranges for least-altered and altered mafic and ultramafic rocksgenerally overlap, a trend of increasing density in altered ultramafic rocks with carbonatealteration was indicated (Fig. 4.5b, and Chapter 2). This suggests that where ultramaficrocks are knowi to dominate within an area, it may be possible to identify carbonate-altered rocks using density in addition to susceptibility.Resistivity and chargeabilityResistivity values measured in the lab are not consistently representative of largerscale measurements since there can be large scale features in the rock controllingresistivity that are not present at the hand sample scale (www.zonge.com!LabIP.html). Forinterpreting resistivity data and relationships to geology, sample measurements are bestcompared to one another on a relative scale. From Hislop physical property studies,resistivity was determined to be partly controlled by rock texture, specifically porosity andschistosity. Low resistivity (or high conductivity) values associated with sheared and145porous ultramafic volcanic rocks may distinguish them from other Hislop rock types,which otherwise have higher, overlapping ranges of resistivity (Fig. 4.6). A pattern ofincreasing resistivity with carbonate-alteration occurs in ultramafic rocks. The increasedresistivity ranges related to carbonate-altered ultramafic rocks, however, begin to overlapwith the resistivity ranges for other rock types.43264286Although most sulfides are known to be conductive (Telford et al., 1990), therewere no significant correlations observed between pyrite abundances derived from XRI)(Rietveld) analyses and resistivity during physical property work (Chapter 2, Appendix2G).Compared to resistivity measurements, chargeability measurements made in thelab are less inconsistent with larger scale measurements, and can thus be trusted to betterrepresent in-situ chargeability. Chargeability values do not distinguish between differentUltramafic volcanic rocks64242Mafic volcanic rocksIntermediate dikeslear! 2816___I___Mean 28432n 9759Porphyritic rhyolite dikes Mean 1153410100 1000 10000 100000 1000000Resistivity (Ohm-rn)Figure 4.6. Resistivity histograms for Hislop deposit rocks. Data indicates lower overallresistivities for ultramafic volcanic rocks from Hislop.146rock 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 thepresence of sulfide minerals (Telford et al., 1990), physical property studies at Hislopindicate only a weak trend between pyrite abundance and chargeability, and only for felsicrocks (Fig. 4.8). Chargeability studies at Hislop also suggest that porosity may decreasechargeabilities, complicating relationships between this physical property and maficvolcanic rocks (Fig. 4.9, Chapter 2). Despite the lack of direct correlation between sulfidesand higher chargeabilities for Hislop drillhole and surface samples, induced polarizationhas been used successfully to target sulfides in previous exploration efforts in similargeological settings (Johnson et al., 1989; Bate et al., 1989; Hallof and Yamashita, 1990).4.1.4. Inversion backgroundGeophysical inversion methodology is regularly used throughout industry,government, and academia, to investigate the Earth’s subsurface geology and explore formineral deposits (Oldenburg et al., 1998). Geophysical inversion can be considered theopposite of geophysical forward modeling processes. Whereas forward modeling involvescalculation of a geophysical response from a known, or hypothetical subsurface physicalproperty model, geophysical inversion involves a calculation of the subsurfacearrangement of physical properties, based on surface measurements, that is capable ofcausing an observed dataset.147a-Cu01086426424210.00Chargeability (ms)Figure 4.8. Chargeability plotted against pyrite abundance for Hislop samples. A weakpositive correlation exists between pyrite abundance and chargeability, however the trendis mainly controlled by porphyritic rhyolite dike and syenite samples. There is no evidenceof a consistent relationship between chargeability and pyrite abundance for intermediate toultramafic volcanic rocks.Ultramafic volcanic rocks[n 6.9LL- Mafic volcanic rocks:Mean 31.32rIntermediate dikes -rrlean46- Syenite intrusives -6 -Porphyritic rhyolite dikes_2_______Mean 10.47110 100 1000Chargeability (milliseconds)Figure 4.7. Chargeability histograms for Hislop deposit rocks. Chargeability ranges for theindividual rock types overlap and are not unique.1010Rock Types0 Ultramaflc rocks•Mafic rocksx Intermediate dikesAASyenite intrusiveso Rhyolite porphyriesAAA0001.00 100.0014810Mafic rocks(j)l00000 I1.00 10.00 100.00 1000.00Chargeability (ms)Figure 4.9. Chargeability versus porosity for mafic rock samples from Hislop. A negativecorrelation between chargeability and porosity in this plot indicates that increases inporosities of mafic volcanic rocks at Hislop may hinder the ability for metallic minerals tobecome charged.One of the limitations of inverting geophysical data is that the solution is nonunique. Due to the fact that there are a greater number of unknowns (i.e. cells in thediscretized model volume), than there are data, the problem is underdetermined. There aremany distributions of physical properties that can cause the same observed data set. Toalleviate this non-uniqueness, a model objective function, or model ‘goal’, is defined, sothat the model outcome is consistent with expected geology. Additionally, a specific misfitmust also be achieved between observed data and predicted data calculated from therecovered model. The inversion process is an iterative process. The model will be recomputed numerous times in an attempt to minimize differences between the predictedand observed data sets, and to satisfy the terms of the model objective function. Detailedinversion calculations can be found in Li and Oldenburg’s (1996, and 1998) papers on 3Dgravity and magnetic inversions.Where geology is better understood, and/or where physical property data isavailable (published data, downhole data, drill core, or outcrop measurements) inversionscan be more thoroughly constrained. Physical property data, or reference models areincorporated into the descretized volume of interest. The model then has to be estimated149such that the incorporated data is honored. Physical property bounds can be specified tolimit 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 intothe inversion algorithm (alpha weightings, cii, c’ and ct) can invoke geologicaldirectionality.Constraining inversions should result in more accurate models, and betterestimated physical properties throughout the model (Phillips, 2002; and Williams, 2006;also Chapter 3). It also further reduces non-uniqueness. Generating multiple inversionmodels 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 APPROACH4.2.1. General strategyUnconstrained inversions of airborne magnetic data, airborne gravity data, and DCresistivity and IP data, were completed over the Hislop deposit area. As there exists asignificant amount of magnetic susceptibility data for Hislop deposit area rocks, and thereare well-established relationships between magnetic susceptibility and geology, magneticinversions are also inverted with constraints incorporated via reference models built usingWilliam’s (2008) GiFtools ModelBuilder software. Constraining data include downholesusceptibility measurements, and surface sample susceptibility measurements.Additionally, cells within the model mesh where the physical property values can beestimated based on interpreted geology are also constrained. Further details onModelBuilder applications are given in section 2.5. Both unconstrained and constrainedmagnetic inversion results are presented herein. Gravity, DC resistivity, and IP inversionsare 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 referencemodels, and bounds on physical property ranges, were used successfully to improveinversion results in synthetic modeling studies (Chapter 3). Only constrained gravity, DC150resistivity, and IP inversion results are presented here, although all results are included inAppendix 4A.Inversion results are interpreted with respect to mapped and interpreted surfacegeology, and geology logged from drill core. Recovered models are interrogated throughquerying 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 inversionsA high resolution airborne magnetic survey completed in 2002 covers an area ofroughly 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. Themagnetic data extents are shown in Figure 4.10, and the magnetic data are given in Figures4.11 to 4.13. The datasets to be used in the inversion must include estimated standarddeviations. 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 parameterdetails 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 x1.5 km). Cells making up the core volume of interest in the regional, local, and depositscale models are 200 m2,50 m2,and 25 m2,respectively. Inversion parameters are detailedin Table 4.3. Local and deposit scale inversions are completed from magnetic data that hashad larger scale magnetic signatures removed using regional removal methods describedby Li and Oldenburg (1998). There is not sufficient data coverage to remove any largerscale 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 basedon maximum coverage required to explain any subsurface features that might occur withinthe core volume of interest. Padding cells were added along the perimeters of the inversion151Figure 4.10. Extents of magnetic data used in the deposit-, local-, and regional-scalemagnetic inversions (red outlines), of gravity data (blue outline) used in the regional-scalegravity inversion, and of DC Resistivity and IP data used in corresponding deposit andlocal scale inversions (yellow outline). Geological map from Power et al., 2004. SeeFigure 4.2 for geology legend.IRegional Scale FI71Z[lEL16.5 km152I496139285371507_Ct567753_5356493 -1238nT547158 550678 554198 557719 561239 564759 568280Easting (m)Figure 4.11. Data used in regional-scale magnetic inversion. Local-scale magnetic datasetoutlined. Refer to Figure 4.10 for corresponding geology.5382767_Observed Magnetic Data12355data, 1= 75, D-125379014_5375260_5364000 -5360246- 2895- 1862828.5-204.6153Figure 4.12. Data used in local-scale magnetic inversion. Deposit-scale magnetic datasetoutlined. Refer to Figure 4.10 for corresponding geology.Figure 4.13. Data used in deposit-scale magnetic inversion. Refer to Figure 4.10 forcorresponding geology.Observed Magnetic Data27976data, 1=75, D=-125376170 580646345374274- 3461572378- 2288570482 -11155368586-57.75366691_ - -1231I nT547067I I I548932 550798 552663 554528 556393 558259Eastinq Cm)Observed Magnetic Data92O9data, 175, D-125373140_ 2708- 21265372440154357174020)571040- 378.45370340- -2045369640 -786.5I I I I InT550665 551325 551984 552644 553303 553963 554622Easting (m)154Table 4.2. Survey parameters.Magnetics x: 547067 - 568259(local) y: 5366691 - 5376170Gravity x: 545003 - 561995y: 5365060 - 5377899x: 550100 - 555300y: 5370100 - 5372400IP (local) x: 550100 - 555300y: 5370100 - 5372400IP (deposit) x: 551100-554300y: 5370400 - 5372400N-S margins margins draped 27982 2%;(lOOm) (100) 50m floorcentre centre l4OnT(—50 m) (—50)N-S 50 m 25 m draped 9209 2%;50 m floor5OnTE-W 500 m 250 m constant 1850 0.01468 m mGalfloorSW-NE lOOm 20m ground 6545 10%max.VoltageSW-NE lOOm 20m ground 4576 10%max.VoltageSW-NE 100 m 20 m ground 6545 10%max.VoltageSW-NE 100 m 20 m ground 4576 10%max.Voltagelnclination:75°;Declination:-12°;Strength: 57478 nT2002 lnclination:75°;Declination:-12°;Strength: 57478 nT2002 lnclination:75°;Declination:-12°;Strength: 57478 nT20031996 Realsection survey -5 Tx spacings: 1000m, 1500 m, 2000 m,2400 m, 3200 m (26lines)1996 Realsection survey -5Tx spacings: 1000m, 1500 m, 2000 m,2400 m, 3200 m (20lines)1996 Realsection survey -5Tx spacings: 1000m, 1500 m, 2000 m,2400 m, 3200 m (26lines)1996 Realsection survey -5Tx spacings: 1000m, 1500 m, 2000 m,2400 m, 3200 m (20lines)Model Data area (UTM) Lines Line Station Height # Data DataYear Other informationspacing Spacing errorsMagnetics x: 547158 - 568280 N-S 200 m 200 m draped 12361 5%; 2002(regional) y: 5356493 - 5382767 50 m 300nTMagnetics(deposit)floorx: 550665 - 554622y: 5369640 - 5373140DCResistivity(local)DCResistivity(deposit)x: 551100-554300y: 5370400 - 5372400Table 4.3. Inversion parameters.Inversion # Data12361Inversion core extents(UTM)x: 541700 - 563500y: 5381350 - 5361550z: 500 - (-)4700Magnetics 27982 x: 550050 - 555150(local)y: 5369780 - 5373080z: 450 - (-)2150Magnetics 9209 x:551630-553630(deposit) y: 5370640 - 5372140z: 450 - (-)550Gravity 1850 x: 547500 - 559500y: 5367400 - 5375400z: 500 - (-)1500IP (local) 6545 x: 550700 - 554700y: 5370100 - 5372400z: 400 - (-)800IP (deposit) 4576 x: 551500 - 553900y: 5370400 - 5372400z: 400 - (-)200Topography used in all modelsunconstrained model: default alphas;constrained model: using reference modelbuilt in GIFtools (Tab. 4.4); L=2009013/ unconstrained model: default a values;9142 constrained model: using reference modelbuilt in GlFtools (Tab. 4.4); L1001892/ unconstrained: L=500, L=4001856constrained: L750, L=6006529/ Near-surface cell weightings applied;6455 unconstrained: default alphas;constrained: L200; L=100;reference value 0.00015 S/rn4670/ Near-surface cell weightings applied;4486 unconstrained: default alphas;constrained: L=1 00; L=50;reference value = 0.00015 S/rn6369/ Near-surface cell weightings applied;6445 unconstrained: default alphas;constrained: L200; L= 100;reference value = 0.031 ms4412/ Near-surface cell weightings applied;4519 unconstrained: default alphas;constrained: L1 00; L=50;reference value = 0.031 rnsMagnetics(regional)# Core Core cell # Padding Achieved Othercells size cells misfit280566200rn3178101 12321 unconstrained:a=0.0001,a=16;165312 centre50 rn3margins100 m3116928 29190(unconstr.) I28452(constr.)DCResistivity(local)DCResistivity(deposit)6545 x: 550700 - 554700y: 5370100 - 5372400z: 400 - (-)8004576 x: 551500 - 553900y: 5370400 - 5372400z: 400 - (-)20019200025m323606415360250 rn356460(xy);200 rn2 (z)5152050m315968018432025rn32498565152050rn315968018432025m3249856cJRelationship of a (alpha weight) to L (length scale): (L)2 = (a/a)2;similar for L, Lvolumes to a distance that is required to explain any features that might occur in thedataset but not directly within the volume of interest.Topographical information was used in all inversions. Topography data wasdownloaded from the Shuttle Radar Topography Mission (SRTM) online database. Datawas collected at approximately 90 m spacings.Located constraints applied to magnetic inversions are discussed in Section 2.5.4.2.3. Gravity inversionsAirborne gravity surveys over the northeastern, northwestern, and southernTimmins areas were completed in 2003 as part of the Discover Abitibi Initiative (OntarioGeological Survey, 2004). For the eastern Timmins survey (Fig. 4.10 and Fig. 4.14), linesare 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 aregional scale inversion could be performed with cell sizes of 250 m to correspond withan 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 notextend far enough beyond the chosen regional scale area to effectively remove a regionalsignature. A good correlation between gravity data and mapped geology indicates that alarger regional trend does not contribute strongly to the dataset, and the lack of a regionalremoval should not be detrimental to the inversion result. Gravity data was assignedstandard 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 expectedfor the rocks in the study area, and inversion smoothing weightings were increased.157Figure 4.14. Data used in regional-scale gravity inversion. Refer to Figure 4.10 forcorresponding geology.4.2.4. DC resistivity and IP inversionsA combined DC Resistivity and induced polarization (IP) survey was completedin 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 1000m up to 3200 m along lines, and receiver electrode spacing was 20 m. Measurementswere made in time domain for both DC resistivity and IP surveys. The required dataformat for inversion of DC resistivity data is the potential in Volts normalized by thecurrent (DCIP3D User Manual, version 2.1).5377899Observed Gravity Data1850 data5375331 -15.53p72763 -C7O195_111.35- 7.1765367627_2.9995365060_-1.177I I I545003 547835 550667 553499 556331Easting (m)-5.354559163 561995-9.531mGaI158Figure 4.15. Location of DC resistivity and IP lines used for 3D DC resistivity and IPinversions. 