RESEARCH Nature of the Moho Transition in NW

RESEARCH

Nature of the Moho transition in NW Canada from combined near-vertical and wide-angle seismic- reﬂ ection studies

J. Oueity1,* and R.M. Clowes1 1DEPARTMENT OF EARTH AND OCEAN SCIENCES, UNIVERSITY OF BRITISH COLUMBIA, 6339 STORES ROAD, VANCOUVER, BC V6T 1Z4, CANADA

ABSTRACT

Although the year 2009 marked one century since the discovery of the Moho, or crust-mantle boundary, the exact nature of that boundary and the manner in which it formed remain major uncertainties in lithospheric studies. In northwestern Canada, sharp Moho refl ections at both near-vertical and wide-angle incidence have been imaged beneath the Great Bear magmatic arc. They show a remarkably fl at Moho and do not refl ect the complex tectonic history of the Wopmay orogen, of which the Great Bear arc is one part. In order to understand the origin of these refl ections and the nature of the Moho, we calculated near-vertical and wide-angle synthetic seismograms for a number of crust-mantle transition models using one- and two-dimensional wave propagation algorithms. Only laterally and vertically heterogeneous models can properly simulate the observed seismic signature recorded on both near-vertical and wide-angle refl ection data. The heterogeneity is achieved by either laterally discontinuous layering or a lamellae structure with randomly distributed ellipses. These models suggest that the Moho represents a thermal/metamorphic front, a regional décollement, or both.

LITHOSPHERE; v. 2; no. 5; p. 377–396. doi: 10.1130/L103.1

INTRODUCTION granulite, gabbro, and amphibolite and an upper mantle composed of several varieties of peridotite with some eclogite. The crust-mantle boundary, or the Moho, is one of the most distinct Despite all the geologic studies of the Moho by means of exposed sec- manifestations of a differentiated Earth. Major changes in petrology, min- tions (ophiolites) or xenolith samples, seismic data still remain the main eralogy, chemistry, seismic wave velocity, density, and rheology occur source for our understanding of the structure of the Moho and its develop- across it (Jarchow and Thompson, 1989). Originally, the Moho was defi ned ment. As techniques applied to lithospheric studies have improved con- on the basis of seismic head waves (Mohorovičić, 1910). It was described siderably in terms of seismic data acquisition technologies (e.g., Meiss- as the depth at which the compressional P-wave velocity increases rapidly ner, 1986; Mooney, 1987; Mjelde et al., 1997; Panea et al., 2005) and or discontinuously to 7.6–8.6 km/s (Steinhart, 1967). This suggested a pet- computational capabilities (e.g., Fuchs and Müller, 1971; Spence et al., rologic interpretation of the Moho as a fi rst-order (i.e., zero thickness) dis- 1984; Levander and Holliger, 1992; Zelt and Smith, 1992; Carbonell et al., continuity in rock composition from predominantly mafi c lower-crustal 2002), it has become apparent that the Moho is typically not a uniform, rocks to predominantly ultramafi c upper-mantle rocks. Ophiolite studies sharp, fi rst-order discontinuity, but rather a complex and variable transi- indicate that the lithologic sequence of the oceanic Moho includes pelagic tion zone. From the seismic perspective, a frequently cited model for the sediments, pillow basalts, massive and layered gabbro, and residual ultra- Moho is characterized by a variable-thickness transition zone composed mafi c tectonites of the upper mantle (Casey et al., 1981; Karson et al., of anastomosing layers progressing with depth from mafi c to ultramafi c 1984; Boudier and Nicolas, 1995; Jousselin and Nicolas, 2000; Dilek et rocks (Jarchow and Thompson, 1989; Braile and Chiang, 1986). Alterna- al., 2008). These rock cumulates form laterally discontinuous lenses with tively, the Moho transition may represent a change in scale of vertical lay- thicknesses ranging from 1 cm to several tens of meters and aspect ratios ering and horizontal extent of inhomogeneities (Tittgemeyer et al., 1999; ranging from 10:1 to 100:1. Prominent refl ections from within the oceanic Hurich, 2003; Carpentier and Roy-Chowdhury, 2007). Moho transition zone obtained by seismic-refl ection methods also support Although the Moho is well imaged by near-vertical incidence (NVI) a geologically complex boundary (e.g., Hasselgren and Clowes, 1995; and refraction/wide-angle refl ection (R/WAR) data, the exact nature and Nedimović et al., 2005). characteristics of the transition are not well understood. The wave fi eld Unlike abundant ophiolites, which allow direct observation of the oce- (i.e., waveform, duration, and amplitude) of a seismic signal (refl ected or anic Moho, exposed sections of the continental Moho are rare (Fountain refracted) from the lower crust and upper mantle may contain signifi cant and Salisbury, 1981). Nevertheless, important constraints on the nature information about the structure of the Moho transition zone, its fabric, and of the continental Moho are provided by the small continental data set scale of heterogeneities. Only a few studies have specifi cally addressed of xenoliths (McGetchin and Silver, 1972; Debari et al., 1987; Kopylova this issue. These studies can be grouped into two categories, qualitative et al., 1998). Xenolith samples indicate a lower crust composed of mafi c and quantitative. Examples of qualitative analyses of refl ection patterns near the Moho are those of Gibbs (1986) on Consortium for Continen- *Corresponding author: +1-604-822-5703; fax: +1-604-822-6088; e-mail: joueity@ tal Refl ection Profi ling (COCORP) seismic-refl ection profi les, Meissner eos.ubc.ca. (2000) using Deutsches Kontinentales Refl exionsseismisches Programm

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

(DEKORP) data, and Hammer and Clowes (1997), Cook (2002), and teristics of Moho refl ections by calculating the synthetic seismic signature Eaton (2006) on Lithoprobe refl ection transects that sample regions of for both near-vertical and wide-angle data of a number of crust-mantle diverse age and tectonic history. From these studies, refl ection patterns transition models using one- and two-dimensional wave propagation algo- appear to be divided into three primary classes: (1) a distinct, multicyclic rithms. The models range from a simple fi rst-order discontinuity to more band of refl ectors that are laterally continuous or piece-wise continuous complex, laterally and vertically heterogeneous structures. Unraveling over tens of kilometers, (2) fading out of crustal refl ectivity with depth, the detailed structure of the crust-mantle boundary and the way in which and (3) no clear refl ections in the vicinity of the Moho. These studies did it forms can help us understand some major tectonic processes, such as not involve wide-angle data. Although the comparisons have not clearly crustal growth, accretion, and delamination (Morozov et al., 2001; Car- linked a specifi c tectonic process with a distinct refl ection pattern, sharp bonell et al., 2002). Moho refl ections have been associated with lower-crustal deformational processes including compression, extension, and intrusive features. TECTONIC BACKGROUND Quantitative studies use synthetic modeling to construct the seismic response of the continental Moho using seismic velocity data (laboratory The Wopmay orogen in northwest Canada (Fig. 1) is a Paleoprotero- seismic-refraction results) together with petrologic information derived zoic assembly of domains that developed during the Calderian orogeny from exposures of the crust-mantle boundary when available. In earlier between 1.92 Ga and 1.84 Ga (Hoffman, 1989). The orogen is divided into studies, the Moho was modeled as a fi nite-thickness laminated zone of four major north-south–trending domains: the Coronation Supergroup, the alternating high- and low-velocity layers (Clowes and Kanasewich, 1970; Great Bear magmatic arc, the Hottah terrane, and the Fort Simpson ter- Hale and Thompson, 1982; Fountain, 1986; Braile and Chiang, 1986). rane. Rifting to the west of the Slave Province at ca. 1.90 Ga initiated More recently, the Moho has been considered as a laterally and vertically deposition of the Coronation Supergroup, a west-facing shelf rise and heterogeneous transition where the heterogeneities are represented by ran- foredeep succeeding prism, which shortly thereafter was intruded by a domly distributed ellipsoids (Carbonell et al., 2002; Tittgemeyer et al., suite of plutons and translated eastward by the Hottah collision to form a 2003). Another representation of the Moho follows a statistical approach, foreland fold-and-thrust belt (Hoffman, 1988). in which the transition zone is characterized by a stochastic heterogeneity The Hottah terrane formed as a magmatic arc and collided with the distribution (Hurich, 2003; Nielsen and Thybo, 2006; Carpentier and Roy- western Slave craton ca. 1.89–1.88 Ga (Hoffman and Bowring, 1984; Hil- Chowdhury, 2007). Only a few of these studies included consideration of debrand and Bowring, 1999). The collision caused compression, shorten- wide-angle data (Carbonell et al., 2002; Tittgemeyer et al., 2003; Nielsen ing, and eastward thrusting of the Coronation Supergroup onto the Slave and Thybo, 2006). craton (Hoffman and Bowring, 1984). The lack of coeval arc magmas on In this paper, we analyzed the near-vertical incidence and wide-angle the western Slave craton indicates that the accretion of the Hottah ter- seismic data acquired along line 1 of Lithoprobe’s Slave–Northern Cordil- rane onto the craton was the result of west-dipping subduction of an lera Lithospheric Evolution (SNORCLE) transect in the Paleoproterozoic– ocean basin. The Hottah terrane consists of Paleoproterozoic sedimentary Archean domains of the Northwest Territories, Canada (Cook et al., 1999; and intermediate volcanic rocks metamorphosed to amphibolite grade Fernández Viejo and Clowes, 2003). We investigated the dynamic charac- and intruded by calc-alkaline granitic plutons during the period 1.940–

Cordillera Wopmay Slave 1100 CS N 64° Figure 1. Tectonic map of the study area and location of Litho- probe’s Slave–Northern Cor- dillera Lithospheric Evolution 1104 Deformed (SNORCLE) near-vertical incidence refl ection and refraction/ 1101 Ancestral 1105 wide-angle refl ection line 1. Stars North Fort identify shot locations; some are America Simpson identifi ed by numbers. Sediments of the Phanerozoic Western Can- N ada sedimentary basin overlie 1112 the Precambrian domains west 1108 of the long dashed line. Short N 60° YK dashed black lines show politi- BC cal boundaries. CS—Coronation Supergroup, SD—Sleepy Dragon, 1113 GSLsz—Great Slave Lake shear zone, AB—Alberta, BC—British Columbia, NWT—Northwest Ter- W 128° 108° ritories, YK—Yukon. Inset shows location of map within Canada. 120° Seismic line 1

