by Anatoly NIKISHIN, Dr. Sc. (Geol. & Mineral.), Head of the Chair of Regional Geology and History of the Earth of the Geological Department of Moscow State University

Today we know reasonably well what the geological structure of our planet Earth is like. Owing to progress in Soviet (Russian) and American space programs, we have at our disposal detailed information on the terrestrial planets. Proceeding from it, we may try to understand how the Earth was formed, and look into some of the stages of its history.

First, let us recall what is known for certain about the structure of our planet. The hard Earth with a radius of 6,378 km consists of three chemical mantles, namely, a mostly iron core in the center (depth, 2,890-6,378 km), a silicate mantle formed of ultrabasic* magmatic rocks (35-2,890 km) and a crust formed of silicates with a composition ranging from basic to acid (0-35 km). On the surface, the lithosphere, there is a stone "mantle" of 50-300 km thick, hard contrary to the underlying asthenos-phere and, therefore, able to move in the horizontal. Besides, the bottom of this drifting "iceberg" passes approximately in the isotherm of 1,250°, i.e. deeper its heated substance may be in for convection.

The earth crust under the oceans and continents (together with shelves) is not uniform. In oceans its thickness is around 6 km, it has a basalt composition and is formed in the axial parts of mid-oceanic ridges, in areas of lithosphere plate shifting as the occurring cracks are filled with mantle melts. The continental crust has a typical thickness of 25-60 km and an andesitic** composition by and large. Besides, its upper part is acid (granitoid) and the lower part, basic (basalt). Its substance is formed by mantle components with the partial melting of basalts.

Nowadays the majority of specialists agree that the lithosphere plate tectonics* is active on the Earth. The foundations of this theory were formed by many geologists and geophysicists beginning with the works of the German scientist Alfred Wegener (1910-1912), Canadian Tuzo Wilson (1965), American Jason Morgan (1968-1972), French Xavier Le Pichon (1968) and Englishman Dan McKenzie (1967). In Russia, a tangible contribution to the development of this science was made by RAS Corresponding Member Leo Sonenschein (Institute of Oceanology of the Russian Academy of Sciences), Leonid Parfenov (Institute of Geological Sciences of the RAS Siberian Branch), both Drs. Sc. (Geol. & Mineral.), Academicians Viktor Khain (Geological Institute of the Russian Academy of Sciences, Moscow), Nikolai Dobretsov (Institute of Geological Sciences of the RAS Siberian Branch) and others. Scientists believe that the Earth lithosphere is broken into plates (about 15 in all), moving relative to one another. In areas where they diverge, an oceanic crust and oceans are formed, and where they draw together, zones of deep-water trenches are formed, with the lithosphere submerging into the mantle (subduction zones).

* Silicate ultrabasic rocks contain below 45 percent SiO₂, basalt, 45-52 percent, and acid ones above 70 percent SiO₂.--Ed.

** Andesite (from the name of the Andes range in South America), a volcanic rock of mean composition and normal alkalinity. SiO₂ content is 52-65 percent.--Ed.

* See: R. Balandin, "Cognition of the Ocean: a New Pace", Science in Russia, No.3, 2011.--Ed.

Model of the Earth's structure at the end of its accretion (A) and the assumed mineral composition of the mantle and the crust formed in course of the crystallization of the 2,000 km-thick magma ocean (B).

According to Bkins-Tanton, 2008.

FROM THE PLANET'S FORMATION ON TO PLATE TECTONICS

The model of the geological evolution of the Earth develops continuously, and this problem is discussed in a vast number of publications. But the following scenario appears to be the most realistic today.

About 4.6 bln years ago in the course of accretion (a process whereby outer space substance falls to a cosmic body) the Earth was formed from a swarm of bodies of silicate and metal composition, gases and comets. The time of its formation is estimated at 10-100 mln years. It increased in size as more and more new parts of different size joined the initial base. Starting from some radius there came a high heating and melting of outer zones. The planet became red-hot, its substance approached the melting point, the outer shell of hundreds of kilometers thick melted to give rise to the so-called magma ocean. In these conditions the solid metal components went under and formed a predominantly iron liquid core, while the silicate components became a mantle. Most of the water was not incorporated into the semimolten body of the Earth, but became part of a dense atmosphere of vapor and gases (mainly carbon dioxide and nitrogen).

At the end of this process of accretion a massive cosmic body is believed to have hit the Earth to make a great crater on its surface as deep as its present-day mantle to cause its heavy ejection into the terrestrial orbit. Most probably, the Moon was formed just from this ejected hot melt, as its mean composition and the ratio of components within the earth mantle are alike.