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 physicalinterfaces within a medium. Chargeability is measured by assessing the decay of voltageover time when the electrical current is shut off (Telford et al., 1990). Data over theHislop deposit was collected over 10 different time windows. Measured data is in mV/Vand a total apparent chargeability is calculated for this work as the sum of the voltagesrecorded for time windows 2 to 8, multiplied by 0.8. The value is divided by 1000 to getdata into the form V/V, the correct units for IP inversion calculations. Standard deviationson DC resistivity data and IP data are assigned at 10%, with a floor of 10% of themaximum voltage to avoid small errors on low data values (Tab. 4.2).When DC resistivity and IP data are displayed as pseudosections, the depth isusually arbitrarily assigned for visualization purposes based on n-spacings, or thedistance between the transmitters in this case (Telford et al., 1990). Pseudosections aresimply a method of displaying the data, and the z-scale does not represent depth. Thepositions and shapes of features also do not likely reflect the true geology. Inverting DCIE00CDC)7100 m159resistivity 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 foreach of the 39 survey lines prior to running 3D inversions. This helped to determineappropriate errors for 3D inversions, and to examine depth of investigation (Oldenburgand Li, 1999). 2D results were also compared to cross-sections through 3D inversionresults, 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 theHislop deposit were focused on for the 3D DC resistivity and IP inversions. The corevolume for the ‘local’ scale inversions was discretized into 50 m2 cells, and the corevolume 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 DCresistivity and IP inversions. A global reference model of 0.00015 S/m was applied to DCresistivity 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 propertyinformation. and ct, were increased relative to ct, to impart known geological fabrics.4.2.5. Constraining magnetic inversions with reference models built in ModelbuilderMagnetic inversions at the local and deposit scales were constrained usinggeological 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 Hislopphysical property study, as well as from a regional physical property study focused on160geology 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 theHislop 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 oroutcrop, can be input into the reference model with the appropriate associated drillholecollar and survey information, and with XYZ locations. These measurements, along withtheir linked geological and alteration information, form the basis of a physical propertiesdatabase for the reference model being created. Any other empirical geologicalinformation, including geological maps, outcrop maps, downhole geology logs, and 3Dgeological volumes, can be painted onto the model cells and subsequently translated intophysical properties by way of the program looking up the average physical property valuecalculated from previous property measurements for the rock type identified in the modelcell. Thus, it is possible to populate an entire layer of cells with physical properties basedon a geological map that covers the area, or to populate all cells intersected by a drillholewithout actual measurements made on the core. Since potential fields geophysics issensitive 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-likegeological setting (Chapter 3) showed that poor susceptibility estimation for near-surfacecells can affect the distribution of susceptibility throughout the whole model. Thus, whilegeological maps are available covering the area surrounding the Hislop deposit, onlyoutcrop observations were used to populate near-surface cells. Geological contacts androcks types from outcrop maps were considered more reliable than larger scale regionallyinterpreted geology maps.It is not uncommon to have more than one physical property data existing withinone cell. For the Hislop deposit inversions, cells are 25 to 50 m. Susceptibilitymeasurements were collected every 5 m on Hislop drillcore, adding 5 to 10measurements to a single cell as a result. Where cells have more than one type of data161(e.g. downhole physical property measurements, plus data assigned based on geologicalmapping or logging) a single representative value must be chosen. The ModelBuilderprogram presents a number of options for choosing the representative value, dependingon 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 bothinformation sources considered equally reliable.Smallness weights are assigned to the constrained cells. These weights relay to theinversion the degree of reliability of its assigned physical properties. If a high smallnessweight is specified, the inversion will attempt to achieve values close to the cell’sreference value. If properties within a cell are expected to be consistent over asurrounding volume, a ‘buffer’ can be designed around the cell. The information withinthe central cell is extrapolated to the cells of the buffer. Buffer cells might be assigned alow smallness weight, having a lower reliability than cells containing measured data.Refer to Table 4.4 for all constrained model parameters chosen for magneticinversions. The resulting susceptibility reference models constrain 10% of the local scalesusceptibility model, and 8.4% of the deposit scale susceptibility models.4.2.6. Inversion model displayFigure 4.16 outlines the surface extents of each of the model results to bediscussed herein. Inversion results are displayed as cross-sections through the recovered3D model for comparison to overlying mapped and interpreted geology. The location ofthe 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 anomalousmaterial in the subsurface, and are interpreted based on previously noted correlations withgeology. It is difficult to show the full 3D distribution of physical properties in a single3D representation. To appreciate the shapes and depths of anomalous areas throughout162Table 4.4. GiFtools ModelBuilder options chosen for building Hislop reference models.Local scale Deposit scaleGiFtools parameters magnetic magneticinversion inversionDefault parametersLowest 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 ofcloseness to reference model values)Property lower bound (default upper bound where no data)Property upper bound (default upper bound where no data)03000x103SI01 (low)0I 000xl 0 SI03000x103SI01 (low)01000x103SlSource dataDownhole property measurementsSurface sample measurementsDrillholes with geological observations and property measurementsDrillholes with geological observationsWeights and boundsBounds assigned to a cell are controlled by the contained propertydata, and are defined by data within the confidence interval of:Representative block size% of block requied to be filled before bounds allowed to be appliedRelative smallness weight (reliability weight) for surfacemeasurementsRelative smallness weight for drilling measurementsRelative smallness weight for drilling geology logsRelative smallness weight for outcrop geology mapBuffersSmooth interpolation across cellsMaximum buffer distance for surface measurementsMaximum buffer distance for drilling measurementsMaximum buffer distance for drilling geologyMaximum buffer distance for outcrop mapStrikeDipPitch% model constrained10 101934 90399.7% 99.7%Models builtReference model & smallness weightsLower & upper bounds modelSmoothness weights103411310345825 m75%25 m75%100 1001005050200 m200 m200 m200 m1005050lOOmlOOmlOOmlOOm115 11590 900 010% 8.4%163the model volume, the models should be viewed with a 3D viewer such as UBC-GIF’sMeshTools3D, or with a Gocad viewer. The models and their associated meshes areincluded, along with a MeshTools3D model viewer, as an appendix on a CDaccompanying the thesis (Appendix 4A). Instructions on how to use the viewer can befound on the UBC-GIF website, http://www.eos.ubc.ca!ubcgif/, under “Softwaremanuals”. Observed versus predicted data for all inversion results are plotted in Appendix4C (on CD).Esity regional volumeFigure 4.16. Extents of inversion model volumes, with cross-section location indicated.1644.3. INVERSION RESULTS AND ANALYSIS4.3.1. Magnetic susceptibility modelsRegional scale 18 km x 20 km model from unconstrained magnetic inversionHigh susceptibilities (>10 x 1 0 SI Units) in the regional magnetic model areassociated with the dark green units on the Hislop area geologic map, which correspondto Fe-rich tholeiitic basalts (Fig. 4.17). The faulting and folding of a series of Fe-richmafic rocks near the center of the map area, seems to thicken this rock package causingthe significant central high susceptibility anomaly. The central faulted package of Fe-richbasalt units that dominate the northern part of the Hislop deposit stratigraphy appear tobottom-out at a depth of -3000 m. The anomaly conveys a steep dip to the southwest. Alow susceptibility zone south of the central Fe-rich basalt package represents volcanicstratigraphy dominated by Fe-poor tholeiitic basalts and felsic volcanic rocks. It is notpossible to distinguish between these two low susceptibility rock types in thesusceptibility model result. High susceptibilities correlating with a series of Fe-richvolcanic flows persist through the southern region of the model, extending to depths ofaround 7000 m.A strong contrast occurs between the central high susceptibility zone, and lowsusceptibility rocks in the north, which is interpreted to be the manifestation of thelocation of the Porcupine Destor Deformation Zone. The belt scale PDDZ, mapped at thesurface to occur along the southern margin of an Fe-poor basalt unit south of thePorcupine and Timiskaming assemblage sedimentary rocks, is indicated to dip about 45° -60° southward beneath the interlayered mafic and ultramafic volcanic strata. Thisstructure may be truncating mafic and ultramafic rock packages at depth. The very lowsusceptibility volume north of the interpreted fault likely represents the sedimentary rocksequences of the Porcupine and Timiskaming assemblages, or a combination ofsedimentary rocks and Fe-poor mafic volcanic sequences.165The isosurface model in Figure 4.18 shows the regional subsurface distribution ofFe-rich mafic and ultramafic rocks recovered by the magnetic inversion.Figure 4.17. North-south cross-section through the regional-scale unconstrained magneticinversion result, with overlying geologic map of the greater Hislop deposit area. Forgeological legend see Figure 4.2. Inset shows extent of model volume and cross-sectionlocation. Figure 4.16 can also be referred to.166Regional scale magneticsusceptibility modelIsosurface cut-off: 29 x iO SI UnitsFigure 4.18. Isosurface model from regional scale magnetic inversion results.UltramaficintrusionLow susceptibility areasdominated by felsicvolcanics and sediments1OONRotation of magneticmafic and ultramaficstratigraphy acrossRoss and Hislopfaults167Local scale 4 km x 6 km modelThis inversion essentially zooms in on the structure of the central highsusceptibility body seen in the regional magnetic inversion result.Unconstrained inversionThe 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 toinversion-related smoothing over smaller distances in accordance with smaller cell sizes.The local scale model indicates the main susceptibility body is more structured thansuggested in the regional model. The central susceptibility bodies extending in segmentsto depth gives the appearance of having once been one coherent unit, that was laterdissected by near-vertical faults. A vertical low susceptibility zone in the south projectsupward 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 ofstructurally 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 withshallow high susceptibility bodies. A high susceptibility body to the north is likely relatedto a mapped ultramafic unit. The associated susceptibilities of this northern body areconsistent with those of talc-chlorite rich ultramafic rocks from the Hislop physicalproperty studies, being somewhat lower than susceptibilities characteristic of Fe-richbasalts (Chapter 2).Low susceptibilities are associated with Fe-poor basalts, rhyolite units, felsicintrusives, and faulted areas. The extremely low susceptibility area north of the PDDZand at depth is presumably reflecting thick packages of sedimentary rocks, which168Figure4.19.North-southcross-sectionsthroughthelocal-scalea)unconstrained,andb)constrainedmagneticinversionresults,withoverlyinggeologicmaps.ThecrosssectionspansthemainorezoneatHislop.ForgeologicalmaplegendseeFigure4.2.Insetshowsextentofmodelvolumeandcross-sectionlocation.Figure4.16canalsobereferredto.UnconstrainedlocalscalemagneticsusceptibilityConstrainedlocalscalemagneticsusceptibilityMagneticSusceptibility(xloSIUnits)100150were mapped north of the PDDZ on the geological map (Fig. 4.2), or Fe-poor mafic volcanicrocks, also mapped in the northern areas.The inferred Porcupine Destor Deformation Zone, mapped just north of the northernultramafic units, separates high and low susceptibility regions. Its dip is slightly shallower in thisresult than in the regional model result.Constrained inversionThe constrained local scale results exhibit some noticeable differences fromunconstrained results (Fig. 4.1 9b). The core high susceptibility body is clearly separated from asmaller susceptibility anomaly closer to the surface. The ultramafic body north of the highsusceptibility Fe-rich basalt units is more clearly disconnected from the basalts, and now has amore wedge-like appearance. Constraining the result also pushes high susceptibility bodies to agreater depth, steepening the dip angle of the inferred PDDZ, making it more consistent with the- 600angle suggested in the regional model results. The steep dip angle of the central maficvolcanic rock package, and additional features not seen in the cross-section are illustrated in theisosurface model in Figure 4.20.Deposit scale 1.5 km x 2 km modelThis inversion focuses on the core portion of the central high susceptibility basaltsinterpreted from the local scale magnetic inversions. The goal is to attempt to uncover morefine scale structure, and to locate narrow low susceptibility syenite and rhyolite dikes, andalteration zones known to be spatially related to gold mineralization.Unconstrained inversionFrom the recovered model (Fig. 4.21a), a sharp gradient is obvious between the mappedFe-rich basalt units, and the gold-related syenite. From synthetic modeling work (Chapter 3), anultramafic rock-syenite dike contact was not detectable at these scales of inversion, and theequivalent contact is not obvious here. The ultramafic rocks south of the central syenite dike are170apparently low susceptibility, which, from physical property studies, could indicate theiralteration to a carbonate-rich assemblage. Rhyolite dikes mapped to intrude the ultramafic unitcould also be lowering the susceptibility here. Low susceptibilities are spatially related, inFigure 4.21, to Fe-poor mafic units, ultramafic units, and faulted rocks.Figure 4.20. Isosurface model from local magnetic inversion results.53728805372155West5370705Tightly foldedmagnetic maficand ultramaficunitsNormal faultdisplacing Fe-richmafic units todepth/Felsicintrusion5369980550050Local scale magneticsusceptibility modelIsosurface cut-off: 101 x iO SI Units//4:/ 5S150171Unconstrained deposit scalemagnetic susceptibilityN— i15OMagnetic Susceptibility(x10 SI Units)50 100 150I I I200Figure 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 spansthe main ore zone at Hislop. For geological legend see Figure 4.2. Inset shows extent of modelvolume and cross-section location.a)Constrained deposit scalemagnetic susceptibilityN172The depth of the anomaly (-l000 m), and dip angle of the bottom of the highsusceptibility body is similar to the local inversion outcome. But, despite cell sizes being smallerin the deposit-scale model (25 m3), there is little more resolution gained. In fact, a separation inthe anomaly apparent in the local result does not occur in the unconstrained deposit model. Theapparent lower resolution at the deposit scale might be explained by the default ci weightings orlength scales (Mag3D User Manual, version 3.0, 2005). Length scales are applied in inversionwork to manipulate smoothing in given directions according to cell size and prior geologicinformation. The default length scales used in magnetic inversions corresponds to cells sizes of50 m (as was used in the local-scale inversions). Since the length scales were not reduced tocorrespond to smaller sizes in the unconstrained deposit-scale model, smoothing in the x, y, andz directions may be excessive.No obvious narrow low susceptibility zones characteristic of felsic dikes or altered rocksare distinguished within the susceptibility anomaly. The 50 m x 25 m spacing of the magneticdata used for this inversion limits the resolution of features smaller than this. In addition, asdiscussed in Chapter 3, smoothing inherent in the inversion result brought about by the choice ofa model norm that gives priority to smooth results, would further obscure small-scale lowsusceptibility zones.Constrained inversionConstrained deposit-scale results indicate there is more complex structure within thehigh susceptibility zones related to the Fe-rich basalts (Fig. 4.21b). Although the irregular shapeof the low susceptibility area within the central high susceptibility area is obviously an artifactof the location of the drillhole, and buffer zones, used to constrain the inversion, rendering theresult not particularly geologically realistic, it indicates the existence of more fine scale structurewithin the central susceptibility anomaly. The internal low susceptibility zones were determinedfrom drill core assessment to be related to the presence of syenites, Fe-poor basalts, andcarbonate-altered rocks. Hislop drill core logs presented in Chapter 2 indicate that significantchanges in rock type and alteration mineral assemblages, and thus susceptibility, can occur at thecentimeter scale. The cell sizes in the inversion limits the ability to resolve these fine scalefluctuations. Nonetheless, constraints can offset some of the smoothing that occurs in the173inversion result and highlight some of these small scale features. The presence of the lowsusceptibility zones forces susceptibility to be redistributed within the model, and its magnitudeto increase in the upper portion of the anomaly.As with the local-scale constrained inversion results in the previous section, a highsusceptibility zone to the north interpreted to be related to ultramafic rocks, now appears to bemore detached from the central high susceptibility Fe-rich basalts. An isosurface model for thedeposit-scale constrained magnetic inversion is shown in Figure 4.22.5372240WestN5371 81 5Distribution of Fe-rich mafic volcanicrocks in the Hislop deposit area -where highs drop away from5371390surface, may be representative offelsic intrusions, alteration or faults5370540SS1 530II15370965Deposit scale magneticsusceptibility modelIsosurface cut-off: 85 x iO SI UnitsPart of• .,.ultramafkunitdl52630Figure 4.22. Isosurface model from deposit-scale magnetic inversion results.1744.3.2. Density modelGravity data spacing over the greater Hislop depositarea is large (500 m x 120 m), assuch only a regional scale inversion was carried out. The correspondinglarge cell sizes (250 m)means only large scale features representing larger volumes dominatedby low density felsic orsedimentary rocks versus high density mafic and ultramafic rocksare resolved.Fe-rich and Fe-poor mafic, and ultramafic volcanic rocks that underliethe central portionof the mapped area, cause a high density zone to dominatethe core of the inversion volume (Fig.4.23). To the north, there exists low density materiallikely related to sedimentary rocks of thePorcupine and Timiskaming assemblages. Low magneticsusceptibilities in the same locationconfirm the dominance of sedimentary rocks at depth. The boundarybetween the central highdensity area and the northern low density area here is notthe same boundary that separates highand low susceptibility rocks in the regional magnetic results (seeFig. 