Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Integrated geophysical modelling of the northern Cascadia subduction zone Dehler, Sonya Astrid 1991

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1991_A1 D43.pdf [ 12.32MB ]
JSON: 831-1.0052854.json
JSON-LD: 831-1.0052854-ld.json
RDF/XML (Pretty): 831-1.0052854-rdf.xml
RDF/JSON: 831-1.0052854-rdf.json
Turtle: 831-1.0052854-turtle.txt
N-Triples: 831-1.0052854-rdf-ntriples.txt
Original Record: 831-1.0052854-source.json
Full Text

Full Text

INTEGRATED GEOPHYSICAL MODELLING OF THE NORTHERN CASCADIA SUBDUCTION ZONE by Sonya Astrid Dehler B.Sc. (Geophysics), University of Calgary M.Sc. (Geophysics), University of British Columbia A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF GEOPHYSICS AND ASTRONOMY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA January 1991 © Sonya Astrid Dehler 1991 In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Geophysics and Astronomy The University of British Columbia 129-2219 Main Mall Vancouver, Canada V6T 1W5 Date: ABSTRACT The northern Cascadia subduction zone involves convergence of the Explorer Plate and northern part of the Juan de Fuca Plate with the North American Plate along a margin lying west of Vancouver Island, Canada. A wide accretionary complex which underlies the continental slope and shelf has been formed. Two allochthonous terranes, the Crescent Terrane of Eocene oceanic crustal volcanics and the Pacific Rim Terrane of Mesozoic melange sedimentary rocks and volcanics, lie against the Wrangellia Terrane backstop beneath the west coast of Vancouver Island and outcrop on the southern tip of the island. The intrusive Coast Plutonic Complex underlies the westernmost part of the British Columbia mainland east of Vancouver Island and marks the location of the historic and modern volcanic arcs. An integrated interpretation of geophysical and geological data has been conducted for the northern Cascadia subduction zone. Regionally extensive gravity and magnetic anomaly data have formed the basis of the interpretation, while surface geology, physical properties, and seismic reflection, refraction, heat flow, borehole, magnetotelluric, and seismicity data have provided constraints on structure and composition. Horizontal gradient and vertical derivative maps of the potential field data were calculated to provide additional control on the locations of major faults and lithologic boundaries. Iterative forward modelling of the gravity and magnetic anomaly data was conducted along three offshore multichannel seismic reflection lines and their onshore extensions. The two-and-a-half-dimensional (2.5-D) models extended from the ocean basin across the ac-cretionary complex and Vancouver Island to the mainland along lines perpendicular to the major structural trends of the margin and revealed lateral changes in the location of several structural components along the length of the margin. The interpretations were extended ii laterally by moving the original models to adjacent parallel positions and perturbing them to satisfy the new anomaly profile data and other constraints. The models thus formed were moved to the next position and the process repeated until a total of eleven models was devel-oped across the margin. A twelfth line across a gravity anomaly high on southern Vancouver Island was independently modelled to examine the source of this feature. A n average density model for the southern half of the convergent margin was constructed by averaging the models and profiles for seven lines at 10 km spacings. This process removed anomalies due to small source bodies and concentrated on the larger features. Finally, a regional density structural model was developed by linearly interpolating between all eleven cross-margin lines to construct a block model which could then be 'sliced' open to examine the internal structure of the margin at any location. The final models allow the Pacific Rim and Crescent Terrane positions to be extended along the offshore margin from their mapped locations. The Pacific Rim Terrane appears to be continuous and close to the coastline along the length of Vancouver Island, while the Crescent Terrane either terminates halfway along the margin or is buried at a depth great enough to suppress its magnetic signature. The location of the Westcoast Fault, separating the Pacific Rim and Wrangellia Terranes, has been interpreted to lie west of Barkley Sound at a position 15 km west of its previously interpreted position. Beneath southern Vancouver Island and Juan de Fuca Strait, the Crescent Terrane appears to have been uplifted into an anticlinal structure, bringing high density lower crustal or upper mantle material close to the surface and thereby causing the observed gravity anomaly high. The western part of the Coast Plutonic Complex has been interpreted as a thin lower density layer extending from its surface contact with Wrangellia to a position 20 to 30 km further east where the unit rapidly thickens and represents the main bulk of the batholith. The complexity of the thermal regime and its effects on density in this region allows for other interpretations. i i i Finally, a comparison of the models along the length of the margin reveals that the crust of Vancouver Island appears to thin toward the north above the shallower Explorer Plate and the complex low - high density banding used in the southern Vancouver Island models is replaced with a single high density unit on the northernmost line. iv TABLE OF CONTENTS ABSTRACT ii LIST OF TABLES ix LIST OF FIGURES x Acknowledgements xiii 1 PRELUDE 1 1.1 Introduction 1 1.2 Tectonic Overview 2 1.3 Major Geologic Units 6 1.3.1 Wrangellia Terrane 7 1.3.2 Pacific Rim Terrane 11 1.3.3 Nanaimo Group 14 1.3.4 Crescent Terrane 15 1.3.5 Carmanah Group 17 1.4 Overview of Procedure 17 2 PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 19 2.1 Data Acquisition and Preparation 19 2.2 Gravity and Magnetic Field Maps 23 2.2.1 The gravity anomaly map 23 2.2.2 The magnetic intensity map 25 v 2.3 Filtered Potential Field Maps 27 2.3.1 Horizontal gradient of gravity 27 2.3.2 Vertical derivatives 30 2.4 Results of Potential Field Processing 33 3 GEOPHYSICAL CONSTRAINTS 37 3.1 Constraints on Subsurface Structure 37 3.1.1 The ocean basin 37 3.1.2 The continental shelf and slope 39 3.1.3 The accretionary prism 40 3.1.4 Continental crustal structure 41 3.1.5 The subducting oceanic lithosphere 50 3.2 Previous Modelling Studies 51 3.2.1 Gravity modelling 51 3.2.2 Magnetic modelling 55 3.3 Constraints on Density and Magnetic Susceptibility 56 4 POTENTIAL FIELD MODELLING 58 4.1 Overview of Procedure 58 4.1.1 The modelling algorithm 58 4.1.2 The data profiles 60 4.2 Primary Models 66 4.2.1 Line 1 66 4.2.2 Line 2 84 4.2.3 Line 4 88 4.3 Secondary Models 94 4.3.1 Line IAS 94 vi 4.3.2 Line 1A 97 4.3.3 Line IB 99 4.3.4 Line IC 99 4.3.5 Line ID 104 4.3.6 Line IE 106 4.3.7 Line 2A . . . 107 4.3.8 Line2B 109 4.4 Southern Vancouver Island Anomaly I l l 4.4.1 LineS I l l 5 FINALE 117 5.1 Model Summary 117 5.1.1 Average density model 117 5.1.2 Regional density model . 122 5.2 Results and Implications 126 5.2.1 The Crescent Terrane 126 5.2.2 The Pacific Rim Terrane 128 5.2.3 The CPC contact 129 5.2.4 The southern Vancouver Island anomaly 130 5.2.5 The reflective zones 131 5.2.6 The Explorer Plate region 131 5.3 Summary 132 REFERENCES 135 APPENDIX A 145 vii APPENDIX B 149 viii LIST OF TABLES B . l Densities measured from rock samples 150 B.2 Susceptibilities measured from rock samples 151 ix L I S T O F F I G U R E S 1.1 Geometry of the Cascadia subduction zone 3 1.2 Generalized physiography of the study area 5 1.3 Generalized geologic map of Vancouver Island 8 2.1 Location map of anomalies 22 2.2 Gravity anomaly map 24 2.3 Magnetic anomaly map . 26 2.4 Gravity horizontal gradient map 29 2.5 Power spectra of potential field data 31 2.6 Gravity first vertical derivative map 32 2.7 Magnetics first vertical derivative map 34 2.8 Magnetics second vertical derivative map 35 3.1 Location map of major geophysical studies 38 3.2 VISP 80 along-island refraction models 42 3.3 VISP 80 onshore-offshore refraction profile . 44 3.4 Line drawings of LITHOPROBE 1984 seismic sections 46 3.5 Cross-section across the margin 48 3.6 Location map of previous gravity and magnetic modelling studies 52 4.1 Locations of data profiles and models 61 4.2 Gravity anomaly profiles 63 4.3 Magnetic anomaly profiles 64 x 4.4 Reflection section 85-01 and 84-01 67 4.5 Line 1 initial and intermediate models 72 4.6 Line 1 intermediate models 73 4.7 Line 1 final model 75 4.8 Proposed Westcoast Fault location 78 4.9 Anomaly signatures of model 1 structure 83 4.10 Reflection section 85-02 85 4.11 Line 2 final model 87 4.12 Reflection section 85-04 89 4.13 Line 4 final model 91 4.14 Line IAS final model 95 4.15 Line 1A final model 98 4.16 Line IB final model 100 4.17 Line 1C final model 101 4.18 Line ID final model 103 4.19 Line IE final model 105 4.20 Line 2A final model 108 4.21 Line 2B final model 110 4.22 Structural cross-section across southern Vancouver Island 112 4.23 Line S final model 113 5.1 Average gravity anomaly profile 118 5.2 Average density model 119 5.3 Comparison between previous model and results of this study 121 5.4 Density model locations in regional block model 123 5.5 Regional block model 124 xi 5.6 Vertical slices through regional block model 125 5.7 Tectonic summary map 127 A . l Contour gravity anomaly map 146 A.2 Contour magnetic anomaly map 147 A.3 Contour magnetic anomaly map 148 xii Acknowledgements I would like to thank Dr. Ron M. Clowes for his supervision and guidance during the course of this study. It was a pleasure working with Ron and sharing his enthusiasm for tectonic research. I am also grateful to the other members of my committee, Drs. Doug Oldenburg and Garry Clarke of U.B.C. and Dr. Roy Hyndman of the Pacific Geoscience Centre, for then-many helpful comments which considerably improved the results of this study. Discussions with scientists from the Geological Survey of Canada, Vancouver, and the Pacific Geoscience Centre have been very helpful and I would in particular like to thank Ralph Currie, Chris Yorath, David Seemann and Jack Sweeney for the information they have provided. I am indebted to John Amor for his expertise in managing and supporting the Geophysical Research Processing Facility in the Department of Geophysics and Astronomy, and for his considerable assistance in program development and modification. I am also grateful to the many others who endured endless discussions on the various aspects of this research and contributed insight and coffee. I wish to thank the Science Council of B.C. for funding provided through a Graduate Research in Engineering and Technology (G.R.E.A.T.) Award held from 1988 to 1991, and Shell Canada Ltd. for their substantial contributions to research costs over this same pe-riod. Funding for research costs was also provided by Energy, Mines and Resources Canada through Research Agreements 88-04-225, 89-04-222 and 90-04-61 (EMR/NSERC). Addi-tional funding was obtained through NSERC Operating Grant OGP-7707 to R.M. Clowes. My sincere thanks to the members of the Department of Geophysics and Astronomy for making this such an enjoyable place to study, and to friends and family for years of encouragement and support. xiii CHAPTER 1 PRELUDE 1.1 Introduction Geophysical methods provide the interpreter with many tools with which to remotely probe the interior of the Earth. Each method evaluates a particular feature of the Earth's composition and allows one to propose a structure that could give rise to that feature. In most cases, however, the derived structural interpretation is highly nonunique and it is only by combining methods and interpretations that one can hope to achieve the unique solution. Nettle ton (1971) compared the geophysical exploration problem with a detective story: "In a well-constructed story the "whodunit" has many possibilities at the beginning. As more and more clues are developed the number of possibilities becomes more and more limited until finally only one remains. In the exploration analog, the "criminal" to be found and identified is the true geological situation. With a single exploration method (surface geology or one of the geophysical methods) the number of geological possibilities is relatively large. As more "clues" are brought to bear by application of more geological and geophysical techniques the number of possibilities is reduced until finally, hopefully, the actual geological situation is revealed by finding the set of geological circumstances which is consistent with all the information available." (Nettleton 1971, p. 117). The objective of this study is to search for the "most consistent" geotectonic represen-tation of the northern Cascadia subduction zone by combining and integrating an extensive 1 CHAPTER 1. PRELUDE 2 collection of geophysical data consisting of gravity, magnetic, seismic refraction and reflec-tion, surface geology, heat flow, magnetotelluric, borehole and seismicity data. The study area encompasses Vancouver Island and the offshore continental margin, which includes a large accretionary complex of sedimentary material and allochthonous terranes emplaced above the subducting plate. The result derived from this study is a regional structural model that provides information about the geometry of the subducted plate, the orientation and lateral extent of the accreted terranes, and the location of boundaries separating the major tectonic and stratigraphic units. 1.2 Tectonic Overview The Cascadia subduction zone extends northward, along the western continental margin of North America, from approximately 40°N to 51°N latitude and involves the subduction of the oceanic Gorda, Juan de Fuca, and Explorer Plates beneath the continental North American Plate (figure 1.1). These oceanic plates are the remnants of the large Farallon Plate, which has been converging in a predominantly northward direction with the North American Plate for at least the past 100 million years (Riddihough 1984). During this time the margin has been oriented in a generally north to northwest direction such that the plate interaction has been episodically expressed as both northeast convergence and northwest transform motion or oblique convergence, allowing exotic material to be brought in, accreted to the margin, and then sheared northward (Riddihough 1982b). Present day plate motions relative to North America are approximately 47 mm/a at N56°E for the Juan de Fuca Plate, 21 mm/a at N50°E for the Explorer Plate, and either 42 mm/a northward (Riddihough and Hyndman 1990) or 33 mm/a northeastward (Engebretson et al. 1985) for the Gorda Plate. Left-lateral strike slip motion of 30 mm/a has been estimated for the Nootka Fault Zone which separates the Juan de Fuca and Explorer Plates (Hyndman et al. 1979). Figure 1.1: Geometry of the Cascadia subduction zone and related features {after Hyndman et al. 1990; Gabrielse and Yorath 1989). The boundaries of the major tectonic belts in the northern part are indicated by dashed lines. The box encloses the area of this study. CHAPTER 1. PRELUDE 4 This study focuses on the northern part of the subduction zone from 123°W to 130°W longitude and north of latitude 48°N. These boundaries enclose Vancouver Island, part of the Coast Plutonic Complex (Coast Belt) on the mainland of British Columbia, and the northern edge of the Olympic Peninsula of Washington State. The present day continental margin represents the western limit of a large group of "suspect terranes" that accreted to the North American craton during the Mesozoic and early Cenozoic eras and now compose more than two-thirds of the North American Cordillera (Coney et al. 1980). The study area is dominated by the Insular Superterrane, the Coast Plutonic Complex (CPC) which straddles the Insular - Intermontane Superterrane boundary, and the offshore continental margin. The collisional model of emplacement suggests that the Intermontane Superterrane was already attached to North America when the Insular Superterrane accreted during the Middle Cretaceous in an ocean-continent subduction zone, closing the ocean basin and creating the CPC at the collisional suture zone (e.g. Monger et al. 1982). An alternate explanation argues that the two superterranes were part of the same "megaterrane" that accreted to North America during the Middle Jurassic and was broken into two fragments corresponding to the superterranes during a Late Jurassic rifting stage. The intra-terrane rift zone and associated marine basins closed during the Late Cretaceous and were partially overprinted by the CPC (van der Heyden 1989). Within the study area, the surface contact between the CPC and the Insular Superterrane occurs near the eastern edges of the Strait of Georgia and Johnstone Strait (figure 1.2). Vancouver Island is almost totally composed of Insular rocks, the exceptions being the southern tip of the island and a narrow strip along the west coast. The pre-Tertiary rocks along the south and west coasts may have been removed from the margin by dextral transform motion during the Late Cretaceous (Johnson 1984). The southern tip of the island was formed by the emplacement of the Pacific Rim and Crescent Terranes against and beneath the Insular Superterrane in the Early Eocene. Outcrops of Pacific Rim Terrane rocks also CHAPTER 1. PRELUDE 5 Figure 1.2: Generalized physiography and tectonic features of the study area. Oceanic plates are labelled as: E X = Explorer Plate, JF = Juan de Fuca Plate. Faults are abbreviated as: SMF = Survey Mountain Fault, SJF = San Juan Fault, WCF = Westcoast Fault, L R F = Leech River Fault, TF = Tofino Fault, and HRF = Hurricane Ridge Fault. Locations of exploratory wells on the continental shelf are marked (solid circles). Linear features in Winona Basin denote ridges as labelled. The inferred boundary between Winona Basin and the Explorer Plate is shown (marked as - ? - ) . CHAPTER 1. PRELUDE 6 occur along the west coast of Vancouver Island (e.g. Barkley Sound to Clayoquot Sound, Brooks Peninsula) and the Scott Islands at the northern tip and may mark the former transform boundary. Crescent Terrane rocks outcrop in the Olympic Peninsula of Washington State and are believed continuous beneath Juan de Fuca Strait (MacLeod et al. 1977). Boreholes on the continental shelf west of Barkley Sound (Shouldice 1971) mark the northernmost mapped extent of this terrane. The modern accretionary complex has been steadily building against the margin since the Eocene, attaining a width of over 100 km west of Barkley Sound. The wedge narrows north of the Nootka Fault, over the Explorer Plate, to just 60 km width southwest of Brooks Peninsula. The Explorer Plate was formerly the youngest part of the Juan de Fuca Plate but has been moving independently for the past 4 Ma (Riddihough 1984). A large fore-arc basin, extending from Brooks Peninsula to Juan de Fuca Strait, has formed above the subsiding wedge sediments. Exploratory wells and continuous seismic profiles have shown that Torino Basin contains in excess of 4000 m of Tertiary sediments above a basement of deformed Mesozoic wedge material (Shouldice 1971; Tiffin et al. 1972). Another major feature of the subduction zone is the accompanying volcanic arc. Episodic igneous activity has been associated with the margin since Late Paleozoic time and orogenic volcanism dates back to the latter half of the Mesozoic. The modern volcanic chain, the High Cascade Range, began forming during the Pleistocene (McBirney 1978). It consists of a series of large composite andesite volcanoes that extend from northern California to southern British Columbia about 300 km inland from the subduction trench (figure 1.1). 1.3 Major Geologic Units This study will derive structural models of the margin by utilizing two physical properties, namely density and magnetic susceptibility, of the rocks and sediments that compose the CHAPTER 1. PRELUDE 1 major geologic units. It is useful, and necessary, to have a basic understanding of the composition and distribution of each of the major groups and formations that compose the island and neighbouring regions. The generalized geologic map of Vancouver Island is shown in figure 1.3. Vancouver Island geology is dominated by rocks of the Wrangellia Terrane and extensive igneous intrusions. This section briefly summarizes the composition and distribution of the major geologic units. 1.3.1 Wrangellia Terrane Wrangellia Terrane extends from south-central Alaska along the Pacific margin through Chichagof Island, the Queen Charlotte Islands, and Vancouver Island, and may also occur in the Hells Canyon region of eastern Oregon and western Idaho (Jones et al. 1977). It is believed to have moved to its present location from a former position of 15° north of the paleo-equator during the Middle Cretaceous (Smith and Tipper 1985). On Vancouver Island the terrane is characterized by igneous sequences of Devonian, Upper Triassic and Lower Jurassic age separated by associated Late Paleozoic and latest Triassic sedimentary successions. Together they represent the superposition of marine volcanic arc, oceanic rift, and marine to subaerial volcanic arc assemblages (Yorath et al. 1990). Sicker Group The Sicker Group is of Middle to Late Paleozoic age and was formed in a marine volcanic arc setting on oceanic crust (Muller 1977a). The group is exposed in narrow, fault-bounded structures: the Buttle Lake Uplift, the Cowichan - Home Lake Uplift, and the Nanoose Uplift (figure 1.3). Three subdivisions were originally used to group the rocks: (a) a lower volcanic formation containing 1000 to 3000 m of breccia, tuff, and basaltic to rhyolitic flows; (b) a middle greywacke-argillite formation containing 600 m of thin graded beds of argillite CHAPTER 1. PRELUDE Figure 1.3: Generalized geologic map of Vancouver Island (after Muller 1980b). CHAPTER 1. PRELUDE 9 and siltstone and thick beds of greywacke sandstone; and (c) an upper limestone forma-tion (Buttle Lake Formation), up to 760 m thick, containing at least 320 m of interbedded crinoidal limestone and chert (Muller 1977a, 1977b). A new grouping and naming of subdi-visions subsequently proposed by Muller (1980b) divided the Sicker Group into the Nitinat Formation, Myra Formation, Sediment-Sill Unit (transitional), and Buttle Lake Formation. Vancouver Group Karmutsen Formation The thick tholeiitic basalts of the Middle to Upper Triassic Karmutsen Formation are one of the major unifying and identifiable members of the Wrangellia Terrane. The volcanics formed on the rifting marine arc, first creating a broad oceanic plateau of pillow lavas in a deep rift basin and then infilling the basin with aquagene tuff and breccia. Massive subaerial flows complete the succession. The Karmutsen Formation has also been subdivided into three parts: (a) a lower member of pillow lava, 2600 m thick; (b) a middle member of broken and whole pillow breccia and well-bedded aquagene tuff, 610 to 1070 m thick; and (c) an upper member of massive flows with minor interbedded pillow lava, breccia and sedimentary layers, 2900 m thick (Carlisle and Susuki 1974). Quatsino Formation The Quatsino Formation conformably overlies the Karmutsen Formation and represents a shallow-water or carbonate-platform deposit (Jones etal. 1977). The unit contains primarily light grey massive limestone of an average 300 m thickness. It is coarse and richly bioclastic with fossils of Upper Triassic (Karnian, -227 Ma) age (Carlisle and Susuki 1974). CHAPTER 1. PRELUDE 10 Parson Bay Formation The Quatsino limestones are succeeded by the Parson Bay Formation which formed in lower energy near- and off-shore basins (Muller 1977b). The formation begins with black laminated siliceous limestones and carbonaceous shales of upper Karnian age. Interbedded calcareous grey wacke and sandy to shaly limestone of Upper Triassic (Norian, -214 Ma) age form the upper Parson Bay Formation (Carlisle and Susuki 1974). The formation thickness ranges from 300 to 600 m on Vancouver Island (Muller 1977b). Bonanza Group The Bonanza Group consists of thick accumulations of lava, tuff and breccia with intercalated beds and sequences of marine argillite and greywacke. The group is exposed on the southwest and northwest parts of Vancouver Island and a section thickness exceeding 2500 m has been measured in the northwest (Muller 1977b). The deposits represent a return to arc development and explosive volcanism in the Early Jurassic (Yorath et al. 1990) and are the final depositional element in the composition of Wrangellia. Island Intrusions The Island Intrusions include both batholiths and stocks of Early to Middle Jurassic age granitoid rocks that intrude Sicker, Vancouver and Bonanza Group rocks. The intrusives range from quartz diorite to granite (Muller 1977b) and are exposed primarily on west-central Vancouver Island. Metamorphic Complexes The Westcoast Crystalline Complex is exposed along the west coast of Vancouver Island from Barkley Sound to Brooks Peninsula (figure 1.2). The complex is a heterogeneous assemblage CHAPTER 1. PRELUDE 11 of gneiss, amphibolite, agmatite, and quartz diorite or tonalite. It is believed to have been derived from Paleozoic and Mesozoic supracrustal rocks and early solidified magmas, and converted in a deep Jurassic magmatic arc environment to migmatites and metamorphic rocks (Isachsen 1987). The complex may also be interpreted as the deeper crustal equivalent of the more differentiated Island Intrusions and Bonanza volcanics. The mafic Wark Gneiss and silicic Colquitz Gneiss together compose a metamorphic complex exposed near Victoria on Vancouver Island. Wark Gneiss is a dark, massive to well foliated hornblende-plagioclase rock. The lighter coloured Colquitz Gneiss is irregu-larly foliated biotite-quartz-plagioclase gneiss, in places interlayered with Wark-type bands (Muller 1983). The Wark-Colquitz Complex is essentially indistinguishable from the West-coast Complex and a similar history is inferred (Isachsen 1987). 