Elsevier Encyclopedia of Geology

Подождите немного. Документ загружается.

DESERTS

See SEDIMENTARY ENVIRONMENTS: Deserts

DIAGENESIS, OVERVIEW

R C Selley, Imperial College London, London, UK

Introduction

Lithification is the process by which unconsolidated

sediment is turned into solid rock. In 1888, von

Gumbel proposed the word ‘diagenesis’ to describe

the processes of lithification. Subsequently, in 1957,

Pettijohn wrote:

Diagenesis refers primarily to the reactions which take

place within a sediment between one mineral and an-

other, or between several minerals and the interstitial or

supernatant fluids.

In 1968, Dunoyer de Segonzac gave a detailed ac-

count of the history of the concept of diagenesis, and

pointed out that, in addition to the chemical processes

mentioned by Pettijohn, some physical processes, such

as compaction, are important in diagenesis. Thus, two

types of diagenesis are now recognized: physical and

chemical. These will be described in turn, but first it is

appropriate to define the boundaries of diagenesis.

The Boundaries of Diagenesis

Diagenesis begins as soon as sediment is deposited.

Fossilized beer bottles and other anthropogenic de-

tritus found in modern ‘beach rock’ limestone pic-

turesquely illustrate this. Ancient evidence of early

diagenesis is further confirmed by the occurrence of

intraformational conglomerates containing clasts,

not only of contemporaneous limestone, but also of

siderite-cemented sandstones and shales. As sediment

is buried more deeply, temperature and pressure in-

crease and, ultimately, diagenesis merges into meta-

morphism, with shale becoming slate, sandstone

becoming quartzite, and limestone becoming marble.

Field observation and laboratory experiments dem-

onstrate that the boundary between these rock

types, and hence diagenesis and metamorphism, is

gradational. The sequence, deposition ! diagenesis

! metamorphism, is not a ‘one-way street’, however.

At any time while sediment is on its way to meta-

morphism, it may be uplifted to the surface again.

Rocks returned to the surface show a reversal of the

trend of porosity decreasing with burial, and its en-

hancement by both physical and chemical processes.

The term ‘epidiagenesis’ was applied by Fairbridge in

1967 to the diagenesis resulting from uplift and

weathering. Epidiagenesis is of little significance in

shales. It is, however, of great importance in sand-

stones and carbonates, because of the way in which it

restores porosity and permeability to rocks that had

previously lost these features. When buried beneath

an unconformity, these epidiagenetically enhanced

zones may provide excellent petroleum reservoirs.

Epidiagenesis is also well known to mining geologists,

being responsible for the ‘gossan’ sulphide ore bodies,

such as those of Rio Tinto, Spain. Figure 1 delineates

the boundaries of diagenesis.

Physical Diagenesis

The main processes of physical diagenesis are the

expulsion of connate waters concomitant with the

compaction of sediment. There is abundant evidence

of compaction in sedimentary rocks, notably obser-

vations in thin sections of warped mica flakes and

buckled fragmented shells. Arguably much of this

compaction occurs during shallow burial, aided by

seismic (including microseismic) shocks. Vibration

causes the packing of sediment to tighten as the fabric

adjusts and grains snuggle up close to each other.

Figure 1 Diagram to show the interfaces and pathways of

diagenesis.

DIAGENESIS, OVERVIEW 393

Theoretical studies show that loosely packed sand

may have a porosity of nearly 50%, but this is

reduced to 27% for the tightest packing possible, a

loss of 23%. Experiments reveal that poorly sorted

muddy sediments compact more than clean well-

sorted sands. This is because of the ease with which

clay and finer particles may be squashed into the

pores between the larger framework grains. Compac-

tion gradually increases with burial depth, resulting

in a concomitant loss of porosity. The exception to

this general statement is the condition termed over-

pressure, in which the fluid pressure within pores

diminishes the pressure at grain contacts and supports

the sediment load. Once sediment is lithified, how-

ever, compaction ceases, although rocks have a degree

of elasticity.

Graphs of porosity against burial depth for clay

and sand are different. Clay burial graphs are curved,

with the rate of porosity loss declining to become

linear at depth. Sand burial curves, on the other

hand, tend to be linear throughout. Intuitively, this

suggests that clay loses porosity quickly by dewatering

and compaction, whereas sands lose porosity more

by cementation. Not all physical diagenesis results in

porosity loss. Fracturing is an important process that

takes place only in lithified rock: sedimentary, igneous,

and metamorphic. Fracturing may occur in several

ways, most commonly tectonic, but also, of course,

anthropogenically around a borehole. Fracturing also

occurs due to the release of stress when rocks are

uplifted and the overburden pressure is diminished.