But let us retrace the history further. About 4.4-4.3 bln years ago the terrestrial surface started cooling due to a considerable decrease in the number of cosmic bodies hitting our planet, and the magma ocean was being covered with a solid crust (possibly of basalt and ultrabasic composition), on which the continued meteorite bombardment left numerous craters of different size, with lavas erupting in areas where asteroids fell. As soon as the temperature of the planet's upper shell fell below the boiling point of water, an immense amount of water was condensed from the atmosphere, and the World Ocean was formed. A part of its moisture was liberated from the mantle during its cooling and crystallization. As a result, about 4.3 bln years ago conditions were available for the origin of primitive forms of life. It is not clear*, whether it was imported in one way or another from outer space (this is much more likely) or it originated in the course of accidental chemical reactions on the Earth.

The plate tectonics process started about 4.3-4.0 bln years ago owing to the combination of different factors, such as convection currents in the mantle, the floating to the surface from its lower layers of bodies with a relatively high

* See: M. Fedonkin, "Modem Science About the Origin of Life", Science in Russia, No. 6, 2011.--Ed.

temperature and a lower viscosity (plumes), the existence of lavas of the ultrabasic composition (komatities) creating consolidated sections capable of sinking into the asthenos-phere (the mantle layer of a reduced hardness and viscosity). This can be regarded as a precursor of the subduction* process. One way or another, it is just then that the lithosphere of the Earth broke into plates, with a young oceanic crust being formed in areas of their separation, and zones of crowding were isolated in areas of convergence. Above them, magma of the andesite composition was being formed as a framework for the crust of future continents.

In the Archean (4.0-2.5 bln years ago) plate tectonics and mantle plumes were predominant. The primary crust sunk completely into the mantle or was remelted. Volcanic arcs with basalts and andesites were formed above the subduction zones. Their collision gave rise to the first sections (cores) of the continental crust. At that time at least 50-70 percent of the substance of modern continents with shelves was formed. The mantle was hotter than the present one, and the lithosphere sinking into it heated up quickly without penetrating the lower layers. A great number of lavas of the ultrabasic composition, or komatities, melted out of the red-hot mantle. The Earth surface was mostly in the form of a water ocean with a crust of the basalt-komatite composition and numerous protruding volcanic arcs and uplands (both intraplate and interplate).

Primeval life was born in the ocean. For example, isotopes formed in the biological cycles are known in rocks of ~3.8 bln years of age, and traces of the biota are found in rocks younger than 3.5 bln years. Large continents (possibly one, the first supercontinent Kenorland) had been formed by the end of the Archean, or 2.5 bln years ago, in the course of collision of numerous continental "cores" with the continental crust and large-scale granitoid magmatism. This scenario is supported by the age of the earliest lacustrine-alluvial systems observed in South Africa (Witwatersrand series) and estimated at 2.8-2.9 bln years. But what about the climate? By that time it was comparable to the Cainozoic one. If in the early Archean the ocean water was hot, then the period of 2.4 bln years ago is noted for the first known global Hudsonian glaciation.

In the Proterozoic-Phanerozoic (2.5-0 bln years ago) the general cooling of the Earth caused the substance sunken into the mantle (subducted) to reach its lower layers, thus ordering the whole system of convection, and promoting long-term tectonic supercontinental cycles. How did it happen? The continents were breaking apart and coming together, the oceans now opening, now closing. But against the apparent chaos the relatively stable supercontinents were formed every 750 mln years. The composition of the oceanic crust changed from komatite-basalt to basalt, and its thickness decreased. Mountains with their maximum height gradually increasing were being formed constantly in zones, where the continents collided and also on their active margins.

Meanwhile life was in progress. Let us note the most important milestones of its evolution. About 2 bln years ago came cells with nuclei, and multicellular organisms; 630-600 mln years ago (the Vendian period) saw the wide occurrence of multicellular marine askeletal organisms; 542 mln years ago (beginning of the Cambrian period) there appeared skeletal animals of different types; and the terrestrial flora and fauna emerged in the Silurian.

In the last 2.5 bln years the climate has been relatively stable. The epochs of global warming were followed by the shorter periods of global glaciations and vice versa (0-10, 290-325, 445-455, 625-700 mln years ago and earlier). The atmosphere in the late Proterozoic (as of 1 bln years ago)

* Subduction, a sinking of lithosphere plates and the oceanic crust into the Earth mantle to the depths of hundreds and more kilometers.--Auth.

Model of the Moon's origin: A-a big cosmic body hitting the Earth; B-a new planet formed from the Earth mantle ejected by impact to its orbit.

The assumed section of the Earth about 4.3 bln years ago and a schematic representation of how the first water ocean was formed.

came to contain appreciable amounts of oxygen and in its composition was approaching the present one.

