Sie sind auf Seite 1von 36

4 EXERCISES

4.1 Geochronology

1. How many half-lives must elapse before the activity of a radioactive isotope decreases
to 1% of its initial value? How long is this time for 14C, which has a decay rate of
1.21 10-4 yr-1?

Radioactive decay is described by the equation


A = A0 e t

where A0 is the initial activity when the isotope was formed, A is the activity after time
t, and is the decay rate. The half-life t1/2 is the time at which A has decayed to one half
of its initial value:
0.5 A0 = A0 e t
ln(2)
t1/ 2 =

When the activity has decayed to 1% of its initial value, A = 0.01 A0. Inserting this in
the decay equation,
0.01 A0 = A0 e t
ln( 0.01) ln(100 )
t 0.01 = =

ln(100) ln( 2) ln(100 )
t 0.01 == = t1/ 2 = 6.644 t1 /2
ln(2) ln(2)

In the case of 14C, with a decay rate = 1.21 10-4 yr-1


ln(2) 0.69315
t1/ 2 = = 4
= 5728.5 5730 yr.
1.21 10
14
The time for the activity of C to decay to 1% of its initial value is
t 0.01 = 6.644 t 1/2 38,100 yr.

77
2. Radiocarbon dating of a sample of wood from the tomb of an Egyptian pharaoh gave
concentrations of 9.83 1015 mol/g for 14C and 1.202 102 mol/g for 12C. Assuming
that the initial 14C/12C ratio in the sample corresponded to the long-term atmospheric
ratio of 1.20 1012, determine the age of the tomb, the percentage of 14C remaining,
and the original 14C concentration in the wood.

Let the isotopic carbon ratio at present be R = [14C/12C] and let the initial ratio when
the tomb was made be R0 = [14C/12C]0. The decay of radioactive carbon changes the
isotopic ratio with time according to the equation
R = R0 e t

Due to radioactive decay of 14C, its proportion has decreased relative to the stable 12C.
The isotopic ratio is now R = (9.83 1015)/(1.202 102) = 8.178 1013; the initial
ratio is assumed to have been R0 = 1.20 1012.

Setting these values in the equation and using the decay rate = 1.21 104 yr-1 :
R
ln = t
R0

1 R0 ln(1.20 10 12 / 8.178 10 13 )
t= ln = 3,170 yr. This is the age of the
R 1.21 10 4
tomb.

The factor e t describes the proportion of the radioactive isotope remaining after
time t. After 3,170 yr the product (t) = 0.3836. The fraction of 14C that remains after
3,170 yr is therefore e0.3836 = 0.6814; the percentage remaining is 68.1%.

Applying the decay equation to the radioactive 14C component only, we can write
14
C= ( C)
14
0
e t =0.6814 ( C)
14
0

The present measured concentration of 14C in the wood is 9.83 1015, so the initial
concentration was (9.83 1015 /0.6814) = 1.44 1014 mol/g.

78
3. The decay constants of 235U and 238U are 235 = 9.85 1010 yr1 and 238 = 1.55 1010
yr1. Calculate the half-lives of these uranium isotopes.

The half-life t1/2 of a radioactive isotope with decay constant is given by


ln(2)
t1/2 =

For the uranium isotope 235U the decay constant is 235 = 9.85 1010 yr1 and the half-
life is
ln(2) ln(2)
235
t1/2 = = = 7.037 10 8 yr
9.85 10 10
The half-life of 235U is t1/2
235
= 704 Ma

For the uranium isotope 238U the decay constant is 238 = 1.55 1010 yr1 and the half-
life is
ln(2) ln(2)
238
t1/2 = = = 4.472 10 9 yr
1.55 10 10
The half-life of 238U is t1/2
238
= 4.47 Ga

79
4. Assuming that the isotopes 235U and 238U were created in equal proportions in a
common event, such as a supernova, and given that their abundances are now in the
ratio 235U/238U = 1/137.88, calculate how long ago they were created.

The decay equation for the radioactive uranium isotope 235U is


235
U= 235
U 0 e 235 t ,
where 235
U and 235
U 0 are the present and initial concentrations and 235 is the decay
constant, equal to 9.85 1010 yr1.

The corresponding decay equation relating the present and initial concentrations of 238U
with decay constant 238 equal to 1.55 1010 yr1 is
238
U= 238
U 0 e 238 t .

Dividing the second equation by the first:


238
U 238
U 0 e 238 t
=
235
U 235
U 0 e 235 t

137.88 = e( 235 238 )t


ln(137.88) = ( 235 238 ) t
ln (137.88 ) ln (137.88 )
t= = = 5.940 10 9 yr.
( 235 238 ) ( 9.85 1.55 ) 10 10

The age of the creation of the isotopes would be 5.94 Ga.

80
5. The analysis of strontium and rubidium isotopes in whole rock samples from a granitic
batholith gave the following atomic concentrations in p.p.m.:

87 87 86
Sample Sr Rb Sr

A 2.006 1.656 2.751

B 0.349 0.726 0.450

C 0.984 2.776 1.232

D 0.720 2.515 0.871

(a) Calculate the 87Rb/86Sr and 87Sr/86Sr isotopic ratios for these samples.
(b) Determine the age of the batholith and the initial 87Sr/86Sr ratio.