Extent of Phanerozoic cover Shot points

378 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

1.902 Ga (Hildebrand et al., 1987). To the west, another oceanic plate was parameters are given in Cook et al. (1999). For such data, the nominal converging against the western Hottah terrane, generating an east-dipping vertical resolution is given by λ/4, where λ is the wavelength, and the subduction zone below the Hottah terrane and Coronation Supergroup. horizontal resolution is given by the radius r of the fi rst Fresnel zone, It led to the formation of the Great Bear magmatic arc during the period rz=λ/2, where z is the depth (Yilmaz, 2001). Thus, at a Moho depth between 1.875–1.843 Ga (Hildebrand et al., 1987; Gandhi et al., 2001). of ~32 km where the seismic velocity is ~6.8 km/s (Fernández Viejo and Between the Coronation Supergroup and the Hottah terrane, there lies Clowes, 2003) and the dominant frequency determined from Moho refl ec- the 100-km-wide Great Bear magmatic arc, which consists of volcanosedi- tions is ~21 Hz, the vertical resolution is ~80 m and the horizontal resolu- mentary sequences and plutonic rocks (Hildebrand et al., 1987). Aeromag- tion is ~2.2 km. netic data indicate that rocks of the Great Bear arc extend from north to Refraction/wide-angle refl ection data were recorded from 13 explosive south beneath the younger Proterozoic and Paleozoic cover for a total length shots along the same SNORCLE line 1. The total number of seismographs of ~900 km (Coles et al., 1976). At its southern end, the magmatic arc is was 594, with station spacing of 1.2 km. Further details of the fi eld acqui- ~2.0–4.5 km thick, as interpreted from seismic-refl ection data along SNOR- sition can be found in Fernández Viejo and Clowes (2003). Interpreta- CLE line 1 (Cook et al., 1999). These data show a strong refl ection from the tion of the R/WAR data followed standard procedures (Zelt and Smith, top (~0.2 s) of the arc, which has been drilled, and another strong refl ection 1992) and provided information on the two-dimensional (2-D) seismic at ~1.5 s from its interpreted base. Whether the thin basin-like feature rep- velocity structure. The estimated uncertainties in velocities based on pro- resents the complete Great Bear section at this latitude or whether there are cedures discussed by Fernández Viejo and Clowes (2003) are given in additional Great Bear rocks below is not known. In either case, development their Table 3. For the WAR data, traveltime modeling indicated vertical of the arc does not appear to have disrupted the older structure beneath it, at resolution at lower-crustal depths on the order of 1–2 km and horizontal least within the resolution of the refl ection survey. One possible explanation resolution between 60 and 80 km (Fernández Viejo and Clowes, 2003). for the intact structure underneath is that the arc magmas migrated through the crust along relatively narrow preexisting zones of weakness without dis- Processing of the Refl ection Data rupting the primary crustal structure (Cook et al., 1999). The entire Great Bear–Hottah–Slave assembly is deformed by a set The refl ection data were reprocessed following standard procedures of brittle conjugate faults, which have been interpreted as the result of with parameters (Table 1) that are similar to, but not identical with, those the terminal collision of the Fort Simpson exotic terrane with the western used by Cook et al. (1999). The main difference in our procedure is with margin of the Hottah terrane (Hoffman and Bowring, 1984). The collision handling noise. Incoherent noise present in the data corrupts the quality of the Fort Simpson terrane with the Hottah terrane occurred after the of the signal and may lead to misinterpretation. Thus, separation of signal formation of Fort Simpson magmatic rocks, ca. 1.845 Ga (Villeneuve et and noise is an important step in data processing, particularly in the lower al., 1991), and pre–1.71 Ga mafi c intrusions of sedimentary rocks (Abbott crust where the signal-to-noise ratio (SNR) is low. Traditional seismic et al., 1997). Formation of the Fort Simpson basin, identifi ed as such on processing takes advantage of the redundant (multifold) data to improve SNORCLE line 1 by seismic-refl ection and seismic-refraction data (Cook SNR. In this study, we adopted an additional and new technique, based on et al., 1999; Welford et al., 2001), was the result of lithospheric exten- the curvelet transform (Hennenfent and Herrmann, 2006; Neelamani et sion, which followed the collision of the westernmost terranes with the al., 2008), to suppress incoherent noise and improve the resolution of the older Paleoproterozoic domains. From the refl ection data, the basin is seismic data at deeper levels, including the Moho (e.g., Fig. 2). interpreted as a monocline with a base that dips westward from a depth of The curvelet transform belongs to a family of multiscale and multi- ~3 km at the eastern end to a depth of ~24 km at the western end, revealing directional transforms. A curvelet is localized in frequency and pseudo- at least 20 km of subsurface relief (Cook et al., 1999). localized in space (rapid spatial decay). In the physical domain, it looks The enlarged continental assembly was affected by a number of crustal like a small plane wave that is oscillatory in one direction and smooth extension phases that continued to Neoproterozoic time (Thompson et al., in the perpendicular direction. The signal and noise have minimal over- 1987). The Proterozoic rocks of the Wopmay orogen along SNORCLE lap in the curvelet domain (Neelamani et al., 2008), making the curvelet line 1 are overlain by Phanerozoic sedimentary rocks that thicken from transform an ideal choice for our purpose. The underlying assumption their feather edge near the eastern boundary of the Great Bear magmatic of constructing curvelets is that any object with wavefront-like structure arc to ~1000 m over the Nahanni domain (Fig. 1; Cook et al., 1999). The (e.g., seismic images) can be represented by a relatively few signifi cant Nahanni domain, west of the Fort Simpson terrane, is the westernmost transform coeffi cients (Candès et al., 2006). component of the Wopmay orogen in the study area. It has no surface exposure and is only distinguished by its magnetic signature. TABLE 1. PROCESSING PARAMETERS FOR NEAR-VERTICAL SEISMIC EXPERIMENT AND DATA INCIDENCE (NVI) DATA Parameter Value SNORCLE Line 1 Crooked line geometry 30 m × 2000 m bin Static correction The main purpose of the SNORCLE transect studies was to investigate Trace balance Average amplitude scaled to be 1.0 the various processes involved in the westward growth of North America Band-pass fi lter 10–40 H z from the Archean to the present using a multidisciplinary approach (Cook Deconvolution Filter length = 150 ms; and Erdmer, 2005). Vibroseis near-vertical incidence refl ection data were gap length = 2.0 ms Automatic gain control 500 ms window acquired along SNORCLE seismic line 1, one of four such lines in the Normal moveout (NMO) correction transect. Line 1 is 725 km long and crosses the southwest Slave craton and on common midpoint gathers Wopmay orogens (Fig. 1). In total, 404 geophones per record with a sta- Stack Nominally 189-fold tion spacing of 60 m were used. Shot points were every 90 m with sweep Curvelet denoising 5 scales, 64 angles frequencies ranging between 10 and 80 Hz. Complete fi eld acquisition Plot

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 379

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

SNORCLE Line 1 N

SW Great Bear magmatic arc NE

Figure 2. The seismic-refl ection profi le beneath the Great Bear magmatic arc along Lithoprobe’s Slave–Northern Cordillera Litho- spheric Evolution (SNORCLE) line 1. Strong Moho refl ections are visible at ~11.0 s between 15 and 60 km and at ~11.5 s between 0 and 15 km. Block arrows bound a region of dipping refl ectivity in the middle and lower crust.

Time (s) Time Curvelet denoising (see text) has been applied to the stacked seismic section; no migration. Inset shows location of the proﬁ le (white) with respect to line 1.

Moho Moho 12 10 8 6 4 2 0 2 4 6 8 10 12 Moho 0 30 60 Distance (km)

The noisy seismic data (y) can be represented in terms of curvelets as: lation of synthetic seismograms from model representations of the Moho. To simulate the stacked section accurately, calculation of a large number y =+CxT n, (1) of shot gathers along a high-resolution 2-D velocity model is required. This process is computationally very expensive, and a high-resolution 2-D where x is the curvelet transform vector, CT is the curvelet transform syn- velocity model is not available. In addition, some of the processing steps thesis operator, and n is the noise. The denoising problem can be solved by preceding the stack may introduce some artifacts to the data. a series of the following unconstrained optimization problems (Herrmann An alternative approach is to identify Moho refl ections on individ- and Hennenfent, 2008): ual shot gathers and then attempt to simulate these through synthetic seismogram modeling. Previously, this has not been attempted because xyCxx =−+arg min T λ (2) individual shot gathers are characterized by low SNR ratios, and deeper x 2 1 refl ections, particularly at the base of the crust, are usually obscured where x is the estimated curvelet transform coeffi cient vector, and λ is the by random noise. We overcame this problem by utilizing the aforemen- regularization parameter that determines the trade-off between data con- tioned curvelet denoising technique to attenuate incoherent noise on the sistency and the sparsity in the curvelet domain. We start with high λ and shot gathers. Our results show remarkable refl ections at ~11.0 s that we decrease its value until yCx−≈T ε, where ε is proportional to the noise interpret as the Moho. The refl ections are characterized by a narrow 2 level (Elad et al., 2005). This step corresponds to the solution of the opti- band of 5–9 cycles that are laterally variable and piecewise continuous mization problem. The fi nal, noise-free reconstructed image is given by m for several kilometers or more (Fig. 3). = CTx. Further information and discussion concerning curvelet denoising and its application to seismic-refl ection data are given in Kumar (2009) Processing of the Refraction Data and from V. Kumar (2010, personal commun.). The stacked and denoised section of the seismic image beneath the Processing of the refraction data involved editing/killing noisy traces Great Bear magmatic arc shows a distinct Moho transition at ~11.0 s from and band-pass fi ltering (3–15 Hz). To offset distances of ~160 km, the ~15 to 60 km and a slightly deeper Moho at 11.6 s from 0 to 15 km distance crustal refracted phase, Pg, is the fi rst arrival (Fig. 4). Prominent second- (Fig. 2). The transition consists of a multicyclic band of refl ections sepa- ary wide-angle refl ections from the crust-mantle boundary (PmP) beneath rating a highly refl ective crust from a relatively transparent upper mantle. the Great Bear magmatic arc are observed on three shots (1104, 1105, and Ideally, we would like to simulate this Moho transition response by calcu- 1108). They are characterized by a prominent reverberatory pattern with

380 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

Offset (km) Offset (km ) -10 -5 0 5 10 -10 -5 0 5 10 0 0

5 5

(

10 10

11 11

0 km 5 0 km 5

12 12 -10 -5 0 5 10 -10 -5 0 5 10 0 0

5 5

(

10 10

11 11

0 km 5 0 km 5 12 12

Figure 3. Different shot gathers after curvelet denoising (see text). Note the distinct band of reﬂ ections at ~11.0 s. Enlargements of Moho reﬂ ections are shown in the insets.