Thus, rocks subjected to epidiagenesis become frac-

tured, and fracturing is commonly well developed in

truncated strata beneath an unconformity.

Fractures are extremely important, not so much

because of the way in which they increase the porosity

of a rock, which may be minimal, but because they

cause a dramatic increase in the permeability of a pre-

viously impermeable rock. The resultant fractures

may permit the flow of petroleum, water, and miner-

alizing fluids, which may, in turn, re-cement the very

fractures that permitted their invasion.

Chemical Diagenesis

Chemical diagenesis includes mineral transformation,

recrystallization, cementation, and dissolution. A

wide range of minerals occur as cements in sediment-

ary rocks, precipitating in the pore spaces between the

framework grains. The principal pore-filling cements

are quartz, carbonate (calcite, siderite, and dolomite),

and clay. Cement is obviously a major destroyer of

porosity and permeability.

Chemical diagenesis, however, also includes the

dissolution of grains, matrix, and cement by corrosive

fluids. The grains most commonly dissolved are bio-

clasts, peloids, and ooids composed variously of ara-

gonite (during shallow burial of limestone) and

calcite. The matrix that is most commonly removed

is lime mud, micrite. Similarly, the cement that is most

commonly leached out is calcite, and other carbon-

ates, rather than quartz. These observations demon-

strate that dissolution results from the invasion of

a sedimentary rock by acidic rather than alkaline

pore fluids. Two sources of acidic fluid have been

suggested. It is known that carbonic acid is expelled

from shales prior to the emission of petroleum. It has

been suggested that the carbonic acid serendipitously

enhances the porosity and permeability of the poten-

tial reservoir, ahead of petroleum invasion. It is also

noted that there is nothing new in acid rain. Sedi-

mentary rocks exposed to epidiagenesis are subjected

to flushing by acidic meteoric water. This leaches out

carbonate grains, matrix, and cements. In limestones,

it will enlarge existing fractures, and generate exten-

sive zones of mouldic, vuggy, and cavernous porosity.

The caves may synchronously fall in to form collapse

breccias.

Dissolution, of whatever origin, will, of course,

increase the porosity and permeability of sedimentary

rock. Renewed burial will lead to renewed cementa-

tion, and the sedimentary rock will continue on its

downward path to metamorphism.

Summary

Diagenesis is the term that describes the physical and

chemical processes that take place in sediment after

deposition and before it reaches the threshold of

Figure 2 Diagram to illustrate diagenetic pathways and their

petrophysical responses.

394 DIAGENESIS, OVERVIEW

metamorphism. Physical processes include the loss of

contained fluid and compaction, which destroys po-

rosity, and fracturing, which enhances permeability.

Chemical processes include porosity-destroying ce-

mentation and porosity-enhancing dissolution. The se-

quence, deposition ! diagenesis ! metamorphism, is

not a ‘one-way street’, however. At any point along this

route, the sediment may be uplifted to the surface

again. Rocks returned to the surface show a reversal

of the trend of porosity decreasing with burial, and its

enhancement by the physical and chemical processes of

fracturing and dissolution, respectively. Figure 2 sum-

marizes the diagenetic pathways of sedimentary rocks.

Diagenesis is of great economic importance. Diage-

netic studies form an integral part of the analysis of

the evolution of porosity and permeability in aquifers

and petroleum reservoirs. Diagenetic studies are also

an important aid to understanding the formation of

many mineral deposits.

Further Reading

Burley SD and Worden RH (2003) Sandstone Diagenesis

Recent and Ancient. Oxford: Blackwell Science.

Dunoyer de Segonzac G (1968) The birth and development

of the concept of diagenesis. Earth Science Reviews 4:

153–201.

Fairbridge RW (1967) Phases of diagenesis and authigen-

esis. In: Larsen G and Chilingar GV (eds.) Developments

in Sediments, pp. 19–28. Amsterdam: Elsevier.

Giles MR (1997) Diagenesis: A Quantitative Perspective.

Dordrecht: Kluwer Academic Publishers.

Leeder MR (1999) Sedimentology and Sedimentary

Basins: From Turbulence to Tectonics. Oxford: Blackwell

Science.