1.3.2 Pacific Rim Terrane The Pacific Rim Terrane comprises a thick sequence of Lower Cretaceous olistostromal melanges (sedimentary chaotic deposits of mixed heterogeneous materials) and a base unit of Lower Mesozoic arc-volcanic rocks. Pacific Rim rocks are exposed as a narrow series of fault-bounded outcrops along the west coast of Vancouver Island (Pacific Rim Complex and Ucluth Formation) and as a wider belt on southern Vancouver Island (figure 1.2) between the San Juan - Survey Mountain and Leech River Faults (Pandora Peak Unit and Leech River Formation). The terrane was previously thought to be a subduction complex that formed along the western margin of Wrangellia during the Late Mesozoic (e.g. Muller 1977a; Jones et al. 1977). Fairchild and Cowan (1982) favoured eastward emplacement of the Leech River Formation by left-lateral motion along the San Juan Fault. Others (e.g. Johnson (1984); Brandon (1989b)) subsequently proposed that the terrane was offset from the San Juan Islands region after formation and emplaced by right-lateral strike-slip faulting along CHAPTER 1. PRELUDE the San Juan and Westcoast Faults during the latest Cretaceous or Early Tertiary. Pacific R i m Complex The sediment-rich melanges of the Pacific Rim Complex have been divided by Brandon (1989a) into three types: (a) a 600 m thick matrix of interbedded mudstone, chert, sand-stone, and green tuff surrounding exotic blocks of igneous rocks and limestone; (b) 1000 - 1500 m thick mudstone-rich depositional sequences with interbedded mudstone, turbidite sandstone, conglomerate, pebbly mudstone, and rare chert; and (c) a 500 - 1500 m thick sandstone-rich depositional sequence of thick-bedded and massive turbidites, which grada-tionally overlies the other two laterally-related melanges. Total complex thickness as exposed on the west coast exceeds 2 km. The melanges underwent low-temperature, high-pressure metamorphism after Lower Cretaceous formation and prior to Early Eocene plutonism (Bran-don 1989a). The complex is bounded on the northeast side by the Westcoast Fault, which probably formed sometime during the latest Cretaceous or Early Tertiary, at least 45 Ma after formation of the melanges, and on the west side by the Eocene Torino Fault (see figure 1.2). Ucluth Formation The Lower Cretaceous melanges of the Pacific Rim Complex are in depositional contact with the underlying Lower Mesozoic Ucluth Formation. The formation is composed primarily of Upper Triassic unstratified fragmental volcanic rocks, interbedded limestone, and small diorite stocks and dikes (Brandon 1989b). A group of Lower Jurassic volcanic rocks (fragmental tuff, chert and minor pillow lava) has been tentatively included. Brandon (1989b) favours a calc-alkaline volcanic origin for the Ucluth Formation. CHAPTER 1. PRELUDE 13 Pandora Peak Unit The sedimentary and volcanic Pandora Peak Unit is juxtaposed against the fault-truncated igneous and metamorphic rocks of Wrangellia at three small outcrops on southern Vancouver Island. The unit is composed primarily of black mudstone or argillite, greywacke, radiolarian chert, tuff, greenstone, and minor pebbly mudstone and limestone blocks, and probably formed on the innermost part of a submarine fan or on the lower slope (Rusmore and Cowan 1985; Yorath et al. 1990). Stratal disruption and deformation occurred prior to complete lithification and low-temperature, high-pressure metamorphism occurred after formation and prior to emplacement against Wrangellia. The Late Jurassic - Lower Cretaceous Pandora Peak Unit is lithologically and temporally correlated with the Pacific Rim Complex of western Vancouver Island and the Constitution Formation of the San Juan Islands (Rusmore and Cowan 1985). Major faults separate the Pandora Peak rocks from the more highly deformed metasedimentary rocks of the Leech River Formation to the south. Leech River Formation The Leech River Formation of southern Vancouver Island comprises metamorphosed and deformed sedimentary and volcanic rocks of Late Jurassic to Cretaceous age. Muller (1983) divided the formation by lithology and decreasing age into three units which may represent the original stratigraphic sequence: (a) a volcanic-rich unit of sheared metavolcanics, schist, ribbon chert, and argillite which lies adjacent to the San Juan and Survey Mountain Faults; (b) a unit of thinly-bedded greywacke and argillite, converted to schist by shear folding, which underlies much of the area between the San Juan and Leech River Faults; and (c) a metagrey wacke unit which includes thick metasandstones and is exposed in east-west strips of a few kilometres width. The formation may represent a series of turbidites with associated minor volcanic rocks and thick metasandstone bodies from former distributary channels CHAPTER 1. PRELUDE (Rusmore and Cowan 1985). Two stages of deformation in the Early Tertiary produced large and small scale folding, faulting, and transposition of layering. High-temperature, low-pressure metamorphism began during the first phase of deformation and ended shortly after the second (Muller 1983; Yorath et al. 1990). Both deformation and metamorphism increased in intensity from the north toward the Leech River Fault at the southern boundary. Intrusion of quartz diorite to granite bodies preceded and accompanied deformation and metamorphism (Rusmore and Cowan 1985). Potassium-argon data yield an age of 39—41 Ma for the completion of deformation, synkinematic metamorphism and related intrusive activity (Fairchild and Cowan 1982). 1.3.3 Nanaimo Group The Upper Cretaceous Nanaimo Group comprises five transgressive clastic, upwardly-fining sedimentary cycles that were deposited in a forearc basin between the active volcanic arc of the Coast Plutonic Belt and the Insular Belt (Muller 1977b). The first four cycles have each been divided into a lower, fluvial to deltaic formation of conglomerate, sandstone, shale, and coal, and an upper, marine formation of marine sandstone, shale and siltstone (Muller 1977b, 1983). The fifth cycle contains a deltaic formation only. Nanaimo Group rocks are exposed on the east side of Vancouver Island and on adjacent Gulf Islands. Their structure is characterized by a few northwest-trending open synclines and anticlines separated by steep, northwest-striking normal and/or strike-slip faults and several northeasterly cross-faults (Muller 1983). CHAPTER 1. PRELUDE 15 1.3.4 Crescent Terrane The Early Eocene volcanic Crescent Terrane includes the Metchosin Igneous Complex of southern Vancouver Island, the Crescent Formation of Washington, and other correlative vol-canics of the Coast Range Basalt Province in southwestern Washington and Oregon (Yorath et al. 1990). The tectonic setting of the terrane has received considerable debate and both spreading ridge and intraplate seamount chains have been considered in addition to a marginal basin setting. The following history of formation based on a marginal basin setting was pre-sented by Massey (1986, 1990). The basin was opened behind the trailing edge of the Kula Plate block that moved northward past Vancouver Island and truncated the west coast along a transcurrent fault. The Coast Range basalts formed as new crust within this basin until ex-tension stopped around 52 Ma ago, although basinal sedimentation and intraplate volcanism continued past this date. A shift in Pacific - North American relative plate motions about 43 M a ago established the present convergent regime and the Crescent Terrane was detached from the Kula Plate and emplaced beneath the continental margin. Crescent Formation The Crescent Formation outcrops in a horseshoe-shaped curve on the Olympic Peninsula of Washington. The basalts are dominated by tholeiitic compositions with minor amounts of alkali basalt also present. Three mappable units compose the formation (Glassley 1974): (a) an upper basalt member of columnar to massive flows of 1 to greater than 10 m thickness, with occasional pillow basalts at the base of the flows and no intrusive rocks; (b) a lower basalt member of very dense, closely packed pillow basalts (60% by volume), massive basalts and intrusive basalt, diabase and gabbro dikes (20%), volcanoclastic rocks (15%), and the remainder chert, limestone and fine-grained sediments; and (c) a sedimentary member containing fine-grained sediments, sandstones and rare conglomerates, commonly sheared CHAPTER 1. PRELUDE 16 to phyllitic grade. The upper basalt member is unsheared and unmetamorphosed while the lower is intensely sheared and post-kinematically metamorphosed. A zone of fractured pillow basalts in a matrix of fragmented basalt represents the tectonic contact and metamorphic facies boundary between the two basalt members. Glassley (1974) inferred that the lower basalt member represents upper oceanic crust which was subducted during the Oligocene to at least 4 km depth, faulted away from the remaining oceanic crust, and then juxtaposed through faulting against the upper member extrusive basalts during Miocene uplift of the Olympic massif. Metchosin Igneous Complex The Metchosin Igneous Complex covers southernmost Vancouver Island between the Leech River Fault and Juan de Fuca Strait. Three components comprise the complex: (a) basaltic volcanics (Metchosin Volcanics) of tholeiitic composition, divided into a 1500 m thick lower submarine unit of pillow lavas with interbedded pillow breccias and aquagene tuffs, and an upper subaerial unit, 1000 m thick, of layered, commonly amygdaloidal basalt flows; (b) a sheeted dike complex, of perhaps 300 - 500 m thickness, of dominandy fine- to medium-grained diabase but ranging to feldspar diabase and fine-grained basalt; and (c) gabbroic stocks (formerly Sooke Gabbro) of isotropic and layered gabbros with minor stocks of quartz diorite and tonalite (Muller 1977c, 1980a; Massey 1986, 1990). The complex has been interpreted as a partial ophiolite and the combined 2 km thick submarine volcanics and sheeted dike portions are comparable in thickness with ocean crust in other ophiolites (Muller 1977c). An average gabbroic layer thickness of 4.8 km has been determined from seismic studies of lower oceanic crust and measurements of ophiolites (e.g. Christensen and Smewing 1981). Christensen and Salisbury (1975) further report that the layer may be as thin as 3.0 km at the time of formation and may gradually thicken over a 40 Ma period. If CHAPTER 1. PRELUDE 17 this age-thickness relationship holds for the Metchosin Igneous Complex, it implies that an upper mantle peridotite layer could be lying close to the surface beneath southern Vancouver Island. 1.3.5 Carmanah Group The Carmanah Group represents the entire sequence of Tertiary strata which overlie south and west Vancouver Island. Late Eocene to Pliocene rocks of the Carmanah Group overlie deformed Mesozoic metamorphic, volcanic, plutonic, and sedimentary rocks, as well as the Early Tertiary Metchosin Igneous Complex, and represent the outcrop limit of the seaward-thickening wedge of sediments which underlies Torino Basin (Yorath 1980). Correlative strata outcrop in the uplifted Olympic Core of western Washington. The group has been divided into three upward successive formations: (a) the Escalante Formation, containing approximately 140 m of sandstone with minor conglomerate lenses; (b) the Hesquiat Formation, containing 1100 m of interbedded grey shale, sand shale and laminated sandstone which has been stacked up to 3500 m thick in some outcrops; and (c) the Sooke Formation of pebbly conglomerate, coarse gritty sandstone, and volcanic, intrusive, shale, and sandstone pebbles (Muller et al. 1981). 1.4 Overview of Procedure This study uses two physical properties of the major geologic units, namely density and magnetic susceptibility, to derive structural models for the margin from regional gravity and magnetic anomaly data. The initial models are located coincident with major offshore deep seismic reflection and refraction lines to maximize constraints on crustal structure. The interpretations extend outward from the initial models in a "slice-wise" manner, making use of all available geophysical or geological constraints. These colinear model slices are then CHAPTER 1. PRELUDE combined to form a mree-dimensional structural model that is representative of the margin. CHAPTER 2 PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 2.1 Data Acquisition and Preparation Gravity data covering the study region were obtained in digital format and as maps from the Geophysical Data Centre of the Geological Survey of Canada. The digital data included complete Bouguer and free-air anomalies, corresponding bathymetry or elevation, and ge-ographic location for each of 16375 scattered points. Point spacing averaged 3 km along marine survey lines and 10 km between lines and land stations. A limited magnetic data set, at the time only providing coverage over the seafloor and western part of Vancouver Island, was obtained from the Pacific Geoscience Centre, Energy, Mines and Resources Canada (R.G. Currie 1990, personal communication). The total-intensity magnetic anomaly data, consisting of marine and aeromagnetic data reduced by the applicable International Geomag-netic Reference Field (IGRF) and continued to sea level and merged, had been gridded as an irregularly-shaped array of 28668 data points with grid cells measuring 0.025° longitude by 0.02° latitude (approximately 1800 m x 2200 m). In order to prepare the data for digital processing it was first necessary to develop an appropriate grid system and interpolate both the scattered gravity data and the magnetic data onto the grid. The ellipsoidal Transverse Mercator (TM) projection was chosen for presenting the fairly large (3°x7°) map area as it balances conformality, i.e. the preservation of angles and the shapes of small objects, with the minimization of linear distortions and hence preserves angles and scale over much of the map area. The result is a map with two 19 CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 20 transverse curves, approximately 180 km east and west of a central meridian, along which the principal scale is preserved and distortion is zero. Reductions in scale increase away from the standard parallels; the maximum error is along the central meridian where scale is 0.9996 of true scale, an error of 1 m in 25 km (Maling 1973). The T M projection for this study area was constructed using the Clarke 1866 ellipsoid (Snyder 1983) and a central meridian of 126.5° W, the longitudinal centre of the study area. The lines of true scale are approximately 124° and 129° W longitude. Transforming geographic locations in latitude / longitude pairs to the T M projection gives x (east-west) and y (north-south) coordinates in metres. The point (126.5° W, 0° N) corresponds to (0 m, 500000 m) and values increase to the east and north. Spacing between grid lines is dependent on the original sampling intervals of the data sets, and values of 3000 m and 1500 m were chosen for the gravity and magnetic data, respectively, to minimize spatial aliasing. A quadratic weighted averaging function was used to interpolate the scattered data onto the grid points. The function is given by r( N YJtLim(x,y)gi(xi,yi) 1 fix, y) = — = 3 ? > mix, y) = , 2 where (Xi,yi) are the original data locations relative to grid point (x,y) and gixi.yi) is the original data set (Sprenke 1982). The five scattered data points closest to each grid point, within a maximum distance limit, were weighted by their distances from the grid point and used to calculate the value assigned to the point. If a scattered data point exactly corresponded to a grid point location it was used instead of the averaged value. Grid point values based on fewer than three qualifying data points were rejected and not used in contouring or other calculations. The gridded data were either line contoured, using a routine based on CONTUR (U.B.C. Computing Centre) or contoured by matching a colour to each contour interval and filling grid spaces with the appropriate colours. The completed line contour maps were compared with maps plotted by the Geophysical Data Centre for accuracy and CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 21 no discrepancies were noted. The coastal outlines for Vancouver Island and the adjacent mainlands of British Columbia and Washington were obtained in digitized format from the World Data Bank II (WDBIJ) files, again through the Geophysical Data Centre, and superimposed on the data maps. The plotting routine developed for the study also allowed for additional information, such as seismic line and borehole locations, to be plotted and labelled on the maps as required. Sources of error associated with the final contoured potential field data could have been introduced at three stages: first, during data acquisition and subsequent reduction to anomaly stage, including instrumentation and measurement errors; second, during the gridding stage by application of the weighted averaging function; and third, during contouring, which is highly dependent on the algorithm used. A fourth source of error could be introduced later when profiles were manually obtained from the contour maps and great care was taken with these readings to minimize such errors. A simple test was made to compare the weighted averaging function with other methods of gridding the data. A profile obtained from a contour map gridded with the weighted averaging function was compared with coincident profiles interpolated from the original scattered gravity anomaly data (B. Isbell 1990, pers. comm.) using a simple mean function, the weighted averaging function, and a polynomial interpolation. The contour map profile compared favourably with the other profiles and reproduced all of the features of the original data, with some smoothing of sharp peaks and troughs dependent on the contour interval of the map from which the profile was obtained. Therefore the weighted averaging method was deemed a suitable and rapid technique for gridding the data. The base map of the study area, showing bathymetry contours and physiographic features, is presented in figure 2.1. Locations of gravity and magnetic anomalies discussed in the following sections have been highlighted on this map. Both this map and the generalized geological map of Vancouver Island (figure 1.3) should be used for reference when examining CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 22 Figure 2.1: Map showing the locations of major physiographic features and potential field anomalies in the study area. CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 23 the following potential field maps. Larger size line contour maps of the gravity and magnetic anomaly fields are provided in Appendix A for examination of detailed features. 2.2 Gravity and Magnetic Field Maps 2.2.1 The gravity anomaly map The colour-contoured gridded gravity anomaly map is shown in figure 2.2. The free-air anomaly is used over the Pacific Ocean and the straits between Vancouver Island and the mainland, and the complete Bouguer anomaly is shown over all land masses. A Bouguer reduction density of 2670 kg/m 3 was used. A broad examination of the anomaly map reveals the northwestward-trending double-band pattern which characterizes the Cascadia subduction zone: a slightly positive anomaly seaward of the subduction trench, a negative anomaly over the trench, continental slope and shelf, divided in the southern region by a small positive anomaly over the shelf-slope break, a stronger positive anomaly over Vancouver Island, and a large negative anomaly beneath the Coast Mountains. Closer examination of the anomaly map allows several anomalous regions to be located and the probable sources identified. For example, the outlines of the continental shelf and slope, and hence the contribution of seafloor topography to the free-air anomaly, are easily identified, especially where the slope broadens toward the south. The effects of the topography associated with the offshore Nootka fault zone and the northern end of the Juan de Fuca spreading ridge are also visible in the southwest corner of the map. A large negative anomaly along the northern end of the shelf marks the position of Winona basin, which seismic refraction studies have shown to be filled with thick deposits of sediments (Clowes et al. 1981; Davis and Clowes 1986). The southern tip of Vancouver Island is marked by an intense positive anomaly which roughly outlines the mapped surface outcrop of the Metchosin volcanics of the Crescent terrane. Central Vancouver Island shows a decrease CHAPTER 2. PROCESSPNG AND ANALYSIS OF POTENTIAL FIELD DATA 25 in the anomaly to near-zero levels, and then the value rises again toward a local high off the northwestern tip of the island. A broader negative anomaly lies over the Fraser Delta west of the city of Vancouver. Finally, the Strait of Georgia between Vancouver Island and the British Columbia mainland is marked by a local anomaly high centred between Texada and Lasqueti Islands. This is probably associated with the vast iron ore deposit (primarily magnetite, average density 5120 kg/m3) that is currently being mined on Texada Island. 2.2.2 The magnetic intensity map The magnetic anomaly map is shown in figure 2.3. The data set shown represents the most complete reduced coverage available at the time of this study. The most noticeable feature on the map is the offshore stripe pattern associated with the oceanic crust. The gradual suppression of the pattern near the edge of the continental shelf is undoubtedly due to the increased burial depth of the oceanic crust beneath the thick sedimentary wedge. Several prominent magnetic highs are identifiable over the continental shelf. The largest of these parallels the west coast of Vancouver Island and extends from near the tip of the Olympic Peninsula of Washington to west of the Clayoquot Sound region (figure 1.2). This anomaly has previously been termed the "Prometheus anomaly" and is associated with the offshore location of the Crescent Terrane (MacLeod et al. 1977). A magnetic low separates this high from Vancouver Island. The low appears to continue onshore over the mapped Leech River schist of the Pacific Rim Terrane (figure 1.3). Landward of this magnetic low is another linear high that may be mapped from the San Juan Fault on southern Vancouver Island, across Barkley Sound and back onshore near the Westcoast Fault, and then continues northwestward up the coast, remaining offshore with minor exceptions, to pass the Scott Islands at the extreme northern tip of Vancouver Island (figure 1.2; see also figure A.2). This low-high couple may mark either the upper bounding fault of the Pacific Rim terrane CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 21 or the seaward extent of the diorite unit that has been mapped along much of the coast. Several long linear onshore magnetic highs also parallel the coast from Brooks Peninsula northwestward into Queen Charlotte Sound. Finally, a sinuous magnetic high on southeastern Vancouver Island correlates well with a mapped granodiorite body (Roddick et al. 1979) of the mid-Jurassic Vancouver Island Intrusives. 2.3 Filtered Potential Field Maps In an effort to extract additional information from the basic potential field maps, various filtering techniques were applied. This effort proved very useful as the resulting maps allowed successful mapping of faults and structural boundaries beneath the shelf and on Vancouver Island. Processing was carried out on a Microvax II minicomputer with high resolution colour monitor, part of the Geophysics Research Processing Facility at U.B.C., and all programs were written in Fortran. 2.3.1 Horizontal gradient of gravity The horizontal gradient of the gravity anomaly data was calculated to assist in delineating boundaries and edges of source bodies. The gradient terms are the rates of variation of the acceleration due to gravity, g, in the horizontal directions x and y and are two of the six second derivatives of the gravity potential V. (Note that g is the first derivative of V in the vertical direction and is the only non-zero first derivative.) The gradient function f(x, y) is determined as: CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 28 (Tsuboi 1983). The horizontal components of the gravity gradient may be estimated by finite differences as: 3g _ Si+ij-gi-ij 3g ^, gij+i -gij-i dx ~ 2 Ax ' dy ~ 2Ay for g(x,y) defined at grid point as gij (Cordell and Grauch 1985). The horizontal gradient will have maximum amplitude over rapid variations in the gravity field and hence the map display will show sharp peaks over abrupt lateral changes in mass density such as might occur across lithologic contacts where there is a density contrast. Figure 2.4 shows the horizontal gradient map for the study area. Several large linear peaks in the northern shelf region correlate well with the Paul Revere and Winona ridges of Winona Basin and the edge of the continental shelf. A lengthy linear feature parallels the coastline from Brooks Peninsula to Barkley Sound, at which point it divides into a southern strand leading into the Olympic Peninsula and a northern strand which crosses onto Vancouver Island in the approximate location of the Leech River fault. South of this a series of moderate peaks in Juan de Fuca Strait are attributed to a combination of the southern edge of the large gravity high of southern Vancouver Island and the local effects of bathymetry in the strait. On northern Vancouver Island the contact between the andesitic rocks of the Bonanza Formation and the higher density basaltic Karmutsen Formation shows up clearly as a northwest-southeast lineation approximately 30-40 km inland of the west coast. Other features on Vancouver Island appear to correspond with mapped faults and geologic contacts. The landward boundary of the Pacific Rim Terrane, marked by the West Coast Fault, is one of these features. Sharp peaks in the Strait of Georgia now characterize the previously noted area of positive gravity anomaly and close examination allows one to speculate that at least some of these linear features may mark the intrusive contact and mineralization zone with which the iron ore is associated. CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 30 2.3.2 Vertical derivatives Vertical derivatives of potential field data provide information about the curvature of the potential surface. The first vertical derivative, dg/dz, resolves lithologic boundaries since the vertical gradient is more sensitive to changes in rock densities occurring near the ground sur-face than to changes at depth. The second vertical derivative, cPg/dz2, emphasizes gradients and enhances high frequencies, allowing better resolution of anomalies of small areal extent. The vertical derivative filters were applied to the data in the wave-number domain. Trans-formation of the data from the spatial to the wave-number domain required prehandling of the data to minimize truncation and other transformation effects. First, the mean of the data was calculated, removed from the data, and stored for later use. Next, to avoid introducing short wavelength noise, the edges of the data set were tapered using a cosine bell of the form: reduced smoothly to zero over a taper width of 5 points which resulted in the outermost 12 km of the filtered gravity maps (given the grid spacing of 3 km) and 6 km of filtered magnetic maps being unsuitable for analysis. Padding of the data set to over twice its original dimensions in both the x and y directions completed the prehandling process and the data were transformed into the frequency domain using a Fast Fourier Transform (FFT) and unfolded. The magnetic field data were then reduced-to-pole, a procedure which removes the effects of the direction of magnetization and the direction of measurement and reduces the field to an equivalent vertically-acting field (Baranov 1957). The striped oceanic magnetic anomalies, which possess strong components of remanent magnetization, were not treated separately and hence were not included in the interpretation. Due to the steep angle of inclination of the present field in the study area ( 70°) the reduced-to-pole map (inclination where R, •max is the desired maximum filter length (Fuller 1967). The data values were CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 31 GRAVITY ANOMALY DATA MAGNETIC ANOMALY DATA 0 0.02 0.04 0.06 0.08 0.1 0 0.02 0.04 0.06 0.08 0.1 Wavenumber ( k m *) f a v e n u m b e r ( k m - 1 ) Figure 2.5: Radially-averaged log power spectra for (a) gravity data and (b) magnetic data. 90°) is very similar to the original map (figure 2.3) and need not be shown. The vertical derivatives were obtained by multiplying the transformed data by filters with the appropriate responses (Gunn 1975; Boland 1989). The filters were first optimized using Wiener filter theory, which involved separating the power spectra of the data into signal and noise contributions (Clarke 1969). The filtering algorithms of Boland (1989) were used to obtain the vertical derivatives. Radially-averaged power spectra for the gravity and magnetic data (figure 2.5) were examined to select white noise estimates. Values of 1.5 and 5.0 were chosen for the gravity and magnetic data respectively and used in the optimization term. After filtering, the data were refolded, transformed back to the spatial domain using an inverse FFT, and clipped back to the original dimensions before presentation in contour map form. Vertical derivative map (Gravity) CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA 33 The map in figure 2.6 shows the first vertical derivative of the gravity anomaly field. Features that are enhanced by this filtering technique are Winona Basin in the north, the edges of the continental shelf and slope, and the Coast Ranges. The faults associated with the Crescent Terrane on land and beneath the shelf, and the anomalous region in the Strait of Georgia are also highlighted. Use of this map in conjunction with the horizontal gradient map provides good information about most of the anomalous features in the area. Vertical derivative map (Magnetics) The first vertical derivative map for the magnetic field data (figure 2.7) heightens the contrast between peaks and troughs and hence appears to "sharpen" the anomalies. This is particularly evident for anomalies, such as the Prometheus, that parallel the west coast of Vancouver Island. Also, the offshore extensions of the anomalies associated with the Leech River and San Juan Faults are considerably enhanced. Second vertical derivative map (Magnetics) Figure 2.8 shows the second vertical derivative map of the magnetic field. While the anoma-lies on this map are slightly sharper than those in the first vertical derivative map (figure 2.7), the relatively higher noise levels interfere with and degrade the outlines of the anomalies. This map does not contribute significantly to the interpretation but illustrates the limits of this type of filtering. 