by Academician Viktor KHAIN, Institute of the Lithosphere of Fringing and Inland Seas, Russian Academy of Sciences

REVOLUTION IN GEOLOGY: PLATE TECTONICS

Late in the 1960s and early 1970s the solid earth sciences ceased to be descriptive and got down to explaining geological processes by means of physical and mathematical laws. The new trend in research - geodynamics - united efforts of geologists, geophysicists and geochemists who focused on the nature of forces and processes occurring on our planet. In particular, those responsible for changes in its outer shells, including the lithosphere and the crust. There came a theory describing the development of the lithosphere as interaction of an ensemble of relatively monolithic plates, or the plate tectonics theory which wrought a virtual revolution in the earth sciences. The new theory relied on novel methods of research: geophysical (seismic, paleomagnetic, geothermal), isotope-geochemical, geological (experimental mineralogy, petrology). Upgraded, these and other innovative techniques (such as deep-water drilling, exploration of the earth from outer space) confirmed the basic principles of plate tectonics, for one, the vast importance of major lateral displacements (shifts) of earth rock.

In keeping with the plate tectonics theory, the lithosphere of the earth falls into a small number of plates, of large and medium size, always in motion with respect to one another. These shifts are of triple nature-the plates may diverge move apart to form gaps that come to be filled with basaltic magma (the process of sea-floor spreading along mid-ocean ridges) or converge move toward one another, with one of them, an oceanic plate, underthrusting another (the process of subduction, "pushing under", on the periphery of continents or along island arcs); or else the plates may slide relative to one another along vertical rifts. Plate shifts obey the theorem of Leonard Euler, a great savant of the 18th century: a relative motion of conjugate points on a spherical surface proceeds along circumferences around some pole of rotation - the projection on the surface of the axis passing through the center of the earth. Lithospheric plates travel along the divide line between the lithosphere, which is a brittle solid, and the underlying asthenosphere, by comparison a weak, viscous and fluent layer. Lithospheric plates are drawn in by its flows triggered by heat convection in the mantle. These flows move from the axes of mid-ocean ridges (zones of sea-floor spreading) to the axes of deep-water trenches on the ocean periphery, toward the subduction zones. Spreading and subduction offset, neutralize each other: as much of the new oceanic crust is born in the zones of spreading, so much of it is consumed in the subduction zones; therefore, according to the plate tectonics theory, the overall mass of the earth stays constant.

This theory not only explained the origin of oceans-postulating the identity of what is called the ophiolites* of folded systems with the

* Ophiolites - a complex of ultrabasic and basic intrusive, effusive and sedimentary rocks thought to be relicts of the oceanic crust of the geologic past and transferred to continental margins. - Ed.

Pages. 83

oceanic type crust, it could also explain the formation of such systems and continents as a whole, formerly described in rather abstract and hazy terms from positions of the geosyncline-orogeny theory.

PLATE TECTONICS IN THE LIGHT OF PRESENT DATA

Since 1967 - 1968, when fundamentals of the plate tectonics theory were made public, new facts have come to light to confirm this theory. It has been gaining ever new terrain in geology and earth sciences in general. Yet at the same time definite bits of evidence that, though augmenting this theory, did not agree with some of its postulates. Actually certain facts were available at the very beginning, when plate tectonics was still in its cradle - e.g. intraplate volcanism and magmatism in general. This did not fit the plate tectonics pattern and thus brought to life a hypothesis on "hot spots", that is the upward mantle jets (plumes) rising from the lower depths of the mantle and kind of burning, piercing through the crustal plates that keep in motion under the effect of convective flows. However, the interdependence of these two essentially different types of mantle jets - advection (plumes) and convection - was still obscure.

There were other weak points, too. First, the mechanism of plate displacement was not explained well enough. It involved something more than the asthenospheric drag for that matter. At least another two mechanisms must be implicated. One, the dislocation of oceanic plates from the axes of sea-floor spreading in mid-ocean ridges under the effect of gravity pull when these ridges rise, upthrust above submarine plains, and when basaltic magma rushes into rift crevices. The other mechanism involves the drawing of oceanic plates into subduction zones in deep-water trenches, for these plates gain in weight during their departure from the axes of sea-floor spreading. They cool during this movement, while the crevices come to be "patched up" by minerals sedimented from sea water.

