FUNDAY

664 New Concepts in Global Tectonics Journal, V. 4, No. 4, December 2016. www.ncgt.org

A History of the Earth’s Seawater: Transgressions and Regressions
Karsten M. Storetvedt
Institute of Geophysics, University of Bergen, Bergen, Norway
karsten.storetvedt@uib.no
“We do not have a simple event A causally connected with a simple event B, but the whole background of the system in which the events occur is included in the concept, and is a vital part of it”
P.W. Bridgman, in: The Logic of Modern Physics, p. 83
Synopsis
The origin of sea water is viewed in the context of the model of a slowly degassing Earth – an Earth which is more than likely being far from having acquired thermo-chemical equilibrium. The internal reorganization of the Earth’s planetary mass has led to changes of its moment of inertia and thereby its rotation characteristics – giving rise to variations in spin rate as well as spatial redistribution of its mass (producing the dynamic phenomenon of true polar wander). These dynamical pulsations are seen as the engine behind the spasmodic behaviour of the principal global geological phenomena. In the continuing restructuring processes of the Earth’s interior, water has been added incessantly (but episodic) to the surface, while hydrous fluids have also played a central role in breaking down the original thick pan-global crust – progressively forming ever deeper ocean basins. Vertical uplifts and subsidence of the evolving deep sea crust, in association with continental transgression-regression cyclicity, are natural consequences of a slowly degassing Earth. Thus, the unceasing transformation of the felsic crust is intimately tied to the history of seawater – along with a multitude of first-order geological, geophysical, environmental and biological occurrences.
Keywords: Origin of seawater, sea-level changes, dynamic drivers, geological history
(Received on 6 December 2016. Accepted on 31 December 2016)
Problem outline
o a large extent, the history of the Earth’s dynamo-tectonic development is related to the origin of the oceanic water masses and their surface oscillations – characterized by the advances and retreats of epicontinental oceans. During major parts of post-Precambrian time, the present land surface was extensively covered by shallow seas, while today the continents are dryer than at any time during the last 570 million years (Phanerozoic). During the late Mesozoic, the continental flooding was nearly as widespread as that of the Lower-Middle Palaeozoic, though the highest sea-level may not have been higher than 200-400 metres above the present shore line (cf. Miller et al., 2005). “Today, a similar rise would inundate less than half the area that was flooded in the Cretaceous, because our continents stand high above the sea, whereas the Mesozoic lands were low and flat” (van Andel, 1985). The same low and flat continents were apparently the norm during the Palaeozoic as well as in Precambrian time; the elevation of our continents and continental mountain chains, as well as the mid-ocean ridges, seems to have a quite recent origin – having basically occurred during the last 5 million years of Earth history (cf. Storetvedt, 2015 for references and a compilation of evidence).
Fluctuations in global sea-level result either from changes of the volume of sea water on the planet’s surface or of deep sea basins. Today, the growth and decay of major continental ice fields are thought to be the most likely causal mechanism for sea-level changes, but vertical oscillations of the deep sea crust, caused by variable rates of hypothesized seafloor spreading, has also been proposed – but without real success. More recently, attempts have been made to relate sea-level changes to climatic control (Miller et al., 2005; Zhang, 2005), but none of these ideas seem to account satisfactorily for the large number of recorded sea-level changes – with time scales varying from long-term super-cycles over hundreds of millions of years to rapid changes in the order of tens of thousands of years or less. Modern compilations of Phanerozoic sea-level trends have been given by a number of authors (e.g., Haq et al., 1987; Hardenbol et al., 1998; Haq and Shutter, 2008), but the description the overall sea-level pattern has not changed significantly since the early work of Sloss (1963) and Vail et al. (1977).
It is natural to think that seawater is intimately associated with the Earth’s internal chemical reconstitution and degassing, but when did the bulk of surface water accumulate? During Precambrian times, there is no factual evidence for the existence of deep sea basins, and the volume of surface water was apparently modest – but there is ample evidence of deposition in shallow marine waters within greenstone belts (Figure 1) along with indications of fluvial activity (cf. Windley, 1977). Perhaps an appropriate description of the surface conditions in the late Precambrian can be unveiled by the Grand Canyon sedimentary system – as described by Dunbar (1949, p. 93-94):
“The Grand Canyon system is essentially unmetamorphosed, thus contrasting in the most striking manner with the underlying schists, which must be vastly older. The system begins with a basal conglomerate resting on a peneplaned surface of the Vishnu schists. Following this come limestone and then limy shale and sandy shale and quartzites. The limestones, and probably a larger part of the shales and quartzites, were deposited in shallow marine water, but parts of the sandy shale and sandstone are bright red and are so commonly mud **** as to suggest deposition on a broad floodplain. The region was probably part of a great delta plain in which submarine and subaerial deposition alternated. And since these strata were formed near sea-level, the region obviously subsided slowly [...] while deposition was in progress.”
The Archaean aeon, which was characterized by features such as the relative abundance of komatiite extrusions and a relative scarcity of redbeds and carbonates, was succeeded by the much more diversified geological record of the Proterozoic – progressively distinguished by large sedimentary basins with primitive living forms more abundantly recorded in surface carbonates. Contrasting strongly with the Proterozoic situation, the Cambrian experienced a general “transgression onto cratons, with a classic orthoquartzite-to-shale sequence resting unconformably on Precambrian and overlain by carbonates” (Hallam 1992). And suddenly, a diversity of complex life forms, dominated by trilobites and brachiopods, appear in abundance at the base of the Cambrian; this remarkable biological explosion was probably a direct consequence of the rapidly increasing volume of surface water. Thus began a major Lower-Middle Palaeozoic submergence of the apparently flat and low-lying continental masses that was accompanied by a rapid development of sea-living creatures. According to Dunbar (1949, p. 155), “the Early Cambrian oceans seem to have been somewhat openly connected, so that intermigration was easy and the leading types of life are much alike in various parts of the world”. Thus, the Cambrian eustatic transgression probably represents the first major supply of water to the Earth’s surface – degassed from the interior of the Earth; the explosive prevalence of marine fauna at that time is likely to have been a consequence.
Figure 1. Depiction of a transect across a block of late Archaean greenstone belts – subsiding rift basins developing along one of the pre-existing orthogonal fracture systems, with associated volcanism and shallow water sedimentation. Diagram is based on Cloos (1939).
For post-Precambrian time – ranging from 570 My to the Present, the stratigraphic record is generally well exposed due to the fact that epicontinental seas repeatedly covered substantial parts of the present land surface. Based on geological maps, depicting the distribution of shallow marine deposits, it is possible to
evaluate the fluctuations of sea level with time. However, compared with younger geological periods, Cambrian stratigraphy is poorly known so, for these early times, eustatic correlations of sea level are rarely possible (Hallam, 1992). Nevertheless, since the late 19th century, a cogent picture has emerged that suggests that during post-Precambrian time, significant regions of the present continents have repeatedly been engulfed by longer term shallow seas, despite the fact that the present volume of marine surface water seemingly is larger now than ever before (cf. Storetvedt, 2003).
Traditionally, it has been accepted that the major proportion of the Earth’s seawater is the product of an early stage internal differentiation and degassing while the planet was in a much hotter state than now – giving rise to the hypothesis that the Precambrian Earth was blanketed by a more or less pan-global ocean concealing a rather featureless granitic crust (Süss, 1893). However, Rubey (1951) argued that the greater part of the surface water is unlikely to have been exhaled through processes of primordial segregation, it being more likely to have accumulated through slow but progressive degassing via volcanic action. He argued that the widespread epicontinental seas of the past should not be confused with the aspect of seawater volume, because “if the ocean basins have been sinking relative to the continental blocks, then one must look largely to the ocean floor, rather than to the continents, for evidence of a growing volume of seawater”. In other words, Rubey was apparently adhering to the crustal oceanization model of Barrell (1927).