2.4 Results of Potential Field Processing Processing the potential field maps has provided constraints on regional modelling by locating or extending faults and terrane boundaries, in particular the Westcoast Fault or Pacific Rim Terrane boundary. The horizontal gradient map (gravity) and the first vertical derivative map Figure 2.7: Map showing first vertical derivative of reduced-to-pole magnetic anomaly. Contour interval is 5 nT/km. CHAPTER 2. PROCESSING AND ANALYSIS OF POTENTIAL FIELD DATA (gravity and magnetics) have proven to be the most useful in this region. The information provided by these maps will be added to the constraints provided by other geophysical data which will be discussed in the next chapter. C H A P T E R 3 G E O P H Y S I C A L CONSTRAINTS 3.1 Constraints on Subsurface Structure Several comprehensive reviews of geophysical analyses of the northern Cascadia subduction zone have been published in recent years (e.g. Keen and Hyndman 1979; Couch and Rid-dihough 1989; Hyndman et al. 1990). An extensive data set has been compiled over the past three decades of geophysical investigation of the margin. The locations of several major studies and data sets are shown in figure 3.1. The results of these surveys have been used to constrain the models developed in this study, and this section will recount some of the major constraints. Gravity and magnetic anomaly studies will be considered in the section following. The results of these earlier studies will be discussed and re-evaluated in Chapter 5 in conjunction with the results of this study. 3.1.1 The ocean basin Several seismic surveys have been conducted over the ocean basin west of the accretionary margin. North of Brooks Peninsula, seismic refraction surveys and continuous seismic profil-ing have imaged the thick sedimentary sequences within Winona Basin. The sequences have been interpreted as high velocity, high density Pleistocene turbidites, and both refraction and gravity modelling support thicknesses of up to 6 km in the centre of the basin (Clowes et al. 1981; Davis and Riddihough 1982; Davis and Clowes 1986). Oceanic crustal thicknesses of approximately 8 km were also interpreted. 37 CHAPTER 3. GEOPHYSICAL CONSTRAINTS Figure 3.1: Location map of major geophysical studies used to constrain margin structure. Heavy solid lines are multichannel seismic reflection profiles, light solid lines are CSP sur-veys, dashed lines are refraction profiles, circles are boreholes. Numbered lines are onshore LITHOPROBE (84-) and offshore FGP (85- ) reflection profiles. CHAPTER 3. GEOPHYSICAL CONSTRAINTS A refraction study across the Nootka Fault zone detected crustal thicknesses of 6.4 to 11.2 km (including a 1 km-fhick sedimentary layer) for 1 to 3 Ma old crust, and a sharp 1.6 km increase in thickness northward across the fault zone (Au and Clowes 1982). An adjacent refraction profile, acquired in 1980 as part of the Vancouver Island Seismic Project (VISP), also found crustal thicknesses in excess of 11 km just south of the fault zone near the end of the Juan de Fuca spreading ridge (White and Clowes 1988, 1990). Two multichannel seismic reflection lines of the 1985 Frontier Geoscience Program (FGP) survey (lines 85-07 and 85-09) imaged the sedimentary sequences and underlying oceanic crust off southern Vancouver Island between the accretionary prism and the Juan de Fuca ridge. These lines revealed average crustal thicknesses of 6 to 7 km, with some pronounced lateral variations, and showed a general increase in depth-to-basement and sediment thick-ness eastward toward the trench (Hasselgren 1991). 3.1.2 The continental shelf and slope The depositional sequences underlying the continental shelf were sampled by six exploratory wells drilled between 1967 and 1969 by Shell Canada Limited (see figure 3.1). Up to 3000 m of Miocene to Recent marine clastic rocks were encountered, and two wells (1-65 and 1-87) reached Eocene sedimentary rocks (Shouldice 1971). The Zeus D-14 and Prometheus H-68 wells entered Eocene volcanics of the Crescent Formation at about 2200 and 1800 m respectively. Interval velocities were measured from copies (Riley's Reproduction) of borehole-compensated sonic logs for the six wells. These in turn provided density estimates for Tofino Basin strata and uppermost Crescent volcanics. Numerous continuous seismic profiles (CSPs) and multichannel reflection sections have been recorded over the continental shelf and slope west of Vancouver Island (see figure 3.1 for locations). Sediment thicknesses and depths-to-basement were estimated from CSPs CHAPTER 3. GEOPHYSICAL CONSTRAINTS 40 published by Davis and Seemann (1981) and Tiffin etal. (1972), and from seismic reflection sections and line drawings presented by Barr (1974) and Yorath (1980). Igneous basement depths calculated by Barr (1974) from magnetic anomaly data range from 3—4 km at the foot of the continental slope to greater than 10 km beneath the shelf-slope break. A 24-channel reflection profile acquired by the U.S. Geological Survey across the margin west of the Juan de Fuca Strait provided clear images of sediment thickness and structure beneath the shelf (Snavely and Wagner 1981). A characteristically rough reflector underlying a 10 km-wide portion of the shelf sediments was interpreted as the top of the volcanic Crescent Terrane. This interpretation was supported by a 150 nT anomaly observed on a coincident magnetic anomaly profile. The shelf and slope structure was sampled in detail by several multichannel seismic reflection lines acquired in 1985 under the Frontier Geosciences Program (FGP) of the Ge-ological Survey of Canada (Yorath et al. 1987). These high resolution data also imaged the oceanic crust underlying Cascadia Basin, at the foot of the continental slope, and the accretionary wedge, and provided important information on the structure of the accretionary prism. These lines were of primary importance in the development of of structural models in this study and their interpretation will be discussed in detail in Chapter 4. 3.1.3 The accretionary prism The velocity structure of the accretionary prism off southern Vancouver Island was derived in the offshore component of the VISP 80 refraction study (Waldron 1982; Waldron et al. 1990). Sediment velocities increased from 1.9 to 2.6 km/s landward, presumably due to compaction effects, and much higher velocities (> 4.8 km/s) were observed for the deeper part of the wedge. The calculated dip of the underlying Juan de Fuca Plate increased from 1° west of the margin to 2° beneath the outer slope and 8° beneath the middle slope. Taber CHAPTER 3. GEOPHYSICAL CONSTRAINTS and Lewis (1986) obtained similar results for the margin of Oregon. They modelled oceanic lithosphere dipping at 9° beneath an accretionary wedge in which velocities ranged from 2.4 to 6.0 km/s. Five multichannel seismic reflection lines of the 1985 FGP survey imaged structures within the accretionary complex and the top of the underlying oceanic crust. The sections showed tight folds and offset blocks marking the upper surface of the accretionary prism, and reflectors associated with faults leading toward the oceanic crust (Hyndman et al. 1990; see also Chapter 4). At the front of the wedge, thrust faults extending to the top of the underlying crust marked recent sediment deformation (e.g. Davis and Hyndman 1989). The top surfaces of the Crescent and Pacific Rim Terranes within the accretionary complex were also imaged on several of the lines (85-01, 85-02 and 85-05; see Hyndman et al. 1990). The depth extent of the terranes was indeterminate on the sections and various configurations and emplacement methods were proposed (e.g. Hyndman et al. 1990; Clowes et al. 1987b). 3.1.4 Continental crustal structure The continental crust beneath Vancouver Island was sampled by two refraction lines of the VISP 80 study (see figure 3.1). McMechan and Spence (1983) interpreted the 330 km-long profile extending the length of the island and constructed a velocity model for the crust through ray tracing techniques. The upper crust was laterally homogeneous except for a high velocity reflector apparently dipping northward from 15 to 20 km beneath the southcentral part of the island (figure 3.2a). Mantle velocities were reached at about 23 km depth beneath this feature. The lower crust was modelled as a negative gradient low velocity zone (7.0-6.2 km/s) to 37 km depth, beneath which mantle velocities (7.46 km/s) were encountered. This model was later revised to incorporate additional constraints provided by multichan-nel seismic reflection lines acquired on Vancouver Island in 1984 by LITHOPROBE (figure CHAPTER 3. GEOPHYSICAL CONSTRAINTS NORTH (a) N A 42 0-10-E 20-X 1-0. IU 30-Q 40-50-5.5" 6.5-X SOUTH F "5.5-- 6.5- 6.6 •7.5 '••'.'.;-7.8-- : ""7.6"' 6.2, 7.46 : 6.7S, 6.95 6.96 •••6.6 100 200 300 North DISTANCE (KM) South , . s N A E-W Una r \P) 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 0-X t— 0_ UJ Q (c) N r 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 I— CL UJ Q Figure 3.2: VISP 80 along-island seismic refraction line models. Velocities are in km/s. Shotpoints are marked N, A, F and X marks the intersection with the East-West line (see figure 3.1 for locations), (a) Model of McMechan and Spence (1983); solid boundaries indicate highest model reliability, dotted lines indicate poorest, (b) Model of Drew (1987); P marks the top of the subducting plate, (c) Model of Drew (1987) modified to correspond to a straight line from shotpoints N to F. {After Drew and Clowes 1990). CHAPTER 3. GEOPHYSICAL CONSTRAINTS 43 3.1). The revised model (Drew 1987; Drew and Clowes 1990) included a subducted oceanic plate at 35 km depth (39 km in the north), separated from the continental crust by a thin (< 3 km) wedge of mantle material (figure 3.2b). The upper crustal structure is similar to the previous interpretation but the lower crust was revised to include two thin (2-3 km) low velocity bands, correlated with bands of high reflectivity on the reflection profiles, and two wider (10 km) bands of high velocity. These four layers were considered to be part of the Cenozoic subduction complex underlying the 10 to 18 km-thick Wrangellia Terrane (Drew and Clowes 1990). When the model was corrected for the hinged geometry of the refraction line, the thin mantle wedge at 38 km depth was no longer required and a thicker low velocity band was used in the lower crust (figure 3.2c). The 350 km-long onshore-offshore VISP 80 refraction line extended from the deep ocean basin, across Vancouver Island and the Strait of Georgia, to Jervis Inlet on the mainland (see figure 3.1). The derived velocity model of Spence et al. (1985) incorporated the offshore model of Waldron (1982) and was constrained by the along-island interpretation of McMechan and Spence (1983). The model included a subducting oceanic lithosphere, which dipped at 3° beneath the outer melange unit (Waldron 1982) and increased to 14-16° beneath western Vancouver Island. The crustal structure was generally consistent with that derived for the along-island profile but included a wedge of material of mantle-type velocity (7.7 km/s) at approximately 20-25 km depth beneath western Vancouver Island and the continental shelf (figure 3.3(a)). Spence et al. (1985) speculated that this and the overlying low velocity sliver might represent the stranded remnant of a subducted oceanic slab. The adjacent negative gradient low velocity zone in the lower continental crust was interpreted as possibly being composed of pockets of subducted or underthrust material from a number of terranes. This model was also revised to include the more recent 1984 LITHOPROBE onshore reflection data (Drew 1987; Drew and Clowes 1990). The upper 15 km of continental crustal CHAPTER 3. GEOPHYSICAL CONSTRAINTS Figure 3.3: VISP 80 onshore-offshore seismic refraction line models. Velocities are in km/s. Bold arrows mark coastlines, (a) Model of Spence et al. (1985). The continental Moho is marked by M . (b) Model of Drew (1987). (After Drew and Clowes 1990). CHAPTER 3. GEOPHYSICAL CONSTRAINTS 45 structure were unaffected. However, the low velocity zone with negative gradient in the lower crust was replaced with a positive gradient low velocity zone and maginally higher average velocity; slightly lower velocities in the upper mantle were used to compensate (figure 3.3b). The imbedded block of 7.7 km/s material was replaced with a four-layer structure of two low velocity layers, corresponding to the two highly reflective zones, alternating with two high velocity layers to match the geometry indicated by the reflection section and the revised along-island model. Minor changes were made in the location of the top of the subducting plate to match the associated reflection but the average dip beneath western Vancouver Island remained unchanged at 14—16°. The refraction and reflection data were used as the basis for an interpretation exercise at a workshop of the International Association of Seismology and Physics of the Earth's Interior (IASPEI) Commission on Controlled Source Seismology held in 1987 (Green 1990). Several alternate models were developed that, in general, confirmed the major features of the original interpretations. This served to provide bounds on the range of structures that would satisfy the data. The 1984 LITHOPROBE experiment on Vancouver Island recorded a total of 205 km of deep seismic reflection data along four profiles (see figure 3.1). Line 84-01 crossed the island almost coincident with the VISP 80 refraction line and close to a gravity profile of Riddihough (1979). Line 84-03 was a shorter line to the southeast, and lines 84-02 and 84-04 crossed over the terrane-bounding faults on southern Vancouver Island. Two zones of high-reflectivity ( ' C and 'E ' ) dip across all sections approximately parallel to the direction of relative plate convergence (figure 3.4). Zone ' C dips from 11 to 20 km depth and is approximately 10 km thick. It is roughly coincident with the top of the high velocity block of Spence et al. 1985 and was interpreted as marking the top of a tectonically underplated zone (Clowes et al. 1987a). Zone ' E ' is 5 km thick and dips from 23 to 34 km depth at 9-13° northeast along most of line 84-01, although the dip may steepen at the northern end. CHAPTER 3. GEOPHYSICAL CONSTRAINTS 46 (a) sw LINE 84-01 N E 4-- 6 >- 8 < 5 12 a-. 5^ HIGH VELOCITY REGION E 2 - — VE=0.9 km - 3.0 6.3 9.5 12.8 16.2 19.8 23.7 o-LU 27.2 -30.5 33.7 36.9 40.8 44.4 48.0 52.0 Figure 3.4: Line drawings of LITHOPROBE 1984 multichannel seismic reflection sections. Reflections are labelled as: A = Survey Mountain Fault, B = Leech River Fault, C = Hurricane Ridge Fault, F = top of subducting plate. Zones of high reflectivity are marked as DI - D2 (or C - D2) and E l - E2. (a) Line 84-01. (b) Line 84-02. (c) Line 84-04. (From Clowes etal. 1987a). CHAPTER 3. GEOPHYSICAL CONSTRAINTS 47 It may mark a layer of accreted marine sedimentary and igneous rocks (Clowes et al. 1987a; Green et al. 1987), indicate fluids trapped at a metamorphic phase boundary within the lower continental crust (Hyndman 1988), or represent a shear zone developed in the crust above the subducting oceanic plate (Calvert and Clowes 1990). The 10-15 km thick poorly reflective zone between ' C and ' E ' may be composed of mafic or ultramafic rocks that were added to the underplated zone from the downgoing oceanic plate. Both episodic accretion of a single slice of crust and mantle, and ongoing accretion of imbricated slices of mafic rock derived from the top of the plate have been proposed as possible methods of emplacement (Clowes etal. 1987a). The reflection data also imaged the major terrane-bounding faults at depth (Clowes et al. 1987a). The Survey Mountain Fault appeared on line 84-02 as a low-angle, northeast-dipping fault extending to at least 10 km depth. The San Juan Fault was not imaged on line 84-04, which crossed over the surface position, and was interpreted by others as a high-angle structure which would be consistent with the lack of seismic expression (Hyndman et al. 1990). It may be a listric thrust that becomes parallel to or joins with the Leech River Fault at depth. Mayrand et al. (1987) applied special processing and migration techniques tO line 84-04 and tentatively interpreted a coherent event as a reflection from a very steeply dipping near-surface segment of the San Juan Fault. The Leech River Fault was imaged on lines 84-04 and 84-02 as dipping to the northeast at 35-45° to approximately 10 km depth at the northern end of the record sections. The acoustically transparent unit underlying the fault was interpreted as corresponding to the Metchosin basalts and the underlying fault as possibly the extension of the Hurricane Ridge thrust fault that underlies the Crescent Formation in the Olympic Peninsula (Clowes et al. 1987a). A band of enhanced electrical conductivity below 20 km depth was interpreted by Kurtz et al. (1986) to be approximately coincident with the ' E ' reflection zone (figure 3.5). Two-dimensional modelling by Kurtz et al. (1990) showed a conductive layer dipping to the CHAPTER 3. GEOPHYSICAL CONSTRAINTS 48 Figure 3.5: Cross-section along the LITHOPROBE line 84-01 corridor relating electrical conductivity (stippled band), high reflectivity (short lines), surface heat flow (data points on top curve), crustal temperature isotherms (solid curves), seismicity (dots) and subduction geometry (from Hyndman 1988). Heavy lines define oceanic crust and dotted line marks lithosphere—asthenosphere boundary. CHAPTER 3. GEOPHYSICAL CONSTRAINTS northeast from the Pacific Ocean to beneath Vancouver Island and in electrical contact with the mainland conductive region. The dipping high conductivity layer was attributed to the presence of saline fluid within the pore spaces of sedimentary and mafic materials of 3.6% porosity, or less for salinities higher than sea water (Kurtz et al. 1986). Hyndman (1988) estimated an average porosity of 1-2% in the conductive zone using the electrical model in combination with results of seismic reflection, refraction, heat flow and gravity. Heat flow measurements compiled by Lewis et al. (1988) revealed a low-high heat flow anomaly centred northeast of the Strait of Georgia (figure 3.5). The heat flow pattern averaged 110-140 mW/m 2 over the ocean basin, decreased to 50 mW/m 2 across the continental shelf, decreased further to 30 mW/m 2 about 30 km west of the volcanic arc, increased sharply over just 20 km distance to greater than 100 mW/m 2 approximately 20 km west of the arc, and maintained these values across the arc before eventually decreasing further inland to the somewhat uniform 75 mW/m 2 values that are characteristic of the Intermontane Belt (Davis and Lewis 1984). Assuming that heat is flowing conductively in the upper crust, the 10 to 15 km half-width of the sharp increase before the volcanic arc requires intrusive thermal sources at middle to upper crustal levels (Lewis et al. 1988). From the heat flow pattern a temperature model of the crust was developed (Hyndman 1988) and it was observed that under Vancouver Island the 450° C isotherm lies approximately parallel to and between the ' E ' zone and the Benioff-Wadati zone (figure 3.5). This isotherm marks the approximate boundary between blueschist and greenschist facies metamorphism (Hyndman 1988) and supports the hypothesis that metamorphic reactions are occurring along the ' E ' zone and may be buffering the 450° C temperature (Lewis et al. 1988). Blueschist conditions above this boundary could give rise to the high velocities observed on the refraction data (e.g. Spence et al. 1985) and high densities required by gravity data (e.g. Riddihough 1979). The 600° C isotherm approximately marks the depth limit beneath which materials lose their magnetic susceptibility (i.e. the Curie point). This occurs at approximately the CHAPTER 3. GEOPHYSICAL CONSTRAINTS 50 base of the crust beneath Vancouver Island and at much shallower depths (approximately 12 km) beneath the Coast Plutonic Complex. 3.1.5 The subducting oceanic lithosphere The position of the descending slab of oceanic lithosphere beneath Vancouver Island and the B.C. mainland has been estimated from seismicity studies (e.g. Rogers 1983, Rogers et al. 1990; Crosson and Owens 1987). A cross-sectional projection across Vancouver Island was obtained by Rogers et al. (1990) using microearthquakes that occurred between 1982 and 1989 within a 30 km-wide corridor centred on LITHOPROBE line 84-01. A shallow diffuse suite of events extending to 20-25 km maximum depth defines the continental crust The deeper band of earthquakes lies below 25 km depth beneath the west coast of Vancouver Island and dips an average of 20° beneath central Vancouver Island before steepening to reach 65 km below the Strait of Georgia (figure 3.5). Assuming that these earthquakes are occurring within the cooler, uppermost 5 to 10 km of the oceanic lithosphere, the events mark the position of the subducting slab beneath Vancouver Island. A similar cross-section constructed for the west coast of the island and continental shelf reveals a seismic zone dipping at 14° at 25 km depth beneath Barkley Sound (C. Dimate 1990, pers. comm.). This value is consistent with the 14—16° dip discussed earlier for the seismic refraction models. Crosson and Owens (1987) studied the seismicity of the northern Olympic Peninsula and southern Vancouver Island and calculated and contoured depths to the oceanic crust -mantle interface beneath this region. The contours clearly show an anticlinal eastward bulge in the subducted plate beneath the Olympic Peninsula that was attributed to the bend in the subduction zone. An additional constraint on the depth of the plate is provided by the positions of the Quaternary volcanoes, beneath which the plate is assumed to have reached an approximate depth of 100 km based on geochemical studies of the basalts (Dickinson CHAPTER 3. GEOPHYSICAL CONSTRAINTS 51 1970). 3.2 Previous Modelling Studies Figure 3.6 shows several profile locations along which gravity or magnetic anomaly profiles have been modelled in previous studies. Several qualitative analyses of the gravity and magnetic signatures of the margin have also been conducted (e.g. MacLeod et al. 1977; Arkani-Hamed and Strangway 1988). These have been useful in extending the quantitative interpretations derived in this study along the length of the margin. This section briefly reviews the main features of the previous models. 