STRUCTURE OF THE EARTH
Observations
|
Mean radius = 6371 km |
Mass = 5.97 x 1024 kg |
|
Density of surface rocks = 2-3 x 103 kg m-3 |
Average density = 5.52 x 103 kgm-3 |
The Earth has a radius = 6400 km. The deepest man made hole is about 12 km. Volcanoes bring material to the surface from depths of ~100 km. Diamonds may come from ~400 km. The rest of our information about the Earth's interior must come from study of Earthquake (or nuclear) shock waves
By studying how long it takes waves to pass through Earth (travel times) can work out a model for the internal structure of the planet.
In a planetary size body can find four types of seismic waves:
- two are body waves that travel through the Earth's
interior,
- two are surface waves.
Body waves - "P-waves" and "S-waves"
P-waves are longitudinal waves, particles move in the direction of travel of wave. P = primary (travel fastest) or Push-Pull. Sound in air travels as a longitudinal wave.
S-waves are transverse waves, that shear the rock.
Particles move at - to motion of wave.
Vp = [K + 4/3µ / r] 0.5
Vs = [µ / r] 0.5
Vp,s = velocity; K = bulk modulus, a measure of how material compresses under P; µ = shear modulus; r = density
Since K > 0, Vp > Vs, P = primary. S = secondary or shear
Surface Waves
In these waves the amplitude of wave decreases with depth in the Earth.
Love waves (L-waves) are shear waves, with displacement within plane of surface.
Rayleigh waves (R-waves) like ocean waves, particles in medium through which wave travels describes an elliptical path.
Typical P wave velocity near surface is approx. 5-6 kms-1, S waves travel at approx. 3-4 kms-1. R and L waves are slowest of all.
Seismic Structure of the Earth
In Earth, seismic velocity (V) varies with depth (Z). Seismic waves can be reflected and refracted at interfaces (cf light). The surfaces between layers are curved, but principles of refraction and reflection can be used to infer how V varies with Z, and the depths of layers, etc [see Lowerie 3.7 and Fowler 4.3].
Major result is a V v Z plot obtained from travel-time plot.
Vp
and Vs tend to vary smoothly over large
part of Earth but have a number of DISCONTINUITIES.
At depth of between 10-50 km have Mohorovicic discontinuity - marks CRUST/MANTLE boundary (upper mantle).
At 50-200 km have a LOW VELOCITY ZONE.
At 400-670 km have a number of sharp increases in Vp and Vs - TRANSITION ZONE OF MANTLE.
670 km marks boundary between T.Z. and the LOWER MANTLE.
670-2885 km monotonic increase in Vs and Vp.
2885 km, Vs = 0, Vp drops, marks the CORE-MANTLE boundary (CMB).
Outer core liquid (S-waves not possible).
5145 km, Vp
increases and Vs inferred > 0, -> solid inner core.
From the seismic data, it is also possible to work out the density of the Earth as a function of depth. This shows that the density increases from about 3.3 g/cc in the upper mantle, and reaches about 5 g/cc at the CMB. Here there is a major discontinuity with a jump to ~ 10 g/cc in the outer core and which rises to ~13 g/cc at the centre of the Earth. The average density of the core is approximately 10.8 g/cc.
The inferred V and r curves are average values for a given depth. More detailed 3 dimensional studies, based on seismic tomography, reveal local fluctuations in V and r due to variations in chemical composition or thermal structure. These will be discussed in greater detail later in the course.
THE COMPOSITION OF THE EARTH
Seismology shows that the Earth is layered, that it is largely solid and crystalline (LVZ close to melting, outer core liquid), and that it has a complex but well defined density structure.
What are the chemical and mineralogical make up of the following layers:
|
CRUST |
- |
continental |
|
MANTLE |
- |
upper |
|
CORE |
- |
outer |
Crust
In general:
- continental - inhomogeneous, high SiO2, t = 35 km (25-70
km), mixed ages (oldest 3,800 my)
- oceanic - layered basalts and gabbros, t = 6 km, orderly in age and
structure, young (? 200my)
Mantle
The Upper Mantle
Upper mantle can be sampled directly via:-
(i) Ophiolites
and tectonic slices.
(ii) Inclusions brought up in volcanics, kimberlites, etc.