4.17). This represents thecontact between dense Fe-poor tholeiitic basalts northof the PDDZ, and the adjacentsedimentary assemblages. This may explain the difference in theapparent dip of the geologicunits composing the central Hislop area between the magnetic andgravity inversion results.Low density areas in the southern regions of the model correlatewith a package ofrhyolitic volcanic rocks (pale yellow unit in inset of Fig. 4.23)and sedimentary rocks thatextend southeastward out of the section. The 3D distributionof mafic and ultramafic rocksversus felsic and sedimentary rocks can be observed fromthe isosurface model Figure 4.24.4.3.3. Resistivity modelsDC resistivity and IP inversion modeling was completedat two scales (Tabs. 4.2 and4.3), however, since results are similar, onlyfigures corresponding to the deposit-scale resultsare shown.175Figure 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 areaoverlies the model. For geological legend see Figure 4.2. Inset shows extent of the modelvolume and cross-section location.Regional scale densitywith global constraintsIi(r\r. _•—Density(glcm3)2.7 2.75 2.8 2.85 2.9 2.95I I I I__1765375150Figure 4.24. Isosurface density model from regional-scale gravity inversion results.From physical property studies, high conductivities (low resistivities) were found tobe associated with metamorphosed ultramafic rocks (talc-chlorite schists), whereas otherrock types were less conductive. Higher conductivities in the DC resistivity inversionresult,as expected, are associated with ultramafic rocks in the northern and central parts of themodel (Fig. 4.25a). High conductivities however, may be instead, or additionally, correlatedwith Hislop deposit sulfides near the center of the map, or with interpreted faults. Twosignificant 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 normallyN5371 400Low density areas aredominated byr”sedimentary unfts../• .and felsic volcanicrocks and intrusives..1//Regional scale gravity modelIsosurface cut-off: 2.83 g/cm3177Figure 4.25. North-south cross-section through the deposit-scale a) DC resistivity andb) IPinversion results, both inverted with non-located constraints. Hislop area geologicmap overliesmodels. For geological legend see Figure 4.2. Inset shows extent of the model volume and crosssection location.178b) Deposit scale chargeability modelIsosurface cut-off: 0.031Figure 4.26. Isosurface models for deposit-scale a) conductivity, and b) chargeability results.NArea dominated byultramafic rocksjConductivity anoma_,,,_..1weakly associatedwith syenite dikeSoutha) Deposit scale conductivity modelIsosurface cut-off: 0.00064 S/rnnorthArea where a NW-SEfault, west of the Rossfault, intersects the N.Arrow fault.NCorrelating anomaliesnear felsic intrusive!syenite dikeSouthChargeability anomalyfollows syenite dike554300179resistive, these anomalies may reflect the presence of sulfides. Alternatively, highconductivities may be associated with a conductive overburden, a feature indicated ininversion models generated for this area by Mueller et al. (2006). Outside of ultramaficrock-dominated areas and faulted areas, where geology is dominated by mafic and felsicrocks, lower conductivities (higher resistivities) occur.Depth of investigation tests (Oldenburg and Li, 1999) were conducted for selectedlines during 2D inversion work on the Hislop DC resistivity and IP datasets (Appendix4B), and indicate that subsurface features are generally not resolvable below about 400-600 m depth.4.3.4. Chargeability modelsA distinct high chargeability zone, likely to represent the presence of sulfides,occurs in the subsurface beneath the mapped syenite dike (slightly obscured in figure) atHislop, and extends to depth, dipping slightly to the southwest (Fig. 4.25b). This anomalydisperses horizontally at depth. The horizontal displacement does not correspond to anyknown features and is similar to artifacts in synthetic inversion models for IP data inChapter 3. Additionally, from depth of investigation tests, it is suspected that the model isessentially unreliable at these depths. An additional small anomaly occurs just north ofthe central chargeability anomaly in proximity to some interpreted faults, and overlappingwith a conductivity high.When the 3D model is viewed, the central chargeability anomaly extends a fewhundred meters to the northwest and to the southeast following the syenite dike, before itdetaches from the surface and moves to depth (Fig. 4.26b). Northwest of the West Areaopen 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 the180northeast in mafic rocks north ofthe PDDZ. The correlation between thetwo physicalproperties could indicate sulfide-richrocks.4.4. QUERYING COMBINEDINVERSION RESULTSData from Hislop 3D inversion modelswere combined using Gocad 3DGISsoftware with Mira Links add-ons,and the resulting ‘common earth models’queried inan attempt to define spatialextents of rock units, and potentiallyprospective areas forexploration targeting. It is importantto note that the local and deposit-scalemagneticsusceptibility models usedare from constrained inversions, whereasall other inversionresults making up the common earthmodels are unconstrained.This process involved projectingproperties from the different inversionresultsonto one discretized mesh. ForHislop, three common earth models werecreated, aregional scale model where susceptibilityand density were projected ontoa mesh with200 m cells, a local scale model,where susceptibility and chargeabilityfrom local scaleinversions were projected onto a mesh with50 m cells, and a deposit scale model,withsusceptibility and chargeabilitydata held in 25 m cells. Duringdata projection, each‘client’ cell in the common earthmodel grid takes on the value ofthe closest ‘server’(inversion) cell center.Physical property cut-offsused to query the common earth modelweredetermined using descriptive statistics calculatedduring Hislop physical propertystudies.In essence, susceptibilityand density are queried at the regionalscale with expectationsof modeling lithological units, orsignificant packages of rocks,and susceptibility andchargeability are used at the localand deposit scales to find sulfide-bearingfelsicintrusives, and carbonate-alteredrocks. Conductivity values do notuniquely defineprospective rock types, or hydrothermalalteration (Tab. 4.2), and is thusnot used in thequeries. High conductivites can indicatethe presence of faults that actas important181structural traps for gold mineralization, or sulfide-rich rocks, but high conductivities canalso be related to least-altered and likely unmineralized ultramafic volcanic rocks.The cut-off values used for common earth model queries are based on physicalproperty ranges characterizing rock types and alteration at Hislop (Tab. 4.2). The queryresults are presented in plan-view in the corresponding figures, with a transparentgeologcal 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 theHislop susceptibility versus density plot (Fig. 4.4): 1. low susceptibility-low density felsicrocks (Fig. 4.27), 2. high susceptibility-high density least-altered mafic and ultramaficrocks (Fig. 4.28), and 3. low susceptibility-high density carbonate-altered or Fe-poormafic 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, andsedimentary rocks mainly associated with the Porcupine and Timiskaming assemblagesin the northeast map area are isolated (Fig. 4.27). Felsic intrusive bodies near the centerof the mapped area overlying this model are not detected by this query. This might relateto the large cell sizes used and the overwhelming of smaller lower susceptibility anddensity zones by the more abundant susceptible and dense mafic and volcanic units, aneffect noticed in synthetic modeling results (Chapter 3).A query targeting high susceptibility (>5 x iO SI Units), and high density (>2.8g/cm3)cells in the regional scale common earth model targets areas dominated by Fe-richbasalts and ultramafic volcanic rocks in the central and southern parts of the map area182(Fig. 4.28). High susceptibilities and densities stop abruptly at the mapped location of thePDDZ.Figure 4.27. Result for a physical property query targeting low magnetic susceptibility-low density cells within the regional-scale common earth model. Anomalous zonesextend to greater than 2000 m depth. Plan view with transparent geology. Geologicallegend in Figure 4.2.A low susceptibility (<3 x SI Units) and high density (>2.8 g/cm3)queryidentifies 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, underlyingmapped Fe-rich basalts and ultramafic rocks, may warrant further inspection as the lowsusceptibilities here could indicate carbonate alteration of these normally highsusceptibility rocks. A northern zone of low susceptibility-high density cells extends froma sequence of mafic rocks just north of the PDDZ, into the mapped Porcupine assemblagesedimentary rocks. Sedimentary rocks elsewhere have typically low densities, and thisanomaly might indicate that the contact is interpreted incorrectly.Showsallcellsthatmeetquerycriteria—çromsu 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 544I — S S S S S S S S S S ••eaa-. a L- a a a a a a a a a a seSaseeSa.... — see— a — — a a a . • a a a — — a s ——• • — 5 esa a a a 5a.—...—Regional scalelow susceptibility -low density query/11111111ulla_IlIl.Ia III III ••a..-. 11111111 1 km183Figure 4.28. Result for a physical property query targeting high magnetic susceptibility -high density cells within the regional-scale common earth model. Anomalous zonesextend to greater than 2000 m depth. Plan view with transparent geology. Geologicallegend in Figure 4.2.4.4.2. Local scale query (susceptibility, chargeability)Cells in the local scale common earth model containing low susceptibilities (<3 xi03 SI Units) combined with high chargeabilities (>0.12) were targeted to identifypotentially prospective felsic intrusive rocks, carbonate altered zones, and sulfide-richareas.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-richbasalt unit), and especially where the contacts are faulted, or where two or more faultsintersect. Along the southern central syenite dike contact the highlighted zones are184%%%%‘a -aI • aaaaaa •‘a —Regional scalehigh susceptibility -high density query• S S S SS S SøØ-a......• alaaaa 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..Sa. a a ..weeS —• aaaaa_a S — a .aaaaasa aanaaa a — — — a a a a a a a a a a a a a a a a — a — a aaaaas.. 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 — aa 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 am,vIi1$.%ja I II a.a 00’ a aPWIØWFYIOa.. -:a.I lila. at aa. at a a a laa a at%%%aaa1kmaaaa.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 zonesextend up to 1500-2500 m depth. Plan view with transparent geology. Geological legendin Figure 4.2.coincident with some high gold abundances (Fig. 4.30). Returned anomalies also coincidewith higher gold concentrations near the northern Arrow Fault, and where the northernArrow Fault intersects the north-south trending fault west of the Ross Fault. An anomalynorthwest 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, neara felsic intrusion, is in an area of minimal to no drilling.Some high gold values occur in association with a drilled area southwest of theHislop deposit, near the Hislop Fault. Only about half of this drilled area is containedwithin the common earth model. The query did not identify prospective rocks here.Referring back to the regional susceptibility-density queries however, this area correlateswith low susceptibilities. It is possible that the gold here is not associated with sulfidesRegional scalelow susceptibility -high density querya. a .d’øa a a a• .ø•sm—a• a.• . .. a.%%’•••a aa a a•• . a a a . .•a. • • • • • •• .,at sat.......•...aa...•.•• .5S... •S•I a.....p,... .•S. SS •••S• •.••.•.a... ..S.. .t....t.I....u...aa..a.....aS • 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 — SS 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 aa.aaaatt•.,S. •...a.att..as Ia••.S•S•s-,. a 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. %_ •:.0.11kmr’.185and thus chargeability values are not anomalous at this location. This area is currentlybeing explored by Stroud Resources (www.stroudresourcesltd.com), and the endeavor isconfusingly 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 zonesextend up to 3 00-600 m depth. Anomalous downhole gold assays are indicated. Planview 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 andhigh chargeability criteria, returns several zones with the desired characteristics, whoselocations are generally consistent with local scale query results (Fig. 4.31). There isslightly more detail compared with the local scale results, with some additional smallregions highlighted, and others eliminated. Again, most prospective areas are spatially1 km..fl..flfl...S5%‘!N”‘N, •.. “‘.::::::::, N. Arrow Fault.::::::::..\‘ “• ...••:.•H 4S.ArrOW_Local scalelow susceptibility -highchargeability queryGold(ppm)4 6 8 10 12 14 16186associated with areas of complex faulting, and are also proximal to felsic intrusives anddikes (some narrower dikes south of the main syenite dike are obscured by the anomaliesand plotted gold assays, but can be see more clearly in Fig. 4.2). As with the local-scaleresults, there are areas of high gold concentrations not detected by the query that mayrepresent 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 zonesextend up to 60-500 m depth. Anomalous downhole gold assays are indicated. Plan viewwith transparent geology. Geological legend in Figure 4.2.The query results are not expected to detect all subsurface areas meeting thecriteria. Constrained deposit scale magnetic inversion results indicate that there are smallscale low susceptibility zones that can be masked by smoothing of high susceptibility1 km..4... flu a....•:..:::.n :::::.1iF:EE...jJEra.