Printer Friendly

New seismic images of the oceanic crust.

How do we know what lies beneath the seafloor at an oceanic spreading center? There are a variety of approaches. The topography of the spreading axis must be an expression of processes occurring along the ridge, so by constructing detailed bathymetric maps of the seafloor it is possible to infer something about the structure of the underlying crust. Variations in the gravity and magnetic fields over a spreading center can also be used to map crustal and upper-mantle properties. The chemistry of lavas erupted at the seafloor and the composition of rocks dredged from steep submarine scarps provide a different type of window into the workings of a spreading center. However, all of these approaches are indirect in that they do not allow us to "see" the internal structure of a spreading ridge.

The closest we can come to directly imaging the structure of a spreading center is with images that the seismic reflection technique produces. Marine seismologists create these images using the same basic principles that allow obstetricians to obtain "pictures" of an unborn child in a mother's womb. In both cases, sound, rather than light, is used to create the image. Very-high-frequency sound is required to construct a sonogram of the fetus. Seismologists employ much lower frequency sound to record "echograms" from the boundaries between layers of rock deep within Earth's crust.

Because seismic energy travels as an elastic wave, it is both reflected from and penetrates through the seafloor and interfaces beneath. This enables us to simultaneously "see" the seafloor and the structure below. These reflections usually arise from relatively abrupt changes in seismic velocity and/or density. These contrasts may be caused by changes in rock composition (that is, from basalt or gabbro to peridotite), bulk porosity (between fractured and unfractured rock), or the physical state of the rock (molten to solid basalt). This reflected energy travels back to the surface, where it can be detected and recorded. These echoes are usually extremely weak, but the signal can be amplified by recording a large number of reflections from the same point on the seafloor using a long string of receivers arranged in groups. By measuring the time between the outgoing pulse of sound and the returning echoes, it is possible for seismologists to construct highly detailed acoustic "pictures" of the crust below the seafloor.

This technique, known as multichannel seismic reflection profiling, is widely used in the exploration industry to locate oil and gas in sedimentary basins both on land and along continental margins. Marine seismologists at universities and research laboratories have been using these same techniques since the mid-1970s to investigate the structure of igneous crust lying below the seafloor, and the tectonic and volcanic processes associated with the great shifting plates that make up Earth's outer layer. The world-encircling mid-ocean ridge system, where new oceanic crust is being formed as plates slowly slide apart, is one of the most difficult areas to conduct multichannel seismic studies. Because the seafloor is very rugged along this mountainous ridge system and little or no sediment covers the fresh, young basaltic lava, reflections and side echoes tend to obscure the structure of the crust below the seafloor. Despite these problems, several recent multichannel reflection studies have been conducted along the mid-ocean ridge and over older crust on the ridge flanks. They have yielded spectacular images of the internal structure of the oceanic crust that have changed some long-standing ideas about the origin and structure of oceanic crust.

Creating a Reflection Image of Oceanic Crust

Modern seismic reflection profiling requires a specially equipped research vessel to acquire the original records, and sophisticated computer processing to construct the final acoustic images from hundreds of thousands of individual seismograms. The vessel used by most academic marine seismologists is R/V Maurice Ewing operated by Columbia University's Lamont-Doherty Geological Observatory. An extremely powerful sound source is required on vessels like this in order to image the entire oceanic crustal section. The base of the crust, marked by a boundary known as the Mohorovicic discontinuity (or Moho for short), generally lies at least 10 kilometers below the sea surface. Since high-frequency sound is rapidly attenuated in the earth, the source must be able to emit sound at very low frequencies, usually in the range of about 6 to 60 hertz. The initial sound pulse must also be relatively sharp, or the returns from closely spaced reflectors will overlap and blur the final image. The most commonly used sound source is the air gun, a device that suddenly expels a small volume of air under high pressure. The rapid expansion of this air bubble creates the initial pulse of sound. Rather than use one large air gun, an array of guns of different sizes, some quite small, are tuned to produce a sharp, powerful signal. In a typical Ewing array, 20 airguns are deployed from the stern A-frame and two booms extending about 10 meters out on either side of the ship. These guns release a total of 8,000 cubic inches of high-pressure air (approximately 2,000 pounds per square inch) every 20 seconds, creating a sound that can penetrate as much as 15 kilometers into the earth.

The weak echoes coming back from reflecting surfaces deep within the ocean crust are detected by sensitive hydrophones located in a long plastic tube or "streamer" towed behind the ship. This streamer, which can be thought of as a long acoustic "antenna," is filled with a low-density fluid and towed about 10 meters below the sea surface to isolate it from the acoustic noise produced by waves at the sea surface. The Ewing streamer is up to 4 kilometers long and has 4,000 individual hydrophones located at about one-meter intervals. These hydrophones are grouped together electrically to form up to 240 separate receiving channels spaced 12.5 or 25 meters apart. The acoustic return detected by the hydrophones in each channel is converted to an electrical signal, amplified, digitized, and transmitted up the streamer to the ship where it is recorded. Each channel is stored separately since each records a seismic signal from sound that has traveled along a slightly different source-to-receiver path, and bounced off different reflector points within the crust.

