Click here to go to 'Figures/Overheads' section.


Click here to return to course outline.



    The heat of the Earth is composed of three components: a one-time thermal increment derived from the release of gravitational energy during the early accretionary stage of formation of the Earth, a one time thermal increment during the formation of the Earth's core, and a time integrated component released by the transformation of radioactive elements contained within the Earth..

Heat produced within the earth is transferred to the Earth's surface by three processes - conduction, radiation, and convection.


    Within the relatively thin insulating outer part of the Earth ( the lithospheric shell/lid) heat is transferred by conduction. However, within the deeper mantle (asthenosphere) heat transfer is effected by thermal convection; in other words, by the physical transport of hot solid but plastic (viscous) material towards the Earth's surface.


    As the average temperature of the Earth decreases, the lithosphere grows downwards and becomes more effective as a thermal insulator. For this reason the temperature gradient within the Earth decreases to a self-regulated minumum value. It is currently estimated that although the rate of radioactive heat production 3 billion years ago was twice the rate it is today, the mean temperature of the mantle at that time was only 150 degree K higher than its present value.

    As convection currents approach the Earth's surface, the mantle, composed of about 60% olivine and lesser amounts of pyroxene and spinel (which nevertheless contain all the Al2O3, Ca, and Na in the mantle), undergoes isothermal decompression melting,


and the basaltic magma so formed leaks to the Earth's surface at thermal 'hot spots' and along the mid-ocean ridges of ocean basins. The formation of a basalt melt can be envisaged as the removal of incompatible elements from a mantle composed of oxygen maintained in a solid state by dissolved atoms of silicon and magnesium. The decompression melting of the mantle in this manner represents a first-order material differentiation event, and the basaltic magma produced constitutes the primary material from which all other rocks are formed by secondary fractionation processes - collisional or hot spot igneous melting to form granites, metamorphism (e.g. greenschists, blueschists, granulites), hydrothermal alteration, weathering and sedimentation. (Over time fractionation has tended towards the building of continental crust enriched in Ca, Na, K, and Al; a hydrosphere enriched in alkalies; and an atmosphere enriched in oxygen..)

    MANTLE CONVECTION - Forte  (click here)


    As the basalt formed at a mid-ocean ridge cools, part of its heat energy is lost by conduction to the overlying sea water and part is converted to chemical energy by endothermic reaction of the basalt minerals (Cpx, Opx, Plagioclase) with sea water to form amphibole, epidote and haematite (Fe2O3). These minerals are then transported by the process of 'sea-floor spreading' to zones of subduction where they pass back into the mantle. At depths of the order of a 100 km, the amphibole-epidote rocks undergo an exothermic dehydration reaction to form an anhydrous assemblage of pyroxene (jadeite) and garnet (the high pressure eclogite assemblage) and a hydrous phase highly charged with alkali (Ca, Na, K), Si,Al, and even Ti cations. The water and heat released by the exothermic reactions rises into the overlying mantle, which consequently melts to produce an oxidized magma which rises to the surface to form the island arcs found adjacent to subduction zones.

    Minerals and Tectonic Environments

    The existence of a given mineral assemblage in a rock is controlled by the chemical composition of the rock and the physical conditions (temperature, pressure) at the time of its crystallization. Particular mineral assemblages may therefore characterize specific tectonic environments. As explained above, the assemblage pyroxene (jadeite) and garnet forms in rocks of basaltic composition (eclogite) under metamorphic conditions representative of deep subduction slip zones or continental collision zones, whereas olivine rich assemblages (peridotite) characterize high temperature mantle or igneous environments, and rocks composed of quartz and kaolinite (quartzites and clay shales) form at the Earth's surface under conditions of intense chemical weathering. The following table illustrates this concept:

Igneous minerals found in rocks formed at oceanic spreading centres and island arcs