Following Rubey’s reasoning, the formation of the world’s ocean basins by internal mechanisms would be associated with a major release of juvenile water. The fact that to some extent the Earth continues to be volcanically active leads to the conclusion that, unless there is a mechanism by which water can recirculate back into the mantle – for which there is hardly any evidence, the volume of surface water is larger now than ever before despite the fact that the present continental landmasses are significantly dryer now than during earlier Phanerozoic history.
However, this long-term drying-up of the continents has been discontinuous. Thus, major eustatic inundations during the Lower-Middle Palaeozoic were followed by a sharp drop of the sea level in the late Permian, and then another major inundation resumed during the Mesozoic – culminating in the late Cretaceous. On top of this mega-scale sea-level trend, it was recognized early on that many shorter fluctuations of the shore line had been superimposed on the longer trends. This Pulse of the Earth – through its intimate association with the history of surface water and the record of oscillating sea-level changes – is the theme of this paper.
The level of the sea will be lowered when a substantial volume of surface water is being stored in major ice fields, as during the Quaternary ice age of the Northern Hemisphere. But during post-Precambrian time, such global cold spells seems to have been rare and of too short duration and the ice fields of too limited an extent, to have had an appreciable effect on the longer term trend of the global sea level. In this context, van Andel (1985, p. 155) wrote: “The rate of sea-level change for glaciations and deglaciations is measured in metres per 1,000 years, much too fast [for explaining Phanerozoic sea-level variations], and we are quite certain that there were no ice ages during the Mesozoic. Clearly, non-glacial eustatic changes cannot be explained by changing the volume of water”. The recorded inundations (transgressions) of the sea could be equated with either an increasing volume of seawater or sinking of the land, and sea-level retreats (regressions) with the reverse. However, the history of seawater and the pulsating eustatic sea-level stand seems to be closely associated with the rest of the Earth’s evolutionary pattern – including the episodicity of principal global tectonic events along with the associated diversity of geological and environmental phenomena. In other words, any sensible explanation of seawater and sea-level oscillations demands a realistic system understanding of Earth’s geological history.
Historical snapshots
Around the middle of the 19th century, it was commonly thought that the Earth had cooled and chemically differentiated from an original fluid magmatic state, and that seawater was a natural product of the planet’s primeval de-volatilization. In the view of the Austrian geologist Eduard Süss [1831-1914], the expelled water had originally spread across a fairly featureless crustal surface, but progressive cooling and planetary contraction had produced crustal warping and fracturing leading to redistribution of the surface water. Consistent with this view, Süss (1893) conceived of an early Precambrian Earth covered by a shallow pan-global ocean – an idea which Alfred Wegener later took for granted (Wegener, 1912 and 1924). In Süss’ opinion, continental and oceanic crust were compositionally similar and interchangeable; he opined that during the presumed planetary contraction, large areas of the surface had collapsed to become deep sea depressions into which water masses previously residing within the continental crust had drained. The volume of surface water was considered to be constant, but it was well known that the sea level relative to the continents had displayed rhythmic variations: the sea advancing over low-lying lands – transgressions, had alternated with sea-level retreats – regressions, forming a pulsating global shore line succession for which Eduard Süss coined the term eustatic. His dynamic driving force was the Earth’s cooling and contraction.
In the Süssian tectonic system, the oceans had been growing for a long time (i.e. an increasing part of the global crust had been down-warped by the forces of thermal contraction) at the expense of upstanding continents, thereby accounting for the geological fact that the land masses – with their extensive cover of ancient marine sediments – had been subjected to an overall progressive drying-up during post-Precambrian time. However, the common transgressive pulses were explained by a net reduction in storage capacity of the developing oceanic basins – attributed to the accumulation of transported terrigenous material from surrounding regions of the crust. Regressions, on the other hand, were ascribed to the increasing volume of the deep sea basins formed as a consequence of contraction. Thus, transgression and regressions were ascribed to different and seemingly independent causes – which therefore smacked of an ad hoc escape. Furthermore, the generally slow build-up of the transgressive phases, as compared to the much more rapid regressions, was an even greater problem for the Süssian theory.
The distribution of land and sea in the Upper Palaeozoic, according to Eduard Süss, is shown in Figure 2. Süss postulated the Gondwana palaeo-continent – an ancient mega-continent that in the late Palaeozoic had united all southern land masses. When a major part of Gondwana subsequently subsided (by the presumed forces arising from planetary contraction), previous biological migration routes had been broken. In this way, the Süssian theory – representing vertical contraction-enforced oscillations of the crust – could explain biological similarities between continents now widely separated by oceanic barriers, for which the competing American contraction model of James Dana (1873 and 1881) offered no solution. Dana’s contraction hypothesis was strongly discrepant with that of Eduard Süss. According to Dana the overall physical state of the Earth had not changed significantly during most of its history: its internal state and surface structures were assumed to be static, including the configuration of continents and deep sea basins.
The ultimate product of the American version of the contraction theory was a very slow- to-moderate episodic growth of the continents through accretion along their margins. A central theme in Dana’s model was the formation and episodic deformation of fold belts; in his view, incessant crustal contraction had produced recurrent down-warping, sedimentary accumulation, compression, and then uplift. But why were the Rocky Mountains located so far inland, and how had the intracontinental Alpine-Himalaya tectonic belt formed? With regard to the long-standing problems of the widespread marine deposits blanketing the continents, Dana suggested that the primeval oceans had been too shallow to accommodate the expelled primordial water masses, implying that the present lands, in their early history, had been submerged by epeiric seas which had then drained into the subsequently-formed deep sea basins. But this proposition did not readily fit such geological facts as that the Lower-Middle Palaeozoic marine deposits blanketed significant parts of North America, the extensive and long-lasting Tethyan Sea had been a characteristic feature across southern Eurasia for most of post-Precambrian time, and the late Cretaceous (‘Cenomanian’) transgression had apparently covered considerable parts of the continents (discussed later). It seemed, therefore, that the North American contraction-based evolutionary scheme was unable to account even for most prominent surface geological features.
GONDWANALAND
Figure 2. A sketch of the suggested distribution of land (white) and sea (light blue) in the Upper Palaeozoic – according to the palaeogeographic model of Eduard Süss; in this synthesis, Gondwana was an Upper Palaeozoic southern land mass which, during subsequent contraction and vertical crustal down-warping, had been turned into the present ocean-continent arrangement. Note also the extensive intracontinental seaway, running E-W across southern Eurasia and northern Africa, which Süss (1893) named Tethys.
It became evident that the contraction theories did not provide satisfactory explanations for the uneven distribution and global tectonic interrelationships of the various tectonic belts. Thus, even more than a century ago, the time was ripe for re-thinking the basic concepts. Essential aspects like the origin of the oceanic water masses, eustatic sea-level fluctuations, along with their natural link-ups with other prominent geological phenomena, such as tectonic belt formation – a necessity for a functional global geological theory, had no ready explanation. In many ways, geology was, and still basically is, a fact-gathering enterprise without a realistic and functional global mechanism. With respect to the state-of-the-art at the end of the 19th century, the following short-list of principal problems may suffice to explain the lack of a satisfactory explanation for the geological facts as seen at that time (Storetvedt, 2003):
# The extensive periodic flooding and subsequent long-term draining of the land masses in post-Precambrian time, that left behind a blanket of shallow marine sediments, had no satisfactory explanation.
# Recurrent, but variable, sea-level fluctuations were well established, but the origin of the internal processes that produced these transgression-regression cycles, and how these sea-level pulses tied to the overall long-term draining of the present continents, remained unknown.
# Periods of transgression were much longer than the relatively short and distinct periods of regression. What was the cause of this discrepancy?
# A genetic relationship between oceanic depressions and high-standing continents was likely, but how was this connection to be understood and explained?