Multichannel seismic profiling generates prodigious amounts of data. Using the system described above, 240 separate data channels will be recorded every 20 seconds for each shot. Each of these records is usually at least 8 seconds long to ensure that echoes from deep within the crust are detected. The data from each channel is typically digitized at 250 samples per second (4 bytes per number) resulting in about 2 megabytes of data every 20 seconds. A 50-kilometer reflection profile is comprised of approximately 1,000 individual shots, and thus records about 2 gigabytes of data. That is about 25 times the capacity of the hard disk on an average desktop PC! A typical marine multichannel seismic survey may record some 3,500 kilometers of reflection data in a single cruise, resulting in nearly 150 gigabytes of data recorded on hundreds of magnetic tapes.

Extensive computer processing of the recorded data is required when the ship returns from sea in order to construct a reflection image of the oceanic crust. The reflections from deep crustal layers are usually very weak. Only a fraction of the original sound pulse is reflected back to the surface, and this echo is attenuated by its passage through many kilometers of rock. This signal can also be obscured by acoustic "noise" caused by towing the streamer through the water, wave action, or the sound of the ship's engines. This noise, however, is generally random while reflections have a coherent waveform from trace to trace. It is possible to cancel this noise by adding the signals from many different records with the same reflection point on the seafloor using a technique known as common-midpoint profiling. A general rule of thumb is that the signal-to-noise ratio increases as the square root of the number of traces added together. Thus if the 240 separate channels recorded by Ewing's streamer are all used, the amplitude of weak crustal reflectors may be increased by as much as 15 times. The thousands of traces generated by this process are plotted side by side to produce a final image like the ones shown above.

These reflection images look like a vertical section or slice through Earth's crust. Strong reflections appear on these sections as dark and light bands. By mapping these events, geologists have been able to construct detailed models of what the sub-seafloor structure looks like. The following examples illustrate the unique contributions reflection imaging has made to our understanding of oceanic crustal structure and the processes of seafloor spreading.

Crustal Structure Varies with Spreading Rate

Spreading rates along the global mid-ocean ridge system vary from less than 15 to more than 150 kilometers per million years. These variations are typically accompanied by a systematic variation in topography at the rise axis. Where opening rates are slow, as along the Mid-Atlantic Ridge, the spreading axis is usually associated with a deep rift valley flanked by shallow, rugged rift mountains. At faster opening rates, like those along the East Pacific Rise, the rift valley disappears and the spreading axis is usually associated with a linear volcanic ridge that has relatively subdued flanking topography.

Conventional seismic refraction studies have shown that the thickness and gross seismic velocity structure of oceanic crust formed at fast- and slow-spreading ridges are similar, leading to the view that the structure of oceanic crust does not vary with spreading rate. However, images of the oceanic crust constructed from seismic reflection data challenge this assumption. Reflection profiles of crust created at fast-spreading ridges, like the East Pacific Rise, typically reveal an acoustically transparent crust with a strong, quasi-continuous reflection of about 2 seconds duration (some 6 kilometers) below the seafloor that is thought to be a reflection from the base of the crust. In contrast, the crust created at slower spreading ridges, like the Mid-Atlantic Ridge, is characterized by a wide variety of intracrustal reflecting horizons. Distinct, isolated, sub-horizontal reflectors and steep dipping reflectors occur in the shallow crust while the mid-crust is often almost acoustically transparent. The lower crust is associated with a diffuse background reflectivity and distinct banded patterns of strong, linear or arcuate dipping reflectors with highly variable spacing. The base of the crust is not marked by a strong Moho reflection as at fast-spreading ridges, but is a comparatively indistinct boundary that is absent in many places.

The origin of this variation in crustal reflectivity is not well understood, but it suggests that important differences may exist in the architecture of oceanic crust created at ridges with different spreading rates.

Imaging Faults Within the Oceanic Crust

The separation of two plates along a mid-ocean ridge is accompanied by both volcanic activity, which creates new oceanic crust, and stretching and faulting of the newly forming lithosphere. The relative importance of faulting is greatest along slowly spreading ridges, like the Mid-Atlantic Ridge. High-resolution bathymetric mapping of the Mid-Atlantic Ridge reveals numerous small fissures and faults on the rift-valley floor and steep, linear slopes or scarps in the rift-valley walls and flanking rift mountains. The depth to which these faults extend in the crust and the amount of relative motion they accommodate are of considerable importance in understanding the structure of crust formed at slow-spreading ridges. The inferred dip and spacing of most faults suggests that they probably do not penetrate more than a few hundred meters into the crust. However, teleseismic earthquakes and microseismicity studies show that some faults rupture the entire thickness of the crust down to depths of 8 to 10 kilometers (see "Mid-Ocean Ridge Seismicity," Oceanus Winter 1991/92).