Typical rock a garnet peridotite made of Mg,Fe silicates -
60% olivine (Mg,Fe)2SiO4
18% orthopyroxene (Mg,Fe)SiO3
12% garnet (C,Mg,Fe)3Al2Si2O12
10% clinopyroxene C(Mg,Fe)Si2O6
Garnet peridotite partially melts to give basalt and so is a suitable candidate for upper mantle on petrological grounds as we know basalt liquid comes from upper mantle to form oceanic crust.
Upper mantle heterogeneous (with eclogite, dunite, etc.), because of melting to give basalts and residual rocks.
Density, Vp and Vs of minerals give an excellent fit to density, Vp and Vs of upper mantle, so garnet peridotite also satisfies geophysical constraints.
LVZ may be due to geotherm approaching the solids of slightly hydrous peridotite. Pre-melting gives rise to anomalous properties.
Transition Zone
At 400 kms have discontinuity in Vp and Vs. Density increases, this could be due to:-
(1) same
minerals but with higher molecular weight (i.e. more Fe, less Mg).
(2) structural phase change to a more densely
packed structure.
(3) a combination of (1) and (2).
What happens to olivine if it is subjected to P + T of transition zone (P about 120 kbars, T about 1400°C)?
At about 120 kb + 1400°C (400 km)
Forsterite -> Beta-Mg2SiO4 (wadsleyite)
At P, T about 550 km
Beta-Mg2SiO4 -> Spinel-Mg2SiO4 (ringwoodite)
Forsterite transforms to a denser polymorph at
high P. Beta-phase and spinel-Mg2SiO4 are the
minerals of the transition zone. P/T of transformation match
those of seismic discontinuity. Density, Vp
and Vs of beta- and spinel-Mg2SiO4 are exactly compatible
with transition zone seismic data.
Also find in this P/T zone
pyroxene -> garnet
Lower Transition Zone about 60% spinel, 40% garnet structured (Mg, Fe) silicates.
Lower Mantle
More difficulty to be sure what is responsible for 670 km discontinuity. Reasons
P about 250 kbar )
Difficult to achieve
T about 1800° C ) by experiment
Can be obtained using a Multi Anvil Cell or the Diamond Anvil Cell and Laser heating (P > 1 Mbar; T > 3000 K). Both of these techniques are used in research in UCL-Bbk.
Problem, very small sample volume (<10-2 mm3 for DAC).
High P generated by very small area of diamond tip (c.f. stiletto heels). At P/T of 670 km discontinuity have:
Mg2SiO4 -> MgSiO3
+ MgO
spinel
perovskite
periclase
Nb Si coordination change: Si [IV]
Si [VI]
Spinel structure polymorph disproportionates to perovskite structure MgSiO3 + MgO. Also
Garnet -> Perovskite
Lower mantle composed of (Mg,Fe)SiO3 perovskite plus (Mg,Fe)O - magnesiowustite, plus minor phases such as CaSiO3-perovskite. Because it is so difficult to make, we do not know Vs, Vp for MgSiO3 perovskite very well. 670 km discontinuity could be due to isochemical phase transformation or lower mantle could be richer in Fe or Si than transition zone. Still not sure.
Core
Believed to be Fe rich on basis of
(1) Cosmic abundances.
(2) Iron meteorites.
(3) High density in close accord with that expected for Fe.
(4) Metallic to give magnetic field.
P + T of core very high. P
> 1.5 Mbar, T about 5000-6000 K. Since T is
not well known difficult to be too precise about core composition.
Vp, Vs of Fe at these pressures not easy to determine. Can be obtained from Shock-Wave experiments.
Pure Fe considered too dense for outer core. Must be alloyed with lower density elements - Si, S, C or O?
S found in iron meteorites. Fe-S outer core fits density data for 9-12% S by wt.
Is the outer core - inner core boundary isochemical or is there any compositional change?
Shock data suggest IC denser than pure Fe. Could be an Fe-Ni alloy (if meteorites). In this case the OC/IC boundary is a chemical discontinuity.
Details of the core not well established and still open to revision (see
recent paper by Price on Fe in the core).
Click here for more detailed notes of data analysis of deep Earth seismic
waves and here for more
on structure.
Click here for a practical on
Seismology & Earth Structure.