Pettijohn FJ (1957) Sedimentary Rocks, 2nd edn. New

York: Harper Geoscience.

Selley RC (2000) Applied Sedimentology, 2nd edn. San

Diego: Academic Press.

Tucker ME (1991) Sedimentary Petrology, 2nd edn.

Oxford: Blackwell Scientific Publications.

DINOSAURS

See FOSSIL VERTEBRATES: Dinosaurs

DIAGENESIS, OVERVIEW 395

EARTH

Contents

Mantle

Crust

Orbital Variation (Including Milankovitch Cycles)

Mantle

G J H McCall, Cirencester, Gloucester, UK

Introduction

The mantle is the middle of the three primary concen-

tric zones of the Earth, making up together with the

core more than 99% of the planet’s volume. It extends

from the Mohorovicic Discontinuity (the base of the

crust, which varies in depth from 25–35 km below the

continents to 6–11 km beneath the ocean floor)

(Figure 1A) to its boundary with the outer core at a

depth of 2890 km.

The mantle itself is subdivided into an upper mantle,

extending from the Mohorovicic Discontinuity to a

depth of 670 km, and a much thicker lower mantle

beneath. The upper mantle is further divided into a

rigid upper zone, extending to a maximum depth of

about 100 km, which together with the crust comprises

the lithosphere, and a weaker ductile lower zone,

which forms the asthenosphere. The ductile lower

zone is commonly delineated at its base by a transi-

tion zone, extending from 400 to 670 km; the change

to the lower mantle is transitional rather than abrupt.

The lower mantle is delineated by a more rapid in-

crease in density, amounting to 20%, as shown by an

increase in the velocities of seismic waves. It is cus-

tomary to consider the thickness of the lithosphere

as varying from a maximum beneath the continents

to zero beneath the mid-ocean ridges, the sites of

eruptivity leading to sea-floor spreading (Figure 1A).

Direct Sampling of the Mantle

is not Possible

The mantle, like the core beneath it, cannot be

sampled directly. We shall never have revealed to us

the actual rock material of which it is composed.

Thus, all our knowledge of the mantle is derived

from indirect methods, and this means that there

remains a considerable degree of uncertainty as to its

exact composition. The indirect methods that have

been used to study the mantle are discussed below.

Seismology

Earthquake waves and similar seismic waves gener-

ated by artificial explosions (e.g. nuclear) provide a

major line of evidence. Their velocities depend on the

physical properties of the materials through which

they pass. Experiments are carried out in the labora-

tory to establish the behaviour of such waves when

passing through a known range of likely materials.

Radial and lateral velocities can then be compared.

Solutions, however, are seldom unique; there is also

the difficulty of replicating the very high pressures of

the deep mantle in the laboratory, so the method is

less satisfactory when considering increasing depths

in the mantle.

Chemical-equilibrium Studies

Experiments at high pressures and temperatures can

examine the exchange reactions between different

minerals and between minerals and magmas, and

the stability fields of minerals at different pressures

and temperatures, replicating the conditions at vari-

ous levels in the mantle (Figures 2 and 3). From such

experiments the possible or probable mineral and

elemental make-up of the mantle can be deduced,

but again there is no certainty in these results.

Peridotites and Oceanic Basalts

Materials erupting at the spreading centres in the

oceans are considered to be virtually uncontaminated

derivatives, ascending rapidly from layers in the

mantle, and valuable conclusions can be drawn from

the study of these. This method is believed to give

reliable information about the mantle to a depth

of about 300 km. Peridotite may well exist in the

deeper mantle, and some authorities entertain a

wholly peridotitic mantle.

EARTH/Mantle 397

Kimberlites

Kimberlites (see Igneous Rocks: Kimberlite) are mica

peridotites, which, unlike oceanic peridotites, are

volatile-rich. They carry mica and carbonate min-

erals. There are two ways in which these rocks,

which are erupted very rapidly and explosively thro-

ugh continental crust in diatremes or pipes, sample the

mantle. They contain diamonds with melt or mineral

inclusions that come from great depths in the mantle,

possibly as deep as the base at 2890 km. The dia-

monds in kimberlites are older than the host rocks,

as shown by isotopic evidence, and have clearly been

picked up during passage of the Kimberlite through

the mantle to the surface. There may also be mineral

or rock xenoliths, accidental fragments also picked

up by the turbulent host magma on its ascent. These

may range from comminuted debris to actual rock

fragments. Kimberlites occur only within the contin-

ents, so they provide information only about the com-

position of the subcontinental mantle. Such accidental

inclusions probably sample the mantle to a depth of at

least 150 km.