Multichannel seismic data collected in the western North Atlantic southwest of Bermuda have shed new light on the nature of faulting at the Mid-Atlantic Ridge. This 150-million-year-old crust preserves a record of the volcanic and tectonic processes that created it millions of years ago. The most striking features observed in reflection data from this area are the bands of dipping reflectors that cut through the crustal section. These events occur with equal frequency on lines shot perpendicular or parallel to the spreading direction. They typically dip 20 |degrees~ to the south on perpendicular profiles, and east at about 30 |degrees~ toward the paleo spreading center on parallel profiles although they often flatten out near the base of the crust. While generally confined to the lower crust, in a few cases these reflectors can be seen cutting through the entire crustal section.

Similar reflectors have now been observed in other areas, such as the Canary Basin in the eastern North Atlantic. These dipping events have been interpreted as the subsurface expression of major fault systems that have ruptured the entire crustal section down to depths of 8 to 10 kilometers, and provide strong evidence for the important role of crustal extension and faulting in shaping the crust formed along the Mid-Atlantic Ridge. The most surprising, and still poorly understood, aspect of these data is the occurrence of these events on lines both parallel and perpendicular to the ancient spreading axis. If faulting occurs primarily along ridge-parallel normal faults, as many simple two-dimensional crustal accretion models predict, then these dipping reflectors should only be seen on ridge-perpendicular lines. The complex lower crustal reflectivity revealed in these data may indicate a more three-dimensional fault geometry with fault surfaces dipping both toward the ridge axis and away from major accommodation zones linking major boundary faults. Alternatively, these events may represent two different classes of faults that formed during different stages of the emplacement and aging of oceanic lithosphere. Resolving the true geometry of these events, and their origin, will ultimately require more detailed seismic imaging of the reflectivity and crustal structure.

Imaging Crustal Magma Bodies at Spreading Centers

Along faster spreading ridges, like the East Pacific Rise, the magma supply is higher than at the Mid-Atlantic Ridge, and spreading is dominated by volcanic rather than tectonic processes. Geologists have believed for many years that fast-spreading ridges are underlain by molten reservoirs or crustal chambers that accumulate magma prior to eruption at the seafloor. Based on studies of ophiolites (fragments of oceanic crust found on land), most geologists envisioned these magma chambers as large bodies, up to 10 or 20 kilometers wide at their bases and several kilometers high, filled with melt (see "Onion and Leaks: Magma at Mid-Ocean Ridges" Oceanus Winter 1991/92).

Magma has a much lower seismic velocity than the solid rock surrounding it. At the roof of the magma chamber, the boundary between rock and melt is likely to be quite sharp and should be detectable in seismic reflection data. In the mid-1970s, a Lamont-Doherty group led by Tom Herron recorded a shallow reflector at the East Pacific Rise at about the same depth where refraction data collected earlier by John Orcutt of the Scripps Institution of Oceanography had indicated a zone of low crustal velocities. In 1985 and again in 1991, we and a Scripps group led by Orcutt and Alistair Harding conducted detailed multichannel seismic surveys of two portions of the East Pacific Rise. These surveys confirmed the presence of a shallow crustal magma body at the rise, but showed that it is far smaller than previously imagined.

Reflection profiles across and along the axis of the East Pacific Rise near 14 |degrees~ 14 |prime~ S show several different events. Two flat-lying events can be traced along the rise axis, one about 150 milliseconds (less than 200 meters) below the seafloor and another about 450 milliseconds (some 1,000 meters) below the seafloor. The shallower event, which occurs at the base of a layer that thickens rapidly off-axis, is not a true reflection but is due to refracted energy turning at the base of a near-surface layer characterized by very low seismic velocities. This surficial layer has been interpreted to be either a lava flow layer overlying a sheeted-dike sequence, or a high-porosity layer overlying lower porosities within the extrusive section. The off-axis thickening of this layer may be due to progressive thickening of the extrusive section by lava overlying the axial summit caldera. A remarkable, and unexpected, result revealed by the reflection data is the uniformity in thickness of this lava-flow layer along the rise axis.