(a) The isotopic ratios for the four samples are obtained by simple division of their
measured concentrations:
87
Sample Rb/86Sr 87
Sr/86Sr

A 0.6020 0.7292

B 1.613 0.7756

C 2.253 0.7987

D 2.887 0.8266

(b) Eqs. (4.12) and (4.14) describe the RbSr dating system. A plot with the 87Rb/86Sr ratios
on the abscissa (x) and 87Sr/86Sr on the ordinate (y) axis gives a straight line as in the
figure.

81
The best-fit straight line is the Rb-Sr isochron. Its equation in this plot is
y = 0.04224x + 0.7049 .
The age of the batholith is found from the slope m = 0.04224 using Eq. (4.14):
1 1010
t= ln(1 + m) = ln(1 + 0.04224) = 2.913 10 9 yr
0.1420

Thus the age of the batholith is, to 3 significant digits, 2.91 Ga.

The initial ratio of 87Sr/86Sr is the intercept on the y-axis where x = 0. From the equation
of the best-fit line the initial ratio is equal to 0.705.

82
6. Argon-argon dating of muscovite in a Late Cretaceous granite gave the following
isotope ratios for the plateau stages during incremental heating:

Maximum Heating
39
Temperature (C) Ar/36Ar 40
Ar/36Ar
750 1852 8855
830 1790 8439
895 1439 6867
970 3214 15380
1030 2708 12970

(a) Calculate the 40Ar/39Ar ratios for each incremental heating step.
(b) A calibration constant J = 0.00964 was determined for the monitor mineral. Using
the 40Ar/39Ar ratios from (a) in Eq. (4.18), calculate the apparent ages at each heating
step.
(c) Draw an 40Ar/39Ar isochron diagram by plotting each 40Ar/36Ar ratio as ordinate
against the corresponding 39Ar/36Ar ratio as abscissa. Draw a best-fitting line the
isochron through the data points, and determine its slope and intercept.
(d) Compute the age of the muscovite from the slope of the isochron. Is the intercept on
the ordinate axis significant?

(a) The 40Ar/39Ar ratios, found by simple division, are given in the following table:

Heating Heating
39 36 40 36 40 39
Temperature Ar/ Ar Ar/ Ar Ar/ Ar Step Age
(C) (Ma)
750 1852 8855 4.781 81.3
830 1790 8439 4.715 80.2
895 1439 6867 4.772 81.1
970 3214 15380 4.785 81.4
1030 2708 12970 4.790 81.4

(b) Eq. (4.18) describes how the 40Ar/39Ar ratio is used to make an age determination. The
age t is given by
40
Ar
t = 1.804 10 9 ln 1 + J
39
Ar

The factor J in this equation is the empirical calibration factor of the equipment,

83
determined for a control sample of known age. J includes, among other terms, the 39Ar
production yield from neutron irradiation of 39K; in the present exercise J = 0.00964.
This gives for the individual heating steps in the experiment the age estimates in the
right column of the above table. Their average age is 81.1 Ma and the standard
deviation of the ages is 0.5 Ma.

(c) The 40Ar39Ar plot is a straight line, the isochron.

(d) The equation of the best-fitting straight line is


y = 4.820x 105

The slope of the line gives the optimum 40Ar/39Ar ratio for all the data, which can be
inserted in Eq. (4.18) to get the age:
t = 1.804 10 9 ln (1 + 0.00964 4.820 ) = 8.19 10 7 yr

The age of the muscovite by this method is 81.9 Ma.

The non-zero intercept suggests that the measured 40Ar/36Ar ratios are systematically
low by an amount equal to 105 p.p.m. This causes the estimated age at each temperature
to be low. Consequently, the mean of the estimated ages, 81.1 Ma, is 0.8 Ma lower than
the best-fit result.

84
7. The following isotopic ratios were measured in U-Pb age determinations on three zircon
grains extracted from a granite:

207
Sample Pb/235U 206
Pb/238U
zircon 1 27.4 0.600
zircon 2 33.3 0.680
zircon 3 37.9 0.740

(a) Using the values listed in Table 4.2, plot a concordia diagram on graph paper or with
a plotting routine. Enter the measurements from the above table on the graph, and draw
the straight discordia line through the points.
(b) Determine the coordinates of the intersection points of the concordia and discordia
lines.
(c) Using the coordinates of the upper intersection point together with Eq. (4.19) and
Eq. (4.20), calculate the age of formation of the zircons.
(d) Calculate when loss of lead occurred in the zircons.

(a) The values listed in Table 4.2 give the curved concordia line on the following
plot. The curve may be plotted by hand on graph paper, or the values may be
entered in a spreadsheet-style computer program (e.g., MS Excel), from which a
plot of the data can be obtained.

85
The x-coordinate is the 207Pb/235U ratio, which at time t is given by
207
Pb
x= 235
= e235 t 1 (1)
U

where 235 = 9.8485 1010 yr1 is the decay constant of 235U.

The y-coordinate is the 206Pb/238U ratio at time t, given by


206
Pb
y= 238
= e238 t 1 (2)
U

where 238 = 1.5513 1010 yr1 is the decay constant of 238U.


The numbers in italics on the concordia curve are the corresponding ages in Ga (109 yr).

The data points for the three zircons given in the table are shown as small squares. A
best-fit straight line through the three points has equation
y = 0.2348 + 0.01334x (3)

(b) The straight line is the discordia line; it intersects the concordia curve at two points. The
coordinates of the intersection points cannot be found analytically but must be obtained
numerically. There are mathematical methods for doing so, but they are beyond the
scope of this book. As a first attempt, however, a graphical solution may be attempted.
This gives an approximate solution for the coordinates of the upper intersection point as
(x = 44, y = 0.8), and the lower intersection point as (x = 4, y = 0.3).

