Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A seismic refraction study of the Queen Charlotte fault zone Dehler, Sonya Astrid 1986

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

Item Metadata


831-UBC_1986_A6_7 D43_6.pdf [ 5.18MB ]
JSON: 831-1.0052982.json
JSON-LD: 831-1.0052982-ld.json
RDF/XML (Pretty): 831-1.0052982-rdf.xml
RDF/JSON: 831-1.0052982-rdf.json
Turtle: 831-1.0052982-turtle.txt
N-Triples: 831-1.0052982-rdf-ntriples.txt
Original Record: 831-1.0052982-source.json
Full Text

Full Text

A SEISMIC R E F R A C T I O N S T U D Y OF T H E QUEEN CHARLOTTE FAULT ZONE  by SONYA ASTRID  DEHLER  B.Sc.(Geophysics), The University of Calgary, 1983  A T H E S I S S U B M I T T E D IN P A R T I A L F U L F I L L M E N T  OF  THE REQUIREMENTS FOR THE D E G R E E OF MASTER OF SCIENCE in T H E FACULTY OF G R A D U A T E  STUDIES  Department of Geophysics and Astronomy  We accept this thesis as conforming to the required standard  T H E UNIVERSITY OF BRITISH  COLUMBIA  A p r i l 1986 (£> Sonya A s t r i d Dehler, 1986 7  In p r e s e n t i n g  this  thesis i n partial  f u l f i l m e n t of the  r e q u i r e m e n t s f o r an a d v a n c e d d e g r e e a t t h e U n i v e r s i t y of B r i t i s h Columbia, I agree that it  freely  t h e L i b r a r y s h a l l make  a v a i l a b l e f o r r e f e r e n c e and study.  agree that permission f o r extensive for  financial  copying or p u b l i c a t i o n of t h i s  gain  Department  of  I ti s thesis  s h a l l n o t be a l l o w e d w i t h o u t my  permission.  oKOPHYSTr.s & ASTRONOMY  The U n i v e r s i t y o f B r i t i s h 1956 Main M a l l V a n c o u v e r , Canada V6T 1Y3  (3/81)  thesis  s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e h e a d o f my  understood that  E-6  further  copying of t h i s  d e p a r t m e n t o r by h i s o r h e r r e p r e s e n t a t i v e s . for  I  Columbia  written  ii  ABSTRACT The margin between the continental North American and oceanic Pacific plates west of the Queen Charlotte Islands is uniquely marked by an active transform fault zone. The region is the locus of oblique convergence between the two lithospheric plates. West of the fault zone the absent continental shelf is replaced by a 25 k m wide scarp-bounded terrace at 2 km depth which separates the oceanic and continental crust. A n onshore-offshore seismic refraction survey was carried out in 1983 across the Queen Charlotte Islands region. T h i r t y - t w o explosive charges and several airgun lines were recorded on eleven land-based and six ocean-bottom instruments. A subset of the resulting d a t a set was chosen to study the structure of the Queen Charlotte Fault zone and adjacent terrace. Two-dimensional ray tracing and synthetic seismogram modelling produced a velocity structural model of the fault region. Underlying the deformed terrace sediments is an upper 3 k m thick faulted unit with velocity 3.8 km/s and a high gradient. The lower crustal region is on average 10 km thick and has velocity 5.3 km/s and a slightly lower gradient. Beneath this unit the Moho increases in dip from 5° to 19° eastward. The terrace velocities are anomalously low compared to the adjacent oceanic and continental crustal structures. Velocities of the oceanic crust are consistent with those observed in ophiolite sequences. The velocity structure of the continental crust is not well-defined; however, vertical offset of 1.1 k m is seismically observed on the Rennell-Louscoone Fault on Moresby Island. T w o tectonic mechanisms are proposed to explain the anomalous terrace structure. Subduction of the oceanic lithosphere beneath the terrace would accrete sediments to the seaward edge of the terrace and subduct material beneath it. Upthrusting of terrace  iii material along near-vertical fault planes would result from compression at the transform fault above an inactive subduction zone.  A combination or alternation of the two  mechanisms would explain the observed fault zone structure.  iv T A B L E OF  CONTENTS  ABSTRACT  ii  LIST OF T A B L E S  vi  LIST OF FIGURES  vii  ACKNOWLEDGEMENTS  ix  CHAPTER I  1  INTRODUCTION  1.1 Regional Geotectonic History 1.1.1 Terrane accretion 1.1.2 Plate interactions 1.2 Present Tectonic Regime 1.2.1 Geologic setting 1.2.2 Geophysical studies 1.3 E x p e r i m e n t a l Objectives C H A P T E R II  DATA ACQUISITION A N D PROCESSING  2.1 The Refraction Experiment 2.2 Instrumentation 2.3 D a t a Processing 2.3.1 Initial data reduction 2.3.2 Shot and receiver positioning 2.3.3 Depth determinations 2.3.4 T i m i n g calculations and corrections 2.3.5 Filtering 2.3.6 A m p l i t u d e corrections C H A P T E R III  DATA ANALYSIS  3.1 Initial Observations and Interpretation 3.1.1 The airgun data set 3.1.2 The three-layer case 3.1.3 The explosion data set 3.1.4 First break analysis 3.2. The Modelling M e t h o d of Interpretation 3.2.1 The modelling algorithm 3.2.2 M o d e l l i n g techniques  1 1 4 7 7 10 16 17 17 19 20 20 21 23 23 25 29 30 30 30 31 33 34 35 36 36  V  C H A P T E R IV INTERPRETATION OF AIRGUN DATA 4.1 Initial Constraints 4.2 Profile O B S W 2 4.3 Profile O B S W l E A S T 4.4 Profile O B S W l W E S T 4.5 The F i n a l M o d e l . . .  38 38 40 44 48 51  CHAPTER V INTERPRETATION OF EXPLOSION DATA 5.1 Initial Constraints 5.2 Receiver 010 5.3 Receiver 020 5.4 Receiver 030 5.5 Receiver 060 5.6 Receiver 100 5.7 Shot 13 5.8 Shot 25 5.9 The F i n a l Velocity M o d e l  52 52 53 58 62 66 67 70 76 79  •  C H A P T E R VI DISCUSSION A N D CONCLUSIONS 6.1 Stratigraphic Interpretation 6.2 Tectonic Implications 6.3 Conclusions  82 82 85 93  BIBLIOGRAPHY  94  APPENDIX A  Airgun Data  APPENDIX B  Explosion D a t a  98 102  vi LIST OF  TABLES  I. Receiver numbers and locations of seismographs  20  II. T i m i n g uncertainties  25  III. Velocity and depth estimates for terrace region  33  I V . Stratigraphic interpretation of velocity units  83  Vll  LIST OF  FIGURES  1. Plate tectonics off Canada's west coast  2  2. Suspect terranes of the Q C I region  3  3. Past plate interactions off western C a n a d a  6  4. Physiographic location map of the Q C I  9  5. Line drawing of C S P profile across terrace region  11  6. Lithospheric flexural model of the Q C I region  13  7. B a t h y m e t r y west of the Q C I  14  8. Velocity structural model from previous study  15  9. Location of receivers and survey lines  18  10. Periodograms for receiver 010  26  11. Periodograms for receiver 100  27  12. Comparison of unfiltered and filtered data  28  13. Three-layer interpretation model  32  14. F i r s t arrival curves for explosion data  35  15. Final velocity model from airgun data  39  16. O B S W 2 ray tracing model and traveltimes  41  17. O B S W 2 seismograms  42  18. O B S W l E A S T ray tracing model and traveltimes  45  19. O B S W l E A S T seismograms  46  20. Detail from O B S W l E A S T data section  47  21. O B S W l W E S T ray tracing model and traveltimes  49  22. O B S W l W E S T seismograms  50  23. Final velocity model  54  viii  24. Receiver 010 model and traveltimes  56  25. Receiver 010 seismograms  57  26. Receiver 020 model and traveltimes  60  27. Receiver 020 seismograms  61  28. Geological cross section of Moresby Island  63  29. Receiver 030 model and traveltimes  64  30. Receiver 030 seismograms  65  31. Receiver 060 model and traveltimes  68  32. Receiver 060 seismograms  69  33. Receiver 100 model and traveltimes  71  34. Receiver 100 seismograms  72  35. Shot 13 model and traveltimes  74  36. Shot 13 seismograms  75  37. Shot 25 model and traveltimes  77  38. Shot 25 seismograms  78  .  39. Seismic profiles across the Aleutian Trench  88  40. Tectonic model of the Q C Fault zone  89  41. Conceptual diagram of an upthrust welt  90  42. Proposed tectonic processes at the Q C Fault  92  ix  ACKNOWLEDGEMENTS To my parents with gratitude for their years of love, support, and encouragement. I would like to thank my supervisor, D r . Ron M . Clowes, for his support and guidance over the years and the many helpful discussions which kept me "on track". I would also like to thank D r . Robert M . Ellis for critically reviewing this thesis. I a m greatly indebted to B o b M e l d r u m for his patient and thorough explanations of the more technical (and mystical) aspects of this study. M y thanks also go to D a v i d Mackie for his invaluable assistance with the processing, and to many others who helped w i t h the thesis, in particular Joane Berube, Jeff Drew, Chris Pike, A l e x von Breymann, and B a r r y Zelt. The members of the department (faculty, staff, and students) have made this an enjoyable place to study. M y thanks to a l l of them for their friendship. Financial support for d a t a acquisition and analysis was provided by N S E R C through O p e r a t i n g G r a n t A7707 and Strategic (Oceans) G r a n t G0738, and by D.S.S. contracts 05SU.23235-3-1089 from the E a r t h Physics B r a n c h , E.M.R., and 06SB.23227-4-0904 from the Pacific Geoscience Centre, E . M . R . Also, E . M . R . Research Agreement 287 provided support in 1983/84.  Additional support was provided by Chevron Canada  Resources L t d . and Shell C a n a d a Resources L t d .  1  CHAPTER I  INTRODUCTION  The Queen Charlotte Fault zone marks the active boundary west of central British C o l u m b i a , Canada, between the North American and Pacific lithospheric plates. Coincident w i t h the fault zone between 52° and 54° is the western margin of the Queen C h a r l o t t e Islands (figure l ) . The transition across the margin is structurally complex, a product of the long and complicated geologic history of the region. A brief review of this history will provide a tectonic framework for newly developed structural models.  1.1 Regional Geotectonic History 1.1.1 Terrane accretion M u c h of the N o r t h American Cordillera is composed of "suspect" terranes of uncertain paleogeograp hie origins (figure 2). Most of these terranes are allochthonous to N o r t h A m e r i c a and were accreted to the cratonic margin as a result of plate interactions (Coney et al., 1980). T h e Queen Charlotte Islands and immediate vicinity are underlain by two such terranes, Alexander and Wrangellia. The Alexander terrane is a continental edge fragment (Coney et al., 1980) consisting of a complex assemblage of volcanic, clastic, and limestone rocks of primarily Late P r e c a m b r i a n to Late Paleozoic age (approximately 650 to 250 M a ) . Wrangellia is a submarine arc complex composed of M i d d l e to Upper Triassic tholeiitic basalts and calcareous sedimentary rocks and Early to M i d d l e Jurassic terrigenous elastics and volcanics (Coney et al., 1980). Wrangellia terrane rocks are represented on the Queen C h a r l o t t e Islands by the Karmutsen, Kunga, Maude, and Yakoun Formations (Yorath and Chase, 1981; Sutherland Brown, 1968).  2  F i g u r e 1. P l a t e tectonics off Canada's west coast (after Riddihough, 1982a).  3  J'  ' | CHUGACH TERRANE — Cretaceous sediments and melange  1 TRACY ARM TERRANE — Central Gneiss complex: Late MesozoJc 1 and Cenozolc plutons of Coast Plutonic Complex GRAVINA - NUTZOTIN BELT — Jurrasslc and Cretaceous arc • volcanlcs and sediments H H H TAKU TERRANE — Upper Paleozoic to Trlassic volcanlcs, sediments I I  WRANGELLIA — Late Trlassic to Jurassic volcanlcs, sediments. Paleozoic basement STIKINE TERRANE — Late Paleozoic to Jurassic volcanlcs, sediments ALEXANDER TERRANE — Paleozoic volcanlcs, sediments, plutonic rocks -CD Faults — 1. Chatham Strait Fault 2. Clarence Strait Fault 3. Queen Charlotte Fault 4. Beresford Bay and Langara Faults S.Sandspit Fault  F i g u r e 2.  6. Rennell Sound Fault 7. Louscoone Inlet Fault 8. Grenville Channel Fault 9. Principe Laredo Fault 10. Kltkalta Fault H.Holberg Fault  Suspect terranes of the Queen Charlotte Islands region (after  Yorath and Chase, 1981).  4 B o t h terranes are believed to have moved northwards from their places of origin before accreting to N o r t h America. Paleomagnetic studies (van der Voo et al., 1980) indicate that Alexander terrane moved 1800 k m northwards with respect to cratonic North A m e r i c a during Late Carboniferous and Triassic time (300 to 180 M a ) . Wrangellia is believed to have originated within approximately 15° of the paleo-equator (in either the northern hemisphere or the southern hemisphere, the latter more favoured by Irving, 1979,  and Hilde et al., 1977) and subsequently travelled between 3000 and 6300 km  northwards. Wrangellia collided w i t h the western edge of Alexander terrane during Middle to Late Jurassic time (170 to 130 M a ) along a suture zone marked by the present Sandspit Fault on the Queen Charlotte Islands and Rennell Sound Fault through Hecate Strait (figure 2). D u r i n g the collision the terranes were uplifted and intruded by plutons dated at 144 M a (Young, 1981). The amalgamated superterrane collided with the westward moving N o r t h A m e r i c a plate in the Late Cretaceous or Early Tertiary period (90 to 40 M a ) resulting in plutonic uplift of the Coast Mountains and a westward j u m p of the active continental margin to the seaward edge of the terrane (Yorath and Cameron, 1982).  1.1.2 P l a t e interactions The concept of seafloor spreading has enabled researchers to reconstruct plate motions and interactions off western North A m e r i c a using magnetic anomalies, hot spot traces, and paleomagnetism (eg. d i h o u g h , 1982b).  Atwater,1970; Coney, 1970; Stone,1977; and R i d -  A l l models propose convergence and strike-slip faulting along the  5 margin for the past 100 M a , although the details of the interactions do not always agree. A generalized plate motion reconstruction presented by Riddihough (1982b) shows the types and rates of past plate interactions off western Canada (figure 3).  Three  plates interacted w i t h the A m e r i c a (A) plate: the Pacific (P) plate, the K u l a (K) plate which once existed north of the Pacific plate, and the large Farallon (F) plate east of the Pacific plate. D u r i n g the M i d d l e to Late Cretaceous (100 to 80 Ma) the Farallon plate interacted in a generally north to northeast direction with the A m e r i c a plate at an estimated rate of 60 to 120 m m / a . B y Late Cretaceous to Eocene time (80 to 40 M a ) the rate had increased to 120 to 170 m m / a in a northeast direction. The K u l a plate was obliquely converging at 100 to 150 m m / a w i t h the A m e r i c a plate at this time and the K - F - A triple j u n c t i o n was migrating either southwards along the coast from A l a s k a (Stone, 1977) or northwards from southern California (Atwater, 1970) to a position near Vancouver Island. In Late Eocene time (45 to 40 M a ) the Pacific plate changed its direction of motion by 50°, as indicated by the bend in the E m p e r o r - H a w a i i chain, resulting in a reduction of all interaction rates across the North American margin by one-half (Riddihough, 1982b). D u r i n g the Oligocene to Early Miocene (40 to 20 M a ) a number of plate adjustments occurred as a southern part of the P - F spreading ridge started subducting beneath California, and transcurrent motion was initiated between 30 and 20 M a along the San A n d r e a s Fault (Atwater, 1970). The northern part of the Farallon plate broke off and began to move independently as the Juan de Fuca (J) plate around 20 M a ago. The K-F-A  triple junction had meanwhile continued migrating rapidly northwards along  the margin and the remainder of the K u l a plate was subducted during this time.  6  F i g u r e 3. Plate interactions off western C a n a d a during the last 100 M a ( R i d d i h o u g h , 1982b). P - Pacific plate, A - A m e r i c a plate, K - K u l a plate, and F - Farallon plate. Arrows indicate relative directions of motion; rates of motion in m m / y r are shown.  