“ ...Deposit scalelow susceptibility -high chärgeability queryS.ArroW9itGold(ppm)2 4 6 8 10 12 14 16I I I I I187values in the inversion result. There is likely more detail in high susceptibility rocks thatcannot be resolved using inversion, or querying techniques at this scale.4.5. SUMMARY AND DISCUSSIONGeophysical inversion of a series of geophysical datasets over the Hislop golddeposit in the south-central Abitibi greenstone belt, was carried out at a range of scales ofinvestigation. Results show that inversion is a useful tool for detecting specific lithologicpackages, and altered and mineralized rock, and for interrogating their 3D subsurfacedistribution.Regional scale inversion results highlight large scale structures, and lithologicalboundaries. Magnetic results show packages of susceptible rocks, dominated by Fe-richtholeiitic basalts, extending to depth in the crust up to about 7 km. This depth isconsistent with published depths for crustal rocks above granitic basement rocks in theAbitibi greenstone belt (Reed et al., 2005). Magnetic inversion traces the crustal scalePorcupine Destor Deformation Zone, a regionally important gold-related structure, intothe 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 volcanicrock. This southward dip is consistent with results for recent seismic work, and magneticand 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 itstrace, however, changing from 45°- 65° in the Hislop area to steeper angles closer to theOntario-Quebec border (Berger, 2002).At the regional scale, major lithologic units and domains are mapped usingcombined magnetic and gravity inversion results. Querying the combined results revealedthree petrophysically distinct lithological packages, and importantly, allowed felsic andsedimentary rocks (low susceptibility-low density) to be distinguished from Fe-poortholeiites and potentially carbonate-altered mafic and ultramafic rocks (low188susceptibility-high density). Exploration at the regional scale might focus generally onthe range of low susceptibility regions, which are expected to contain dominantly felsicrocks, and carbonate-altered rocks. Associated lithogeochemical studies testing variousalteration indices (Davies et a!., 1990; Eilu et al, 1995; Piche and Jebrak, 2003) could behelpful in further distinguishing least-altered Fe-poor mafic rocks from carbonate-alteredrocks. Regional scale inversions may be more appropriate for mapping geology in areasof poor outcrop than for generating targets directly. Although high susceptibility and highdensity rocks are likely to reflect mainly least-altered mafic and ultramafic rocks, it ispossible that smaller zones of prospective low susceptibility-low density rocks withinthese larger rock packages are not being detected.At the local and deposit scales of magnetic inversion, more detail is resolvedwithin the subsurface. Distributions of Fe-rich basalt, versus Fe-poor basalt and felsicrocks are better defined, and locations and orientations of near-vertical faults dissecting acentral package of high susceptibility Fe-rich basaltic rocks are discernible. There is not asignificant increase in detail visible in 25 m3 cell deposit scale inversion over the 50 m3cell local scale inversion. Features smaller than 25 m, which might constitute narrow,mineralization-related felsic dikes and alteration zones, are simply not detectable at thesescales, with magnetic data spacing limited to 50 m x 25 m, and typical inversion-relatedsmoothing occurring. More detail was indicated however, when constrained magneticinversions are completed at the deposit scale, and it becomes apparent that there are smallscale heterogeneities in the physical property distributions that are not being detected byunconstrained inversions. Synthetic modeling studies showed that narrow lowsusceptibility and low density zones can be imaged down to at least a few hundred metersat 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 dataspacing, 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.189Unfortunately gravity data available for this work was extremely sparse and could not beused to model the subsurface at the local or deposit scales. Having both high resolutionmagnetic and gravity data available for deposit-scale exploration would be very useful forgeological mapping, and detecting gold-related rock types at smaller scales ofinvestigation.DC resistivity inversions image some ultramafic volcanic units in the Hislop area,as was predicted by the consistently low resistivities (high conductivities) of thesecommonly sheared rocks indicated from physical property studies on hand samples(Chapter 2). Through inversion work, it was shown that high conductivities are alsocorrelated spatially with faults. Whether this is related to fluid content and porosity whichincreases conductivity (Telford et al., 1990) or to the presence of sulfides was notdetermined. Conductivity was not used in common earth model queries. Conductivitydoes not consistently detect prospective rocks in the Hislop area. High conductivities mayindicate faults which do not necessarily host mineralization, or conductive talc-chloriteschists which are not typically mineralized. Induced polarization inversions locatechargeability anomalies that are interpreted to be due to the presence of sulfides. Theseanomalies align with the immediate location of the Hislop deposit proximal to the centralnorthwest-southeast trending syenite dike.Again, by combining the results of different inversions, the most information isgained. At the local and deposit scales, queries of combined susceptibility andchargeability data were used to try and locate sulfide-rich felsic dikes and carbonate-altered rocks. Query results highlighted zones focused along the mainly faulted contactsbetween Fe-rich basalts and other rocks, and near cross-cutting faults. These areashighlighted by the queries in the area of the Hislop deposit are geologically ideal goldtargets, with faults providing conduits and structural traps for hydrothermal fluids, andnearby Fe-rich rocks that promote sulfidation processes leading to gold precipitation(Mikucki, 1998). Some areas where high gold grades were intersected during drillingwere targeted by the queries, confirming prospectivity.190Although gold mineralization hosted in greenstone facies rocks does not usuallyhave a strong geophysical signature due to its typically low grades, it is still possible toremotely target Archean orogenic gold deposits using alternative exploration vectors suchas hosting lithology, and alteration mineral assemblages. Geophysical inversion not onlyallows detection of prospective gold-related rocks but can indicate the spatial extent ofthese rocks in the subsurface. Geophysical based mapping of geology, and explorationtarget generation is so valuable in Archean greenstone terranes since they are oftencharacterized by low percentages of outcrop.The key to getting the most from inversions is by understanding relationshipsbetween physical properties in the geological environment or mineral deposit setting ofinterest. With this prior knowledge inversions can be constrained to yield resultsconsistent with geological observations, inversions will be more confidently interpretedoverall, and queries can be developed to target model cells with the desired combinationof physical property characteristics representative of mineralized rocks.191REFERENCESAyer, 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 andsedimentation: Precambrian Research, v. 115, p. 63-95.Bate, S.J., Thorsen, K.R., and Jones, D., 1989, The Casa Berardi area; an exploration casehistory: 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, OpenFile 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 ofprogressive hydrothermal carbonate alteration in Archean metabasalts andmetaperidotites: Mineralium Deposita, v.25,p.65—72.DCIP3D User Manual, version 2.1, A program library for forward modelling andinversion of DC resistivity and induced polarization data over 3D structures: UBCGeophysical 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 Archaean192lode-gold deposits, in Pasava, J., Kribek, B., and Zak, K., eds., Mineral Deposits; fromtheir Origin to their Environmental Impacts, Third Biennial SGA Meeting, Prague, CzechRepublic, p. 113-116.Farquharson, C.G, Ash, M.R., and Miller, H.G, 2008, Geologically constrained gravityinversion 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-bearingsulfide 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 ofExploration 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 spectralIP 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 hostedmesothermal 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 InternationalExposition and72ndAnnual Meeting, 4p.Li, Y., and Oldenburg, D.W., 1996, 3-D inversion of magnetic data: Geophysics v. 61,p.394-408.193Li, Y., and Oldenburg, D.W., 1998, 3D inversion of gravity data: Geophysics, v. 63,p.109-119.Mag3D User Manual, version 4.0, 2005, A program library for forward modelling andinversion 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 presentationof geophysical data for geoscientific profiles in the Timmins—Kirkiand Lake area:Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6189 , 28p.,15sheets.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 electricalresistivity and induced polarization methods: Ontario Geological Survey, MiscellaneousRelease—Data 181.Mira Geoscience Limited, 2005b, Detectability of mineral deposits with potential fieldmethods: Ontario Geological Survey, Miscellaneous Release — Data 177.Oldenburg, D.W., and Li, Y.,1999, Estimating depth of investigation in dc resistivity andIP 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 mineralexploration problems: The Leading Edge, v. 17,p.461 - 465.194Oldenburg, D.W., Li, Y., and Ellis, R.G., 1997, Inversion of geophysical data over acopper 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 RockProperty 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, GeophysicalData 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 BritishColumbia, 237p.Piche, M., and Jebrak, M., 2003, Normative minerals and alteration indices developed formineral exploration: Journal of Geochemical Exploration, v. 82,p.59-77.Poulsen, K.H., Robert, F., and Dube, B., 2000, Geological classification of Canadian golddeposits: Bulletin - Geological Survey of Canada, Report 540, 106p.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., andArchibald, N. J., 2004, Geoinformatics evaluation of the eastward extension of theTimmins Gold Camp: Geoinformatics Exploration Inc., Unpublished report for StAndrew Goldfields Ltd.Prest, V.K., 1956, Geology of the Hislop Township: Ontario Department of Mines,Annual Report, 1956, v. 65, 51p.195Reed, L. E., Snyder, D. B., and Salisbury, M. H., 2005, Two-dimensional (2D) reflectionseismic surveying in the Timmins-Kirkiand Lake area, northern Ontario; acquisition,processing, interpretation: Discover Abitibi Initiative, Ontario Geological Survey, OpenFile Report 6169, 68p.,10 plates.Reed, L. E., 2005, Gravity and magnetic three-dimensional (3D) modeling: DiscoverAbitibi Initiative, Ontario Geological Survey, Open File Report 6163, 40p.,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, SecondEdition: Cambridge University Press, 770p.UBC-GIF Inversion for Applied Geophysics CD-ROM, 2000-2006, a teaching andlearning CD-ROM, by Oldenburg, D.W., and Jones, F.M.: University of BritishColumbia, Geophysical Inversion Facility.Williams, N.C., 2006, Applying UBC-GIF potential fields inversions in greenfields orbrownfields exploration: Australian Earth Sciences Convention, 2006, Melbourne,Australia, extended abstract, 10p.Williams, N.C., 2008, Geologically-constrained UBC—GIF gravity and magneticinversions with examples from the Agnew-Wiluna greenstone belt, Western Australia:Unpublished Ph.D. Thesis, The University of British Columbia, 479p.www.standrewgoldfields.com, website for St. Andrew Goldfields, Ltd.196www.stroudresourcesltd.com, website for Stroud Resources Ltd.www.zonge.comlLablP.html, website for Zonge Engineering and Research Organization,IP and resistivity measurements.197Chapter 5: Summary and future work5.1. SYNTHESIS OF RESEARCH PRESENTEDThe goal of this research was to apply an understanding of the characteristicgeology and physical properties of a typical Archean orogenic gold deposit togeophysical inversion for improved mapping of geology and delineation of gold-relatedrocks in the subsurface. The Hislop deposit of eastern Ontario was used as an example ofthis deposit type in the case study.Key relationships between geology and alteration, and physical properties wereestablished for Hislop, and ranges of physical properties representing more prospectivegeology were identified. Physical property information was eventually used to improveinversion results through their incorporation as inversion constraints. Synthetic modelingrevealed the sizes and depths, and necessary physical property contrasts required toimage petrophysically distinct gold-related features in the subsurface. It allowed depthsof investigation to be roughly determined, and allowed certain inversion artifacts to beidentified. Preliminary physical property, and synthetic forward and inverse modelingwork contributed strongly to how eventual inversion models for the Hislop deposit wereinterpreted.At larger scales of investigation, magnetic and density data can be used formapping geology, and for determining regional-scale exploration targets based on thedistribution of the geologic units and structures modeled. Results are especially useful inthe parts of the Abitibi greenstone belt that were modeled during this study, as outcroppercentages are low. At deposit-scales of investigation, induced polarization (IP)inversion methods were effective in detecting sulfide-rich rocks, with chargeabilityanomalies correlating well with known mineralization. DC resistivity inversion resultswere not easily interpretable due to the variable behavior of conductivity. Somecorrelations between chargeability and conductivity anomalies in select areas surrounding198the Hislop deposit, however, may suggest the presence of potentially gold-bearing,sulfide-rich rocks. Magnetic inversions at the local and deposit scales identified acomplex distribution of faults characterized by low susceptibilities possibly brought onby magnetite-destructive carbonate alteration. Smaller scale features, such as the gold-related syenite dike at Hislop, were obscured by smoothing of higher susceptibilitieswithin the inversion volume. Synthetic modeling work has indicated that more detail canbe derived from inversion, permitting better resolution of the narrow features thatcharacterize typical Archean orogenic gold deposits. But to attain this detail, it isnecessary to focus on a small area, collect closely-spaced data, and to use small inversioncell sizes. The density data available for the Hislop area was very widely spaced. Fromphysical property studies and inversion investigations, it is expected that smaller scaledensity data in combination with closely spaced magnetic data would be effective inestablishing geological contacts and rejecting least-prospective rocks at the deposit scale.5.2. SIGNIFICANCE AND CONTRIBUTIONS TO THE FIELDThere is limited published information detailing geophysical inversion modelingefforts in Archean orogenic gold environments. The work presented in this thesisprovides a comprehensive case study focused on the application of inversion methods fororogenic gold exploration, and may act as a reference point for others embarking onusing inversion to explore in this deposit setting.As previously discussed, an understanding of physical properties lays thegroundwork for applying geophysics or geophysical inversion as exploration tools. Theextensive physical property work completed constituted a major component of this thesisand is an important contribution to geophysics-based exploration. A significant amount ofphysical property data was generated for Hislop deposit rocks, and physical propertyranges for typical host rock types and for prospective rocks were delineated. This datamay eventually be contributed to a regional or national physical property databases,199enhancing the sources on which to draw for geophysics-based exploration in similar areaswhere little sampling or physical property reconnaissance has been done.Synthetic modeling of a typical gold deposit provided insight into the features thatwill, and will not be imaged for a given survey design and mesh discretization. It alsoallowed application of various basic constraints to be tested to assess their influence onrecovered models. This compilation might provide some guidance for geophysicalsurvey, or inversion design, in a similar setting.Inversion of the range of geophysical data available over Hislop, at a range ofscales made for a unique case study with significant breadth. Querying combinedphysical property models was shown to be a valuable application of inversion results. Itwas demonstrated that the combination of magnetic susceptibility and density modelswere useful for distinguishing sedimentary and felsic rocks from Fe-rich mafic andultramafic rocks, and Fe-poor mafic and ultramafic rocks, and for outlining their 3Dsubsurface distributions at the regional scale. The queries used constitute importantmapping tools in areas of poor outcrop in this part of the Abitibi greenstone belt. Atsmaller scales, prospective areas can be distinguished by combining chargeability andsusceptibility results, as was indicated by correlation between known mineralization, andhigh chargeability-low susceptibility anomalies. Physical property studies (Chapter 2)indicated similarities in local and regional scale physical property ranges anddistributions. As such, these