The intersection points must satisfy the discordia line and the concordia curve
simultaneously. Starting from the rough values obtained graphically, an iterative
method may be used to find more accurate values that satisfy both equations.

Iterative solution:
Starting with a value x = 43.0 slightly below the expected intersection value, the
corresponding y-value is calculated for the discordia line, Eq. (3); it equals 0.8086.
Inserting the same x-value in Eq. (1) we solve for the time t:
1 1010
t= ln(1 + x) = ln(1 + 43) = 3.842 10 9 yr.
235 9.8485
This value is inserted in Eq. (2) to obtain the y-value of the concordia line at x = 43.
y = e238 t 1 = e(0.15513 3.8424 ) 1 = 0.8149 .

The value of x is now incremented by a small amount (0.5) and the procedure is
repeated; further repetition gives a table of results like the following:

86
x-value 43.0 43.5 44.0 44.5 43.95
discordia y 0.8086 0.8152 0.8219 0.8286 0.8212
concordia t 3.842 3.854 3.865 3.876 3.864
concordia y 0.8149 0.8182 0.8214 0.8245 0.8211
y-difference 0.0064 0.0029 0.0005 0.0040 0.0002

The differences in y-values show that the discordia line crosses the concordia curve
at an x-value slightly less than 44.0. Thus, a better starting value for the iteration would
be around x = 43.9. The procedure is now repeated with smaller increments of x to
determine the intersection point more exactly.
Using this iterative approach, the coordinates of the upper intersection of the
discordia and concordia lines are found to be: x = 43.95; y = 0.821.
The same iterative method can be applied to the lower intersection. Its
coordinates are: x = 3.986; y = 0.288.

(c) The age of formation of the zircons is given by the upper intersection point of the
discordia and concordia lines. As evident from the table for the iterative solution,
their age of formation is 3.86 Ga.

(d) The departure of the lower intersection point from zero implies loss of lead in an
event that occurred at the time corresponding to the lower intersection point. The x-
coordinate of the point, found by iteration, is x = 3.986. Inserting this value in
Eq. (4.20) gives the age of the event:
1 1010
t= ln(1 + x) = ln(1 + 3.984) = 1.631 10 9 yr.
235 9.8485

Rounding the answer to fewer significant digits gives the age as 1.63 Ga.

87
8. The following isotopic ratios were measured in a K-Ar age determination on an
ignimbrite as part of a combined radiometric-paleomagnetic study of geomagnetic
polarity.

40
Sample K/36Ar 40
Ar/36Ar
A 4,716,000 822
B 8,069,000 1200
C 12,970,000 1730
D 27,670,000 3280

(a) Plot the isotope ratios, draw the isochron, and compute its slope and intercept.
(b) Calculate the isochron age of the ignimbrite.
(c) Correct the observed 40Ar/36Ar ratios for the initial 40Ar/36Ar concentration, and
compute the individual sample ages.
(d) Calculate the mean age and its standard deviation. Compare the mean age with the
isochron age.
(e) With reference to the radiometric time scale in Fig. 5.74, what magnetic polarity
would you expect the ignimbrite samples to have?

(a) The values listed in the above table give the following isochron plot:

The best-fit straight line has equation


y = 1.0673 10 4 x + 332.5

88
(b) The isochron age is obtained from the slope m of the straight line, using Eq. (4.17):
40
Ar
t = 1.804 10 ln 1 + 9.54
9


40
K

In this equation 40Ar refers to the pure in-situ generated component. Both the ordinate
and abscissa in the plot contain the same 36Ar so the ratio 40Ar/40K needed for the age
determination is equal to the slope of the isochron. The age is then
t = 1.804 10 9 ln (1 + 9.54 1.0673 10 4 )= 1.83 Ma

(c) The non-zero intercept on the isochron plot gives an initial 40Ar/36Ar concentration of
332.5. Subtracting this from each of the measurements of 40Ar/36Ar gives the following
table with corrected 40Ar/36Ar ratios:

40
Sample observed corrected Ar/40K Age
40
K/36Ar 40
Ar/36Ar ( 104) (Ma)
A 4,716,000 489.5 1.038 1.785
B 8,069,000 867.5 1.075 1.849
C 12,970,000 1397.5 1.077 1.853
D 27,670,000 2947.5 1.065 1.832

(d) The mean of the individual ages is 1.83 Ma and its standard deviation is 0.03 Ma. This
agrees well with the age determined from the isochron.

(e) The magnetic polarity time scale in Fig. 5.74 consists of a composite scale determined
from radiometric measurements and a scale determined from the spacing of marine
magnetic anomalies. In both scales the age 1.83 Ma places the ignimbrite within the
Olduvai normal polarity event.

89
4.2 The Earths Heat

9. List and compare the various factors that may influence the measured temperature
gradient at a depth of 5 m in (a) a deep drillhole in oceanic sediments and (b) a
continental well that encounters the groundwater table at 2 m depth.

(a) Oceanic sediments are shielded from temperature changes at the ocean surface by the
great depth of water. The effects of seasonal and climatic temperature changes can be
neglected. The ocean bottom is generally flat except near to an oceanic ridge, and so
only in this vicinity are corrections for bottom topography needed. In the undisturbed
sediment column there is little or no circulation of sea water, so thermal conditions are
stable. When oceanic heat flow is measured with a deep-sea corer as in Fig. 4.26, the
temperature is measured with thermistors mounted a few centimeters from the barrel of
the corer. The temperature measurements are made as soon as possible after penetration
of the sediment, before thermal disturbance from the coring can reach the sensors.
When a drillhole is drilled, the thermal stability in the sediments is disturbed by the heat
generated from the friction of the drilling process and this must be compensated, for
example by monitoring temperature over a length of time.