The deeper event observed at the rise axis is a reflection off the roof of a shallow magma body about 1 kilometer below the seafloor. This body is quite continuous along the rise axis, but extremely narrow in the cross-axis direction. In this area it is typically less than 1 kilometer wide, and elsewhere along the East Pacific Rise it is never more than 3 to 4 kilometers wide. Analysis of the character of this reflector and coincident refraction data suggest that the predominantly molten part of this magma body is also quite thin--probably less than a few hundred meters thick and perhaps as little as a few tens of meters thick. Thus we have begun to refer to this body as a magma lens or melt sill. Whatever it is called, it is clearly much different than the large, entirely molten magma chambers once envisioned to exist at mid-ocean ridges. While relatively constant in depth along the rise crest, the melt lens is observed to shallow significantly at one location along the southern East Pacific Rise thought to be recently volcanically active. The combined thickness of the overlying extrusive and sheeted-dike section thins by several hundred meters in this area, providing an important new constraint on how the formation of the shallow crust section is related to magmatic activity.

These observations, recent seismic refraction studies at the East Pacific Rise (see "Tomographic Imaging of Spreading Centers," Oceanus Winter 1991/92), and petrological studies of volcanic and plutonic rocks from mid-ocean ridges have led to a new concept of what ridge-crest magma chambers look like. In this new view, the crust beneath the spreading center is mostly solid. The predominantly molten part of the magma chamber is a sill-like body 1 to 2 kilometers below the seafloor. This melt sill is very narrow in cross section (typically less than 1 to 2 kilometers wide) and quite thin (tens of meters of thick), but relatively continuous along the ridge axis. It grades downward into a partially solidified, crystal mush zone that is surrounded by a transition zone of solidified, but still hot and ductile, lower crustal rocks. Mid-ocean ridge magma chambers along fast-spreading or high-magma-supply ridges are thus seen as volumetrically small, composite bodies consisting of both melt and mush that are confined to the mid-crust beneath the rise axis.

This is very different from the old ophiolite-based models that viewed ridge-crest magma chambers as large, well-mixed, steady-state, essentially molten reservoirs. This new composite magma chamber model has broad implications for the composition of mid-ocean ridge basalts and how they vary along the ridge crest, the structure and formation of the lower oceanic crust, the circulation of hydrothermal fluids, and the longevity of mid-ocean ridge hydrothermal systems.

The Future

Multichannel profiling has proven to be a powerful tool for investigating the structure of oceanic crust and geologic processes occurring at divergent plate boundaries, especially along fast-spreading ridges. As more powerful sound sources and receiving systems are introduced and more advanced data-processing techniques become available, we can expect to see an improvement in our images of the oceanic crust, and we may be able to begin to resolve structure within the underlying mantle.

The reflection technique has been less successful in imaging crustal structure at slow-spreading ridges, primarily because of the rugged topography in that tectonic setting. In theory, given knowledge of seafloor topography and an accurate velocity model of the crust, it is possible to produce good images of the crust with appropriate processing; however, these approaches are so computationally intensive that they have not been widely used. With expected advances in computer technology, it is likely that we will see renewed efforts to image crustal structure using reflection methods in areas like the Mid-Atlantic Ridge.

The reflection method is also limited to mapping the seismic response to fairly sharp boundaries in the crust. Hence it is not very likely to tell us much about important questions such as the relative distribution of melt and mush in a crustal magma body or the size of the transition zone to a solidified lower crust. In contrast, seismic refraction and tomographic techniques are well suited for constraining vertical and lateral variations in seismic velocity, but generally lack the spatial resolution of seismic reflection techniques. Increasingly, we expect to see these techniques used in combination in experiments employing arrays of ocean-bottom instruments and dense two- or three-dimensional grids of seismic reflection profiles. The seismic images of spreading centers and oceanic crustal structure obtained in these experiments will undoubtedly lead to new insight into the tectonic, magmatic, and hydrothermal processes responsible for the formation of the oceanic crust that covers two-thirds of our planet's surface.

Robert S. Detrick is a Senior Scientist in the Department of Geology and Geophysics at the Woods Hole Oceanographic Institution. He has been studying mid-ocean ridges for most of the past 18 years and is currently Chairman of the Steering Committee of the U.S. RIDGE Program, a major, decade-long interdisciplinary research effort aimed at gaining a better understanding of the geology, physics, chemistry, and biology of the global mid-ocean ridge system.

John C. Mutter is a Professor of Geology at Columbia University and a Senior Research Scientist at Lamont-Doherty Geological Observatory, where he heads the multichannel seismics group.
COPYRIGHT 1992 Woods Hole Oceanographic Institution
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Detrick, Robert S.; Mutter, John C.
Date:Dec 22, 1992
Previous Article:From the Gobi to the bottom of the North Pacific.
Next Article:Marine seismology.

Related Articles
Connect the dots in the ocean crust.
Magma reservoir seen under ocean ridge.
First look at the base of the plates.
A new mandate for deep-ocean drilling.
Oceanic crust and mantle structure.
DSDP/ODP downhole measurements in Hole 504B.
How underwater earthquakes form.
Exploring volcanoes, earthquakes, and mounting building under the Atlantic Ocean.
Discovery of "megamullions" reveals gateways into the ocean crust and upper mantle.
The big MELT.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters