Tải bản đầy đủ - 0 (trang)
Chapter 8.3 Ancient to Modern Earth: The Role of Mantle Plumes in the Making of Continental Crust

Chapter 8.3 Ancient to Modern Earth: The Role of Mantle Plumes in the Making of Continental Crust

Tải bản đầy đủ - 0trang


Chapter 8.3: Ancient to Modern Earth: The Role of Mantle Plumes

lower crustal delamination may have played a major role in parts of eastern Asia, Europe

and western North America. These delamination processes, which produced widespread

alkaline volcanism, are associated with destruction, at least in eastern Asia, of continental


The Archean and Proterozoic Eons are characterised by a predominance of greenstone

belts (Archaean), and the development of epicontinental sedimentation in basins and of

continental rifting (Proterozoic), whereas the Phanerozoic Eon is essentially based on the

first appearance of exoskeletal invertebrates and the radiation of life. However, these traditional, and indeed convenient divisions, are not entirely coincident with global thermal

events. Condie (1997) considered age distributions in juvenile continental crust and related

them to time-integrated evolution of thermal regimes in the mantle. It was proposed that

these thermal regimes began with layered convection and buoyant subduction (Stage 1),

and evolved to catastrophic mantle overturns between 2.8 and 1.3 Ga (Stage 2), and finally to whole mantle convection (Stage 3). Thus, it appears that the crustal history, with

its tectonic and magmatic record, is closely linked with mantle dynamics. In this paper,

it is suggested that the role of mantle plumes in providing juvenile material to the crust

decreased from the Archaean through the Proterozoic to the Phanerozoic, and that mantle

dynamics can be considered within the framework of five evolutionary stages: (1) >3.0 to

ca. 2.8 Ga; (2) 2.8–2.4 Ga; (3) 2.4–1.0 Ga; (4) 1.0–0.6 Ga and (5) 0.6–present.


Based on the velocity distribution of seismic waves, the present-day continental crust

has a layered structure with two main subdivisions (Fig. 8.3-1):

• upper crust (compressional wave velocities, Pv, of 5.9 to 6.3 km s−1 )

• lower crust (Pv of between 6.5 and 7.6 km s−1 ).

The boundary between the upper and lower crust is widely displayed as the Conrad discontinuity at a depth of about 20 km (Wedepohl, 1995), although this does not appear to be

present everywhere. Recent estimates of the mean composition of the upper crust indicate

that it has a granodioritic to quartz dioritic composition. The lower crust is characterised

by strong seismic reflections, probably caused by layered mafic and mafic–ultramafic intrusions (Percival et al. 1989). In the western Alps, the Ivrea-Verbano zone is a 10 km thick

mafic igneous complex that is considered an exposed slice of lower crust and possibly an

exposure of the crust–mantle boundary (Sinigoi et al., 1995). The higher seismic velocities

in the lower crust suggest either a more mafic composition, and/or that it has high-pressure

mineral assemblages. Following early models of basaltic or gabbroic compositions, there

are three possibilities for the composition of the lower crust (see Pirajno, 2000 for overview

and references). In one, the lower crust is “dry” (there are only mineral phases that do not

contain water, such as pyroxene) and formed by felsic to intermediate rocks (granitic to

dioritic) that have been subjected to high pressure modifications to granulite facies. The

second is that the lower crust is “wet” (contains water-bearing mineral phases, such as

8.3-2. The Continental Crust


Fig. 8.3-1. An idealised section of modern continental crust. After Rogers and Santosh (2004).

amphibole) and is predominantly made up of amphibolite, with a mineral assemblage consisting of amphibole, plagioclase, epidote and Fe-rich garnet. The third possibility is that

the lower crust is composed of gabbroic anorthosite (predominantly feldspar, lesser pyroxene, with minor garnet, quartz and kyanite).

It is probable, however, that the composition of the continental crust is more complex

than the above-mentioned possibilities and that in reality it is laterally heterogeneous, with

one or the other of the above cases dominating in a given region. Measurements of crustal


Chapter 8.3: Ancient to Modern Earth: The Role of Mantle Plumes

Poisson’s ratio σ (the ratio of P to S waves velocity) by Zandt and Ammon (1995) revealed

a general increase of σ with the age of the crust. This finding supports the presence of a

mafic lower crust beneath cratons (Zandt and Ammon, 1995).

