next up previous contents
Next: Upper mantle discontinuities Up: diss_01 Previous: Introduction   Contents

Earth's upper mantle

The layered structure of the Earth is known since the beginning of the 20th century. The coarse layered structure is described by the crust, the mantle and the core. The boundaries between these layers are characterized by discontinuous changes of seismic velocities and density. These seismic discontinuities reflect the changes of different parameters of the material, e.g. composition, seismic velocity, density or chemistry, between the different sections of the Earth. Since the beginning of seismology the structures within this rough framework have been refined and regional differences have been noted.
This thesis concentrates on structures in the Earth's mantle. The mantle extends from the Mohorovicic discontinuity (Moho) (at depths of $\sim$ 6 km in oceanic regions and up to 60 km beneath continents) to the core-mantle boundary at a depth of approximately 2889 km. The mantle is divided into the upper and the lower mantle by a seismic discontinuity at approximately 660 km.
The increase of P-wave velocity from the Moho to the 660-km discontinuity is approximately 35% (after IASP91, Kennett and Engdahl, 1991). Within this continuous increase several discontinuous changes of seismic velocities and/or density have been discovered. A summary of some detected discontinuities together with the depth and the method of detection can be found in Table 1 (after Bina, 1991).

Table 2.1:
Summary of detected discontinuities at different depths following Bina (1991)
Depth(km) Study Reference Characteristics
67 S Revenaugh [1989] 5-8% impedance contrast
86 S Revenaugh and Jordan [1989] observed, W. Pacific
75-225 S Graves and Helmberger [1988] LVZ, old Pacific
upper 100 S Revenaugh [1989] LVZ onset, oceanic and tectonic, base not detected
165-215 P LeFevre and Helmberger [1989] LVZ, Canadian shield
200 P Bowman and Kennett [1990] 3.3% V$_p$ increase, NW. Australia
$\sim$220 P,S Shearer [1990] not detected
$\sim$220 S Revenaugh and Jordan [1989] not detected, W. Pacific
232 S Revenaugh [1989] 4% impedance contrast, NWC. Australia
300 P'P' Wajeman [1988] Eurasia
400 P Bowman and Kennett [1990] 5.6% V$_p$ increase, NW. Australia
405 P LeFevre and Helmberger [1989] 5% V$_p$ increase, Canadian shield
405 S Graves and Helmberger [1989] 3.6% V$_s$ increase, old Pacific
410 P,S Shearer [1990] topography < 20 km
414 S Revenaugh and Jordan [1987, 1989, 1991a,b] topography $\sim$12 km
415 P'P' Nakanishi [1988,1989] intermittent
520 P,S Shearer [1990] 3% impedance constrast
520 S Revenaugh and Jordan [1991b] reflecton coefficient = 0.014
630 P Bowman and Kennett [1990] 3.9% V$_p$ increase, NW Australia
655 P'P' Nakanishi [1988, 1989] sharp
659 S Graves and Helmberger [1989] 6.8% V$_s$ increase, old Pacific
660 P LeFevre and Helmberger [1989] 4% V$_p$ increase, Canadian shield
660 P,S Shearer [1990] topography < 20 km
660 S Revenaugh and Jordan [1987, 1989, 1991a,b] topography $\sim$12 km
670 P'P' Davis et al. [1989] not detected $\Rightarrow$ more than 10 km topography
670 P'P' Wajeman [1988] 7.5% V$_p$ contrast, Eurasia
670 P to S Paulssen [1988] sharp
660-680 S to P Richards and Wicks [1990] Tonga
710 S Revenaugh and Jordan [1989, 1991a] tentative
840 P,S Shearer [1990] tentative
900 S Revenaugh and Jordan [1989, 1991c] tentative
2546 P Baumgardt [1989] 2.75% V$_p$ increase
2600 P Davis and Weber [1990] 3% V$_p$ increase, N. Siberia
2610 S Young and Lay [1987] 2.75% V$_s$ increase, Indian Ocean
$\sim$2610 S Revenaugh and Jordan [1989] not detected, W. Pacific

The lower region of the upper mantle is characterized by the mantle transition zone from a depth of 410 km to 660 km. This zone is marked by a strong gradient of the depth-velocity profile.