# The Earth’s hydrosphere had either formed during its early history or accumulated progressively, through internal degassing and volcanic action, since the birth of the planet. But if the Earth began as a red-hot molten body, as was commonly taken for granted, would it not be reasonable to think that degassed light hydrogen and hot water vapour would largely have escaped into space?
# In general, palaeo-biological problems were inexplicable within the context of the present-day continental configuration. Any functional global theory had to account for faunal and floral similarities between continents now separated by deep oceanic barriers, in addition to cases of endemism.
# It was gradually realized that major mountain chains had formed in very recent geological time, regardless of the ages of underlying tectonic disturbances. So were the deformation of pre-existing sedimentary troughs (geosynclines) and their fairly recent topographic uplift really closely connected phenomena – as had been commonly assumed?
# During post-Precambrian times, the climatic zones had had quite different orientations from those of today. In extreme cases, the present polar regions had been tropical and vice versa. What dynamic mechanism could have caused this profound shift of the global climate system?
# If the shifts of climate belts were of global extent, the old speculation of changes of the Earth’s body relative to the Sun (a notion now called True Polar Wandering), originally discussed by the famous German philosopher Johann Gottfried von Herder (in 1785: see Schwarzbach 1963), would gain strong evidence in its favour. Furthermore, if spatial changes of the Earth’s mass were a reality, how would its ellipsoidal shape affect sea-level oscillations across the globe?
# The Caledonian, Hercynian and Alpine tectonic belts running across Eurasia form a southward progression with decreasing age – probably defining globe-encircling great-circle structures. What was the cause of this dynamo-tectonic shift, and how was it connected with the rest of Earth history – including aspects such as (1) the relatively short-lived geological cataclysms characterizing the principal geological time boundaries and (2) the predominant sea-level super-cycles, with their superimposed shorter period transgression-regression cyclicity?
As a result of the multitude of unsolved problems, central European geophysicists began to explore new directions in global tectonics. By integrating palaeo-climatology and geophysics, the old notion of True Polar Wandering was substantiated – notably by Kreichgauer (1902), while arguments in favour of continental mobility were expounded. Thus, Damian Kreichgauer argued for a westward rotation of the whole crust without altering the relative continental positions, and Wettstein (1880) followed Eduard Süss by suggesting that deep sea basins were sunken parts of former land masses. Kreichgauer was apparently the first to suggest a close dynamic link between tectono-magmatic belts and the Earth’s rotation; later, Wegener (1912, 1915 and 1929) gave Kreichgauer the credit for having discovered the pole-fleeing force (later named the Eøtvøs force). Wegener followed Kreichgauer by postulating the tectonic effect from the tidal torques from the Sun and Moon – the pole-fleeing force (Pohlflucht) and the Coriolis Effect as possible driving mechanisms; these forces are indeed directed westward and towards the time-equivalent equator. However, the vast global-extent invasion of epicontinental seas during a major part of the Palaeozoic and then again during the Upper Mesozoic remained a puzzle for both Wegener and other central European geophysicists.
In his discussion of polar wandering and its possible consequences for sea-level changes, Wegener (1929, p. 159) wrote that “Many authors […] have already discussed the fact that internal axial shifts must be tied up with systematic transgression cycles; this is because the earth is ellipsoidal and because there is a time lag while it adjusts itself to the new position of the axis, whereas the sea follows at once. Since the ocean follows immediately any re-orientation of the equatorial bulge, but the earth does not, then in the quadrant in front of the wandering pole increasing regression or formation of dry land prevails; in the quadrant behind, increasing transgression or inundation [is the consequence]”.
Thus, Wegener interpreted sea-level changes as being intimately tied to resettings of the equatorial bulge; however, in his scheme, transgressions and regressions did not affect all continents simultaneously – they were quadrant-dependent. This view markedly contradicted the stratigraphic observations which Eduard Süss and later workers regarded as evidence for eustatic (global) sea-level changes.
During the following decades, a number of prominent geologists paid special attention to the chronological distribution of long-term changes of sea-level at variable scales (e.g., Barrell, 1917; Stille, 1924; Joly, 1925; Bucher, 1933; Umbgrove, 1939). Thus, Umbgrove (1942) wrote that “a great number of major transgressions took place each being separated by periods of widespread emersion [sic] of the continents. There was a rhythmic advance and retreat of the sea. We can therefore only conclude that the transgressions and regressions on the continents must be ascribed solely to a world-embracing cause. Stille expressed the synchronism of the great trans- and regressions in his law of epeirogenic synchronism, which Bucher formulated as follows: – “In a large way the major movements of the strandline, positive and negative, have affected all continents in the same sense at the same time”.
Though Umbgrove proposed that the eustatic movements were a major rhythmic phenomenon throughout post-Precambrian time, the cause of these oscillating motions were referred to unspecified vertical pulsation processes in the mantle. In addition to the global cyclicity and synchronicity, it had been known, since the time of Eduard Süss that superimposed on the ‘first order’ post-Precambrian eustatic changes there were regional-scale movements of the strandline. As we have seen above, Rubey (1951) took an unconventional look at this problem suggesting that the hydrosphere had been exhaled by episodic internal processes in connection with sub-crustal thinning of continental crust thus trending towards an oceanic
mode – an idea closely related to the oceanization model of Barrell (1927). However, sedimentation on the ocean floor has not been continuous; numerous Deep Sea Drilling Project (DSDP) cores show that sedimentation and erosion are typically episodic phenomena. Thus, Rona (1973) described hiatuses of up to tens of millions of years in the late Mesozoic to Middle Tertiary stratigraphic record of every principal ocean basin – expressed by intervals of non-deposition and/or erosion, which he tentatively associated with the transgression-regression cyclicity on shallow continental crust. Nevertheless, the ultimate question remains: which dynamo-tectonic mechanism stands behind the eustatic sea-level changes and the associated multitude of episodic surface geological phenomena?
An intermittently degassing Earth
Against the prevailing view of the 19th century – of an initially hot and molten planet, there was indeed considerable surface evidence for the presence of an assortment of discharged internal gases – discussed by authors like Reyer (1877), Guenther (1897) and Chamberlin (1897). The enormous gas blowouts of the 19th century – Mt. Tambura in 1815 and Krakatoa in 1883 – may have been reminders in this respect. Chamberlin (1897) – struggling with the many unsolved problems in global geology – took a completely new starting point by proposing that the terrestrial planets had formed by aggregation of rocky dust particles. He suggested that the early Earth probably began as a very cold body (with temperatures near 0˚K) which subsequently, as a consequence of entrapped radioactive materials, gradually heated up. On this basis, the solid material of an initial cold Earth could well have maintained at least part of its primordial heterogeneity and, therefore, could still be in a state of internal differentiation with associated degassing. Adding to this untraditional view, Hixon (1920) suggested that tectonic processes were diapiric phenomena caused by the release of internal gases, and Ampferer (1944) discussed the possibility of subsurface gas pressure powering vertical tectonic processes. In addition, the closely related theory of Earth Pulsation (e.g., Stille, 1924; Bucher, 1933), implying distinct global and synchronous tectonic events alternating with much longer periods of tranquility, was reiterated by Umbgrove (1942 and 1947).
Cosmo-chemist and planetologist Harold Urey (Urey, 1952) – the 1934 Nobel laureate in chemistry for the discovery of deuterium – restated the old view of Pierre-Simon Laplace and Immanuel Kant (late 1700s) that the planetary system had formed by aggregation of material from a flattened nebular disk surrounding the Sun, comprising a cold mix of predominantly hydrogen gas and particulate matter. On this basis, Urey argued that differentiation of the Earth, into a metallic core and silicate shells, could well be incomplete and therefore still in progress. Consistent with this thinking, Turekian (1977) argued that the present volume of surface water is considerable less than what might be expected if all water had been driven off – that is, if the Earth at an early state had been a hot molten body. Karl Turekian pointed out that, if the chemical composition of the original mantle was like that of average carbonaceous chondrites (and the early Earth in a hot molten state) – as is generally believed, the surface should contain at least 20 times more water than is presently the case.
It can be envisaged that continuous planetary degassing and related reorganization of the Earth’s interior mass has modified both the internal and the outer regions of the Earth progressively since early Archaean time – transforming an initially thick proto-crust as well as progressively, and episodically, increasing the volume of surface water (cf. Storetvedt, 2003 and 2011). The gradual accumulation of fluids and gases in the upper mantle and lower crust must have led to a considerable increase in the confining pressure at these levels. At each depth level, rocks and fluids would naturally be subject to a common pressure – producing a kind of high pressure vessel situation – with fractures being kept open just like those in near-surface rocks at low pressures (Gold’s pore theory, see Hoyle, 1955; Gold, 1999). This principle is well demonstrated in the Kola and KTB (S. Germany) deep continental boreholes (which reached maximum depths of 12 and 9 km, respectively) where open fractures filled with hydrous fluids were found throughout the entire sections drilled (e.g., Möller et al., 1997; Smithson et al., 2000); brines were seen to coexist with crustal rocks and, in the KTB site, the salinity of the formation water turned out to be about twice that of present-day normal sea water (Möller et al., 2005). In both drill sites, a variety of dissolved gases and fluids was found; primitive helium was observed at different depth levels indicating that the fluids were of deep interior origin (Smithson et al., 2000). As there is no observational evidence that deep oceanic depressions existed prior to the middle-late Mesozoic (see below), the bulk of present-day surface water must, in fact, have been exhaled from the deep interior during later stages of the Earth’s history. Nevertheless, there are reasons for believing that most of the planet’s water is still residing in the deep interior.
In view of the extremely limited information on the physical state of rocks even at shallow depths, modern studies of the Earth’s internal constitution must rely on geophysical inversion techniques, based primarily on seismological and geodetic observations, supplemented by high-temperature, high-pressure mineral physics and chemistry experiments. Nonetheless, inversion techniques have no unique solutions so inferences about the planet’s inner state and chemical constitution must necessarily be strongly model-dependent – resting on hypothetical scenarios of primordial accretion, temperature development, and mass/energy transfer processes. Therefore, to a large extent, the picture of the Earth’s interior has changed according to the needs of whatever particular theories have been/are invoked to explain surface geological phenomena. Regrettably, purely speculative ideas from time to time have become immaculate facts in all the sciences, and so it has been with regard to the interior of the Earth. For example, in recent decades deep continental drilling (Kola and KTB, S. Germany) has demonstrated that the physico-chemical constitution and structural state at even near-surface levels differ markedly from long-held conventional views – albeit without having had any noticeable effect on currently ingrained and popular views (cf. Storetvedt, 2013). Or as expressed by Wilfred Trotter (Trotter, 1941): “a little self-examination tells us pretty easily how deeply rooted in the mind is the fear of the new”.
If we accept that the Earth formed by aggregation of cold gases and rocky dust particles, the early planet must have been left in a relatively undifferentiated state. It follows that chemical elements must have experienced differentiation as the body evolved toward some lower-energy state. Within the gas-filled proto-planet, incremental coalescence of ferromagnetic planetesimals can be expected to have led to heavier concretions for which the gravitational influence outbalanced the centrifugal effect. Thus, the heavier Fe-rich masses settled inwards through the relatively less dense (presumably) gaseous mass – gradually building up a high-density central core (see Tunyi et al., 2001). As lighter elements like sulphur, carbon, silicon, hydrogen and oxygen easily dissolve in high-pressure metallic mixes (cf. Stephenson, 1981; Hunt et al., 1992; Okuchi, 1997), such lighter constituents can be expected to have followed iron alloys into the core giving rise to the well-established density deficit of the central body. According to Gottfried (1990), the core must be the host of a significant amount of hydride-metal compounds while the present silicate-rich lower mantle must include an appreciable volume of silicides – notably, silicon carbide. According to Stevenson (1981) and many others, the core is not in equilibrium with the mantle, and the presence of an irregular ‘topography’ of the core-mantle boundary (CMB) region (cf. Morelli & Dziewonski, 1987) gives further evidence of a thermo-chemically active and heterogeneous zone. It follows that the CMB region may represent the fundamental trigger of endogenous energy – this eventually leading to the observed range of geodynamic and surface geological phenomena – including surface accumulation of water.
Lighter elements, originally entrapped in the relatively cold (but slowly heating up) deep interior must have begun their upward voyage at an early stage – necessarily taking part in a number of phase changes en route. With the many lighter elements now regarded as possible constituents of the deep Earth (cf. Storetvedt, 2003 for references and discussion), it is of paramount importance to consider the geodynamic and geological consequences of buoyant volatiles – including a range of hydrocarbon compounds which may provide the most important mechanisms for internal mass transfer (Gold, 1979 and 1999). For a planet undergoing irregularly-distributed degassing (both temporally and spatially), one would expect lateral variations of density arising from temperature differences, irregular fracture distribution, and compositional heterogeneities. It is significant that sub-oceanic and sub-continental mantle sections display a relatively clear seismic difference – notably in the outer few hundred kilometres (e.g., Dziewonski 1984; Dziewonski and Woodhouse, 1987; Forte et al., 1995).
An important observation in this respect is that, when projected onto the Earth’s surface, upstanding regions of the CMB correspond to deep oceanic basins. Figure 3 demonstrates this CMB-planetary surface relationship – suggesting that processes at the outer core release energy and buoyant masses that on the surface have led to the formation of deep sea basins (see Morelli and Dziewonski, 1987) as well as, apparently, the whole range of principal geodynamic and surface geological phenomena (Storetvedt, 2003). Ruditch (1990), studying the distribution of shallow water sediments in more than 400 deep sea drill holes in the Atlantic, Indian and Pacific oceans, submitted that, since the Jurassic, oceanic depressions have formed as a result of large-scale chemical transformation and subsidence of an initial thick continental crust; he argued that the world oceans had evolved from separate and initially isolated basins – like those currently observed on the continents.
Figure 3. The diagram illustrates the estimated topography of the core-mantle boundary region obtained by PcP and PKP residuals combined – simplified after Morelli & Dziewonski (1987). Note that when projected onto the Earth’s surface, the upstanding regions of the core-mantle interphase (cf. coloured scale) correspond to deep oceanic depressions.
The fact that deep oceanic depressions apparently did not exist prior to the late Mesozoic and that most seawater seems to have accumulated during late Phanerozoic time suggests that both planetary outgassing and the vertical transfer of internal mass have been extremely slow – albeit markedly accelerating during the Mesozoic. The irregular CMB topography, as outlined by Morelli and Dziewonski, suggests that the core-mantle boundary zone is a thermo-chemically active and heterogeneous region. Whatever buoyant phases arise from the CMB region, the implications of the broad regions of diapiric upwelling, aided by hydrocarbons and hydrous fluids, are crustal thinning – through eclogite formation and associated gravity-driven delamination of the crust from its base upward. Hence, isostatic subsidence and development of surface depressions would ensue. Eclogitization commonly propagates along fractures and shear zones, and the metasomatic front often defines bands of eclogite trending along fractures – showing an abrupt transition from granulite to eclogite facies. Granted the availability of sufficient hydrous fluid, and with pressure conditions being satisfied, the reaction to eclogite will predictably proceed rapidly (Austrheim et al., 1996).
It has been demonstrated that natural occurrences of the granulite-to-eclogite transition are strongly impeded when hydrous fluids are absent (e.g., Austrheim, 1987 and 1990; Walther, 1994; Leech, 2001; Austrheim et al., 1997). Thus, Austrheim (1998) argues that hydrous fluids are much more important than either temperature or pressure, and Leech (2001) concluded that gravity-driven sub-crustal delamination (through eclogite formation) is strongly controlled by the availability of water. According to Austrheim et al. (1997), the eclogitization process brings about material weakening which make eclogites deform more easily than their protoliths – the degree of deformability being further increased in the presence of water. Thus, the large density increase consequent upon eclogitization destabilizes the lower crust and makes it detach from the relatively unaffected crust above (Leech, 2001). Figure 4 gives an illustration of this sub-crustal thinning process – advancing upward and eventually forming deep sea basins.
