Continental Assembly, Stability, and Disassembly : USArray: An Earth Sciences Initiative to Investigate the North American Continent

Alan Levander, Gene Humphreys, Goran Ekstrom, Anne Meltzer, and Peter Shearer

Introduction

The past decade and a half have seen major advances in structural seismology, i.e. in our ability to image complex Earth velocity and impedance structures and from them make valid inferences on the physical state within the Earth. Similar advances in the other solid Earth sciences have poised us for a major advance in our understanding of continental dynamics and evolution. USArray is a proposed project that will permit a three-dimensional systematic seismic investigation of North America that will improve the resolution of lithospheric images by an order of magnitude. A number of factors suggest that North America is particularly suited for this project, including the states of current knowledge and technology, the availability of a sophisticated infrastructure, organization in the seismological community, scientific economy, and widespread scientific interest. In the past whenever seismic resolution is dramatically improved the Earth sciences have made major advances in understanding the dynamics of our planet, changing the way we think about geologic processes. As examples we cite new understanding of the orogenic cycle based on deep crustal reflection images of crustal thickening and collapse, and evidence for whole mantle convection in the recognition of deeply subducted plates in global tomograms.

In this article we present a concept, USArray, in the context of the tectonic elements of the continent and developments in seismic imaging which are important for understanding continental evolution. A companion article (Shearer et al., this issue) presents an outline of the proposed USArray seismic facility and how it relates to current seismic networks. A key part of the USArray experiment is a transportable seismic array that will be fielded in a series of temporary deployments to study Earth structure and seismicity across the whole of the lower 48 United States. The plan for USArray is evolving as we seek and respond to community input. A primary goal is to design a strategy for studying the structure and ongoing deformation of North American. The plan, in its current formulation, has three seismology components and an education and outreach component. The seismology components consist of (1) an upgrade of the permanently installed US National Seismic Network in coordination with the US Geological Survey and US regional networks, (2) a set of 100 to 200 portable telemetered seismographs which will be systematically deployed to cover the entire lower 48 United States, and (3) additional broadband and short-period instruments for complementary experiments. These three components provide an adequate set of reference stations, systematic coverage of the continent with high resolution, and very high resolution of the continent in tectonically important areas, respectively. The data from the transportable array and the fixed stations would be available to the community in near real time, ensuring timely analysis of the data.

An education and outreach component will capitalize on the attention that will be focused on one region of the country after another, linking schools and the general public with specific area- related geologic issues and making the roving array an exciting affair nationwide. Specific programs will be developed for all levels of education and the media. The goal is to use US Array to focus public attention on the geological sciences, thereby increasing public awareness and interest in geology and in science in general.

Finally another intent of USArray is to provide a framework for other branches of the geological sciences to join together in addressing the important geological problems in different regions of our continent, be they the investigation of the San Andreas fault, orogenesis in the western US, the structure of the craton, or the assembly of the continent. US Array will provide a framework of seismic information and an organizational structure that can focus the Earth sciences in a systematic investigation of the North America.

Investigating the Architecture of the North American Continent

USArray is designed to provide a 3-D look at a significant part of the North American continent with resolution an order of magnitude better than global tomograms, similar to many regional 2-D PASSCAL experiments. North America has a diversity of tectonic provinces (Figure 1): The interior is constructed of a series of granite and greenstone belts of Precambrian age. These belts are known to have experienced a complex Precambrian tectonic history, with much in common with modern plate tectonics, but also with some notable differences. The center of the continent is largely Archean, while the south/central portion of the continent consists of a succession of Proterozoic terranes some of which are island arcs accreted to the craton by Proterozoic collisions. Around the periphery of the craton are a series of mobile belts containing a record of Phanerozoic collisional and extensional tectonics. In the eastern quarter of the continent the Appalachian- Caledonide continent-continent collision zone was a major orogenic belt formed during construction of the Pangean super- continent. Passive margins and rift basins formed on the Eastern and Gulf coasts during Pangean disaggregation and formation of the Atlantic Ocean. In the west a succession of oblique island arc- continent collisions left a complicated history of orogenic uplift and terrane migration. The Mesozoic-Cenozoic ocean-continent collisional system that girdles half the globe along the western Americas has particular features of scientific and societal relevance in North America. This margin consists of two major strike slip boundaries (the San Andreas and Queen Charlotte-Fairweather systems), and three trenches. All but the Queen Charlotte system are seismically hazardous and pose threats to major urban centers. The larger plate boundary system, which space geodetic data shows extends inland as far as the Rocky Mountain front, contains one of Earth's three great orogenic plateaus, two active rift systems, an orogenic collapse structure, and one of a handful of global hotspots. The eastern ranges of the Canadian Cordillera are the type locales for thin- skinned fold and thrust belts, whereas the Laramide are the type locales for thick-skinned tectonics. Although many of these tectonic features are presently being studied, the sytems are so large as to require a more systematic effort than can be mounted by individual investigators or small scientific working groups.