The crust is separated from the subcontinental lithospheric mantle by the Mohorovicic

discontinuity, commonly called Moho (Fig. 8.3-1). Based on various considerations, the

nature of the Moho beneath the continents is likely to be a boundary between the silicarich rocks of the crust and the ultrabasic rocks of the upper mantle. The nature of the

Moho is considered to reflect either a physical discontinuity (phase change, within a single

composition) or a chemical change (the lithospheric mantle below the Moho is of ultrabasic

composition, i.e. dunite or peridotite) (Wyllie, 1971)

8.3-2.1. The Growth of Continental Crust

As mentioned above, the origin of the continental crust is not completely understood.

The general consensus is that much of the crust would have formed in the earliest part

of the Earth’s geological evolution (Fig. 8.3-2), following initial planetary accretion, and

was formed by differentiation of magma from the mantle (the secondary and tertiary

crusts of Taylor and McLennan (1985)). The debate as to whether the continental crust

grew gradually through geological time or whether it was, and is, being maintained in

steady-state processes of addition and destruction is still very much an open question

(Hofmann, 1997). Sylvester et al. (1997) used the Nb/U ratios of crust and mantle in an

attempt to solve the problem. They compared the Nb/U ratio of 30 in stony meteorites

Fig. 8.3-2. Crustal growth models. After Taylor and McLennan (1985 and references cited therein)

and Condie (2004). Most of these models suggest rapid rates of crustal growth before 2.5 Ga. See

text for details.

8.3-2. The Continental Crust


to that of the primitive mantle, with the ratio of 47 in present-day “residual” mantle.

These authors related this difference to the creation of continental crust, which has a

Nb/U ratio of 10. This means that Nb tends to remain in the residual mantle, whereas

U is transferred to the crust, via melt transfer of differentiated magmas. Sylvester et al.

(1997) found that the Nb/U ratios of 2.7 Ga lavas are close to 47, implying that a similar amount of continental crust had been formed since the Archaean as compared with

modern day Earth. As detailed later in this contribution, remnants of the earliest crust can

be found in the Archaean Na-rich tonalite-trondhjemite-granodiorite (TTG) suite. Subsequent additions to this original Archaean crust were made during successive igneous

and tectonic events that are still occurring today. Recent views propose that Archaean

continents grew by the accretion of oceanic plateaux and island arcs, so that a supercontinent became established by about 2.7 Ga (e.g., Percival, this volume; Rogers and

Santosh, 2004).

It is generally recognised by most planetary scientists that early thermal instabilities in

the mantle, caused by accretionary energy and the decay of radionuclides, resulted in a

vigorously convective mantle in the early Earth (cf. Davies, 1999). Models of the geotectonic evolution of planet Earth, from the Hadean to present-day, are based on the thermal

regimes of the mantle, its dynamics and the role of convection in the establishment and the

making of continental crust. The tectonic regimes of the primitive Earth were substantially

different from those of later ages, due to progressive cooling and related changes in the

convective patterns of the mantle (Schubert et al., 2001).

The Hadean began with the accretion of the Earth until the oldest known rocks, the ca.

4.03 Ga Acasta Gneiss, and any direct geological evidence for the period between accretion and the oldest rocks is almost completely lost from the geological record, presumably

because of intense asteroid bombardments (Martin, 2005: but see Kamber, this volume;

Cavosie et al., this volume: Iizuka et al., this volume). Various growth models suggest that

from the Hadean to the end of the Archaean, 70–75% of juvenile continental crust was

generated and extracted from the mantle (Martin, 2005) (Fig. 8.3-2). The Archaean mantle was hotter, perhaps vigorously convecting, and with abundant mantle plume activity,

as recorded by komatiite fields, which are common in greenstone belts. Indeed, some Archaean greenstone belts are considered as a kind of large igneous province (LIPs), related

to mantle plume events (Eriksson and Catuneanu, 2004; Van Kranendonk and Pirajno,

2004; Van Kranendonk et al., this volume).