7 D u r i n g the Miocene (20 to 10 Ma) the Juan de Fuca plate continued subducting beneath the A m e r i c a plate at 40 to 60 m m / a in a northeastward direction and the Pacific plate moved northwestward at 50 to 60 m m / a , producing transcurrent motion along a paleo- Queen Charlotte Fault at the A m e r i c a plate margin. The P - J - A  triple  junction migrated southeastwards along the margin during this time and may have been responsible for the initiation of rifting in Queen Charlotte Sound 17 M a ago which eventually shifted and rotated the Queen Charlotte Islands along reactivated faults to their present position (Yorath and Chase, 1981; Horn, 1982). Alternatively, the America plate may have overridden a hot spot or mantle plume which caused the rifting and produced the A n a h i m volcanic belt of central B r i t i s h C o l u m b i a (Yorath and Cameron, 1982; Yorath and Chase, 1981). Finally, the Pliocene and Pleistocene (10 to 0 M a ) saw the stabilization of the P - J - A triple junction off northern Vancouver Island and a decrease in the J - A convergence rate to 40 m m / a as the spreading direction at the Juan de Fuca ridge changed (Riddihough, 1977).  New spreading ridges were formed at Dellwood Knolls 1 M a ago and Tuzo  W i l s o n Knolls 0.5 M a ago and joined to the stable triple junction by transform faults (Riddihough et al., 1980).  Spreading at the Tuzo Wilson Knolls caused the Queen  C h a r l o t t e Fault to j u m p landwards to its present location where it is currently active ( H y n d m a n and Ellis, 1981).  1.2 P r e s e n t T e c t o n i c 1.2.1 G e o l o g i c  Regime  setting  The Queen Charlotte Islands were divided by Sutherland Brown (1968) into three physiographic units: the Queen Charlotte Ranges, Skidegate Plateau, and the Queen  8 Charlotte Lowlands (figure 4). The regions differ not only in physiographic features but also in geologic composition, as the 1100 m high Ranges are underlain by granitic and Tertiary volcanic rocks, the 600 m high Plateau by slightly-tilted Jurassic volcanic, Cretaceous sedimentary, and Early Tertiary basaltic rocks, and the 150 m high Lowland by complex sequences of rocks from Jurassic to Pleistocene age (Sutherland B r o w n , 1968). The boundaries of the physiographic units generally follow two major northwesttrending fault systems, the Rennell Sound—Louscoone Inlet Fault and the Sand spit Fault (figure 2). The Rennell-Lou scoone Fault strikes N25°W in the south and N 5 0 ° W north of Louise Island. The fault zone is composed of several associated strands that dip between 75° northeastward and 90° vertical. Total fault displacement since L a t e Jurassic is estimated as 600 to 3000 m downwards and 19 to 93 k m southward for the east block relative to the west block (Sutherland B r o w n , 1968). The Sandspit Fault strikes uniformly N 3 7 ° W and dips at 65° northeastward. The fault zone separates older rocks in the west block from younger formations in the east and cuts through plutons of Cretaceous age. Total fault displacement is estimated as thousands of metres downward for the east block relative to the west and unknown but presumably considerable horizontal offset (Sutherland Brown, 1968). The major faults on the islands are colinear with the Queen Charlotte Fault located at the foot of the continental shelf just offshore. The continental shelf is generally less than 5 k m wide west of the Queen Charlotte Islands and is little more than a notch in the wide, steep continental slope. The slope is composed of three parts: a steep upper scarp, a middle terrace plateau, and a lower scarp.  The upper scarp has a gradient of up to 20° and includes the  9  Figure  4.  Physiographic location map of the Queen Charlotte  (Sutherland Brown, 1968).  Islands  10 truncated spurs of the Insular Mountains at the coast.  The terrace is at an average  depth of 2 km and has variable topography which becomes more rugged towards its termination off southern Moresby Island.  The lower scarp slopes as steeply as 23°  (Chase and Tiffin, 1972). The physiography and geology of the continental slope and shelf appear to be linked to past and present interaction between the Pacific and America plates at the Queen C h a r l o t t e Fault zone.  1.2.2 G e o p h y s i c a l studies Present motion at the Queen Charlotte Fault zone is primarily dextral strike-slip at an average rate of about 55 m m / a (eg. Atwater, 1970; Keen and Hyndman,1979; H y n d m a n and Weichert, 1983). The strike of the fault zone west of Moresby Island, as obtained from its morphology and seismicity, is over 10° different from the more easterly relative plate motion predicted by global plate models (Minster and Jordan, 1978). This suggests that convergence at 10 to 20 m m / a is occurring at the fault zone as subduction of the Pacific plate beneath the Queen Charlotte Islands and/or as crustal deformation of the islands and adjacent region (Hyndman and Ellis, 1981; Yorath and H y n d m a n , 1983). The active fault has been located by recent microseismicity studies (Hyndman and Ellis, 1981; Berube, 1985) as being beneath the upper fault scarp of the continental slope. Composite P - n o d a l fault plane solutions for earthquake clusters indicate primarily strike-slip motion along the fault west of G r a h a m Island and some thrusting and vertical faulting with ocean side down west of Moresby Island (Berube, 1985). This is consistent with studies of large earthquakes which determined strike-slip motion with a  11 small thrust component for the 1949 magnitude 8.1 earthquake west of Graham Island and a strike-slip mechanism w i t h a significant thrust component along a fault plane d i p p i n g 50° eastward for the 1970 magnitude 7.0 earthquake just south of Moresby Island (Rogers, 1983). Seismic reflection profiles across the fault zone reveal the effect of compression and strike-slip movement on the continental margin (eg. Chase and Tiffin, 1972; Chase et al., 1975; P.D. Snavely, unpublished data). The terrace region of the continental slope is underlain by grabens and synclines infilled with deformed sediments (figure 5). The steep upper scarp is the currently active Queen Charlotte Fault and the outer scarp is presumed to be the former fault location. Seismic reflections from oceanic basement end abruptly at the outer scarp, as do magnetic lineations from the oceanic crust (Currie et al., 1980).  Tectonic faulting, possibly associated with trench formation, could be  destroying the remnant magnetization of the crustal rocks at the edge of the terrace (Srivastava, 1973). T  1  1  1  1  1  KILOMETERS  F i g u r e 5. Line drawing interpretation of a C S P profile west of northern Moresby Island (Chase and Tiffin, 1972).  10  12 Heat flow and gravity profiles of the oceanic crust also show dramatic changes as they cross over the fault zones and terrace to the continent. A sharp decrease in heat flow at the seaward edge of the terrace was observed by Hyndman et al. (1982). They also noted a general decrease in heat flow by a factor of three from the axis of the Queen C h a r l o t t e Trough to 40 km inland, consistent w i t h measurements in many subduction zones.  The low terrace heat flow was attributed to tectonic thickening of sediments,  possibly by accretion, and the underthrusting of cold oceanic lithosphere (Hyndman et al., 1982). Free-air gravity data indicate a large negative anomaly at the base of the continental slope and a positive anomaly on the landward side of the continental shelf (eg. Couch, 1969; Dehlinger et al., 1970; Srivastava, 1973; Currie et al., 1980). The gravity low over the terrace may be due to a thick prism of sediments or unusually thick crust (Currie et al., 1980). The gravity profile is again characteristic of subduction zones (Keen and H y n d m a n , 1979). A lithospheric flexural model for the Queen Charlotte Islands region, developed by Y o r a t h and H y n d m a n (1983), agrees well with most of the geophysical and physiographic observations (figure 6). Uplift at the edge of the continental lithosphere and depression of the adjacent area inland, a flexural response to underthrusting, have been observed and described by Riddihough (1982a) in the Queen Charlotte Islands region.  The  corresponding flexural bulging of the oceanic lithosphere is observed on bathymetric maps as the Oshawa Rise (figure 7). The subduction process would produce a thick accretionary wedge of sediments in the vicinity of the continental slope terrace. These sediments would be faulted and deformed by the additional transform motion, as would the crust beneath them.  13 OFFSHORE  TERRACE ACCRETIONARY SEDIMENTARY WEDGE  (Yorath and H y n d m a n , 1983).  Seismic refraction models developed for the region support underthrusting of the Pacific plate beneath the Queen Charlotte Islands as a possible convergence mechanism (Mackie, 1985; Horn et al., 1984; Horn, 1982; B i r d , 1981). T h e model of Horn et al. (1984) indicates anomalously low velocities and high gradients in the terrace region separating the oceanic and continental blocks (figure 8); the model is consistent with an accretionary wedge of compressed sediments overlying a subducting oceanic crust. The subduction and transform motion at the continental margin and the variable geotectonic history appear to have created a unique and complex structure at the fault zone.  14  F i g u r e 7. B a t h y m e t r y west of the Queen Charlotte Islands. The locations of the C S P linedrawing of figure 5 and the refraction model of figure 8 are shown.  15  Distance 40  60  (km) 80  F i g u r e 8. Velocity structural model of Horn et al. (1984) from a seismic profile off southern Moresby Island.  16 1.3 E x p e r i m e n t a l  Objectives  A s the previous discussion has indicated, the structure at the junction of the Pacific and A m e r i c a plates is very complex.  Geophysical observations across the continental  margin indicate that a thick wedge of sediments may be accreting to the margin as a result of subduction of the Pacific plate. A t the same time, dextral transform motion along the active fault plane is pushing the accretion zone northwards and deforming the sediments and underlying crust. A seismic refraction experiment was designed to study the structure of the terrace region that marks the active plate boundary west of the Queen Charlotte Islands, and to try to determine the origin and composition of the terrace. To achieve these objectives it is necessary to look at the structural transition from the oceanic crust west of the terrace to the continental crust beneath the Queen Charlotte Islands.  17 CHAPTER 2.1 T h e R e f r a c t i o n  II  DATA ACQUISITION  AND  PROCESSING  Experiment  A major marine-land seismic refraction survey was carried out in August 1983 by the University of B r i t i s h C o l u m b i a ( U B C ) in collaboration with the Earth Physics Branch, O t t a w a , and the Pacific Geoscience Centre, Sidney.  A 330 km profile was recorded  from the deep ocean across northern Moresby Island and Hecate Strait to the mainland of British C o l u m b i a (figure 9).  The experimental objectives were to determine the  structure of the sediment, crust, and lithosphere ( l ) of the Queen Charlotte transform fault zone, (2) beneath the Queen Charlotte Islands, and (3) below Hecate Strait. A total of seventeen seismographs were deployed: two ocean bottom seismographs (OBSs) west of the Queen Charlotte Islands, eight land-based seismographs (LBSs) on the Islands, four O B S s in Hecate Strait, and three additional L B S s on the mainland. T w o types of energy source were used.  The pelleted T N T explosive N I T R O P E L ^ 1  1  packed into drums with primacord and C.I.L. Procore P r i m e r s ^ was detonated every 3.4 k m along a 110 km line bearing southwest from northern Moresby Island. Twelve 540 kg and twenty 60 kg shots were fired, with two of the small shots following each large shot.  Three additional refraction profiles, one over the western O B S s and two  in Hecate Strait, were shot using a 32 litre ( 2000 in  3  ) airgun with a shot spacing of  approximately 0.18 k m . This study of the structure of the Queen Charlotte Fault zone will focus on the airgun profile west of the Queen Charlotte Islands and the explosion profile as recorded by the ten westernmost receivers.  18  Figure 9. Location of receivers and survey lines in the 1983 refraction experiment. Three of the 540 kg explosion locations ( l , 16, 33) are indicated as are receiver locations 010, 100, and 170.  19 2.2 I n s t r u m e n t a t i o n Four types of seismographs were deployed during the survey (table I). The six O B S s , U B C - m o d i f i e d versions of Atlantic Geoscience Centre instruments (Heffler and Barret, 1979), consist of a frame holding a flotation sphere, an instrument pressure case, a hydrophone, recovery devices, and a detachable weight to take the package to the ocean b o t t o m . Inside the pressure case are gimbaled 4.5 Hz horizontal and vertical component geophones, primary and secondary clocks, a microprocessor, a 4-channel slow-speed direct record analog tape recorder, and the necessary power supplies. The outputs from the geophones and hydrophone were recorded together with an internally-generated 10 H z amplitude-modulated time code directly onto cassette tapes. Six E a r t h Physics B r a n c h ( E P B ) M a r k II digitally recording "Backpack" seismographs were deployed on Moresby Island and the mainland. The Backpacks sampled the output from a 2 Hz vertical component seismometer at 60 samples per second (sps) and recorded on digital cassettes. A n internal clock provided timing for the data. Four digitally recording U B C Microcorders were deployed on Moresby Island. The output from a 1 Hz vertical component seismometer was sampled at 60 sps and recorded on digital cassette tape. T i m i n g was by an internal clock. The last site on Moresby Island was occupied by a U B C F M analog recorder.  It  contained 1 Hz vertical and horizontal component seismometers and recorded their output at high and low gain settings together with W W V B time code on five parallel tracks of 7-track analog tape. The seismographs successfully recorded all shots during the experiment.  20  INSTRUMENT TYPE OBS  RECEIVER NUMBER  010 020 110 120  ISO  140 MCR  050 070 080  100 FMA EPB  090 OSO  040 060 150 160 170  LOCATION offshore terrace offshore terrace Hecate Strait Hecate Strait Hecate Strait Hecate Strait Moresby Island Moresby Island Moresby Island Moresby Island Moresby Island Moresby Island Moresby Island Moresby Island mainland region mainland region mainland region  T a b l e I . Receiver number and location of seismographs used in the experiment. Instrument types are: O B S - ocean b o t t o m seismograph, M C R microcorder, F M A - F M analog recorder, and E P B - E a r t h Physics Branch " B a c k p a c k " recorder.  