Figure 1: Shaded relief map of the North America showing current physiography and tectonic regimes. The area of North America shown in figures 1 and 2 is identical.

The geologic division of the continent into a stable cratonic center with peripheral mobile belts is reflected in global tomograms of upper mantle structure. Figure 2 is a shear wave velocity perturbation model of North America at a depth of 100-175 km (Grand, 1994). The S-wave shows that North American upper mantle, like the surface geology, is broadly divided into two domains, with the blue anomalies representing fast, cold stable lithosphere, the tectosphere, underlying the craton, and red anomalies representing hot, slow, presumably mobile mantle underlying the tectonically active west. The high velocity anomaly centered on Hudson Bay, extending to at least 400km depth, is one of the largest positive velocity anomaly on Earth. Equally dramatic is the velocity gradient from the continental interior to the western margin, where the unusually low velocities beneath the orogenic plateau extend as far north as the Mendocino Triple Junction, as far west as the Rio Grande rift, and continue southward to underlie the Gulf of California and the East Pacific Rise. Images of the deeper mantle under North America show that the subducted Farallon plate is now beneath the midcontinent as it is recycled by the Earth's convection system.

Figure 2: S-wave velocity anomalies at 100-175km depth under North America (from Grand, 1994). The interior of the continent is underlain by cold fast tectospheric mantle, whereas the peripheries, and particularly the western US are slower. Slow velocities under the western US, corresponding to regions of Mesozioic-Cenozoic tectonic activity, may represent young mobile mantle or North American mantle thermally and chemically altered during Farallon subduction.

Seismic imaging of the Earth has been enhanced in the past decade with developments in imaging theory, lithospheric-scale active source seismology, global tomography using the Global Seismic Network operated by IRIS and its international partners, and a revolution in seismic experimentation resulting from deployment of portable IRIS PASSCAL broadband seismographs in remote corners of the globe. Global tomography images now have resolution of 500 to 2000 km, depending on a variety of factors including stations spacing and analysis methods(Figure 2). Data from a number of broadband PASSCAL experiments, coupled with observations from regional networks for monitoring seismicity in western North America have been utilized for the upper mantle P-wave anomaly map at 100 km depth shown in Figure 3 (Humphreys and Dueker, 1994). This image, which has lateral resolution which ranges from 50 to 500 km has a number of notable features. Trends in much of the surface geology are related to trends in the mantle structure, e.g., the linear Sierra Nevada batholith and Great Valley forearc basin appear in California with colinear mantle anomalies, the mantle beneath the volcanic Snake River plain and Yellowstone hotspot is slow, and the Colorado plateau has relatively fast mantle ringed by low velocities in the Basin and Range and Rio Grande rift system. Less obvious are the trends of mantle features in the southwestern US where slow anomalies are aligned with a preferred northeast- southwest orientation along sutures dating to Proterozoic construction of the southwest. Areas in the tomogram with little mantle structure have few or no seismic observations. The upper mantle at lateral scales of 100km shows far more variation and structure than previously suspected, however this image suffers from uneven sampling.

Figure 3: Composite image of upper mantle P-wave seismic structure at 100 kmdepth beneath the western US. Blue is high velocity mantle and red is low velocity mantle. Resolution of this image is variable, but is about 300 km. Actual total range in resolved P-wave velocity anomaly is about 8%. Using standard scaling relations, red regions are partially molten and blue regions are subsolidus (modified from Humphreys and Dueker, 1994; and Grand, 1994).

Complementing the tomographic images are converted wave seismic sections that show the Earth's impedance structure variations in great detail. Figure 4 (Dueker and Shehan, 1997) is a seismic section from the PASSCAL Snake River Plain experiment showing a converted S-wave image of mantle discontinuity structure centered on the axis of the Snake River plain and extending to 700 km depth. The data have been processed in a manner similar to a common midpoint reflection image using forward scattered rather than backscattered energy. The image has a vertical resolution of ~30 km. The prominent events between 400-500, and 600-700km correspond to the top and bottom of the mantle transition zone. The topography on the 410 and 660 discontinuities is pronounced directly beneath the Snake River plain, a recent volcanic center. Much image formation research remains to be done before we can interpret this type of section with the same confidence that we treat CMP reflection sections, yet this development is an important step forward.