Olivine         Forsterite - Fayalite   Mg2SiO4 - Fe2SiO4 [(Mg,Fe)2SiO4]
Clinopyroxene   Diopside - Hedenburgite CaMgSi2O6 - CaFeSi2O6 [Ca(Mg,Fe)Si2O6]
Orthopyroxene   Enstatite -Ferrosilite  Mg2Si2O6 - Fe2Si2O6 [(Mg,Fe)2 Si2O6
Plagioclase     Anorthite - Albite      CaAl2Si208 - NaAl Si3O8
K-feldspar      Orthoclase, Microcline  KAlSi3O8

Felspathoids    Nepheline               NaAlSiO4
                Leucite                 KAl Si2O6

Spinel          Hercynite - Chromite    MgAl2O4 - FeCr2O4
                 - Magnetite            FeFe+32O4; (all mutually soluble = (Mg,Fe)(Al,Cr,Fe+3)2O4
Ilmenite                                FeTiO3
Apatite                                 Ca5(PO4)3.OH,F,Cl (O=24; F,Cl,OH =1; Ca=10; P=15)
Pyrite                                  FeS2
Pyrrohtite                              Fe(1-x)S

Minerals formed from Igneous Minerals during Metamorphism, Diagenesis, and Weathering.


Amphibole       Tremolite-Actinolite    Ca2Mg5 Si8022(OH)2 - Ca2Fe5 Si8O22(OH)2
                                        (Ca=4; Mg,Fe=10; Si=32; H=2; O=48)
                Hornblende              Ca2(Mg,Fe)4Al2 Si7O22(OH)2 (coupled substitution)
                Anthophyllite (not common) (Mg,Fe) 7 Si8O22(OH)2
Epidote         Epidote-zoisite         Ca2Al3 Si3O12 (OH)
Talc                                    (Fe,Mg)3 Si4O10(OH)2
Serpentine                              (Fe,Mg)3 Si2O5(OH)4
Chlorite                                (Fe,Mg)2Al2 SiO5(OH)4 (note the coupled substitution)
Mica            Muscovite               KAl3 Si3O10(OH)2
                Phlogopite -Biotite     KAlMg3 Si3O10(OH)2 - KAlFe3 Si3O10(OH)2


Titanite (Sphene)                       CaTiO3

Metamorphic minerals formed from oceanic crust during subduction (relatively high pressure)

Glaucophane                             Na2(Mg,Fe)3Al2 Si8O22(OH)2
                                        (O=48; Na =2; Mg,Fe = 6; Al = 6; Si = 32; H= 2)
Lawsonite                               CaAl2 Si2O8.2H2O

Clinopyroxene   Jadeite (Omphacite)     NaAl Si2O6 (cf Albite)
                Ca-Tschermaks Molecule  CaAl2 SiO6 (cf Anorthite) (coupled substituion)
Garnet          Pyrope - Almandine -    Mg3Al2 Si3O12 - Fe3Al2 Si3O12 - 
                Grossularite            Ca3Al2 Si3O12
Rutile                                  TiO2
Aragonite                               CaCO3

Minerals formed by weathering of feldspars:

Clay Minerals

Kaolinite (e.g. weathering of plagioclase) Al2 Si2O5(OH)4 (2-layer clay)
Pyrophyllite                               Al2 Si4O10(OH)2 (3-layer clay)
Smectite                                  (Mg.5,1.75Al)(Si3.75,Al.25)O10(OH)2 (3 layer) (coupled substitution)
Gibbsite                                   Al(OH)2

Carbonate Minerals
Calcite, Dolomite, Siderite, ankarite   CaCO3, CaMg(CO3)2, FeCO3, Ca(Fe,Mg)(CO3)2

Minerals formed by metamorphism of clays

Andalusite, kyanite, sillimanite         Al2 SiO5
Mullite                                  Al6 SiO11
Staurolite                               Fe2Al9 Si4O23(OH)
Chloritoid                               FeAl2 SiO5(OH)2
Cordierite                              (Mg,Fe)2Al4 Si5O18
The Primary Rock Categories

    Rocks are generally subdivided into four types: mantle, igneous, sedimentary, and metamorphic rock. Since metamorphic rocks form from  rocks that were initially of  igneous or sedimentary origin, the supposed original rock type is referred to as the protolith of the metamorphic rock. For example, the protolith of a greenschist in which it is still possible to recognise flattened or stretched remnants of pillow structure would be basalt. In this case the greenschist may legitimately be referred to as 'meta-basalt'.