Apart from changes in their shape, estimated crustal growth curves show continuity (Fig. 8.3-2), but there is also evidence that there were major episodes of continental

crustal growth at 2.7–2.6 Ga, 1.9–1.8 Ga, 0.5–0.2 Ga, while peak mantle plume events

are recorded at 3.0 Ga, 2.8 Ga, 2.5 Ga, 1.8–1.7 Ga, 1.1–1.3 Ga, 0.4 Ga, and 0.25–0.1 Ga

(Condie, 2000; 2004; Abbott and Isley, 2002; Groves et al., 2005). A major change is

recorded at ca. 2.5 Ga, at the Archaean–Palaeoproterozoic boundary, when a mantle plumedominated tectonic regime was replaced by a tectonically quieter period (Barley et al.,

2005), when mantle plumes were perhaps less numerous but more intense and extensive.

Proterozoic mantle plumes impinged onto Neoarchaean supercontinents, leading to cycles

of breakup and assembly (Condie, 2004; Rogers and Santosh, 2004). Tectonic regimes


Chapter 8.3: Ancient to Modern Earth: The Role of Mantle Plumes

changed across the Neoproterozoic–Cambrian boundary, with mantle plumes still active,

but with marked changes in the extent and frequency of subduction zones, which tend to

dominate this period. The remainder of juvenile continental crust (ca. 30%) was added

during Neoproterozoic and Phanerozoic times, some of which was extracted from mantle

plumes, some from asthenospheric mantle upwelling during lithospheric delamination, and

some extracted in subduction zones.


Mantle convection is driven by three fundamental processes: heat loss from the core

(about 20%), internal heating from radioactive decay (about 80%) and cooling from above

(sinking of lithospheric slabs) (Condie, 2001). Mantle convection is responsible for the

movements of lithospheric plates, earthquakes, magmatism, surface volcanic activity and

indeed most of the geological and tectonic processes manifested in the crust. Other surface

phenomena or physical manifestations that can be related to a convective mantle are geoid,

gravity, and heat flow anomalies. Details on the principles and concepts of convection in

the Earth’s mantle can be found in Davies (1999). Mantle convection is time-dependent

and as the rate of heat production decreases, the planet cools and eventually convection

slows or stops altogether. Convection is driven by buoyancy anomalies that form at thermal boundary layers (Campbell and Davies, 2006). There are two boundary layers in the

Earth’s mantle; an upper boundary layer (lithosphere) at ca. 660 km, and the core-mantle

boundary (CMB) at ca. 2890 km. Temperature differences at the boundary layers result

in convective upwellings, which laboratory experiments show that are likely to be columnar with a tail and a head developing as the column moves through the mantle (Campbell

and Davies, 2006). This constitute what is known as a mantle plume. The plume head is

cooler than the tail, because it contains entrained material from the surrounding cooler


Current ideas on mantle plumes posit that they are “jets”, “narrow upwelling currents”,

or “narrow cylindrical conduits” of hot, low-density material originating either from the

CMB (or D thermal boundary layer: one-layer mantle model), and/or from the 660 km

discontinuity (two-layer mantle model) (Davies, 1999; Schubert et al., 2001; Campbell

and Davies, 2006) (Fig. 8.3-3). The general consensus is that most large and long-lived

plumes originate from the CMB, at the D thermal boundary layer, and are caused by

heat from the outer core and ensuing thermal instability. Some researchers suggest that

this thermal instability is in response to the sinking of cool lithospheric slabs that cascade

through the 660 km discontinuity into the lower mantle and accrete onto the CMB (mantle

avalanches; see Schubert et al. (2001, pp. 454–456)). Kellogg et al. (1999) suggested that

mantle plumes originated as a result of slab avalanches, and may be carrying the geochemical and isotopic signatures of recycled slab material. This is supported by evidence from

seismic tomography of slab avalanches to 2700 km depth (Grand et al., 1997).

Courtillot et al. (2003) identified three types of plumes: (1) primary, or deep, plumes,

originating from the D layer; (2) secondary plumes, originating from the top of large

8.3-3. Mantle Plumes


Fig. 8.3-3. (A) Model of mantle convection and plumes in the Earth; based on Davies and Richard

(1992) and Jacobs (1992). (B) Types of mantle plumes in the modern Earth, according to Courtillot

et al. (2003); (1) deep mantle plumes rise from the core–mantle boundary, forming hotspots such

as those of Hawaii and Reunion or large plumes (superplumes), such as those under Africa and

the Pacific; (2) secondary plumes may emerge from these large plumes forming discrete igneous

provinces; (3) shallow mantle plumes may rise from the upper mantle, as a result of lithospheric

delamination, following the collapse of collision orogens, or in back-arc rifts. (Continued on next


domes of deep plumes or superplumes; (3) tertiary, or Andersonian, plumes, originating from near the 660 km discontinuity and linked to tensile stresses in the lithosphere

(Fig. 8.3-3(b)). The superplumes of Courtillot et al. (2003) are located under Africa and

the central Pacific Ocean, where two massive mantle upwellings are evidenced by high

crustal elevation (superswells) and by corresponding regions of low shear wave (Vs) velocity anomalies in the mantle (Gurnis et al., 2000).