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 381

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

10 1.0 9 A 0.8

(

0 8 0.6

8 / 0.4

- 7 PmP

Amplitude e 0.2

i 6 SP 1104

T) 0.0 Pg -160 -140 -120 -100 -80 5 SP 1104 1.0 4 180 160 140 120 100 80 0.8 Offset (km) 0.6 0.4

10 Amplitude 0.2 SP 1105 0.0 9 -200 -180 -160 -140 -120 -100 -80

(

0 8 . 1.0

X PmP 0.8

- 7

e Pn 0.6

i 6 0.4 T) Pg

Amplitude 0.2 5 SP 1108 SP 1105 0.0 80 100 120 140 160 4 220 200 180 160 140 120 100 Offset (km) Offset (km)

10 B 1.0 9

) Offset (km)

s 0.8

(

100

. 8 120

8 / PmP 0.6

X 140 - 7

m 0.4 i 6 Pg

T Pn

Amplitude

5 0.2 SP 1108 4 0.0 100 120 140 160 180 200 220 0 5 10 15 20 Offset (km) Frequency (Hz)

Figure 4. Record section for shots 1104, 1105, and 1108. Data are band- Figure 5. (A) Amplitude versus offset of the observed PmP phase on shots pass fi ltered from 3 to 15 Hz and displayed as trace-normalized ampli- 1104, 1105, and 1108. Amplitude values are normalized to unity. Amplitude tudes. Pg is a crustal, fi rst arrival, refracted phase. PmP is a prominent, behavior is erratic, and no common peak is observed among the shots. second arrival refl ection from the Moho observed at offsets >80 km, fol- (B) Normalized amplitude spectra of the observed PmP phase. The spec- lowed by ~0.5-s-long coda. Pn is a weak, refraction phase in the upper tra are calculated on single traces at 100, 120, and 140 km offsets and over mantle observed at offsets >160 km. a 0.5 s window.

duration of ~0.5 s between 80 and 220 km offset (Fig. 4). At offsets greater horizontal refl ections correspond remarkably well with the horizontal than 160 km, the refraction phase Pn traveling within the uppermost man- Moho refl ections observed on the NVI data. tle becomes the fi rst, but rather weak, arrival. Because of the ambiguity of For amplitude analysis, we calculated the PmP amplitudes within a 0.5 the Pn phase, further amplitude/frequency analysis of this phase was not s window length after each traveltime pick for this phase (Fig. 5A). Only pursued. A normal-moveout–corrected and stacked section of shot gathers amplitudes associated with pick uncertainties <0.15 s were considered on which the secondary phase PmP was observed showed clear, horizontal (Table 2). All trace amplitudes were initially scaled such that “true relative refl ections at ~11 s that extend laterally over ~200 km below the Hottah amplitude” between data of various instrument types was achieved; see terrane and Great Bear magmatic arc (Fernández Viejo et al., 1999). These Ellis et al. (2002) for details. The amplitude versus offset plots for the

382 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

TABLE 2. AVERAGE UNCERTAINTY ASSOCIATED terms, Qs is about one half the value of Qp (Fowler, 2005). Although Qs WITH THE PICKS OF PMP AND PN PHASES must be specifi ed within the algorithm, this study does not include shear- IN SHOTS 1104, 1105, AND 1108 wave arrivals, so the value specifi ed is not signifi cant. Values of the qual- Shot point Phase Average pick uncertainty ity factor for the uppermost mantle are generally considered higher. The (s) generic value of Qp = 1000 is consistent with that shown by Fowler (2005) 1104 PmP 0.125 for a general Earth model. It should be noted that quality factors of 1000 Pn 0.15 or more are indicative of very limited attenuation. Moreover, changes in 1105 PmP 0.125 Pn 0.14 the values of Q by factors of one half or two do not materially affect the 1108 PmP 0.14 dynamic characteristics of the seismograms, as established by additional Pn 0.15 calculations using different Q values. For the wide-angle data, a zero-phase Ricker wavelet was used as the source signal with a bandwidth of 4–16 Hz to simulate an explosive source. For the Vibroseis data, which is characterized by higher frequen- three shots show high variability and no obvious distance at which a com- cies, a zero-phase Ricker wavelet with a 20 Hz central frequency of a two- mon amplitude peak is observed. octave wavelet (10–40 Hz) was used as the input source.

SYNTHETIC SEISMIC MODELING Synthetic Models

Methodology As represented by WAR and NVI data sets, the characteristic features of the observed wave fi eld of the Moho below the Great Bear magmatic Both near-vertical and wide-angle refl ection seismic data show very arc are: distinct Moho refl ections from beneath the Great Bear magmatic arc. The (1) prominent PmP phase between 80 and 220 km offsets, followed by dynamic characteristics of these refl ections contain valuable informa- ~0.5-s-long coda, and the absence of this phase at near-vertical incidence tion that can help us shed some light on the structure of the crust-mantle angles (Fig. 4); transition (Moho). In order to investigate wavefi eld characteristics (trav- (2) amplitude behavior of the PmP phase with offset that is highly eltime, duration, amplitude, frequency, etc.), we calculated the synthetic erratic and no common amplitude peak observed among the different seismic signature of different crust-mantle transition models using 1-D shots (Fig. 5A); and 2-D wave propagation algorithms. We examined a suite of models (3) variable dominant PmP frequency, measured on individual traces, ranging from simple to more complex transitions suggested by the many at different offsets between 10 and 13 Hz (however, no clear relation studies mentioned earlier. These include: fi rst-order discontinuities, gra- between amplitude spectra and offset is observed; Fig. 5B); dient transitions, layered transitions, and laterally heterogeneous transi- (4) a very weak amplitude Pn phase at offsets >160 km offset; and tions. The key parameters of these models, which control the seismic (5) a Moho signature on the NVI data characterized by a very distinct, signature of the synthetic seismograms, are the total thickness of the mutlicyclic band of refl ectors that are laterally variable and piecewise con- transition zone, the velocity contrast between the transition zone and the tinuous over several kilometers (Fig. 3). Amplitude spectra measured over surrounding material, and the correlation length of the lateral heteroge- 25 traces show a dominant frequency centered at ~21 Hz. neities. By modifying these parameters, synthetic seismograms that best We calculated the synthetic seismograms of 1-D and 2-D velocity simulate the real data can be obtained. To examine the seismic signa- models to determine the extent to which these models can replicate the ture of the fi rst-order discontinuity, gradient, and layered transitions, we wave-fi eld characteristics noted here. Both early studies of the proper- used a 1-D refl ectivity modeling algorithm (Fuchs and Müller, 1971). ties of the Moho (e.g., Clowes and Kanasewich, 1970) and more recent For the laterally heterogeneous models, a 2-D viscoelastic fi nite differ- ones (e.g., Carbonell et al., 2002) have demonstrated that representation ence algorithm (Robertsson et al., 1994) was required. The 1-D and 2-D of the crust-mantle transition by a fi rst-order discontinuity, or a linear velocity models that were incorporated into the 1-D refl ectivity and 2-D increase in velocity from lower-crustal to upper-mantle values over a viscoelastic modeling, respectively, were based on the interpretation of limited depth interval, does not generate synthetic seismograms that the R/WAR study along SNORCLE line 1 (Fernández-Viejo and Clowes, replicate observed wave-fi eld characteristics. Our studies of such Moho 2003). The crustal P-wave velocities increase from 5.9 km/s at the surface models, not discussed herein, further confi rmed these results. Accord- to 6.8 km/s at the crust-mantle boundary (~32 km depth). The velocity ingly, we initiate our discussion with layered transition models. of the uppermost mantle is 8.1 km/s and linearly increases to 8.4 km/s at 45 km depth. Corresponding S-wave velocities were calculated using 1. Layered Transition the ratios Vp/Vs = 1.78 and Vp/Vs = 1.82 for the crust and upper mantle, The crust-mantle transition is represented by alternating high- and respectively (Fernández-Viejo et al., 2005), whereas density values were low-velocity layers with a velocity contrast of 0.6 km/s (Fig. 6). Alter- based on laboratory measurements by Christensen and Mooney (1995) native models with different/variable velocity contrasts (0.2–0.8 km/s) and Christensen (1996). were also examined. Following Carbonell et al. (2002), these layered

The P- and S-wave quality factors, Qp and Qs, respectively, were more models are characterized by two parameters, the total thickness of the difﬁ cult to assign because no attenuation studies have been carried out in transition zone (H) and the thickness of the internal layers (h). Initial

the region. Values of Qp = 500 and Qs = 200 for the crust and Qp = 1000, modeling suggests that H controls the length of PmP coda, whereas the

Qs = 400 for the upper mantle were used for generating the synthetic seis- number of wiggles is controlled by h, as well as the velocity contrast, mograms displayed. These are “generic” values typically used within the similar to the results obtained by Carbonell et al. (2002). We kept the

algorithms by other scientists who have used it for crustal studies. The Qp total thickness of the transition zone ﬁ xed at H = 3 km, which produces value for the crust is consistent with values of 400 ± 200 determined for an ~0.5 s coda length, comparable to the observed data, and we var- the continental crust in England (Scheirer and Hobbs, 1990). In general ied the internal thickness (h) between 100 and 500 m. Layered models

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 383

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

3 h 56789 2.5 3 3.5 × 10 H 0 0 B 5 5 10 10

)

m 15 15

(

10 h 20 20

e 25 25

D 9 30 30 35 35

40 8 40 Velocity (km/s) Density (kg/m3)

PmP Time - X/8.0 (s) X/8.0 - Time 6 10 Pn 5 9

4 8 220 200 180 160 140 120 100

Offset (km) 7 PmP

Time - X/8.0 (s) X/8.0 - Time 6 Pn

5 Figure 6. Synthetic and observed seismograms for a layered transition. (A) Layered velocity model (H = 3 km, h = 500 m) for P-wave and corresponding densi- 4 ties. (B) Observed PmP and Pn phases (left) and corre- 220 200 180 160 140 120 100 sponding synthetic seismogram (right) for SP 1105. All seismograms are displayed in true relative amplitudes Offset (km) and were band-pass ﬁ ltered between 3 and 15 Hz.

produce prominent, reverberatory PmP phases beyond 80 km offset but The synthetic NVI seismograms feature a distinct Moho character- rather weak Pn phases at around 140 km offset (Fig. 6). In a test of two ized by sharp, mutlicyclic bands of refl ectors (4–9 bands) depending on models with the same H = 3 km, but one featuring h = 300 m and the the thickness (h) of the internal layers (Fig. 10). Higher velocity con- other with h = 500 m, the latter generates a phase characterized by a trasts between the internal layers result in higher amplitude refl ections more reverberatory pattern (Fig. 7). For very thin internal layers (h ≤ and vice versa. The dominant frequency of these refl ections, measured 100 m), the Moho signature is very similar to that generated by a fi rst- from averaging the amplitude spectrum over 25 traces and for different order discontinuity model. From models with varying velocity contrasts, layer thicknesses, ranges between 19 and 23 Hz, compared with the 20 we note that smaller values produce fewer numbers of wiggles. Hz central frequency of the source wavelet. This is due to the fi ne-tuning Amplitude-versus-offset (AVO) curves are relatively simple com- effect of the varying thicknesses of the internal layers (Fig. 11). We pared to the observed PmP amplitudes, regardless of the internal thick- notice that thicker layers produce more spiked-shape spectra, whereas ness of the layers. They are characterized by a clear peak amplitude layers <100 m thick produce a smoother spectrum similar to that of the located at ~120 km offset that corresponds, as expected, to the critical input source. distance (Fig. 8). Beyond 120 km, there is a nearly steady decrease in Previous results and those from our study indicate that the crust- amplitudes with offset. mantle transition is not a simple, uniform discontinuity in which veloci- The amplitude spectrum of the PmP phase was calculated on single ties increase suddenly or linearly from crustal to mantle values, but traces and over a 0.5 s window (here and in all subsequent models). The rather a complex and variable transition zone. One possible scenario spectrum features a varying dominant frequency between 7 and 12 Hz at that more closely replicates the observed data is a transition zone that different offsets. There is no correlation between the dominant frequency consists of alternating high- and low-velocity layers, as described in and varying offsets (Fig. 9A). Also, the internal thickness of the layers has the preceding based on 1-D modeling. However, a 1-D model does a limited effect on the shape of the spectrum (Fig. 9B). not generate the lateral variability that is observed, particularly on the

384 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

h=300 h=500 H H

A B

10 10

9 9

8 PmP 8 PmP

7 7

6 6

Time - X/8.0 (s) X/8.0 - Time 5 5

4 4 200 180 160 140 120 100 80 200 180 160 140 120 100 80 Offset (km) Offset (km) Figure 7. Two synthetic seismograms resulting from a layered velocity model with the same total thickness, H = 3 km, but a different internal layer thickness h. (A) Internal layer thickness h = 300 m. (B) Internal layer thickness h = 500 m. Notice the latter produces more reverberations in the PmP phase.