(b) Annual fluctuations of surface temperature have a penetration depth of around 4 m in


continental rocks. At a depth of 5 m in a continental well, these fluctuations can be
ignored. However, longer term temperature changes, such as those associated with
paleoclimatic variation, penetrate much deeper and must be corrected. At 5 m a sensor
is probably in the groundwater, which may be in motion; this can cause time-dependent
(e.g., seasonal) fluctuations in temperature at that depth. If the surface topography
around the well is significant, a correction must be made for its effect on the shallow
subsurface temperature profile. The influences of vegetation, hydrology, latitude and
sun angle, as well as seasonal snow cover must be taken into account where necessary.
Finally, as mentioned in (a) above, if the well was drilled recently, residual heat from
the drilling process may influence the temperature profile.

90
10. A shallow circular pond 100 m in diameter freezes solid during a very cold night. The
pond is in a geothermal area in which the temperature at 200 m depth is 40C higher
than at the surface. The thermal conductivity of the intervening rock is 3.75 W m1 K1
and the latent heat of fusion of ice is 334 kJ kg1. Neglecting other heat sources,
calculate the mass of ice that melts per hour due to the geothermal gradient.

The temperature gradient beneath the pond is (40/200) = 0.2 C m1, which is
equivalent to 200 C km1. Combining this gradient with the thermal conductivity
k = 3.75 W m1 K1 gives the local heat flow q:
dT
q=k = 3.75 0.2 = 0.75 W m2
dz

The surface area A of the circular pond with diameter d = 100 m is


2
A= d = (100 ) = 7.854 10 3 m2
2

4 4

The thermal energy Q flowing into this area per hour (3600 s) is
Q = qAt = 0.75 7.854 10 3 3600 = 21, 206 kJ

This energy is used to melt an unknown mass m of ice. The latent heat of fusion L of
a material is defined as the amount of heat energy needed to melt 1 kg of the material.
For ice, L = 334 kJ kg1. The mass of ice melted by the thermal energy entering the
pond per hour is
Q 21, 206
m= = = 63.5 kg hr1
L 334

91
11. Assuming a constant geothermal gradient of 30 C per kilometer, estimate what
percentage of the Earths volume is hotter than the temperature of molten lava at
atmospheric pressure. Why is the deeper interior of the Earth not entirely molten?

At atmospheric pressure molten lava has a temperature around 1200 C when it


flows and a temperature around 800 C when it solidifies. If the geothermal gradient is a
constant 30 C per kilometer, this tmperature is reached in a depth of 40 km. The radius
of the Earths interior that is hotter than 1200 C is therefore (6371 40) = 6331 km.
The volume of this spherical region is proportional to the cube of its radius, and so it
can be expressed as a fraction of the Earths total volume by the ratio
3
6331
= 0.981
6371
More than 98% of Earths volume is hotter than molten lava at atmospheric pressure.

Despite high temperatures the Earths mantle is not molten because the increase of
pressure with increasing depth causes the melting point to increase faster than the
increase of temperature. Throughout the crust and mantle the temperature is below the
local melting point and so the material remains solid.

At the core-mantle boundary there is a change of composition from silicate mantle to


iron core. The temperature of the outer core is above the local melting point of iron and,
as a result, the outer core is molten. With increasing depth in the core the melting point
(solidus) curve rises more steeply than the actual temperature and eventually crosses it.
This gives rise to an inner iron core that is solid.

92
12. The mean global heat flow at the Earths surface is 87 mW m2. Calculate the time in
years needed for the mantle and core to cool by 100 C, with the following assumptions:
(i) the Earths mantle and core cool as a homogeneous unit, (ii) 20% of the observed
heat flow at the Earths surface is from the mantle, (iii) the lithospheric thickness is 100
km, (iv) thermal effects from the lithosphere itself may be ignored. Relevant properties
of the mantle and core are: mean density 5650 kg m3, specific heat 400 J kg1 K1.

The total amount of heat flowing out of the Earth per second (dQ/dt) is equal to the
mean heat flow per square meter (q), multiplied by Earths surface area (A):
dQ
= qA
dt

The mean global heat flow is 87 mW m2.

( )
2
Earths surface area: A = 4 R 2 = 4 6.371 10 6 = 5.101 1014 m2.

dQ
Total heat loss per second: = 87 10 3 5.101 1014 = 4.438 1013 J s1
dt

In the exercise, 20% of this heat comes from below the lithosphere and thermal effects
of the lithosphere may be neglected. The mantle heat loss is thus 8.875 1012 J s1.
The mantle and core cool as a unit, with radius equal to (6371100) = 6271 km.
The volume of this cooling sphere is 43 R 3 = 1.033 1021 m3.

Assuming the given mean density of 5650 kg m3, the mass of mantle and core is
5.836 1024 kg.

The specific heat is 400 J kg1 C1. For an object of mass m with specific heat c that
cools through a temperature difference T, the heat loss Q is given by
Q = cm T
For a temperature drop of 100 C, the Earths heat loss is
Q = 400 5.836 100 10 24 = 2.335 10 29 J

If the heat is lost at a constant rate, the length of time required for this temperature
change is
Q 2.335 10 29
t= = = 2.63 1016 s.
(dQ / dt) 8.875 10 12

This is equivalent to a cooling time of about 830 Ma.