Figure 4: Mantle discontinuity structure under the eastern Snake River Plain from receiver-function common depth point stack with S-wave tomographic velocity perturbations superimposed as a gray scale (Dueker and Sheehan, 1997). Converted shear waves recorded from teleseismic body waves are sorted by depth and location of conversion and stacked for signal enhancement. The positive velocity gradients at the 410 and 660 discontinuities are clearly visible with topographic relief beneath the axis of the Snake River Plain. The event at 250km depth may be a reverberation, the weak event at 525km depth is unexplained.

Advances have also been made in deep reflection profiling of the continental lithosphere, notably data quality is significantly improved so that all modern data can be wavefield depth migrated, and signal penetration has improved to the point that continuous upper mantle reflections from depths of 100 or more kilometers are now often seen. This is nowhere more spectacularly shown than in the recent Lithoprobe SNORCLE transect across the Slave province and the Proterozoic Wopmay orogen in the Northwest Territories, which shows continuous mantle reflections at depths of 80km and possibly 200km beneath the Wopmay orogen connected to a paleosubduction zone(Bostock, 1998). Among the many things that these data suggest is that mantle accretion and shortening beneath the Moho in developing mountain belts is as important to lithospheric dynamics as terrane accretion and crustal scale shortening is above the Moho.

Crust-Mantle Processes

Taking the view that Earth structure is the result of successive processes described by differential or integral equations, modern seismic images provide a number of important parameters for use in process description. They provide geometry, and they can be interpreted at least partly quantitatively for strain regime and rheology. Mantle images provide a proxy for temperature given suitable other constraints. Although we can not yet solve the geodynamic inverse problem, we can constrain aspects of the inverse problem, as well as provide boundary conditions and geometrical constraints for the forward problem (e.g. Willett et al., 1993).

While it is not surprising that the mantle controls many crustal processes, the forms of this control are often not obvious. For example, in the San Andreas transform system active source seismic imaging shows that the strike slip faults deform the entire crust to the Moho (Henstock et al, 1997), and both active seismology and regional P-wave tomography shows that mobile upper mantle associated with Gorda slab removal lies within the fault system where modern seismicity and volcanism are concentrated (Levander et al., 1998; Benz, personal communication). We infer that upper mantle structure controls crustal faulting in the San Andreas system at 0-2 Ma ages. Similarly, in southern California active source seismology has demonstrated that the Sierra Nevada batholith has an insufficient crustal root to support its elevation (Wernicke et al., 1996), but buoyant mantle beneath the eastern Sierra Nevada and western Nevada provide support. The crust can also control mantle structures, or else crust-mantle structures act in concert through time. As we pointed out previously the northeast trending low velocity upper mantle anomalies in the southwest US follow Proterozoic age crustal sutures, apparently providing long lived controls on modern tectonics (Karlstrom and Bowring, 1988). The crust can also provide indirect control on mantle processes, for example dehydration of subducting oceanic crust at ocean-continent collisional zones mechanically weakens the overlying continental mantle lithosphere. This has been offered as one explanation for the upper mantle low velocity zone observed beneath the Colorada Plateau and southern Rocky Mountains: Low angle Farallon subduction resulted in widespread aqueous fluid emplacement in the overlying North American mantle. With these few examples we have tried to demonstrate that seismic imaging across a broad range of scales, which requires a number of different seismic probes, allows us to infer tectonic processes better than any single tool can. Not only are the images consistent, but they are complementary to one another. USArray offers an opportunity to investigate the structure of the continent and the processes that formed it across a broad range of scales.

USArray

USArray is a seismic facility envisioned to systematically investigate the three-dimensional seismic structure of the US North American lithosphere in a ten year period. Most seismic imaging done in PASSCAL experiments now is two-dimensional; available instrumentation is inadequate for high resolution three-dimensional imaging. USArray has several components, including a telemetered array of portable, observatory quality, broadband seismographs deployed with constant instrument spacing and therefore array characteristics. The technology to construct and field USArray exists today, although developments in global satellite communication systems may decrease the anticipated telemetry costs significantly. Deployed repeatedly for 6 to 12 month intervals, this array will cover the entire lower 48 states to provide resolution of upper mantle structures at the scale of ~100-150 km. (see Figure 2 of Shearer et al., this issue). These data will be available to the Earth science community in near-real time. The second element is another group of instruments used as a flexible portable array for making detailed measurements of specific targets within the footprint of the first mobile array. Regardless of the array configuration employed, there are common underlying sampling criteria that relate array design to resolution of Earth structure. The flexible component of the transportable array can be used to improve resolution for specific crust-mantle features. Details of implementation of this aspect of USArray remain to be worked out.

The third element of USArray is an enhancement of the US National Seismic Network permanent station network to produce a national grid of broadband seismographs with an array spacing of 300 to 400 km. The permanent array will provide the baseline measurements for the portable array.

Finally, an integral part of the USArray plan is a public Earth sciences education program to accompany the array as it moves around the United States.

The Earth Sciences and USArray

The great strength of Canada's highly successful Lithoprobe program (Clowes et al., 1998) has been the integrated multidisciplinary effort to understand the structure of the Canadian crust using their extensive seismic reflection profiling program as a catalyst for other studies. We feel that equally important to the success of the systematic seismic are a number of other investigations : 1) Individual PI active and natural source PASSCAL experiments and other geophysical experiments coordinated with USArray which target specific tectonic problems within the USArray study area. These experiments will shrink the scale resolution locally by 1 to 2 orders of magnitude. 2) Coordinated geological, geodetic, and geochemical studies to systematically investigate North America which will make the seismic images interpretable in terms of tectonic development of the continent. 3) Partnership with our Canadian and Mexican colleagues to truly address the tectonic structure of the entire continent. We note here that a US marine science initiative to build a fleet of Ocean Bottom Seismographs is being considered and may be able to contribute a continental margin component to USArray.

The US has several community organizations which can cooperate to produce an integrated geophysical field program such as USArray: IRIS, UNAVCO, EMSOC, and CSEDI. Organized input describing the needs of the geological and geochemical communities for this continent-scale Earth sciences investigation are necessary to fully exploit the opportunities USArray presents. Clearly the project we envision will require a significant community driven organizational effort. Again we point to the successes of our colleagues to the north who undertook this same task some ten years ago.

Conclusions

Advances in seismic instrumentation, and interpretation and imaging techniques in the past decade have made it possible to develop highly resolved crust-mantle images which constrain many processes of lithospheric formation and evolution. We are advocates of a program and facility to systematically investigate the North American lithosphere which is a seismological equivalent of the Hubble space telescope for the Earth sciences, with related geological, geodetic, and geochemical investigations. We are holding a workshop from March 15-17, 1999 in Albuquerque, New Mexico which will consider the organizational and technical details of this undertaking.

Interested parties can register through the URL http://www.iris.edu/USArray.html, or can contact the authors of this paper.

Acknowledgements

The idea for USArray was presented to the seismological community at the NSF-AFOSR sponsored ILIAD workshop in Taos, New Mexico in November 1994, by Art Lerner-Lam of Lamont Doherty Earth Observatory. The idea was developed further by an ad-hoc working group which included the authors and Jeffrey Park, George Zandt, Paul Silver, Rob van der Hilst, and Tom Owens. The USArray concept has been discussed at a PASSCAL Instrumentation workshop , and at the 1998 IRIS Annual Workshop.

References

Bostock, M. G. (1998), " Mantle stratigraphy and evolution of the Slave province." J. Geophys. Res., 103, 21183-21200.

Clowes, R. M., F. A. Cook, et al. (1998), "Lithoprobe Leads to New Perspectives of Continental Evolution." GSA Today, 8, 1-7.

Dueker, K. G. and A. F. Sheehan (1997), "Mantle discontinuity structure from midpoint stacks of converted P and S waves across the Yellowstone hotspot track." J. Geophys. Res., 102, 8313-8326.

Grand, S. P. (1994), " Mantle shear structure beneath the Americas and surrounding oceans." J. Geophys. Res., 99, 11591-11621.

Henstock, T. J., A. Levander, et al. (1997), "Deformation in the lower crust of the San Andreas Fault system, northern California." Science, 278, 650-653.

Humphreys, E. D. and K. G. Dueker (1994), "Western U.S. Upper Mantle Structure." J. Geophys. Res., 99, 9615-9634.

Karlstrom, K. and S. A. Bowring (1988), "Early Proterozoic assembly of tectonostratigraphic terranes in southwestern North America." J. Geology, 96, 561-576.

Levander, A., T. J. Henstock, et al. (1998), "Fluids in the lower crust following Mendocino triple junction migration, Active basaltic intrusion?" Geology, 26, 171-174.

Wernicke, B., R. Clayton, et al. (1996), "Origin of High Mountains in the Continents, The Southern Sierra Nevada." Science, 271, 190-193.

Willett, S., C. Beaumont, et al. (1993), "Mechanical model for the tectonics of doubly vergent compressional orogens." Geology, 21, 371-374.