2.3 D a t a  Processing  2.3.1 I n i t i a l d a t a r e d u c t i o n The initial stage in the data reduction process was to standardize the data recorded by the different seismograph systems. First, the analog data from the O B S s and F M recorder were converted to digital format on 9-track magnetic tapes using a P D P 11/34 -based digitization and editing package. Digitization of the O B S data was a two-step process involving the transfer of field d a t a , recorded at 0.2 m m / s , to high quality 1/4 inch F M tape at 6.0 m m / s . The  21 resulting 30-time speed-up was necessary to overcome digitization system limitations for the slow-speed data recordings. The 4-channel F M tapes of O B S data were then digitized at 120 sps using the 10 Hz time code signal recorded on one of the tracks as a trigger to avoid timing problems associated with tape stretch and recorder speed variations. A detailed description of the O B S digitization procedure may be found in M a c k i e (1985). The F M recorder analog data tapes were digitized directly at 120 sps using an external frequency generator as a sampling trigger. T i m i n g errors associated with tape stretch and tape speed variations were negligible. B o t h the O B S (explosion only) and F M analog data were later reduced to 60 sps recordings for plotting compatibility with the digitally recorded data from the other seismographs. D a t a from the E P B "Backpack" systems were transfered from digital cassettes to 9-track magnetic tapes by personnel at the E a r t h Physics B r a n c h . A similar transfer of U B C Microcorder data was done at U B C . Once the entire data set was in digital format on magnetic tapes, editing of shot recordings and demultiplexing was done using the P D P 11/34 system. A l l explosion recordings and randomly-selected airgun shots were plotted to monitor data quality before proceeding with further processing.  2.3.2  Shot and receiver positioning The position of the ship was determined from Loran C navigation with a relative  accuracy of ± 200 m and an absolute accuracy of ± 300 m (Hyndman et al., 1979). The accuracies were increased by updating the Loran C system with satellite fixes when available. The hyperbolic Loran C measurements were converted to geographic  22 coordinates and a linear interpolation between measurements was used to determine the ship's position for each airgun shot. The location of the explosive shots was determined from the ship's position at drop time. Similarly, the O B S positions were measured at the time of deployment and assumed a vertical fall of the instrument to the ocean floor. T h e L B S positions were directly obtained from 1:50,000 scale topographic maps with absolute accuracies of approximately 150m. Shot-receiver distances were computed from the geographic coordinates of the various shots and receivers. A geometric factor was applied to the airgun data to correct for the offline location of O B S W l (receiver 010) and shift the O B S position to the shooting line. This correction ranged from 2.4 k m for the closest shot to 0.2 km for the shot 15 k m distant. A similar correction incorporating receiver depth as well was applied to the corresponding travel times. A positioning error was noticed on recordings of explosive shots at O B S W l .  Hy-  drophone component plots showed that direct water wave arrivals did not pass through the origin but instead intersected the distance axis at a point 1.78 k m away. When the bathymetric profile was compared to the depth profile of the overlapping airgun survey a discrepancy of approximately 2 k m was observed for major physiographic features. T h e result was a 1.78 k m correction applied to shot-receiver distances for the O B S W l explosion d a t a set. The error may have resulted from an incorrect log entry during data acquisition; the above correction has hopefully restored the true positions and distances.  23  2.3.3 Depth determinations Water depth was continuously recorded through the ship's depth sounding system on an E P C line scan recorder during the survey. The depth of each O B S was determined from these recordings at the time of deployment. The elevations of the L B S s were read from the topographic maps on which their positions were located. T w o methods were used to calculate explosion shot depths. The geometrical method, described in Horn (1982), uses the ship speed, ditch time, water depth, and the time lag between the direct and bottom reflected water waves as recorded on a trailing hydrophone to compute shot depths. This method provided a good estimate of depth for the more distant shots beneath which the seafloor was relatively flat and horizontal. The bubble pulse method relies on the relationship between explosion depth and the period of first oscillation of an expanding explosion-generated gas bubble which is forced to collapse due to hydrostatic pressure. The R a y l e i g h - W i l l i s bubble formula, based on work by L o r d Rayleigh (Strutt, 1917) and Willis (1941), was used to compute depths for shots over the irregular topography of the terrace. B o t h methods have a maximum estimated error of ± 30 m. The airgun maintained a relatively constant depth of 17 m throughout the survey as controlled by the length of the towing cable and the ship speed, thereby obviating the need for complicated depth calculations.  2.3.4 Timing calculations and corrections The W W V B time signal received aboard ship was used as absolute time throughout the survey. Readings from other clocks and instruments were converted to W W V B time during processing to standardize time measurements.  24 Shot origin times for explosions were calculated from the shot depth and distance previously computed, using  where d =shot depth, t<i =arrival time of direct water wave, x —ship-to-shot distance, and V  w  ^water velocity at 1.49 k m / s (Horn, 1982).  The airgun firing time was controlled by a trigger clock for which the signal was recorded with W W V B code on both a 2-channel chart recorder and a 4-channel tape recorder. A linear clock drift correction was subsequently applied to airgun shot origin times based on these recordings. A l l instruments with internal clocks ( O B S s and Microcorders) were rated against either W W V B (OBSs) or W W V (Microcorders) immediately before and after use. Internal clock readings in code were recorded directly on to the tapes. Recording times could thus be related to W W V B after applying any necessary corrections for the linear drift of the internal clocks. The F M analog system recorded W W V B directly onto a parallel channel thereby avoiding timing problems. Clock drift for the E P B Backpacks is unknown as no information was provided by E P B personnel. Corrections applied to times of first sample include the geometrical offline corrections mentioned in section 2.3.2 and shot depth corrections. The explosions and airgun shots were vertically shifted to the ocean surface using the shot depths and a water velocity of 1.49 k m / s . The m a x i m u m uncertainties due to shot origin times, clock drift, shot depths and travel time picks are summarized in table II. The time errors due to uncertain travel time picks are considerably larger than other t i m i n g uncertainties in most cases.  25  CORRECTION Origin Time (EX) Origin Time (AG) Clock Drift Shot Depth Time Picks  OBS (s) O.OSO 0.010 0.0 IS 0.020 0.01-0.05  MCR (s) O.OSO 0.010 0.006 0.020 0.01-0.15  FMA (s) O.OSO 0.010 0.006 0.020 0.01-0.15  EPB (s) O.OSO 0.010 0* 0.020 0.01-0.05  T a b l e I I . T i m i n g uncertainties due to various corrections and traveltime measurements. E X - explosives, A G - airgun; * = value assumed.  2.3.5 F i l t e r i n g Spectral analysis of airgun and explosion data was performed for representative signal and noise recordings. Figure 10 shows periodograms for an O B S recording of an explosion. T h e seismic signal of the O B S extends from 0.5 to 15.0 H z with significant noise in the 18 to 27 H z range. Periodograms for a Backpack recording of the same explosion (figure 11) indicate the presence of noise below 2 H z and strong signal above 2.5 H z . The characteristics of these periodograms are typical of all L B S recordings which show low frequency signal and a clear noise-signal separation leading to relatively good signal-to-no ise ratios. A n eight-pole zero-phase Butterworth bandpass filter was applied to the explosion d a t a to reduce the noise content.  A 0.5-15.0 H z bandpass was used for the O B S s  and a 2-12 H z bandpass filter was applied to most of the L B S data. Filtering visibly enhanced the data especially on receivers 090 and 100 where noise had obscured first arrival energy (figure 12). The filtered d a t a sections were used to approximately locate the first arrivals; more accurate time picks were then made on the unfiltered sections.  26  NOISE  (010)  o_ co. d  Frequency (Hz)  o  ,  SIGNAL (010); ' ,  co. d  Frequency (Hz) F i g u r e 10. Periodograms computed for segments of noise and signal + noise recorded on the vertical geophone of receiver 010, an O B S .  27  NOISE  (100)  oo.  6  0 10  20 15 Frequency (Hz)  SIGNAL  '0  5  10  25  30  25  30  (100)  15 20 Frequency (Hz)  F i g u r e 1 1 . Periodograms computed for segments of noise and signal + noise recorded on receiver 100, a microcorder on Moresby Island.  28  200  180  160  140 Receiver 100 (2  120 12 Hz)  100  F i g u r e 1 2 . A comparison of unfiltered and filtered data sections for receiver 100 on the eastern side of Moresby Island. The horizontal axis indicates shot-receiver distance in km.  29 Spectral analysis of the airgun data was performed and a suitable bandpass filter was tested. N o significant improvement in the signal resolution of the data was achieved, however, so the data remain unfiltered.  2.3.6 Amplitude corrections A m p l i t u d e variations from trace to trace due to different charge sizes and receiver responses were expected.  O'Brien (i960) determined that the amplitude of a T N T -  generated wave varies as W I 2  Z  where W is the weight of the charge in pounds. A c -  cordingly the trace amplitudes were multiplied by W~ / to account for the charge size 2  3  differences. This process overcorrected large shot amplitudes for close (<30 km) shots so an additional correction was applied to make large shot amplitudes consistent with the amplitudes of adjacent small shots for near distances. The explosion data were also corrected for spherical spreading by multiplying the trace amplitudes by d where d is the shot-receiver distance normalized to 200 km. T h i s 2  correction enhanced the farther arrivals with respect to the closer ones. A spreading correction of d where d is normalized to 20 k m gave best results for the airgun data. 1  A m p l i t u d e variations due to different receiver responses were corrected by relating the velocity sensitivities for the various instruments and computing the necessary gain factors to apply to the data. The application of the above processing methods resulted in data of generally high signal-to-no ise ratio in a format ready for interpretation.  30 CHAPTER  III  DATA  ANALYSIS  3.1 I n i t i a l O b s e r v a t i o n s a n d I n t e r p r e t a t i o n C o m m o n receiver gathered sections of the airgun and explosion data are shown in appendices A and B. The filters and amplitude corrections mentioned in the previous chapter have been applied as noted with each section. Further discussion of these as well as other explosion d a t a from the experiment may be found in an open file report (Clowes, 1984).  3.1.1 T h e a i r g u n d a t a set The three airgun sections shown in appendix A represent a single profile west of receiver 010 ( O B S W l W E S T ) and a reversed profile between receivers 010 and 020 ( O B S W 1 E A S T and O B S W 2 ) . The data have been plotted with a reducing velocity of 4.0 k m / s ; hence a line parallel to the distance axis is equivalent to an apparent velocity of 4.0 k m / s . The oscillatory waveforms heading diagonally away from the origin on all three sections are the direct water wave arrivals. The non-linearity of their traveltime characteristic is a result of the highly irregular topography in the region and causes them to interfere with early refraction arrivals. A t least two refraction branches can be seen on each section. A n early branch which interferes w i t h the irregular direct wave arrivals on O B S W l W E S T intersects sharply w i t h a strong second refraction branch that continues linearly across the section to the m a x i m u m shot-receiver distance of 15 km. The transition between refraction branches is much smoother for the other two sections and the gentle curvature of O B S W 2 ' s arrivals suggests several ever steeper branches are involved. The refraction arrivals for reversed profiles O B S W 2 and O B S W l E A S T undergo a severe amplitude loss just past 3 km  31 and 10 k m model distance respectively; this corresponds to 10 km shot-receiver in both cases, suggesting a common loss mechanism. The location of arrivals after this distance is uncertain. Additionally the arrivals on O B S W l E A S T show amplitude fluctuations over their entire length suggesting structural complications.  3.1.2 T h e t h r e e - l a y e r case A preliminary velocity-depth interpretation of the airgun data sections was done using three-layer horizontal interface equations (Sheriff and Geldart, 1982) which have been rederived to consider a surface source and a receiver positioned at the first interface (figure 13). Refraction along first interface:  t =  hi V^cosflx x  —  1  V  x — hi tan 0\  +  -  V  h\  x COS  V\  h  —  Vi  2  2  V  2  t] 1  cos V tx9i x  hy Refraction along second interface:  t  hi  -  Vi  + cos 4>i  2h  x — hi tan <j>i — 2h  2  V cos 9 2  2  x hi h x — + — cos 4>i + 2 — cos 6 = 2  2  2cosfl  ^ 2  2  -  ^-COS^!^  rr +  t  2  2  tan 0  2  32 where t =traveltime of ray, h  n  =thickness of layer n, V  n  ^velocity of layer n, t  n  =inter-  cept of n-th refraction branch at zero offset, x = horizontal component of shot-receiver distance, and 0  n  ^critical angle for layer n.  F i g u r e 1 3 . Ray paths for critically refracted waves in a three-layered medium travelling from a surface source to a receiver at the first interface.  The equations assume horizontal interfaces; this assumption is invalid for interface 1 (water—sediment) because of the known irregular topography. However, the depth estimates will provide at least a general structure which can later be refined through additional interpretation. Layer thicknesses were computed for all three data sections by measuring the velocities and times on the sections and using V = 1.49 km/s as water velocity. The results t  33  PROFILE OBSWl OBSWl  Vi  hy  WEST  1.49  EAST  1.49  2.24 1.14 0.44  0BSW2  1.49  v  h  2  2  v  3  2.29  1.7S  1.64  4.4S  1.20  4.00  2.05  1.01  8.84  Table III. Velocity and layer thickness estimates from three-layer interpretation of airgun data.  are summarized in table III. These values will form the preliminary model for further interpretation.  3.1.3 The explosion data set The gathered sections for receivers 010 to 100 found in appendix B have been plotted w i t h an 8.0 k m / s reducing velocity. It should be noted that the high amplitudes resulting from large shots at near offsets and close shots in general have been over—compensated for in the correction routines and hence their relative amplitudes are misrepresented in these sections. A l l sections show distinct primary arrivals; some also have lower velocity higher a m p l i t u d e secondary arrivals. entire section.  Receiver 010 recorded both sets of arrivals across the  Noise obscures arrivals at receiver 020 beyond 100 k m shot-receiver  distance; both refraction branches are present to that point.  