93
13. A temperature gradient of 35 C km1 is measured in the upper few meters of sediments
covering the ocean floor. If the mean thermal conductivity of oceanic sediments is 1.7
W m1 C1, calculate the local heat flow. How far do you think the sampling site is
from the nearest active ridge?

By definition, heat flow = thermal conductivity temperature gradient


dT
q=k
dz

The temperature gradient is given as 35 C km1 = 0.035 C m1. Combining this values
with the thermal conductivity k 1.7 W m1 gives the local heat flow:
dT
q=k = 1.7 0.035 = 0.0595 W m2,
dz

i.e., the heat flow is 59.5 mW m2.

The relationship between heat flow (in mW m2) and distance from a spreading ridge
depends on the age t of the ocean crust. Two relationships between heat flow and age
are given in Eq. (4.62). Assuming the simplest, which is applicable to ages younger than
55 Ma:
2
510 510
q= mW m2, from which t = Ma. The heat flow of 59.5 mW m2
t q
2
510
2
510
corresponds to a crustal age of t = = = 73.5 Ma. This is older than the
q 59.5
range for which the assumed age-heat flow equation is valid. The alternative equation
for ages older than 55 Ma is
q = 48 + 96e0.0278t mW m2.

Inserting the calculated heat flow in this equation gives an age


1 96 1 96
t= ln = ln = 76.3 Ma.
0.0278 q 48 0.0278 59.5 48

The distance of a location with this age from a ridge axis depends on the spreading rate.
Assuming a low half-spreading rate of 1 cm yr1 (as in the North Atlantic Ocean) the
distance from the ridge is about 760 km. With a fast spreading rate of 6 cm yr1 (as on
the Pacific-Nazca ridge), the distance from the ridge could be around 4500 km.

94
14. What heat flow values would you expect at the locations of the oceanic magnetic
anomalies with numbers C5N, C10N, C21N, C32N, M0? Interpret the ages of the
anomalies from Fig. 5.78 and use the heat-flow model GDH1 (Eq. (4.62)) for the
cooling of oceanic lithosphere.

From the polarity time scale in Fig. 5.78 the ages of anomalies C5N, C10N and C21N
are less than 55 Ma. To compute the oceanic heat flow at the locations of these
anomalies the first part of Eq. (4.62) is used:
510
q= mW m2.
t

To compute the heat flow at the anomalies older than 55 Ma, the alternative equation
must be used
q = 48 + 96e0.0278t mW m2.

Combining the results, the following table of heat flow values is obtained:

Anomaly Age from Fig. 5.78 Heat flow


number (Ma) (mW m2)
C5N 10.5 157
C10N 28.5 96
C21N 47 74
C32N 72.5 61
M0R 121 51

95
15. Using the relationships in Eq. (4.73), estimate the approximate depths of the ocean at
these locations? What is the thickness of the elastic lithosphere and the depth of the top
of the asthenosphere at these locations (see Fig. 2.79)?

Eq. (4.63) gives two relationships between the ocean depth d (in m) and age t (in Ma) of
the ocean crust:
d = 2600 + 365 t for t < 20 Ma
d = 5651 2473exp ( 0.0278t ) for t 20 Ma

Of the anomalies in this exercise, only anomaly C5N is younger than 20 Ma. Using
the appropriate equations the following age-depth relationships are obtained:

Anomaly Age from Ocean


number Fig. 5.78 Depth
(Ma) (m)
C5N 10.5 3780
C10N 28.5 4530
C21N 47 4980
C32N 72.5 5320
M0 121 5570

The base of the elastic lithosphere occurs at temperatures between 350 C and 650 C.
The top of the asthenosphere has a temperature of around 1100 C. Reading the
corresponding depths directly from Fig. 2.79 the following results are obtained:

Anomaly Age Depth to Depth to Depth to


number (Ma) 350 C 650 C 1100 C
(km) (km) (km)
C5N 10.5 8 15 30
C10N 28.5 13 25 49
C21N 47 17 32 63
C32N 72.5 21 40 78
M0 121 27 51 100

96
16. Assuming that the Earth initially had a uniform temperature throughout and has been
cooling by conduction only, use the solution for the 1-dimensional cooling of a semi-
infinite half-space (Eq. (4.57) and Box 4.2) to derive Eq. (4.2) for Kelvins estimated
age of the Earth.

The 1-dimensional model of cooling represented by Eq. (4.57) gives the temperature
T at depth z and time t after a half-space starts to cool as:
T = Tm erf ( ) , in which erf() is the error function (Box 4.3), with argument
z
= . Here is the thermal diffusivity of the half-space, given by
2 t
k
= , where k is the thermal conductivity, c is the specific heat at constant
c
pressure, and is the density.

In order to compute the heat flow we need to know the temperature gradient.
dT dT d
=
dz d dz

Evaluating separately the differentiations on the right side of the equation:



dT d 2T d 2T
erf( ) = 0 eu du = 0 e , where T0 is the initial temperature;
2 2
= T0
d d d 0
d d z 1
= =
dz dz 2 t 2 t

Combining these expressions gives the temperature gradient at depth z and time t:
2
dT T0 2 T0 4z t
= e = e
dz t t

The surface temperature gradient at z = 0 at time t after the start of cooling is:
dT T0 c T0
= =
dz z = 0 t k t

This is Eq. (4.2) from which Kelvin estimated the age of the Earth.