Figure 4. Geological interpretation of a N-S seismic profile across the North Pyrenean Fault Zone of the inner Bay of Biscay. Gravity-driven eclogitized lower crust delaminates from the lower crust and sinks into the upper mantle giving rise to the Parentis Basin. Illustration is a simplified version after Pinet et al. (1987). It is suggested that during Earth history a presumed thick proto-crust has been progressively thinned and chemically transformed – gradually implanting the present Moho interface.
Throughout its history, the Earth must have lacked thermochemical equilibrium, so in the process of reaching internal stability, mass reorganization – aided by buoyant volatiles – seems to have been at work to produce a progressively evolutionary course of crustal thinning and intermittent geological activity along with episodic accumulation of the present volume of seawater (cf. Storetvedt, 2003). The discharge rate of juvenile water seems to have accelerated greatly in Cretaceous and Tertiary times. Though shallow seas may have existed in the Precambrian, Truswell and Eriksson (1975) have argued that their tidal amplitudes were only modest.
As a consequence of the Earth’s degassing and associated internal mass reorganization, changes of its moment of inertia would be a natural consequence – producing secular changes of the globe’s rate of rotation as well as episodic, but generally progressive, changes of its spatial orientation (true polar wander). A method for studying the Earth’s spin rate (length of day, L.O.D.) for the geological past was introduced by Wells (1963 and 1970): by counting presumed growth increments in recent and fossil corals, he estimated the number of days per year back to the Lower Palaeozoic. A famous result from this study was that Middle Devonian corals gave some 400 daily growth lines per year – suggesting a pronounced slowing of the Earth’s spin rate over the past 380 million years. Subsequent studies of skeletal increments in marine fossils back to the Ordovician were generally consistent with a higher rotation rate also in the Lower Palaeozoic (Pannella et al., 1968). Creer (1975) and Whyte (1977) summarized the palaeontological length of day data available by the mid-1970. Figure 5 shows the graph of presumed number of days during post-Precambrian time given by Creer. A subsequent compilatory L.O.D. study by Williams (1989) gave closely similar results – in addition to presenting fossil clock data for the Mesozoic. More recently, a study by Rosenberg (1997) concluded that at Grenville time (some 900 million years ago) the year had 440 days.
From the zig-zag appearance of Figure 5, it is remarkable how closely the established break-points of the L.O.D. curve (numbered 1-4) – separating periods of deceleration from periods of acceleration – correspond to times of global tectonic events. These tectonic revolutions are: 1, the Alpine climax at around the Cretaceous-Tertiary boundary; 2, the Appalachian-Palatinian event near the Permian-Triassic boundary; 3, the late Devonian Acadian disturbance; and 4, the late Ordovician Taconian event. As will be outlined below, the inferred close relationship between changes in the Earth’s rotation and global tectonics is additionally associated intimately with prominent regressive sea-level events and biotic mass extinctions.
Phanerozoic sea-level changes
The volume of sea water in the late Precambrian has remained speculative, and relatively little is known about marine stratigraphy and eustasy in the early Cambrian. Nevertheless, a general Cambrian transgression onto progressively drowned cratons (Matthews and Cowie, 1979) begins with a classic orthoquartzite-to-shale succession followed by carbonites (cf. Hallam, 1992 and references therein). From the modest sea water incursion in the early Cambrian, the late Cambrian epicontinental coverage of North America had increased by some 75 %, while in the late Ordovician to Middle Silurian the shallow sea had enlarged to around 90 % or more (Dott and Batten, 1976; Dott and Prothero, 1994). Thereafter, sea level fell gradually to even below its present level at around the Permian-Triassic boundary. Figure 6 shows the global distribution of the Lower Silurian epicontinental seas.
Figure 5. Compilation of presumed days per month during the Phanerozoic – based on growth rings in fossil shells – simplified after Creer (1975). Numbers refer to break-points which in turn represent prominent tectonic events corresponding to the principal geological time boundaries.
Figure 6. A sketch map of the overall distribution of Lower Silurian epicontinental seas (blue) superposed on the current land masses. Note the relatively modest areal extent of dry land (green). Due to the low and fairly flat global surface, the overall shallow-water cosmopolitan faunas were widespread. The diagram is simplified after Boucot and Johnson (1973). At that time, the present oceanic domains are likely to have had thick continental crust so these regions too are likely to have been dominated by shallow epicontinental seas (cf. Storetvedt, 2003).
Cambrian stratigraphy is poorly known, and so are eustatic sea-level variations during that era (cf. Hallam, 1992), though a widely accepted transgression onto cratons is demonstrated by the Exxon sea level curve (see below, and Figure 7). Illuminating studies in North America (Bond et al., 1988) show consistent sea-level changes for certain specific regions: in North America, an overall eustatic rise in the Cambrian-early Ordovician is followed by a marked sea-level fall in Ordovician-Silurian time. The progressive Cambrian flooding of the cratons probably represents the first major influx of water to the Earth’s surface (as a result of degassing from the interior) – the principal factor behind the explosion of marine life at that time. In addition, world maps of the maximum degree of shallow marine inundation (Strakhov, 1948; Termier and Termier, 1952) demonstrated a similar eustatic high sea-level during the Lower-Middle Palaeozoic.
The more detailed sea-level curve of the Exxon group (Vail et al., 1977), based on onshore North American data, gave five asymmetric sea-level cycles – each representing a relatively slow transgression followed by a sharp basin deepening and a related regressive event. Figure 7, showing the Exxon sea-level curve for the Palaeozoic based on North American sequence stratigraphy, demonstrates an obvious oblique saw-tooth-shaped sea-level variation from the Silurian onwards, and an overall regression culminates in a marked Permian low-stand. In an attempt to eliminate any regional tectonic effects, Hallam (1992) proposed a generalized eustatic sea-level curve as depicted in Figure 8. For the time range concerned, the two curves are remarkably similar.
Figure 7. The Palaeozoic section of the Exxon sea-level curve – after Vail et al. (1977). Note the sharp regressive events compared with the preceding and slower transgressive periods, and the overall progressive continental draining after Silurian time.
Figure 8. Generalized eustatic sea-level variations for the Phanerozoic – after Hallam (1992). Star symbols mark the six principal events of marine extinctions; note that these biotic catastrophes correspond to times of sea-level minima (distinct regressive events).
On the basis of a progressively degassing Earth, the inferred reorganization of the internal mass would have dynamic implications – periodically altering the planet’s moment of inertia producing events of polar wander and variations in spin rate (Storetvedt, 1997, 2003 and 2011).
These intermittent changes of planetary dynamics would naturally affect the inventory of gasses and volatiles accumulated at the outer levels of the Earth and trigger a range of tectono-magmatic and surface environmental processes – including crustal transformation and variations in the mass distribution of seawater. This interlinking of geological phenomena, influencing the Earth’s progressive, variegated and episodic history, is the cornerstone of my Global Wrench Tectonics theory. By postulating the proto-Earth as a relatively cold and, hence, rather undifferentiated planetary body (cf. Storetvedt, 2011), its early history could be expected to have encompassed slow volatilization that progressively would have added gases and fluids to the developing upper mantle and crust, as well as the hydrosphere and atmosphere – besides continuously changing the planet’s internal constitution. In this way, geological evolution as well as the Earth’s seawater history became intimately associated with intermittent changes in planetary rotation which, in the surface record, is expressed by stratigraphic upheavals seen between the major geological time boundaries.