The fact of the dislocation mechanism is corroborated by the orientation of compressional forces in plates on either side of spreading within oceans and on continents. These forces come into play in seismic foci and can be measured directly in prospecting holes.

But there is a deeper, in the literal sense, cause of all these processes, and this is the mantle convection. For a long time the very idea-put forward all the way back in the early 20th century-lingered in the shadows as abstract and speculative. For this reason the very concept of plate tectonics was said to be but kinematic.

The tables were turned in the 1980s with the appearance of seismic tomography which helped to spot heated and cooled areas alternating in the mantle down to its boundary with the earth core. The reality of convection and subduction of ocean plates was proved by actual evidence. One of the chief arguments of plate tectonics critics was that the earth mantle had not been studied well enough.

According to seismic tomography data, the distribution balance for convective flows, zones of spreading and subduction registered next to the terrestrial surface does not go deeper than 300 - 400 km. It came out that the plate tectonics scenario extends only to the crust and upper mantle, or lithosphere and asthenosphere (justly unified into the notion of a tectonosphere). Other, different processes occur lower down, which means that plate tectonics is global only in two dimensions, but not in a third, plutonic one.

NEW VISION OF MANTLE PROCESSES

Clearly, there are limits to the effect of plate tectonics in a fourth, time-related dimension, i.e. in retrospect. The complex of present-day geological processes, subduction in particular, must have taken body and form only as of the Late Proteozoic, or a billion years ago. Earlier, as far back as 3 bn years ago, the plate tectonics mechanism was different to what we are having today. But we cannot tell what this mechanism was like earlier than 3 bn years ago, its very existence is called in question, though some of its elements were there.

The authors of the plate tectonics theory in its original form did not allow for the periodic intensification and weakening of the earth's endogenic (inner) activity, nor did they consider essential restructurings in the lithosphere's plate architecture. It is in these past ten years that we have learned about periods in the history of the globe when all its continents came together into one supercontinent and then broke apart to give rise to secondary oceans. This pattern recurred every 500 - 600 mn years. Such megacycles must have been due to changes in the character of mantle convection. Drs A. Monin and D. Sorokhtin suggest that during the formation of supercontinents unicellular convection substituted for the bi- or multi-cellular one, with the ascending flows being under the oceanic (western), and the descending-under the continental hemisphere.

Thus, although in these twenty-five years the plate tectonics theory has offered a satisfactory interpretation for most of the tectonic and endogenic processes taking place in the upper layers of the earth in the last billion years, it cannot be regarded as a universal, coverall theory.

There are several reasons for that. First, one could not explain the phenomenon of intraplate magmatism allowing for the other form of heat-and-mass transport deep in the bowels of the earth in the form of advection of mantle jets, the plumes. Second, plate tectonics is at work only in the upper solid shells of the earth, its tectonosphere, and it went into full swing only in the Late Proterozoic. The classical concept of plate tectonics also left out the periodicity of the endogenic activity of the earth and its restructurings.

Using new research methods, geologists and geophysicists in this past decade have zeroed on two key areas: plutonic geodynamics and early history of the earth. Now, plutonic geodynamics is concerned with pro-

Pages. 84

cesses in the middle (to a depth of 400 - 670 km) and upper mantle, and also in the earth kernel; such studies rely on methods of seismic tomography, experimental mineralogy and mathematical modeling. Researchers are mostly involved with the upper (670 km) and lower (2,900 km) boundaries of the lower mantle. The upper boundary is a divide between the interacting middle (transitional) and the lower mantle. Here the main question is how "permeable" this boundary is for convective flows, subduction and ascending, upward mantle jets, the plumes.