Receivers 030 to 100  recorded low amplitude first arrivals for shot-receiver distances greater than about 85 k m and much higher amplitude first and/or second arrivals for distances less than 85 km.  34  3.1.4 First break analysis Figure 14 is a plot of first arrivals for shots 1 to 16 recorded by receivers 010 to 100. First arrival times were determined from unfiltered data sections used in conjunction w i t h the filtered sections which more clearly showed overall phase coherency.  Shot  positions are fixed for all arrivals and receivers are located at various distances to the right (east) of the plot. P l o t t i n g the arrival times above fixed shot positions provides a "common factor" for easy comparison of traveltime curves.  Traveltime differences  resulting solely from increased receiver offset will delay first arrival times but will not alter the shape of the curve. Structural variations encountered by rays travelling to different receivers w i l l however be reflected as changes in the shape of the curves. Several trends between curves immediately become apparent. There is an average time delay of 0.6 s between arrivals from shots 4 and 5 at all receivers. T h i s indicates a major lateral change in subsurface structure beneath these points which are located above the outer edge of the terrace. T h e first arrival curves are all approximately parallel w i t h the exception of receiver 010 ( O B S W l ) which has a much lower apparent velocity. Energy travelling east of O B S W l towards the other receivers appears to be passing through material of different composition.  The remaining arrival curves successively  increase in time w i t h increasing receiver distance except for arrivals at receiver 020 ( O B S W 2 ) which are delayed by a m i n i m u m of 0.2 s. This delay suggests the existence of lower velocity material beneath O B S W 2 than beneath the Queen Charlotte Island receivers. The general information provided by these curves will be combined w i t h available constraints to construct the preliminary velocity structural model for interpretation.  35  60.0  75.0  90.0  DISTANCE F i g u r e 14. First arrivals for shots 1-18 at receiver 010. Shot locations model and receivers are located to corrections have offset the shots for interpretation method.)  105.0 (KM)  120.0  135.0  1—16 at receivers 020 to 100 and shots are fixed to their positions in the final the right (east) of the plot. Distance receiver 010 by 1.78 km. (See text for  3.2 The Modelling Method of Interpretation The general method of interpretation used in this study involves the comparison of vertical component observed data w i t h theoretical seismograms. A ray tracing algorithm  36 is applied to a representative velocity model and synthetic seismograms are computed from the resulting traveltimes and amplitudes. The solutions determined by this forward modelling technique are non-unique so in this study every attempt was made to generate the most geologically reasonable model within the available constraints.  3.2.1  The modelling algorithm  A n algorithm based on asymptotic ray theory (Spence, 1984; Spence et al., 1984) was used to calculate the synthetic seismograms. The ray tracing routine is an extension of the W h i t t a l l and Clowes (1979) algorithm which calculates traveltimes for refracted and postcritically reflected rays and "pseudo" head waves traced through a two-dimensional velocity model. The extension enables tracing of precritically and multiply reflected rays through the model. A m p l i t u d e s are calculated by asymptotic ray theory approaches and synthetic seismograms are generated after superimposing the displacements of all arrivals at a particular distance.  3.2.2  Modelling techniques  M o d e l s to be used with the ray tracing routine are constructed of polygonal blocks, each w i t h its own velocity and linear velocity gradient. The velocity is defined on the uppermost boundary of each block and is constant along the line segment. The gradient is then oriented normal to this boundary. Rays are traced through the model and the computed traveltimes are compared to observed times. The model is changed by repositioning boundaries and blocks or varying the defined velocity and gradient. The procedure is repeated until a suitable traveltime fit is attained. The calculated amplitudes must also agree with the observed data; a  37 model that fits the traveltimes may not provide a good amplitude match. The forward modelling procedure must thus be continued until satisfactory agreement with both the traveltimes and amplitudes is obtained and the model meets the initial constraints.  38  C H A P T E R IV  I N T E R P R E T A T I O N OF A I R G U N DATA  4.1 Initial Constraints The airgun line A G 2 was shot over the two O B S s to provide information on the sediment and upper crustal structure of the terrace west of the Queen Charlotte Islands. Several continuous seismic reflection profiles (CSPs) in the region (Chase and Tiffin, 1972; Chase et al., 1975; Davis and Seemann, 1981) indicate that the terrace subsurface is composed of faulted and tightly folded sediments partially infilled with stratified sediments showing increasing deformation with depth.  Analyses of terrace sediment  velocities south of the present study area (Kirstiuk, 1981; Tjaden, 1981) indicate extreme lateral inhomogeneity in the shallow subsurface above 4 k m depth. A structural model developed for a seismic refraction profile west of southern Moresby Island (Horn, 1982; Horn et al., 1984) provides structural and velocity information for the deeper structure of the terrace region. The velocities and layer thicknesses estimated from the data sections (table III) and constrained by these other studies were combined with the bathymetry profiles to form initial models for use w i t h the ray tracing routine. Continuity between receivers was provided by the Horn et al. (1982) model.  The final composite velocity model was  derived from analysis of data recorded at O B S W l for shots west of the O B S location (segment I) and from reversed data recorded at O B S s W l and W 2 for shots between the two O B S s (segment II). The model is presented in figure 15 for reference during the following modelling discussions of the individual profiles.  39  W o  DISTANCE 4.0  0.0  0.0 I  4.0 I  8.0  8.0 I  12.0  DISTANCE 12.0 I  (KM) 16.0  (KM) 16.0 I  E 20.0  20.0 I  24.0  \  28.0  W  24.0 I  F i g u r e 1 5 . F i n a l velocity model from ray tracing of airgun data. Velocities are in km/s, velocity gradients are in k m / s / k m . The upper model shows the division into individual profile models. Section I was modelled as O B S W l W E S T and section II was modelled as O B S W 2 and O B S W l E A S T .  ^  28 I  40 4.2 P r o f i l e  OBSW2  Sections O B S W l E A S T and O B S W 2 provide a reversed profile between O B S W l on a mid-terrace topographic high and O B S W 2 at the edge of the narrow continental shelf. The major amplitude decrease on both sections severely limits the extent of the region sampled by both data sets; good control of depth and dip in the central portion of the model is however still expected. Figure 16 shows rays traced through the final O B S W 2 model and compares the resulting traveltimes w i t h picks from the data. The model shows a two-layer structure w i t h low velocity sediment above a higher velocity unit. Lateral velocity variations in both layers are seen across the proposed Queen Charlotte Fault ( Q C F ) location. First arrival time picks indicate that the sediment beneath the continental shelf is of fairly high velocity (2.35 km/s). N o clear reflections from the top of the underlying structural unit were observed on the data section, suggesting a partial loss of energy transmitted through the sediments. Several flat-lying units of increasingly higher velocity were thus used to model the sediment layer, although it should be recognized that the energy loss may be from passage through deformed and faulted sediments instead of through several reflecting interfaces.  The arrivals between 27 and 23 k m in figure  16b are the modelled refractions through the shelf sediments with some later reflected arrivals from the underlying unit. A lower velocity (1.65-1.80 km/s) was required by traveltimes for the deformed sediments of the terrace itself. The complex structure observed on the C S P s was not compensated for in the model. High velocity gradients are required for the sediments in both regions.  41  15.0  21.0  DISTANCE  F i g u r e 1 6 . Ray tracing model (a) and traveltimes (b) for profile O B S W 2 . C o m p u t e d traveltimes are marked as x's, observed traveltimes are connected points. The distance scale and velocity structure correspond to those given for segment II in figure 15. O B S W 2 is at 28 k m as indicated.  42  29.0 LINE AG2  OBSW2  (b)  F i g u r e 17. Seismograms for profile O B S W 2 . (a) synthetic seismograms corresponding to figure 16. (b) vertical component observed data. The source wavelet used for all airgun synthetic seismograms is from data section O B S W l W E S T . Distance scales are model distances.  43 Arrivals from 23 to 16 km were modelled as refractions through the lower structural unit, which is composed of several blocks with high-angle boundaries modelled as vertical faults. A good fit to the undulating observed first arrival curve (figure 16b) was achieved by varying the velocity and depth of each individual block. The faulted unit beneath the continental shelf has a velocity of 4.1 km/s and a fairly high gradient of 0.40 k m / s / k m . This unit is separated from the main Q C F by a narrow block of much lower velocity (3.4 km/s) and higher gradient. West of the Q C F lies a slightly deeper unit of 3.8 k m / s velocity and moderate gradient. Figure 17 compares the synthetic seismograms with the processed data section. The computed amplitudes for refractions between 27 and 23 k m are much higher than the observed ones.  T h i s is presumably because of interference between these refractions  and direct waves which have not been modelled, and attenuation due to near-surface sediments which cannot be modelled with the program used.  Modelled refractions  through the deeper unit are seen as a smooth line of increasingly higher amplitude arrivals to approximately 17 k m model distance.  The reason for the sudden loss of  amplitude for corresponding first arrivals at about 18 km model distance on the data section is uncertain.  The amplitude loss was successfully reproduced with a model  that incorporated a low velocity zone ( L V Z ) east of the Q C F below 3 k m depth. Rays travelling upwards through the L V Z were reflected back at its upper boundary. However, the thickness of the L V Z could not be determined from the available data. This factor would have introduced a large degree of traveltime uncertainty into the explosion model for O B S W 2 and created a local anomaly within the regional structure so the L V Z model was rejected. The Q C F itself remains the most probable cause of the amplitude loss.  44  4.3 Profile O B S W l  EAST  The ray tracing model and traveltimes for arrivals at O B S W 1 are shown in figure 18. The model shown is the same as previously discussed for O B S W 2 .  The reversed  receiver position provides additional constraints for the western and central subsurface structure. Arrivals to 20 k m model distance have been modelled as refractions through the upper sediment layer.  A single unit of velocity 1.65 - 1.80 km/s and high gradient  (0.53 k m / s / k m ) was used to represent the sediment layer. The sediment deformation seen on C S P sections could not be duplicated at the scale of the model but is expected to reduce amplitudes of refractions through the sediments and reflections from the underlying unit. T h i s may explain why the triplication branch (a to b in figure 18b) from modelled reflections is not observed on the data section. Refractions through the lower structural unit generate the remaining first arrivals east of 20 km distance. The O B S W l  EAST  data section and corresponding synthetic seismograms are  shown in figure 19. Amplitudes for rays refracted through the upper sediment to 19 k m distance are higher than those observed on the data as expected. Arrivals for rays refracted through the lower unit show uniformly low amplitudes on the synthetic seismogram plot.  The corresponding arrivals on the data section have moderately high  amplitudes to 22.3 k m distance followed by a sudden decrease to the modelled amplitude level.  The arrivals continue with low amplitude to 24.5 k m where they appear  to stop. Very weak arrivals between 25 and 27 k m (figure 20) may be the attenuated continuation of these low amplitude refractions. A band of high amplitude arrivals with concave upward characteristics is seen on the data section between 22.5 and 25.5 km (figures 19b and 20). Their nature suggests  45  F i g u r e 1 8 . Ray tracing model (a) and traveltimes (b) for profile O B S W l E A S T . O B S W l is at 15 km model distance as shown.  46  F i g u r e 1 9 . Seismograms for profile O B S W l E A S T , (a) synthetic seismograms corresponding to figure 18. (b) observed data. The receiver is located at 15 km.  47  LINE AG2  OBSWl EAST  F i g u r e 2 0 . Detail from figure 19 showing low amplitude arrivals.  that they may be reflections but a suitable reflector position could not be modelled. Similarly, the inclusion of multiply-reflected rays in the model did not provide the necessary traveltime match. The arrivals have tentatively been attributed to a local focussing of rays.  48  4.4 Profile O B S W l  WEST  Profile O B S W l W E S T is a westward continuation of the previously modelled data recorded at O B S W l .  The profile is unreversed, affording no dip control, but lateral  continuity of segment II model features (figure 15) across the receiver location will provide velocity constraints. A two-layer structure has again been used to reproduce the data traveltimes (figure 21). The upper layer is a continuation of the sediment unit from the O B S W l model and as before does not compensate for sediment deformation.  EAST  The lower layer  has the same velocity as the adjacent unit in the O B S W l E A S T model but is located an average 1 km deeper to satisfy the traveltimes.  A slight eastward dip is required  for this unit; a horizontal interface and higher velocity would also have reproduced the traveltimes but would have required a lateral velocity discontinuity between sections. Ray tracing has duplicated the observed traveltimes well (figure 21b). Refractions through the upper sediment layer generated first arrivals from 12 to 8 k m model distance. A narrow triplication branch from the modelled interface reflections (a to b in figure 21b) was not observed on the data section; the discrepancy results from the difficulty in representing deformed sediments in the model. The interface may also be of a more gradational nature than the sharp boundary in the model; the data did not enable such fine stratigraphic detail to be determined. A small vertical offset and change in dip of the lower unit reproduce the slight bend at 4.5 km in the smooth first arrival time curve. The synthetic and observed seismograms for this section are shown in figure 22. Upper layer refractions again show larger amplitudes on the synthetic data than on the observed set for the same reasons mentioned earlier. As just discussed, the synthetic  49  DISTANCE 0.0  2.0  4.0  6.0  8.0  (KM) 10.0  12.0  14.0  16.0  16.0  F i g u r e 2 1 . Ray tracing model (a) W E S T . O B S W l is at 15 km model 16 k m overlaps the previous model distances correspond to section I in  and traveltimes (b) for profile O B S W l distance. The section between 15 and (figure 18). The velocity structure and figure 15.  