Meteoritic Analogy

Geochemical modelling has been based on analogies

with chondritic meteorites (see Solar System: Meteor-

ites) because

they are the most abundant extraterrestrial objects

that strike the Earth (as measured by fall rates and

excluding finds);

they are essentially similar to terrestrial ultrabasic

rocks, consisting of olivine-pyroxene assemblages

with similar Fe/Mg ratios; and

the stable-isotope data is in agreement with such

modelling.

However, there are difficulties in accepting the ‘chon-

drite Earth model’, a major one being the difference

in alkaline-element contents, and there is another

problem involving the

C isotopic ratio, which

is low in common chondrite meteorites and terrestrial

basalts, but high in carbonaceous chondrites, terres-

trial carbonates, and diamonds from kimberlite pipes

(the latter must come from the mantle). Thus, a

number of authorities believe that the mantle is de-

rived from the accreted material of carbonaceous

chondrites, rather than common chondrites. This

conveniently provides a source for the water and

other volatiles that are known to reside in the mantle

(e.g. according to the Rubey model). The fact that

Figure 2 Pressure variation curve for the interior of the Earth

(reproduced with permission from Mason B (1966) Principles of

Geochemistry.

New York, London, Sydney: John Wiley and Sons,

Inc.).

Figure 1 (A) Cross-section from a mid-ocean ridge to the inter-

ior of a continent, showing the relationship between the crust,

lithosphere, Moho, and zones in the upper mantle. (B) The rela-

tionships between the upper mantle, lower mantle, and core

(after Hancock PL and Skinner BJ (eds.) (2000)

The Oxford

Companion to the Earth

. Oxford: Oxford University Press).

398 EARTH/Mantle

carbonaceous chondrites fall to Earth extremely

rarely compared with common chondrites can be

explained both by their fragility and by the fact that

common chondrites, according to theory, are derived

from them by metamorphic processes. However,

doubts must remain about whether the material of

chondrites, which accreted as asteroids in the interval

between Mars and Jupiter at about 4500 Ma, is really

representative of the material from which the Earth’s

mantle accreted at about the same time. There is an

assumption of a homogeneous dust cloud from the

Sun outwards.

The metal forming the core is supposed to have

melted and gravitated to the centre of the planet

(Figure 4). Whether accretion of the mantle from the

dust cloud was homogeneous or heterogeneous

(layered) remains undecided – both possibilities are

shown in Figure 4.

Mantle Composition

Concepts concerning the composition of the mantle

have inevitably changed since the acceptance of the

plate-tectonic paradigm. It is relevant here to go back

to the earlier statements. Those of Arthur Holmes and

Brian Mason are concise summaries of what was then

believed.

Holmes, in a 1965 reissue of his famous book

Physical Geology, wrote

It has long been thought that meteorites, stony and iron,

might be direct clues to the nature of the Earth’s mantle

and core. Stony meteorites are like our terrestrial peri-

dotites in many ways and, moreover a few varieties have

compositions not unlike some of our basalts. For these

and other reasons a similar range of composition for

the mantle, with ultrabasic materials predominating,

has seemed to be a plausible guess. The sudden change

of seismic velocities at the Moho indicates either a

change in composition (e.g. from basic rocks above to

ultrabasic below) or a change in state (from, say, gabbro

or amphibolite to a high pressure modification, which,

for convenience, we may refer to as eclogite), without

much change in composition. Olivine-rich peridotites

and garnet-bearing varieties, at the appropriate tempera-

tures and pressures, give seismic velocities that are about

right for most parts of the upper mantle. Eclogite would

do equally well, and various associations of the rocks

would match seismic and density requirements. That

eclogite does occur in the mantle is proved by the fact

that it occurs in diamond pipes as inclusions.... More-

over, basaltic magma ascends from the mantle and

Figure 3 Temperature (degrees C) variation curve for the interior of the Earth. Lines are isotherms (reproduced with permission

from Mason B (1966)

Principles of Geochemistry. New York, London, Sydney: John Wiley and Sons, Inc.).

EARTH/Mantle 399

one very probable source of basalt supply would be the

melting of eclogite.