3.2.1 Gravity modelling One of the first quantitative gravity studies of the northern Cascadia subduction zone was conducted by Couch (1969). The model derived for a profile across the margin at southern Vancouver Island (figure 3.6) showed a 5 km-thick oceanic crust (excluding sediments) thickening to a 25-35 km-thick crust beneath the island and Coast Mountains. A subducting slab geometry was not considered in this study or a regional study of Dehlinger et al. (1970). Stacey and Stephens (1969) qualitatively reviewed the gravity anomaly field over Vancouver Island and noted correlations between high gravity values and Karmutsen rocks (sample mean p = 2960 kg/m3). Walcott (1967) examined the gravity high at the southern tip of Vancouver Island and concluded that only 30 mGal could be accounted for by the Eocene volcanic rocks (2880 kg/m3) and attributed the remainder (- 100 mGal) to a decrease in average crustal density south of the Juan de Fuca Strait and the probable presence of a major left-lateral fault beneath the strait. Stacey (1973) analysed a swath of data across the island and Canadian Cordillera (49 -51°N) and used the constraint of isostatic balance to construct models of density contrast. CHAPTER 3. GEOPHYSICAL CONSTRAINTS 52 Figure 3.6: Location map of previous gravity and magnetic modelling studies. Triangles denote volcanic peaks. Solid lines denote gravity profiles, dashed lines mark magnetic profiles. Abbreviated references are indicated: B H = Brown and Hanna (1971), C = Couch (1969), CC = Coles and Currie (1977), DR = Davis and Riddihough (1982), R = Riddihough (1979), S = Stacey (1973). Profile R* is shown in figure 5.3. CHAPTER 3. GEOPHYSICAL CONSTRAINTS 53 He expected from the density data that the Bouguer anomaly values would be lower over granitic batholiths and areas of sedimentary rocks rather than over metamorphic and volcanic rocks; this indeed proved to be the case. Stacey and Stephens (1969) had noted that local anomalies on Vancouver Island were commonly associated with density contrasts within surface rocks, and Stacey (1973) added that these were distorted by large regional gradients in the vicinity of the island. Sediments on the continental shelf corresponded with negative Bouguer anomalies, but any gravitational effect due to sediments at the foot of the slope was obscured by a rapid increase of the Bouguer values across the slope. A model for the Bouguer anomaly profile at 50°N across the Canadian Cordillera gave good agreement with observed gravity anomalies and seismic depths by allowing either thick (50+ km), dense (3090 versus 2920 kg/m3) crust, or thin crust (20 km) and low density mantle (3230 versus 3310 kg/m3) to exist beneath Vancouver Island. Stacey (1973) discussed the possibility that the low density material under Vancouver Island might be a remnant plate of oceanic lithosphere, but did not include a subducting slab in the models. Srivastava (1973) used rectangular blocks of density contrasts to model residual Bouguer anomalies across the margin from the southern tip of the Queen Charlotte Islands to northern Vancouver Island. A profile at the northern tip of Vancouver Island indicated a basement dip of 10° eastward beneath the west side of the offshore basin. Based on his gravity data models, and supported by the faulted nature of the margin and the folded and faulted sediment at the base of the continental slope off Vancouver Island, he favoured the possibility of oceanic basement dipping under the continent. Riddihough (1979) was the first to specifically use a subduction model to interpret the gravity signature of the margin. He satisfied the existing geophysical data (refraction, seis-micity, heat flow) using very simple two-dimensional structural models based on topographic, geological, and seismic control and adjusted using density and isostatic balance constraints. He also analysed the main contributions to the low-high couple that characterized the gravity CHAPTER 3. GEOPHYSICAL CONSTRAINTS 54 field. The high came from a combination of the ocean-continent 'edge effect' and the 20-30 km 'crustal' thickness in the arc-trench gap and was thus almost entirely the product of the geometry of the 'crust-mande' density boundary. This implied that, while the density of the oceanic crust should increase to mantle densities by 20-30 km depth, there was no need for abnormally high mantle densities in the descending oceanic crustal material. Instead, Riddihough (1979) used a wedge of material of 3300-3285 kg/m3 overlying the downgoing slab (see figure 5.3). The wedge was speculatively interpreted to consist of phase-changed oceanic crust and overlying crust or mantle affected by the unique arc-trench gap processes of hydration and pressure. Asthenospheric densities varied from 3295 kg/m3 beneath the oceanic section to 3300 kg/m3 beneath Vancouver Island and 3285 kg/m3 beneath the continental section east of the volcanic arc. Riddihough (1979) proposed a change in the dip of the downgoing plate based on seismic data and a slab depth of 100 km near the volcanic arc. He suggested that the upper portions of the plate dip at 5 to 13° and the lower portions at 20, 30, 40, and 50° respectively moving southward along the Cascadia margin. The position of the bend in the models was beneath the Strait of Georgia and Puget Sound, 90 to 160 km forward of the volcanic arc. — Following Riddihough's (1979) work, seismic data have played an important role in con-straining gravity interpretations. Clowes et al. (1981) converted their refraction velocity model for the deep-water sedimentary Winona Basin to a density model and successfully matched the observed -130 mGal free-air anomaly. Davis and Riddihough (1982) calcu-lated cross-sections based on numerous reflection sections to interpret three free-air residual profiles across the basin. They concluded that the crust beneath the basin was of uniform thickness (7 km), depth to basement was similar along the length of the basin, and the density in the basin (2200-2600 kg/m3) was symmetric except for the northwestern part. Spence etal. (1985) quantitatively modelled an observed gravity profile close to the VISP CHAPTER 3. GEOPHYSICAL CONSTRAINTS 55 onshore-offshore refraction line, utilizing the new constraints provided by the seismic model but otherwise adhering to the simple structure and density values of Riddihough (1979). They interpreted the main gravity high over Vancouver Island as possibly being due to imbedded pockets of high-density material in the crust, including a 9 km-thick sliver of high density, low velocity material beneath eastern Vancouver Island and part of the mainland. This feature caused a disagreement between the velocity model and the density model. The other sliver (7.7 km/s, 3280 kg/m3) between 20 and 25 km depth beneath western Vancouver Island agreed with both data sets. The sliver, interpreted as mantle material or possibly high velocity lower crustal material, could have been a large fragment equivalent to the detached slab of oceanic crust (Crescent Formation) beneath the shelf. Thus one of the main conclusions from previous gravity modelling was that special, relatively high density, low velocity material is required in some portions of the lower crust. This could be attributed to anomalous material formed under conditions of high pressure and low temperature in a hydrous environment above the subducting crust (Riddihough 1979) or to material formed elsewhere, transported to the region as part of an accreted terrane, and thrust under the stack of terranes already in place (Spence et al. 1985). 3.2.2 Magnetic modelling Very little quantitative modelling of magnetic anomalies has been done in the region, although several qualitative analyses have been published (e.g. Coles and Currie 1977; MacLeod et al. 1977; Arkani-Hamed and Strangway 1988). Brown and Hanna (1971) examined a northwest-trending profile over the Clallam syncline, northern Olympic Peninsula and Juan de Fuca Strait (see figure 3.6). They considered average total magnetizations, which are equal to the vector sum of the induced (magnetic susceptibility x external magnetic field) and remanent magnetizations. They concluded that, given the formation of the magnetic CHAPTER 3. GEOPHYSICAL CONSTRAINTS 56 rocks during an epoch in which the field was both normal and reversed and the probability that this resulted in mixed and hence random magnetizations, the average total magnetization of the Crescent Formation on a crustal scale was approximately parallel with the ambient field. The resulting model involved an asymmetric anticline of volcanic rocks of 0.0015 cgs units (0.0188 SI) susceptibility and thicknesses of 9000 m in the south limb and 4500-6000 m in the north limb. A horizontal mass of these rocks approximately 4500 m thick underlay Juan de Fuca Strait. A 70° northward dipping highly magnetized body beneath the strait (0.0030 cgs units (0.037 SI), 6000 m thick), was interpreted as a gabbroic body similar to the Sooke intrusives mapped on Vancouver Island. Coles and Currie (1977) developed quantitative models for an east-west trending smoothed profile crossing Vancouver Island and the Coast Mountains between 50 and 51°N. They found that the Insular Belt (i.e. Vancouver Island) was regionally magnetically quiet compared with the western portion of the CPC which was outlined on its western margin by strong (< 900 nT) positive anomalies. A broad scale qualitative agreement existed between higher residual magnetic field values and higher susceptibilities within the CPC. They also examined the natural remanent magnetizations (NRM) of many rock samples and concluded that stable "coherent remanent magnetization did not contribute significantly to the magnetic anomaly field. 3.3 Constraints on Density and Magnetic Susceptibility Measurements of density and magnetic susceptibility have been made by various researchers on numerous rock samples from the Insular Belt and Coast Plutonic Complex. Compilations of these measurements are presented in Appendix B . Average densities for the major litho-spheric and asthenospheric model components were obtained from previous gravity modelling studies, notably those of Riddihough (1979) and Couch (1969). Other estimates of crustal CHAPTER 3. GEOPHYSICAL CONSTRAINTS 57 densities were calculated by converting seismic refraction models to density using the rela-tionship of Nafe and Drake (1963) or Barton (1986). The compiled density measurements in Appendix B were used to temper the average upper crustal and sedimentary unit densities to correspond with mapped surface geology. C H A P T E R 4 P O T E N T I A L F I E L D M O D E L L I N G 4.1 Overview of Procedure A total of twelve two-dimensional density and magnetic susceptibility structural models were developed to explain the observed geophysical characteristics of the northern Cascadia subduction zone. A series of eleven models provided coverage, from the ocean basin to the volcanic arc, over a 230 km-long swath from south of Barkley Sound to Brooks Peninsula in the north. These models were subsequently interpolated to construct a block model of this segment of the margin. A northward trending twelfth model examined the source of the gravity high on southern Vancouver Island. The modelling proceeded in three parts. First, three primary models were constructed along seismic reflection/refraction profiles, providing well-constrained "anchors" for the re-gional model. Next, secondary models were developed by perturbing the base models to satisfy observed data profiles offset from the primary lines. Finally, a three-dimensional model "cube" was built from the set of 2-D models to provide a regional structural repre-sentation. 4.1.1 The modelling algorithm Model responses were calculated using the interactive Fortran program SAKI (Webring 1985). The program was modified (in collaboration with J.A. Amor) to be used on either the mi-cro V A X computer or the SUN workstations available in the Department of Geophysics and 58 CHAPTER 4. POTENTIAL FIELD MODELLING 59 Astronomy at U.B.C. , and to improve the interactive editing procedure. The program cal-culates and compares theoretical gravity and magnetic responses of a structural model with profiles of observed data. The degree of misfit is quantified by the calculated root-mean-square (rms) error, and the user can interactively modify the model to reduce the error. Each model was constructed as an ensemble of polygonal prisms of finite rather than infinite lateral extent, thereby forming a 2.5-dimension (2.5-D) model. Every polygon was defined by a density contrast, magnetic properties, and a series of polygon vertices; these were the param-eters that could be varied in each iteration. Density contrasts were referenced to a reducing density of 2.67 g/cm3 (2670 kg/m3), equivalent to the Bouguer reducing density at which the field maps were constructed. Magnetic susceptibilities were assigned and multiplied within the program by the Earth's field strength to obtain the induced magnetization. A remanent magnetization vector could also be added to the calculation. The strength, inclination, and declination of the Earth's field were calculated for each model position using the G E O M A G program (Bill Flanagan, National Geophysical Data Centre, NOAA) and the 1985 IGRF ref-erence model. This vector was then corrected within SAKI for the orientation of each model profile. A floating" datum level was used for comparing the calculated anomalies to the observed data. This program feature compensated for and removed any regional shifts that had not been considered, such as long wavelength (>600 km) anomalies due to very deep or broad scale structures. However, the dipping structure of the subduction zone produced a sloping regional anomaly that could not be removed by this method. Hence it was important to estimate, or model, as accurately as possible the orientation of the subducting plate. A display model length of 350 km was used to examine margin features from the ocean basin to the approximate position of the volcanic arc on the mainland. Similarly, a depth of 50 km included the subducting oceanic slab below the depth of the continental Moho to approximately the limit of model resolution of its position. The top surface of the model CHAPTER 4. POTENTIAL FIELD MODELLING was set to sea level. The models were extended to ± 2000 km length and 200 km depth, well beyond the range of interest, in order to reduce edge effects. Polygon widths (strike lengths) were similarly extended to a minimum of five times the depth of the bodies and, if possible, to match the mapped extent in the case of near surface bodies. The orientation of the models and anomaly profiles perpendicular to the main tectonic and structural elements should satisfy the requirements of 2.5-D modelling. Mass balance calculations were made to ensure that the modelled structure satisfied the constraints of isostasy outside the trench-arc gap. Within the gap, Riddihough (1979) pointed out that dynamic, rather than isostatic, balance was probably more realistic. In particular, the high heat flow values observed near the volcanic zone would imply considerable crustal buoyancy that is not isostatically compensated. The oceanic column, formed with standard water depths and oceanic crustal thicknesses and used to extend the model to —2000 m distance, was balanced against the continental column of uniform crust of Woollard (1969; p=2920 kg/m3) used east of the Yalakom Fault, from approximately 460 to 2000 km model distance, by equating density-thickness products (as per Riddihough 1979). This procedure served to satisfy the constraints of isostasy and to provide an estimate of the regional gradient "outside the display model range. 4.1.2 The data profiles Observed data profiles were obtained manually from the regional field anomaly contour maps plotted at 1:500000 scale (refer to Chapter 2 for processing and plotting details). Points outside the study region were obtained from a published 1:2,000,000 scale gravity anomaly map (Riddihough and Seemann 1982). Figure 4.1 shows the locations of the profiles along which models were constructed to match the observed data. The locations of multichannel seismic reflection lines 85-01, 84-01, 85-02, and 85-04 have been highlighted. These CHAPTER 4. POTENTIAL FIELD MODELLING 61 Figure 4.1: Locations of profiles along which models were constructed to match the observed gravity and magnetic anomaly data. Line numbers correspond to model numbers. Also shown are locations of seismic reflection lines along which primary models 1, 2 and 4 were constructed. CHAPTER 4. POTENTIAL FIELD MODELLING seismic lines were used, in conjunction with other constraints, to build the three primary models, lines 1, 2 and 4. Secondary models have been named to indicate the primary models from which they were derived. Line S across the southern tip of Vancouver Island was considered independently. Note that line 1 was constructed in three segments. The model line first followed the offshore seismic line 85-01 to Barkley Sound, then was offset 15 km to the southeast to the beginning of line 84-01. A straight-line projection of 84-01 was followed across the island to the east coast, whereupon the model line returned to the original N52°E azimuth and continued across to the Coast Mountains. It was felt that the benefits of seismic control across the island outweighed the complexity of a bent model line with its resulting geometric corrections. The remaining lines were constructed parallel to the N54°E trend of line 85-02, approximately equal to the N56°E plate convergence azimuth, with the exception of line 4 at the northern end, which followed the seismic line 85-04 azimuth of N48°E. Geometric corrections for the different azimuths were made to the structures of lines 1 and 4 to match the remaining models. A comparison of observed data for the various profiles illustrates the anomaly features that are common to all lines, and should therefore be included in all models, and smaller localized features. Gravity anomaly profiles for lines IAS to 4 are shown in figure 4.2. Prominent regional features on the gravity profiles are the gravity lows of the outer slope and continental shelf, at 20 and 110 km respectively on line IAS, which gradually reduce in amplitude and converge northward to line 2B. A single large gravity low in this position on line 4 corresponds to the southern end of Winona Basin. Landward of the lows on all lines is an approximately 40 km-wide peak which lies over the westernmost third of Vancouver Island. The remainder of the island is characterized by a broad peak of lower amplitude, with local variations, that continues into the Strait of Georgia on the southern lines before beginning a general decrease to the large negative anomaly associated with the CHAPTER 4. POTENTIAL FIELD MODELLING GRAVITY ANOMALY PROFILES 400 200 0 Line 4 Line 2B — Line 2A Line 2 Line IE Line ID Line IC Line IB Line 1A Line 1 L ine IAS BAIIN1 T O F I N O BASIN 0 100 200 Model Dis tance (km) 300 Figure 4.2: Gravity anomaly profiles for lines shown in figure 4.1. Line positions are offset 20 mGal (positive northward) per 10 km with respect to line 1 position to reduce overlap. CHAPTER 4. POTENTIAL FIELD MODELLING MAGNETIC ANOMALY PROFILES 4000 ~1 T" T Line 4 3000 Line 2B 2000 1000 0 Line 2A Line 2 Line IE Line ID Line IC Line IB Line 1A Line 1 Line IAS Prometheus Magnetic Anomaly 1000 _OCEAN BASIN VANCOUVER ISLAND i i i I i ' i 0 100 200 Model Dis tance (km) 300 Figure 4.3: Magnetic anomaly profiles for lines shown in figure 4.1. Line positions are offset 160 nT (positive northward) per 10 km with respect to line 1 position to reduce overlap. CHAPTER 4. POTENTIAL FIELD MODELLING 65 Coast Mountains. A sharp maximum superimposed on line 1 and lines IB to 2 at 240 to 260 km is centred between Lasqueti and Texada Islands in the strait. The decrease to the negative anomaly of the Coast Mountains begins at the mapped or assumed contact with the lower density intrusives of the Coast Plutonic Complex. A moderate peak, approximately 40 km-wide at 280 km model distance, is superimposed on the decrease on several lines. Modelling of this feature, as will be discussed, has suggested that the CPC does not achieve appreciable thickness until tens of kilometres east of its surface contact. The magnetic anomaly profiles are shown in figure 4.3. Peaks and troughs west (seaward) of 80 km (110 km on line 4) are the magnetic "striping" pattern associated with oceanic crust. The next prominent peak, at approximately 120 km, is the Prometheus anomaly marking the Eocene volcanics of the Crescent Formation. This peak is largest beneath lines 1A and IB and weakens northward to apparendy die out completely by line 2B. A pronounced trough separates this peak from a generally higher amplitude peak that is centred between 140 and 160 km and marks the metamorphosed diorite complex that lies near the west coast of Vancouver Island (see also figures 1.3 and 2.3). Other prominent peaks are associated with outcrops of the Island Intrusions on Vancouver Island. In particular, a large peak at approximately 220 km corresponds on lines IAS, 1A and IB with a mapped granodiorite / quartz diorite exposure on eastern Vancouver Island. It would appear from the magnetic anomaly data that this body continues underneath the overlying Nanaimo Formation sedimentary rocks to at least the line 2 position, beyond which data are not available in this area. This interpretation was confirmed by modelling. CHAPTER 4. POTENTIAL FIELD MODELLING 4.2 Primary Models 4.2.1 Line 1 Line 1 was the best constrained of all lines modelled in this study and as such provided the basic interpretation that was modified and verified on the other lines. Constraints on structure and composition were provided by the two multichannel seismic reflection lines (85-01 and 84-01), the across-margin VISP 80 refraction line, a gravity profile model of Riddihough (1979), and several other geophysical studies in this well-studied corridor. Seismic line 85-01 corresponded to model distances 0-138.8 km, and the straight-line projection of 84-01 to the range 146.6-220.5 km, with the line offset occurring at 140.5 km. The gravity data along the two segments were merged at 136.1 km, and the magnetic data at 132.0 km, positions at which the contours were parallel to the merge direction (i.e. perpendicular to the straight-line projection of line 84-01) and of the same magnitude. Interpretations of the reflection lines have previously been published (e.g. Hyndman et al. 1990; Davis and Hyndman 1989; Clowes et al. 1987b, 1987a; Green et al. 1987; Yorath et al. 1985a, 1985b). The sections were re-examined in light of the potential field data and a modified interpretation is discussed herein. The stacking velocity information accompanying the seismic sections, in conjunction with the nearly coincident refraction model, permitted the reflection times to be converted to depths. In addition, density estimates for the sediments and deeper structure were obtained from the processing interval velocities and refraction model velocities using the velocity -density relationship of Nafe and Drake (1963). Seismic interpretation Reflection sections 85-01 and 84-01 are shown in figure 4.4, together with an interpretation of the main features. Line 85-01 covers part of Cascadia Basin, the continental slope, and most of the continental shelf. The sediments in Cascadia Basin, divided into an upper LINE 85-01 Figure 4.4: Multichannel seismic reflection sections 85-01 (migrated) and 84-01 (unmi-grated). Interpretation, shown in lower panel, has been modified from others (refer to text) to include the results of this ;study. as CHAPTER 4. POTENTIAL FIELD MODELLING CHAPTER 4. POTENTIAL FIELD MODELLING 69 turbiditic and a lower hemipelagic unit, are relatively flat lying until they encounter the deformation front at the eastern edge (shotpoint 700), where they are disrupted by thrust faults that appear to be detaching all material above the oceanic crust. The reflection character of the adjacent accretionary wedge is chaotic, with only a few landward-dipping faults detectable within the structure. The upper surface of the wedge is characterized by tight folds and blocks showing fault offset. Between shotpoints (SP) 2050 and 2400, the rough character of the wedge reflector implies the presence of volcanics, an interpretation in agreement with the Crescent Formation basalts encountered at 1800 m depth in the Prometheus H-68 borehole near SP 2280. A disrupted reflector to the east, between SP 2450 and 2700, has been interpreted as the top of the Pacific Rim Terrane. Tofino Basin sediments underlying the shelf have been gently folded, and in places faulted, to echo the wedge deformation. A reflector associated with the top of the subducting oceanic crust is clearly imaged from 5 to 6 s beneath the Cascadia Basin sediments and then is discontinuously, but clearly, imaged dipping beneath the wedge to reach 9 s, or approximately 25 km depth, at SP 2881 near the west coast of Vancouver Island. The oceanic crustal reflection is extended by a short segment at 10 s (35 km depth), 10 km to the east on line 84-01. Approximately 3 s above this, a dipping 2 s-thick band of high reflectivity (the ' E ' zone) can be followed two-thirds of the way across the section before it fades. Reflective zone ' C , of similar thickness and length, lies between 4 and 5 s. Various reflective bands within the uppermost 4 s have been attributed to lithologic and intrusive contacts within Wrangellia. Several eastward dipping faults appear to cut across the upper structure beneath western Vancouver Island. A fault (labelled "?") that begins at 0.5 s at the western end of the line (SP 1420) and appears to bisect the ' C reflector has been previously interpreted as the Westcoast Fault (e.g. Hyndman et al. 1990), which marks the upper bound of the Pacific Rim Terrane, or as the Tofino Fault (e.g. Yorath et al. 1985a,b; Clowes et al. 1987a; Green et al. 1987), beneath which lies the Crescent Terrane. However, the additional CHAPTER 4. POTENTIAL FIELD MODELLING information provided by the potential field data analysis in this study indicates that both faults probably surface further westward beneath the continental shelf. This interpretation will be further examined during the discussion of the line 1 final model. CHAPTER 4. POTENTIAL FIELD MODELLING 71 Line 1 model development The model for line 1 was constructed using the geometry obtained primarily from seismicity, seismic refraction, and reflection studies. Figure 4.5a shows the seismic refraction model of Drew (1987; see also figure 3.3b) converted to a density model using the Barton (1986) velocity-density relationship. (Barton (1986) provides both the mean and the maximum / minimum bounds of the measurements used for the Nafe and Drake (1963) relationship. The limits shown by Barton for seismic velocities above 6.5 km/s suggest slightly higher densities than would be obtained with the Nafe-Drake relationship.) The misfit between the observed and calculated gravity anomalies is considerable, indicating an excess of mass in the shelf region of the model and a mass deficiency beneath Vancouver Island. The next modelling stage began by simplifying the crust and offshore sedimentary struc-tures, and incorporating detailed bathymetry and the interpreted features of seismic reflection line 84-01. Density values for the oceanic lithosphere, asthenosphere and continental mantle were based on those of Pviddihough (1979) and Spence etal. (1985). Magnetic susceptibility values for the uppermost continental crust and offshore terrane extensions were based on the compiled measurements listed in table B.2, supplemented by values obtained from standard tables (e.g. Telford et al. 1976). Values for other units and structural components were based on assumed composition and are not well constrained. The thin low velocity / high density ' C and ' E ' bands have been removed from the model (figure 4.5b) and a single, very high density block was used to try to generate the gravity anomaly high over western Vancouver Island. The simplified continental crust model east of the island is clearly too high in density as shown by the anomaly curves. The lower continental crust was restructured in the next stage (figure 4.6a) to return to the structure of the initial model. Small, near-surface bodies were introduced, as per surface geology, to model local variations in the anomaly curves. A lower density crustal CHAPTER 4. POTENTIAL FIELD MODELLING 72 Figure 4.5: Line 1 initial model (a), based on velocity model of Drew (1987), and first intermediate model (b). Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for blocks. Vertical exaggeration is 1.3x. Symbols represent observed data and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 73 Figure 4.6: Line 1 second (a) and third (b) intermediate models. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for blocks. Vertical exaggeration is 1.3x. Symbols represent observed data and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 74 block (2850 kg/m3) was emplaced beneath the mainland to represent the granitic rocks of the Coast Plutonic Complex and produce the observed gravity anomaly low. In addition, the subducting oceanic crust was increased to mande density over the 35-70 km depth range as per Riddihough (1979) to represent a metamorphic phase change of the basalt. The next stage of modelling saw the introduction, based on surface geology and mod-elling results, of greater complexity in the upper crust beneath Vancouver Island to represent the gravity and magnetic influence of components such as Sicker Group rocks and various intrusive units (figure 4.6b). The geometry of the low density ' C and ' E ' zones was varied slighdy to include only the most highly reflective regions on seismic line 84-01. These zones were extended eastward as dipping layers that were truncated at or joined the base of the continental crust. The crust beneath the mainland was subdivided into an upper unit and a higher density lower unit to allow better control on the shape of the calculated anomaly. The final stage of line 1 modelling saw a few changes in shallow continental crustal structure to satisfy the magnetic anomaly data, and a shift in the position of the ' C zone extension. Line 1 final model The final derived structural model for line 1 is shown in figure 4.7. The model has three main components: the subducting oceanic plate, the continental framework, and the modern accretionary complex. Above the model lie the gravity and magnetic anomaly curves. Ob-served values, denoted by cross (+) and crisscross (x) symbols respectively, are compared with the calculated solid line curves. Magnetic susceptibility values, labelled in italic type on the model, are shown in brackets where poorly constrained and are omitted where their contribution to the calculated curve is considered negligible. Remanent magnetization was not included in this or the other models as units with measured remanence vectors also Figure 4.7: Line 1 final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data and J-J solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING showed signs of considerable displacement and rotation, and most units had an uncertain magnetization history. The offshore basin sediments, divided into the upper turbidite (p = 2000 kg/m3) and lower hemipelagic (2350 kg/m3) layers, extend to the foot of the slope at 41 km model distance. The combined sediment thickness is approximately 4 km, increasing to 5 km near the deformation front. The wedge sediments have been divided into three units with density increasing landward (2290, 2410 and 2610 kg/m3) to represent the effects of compaction. The third unit abuts against the dipping slab (2780 kg/m3) that represents the Crescent Terrane. The slab has been modelled as being 4 km thick, with a slight bend and eroded surface at the upper end to match the seismic expression and observed +100 nT magnetic anomaly. The depth of the top of the slab was determined from the Prometheus H-68 borehole at 109 km model distance. The slab in the model dips at approximately 25° which brings it to a depth of 23 km near the end of the ' E ' zone reflector imaged on line 84-01. Landward of the Crescent slab lies an 8-10 km-thick unit representing the Pacific Rim Terrane melanges and volcanics. A relatively high density (2900 kg/m3) was required in this region to satisfy the observed data. This is higher than surface measurements of the Leech River Formation (2670-2750 kg/m3) and could imply either the presence of higher density material at depth, or an increase in the melange density due to metamorphism (blueschist facies). Overlying the wedge sediments, Crescent Terrane, and Pacific Rim Terrane, the sediments of Tofino Basin have been modelled with an average density of 2520 kg/m 3. A small wedge of sediments (2530 kg/m3) extends downward from the basin into the gap between the tops of the Crescent and Pacific Rim Terranes to match the seismic observations. This unit contributes -10 mGal to the calculated gravity anomaly over a width of 15 km. The continental crust of Vancouver Island has been subdivided into an upper layer, ap-proximately 10 km thick, a lower layer 5-15 km thick, and an underlying region which contains the ' C and ' E ' zones and high velocity material. The upper crust has been further CHAPTER 4. POTENTIAL FIELD MODELLING 11 subdivided laterally into reasonably large blocks according to the mapped surface geology. These blocks have been assigned densities that vary from the average 2930 kg/m 3 used for the Karmutsen volcanics according to the amount of other material, such as granodiorite in-trusives, believed to be present on the basis of surface geologic trends. That is, these blocks do not necessarily represent structural divisions but are averaged density and susceptibility blocks. Smaller blocks have been used where necessary to satisfy the magnetic anomaly data. One of these blocks plays a prominent role in the interpretation of the magnetic anomaly. An intrusive unit of metamorphosed diorite has been mapped along much of the west coast of the island (see figure 1.3). The contact between this unit of high susceptibility and the Pacific Rim Terrane, as mapped across the Westcoast Fault north of Barkley Sound, is a sharp increase of >200 nT over a 5 km distance (see also figure 4.3). On the magnetic anomaly field map (figure 2.3), this anomaly can be followed almost continuously from the vicinity of the San Juan Fault along the coast to just north of Brooks Peninsula. Figure 4.8a shows a detail of the magnetic anomaly map in the vicinity of Barkley Sound. Superimposed on the map are the locations of Pacific Rim Terrane outcrops and several nearby intrusive units (from Muller 1977b) that affect the anomaly. The inferred offshore position of the Westcoast Fault would approximately follow the -50 nT contour, continuing the trend observed on land. Accordingly, the Pacific Rim Terrane has been modelled as forming the seaward unit at this contact for all models in this study. Figure 4.8b shows the calculated anomalies resulting from the geometry and physical parameters used in the line 1 model for the Pacific Rim and diorite units (curve (a)). For comparison, additional anomaly curves are shown, one in which both units assume the properties of the Pacific Rim Terrane (curve (b)) and the other in which both units have the properties of the diorite unit (curve (c)). Clearly a major change in properties must occur near the location shown, and thus this position has 78 Figure 4.8: Proposed Westcoast Fault location, (a) magnetic anomaly map in the vicinity of Barkley Sound. Geology simplified from Muller (1977b); refer also to figure 1.3. (b) detail of the line 1 final model and the calculated anomalies for: a - properties shown; b - both bodies Pacific Rim Terrane; c - both bodies Westcoast Complex (diorite). Symbols represent observed data. CHAPTER 4. POTENTIAL FIELD MODELLING been interpreted for the Westcoast Fault. If the Pacific Rim Terrane is not present, then an equivalent unit of low-susceptibility, but relatively high density, material must be present. This specifically excludes the Crescent Terrane and other intrusive units of Wrangellia. A density of 2870 kg/m 3 and susceptibility of 21400 SI units has been used to characterize the diorite unit on line 1 between the interpreted Westcoast Fault at 133 km and the surface contact with Karmutsen basalts at 147 km model distance. The source of the broad high amplitude magnetic anomaly peak (500 nT) at 210 km model distance was also examined. The low susceptibility, sedimentary Nanaimo Formation has been mapped at the surface in this region, so a deeper source was sought. A linear granodiorite body, with a corresponding prominent magnetic anomaly signature, has been mapped 5 km to the south of line 1 up to the edge of the Nanaimo deposits. Therefore, it was assumed that this intrusive body continued at depth below the profile and a 20 km-wide unit, with a westward-dipping upper surface at approximately 2 km depth and high susceptibility (62800 SI units), was used to generate the calculated anomaly. The remaining structure was developed to satisfy the gravity anomaly profile. The profile starts with a trough over the Cascadia Basin sediments and slightly landward dipping oceanic plate (1-2°). This changes to a broad peak centred at 65 km model distance which was attributed to the decreasing water depth over the slope and shelf. A second trough then reflects the steepening of the oceanic plate as it turns downward more sharply beneath the accretionary wedge. The large increase in near-surface density, as the sedimentary accretionary wedge meets the interpreted Crescent and Pacific Rim Terranes and continental backstop of Wrangellia, initiates the second peak which is quite broad and extends over the width of Vancouver Island. It is approximately 20 mGal higher over the western part of the island than over the east, and this has been modelled by using a higher density at depth. The low/high velocity doublets observed on the seismic reflection and refraction lines have been converted to low/high density bands using the Nafe-Drake relationship and incorporated into CHAPTER 4. POTENTIAL FIELD MODELLING 80 the model between 150 and 220 km distance. This combination has successfully matched the observed gravity anomaly profile, although it should be noted that the individual low density bands are too deep and too thin to be resolved through gravity modelling, and other configurations having the same average density would probably reproduce the results. The ' C zone band was extended across the model at slightly higher density (2850 kg/m3) to pass through a band of high reflectivity observed on seismic line 88-16 just east of the Strait of Georgia (Clowes 1990). Beneath the ' C zone, the high velocity / high density layer was also extended, although at a slightiy reduced density of 3000 kg/m 3, to fill in above the interpreted Moho at < 39.5 km depth. The ' E ' zone unit was terminated at the Moho, although a modified model in which the ' E ' zone extended into the continental mantle also satisfied the observed gravity data. The space beneath the ' E ' zone and subducting oceanic crust required a unit of quite high density (3030 kg/m3) which, upon conversion, approximately matched the velocity used in the refraction model of Drew and Clowes (1990) for the lowermost crustal unit. A local negative anomaly at approximately 230 km distance was modelled using a max-imum 1 km thickness of Nanaimo Formation (2550 kg/m3) and the almost coincident Strait of Georgia. The anomaly increases sharply to slightly positive values starting at 238 km distance. Most of this 20 km-wide peak" can be matched simply by using Karmutsen or equivalent rock densities. However, this peak has a secondary component represented by the sharper, slightly higher peak between 239 and 247 km distance. The secondary peak is centred over the strait between Texada and Lasqueti Islands and requires slightly higher density material at shallow depths. A large iron ore body, consisting primarily of magnetite with smaller amounts of chalcopyrite, has been mined at the north end of Texada Island for almost a century. The high densities associated with these minerals (5120 and 4200 kg/m 3 respectively; Telford et al. 1976, p. 28) are probably generating the observed anomaly. The ore body is herein considered a local feature and was not included in the model. East of this peak, a second local negative anomaly was not satisfied with simply the CHAPTER 4. POTENTIAL FIELD MODELLING 8 1 w a t e r a n d s e d i m e n t s o f t h e e a s t e r n S t r a i t o f G e o r g i a b u t r e q u i r e d a r e d u c t i o n i n t h e d e n s i t y o f t h e u p p e r m o s t c r u s t . T h i s s e e m e d r e a s o n a b l e g i v e n t h e c l o s e p r o x i m i t y o f t h e m a p p e d g r a n i t i c C o a s t P l u t o n i c C o m p l e x , a n d s o a 4 k m t h i c k n e s s o f 2 8 0 0 k g / m 3 m a t e r i a l w a s a d d e d , t h i n n i n g e a s t w a r d t o a c c o m m o d a t e t h e p o s i t i v e i n c r e a s e i n t h e o b s e r v e d a n o m a l y t o a l o c a l m a x i m u m a t 2 8 0 k m . T h e a n o m a l y c u r v e t h e n s h o w e d a m a j o r d e c r e a s e w h i c h w a s m o d e l l e d b y r e d u c i n g t h e d e n s i t y o f t h e u p p e r m o s t 2 0 k m o f c r u s t t o 2 8 0 0 k g / m 3 t o r e p r e s e n t t h e b u l k o f t h e b a t h o l i t h . T h i s w a s u n d e r l a i n b y 1 0 k m o f 2 8 5 0 k g / m 3 m a t e r i a l w h i c h i n c l u d e s t h e ' C z o n e e x t e n s i o n . T h e c o n t i n e n t a l c r u s t w a s a l s o t h i c k e n e d s l i g h t l y t o p l a c e t h e M o h o a t 3 9 . 5 k m b e n e a t h t h i s p o i n t . T h e o c e a n i c p l a t e w a s m o d e l l e d a s d i p p i n g a t 1 ° s e a w a r d o f t h e c o n t i n e n t a l s l o p e , s t e e p -e n i n g s u c c e s s i v e l y t o 4 a n d 1 0 ° b e n e a t h t h e w e d g e , a n d finally t u r n i n g d o w n w a r d t o 1 8 ° a t 1 5 k m d e p t h . T h e d e n s i t y o f t h e o c e a n i c c r u s t , o r i g i n a l l y 2 8 8 0 k g / m 3 , w a s a l l o w e d t o g r a d u a l l y i n c r e a s e t o m a n t l e - r a n g e d e n s i t y ( 3 3 1 0 k g / m 3 ) b e t w e e n 3 3 a n d 7 0 k m d e p t h i n r e s p o n s e t o t h e p o s t u l a t e d b a s a l t - e c l o g i t e p h a s e c h a n g e ( G r e e n a n d R i n g w o o d 1 9 6 7 ) a t i n -c r e a s e d t e m p e r a t u r e s a n d p r e s s u r e s . A s a r e s u l t , t h e d i p o f t h e p l a t e b e n e a t h a p p r o x i m a t e l y 5 0 k m c o u l d n o t b e w e l l r e s o l v e d b y t h e g r a v i t y m o d e l l i n g a n d u n c e r t a i n t i e s i n d i p a n d p o s i t i o n w i l l i n c r e a s e w i t h d e p t h b e n e a t h t h i s l e v e l . ~ Model evaluation T h e final m o d e l r e p r e s e n t s a n a t t e m p t t o m a t c h t h e o b s e r v e d g r a v i t y a n d m a g n e t i c a n o m a l y d a t a a l o n g p r o f i l e l i n e 1 w h i l e s a t i s f y i n g t h e m a n y c o n s t r a i n t s p o s e d b y t h e o t h e r g e o p h y s i c a l a n d g e o l o g i c a l d a t a . S o m e o f t h e f e a t u r e s s h o w n i n t h e m o d e l w e r e i n c l u d e d p r i m a r i l y t o s a t i s f y o t h e r d a t a a n d a r e n o t w e l l - c o n s t r a i n e d b y t h e m o d e l l i n g i n t h i s s t u d y . T h e s e i n c l u d e t h e t h i n ' C a n d ' E ' z o n e s ( 2 8 0 0 k g / m 3 ) w h i c h c a n n e i t h e r b e d i s t i n g u i s h e d n o r r e s o l v e d s i m p l y o n t h e b a s i s o f g r a v i t y a n d m a g n e t i c a n o m a l y m o d e l l i n g . H o w e v e r , t h e c l e a r i m a g i n g CHAPTER 4. POTENTIAL FIELD MODELLING 82 of these features on several seismic reflection lines and, in the case of the ' E ' zone, by other methods such as magnetotelluric measurements, lends support to their existence and they have thus been incorporated into the model. The point to be stressed is that, while the models represent the best fit to the observed data, they are not necessarily the only models that wil l fit the data satisfactorily. The line 1 model was divided into four major tectonic components and the individual anomalies were calculated to determine the contribution of each component to the total grav-ity and magnetic anomalies. Figure 4.9 shows the model divisions and the corresponding calculated anomalies. The four divisions represent: (1) the oceanic and accretionary prism units, including the Crescent and Pacific Rim Terranes; (2) the upper and middle continental crust, extending to and including the ' C zone; (3) the lower crust and/or underplated region; and (4) the subducting oceanic lithosphere. Note that the oceanic asthenosphere and conti-nental mantle were not included in the calculations as their gravity anomalies are of much higher amplitude and would make the comparisons among curves difficult. The anomaly curves represent the total calculated effect of each component; a floating datum level was not removed from the anomaly values as for other model plots. It is readily apparent that the major contribution to the observed magnetic anomaly comes from the upper continental crustal structure (curve 2, figure 4.9). The only anomaly not satisfied by this structure is the offshore peak attributed to the Crescent Terrane and included in the curve 1 calculation. The use of low susceptibility values for much of the subducting slab and lower continental crust therefore seems appropriate. The observed gravity anomaly receives its most significant contributions from the off-shore/ accretionary units (curve 1) and the upper continental crust units (curve 2). The high density wedge in the lower continental crust contributes to the broad gravity anomaly over the western part of Vancouver Island (curve 3). The subducting slab generates a long wave-length anomaly (curve 4) that smoothly decreases by 150 mGal as the slab dips beneath the CHAPTER 4. POTENTIAL FIELD MODELLING 8 3 i i i i i i i i i i i | i i i i | i i i i | i i i i | i u i ' I ' ' '' I ' ' ' 1 • i o o o o o -tf C\2 i L u o o co o o CV 0) o rd T3 O O O -4} <p ! g & t : i i 12 60 ^ I i i "i i I i i i t I J _ l _ O O o o o LO ' " 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1111 o o o o o o H W O ^ I f i CHAPTER 4. POTENTIAL FIELD MODELLING 84 accretionary wedge and shows little variation to the east where the plate is below 50 km depth. The line 1 model provided the "standard" against which subsequent models were com-pared. Differences in observed anomalies between this and other lines was at first attributed to near surface compositional changes as indicated by lateral variations in mapped surface geology. Changes to the lower part of the model were considered only if modifications to the upper layers (within the bounds posed by composition) failed to satisfy the observed anomaly. Thus, the greater the number of models that could be satisfied by the general line 1 structure, the greater the support for the appropriateness of the model. 4.2.2 Line 2 Line 2 lies approximately 60 km northwest of line 1 and crosses Vancouver Island just south of Clayoquot Sound. The Pacific Rim Terrane outcrops along the west coast in this region, and this line provides control on the geophysical signature of this unit for use on the other lines. Seismic interpretation _ The reflection section for line 85-02 is shown in figure 4.10. The line was located approxi-mately 60 km northwest of line 85-01 and extended from the abyssal plain across most of the continental slope and shelf. Previously published interpretations have been re-evaluated to include the potential field data and modelling results. The imaged structure is very similar to that of line 85-01 although the accretionary prism shows a different deformational style. A single frontal thrust fault at SP 1950 marks the start of deformation, beyond which the continental slope is terraced into two tiers separated by steep grades. Several thrust faults appear to lead downward from the terrace features to the top of the oceanic crust. The Figure 4.10: Multichannel seismic reflection section 85-02 (migrated). Interpretation, shown in lower panel, has been modified from others (refer to text) to include the results of this study. CO CHAPTER 4. POTENTIAL FIELD MODELLING 86 surface of the accretionary prism is gently undulating and the overlying sediments of Tofino Basin are flat-lying except where they are disrupted by a diapiric structure near the eastern end. The characteristic rough reflector associated with the Crescent volcanics appears absent on the seismic section. The location shown, just landward of the diapiric structure, was derived from the potential field interpretation of this study. The strong reflection from the subducting oceanic crust dips strongly beneath the Cascadia Basin sediments, is broken into short segments beneath the continental slope, and continues uninterruptedly beneath the shelf to a depth of approximately 18 km (7 s) at SP 101, 20 km west of Vancouver Island. Line 2 final model The final model for line 2 is shown in figure 4.11. The basic structure of this model is very similar to that of the line 1 model, with minor differences in plate depth and upper continental crustal structure. Values of density and magnetic susceptibility are consistent with those used in line 1 for all regions except the upper continental crust, which was allowed to vary based on surface geology. The Crescent Terrane has been modelled as dipping more steeply (34°) in this region, as has the Pacific Rim Terrane unit, in order to preserve the geometry of the continental block while satisfying the observed data. The diapiric structure imaged on the seismic reflection line is centred above the seaward boundary of the Crescent slab, suggesting that if the diapirism was caused by fluid flow, as seems reasonable, the fluids may have been guided to the surface along this impermeable boundary. The presence of fluids seems highly likely given the proximity to the subducting slab and the probable dewatering of the downgoing sediments and upper oceanic crust. Higher density than on line 1 was required in the upper continental crust adjacent to the Pacific Rim Terrane to match the positive gravity anomaly between 150 and 180 km distance. The higher density was "created" by having Karmutsen Formation rocks replace 0 STR. OF OCEAN BASIN SLOPE SHELF i VANCOUVER ISLAND i GEORGIA , MAINLAND Figure 4.11: Line 2 final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities; in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data co and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING the lower density Bonanza Group material of the equivalent unit on line 1, as per surface geological indications. The remaining shallow crustal structure was developed to satisfy the gravity and magnetic anomaly data while adhering to compositional constraints imposed by the surface geologic maps. The west coast diorite body of line 1 was thinned and widened accordingly to match the broader, lower amplitude peak of the observed magnetic data. Similarly, the intrusive body beneath the east coast was modified to fit the observed peak, which unfortunately is incomplete due to data unavailability. The overlying Nanaimo Formation sediments have been thinned to 300 m thickness. The contact between Wrangellia and the Coast Plutonic Complex has moved slightly eastward, with the major thickening of the CPC now occurring at 310 km model distance, almost 50 km east of the surface contact. 