queries could be applied to other inversion results regionally.5.3. LIMITATIONS OF THE THESIS RESEARCHDue to their ease of collection, it was possible to amass a large number ofmagnetic susceptibility and density measurements for Hislop samples. An equivalentnumber of measurements for resistivity and chargeability were not generated, asequipment was not available to make the measurements in-house. Measurements had tobe completed at the physical properties laboratory at Zonge Engineering and Research200Organization, 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 athorough assessment of the effects on alteration on these two electrical properties, and inaddition, 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 andinversion work could both be expanded on. For example, synthetic modeling was onlycarried out at one representative scale of exploration, the deposit-scale, with datacollected only on a 50 m x 10 m grid. With anticipation of completing regional inversionsit would be beneficial to model the deposit at a more regional scale. Only selectvariations on the geological setting of the modeled gold deposit were considered duringsynthetic modeling studies, and constraints only demonstrated for a subset of thesescenarios. There are obviously many different scenarios that can be tested, but it wouldtake considerable time to assess them all. Similar expansions on work could be applied toinversion of actual data over Hislop. Different combinations of constraining informationcould 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 alsovary dramatically when geophysical data are scaled differently, and when errors arechanged. These parameters might also be investigated more extensively throughadditional inversions. With the array of possible modifications, it is feasible that there is abetter model to be generated in each case.An additional limitation of the research relates to the previous comments. One ofthe interesting challenges encountered in completing this project was dealing with therapid rate at which inversion concepts and methods are developing. At times, a series ofmodels would be completed only to discover that there was a newer version of theinversion code available! This is a relatively new field, and the Geophysical InversionFacility at UBC are at the forefront of it. The UBC-GIF has developed robust inversioncodes that are used worldwide, and the codes are constantly being updated to adapt to themodeling needs of exploration and environmental communities. This means that more201effective codes, or programs with increased functionality, are becoming available on aregular basis, and that the models presented herein might be improved on withapplication of newer software.5.4. RECOMMENDATIONS FOR CONTINUED WORKThere were several ideas proposed during the course of this research that were notfollowed up on. Some of the ideas worthy of further investigation are listed here, alongwith additional suggestions.More resistivity and chargeability data is needed to better define relationshipsbetween geology and physical properties. Since IP methods are so effective in delineatingsulfides at Hislop, and have been shown to be effective in detecting mineralization forother gold deposits, more chargeability data would be useful. It would be beneficial to doa more in depth analysis of relationships between chargeability to sulfides types, sulfidetextures and abundances, as well as attempt to define a relationship between gold andchargeability.As chargeability data was collected at multiple time windows during IP work bothin the field, and in the laboratory, there is potentially more information to be gained. Tocalculate chargeability for this thesis, the value representing the voltage decay over thesetime windows was chosen to be 80% of the sum of voltages over eight of the timewindows. This choice of representative value is somewhat arbitrary, and there exist otherstandard measurements in the industry. The consistency of measurement methods for asuite of data is of more importance than choice of calculation. By assessing the entiredecay curve, or looking at voltages from individual time windows, instead of calculatinga representative value, relationships between chargeability and mineralization notpreviously identified may be revealed.202It may be constructive to automate the synthetic modeling process. Constructingthe range of synthetic models and testing them was time-consuming, and only selectscenarios were represented. Such a program could automatically vary geometry andphysical property contrasts of a target feature for given survey parameters and inversioncell sizes, and assesses model difference values (Chapter 4) to determine conditionswhere the difference between true and recovered models are low. This may be aneffective way to know more accurately and efficiently when a feature is too small or toodeep, or has too low of a contrast from host rocks, to be imaged.From a data management standpoint, another program might be devised to helpmanipulate the typically large datasets to be used in inversions. Some basic unofficialprograms exist, but a formal one could be made. The program should be able to cut aspecific range of data from a dataset that covers a larger area, and decimate data to getspacing to correlate with inversion cell sizes, perhaps allowing more dense data at thecore and sparse data in outer regions. A formal program that reorganizes DC resistivityand chargeability into an inversion-friendly format would also save time.Regarding the inversion models, some may be rerun to test application of variouscombinations 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 addedto 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 physicalproperties, 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 becreated for use as a reference model for inversions, and for general comparison toinversion results and incorporation into common-earth models. The model was initiated,based on cross-sections drawn from select drillholes, however it was not completed. Theprocess of building a 3D geological model requires significant time, reasonable203experience in GIS modeling, and a thorough understanding of geology. There was simplynot enough information collected during this study to build anything but a very simplemodel that is extensively interpreted. There is, however, potential for a 3D model to bebuilt for the Hislop deposit in the future, as there is a wealth of information from themany drillholes that were logged in this area, and now there are geophysical modelswhich can help with geological interpretations at depth. The geological model must becompleted with contribution from geologists that are well-familiarized with the geologyand structure of the depositThe Hislop common-earth model can be further developed with the addition of a3D geological model, and with the contribution of other existing data. Data from a largescale 3D model of the area created in the Fracsis GIS program by GeoinformaticsExploration Inc., including fault and geological contact surfaces, can be converted toforms usable in Gocad. A large quantity of drilling information, along with gold assays,and geochemical information collected by numerous workers throughout Hislop’sexploration history can be incorporated into the model for the purposes of mapping andtarget generation. At the start of this project lithogeochemical data was obtained with theanticipation that there may be relationships existing between this data and physicalproperty data which would allow chemistry to be used to predict physical properties.Unfortunately, no statistically relevant trends emerged. Although geochemical data donot 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 earthmodels of Hislop. Anomalous abundances of elements reflecting carbonate, muscovite(or sericite), and albite-dominated alteration, such as C02,K, and Na, would act as aadditional exploration criteria for querying along with geophysical inversion results.5.5. FUTURE DIRECTIONS OF THE FIELD OF STUDYThe field of geophysical inversion-based exploration is young. The inversioncodes developed at UBC are constantly being updated and refined in order to allow more204flexibility with respect to incorporating geological information. They will continue todevelop 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 thatallows all prior geological and physical property knowledge to be input into inversions asconstraints. The influence of the input data on the model is determined by the user basedon the confidence the user has in the data.Additional programs to help input geological information into inversions, or makethe results consistent with expected geology, are in progress. Diagonal dips and structuraltrends outside of north, south, east, and west directions can now be input using codesbeing developed by Lelievre et al. (2008). This is would be of use for inversions in anygeologic setting, however could be especially useful in Archean greenstone terraneswhere 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 lowand high anomalies. This may not be considered representative of the true geological orphysical property situation. Phillips et al., (2007) initially introduced a method thatrestricts ranges of physical properties allowed to be taken up by model cells. This wouldbe a useful tool where geology is simple, with only a few rock types present, and specificphysical property ranges are expected. Lelievre et al. (2008) demonstrate how thisapplication can be used. This technique might be constructively applied to inversions ingreenstone belts where smoothing in inversion results can obscure important contactsbetween petrophysically distinct mafic and felsic units.The importance of physical property data collection is being increasinglyrecognized, especially in light of the need to use geophysics to explore for deeper mineraldeposits. Large scale, publicly accessible physical property databases will become morecommon in the future, allowing geoscientists to cull physical property information fromspecific geographic areas, geologic regimes, and deposit types, to fortify geophysicalwork. A large data collection effort initiated by the Ontario Geological Survey (2001) in205the central Abitibi greenstone belt was mentioned in Chapter 2. A national physicalproperty database is currently being compiled by the Geological Survey of Canada andMira Geoscience Ltd. (Parsons and McGaughey, 2007).The best inversion results are generated when geologists and geophysicistscollaborate on the problem. Geologists and geophysicists need to combine efforts toresearch or investigate physical properties in a given environment prior to inversion.Geologists can play a larger role in geophysical investigations, and will benefit theexploration effort by doing so. Geologists can provide insight when surveys are beingdesigned, and can aid the inversion process by contributing prior geologic informationincluding dominant structural fabrics, typical stratigraphic thicknesses, proportions andvolumes of rock types or alteration present, and shapes and sizes of typical orebodies.Significant geologic information can be incorporated into inversions, by directlymanipulating basic input parameters or with a complex reference model building programlike that of Williams (2008).Recent collaborations between geologists and geophysicists for the greaterunderstanding of a geological region took place during the Discover Abitibi Project.Greenstone architecture and mineral deposit settings were investigated indepth using acombination 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 ofcollaboration between the two disciplines, case studies need to be presented in moregeneral forums or as short courses that will attract members from both fields.206REFERENCESAyer, 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: DiscoverAbitibi Initiative, Ontario Geological Survey, Open File Report 6154, 146p.,3 sheets.Lelievre, P., Oldenburg, D., and Williams, N., 2008, Constraining geophysical inversionswith 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 presentationof geophysical data for geoscientific profiles in the Timmins—Kirkland Lake area:Discover Abitibi Initiative, Ontario Geological Survey, Open File Report 6189 , 28p.,15sheets.Ontario Geological Survey 2001, Physical rock property data from the Physical RockProperty 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 ofExploration ‘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 the3D integration of exploration data: KEGS Inversion Symposium, PDAC 2007, extendedabstract, 9p.207Reed, L. E., 2005, Gravity and magnetic three-dimensional (3D) modeling:DiscoverAbitibi Initiative, Ontario Geological Survey, Open File Report 6163,40P.,4 sheets.Reed, L. E., Snyder, D. B., and Salisbury, M. H., 2005, Two-dimensional(2D) reflectionseismic surveying in the Timmins-Kirkland Lake area, northern Ontario;acquisition,processing, interpretation: Discover Abitibi Initiative, Ontario GeologicalSurvey, OpenFile Report 6169, 68p.,10 plates.Williams, N.C., 2008, Geologically-constrained UBC—GIF gravityand magneticinversions with examples from the Agnew-Wiluna greenstone belt,Western Australia:Unpublished Ph.D. Thesis, The University of British Columbia, 479p.208APPENDIX 2A - LIST OF ABBREVIATIONSRock Type MineralsIF feisic intrusive ab albiteIFP feldspar-phyric rhyolite dike act actinoliteIQFP quartz-feldspar-phyric rhyolite dike al aluniteII intermediate dike an anataselix brecciated intermediate dike ank ankeriteIM mafic dike ap(hy) hydroxyl apatiteKMXmag brecciated K-fsp vein in magnetic mafic volcanic rock au augiteL lamprophyric dike bt biotiteML)( multi-lithic breccia cal calciteQMX brecciated quartz vein in mafic volcanic rock clz clinozoisiteQUX brecciated quartz vein in ultramafic volcanic rock dc clinochloreS syenite dike chi chloriteSeds sedimentary rocks dol dolomiteT volcanic tuff ep epidoteVM mafic volcanic rock Fe-cb Fe-carbonateVMX brecciated mafic volcanic rock Mg-cb Mg-carbonateVMmag magnetic mafic volcanic rock Fecb ankerite+dolomite+sideriteVMXmag brecciated magnetic mafic volcanic rock hem hematiteVMP pillowed mafic volcanic rock hbl hornblendeVMPX brecciated pillowed mafic volcanic rock ksp potassium feldsparVU ultramafic volcanic rock mag magnetiteVUX brecciated ultramafic volcanic rock mns magnesitemc(int) microcline (intermediate)mc(or) microcline(ordered)Alteration ms muscoviteB carbonate+muscovite alteration (bleached) ms(Mg) muscovite (magnesium)B+P carbonate+muscovite+albite alteration mus(tot) total muscoviteC chlorite or orthoclaseCB chlorite+carbonate+sericite par pargasiteCH chlorite+hematite pnt paragoniteCs chlorite+sericite per peridlaseF carbonate+fuchsite alteration ph phlogopiteFC Fe-carbonate alterationp1plagioclaseFC+H Fe-carbonate+hematite alteration py pyriteFC+H+S Fe-carbonate+hematite÷sericite alteration qtz quartzFC+Q Fe(Mg?)-carbonate+quartz alteration rut rutileFC+5 Fe(Mg?)-carbonate-’-muscovite/sericite alteration ser sericiteH hematite alteration sid sideriteS muscovite/sericite alteration sm smithsoniteS+Q sericite+quartz alteration sp sphaleriteT talc-chlorite metamorphic assemblage tc/tlc talcU generally unaltered wt witherite209APPENDIX 2B - HISLOP DRILLCORE LOGS, CROSS-SECTIONS, ANDOUTCROP MAPSLDO — late diorite/doleriteSSG - greywackeSLO — mudstone - siltstoneS00 — sediment, undividedIFD/lFO — felsic intrusive dykelfelsic intrusive undivided100 — intrusive, undividedI I ISO — syenite intrusive, undivided______VFO — felsic volcanic, rhyolite, rhyodaciteI I VUO — ultramafic volcanic, undivided>VMF — magnetic mafic volcanicI I VMO — mafic volcanic, basalt, andesiteFigure 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 forthis study, five geologic cross-sections compiled from drill logs (1-5), and 2 outcrops (Aand B)./1 Modified from Power et al., 2004NSLO500m 4•H9711I H9Th7H2AnuwFauftLDOH9606oP7V’H9605N1i97ó8 (N)4EXT20VMO\/VMOoiGK 204210Hislop drill log legend.AlterationIJ Weak to moderate Fe-cb+msL_iStrong Fe-cb+msFe-cb + abTic-chl metamorphic assemblageMg-Cb (magnesite) + ms (fuchsite)ChloriteESericiteFe-cb + ab(intermediate dikes)Ms/ser (syenite and rhyolite dikes)Fe-cb (syenite and rhyolite dikes)Pink Fe-rich dol veinsEpidote veinsHematite - pervasiveHematite along fracturesMagnetite?nIIHHIIIILithologyMulti-lithic volcanic brecciaLamprophyric dikelritermediate-mafic dikePorphyritic rhyolite dikeSyenite intrusiveMafic volcanic rockUltramafic volcanic rockUExample of column layout:750B750_r775 7758001 280OColumn 1: LithologyColumn 2: AlterationColumn 3: Magnetic Susceptibility (x103 SI Units)Column4*:________Au grades between 0.15- 1 ppm______Au grades between 1 - 5 ppmAu grades > 5 ppm*notall intervals sampledFaultAbbreviations: see first page of appendices211Loo for H960125 2550 0075 -75100 -100125 125150 150170 ç 175200 200225 225250 250275- 275300 -- - 300325_— 325350 -350____/--375 375400 - 400425 425455 - 450475 : 475500 - 500525 525550 - 550575” -575- 4600__________- -605625 6250Lo!p[ H96020I 1 -2oj.JL-2:75-100125-150175r -200-225-250275-, 300[I325H350400425450b47513500700725750800>-i0 32—650675700725750775800212LogforH96O60255575100125150175250225280275350/626375//650400425700450725-475/750500 1775525 1850650157510Lo for H9605254V2550 50/A100____ 100125 - 125150- -150175 -/ 175200 200225 225250 250275 275305 3000 15302550•75-100 -125150175200225250275300325350400:425450 -525575- 600- 625:650675700725750775:5550 149213oLog for H9707 Lo9 for H970825 —25 25 -2550L5050 .