Volatiles have a high vapour pressure so, if they are incorporated into solid or liquid material during their transport outwards, they will have a tendency to escape, atom by atom, from their host compounds thereby increasing the local gas pressure: at near-surface levels, the gases contributing to the enhanced pressure may include methane and other alkanes, carbon dioxide, carbon monoxide, hydrogen sulphide, hydrogen, nitrogen, helium, and water – as vapour (see summaries by Gold, 1987 and 1999). Thus, the continuing build-up of pressure from volatiles in the outer levels of the Earth can be predicted to have triggered eclogitization and associated gravity-driven sub-crustal attenuation, giving rise to isostatic subsidence and basin formation. This process naturally began as continental depocentres, but progressive delamination of the lower crust (accelerating during the Phanerozoic), along with degassing-related magmatic processes, eventually led to a thin and basaltic deep sea crust as well as accumulating surface water (cf. Storetvedt, 1997 and 2003). Thus, the slow build-up of hydrostatic pressure beneath the evolving deep sea basins would naturally provide a lifting power for the attenuated and mechanically-weakened oceanic crust; this, in turn, would lead to accumulated seawater that would gradually transgress low-lying continental regions. Subsequently, associated sub-crustal eclogitization and delamination would lead to basin subsidence and eustatic regression – in addition to new supply of pristine water from the interior. As demonstrated by Figures 7 & 8, the long-term eustatic sea-level changes, caused by vertical motions of the evolving and progressively thinned oceanic crust, has been an ongoing process notably since Cambrian time. The important question is what dynamic mechanism led to the relatively rapid influx of surface water during the Palaeozoic?
The long-term build-up of fluids and gases in the upper mantle and lower crust can be inferred to have led to a considerable increase in the confining pressure at these levels setting off a chain of related dynamo-tectonic and environmental processes. Those parts of the upper mantle that received the greater amount of degassing volatiles – the oceanic regions to be – underwent long-term uplift, whereby the remaining continental blocks were affected by transgressive super-cycles along with superimposed events of higher frequency sea-level changes. In response, sub-crustal eclogitization and associated delamination caused broad regions to undergo overall progressive subsidence, while corresponding regressive events affected less attenuated (higher standing) crustal blocks. Dynamically, the episodic widespread inward loss of heavier eclogitized sub-crustal sections led to periodic planetary acceleration which, in turn, gave rise to events of inertia-driven torsion of the increasingly fragmented brittle shell. Hence, wrench tectonics processes were set in action.
According to present geological and palaeomagnetic evidence, the late Proterozoic-early Cambrian equator is only exposed in two continental regions: (1) the Adelaide Geosyncline and Warburton-Georgina-Bonaparte basins of Central Australia (Brown et al., 1969) – with the continent in its pre-late Cretaceous/early Tertiary orientation (see Storetvedt & Longhinos, 2014a& b; Storetvedt 2015b) and (2) the Arctic Canada-Baffin Bay-Davis Strait-Labrador Sea sector. The remaining part of the topmost Precambrian palaeoequator cuts across present-day oceanic regions (see Storetvedt, 2003 for discussion). Consistent with this palaeo-equatorial orientation, the Lower Cambrian Bradore Sandstone of northern Newfoundland and Labrador shows near-horizontal remanence inclinations – suggesting a palaeo-equatorial location (Rao & Deutsch, 1976). From a more extensive palaeomagnetic and geological database, it has been inferred that the North American craton resided at low palaeolatitudes throughout the Upper Proterozoic (e.g., Link et al., 1992; Storetvedt, 2003). Furthermore, palaeomagnetic data indicate a palaeo-equatorial setting for the late Precambrian of Australia (Embleton & Williams, 1986). The occurrence of redbeds at various horizons of the Adelaide Geosyncline and the widespread accumulation of carbonates, including stromatolitic reef sequences, provide further evidence that Australia, during the greater part of late Precambrian and Lower Palaeozoic times, experienced tropical to sub-tropical conditions.
Palaeomagnetic data show that the Northern Appalachian foldbelt – of late early Lower-Middle Palaeozoic age, strikes across Newfoundland in a NE-SW direction and follows along the corresponding palaeo-equatorial zone. Thus, in the Labrador Sea region, the two palaeo-equatorial zones (late Precambrian and Lower Palaeozoic, respectively) intersect each other at a fairly steep angle, signifying an important spatial resetting of the globe (an event of polar wander) in the early Palaeozoic. In the wrench tectonic system, the equivalent anti-podal palaeo-equatorial crossing corresponds to the Tasman-Adelaidean junction in the Australia region; in the pre-late Cretaceous setting of the continents, the Caledonian-Appalachian foldbelt formed a great-circle girdling the globe along which the Tasman-New England tectonic zone was located (see Storetvedt, 2003). Inferentially, the major event of polar wander in the early-middle Cambrian – resetting the palaeo-equatorial bulge and the corresponding polar flattening – must have caused a significant hydrostatic pressure increase affecting the gas- and fluid-rich upper mantle thereby triggering a number of geological processes – such as sub-crustal eclogitization and associated gravity-driven crustal loss to the upper mantle, as well as ‘beginning’ isostatic basin subsidence, surface volcanism driven by high-pressured volatiles, expulsion of a significant volume of endogenous hydrous fluids to the surface – along with gases including methane, hydrogen, helium, hydrogen sulphide, hydrogen, etc. (cf. Gold, 1999; McLaughlin-West et al., 1999; Lupton et al., 1999, and many others).
According to Figure 8, marked eustatic regressions characterize principal geological time boundaries – which are thought to correspond to times of sub-crustal attenuation and isostatic basin subsidence, each event resulting in a distinct tectono-magmatic upheaval caused by changes in the Earth’s moment of inertia and thereby its rotation characteristics (Storetvedt, 1997 and 2003). The late Cambrian transgressive-regressive event was followed by subsequent sea-level rises during the Palaeozoic – culminating in regressive occurrences at the Ordovician-Silurian, Silurian-Devonian, Devonian-Carboniferous and Permian-Triassic boundaries. Thus, during the Palaeozoic, the rudimentary sea basins of the late Cambrian were deepened and laterally extended; although juvenile water from the interior was periodically added to the surface, the overall global sea-level fell ending in a marked low-stand at around the Permian-Triassic boundary. Thus, during the Palaeozoic, due to dynamo-tectonic processes, a substantial volume of seawater was added, but at the same time the capacity of the developing oceanic basins had grown so that the much less affected continental block was significantly drained. In fact, the deep regression at the Permian-Triassic boundary left more dry lands than existed prior to the major influx of seawater during the Cambrian; a rudimentary outline of the modern continents had thereby been established.
A number of studies have demonstrated that during Phanerozoic time, there was a strong correlation between distinct regressive episodes and events of mass extinction – particularly of marine faunas (e.g., Bayer & McGhee, 1985; Jablonski, 1986; Raup and Sepkoski, 1982; Hallam, 1989; Hallam and Wignall, 1999). Thus, Hallam and Wignall (1999) concluded that “Rapid high amplitude regressive-transgressive couplets are the most frequently observed eustatic changes at times of mass extinction, with the majority of extinctions occurring during the transgressive pulse when anoxic bottom waters often became extensive”.
The six main events of marine mass extinction, corresponding to marked regressive events at principal geological time boundaries, are shown in Figure 8. The sea-level high during most of the Palaeozoic – reaching its maximum in late Ordovician and Silurian times – was punctuated by a number of regressive events. The most distinct sea-level falls occur at principal geological time boundaries corresponding in turn to events of crustal loss to the upper mantle, progressive isostatic subsidence and cumulative development of oceanic basins, as well as a range of environmental events. In this way, eustatic sea-level variations are intimately tied to the range of first-order events in the Earth’s history. By the end of the Permian, the accumulated high volatile pressures in the upper mantle had eventually been ‘exhausted’. During the Palaeozoic, the flooded land masses had been subjected to a number of distinct regressive events, each supposedly related to stages of the progressively evolving deep sea basins, but the deep late Permian regression exposed more dry land than since the Precambrian. By now the evolving oceanic basins were in a rather unfinished state, but the increasing eustatic transgression during the Mesozoic, reaching its peak in the Upper Cretaceous (Figure 9) and followed by a sharp regression at around the K/T boundary, eventually gave rise to the modern deep sea basins. During the predicted long-lasting crustal oceanization – that gradually and episodically turned the once global-extent thick continental crust into the present land-deep sea mosaic – the volume of surface water must have increased exponentially, but the capacity of the deep sea containers had clearly expanded even more so that, today, we have more dry land than since the early Cambrian.