4.2.3 Line 4 Line 4 lies just south of Brooks Peninsula and is the only line examining the Explorer Plate. It is not as well constrained as its neighbours to the south, relying primarily on seismic line 85-04 and geologic surface mapping. A complete interpretation of line 85-04 has not been published to date, although Davis and Hyndman (1989) analysed the structure of the deformation front. Therefore a complete interpretation of the line was done as part of this study and to provide a starting point for the modelling. Seismic interpretation Seismic line 85-04 (figure 4.12) was located over the Explorer Plate, just south of Brooks Peninsula. In this region of the subduction zone, the continental shelf is quite narrow and the sedimentary basin at the foot of the slope, a southward continuation of the Winona Basin, is filled with thick accumulations of Pleistocene turbiditic deposits (e.g. Davis and Clowes 1986). The line crossed over the southern extremus of the Haida Ridge (SP 600), a feature CHAPTER 4. POTENTIAL FIELD MODELLING CHAPTER 4. POTENTIAL FIELD MODELLING 90 apparently formed by the compressive upthrusting of a thick sequence of sediments. The oceanic crust does not appear to be involved in the ridge structure, although the top of the crust appears quite rough with an offset of 0.5 s at SP 450 just west of the ridge. Several thrust faults, dipping both eastward and westward, cut through the basin sediments to the top of the oceanic crust, including several west of the ridge. Landward of SP 850, the style of deformation changes and the sedimentary sequence has been compressed into several tight folds. The continental slope has a very rough surface with much evidence of slumping and mass wasting, as well as what appear to be faults leading down into the accretionary wedge. Few deep reflections are visible below 3 s beneath the slope and shelf. A very weak 2-3 cycle reflection that infrequently appears (e.g. 5.5 s at SP 1850) is similar to the reflection from the oceanic crust beneath the shelf on lines 85-01 and 85-02 and has thus been interpreted as the top of the oceanic crust. The location of the crustal reflector beneath much of the slope can only be approximated. A disrupted reflection near the shelf-slope break (SP 1650, 1.4 s to SP 1730, 0.9 s) is similar in character to the reflection from the top of the Pacific Rim Terrane on line 85-01 and was tentatively interpreted as such. The results of the potential field-modelling and the proximity to an outcrop at the western tip of Brooks Peninsula appear to support this interpretation. Line 4 final model The final model for line 4 is shown in figure 4.13. This model differs considerably from the line 1 and 2 models, both in the position of the subducting plate and the structure of the continental crust. The initial attempt at modelling, which incorporated the shallower depth of oceanic crust observed on the seismic section while retaining much of the line 1 structure, provided a poor fit to the data. Much higher density was required in the continental region (or, equivalently, lower density in the offshore oceanic region) to match the observed profile. Figure 4.13: Line 4 final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 92 Therefore, it was necessary to introduce major changes both in the position of the plate as it passed beneath the accretionary wedge and edge of the continent, and to the continental structure and density. The shallower depth of oceanic crust observed on the seismic section has thus been continued beneath the accretionary wedge to satisfy the observed gravity data, and as a result the slab remains in contact with the overlying continental fabric until 180 km distance, 20 km further east than for lines 1 and 2. The continental Moho is much shallower at this point, approximately 26 km rather than 33-34 km as in the south. The shallower plate location is not too surprising, given the slower convergence rate and warmer temperature, therefore increased buoyancy, of the Explorer Plate. However, this means that the oceanic crust is occupying the position previously modelled as the ' E ' zone and underplated material, and these two units have been displaced from the model. In addition, a considerable increase in continental crustal density was required to compensate for the shallow lower density oceanic crust, with the result that the ' C zone density of 2800 kg/m3 was replaced with the overlying crustal density of 2930 kg/m 3. The ' C zone is thus indistinguishable from the overlying material (lower Wrangellia) and is shown in the model simply for comparison with lines 1 and 2. The eastern extensions of the ' C zone and high velocity zone in the lower crust were also replaced with higher density material, 2950 and 3140 kg/m3 respectively, to avoid what would otherwise have been a —30 mGal misfit in the 200-300 km distance range. The position of the subducting slab beneath this range was varied to try to avoid the drastic changes in crustal density, but to little avail as the plate, with its increasing crustal density due to the postulated phase change, has little influence on the anomaly beneath approximately 50 km depth. A thinner continental crust is another method of increasing the calculated gravity anomaly, as this would replace the lower crustal material with higher density mantle material. A crustal thickness of approximately 32 km would provide the desired effect and would move the CHAPTER 4. POTENTIAL FIELD MODELLING 93 Moho to just beneath the 4 C zone east of 260 km, shallowing westward to just above the position shown at 180 km. The present data do not allow a clear choice between the two alternatives. However, it is clearly evident that significant differences exist between the lower crustal structure ( ' C zone and lower) of southern Vancouver Island and the equivalent crustal structure in the north. The large magnetic anomaly peak at 145 km is associated with the metamorphosed diorite unit that outcrops on most of Brooks Peninsula to the north and at several other locations along the west coast of Vancouver Island. A second peak at 180 km corresponds to a quartz diorite unit (2770 kg/m 3, 21400 SI units) that is crossed by the line. Note that only the positions and approximate shapes of these units were modelled as there was no other information to constrain their size and orientation. The trough between the two peaks represents the considerable volume of lower susceptibility Bonanza Formation rocks that outcrop along the northwestern part of the island. A gravity low centred at 180 km distance also corresponds with the combined Bonanza Formation and intrusive rocks. The gravity anomaly over the remainder of Vancouver Island was satisfied with a single block of 2930 kg/m 3 density and minor low density units of Quatsino Formation sedimentary rocks (208 km distance) and Johnstone Strait water (238 km distance). The contact to the east with the Coast Plutonic Complex remains similar to that previously modelled. Offshore to the west, high densities have been used for the sediments of Winona Basin, agreeing with previous modelling results of Clowes et al. (1981) and Davis and Riddihough (1982) and also consistent with the high measured velocities of those sediments (Davis and Clowes 1986). The Pacific Rim Terrane is present, with its previously modelled properties, lying seaward of the diorite unit at 138 km distance and having a modelled width of 15 km. There is no evidence to suggest that the Crescent Terrane is present and it has not been included in the model. CHAPTER 4. POTENTIAL FIELD MODELLING 94 4.3 Secondary Models Potential field modelling along the three seismic lines provided good, but restricted, infor-mation on the structure of the margin. One of the aims of this study was to see if the derived interpretations could be extended laterally along the margin so as to extend the interpreted region. In general, this meant moving from a well-constrained area into one of fewer, if any, structural or compositional constraints. Therefore, it was decided to use the well-constrained primary models, corrected for local bathymetry, surface geology, and depth variations due to dipping features, as initial models for adjacent profile lines and then determine through mod-elling the variations required to satisfy the observed data. The finished "secondary" model then served as the initial model for the next profile location, and the procedure was repeated until the desired coverage was obtained. Various line spacings were used to determine the optimum distance for laterally extending an interpreted model. The areas chosen for this "slice-wise" interpretation were: (a) the region south of line 85-01, a one line extension 20 km to the south into the region of curvature of the subduction zone, denoted as line IAS; (b) the region between seismic lines 85-01 and 85-02, at 10 km spacings, numbered as lines 1A to IE; and (c) the region between lines 85-02 and 85-04, at 40 km spacings, denoted as lines 2A and 2B. The secondary models were oriented parallel to line 2 (azimuth N54°E) as this was both the best-constrained unbroken model line and a good approximation to the direction of plate convergence so that deeper subduction-related features were assumed to occupy approximately the same position on each line and not require any geometrical corrections. 4.3.1 Line IAS The observed data for line IAS were easily satisfied with a slightly modified version (figure 4.14) of the line 1 model. The 10 km offset between the lines saw an increase in the thickness Figure 4.14: Line IAS final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data vo and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 96 of the Torino Basin sediments, especially beneath the western half of the continental shelf, and an increase in the size of the anticline above the deformation front. Both of these changes were imaged on the USGS seismic reflection line which was shot 5-20 km south of the line IAS position. Both the Crescent and Pacific Rim Terrane units appear similar on both lines, except that the upper surface of the Crescent Terrane has been flattened and lengthened on line IAS as per the seismic reflection data, a change which provides an excellent fit to the magnetic anomaly data. The upper crustal structure beneath Vancouver Island has been modified to account for changes indicated by surficial geology, with the most noticeable changes being to the highly susceptible intrusive units which now outcrop at the surface (near 165 and 202-217 km) and contribute significantly to the magnetic anomaly signature. The sharp gravity anomaly peak observed near 240 km distance on line 1, and attributed to a shallow source near Texada and Lasqueti Islands, is understandably absent on the line IAS anomaly curve as, due to the segmented geometry of line 1, the two models being compared here actually diverge across the island so that their separation increases from 3 km at the west coast to 38 km near the east coast of Vancouver Island. This variation has been considered in the lower part of the model and differences in, for example, the shape of the high density wedge (3140 kg/m3) are direcdy due to geometrical corrections for the differences in line azimuths. Another difference between the two models is in the shape of the Wrangellia - Coast Plutonic Complex boundary. Whereas on line 1 the CPC was modelled with a 15 km-wide "arm" of 1-4 km thickness before the bulk of the lower density structure was encountered, on line IAS an 8-10 km thickness is required beneath the surface contact. This thicker section continues uniformly for 17 km until the unit abruptly thickens to approximately 20 km depth as per line 1. The structural differences are reflected in the gravity anomaly curves, which for line IAS decreases monotonically without the local peak (above the thinning arm) that is observed on line 1 at 280 km and on several other lines). The lower crustal structure and subducting plate structure are equivalent for the two CHAPTER 4. POTENTIAL FIELD MODELLING 97 lines, with allowances for geometrical corrections. 4.3.2 Line 1A Line 1A is located 10 km northwest of line 1 at its seaward end, and 20 km northwest of line IAS, but is crossed by line 1 halfway across Vancouver Island. At the east coast it lies equidistant between line 1, 20 km to the north, and line IAS to the south (see figure 4.1). The observed data and calculated model (figure 4.15) are remarkably similar to the line 1 model and data, with most of the differences concentrated at the eastern end near the Strait of Georgia. The Texada - Lasqueti Island gravity anomaly does not yet appear on the observed data and the observed peak at 260-280 km model distance is nicely matched with just typical Karmutsen material between the channels of the strait. The contact with the Coast Plutonic Complex is much steeper and more abrupt on line 1A than on line 1, an observation in keeping with the contact modelled on line IAS to the south. A block of lower density granodiorite has been emplaced within the predominantly quartz diorite CPC to produce the steep-sided local negative gravity anomaly at 310 km. The surface dimensions of this block match a mapped surface exposure (north of Howe Sound); however the shape and density contrast are poorly constrained and this level of detailed modelling will not be pursued within the CPC other than for lines affected by this interpreted body. The calculated magnetic anomaly over the Crescent Terrane at 110 km distance underfits the observed anomaly by 50 to 80 nT. The calculated response could be increased by moving the upper surface of the basaltic slab closer to the surface, by thickening the upper part of the slab, or by increasing the susceptibility of the unit. The first two corrections would also increase the calculated gravity anomaly, which is already too high, while the third would alter a value that has been considered constant for all models. The slab has therefore been left at its present position pending additional information on its position or magnetic properties. Figure 4.15: Line 1A final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data \o and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 99 Other large magnetic anomalies have been matched reasonably well based on their mapped surface extent and geometries previously explored on lines 1 and IAS. 4.3.3 Line IB Line IB lies 10 km north of and parallel to line 1A. The observed magnetic anomaly profiles for the two lines are very similar and similar source bodies have been used to derive the calculated anomalies (figure 4.16). The increased separation between the Crescent Terrane and west coast diorite magnetic anomaly peaks has been modelled by increasing the width of the lower susceptibility Pacific Rim Terrane to 25 km beneath the shelf sediments. This gives a good fit to the observed gravity anomaly as well. Minor variations in the shape of the observed gravity profile occur over the continental slope, which displaces almost 15 km of the shelf compared with line 1A and hence reduces the gravity anomaly. The peak at 250 km model distance is similar to that observed on line 1 and the minor misfit represents the southern end of the unmodelled anomalous density structure beneath Texada and Lasqueti Islands. East of this peak a pronounced gravity low has been modelled as a sharp transition to a 5 km-thick block of CPC intrusives that thins rapidly eastward to less than 1 km thickness to produce the peak near 280 km distance, and then thickens abruptly to 20 km thickness td mark the main plutonic body. The lower density granodiorite body has again been included, this time with the uppermost 3 km replaced by standard CPC density, and gives a good fit to the observed anomaly. The lower crustal structure is consistent on lines 1A and IB. 43.4 Line 1C Line 1C lies 10 km to the north of lines IB and its continental mainland component is also coincident with that part of line 1. The observed gravity anomaly continues to respond to the widening continental slope, narrowing shelf configuration by flattening the seaward part Figure 4.16: Line IB final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data and solid line is calculated anomaly. Figure 4.17: Line 1C final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data o and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 102 of the positive anomaly peak (figure 4.17). The widening of the second negative anomaly, at 120 km distance, appears to reflect an increase in the thickness of sediments within Tofino Basin and an increase in the depth of the higher density Crescent slab. The decreasing amplitude of the observed magnetic anomaly, from 200 to 150 nT, also reflects a deepening of the volcanic material. The Pacific Rim Terrane is again quite wide on this line and outcrops on the west coast of Vancouver Island between 141 and 143.5 km model distance. The large density contrast between this unit (2900 kg/m3) and the adjacent water and sedimentary material (1030 and 2520 kg/m 3 respectively) produces a steep positive slope near 140 km on the observed gravity profile. An adjacent lower density unit (2650 kg/m 3, Sicker Group and/or quartz monzonite) produces a sharp gravity "plateau" between 140 and 150 km distance that is well-matched by the geometry shown. Note that the observed magnetic anomaly profile is quite flat over the Pacific Rim Terrane outcrop and the next peak begins several kilometres to the east over the diorite unit. This observation is important in the re-interpretation of the 84-01 seismic reflection section and will be discussed further in the next chapter (refer also to section 4.2.1). The model crustal structure beneath Vancouver Island is quite similar to that of the line IB model, with a decrease in the size and magnetic anomaly amplitude of the intrusive body beneath eastern Vancouver Island, which is no longer exposed at the surface but is inferred to lie beneath the Nanaimo Formation. The magnitude of the gravity anomaly centred over Texada and Lasqueti Islands in the Strait of Georgia at 250 km has also increased, but as previously mentioned the presumed iron ore source body has not been included in the model. The shape of the CPC contact matches that used for line 1, since the lines are at this point coincident, and gives a reasonable fit to the observed gravity anomaly. The lower continental crustal structure remains consistent from line IB to line 1C. Figure 4.18: Line ID final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 104 4.3.5 Line ID The model and anomaly data for line ID are shown in figure 4.18. The offshore gravity anomaly data and model structure are virtually identical to those of line IC 10 km to the south. The magnetic anomaly data reveal the one difference, which is a further 50 nT reduction in the amplitude of the magnetic anomaly at 115 km and the resultant thinning of the Crescent Terrane slab that was employed with moderate success to try to generate a matching anomaly. Although the Pacific Rim Terrane again outcrops along the west coast of the island, the gravity anomaly data require significantly lower densities just beneath the surface. Accord-ingly, a 6 km-thick block of Sicker Group rocks (2770 kg/m3) has been emplaced beneath and landward of an overlapping slice of Pacific Rim material, satisfying both the magnetic and gravity anomaly data. Further eastward, a 4 km-thick unit of quartz diorite reproduces the observed 150 nT magnetic anomaly at 160 km model distance. A similar, but thicker, body produces the steep-sided 200+ nT anomaly at 200 km. The large amplitude magnetic anomaly at 225 km has decreased in amplitude and broadened since line IC, and a matching calculated anomaly has been derived by widening the source body and increasing its surface depth beneath eastern Vancouver Island. The remaining structure is essentially the same as for line IC. The Strait of Georgia gravity anomaly at 255 km is of the same order of magnitude although it is slightly narrower and the centre-of-mass of the source body presumably lies in the 10 km range between lines IC and ID. The contact with the CPC is of similar shape as on lines 1 and IC, although the western arm has thickened slightly to 5 km. Once again, the lower crustal structure remains the same as on previous lines. Figure 4.19: Line IE final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data o and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 106 4.3.6 Line IE Line IE lies 10 km south of line 2 and was derived from both the line 1 and line 2 models. The continental shelf has widened again slightly, and the slope has narrowed, giving a larger amplitude gravity anomaly (figure 4.19) than was observed on neighbouring lines to the south. The line crosses over the crest of an anticline on the shelf that ties in with the diapiric structure observed on seismic line 85-02. Three boreholes also lie within 5 km of the line (figure 3.1) and provide good constraints on sediment thickness and depth to the top of the Crescent slab. The small local gravity anomaly at 110 km could not be modelled adequately using just the seafloor topography over the anticline, so a density increase, from 2520 to 2750 kg/m3, was used in the diapiric unit to produce the calculated anomaly. The diapir/anticline lies above the seaward edge of the basaltic Crescent Terrane slab, as previously mentioned and confirmed by the gravity and magnetic anomaly data, and a process of trapped and guided fluid flow has been suggested in a previous section. The Pacific Rim Terrane is again modelled as a fairly wide ( 20 km) unit with a narrow outcrop which overlaps a lower density (2650 kg/m3) unit and abuts against a thin, higher susceptibility unit composed of the diorite and quartz diorite bodies that outcrop along the line. Another quartz diorite body produces the 200 nT anomaly seen at 200 km near central Vancouver Island, and a narrower version of the intrusive body beneath the eastern part of the island satisfies the 400 nT anomaly at 225 km model distance. A negative gravity anomaly at 245 km distance seems to require an increased thickness of Nanaimo Formation sedimentary rocks, over 1 km, beneath the Strait of Georgia. The sharp anomaly east of this is the Texada - Lasqueti Island anomaly, which has again not been modelled due to a total lack of constraints, and a broad negative on its east side marks the contact with the Coast Plutonic Complex beneath Malaspiha Strait east of Texada Island. A thin (2 km) upper arm, similar to that on line 2, is used for the CPC until the abrupt CHAPTER 4. POTENTIAL FIELD MODELLING 107 thickening and corresponding decrease in the gravity anomaly occur near 300 km model distance. 4.3.7 Line 2A Line 2A lies 40 km north of line 2 (see figure 4.1). Despite the considerable offset, the observed anomaly data and derived models are quite similar for both lines. The oceanic crust was placed at a slightly shallower depth on line 2A (figure 4.20), in keeping with the trend observed between the three seismic reflection lines and the CSP profiles. A thinner accumulation of accretionary material lies beneath the continental slope, and the gravity profile is characterised by a much broader negative anomaly than on the lines to the south. Within the wedge, the Crescent Terrane still appears to be present, although the corresponding magnetic anomaly has been reduced an additional 50 nT and is now only 50-100 nT above the background level. The anomaly is also narrower than on line 2, so accordingly the basaltic slab unit was thinned and the upper surface placed at greater depth. The adjacent negative anomaly suggests that the Pacific Rim Terrane is also present, although its width has been decreased slightly to 12 km. A small magnetic anomaly at the west coast of Vancouver Island (145 km model distance) characterises the surface-mapped quartz diorite unit, with the shape of the curve affected by the water within the inlet that is crossed by the fine. The differences in the continental gravity anomaly curves for lines 2 and 2A were easily satisfied by consulting the geologic map. A negative anomaly at 205 km was satisfied by the surface exposure of the lower density (2770 kg/m3) Sicker Group rocks. A return to Karmutsen-type densities (2930 kg/m3) restored the higher anomaly levels. The next anomaly low, at 250 km, lay over the Nanaimo Formation and an 800 m thickness of this unit successfully modelled the anomaly. The adjacent positive anomaly was not matched by a return to Karmutsen material beneath the Strait of Georgia, and the source of the Texada Figure 4.20: Line 2A final model. Line location is shown in figure 4.1. Densities in kg/m3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 109 - Lasqueti Island anomaly may still be affecting the gravity anomaly field. The onset of the Coast Plutonic Complex was modelled with an upper crustal arm of up to 6 km thickness, compared with 2 km on line 2, and the bulk of the complex was east of 315 km. Once again, the lower crustal structure satisfied the gravity data without changes being required. 4.3.8 Line2B Line 2B lies 40 km north of line 2A and is almost coincident with the Nootka Fault zone, thereby becoming the final model on the Juan de Fuca Plate (see figure 4.1). The observed gravity and magnetic anomaly data differ considerably from the data observed further south. The offshore gravity anomaly is characterized by a broad trough with the former shelf edge high limited to a 20 mGal anomaly at 105 km (figure 4.21). A smaller anomaly at 65 km marks a small anticline on the lower continental slope. Thick sedimentary sequences underlie the continental shelf, as confirmed by the Apollo J-14 borehole approximately 8 km south of the line at 112 km model distance. A higher density was used for the sediments than on previous lines (2600 and 2650 versus 2520 kg/m3). It is quite possible that the Apollo well has given anomalous thickness values, as it sampled an anticline and not the typical layers of sediments, and a thinner, lower density sedimentary sequence may be present. The magnetic anomaly previously associated with the Crescent Terrane is absent on line 2B. A slab unit was included in the model to test whether the terrane may in fact be present but at a burial depth great enough to suppress the magnetic signature. The depth and geometry shown would generate a +30 nT anomaly (not shown) using the susceptibility (18800 SI units) previously assigned to the unit, so a greater burial depth would be required. However, the density used for the terrane on the other lines (2780 kg/m3) generates a gravity misfit of -10 mGal and this would be increased further if the body were more deeply buried and low density sediments remained the overlying material. Therefore, the properties of the Figure 4.21: Line 2B final model. Line location is shown in figure 4.1. Densities in kg/m 3 and magnetic susceptibilities in SI units (italic type, in brackets where poorly constrained) are shown for each block. Vertical exaggeration is 1.3x. Symbols represent observed data and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 111 Pacific Rim Terrane material (2900 kg/m 3, 377 SI units) have been assigned to the unit in the model shown, effectively terminating the linear Crescent Terrane unit that has been followed along the coast. It is stressed that the results are inconclusive and the Crescent Terrane may be present but, if so, it is probably more deeply buried here than in the southern region. A 300+ nT magnetic anomaly marks the contact of the diorite unit at the west coast of Nootka Island. The gravity anomaly over Vancouver Island is markedly different than on the southern lines, showing instead a more subdued character with a 60 km-wide low over the western half of the island that appears related to a considerable quantity of intrusive granodiorite and quartz diorite beneath this part of the island. A return to Karmutsen rocks, as indicated by the surface geology, restores the anomaly curve to its higher level. The contact with the Coast Plutonic Contact, which for once is exposed at the surface, marks the start of the general decrease in gravity anomaly values. An initial CPC thickness of 2 km, thickening rapidly to 10 km, and the abrupt increase to a full 15 km thickness at 318 km, gave a good fit to the observed anomaly curve. A 20 mGal misfit just west of the exposed CPC contact may have a similar source to the positive anomaly observed on most of the lines over the central Strait of Georgia. 4.4 Southern Vancouver Island Anomaly 4.4.1 Line S Previous interpretations and constraints A model (line S) was constructed crossing from the Olympic Peninsula to the CPC east of Vancouver Island, over the large positive Bouguer anomaly on southern Vancouver Island to try to obtain quantitative control on the source of this major anomaly. The model line followed a profile previously interpreted by Brandon et al. (1984) and revised by Clowes et CHAPTER 4. POTENTIAL FIELD MODELLING 1 Figure 4.22: Structural cross-section from the Olympic Mountains to southern Vancouver Island (modified from Clowes et al. 1987a; refraction velocities are from Taber 1983). al. (1987a) to incorporate the interpretation of seismic reflection line 84-02. A geologic depth section illustrating the Clowes et al. (1987a) interpretation is shown in figure 4.22. The basic near surface units that have been included, based on the previous interpretation of Brandon et al. (1984), are the accreted Core Rock complex of the Olympic Peninsula, the Eocene volcanics of the Crescent Terrane, the metasedimentary Leech River Schist of the Pacific Rim Terrane, and the continental framework represented by the Wrangel-lia Terrane. Clowes etal. (1987a) added their interpretation of seismic line 84-02, which as shown has been revised to place the Juan de Fuca Plate beneath the ' F ' reflector as more re-cent reinterpretations have indicated. Velocity values from a seismic refraction study (Taber 1983) help constrain the lower crustal structure beneath the Olympic Peninsula. Figure 4.23: Line S final model. Line location is shown in figure 4.1. Densities in kg/m 3 are shown for each block. Solid triangle marks the location of the Quaternary volcano Mt. Garibaldi. Vertical exaggeration is 2.6x. Symbols represent observed data and solid line is calculated anomaly. CHAPTER 4. POTENTIAL FIELD MODELLING 1 Line S final model The gravity model extended from the oceanic basin west of Oregon to the British Columbia mainland east of central Vancouver Island and passed directly over Mt. Garibaldi to locate the volcanic arc. The orientation of the line places it approximately perpendicular to the strike of the Crescent Terrane outcrops and bounding faults on both southern Vancouver Island and the Olympic Peninsula. The line orientation used in the previous models is oblique to these features and 2.5-D modelling would not have been representative of the structure along such a profile. The location of the previous structural model (figure 4.22) corresponds to distance range 0-129 km in this model. The offshore gravity structure was based on bathymetry and utilized the three density units used on the other models (2290, 2410 and 2610 kg/m3) to provide a good fit to the anomaly curve (figure 4.23). A density of 2770 kg/m 3 was used to represent the Core Rocks of the Olympic Peninsula, which was not subdivided into smaller units. The Coast Plutonic Complex was constrained by other models which were crossed by this north-northeastward trending line. A seismic refraction study in the Strait of Georgia (White and Clowes 1984) provided estimates of sediment thickness for this portion of the line. The subducting slab was constrained by a seismic refraction model across Washington State (Taber and Lewis 1986), seismic line 84-02 on southern Vancouver Island, plate depth contours based on seismicity (Crosson and Owens 1987), and the position of the volcanic arc. Beneath Vancouver Island the model structure was primarily derived from seismic line 84-02 and extended to match the other models to the north. The interpreted Survey Mountain and Leech River Faults constrained the Leech River schist unit (2900 kg/m3) to a thin band dipping at approximately 27° to 14 km depth at which point it is truncated as per Clowes et al. (1987a). The Metchosin basalts (Crescent Terrane) have been interpreted as a wedge CHAPTER 4. POTENTIAL FIELD MODELLING 1 model is simply an "average density" representation. Beneath the highly reflective zone lies the non-reflective high velocity zone of Clowes et al. (1987a), which is interpreted in the model as a unit of 3140 kg/m 3, consistent with the other models to the north. Beneath this block lies another highly reflective zone, labelled as the ' E ' zone and probably the same zone as imaged on seismic line 84-01 across Vancouver Island. However, on this line it is modelled as a block overlying the subducting plate and having a density of 2850 kg/m 3. Its shape and density are not constrained by the gravity model and could be traded off against each other to produce other geometries. The crustal structure east of the seismic-constrained region consists of a 10 km-thick upper unit of density 2860 kg/m 3, representing an average of the Colquitz and Wark gneisses, the Sicker Group, and the granodiorite bodies that are crossed by the line. A standard crustal density (2930 kg/m3) was used beneath this, terminating downward against the extrapolated high reflectivity ' C zone. Small units near the surface represent the Nanaimo Group (2550 kg/m 3), deltaic sediments (2400 kg/m3), and the Strait of Georgia (1010 kg/m3). The derived model, while poorly constrained in the lower crustal region, nonetheless shows that the anomalous Bouguer gravity anomaly observed over southern Vancouver Island can be satisfied by assuming "that the portion of the Crescent Terrane that underlies this region includes lower oceanic crustal and/or upper mantle material. The total thickness of the oceanic slab as modelled is approximately 6 km, of which the uppermost 2 km or so represent pillow basalts and other lower density (2880 kg/m3) material. The lower unit has been modelled with a single density of 3200 kg/m 3 but represents both the gabbroic lower crust and peridotite upper mantle components. The very good fit to both the amplitude and width of the observed anomaly suggests that the configuration shown is probably a good representation of the shape and relative composition of the Crescent Terrane in this region. CHAPTER 4. POTENTIAL FIELD MODELLING 115 tapering from a thickness of over 6 km on the southern half of the seismic line to less than 3 km at the northern end, with the uppermost surface resting against the Leech River Fault (Clowes et al. 1987a). In the model this forms the northern arm of an anticline that crests at the southernmost tip of Vancouver Island and dips slightly downward beneath the Juan de Fuca Strait (figure 4.23). The unit turns upward again south of the Clallam Syncline on the northern Olympic Peninsula and is exposed at the surface as the Crescent Formation before being truncated to the south by the Hurricane Ridge Fault. The Crescent Terrane has been divided into two units which together reproduce the major gravity anomaly very well. The upper unit (2800 kg/m3) corresponds to the basaltic members mapped on Vancouver Island and the Olympic Peninsula as the Metchosin basalts and Crescent Formation respectively. This unit has a maximum thickness of 2 km in the model. Beneath them, and exposed at the crest of the anticline (Sooke gabbro) on the tip of Vancouver Island, is a 4-5 km-thick unit interpreted as lower oceanic crust / upper mantle and assigned a density of 3200 kg/m 3 which is slighdy above the measured range for gabbro but is well within the range for peridotite, constituents of lower crust and upper mantle, respectively (Christensen and Smewing 1981). Beneath the Crescent Terrane units lies the highly reflective zone, denoted the ' C - D 2 ' layer by Clowes et al. (1987a) and interpreted by them as possibly being part of the Core Rocks. The gravity model suggests that a fairly high density (2900 kg/m3) is required for this unit, which seems slightly high for sedimentary rocks at a reasonably shallow depth. It may be that the zone is a combination of higher density oceanic crustal material and accreted sediments. The zone is constructed in the model so as to lead into the ' C zone (2850 kg/m3) beneath the Coast Plutonic Complex at 215 km distance. However, it seems highly unlikely that the two zones as shown are related in this manner, and a more complicated structure is probably present. Note that the lower crustal geometry shown was based solely on the seismic reflection data and is otherwise not constrained. Therefore the derived lower crustal CHAPTER 5 FINALE 5.1 Model Summary This study has examined the structure of the northern Cascadia subduction zone by modelling the gravity and magnetic anomaly fields and concurrently satisfying other geophysical and geological constraints. While it must be stressed that the developed models are not unique and other models may also satisfy the data, the number of data that can be satisfied by these models lends credence to their reliability. The density models, being better constrained than the susceptibility models, have been examined and are summarized in two ways: as an "average" model across the most closely modelled portion of the margin, and as a three-dimensional regional model that shows the transitions along the length of the margin. The average model was constructed from seven parallel models (lines IAS to 2, excluding line 1) to satisfy a gravity anomaly curve generated by averaging the corresponding observed data profiles. The regional model was formed by interpolating between the eleven across-margin profiles to produce a gridded data "cube" representative of a structural block of the margin. 5.1.1 Average density model Figure 5.1 shows the average gravity anomaly profile and the seven profiles from which it was created. The averaging process, in a manner similar to the stacking of seismic traces, reduced the magnitude of small-scale variations and amplified features that were common to 117 CHAPTER 5. FINALE 1 AVERAGE GRAVITY ANOMALY PROFILE — 150 L—i—i—i—i—I—i—i—i—i I i i i i I "i i" 0 100 200 300 Model Distance (km) Figure 5.1: Average gravity anomaly profile (solid line) obtained by averaging profiles for lines IAS to 2 (dashed lines), excluding line 1. all curves. Minor anomalies associated with the topography of the continental shelf and slope have been averaged out and the sharp anomalies associated with the Nanoose Uplift on lines IAS and 1A and the gravity high near Texada and Lasqueti Islands in the Strait of Georgia region on all other lines have been combined into a single anomaly of reduced amplitude (15 mGal) at 250 km model distance. Note that this procedure could not be applied to the magnetic data for several reasons, such as the oblique orientation of the near-surface source bodies with respect to the profiles which would tend to produce smeared anomalies rather than a representative anomaly profile if the averaging process were applied. The average density model is shown in figure 5.2 together with the calculated and ob-served gravity anomaly curves. This generalized model has the same sedimentary and lower crustal features as used on the individual models, and the upper crust beneath Vancouver Is-land has been simplified to five basic blocks representing the dominant geologic components. Densities generally increase eastward, as the lower density intrusives are concentrated at the Figure 5.2: Average density model across the margin between lines IAS and 2. Densities are shown in kg/m3. Solid line is calculated anomaly, symbols are averaged observed data. Interpretation of various density units is shown in lower model. CHAPTER 5. FINALE 120 western edge. The units are, moving eastward across the island region: metamorphosed dior-ite (p=2860 kg/m3), responsible for the west coast magnetic anomaly; a predominantly upper Wrangellia unit of Bonanza and Karmutsen rocks (2900 kg/m3); a lower Wrangellia unit of primarily Sicker Group rocks (2870 kg/m3), representing the crest of the Cowichan Uplift; and predominantly Karmutsen Formation material (2930 kg/m3), leading to the contact with the Coast Plutonic Complex beneath the eastern edge of the Strait of Georgia. Within this last unit is a block of granodiorite / quartz diorite (2900 kg/m3), which generates a large magnetic anomaly. Thin overlying units of water (1010 kg/m3) and Nanaimo Formation sedimentary rocks (2550 kg/m3) generate local negative anomalies near 240 and 260 km model distance. The small combined gravity anomaly peak separating them at 250 km could not be adequately matched by a generalized model. Figure 5.3 compares the average model, to 200 km depth, with a generalized model developed by Riddihough (1979) (location shown in figure 3.6). The constraints used by Riddihough and shown as solid bars in figure 5.3b have also been superimposed on the average model. The subducted plate position beneath approximately 70 km cannot be re-solved through gravity modelling without additional information and is therefore not well constrained on either model. Note that whereas Riddihough used density variations within the upper mantle to model the observed anomaly, this study concentrated anomalous struc-ture in the lower crust. Riddihough's (1979) model produced a good fit to the data, and the subsequent availability of additional constraints has allowed a more detailed model to be developed in this study. This model in turn will undoubtably be refined in the future as analysis of the subduction zone continues and newer data become available. The assumptions of two-dimensional structure and the appropriateness of 2.5-D modelling in this region have been validated by the good fit between the model and the average gravity anomaly profile. This average model is a good representation of the margin structure between Clayoquot Sound and Pachena Point just south of Barkley Sound. Figure 5.3: Comparison between the average density model, (a), and the coincident simple model of Riddihough (1979), (b). CHAPTER 5. FINALE 122 5.1.2 Regional density model A three-dimensional density model representing a 230 km-long swath of the margin was constructed from the ten models (lines IAS to 4, line 1 omitted) oriented perpendicular to the margin. A data "cube" was constructed with 5x5x2 km grid sizes for a total overall size of 350x230x50 km and a 2.5x vertical exaggeration. The individual models were sampled every 5 km in distance at 2 km depth increments. In order to maintain equal scale in the surface dimensions (length and width), a simple linear interpolation routine (J. Amor 1990, pers. comm.) was used to "create" models at 5 km intervals between the actual model lines which were 10 to 80 km apart. The grid space and original models are shown in figure 5.4. The "cube" extends north-westward so that the line IAS model is on the front face and line 4 is at the back. The grey-scale palette uses darker shades for lower densities with the exception of the water, which is shown at a lighter shade so as not to obscure the sedimentary structures. The lightest feature shown is therefore the oceanic mantle at a density of 3340 kg/m 3. This figure may be compared with the individual models in Chapter 4 to identify the various density units. Figure 5.5 shows the complete model "cube" to examine density variations and trends along the outer surfaces. The upper surface reveals the shelf and slope growing narrower towards the north, the northwest-trending density structures at the surface of Vancouver Island, and the surface contact of the Coast Plutonic Complex. The east side of the block (right face) shows a relatively uniform increasing-downward density structure along the margin, with the four layers representing the CPC, the underlying crust, the high density unit bounded by the thin low density layers, and the continental mantle. The model block can be "sliced" along any plane to examine the structure at any depth or distance. Figure 5.6 shows three vertical slices through the model, approximately 100 km apart, to show structural variations along the length of the margin. It can be seen that the first CHAPTER 5. FINALE 123 Figure 5.4: Locations of ten derived density models within the regional model block. (Refer to figure 4.1 for line locations.) Line 1 was omitted due to crooked line geometry. Colour scale ranges from blue for lowest density through yellow to dark brown for highest density. Vertical exaggeration is 2.5x. CHAPTER 5. FINALE 124 Figure 5.5: Regional block model for the margin along the northern Cascadia subduction zone. Colour scale ranges from blue for lowest density through yellow to dark brown for highest density. Vertical exaggeration is 2.5x. CHAPTER 5. FINALE 125 Figure 5.6: Vertical slices through the regional block model showing lateral variations in density along the margin. Colour scale ranges from blue for lowest density through yellow to dark brown for highest density. Vertical exaggeration is 2.5x. CHAPTER 5. FINALE 126 two lines are quite similar, especially in the lower continental crust, but the northernmost line differs considerably. The subducting slab (Explorer Plate) enters the accretionary zone at a much shallower depth on this line (line 4) than in the south (Juan de Fuca Plate) and this both affects the geometry of the accretionary wedge and displaces part of the lower continental units. The upper continental crustal structure (above the ' C zone position) and Coast Plutonic Complex remain unaffected. 5.2 Results and Implications The derived model has revealed and examined several interesting features related to the structure of the margin. These include the offshore extension of the Crescent Terrane, the contact between the diorite unit at the west coast and a lower susceptibility unit (Pacific Rim Terrane (?)), the geometry of the Wrangellia — Coast Plutonic Complex contact, the source of the gravity high of southern Vancouver Island, and the nature of the material above the subducting slab as imaged by seismic reflection lines. The tectonic summary map (figure 5.7) shows the lateral extent of the Crescent and Pacific Rim Terranes, as determined in this study, along the accretionary margin of the northern Cascadia subduction zone. Also shown is the derived boundary of the batholith representing the bulk of the Coast Plutonic Complex. 5.2.1 The Crescent Terrane The magnetic anomaly associated with the Crescent Terrane can be followed just to the north of line 2A, at which point the peak is no longer detectable, above background levels. Magnetic susceptibility modelling has confirmed the presence of a unit possessing appropriate susceptibility beneath the shelf sediments to this point. The magnitude of the calculated anomaly was dependent on both the thickness and burial depth of the unit, while the anomaly shape was dependent on the dip and topography of the upper surface of the unit. The CHAPTERS. FINALE 111 Figure 5.7: Tectonic summary map showing the lateral extent of various geotectonic units as determined in this study. Mapped or interpreted faults are marked as bold solid lines (dashed where inferred). Fault abbreviations are listed in the figure 1.2 caption. The short dashed lines indicate the mapped Wrangellia - Coast Plutonic Complex surface contact; longer dashes (marked "BATHOLITH") show the position at which the complex extends to crustal thickness. CHAPTER 5. FINALE 128 presence of the Crescent Terrane beneath line 2B could neither be confirmed nor denied through modelling, as the terrane may have been too deep to produce a detectable anomaly. The calculated gravity anomaly gave a better fit if the Crescent unit was not included in the line 2B model but this was not conclusive as near surface material had a much stronger influence on the gravity anomaly and the misfit was not large. Therefore it may be concluded that the Crescent Terrane extends northwestward along the margin to at least 49°N latitude and, if present very much north of this, must be at a burial depth in excess of 4 km. The geometry of the terrane along this part of the margin resembles a thin linear slab that dips to the northeast at 20-35° beneath the adjacent Pacific Rim Terrane and western edge of Wrangellia. In the vicinity of Juan de Fuca Strait, the slab has been uplifted into a broad anticlinal structure that produces a 50 km-wide outcrop across the northern Olympic Peninsula and southern Vancouver Island and brings lower crustal / upper mande rocks to very shallow depths. If it is assumed that the Crescent Terrane is continuous between all lines on which it was modelled and the surface outcrops which extend at least to southwestern Oregon (Johnson et al. 1984), then the Crescent Terrane comprises a slice of oceanic crust at least 700 km long. This suggests minimum limits on the dimensions of the tectonic environment, such as a marginal basin setting (Massey 1986), in which the terrane was formed. 5.2.2 The Pacific R i m Terrane The Pacific Rim Terrane was included on all model lines as a unit of 2900 kg/m 3 density and width averaging 15 km, increasing in width to 25 km where exposed along the west Coast of Vancouver Island between Barkley Sound and Clayoquot Sound. The outboard Tofino Fault remains relatively straight past this region but the Westcoast Fault moves sharply inland near the northern edge of Barkley Sound, .resulting in the increased width of the unit. The unit CHAPTER 5. FINALE 129 was extended to a depth of approximately 15 km, which brought it to the vicinity of the 'C zone, and the westernmost part of the high velocity / high density block was emplaced beneath it. The gravity anomaly generally showed a steep positive slope over the unit as it marked the transition from the lower density accreted sediments to higher density material. The magnetic anomaly was either null or slightly negative, including over the exposed volcanic Uclufh Formation north of Barkley Sound. The contact between the terrane and Wrangellia coincided with a sharp magnetic anomaly that was attributed to the metamorphosed diorite complex that was mapped along much of the west coast of Vancouver Island. Thus the combination produced a magnetic low/high anomaly pattern that could be followed along the entire west coast of Vancouver Island past the Scott Islands at the northern end, and leads to the conclusion that the Pacific Rim Terrane is continuous along the margin. The geometry of the modelled contact between the terrane and Wrangellia is strongly supportive of the interpretation (e.g. Johnson 1984; Brandon 1989b) of emplacement of the Pacific Rim Terrane by strike-slip faulting along the margin. 5.2.3 The C P C contact .. The nature of the contact between the units used to represent the Wrangellia Terrane and the lower density unit used to represent the Coast Plutonic Complex has proven to be very interesting. A good fit to the gravity anomaly curve is attained if only a thin layer of lower density CPC material is used for the first 20-30 km east of the surface contact and then the CPC unit is abrupdy thickened. Additional support for the interpreted geometry of the contact comes from a study of seismic events recorded at a station located near Egmont on the British Columbia mainland 30 km northeast of Texada Island. Cassidy (1991) analysed the associated teleseismic receiver functions and interpreted a boundary at 6.1 km depth, CHAPTERS. FINALE 130 dipping 15° to the southwest (S30°W). The station lies midway between model lines 1C and ID at 289 km distance. The density unit boundary is at 1.5 to 2 km depth at this point, dipping to the southwest at 7.5° and increasing in thickness to 5 km within 20 km distance. Although the agreement is not exact, the arrival at a similar interpretation by two independent methods is supportive of the existence of such a boundary. The location of the bulk of the CPC batholith, near 300 km on most models, also corresponds closely with the sudden increase in heat flow measured across the margin. Lewis and Bentkowski (1990) contoured the heat flow pattern for the southwestern margin of British Columbia and the sudden increase in heat flow observed seaward of the volcanic arc matches quite well with the modelled location of the edge of the batholith. This correlation may be purely coincidental, or may be related to differences in thermal conductivity or fluid circulation between the different materials. The batholith may possess latent heat from its emplacement or more recent activity and hence the correlation between higher heat flow and the density model may represent actual structural and compositional boundaries. 5.2.4 The southern Vancouver Island anomaly The results of gravity modelling across the high amplitude Bouguer gravity anomaly at the southern tip of Vancouver Island are strongly indicative of lower crustal (gabbroic) and upper mantle (peridotite) layers of the ophiolitic Crescent Terrane lying at shallow depths beneath the surface. Together, the mapped presence of a middle crust sheeted dike complex and gabbroic stocks, the inferred terrane thickness derived by Massey (1986), and the interpretation of Brown and Hanna (1971) of a high susceptibility, gabbroic-like body underlying Juan de Fuca Strait all support the interpretation shown in figure 4.22. The slab of Crescent Terrane lying under the southern tip of Vancouver Island appears to have been uplifted from the modelled dipping slab geometry of the continental shelf region into an CHAPTER 5. FINALE 131 anticlinal structure that dips beneath the surface to the northeast along the Leech River Fault. The southern limb of this broad structure forms the northern part of the Clallam Syncline beneath the northern Olympic Peninsula. 5.2.5 The reflective zones The nature of the highly reflective low velocity bands, or even their corresponding existence as low density zones, could not be resolved through gravity modelling, primarily because of their considerable depth with respect to their size. One important point should, however, be pointed out with respect to the shallower ' C zone. The revised location of the Westcoast Fault derived on the line 1 and adjacent models in this study implies that the terrane-bounding fault no longer cuts through the ' C zone, as previously interpreted (e.g. Hyndman et al. 1990), but leads into the western end of the zone as imaged on reflection line 84-01. The ' C zone may thus be a feature related to a single terrane or other tectonic unit. In fact, one can now speculate that the ' C zone marks the base of the Wrangellia Terrane and the higher velocity/density unit beneath it represents mantle-derived material or blueschist metamorphic facies material subcreted to the base of the terrane. This would explain the existence of the ' C zone over a wide region, as for example, its interpreted presence on the seismic line 88-16 (Clowes 1990) at the western edge of the British Columbia mainland. 