Although eclogites might represent primordial

mantle or subducted oceanic crust, one problem with

an eclogitic upper mantle is that eclogite xenoliths are

scarce in kimberlites, whereas peridotite xenoliths

are common. Mason, in a textbook reissued in 1966,

stated that the alternative of a chemical explanation

for the Moho was favoured by Clark and Ringwood

in 1964; they preferred an ultrabasic model for the

upper mantle, with the overall composition being one

part basalt to three parts dunite (an olivine rock). This

composition would equate to that of a pyroxene-

olivine rock, which they called ‘pyrolite’. This is a

hypothetical composition, not a rock type. Fractional

melting of this material would yield basaltic magma,

injected into the crust over geological time, leaving

residual dunite or peridotite. The pyrolite could

crystallize in four ways, depending on the pressure

and temperature in the mantle: olivine and amphi-

bole;. olivine, aluminium-poor pyroxene, and plagio-

clase; olivine, aluminium-rich pyroxene, and spinel;

or olivine, aluminium-poor pyroxene and garnet.

Because of pressure and temperature differences, the

compositions of the upper mantle beneath continents

and oceans would differ. The proposed compositions

under the continents and oceans are shown in Figure 5.

Polymorphism of minerals was considered to be

Figure 4 Two models for the accretion of the mantle and core (reproduced from Van Andel 1994).

400 EARTH/Mantle

an important mantle process: olivine converting to an

isometric spinel structure at higher density, pyroxene

to an ilmenite structure, and quartz to the high-

pressure polymorph that occurs naturally in impact

structures – stishovite with rutile, stable at 130 kb and

1600



C. Transformations into closer-packed forms

were suggested to occur in the transition zone. The

lower mantle was believed to consist of a mixture of

(Mg,Fe)SiO

with an ilmenite structure and periclase

(MgO), but the possibility of the ilmenite structure

converting to that of perovskite (even denser) was

suggested, as was the possibility of that then

converting to a CaCl structure (though experimental

evidence for these changes was lacking).

The pyrolite hypothesis, which was first stated by

Clark and Ringwood in 1964, has survived the app-

earance of the plate-tectonic paradigm and was

brought up to date in a summary by Agee published

in 2000. He noted that a major point at issue is

whether the mantle is homogenous or heterogeneous.

The abrupt jumps or discontinuities in seismic proper-

ties within the transition zone are attributed to changes

of minerals to denser, more closely-spaced phases; oliv-

ine is converted at a depth of 400 km to wadsleyite, an

orthorhombic form found in meteorites, and at 520 km

to ringwoodite (a gamma spinel form). At approxi-

mately 670 km olivine breaks down into perovskite

([Mg,Fe]SiO

) and magnesiowustite ([Mg,Fe]O). Pyr-

olite has been given a hypothetical composition of

57% forsteritic olivine, 17% enstatitic pyroxene,

12% diopsidic pyroxene, and 14% pyrope garnet.

Pyrolite is a hypothetical analogue of the primor-

dial composition of the mantle that existed after plan-

etary accretion and core formation (the rigid part of

the upper mantle, above the asthenosphere, is thus

termed the ‘residual mantle’)(see Earth Structure and

Origins). The shallowest part of the mantle is believed

to have been continuously depleted in basaltic melt

throughout geological time.

An up-to-date diagram of the likely composition

of the mantle at various depths under the continents

and oceans is given in Figure 6.

It has been noted above that there are two possibil-

ities for the original accretion of the mantle from

the dust cloud – homogeneous and layered. D L

Anderson and others favour a compositionally

layered mantle rather than the model of progressive

downwards change under increasing conditions of

pressure and temperature. The plate-tectonic para-

digm requires the solid upper mantle to be rheid and

slowly convecting as shown in Figure 7. Advocates of

Figure 5 Suggested compositions of the mantle according to

the original pyrolite hypothesis (reproduced with permission

from Mason B (1966) Principles of Geochemistry. New York, London,

Sydney: John Wiley and Sons, Inc.).

Figure 6 Suggested composition of the mantle at various

depths. The verticle dimension is depth, the horizontal dimension

is minerals (modified from Agee CE (2000) Mantle and core com-

position. In: Hancock PL and Skinner BJ (eds.)

The Oxford Compan-

ion to the Earth

, pp. 654–657. Oxford: Oxford University Press).

EARTH/Mantle 401

the compositionally layered mantle believe that the

discontinuity at the base of the upper mantle, at ap-

proximately 670 km, is not only a phase-change

boundary but also a compositional boundary. The

layered model requires layered convection with little

or no mass transport across this discontinuity. The

pyrolite hypothesis is retained, but, whereas the

upper mantle is taken to be composed of olivine-rich

peridotite similar to pyrolite, the lower mantle is

believed to be enriched in perovskite. This requires a

lower mantle that is enriched in silica compared with

the upper mantle, something that is in agreement

with the chondritic Earth model. The compositional

layering of the mantle could stem from the accretion

phase, but could also be a product of an early melting

stage, with the mantle crystallizing like a giant

layered igneous intrusion.

The Problem of Subducted Slabs

It is a requirement of the plate-tectonic paradigm

that lithospheric slabs descend into the mantle in

subduction zones. One hypothesis is that they sink

right through the mantle to a ‘slab graveyard’ at the

core-mantle boundary. There is some seismological

evidence for this in the basal 200 km of the mantle

(the D

zone). However, the seismological anomaly

here may be explained in a number of ways, and

advocates of the compositionally layered mantle be-

lieve that there is a high enough contrast of density

and viscosity across the 670 km boundary to prevent

any passage of material through it and that the slabs

must remain in the upper mantle, the ‘graveyard’

being situated in the transition zone and affecting its

composition. The sedimentary cover in these slabs

contributes to the calc-alkaline eruptives in the crust

above the descending plate (e.g. the Peruvian granit-

oid batholiths). There is no real certainty about what

happens to the sediments descending into the mantle.

Tomography

Tomography delineates in three-dimensions regions of

fast and slow seismic-wave in the Earth’s interior, both

radially and laterally. These may simply represent

colder downwellings and hotter upwellings. They

may alternatively represent compositional differences

in a primordially layered mantle being complicated

and eradicated by convection. Tomographical irregu-

larities are invoked by advocates of slab descent to the

base of the mantle and of plumes ascending through

the 670 km discontinuity from that region (see Earth

Structure and Origins), whereas advocates of ‘closed

upper mantle circulation of plate tectonics’, such as

Warren Hamilton, believe that plumes could not pene-

trate the 670 km discontinuity and that convection

cells are restricted to the upper mantle.

What Drives Plates and Initiates

the Pattern?

The plate-tectonic paradigm has been with us for four

decades, but there is no certainty at all about what

drives the plate motion. What actually starts the pro-

cess? Is the rise of magma from the already convecting

mantle at the site of the future mid-ocean ridge the

splitting initiator? Or is initiation controlled by litho-

spheric irregularities above the mantle before convec-

tion begins? That is, is the primary drive exerted by

the mantle or by the lithosphere? Though there has

been much discussion of the driving forces of plate

movement, none of the theories suggested seems ad-

equate. However, we do have a clue in the Rift Valley

of East Africa (see Africa: Rift Valley). This is mani-

festly an aborted ocean, a site of magmatism and

rifting, which has, nevertheless, not developed into a

spreading ocean – it has been ‘halted in its tracks’.

This surely is evidence that the process starts at the

spreading centre, before any subduction takes place. It

would seem to invalidate slab pull or the passive sink-

ing of oceanic lithosphere, as favoured by Warren

Hamilton, as the prime driving mechanism; rather,

either already commenced mantle convection forces

a split or the whole process is started by a lithospheric

split along a favourable line of weakness in the crust–

both mechanisms seem to be consistent with this

evidence.

Conclusion

The mantle is of critical importance in understanding

geological processes, yet remains elusive. Numerous

models have been erected, and no doubt many more

will be derived in the coming decades. An aspect that

has not been touched on here, the variations in iso-

topic composition in basalts derived from the mantle,

Figure 7 Mantle convection under a spreading ocean (repro-

duced from Van Andel 1994).

402 EARTH/Mantle

illustrates the complexity of the problems. It is not

possible to explain all the isotopic complexities by

simple mixing of one enriched and one depleted res-

ervoir, and the answer would seem to be that, even

taking into account crustal contamination on ascent,

there are a number of different reservoirs, a fact that

argues for a heterogeneous mantle. Yet the exact

petrological, spatial, and physical nature of the source

regions remains as elusive as ever.

Elsevier Encyclopedia of Geology - vol I A-E

Подождите немного. Документ загружается.