50  16.0  (b)  F i g u r e 22. Seismograms for profile O B S W l W E S T , (a) synthetic seismograms corresponding to figure 21. (b) observed data, The receiver is at 15 k m distance as indicated.  51  reflection branch is not seen on the data section but could not be avoided in the model. Refractions through the deeper layer are seen as a smooth set of increasingly higher amplitude arrivals extending to the end of the synthetic section and match the observed amplitudes well.  4.5 T h e F i n a l  Model  The composite model combining the results for the entire airgun data set was shown in figure 15. The two-layer structure was interpreted as a variable sediment layer overlying a complex lower unit. Sediment thicknesses range from 0.4 to 1.6 km and velocity increases rapidly with depth. The lower unit shows a graben- and horst-like feature and has different velocity structure on either side of the Queen Charlotte Fault. A m plitude loss on profiles O B S W 2 and O B S W l E A S T suggest that an unmodelled, highly attenuative feature may exist in the inner terrace region. A n attenuation zone located at depths greater than 3 k m along the vertical fault would satisfy the requirements. Such a zone might be due to extensive shearing and deformation along the fault plane and could permit passage of energy from lower frequency explosive sources while attenuating the higher frequency airgun energy. M o d e l l i n g of the explosion data set should provide a look at the deeper structure and inhomogeneities of the terrace region.  52 CHAPTER V  INTERPRETATION  OF E X P L O S I O N  DATA  5.1 I n i t i a l C o n s t r a i n t s The explosive shots were detonated along a 110 km profile extending into the Pacific Ocean from the western edge of the terrace.  Energy generated by the 33 shots was  recorded by the colinear array of seismographs after it passed through the ocean crust, terrace, and continental material beneath the Queen Charlotte Islands. Initial constraints for a velocity model of the sampled region were obtained from numerous sources.  Several C S P s (Chase and Tiffin, 1972; Chase et al., 1975; Davis  and Seemann, 1981) and a nearby high density reflection seismic profile (P.D. Snavely, unpublished data) provided estimates of ocean sediment thickness and seismic basement dip.  Velocities of ocean sediment and upper crust were approximated from previous  studies to the north (Snavely et al., 1981) and south (Clowes and Knize, 1979) of the present study area. The model developed for the airgun data (Chapter I V ) defined the shallow structure of the terrace region. Constrai nts for the deeper structure were based on the Horn et al. (1984) refraction model. A study by Mackie (1985) provided initial estimates of depth to the Mohorovicic discontinuity beneath the survey region. A somewhat iterative procedure was used to develop the final model. A n initial model was constructed and modified through forward modelling to fit the first common receiver data section (receiver 010). This model was then expanded and altered to fit the second data section (receiver 020) and then further modified to provide a good fit for both data sections. Modelling continued in this manner with each subsequent model having to satisfy previous data sections within reasonable limits until the final model satisfied all sections.  Model reversal was then achieved by tracing rays from  various shots to all receivers through the previously modelled regions and comparing  53 the traveltimes and amplitudes with common shot record sections that show the data recorded by all receivers for each particular shot. The development of a single model degraded the computed traveltime and amplitude fits for each i n d i v i d u a l data section; however, the final model satisfies the total data set so its features are considered more representative of the entire region than those of the i n d i v i d u a l models. Five common receiver gathered sections were used to develop the final model: the two O B S d a t a sections, and three of the generally similar L B S sections recorded on Moresby Island (see appendix B ) . Two common shot gathered sections for large explosions were also used. The final velocity model is shown in figure 23 for reference during the following discussions of individual record section modelling. The main components of the model are: (a) an oceanic segment from 0 to 114 km; (b) the terrace region from 114 to 135.8 km; and (c) the Queen Charlotte Islands from 135.8 to 200 km distance.  5.2 R e c e i v e r 010 The first data modelled were the arrivals at receiver 010 ( O B S W l ) on the terrace. The initial model was constructed as a 130 km long segment extending from the ocean to receiver 010 and incorporated the model for airgun profile O B S W l W E S T between 115 and 130 k m distance. West of the terrace a uniform oceanic structure was inserted based on the velocities and depths of Horn et al. (1984). This structure was extended beneath the terrace region to the end of the model. Several problems with this initial model became apparent when ray tracing commenced.  A much lower velocity was required for the lower terrace and an eastward  d i p on the adjacent oceanic segment was necessary to sufficiently delay the computed  54  PA plate w _o.o  terrace  OISTRNCE IKMB100.0 ° 120.0 J i y__i 2  80.0  NA plate  u O140.0 *i  160.0  ^  160.0  200.0  V E 1:1  .0.0  20.0  o  0< * > 1 —  O  40.0  JL  O  ^DISTANCE (KM)  60.0 _l  80.0  100.0 _l  120.0  y  CM O  T  140.0  1_  O CO  O <o  y  y  o  O O  o  160.0  180.0  r  200.0  VELOCITY (km/s) ; GRADIENT (km/s/km)  1| | 1 .49; 0 001  5  3.80 ; 0.350  2jj  2.20 ; 0.530  6  3.40 ; 0.500  2.35; 0 700  7|p]  4.10 ; 0.400  3.80 ; 0 350  8g  5.00 ; 0.180  3S  9+ + 10ft 11&  6 .70 ; 0 012 6 6  CO  ; o 200 50 ; o 030 10 ;o 095 90 ;o 005  5 30  7  F i g u r e 2 3 . The final velocity model derived from airgun and explosion d a t a modelling. The positions of shots and receivers used for modelling are shown. T h e upper 5 k m between 115 and 143 k m correspond to figure 15. The upper figure shows the model w i t h no vertical exaggeration.  55  traveltimes to match the observed values. These and other parameters were varied until satisfactory traveltime and amplitude fits were attained and ultimately a final model was produced. Figure 24 shows the final ray tracing model for receiver 010 and compares the computed traveltimes with the observed times.  The earliest arrivals at near offsets are  refractions and reflections from the sediment and upper crust of the terrace. The calculated traveltimes for these rays are slightly slower than the observed times. However, the structure was well defined by airgun modelling so alterations were not made to the model to improve the traveltime fit. The boundary between the upper and lower crustal units was positioned so as to limit the lateral extent of upper crustal refractions to the region observed on the data. The oceanic boundary (unit 4 in figure 23) is vertically downdropped 2 to 3 k m with respect to the upper terrace (unit 5) to match the observed time delay of 1 s at 110 k m .  Traveltimes for refractions through the oceanic  crust require a thickening and eastward dip of the upper crust and overlying sediment between 80 and 113 k m . This approximately corresponds to the location of the Queen C h a r l o t t e Trough observed on bathymetry (figures 24a and 17). A low gradient (0.012 k m / s / k m ) is required for the lower oceanic crust (unit 9 in figure 23) to permit refractions to travel across the entire region; the velocity was adjusted accordingly to provide the best traveltime fit. The refractions appear as secondary arrivals (a to b in figure 24b) of apparent velocity 6.8 k m / s across most of the section. Reflections from the Moho (arrivals c to d) constrained its depth between 70 and 113 k m distance. A dip of 2.3° to 4.5° is required in this region. No depth or dip inform a t i o n is available west of 60 km so the Moho and shallower interfaces were extended horizontally to the end of the model. The effect of adding a component of eastward dip to these units, especially the Moho, would be an increase in the apparent velocity of  56  57  Receiver 010  (0.5-15.0 Hz)  (b) F i g u r e 2 5 . Seismograms for receiver 010. (a) synthetic seismograms corresponding to figure 24. (b) observed data. The three-cycle source wavelet used for the synthetic seismograms was taken from the data. The receiver is located at 130 km model distance. Low amplitude mantle refractions are indicated with arrows in (b).  140  58 mantle refraction arrivals. These arrivals (e to f in figure 24b) presently agree well with the observed arrivals of apparent velocity 7.9 km/s. The synthetic seismograms and observed data for receiver 010 are shown for comparison in figure 25. Rays which travelled through the upper terrace region have generally moderate amplitudes that are slightly lower on the data section for presumably the same reaons mentioned earlier. The amplitudes of ocean crust refractions gradually increase across the section from their initially moderate level near the outer fault; amplitude fluctuations  occur near 40 km as M o h o reflections interfere with the crustal refractions.  M a n t l e refractions have very weak amplitudes on the synthetic seismograms; they are barely visible on the data west of 40 k m because of the noise level. The amplitudes of the arrivals on the synthetic seismograms increase slightly w i t h increasing shot-receiver distance. This trend cannot be confirmed on the data section.  5.3 R e c e i v e r 0 2 0 Receiver 020 ( O B S W 2 ) was situated on the continental shelf just east of the terrace and across the Queen Charlotte Fault ( Q C F ) from O B S W l .  The initial model from  receiver 010 was extended across the remaining terrace region by inserting the reversed airgun profile O B S W l E A S T — O B S W 2 (segment II in figure 15) between 130 and 143 k m and assuming starting velocities for the region beneath it. Crustal velocity estimates and thicknesses were later refined by modelling. The velocities of the sediments and upper terrace crust between 130 and 144 k m were determined from the airgun data (figure 15). Ray tracing through the upper terrace region determined approximate locations for upper/lower crustal boundaries.  Lower  crustal velocities are not well constrained as these units represent only a small portion  59 of the rays' travelpaths.  A velocity of 5.3 km/s was chosen for the block west of the  Q C F (block A in figure 26) to be consistent with the adjacent terrace blocks (figure 23). T h i s required a rather high velocity of 6.5 km/s for the block east of the Q C F (block B in figure 26) to achieve a good traveltime fit. The vertical boundaries seen within this block are used only to define velocity gradients and do not represent structural features. Three groups of rays were used to model the observed traveltimes and amplitudes. The arrivals extending from 123 to 110 k m as first arrivals represent refractions through the upper terrace. Secondary arrivals of apparent velocity 7.2 km/s (a to b in figure 26b) represent crustal refractions through the lower terrace and ocean crust. The time delay at 110 km again clearly shows the transition from terrace to oceanic structure. The refractions through the ocean crust in the model end abruptly at 69 k m when the rays encounter the Moho.  Reflections from the M o h o are focussed into a narrow  group of secondary arrivals between 102 and 112 km (c to d in figure 26b). The arrivals indicate a dip of 19° for this portion of the M o h o .  Refractions through the mantle  are the first arrivals west of 100 km and extend across the modelled section w i t h an apparent velocity of 8.1 km/s (arrivals e to f in figure 26b). The synthetic seismograms and observed data for receiver 020 are presented in figure 27. The large amplitudes of the crustal refractions terminate abruptly at 69 km on the synthetic seismograms.  A fairly low signal-to-noise ratio makes it difficult to locate  the refraction arrivals on the data section beyond 90 km but they appear to continue to perhaps 55 k m . Extending the modelled refractions to this distance would require a deeper M o h o or lower terrace gradient, both of which would degrade the fit of the model to other data sets. Refractions through the mantle are of equally low amplitude on both sections. A n increase in amplitude with increasing shot-receiver distance is observed for the synthetic  60  DISTANCE o  0  20 0  0  40.0  60.0  80.0  (KM) 100.0  120.0  140.0  (a)  F i g u r e 2 6 . Ray tracing model (a) and traveltimes (b) for receiver 020, located at 143.1 km model distance. Numerous vertical lines in the model define velocity gradient directions and are not structural features.  160.0  61  o oo"  U °  LlJcd"  o CO  o rsi" 0.0  20,0  40.0  T  60.0  80.0  DISTANCE  100.0  120.0  140.0  (KM)  (a)  20  40  60 Receiver 020  80 (0.5-10.0 Hz)  100  120  (b) F i g u r e 2 7 . Seismograms for receiver 020. (a) synthetic seismograms corresponding to figure 26. (b) observed data. The receiver is located at 143.1 k m model distance. The arrows in (b) indicate low amplitude mantle refractions.  140  62 arrivals; the noise-laden signal on the data section unfortunately limits observations of amplitude characteristics at far offsets.  5.4 R e c e i v e r 030 Receiver 030 is a Backpack recorder located near the western shore of Moresby Island. Rays travelling to the receiver from all shots must pass through the terrace and the continental crust beneath the island. The velocity structure of the island is poorly denned as few studies have sampled the region. Extensive m a p p i n g of surficial geology provided the basis for a stratigraphic and structural analysis of the Queen Charlotte Islands (Sutherland B r o w n , 1968). Several geologic cross sections were constructed in a southwest-northeast direction across the islands and one of these is fortunately located within 3 km of the L B S s on Moresby Island. T h i s profile (figure 28) can be divided into three major sections. West of the Louscoone Fault the subsurface is composed primarily of Triassic lavas of the Karmutsen Formation overlain in regions by Upper Triassic limestones of the K u n g a Formation. Severe deformation and faulting has occurred between the Louscoone and Rennell Sound Faults. Near-surface units within the fault zone are primarily Jurassic and Cretaceous in age and generally consist of volcanics of the Yakoun Formation, calcareous siltstone of the L o n g a r m Formation, and sandstones of the Haida and Honna Formations. East of the Rennell Sound Fault lie volcanics of the Yakoun Formation which are overlain by H a i d a Formation sandstones in the western half. Vertical offset across the RennellLouscoone Fault zone is estimated as 600 to 3000 m with the east block dropped down (Sutherland B r o w n , 1968). The stratigraphic interpretation of the profile was used to obtain crude estimates of upper crustal thickness and velocity for the model.  