OCCURRENCE OF WATER

It is all-important to realize, wherefrom water has appeared on the surface of our planet. We have already mentioned two theoretically possible alternatives widely discussed in the literature, namely, deposition from the ancient atmosphere upon the completion of the hot accretion of the Earth, and emission to the atmosphere from magma, above 2,000 km thick, with its crystallization. Let us go into these two scenarios, as the formation of the water ocean and the development of life in it predetermined in many ways the further geological history.

Thus, in theory the mantle substance can contain water. For example, in the transitional area (depth, 410-670 km) the content of H₂0 reaches 2-3 wt% and in the lower layers it is equal to 0.15 wt%, but according to different experts, the maximum mass of moisture is equal here to 2.5 masses of the World Ocean. Furthermore, in the course of accretion water could not be a part (at least, in large quantities) of the hot planet, but remain, to a great extent, in the atmosphere brought in initially in the bodies of ice comets and meteorite components. And we may well assume that water came to the ocean from the atmosphere and the mantle.

According to geological data, water appeared on the earth surface more than 4 bln years ago. For example, as shown by an isotope analysis of δ¹⁸O and Hf (hafnium), 4.1-4.3 bln year-old zircons* were formed in rock with the implication of H₂0. A submarine accretion complex was formed above the subduction zone from the earliest Isua (western Greenland) sediments on our planet, which have been in existence for 3.8 bln years and containing organic carbon. Therefore, the plate tectonics and the ocean with insipient life were already there. The available models prove that 4 bln years ago the volume of water reached its all-high in the history of the Earth.

The primary crust, which separated during magma ocean crystallization, has not preserved anywhere (though the origin of 4.0-4.4 bln year-old detrital zircons is not clear; some of them were possibly formed in granitoid bodies of the crust), but it is clear that it could not have large tectonic depressions of the oceanic basin type. Deep basins occurred only in impact depressions (giant meteorite craters). But plate tectonics changed the situation and created a reservoir for a water ocean. However, the time of its origin is still debated. When we described the scenario of the Earth's geological history, we mentioned its

* Zircon, zirconium silicate (ZrSiO₄). This mineral contains, as a rule, 1-4 percent hafnium, which substitutes isomorphously for zirconium in a crystal lattice.--Ed.

time-about 4.3-4.0 bln years ago. But what evidence argues for this point of view?

PLATE TECTONICS: REFERENCE POINT

The Isua sedimentary and volcanic complex was formed 3.8 bln years ago during the sinking of the oceanic crust in the frontal part of the overlying plate (specialists call it the situation of an accretion prism). Isotope and geochemical analysis data on rare-earths show that in the basic and ultra-basic rocks the ratio of Nd/Th (neodymium/thorium) and Th/U (thorium/uranium) elements changed dramatically about 3.6 bln years ago, which is indicative of the presence of cold and heavy edges of lithosphere plates subject to subduction at that period. By geochemical evidence as early as 3.5 bln years ago there were cratons, or sections of the continental earth crust not subject to considerable folded deformations, but having thick lithosphere "roots" (foreign colleagues use a poetical name, "mantle keels"). Thus, already at that time the process of lithosphere thickness doubling was active as one lithosphere plate underthrust another one.

Some scientists (American geologist Kent Condie, Prof. Alfred Kroner of the University of Mainz, Dr. Alexander Slabunov from the Institute of Geology of the Karelian Scientific Center of the Russian Academy of Sciences and Dr. Andrei Shchipansky of the Institute of Geology of the Russian Academy of Sciences) brought together data on the time of origin of plate tectonics and proved that the age of the earliest structures interpreted as a consequence of that process exceeded 4 bln years. For example, if the typical ophiolites (rock units recognized by all scientists as formations of the oceanic crust) are less than 1 bln years old, then their Archean analogs, such as parts of the Isua and Nuwuagituk complexes (vicinity of the Hudson Bay), should be about 3.8 bln and 4.3 bln years old, respectively.

Paleomagnetic studies demonstrate that the displacements of cratons relative to one another took place 2.4-2.7 bln years ago (accurate data on earlier periods not available). By geochemical evidence, the temperature of the Archean mantle was by 100-250 °C higher than that of the modern mantle and consequently, the character of plate tectonics at that time differed from that of the Phanerozoic period. Besides, as late as 3.45 bln years ago there was a magnetic field with the intensity of 30-50 percent lower than that of the present one. Thus, on the whole the general geodynamics of the Earth was similar to what it is now.

Finally, according to numerical computer modeling data, lithospheric subduction is impossible if the temperature of the upper mantle is by 250 °C above the present one. In other words, this process could not happen physically on the early Earth soon after its accretion.