Kelvin assumed an initial temperature of 7000 F (4044 K) and a surface temperature


gradient of 0.036 K m1. Combining these with representative values for surface rocks
(e.g, k = 3.0 W m K1, = 2700 kg m3, cp = 800 J kg1 K1), we get a cooling time
2
c 2700 840 4044
2
T0
t= = = 3.19 10 , which equals 101 Ma.
15

k ( dT dz )0 3.0 0.036

97
17. The temperature in the near-surface layers of the Earths crust varies cyclically with
daily, annual and longer periods. For a surface temperature variation given by
T = T0 cost, the temperature variation at depth z and time t is described by:
z z
T (z,t) = T0 exp cos t
d d
2 k 2
d= ; = ; =
c p
where is the period of the variation, k is the thermal conductivity, cp is the specific
heat, and is the density. For surface sediments assume k = 2.5 W m K1, cp = 1000
J kg1 K1, and = 2300 kg m3.
(a) Calculate the phase difference (in days) between the temperature variation at the
surface and at depths of 2 m and 5 m, respectively. Perform the calculations for both the
daily and annual temperature fluctuations.
(b) Assuming that the range in surface temperatures between summer and winter is
40 C, calculate the depth at which the annual temperature range is 5 C. How large (in
weeks and days) is the phase difference between the surface temperature and the actual
temperature at this depth?

(a) The expression for temperature at depth z and time t can be written
z z
T (z,t) = T0 exp cos t = T1 T2 , where
d d
z
T1 = T0 exp describes how the amplitude of a temperature change at the
d
surface decreases with depth, and
z z
T2 = cos t = cos t = cos ( t t 0 ) describes how temperature varies
d d
with time at a given depth z. The function (t t0) is the phase of this variation and t0 is
the phase difference. t0 is the amount of time by which the fluctuation at depth z is
delayed with respect to the surface variation; it is sometimes referred to as the time-lag.
z z
t0 = = , where is the period of the fluctuation.
d 2 d
Note: If the period is expressed in days, the phase difference will also be in days.

Using the given values for surface sediments:


k 2.5
 = = = 1.087 10 6 m2 s1
c p 2300 10 3

98
2
The penetration depth is d = = . In order to compute this depth the period

of the signal must be expressed in seconds.

Daily temperature variation

For the daily temperature variation t = 86,400 s. This gives a penetration depth of

86400 1.087 10 6
d= = = 0.173 m.

Inserting this in the expression for the phase difference, and expressing the times in
days, we find the phase difference of the daily temperature variation at depth 2 m to be
z 1 2
t0 = = = 1.8 days
2 d 2 0.173
Similarly, the phase difference of the daily temperature variation at depth 5 m is
1 5
t0 = = 4.6 days
2 0.173

Annual temperature variation

The period of the annual variation is 31,536,000 s and the penetration depth is

3.1536 10 7 1.087 10 6
d= = = 3.3 m.

The phase difference of the annual variation at a depth of 2 m is


z 365 2
t0 = = = 35 days, or 5 weeks, and
2 d 2 3.3
the phase difference of the annual variation at depth 5 m is
z 365 5
t0 = = = 88 days, or 12 weeks and 4 days.
2 d 2 3.3
(b) To find the depth at which an annual surface temperature range of 40 C is reduced to
5C, we use the expression T1 above, and the annual penetration depth of 3.3 m for d:
z
T1 = T0 exp
d

T 40
z = d ln 0 = 3.3 ln = 6.9 m
T1 5

At this depth the phase delay of the annual variation is


z 365 6.9
t0 = = = 121 days = 17 weeks 2 days
2 d 2 3.3

99
18. The daily average temperature in northern Canada is +10 C in July and -20 C in
January. Using the heat conduction equation calculate the thickness of the permafrost
layer (below which the ground is permanently frozen). Relevant physical properties of
the ground are: thermal conductivity k = 3.0 W m K1, specific heat cp = 840 J kg1 K1,
density = 2700 kg m3.

Taking the average of the January and July temperatures as the mean value for a
year, the mean is found to be 5 C and the temperature range is 15 C about this
value. The permafrost level is the depth below which the temperature is always less than
0 C so that the ground remains frozen. For this problem we can ignore the geothermal
gradient (about 30 C km1) as its effect would not be significant.
The penetration depth for the annual temperature change at the site can be computed
with the given parameters and the equation

k 3.0 3.1536 10 7
d= = = = 3.64 m
c 2700 840

The amplitude of the temperature variation at the surface is 15 C; the mid-value of


the temperature fluctuation is 5 C. We need to find the depth z at which the amplitude
of the temperature change is 5 C; the summer maximum of +5 C will then equal the
mean temperature of 5 C and permafrost will (just) melt. Using the surface amplitude
T0 = 15 C, desired amplitude T1 = 5 C, penetration depth of 3.64 m, and the equation
from the previous exercise
T 15
z = d ln 0 = 3.64 ln = 4.0 m.
T1 5

This is the permafrost depth at the particular location with the given ground
parameters; at greater depths the ground will remain frozen. Note that the permafrost
depth depends on the local thermal parameters and will vary with the type of ground
(e.g., soil, sediment, hard rock, etc.)