Canonical seismology visualized the subsidence of oceanic lithospheric plates only as deep as the roof of the lower mantle, for no seismic foci were observable deeper than that. Seismic tomography, however, enables explorers to track down the plates submerged in subduction zones underneath this boundary as well. But research findings thus obtained have proved ambivalent. In some cases (off the islands of Japan for example) submerging lithospheric plates curve on reaching this boundary in the direction of their slope. In other cases a subducted plate sinks below the 670 km boundary where it bends and curves as if coming up against the underlying mantle matter. This is what we see opposite the island arc of the Tonga Archipelago in the Pacific. And quite recently, in the Zonde trench (Malay Archipelago) off North America, lithospheric plates were found to subside as deep as the mantle/core boundary. This means that the 670 km boundary, even though being an obstacle to all-mantle convection, is overcome under definite conditions. It is quite possible therefore (as French researchers and RAS corresponding members P. Manchetel and P. Weber first noted), there is an alternating pattern to convection enveloping the entire mantle but effective only within its upper or lower layers. But even allowing for such segregated convection modeled by Drs A. Dobretsov and A. Kirdyashkin of Novosibirsk, lower mantle flows should induce (possibly opposite in sign) similar movements of matter in the upper mantle as well, as shown by mathematical modeling carried by Drs P. Pushcharovsky, V. Fadeyev and coworkers at the RAS Institute of Geology.

No less intriguing is another plutonic boundary, that between the earth mantle and the core. Prominent here is the border layer D, 200 to 300 km thick. Here heat-and-mass transfer is probable between the cooler silicate-oxide lower mantle and the molten metallic (iron and nickel in the main) central kernel. This confirms the supposition that it is from these depths that mantle jets, the plumes, rise to the lithosphere and even to the surface of the earth.

This is especially true of superplumes which generate "hot spots" and even whole "fields" in the lithosphere (as described by L. Sonnenschein and M. Kuzmin, RAS Corresponding Members). One such "hot field" appeared in the west-central Pacific in the middle of the Cretaceous to form the submarine Darwin Rise. The fact that this event concurred with a long period when inversions of the earth's magnetic field happened to be absent points to a connection between magnetic inversions and origination of plumes at the mantle/core boundary, though the mechanism of such dependence is interpreted differently. Acad. Ye. Milanovsky has identified a correlation between these events and phases of enhanced tectonic activity. But not every plume is of plutonic origin- some may be born also at the 670 km boundary from submerging lithospheric plates that halt there for a time, as Dr. A. Ringwood of Australia sees it.

AT THE DAWN OF THE EARTH'S HISTORY

Now what concerns another research trend, the early history of the earth. Its rapid progress was spurred above all by achievements in isotope geochronology capable of dating the age of rock for epochs of more than 3 bn years ago to within the first few million years.

The presence of typical ophiolites in some of the old shields of the Lower Proterozoic as well as festoon island volcanites and granite batholiths is clear evidence for plate tectonics as it was manifested in the early history of the earth. But other tokens like high-pressure and low-temperature metamorphites of such age (and older) are not known yet. Apparently they were replaced by more high-temperature rock. Plate tectonics of the Early Proterozoic differed from that of later epochs in a larger number of small crustal plates and, consequently, in longer axes of sea-floor spreading in between.

Granite-greenstone regions were formed at the end of the Archean, or 2.7 - 2.5 billion years ago, from the consecutive accretion of numerous island arcs (festoon islands) at the cores of the older protocontinental crust built up by gneisses of the Early Archean. Consequently, given that the tectonics of the Early Proterozoic used to be "multiplate", the Archean one could be designated as that of "multiple island arcs". Geodynamic processes could be essentially different in the Middle Archean, not to speak of the Early (Lower) Archean. The volcanism of the greenstone belts of the Middle Archean, characterized as it was by basic and acidic volcanites in the absense of intermediate andesites, was of riftogenic rather than island arc origin. None the less even the so-called gray gneisses of the Early and Middle Archean aged more than three billion years look quite like magmatites formed under the zones of subduction of the young oceanic crust. We may admit therefore that as early as the beginning of the Archean, i.e. about 4 bn years ago, some of the features of plate tectonics were already there.

New research findings on abyssal geodynamics and on the early geological history of the earth have prepared the ground for incorporating plate tectonics into a more general global model.

GLOBAL GEODYNAMIC MODEL ON EARTH...

In an article published by the journal Doklady AN (Proceedings of the Academy of Sciences), I listed princi-

Pages. 85

pies that should underlie a genuinely global geodynamic model. This is above all the multilayered structure of the earth, convection restricted to particular shells (layers) alongside the mutual impact of convective processes in adjacent geospheres and last, the recurring pattern of change in their interaction.