At times of major volcano-tectonic upheavals, including mass extinctions of marine fauna, the anoxic conditions discussed by Hallam and Wignall (1999), may easily have entered the seawater column. For example, some authors have suggested that the combination of massive gas-driven volcanism, associated ocean anoxic events and bursts of methane release may be responsible for three major biological catastrophes – at 250, 200, and 65 million years respectively, while Max et al. (1999) considered methane gas blow-outs as the actual source of fuel for the global firestorm recorded by soot layers at the K/T boundary. For the end of the Permian mass extinction – corresponding to a deep regression and the loss of as much as 95 % of all species on Earth, Erwin (1994) and Benton and Twitchett (2003) considered widespread volcanic activity to be the most likely cause. They concluded: “The extinction model involves global warming by 6˚C and [a] huge input of light carbon into the ocean-atmosphere system from the eruptions, but especially from gas hydrates, leading to an ever-worsening positive-feedback loop, the ‘runaway greenhouse’”. A global carbon isotope excursion behind the catastrophic die-off of terrestrial vegetation at the Permian-Triassic boundary was noted and discussed by Ward et al. (2000), and Michaelsen (2002) - studying the peat-forming plants across the northern Bowin Basin, Australia - concluded that about 95% of the plants disappeared rapidly at that time.
Figure 9. Part of the world map depicting the distribution of shallow seas across the present-day continents in the Upper Cretaceous. Diagram is based on Umbgrove (1942).
Hesselbo et al. (2000) presented evidence that, in the early Jurassic, isotopically-light carbon dominated all the upper oceanic, biospheric and atmospheric carbon reservoirs. They suggested that the observed patterns were produced by voluminous release of methane from marine deposits of gas hydrates, which would be a natural consequence of the Earth’s internal degassing (cf. Gold, 1999; Storetvedt, 2003). A similar dissociation of oceanic methane hydrate has been suggested for the isotope excursion at the Palaeocene-Eocene boundary (Dickins et al., 1995; Katz et al., 1999). Thus, throughout the post-Precambrian at least, the emission of major amounts of mantle-derived methane is liable to have raised global atmospheric temperature, notably at times of rapid eustatic excursions. The occurrence of soot in and immediately above the K/T boundary and extinction zone has been associated with a global firestorm (Wohlbach et al., 1988), an observation that Gilmour and Guenther (1988) referred to as “an incomplete combustion of methane” – a conclusion with which Max et al. (1999) also concurred.
The Upper Cretaceous transgressive peak was interrupted by a number of shorter-period sea level oscillations – presumably interlinked with progressive sub-crustal attenuation, changes in planetary rotation rate, and gas-driven volcanic activity in many regions of the world. However, since then the seas have gradually retreated from the continents. In oceanic regions, this ‘multifarious’ global pulsation – often referred to as the Alpine tectonic revolution – is well imprinted into the geological record, either as horizons of erosion or non-deposition (formed by stages of uplift of the developing oceanic crust), and/or events of volcanic activity (cf. Storetvedt, 1985). During the Upper Cretaceous, widespread distribution of thinly-crusted deep oceans appeared for the first time in Earth history. The deep sea basins that had existed during the early-mid Mesozoic were only of limited extent, consisting of circular to oval-shaped depressions surrounded by a mosaic of sub-aerially exposed continental masses less affected by sub-crustal attenuation. Within the deep oceans, many fragments of former land can still be recognized by a multitude of submerged aseismic ridges and plateaus with anomalously thick crust. Thus, throughout most of the Mesozoic, there existed land connections between the remaining continental blocks, providing relatively free exchange of biota, though – due to the accelerated loss of eclogitized crust (to the upper mantle) by the end of the Cretaceous – the developing ‘asthenosphere’ had reached a more ‘mature’ stage: the irregular brittle crust had become mechanically weakened as well as more easily detachable from the underlying soft asthenosphere.
A dynamical consequence of heavier (eclogitized) crust sinking into the deformable upper mantle was an increase in planetary spin rate and/or events of polar wander – triggering latitude-dependent wrench deformation of the inhomogeneous crust (Storetvedt, 2003). Thus, for the first time, the modern continental masses were separated by thin and deformable oceanic crust and, due to an increasing planetary rotation, the land masses became subjected to relative motions in situ. For the larger continental blocks, these inertial rotations were only minor. Figure 10 gives a sketch of the suggested overall Upper Cretaceous palaeogeography – immediately before the onset of the global wrench tectonic revolution at around the K/T boundary which moderately changed the azimuthal orientations of the major continents.
An overall regressive sea-level trend prevailed during the Lower Tertiary, but by the beginning of the Miocene this tendency was put in reverse. It may be argued that the second eustatic sea-level super-cycle of the Phanerozoic, having been initiated in the early Triassic, eventually came to a close in the late Oligocene (cf. Figure 8); it had lasted for some 220 million years and had included many minor eustatic rises and falls in combination with tectono-magmatic pulses, some of them accompanied by pronounced biological and environmental consequences. Thus, a sharp event of polar wander took place at around the Eocene-Oligocene boundary (ca. 35 million years ago), amounting to an angular shift of 35˚ of the equatorial bulge, bringing the Earth to approximately its present spatial orientation. Thus, for the first time in Phanerozoic history, the North Pole became positioned in the land-locked present-day Arctic Basin, and the South Pole was displaced a corresponding distance from its early Tertiary position in the South Atlantic, onto the Antarctic continent. This polar wander event marks the beginning of the well-established onset of the present Antarctica ice cap; in Europe, the major latitudinal shift is well demonstrated by palaeontological and palaeoclimatological evidence (cf. Pomerol, 1982) – associated with a drastic cooling (e.g. Buchardt, 1978).
Figure 10. Sketch map of the suggested palaeogeography of the Earth by the end of the Cretaceous, prior to the subsequent wrench tectonic continental rotations which disrupted former trans-oceanic land ‘bridges’. In comparison with present-day geography, it may be noted that the wrench rotations of the Atlantic continents (their separation as well as their azimuthal orientation) were only minor. Dark blue colour indicates deep sea basins, while light blue represent ‘Cenomanian’ transgressive seas. Diagram is based on Storetvedt (2003).
In continental settings, the Eocene/Oligocene dynamic transition triggered the eruption of the Ethiopian flood basalts (36.9 ±0.9 My), and a number of volcanic gas blow-outs took place at that time – e. g., the Mistastin and Wanapitei Lake craters in Canada, and the Popigai crater in Russia. In an Ar/40-Ar/39 age study of the 100 km diameter Popigai structure, Bottomley et al. (1997) noted the close match between the obtained age (36.9 ±0.2 My) and that of the North American tektites which had been associated with the 85 km diameter Chesapeake Bay crater off the eastern U.S. coast, with an age of 35.3 ±0.2 My (Poag et al., 1994; Poag and Aubry, 1995). Adding to the diversity of global geological phenomena occurring at this time, can be cited by the volcanic ashes in the Massignano stratigraphic section of Italy, dated at 35 ±0.4 My, which contain a distinct Ir peak – in association with shocked quartz (Montanari et al., 1993).