1.0

Figure 8. Comparison between 0.8 synthetic (solid line) and observed (dotted line) amplitude versus off- 0.6 set curves of the PmP phase. Syn- thetic amplitudes are for a layered 0.4 transition model with h = 300 m.

Amplitude Amplitude values are normalized 0.2 to unity. SP 1104 0.0 -160 -140 -120 -100 -80 Offset (km)

A 1.0 B 1.0 Offset (km) Thickness (m) 0.8 0.8 100 100 300 120 0.6 0.6 500 140

0.4 0.4

Amplitude Amplitude 0.2 0.2

0.0 0.0 0 5 10 15 20 0 5 10 15 20 Frequency (Hz) Frequency (Hz) Figure 9. Normalized amplitude spectra of synthetic PmP phase produced by layered discontinuity (h = 300 m). (A) The spectrum is calculated on single traces at 80, 100, 120, and 140 km offsets and over a 0.5 s window. (B) The amplitude spectrum is calculated at 120 km offset on single traces over a 0.5 s window and for different layer thicknesses, h = 100, 300, and 500 m.

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 385

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

AB10.5 10.5

11.0 11.0 Time (s) (s) Time

11.5 11.5 04Distance (km) 0 Distance (km) 4 Figure 10. (A) Synthetic near-vertical reﬂ ection data for a layered transition model with h = 300 m and (B) for a layered transition model with h = 500 m.

NVI reﬂ ection gathers and stacked sections (Figs. 2 and 3). In order to 2% standard deviation. The ellipses are characterized by their correlation

account for all the characteristics of the wave fi eld, a 2-D wave propa- length, which has a horizontal (ax) and vertical (az) component. For the gation algorithm is needed, one that can handle both vertical and lateral WAR data, we calculated the synthetic seismograms for three shots that velocity variations. simulate shots 1104, 1105, and 1108. In the case of NVI data, several synthetic seismograms simulating normal shot gathers are generated. For all 2. Heterogeneous Transitions cases, the velocity structure on either side of the region of the crust-mantle To study the effects of both lateral and vertical heterogeneity at the transition is derived from the refraction analysis of line 1 data (Fernández crust-mantle boundary, we investigated a range of velocity models that Viejo and Clowes, 2003). represent two end members. In one model, we use laterally discontinuous layers with variable thicknesses and velocities. This represents an anas- Laterally Discontinuous Layers tomosing layered structure with juxtaposed rocks from the lower crust The Moho transition consists of a complexly layered velocity struc- and upper mantle (e.g., Jarchow and Thompson, 1989). In the second ture over a 3-km-thick zone. Individual layers have thicknesses varying approach, the velocity model is assumed to be made up of two compo- between 100 and 500 m, and their horizontal extent ranges between 5 and nents (Mereu and Ojo, 1981): (1) a deterministic background component 20 km, with varying P-wave velocities of 6.9, 7.5, and 8.1 km/s (with 2% in which the velocity is increasing with depth and (2) a random velocity standard deviation), characteristic of lower-crust and upper-mantle values component consisting of a fi eld of small elliptical refl ectors embedded (Fig. 12A). in the background velocity fi eld. Random numbers are generated, which Synthetic seismograms generated by the viscoelastic fi nite-difference in turn are used to distribute the ellipses throughout the transition zone algorithm display a prominent PmP phase followed by an ~0.5-s-long and to assign velocity contrast values to each ellipse (Mereu, 2003) with coda, in agreement with the observed data (Figs. 12B and 4). Through a variety of tests, layers as short as 3 km long still produce a PmP phase with similar dynamic characteristics. On the other hand, an increase in the total thickness of the transition zone to 6 km results in a much longer coda 1.0 (~1.2 s), which is not observed. In general terms, the synthetic seismograms of Figure 12 compare favorably with the observed data of Figure 4,

0.8 50 m including variability of the PmP phase and a weak Pn phase beginning 200 m at ~170 km offset. The PmP amplitude analyses show complex behavior 300 m with offset in all three shots (Fig. 13). This resembles the amplitude analy- 500 m

e 0.6 sis results for the observed data (Fig. 5). Amplitudes were corrected for d

u ½

t geometrical spreading by multiplying all traces by r (r = source-receiver

i l

p distance), assuming a point source in 2-D as equivalent to a line source in 0.4 m three dimensions (3-D). A On the synthetic NVI data, the Moho signature features a strong,

0.2 multi cyclic band of reﬂ ectors. However, only models that have layers with shorter horizontal extent (3–7 km) display laterally variable and piecewise continuous reﬂ ections, similar to the observed data (Fig. 14). 0.0 0 1020304050Lamellae Structure Frequency (Hz) The Moho is represented by a 3-km-thick transition zone characterized by randomly distributed ellipses with correlation lengths of a = Figure 11. Normalized amplitude spectra of synthetic near-vertical x 1000 m and a = 100 m and velocities ranging between 7.1 and 7.8 km/s data produced by layered discontinuity. Each spectrum is aver- z aged over 25 traces. Complexity of amplitude spectra increases (Fig. 15A). This velocity structure produces a strong and reverberatory with increasing thickness of the layers. Layer thicknesses are 50, PmP phase characterized by ~0.5-s-long coda, comparable to the observed

200, 300, and 500 m. data (Fig. 15B). Increasing the dimensions of these ellipses (ax = 3000 m

386 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

SP 1108 SP 1105 A 0

40 VE≈2:1 0 100 200 Distance (km) 5.9 6.2 6.8 8.1 8.4 VE=1:1 Velocity (km/s) 0 10 km

B 10

7 PmP

Time - X/8.0 (s) X/8.0 - Time 6 Pn

5 Pg SP 1105

4 200 180 160 140 120 100 Offset (km)

PmP

7 ime - X/8.0 (s) X/8.0 - ime

T 6 Pn PiP

5 Pg SP 1108

4 100 120 140 160 180 200 220 Offset (km) Figure 12. Synthetic wide-angle seismograms for laterally discontinuous layers. (A) Two-dimensional (2-D) velocity model characterized by discontinuous layers with lengths from 5 to 20 km, thicknesses ranging between 100 and 300 m, and alternating velocities of 6.9, 7.5, and 8.1 km/s. (B) Synthetic seismograms simulating SP 1105 and SP 1108. Phases Pg, PmP, and Pn are explained in the text. PiP is a wide-angle phase from a velocity contrast at ~23 km. All seismograms are displayed in true relative amplitudes and were band-pass ﬁ ltered between 3 and 15 Hz. Compare with Figure 4. VE—vertical exaggeration.

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 387

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

1 1

0.8 0.8

0.6 0.6

0.4 0.4 Amplitude SP1105

SP1104 Amplitude 0.2 0.2 -160 -140 -120 -100 -80 -200 -180 -160 -140 -120 -100 -80

0.8 Figure 13. Amplitude versus offset of the synthetic PmP phase on shots 0.6 1104, 1105, and 1108. Amplitude values are normalized to unity. Ampli- tude shows erratic behavior, and no common peak is observed among 0.4 SP1108 the different shots. Amplitude 0.2 80 100 120 140 160 Offset (km)

Offset (km)

10 5 0 510 0.0

5.0

Time (s)

10.0

15.0 11.0

Time (s)

12.0 5 0 km

Figure 14. Synthetic near-vertical shot gather for laterally discontinuous layers. Inset shows Moho reﬂ ections that are multicyclic, laterally variable, and piecewise continuous over several kilometers.

388 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

SP 1105 SP 1104

A 0

)

(

e D

VE≈2:1

VE=1:1 Distance (km) 5.9 6.2 6.8 8.1 8.4

Velocity (km/s) 010 km B 10

7 PmP

Time - X/8.0 (s) X/8.0 - Time 6 Pn

5 Pg PiP SP 1104

4 180 160 140 120 100 80 Offset (km)

7 PmP Time - X/8.0 (s) 6 Pn

5 Pg SP 1105 4 200 180 160 140 120 100 Offset (km) Figure 15. Synthetic wide-angle seismograms for lamellae structure transition zone. (A) Two-

dimensional (2-D) velocity model with randomly distributed ellipses (correlation lengths ax =

1000 m and az = 100 m) and velocities ranging between 7.1 and 7.8 km/s. (B) Synthetic seismograms simulating SP 1104 and SP 1105. Phases Pg, PmP, and Pn are explained in the text. PiP is a wide-angle phase from a velocity contrast at ~23 km. All seismograms are displayed in true relative amplitudes and were band-pass ﬁ ltered between 3 and 15 Hz. VE—vertical exaggeration.

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 389

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

and az = 300 m) or replacing their varying velocities with one ﬁ xed value ~3.3-km-thick zone, which is in good agreement with the results from the at 7.2 km/s does not change the dynamic characteristic of the PmP phase synthetic WAR data.

substantially. On the other hand, when ax < 700 m, the synthetic seis- Only a laterally and vertically heterogeneous crust-mantle bound- mograms feature a simple PmP phase lacking a visible coda, similar to ary can simulate all the dynamic features observed on the wide-angle that generated by a ﬁ rst-order discontinuity. A velocity contrast as small and near-vertical data (Figs. 12 and 15). This transition consists of dis- as 0.1 km/s still produces a reverberatory PmP phase. Distributing the continuous layers/ellipses characterized by variable correlation lengths ellipses over a 6-km-thick zone produces a distinct PmP phase with a very and velocities. Synthetic models for both WAR and NVI data allow long coda (~1.5 s), whereas a 1-km-thick transition zone produces a PmP us to place lower and upper limits on the values for such correlation phase without a coda. Table 3 summarizes all the different combinations lengths and velocities. A minimum thickness of 100 m and a 0.2 km/s of parameters that were tested. All of the scenarios generate a weak Pn velocity contrast are needed for the layers/ellipses in order to reproduce phase observed at offsets >160 km. As in the previous model, the PmP the reverberatory pattern of the PmP phase. Models with layers/ellipses amplitude curves of all the different shots and for different correlation thicker than 600 m do not simulate the multicyclic band of reﬂ ections lengths vary considerably with offset (Fig. 16). observed on the NVI data. On the other hand, layers/ellipses that are at least 1 km but no more than 6–7 km long are required to generate a distinct PmP coda and to achieve the lateral variability seen on the NVI data. TABLE 3. DIFFERENT COMBINATION OF ALL THE PARAMETERS Calculated amplitude versus offset curves for the PmP phase from CHARACTERIZING THE LAMELLAE TRANSITION ZONE the layered 1-D velocity model are characterized by one prominent peak Velocity (km/s) Correlation length Thickness PmP characteristics amplitude at around 120 km. In contrast, amplitude curves calculated (km) (km) from the heterogeneous models have more complicated shapes with

7.1–7.8 ax = 3, az = 0.3 3 Strong and reverberatory phase, multiple high amplitudes, in agreement with the calculated amplitudes 7.1–7.8 a = 1, a = 0.1 3 coda length = ~0.5 s x z from the observed data (Figs. 8, 13, and 16). The deviation between 7.2 a = 3, a = 0.3 3 x z these curves and the ones calculated for simple 1-D velocity models 7.2 ax = 1, az = 0.1 3 8.0 a = 3, a = 0.3 3 is most probably the result of constructive and destructive interference x z effects due to the lateral heterogeneity as well as the vertical complexity 8.0 ax = 1, az = 0.1 3

7.1–7.8 ax = 3, az = 0.3 6 Strong and reverberatory phase, of the crust-mantle boundary. coda length = ~1.5 s 7.1–7.8 ax = 1, az = 0.1 6 Based on the spectral analysis, the dominant frequency of the PmP 7.2 a = 3, a = 0.3 6 x z phase varies randomly between 7 and 12 Hz at different offsets. Thus, 7.2 a = 1, a = 0.1 6 x z no relation can be inferred between frequency and offset. In addition, the 7.1–7.8 ax = 0.5, az = 0.1 3 Strong phase lacking visible 7.2 a = 0.5, a = 0.1 3 coda wide-angle data are less sensitive to the ﬁ ne structure of the crust-mantle x z boundary (e.g., internal thickness of the layers) than the near-vertical 7.1–7.8 ax = 3, az = 0.3 1

7.2 ax = 3, az = 0.3 1 data (Figs. 9B and 11). Unfortunately, spectrum analysis of the reﬂ ection waveforms is insufﬁ cient to distinguish between different heterogeneous models.

Multigenetic Origin for the Continental Moho On the NVI data, the Moho resulting from this lamellae structure is imaged as a narrow (0.4–0.7 s two-way traveltime [TWTT]) band of 4–8 The Wopmay orogen is an amalgamation of Paleoproterozoic domains cycles of refl ections (Fig. 17), which is comparable to the observed data that developed and accreted onto the Slave craton between 2.1 Ga and (Fig. 3). These refl ections are laterally variable and piecewise continuous 1.84 Ga by complex geologic processes (Hoffman, 1989). Thus, in this over ~10 km distance. Our results show that increasing the total thickness region, refl ection patterns near the crust-mantle boundary are expected to of the transition zone from 3 to 6 km would result in a much longer (~0.9 s be similarly complex (e.g., Clowes et al., 1987). However, observations TWTT) band of 12–13 cycles of refl ections. Different correlation lengths along SNORCLE line 1 reveal simple Moho refl ections characterized by

change the number of cycles, and even with values as small as ax = 300 m subhorizontal surfaces and relatively uniform arrival time (~11.0 s) over

and az = 30 m, distinct bands of refl ections can still be detected, attesting the entire profi le, even beneath regions with different ages such as the to the higher resolution of the NVI data. However, a minimum velocity Archean Slave Province and the Paleoproterozoic Wopmay orogen (Cook contrast of 0.2 km/s is needed between the ellipses and the surrounding et al., 1999). This uniformity implies that postaccretion, regionally exten- material in order to produce visible refl ections. sive processes have effectively modifi ed rocks and reset the refl ection Moho to a uniform depth (~33 km). Therefore, a complete interpretation DISCUSSION for the development of the Moho must consider all possible thermal and/ or mechanical processes in order to account for its formation and evo- Synthetic Models lution from initial complexities during accretion to increasing coherency over time (Cook, 2002). From the results of synthetic modeling and based on the duration of Several hypotheses have been proposed for the formation and develop- the PmP coda, a 3-km-thick Moho is required to generate 0.5-s-long coda. ment of the continental Moho that can be classifi ed under either igneous or Thicknesses greater than 5 km or less than 2 km produce either a lon- nonigneous origins (Eaton, 2006). Igneous origins include thermal front ger coda (~1.0 s) or a PmP phase lacking any visible coda, respectively. (e.g., partial melt and magmatic intrusion) and magmatic underplating, Furthermore, observed NVI data at Moho levels feature a distinct band whereas nonigneous origins include metamorphic or metasomatic front of refl ections over 0.45 s traveltime (0.9 s TWTT). Considering average (e.g., mafi c granulite to eclogite) and regional décollement (Eaton, 2006; seismic velocity at this level to be ~7.4 km/s, this band translates to an Cook et al., 2010).

390 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

A B

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

Amplitude 0.2 0.2 SP 1104 SP 1104 0.0 0.0 -160 -140 -120 -100 -80 -160 -140 -120 -100 -80

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

Amplitude 0.2 0.2 SP 1105 SP 1105 0.0 0.0 -200 -180 -160 -140 -120 -100 -80 -200 -180 -160 -140 -120 -100 -80

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

Amplitude 0.2 0.2 SP 1108 SP 1108 0.0 0.0 80 100 120 140 160 80 100 120 140 160

Offset (km) Offset (km)

Figure 16. Amplitude versus offset of the synthetic PmP arrivals on shots 1104, 1105, and 1108 produced by the lamellae struc-

ture. Amplitude values are normalized to unity. (A) ax = 1000 m, az = 100 m. (B) ax = 3000 m, az = 300 m.

The spatial correlation between the Moho and crustal refl ections Regional Décollement provides important constraints on the origin of the crust-mantle bound- Many deep seismic profi les reveal lower-crustal refl ections that are lis- ary. For instance, lower-crustal refl ections that are listric into the Moho tric into the Moho (Cook et al., 1992, 1999; Calvert et al., 1995; Kukkonen suggest that the crust-mantle boundary is most probably a regional et al., 2008). This suggests that the transition zone near the crust-mantle décollement (Cook and Vasudevan, 2003). The transition zone acts as boundary behaves as a mechanically weak layer (more plastic) into which a rheologically weak layer and is underlain by a stronger layer. On the upper-crustal detachments sole (Eaton, 2006). This process could be asso- other hand, truncation of lower-crustal refl ections at Moho levels could ciated with either regional extension and/or contraction (Cook, 2002). be caused by thermal or metamorphic fronts (e.g., Carbonell et al., 2002; Within the lithosphere, the strength of minerals decreases consider- Cook et al., 2010). Interestingly, seismic data from the lower crust of the ably from olivine, amphibole, and garnet to feldspar and quartz, the lat- Great Bear magmatic arc, part of SNORCLE line 1, exhibit both types of ter being the weakest mineral (Austrheim et al., 1997). The lower crust spatial correlation between the Moho and lower-crustal refl ections near of the Hottah terrane consists primarily of amphibolite and paragranulite 10–11 s (Fig. 2). Thus, in our synthetic models, the lamellae structure (Fernández-Viejo et al., 2005). Consequently the rheology is presumably with randomly distributed ellipses represents a thermal/metamorphic controlled by amphibole and granulite, whereas olivine controls the rheol- front, whereas the laterally discontinuous layering represents a regional ogy of the mantle. Therefore, at the crust-mantle boundary, one expects décollement. a distinct strength contrast that may tend to localize detachment surfaces

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 391

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

A 10.5

11.0 Time (s)

11.5 0 5 10 Distance (km)

B 10.5

11.0 Time (s)

11.5 0 5 10 Distance (km)

C 10.5

11.0 Time (s)

11.5 0 5 10 Distance (km)

Figure 17. Synthetic near-vertical seismograms for a lamellae structure transition zone. (A) Correlation length ax = 1000 m and

az = 100 m. (B) ax = 3000, az = 300 m. (C) ax = 300 m, az = 30 m.

(Cook, 2002; Cook et al., 2010). Based on published geotherms and litho- Moreover, improved seismic images of the same profi le (fi g. 3d in Cook spheric composition, Meissner and Mooney (1998) estimated the depth of and Vasudevan, 2003) show distinct upper-mantle structures, which were low-viscosity (weak) zones within the crust and upper mantle along which interpreted as a synform fabric overlain with angular discordance by the decoupling may occur. They found these weak zones at three depths: (1) fl at Moho. Such observations support the interpretation of the Moho as a the base of the felsic upper crust; (2) just above the Moho; and (3) some structural boundary along which lower-crustal and upper-mantle rocks are tens of kilometers below the Moho. displaced, sheared, and blended. However, as Cook and Vasudevan (2003) In the lower crust beneath the Great Bear magmatic arc, a series of explained, the Moho here cannot be regarded only as a regional décolle- discrete, northeast-dipping refl ections (Fig. 2) is listric into the Moho. ment for the following two reasons: (1) the crust is much more refl ective

392 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

than the upper mantle, and with a simple detachment, the rocks above the lenses ranging in size from decimeters to several hundreds of meters in detachment surface should be equivalent to, but dislocated from, rocks length embedded in amphibolites (Austrheim, 1987). This is very similar below; and (2) a detachment surface does not explain how rocks below the to the lamellae structure we tested in our synthetic models. Although in the Moho have signiﬁ cantly higher seismic velocities (~8.1 km/s; Fernández Western Gneiss region, most eclogites are mylonitic, examples of partial Viejo and Clowes, 2003) than those above it (~6.8 km/s). eclogitization are abundant. Therefore, in our model, the wide range of From the preceding discussion, consideration of the Moho as a velocities/densities characterizing the randomly distributed ellipses could regional décollement represents only one stage of its development. A be explained by different degrees of eclogitization, depending on the rock logical sequence of stages or events is probably necessary to explain the composition and availability of water. varying characteristics of the crust-mantle boundary observed on both NVI and WAR data. CONCLUSIONS

Thermal Front (e.g., Magmatic Intrusion and Partial Melting) Observed wide-angle and near-vertical seismic-refl ection data and 1-D Mafi c igneous rocks may be injected into the lower crust and upper and 2-D synthetic seismogram modeling for the crust-mantle boundary mantle in the form of sills during the amalgamation of island arcs at col- offer different but complementary information about its nature and forma- lisional orogens (e.g., Wopmay orogen). These sills are characterized tion. A unifi ed interpretation, in which both wide-angle and near-vertical by relatively high density and high seismic velocity (>7.2 km/s). The data sets derived from the same Moho location, can provide a more com- intermediate-composition crust above acts as a density fi lter for these plete explanation concerning the structure of the Moho, one with higher intrusives, and hence they pond within the lower crust and along the resolution and perhaps less ambiguity. crust-mantle boundary, producing seismically laminated lower crust and a Our synthetic models lead to the conclusion that beneath the Great sharply defi ned Moho (Furlong and Fountain, 1986; Nelson, 1991). Bear magmatic arc, an ~3-km-thick, laterally and vertically heterogeneous Partial melting can modify the chemical and physical properties of the crust-mantle boundary properly simulates the dynamic characteristics of intermediate-composition rocks at elevated lower-crustal temperatures the observed wave fi elds recorded on the near-vertical and wide-angle and pressures by preferentially removing lower-density minerals such as refl ection data. The heterogeneity can be achieved by either laterally quartz and feldspar. This will leave mafi c residuals with higher density discontinuous layering or a lamellae structure with randomly distributed and higher seismic velocity (>7.5–7.8 km/s) appropriate for the upper ellipses. These two models may represent different re-equilibration mech- mantle (Hynes and Snyder, 1995; Rudnick and Fountain, 1995). The soli- anisms, one characterized by thermal/metamorphic fronts and another that dus temperature for these restites is considerably elevated, and any further features a regional décollement. Due to insuffi cient details on the seismic- partial melting requires substantially higher temperatures. As a result, the refl ection image, it is diffi cult to establish the exact geometric relation- Moho could be “frozen” unless signifi cantly higher temperatures occur ships between the Moho and lower-crustal layers and consequently the (Cook 2002). Because partial melting might not remove all of the lighter re-equilibration process. However, assuming that our images of the Moho fraction from the lower crust, structural geometry near the Moho can be beneath the Great Bear magmatic arc represent “snapshots” of the crust- preserved (Cook et al., 2010). mantle boundary “taken” at different stages of its evolution (Keller et al., As mentioned previously, termination or truncation of some of the 2005), we believe that the Moho has formed as a combination of the afore- dipping structures in the lower crust of the Great Bear magmatic arc by mentioned processes. the Moho, in addition to the discordance between the underlying synform Here, we propose a scenario for the evolution of the Moho (Fig. 18) fabric and the Moho (Cook and Vasudevan, 2003), strongly suggests that that builds on, but differs somewhat from, a previous study (Cook and the Moho postdates the synform and lower-crustal structures. Thus, it may Vasudevan, 2003) as follows: (1) Prior to 1.9 Ga, the Moho represented have been superimposed on them by thermal processes. The approach and the base of the Hottah terrane crust, which was generated as a magmatic subsequent subduction of oceanic lithosphere beneath the Hottah terrane arc on cryptic 2.4–2.0 Ga crust (Hoffman and Bowring, 1984). The depth was probably the source of thermal activity responsible for partial melting to the Moho is unknown but was perhaps at ~35 km. (2) The collision of and magmatic intrusion in the study area, which in turn was associated the Hottah terrane onto the Slave craton caused substantial compression with the deposition of the Great Bear magmatic rocks between ca. 1.88– and led to the formation of the listric structures that fl atten into the Moho. 1.84 Ga (Cook et al., 1999; Oueity and Clowes, 2010). This implies that the Moho at this stage acted as a ductile decoupling zone or a detachment surface. (3) Between 1.88 and 1.84 Ga, the Great Bear Metamorphic Front (Eclogite Phase Transition) magmatic arc was formed as a result of east-dipping subduction of oceanic Since the work of Green and Ringwood (1967) and Ito and Ken- lithosphere beneath the Hottah terrane and the Slave craton generated by nedy (1971), it has been recognized that under relatively high pressure- incipient of the Fort Simpson terrane. The thermal activity associated with temperature (P-T) conditions, mafi c rocks (e.g., mafi c granulite, gab- the previous subduction produced partial melt and magmatic intrusions bros) might transform into eclogite at depths of ~35 km. Hence, the at the base of the crust. A horizontal Moho was developed as a thermal/ Moho could be interpreted as a metamorphic front associated with a metamorphic boundary. phase change. Eclogitization results in an increase in the density and seismic velocity of the lower crust such that it becomes indiscernible ACKNOWLEDGMENTS from the uppermost mantle (e.g., Austrheim, 1987; Mengel and Kern, 1992; Fischer, 2002; Kukkonen et al., 2008). Consequently, lower- We wish to thank Phil Hammer for stimulating discussions and construc- crustal eclogitization has been invoked to explain the observation of a tive comments. We thank the editor for comments that helped clarify some relatively shallow Moho at several places (e.g., the Scottish Caledonides parts of the text. The data were acquired as part of Lithoprobe, which was [Hynes and Snyder, 1995] and the Appalachians off northeastern New- funded through a Research Networks grant from the Natural Sciences and foundland [Chian et al., 1998]). Engineering Research Council of Canada (NSERC) and the Geological Along the southwestern coast of Norway, eclogites are exposed over Survey of Canada. This study was supported by an NSERC Discovery a large area in the Western Gneiss region (WGR). They form pods and grant (7707-2009) to Clowes.

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 393

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

Figure 18. Schematic evolutionary model for the development of the Moho visible beneath the Great Bear arc along Lithoprobe’s Slave–Northern Cordillera Lithospheric Evolution (SNORCLE) line 1. (A) Formation of the Hottah terrane as a magmatic arc west of the Slave craton. Pre–1.9 Ga tectonism may have produced the internal structure within the lower crust. (B) Accretion of the Hottah terrane onto the Slave craton as a result of west-dipping subduction. The collision caused folding and faulting within the crust, and the Moho acted as a detachment surface. This probably led to the formation of the listric fabric in the lower crust. (C) Formation and deposition of the Great Bear magmatic arc on both the Hottah terrane and the Coronation Supergroup as a result of west-dipping subduction. The discordance between the arc rocks and the fold structures below implies that the crustal deformation, and consequently Moho detachment, occurred prior to arc mag- matism (Cook and Vasudevan, 2003). We suggest that partial melt and magmatic intrusive rocks at the base of the crust, produced by this thermal activity, are responsible for the Moho development as a thermal/metamorphic front.

394 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 Nature of the Moho transition in NW Canada | RESEARCH