100
19. The half-spreading rate at an oceanic ridge in the middle of a symmetric ocean basin
bounded by subduction zones is 44 mm yr1. The ridge is 1,000 km long and the
distance from the ridge to each subduction zone is 2,000 km. If the oceanic heat flow
varies with crustal age as in Eq. (4.62), calculate how much heat is lost per year from
the ocean basin.

The ridge is in the middle of the symmetric ocean basin, so the total heat lost is
double that lost from each half of the basin. At a half-spreading rate of 44 mm yr1, the
age tSZ of oceanic crust at a subduction zone 2,000 km away is
2000 10 3
t SZ = = 45.45 Ma.
0.044

The heat flow q from oceanic crust younger than 55 Ma after a cooling interval t (in
Ma) is given by the first part of Eq. (4.62):
510 0.51
q= (mWm 2 ) = (Js 1m 2 )
t t
0.51 0.51
The heat flow at the subduction zone is qSZ = = = 7.565 10 2 J s1 m2
t SZ 45.45

Consider the heat loss through a narrow strip of crust parallel to the ridge axis at a
distance from the ridge corresponding to age t. The area of the strip is dA =(Lv)dt,
where L is the length of the ridge and v the spreading rate. The heat loss dQ is
0.51
dQ = qdA = q ( Lv ) dt = ( Lv ) dt
t

The total amount of heat lost from one half of the basin is found by integrating this
expression with respect to time from the ridge axis to the subduction zone:
0.51
( Lv ) t dt = 2 ( Lv ) 0.51
t SZ t SZ
Q= t
0 0

0.51
Q = 2 ( Lv ) 0.51 t SZ = 2 ( Lv ) t SZ = 2 ( Lv ) qSZ t SZ
t SZ

Here the units of (Lv) are m2 yr1, qSZ is in W m2, and the age tSZ is in Ma (=106 yr).
Correcting units, the heat loss per second from each half of the basin is
( ) (
Q = 2 1000 10 3 44 10 3 ( 0.07565 ) 45.45 10 6 = 3.03 1011 J )
To find the annual heat loss from the ocean basin, this result must be doubled and
multiplied by the number of seconds in a year (3.1536 107), which gives
Q = 3.03 1011 2 3.1536 10 7 = 1.9 1019 J yr1

101
4.3 Geoelectricity

20. At the interface between two layers with electrical resistivities 1 and 2, as in the figure
below, the electrical boundary conditions are: (i) the component of current density Jz
normal to the interface is continuous, and (ii) the component of electric field Ex
tangential to the interface is continuous. A current flow-line makes angles 1 and 2
before and after refraction, respectively.
z
J1z = J2z
J1
E1x = E2x
1
1
x
2
2

J2

tan 1 2
Derive the electrical law of refraction given by Eq. (4.102): =
tan 2 1

Ohms Law relating current density J, electric field E, and resistivity (Eq. 4.80) is
E = J .
The normal component of current density above the interface is J1z = J1 cos1 ; below
the interface it is J 2 z = J 2 cos 2 . The tangential component of electric field above the
interface is E1x = E1 sin 1 ; below the interface it is E2 x = E2 sin 2 .

Equating the normal components of current density as in the figure we get


J1 cos1 = J 2 cos 2
Equating the tangential components of electric field we get E1 sin 1 = E2 sin 2 ,
which can be rewritten using Ohms law as
1 J1 sin 1 = 2 J 2 sin 2 . Thus,
J cos 2
1 2 sin 1 = 2 J 2 sin 2
cos1

sin 1 sin 2
1 = 2
cos1 cos 2
tan 1 2
=
tan 2 1

102
21. What is the effective resistivity of a slab of thickness L composed of two half-slabs each
of thickness L/2 and with resistivities (2) and (/2), respectively, as in the diagram?

0 L/2 L

2 /2

The resistance R of a conductor of length l and cross-sectional area A, made of


material with resistivity , is given by
l
R=
A

Consider the flow of electrical current along a path of arbitrary cross-sectional area A
normal to the interfaces of the given slab. The resistance of the left half of the slab is
(L / 2) L
RL = ( 2 ) =
A A

The resistance of the right half of the slab is


(L / 2) L
RR = ( / 2 ) =
A 4A

The total resistance of the slab is the sum of these two parts: R = RL + RR
L L 5 L
R= + =
A 4A 4 A

The effective resistivity eff is defined by


L
R = eff
A
5
Thus, the effective resistivity of the slab is eff =
4

103
22. Seawater is contaminating an aquifer that is the source of drinking water for a seaside
town. The following measurements of apparent resistivity (a) were made at various
electrode separations (a) with the expanding-spread Wenner method to investigate the
leak.

a a a a a a
(m) (m) (m) (m) (m) (m)
10 29.0 140 19.8 280 8.7
20 28.9 160 18.0 300 7.8
40 28.5 180 16.3 320 7.1
60 27.1 200 14.5 340 6.7
80 25.3 220 12.9 360 6.5
100 23.5 240 11.3 400 6.4
120 21.7 260 9.9 440 6.4

(a) Estimate the electrical resistivity of each layer.


(b) Divide the apparent resistivity at each position by the resistivity of the upper layer,
then plot the normalized resistivity against electrode separation on a log-log diagram on
the same scale as the model curves in Fig. 4.53.
(c) Match the measured curve with the model curves and estimate the depth to the
interface.

(a) The apparent resistivity varies with electrode separation as follows:

The curve is characteristic of a two-layer structure (see 4.3.5.6), in which a good


conductor (the aquifer) lies beneath a more resistive near-surface layer.