63 MITCHELL Sol lit  BOTTLE  Point  HARBOUR  OOUGLAS  INLET  INLET  UUOGE  INLET  4 000'-  TM "EEL  KHo  "RKA  INLET  Louscoone  040  Fault;  GILLATT  x"RKu  "RKu.JY  050  060  KL  TM  LAKE  COPPER  RIVER  KL  UKA  "RJKo  Sandspit SKIOEGATE  Rennell  ARM  COPPER  Sound  070  KL KHo  RJKU  Fault  KHA  "RKA  Fault BAY  HECATE  STRAIT  -4000  100  Sea  2  JY  TERTIARY TM M a s s e t t Fm:  b r e c c i a s , columnar  mi  basalts, rhyolites  CRETACEOUS AND TERTIARY KT P o s t T e c t o n i c P l u t o n s : diorite,  quartz,  granite  CRETACEOUS Queen C h a r l o t t e  Group  KS S k i d e g a t e Fm:  shale,  siltstone,  sandstone  KHo Honna Fm:  conglomerate, f e l d s p a t h i c  KHA  shale,  H a i d a Fm:  KL Longarm Fm:  sandstone  sandstone  calcareous  siltstone,  lithic  sandstone  JURASSIC JY Y a k o u n Fm:  volcanic  sandstone, t u f f ,  TRIASSIC A N D JURASSIC JK , Kunga Fm: f l a g g y a r g i l l i t e T/jKfj Kunga Fm: f l a g g y and m a s s i v e  sill  (  limestone  TRIASSIC  TpK  A  Karmutsen  Fm:  limestone,  p i l l o w lavas,  flows  F i g u r e 2 8 . Geologic cross-section across Moresby Island near the receiver locations (after Sutherland B r o w n , 1968). Approximate receiver positions for this experiment are indicated.  Level  64  DISTANCE  (KM)  (a)  oo  F i g u r e 2 9 . Ray tracing model (a) and traveltimes for receiver 030, located at 154.5 k m model distance.  65  o  00  140.0  0  20  40  60 Receiver 030  80 (1.5-8.5 Hz)  100  120  F i g u r e 3 0 . Seismograms for receiver 030. (a) synthetic seismograms corresponding to figure 29. (b) observed data. The receiver is located at 154.5 k m model distance.  140  66  The ray tracing model and traveltime plots are shown in figure 29. Three groups of rays were again involved in the interpretation. Refractions through the lower crust of the terrace are seen as first arrivals from 122 km until the outer fault offset at 110 km.  A r r i v a l s from 108 to 87 km with apparent velocity 7.6 km/s (a to b in figure  29b) are from rays that continued through the fault into the lower oceanic crust until they were stopped at the M o h o .  Reflections from the Moho are again limited to a  branch of secondary arrivals from 110 to 114 km (c to d) and confirm the previously determined M o h o dip of 19°. Deeper travelling rays were refracted at the Moho and travelled through the mantle to form the 8.2 km/s apparent velocity branch (e to f) that intersects the lower crustal refractions at 103 k m on the traveltime plot. The synthetic and observed seismograms are shown in figure 30. The high amplitude arrivals seen on the d a t a between 110 and 90 km correspond well to the modelled refractions through the lower oceanic crust. Additional energy is provided by the slightly later M o h o reflections between 103 and 113 km. The much weaker mantle refractions continue across both sections from 100 km and increase slightly in amplitude with distance.  5.5 Receiver 060 Receiver 060 is a Backpack recorder located midway across Moresby Island in the Rennell-Louscoone Fault zone (figure 28). The surface geology indicates that vertical displacement of units adjacent to the fault zone has certainly occurred. However, rays arriving at receiver 060 are apparently not affected so there is no major velocity contrast across the fault at the depth concerned.  67 The ray tracing model and corresponding traveltimes are shown in figure 31. The only change from the previous model is the extension of the island structure to the receiver 060 location. The fault and vertical offset shown at 172 km have no effect on ray paths or traveltimes and were included in the final model to satisfy the receiver 100 d a t a discussed in the next section. The results of the ray tracing are very similar to those for receiver 030.  Crustal  refractions, M o h o reflections, and mantle refractions are again the predominant ray groups involved.  The high terrace gradient and M o h o depth have further restricted  the range to which crustal refractions can occur and all arrivals modelled west of 100 k m are mantle refractions. The eastward dip of the M o h o between 60 and 110 km is clearly evident in these refraction arrivals as they show an unrealistically high apparent velocity (8.8 km/s) in this region before assuming the more reasonable 7.9 km/s for the horizontal region. The seismograms (figure 32) show the concentration of high amplitude crustal refraction arrivals east of 98 km and the continuous low amplitude mantle refractions extending westward across the section. Arrivals on the data section are very clear to large shot-receiver distances, especially for the big shots, and show well an amplitude increase w i t h distance of the theoretical mantle refractions.  5.6 Receiver 100 The last common receiver gather modelled was for receiver 100, a microcorder located near the eastern shore of Moresby Island. Rays travelling to the receiver from the shots will be passing through lower crustal material east of the Rennell-Louscoone Fault zone and will be sensitive to the thickness of the crustal blocks. Combining the  F i g u r e 3 1 . Ray tracing model (a) and traveltimes (b) for receiver 060, located at 174.8 k m model distance.  69  0.0  20.0  80.0  60.0  DISTANCE  100.0  120.0  140.0  (KM)  (a)  60 Receiver 060  80 (2.0-12.0 Hz)  (b) F i g u r e 3 2 . Seismograms for receiver 060. (a) synthetic seismograms corresponding to figure 31. (b) observed data. The receiver is located at 174.8 k m model distance.  140  70 constraints of this model with those of receiver 060 should provide an estimate of vertical displacement at depth across the R L F zone. The island structure of the receiver 060 model was extended uniformly to receiver 100; the upper crustal structure east of the R L F was then varied in subsequent modelling. The model that provides the best overall fit is shown with corresponding traveltimes in figure 33. Traveltimes for crustal refractions, seen on the time plot between 115 and 104 k m , are satisfied by downdropping the easternmost crustal block 0.4 to 2.0 k m . The best fit for all arrivals was achieved with a vertical displacement of 1.1 km. The remaining arrivals are attributed to mantle refractions and a narrow band of M o h o reflections between 105 and 116 km (c to d in figure 33b). The refraction arrivals (e to f) have an apparent velocity of 11.1 km/s for the dipping M o h o region and 8.06 km/s through the horizontal boundaries to the west. The amplitudes of the synthetic seismograms (figure 34) match the observed seismograms well. The crustal refractions yield high amplitudes which are augmented by the M o h o reflections. Low amplitude mantle refractions lead away from the the high amplitude arrivals starting at 105 k m and continue westward across the section with slightly increasing amplitudes.  5.7 Shot 13 Shot 13 was a 540 kg explosion detonated approximately 48 km west of O B S W l at 81.7 k m model distance. The ray tracing model and traveltimes are presented in figure 35.  Rays were traced from the shot position to the surface through the previously  F i g u r e 3 3 . Ray tracing model (a) and traveltimes (b) for receiver 100, located at 194.3 k m model distance.  72  CO  o 1  oJn  0.0  20.0  1  40.0  i  60.0  DISTANCE  i  80.0  i  100.0  i  120.0  140.0  (KM)  (a)  Receiver 100 (2.0-120 Hz) (b)  F i g u r e 3 4 . Seismograms for receiver 100. (a) synthetic seismograms corresponding to figure 33. (b) observed data. The receiver is located at 194.3 k m model distance.  140  73 derived model. O B S positions were time-corrected to the surface for easy comparison of observed and computed traveltimes. The terrace region was well sampled at all depths by the various ray groups, which consist of refractions through the oceanic crust and terrace, reflections from the oceanic M o h o , and refractions through the mantle that surface through the terrace and island crust. The time plot shows good agreement between the theoretical and observed traveltimes. The computed first arrivals at all receivers are from rays refracted through the mantle. The secondary arrivals picked for receivers 010 and 020 (the OBSs) correspond to M o h o reflections and crustal refractions respectively.  The most distant reflection  from the Moho surfaced at approximately 138 km so O B S W l is the only receiver in this model that could record Moho reflections from this shot. Figure 36b shows the observed d a t a recorded by the ten receivers. Amplitude corrections were made for the different types of seismographs (Chapter II) but did not consider the siting of the instruments. While the L B S s were carefully sited on bedrock, the O B S s were presumably dropped into soft ocean bottom sediments. The resulting poor coupling would reduce the amplitudes of the recorded signals and make theoretical amplitude comparisons difficult. The synthetic seismograms are presented in figure 36a. The high amplitude arrivals between 120 and 135 km are from Moho reflections. No corresponding high amplitudes were observed on the data recorded at receiver 010 ( O B S W l ) . Similarly, the moderate amplitude crustal arrivals seen on the synthetic seismograms between 140 and 150 km were not observed on the O B S W 2 data. Presumably the generally low amplitudes of these two observed seismograms are a result of their location on the soft ocean bottom. The mantle refractions observed across the remainder of the section generally match the arrivals recorded by the L B S s .  74  F i g u r e 3 5 . Ray tracing model (a) and traveltimes (b) for shot 13, located at 81.7 km model distance. Receiver locations are indicated.  75  LUio" LT) O CO  a.  120.0  130.0  _! 140.0  (  150.0  ! 160.0  DISTANCE  !  (KM)  170.0  180.0  190.0  200.0  (a)  130  120  140  150  160 170 Shot 13 (2.0-12.0 Hz)  180  190  •(b) F i g u r e 3 6 . Seismograms for shot 13. (a) synthetic seismograms corresponding to figure 35. (b) observed data, from receiver 010 at left to receiver 100 at right. The shot is located at 81.6 k m model distance.  200  76  5.8 Shot 25 Shot 25 was a more distant 540 kg explosion detonated approximately 91 km west of O B S W l at 38.4 km model distance. Rays traced from this shot to the surface near the 10 receivers sampled the ocean crust west of the terrace, the terrace itself, and the upper mantle and island crust. The model and traveltime plot are shown in figure 37. Three groups of rays were again used to reproduce the observed traveltimes. The first group was refracted through the oceanic crust and travelled near-horizontally before being refracted sharply upwards to the surface by the high velocity gradient within the terrace. These rays are seen as secondary arrivals between 124 and 138 k m on the traveltime plot (a to b in figure 37b). The second ray group was reflected from the shallowly dipping M o h o west of the terrace and travelled upwards through the oceanic and terrace regions.  The corresponding  traveltime arrivals begin at 80 k m and become successively later to join the crustal refraction arrivals at 133 km (c to d). The final group represents rays that were refracted into the mantle at the oceanic M o h o and were subsequently refracted back to the surface through the various crustal structures at the dipping Moho beneath the terrace. These rays are the first time arrivals between 100 and 200 km (e to f) and fit the observed traveltimes well. The " V " shape of the arrivals is due to the dip of the M o h o and the velocity structural contrast between the terrace and the island. The synthetic and observed seismograms are shown in figure 38. The low overall amplitudes and the noise on the receiver 010 and 020 traces makes amplitude comparison difficult but there appears to be good general correlation between the sections. C r u s t a l refractions and Moho reflections are the source of the high amplitude secondary arrivals between 120 and 135 k m . The reduced amplitude on the synthetic trace at 130  F i g u r e 3 7 . Ray tracing model (a) and traveltimes (b) for shot 25, located at 38.4 k m model distance.  78  LlJlD" CD  O 00  —  130.0  120.0  l 140.0  150.0  160.0  170.0  180.0  160 170 Shot 25 (2.0-12.0 Hz)  180  DISTANCE  (KM)  ~l 190.0  200.0  (a)  W  q CO  G  120  130  140  150  190  F i g u r e 3 8 . Seismograms for shot 25. (a) synthetic seismograms corresponding to figure 37. (b) observed data, receivers 010 to 100. The shot is located at 38.4 k m model distance.  200  79 k m is the result of destructive interference between the two types of arrivals. Low amplitude mantle refractions can be followed across the section and reproduce the observed amplitudes quite well.  5.9 T h e F i n a l V e l o c i t y  Model  The final model based on the complete ray tracing interpretation was presented in figure 23. The model can be divided into three major structural components (ocean, terrace, and island) which will be discussed individually. The region extending from the beginning of the model to the edge of the terrace at 114 k m comprises the oceanic segment. A n average 1 k m of sediments (unit 2) overlie a 6.2 k m thick oceanic crust which has been subdivided into a lower velocity 1.3 k m thick upper unit (unit 4) and a higher velocity lower unit (unit 9). A low velocity gradient (0.012 k m / s / k m ) is required for unit 9 to enable crustal refractions from the distant shots to reach all the receivers. Beneath the crust lies the oceanic mantle (unit 13) w i t h a velocity of 7.9 km/s. West of 60 km a horizontal Moho at 10 k m depth satisfied the available data. East of 60 km the Moho begins to dip towards the mainland, first at a gentle 2.3° and then at 4.5° to the edge of the terrace. The depth of the Moho in this region is well constrained by reflections and the cessation of crustal refractions. The terrace region extends from 114 km to the Q C F at 135.8 km. A l t h o u g h it too is composed of three major units, it is structurally much more complicated than the oceanic region.  The uppermost layer consists of low velocity deformed sediments of  irregular surface topography and variable thickness. The upper crustal unit has been separated into three blocks (unit 5) by vertical faults to satisfy the airgun data, although  80 a more complicated structure may be present. The vertical extent of the faults into the adjacent units could not be determined; surface topography however suggests that the fault at 130 km extends to the ocean floor. The lower crustal unit (unit 10) has an anomalously low velocity and fairly high gradient as required by the crustal refractions. It has been modelled as a homogeneous unit although faults may complicate the structure. The M o h o dips beneath the terrace at 19° as determined from mantle refractions and a few unreversed M o h o reflections. The M o h o dip and mantle depth are dependent on the velocity structure of the terrace and would vary if it were altered. G o o d agreement is found, however, w i t h the 20° dip determined by Horn et al. (1984) off southern Moresby Island, although their model was more poorly constrained at this level than that of the present study. Mackie (1985), in a seismic refraction study of the Queen Charlotte Islands/Hecate Strait region, determined a M o h o dip of 5° extending from the outer edge of the terrace to the mainland of British C o l u m b i a and a 2° dip beneath the oceanic crust to the west. The simplified velocity structure that he used for the terrace and Q C Islands will permit small variations in these values, but the study nonetheless indicates a shallower regional Moho dip than the 19° value obtained in this study for the limited region beneath the terrace. T h i s suggests that the steeper dip beneath the terrace may be local and the dip may decrease beneath the Q C Islands. The island region east of the Q C F to 200 km represents the edge of the continental structure of North America. Sampling density decreases towards the end of the model and the structure can only be roughly approximated. The region adjacent to the Q C F appears distinct from the other island structure at all model depths. A 1 km thick unit of stratified, 2.35 km/s sediments (unit 3) overlies the crustal units. The upper crust again appears to be faulted and is divided into two distinct velocity units at 137 k m .  