% ‘- 075 -75 75-—) 75100 —100100 100125 —125125 125150 —150150 150175_=175170 175r200 200200 - 200225 —225225 225250 — 250250;250275 -- 275— 275 ‘- 275300 300300)300325--325 -— 325 325350 :350350 .- 350375 -375S400 - - 400 -400 400425 — 425425=425625 - 625F_:::::>550 50= 575 75600 —600600-214250275300325350375400-for EXT 280250-275-3003253754000 49 -0-25-50-75100125150175200225250275300325350375400425450475500525550Log for H9711/0 169-0:25•5075100125150-175:225:250-275300325300375-400-425450-475525550215Legend applies to all following cross-sectionsMulti-lithic Volcanic BrecciaLamprophyric DikeIntermediate DikePorphyritic Rhyolite DikeSyenite IntrusiveJFe-poor Mafic Volcanic RockFe-rich Mafic Volcanic RockUltramafic Volcanic Rock. Fault.. —. Drill trace216k)Cross-section 2.s--Ir100 mRl.Cross-section 3. DDH Ext 280, GK 280, and H96052184.L’Jc. KI5OmjcCross-SEsw-I200 mRL100 mRLDDH H9708 DDH H9707/-200 mRil.1øJCOutcrop map A.47to643 (4)167 to347 (1)35to167 (12)• 4to 35 (7)• Oto 4(30)Magnetic susceptibilityJd3-3-Skintt3 —Carbonate altered maficvolcanic rockSyeniteDarker colors represent outcrop+Outcrop map B.•OverburdenUnaltered variolitic mafic volcanic•107b0173 (7)5Oto 107 (24)Bleached and oxidized variolitic mafic volcanicot 50 (14)• 8to 20 (16)• Oto 8 (33)Magneticsusceptibility(xl 0-3SI units)221APPENDIX 2C - DETAILEDAND EXPANDED METHODSXRD Analysis - Reitveld analysisThe standardless Rietveld refinementmethod was used to determine mineralabundances for Hislop samples.Samples were prepared andrun, and data was analyzed,by Elizabetta Pani at the Universityof British Columbia. X-ray diffraction(XRD)analyses are first run on powderedbulk rock samples. The samplemust first be groundsuch that particle sizes are <10pmto avoid inaccurate diffractionpeak intensities andpreferred orientation of grains.Samples are ground in ethanolusing a McCroneMicronising Mill with corundumelements. The sample is placedinto a back-loadingmount. The top of the mount isfit with a textured glass to minimizepreferred orientationon the surface of the sample, andis removed before analysis.A modified razor blade maybe used additionally to create texture inthe top of the sample. Standard X-raydiffractionpatterns are collected for samplesusing a Siemens D5000 diffractometer.X-raydiffraction data are collected inincrements of 0.04°, from 3°to70029. The counting timeis 2 seconds/step, and CuKa radiationis used. The diffractometer usedincludes adiffracted-beam monochromator,10divergence and anti-scatter slits,a 0.6 mm receivingslit, and an incident-beam Sollerslit which was removed.A long-fine-focus Cu X-raytube is used and operatedat 40kV and 4OmA, with a take-offangle of 6°. The mineralphases are determined using conventionalsearch-match procedures.The XRD data are analyzedby Rietveld refinement using the programTopas 2.0(Bruker AXS 2000). Forthis method, information regardingthe crystal structure ofalldetected phases is usedto calculate a diffraction pattern foreach phase present. Thesepatterns are summed and thenfitted to the collected diffractionpattern using a leastsquares refinement. Numerous parametersare considered in the refinementincluding aseries of global parameters (e.g.background, radiation wavelength,correction for themonochromator crystal),and mineral phase-dependantparameters (e.g. atomiccoordinates, size and shape ofthe unit cell, site-occupancy). UsingRietveld methods, therelative masses for each phasecan be calculated by consideringthe scaling factor222determined when observed and calculated data were being fit, the number of formulaunits per unit cell, the mass of the formula unit, and the volume of the unit cell. Detailedmethods are found in Raudsepp and Pani (2003).Magnetic susceptibility correctionsCore diameter correctionsAn Exploranium KT-9 Kappameter was used to collect magnetic susceptibilitydata from 3.6 cm diameter drill core. Since the meter can only be set to takemeasurements from drill core with a diameter that is a whole number, the KT-9 was set totake readings for 4 cm drill core, the closest whole number to the diameter of the drillcore being used. Through some experimentation, it was determined how to correct themagnetic 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 datapoints (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 magneticsusceptibility values and changes in diameter. First, diameter was plotted againstmagnetic susceptibility for select sample points to determine a relationship (for anexample of this see Fig. 2C.2; for the full experimental dataset, see the spreadsheetlabeled Appendix 2C on accompanying CD). It was noted that this relationship changeswith 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 thesusceptibility at a given point. Coefficients a and m calculated from select sample pointson four different samples were plotted against susceptibility measured on the 4 cmdiameter setting (Figs. 2C.3 and 2C.4). From this it was determined that there were linearrelationships between m and magnetic susceptibility and a and magnetic susceptibility. Itwas concluded that readings at 4 cm can be plugged into the equations m = 3.2295x- 33.133and a = -0.0004x - 0.6506 (where x = magnetic susceptibility at 4 cm) to get m and a, andthen m, a and the diameter we want to correct to (3.6 cm) can be input back into the223equation y = mx, where x is the new magnetic susceptibility (at 3.6 cm). This correctionwas applied to all of the whole core data collected in the field.Split core correctionsSome of the drill core was split lengthwise for sampling purposes, and only onehalf of these particular intervals was available to test. The KT-9 meter setting was kept on4 cm diameter for these intervals. It was noticed that there existed discrepancies betweenthe magnetic susceptibilities of whole and split core of similar rock types and it wasnecessary to correct for this. Similar experiments to the core diameter tests werecompleted, and susceptibility readings taken along whole core at particular data points (atvarious diameters), after which the core was split lengthwise using a rock saw andreadings taken along the split pieces at the same designated data points. The datacollected at each point was compared between the different diameters (this can be seen inspreadsheet 4 in Appendix 2C on the accompanying CD) with the anticipation that thechange in values between whole and split core was simply proportional. For each samplethe average ratio between whole core and split core values is consistent between variousdiameter settings. However, from one sample to the next (samples range from felsic rockswhich have the lowest magnetic susceptibilities, to ultramafic rocks with higher magneticsusceptibilities) the ratio changes slightly (ranging from about 0.83 to 0.89). Lowmagnetic susceptibility samples do not consistently change by a different ratio than dohigh magnetic susceptibility samples when split. From spreadsheet 4, an attempt wasmade to determine if there was a relationship between susceptibility and this ratio. Thereappears to be a weak trend that indicates that a higher ratio can be used to correct for lowmagnetic susceptibility samples while a lower ratio can be used to correct for highermagnetic susceptibility samples (Fig. 2C.5). But the trend is not good, and appears tobreak 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 correctionfactor of 1.15 was applied to split core samples to get approximate whole coresusceptibilities.224Figure 2C. 1. Data pointsalong a whole piece of drillcore. Magnetic susceptibilityreadings were taken ateach of 16 data points at variousdiameter settings (2.54,3, 4, and5 cm).225H-97-07 456m350300 - ..— -0.7249 -________________E— -0.7101y — 452.02x —Power(d5)100 ---- — —Power(d7)y = 451 .85x°7298—Power (d9)50 •-.00 1 2 3 4 5 6diameterFigure 2C.2. Readings for five sample points (data points dl, d3, d5, d7, d9) for sampleH9707-456 with meter set at 2.52, 3, 4, and 5 cm diameters. For a particular sample, thereis not a distinct relationship between susceptibility and diameter, i.e., there is not simplyone equation. The equation (y=mxj changes with variations in magnetic susceptibilitybetween the different points tested.226Figure 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 differentsamples are obvious as the four clusters of data occurring within narrow susceptibilityranges.E • Series II — Linear (Series I)mag sus2270Figure 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 differentsamples are obvious as the four clusters of data occurring within narrow susceptibilityranges.40100 200 300 400-0.1y= -0.0004x- 0.6506• Seriesi— Linear (Series 1)-0.9mag sus2280.8700.86. 0.8500.84sus (of split core)Figure 2C.5. Possible relationship between the susceptibility of split core and the ratiobetween split core and whole core values. Relationship seems to break down at lowsusceptibilities.0.900.890.881‘11‘1Iy = -9E-05x + 0.8583• Seriesi—Linear0.830.820.810 100 200 300 400 500229Comparison of hydrostatic and geometric calculations of densityDensity measurements for Hislop samples were made using the hydrostaticmethod described in Chapter 2, section 4.2.2. To test if this method was generatingreliable density measurements, a geometric method was applied to select samples forcomparison. Density values are determined by dividing the mass of the sample by itsmeasured volume. The volume of the sample was determined by measuring the lengthand diameter of a piece of whole drill core using a Mitutoyo caliper (l2in!300mm). Thedrill core was first cut as evenly as possible on each end. Multiple measurements weremade of the diameter and length to account for slight irregularities, and the average valuewas used to calculate the samples volume. Figure 2C.6 shows a plot of densitymeasurements made by the geometric method versus density measurements made usingthe hydrostatic method. A strong correlation indicates values are accurate.Figure 2C.6. Density data calculated using the hydrostatic method versus the geometricmethod.U0)(CCC)•0UE0C)02.952.92.852.82.752.75 2.80 2.85 2.90 2.95Hydrostatic density (glcm3)230APPENDIX 2D - X-RAY DIFFRACTION ANALYSES_____dol ep Fecb hem hbl ksp maci mns20.718.70.514.922.42.14.028.018.71.146.833.21.00.90.43.75.13.2Sample Rock Altn act al an ank an(hv’ au bt cal clz jjflH9604-57 VMP U 37.1 0.8 14.7 17.34THD8 VMP U27.1 30.7 10.7 22.04THDIO4 VMP U 38.4 4.223.4 15.5 12.2H9601-439 VM B 44.0 5.2H9601-122 VM U 49.6 12.5H9604-28 VM U 27.6 1.0 0.5 11.1 7.2 30.6H9604-122 VM B 3.1 0.5 14.9H9605-111 VMmag U 8.0 2.0H9707-373 VMmag U 34.9 29.6 8.3 1.9 0.9 10.84THDII7 VMmaq U41.7 0.6 2.1 7.1 7.1 2.14THDIOO VMmag U36.4 15.6 7.7 13.84THDII6 VMmag B34.3 16.1 16.14THDII4 VMmag U16.2 19.1 21.7 15.63THDI5 VMmag U35.6 13.8 21.4 6.8 3.2 7.9H9711-281 VU B 2.6 16.7 22.9 22.9H9606-179 VU F 2.0 35.0 19.0 19.0H9601-200 VU T 4.0 3.3 11.7 42.1 6.5 6.5H9601-252 VU U 36.6 23.0 23.0H9604-379 VU T 10.4 49.7 5.2 5.2H9601-396.5VU B 1.2 5.4 20.6 20.63THDI VU B17.3 4.0 6.1 20.5 20.5H9601-410 S U 59.3 0.8 2.2 2.2H9605-176 S FC 63.0 0.44THDII5 S FC+S 37.2 0.5 0.5H9601-417 S BH9605-210 S FC+S 59.4 8.1 0.6 8.1 16.23THD6 S U48.5 1.0 1.0H9601-406 S S 58.6 5.2 5.2H9601-298 IQFP U 64.8 0.4 1.2 1.2H9604-214 IFP 5 54.0 3.2 2.2 3.2 6.4H9601-322 IQFP U 62.8 5.7 5.7H9601-302 IQFP U 59.8 0.8 0.8H9707-137 IM U 32.2 1.7 32.8 11.6 11.6H9606-66 II F 2.5 21.1 44.2H9606-230 II FC 33.0 20.2 24.3H9606-173 II B+P 45.8 14.2 14.2H9604-444 L U 14.3 20.7 25.4 25.66.08.22.51.8 4.62.944.20.6 31.23.6 3.11.50.6Sample RockH9604-57 VMP4THD8 VMP4THDIO4 VMPH9601-439 VM BH9601-122 VM UH9604-28 VM UH9604-122 VM BH9605-111 VMmag UH9707-373 VMmag U4THDII7 VMmag U4THDIOO VMmag U4THDII6 VMmag B4THDII4 VMmag U3THDI5 VMmag UH9711-281 VU BH9606-179 VU FH9601 -200 VU TH9601-252 VU UH9604-379 VU TH9601 -396.5 VU B3THDI VUH9601-410 SH9605-176 S4THDII5 SH9601-417 SH9605-210 S3THD6 SH9601-406 SH9601-298 IQFPH9604-214 IFPH9601 -322 IQFP UH9601-302 IQFPH9707-137 IMH9606-66 IIH9606-230 IIH9606-173 IIH9604-444 LAltnUUUUUUFFCB+P9.94.91.014.43.32.73.1 22.84.832.918.4 17.323.7 38.035.06.71.819.1 9.333.51.432.917.338.013.1 48.08.311.123.315.45.828.617.715.36.2qtz23.63.82.00.9 11.1 1.0 7.211.2 1.319.434.0 1.7 17.00.5 19.7 10.82.4 34.01.67.9 27.315.10.530.529.217.325.627.11.4 1.21.3 3.50.90.643.0 1.1 7.92.9 1.21.21.7 11.20.3 26.20.3 25.01.0 30.40.5 30.61.1 6.521.09.34.6 7.42.9 15.1_______par pgt per ph(1M) p1py rut sid sm sp tc(1A) wI1.3 1.8mc(int)mc(ord) ms(2M1) ms(1M,Mg) mus(tot) or3.5 3.54.7 4.74.26.71.82.60.93.1 33.524.022.84.8 3.1BUFCFC+SBFC÷SUSUS0.3 1.112.68.425.617.619.538.36.78.42.312.76.90.45.80.423.14.10.31.614.221.4APPENDIX 2E - PHYSICAL PROPERTIES OF HISLOP DEPOSITROCKSSample No. HolelD From To Rock type Gr. Altn notesMS Den Chrg Res Porsize(x104SI) (glcm3)(ms) (Ohm-rn) (%)FC 2.793THDIO hndsmp IM equigranular 4.153THDII hndsmp VU sheared24.803THD12 hndsmp VUX B “carbonate breccia”0.963THD13 hndsmp Vmmag? U17.203THD14 hndsmp VMmag? U50.403THD15 hndsmp VMmag? U88.803THDI6 hndsmp VMmag? U82.303THDI7 hndsmp VMmag? B+P north of main mine syenite12.403THDI8 hndsmp S FC+S0.273THDI9 hndsmp Vmmag U94.703THDIA hndsmp VU B breccia; Fe-carbonateand qtz matrix, 11.20disseminated py3THD2 hndsmp S FC+S0.163THD2O hndsmp Vmmag? U41.603THD21 hndsmp VMP U pillow tops to NE (060),10cm to 0.5 m, 15.60some disseminated py3THD21 hndsmp VMP disseminated py15.603THD3 hndsmp IM FC equigranular; disseminatedpy and cpy 0.863THD4 hndsmp L massive12.403THD5 hndsmp IM FC syenite “veins”, cut by Fe-carbonate veins 6.212.733THD6 hndsmp S U 0.17 2.643THD7 hndsmp VMmag? U syenite “veins”; disseminatedpy and cpy 203.00 3.073THD9 hndsmp IF FC carb and disseminatedpy/cpy fills fractures 0.41 2.674THDIO hndsmp VM fU 0.97 2.924THD100 hndsmp VMmag U1.14 3.004THD1OI hndsmp VMP f U pillows 20cm to Im; chlorite selvedges;tops 0.55 2.94to 0754THDIO2 hndsmp3.064THD1O3 hndsmp VM2.894THD1O4 hndsmp VMP U epidote in selvedges2.984THD1O5 hndsmp VMP f U epidote in selvedges2.974THDIO6 hndsmp VU F Royal Oak pit?0.62 2.954THDIO7 hndsmp VM B Royal Oak pit? 0.372.884THDIO8 hndsmp VMP U Royal Oak pit?0.52 2.823.013.012.942.98 20.87 2548.92.872.79 11.17 220522.78 9.83 4757.62.67 10.57 5852.62.892.7713.40 6507.115.17 5034.16.63 10687 0.288.50 33638 0.23IM UU33.0069.5012.80alteration along fracturessize(x1O3SI)hndsmp Seds U Porcupine sedimentsSample No. HolelD From To Rock type Gr. Altn notesMS Chrg Res Por(ms) (Ohm-rn) (%)9.73 3478.5 0.54Den(cm)4THDIO90.21 2.734THD11O hndsmp Seds U Porcupine sediments0.28 2.764THD111 hndsmp IM c U latedike?35.10 3.06 36.60 164724THD112 hndsmp IM c U latedike?26.20 3.044THD1I3 hndsmp VMX U0.43 2.844THDI14 hndsmp Vmmag U74.10 2.92 37.73 58754 0.134THD115 hndsmp S c FC+S0.20 2.69 6.27 3343.84THDII6 hndsmp VMMag B0.15 2.81 7.57 2130.94THD1I7 hndsmp VMMag U54.30 2.78 28.18 30754THD2 hndsmp VMP c U0.68 2.914THD3 hndsmp VU B similar to rock from Royal Oak pit; Fe-carb1.44 2.88alteration along fractures4THD4 hndsmp II U0.24 2.804THD5 hndsmp VM U disseminatedpy 0.35 2.764THD6 hndsmp VM B similar to rock from Royal Oak pit; Fe-carb0.38 2.814THD7 hndsmp VMP U variolitic, flow-banded pillow basalt;0.72 2.90coalescing varioles4THD8 hndsmp VMP U pillows 30 cm; epidote in selvedges0.88 2.99 3.38 42488 0.574THD9 hndsmp VM U epidote in veins 32.40 2.92Ext280-258 EXT 280 258.25 258.45 VU f T sheared, fractured; qtzlFe-carb vein21.06 2.84Ext280-274 EXT 280 274.25 274.5 VU f T rare veins0.55 2.86Ext280-275 EXT 280 275 275.2 VU f T in-situ fragmented and sheared 27.762.86 7.80 283.3 1.49Ext280-287 EXT 280 287.45 287.65 VU f T massive to sheared, abundant qtz 26.90 2.87amygdules, <1mmExt280-306 EXT 280 306.65 306.9 IFP f S 20% fsp phenocrysts average <1 to 1mm 0.16 2.66Ext280-308 EXT28O 308 308.2 IFP f S phenocrystsdiffuse0.12 2.65Ext280-319 EXT 280 319.3 319.55 IFP f U unaltered phenocrysts clearly visible 0.16 2.64Ext280-341 EXT28O 341.2 341.45 IFP f B 0.16 2.61Ext280-353 EXT 280 353.3 353.55 IFP f B 0.132.66Ext280-358 EXT 280 358.2 358.4 II f U sharp contacts0.39 2.86Ext280-365 EXT 280 365.2 365.4 VMmag m B1.28 2.67Ext280-365.5 EXT 280 365.45 365.7 IFP f S 0.12 2.88Ext280-372 EXT 280 372.3 372.55 VMmag m B0.55 2.89Ext280-375 EXT 280 375.65 375.85 II f H 3.29 2.84L’JExt280-388 EXT 280 387.75 388Ext280-401 EXT 280 401.45 401.7Ext280-402 EXT 280 402.1 402.3 VUX2mm to 4cmExt280-406 EXT 280 406 406.25 II fExt280-407 EXT 280 407.4 407.6 II fExt280-410 EXT 280 410.45 410.7 II fExt280-411 EXT28O 411.35 411.55 II fGK204-101 GK204 101.50 101.8 S cGK204-109 GK204 108.97 109.27 S cGK204-116 GK204 116.13 116.43 S cGK204-128 GK204 127.71 128.02 VMmag fGK204-129 GK204 129.08 129.39 MLXGK204-168 GK204 168.71 168.86 VU fGK204-25 GK204 24.99 25.3 VMmag mGK204-60 GK204 59.74 60.05 II mGK204-74 GK204 73.80 74.1 IF cGK204-76 GK204 75.90 76.2 IFcFC+SH hornblend