The year 1994 saw a full-scale global geodynamic model designed as a substitute for plate tectonics with all its limitations. It was suggested by a group of Japanese geophysicists and geologists who published a series of related articles in the journal of the Japanese Geographical Society that had marked its birth centenary. Its authors (Drs S. Maruyama, M. Mumazawa, S. Kawakami et al.), though not refuting the plate tectonics theory, pointed to its limitations. They proceeded from the assumption that the solid earth is in three spheres-the crust and the upper mantle (tectosphere), the lower mantle and the central core. In each of them geologic processes occur differently. Plate tectonics applies only to the tectosphere, whereas in the lower mantle, the authors argued, plume tectonics is predominant, while the core is characterized by "growth tectonics" expressed in the growth of the inner core at the expense of the outer one. The chief process, they said, is in the subsidence of the cooled lithospheric plates into subduction zones (earlier Dr. L. Sonnenschein pointed exactly to that). Upon reaching the upper boundary of the lower mantle, a subducted plate is arrested for a time, which results in the accumulation of matter; and when a sufficiently large amount of it is built up (which takes 500 mn years or so), the lithospheric matter sinks into the lower mantle, down to the core boundary. This mass interacts with the constantly convecting matter of the core to cause a powerful ascending plume on its other side. At the 670 km boundary this superplume breaks into several stripes which give rise in particular to the axes of spreading of mid-ocean ridges. Such is the scenario of transition from plume tectonics to plate tectonics. By seismic tomography data, today there are two super-plumes - one in the south Pacific, and the other - in East Africa. These are offset by a descending cool plume under the central part of the Asian continent. The axes of spreading, arising first under the "filial" plumes in the tectosphere (in the Atlantic for example) may move away then, as seen in the example of the Indian and Pacific Oceans.

This transition from plume to plate tectonics began as early as the beginning of the Archean. But the earliest process to take place before the Archean was the "growth tectonics", or the formation of the earth core. The segregation of this central kernel into inner and outer parts could be completed by the onset of the Proterozoic and even later. Close to the earth surface and above the growing core was a "magmatic ocean" formed under the action of heat from solar-lunar tides (proximity of the moon was a contributing factor) and the greenhouse effect of the primordial atmosphere.

...AND ON OTHER PLANETS...

Drawing upon comparative planetary science data over the recent years, the authors of the new theory apply it to the general evolution of the terrestrial planets. They say that all the planets had to go through the "growth tectonics" stage at the very start, i.e. during core accretion and formation. Venus apparently is living through the stage of predominant plume tectonics with initial elements of plate tectonics already present.

Our planet must have passed a stage like that in the pre-Archean times, while now it is at the stage of plate tectonics, with plume tectonics processes on in more abysmal depths. As the Japanese scientists see it, Mars and Mercury have entered the next stage, that of contraction tectonics, when the lithosphere solidifies into a rigid shell which does not break into plates but contracts because of the cooling interior. At this stage, however, the mantle may melt in part, with magma upswelling to the surface in large volcanic structures similar to those observed on Mars.

The author of a work that came out in the United States after the publication of the Japanese model postulates, on the basis of Martian relief studies, that Mars has already gone through the plate tectonics phase. This proves the validity of an evolutionary series outlined in the Japanese model, with the moon and smaller planets placed at the end as being in the stage of terminal tectonics, when exogenic factors supersede endogenic activity which is confined only to the formation of rifts giving vent to volatile substances released in the process of cooling and the tidal heating of deep mantle matter.

Such are the main points of the geodynamic approach suggested by the Japanese scientists who have made a major stride toward a genuinely global model of the earth and the terrestrial planets of the solar system in general two hundred years after the Scotch savant James Hutton, one of the founders of the science of geology, published his Theory of the Earth, With Proofs and Illustrations (1795). In some respects the Japanese model is not quite waterproof. For one, the Japanese scientists do not view the earth as an open system acted upon by cosmic factors, moving as it is along the galactic orbit. But every new theoretical model is a reference point for further studies. Next on the list is a critical analysis and further perfection of the suggested geodynamic model.

Zemlya i Vselennaya (Earth and Universe), No. 5, 1995

Permanent link to this publication: