The LITHOPROBE Trans-Continental Lithospheric Cross-Section: Imaging the Internal Structure of North America
Crustal-Scale Images of the Continent-Ocean Transition Across the Eastern Canadian Margins
The acquisition and analysis of ~10, 400-500-km-long, deep MCS reflection and wide-angle reflection/refraction (WAR/R) profiles across the eastern Canadian continental margins from Nova Scotia to Baffin Is. have been accomplished over the past 20 years during a number of joint Canadian and international programs. The combination of both reflectivity and velocity images from separate MCS and WAR/R profiles have detailed the large-scale patterns of crustal extension, mantle serpentinization and exhumation, and ocean crustal formation both within and between rifted segments from full thickness continental crust to oceanic crust produced by sea-floor spreading. A number of striking features are documented by these crustal-scale sections. In particular, a wide transition region with very thin seismic crust is delineated by a well-defined upper mantle zone with reduced velocities interpreted as partially serpentinized peridotite. The geological nature of the transitional crust is quite complex and may consist of various regions dominated by highly stretched continental crust, highly serpentinized continental mantle or thin ultra-slow spread ocean crust. It is difficult to define the nature of this region from its velocity structure alone, however, since it is only poorly resolved by standard travel-time methods. One robust characteristic that is generally observed is an abrupt change to typical ocean crust at the seaward edge of the transition zone. This boundary shows characteristic and coincident variations in both velocity structure and basement morphology. New results from the eastern margin of Flemish Cap demonstrate such a pattern particularly well. This observation suggests that once melt begins to form it causes an abrupt shift from a diffuse pattern of lithospheric extension to a focused zone of melt formation. Based on our profiles, we suggest that such transitions have occurred at a number of discrete pulses, which progress in age from south to north and may be linked to previously documented pulses of continental extension in adjacent regions. Future improvements in characterizing the nature of the transition zone and its boundaries will require new detailed seismic profiles in which reflection and refraction data are analyzed simultaneously, including use of full waveform techniques. This will necessitate use of longer towed MCS arrays (10-15 km) and a greater number of ocean bottom receivers (50) spaced at smaller intervals (2-5 km). These requirements are well within our present capacity, but will require new international initiatives with significant new funding. Since the nature of the transitional crust has important implications for both resources and political boundaries, it is hoped that such potential can be realized within the next 5-10 years.
Investigating Variations in Rifting Style Along the Southern Margin of Flemish Cap, Offshore Newfoundland: Results from the Erable Multichannel Seismic Reflection Experiment
In 1992, the Erable project was undertaken by the Geological Survey of Canada and Ifremer to acquire multiple 2-D multichannel seismic reflection profiles in the Newfoundland Basin and along the margins of Flemish Cap. We present four multichannel seismic reflection profiles from the project collected over the southern margin of Flemish Cap and extending into the Newfoundland Basin. These profiles are between and sub- parallel to lines 1 and 2 from the 2000 SCREECH seismic experiment and provide more comprehensive data coverage over the region. We combine these data with the SCREECH seismic profiles, two ODP drill sites, and other geophysical data to map distinct zones of continental, transitional, and oceanic crust in this region. Just as has been evidenced from the mapped crustal boundaries on their conjugate Galicia Bank and Iberian margins, the Flemish Cap and Newfoundland margins show significant along-margin variability in terms of rifting structures and styles. This along-margin variability is superimposed on the overall asymmetry of the conjugate pairs highlighting the complexity of the margins and the importance of considering three- dimensional influences on rifting evolution. In particular, the hypothesized clockwise rotation and southeastward motion of Flemish Cap and the transfer zones that would have accommodated such movement appear to have affected the distribution of extension along the margins as rifting propagated northward. Meanwhile, activity at the North Atlantic triple junction immediately to the east of Flemish Cap may have initiated slow seafloor spreading while rifting was still active to the south as evidenced along the nearby Erable profiles. While simple two-dimensional rifting models may be appropriate for interpreting individual seismic profiles, three-dimensional rifting models are clearly needed to adequately explain the evolution of Flemish Cap and Galicia Bank relative to the margins to the south. These rifting models must incorporate the influences of microplate reorganization on both sides of the North Atlantic as well as transfer zones and the North Atlantic triple junction.