phenocrysts, 1-3mm, 5-10%SFC+SCFC+SFC+SB brecciaB+P mylonite?TU equigranularUSFC+Q0.14 2.683.30 2.790.31 2.760.13 2.690.07 2.680.14 2.700.18 2.740.41 2.860.53 2.9414.12 2.81182.35 3.0067.50 2.790.30 2.670.24 2.67GK204-84 GK204 84.58 84.89 VMmag mGK204-93 GK204 92.96 93.27 vMmag mGK204-97 GK204 97.08 97.38 ScH9601-107 H9601 106.85 107.05 II mH9601-122 H9601 121.95 122.2 VM mH9601-126 H9601 126.5 126.8 VU mH9601-132 H9601 132.4 132.6 VU mH9601-133 H9601 133.55 133.75 IM mH9601-135 H9601 135.6 135.8 IM mH9601-150 H9601 150.4 150.6 vu mH9601-156 H9601 156.4 156.6 vu mH9601-164 H9601 164 164.2 vu mH9601-172 H9601 171.8 172 IM cH9601-186 H9601 186.55 186.75 IQFP fH9601-200 H9601 200.65 200.85 VU mH9601-210 H9601 210.45 210.6 IQFP fCBFC+S sulfidesUUu spinifexTUUT spinifexTUFCu crowded porphyryT spinifex67.20 2.832.06 2.850.14 2.840.70 2.9418.70 2.86 158.17 5400 0.630.71 2.831.56 2.813.20 2.7953.60 2.7521.60 2.8552.60 2.8512.50 2.8711.20 2.772.45 2.67 20.40 59694.42 2.82 20.40 1422.4 1.13Sample No. HolelD From To Rock type Gr. AItn notesMS Den Chrg Res Porsize(x103SI) (qlcm3)(ms) (Ohm-m) (%)VU f T sheared; qtz and Fe-carb veins2.77 2.84f H0.49 2.85T sheared; qtz and hem(?) altered fragments, 1.18 2.85B crowded porphyry 0.18 2.75H9601-229 H9601 228.8 229.05IQFPH9601-24 H9601 24.7 24.9VUH9601-241 H9601 241.25241.45 IQFPH9601-252 H9601252.6 252.8 VUH9601-258 H9601 258.4 258.6MLXH9601-266 H9601 266.05 266.25IQFPH9601-286 H9601 286.4 286.6QUXH9601-294 H9601294.6 294.85 VUXH9601-298 H9601 298.2298.4 IQFPH9601-302 H9601 302.1 302.2IQFPH9601-308 H9601 308.7308.9 QUXH9601-314 H9601 314.6 314.8VUH9601-32 H9601 32.7532.95 VUH9601-320 H9601 320.6320.8 IQFPH9601-322 H9601 322 322.2IQFPH9601-324 H9601 323.9324.1 VUXequigranularcrowded porphyrym U equigranularB chaotic, multi-lithicbrecciaf FC weakly qtz andfsp porphyriticB ultramafic breccia,quartz clasts?f T brecciated,Fe-carb matrixf U1.23 ppm Au from coreboxqtz fragment breccia;2.69 ppm Au5.16 ppm Au from core box3.15 ppm Au from core boxqtz veins0.57 2.890.16 2.702.24 2.84 5.25 1062.50.7216.30 2.800.28 2.730.65 2.854.71 2.860.06 2.64 2.23 78560.05 2.63 2.53 89761.93 2.8513.76 2.870.61 2.850.06 2.64 2.90 95800.07 2.65 4.20 23970H9601-331 H9601 331.4331.6 VUH9601-338 H9601338.1 338.3 IIH9601-361 H9601361 361.2 VUH9601-371 H9601 371.2371.5 VUH9601-383 H9601383 383.15 IIH9601-393 H9601 392.85393.05 VU fH9601-396 H9601 395.95396.15 VM fH9601-396.5 H9601 396.6396.8 VU fH9601-405 H9601 405 405.2SH9601-406 H9601 405.9406.1 SH9601-410 H9601 410.4 410.6SH9601-417 H9601 417.3 417.5SH9601-419 H9601 419 419.2VMmag fH9601-422 H9601 422.3422.5 VMmag fH9601-43 H9601 43.3 43.55VUXH9601-436 H9601 435.9 436.1VMmag fH9601-439 H9601 439.1 439.4VM fm T equigranularf FCm Tm U spinifexf FC hbl andfsp phenocrystsTB+P fracturedB fracturedFC-S 2.13 ppm Au from coreboxFC-S 5.89 ppm Au from coreboxUBB+PBBUB strong mag sus contrast betweenbleached0.48 2.790.26 2.6721.41 2.870.74 2.87 4.33 698.390.51107.00 2.860.44 2.820.28 2.770.78 2.922.31 2.86 27.40 7872.40.18 2.730.16 2.70 25.80 9400.20.12 2.74 15.67 5703.60.46 2.822.12 2.930.41 2.9160.70 2.86?? 33.23 49812Sample No. HolelD FromTo Rock type Gr. AItnnotesMS Den Chrg ResPorsize(x103SI) (gicm3) (ma) (Ohm-rn) (%)f B crowded porphyry0.20 2.73m Uf Uf UBf Tm Uf Uf UT sheared, brecciated to massive;Fe-carb and 11.07 2.84and unbleached areasH9601 -474 H9601H9601-491 H9601H9601-496 H9601H9601 -507 H9601H9601-516 H9601H9601-529 H9601H9601-541 H9601H9601-55 H9601H9601 -551 H9601H9601 -571 H9601H9601 -581 H9601H9601 -600 H9601H9601 -614 H9601H9601-79 H9601H9601-95 H9601H9602-103 9602H9602-119 9602H9602-141 9602H9602-143 9602H9602-162 9602H9602-188 9602H9602-203 9602H9602-217 9602H9602-234 9602H9602-242 9602H9602-247 9602H9602-254 9602H9602-272 9602H9602-292 9602H9602-295 9602H9602-303 9602H9602-320 9602H9602-321 9602474.35 474.55 VMmag f491.45 491.65 VMmag f B496.3 496.5 VMXmag m B+P507.4 507.6 VMPX m B+P516.6 516.8 VMmag m B529 529.2 VMmag m U541.6 541.8 VMmag m B54.95 55.15 VMP551.55 551.7 VM614 614.2 II f U78.75 79 VMP f U95.4 95.6 VMP f U103.05 103.25 VMP f U119.7 119.9 VM m U141.15 141.35 VMmag m FC sulfides143.2 143.4 VMmag m U161.8 162 IM c U188.6 188.9 VMP f U203.25 203.45 VMmag m U217.2 217.45 VU f T234.6 234.8 VMmag f U241.95 242.2 IFP f S247.75 247.95 vMxmag f B254.2 254.4 S c C+H sulfides271.85 272.05 VU T292.15 292.4 VU295.7 295.9 vu303.25 303.45 vu320.1 320.35 vu321.7 321.9 IM f C0.56 2.850.56 2.860.51 2.810.55 2.831.54 2.790.48 2.800.66 2.770.76 2.881.06 2.7513.90 2.930.41 2.72 15.40 161600.57 2.790.72 2.831.74 2.963.95 2.9027.80 2.8683.70 2.8437.88 2.9723.20 2.940.65 2.9325.80 2.9752.10 2.8514.59 2.71132.94 2.870.16 2.7152.70 2.871.01 2.940.52 2.800.73 2.970.54 2.980.54 2.81Sample No. HolelD From To Rock type Gr. AItn notes MS Den ChrgRes Porsize(x103SI) (glcm3)(ms) (Ohm-rn)(%)B 0.51 2.82571.2 571.4 VMXmag581.25 581.55 II599.9 600.1 vuf U mafic phenosm B+P pink rhodachrosite or dolomite carbonateveining?Bf Um TL’Jm TC Tf TTsulfidesspinifexH9602-326 9602H9602-343 9602H9602-364 9602H9602-450 9602H9602-47 9602H9602-476 9602H9602-479 9602H9602-498 9602H9602-56 9602H9602-73 9602H9604-105 H9604H9604-108 H9604H9604-113 H9604H9604-117 H9604H9604-122 H9604H9604-126 H9604H9604-14 H9604H9604-150 H9604H9604-173 H9604H9604-182 H9604H9604-190 H9604H9604-201 H9604H9604-204 H9604H9604-205 H9604H9604-214 H9604H9604-229 H9604H9604-232 H9604H9604-260 H9604H9604-279 H9604H9604-28 H9604H9604-292 H9604H9604-303 H9604H9604-312 H9604H9604-327 H9604326.4 326.6 VU343.6 343.8 IM363.8 364 vu450.25 450.45 VU47.1 47.35 VM476.35 476.55 IFP479.45 479.65 vu498.4 498.6 VMP56 56.25 VM73.75 73.95 VMP104.95 105.15 VM108.6 108.8 VM113.15 113.3 VM117.6 117.8 VM122.7 122.9 vM126.3 126.45 VM14.4 14.6 VM150.6 150.85 VM173.15 173.25 VU181.8 182 VM189.9 190.1 VM201.05 201.25 vu204.2 204.35 IFP205.1 205.3 IFP214.2 214.4 IFP229.4 229.65 IFP232.45 232.65 IFP260.7 260.9 IFP279.3 279.5 vu28.25 28.45 VM291.85 292.05 IFP303.5 303.7 IFP312.3 312.5 vu327 327.2 SU hbl phenocrystsm TTm LI epidotef BC Tf uuf Uf BBf B varioles?f B varioles?m Uf Bf Ff Uf Bf T strongly shearedf Sf Uf Sf Sf Sf Uf Tf Lif Uf S+QC U85.90 2.7981.10 2.7680.10 2.806.99 2.870.08 2.650.52 2.770.55 2.871.63 2.9269.90 2.920.44 2.830.51 2.843.67 2.826.10 2.950.89 2.9539.40 2.870.60 2.880.49 2.880.57 2.800.60 2.8818.35 2.870.08 2.570.130.09 2.640.140.18 2.6771.80 2.826.54 2.880.21 2.712.18 2.61109.00 2.870.55 2.654.30 405.27N/A >70000 0.25Sample No. HolelD From To Rock type Gr.Altn notesMS Den Chrg Res Porsize(x1O3SI) (glcm3)(ms) (Ohm-rn)(%)f C0.55 2.79fffBBsulfide-filled amygdules?varioles?m icrofragmental2.97 3994.3 0.40fragmental, angular fragments0.67 2.93 5.73 18578 0.632.07 541.24 1.5410.70 113570.13 2.68 10.70 14759sheareddark black mineral (?) fills amygdulesm T mottled texture, patches of black mineralsGoSample No. HolelD From To Rock type Gr.Altn notes MS Den ChrgRes Porsize(x104SI) (glcm3) (ms)(Ohm-rn) (%)c U 0.75 2.84H9604-339 H9604 338.9 339.1 IFH9604-357 H9604 357.1 357.3 VU f TH9604-379 H9604 379 379.25 VU f TH9604-405 H9604 405.7 405.9 II fUH9604-407 H9604 407.65 407.85 LUH9604-414 H9604 414.45 414.65 II m UH9604-425 H9604 425.5 425.7 S c FCH9604-428 H9604 428.3 428.5 S c UH9604-433 H9604 432.95 433.15 S c SH9604-442 H9604 441.85 442.05 S cSH9604-444 H9604 444.3 444.5 L UH9604-447 H9604 447 447.25 S c FCH9604-475 H9604 475.65 475.85 S cSH9604-504 H9604 504.5 504.7 S c S+QH9604-512 H9604 512 512.2 II FCH9604-517 H9604 517.55 517.75 VMmag f BH9604-532 H9604 531.8 532 S cSH9604-545 H9604 545.5 545.7 VMmag f UH9604-555 H9604 555.05 555.25 VMmag f UH9604-57 H9604 57.7 57.9 VMP f UH9604-574 H9604 574.2 574.4 II mUH9604-585 H9604 585.45 585.65 VMmag f BH9604-595 H9604 595.6 595.8 VMmagf BH9604-603 H9604 603 603.25 VMmag f BH9604-609 H9604 609.6 609.8 VMmag f B+PH9604-615 H9604 615.45 615.65 VMmag f HH9604-625 H9604 625.25 625.45 VMmag mUH9604-635 H9604 635.5 635.75 VMmag mUH9604-673 H9604 672.75 673 VMmag m UH9604-702 H9604 702.7 702.95 VMmag fUH9604-716 H9604 716.7 716.9 VMmag f C+BH9604-719 H9604 719 719.2 VMmag f BH9604-727 H9604 727.6 727.8 IFP fH9604-738 H9604 738.6 738.8 II f FCsome brecciation 84.40 2.833.37 197.89 2.09some brecciation 41.20 2.744.18 111.16 3.09hbl and fsp phenocrysts 69.41 2.8149.30 2.8767.65 2.950.25 2.670.20 2.670.20 2.700.15 2.680.93 2.69 15.58 196640.26 2.59intensely altered 0.132.670.18 2.71 15.67 7245.21.94 2.6998.82 2.890.42 2.69106.00 2.78mafic phenocrysts 140.00amygs increase at pillow margin0.59 2.79 15.98 155090 0.26135.29 2.89 115.57 2313.6148.24 2.850.72 1.9419.41 2.92 12.43 143463.91 2.825.24 2.7921.50 2.85 141.53 49590 0.22soft green mineral filling fractures, with pink 0.783.03carbonate corepink mineral in vein (dolomite?) 0.6146.20 2.770.58 2.880.48 2.83FC sparce phenocrysts 1.08 2.78 46.2010657L’)0.47 2.77H9605-142 H9605 142.5 142.7 VMmagc B+PH9605-144 H9605 144.7 144.9 VMXmagfH9605-145H 9605-150H 9605-157H 9605-162H9605 145.35 145.55 SH9605 150.5 150.75 SH9605 157.3 157.5 SH9605 161.8 162 II0.24 2.730.14 2.640.15 2.65 14.43 2630.60.51 2.83 10.13 4354.6Sample No. HolelD From ToRock type Gr. AItn notesMS Den Chrg Res Porsize(x103SI) (glcm3) (ms) (Ohm-rn) (%)53.18 2.90H9604-745H9604-757H9604-763H9604-772H9604-776H9604-784H9604-792H9604-90H9604-96H9605-1 11H 9605-128H 9605-135H9605-1 39H9604 745.3 745.5 VMmagf UH9604 757.5 757.75 VMmagf B+PH9604 763.25 763.45 IIc BH9604 772.2 772.4 VUm TH9604 776 776.2 II cBH9604 784.1 784.3 VU mTH9604 792.1 792.3 IIm FCH9604 90.3 90.5 VM mBH9604 96.2 96.4 IIFCH9605 111.5 111.7 vMmagc UH9605 128.1 128.3 VMmagc BH9605 135.7 135.9 IIc B+PH9605 139.65 139.9 II f0.59 2.811.55 2.8011.29 2.890.32 2.8738.00 2.8257.70 2.770.52 2.780.35 2.84“syenite” (K-spa-rich) veins115.00 2.85syenite” (K-spa-rich) veins2.00 2.790.28 2.78FC f-gr inrusive, purple color, possiblyassd 0.18 2.67w/nearby syenite0.71 2.80B bleached clasts, average 1cm; bleachedand 0.62 2.97chaotic matrixc FC+Sc FC+Sc FC+Sf B mafic phenocrysts, <1to 2mm, 3%, slightlyelongateH9605-167 H9605 167 167.2 Sc FC+S0.14 2.68H9605-176 H9605 176.4 176.6 Sc FC0.21 2.68 7.65 4150.6H9605-180 H9605 180.2 180.4 Sc FC0.24 2.66H9605-20 H9605 19.8 20 vMmagf B14.40 2.80H9605-210 H9605 210.7 210.9 Sc FC+S0.22 2.69 10.28 14889H9605-217 H9605 217.3 217.55 c FC+S0.21 2.68H9605-224 H9605 224.15 224.35 VUf B sheared, qtz andFe-carb veining 0.48 2.806.97 5820.4 0.32H9605-238 H9605 237.85 238.05 vuf B sheared, qtz and Fe-carb veining5.96 2.85H9605-264 H9605 263.85 264.05 VUf T massive21.18 2.87H9605-266 H9605 266.5 266.75 VUf T in-situ brecciation, angularclasts, 2cm 22.00 2.83 4.93366.1 1.08H9605-27 H9605 27.7 27.95VMmag f U34.70 2.76H9605-272 H9605 272.2 272.4Il f U mafic (hbl?) phenocrysts,average 1mm, 91.182.8510%, elongateL’JH9605-278 H9605 278278.2 VMH9605-28 H9605 28.428.6 VMmag fH9605-301 H9605 301.3301.5 VU fH9605-58H9605-66H9605-73H9605 58.4 58.65 VMmagmH9605 66.65 66.85IM CH9605 72.8 73 VMXmagUBUTFFF mylonite?B+PFCFTFFCBBFCCBF multi-lithic mylonite?FCFCFC+HB+PTFC+H0.540.310.930.990.150.950.613.85180.001.182.90 6.45 2077.9 0.402.78 13.60 8868.92.802.852.762.932.952.672.882.88Sample No. HolelD FromTo Rock type Gr. Altn notesMS Den Chrg ResPorsize(xlO3Sl) (glcm3) (rns) (Ohm-rn) (%)f U0.96 2.90U51.10 2.87T in-situ brecciation,sheared, fragments 11.882.82stretchedUFC+SB+P in-situ brecciation, alteredfragmentsaverage 0.5-1cm327.00 3.040.82 2.870.43 2.8229.40 2.760.14 2.680.59 2.8137.00 2.850.32 2.760.68 2.950.62 2.920.16 2.77 9.45 42430.59 2.920.54 2.85 9.88 5135.40.310.51 2.83H9605-8 H9605 8.258.45 VMmag fH9605-99 H9605 99.4599.65 IF CH9606-104 H9606 104.4104.6 LH9606-119 H9606118.85 119.05 VU mH9606-147 H9606 147.1147.3 II fH9606-152 H9606 151.75152 II fH9606-154 H9606 154.4154.6 VUX mH9606-173 H9606 173.2 173.45II fH9606-174 H9606 174.5 174.7II fH9606-179 H9606179 179.25 VU mH9606-206 H9606 206.15206.4 VU mH9606-22 H960622.35 22.6 VUX mH9606-230 H9606 230.25230.45 II fH9606-232 H9606 231.85232.05 II fH9606-233 H9606 233.5233.7 II fH9606-247 H9606 247.7247.9 II fH9606-270 H9606 270270.2 VUH9606-276 H9606 276.4276.6 IIH9606-278 H9606278 278.2 IIH9606-295 H9606 295.5295.7 VUXH9606-305 H9606 305305.2 II mH9606-323 H9606 323.5 323.7IIH9606-327 H9606 327.75327.95 VMmag fH9606-338 H9606 338.05338.25 VMmag fH9606-357 H9606357.3 357.5 II cH9606-370 H9606 370.05370.3 VMmag mH9606-40 H9606 40.4 40.6VUX mfragmental, mylonite?1.00 2.91 8.97 8545.90.340.61 2.860.64 2.900.45 2.841.24 2.935.98 2.83T sheared, fragmenta’H9606-411 H9606 411.05 411.3VMmag mH9606-42 H9606 41.85 42.05II mH9606-421 H9606 421.55421.75 MLXH9606-424 H9606 424.3424.5 IIXH9606-430 H9606 429.8430 II mH9606-452 H9606 452.3 452.5VMmag mH9606-459 H9606 459.75459.95 VU mH9606-465 H9606 465.75 465.95vu mH9606-471 H9606 471.5471.7 VU mH9606-496 H9606 496.35496.55 MLXH9606-501 H9606 500.9501.1 vu mH9606-537 H9606 537.05537.25 IIH9606-538 H9606 538538.2 II fH9606-539 H9606 539.5 539.7vu mH9606-569 H9606 569 569.25vMmag fH9606-59 H9606 59.559.7 vu mH9606-604 H9606 604.4604.6 VMmag fH9606-619 H9606 618.9619.1 VMP fH9606-628 H9606 628.25 628.45II mH9606-629 H9606 629 629.25IIH9606-630 H9606 630.15630.35 IIH9606-631 H9606630.85 631.05 VM fH9606-649 H9606 649.3649.55 II mH9606-66 H9606 65.866 II fH9606-660 H9606 660.2 660.4VMP fH9606-675 H9606 675675.2 VMP fH9606-675.5 H9606 675.7 675.9IIH9606-706 H9606 706.7 706.95VMP fH9606-708 H9606 708.15708.4 QMXH9606-713 H9606 713.55713.75 VM fH9606-718 H9606 718.15718.4 VMP fH9606-721 H9606 721.7721.9 VMP fH9606-725 H9606 725.25725.5 VMP mH9606-745 H9606 745.1745.35 VMP mHC+S multi-lithic breccia; large variationin susB buff and purple-coloredclastsBFC+HC massiveU massiveBBUBi-PFCB shearedUF massiveUUFC fsp phenocrystsFCFCBUFU qtz amygdules near pillow marginsBBBBBBBBB weak fabric39.53 2.7814.82 2.850.74 2.940.75 2.88124.71 2.8924.00 2.8227.65 2.8236.82 2.857.59 2.843.24 2.86 15.97 3752.20.704.95 2.920.25 2.731.06 2.8325.80 2.610.54 2.96 3.371009.04 0.5053.20 2.830.63 2.780.74 2.830.45 2.770.68 2.832.48 2.8318.40 2.750.61 2.920.60 2.790.59 2.810.42 2.760.67 2.880.35 2.870.55 2.890.54 2.880.400.39 2.800.48 2.900.55 2.900.33Sample No. HolelD From ToRock type Gr. AItn notesMS Den Chrg ResPorsize(x103SI) (glcm3) (ms) (Ohm-rn) (%)FC+H4.85 2.9011.30 22613L’)H9606-753 - H9606H9606-757 H9606H9606-784 H9606H9606-79 H9606H9606-89 H9606H9606-90 H9606H9608-102 9708H9608-118 9708H9608-129 9708H9608-145 9708H9608-176 9708H9608-190 9708H9608-192 9708H9608-198 9708H9608-209 9708H9608-217 9708H9608-217 9708H9608-224 9708H9608-24 9708H9608-258 9708H9608-259 9708H9608-27 9708H9608-272 9708H9608-306 9708H9608-327 9708H9608-346 9708H9608-361 9708H9608-37 9708H9608-378 9708H9608-39 9708H9608-405 9708H9608-419 9708H9608-422 9708H9608-444 9708H9608-463 9708752.8 753.05 II757.1 757.3 QMX784.1 784.35 VMP79.25 79.45 II89.2 89.45 VU90.1 90.3 II102.55 102.8 IF117.85 118.05 VM129.45 129.65 VM145 145.2 VM176 176.2 VMP190 190.2 VMP192.45 192.65 IM198.2 198.4 IM209.4 209.6 T217 217.3 MLX217.3 217.4 MLX224.65 224.85 VMP24.3 24.5 VMP258.4 258.6 VM258.8 259 VM27.55 27.75 T272 272.2 VMPX306.4 306.6 VM327.35 327.55 VM346.15 346.4 VM361.2 361.4 VMPX37.15 37.35 VM378.5 378.75 VMP38.8 39 VM405 405.2 VMP419.45 419.65 VMP422.65 422.85 VMP444.6 444.8 VMP463.3 463.5 VMPf Bf Bf Uf CFCBFCBuf Uf CBuuCm Cm uUf UUf BBf Cf Bf Bf Bf u0.46 2.860.60 2.800.40 2.8225.18 2.840.80 2.822.47 2.850.71 2.800.51 2840.65 2.840.67 2.7931.53 2.7967.53 2.801.00 2.890.65 2.840.61 2.811.52 2.812.54 2.8370.50 2.910.45 2.851.57 2.9118.50 2.970.72 2.8312.50 2.8514.80 2.880.620.60 2.8834.50 2.950.68 2.9214.10 2.930.91 2.810.54 2.840.58 2.830.62 2.930.53 2.77Sample No. HolelD From To Rock type Gr.Altn notesMS Den Chrg Res Porsize(x104SI) (glcm3) (ms) (Ohm-rn) (%)m FC0.55 2.91f UFTf TC Bf BB qtz-carbonate breccia; shearedspinifexin-situ brecciationpurple mineral (a carbonate?) in veinepidote and cal veinsunaltd pillow basalt; cal veinsI’JH9608-508 9708H9608-522 9708H9608-537 9708H9608-554 9708H9608-562 9708H9608-573 9708H9608-600 9708H9608-639 9708H9608-650 9708H9608-657 9708H9608-674 9708H9608-68 9708H9608-712 9708H9608-85 9708H9608-98 9708H9707-102 H9707H9707-111 H9707h9707-122 H9707H9707-137 H9707H9707-141 H9707H9707-144 H9707H9707-160 H9707H9707-162 H9707H9707-171 H9707508.25 508.45 VMP522.2 522.45 VMP537.6 537.8 VMP554 554.2 VMP562.3 562.5 VMP573.3 573.55 VMP600.1 600.35 VMP639 639.2 VMP649.9 650.1 VMP657.3 657.5 VM673.9 674.1 VM68.5 68.7 VMP712.7 712.9 VM85.25 85.45 VM97.9 98.1 VM102.5 102.7 VU111.2 111.4 IF121.85 122.1 IF137.3 137.5 IM141.6 141.8 VU144.35 144.55 IF160.75 160.95 II162.65 162.85 VU171.2 171.4 MLXf B in-situ brecciationf Bf Bffff FC Um UC Um Tm FCffchlorite amygduleschlorite amygduleschlorite amygduleschlorite amygdules0.41 2.860.41 2.860.34 2.840.45 2.870.42 2.900.53 2.840.38 2.860.46 2.890.47 2.860.38 2.9044.30 2.890.54 2.904.03 2.861.29 2.791.66 2.900.40 2.800.16 2.66to sub-angularH9707-179 H9707H9707-188 H9707H9707-194 H9707H9707-204 H9707H9707-219 H9707H9707-241 H9707H9707-265 H9707H9707-273 H9707H9707-301 H9707H9707-307 H9707179.7 179.9 VMmag fU188.15 188.4 VMXmag fC194.65 194.85 II f B204.7 204.9 VMmag f U219.7 219.9 VMP f U241.1 241.3 VMmag m U265.1 265.3 VMmag m U273.2 273.4 VMmag mU301.05 301.25 VMmag m U307.05 307.3 VMmag m U1-2 cm fragments, clast-supportedamygdules