Thus, the beginning of plate tectonics on the Earth dates back to roughly 4.0-4.3 bln years ago, but we cannot tell

Model of the active start of plate tectonics (A). The passive start of plate tectonics (B).

Section of the typical Phanerozoic oceanic crust and the assumed section of the Archean oceanic crust (with account of data on the structure of the Mesozoic Ontong Jawa oceanic plateau). Arrows indicate the probable levels of strippings (detachments) during compressive deformations.

how it happened. There are two possible scenarios. Some researchers admit that this process started as a single global short-time event, while others consider it improbable since, they say, it originated gradually in different places. There are yet another two versions explaining the plate tectonics phenomenon, namely, the active and the passive one. According to the first version, the mantle convection caused disintegration of the single lithosphere plate and relative displacements of its numerous fragments with the formation of the first linear oceanic depressions. In keeping with the second version, a large amount of heavy komatite magma within the crust increased the weight of the lithosphere which, cooling, sank into the mantle and thus triggered the plate tectonics mechanism.

Accordingly, it would be interesting to trace changes in the composition of the organic crust in course of its evolution. We know that the age of the earliest ophilites similar to those of the Cainozoic period is about 1.95 bln years (like the Jormua complex in Finland, for example). On the other hand, it is postulated that the oceans were there already in the Archean period. In recent years it is commonly agreed to consider sections of many Archean greenstone belts to be fragments of the oceanic crust formed by that geological period. In the Archean period it had the same thickness as the Cainozoic oceanic plateau of the Ontong Jawa* type or the Iceland upland and was formed from the hotter mantle than that of the present-day Earth.

SUPERCONTINENTAL CYCLES

But let us go back to the Proterozoic-Phanerozoic periods, as a vast bulk of scientific literature and many active international projects are devoted to the supercontinental cycles of that time. The gist of the contemporary concept is that in the last 2.5-2.7 bln years the available continental masses have now been coming together, and now a super-continent thus formed has been breaking into multiple small continents. Today there is a sound argument supporting the existence of the protocontinent Pangaea with an integration maximum on the border of the Permian and Triassic periods (about 250 mln years ago). Many authors prove that the Rodinia supercontinent existed on the Earth with a great degree of probability about 1 bln years ago. There are fewer arguments in favor of a similar giant called Columbia about 1.8 bln years ago. And much more modest is the geological evidence on the putative Kenorland supercontinent of 2.5-2.7 bln years ago. If such supercontinents had really been in existence, the duration of a supercontinental cycle could be estimated at 750 mln years. The formative time for those giant continents coincides with the epochs of the maximal accretion of the new continental crust.

Supposing the hypothesis of supercontinental cycles is valid. In this case, in the last 2.5 bln years Earth tectonics

* Ontong Jawa, an atoll in the northern part of the Solomon Islands. Its name was given to a large submarine plateau.--Auth.

Model of present (A) and the Archean (B) subduction and formation of the Archean continental crust.

A shorthand of the Earth's geological history.

has been comparatively uniform, with a newly formed supercontinent breaking apart gradually. Consequently, the global ocean broke into scores of relatively small oceanic bodies divided by continents. Such a scenario was repeated many times, with the volume of the continental crust increasing little by little; meanwhile the total area of the continents kept expanding, and the water area kept contracting.

What with the general cooling of the Earth in the Proterozoic and Phanerozoic, it would be natural to postulate a gradual decrease of temperature in the mantle and, consequently, also a definite role of plume magnetism: the significance of the oceanic plateaus with a thick oceanic crust was decreasing, with the mean ocean depth increasing.

Today scientists do not suppose that the Earth witnessed continental supercycles in the Archean period: at that time, under hot mantle conditions; subduction was not deep, a submerged body did not sink into the lower layers, and general mantle convection was still absent. Its absence caused a more chaotic plate tectonics than in the Proterozoic and Phanerozoic.

Thus, the present geological evidence does not contradict the following theory: the Earth was born in the course of a hot accretion, and its whole history (for one, the appearance of water and of the World Ocean) was connected with the cooling of the planet's body. About 4-4.3 bln years ago, due to incipient lithosphere plate tectonics, the oceans and the cores of future continents appeared, and a new type of crust was taking form. However, in the Archean the ocean area was still predominant, the sinking of plates into the mantle stopped in its upper layers, and the phenomena related to the movement of mantle plumes came to play a larger role in Earth tectonics. But in the Proterozoic and Phanerozoic supercontinental cycles became predominant. Moreover, the Archean oceanic crust on the whole differed from the Phanerozoic one in its greater thickness and more ultrabasic composition. Therefore, the typical depth of the oceans probably did not correspond to the modern oceanic depth either and made up about 1-3 km.