REFERENCES CITED Ellis, R., Clowes, R.M., Spence, G.D., Asudeh, I., Hajnal, Z., and Amor, J.R., 2002, Slave– Northern Cordillera Refraction Experiment (SNORE’97): Field Acquisition and Prelimi- nary Data Processing Report: Vancouver, University of British Columbia, Lithoprobe Abbott, G., Thorkelson, D., Creaser, R., Bevier, M.L., and Mortensen, J., 1997, New correla- Report 83, 14 p., 10 apps. tions among Proterozoic successions and intrusive breccias in the Ogilvie and Wer- Fernández Viejo, G., and Clowes, R.M., 2003, Lithospheric structure beneath the Archean necke Mountains, Yukon, in Cook, F.A., and Erdmer, P., compilers, SNORCLE Transect Slave Province and Proterozoic Wopmay orogen, northwestern Canada, from a Litho- and Cordilleran Tectonics Workshop Meeting, University of Calgary, Calgary, 7–9 March probe refraction/wide-angle refl ection survey: Geophysical Journal International, 1997: Vancouver, Canada, Lithoprobe Secretariat, The University of British Columbia, v. 153, p. 1–19, doi: 10.1046/j.1365-246X.2003.01807.x. Lithoprobe Report 56, p. 188–197. Fernández Viejo, G., Clowes, R.M., and Amor, J.R., 1999, Imaging the lithospheric mantle in Austrheim, H., 1987, Eclogitization of lower crustal granulites by fl uid migration through northwestern Canada with seismic wide-angle refl ections: Geophysical Research Let- shear zones: Lithos, v. 27, p. 29–42. ters, v. 26, p. 2809–2812, doi: 10.1029/1999GL005373. Austrheim, H., Erambert, M., and Engvik, A.K., 1997, Processing of crust in the root of the Fernández Viejo, G., Clowes, R.M., and Welford, J.K., 2005, Constraints on the composition Caledonian continental collision zone: The role of eclogitization: Tectonophysics, v. 273, of the crust and uppermost mantle in northwestern Canada: Vp/Vs variations along p. 129–153, doi: 10.1016/S0040-1951(96)00291-0. Lithoprobe’s SNorCLE transect: Canadian Journal of Earth Sciences, v. 42, p. 1205– Boudier, F., and Nicolas, A., 1995, Nature of the Moho transition zone in the Oman ophiolite: 1222, doi: 10.1139/e05-028. Journal of Petrology, v. 36, p. 777–796. Fischer, K.M., 2002, Waning buoyancy in the crustal roots of old mountains: Nature, v. 417, Braile, L.W., and Chiang, C.S., 1986, The continental Mohorovicˇic´ discontinuity results from p. 933–936, doi: 10.1038/nature00855. near-vertical and wide-angle seismic refl ection studies, in Barazangi, M., and Brown, Fountain, D.M., 1986, Implications of deep crustal evolution for seismic refl ection interpre- L., eds., Refl ection Seismology: A Global Perspective: Washington, D.C., American tations, in Barazangi, M., and Brown, L., eds., Refl ection Seismology: A Global Per- Geophysical Union Geodynamic Series, v. 13, p. 257–272. spective: Washington, D.C., American Geophysical Union Geodynamic Series, v. 14, Calvert, A.J., Sawyer, E.W., Davis, W.J., and Ludden, J.N., 1995, Archean subduction inferred p. 95–106. from seismic images of a mantle suture in the Superior Province: Nature, v. 375, Fountain, D.M., and Salisbury, M.H., 1981, Exposed cross-sections through the continental p. 670–674, doi:10.1038/375670a0. crust: Implications for crustal structure, petrology and evolution: Earth and Planetary Candès, E., Demanet, L., Donoho, D., and Ying, L., 2006, Fast discrete curvelet transforms: Science Letters, v. 56, p. 263–277, doi: 10.1016/0012-821X(81)90133-3. Multiscale Modeling and Simulation, v. 5, p. 861–899, doi: 10.1137/05064182X. Fowler, C.M.R., 2005, The Solid Earth: An Introduction to Global Geophysics: Cambridge, UK, Carbonell, R., Gallart, J., and Perez-Estaun, A., 2002, Modeling and imaging the Moho transi- Cambridge University Press, 685 p. tion: The case of the southern Urals: Geophysical Journal International, v. 149, p. 134– Fuchs, K., and Müller, G., 1971, Computation of synthetic seismograms with the refl ectivity 148, doi: 10.1046/j.1365-246X.2002.01623.x. method and comparison with observations: Geophysical Journal of the Royal Astro- Carpentier, S., and Roy-Chowdhury, K., 2007, Underestimation of scale lengths in stochastic nomical Society, v. 23, p. 417–433. fi elds and their seismic response: A quantifi cation exercise: Geophysical Journal Inter- Furlong, K.P., and Fountain, D.M., 1986, Continental crust underplating: Thermal consider- national, v. 169, p. 547–562, doi: 10.1111/j.1365-246X.2007.03333.x. ations and seismic-petrologic consequences: Journal of Geophysical Research, v. 91, Casey, J.F., Dewey, J.F., Fox, P.J., Karson, J.A., and Rosencrantz, E., 1981, Heterogeneous nature of p. 8285–8294, doi: 10.1029/JB091iB08p08285. the oceanic crust and upper mantle: A perspective from the Bay of Islands ophiolite com- Gandhi, S.S., Mortensen, J.K., Prasad, N., and van Breemen, O., 2001, Magmatic evolution plex, in Emiliani, C., ed., The Sea, Volume 7: New York, John Wiley and Sons, p. 305–338. of the southern Great Bear continental arc, northwestern Canadian Shield: Geochrono- Chian, D., Marillier, F., Hall, J., and Quinlan, G., 1998, An improved velocity model for the logical constraints: Canadian Journal of Earth Sciences, v. 38, p. 767–785, doi: 10.1139/ crust and upper mantle along the central mobile belt of the Newfoundland Appala- cjes-38-5-767. chian orogen and its offshore extension: Canadian Journal of Earth Sciences, v. 35, Gibbs, A.K., 1986, Seismic refl ection profi les of Precambrian crust: A qualitative assessment, p. 1238–1251, doi: 10.1139/cjes-35-11-1238. in Barazangi, M., and Brown, L., eds., Refl ection Seismology: A Global Perspective: Christensen, N.I., 1996, Poisson’s ratio and crustal seismology: Journal of Geophysical Washington, D.C., American Geophysical Union Geodynamic Series, v. 14, p. 95–106. Research, v. 101, p. 3139–3156, doi: 10.1029/95JB03446. Green, D.H., and Ringwood, A.E., 1967, An experimental investigation of the gabbro to eclog- Christensen, N.I., and Mooney, W.D., 1995, Seismic velocity structure and composition of the ite transition and its petrological applications: Geochimica et Cosmochimica Acta, continental crust: A global view: Journal of Geophysical Research, v. 100, p. 9761–9788, v. 31, p. 767–833, doi: 10.1016/S0016-7037(67)80031-0. doi: 10.1029/95JB00259. Hale, L.D., and Thompson, G.A., 1982, The seismic refl ection character of the continental Clowes, R.M., and Kanasewich, E.R., 1970, Seismic attenuation and the nature of refl ecting Mohorovicˇic´ discontinuity: Journal of Geophysical Research, v. 87, p. 4625–4635, doi: horizons within the crust: Journal of Geophysical Research, v. 75, p. 6693–6705, doi: 10.1029/JB087iB06p04625. 10.1029/JB075i032p06693. Hammer, P.T.C., and Clowes, R.M., 1997, Moho refl ectivity patterns—A comparison of Cana- Clowes, R.M., Brandon, M.T., Green, A.G., Yorath, C.J., Sutherland-Brown, A., Kanasewich, dian Lithoprobe transects: Tectonophysics, v. 269, p. 179–198, doi: 10.1016/S0040- E.R., and Spencer, C., 1987, Lithoprobe—Southern Vancouver Island: Cenozoic subduc- 1951(96)00164-3. tion complex imaged by deep seismic refl ections: Canadian Journal of Earth Sciences, Hasselgren, E.O., and Clowes, R.M., 1995, Crustal structure of northern Juan de Fuca plate v. 24, p. 31–51. from multichannel refl ection data: Journal of Geophysical Research, v. 100, p. 6469– Coles, R.L., Haines, G.V., and Hannaford, W., 1976, Large-scale magnetic anomalies over 6486, doi: 10.1029/94JB02941. western Canada and the Arctic: A discussion: Canadian Journal of Earth Sciences, Hennenfent, G., and Herrmann, F.J., 2006, Seismic denoising with nonuniformly sam- v. 13, p. 790–802, doi: 10.1139/e76-082. pled curvelets: Computing in Science & Engineering, v. 8, p. 16–25, doi: 10.1109/ Cook, F.A., 2002, Fine structure of the continental refl ection Moho: Geological Society of Amer- MCSE.2006.49. ica Bulletin, v. 114, p. 64–79, doi: 10.1130/0016-7606(2002)114<0064:FSOTCR>2.0.CO;2. Herrmann, F.J., and Hennenfent, G., 2008, Non-parametric seismic data recovery with curve- Cook, F.A., and Erdmer, P., 2005, An 1800 km cross section of the lithosphere through the let frames: Geophysical Journal International, v. 173, p. 233–248, doi: 10.1111/j.1365- northwestern North American plate: Lessons from 4.0 billion years of Earth’s history: 246X.2007.03698.x. Canadian Journal of Earth Sciences, v. 42, p. 1295–1311, doi: 10.1139/e04-106. Hildebrand, R.S., and Bowring, S.A., 1999, Crustal recycling by slab failure: Geology, v. 27, Cook, F.A., and Vasudevan, K., 2003, Are there relict crustal fragments beneath the Moho?: p. 11–14, doi: 10.1130/0091-7613(1999)027<0011:CRBSF>2.3.CO;2. Tectonics, v. 22, 1026, doi: 10.1029/2001TC001341. Hildebrand, R.S., Hoffman, P.F., and Bowring, S.A., 1987, Tectono-magmatic evolution Cook, F.A., Varsek, J.L., Clowes, R.M., Kanasewich, E.R., Spencer, C.S., Parrish, R.R., Brown, of the 1.9-Ga Great Bear magmatic zone, Wopmay orogen, northwestern Canada: R.L., Carr, S.D., Johnson, B.J., and Price, R.A., 1992, Lithoprobe crustal refl ection cross Journal of Volcanology and Geothermal Research, v. 32, p. 99–118, doi: 10.1016/0377- section of the southern Canadian Cordillera: I. Foreland thrust and fold belt to Fraser 0273(87)90039-4. River fault: Tectonics, v. 11, p. 12–35, doi: 10.1029/91TC02332. Hoffman, P.F., 1988, United plates of America, the birth of a craton: Early Proterozoic assem- Cook, F.A., van der Velden, A.J., and Hall, K.W., 1999, Frozen subduction in Canada’s North- bly and growth of Laurentia: Annual Review of Earth and Planetary Sciences, v. 16, west Territories: Lithoprobe deep lithospheric refl ection profi ling of the western Cana- p. 543–603, doi: 10.1146/annurev.ea.16.050188.002551. dian Shield: Tectonics, v. 18, p. 1–24, doi: 10.1029/1998TC900016. Hoffman, P.F., 1989, Precambrian geology and tectonic history of North America, in Bally, Cook, F.A., White, D.J., Jones, A.G., Eaton, D.W., Hall, J., and Clowes, R.M., 2010, How the A.W., and Palmer, A.R., eds., The Geology of North America: An Overview: Boulder, Col- crust meets the mantle: Lithoprobe perspectives on the Mohorovicˇic´ discontinuity orado, Geological Society of America, The Geology of North America, v. A, p. 447–512. and crust-mantle transition: Canadian Journal of Earth Sciences, v. 47, p. 315–351, doi: Hoffman, P.F., and Bowring, S.A., 1984, Short-lived 1.9 Ga continental margin and its 10.1139/E09-076. destruction, Wopmay orogen, NW Canada: Geology, v. 12, p. 68–72, doi: 10.1130/ Debari, S., Kay, S.M., and Kay, R.W., 1987, Ultramafi c xenoliths from Adagdak Volcano, Adak, 0091-7613(1984)12<68:SGCMAI>2.0.CO;2. Aleutian Islands, Alaska: Deformed igneous cumulates from the Moho of an island arc: Hurich, C., 2003, The nature of crustal seismic heterogeneity: A case study from the Grenville The Journal of Geology, v. 95, p. 329–341, doi: 10.1086/629133. Province, in Goff, J.A., and Holliger, K., eds., Heterogeneity in the Crust and Upper Dilek, Y., Furnes, H., and Shallo, M., 2008, Geochemistry of the Jurassic Mirdita ophiolite Mantle: Nature, Scaling, and Seismic Properties: New York, Kluwer Academic Publish- (Albania) and the MORB to SSZ evolution of a marginal basin oceanic crust: Lithos, ers, p. 299–320. v. 100, p. 174–209, doi: 10.1016/j.lithos.2007.06.026. Hynes, A., and Snyder, D.B., 1995, Deep-crustal assemblages and potential for crustal rocks Eaton, D.W., 2006, Multi-genetic origin of the continental Moho: Insights from Lithoprobe: below the Moho in the Scottish Caledonides: Geophysical Journal International, v. 123, Terra Nova, v. 18, p. 34–43, doi:10.1111/j.1365-3121.2005.00657.x. p. 323–339, doi: 10.1111/j.1365-246X.1995.tb06857.x. Elad, M.J., Starck, L., Querre, P., and Donoho, D.L., 2005, Simultaneous cartoon and texture Ito, K., and Kennedy, G.C., 1971, An experimental study of the basalt-garnet granulite-eclog- image in painting using morphological component analysis (MCA): Applied and Com- ite transition, in Heacock, J.G., ed., The Earth beneath the Continents: American Geo- putational Harmonic Analysis, v. 19, p. 340–358, doi: 10.1016/j.acha.2005.03.005. physical Union Geophysical Monograph 14, p. 303–314.

LITHOSPHERE | Volume 2 | Number 5 | www.gsapubs.org 395

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021 OUEITY ET AL.