at marginssyenite (Kfsp) veins2.21 2.87147.06 2.990.60 2.8857.10 2.620.77 2.8035.80 2.472.87 2.9242.70 2.9457.10 3.00119.00 2.87Sample No. HolelD FromTo Rock type Gr. AItn notesMS Den Chrg Res Porsize(x103SI) (qlcm3) (ms) (Ohm-rn) (%)f C0.56 2.87BBBBf Bm Bm Bf Cm Uf Um Usheared; zoned fragments?49.53 2.83 8.87 758.82 0.76slightly sheared; abund quartzamygdules 10.60 2.83 2.902241.6 0.8830.24 2.76FC abundant py along fracures18.59 2.95T qtz fragments from shearing of veins10.22 2.81FC+H felsic and mafic clasts, < 1cm; sub-rounded14.59 2.75H9707-31 7H 9707-351H9707-36H9707-373H9707-38H9707-388H9707-403H9707-420H9707-456H9707-478H9707-488H9707-494H9707-495H9707-512H9707-51 5H9707-522H9707-540H9707-554H9707-56H9707-575H9707-599H9707-603H9707-614H9707-625H9707-627H9707-637H9707-655H9707-67H9707-677H9707-685H9707-690H9707-84H9707-96H9707 603 603.2 VMmagH9707 614.35 614.55 VMmagfH9707 625.25 625.45 VMmagH9707 627.1 627.3VMmagH9707 637.3 637.5 VMmagfH9707 655.75 655.95 VMmagmH9707 67.15 67.35 IImH9707 677.4 677.6 VMmagmH9707 684.8 685 VMmagmH9707 690.3 690.5 VMmagmH9707 84.4 84.6 IIfH9707 96.65 96.85 IFcFUFUUUUBUUBBFC+SUUUTU2.26 2.890.69 2.8697.5332.12105.6512.5962.901.720.42107.0083.50131.000.690.142.932.913.083.142.992.782.842.873.022.872.68Sample No. HolelD FromTo Rock type Gr. AftnnotesMS Den Chrg Res Porsize(x104SI) (q!cm3) (ms) (Ohm-rn) (%)U60.50 3.00H9707 317.15 317.4 VMmagmH9707 351.15 351.4 VMmagmH9707 36.5 36.7 VUH9707 373.65 373.9 VMmagmH9707 38.25 38.5 IIfH9707 388.75 388.95SH9707 403.7 403.9 VMmagmH9707 419.8 420 VMmagmH9707 456.05 456.25 VMmagmH9707 478.25 478.45VMmag mH9707 488.3 488.5 VMmagmH9707 494.1 494.3 VMmagmH9707 495.5 495.7VMXmagH9707 511.8 512 VMmagmH9707 515.2 515.4 VMXmagH9707 522.45 522.65 VMmagmH9707 540.5 540.7 VMmagmH9707 553.8 554 VMmagmH9707 56.7 56.9 VUfH9707 575.5 575.7 VMmagmH9707 599.65 599.85 KMXmagU syenite (Kfsp) veins130.00 2.88sheared0.81 2.9477.80 2.96 46.50 305560.260.51 2.9130.82 2.69128.00 2.9548.30 2.92141.00 2.972.19 2.5888.24 2.891% cpy interstitialstrongly bleached fragments <1cm,someqtz clastsabundant qtz/Fe-carb veinsshearedU Kspar rich clasts; <2cm frags,may havebeen a veinU in-situ brecciationUU in-situ brecciation <1cm,perlitic; varioles?BUUTUUUFC+S minor qtz amygdules (<1mm)U2.6884.2026.909.4319.4015.100.5546.822.852.822.842.832.802.842.862.87I’—)H9711-106 H9711 106.25 106.45 VMH9711-116 H9711 115.95 116.15 VMH9711-134 H9711 133.9 134.2 VMH9711-149 H9711 148.85 149.05 VMXH9711-149.5 H9711 149.4 149.6 VMXH9711-162 H9711 162 162.2 VMH9711-178 H9711 178 178.2 VMH9711-192 H9711 192.55 192.75 VMH9711-211 H9711 211.6 211.8 VMH9711-247 H9711 247.3 247.5 VMXH9711-250 H9711 250 250.2 VUH9711-252 H9711 252.15 252.35 SH9711-265 H9711 265.45 265.65 VUH9711-281 H9711 281.5 281.7 VUH9711-296 H9711 296.65 296.85 VMmag fH9711-313 H9711 313.5 313.7 VM fH9711-320 H9711 320.6 320.8 VM fH9711-344 H9711 344.2 344.4 VMmag fH9711-368 H9711 368.5 368.8 VM fH9711-378 H9711 378.5 378.7 VM fH9711-401 H9711 401.6 401.8 VM fH9711-430 H9711 430.4 430.6 VMmag fB÷P varioles; bleached clastsB variolesBBm Um U mafic phenocrystsB+PB shearedFH0.560.53 2.830.68 2.740.440.610.611.07 2.800.67 2.810.53 2.760.76 2.890.46 2.720.75 2.87B beige “clay-looking mineral - leucoxene (I.e. 0.69was ilmenite)FC+H+sBB bleached variolesUB variolesB÷PB soft green material in veinsFC+H1.25 2.870.85 2.8824.82 2.882.95 2.891.27 2.931.29 2.9111.06 2.93 N/A >70000H9711-439 H9711 439.1 439.35 VMmag fH9711-446 H9711 445.95 446.15 VMmag fH9711-466 H9711 465.8 466 VMmag fH9711-470 H9711 470.05 470.25 S fH9711-475 H9711 475.75 475.95 VU fH9711-496 H9711 496.4 496.6 VUH9711-521 H9711 521.65 521.85 IFP fH9711-53 H9711 53.05 53.25 VMH9711-551 H9711 551.65 551.85 VUH9711-65 H9711 65.5 65.7 VMB variolesUUFCB shearedTU phenocrysts, 1-2 mm, 5-10%m UT14.35 2.9144.59 2.8226.47 2.820.80 2.843.05 2.8538.60 3.080.22 2.685.6653.80 2.79N/A >70000Sample No. HolelD From To Rock type Gr. Altn notesMS Den Chrg Res Porsize(x104SI) (gicm3)(ms) (Ohm-m) (%)m B÷P 0.49m FC+Sf BffmmfmmB2.87 5.34 5164 0.4932.00 2.94m B 1.19Sample No. HolelDFrom ToRock type Gr. Altn notesMS Den ChrgRes Porsize(x104SI) (g!cm3) (ma) (Ohm-rn) (%)H9711-99 H971199.5 99.7 VMm FC+H0.86 2.72HGP-site I hndsmpVM m9.00 7453.7 0.50APPENDIX 2F - PHYSICAL PROPERTIES- DESCRIPTIVE STATISTICSMagnetic SusceptibilitySUSCEPTIBILITY (x1O3 SI Units)No. Mean Std. Dev. Logmean Median RangeUltramafic volcanic rocks - all82 15.63 22.82 4.293.83 0.41-109Least altered (dol÷chl)8 6.03 9.61 2.15 1.490.57-27.65Talc-chlorite assemblage46 24.12 26.57 9.6 14.610.44-109Fe-carbonate+muscovite altered16 2.73 10.41 1.861.01 0.41-36.82Magnesite+fuchsite altered9 0.75 0.37 0.70.62 0.49-1.66Mafic volcanic rocks - all218 26.11 45.18 3.861.42 0.15-327Leastaltered107 40.09 52.94 9.7821.5 0.35-327Fe-carbonate+muscovite altered75 6.86 25.24 1.030.6 0.15-148.24Fe-carbonate+albite altered14 1.71 3.21 0.840.58 0.28-12.4Intermediate dikes-all59 12.07 29.09 1.210.61 0.13-135.29Least altered11 41.11 47.24 7.0218.4 0.24-135.29Fe-carbonate+muscovite altered10 0.75 0.35 0.680.69 0.32-1.55Fe/Mg carbonate altered22 10.95 25.91 1.260.58 0.13-107Syeniteintrusives-aII34 1.19 5.25 0.250.2 0.07-30.82Least altered12 2.87 8.8 0.420.27 0.16-30.82Muscovitealtered11 0.36 0.65 0.210.15 0.14-2.31Fe/Mg carbonate altered4 0.23 0.021 0.230.23 0.21-0.25Porphyritic rhyolite dikes-all37 1.59 5.41 0.260.16 0.05-30.24Least altered15 0.35 0.61 0.4450.16 0.05-2.45Muscovite altered13 0.48 0.82 5.710.14 0.08-2.47Fe/Mg carbonate altered7 2.39 5.39 2.190.28 0.12-14.5900DensityDENSITY (glcm3)No. Mean Std. Dev. Median RangeUltramafic volcanic rocks-all81 2.86 0.057 2.85 2.74-3.08Least altered (dol+chl)8 2.85 0.023 2.86 2.82-2.89Talc-chlorite assemblage46 2.85 0.061 2.84 2.74-3.08Fe-carbonate÷muscovite altered15 2.87 0.054 2.852.78-3.01Magnesite+fuchsite altered9 2.91 0.035 2.922.85-2.96Mafic volcanic rocks - all205 2.86 0.1 2.87 1.94-3.14Least altered101 2.87 0.094 2.872.47-3.08Fe-carbonate+muscovite altered71 2.85 0.13 2.861.94-3.14Fe-carbonate+albite altered13 2.81 0.051 2.81 2.74-2.93Intermediate dikes - all 592.82 0.078 2.83 2.67-2.95Least altered11 2.83 0.077 2.812.72-2.95Fe-carbonate+muscovite altered10 2.85 0.054 2.86 2.76-2.94Fe/Mg carbonate altered22 2.79 0.088 2.782.67-2.95Syenite intrusives - all 322.7 0.057 2.69 2.59-2.86Least altered10 2.7 0.0572.69 2.64-2.84Muscovite altered11 2.72 0.072 2.692.64-2.86Fe/Mg carbonate altered4 2.67 0.0096 2.682.66-2.68Porphyritic rhyolite dikes - all36 2.61 0.45 2.67 2.63-2.88Least altered15 2.51 0.7 2.67 2.63-2.84Muscovite altered 122.68 0.074 2.67 2.57-2.85Fe/Mg carbonate altered7 2.73 0.082 2.712.65-2.88249ResistivityRESISTIVITY (Ohm-rn)Mean Std.Dev. Log meanMedian2815.99 2451.991574.58 2159.751127.75 1376.22573.95 385.694451.55 1714.493952.65 4757.6028431.94 36820.1613004.05 16462.0017910.02 17881.258344.7016462.00Range111.16-8545.9111.16-3994.31009.04-3217.4541.24-155090541.24-49812Chareabi1itvCHARGEABILITY(ms)Std. Dev. Median4.465.465.89 4.242.45 5.6844.97 13.8011.049.37Range2.9-20.42.9-20.43.37-9.882.07-1 58.172.07-33.23No.2087186Ultramafic volcanicrocks - allTalc-chlorite assemblageCarbonate+muscovitealteredMafic volcanicrocks - allCarbonate+muscoviteor albitealteredIntermediate dikes- allSyenite intrusives- allPorphyritic rhyolitedikes - all6 9758.8510 6759.5111 11534.008025.693528.595359.317193.336027.6110523.106611.756179.8511136.002313.6-226132630.6-148894576-23970No.Ultramafic volcanicrocks - all20Talc-chlorite assemblage8Carbonate+muscovitealtered7Mafic volcanic rocks- all 18Carbonate+muscoviteor albite6alteredIntermediatedikes - all6Syenite intrusives- all10PorphynticyoIite dikes-all11Mean6.916.366.6331.3212.0329.2414.7110.4742.357.0413.1912.4513.924.209.45-115.576.27-27.42.2-46.2250APPENDIX2G - CORRELATIONCOEFFICIENTSFOR PHYSICALPROPERTIESAND XRD (RIETVELD)- DERIVED MINERALABUNDANCESCalculatedusing the statisticalanalysis softwareSPSS Statistics. Correlationcoefficientsare calculated forall physical properties.Not all mineralsare considered,only onesoccurring mostcommonly.Spearman’ s correlationcoefficient calculationsare used as theyare appropriatewheredata do satisfy normalityassumptions.ALL ROCKTYPES (PAGE1)LOG_MAIIMAGSUS GSUSIDEN Cl-IRG RESLOG RES POR ABJCAL CLCSpearmaonrho MAGSUS CorrelationCoefficient1000 1 000-i 444j.278 -.212 .2*4175 -.5131 178- 188Sig. (2-tailed)000 GOt) .034]- 69 07 363003 601 442N365 385 369I 5 58 58 29 32t 19LOG_MAGSUS Correlation Coefficienti.ooor 1.000 -427T-212t-214 175- 513H 176 188Sig. (2-tailed)000000 034 109107 363003 601 442N386 395 36958 59 5028 3211 19DEN CorrelationCoefficient 444W444 1000076 036 - 0383$1 646 405- 030Sig. (2-tailed)013Cr 000 -568 789 778 041000 26 8133N‘9 365 37059 58 5828 32 119CHRG Correlation Coefficient. 78 27s.076 000352 354i 519j-115j030 243Sig. (2-tailed)**34034 .569006 005I 006I560 I 834 348N58 5558 80 8080 27 2610 17RES Correlation Coefficient-212 -212 -0361005 1.000j - 8t3.021 .146 -287Sig. (2-tailed)109 109 7891 00d. 000 000 9*4688 300N58 58 58Ø4 60I60 2728 10 17LOG_RES CorrelahonCoefficient-214 -214 -0381 000 1 000815 020 168- 264Sig. (2-tailed)107 107 778000 000919 643 3i37N58 50 5860 J_,ffl,._60 2728 10 1?POR Correlation Coefficient175 175 —aBr..815 1000-29t I -26t 569Sig. (2-tailed)363 36304t . - -- 000 I 36668 042N29 2999 - - 27--27 27 30 128 13AB Correlation Coefficient:‘-- 64fl - 115 02020 -291 1.000- 74T -395Sig. (2-tailed>900 560 914919 358014j104N42 21j 3228 2828 2 3310L18CAL Correlation Coefficient178 178 405-030 146 168-261 042* 1000 57*Sig. (2-tailed)601 601 2 6934 688 643618 014 I.135N11 II IItO to to6 56 15CLC Correlabon Coefficient-188 -198 -630-243 -267 264569 -395 57 1.000Sig. (2-tailed)442 .442 903348 300 307842 104 139N19 19 1517 17 173 10 8 20DOL Correlation Coefficient 417*417 578 95-067 -013 -060922 -378 -099Gig. (2-tailed>634 fli34 ,)062 385974 954 888001 403 748N26 26- l’ 26 2222 22 825 7 13CB_TOT CorrelationCoefficient 16080 I .6.469* - 68063 065 087- 6’0 301 -037Sig. (2-tailed)332 332625 403 753 745800 SOC 369889N31 3 3127 27 27ii I 30 17HEM Correlation Coefficient.. .7W256’ 072 I 850048 048 -350450 1 000’ 056Sig. (2-tailed>--1ffl 036 878 I 007910 910 6 4310 .913N,-3-,,,7171 8 88 6 73 6MAO Correlation Coefficient- 857’ 957 518083 -083 -083-.800 -19000 07Si9 (2-tailed)407 007 18883 831 831200 651 .873867N8 0 89 9 94 8 58MC_INT Correlation Coefficient-456 -456- ‘84’ 051 - 19 - 191-371 i975 - 725 -367Sig. (2-tailed> 088088 8645 3 513 468- - .$- 193332N15 1545- 14 1414 6 -1$- 6 9MS CorrelationCoefficIent -515-515 -492 -.297-442 - 462 -300-000 -l 090 -429Sig (2-tailed)128 28148 409 200179 624987 1.000 337N10 10 109 10 105 10 27MUS_TOT CorrelationCoefocient -527-527 -592 -297-442 -462 -305-006 - 009 -429Sig. (2-tailed)096 096 055465 200 179624 987 I 000337NIi II Ii10 10 105 10 27PY Correlation Coefficient381 38.494’ 380 -242264 -347-789 600Sig. (2-tailed>132 132044- 159 385 34173 112 .208N17 17 118 15 151 7 56QTZ CorrelationCoefficient -235-235 -163-255 -081 -077308 167 008032Sig. (2-tailed)211 211389 158 693708 330 386983 900N30 30 3026 26 262 29 88FECB_TOT Correlation Coefficient- 411’ 43V 886 .259251 248 000 -c72 -450-242Sig (2-tailed) -i,$2682$- - - 000 233248 2551 500 - 000 310426N- 727 - 27 23 2323 8 257 13Correlation in oigoificnnt at the.51 level (2-tailed)-• . . - . .Relationships addressedin Chapter 2-Correlationis significant at the .05 level (2-tailed)251C)°‘‘2IZCoC)ZCoC)ZCoC)ZCooZCoC)ZCOC)o8C)aC)oC).0-tC)Co 0-—0 -1JVV800 ni 0)z.C)C)0C)C)>gZCoC)ZCoC)ZCoC)ZCoC)ZOOC)aC)8C)c,C)C)tt3;uC)0mzm0Coj:ZCoC)ZCoC)ZCoC)ZCoCo8oC Cl)0 0 C) Co -1 I0-CC)Co.———V000O0rC0)00Co000ob-o-bb--nCo.g.•.CV0)g0)g0)Vm C)——V,V.•VV-—-—-—————-—-___.1•‘0!-‘00’V•V--——VCD!1r1“-V—V———VV:’0’0I.VIVI‘000,(0‘0‘0‘00(00N0(00(N(0(0(N(000(N(000(NV(NV(00)(00(0‘0‘00(0W000fl0’0(0.-0_0_N00fl0-0(000-(-00‘0’‘0‘000)00(0(00-‘0(.000((((000(0‘‘0CD0((08V’0(0’0(fl0((0(000-..0(0(00(00‘00(000(00-0((V(0-0000000‘00((((0W(—————‘0(D(0(0(0(0(((g00g0goo—VVVVV-VVCD0—‘0’0gg’0.0o,oVVV•VVV•VVVVVVVV•VV-V—V—VVVVVrC)CDV00’0_’00’001‘0’00OC)bOfl(aV0-._0)0‘000CD(0(00)(0(00000(0-((0(0(.(0(0(V.CDV(0CDV(00000,CD00(00)00(0((((0(0O((00)‘0‘0‘0(0*00‘00-.0———<VCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCDCD888c88c3(CCC:——0,:0DOD0—IE00O¶1CzC253O CC100tI0000>-ur-r00rgIZ0‘-uI(0d0mI(0-I00C) CzøOz(0ozøOzøozwozøozøozwOzøozwozo,ozwozmozo,ozoozo,ozøoz070zG,ozwnzcnoee2e2eeCCCoenCoCoC)C)C)C)C)C)C)C)C)C)C)C)C)C)C)C)C)(0,..C)-2—----------------)•0 Sg0 C)--——--——,(0,0—---j---,-I-,a Oip-z00CHHrnx3030803030308030308030303080303030303030Soaaaaaaaaaaaaaaaaaaaa0.--,-0 0-----j——j;-:-:0. aE10 C,C CI’JFELSIC ROCKS- SYENITEINTRUSIVESAND RIJYOLITEDIKESFECB_TOT CorrelaSonCoefficient087518. (2-laded).801)N11.999 .)99 .6W.6W -22) -.247.306 .366[-.438.000 001005 .005.373 324282 298 17770 67 1518 18 18It 10ii-.643 565.575 -.770600 119 054.051 0154 71) 12 8opoarrranc cc aco.oCorrelatiOn COOt00eñLone519. (2-laded)N70LOG_MA1 11MAGOUS GSUSDON CI-IRG LOGCoG RES LOG RESAS DOL ICB TOT I MC INTMS I MUS TOTPY LOG Ph’ OTZFECB TOT.730262.097.600LOG_MAGSUSCorrelason 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 .262298 .177 262.600 119056 .651 .015800N70 7067 1818 18 18It 10Ii 44 7 1212 9IIDEN ConretafionCoefficient::::.6 1.000 .655.660 -.096 . 715-.462 .494 -.069250 .400 .667SaP .576 -.73l.402So (2tol d)8 601003 003698 649 153147 040742 800 102048 051 025220N67 67 1818 1818 II ICft4 4 712 129 IiCHRG Correlasn Co ffictent655 I 000 IOOT 095 109600 ISO317 1 000 41)0143 568 508676 2675.9 (21 led)9 903705 667088 651.951 600787 074 074084 488N15 1818 18 189 89 4 49 10 107 9LOG_CHG C I IC Ill I655 1 000 I 000 099169 600 190377 000400 143588 505 679267Sg (21 I d)8 053708 667 088651 406600 787 074074 594489N95 18 1615 169 9 94 4 616 IC 79RES Correla9onCoefficient-223 -.216 -.098-.095 -.0951.005 .99T.233 .033W.850 -200 -.400-.600 .081.091 .107 .833Sig. (2-tailed) 373390 .698 .708.708 0115.546 .013.004 .800 . .600208 803.803 .819.005N19 18 1810 1818 169 89 4 46 10 107 9LOG_RES CorrelationCoetficlenl-247 -.239 -.115-.109 -.109.999 7.000210 814W .879.200 - 405 -.600.036 .036 .162.820Sig. (2-tailed).324 .339.649 .667 .667.000 974014 .002.000 .600.208 920 .920.728 .006N 1818 1818 19 5819 98 9 44 6 tOtO 79AB 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 .260.401NII II II9 9 9 -- 9 1110 114 3 611 119 IlDOL Correla9onCoefficient.366 .366 .494.190 .190 .937. .8l4 -.0181000 .652000 -1000-.700 624 .620-.476 .964Sig. (2-teiled).298 .298.147 .657 .651./2t1) 14. .960. 0691 000 1000 .188.054 .056 233.000N10 1015 864- 8 10 10 tO4 20 tO 108 10CB_TOT C IIi C ffi nt438 438069 317 317$4 8Th 039 6121 600 865500 371 027005 571736Sg (2toI d(177 177 040406 40694 658 915 060200 667 468937 989139 950NII II11 99 9 811 101 4 36 II II8 ItMC_INTConolason Coefficient.739 .739259 1.500 7.000-.209 - 200 -1 SOT.000 -.800 1000 . 1000 .800 .500-.506 .000Sig. (2-tailed).262 .262.742 . ..000 .800.000 1.060200 .. ..200 .200.667 1.000N4 44 4 44 44 4 44 0 24 4 34MS CorrelationCoefficient -400-.400 .400 -400-.400 -.400-400 -1 000 -1.000.500 . 1.000800 . 400-400 1 000500Sig. (2-tailed).600 .600.600 .600 600.600 .660 .0007.050 .667. . .200.605 .603. .667N4 4 44 4 44 32 3 04 4 442 3MUS_TOT Cnrrela8ocCoefficient-.643 -.643667 .143.143 -.600 -.600-.486 -.700-.371 1.000W .800I 000 -.534 -.536.400 . 371Org. (2-ta,ler.719 .119 .102.787 .787 .208.208 .329.198 .468. .200 ..215 .219.600 .490N7 77 6 66 6 65 62 4 77 74 6Ph’ CorrelationCoefficient .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 .061133N12 12 1210 10tO 1011 10 ii4 4 712 12 911LOG_PY CorrelationCoefficient.575 .575 .575.588 .586.691 .036 -.328.620 .005 .850-.450 -.536 .... - 9131317.000 -,669.478Sg (21 I d(051 051 051974 074 803920 325056 989260 600 2151)60 049137N12 12 12In 10 In10 II10 If4 47 - -12 120 IlQTZ Cornela8onCoeffinient-.77-3 - - -.Zl-.679 -.679.107 .162.452 -.476.571 -.5001.000 .400 -.633-.669 1.000-.476Sig. (2-toiled)075 -- 015 -- - - 035.094 .094 .819728 .260 .233.139 .667.8130 .567 049- .233N9 .5 .97 77 78 8 83 2 49 9 98.087.80011Carnela9on is significantat the .01 level (2-tailed).-. Ccrnelason is significant atthe 05 level (2-tailed).Relotrocships addressed inChapter 2.402 .267267 . .. o.aosco -282eon ton.- 000.500 -371 .492.478 -.4761.000.220 .488408 - -- 066401 0011 515t 000 .667 .468.133 .137.23315 99 .9’J-i’ri 13II fll377 4 36 75 II5 11UIONAppendices on accompanyingCD:APPENDIX 3A - OBSERVEDVERSUS PREDICEDDATA FOR SYNTHETICINVERSION MODELSAPPENDIX 4A - HISLOP3D MAGNETIC, 3D GRAVITY,3D DC RESISTIVITY,AND 3D IP INVERSIONRESULTSAPPENDIX 4B- 2D DC RESISTIVITYAND INDUCED POLARIZATIONINVERSION RESULTSFOR HISLOPAPPENDIX 4C- OBSERVED VERSUSPREDICED DATAFOR IIISLOPINVERSION MODELS257

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
Russia 25 0
Canada 22 9
China 20 29
United States 13 22
Finland 5 0
Australia 4 0
India 4 0
Germany 3 10
France 3 0
United Kingdom 3 0
Japan 1 0
Ghana 1 0
City Views Downloads
Unknown 40 58
Shanghai 6 0
Shenzhen 5 29
Penza 5 0
Ashburn 5 0
Sydney 4 0
Mumbai 4 0
Vancouver 4 2
Calgary 3 1
Guangzhou 2 0
St Louis 2 0
Hangzhou 2 0
Brampton 2 0

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

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0052581/manifest

Comment

Related Items