Figure 2.1: Depth, origin and behaviour of upper mantle and transition zone seismic discontinuities (after Revenaugh and Jordan, 1991b). This is a representation of the structure found by the analysis of ScS reverberations. The cross section extends from a continental shield (left) across a subduction zone into an ocean basin (right). On the left, the depth and the division of the Earth after Bullen (1940) is shown. On the right hand side, the possible phase transitions within the upper mantle minerals are displayed. The discontinuities are marked by letters and numbers as described in the text. The letters describe the Hales discontinuity(H), the Gutenberg discontinuity (G), and the Lehmann discontinuity (L). The numbers give the approximated depths of the discontinuities at 410 km, 520 km, and 660 km depth. The dominant minerals are abbreviated as Ol: olivine, Sp: spinel, Gt: garnet, $\beta$: beta phase of olivine, $\gamma$: gamma phase of olivine, Pv: perovskite, Mw: magnesiowüstite, Il: garnet in ilmenite structure.
\begin{figure}
\centerline {\psfig{figure=abb_2.3.eps,angle=-.3,width=16cm,height=9cm}}\hfill
\end{figure}

The schematic structure of the upper mantle for different tectonic regions is displayed in Figure 2.1 (Revenaugh and Jordan, 1991b). The tectonic regions are listed at the top, from a continental shield at the far left to an ocean basin at the right. The discontinuities are marked by solid lines (positive reflectors) or by dashed lines (negative reflectors). A positive reflector is characterized by a velocity change from fast to slow velocities, when the wave travels from the fast velocity region to the slow velocity region, whereas a negative reflector is defined by a velocity jump from slow to fast. On the left hand side, the depth division of the Earth by Bullen (1940) is shown. The right hand side shows boundaries of different mineralogical facies. The cold subducting slab in the centre of the picture influences the discontinuities in different ways.
According to the generally accepted petrological model, the upper mantle consists of the minerals olivine (60%), orthopyroxene (23%), clinopyroxene (2%) and garnet (15%) (Akaogi et al, 1987; Jeanloz and Thompson, 1983; Stöffler, 1997). This assemblage undergoes a sequence of structural phase transitions at different pressures appropriate for the upper mantle. Figure 2.2 shows this composition of a so called lherzolitic mantle (after Stöffler, 1997) and some phase transitions in the different components which could account for seismic discontinuities. In addition to the major transitions shown in the central picture, others can be found, especially at lower pressures. Some of these are listed at the right hand side and approximate depths are given.

Figure 2.2: Composition of lherzolitic mantle (modified after Stöffler (1997)). Displayed are the phase transitions in the upper mantle and the mantle transition zone for a composition of 60% olivine, 25% ortho- and clinopyroxene and 15% garnet. The depth scale on the left is transformed to an approximated pressure scale on the right hand side. Additionally the phase transition in the shallow upper mantle, accounting for the Hales and the Lehmann discontinuity described in the text, can be seen at the right hand side
\begin{figure}
\centerline {\psfig{figure=figure_2.2.eps,angle=0,width=15.5cm,height=9.5cm}}\hfill
\end{figure}

For simplicity, a seismic discontinuity is called in the following a discontinuity. The nomenclature of the discontinuities follows a proposal by Revenaugh and Jordan 1991b) with the Hales discontinuity H (60 - 90 km), the Gutenberg discontinuity G (50 - 150 km) and the Lehmann discontinuity L ($\sim$ 220 km). The discontinuities of the shallow upper mantle and the mantle transition zone are described in the next subsections.



Subsections
next up previous contents
Next: Upper mantle discontinuities Up: diss_01 Previous: Introduction   Contents

2000-09-05