81 The division has been modelled as a vertical fault but is not so constrained. The upper crust (unit 7) has been modelled with a slightly higher velocity (4.1 km/s) than the oceanic crust (3.8 km/s) with the exception of the 1.2 km wide sliver next to the Q C F (unit 6) which has a lower velocity (3.4 k m / s ) . A higher velocity and lower gradient were determined for the lower crust here (unit 11) than for the adjacent lower continental crust (unit 12). However, these values are not well constrained and the two units may be much more similar. The remaining island structure has been roughly divided into two units based on stratigraphic information.  The upper unit (unit 8) has a velocity of 5.0 km/s at the  surface and moderately high gradient.  There is little velocity difference at the 5 km  deep boundary between units but the gradient has been significantly reduced for the lower unit (unit 12). A modelled offset on the Rennell-Louscoone Fault has displaced the eastern half of the upper unit 1.1 k m downwards with respect to the adjacent part. The fault dip cannot be resolved to better than ± 30° nor the vertical displacement to ± 0.8 k m . The mantle was not sampled beneath the island and total crustal thickness is therefore unknown. A geologic interpretation of the velocity structure is necessary to fit the derived model into the tectonic framework for the region. The next chapter will attempt to do this and answer some of the questions raised at the beginning of the study.  82  C H A P T E R VI  DISCUSSION AND  CONCLUSIONS  The application of two-dimensional ray tracing techniques has produced a velocity structural model for the region of the Queen Charlotte Fault zone. The fault—bounded terrace does not appear to be related to either of the adjacent crustal units but seems structurally distinct. A geologic interpretation of the velocity model will help to clarify the relationship between the observed structure and the present tectonic regime.  6.1 Stratigraphic Interpretation The stratigraphic interpretation of the velocity units is summarized in table I V . U n i t 2 is the sedimentary layer covering the terrace and ocean crust. Ocean bottom sediments west of D i x o n Strait were interpreted as Late Miocene to Quaternary deposits (Snavely et al., 1981). Chase and Tiffin (1972) identified trench infill as Late Tertiary pelagic sediments overlain by well stratified Plio-Pleistocene turbidites and recent turbidity current deposits. U n i t 3 is a higher velocity sedimentary unit underlying the narrow continental shelf east of the active fault and may be of terrigenous origin. U n i t s 4 and 9 represent upper and lower oceanic crust respectively. Ophiolite studies indicate that crustal layer 2 (unit 4) consists of an extrusive sequence of pillow lavas and basaltic sheet flows that grades down to dense dike swarms (Christensen and Smewing, 1981). A n associated high velocity gradient is related to the increase of metamorphic grade w i t h depth. Layer 2 velocities range from 3.7 to 6.5 km/s and average 5.5 k m / s (Christensen and Salisbury, 1975). The difference between ophiolite sequence velocities of 5.0 to 6.0 k m / s measured in the laboratory and the unit 4 model velocities of 3.8 to 4.3 k m / s suggests the presence of large fractures within the uppermost crustal section which would significantly lower observed velocities (Christensen and Smewing, 1981).  83  VELOCITY (km/s)  GRADIENT  1  1.49  0.001  water layer  2  2.20  0.530  pelagic and turbiditic sediments  3  2.35  0.700  shelf sediments  4  3.80  0.350  basaltic pillow lavas and sheet  VELOCITY UNIT  (km/s/km)  STRAT1GRAPHIC INTERPRETATION  flows grading to dike swarms 5  3.80  0.350  deformed upper ocean crustal basalts or compressed sediments  6  3.40  0.500  sheared sediments or basalts  7  4.10  0.400  sheared sediments  8  5.00  0.180  tholeiitic basalts, granite, and limestone  9  6.70  0.012  interlayered gabbros and peridotites  10  5.30  0.200  sheared gabbros and peridotites ??  II  6.50  0.030  continental silicic crust  12  6.10  0.095  continental silicic crust  13  7.90  0.005  harzburgite and dunite (mantle)  T a b l e I V . Stratigraphic interpretation of the velocity units of figure 23.  84 Oceanic crustal layer 3 consists of interlayered gabbros and peridotites of amphibolite facies metamorphism. Velocities in the Oman Ophiolite range from 6.7 km/s at the top of the layer to 7.3 k m / s at its base (Christensen and Smewing, 1981). The lower crustal velocity determined by modelling (6.7 km/s for unit 9) agrees well with these laboratoy measurements. U n i t 13 represents the upper mantle which is composed primarily of tectonized harzburgite and dunite (Christensen and Smewing, 1981). The modelled mantle velocity (7.9 km/s) for the 8 M a old oceanic lithosphere is lower than average mantle velocity measurements (8.15 km/s) but is in agreement with other observations of young oceanic lithosphere (Christensen and Salisbury, 1975). Velocity unit 8 represents the upper crustal rocks of the Wrangellia terrane. The K a r m u t s e n Formation, which outcrops at the surface between 190 and 212 km model distance, consists primarily of Late Triassic tholeiitic pillow basalts (Sutherland Brown, 1968). Estimates of ocean basalt velocities range from 3.7 to 6.5 km/s and a velocity of 5.04 k m / s has been measured for basalts at the top of oceanic layer 2 (Christensen and Salisbury, 1975).  A n average velocity of 5.0 km/s is thus representative of the  volcanics of the Karmutsen Formation and the younger rocks of the Yakoun Formation that outcrop to the east. U n i t 12, the lower continental crust, is poorly constrained in the model. Its upper boundary is not a velocity discontinuity but instead marks a reduction in gradient from the upper crust. The reduced velocity gradient required by the ray trace modelling may signify a change in metamorphic facies or composition. The stratigraphic equivalents of the terrace velocity units are not readily apparent. The velocity of unit 5 suggests that it may be upper oceanic crust (unit 4) although de-  85 formation by thickening and faulting has clearly occurred. Alternatively, it may contain highly compressed sediments of oceanic or continental origin. Hyperbolae and disruptions, seismic signatures of irregular surface features, make it impossible to distinguish any subsurface structure on the available C D P data. U n i t 7 may similarly be composed of sediment derived from the upper continental crust. The low velocity and position of the narrow unit 6 suggests that it has been extensively sheared and fractured by motion along the fault and may belong to either of the above mentioned crustal units. Lower crustal unit 11 east of the fault is again not well defined and is probably related to the lower continental crust, unit 12. Unit 10 , however, has a low velocity and high gradient inconsistent with the other lower crustal units. The most acceptable hypothesis is that extensive deformation by compression and shearing has altered normal lower crustal material beyond easy recognition and masked the boundary between upper and lower crustal material.  6.2 Tectonic Implications The fault structure of the velocity model is consistent with a compression al tectonic regime. The g r a b e n - and horst-like appearance of the upper terrace may in fact represent high angle reverse faults and low frequency compressive sediment folds as the fault angles are Only constrained to ± 10°. The downward vertical displacement of the oceanic crust at the outer fault (114 k m distance) agrees with the previously mentioned vertical fault plane solution (Chapter I) of Berube (1985). Hypocentral locations of microearthquakes (Hyndman and Ellis, 1981) require a westward dip of 80° on the inner fault, consistent w i t h the velocity model. The remaining major fault is the Rennell-Louscoone Fault at 172 k m model distance on Moresby Island. Sutherland  86 B r o w n (1968) observed dips ranging from 75° northeastwards to vertical on the associated fault strands. T h i s suggests a tensional mechanism. The rotation and shifting of the Queen Charlotte Islands along this fault zone during the initiation of spreading in Queen Charlotte Sound (Chapter I) would have produced the necessary tensional regime. The ongoing vertical uplift of the western edge of the islands by lithospheric flexure  (Riddihough, 1982a) may produce additional vertical offset on this fault in fu-  ture. Two tectonic mechanisms will be examined to explain the observed terrace structure: accretion associated w i t h active subduction, and structural compression associated with convergence at a transform fault. The sharp increase in model Moho dip from 2.3° west of the terrace to 19° beneath the terrace is supportive of subduction of the Pacific plate beneath the Queen C h a r l o t t e Islands.  Similar values were obtained from refraction  studies for southern Moresby Island (a poorly constrained 20° - Horn et al., 1984) and for the Juan de Fuca P l a t e beneath Vancouver Island (14-16° - Spence et al., 1985). A lower value of 5° (Mackie, 1985) for the Queen Charlotte Islands region implies that M o h o dip decreases beneath the islands but is still consistent with subduction processes (see also discussion in section 5.9). The slow rate of convergence across the plate margin (10 to 20 mm/a) may cause weak coupling between the subducting Pacific plate and the A m e r i c a plate. The expected tectonic processes for such a case are: ( l ) accretion by deposition of sediment and subsequent tectonic incorporation into an accretionary complex; (2) sediment subduction of material ponded in a trench axis; and (3) erosion by failure of the margin front through mass wasting and subsequent incorporation into subducting sediment (von Huene, 1984).  87 Trench sediments seen in C S P profiles (eg. Chase and Tiffin, 1972; Chase et al., 1975) are relatively undisturbed except for minor faulting so it is unlikely that process (2) is occurring on a large scale. However, tectonic erosion of the margin has almost certainly occurred. Strike-slip faulting, an important erosion process in zones of oblique compressional stress (von Huene, 1984), appears to have removed some continental material as evidenced by the truncated spurs and triangular seaward-facing facets of the coastal mountains on the Queen Charlotte Islands (Keen and H y n d m a n , 1979). Sedimentary accretion of material from the subducting ocean crust to the front of the margin could have produced the complex structure of the upper terrace.  Seismic  records across the landward slope of the eastern Aleutian Trench (von Huene, 1984) reveal the presence of 3 k m thick blocks of sediment, separated by steep high angle thrust faults, beneath younger sediments at the continental margin (figure 39). lower sediments may be subducting with the oceanic crust.  The  Accreted sediments and  possibly imbricated oceanic crust were imaged beneath southern Vancouver Island by high density seismic reflection profiling.  Similar structures may compose the upper  Queen Charlotte terrace. The interpretation of the upper terrace as an accretionary sediment wedge would of course require the presence of subducting oceanic crust beneath it.  N o internal  reflections from within the lower terrace were detected, providing no support for a structural division. It is quite possible, however, that the top of the subducting crust is extremely irregular as a result of strike-slip faulting and convergence in the terrace region. Strong reflections would not be generated in such a case and any energy would be severely attenuated by the deformed overlying sediments. The velocity structure of the ocean crust could then have been amalgamated with the overlying terrace structure to produce the single "averaged" velocity unit of the model.  88  SCALES FOR  B, C  F i g u r e 3 9 . Seismic profiles across the Aleutian Trench, illustrating the structures that result from obduction, imbrication, accretion, and subduction of sediments at a convergent margin (after von Huene, 1984).  Alternatively, the complex structure of the terrace may result from compression associated w i t h oblique convergence at the fault zone. H y n d m a n and Ellis (1981) proposed a model in which faulting occurs on the Queen Charlotte strike-slip fault with the shallow portion of the thrust zone locked. Convergence is taken up either by periodically unlocking the upper portion of the thrust or by compressively deforming and shortening the whole oceanic lithosphere beneath the terrace. Hyndman et al. (1982) extended the transcurrent fault into the underlying oceanic lithosphere to be consistent w i t h the microseismicity observed primarily between 10 and 25 km depth. They suggested that the faulting must j u m p seaward at intervals to remain beneath the edge of the continental crust — terrace sediment contact (figure 40). Both models require a sharp downward bend of the oceanic lithosphere beneath the margin, possibly with associated normal faulting at the outer edge of the terrace.  89  0  Deformation Front  Q.C. Trough  Terrace  Coast  =7  Possible old fault or q< E 10- discontinuity ~j NlC — — X rc  A  Q.C. Islands  TERRACE SEDIMENTS  j  USr  CONTINENTAL : : : : CRUST:  H  DE  a.  20-  30-  No vertical exaggeration  Main Transcurrent Fault Fossil Faults ?  ^  F i g u r e 4 0 . A possible tectonic model of the Queen Charlotte Fault zone. Oblique convergence is resolved into strike-slip motion on the fault parallel to the margin and a small component of underthrusting perpendicular to the margin (Hyndman et al., 1982).  Numerous earthquakes recorded in the recent microseismicity study of Berube (1985) were located in the Queen Charlotte fault zone along Moresby Island; a few were located on the terrace and outer fault scarp. The focal depths of the events extend to 25 km although most of the activity was well constrained to depths of less than 15 km. Thus, seismic activity at depths greater than 13 k m would be located within oceanic crust if subducted crust is assumed to underlie the terrace in the velocity structural model of figure 24. The Queen Charlotte fault would thus have to extend into the oceanic crust, as H y n d m a n et al.  (1982) suggested, and essentially "slice" it as transform motion  occurs. The slicing would have to be discontinuous, perhaps by alternating transform  90 motion and subduction, to allow the continued movement of the oceanic material across the fault zone and beneath the continental lithosphere. Lowell (1972) modelled strike-slip movement with a small component of convergence across the fault zone using a clay cake on moving sheets of tin.  