    Rock Attributes

    In many cases the mineralogical composition of a rock cannot be used alone to determine the origin of a rock - for example, a rock composed of quartz and feldspar could be of  igneous, sedimentary or metamorphic origin. (On the other hand, the co-existence of  quartz and kaolinite or kyanite or staurolite might imply that their protolith was a clay-rich sediment formed by tropical weathering of plagioclase-rich rocks such as granite.) Determining the origin of a rock usually also requires evaluation of its textural and structural characteristics.


    Structure refers to the arrangement of texturally and/or compositionaly different parts of a rock, the interrelation of all the parts of a whole. It is usually evident at the hand specimen or outcrop scale, and is the most important attribute of the rock from the point of view of identifying the environment in which it formed.

    Bedding is a planar structure that might be evident because of differences in the composition or grain size of adjacent beds. Some patterns are related to the influence of gravity and the laws of fluid flow during the formation of the rock, as in the case of cross- and graded-bedding in sandstones, or cumulate layering in igneous rocks, or the Bernoulli-like agglomeration of large plagioclase crystals in the centre of some igneous dikes. Other types of structure may reflect the influence of the addition of heat as in the case of hornfelsed metamorphic rocks; rate of loss of heat in the case of pumice; of pressure as in the case of folding, joints, and cleavage; or biotic activity in the case of stromatolitic limestone. Structure is essentially a heterogeneous attribute, e.g.  pillows; concretions; sills; dikes; reefs; sand volcanoes; ball and pillow structures; sand volcanoes.


    Texture refers to the pattern arrangement of the constituent parts of what is otherwise a homogeneous material. Two adjacent beds (structure) with the same proportion of quartz and plagioclase (i.e. same mineralogical composition) may differ because the minerals are coarse grained in one bed and fine grained in the other. The difference between the two beds is one of texture. Igneous rocks of mafic composition that cool fast (basalt) are fine grained, whereas when cooled slowly (gabbro) they are coarse grained. Texture is usually a homogeneous feature. Many textural attributes are apparent only when examination of a rock is carried out at high magnification. Sometimes, as in the case of cleavage, an attribute may be referred to as both a structure (outcrop scale) and a texture (hand specimen or thin section scale). The term fabric is also commonly employed where the minerals exhibit a preferred orientation, e.g. cleavage or linear fabric. If the fabric is the result of deformation the rock is known as a tectonite.

    To some extent the difference between structure and texture is scale dependent, and what is commonly referred to as structure at ground level, e.g. regularly folded bedding, may appear as a texture on a satellite image. The difference between these two attributes is therefore not always clear cut.







Click here to return to beginning.

Click here to return to course outline.

Towards a Grand Unified Model of Earth Dynamics

The quest for a grand unified theory of the four fundamental forces (gravitational, strong and weak
nuclear forces, electromagnetic) in nature continues to elude the efforts of modern theoretical
physicists. This search is motivated by the conviction that all physical phenomena in nature should
ultimately be explained by a single unifying theory. In analogy to this quest for grand unification by
physicists, the Earth Sciences have also seen a concerted effort to develop an integrated theory which
explains the interactions between Earth's component parts and which explains all surface physical
phenomena resulting from these interactions. The goal of this geophysical grand unification is to
develop a theory or model which can successfully reproduce the evolution and dynamics of our planet.