Chapter 8.3: Ancient to Modern Earth: The Role of Mantle Plumes

Fig. 8.3-3. (Continued.)

The “Andersonian plumes” of Courtillot et al. (2003) represent shallow plumes and may

be considered an alternative to the theory of deep mantle plumes. Indeed, a number of researchers do not subscribe to the plume paradigm and maintain that hotspots and flood

basalts are not caused by deep plumes (see Foulger et al., 2005, and references therein).

Shallow plumes may arise as a result of ductile delamination of the lower lithosphere,

which induces upwelling of the asthenosphere, manifested at the surface by uplift, rifting, and the eruption of alkali basalts (Elkins-Tanton, 2005). I return to this mechanism

to explain tectono-magmatic events in the Cenozoic and associated destruction of mantle

lithosphere and lower crust.

8.3-3. Mantle Plumes


Partial melting in the plume head occurs by adiabatic decompression yielding lower

temperature and lower-Mg melts (tholeiitic basalts), whereas melting in the hightemperature tail yields high temperature and high Mg-melts (picrites, komatiites) (Campbell et al., 1989). The surface expression of mantle plumes is typically manifested by

doming of the crust – reflected as topographic swells of 1000–2000 km in diameter and

2000–4000 m elevations (above sea level), by rift basins, and by intraplate volcanism



2001; Ernst and Buchan, 2003). Regions of intraplate anorogenic volcanism are

commonly called “hotspots”, a loose term that essentially refers to the concept of a stationary heat source in the mantle and high heat flow related to magma advection (Wilson,

1963; Schubert et al., 2001). Geodetic data show that several hotspot regions correlate

with rises in the gravitational equipotential surface (geoid high), probably reflecting the

buoyancy of heated lithosphere (Perfit and Davidson, 2000). Uplift is followed by subsidence due to loss of buoyancy of the plume head, or removal of magma from the top of the

plume, thermal decay, or a combination of all three. Subsidence and crustal sagging cause

the formation of sedimentary basins, characterized by the deposition of extensive aprons

of siliciclastics, carbonates and evaporites, commonly overlain by continental flood basalts

and/or transected by related dyke swarms.

In addition to normal plume events, there appear to be major pulses of heat transfer in

the evolution of the Earth, in which a number of plumes impinge on to the base of the

lithosphere. These plume events, called superplumes (Larson, 1991; Ernst and Buchan,

2002), have important implications in terms of possible links with supercontinent cycles

and time–space distribution of metalliferous deposits (Fig. 8.3-5) (Barley and Groves,

1992; Abbott and Isley, 2002). Superplume events have been recognised at various times

in the geological record, and correlate with the growth of continental crust.

The eruption and intrusion of great volumes of mafic and ultramafic melts is attributed to

the rise and impingement of mantle plumes on continental and oceanic lithospheric plates.

These large-scale emplacements of mafic rocks are termed Large Igneous Provinces (LIPs;

Coffin and Eldhom, 1992). The eruptions of these mafic melts form vast fields of lava flows

and associated igneous complexes, up to 7 × 106 km2 in areal extent (e.g., Central Atlantic

Province; Marzoli et al., 1999; Siberian flood basalt province; Nikishin et al., 2002), or

lines of oceanic islands, such as the Emperor-Hawaiian chain, thousands of kilometres

long. In the ocean these vast lava fields are known as oceanic plateaux, such as the Ontong

Java and Kerguelen plateaux (Taylor, 2006; Wallace et al., 2002). Oceanic plateaux probably constitute the greatest contribution to crustal growth, both ancient and modern. Iceland

and Kerguelen, where mantle plumes interact with mid-ocean ridges, provide modern examples of continental crust nucleation processes (Grégoire et al., 1998). In the ancient

record, such a continental nucleus is represented by the East Pilbara Terrane, which was

built by the emplacement of successive volcanic plateaux, as explained below.