5.2.6 The Explorer Plate region Significant differences have been noted between the lower continental crustal structure of south-central Vancouver Island and that of the northern part of the island. The shallower depth of the Explorer Plate as it enters the subduction zone has eliminated the low density ' E ' zone and underlying material from the crustal model. In addition, higher densities are required for the ' C zone layer and its extension, effectively eliminating them from this region CHAPTER 5. FINALE 132 as well. An alternate model would require considerable thinning of the continental crust to reproduce the observed gravity anomaly. In both cases, the transition from the "normal" crustal structure used on the lines to the south occurs between model lines 2A and 4, at or north of the Nootka Fault zone. 5.3 Summary Several generalizations can thus be made about the structure of the northern Cascadia sub-duction zone based on the results of this study. First, it is quite apparent that there is a concentration of high density material beneath western Vancouver Island in comparison with the eastern side. Riddihough (1979) used a higher density mantle wedge between the con-tinental Moho and the top of the subducting slab to generate the required anomaly. The models in this study concentrated the higher densities within the lower continental crust beneath 10 km depth. The cause of the higher densities is undoubtably linked to the subduc-tion process, but the exact source will remain in question until the nature of the anomalous velocity/density/conductivity banding above the subducting plate is resolved. Wrangellia has been considered in this study to include the crustal rocks above the 'C zone beneath Vancouver Island and the Strait of Georgia. This terrane appears to be relatively uniform in both thickness and average bulk density along the length of Vancouver Island. The major structural contribution of this study is related to the accreted material underneath and west of Wrangellia. Johnson (1984) had proposed that a major strike-slip fault truncated the western and southern margins of Vancouver Island during the Late Cretaceous. The Pacific Rim Terrane was then moved northward along the margin from the San Juan Islands region south of the Strait of Georgia by dextral transform motion and emplaced as a narrow strip along the south and west coasts (Brandon 1989b). The steeply-dipping boundary separating the modelled Pacific Rim unit from the blocks of Wrangellia in this study is consistent with CHAPTER 5. FINALE 133 a strike-slip interpretation. The terrane has been modelled as a unit continuous from the San Juan Islands to the northern tip of Vancouver Island, implying fault displacement of at least 600 km. The relatively high density, thick Pacific Rim unit extends to approximately the interpreted base of Wrangellia and probably includes crustal material beneath the Pacific Rim melanges. The Crescent Terrane also appears to have formed during a period of transcurrent motion along the margin. Massey (1986, 1990) proposed that a marginal rift basin opened behind the northward moving Kula Plate and the Eocene basalts formed as new crust within the basin. The mapped or interpreted extent of the Crescent Terrane, from model line 2A west of central Vancouver Island to at least southwestern Oregon (Johnson et al. 1984), suggests that a 700 km or greater length of oceanic crust formed in the marginal basin before extension ceased approximately 52 Ma ago. The shift in relative Pacific - North American plate motions 43 Ma ago re-established convergence along the margin (e.g. Riddihough 1982). This resulted in the detachment of the Crescent Terrane from the Kula Plate and the subsequent emplacement of the terrane by subduction beneath the margin. The modelled geometry of the Crescent Terrane, as a thin landward-dipping slab, is supportive of terrane emplacement in a subduction regime: (It should be noted that underthrusting by younger accretionary wedge sediments has sub-sequently lifted the terrane to a more steeply-dipping orientation.) However, the process by which the Crescent Terrane was brought into position adjacent to the Pacific Rim Terrane is unclear. The Crescent Terrane may have moved northward along a strike-slip fault margin that truncated the seaward edge of the Pacific Rim unit, and then was thrust beneath the margin when the plate tectonic regime changed. Alternatively, the Crescent Terrane may have underthrust the Pacific Rim unit at a highly oblique angle which moved the Crescent Terrane northward during subduction. The modelled boundary between the two terranes dips CHAPTER 5. FINALE 134 at a moderate angle (~ 20°), more typical of a thrust fault than a generally steeper strike-slip fault, but this study cannot rule out the existence of a steep strike-slip fault that was subsequendy "overprinted" by subduction. Material shown in the models as high density lower crust, including the high density / high velocity wedge beneath western Vancouver Island, appears to have formed (possibly by underplating of crustal or mande material) after emplacement of the Pacific Rim Terrane but prior to subduction of the Crescent Terrane. During this time the Kula-Farallon-North American triple junction was moving northward along the margin west of Vancouver Island. As a result, motion along the northern part of the Vancouver Island margin was obliquely convergent (between the Kula Plate and the North American Plate), while the south was experiencing a greater component of convergence and possibly subduction. Hence, this phase of subduction appears to have been initiated at an earlier stage in the south and this difference in subduction history may explain why the thickness of lower crustal material beneath Wrangellia appears greater in the south than beneath northern Vancouver Island. Along the eastern edge of Wrangellia, the models have indicated that the bulk of the lower density Coast Plutonic Complex batholith lies 20-30 km east of the mapped surface contact and that only a thin layer of granitic material underlies the region between:-The good correlation between this interpreted batholith position and the observed rapid rise in surface heat flow at this location (Lewis and Bentkowski 1990; see also figure 3.5) is supportive of a major compositional change (and possible changes in thermal conductivity and heat production) in this vicinity. The use of an integrated data set has clearly contributed to the interpretation of the structure of the margin, as it has resolved some ambiguities and forced some reinterpretations, and the method of laterally extending the interpretations as employed in this study has proven to be a useful technique for maximizing the available information. R E F E R E N C E S A R K A N I - H A M E D , J. and STRANGWAY, D.W. 1988. Interpretation of the aeromagnetic anomalies of southern Vancouver Island. Canadian Journal of Earth Sciences, 25: 801-809. A U , D. and CLOWES, R . M . 1982. Crustal structure from an OBS survey of the Nootka fault zone off western Canada. Geophysical Journal of the Royal Astronomical Society, 168: 27-48. B A R A N O V , V . 1957. A new method for interpretation of aeromagnetic maps: Pseudo-gravimetric anomalies. Geophysics, 22: 359-383. B A R R , S.M. 1974. Structure and tectonics of the continental slope west of Vancouver Island. Canadian Journal of Earth Sciences, 11: 1187-1199. B A R T O N , P.J. 1986. The relationship between seismic velocity and density in the continental crust — a useful constraint? Geophysical Journal of the Royal Astronomical Society, 87: 195-208. B O L A N D , A . V . 1989. A geophysical analysis of the Kapuskasing structural zone. Ph.D. thesis, University of British Columbia, 263 pp. B R A N D O N , M.T. 1989a. Deformational styles in a sequence of olistostromal melanges, Pa-cific Rim Complex, western Vancouver Island, Canada. Geological Society of America Bulletin, 101: 1520-1542. B R A N D O N , M.T. 1989b. Origin of igneous rocks associated with melanges of the Pacific Rim Complex, Western Vancouver Island, Canada. Tectonics, 8: 1115-1136. B R A N D O N , M.T., JOHNSON, S.Y. and STEWART, R J . 1984. Cross section A - A ' Western Washington. In Western North American Continental Margin and Adjacent Ocean Floor off Oregon and Washington. Edited by Kulm, L .D. et al. Ocean Margin Drilling Program Regional Atlas Series, Marine Science International, Woods Hole, M A . Atlas 1, Sheet 19. B R O W N , R.D., Jr. and H A N N A , W.F. 1971. Aeromagnetic evidence and geologic structure, northern Olympic Peninsula and Strait of Juan de Fuca, Washington. Bulletin of the American Association of Petroleum Geologists, 55: 1939-1953. 135 REFERENCES 136 CALVERT, A . and CLOWES, R . M . 1990. Deep, high-amplitude reflections from a major shear zone above the subducting Juan de Fuca plate. Geology, 18: 1091-1094. CARLISLE, D. and SUSUKI, T. 1974. Emergent basalt and submergent carbonate-clastic sequences including the Upper Triassic Dilleri and Welleri Zones on Vancouver Island. Canadian Journal of Earth Sciences, 11: 254-279. CASSIDY, J.F. 1991. Structure of the northern Cascadia subduction zone derived from receiver function analysis. Ph.D. thesis, University of British Columbia (in preparation). CHRISTENSEN, N.I. and SALISBURY, M . H . 1975. Structure and constitution of the lower oceanic crust. Reviews of Geophysics and Space Physics, 13: 57-86. CHRISTENSEN, N.I. and SMEWING, J.D. 1981. Geology and seismic structure of the northern section of the Oman ophiolite. Journal of Geophysical Research, 86: 2545-2555. C L A R K E , G.K.C. 1969. Optimum second-derivative and downward-continuation filters. Geophysics, 34: 424-437. CLOWES, R . M . , B R A N D O N , M.T., G R E E N , A .G. , YORATH, C.J., S U T H E R L A N D BROWN, A. , K A N A S E W I C H , E.R. and SPENCER, C. 1987a. LITHOPROBE - southern Vancou-ver Island: Cenozoic subduction complex imaged by deep seismic reflections. Canadian Journal of Earth Sciences, 24: 31-51. CLOWES, R . M . 1990. LITHOPROBE - Multidisciplinary studies of continental evolution: An example from Western Canada. Proceedings of the Pacific Rim Congress '90, Gold Coasv Queensland-, Australia, 6—12 May 1990. Australasian Institute of Mining and Metallurgy, Parkville, Victoria, Australia, HI: 83-97. CLOWES, R . M . , THORLEIFSON, A.J . and L Y N C H , S. 1981. Winona Basin, West Coast Canada: Crustal structure from marine seismic studies. Journal of Geophysical Re-search, 86: 225-242. CLOWES, R . M . and WHITE, W.R.H. 1990. Juan de Fuca Plate crustal structure: Results from ocean bottom seismograph studies (in preparation). CLOWES, R . M . , YORATH, C.J. and H Y N D M A N , R.D. 1987b. Reflection mapping across the convergent margin of western Canada. Geophysical Journal of the Royal Astro-nomical Society, 89: 79-84. COLES, R.L. and CURRIE, R.G. 1977. Magnetic anomalies and rock magnetizations in the southern Coast Mountains, British Columbia: Possible relation to subduction. Canadian REFERENCES 137 Journal of Earth Sciences, 14: 1753-1770. C O N E Y , P.J., JONES, D.L. and MONGER, J.W.H. 1980. Cordilleran suspect terranes. Na-ture, 288: 329-333. C O R D E L L , L . and G R A U C H , V.J.S. 1985. Mapping basement magnetization zones from aeromagnetic data in the San Juan basin, New Mexico. In The Utility of Regional Gravity and Magnetic Anomaly Maps. Edited by W.J. Hinze. Society of Exploration Geophysicists, pp. 181-197. C O U C H , R.W. 1969. Gravity and structures of the crust and subcrust in the southeast Pacific Ocean west of Washington and British Columbia. Ph.D. thesis, Oregon State University, 179 pp. C O U C H , R.W. and RIDDIHOUGH, R.P. 1989. The crustal structure of the western conti-nental margin of North America. In Geophysical Framework of the Continental United States. Edited by L .C . Pakiser and W.D. Mooney. Geological Society of America, Memoir 172, pp. 103-128. CROSSON, R.S. and OWENS, T.J. 1987. Slab geometry of the Cascadia subduction zone beneath Washington from earthquake hypocenters and teleseismic converted waves. Geophysical Research Letters, 14: 824-827. DAVIS, E.E. and CLOWES, R . M . 1986. High velocities and seismic anisotropy in Pleistocene turbidites off Western Canada. Geophysical Journal of the Royal Astronomical Society, 84: 381-400. DAVIS, E;E. and LEWIS, T.J. 1984. Heat flow in a back arc environment: Intermontane and Omineca crystalline belts, southern Canadian Cordillera. Canadian Journal of Earth Sciences, 21: 715-726. DAVIS, E.E. and H Y N D M A N , R.D. 1989. The structure and tectonic history of the northern Cascadia subduction zone: recent accretion and deformation. Geological Society of America Bulletin, 101: 1465-1480. DEHLINGER, P., COUCH, R.W., M c M A N U S , D.A. and G E M P E R L E , M . 1970. North-east Pacific structure. In The Sea, Vol. IV, Part II. Edited by A . E . Maxwell. Wiley-Interscience Publisher, N Y , pp. 133-190. DICKINSON, W.R. 1970. Relations of andesites, granites and derivative sandstones to arc-trench tectonics. Reviews of Geophysics and Space Physics, 8: 813-860. DREW, J.J. 1987. A re-evaluation of the seismic structure across the active subduction zone REFERENCES 138 of western Canada. M.Sc. thesis, University of British Columbia, 149 pp. DREW, J J . and CLOWES, R . M . 1990. A re-interpretation of the seismic structure across the active subduction zone of western Canada. In Studies of Laterally Heterogeneous Struc-tures using Seismic Refraction and Reflection Data. Edited by A . G . Green. Geological Survey of Canada, Paper 89-13, pp. 115-132. FAIRCHILD, L . H . and COWAN, D.S. 1982. Structure, petrology and tectonic history of the Leech River complex northwest of Victoria, Vancouver Island. Canadian Journal of Earth Sciences, 19: 1817-1835. F U L L E R , B.D. 1967. Two-dimensional frequency analysis and design of grid operators. In Mining Geophysics, Volume II. Society of Exploration Geophysicists, pp. 658-708. GABRIELSE, H . and Y O R A T H , C J . 1989. D N A G #4. The Cordilleran Orogen in Canada. Geoscience Canada, 16: 67-83. G L A S S L E Y , W. 1974. Geochemistry and tectonics of the Crescent volcanic rocks, Olympic Peninsula, Washington. Geological Society of America Bulletin, 85: 785-794. G R E E N , A . G . (editor) 1990. Studies of laterally heterogeneous structures using seismic refraction and reflection data. Geological Survey of Canada, Paper 89-13. G R E E N , A . G . , MILKEREIT, B. , M A Y R A N D , L . SPENCER, C , KURTZ, R. and CLOWES, R . M . 1987. LITHOPROBE seismic reflection profiling across Vancouver Island: results from reprocessing. Geophysical Journal of the Royal Astronomical Society, 89: 85-90. G R E E N , D.H. and RINGWOOD, A .E . 1967. The genesis of basaltic magmas. Contributions to Mineralogy and Petrology, 15: 103-190. G U N N , P J . 1975. Linear transformations of gravity and magnetic fields. Geophysical Prospect-ing, 23: 300-312. H A S S E L G R E N , E.O. 1991. M.Sc. thesis, University of British Columbia (in preparation). H Y N D M A N , R.D. 1988. Dipping seismic reflectors, electrically conductive zones, and trapped water in the crust over a subducting plate. Journal of Geophysical Research, 93: 13391-13405. H Y N D M A N , R.D., RIDDIHOUGH, R.P. and HERZER, R. 1979. The Nootka Fault Zone — a new plate boundary off western Canada. Geophysical Journal of the Royal Astro-nomical Society, 58: 667-683. H Y N D M A N , R.D., YORATH, C.J., CLOWES, R . M . and DAVIS, E.E. 1990. The northern REFERENCES 139 Cascadia subduction zone at Vancouver Island: Seismic structure and tectonic history. Canadian Journal of Earth Sciences, 27: 313-329. ISACHSEN, C E . 1987. Geology, geochamistry, and Cooling history of the Westcoast Crys-talline Complex and related rocks, Meares Island and vicinity, Vancouver Island, British Columbia. Canadian Journal of Earth Sciences, 24: 2047-2064. JOHNSON, S.Y. 1984. Evidence for a margin-truncating transcurrent fault (pre-Late Eocene) in western Washington. Geology, 12: 538-541. JOHNSON, S.Y., NIEM, W.A., NIEM, A.R., B R A N D O N , M.T. and STEWART, R.J. 1984. Geology, Western Oregon and Washington. In Western North American Continental Margin and Adjacent Ocean Floor off Oregon and Washington. Edited by Kulm, L .D. et al. Ocean Margin Drilling Program Regional Atlas Series, Marine Science International, Woods Hole, M A . Atlas 1, Sheet 14. JONES, D.L., SILBERLING, N J . and HTLLHOUSE, J. 1977. Wrangellia - A displaced terrane in northwestern North America. Canadian Journal of Earth Sciences, 14: 2565-2577. K E E N , C E . and H Y N D M A N , R.D. 1979. Geophysical review of the continental margins of eastern and western Canada. Canadian Journal of Earth Sciences, 16: 712-747. KURTZ, R.D., DELAURIER, J .M. and GUPTA, J.C. 1986. A magnetotelluric sounding across the subducting Juan de Fuca Plate. Nature, 321: 596-599. KURTZ, R.D., DELAURIER, J .M. and GUPTA, J.C. 1990. The electrical conductivity dis-tribution beneath Vancouver Island: a region of active plate subduction. Journal of Geophysical Research, 95B: 10929-10946. LEWIS, T J . and BENTKOWSKI, W.H. 1990. Crustal structures near the Lithoprobe southern Cordillera transect (in preparation). LEWIS, T.J., BENTKOWSKI, W.H., DAVIS, E.E., H Y N D M A N , R.D., SOUTHER, J.G. and WRIGHT, J.A. 1988. Subduction of the Juan de Fuca Plate: Thermal consequences. Journal of Geophysical Research, 93: 15207-15225. MacLEOD, N.S., TIFFIN, D.L., SNAVELY, P.D., Jr., and CURPJE, R.G. 1977. Geologic interpretation of magnetic and gravity anomalies in the Strait of Juan de Fuca, U.S. -Canada. Canadian Journal of Earth Sciences, 14: 223-238. M A L I N G , D.H. 1973. Coordinate Systems and Map Projections. George Philip and Son Ltd. (London), 255 pp. REFERENCES 140 M A S S E Y , N.W.D. 1986. The Metchosin Igneous Complex, southern Vancouver Island; ophi-olite stratigraphy developed in an emergent island setting. Geology, 14: 602-605. M A S S E Y , N.W.D. 1990. Stratigraphy of the Sooke region. In LITHOPROBE: Southern Vancouver Island - Geology and Geophysics. Compiled by C. J. Yorath. Geological Survey of Canada, Bulletin (in preparation). M A Y R A N D , L.J. , G R E E N , A . G . and MILKEREIT, B . 1987. A quantitative approach to bedrock velocity resolution and precision: The LITHOPROBE Vancouver Island ex-periment. Journal of Geophysical Research, 92: 4837-4845. McBIRNEY, A.R. 1978. Volcanic evolution of the Cascade Range. Annual Review of Earth and Planetary Sciences, 6: 437-456. M c M E C H A N , G.A. and SPENCE, G.D. 1983. P-wave velocity structure of the Earth's crust beneath Vancouver Island. Canadian Journal of Earth Sciences, 20: 742-752. MONGER, J.W.H., PRICE, R.A. and T E M P E L M A N - K L U I T , D J . 1982. Tectonic accretion and the origin of the two major metamorphic and plutonic welts in the Canadian Cordillera. Geology 10: 70-75. M U L L E R , J.E. 1977a. Evolution of the Pacific Margin, Vancouver Island, and adjacent regions. Canadian Journal of Earth Sciences, 14: 2062-2085. M U L L E R , J.E. 1977b. Geology of Vancouver Island. Scale 1:250000. Geological Survey of Canada, Open File Map 463. M U L L E R , J.E. 1977c. Metchosin Volcanics and Sooke Intrusions of southern Vancouver Island. Geological Survey of Canada, Paper 77-1A (Report on Activities), pp. 287-294. M U L L E R , J.E. 1980a. Chemistry and origin of the Eocene Metchosin volcanics, Vancouver Island, British Columbia. Canadian Journal of Earth Sciences, 17: 199-209. M U L L E R , J.E. 1980b. The Paleozoic Sicker group of Vancouver Island, British Columbia. Geological Survey of Canada, Paper 79-30. M U L L E R , J.E. 1983. Victoria, British Columbia. Geology 1:100000. Geological Survey of Canada, Map 1553A. M U L L E R , J.E., C A M E R O N , B.E.B. and NORTHCOTE, K .E . 1981. Geology and mineral deposits of Nootka Sound map-area, Vancouver Island, British Columbia. Geological Survey of Canada, Paper 80-16.* REFERENCES 141 N A F E , J. and D R A K E , C. 1963. Physical properties of marine sediments. In The Sea. Edited by M . N . Hi l l . Pp. 794-828. NETTLETON, L . L . 1971. Elementary Gravity and Magnetics for Geologists and Seismolo-gists. Society of Exploration Geophysicists, Monograph 1. 121 pp. RIDDIHOUGH, R.P. 1979. Gravity and structure of an active margin - British Columbia and Washington. Canadian Journal of Earth Sciences, 16: 350-363. RIDDIHOUGH, R.P. 1982. One hundred years of plate tectonics in Western Canada. Geo-science Canada, 9: 28-34. RIDDIHOUGH, R.P. 1984. Recent movements of the Juan de Fuca plate system. Journal of Geophysical Research, 89: 6980-6994. RIDDIHOUGH, R.P. and H Y N D M A N , R.D. 1990. The modern plate tectonic regime of the Western Canada continental margin. In The Cordilleran Orogen: Canada. Edited by H . Gabrielse and C J . Yorath. Decade of North American Geology (DNAG), Volume G-2 (in press). RIDDIHOUGH, R.P. and S E E M A N N , D.A. 1982. Juan de Fuca Plate Map: JFP-8 Grav-ity Anomaly, 1:2000000. Earth Physics Branch, Department of Energy, Mines and Resources (Pacific Geoscience Centre) Canada. RODDICK, J.A. 1967. Specific gravity of plutonic rocks in the northern Coast Mountains (Abstract). Meeting of the Geological Society of America, Cordillera Section, Santa Barbara, California. RODDICK, J.A., M U L L E R , J.E. and OKULITCH, A . V . 1979. Map 1386A: Fraser River (British Columbia - Washington), Sheet 92. Geological Survey of Canada, 1 : 1000000 Geological Atlas. ROGERS, G.C. 1983. Seismotectonics of British Columbia. Ph.D. thesis, University of British Columbia, 247 pp. ROGERS, G . C , SPINDLER, C and H Y N D M A N , R.D. 1990. Seismicity along the Van-couver Island Lithoprobe Corridor. Proceedings of the Project Lithoprobe: Southern Canadian Cordillera Transect Workshop, Calgary, 3-4 March 1990, pp. 166-169. RUSMORE, M . E . and COWAN, D.S. 1985. Jurassic-Cretaceous rock units along the southern edge of the Wrangellia terrane on Vancouver Island. Canadian Journal of Earth Sciences, 22: 1223-1232. REFERENCES 142 SHOULDICE, D.H. 1971. Geology of the western Canadian continental shelf. Bulletin of Canadian Petroleum Geology 19: 405-436. SMITH, P.L. and TIPPER, H.W. 1986. Plate tectonics and paleobiogeography: Early Jurassic (Pliensbachian) Endemism and Diversity. PALAIOS, 1: 399-412. SNAVELY, P.D. and WAGNER, H.C. 1981. Geological cross-section across the continental margin off Cape Flattery, Washington and Vancouver Island, British Columbia. U.S. Geological Survey, Open File Report 81, pp. 978-984. SNYDER, J.P. 1983. Map Projections Used by the U.S. Geological Survey, 2nd Edition. U.S. Geological Survey, Bulletin 1532, 313 pp. SPENCE, G.D., CLOWES, R . M . and ELLIS, R . M . 1985. Seismic structure across the active subduction zone of western Canada. Journal of Geophysical Research, 90: 6754-6772. SPRENKE, K .F . 1982. Potential field inversion. Ph.D. thesis, University of Alberta, 301 pp. SRIVASTAVA, S.P. 1973. Interpretation of gravity and magnetic measurements across the continental margin of British Columbia, Canada. Canadian Journal of Earth Sciences, 10: 1664-1677. STACEY, R.A. 1973. Gravity anomalies, crustal structure, and plate tectonics in the Canadian Cordillera. Canadian Journal of Earth Sciences, 10: 615-628. STACEY, R.A. and STEPHENS, L.E. 1969. An interpretation of gravity measurements on the west coast of Canada. Canadian Journal of Earth Sciences, 6: 463-474. TABER, J.J. 1983. Continental structure and seismicity of the Washington continental margin. Ph.D. thesis, University of Washington, 159 pp. TABER, J.J. and LEWIS, B.T.R. 1986. Crustal structure of the Washington continental margin from refraction data. Bulletin of the Seismological Society of America, 76: TELFORD, W.M. , GELDART, L.P., SHERIFF, R.E. and K E Y S , D.A. 1976. Applied Geo-physics. Cambridge University Press (Cambridge), 860 pp. TIFFIN, D.L., C A M E R O N , B.E.B. and M U R R A Y , J.W. 1972. Tectonics and depositional history of the continental margin off Vancouver Island, British Columbia. Canadian Journal of Earth Sciences, 9: 280-296. TSUBOI, C. 1983. Gravity. George Allen and Unwin (Publishers) Ltd. (London), 254 pp. REFERENCES 143 van der H E Y D E N , P. 1989. U-Pb and K - A r geochronometry of the Coast Plutonic Com-plex, 53°N to 54°N, British Columbia, and implications for the Insular - Intermontane Superterrane boundary. Ph.D. thesis, University of British Columbia, 172 pp. WALCOTT, R.I. 1967. The Bouguer anomaly map of southwestern British Columbia. Insti-tute of Earth Sciences, University of British Columbia, Scientific Report 15, 74 pp. W A L D R O N , D.A. 1982. Structural characteristics of a subducting oceanic plate. MSc. thesis, University of British Columbia, 121 pp. W A L D R O N , D.A., CLOWES, R . M . and WHITE, D J . 1990. Seismic structure of a subducting oceanic plate off Western Canada. In Studies of Laterally Heterogeneous Structures using Seismic Refraction and Reflection Data. Edited by A . G . Green. Geological Survey of Canada, Paper 89-13, pp. 105-113. WEBRING, M . 1985. SAKI : A Fortran program for generalized linear inversion of gravity and magnetic profiles. US Geological Survey, Open File Report 85-122. WHITE, D J . and CLOWES, R . M . 1984. Seismic investigation of the Coast Plutonic Complex — Insular Belt boundary beneath the Strait of Georgia. Canadian Journal of Earth Sciences, 21: 1033-1049. WHITE, W.R.H. and CLOWES, R . M . 1987. Juan de Fuca Plate crustal structure: Results from ocean bottom seismograph studies (Abstract). EOS, 68: 1371. W O O L L A R D , G.P. 1969. Regional variations in gravity. American Geophysical Union, Monograph 13, pp. 320-341. Y O R A T H , C J . 1980. The Apollo structure of the Tofino Basin, Canadian Pacific continental shelf. Canadian Journal of Earth Sciences, 17: 758-775. Y O R A T H , C.J., CLOWES, R .M. , G R E E N , A .G. , SUTHERLAND BROWN, A. , B R A N -DON, M.T., M A S S E Y , N.W.D., SPENCER, C , K A N A S E W I C H , E.R. and H Y N D -M A N , R.D. 1985a. LITHOPROBE - Phase 1: Southern Vancouver Island: Preliminary analyses of reflection seismic profiles and surface geological studies. Geological Survey of Canada, Paper 85-1 A , pp. 543-554. Y O R A T H , C.J., CLOWES, R .M. , M c D O N A L D , R.D., SPENCER, C , DAVIS, E.E., H Y N -D M A N , R.D., ROHR, K. , SWEENEY, J.F., CURRIE, R.G., H A L P E N N Y , J.F. and S E E M A N N , D.A. 1987. Marine multichannel seismic reflection, gravity and magnetic profiles — Vancouver Island continental margin and Juan de Fuca Ridge. Geological Survey of Canada, Open File 16,61. REFERENCES 144 Y O R A T H , C.J., G R E E N , A .G. , CLOWES, R .M. , SUTHERLAND BROWN, A. , B R A N D O N , M.T., K A N A S E W I C H , E.R., H Y N D M A N , R.D. and SPENCER, C. 1985b. LITHO-PROBE, southern Vancouver Island: Seismic reflection sees through Wrangellia to the Juan de Fuca plate. Geology, 13: 759-762. Y O R A T H , C.J., HYNDMAN, R.D., SUrHERLAND BROWN, A. , MASSEY, N.W.D. and CLOWES, R . M . 1990. Synthesis. In LITHOPROBE: Southern Vancouver Island -Geology and Geophysics. Compiled by C. J. Yorath. Geological Survey of Canada, Bulletin (in preparation). A P P E N D I X A C O N T O U R M A P S O F G R A V I T Y A N D M A G N E T I C A N O M A L Y F I E L D S Shown are line contour versions of the unfiltered gravity and magnetic anomaly maps shown in figures 2.2 and 2.3. 145 Figure A . l : Contoured free-air and Bouguer gravity anomaly map. Contour interval is 10 mGal. REFERENCES 147 51 50 49 48 130 129 128 127 126 125 124 0 50 k m MAGNETICS 51 50 49 48 123 Figure A.2: Contoured magnetic anomaly map. Contour interval is 50 nT. REFERENCES 1 3 0 1 2 9 1 2 8 1 2 7 1 2 6 1 2 5 1 2 4 1 2 3 Figure A.3: Contoured magnetic anomaly map. Contour interval is 100 nT. A P P E N D I X B DENSITY A N D M A G N E T I C SUSCEPTIBILITY M E A S U R E M E N T S 149 REFERENCES 150 UNIT LITHOLOGY DENSITIES (kg/m3) SOURCE R A N G E No. M E A N INSULAR BELT Sicker Group and,dac,gywk,argl 2730-2850 8 2770 W Karmutsen Fm basalt 2830-3140 74 2960 W Quatsino Fm Is, argl 2690-2850 8 2760 W Bonanza Group and, dac, rhy 2550-2820 13 2690 DO Nanaimo Fm ss, cgl 2460-2630 23 2550 W Leech River Fm gywk 2670-2750 11 2710 W Crescent/Metchosin basalt 2410-3030 >19 2800 SW,P Carmanah Group ss, sh, cgl 2490-2630 2 2560 DO Core Rocks ss, sh — 13 2600 M Plutonic (post-) gd 2560-2820 24 2680 W,DO Plutonic (syn-) di, mtsed 2650-3010 43 2780 W,DO COAST PLUTONIC C. granite g 2550-2700 44 2620 Ro qtz monzonite qm 2550-2760 349 2650 Ro granodiorite gd 2560-2840 1023 2700 Ro qtz diorite qd 2600-2900 1179 2750 Ro diorite dT 2640-2940 378 2820 Ro gabbro gb -— 72 3030 Ro metasediments qtzt, amph 2630-2990 48 2710 W Table B . l : Measured densities of rock samples from the Insular Belt (Vancouver Island) and Coast Plutonic Complex. Source abbreviations are: W = Walcott (1967), DO = Dominion Observatory (in Stacey and Stephens 1969), SW = Snavely and Wagner (1966), P = J.E. Pearl (in MacLeod et al. 1977), M = MacLeod et al. (1977), Ro = Roddick (1967). REFERENCES i 151 UNIT LITHOLOGY SUSCEPTIBILITIES (xl(T* SI) SOURCE R A N G E M E A N INSULAR BELT Sicker Group and,dac,gywk,argl 140-540 340 M Karmutsen Fm basalt 500-50300 11300 CC Quatsino Fm Is, argl <120-2500 630 CC Bonanza Group and, dac, rhy 750-8170 2510 CC Nanaimo Fm ss, cgl 380-1260 750 T Leech River Fm gywk <120-25100 <12500 M Crescent Fm basalt 940-114000 55300 G Metchosin Vole basalt 18800-31400 22600 M , B H Carmanah Group ss, sh, cgl — — Core Rocks ss, sh 750-1260 1000 T Plutonic (post-) gd <27600 — CC Plutonic (syn-) di, mtsed — — COAST PLUTONIC C. granite g — — qtz monzonite qm 2500-26400 6280 CC granodiorite gd 1260-31400 15100 CC qtz diorite qd 1260-46500 18900 . cc diorite di 2510-126000 33900 C C gabbro gb 6280-59100 35200 CC metasediments qtzt, amph — — Table B.2: Measured magnetic susceptibilities of rock samples from the Insular Belt (Vancou-ver Island) and Coast Plutonic Complex. Measurements have been converted to rationalized SI units; divide by 4K to obtain unrationalized electromagnetic cgs units. Source abbrevia-tions are: M = MacLeod et al. (1977), CC = Coles and Currie (1977), T = Telford et al. (1976), G = S. Gromme' (in MacLeod et al. 1977), B H = Brown and Hanna (1971). 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items