The significant spatial shift of the Earth some 35 million years ago must have led to considerable hydrostatic pressure increases in regions of the volatile-rich and irregular asthenosphere. In addition to events of continued crustal delamination, the overpressure within the topmost mantle would create tectonically fractured crustal ‘chimneys’ that served as a form of pressure valves which on the surface would give rise to volcanism and high-pressure blow-outs forming craters. In many ways, the major shift of the equatorial bulge at around the Eocene/Oligocene boundary may be seen as the terminal spasm of the Alpine tectonic revolution which can be related to the widespread tectono-magmatic activity at that time – notably in the oceans. In the Exxon eustatic curve, a regressive event characterizes the Eocene/Oligocene boundary, and the Lower Oligocene transgression terminates in a deep regression in the Middle Oligocene – serving as a marker horizon between the Rupelian and the Chattian epochs (Haq et al., 1987).
In a study of the global distribution of late Lower Tertiary stratigraphic hiatuses in the sea floor record, Keller et al. (1987) found erosion events to have occurred at the Eocene/Oligocene and Oligocene/Miocene boundaries; this is consistent with the general observation of a close link between tectonics and distinct regressive-transgressive couplets linked with geological time boundaries. However, Keller et al. did not find an erosional horizon corresponding to the relatively sharp mid-Oligocene sea-level change in the Exxon curve which is well demonstrated by a DSDP drilling transect of the South Atlantic (see below). On the other hand, they found ‘corresponding’ erosional discordances in both the early and the late Oligocene. In this context, it should be remembered that any major shift of the equatorial bulge and polar flattening (such as that occurring around 35 million years ago) would have been liable to cause regional variations in asthenospheric volatile pressures and related crustal effects, notably at low-to-intermediate palaeolatitudes, thereby masking the true eustatic sea-level variation in some regions.
Overall, the Oligocene showed a regressive tendency indicating ongoing crustal loss to the mantle and related development of deep sea basins. This late stage reconstitution of the crust inevitably led to changes in the Earth’s moment of inertia, increasing the confining pressure within the lithospheric lenses as well as in the melt pockets at higher levels – paving the way for a new round of more forceful tectono-magmatic events. Thus, starting at around the Oligocene/Miocene boundary, ca. 22 million years ago and the sea encroached once more on the land, culminating in an overall high stand in the Lower-Middle Miocene. The Exxon proposal of the post-Oligocene (Neogene) sea-level variations (Haq et al., 1987) is shown in Figure 11. According to this scheme, for the Lower and Middle Miocene – spanning a period of about 15 million years – the global shore-line was raised by some 150 metres. This long-standing transgression was punctured by two short-lived regressive events, around 15 My ago, ending with a major sea-level drop some 8 My years ago – the latter defining the Miocene sea-level minimum. The Miocene Era was terminated by a distinct regressive phase at ca. 5 My ago (end of the Messinian). These sea-level low stands are most likely associated with events of planetary acceleration – being a dynamic response to inward loss of widespread eclogitized lower crustal segments.
In the Atlantic region, the oscillating mid-Miocene regressions, with their related high-pressured volatile-rich asthenosphere, is time-equivalent with the origin of the Columbia River basalts (dated at 16.2 ±1 My) and with the Steinheim and Ries craters in Germany (dated at ca. 15 My) – see Figure 11. Miocene and younger elevations of the deep sea crust, giving rise to continental transgression, affected broader crustal regions of the world oceans. For example, in the islands of the Central Atlantic (Cape Verde Islands, Ascension Island, Madeira and the Azores), Lower-Middle Miocene and younger marine sedimentary horizons are found at heights ranging between 400 and 500 metres above present sea level (Mitchell-Thomé,1976), while Miocene and younger volcanic activity shows widespread distribution in this part of the Atlantic (see Storetvedt, 1985). The Neogene phases of regression are inferred to be related intimately to the youngest phases of oceanization – having transformed particular regions of continental crust into oceanic-type structures. For example, in the Mediterranean a number of isolated circular-to-oval shaped depressions formed during the Messinian – in association with a very thick succession of salt of variable chemistry degassed from the mantle. Wezel (1985), for example, argued that, in the late Miocene, the Tyrrhenian region was the site of an upstanding intra-Alpine continental crust that in Plio-Quaternary time underwent variable sub-crustal thinning and vertical collapse activated by upper mantle processes.
Figure 11. Diagram shows the Exxon eustatic sea-level curve for post-Oligocene (Neogene) time – after Haq et al. (1987).
As we have argued above, periodic vertical motions of the sea floor – reflecting build-up and subsequent release of upper mantle volatile pressures – with related sedimentary discordances and magmatic activity, are likely to have been a persistent global feature and the ultimate cause of the principal events of eustatic sea-level changes. Thus, Figure 12a delineates the significant Miocene depositional break across the South Atlantic, at latitude 30˚S, which inferentially corresponds to the Lower-Middle Miocene transgressive phase shown in Figure 11. The associated flooding of low-lying regions of South America is outlined in Figure 12b. In an extended sedimentary section at DSDP site 355 on the North Brazilian margin, sedimentary hiatuses were recorded in the topmost Cretaceous (Maastrichtian), at around the Eocene-Oligocene boundary, and in the Middle Miocene – supporting the thesis of a close connection between major phases of oceanic crustal uplift and erosion with corresponding events of sea-level rise on low-lying continental regions. Compilation of cored Mesozoic sediments in sites of the western and eastern margins of the Central Atlantic (Arthur, 1979; Storetvedt, 1985) again shows a significant stratigraphic hiatus consistent with the major Upper Mesozoic eustatic transgression.
Figure 12. Diagram (a) shows the ‘Middle’ Miocene sedimentary break of the Deep Sea Drilling Project Leg 3 sites across the South Atlantic at 30˚S (simplified after Maxwell et al., 1970). This trans-oceanic depositional hiatus is regarded here as a segment of a widespread deep sea crustal uplift having produced the Lower-Middle Miocene eustatic sea-level rise. Diagram (b) exemplifies the resulting mid-Miocene sea-level (light blue) of South America (Webb, 1995).
Concluding summary
In this paper, the focus has been on the origin of Earth’s surface water and the cause of sea-level changes for which the crustal product is a continuing, albeit jerky, loss of eclogitized gravity-driven continental material to the mantle – eventually leading to formation of the present-day thin oceanic crust and deep sea basins. As a result of the actual degassing Earth model, today’s continents have, during the Phanerozoic, been repeatedly flooded by slowly rising seas which after sea-level high stands have subsequently retreated to form distinct sea-level lows. It is an observation of paramount importance, long noted by many authors, that the most marked regressive events occur at times of principal geological time boundaries – representing revolutionary episodes in Earth history – in terms of tectonic, magmatic, biological and environmental happenings. In this way, sea-level changes became intimately linked to the rest of the planet’s first-order geological manifestations.
Central in this discussion is that recurrent sea-level low-stands eventually gave rise to ever-growing deep sea basins, and the transgressive-regressive couplets continuously added fresh surface water from the mantle. The first transgressive super-cycle commenced in the early Palaeozoic – being closely linked to the marine biological boom at that time, lasting till the late Palaeozoic. However, a deep sea-level regression at the Permian/Triassic boundary, adding a multitude of toxic gases and fluids to the sea and the atmosphere, led to mass extinction and the most severe crisis in the history of life (Raup, 1979). At this time, the evolving deep sea basins had evolved into a sizeable volume thus draining the continents – leaving more dry land than ever before in post-Precambrian history. But internal gases and fluids continued their upper mantle accumulation and accompanying pressure increase – giving rise to a Mesozoic uplift of the evolving oceanic basement, with an associated overall major sea-level rise that culminated in the Upper Cretaceous. The following regression and upper mantle gas exhaustion led to another major biotic and environmental crisis – at around the K/T boundary. By now the world oceans were nearing their present state and extent, but continued to demonstrate alternating cycles of sea-level changes, with stratigraphic control, suggesting that the deep sea basins are still under development. In addition, it is highly probable that the volume of sea water has increased continuously to this day, and the major part of the planet’s water may probably still be residing in the interior.