An important first step occurred in the early 1960s, with the advent of what is now called the `plate
tectonics revolution'. The theory of plate tectonics is based on the recognition that Earth's crust is
divided into a small number of large, essentially rigid plates which move relative to each other. This
theory provided a framework for understanding the connections between a diverse assortment of
surface phenomena such as the geographic variations in depth of the world's oceans, mountain
formation, earthquakes, volcanism, geographic variations in crustal heat flow, the complex geologic
structures in localities distributed across the globe, and of course the famous horizontal movements or
`drift' of the continents over geologic time. While plate tectonics has helped to clarify our
understanding of the relationship between these many processes, it remains an essentially descriptive
or kinematic theory. It does not provide a model which explains the long suspected causal link between
dynamical processes which may occur deep within the Earth and the resulting surface phenomena (for
example, continental drift) listed above.

Geophysicists have recognized that a unifying dynamical model which can explain the many surface
processes described by plate tectonics should be based on a theory of global scale motions deep in the
solid Earth which occur on very long time scales, ranging from a million years to several billion years.
They have specifically focussed their attention on a region of Earth's interior called the `mantle', which
begins at about 30 km depth, below the crust, and extends down to the top of Earth's molten core at
about 2900 kilometers depth. Geophysicists hypothesize that this thick massive shell of rock is
animated by global scale movements of material across thousands of kilometers with speeds of several
centimeters per year. (This is approximately the speed with which human nails grow!) That this
movement can occur at all, is due the experimentally observed fact that all rocks, if they are sufficiently
hot, can creep or flow very slowly when they are subjected to forces. This displacement of material in
Earth's mantle is called `thermal convection', and it is akin to the convection which occurs in a pot of
water heated on a stove top. The main difference is that the effective viscosity or stiffness of mantle
rocks is vastly greater than the viscosity of water. Thermal convection is in effect a heat engine which
provides an efficient way to transport heat from great depth in the Earth to the surface. (This heat was
originally present when the Earth first formed and it is being released by the cooling of the molten core
and by naturally occurring radioactive elements distributed in mantle rocks.)

Progress in understanding thermal convection in the mantle has been hampered by the obvious difficulty
in directly `seeing' the structure of our planet deep below the surface. A major breakthrough occurred
in the early 1980s when seismologists developed a variety of techniques, called `seismic tomography',
which provide us with direct images of the global three dimensional structures deep inside Earth's
mantle. Seismic tomography uses the waves created by earthquakes which travel in criss-crossing
paths across the mantle, in analogy with X-ray CAT scans used in medical imaging by doctors. The
coverage and resolution of the seismic tomographic images have improved greatly over the past few
years thanks to a signification expansion of the global network of modern digital seismic stations. The
huge increase in the number of high quality seismic data is mirrored by a corresponding explosion in
high precision satellite observations of the global topography, gravity and magnetic fields. There has
also been a a rapid growth in land and space based geodetic measurements which provide continuous
observation of the subtle changes in Earth rotation which occur on a wide range of time scales.
Geologists have also made significant recent progress in analyzing and unraveling temporal and spatial
variations in the surface elevations of continents on very long time scales and also in determining
paleoclimate histories that extend back several million years. This rapid increase in data relating to the
structures in Earth's deep interior and to the many surface phenomena, which may eventually be
understood in terms of the dynamics of these deep-seated structures, has brought us to the threshold of a
profoundly increased understanding of the workings of our planet's internal heat engine.

In an article published in the April 26, 2001, issue of Nature magazine, Alessandro Forte (at the
University of Western Ontario) and Jerry Mitrovica (at the University of Toronto) provide a model of
mantle convection which integrates the most recent data from seismic tomography, from mineral
physics, and from a wide range of global surface geophysical measurements. This multidisciplinary
approach has yielded a unified understanding of mantle dynamics which, to date, is the most
comprehensive model of the link between global data sets (including the tectonic plate motions,
perturbations in Earth's gravity field, variations in dynamic topography at the surface and at the
boundary between the mantle and the core) and the tomographically imaged three dimensional
structures in the mantle. The study published by Forte and Mitrovica also reveals, for the first time, the
presence of a huge increase in the effective viscosity or stiffness of mantle rocks at nearly 2000
kilometers depth and the profound impact of this increased stiffness on the dynamics of the thermal
convective circulation in the mantle. Their analysis also provides the best available maps of the
temperature structure and of the chemical structure in the deep mantle.