Deep Seismic Reflection Imaging of the Queen Charlotte Basin
A deep 2-D seismic reflection survey was shot by the Geological Survey of Canada in Hecate Strait and Queen Charlotte Sound in 1988 using a 45 m shot point interval, 15 m group interval and a maximum offset of 3685 m. Subsequent analysis of these data has focused on the shallower section, and helped constrain, together with earlier industry data, the general evolution of the sedimentary section. North-trending sub-basins are found to the south in Queen Charlotte Sound, but relatively closely-spaced NW-trending sub-basins predominate further north in Hecate Strait. Previous interpretation has demonstrated transtensional tectonics during most of the Miocene with evidence for transpression and inversion of some normal faults in Hecate Strait during the Pliocene. Although a range of origins including hot spot rifting have been suggested for the Queen Charlotte basin, currently favoured explanations focus on the effects of, and changes in, large-scale plate motions in this region. To determine if the deeper sections of the 1988 seismic survey can further constrain models of basin evolution, these seismic data have been reprocessed to 14 s. A thick sequence of subhorizontal reflectors, which extend from 4.5-8.5 s, is identified beneath northern Hecate Strait where rocks of the Alexander terrane are identified at the surface. Further south, discrete 0.5-1.0 s thick reflection packages, some of which may be intrusive in origin, are identified in the mid and lower crust of the Wrangellia terrane. Other reflections occur as deep as 10-11 s beneath the central axis of the basin in Hecate Strait, which could locate them beneath the Moho inferred from wide-angle surveys. Deep seismic imaging is more limited beneath southern Hecate Strait and northern Queen Charlotte Sound, perhaps due to the effects of coherent noise, but reflections can be identified at 6-8 s beneath southern Queen Charlotte Sound. Marine surveys are often degraded by water layer multiples and coherent scattered energy. These problems can be mitigated using swath 3-D marine recording, and if the lateral velocity variation can be determined in the upper few km, a significant improvement in the quality of marine seismic images appears possible.
Waveform Inversion of Marine Seismic Reflection Data from the Queen Charlotte Basin
In order to obtain higher resolution seismic velocity data above approximately 1.5 km depth, two-dimensional waveform inversion was applied to seismic reflection data from the Queen Charlotte sedimentary basin off the west coast of Canada. The inversion, which employs the method of Pratt (1999), was performed using a multi- scale approach in the frequency domain based on the acoustic wave equation. The seismic data are from a survey of 8 lines that was collected in 1988 by the Geological survey of Canada. Data preconditioning, inversion strategies and results from line 88-06 which crosses Tertiary formations in Hecate Strait between Graham Island and Dolphin Island are presented. A starting velocity model, which gives an RMS travel time misfit of 18 ms, is derived from travel-time tomography, and is sufficient to avoid cycle skipping at a starting waveform inversion frequency of 8 Hz. Due to the limited maximum offset of 3700 m the maximum depth of penetration doesn't exceed 1500 m. The relatively high minimum frequency (8 Hz) constitutes a challenge for a successful inversion, as higher frequencies are subject to greater instability. A strategy is presented to mitigate this issue with the inversion implemented in two stages. The first stage consists of enriching at the initial frequency the starting model with intermediate to low wavenumbers, using a filter in the wavenumber domain to ensure continuity in the spectral coverage, and preconditioning the residual by inverting first for the most linear component of the data. The second stage consists of inverting from the initial frequency (8 Hz) up to 15 Hz, using 10 iterations per frequency and following a layer stripping approach to give the same kinematics to all portions of the model. Amplitude scaling and spiking deconvolution constitute the main preconditioning of the data; their effects are also shown. Model appraisal is carried out by comparing synthetic shot gathers obtained by forward modeling with the field-recorded data. This study demonstrates the improved resolution in the velocity model that is possible with marine data through the use of frequency domain waveform inversion even when the starting frequency is relatively large.
New Seismic Approaches to Persistent and New Problems in Deep Crustal Geology
Over 30 years of deep seismic reflection profiling has demonstrated that lower crustal complexity is manifest in diverse ways. Reflection profiling has been critical to relating such complexity to regional tectonic processes, often resolving specific structural questions involving the deep crust. Yet a number of important observations remain as cryptic now as they were when first observed decades ago, even when reflection methods are combined with other seismic and geophysical techniques. Some illustrative examples of persistent problems and future potential: a) The nature of deep crustal fluids: Seismic bright spots have been key features of numerous deep reflection and receiver function studies. Commonly interpreted as magma, the actually identity of the fluids involved remains to be confirmed. Deep brines and supercritical gas are respectable alternatives. Resolving the identity of these unusual reflectors is essential to assessing tectonic processes in the deep crust. Appropriate wide aperture, three component experiments are still lacking to address this problem. b)The 'other' mantle reflectors: Dipping seismic reflectors in the mantle lithosphere that appear to sole into the lower crust are widely interpreted as fossil subduction zones. However, subhorizontal reflectors (and corresponding convertors on receiver functions) are more problematic to the tectonic analyst. Subhorizontal mantle events have been interpreted as the base of the lithosphere (which begs the question as to the nature of this key interface), detachment surfaces, or phase change boundaries. What kind of experiment is needed to identify the nature of such interfaces? c) The layered lower crust: intrusion complexes or strain fabrics? There is evidence to support both common interpretations, and they are far from mutually exclusive. The point raised here is that the utility of such reflectivity as strain indicators has not yet been exploited. S wave reflection methods widely used in resource exploration could be applied to mine 3D vector information from deep reflectivity. d) Is the deep crust 'static'? Of course not: both seismogenic and volcanogenic processes involved temporal variations on human time scales. Time lapse (4D) imaging with both controlled and natural source approaches is an essential tool to probe active tectonics in the lower crust, and represents a key area of common interest between shallow and deep crustal seismologists. Great progress has been made recently in time lapse inversion of earthquake recordings in volcanic regions: where are the 4D controlled source studies?