The time-dependence of mantle convection and hence of plume activity appears to be

substantiated by isotopic (e.g., Nd and Sr) and trace element systematics (Stein and Hofmann, 1994), and by the distribution of ages of juvenile continental crust (Condie, 1997).

Plots of 143 Nd/144 Nd and 87 Sr/86 Sr as a function of the age of orogens tend to follow a

linear growth pattern, from a primitive mantle (low 143 Nd/144 Nd and 87 Sr/86 Sr values), to


Chapter 8.3: Ancient to Modern Earth: The Role of Mantle Plumes

a highly depleted upper mantle (higher 143 Nd/144 Nd and 87 Sr/86 Sr values), with various

scatters being accounted for by mixing of older continental material with more juvenile

upper mantle material (Stein and Hofmann, 1994). Stein and Hofmann (1994) theorised

that mantle flow patterns alternate between two-layer and single-layer (or whole mantle)

convection, which give rise to “Wilsonian periods” and “mantle overturn and major orogenies” (MOMO) periods, respectively.

Condie (2004) proposed that there are two types of mantle plumes: long-lived and shortlived. The former, called shielding superplumes (>200 Ma duration), are the result of the

shielding effect of a supercontinent, which will cause upwelling of mantle plumes followed

by the fragmentation of the supercontinent: these do not appear to be linked to production

of large volumes of juvenile crust. The short-lived plume events (<100 Ma duration), called

catastrophic superplume events, are associated with accretion of volcanic arcs and addition

of juvenile crust.

A decrease in the frequency of mantle plume activity with time is demonstrated by

the progressive decrease in the thickness and geotherms of the sub-continental mantle

lithosphere (SCLM; Groves et al., 2005; O’Reilly et al., 2001) (Fig. 8.3-4). Thus, an evolution from a mantle plume dominated regime in the Archaean, to a more “buoyant” style

tectonics in the Proterozoic, to modern plate tectonics in the Phanerozoic, clearly reflects

the trend of a cooling Earth (Groves et al., 2005).

This trend is possibly linked with a change over geological time from whole mantle

convection of the early Earth to two-layer mantle convection in the Phanerozoic. The latter

is responsible for shallow mantle plumes, which are linked to the upper thermal boundary

layer (Fig. 8.3-3(b)) and result in igneous provinces, dominated by alkaline chemistry.

As mentioned previously, shallow mantle plumes may form as a result of asthenospheric

upwelling due to lithospheric mantle and lower crustal delamination. Tectonic collapse of

collisional orogens is inferred to occur as a result of thinning of negatively buoyant mantle

lithosphere, with detachment and sinking of orogenic roots or of subduction slabs. This

extensional collapse engenders the rise of asthenospheric mantle, during the process of

delamination tectonics (Platt and England, 1993).


The Archaean record is dominated by granite-greenstone terranes, for whose origin

opinions differ from plate-tectonic models dominated by convergent margins and subduction zones to intraplate models that call on vertical tectonics and the activity of mantle

plumes. Granite-greenstone terranes vary somewhat in their nature, structure and geodynamic history, as perhaps best exemplified by contrasts between the East Pilbara Terrane

and West Pilbara Superterrane in the Pilbara Craton (Hickman, 2004; Smithies et al.,

2005b; Van Kranendonk et al., 2007a) and the Yilgarn Craton of Western Australia. Whatever the dominant tectonic regime(s) in the Archaean, there is little doubt that during this

time juvenile continental crust was formed.

8.3-4. The Archaean


Fig. 8.3-4. Lithosphere evolution of the eastern part of the Sino-Korean Craton. (A) subcontinental mantle derived from garnet data. (B) Ordovician palaeogeotherm derived from xenocrysts

in kimberlites, with Archaean mantle thickness. (C) subcontinental lithospheric mantle section

and geotherm derived from xenoliths in Tertiary basalts. CMB crust–mantle boundary, LAB

lithosphere–asthenosphere boundary. After Griffin et al. (1998b) and O’Reilly et al. (2001). Figure

and caption courtesy of S. O’Reilly of Macquarie University, Sydney.

