The Earth is a hot, dense planet in a cold, sparse Universe. At the most basic level the processes that govern the Earth's behaviour are simple physical consequences of this reality, combined with the constraints imposed by the materials that make up the Earth. In particular, heat transfer operates to disperse this thermal energy fluctuation while self-gravitation serves to hold the mass-anomaly together. In order to understand why the Earth is the way it is, and how it evolves in time, it is necessary to understand the consequences of these two, simple, physical processes. The intricate structure of the Earth, at all scales, is a direct consequences of these physical processes, and why it is made of the materials that it is. Moreover, these processes are responsible for the great diversity and beauty of our planet that provides the basic motivation for much of the Earth Science community - a motivation that should not be forgotten.
Dynamicists are concerned with changes (equivalently, motions) accompanying the passage of time in dynamical systems. The system of concern to the geodynamicist is the Earth and our geological perspective focuses our interest primarily on the lithosphere (on which we reside) and, less directly, the asthenosphere. Consequently, geodynamics is primarily concerned with the basic physical processes that modulate the time evolution of the lithosphere; with one of its principal aims being the description of the behaviour of the lithosphere in the modern Earth.
In order to develop an appropriate geodynamic description we need to understand what constitutes utility. To be useful, the dynamical description must reduce some aspect (optimistically, all aspects) of the behaviour of a complex system to a few general statements. In geodynamics, in which our concern is primarily with the motion of the lithosphere and the forces that drive the motion, we may distinguish between descriptions that are concerned solely with the motion or kinematics without regard to the forces or, alternatively, mechanical descriptions concerned with the interaction between forces and motions.
The kinematic description is far less formidable than the mechanical description and has received considerably more attention as illustrated by plate tectonics. Plate tectonics attempts nothing more than a description of the motion of the lithosphere, and as such is purely kinematic. The fundamental tenets of the plate tectonic description are:
Plate tectonics can only provide a useful description if the number of plates remains small; if large numbers of plates are needed to describe the behaviour of the lithosphere, then the description becomes sufficiently complicated to be rendered useless. Does plate tectonics provide a useful description of the behaviour of the lithosphere? Our prejudice concerning the answer to this question - only in part - provides the key to the way in which we present the subject matter. Let us explain! Since every seismic event represents a deformation of the lithosphere, a logical consequence of the plate tectonic description is that a plate boundary should be drawn through the focal point of every seismic event (or, earthquake). In the ocean basins seismicity is restricted to very narrow zones along the mid ocean ridges, subduction zones and transform zones, that are separated by large, essentially aseismic regions. The lack of seismicity over large regions within the ocean basins implies that, to a first approximation, they do behave as rigid, or elastic, plates. In contrast, seismicity in the continents is widely distributed, even in relatively inactive continents such as Australia, with very large areas of intense, distributed seismicity. The most notable regions of distributed seismicity in the modern earth are the Basin and Range Province in the western USA, the Himalaya - Tibet region in Asia and in the Aegean Sea (which is floored by thin continental crust) in Europe. In these locations, active deformation is taking place over vast areas, which are equally as large as the surrounding aseismic regions. In order to adequately account for such deformation the plate tectonic description would need to invoke hundreds of thousands of individual plates each with lateral dimensions of only a few kilometers.
Clearly, while plate tectonics may provide a useful description of the ocean basins it usefulness for many continental regions is doubtful. Indeed, an alternative description of the continents as essentially continuous ductile media (rather than elastic plates) and which therefore effectively ignores the existence of any discontinuities in the strain field such as faults and, more importantly, plate boundaries is useful for many purposes and will be the description we will emphasize here.
At the very outset it is useful to establish our agenda in relation to other Earth science studies, such as structural geology, petrology, sedimentology, geomorphology and seismology. The overlap with these fields is substantial, but significant differences remain. We will be concerned primarily with those processes that give rise to the large-scale, regular (or so-called first-order) features that characterize the modern Earth. For example, we will be explicitly interested in understanding the controls on
We give due warning that we will not generally be concerning ourselves with the small-scale, higher-order variations in topography, strain, composition, and thermal regimes that are very much the stuff of geomorphology, structural geology and petrology, respectively. For example, we will ignore such things as the detailed lithostratigraphic record of individual sedimentary basins, but we will expend considerable effort in trying to understand the controls on subsidence within basins in general. Similarly, the role of assimilation, fractionation and crystallisation that leads to the great variation of magma types that so excite igneous petrologists will be overlooked in favour of the general nature of the processes that lead to large-scale melting events and the way in which these processes naturally give rise to the characteristic chemical signature of the average melt product.
The great challenge and fun of the geological enterprise lies in deciphering the effects of the first-order processes from the higher-frequency geological noise that make the geological record so complex (and, let it be said, at the same time so rich). From the geodynamic point of view the distinction between the first-order processes and the high-frequency noise is readily appreciated once the significance of the lithosphere as the fundamental tectonic entity is acknowledged. This implies for first-order features a natural length scale of the order of 100 km (or greater), as dictated by the thickness of the lithosphere, and natural timescales of the order of 10-100 Ma (at least), as dictated by the thermal response time of the lithosphere.