The model of mantle dynamics presented by Forte and Mitrovica focussed in particular on the
properties of the global scale structures in the deepest portions of the mantle, especially in the bottom
half of the lower mantle. The seismic tomographic images at these great depths clearly show the
accumulation of slabs of cold dense oceanic plates which appear to have plunged down into the mantle
below the deep oceanic trenches around the periphery of the Pacific basin (blue coloured regions,
Figure 1). Especially large slab accumulations may be found deep below the west Pacific margin, in a
broad arc extending from northern Japan to the Indonesian archipelago, and below the east Pacific
margin extending from central North America to South America. The seismic tomographic models also
reveal the presence of mushroom-shaped blobs of presumably hotter than average rock with horizontal
dimensions of several thousand kilometers. These remarkable, immense structures appear to sit at the
bottom of the mantle and they tower upwards to heights exceeding 1000 kilometers (red coloured
regions, Figure 1). There are two such `mega-plumes' of hot rock, one below Africa and one on the
other side of the Earth below the south-central Pacific Ocean. Recent, more detailed tomographic
images clearly show the African mega-plume extending right to the top of the mantle. This mega-plume
thus appears to provide an obvious source for the many active and recently active volcanos which are
found in an arc-shaped region, which includes the East African Rift, and stretches from the Red Sea in
the north to Madagascar in the south. On the opposite side of the globe, the Pacific mega-plume appears
to provide the source region for the numerous volcanic islands throughout the central part of the Pacific
and, in particular, the chain of volcanic islands in French Polynesia and the Hawaiian island chain
further north.

In the development of their dynamical model, Forte and Mitrovica addressed a fundamental unresolved
question which has been hotly debated over the past few years by global geophysicists. The question
concerns whether the the mega-plumes are buoyant and actively rising upwards, like gigantic hot-air
balloons, or whether they are stagnant and sitting passively at the base of the mantle. The answer to this
question has profound implications for the dynamics of thermal convection and for the thermal and
chemical evolution of Earth's interior over geologic time. The seismic tomographic images (Figure 1)
only provide an instantaneous present-day snapshot of the structures in the mantle and therefore they
cannot by themselves indicate whether the mega-plumes are currently moving upwards. Most
geophysicists do agree that oceanic slabs are colder than the surrounding mantle and that they are
actively sinking because they are heavier than the surrounding rocks in the mantle (Figure 2).

There is, however, no consensus on the dynamics of the mega-plumes. In independent studies carried
out over the past few years by other geophysicists, idealized computer models of mantle convection
were used to suggest that the mega-plumes are located in very ancient, primitive mantle which has
remained unmixed and separate from the continuously recycled mantle at shallower depths. It was
therefore argued that the chemical composition of the mega-plumes is different from that of the rest of
the mantle. According to this view, the distinct composition (in particular the iron concentration) of the
mantle rocks in the mega-plumes makes them intrinsically heavy and therefore prevents them from
rising, despite the reduction in weight which results from their hotter temperatures (Figure 2). This
hypothesis of stagnant mega-plumes is in contradiction to work published several years earlier by
Forte and Mitrovica, and more recently by other researchers, who argued that the mega-plumes must be
actively rising upwards in order to explain some basic geophysical and surface geological

The central question tackled in the Nature study by Forte and Mitrovica can be colourfully rephrased as
follows (Figure 2): "Does the Earth's internal heat engine function on only two pistons (the cold
descending slabs) or with a full set of four pistons (descending slabs and rising mega-plumes)?" The
answer to this question also determines whether the Earth's interior is being cooled only from above,
with the descent of cold slabs, or whether it is also being cooled from within, with the active upward
transport of heat in the buoyant mega-plumes. The approach taken by Forte and Mitrovica to determine
if the mega-plumes are light or heavy is to directly determine the weight they must have in order to
explain a number of fundamental geophysical observations such as the small, but measurable,
geographic variations in Earth's gravity which can determined by orbiting satellites.