The relative  motion vector of the model was oriented at 15° to the fault trace, similar to the Queen Charlotte region. The material at the fault zone moved upwards to produce a welt with downward tapering wedges and upthrust margins (figure 41). The steep faults modelled for the terrace region may be upthrust faults formed in a similar manner as opposed to underthrusts formed by accretion.  F i g u r e 4 1 . Conceptual diagram of an upthrust-bounded welt created by convergent strike-slip or transform motion. The welt has downward-tapering wedges and upthrust margins (Lowell, 1972).  91  T h r u s t mechanisms and near-vertical faulting were computed by Berube (1985) for two earthquake clusters at 12 and 9 km depth respectively on the Queen Charlotte Fault zone near the present refraction study area. The shallow focal depths and the clay cake models suggest that thrusting of terrace material above any subducting oceanic crust is occurring as a result of compression in the fault zone. The structure of the terrace can thus be attributed to two different tectonic processes which may occur concurrently or alternate as necessary as shown in figure 42a. Convergent motion may be accommodated by slow, shallow subduction of the oceanic lithosphere beneath the Queen Charlotte Islands. Sediment overlying the oceanic crust would be accreted to the seaward edge of the terrace and/or subducted beneath the terrace. Transform motion is occurring as dextral strike-slip on the Queen Charlotte fault and possibly on associated faults on the Queen Charlotte Islands and beneath Hecate Strait.  Sedimentary and upper crustal material may be thrusting upwards in  the terrace (figure 42b) as subduction is halted at the fault and compression needs to be accommodated. The compressive stress in the lower oceanic lithosphere would continue to build during this phase and may force the resumption of subduction and the subsequent migration of faulting in the downgoing crust. The refraction model developed in this study supports recent subduction but suggests that compression and upthrusting of terrace material presently account for a major part of the component of convergence at the active Queen Charlotte fault zone.  92  O c e a n i c P l a t e  T e r r a c e  C o n t i n e n t a l P l a t e  F i g u r e 4 2 . Proposed tectonic processes in the Queen Charlotte Fault region, (a) subduction - sediments and upper crust deform by accretion and compression as subduction occurs, (b) transform motion - compression in sediments and crust relieved by upthrusting along the fault zone.  93 6.3  Conclusions The structural model developed for the Queen Charlotte Fault zone confirms an  anomalous low velocity terrace region separating oceanic crust to the west and continental crust to the east. The terrace crustal structure consists of a 3 km thick upper unit w i t h velocity 3.8 km/s at the top and fairly high (0.35 km/s/km) gradient, and a lower unit with an average thickness of 10 k m , velocity 5.3 km/s, and gradient 0.20 km/s/km.  The terrace is bounded on the west by an old fault zone and on the east  by the currently active near-vertical Queen Charlotte Fault; steep faulting also extends through at least the upper crustal unit of the terrace. M o h o dip steepens considerably beneath the terrace from 5° to 19°. The structural complexity of the terrace may result from a combination of accretion and deformation by underthrusting associated with subduction, and compression and u p t h r u s t i n g related to oblique convergence at the transform fault. A d d i t i o n a l study is necessary to provide the structural resolution required to distinguish between the two. High resolution multichannel seismic reflection profiling over the terrace would better define the velocities and fault orientations and possibly locate the top of subducted oceanic lithosphere. Similarly, further sampling is necessary to constrain the vertical offset of 1.1 km modelled on the Rennell-Louscoone Fault. Reversed refraction profiles across and along the islands would provide information presently lacking about regional crustal structure and would better constrain M o h o depth and dip. The addition of such future observations to the existing geophysical and geological information will further the understanding of the complicated tectonics and complex structure of the Queen Charlotte Fault zone region.  94  BIBLIOGRAPHY Atwater, T., 1970, Implications of plate tectonics for the Cenozoic tectonic evolution of western North A m e r i c a : Geol. Soc. A m . Bull., 81, 3513-3536. Berube, J . , 1985, A seismicity study of the Queen Charlotte Islands/Hecate Strait region: M.Sc. thesis, U n i v . of British Columbia, Vancouver, 100 pp. B i r d , D . N . , 1981, T i m e - t e r m analysis using linear programming and its application to refraction d a t a from the Queen Charlotte Islands: M.Sc. thesis, Univ. of British Columbia, Vancouver, 79 pp. Chase, R.L. and Tiffin, D.L., 1972, Queen Charlotte Fault-Zone, British Columbia: P r o c . Int. Geol. Congr. 24th, section 8, 17-27. Chase, R.L., Tiffin, D.L. and Murray, J . W . , 1975, The western Canadian continental margin, in Yorath, C.J., Parker, E.R. and Glass, D.J., Eds, Canada's continental margins and offshore petroleum exploration: C a n . Soc. Petr. Geol., Memoir 4, 701-721. Christensen, N.I. and Salisbury, M . H . , 1975, Structure and constitution of the lower oceanic crust: R e v . Geophys. Space Phys., 13, 57-86. Christensen, N.I. and Smewing, J . D . , 1981, Geology and seismic structure of the northern section of the O m a n ophiolite: J . Geophys. Res., 86, B 4 , 2545-2555. Clowes, R . M . , 1984, Acquisition of a crustal refraction profile across the Queen Charlotte Islands and Hecate Strait: E a r t h Physics Branch Open File Report 84-22, Ottawa, 55 pp. Clowes, R . M . , B r a n d o n , M . T . , Green, A . G . , Yorath, C.J., Sutherland Brown, A . , Kanasewich, E.R. and Spencer, C , 1986, Lithoprobe - southern Vancouver Island: Cenozoic subduction complex imaged by deep seismic reflections: Can. J . E a r t h Sci. in press. Clowes, R . M . and Knize, S., 1979, Crustal structure from a marine seismic survey off the west coast of Canada: C a n . J . E a r t h Sci., 16, 1265-1280. Coney, P.J., 1970, T h e geotectonic cycle and the new global tectonics: Geol. Soc. A m . Bull., 81, 739-748. Coney, P.J., Jones, D.L. and Monger, J . W . H . , 1980, Cordilleran suspect terranes: Nature, 288, 329-333. C o u c h , R. W., 1969, G r a v i t y and structures of the crust and subcrust in the southeast Pacific Ocean west of Washington and B r i t i s h Columbia: P h . D . thesis, Oregon State Univ., Corvallis. Currie, R.G., Stevens, L.E., Tiffin, D.L. and Riddihough, R.P., 1980, M a r i n e geophysical survey of the Queen Charlotte Islands (abstr.): Trans. A m . Geophys. Union ( E O S ) , 61, 71. Davis, E.E. and Seemann, D.A., 1981, Compilation of seismic reflection profiles across the continental margin of western Canada: Geol. Surv. Canada Open File Report 751.  95 Dehlinger, P., Couch, R.W., M c M a n u s , D.A. and Gemperle, M., 1970, Northeast P a cific structure in M a x w e l l , A . E . , E d . , The sea, vol. IV, part II: Wiley-Interscience P u b l . , New York, 133-190. Hefner, D.E. and Barret, D.L., 1979, O B S development at Bedford Institute of Oceanography: Mar. Geophys. Res., 4, 227-245. Hilde, T . W . C . , Uyeda, S. and Kroenke, L., 1977, Evolution of the western Pacific and its margin: Tectonophysics, 38, 145-165. Horn J.R., 1982, A snapshot of the Queen Charlotte Fault zone obtained from P wave refraction data: M.Sc. thesis, Univ. of British Columbia, Vancouver, 79pp. Horn J.R., Clowes, R . M . , Ellis, R . M . and B i r d , D.N., 1984, The seismic structure across an active oceanic/continental transform fault zone: J . Geophys. Res., 89, B 5 , 3107-3120. H y n d m a n , R.D. and Ellis, R . M . , 1981, Queen Charlotte fault zone: microearthquakes from a temporary array of land stations and ocean bottom seismographs: C a n . J . E a r t h Sci., 18, 776-788. H y n d m a n , R.D., Lewis, T . J . , Wright, J.A., Burgess, M., C h a p m a n , D.S. and Y a mano, M., 1982, Queen Charlotte fault zone: heat flow measurements: Can. J . E a r t h Sci., 19, 1657-166 9. H y n d m a n , R.D., R i d d i h o u g h , R.P. and Herzer, R., 1979, The Nootka fault zone a new plate boundary off western Canada: Geophys. J . Roy. A s t r . S o c , 58, 667-683. H y n d m a n , R.D. and Weichert, D.H., 1983, Seismicity and rates of relative motion on the plate boundaries of western N o r t h America: Geophys. J . Roy. Astr. S o c , 72, 59-82. Irving, E., 1979, Paleopoles and paleolatitudes of North America and speculations about displaced terrains: Can. J . E a r t h Sci., 16, 669-694. Keen, C . E . and H y n d m a n , R.D., 1979, Geophysical review of the continental margins of eastern and western Canada: C a n . J . E a r t h Sci., 16, 712-747. K i r s t i u k , S., 1981, A preliminary investigation into the structure of the upper crust in the terrace region of the Queen Charlotte fault zone: B.Sc. thesis, Univ. of B r i t i s h Columbia, Vancouver, 77 pp. Lowell, J.D., 1972, Spitsbergen Tertiary orogenic belt and the Spitsbergen fracture zone: Geol. Soc. A m . Bull., 83, 3091-3102. M a c k i e , D., 1985, Subduction beneath the Queen Charlotte Islands? Results of a seismic refraction survey: M.Sc. thesis, U n i v . of British Columbia, Vancouver, 130 pp. Minster, J . B . and Jordan, T . H . , 1978, Present day plate motions: J . Geophys. Res., 83, 5331-535 4. O ' B r i e n , P.N.S., 1960, Seismic energy from explosions: Geophys. J . , 3, 29-44.  96  Riddihough, R.P., 1977, A model for recent plate interactions off Canada's west coast: Can. J . E a r t h Sci., 14, 384-396. Riddihough, R.P., 1982a, Contemporary movements and tectonics on Canada's west coast: a discussion: Tectonophysics, 86, 319-341. Riddihough, R.P., 1982b, One hundred years of plate tectonics in Western Canada: Geoscience Canada, 9, 28-34. Riddihough, R.P., Currie, R.G., and H y n d m a n , R.D., 1980, The Dellwood knolls and their role in triple junction tectonics off northern Vancouver Island: Can. J . E a r t h Sci., 17, 577-593. Rogers, G.C., 1983, Seismotectonics of B r i t i s h Columbia: P h . D . thesis, Univ. British Columbia, Vancouver, 247 pp.  of  Sheriff, R.E. and Geldart, L.P., 1982, E x p l o r a t i o n seismology, vol. I: Cambridge Univ. Press, Cambridge, 253 pp. Snavely, P.D., Wagner, H.C., Tompkins, D . H . and Tiffin, D.L., 1981, Preliminary geologic interpretation of a seismic reflection profile across the Queen Charlotte Island fault system off D i x o n Entrance, Canada - United States: U.S. Geol. Surv. Open File Report 81-299. Spence, G.D., 1984, Seismic structure across the active subduction zone of western Canada: P h . D . thesis, Univ. of British Columbia, Vancouver, 191 pp. Spence, G.D., Clowes, R . M . and Ellis, R . M . , 1985, Seismic structure across the active subduction zone of western Canada: J . Geophys. Res., 90, B 8 , 67546772. Spence, G . D . , W h i t t a l l , K.P. and Clowes, R . M . , 1984, Practical synthetic seismograms for laterally varying media calculated by asymptotic ray theory: B u l l . Seism. Soc. A m . , 74, no. 4, 1209-1223. Srivastava, S.P., 1973, Interpretation of gravity and magnetic measurements across the continental margin of British C o l u m b i a , Canada: Can. J . E a r t h Sci., 10, 1664-1667. Stone, D.B., 1977, Plate tectonics, paleomagn etism, and the tectonic history of the N.E. Pacific: Geophysical Surveys, 3, 3-37. Strutt, J.W., 1917, O n the pressure developed in a liquid during the collapse of a spherical cavity: Philosophical Mag., 34, 94-98. Sutherland Brown, A., 1968, Geology of the Queen Charlotte Islands: B r i t . Dept. of Mines and Petr. Res., Bulletin 54, 226 pp.  Col.  Tjaden, G., 1981, Seismic modelling of refraction data from the shelf off the Queen Charlotte Islands: B.Sc. thesis, Univ. of B r i t i s h Columbia, Vancouver, 50pp. van der Voo, R., Jones, M., Gromme, C.S., Iberlein, G . D . and C h u c k i n , M., 1980, Paleozoic paleomagn etism and northward drift of the Alexander terrane, southeastern Alaska: J . Geophys. Res., 85, 5281-5296.  97 von Huene, R., 1984, Tectonic processes along the front of modern convergent margins — Research of the past decade: A n n . Rev. E a r t h Planet. Sci., 12, 359-381. W h i t t a l l , K.P. and Clowes, R.M., 1979, A simple simple, efficient method for the calculation of traveltimes and ray paths in laterally inhomogeneous media: J . C a n . Soc. Explor. Geophys., 15, 21-29. Willis, H.F., 1941, Underwater explosions, time interval between successive explosions: British Report W A - 4 7 - 2 1 . Yorath, C . J . and Cameron, B . E . B . , 1982, O i l on the West Coast?: G E O S , 11, 13-15. Yorath, C . J . and Chase, R.L., 1981, Tectonic history of the Queen Charlotte Islands and adjacent areas - a model: C a n . J . Earth Sci., 18, 1717-1739. Yorath, C . J . and H y n d m a n , R.D., 1983, Subsidence and thermal history of Queen Charlotte Basin: Can. J . E a r t h Sci., 20, 135-159. Young, I.F., 1981, Geological development of the western margin of the Queen Charlotte basin: M.Sc. thesis, Univ. of B r i t i s h C o l u m b i a , Vancouver, 380 pp.  98 APPENDIX  A  A i r g u n data. Vertical component data sections for airgun line AG2. Distance scales match the distances used in the modelling; shot-receiver distances are also indicated. Spreading corrections have been applied to the data; relative trace amplitudes are otherwise preserved (refer to chapter II).  13.0  Increasing shot-receiver distance (km)  0.0  14.0 LINE AG2  00  OBSWl EAST  Increasing shot-receiver distance (km)  14.0  102  APPENDIX  B  Explosion data. Vertical component data sections for explosion line E X l . Filtering parameters are given with each section. Shot numbers are the central 3 digits of the 5—digit code number plotted with each trace. Distance scales are model distances; shot-receiver distances are also indicated. Spherical spreading corrections and charge size compensations have been applied to the data as described in chapter II. The data plots show relative trace amplitudes after application of these corrections.  co<\j—  01 oo r»  IDI/I  T i n  M  - o o i n r - u )  in  ^ f o r v j - o u i a j ^ a j m ^ r n r v i  Receiver 020 (0.5-10.0 Hz) 143.1  Increasing shot-receiver distance (km)  8.1  (n  CM  o  o  in 7 c i <\i o c n o o r ^ t D in ^ n M - o o m r - i o m v OM C M — — — — — o o o o o CM ( M (M C\J o o o o o o o o o o o o o o o o o o o o o o r n r o r o r o m r n r o r o r o r o r o r o r n r o r o r n m m  —  m  D  r-  cc  o  CM  (M  (M  CM  o o o t n m r o  i  o o  n  o o  i  M o o  o o  O  0 154.5  45 60 75 90 Receiver 030 (1.5-8.5 Hz) Increasing shot-receiver distance (km)  105  120  135 19.5  C  0 174.8  15  30  45 60 75 90 Receiver 060 (2.0-12.0 Hz) Increasing shot-receiver distance (km)  105  120  135 39.8  m  c\j  —  o  o  o  00  0 184.0  oo co  15  co c\i o  r-  <M o  oo oo  30  ID  in  (M <M o o CO 00  CM  o OO  rn  CM  o CO  CM CM  o CO  —  o  o CO  oo  CM CM  o  ifl in o CO  o  o  00 00  o  o  co  coco  N  o  —  o  —  o  o o  o o  o o  o o  o o  o o  o o  o o  o o  c o o o o o a o o o o o o o o o c o o o o o c o  45 60 75 90 Receiver 080 (2.0-12.0 Hz) 1 ncreasin g shot receiv er distance (kin)  105  120  135 49.0  Receiver 090 (1.5-8.5 Hz) 189.5  Increasing shot-receiver distance (km)  135 54.5  


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