As mentioned in the Introduction, if 70–75% of the continental crust was formed before

2.5 Ga, this must mean that the Archaean must have been a period of considerable magmatic and tectonic activity (Martin, 2005). Neodymium isotope systematics (εNd ) clearly

show that most Archaean rocks have εNd > 0, indicating a mantle that underwent extraction of continental crust. Values of εNd < 0 appear at about 2.7 Ga, and are indicative

of greater, or increasing crustal contributions. Archaean terranes essentially consist of a

gneissic basement, greenstone belts (mafic-felsic volcanic and sedimentary rocks), and

late granitic intrusions. The gneissic basement rocks represent the earliest juvenile crust

and are sodic, commonly referred to as TTG (tonalite-trondhjemite-granodiorite; Martin et

al., 2005). TTG typically have initial Sr ratios and εNd 0, suggesting that their parental


Chapter 8.3: Ancient to Modern Earth: The Role of Mantle Plumes

magmas were sourced from the mantle and not from a pre-existing continental crust. The

TTG suites differ from subduction-related granitic batholiths in their limited range of SiO2

abundances and increasing Na2 O with SiO2 , and their Sri and εNd values. There is a vast

literature on TTG and their origin, on which the general consensus is that these intrusions

were derived from the melting of mantle-derived basaltic magmas that were metamorphosed to amphibolite facies (Martin, 2005; Martin et al., 2005, and references therein).

Whether these basaltic magmas were subduction generated, or mantle-plume generated

is an important issue in the context of what was the dominant role for the generation of

juvenile crust (see Van Kranendonk et al., this volume).

Underplating of basaltic magmas, as shown in seismic profiles, is a major contributor

to the growth of continental crust (Condie, 2001). Furthermore, accretion of mantle plume

heads onto the lithosphere is a reasonably well documented process, based on Nd-Sr isotopic systematics for the basalts of the 0.9 Ga Arabian-Nubian Shield, which led to the

conclusion that a plume head became accreted onto the base of the lithosphere (Stein and

Hofmann, 1992, cited in Condie, 2001; Stein, 2003). In Archaean times, oceanic plateaux

may have been common features and I concur with Condie (2001) that it is possible, if not

likely, that the first continents may have been formed through the accretion and/or stacking

of a series of oceanic plateaux, and/or collision of mid-ocean ridges with oceanic plateaux.

For example, the Ksotomusha greenstone belt in the Baltic Shield was an oceanic plateau

that was tectonically accreted to a continental margin, thereby becoming a new segment of

continental crust (Puchtel et al., 1998).

The first postulated supercontinent may have been Ur (Fig. 8.3-5), which stabilised at

around 3.0 Ga and included the Pilbara (Western Australia), Kaapvaal (South Africa), and

Dharwar, Singhbhum, and Bhandara (India) cratons, and Madagascar (Rogers and Santosh,

2004). According to Condie (2004), frequent collisions and accretions in the Archaean

between oceanic plateaux and fragments of continental blocks must have been common

between 2.75 and 2.65 Ga in Laurentia, Baltica and Siberia, and between 2.68 and 2.65 Ga

in Western Australia and southern Africa. Juvenile crust production reached a maximum

at about 2.7 Ga, with the formation of a Late Archaean supercontinent (?Arctica, ?Kenorland) (Fig. 8.3-5). Widespread rifting of Archean crust occurred at about 2.45 Ga, probably

associated with a superplume event, as suggested by the presence of large dyke swarms in

many cratons (e.g., Matachewan dyke swarm in the Superior Province, Canada and the

Widgiemooltha dyke swarm in the Yilgarn Craton, Western Australia). The final breakup

of this Late Archaean supercontinent was at around 2.2–2.0 Ga. I return to this topic in the

section dealing with the Proterozoic.

8.3-4.1. Pilbara Craton, Western Australia

The Pilbara Craton, together with the Kaapvaal Craton in South Africa, may have formed

the Vaalbara continent, which was possibly part of the first supercontinent Ur (Fig. 8.3-5).

The Pilbara Craton consists of five terranes (Van Kranendonk et al., 2006a): (1) East Pilbara

Terrane; (2) Roeburne Terrane; (3) Sholl Terrane; (4) Regal Terrane and (5) Kurrana Terrane. These terranes exhibit different structural and tectonic styles. The 3.53–3.17 Ga East

Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Chapter 8.3 Ancient to Modern Earth: The Role of Mantle Plumes in the Making of Continental Crust

Tải bản đầy đủ ngay(0 tr)