Another fundamental constraint on the weight of the mega-plumes is provided by the extra flattening of
the external shape of Earth's core. This flattening is such that the distance from the center of the Earth to
the point on the core's surface which lies directly below the North Pole is about half a kilometer
shorter than the distance from the center of the Earth to any point on the core's equator. This flattening
is quite small compared to the average radius of the core, about 3480 kilometers, but it can be
accurately determined by analyzing astronomical observations of subtle wobbles, called `nutations', in
Earth's rotation.

Forte and Mitrovica discovered that they could explain the global gravity data, and the measured core
flattening, only if the mega-plumes were lighter than the surrounding mantle. They calculated the
movement in the mantle that would be generated by these buoyant plumes and they found that at depths
halfway down into the mantle, the plumes rise upwards with a speed of about 2 centimeters per year,
nearly the same speed with which the cold slabs sink at this depth. This speed is comparable to the
measured speed of horizontal drift of the continents, about 4 centimeters per year on average.

Using results obtained from the physics of minerals and using the weight of the mega-plumes, Forte and
Mitrovica determined that these plumes do appear to contain some extra iron, which would make them
heavy, but they also discovered that the high temperatures of the rocks in the mega-plumes adds more
than enough lightness to offset the effect of extra iron on their weight. Instead, they found that main
controlling influence on the upward rise of the mega-plumes comes from the enormous increase in
stiffness or viscosity of mantle rocks at about 2000 kilometers depth. This layer of very high stiffness,
discovered by Forte and Mitrovica in their effort to explain the surface gravity observations,
effectively acts as a powerful brake which slows down the speed of thermal convection in the deep
mantle. It is this region of increased stiffness that therefore controls the efficiency with which the
mantle heat engine cools the Earth's interior and the rate of mixing or stirring generated by the mantle

Knowledge of the dynamics of the mega-plumes also holds the key to understanding the remarkable and
surprising impact which the mantle heat engine may have on climatic variations in the geologic past.
The changes in position of the mega-plumes, during their upward rise, can modify the gravitational tugs
exerted on the Earth by the Sun, Moon, and other planets in our solar system, in particular the two
giants, Jupiter and Saturn. These changes in the gravitational forces exerted on the Earth can
significantly perturb the orientation of the Earth relative to the Sun and therefore change the amount of
sunlight received by different locations on Earth's surface. These variations in insolation have long
been understood to be the trigger for the cyclic appearance and disappearance of vast glaciers on
continents. (This effect of mantle convection on paleoclimate was first discovered in a study published
back in 1997, again in Nature magazine, by Forte and Mitrovica. Research is currently being carried
out to verify the previous findings using the most recent dynamical model obtained by Forte and


FIGURE 1. The dominant large scale structure in Earth's lower mantle. The three dimensional seismic
tomographic model shown here was derived by Forte, Woodward, & Dziewonski [J. Geophys. Res.,
vol. 99, pp. 21,857 -- 21,877, 1994]. The structures shown in this image have been mathematically
smoothed and are characterized by horizontal length scales greater than about 4500 kilometers.
Laboratory experiments on rocks show that the waves generated by earthquakes will travel faster in
rocks which are colder and will travel slower in rocks that are hotter. The blue colours in this figure
show the regions of the mantle in which earthquake waves are speeded up, presumably because of the
colder temperatures in these regions. The red colours show the regions of the mantle where earthquake
waves are slowed down, suggesting that these huge mushroom-shaped features or `mega-plumes' have
hotter temperatures.

FIGURE 2. How many pistons in Earth's internal heat engine? This figure is a schematic representation,
in an equatorial slice of the Earth's interior, of the dominant structures shown in Fig. 1.