Jarchow, C.M., and Thompson, G.A., 1989, The nature of the Mohorovicic discontinuity: Nedimović, M.R., Carbotte, S.M., Harding, A.J., Detrick, R.S., Canales, J.P., Diebold, J.B., Annual Review of Earth and Planetary Sciences, v. 17, p. 475–506, doi: 10.1146/annurev Kent, G.M., Tischer, M., and Babcock, J.M., 2005, Frozen magma lenses below the oce- .ea.17.050189.002355. anic crust: Nature, v. 436, p. 1149–1152, doi: 10.1038/nature03944. Jousselin, D., and Nicolas, A., 2000, The Moho transition zone in the Oman Ophiolite; relation Neelamani, R., Baumstein, A.I., Gillard, D.G., Hadidi, M.T., and Soroka, W.L., 2008, Coher- with wehrlites in the crust and dunites in the mantle: Marine Geophysical Researches, ent and random noise attenuation using the curvelet transform: Leading Edge (Tulsa, v. 21, p. 229–241, doi: 10.1023/A:1026733019682. Oklahoma), v. 27, p. 240–248, doi: 10.1190/1.2840373. Karson, J.A., Collins, J.A., and Casey, J.F., 1984, Geologic and seismic velocity structure of Nelson, K.D., 1991, A unifi ed view of craton evolution motivated by recent deep seismic the crust/mantle transition in the Bay of Islands ophiolite complex: Journal of Geo- refl ection and refraction results: Geophysical Journal International, v. 105, p. 25–35, physical Research, v. 89, p. 6126–6138, doi: 10.1029/JB089iB07p06126. doi: 10.1111/j.1365-246X.1991.tb03441.x. Keller, G.R., Karlstrom, K.E., Williams, M.L., Miller, K.C., Andronicos, C., Levander, A.R., Snel- Nielsen, L., and Thybo, H., 2006, Identifi cation of crustal and upper mantle heterogeneity by son, C.M., and Prodehl, C., 2005, The dynamic nature of the continental crust-mantle modeling of controlled-source seismic data: Tectonophysics, v. 416, p. 209–228, doi: boundary: Crustal evolution in the southern Rocky Mountains as an example, in Karl- 10.1016/j.tecto.2005.11.020. strom, K.E., and Keller, G.R., eds., The Rocky Mountain Region: An Evolving Litho- Oueity, J., and Clowes, R.M., 2010, Paleoproterozoic subduction in northwestern Canada sphere: American Geophysical Union Geophysical Monograph 154, p. 403–420. from near-vertical and wide-angle seismic refl ection data: Canadian Journal of Earth Kopylova, M.G., Russell, J.K., and Cookenboo, H., 1998, Upper-mantle stratigraphy of the Sciences, v. 47, p. 35–52, doi: 10.1139/E09-073. Slave craton, Canada: Insights into a new kimberlites province: Geology, v. 26, p. 315– Panea, I., Stephenson, R., Knapp, C., Mocanu, V., Drijkoningen, G., Matenco, L., Knapp, J., 318, doi: 10.1130/0091-7613(1998)026<0315:UMSOTS>2.3.CO;2. and Prodehl, K., 2005, Near-vertical seismic refl ection image using a novel acquisi- Kukkonen, I.T., Kuusisto, M., Lehtonen, M., and Peltonen, P., 2008, Delamination of eclogi- tion technique across the Vrancea zone and Foscani Basin, south-eastern Carpathians tized lower crust: Control on the crust-mantle boundary in the central Fennoscandian (Romania): Tectonophysics, v. 410, p. 293–309, doi: 10.1016/j.tecto.2005.01.009. Shield: Tectonophysics, v. 457, p. 111–127, doi: 10.1016/j.tecto.2008.04.029. Robertsson, J., Balnch, J., and Symes, W., 1994, Viscoelastic fi nite-difference modeling: Geo- Kumar, V., 2009, Incoherent Noise Suppression and Deconvolution using Curvelet-Domain physics, v. 59, p. 1444–1456, doi: 10.1190/1.1443701. Sparsity [Master’s thesis]: Vancouver, University of British Columbia, 58 p. Rudnick, R.L., and Fountain, D.M., 1995, Nature and composition of the continental crust: A lower Levander, A.R., and Holliger, K., 1992, Small scale heterogeneity and large scale velocity crustal perspective: Reviews of Geophysics, v. 33, p. 267–309, doi: 10.1029/95RG01302. structure of the continental crust: Journal of Geophysical Research, v. 97, p. 8797–8804, Scheirer, D.S., and Hobbs, R.W., 1990, Seismic attenuation in the continental crust SW of Eng- doi: 10.1029/92JB00659. land: Geophysical Journal International, v. 103, p. 533–540, doi: 10.1111/j.1365-246X.1990 McGetchin, R.R., and Silver, L.T., 1972, A crustal-upper mantle model for the Colorado Pla- .tb01789.x. teau based on observations of crystalline rock fragments in the Moses Rock dike: Jour- Spence, G.D., Whittall, K.P., and Clowes, R.M., 1984, Practical synthetic seismograms for lat- nal of Geophysical Research, v. 77, p. 7022–7037, doi: 10.1029/JB077i035p07022. erally varying media calculated by asymptotic ray theory: Bulletin of the Seismological Meissner, R., 1986, The Continental Crust—A Geophysical Approach: San Diego, Academic Society of America, v. 74, p. 1209–1223. Press, International Geophysics Series, v. 34, 426 p. Steinhart, J.S., 1967, Mohorovicˇic´ discontinuity, in Runcorn, S.K., ed., International Diction- Meissner, R., 2000, The mosaic of terranes in central Europe as seen by deep refl ection ary of Geophysics, Volume 2: New York, Pergamon Press, p. 991–994. studies, in Mohriak, W., and Talwani, M., eds., Atlantic Rifts and Continental Margins: Thompson, R.I., Mercier, E., and Roots, C.F., 1987, Extension and its infl uence on Canadian American Geophysical Union Geophysical Monograph 115, p. 33–55. Cordilleran passive margin evolution, in Coward, M.P., Dewey, J.F., and Hancock, P.L., Meissner, R., and Mooney, W., 1998, Weakness of the lower continental crust: A condition for eds., Continental Extension Tectonics: Geological Society, London, Special Publication delamination, uplift, and escape: Tectonophysics, v. 296, p. 47–60, doi: 10.1016/S0040 28, p. 409–417, doi: 10.1144/GSL.SP.1987.028.01.25. -1951(98)00136-X. Tittgemeyer, M., Wenzel, F., Ryberg, T., and Fuchs, K., 1999, Scales of heterogeneities in the Mengel, K., and Kern, H., 1992, Evolution of the petrological and seismic Moho—Implications continental crust and upper mantle: Pure and Applied Geophysics, v. 156, p. 29–52, doi: for the continental crust-mantle boundary: Terra Nova, v. 4, p. 109–116. 10.1007/s000240050288. Mereu, R.F., 2003, The heterogeneity of the crust and its effect on seismic wide-angle refl ec- Tittgemeyer, M., Ryberg, T., Wenzel, F., and Fuchs, K., 2003, Heterogeneities of the uppermost tion fi elds, in Goff, J.A., and Holliger, K., eds., Heterogeneity in the Crust and Upper mantle inferred from controlled-source seismology, in Goff, J.A., and Holliger, K., eds., Mantle: Nature, Scaling, and Seismic Properties: New York, Kluwer Academic Publish- Heterogeneity in the Crust and Upper Mantle: Nature, Scaling, and Seismic Properties: ers, p. 299–320. New York, Kluwer Academic Publishers, p. 299–320. Mereu, R.F., and Ojo, S.B., 1981, The scattering of seismic waves through a crust and upper Villeneuve, M.E., Theriault, R.J., and Ross, G.M., 1991, U-Pb ages and Sm-Nd signature of mantle with random lateral and vertical inhomogeneities: Physics of the Earth and two subsurface granites from the Fort Simpson magnetic high, northwest Canada: Planetary Interiors, v. 26, p. 233–240, doi: 10.1016/0031-9201(81)90025-X. Canadian Journal of Earth Sciences, v. 28, p. 1003–1008, doi: 10.1139/e91-091. Mjelde, R., Fjellanger, J.P., Digranes, P., Kodaira, S., Shimamura, H., and Shiobara, H., 1997, Welford, J.K., Clowes, R.M., Ellis, R.M., Spence, G.D., Asudeh, I., and Hajnal, Z., 2001, Litho- Application of the single-bubble airgun technique for OBS-data acquisition across the spheric structure across the craton-Cordilleran transition of northeastern British Colum- Jan Mayen Ridge, North Atlantic: Marine Geophysical Researches, v. 19, p. 81–96, doi: bia: Canadian Journal of Earth Sciences, v. 38, p. 1169–1189, doi: 10.1139/cjes-38-8-1169. 10.1023/A:1004275600265. Yilmaz, O., 2001, Seismic Data Analysis: Tulsa, Oklahoma, Society of Exploration Geophysi- Mohorovicˇic´, A., 1910, Das Beben vom 8. X. 1909: Jahrbuch des Meteorologischen Observa- cists, 2024 p. toriums in Zagreb, v. 9/4, p. 1–63. Zelt, C.A., and Smith, R.B., 1992, Seismic travel-time inversion for 2-D crustal velocity struc- Mooney, W.D., 1987, Seismology of the continental crust and upper mantle: Reviews of Geo- ture: Geophysical Journal International, v. 108, p. 16–34, doi: 10.1111/j.1365-246X.1992. physics, v. 25, p. 1168–1176, doi: 10.1029/RG025i006p01168. tb00836.x. Morozov, I.B., Smithson, S.B., Chen, J., and Hollister, L.S., 2001, Generation of new continen- MANUSCRIPT RECEIVED 22 FEBRUARY 2010 tal crust and terrane accretion in southeastern Alaska and western British Columbia: MANUSCRIPT ACCEPTED 10 MAY 2010 Constraints from P- and S-wave wide-angle seismic data (ACCRETE): Tectonophysics, v. 341, p. 49–67, doi: 10.1016/S0040-1951(01)00190-1. Printed in the USA

Item No. Actual No. Copies Statement of Ownership, Management, and Circulation from Avg. No. Copies of Single Issue (Required by Title 39 U.S.C. 4369) PS Form Each Issue in past Published Nearest 3526 Extent and Nature of Circulation 12 Months to Filing Date Lithosphere (Publication No. 1941-8264) is published bi-monthly by The a. Total No. Copies (Net press run) 628 2000 Geological Society of America, Inc., (GSA) with headquarters and offices at b. Paid and/or Requested Circulation 3300 Penrose Place, Boulder, Colorado 80301 U.S.A.; and mailing address of (1) Sales through dealers and carriers, street vendors, and Post Office Box 9140, Boulder, Colorado 80301-9140 U.S.A. The Publisher is counter sales (not mailed) 0 0 Jon Olsen; offices and mailing addresses are the same as above. The annual (2) Paid or Requested Mail Subscriptions, (Including advertisers) subscription prices are: GSA Members $60; GSA Student / K-12 Teacher proof copies and exchange copies) 321 321 Members $30; non-members $450. The publication is wholly owned by The c. Total Paid and/or Requested Circulation (Sum of b (1) and b (2)) 321 321 Geological Society of America, Inc., a not-for-profit, charitable corporation. No known stockholder holds 1 percent or more of the total stock. There are d. Distribution by Mail (Samples, complimentary, and other free) 0 0 no known mortgagees or holders of other securities. The purpose, function, e. Free Distribution Outside the Mail (Carriers or other means) 0 0 and nonprofit status of The Geological Society of America, Inc., has not f. Total Free Distribution (Sum of d and e) 0 0 changed during the preceding twelve months. The average number of copies g. Total Distribution (Sum of c and f) 321 321 of each issue during the preceding twelve months and the actual number of h. Copies Not Distributed (1) Office use, leftovers, spoiled 307 1679 copies published nearest to the filing date (October issue) are noted at right. This information taken from PS Form 3526, signed 17 September 2010 by (2) Returned from news agents 0 0 the Publisher, Jon Olsen, and filed with the United States Postal Service in i. Total (Sum of g, h (1), and h (2)) 628 2000 Boulder, Colorado. Percent Paid and/or Requested Circulation (c/g x 100) 100% 100%

396 www.gsapubs.org | Volume 2 | Number 5 | LITHOSPHERE

Downloaded from http://pubs.geoscienceworld.org/gsa/lithosphere/article-pdf/2/5/377/3050505/377.pdf by guest on 26 September 2021