Where from comes the noise? At least a large part of the rich and diverse geological noise is modulated by exogenic processes. For example, whereas landscapes are very regular when viewed at the 105 - 106 m-scale (ie., at lithospheric scales), erosion and gullying have natural length-scales that produce very fragmented landscapes on the 101-104 m-scale. Similarly, exogenic forcing, as for example associated with Milankovitch cycles, with natural time scales (102 -106 a) that are very short compared to the thermal time constant of the lithosphere (107 a), influence the third- to fifth-order sequence boundaries that are so obvious in the lithostratigraphic record at the 10-1 - 102 m-scale. It is the rheological anisotropy caused by this short-wavelength lithological variation which is responsible for the structural chaos of orogenic zones when viewed at the map scale, and is the detail which we seek quite deliberately to avoid.
As we have emphasized one of our primary concerns is to provide an insight into the nature and behaviour of lithosphere and, particularly, the processes that contribute to its formation. To begin with we need some common idea of what constitutes the lithosphere. We provide two working definitions (Figure 1.1):
Both definitions allow the distinction of the lithosphere from the underlying asthenosphere, where heat transfer is primarily by convection, and which is, as a consequence, characterized by large vertical gradients in the horizontal component of the velocity field.
The lithosphere comprises the crust as well as a significant part of the upper mantle, down to a depth of 100 - 150 km. The exact thickness of the lithosphere is not easily determined by conventional geophysical methods such as seismology since the lower boundary of the lithosphere, the lithosphere - asthenosphere boundary, is not defined by a prominent compositional discontinuity. Rather it is defined by changes in the rheology of the mantle-forming rock, peridotite. Since the composition of the mantle is essentially constant throughout, mantle density, rm , is dependent primarily on the temperature (through the volumetric coefficient of thermal expansion, a) and the confining pressure (through the coefficient of compressibility, b). An increase in the temperature causes a decrease in density, while an increase in pressure causes an increase in density. At a constant depth and hence at constant pressure small lateral changes in temperature result in variations in the density. The strength of all materials, including peridotite, is a function of temperature; at asthenospheric temperatures (approximately 1300 oC) peridotite ( for peridotite a is about 10-5 oK-1) is sufficiently weak that it will convect in response to stresses resulting from the density variations associated with small lateral temperature gradients whereas in the lithosphere peridotite is sufficiently strong that it does not undergo convection (although it may subduct as an essentially rigid slab). Since the rheological properties of peridotite are likely to be a continuous function of temperature in the range appropriate to the base of the lithosphere there must be a transition zone separating the lithosphere and asthenosphere. This zone is termed the lower thermal boundary layer and is formally included in the thermal lithosphere, but excluded from the mechanical lithosphere (Figure 1.1).
It is important to realise that these definitions make no direct correspondence between the lithosphere and the compositional structure of the earth. The compositional structure of the earth is constrained in a number of ways, most notably by the seismic velocity structure determined by the way in which seismic energy is radiated following large earthquakes. This approach leads to the distinction of three compositionally distinct shells within the earth bounded by prominent discontinuities in the seismic velocity structure, namely the crust, the mantle and the core. The crust is typically about 6-7 km thick beneath the oceans and 30-50 km thick in the continents with maximum thicknesses of about 70 km beneath the great mountain belts. The detailed velocity structure of the continental crust is complex reflecting a complex compositional structure and history. In comparison the velocity structure of the mantle is more regular with weak discontinuities at about 450 and 650 km, consistent with phase transitions in material with the composition of peridotite. The 650 km discontinuity, corresponding to the maximum pressure stability limit of the silicate spinel structure, is commonly taken as the boundary between the upper and lower mantle. A low seismic velocity region is sometimes present in the upper mantle at depths between 100-200 km. In this scheme, the lower boundary of the lithosphere occurs within the upper mantle probably at depths appropriate to the low velocity zone at around 100-200 km. However, it should be noted that the low velocity zone is not universally present and therefore cannot be used to define the base of the lithosphere.
Already we have suggested that we may need two descriptions to encompass the dynamics of the outer part of the earth; one for the oceans and one for the continents. While this is undoubtedly a simplification, it does suggest a natural subdivision of the subject. In Part 2 we discuss the oceans, about which we have a deeper understanding than the continents (Part 3). Our main aim will be to develop a description of the mechanics of the lithosphere, and so we will be concerned with the nature of the forces driving the motion of the lithosphere (which are reviewed in part 4). To do this we will need to incorporate constraints on the mechanism of formation of the lithosphere which requires an understanding of the geochemical relationships between the crust and mantle as well as lithosphere and asthenosphere. Before we do so it is important to review some fundamental principles which are essential to understanding the dynamics of both the oceanic and the continental lithosphere (Part 1).
A vexing question concerning the preparation of the notes is what level of mathematics should be included. Many of the important physical processes in tectonics admit very concise mathematical definitions, and therefore mathematics provides a natural way to describe the processes. However, we will try to describe the basic physics and have only included the most important of the mathematical definitions in the main body of the text. Peripheral mathematical definitions are included either as marginal notes or in the appendices. For those of you who struggle with the language of mathematics we hope that a qualitative appreciation of the nature of the physical process will provide a new avenue to understanding the mathematical description; certainly this is how we have learned to tackle the mathematics.