Studies of convergent margins by U.S. researchers have largely focused on North and Central America. Many of these studies have attempted to map the geometry of the subduction megathrust or to learn more about its physical properties.
A reconnaissance grid of deep-crustal reflection and wide-angle seismic lines in the northeastern northern Gulf of Alaska provide evidence for tectonic underplating along the Trans-Alaska Crustal Transect [ Brocher et al., 1991; Fuis et al., 1991; Fuis and Plafker, 1991; Wolf et al., 1991; Fuis and Clowes, 1993; Brocher et al., 1994]. An additional deep-crustal reflection line crossing the Aleutian Arc and trench imaged mid-crustal reflections interpreted as duplexing of partially subducted sediments accreted during rapid Eocene plate convergence [ Moore et al., 1991]. Wide-angle profiling along these lines defined a regional 6.9 km/s refractor and reflector which Brocher et al. [1991; 1994] correlate to the top of the subducted Yakutat terrane and the Aleutian megathrust. This profiling may represent the first time a mid-crustal asperity on a subduction zone has been mapped using seismic reflection and refraction methods. The three-dimensional structure and inferred geometric controls on the asperity of the megathrust again emphasizes the need for a grid of seismic coverage.
Recent work on the Cascadia margin in the Pacific Northwest has provided evidence for continental growth via tectonic underplating and accretion of large igneous terranes and has addressed the seismic hazards posed by subduction by mapping the megathrust geometry. Seismic imaging of the top of the subducting plate coupled with heat flow measurements and other observations have been used to delineate the seismically active part of the megathrust for seismic hazard evaluation [ Hyndman and Wang, 1993; Dragert et al., 1994]. The teleseismic receiver function method in the vicinity of Vancouver Island provided the first systematic tracking of the subducting Juan de Fuca plate across the forearc [ Cassidy and Ellis, 1991; 1993]. In a creative extension of the teleseismic receiver function method, Nabelek et al. [1993] deployed a dense line of broad band receivers across the Cascadia margin in central Oregon. The processed broad band seismograms, displayed as a seismic record section, clearly image the subducting plate beneath this margin, although seismic tomography had also indicated the presence of this slab [ Harris et al., 1991]. The subducting plate in central Oregon produces so few earthquakes that its geometry can not be defined by seismicity within it. Tréhu et al. [1994] performed onshore-offshore and onshore seismic refraction/wide-angle reflection surveys along a nearby E-W transect, providing evidence for a thick (30-km) basaltic block beneath this part of the Cascadia margin. This coherent block, inferred by Tréhu et al. [1994] to represent an accreted large igneous terrane (oceanic plateau), thins dramatically towards the north and latitudinal variations in its thickness may explain the paucity of seismicity in central Oregon relative to the Puget Sound region of Washington. The igneous terrane thins dramatically to the north in the vicinity of the Puget Sound where seismicity both in the crust and on the megathrust is more commonly observed. Based on a grid of reflection lines offshore the southern half of Vancouver Island, Spence et al. [1991] identified a similar but much thinner igneous body of Eocene oceanic basalts and interpreted it to be a landward dipping backstop to the accretionary prism.
Seismic reflection and refraction evidence for possible examples of accreted ophiolitic slivers within the New Britain accretionary wedge were presented by Berstein-Taylor et al. [1992a,b]. These workers proposed that pre-existing high and low angle normal faults within the igneous oceanic crust, which probably formed in the vicinity of the mid-ocean spreading center, may become reactivated as thrust faults at the accretionary wedge. This study provided additional evidence for the importance of tectonic accretion of oceanic crust as a means of forming the continental crust.
In an exciting study, although not strictly located at a continental margin, the application of three-dimensional seismic reflection profiling to the northern Barbados accretionary prism has allowed the three-dimensional mapping of a major active low-angle thrust (décollement) fault [ Shipley et al., 1994]. The negative-polarity reflections from the fault, produced by a zone of material having a lower seismic velocity (and/or density) embedded within material having a higher seismic velocity (and/or density), were interpreted as originating from a thin, high porosity zone representing an under compacted, high fluid-pressure dilatant section. Positive-polarity sections of the fault plane, produced by a zone of material having a higher seismic velocity (and/or density) embedded within material having a lower seismic velocity (and/or density), were inferred to represent lower porosity sections having high frictional strength and possibly forming asperities in the fault plane. To date, the Ocean Drilling Program has been unsuccessful in drilling the negative-polarity sections of the fault but has succeeded in drilling one positive-polarity fault segment [ Shipley et al., 1994]. The direct imaging of the active décollement and the ability to make inferences about the spatial distribution of seismic asperities using controlled-source methods will contribute to a quantitative assessments of earthquake hazards, and hence are highly relevant to societal needs.
In a related study, two-ship expanding spread profiles allowed Stoffa et al. [1992] to study the distribution of seismic velocities in the sedimentary horizons within the Nankai Trough south of Japan. They reported a low velocity zone beneath the subduction décollement, which is consistent with models for the low wedge taper, high pore-fluid pressure, and negative reflection polarity described by other workers.
Three-dimensional seismic reflection profiling of the Costa Rica accretionary prism yielded evidence for large structural diversity over distances of several hundred meters [ Stoffa et al., 1991; Shipley et al., 1992; McIntosh et al., 1993]. Both duplexing and out-of-sequence faulting were observed within a few kilometers landward of the trench. McIntosh et al. [1993] interpreted seismic reflection data from the Costa Rica accretionary prism as evidence for extension of the margin.
Examination of a paleo-convergent margin beneatrh the Texas Gulf Coast margin was undertaken along a deep crustal transect consisting of Consortium for Continental Reflection Profiling (COCORP) vibrator data [ Culotta et al., 1992] and industry marine deep seismic reflection data [ Weimer and Buffler, 1992]. The COCORP data suggest that the San Marcos Arch involves the Precambrian Grenville basement. In contrast, the Mississippi Fan fold belt, located in the deep Gulf of Mexico, involves Middle Jurassic to Miocene sedimentary strata, but not the underlying basement rocks [ Weimer and Buffler, 1992].