104
At small electrode separations the current lines are confined to shallow depths and
the apparent resistivity is that of the near-surface layer, which is approximately 29 m.
At large electrode separations the current-lines penetrate more deeply, so that most of
the current flows in the conducting aquifer. The apparent resistivity becomes asymptotic
to the resistivity of the aquifer, which, from the values in the table, is 6.4 m.

(b) Normalizing the apparent resistivities a by the resistivity m of the top layer gives:

a a/ m a a/ m a a/ m
(m) (m) (m)
10 1.000 140 0.683 280 0.300
20 0.997 160 0.621 300 0.269
40 0.983 180 0.562 320 0.245
60 0.934 200 0.500 340 0.231
80 0.872 220 0.445 360 0.224
100 0.810 240 0.390 400 0.221
120 0.748 260 0.341 440 0.221

A double logarithmic plot of these values is shown in the following figure:

(c) Using the estimated resistivities of the upper and bottom layers, the normalized k-factor
used in the interpretation of Wenner type-curves (Fig. 4.53) has the value
2 1 6.4 29
k= = = 0.64
2 + 1 6.4 + 29

105
Further interpretation requires use of the type-curves. These are plotted on double-
logarithmic scales. In the figure below the measured data are shown (in blue) against the
relevant part of the type-curves (in red). The ordinate is the observed resistivity
normalized by the resistivity of the upper layer; the abscissa of the type-curves is the
normalized electrode separation (a/d), where d is the unknown depth to the interface
between the layers, and the abscissa of the observed curves is a logarithmic plot of
electrode separation.

The curves must be shifted horizontally until an optimal visual agreement of the
measured curve with the type curves is obtained. This is found for a k-value in the range
0.6 k 0.7, in broad agreement with the value k = 0.64 found above. The fit is not
perfect: for a < 200 m the measured curve fits better the type-curve k = 0.6, while for
a > 300 m it fits better the type-curve k = 0.7.

At the optimal match of the curves, the ratio (a/d) = 1 agrees with an electrode
separation of a 95100 m, so this is also the estimated depth d of the interface.

106
23. In the Schlumberger resistivity method the separation of the current electrodes L is
much larger than the separation a of the voltage electrodes. Suppose that the mid-point
of the voltage pair is displaced by a distance x from the mid-point of the current
electrode pair. Show that, for (L 2x) >> a, the apparent resistivity is given by

V ( L 4x )
2 2 2

a =
4 I a ( L2 + 4x 2 )

The geometry of the problem is shown in the following figure; the electrode
arrangement is asymmetric:

The general formula for a 4-electrode resistivity investigation is given by Eq. (4.89):


V 1 V
= 2 = 2 F
I 1 1 1 1 I

rAC rCB rAD rDB

where F is a geometric factor determined by the electrode array. For the Schlumberger
method shown in the figure the separations of the potential and current electrodes are
L a L a
rAC = x rCB = +x+
2 2 2 2
L a L a
rAD = x+ rDB = +x
2 2 2 2
The geometric factor F is in this case
1 1 1 1 1
=
F rAC rCB rAD rDB
1 1 1 1
= +
La L+a L+a La
x + x x + x
2 2 2 2
1 1 1 1
= 2 +
(L a) 2x (L a) + 2x (L + a) + 2x (L + a) 2x

107
Rearranging the order of the terms, this is equivalent to
1 1 1 1 1
= 2 +
F (L 2x) a (L 2x) + a (L + 2x) a (L + 2x) + a
(L 2x) + a (L 2x) + a (L + 2x) + a (L + 2x) + a
= 2 +
(L 2x)2 a 2 (L + 2x)2 a 2
a a
= 4 + 2
(L 2x) a
2 2
(L + 2x) a
2

If we now apply the condition that (L 2x) >> a, the equation for F reduces to
1 1 1
= 4a + 2
F (L 2x) 2
(L + 2x)
(L + 2x)2 + (L 2x)2
= 4a
(L2 4x 2 )2
(L2 + 4x 2 )
= 8a
(L2 4x 2 )2

The apparent resistivity is therefore given by


1
V V (L2 + 4x 2 )
= 2 g(r) = 2 8a 2 , which, on simplifying, gives
I I (L 4x 2 )2

V (L2 4x 2 )2
=
4 I a(L2 + 4x 2 )

108
24. In the double-dipole resistivity method it is common to keep the separation of the pairs
L an integer multiple n of the distance a between the electrodes in each pair, i.e. L = na.
(a) Rewrite the formula for the apparent resistivity with this assumption.
(b) If L is very large compared to a, modify the formula to show that the apparent
resistivity is proportional to n3.

(a) The apparent resistivity for the double-dipole method is given by Eq. (4.89):

=
2
(
VL L a
2
)

I a2

Substituting L = na,

(
V na n a a
2 2 2
)
=
I 2
V
= n n 1 a
2
( )
a I

(b) If L >> a, then n >> 1, and the term (n2 1) reduces to n2. The expression for the
apparent resistivity becomes
V 3
= na
I

Thus the apparent resistivity is proportional to n3.

109
25. Consider a double-dipole configuration in which the electrode pairs are not collinear but
are broadside to each other (i.e., normal to the line joining them). The electrode
separation is a and the distance between the mid-points of the pairs is L = na. Show
that, for large values of n, the apparent resistivity is given in this case by
V
a = 2 n 3a
I

The geometry of this double-dipole configuration is shown in the following figure:

The inter-electrode distances are:

rAC = L rCB = L2 + a 2
rAD = L2 + a 2 rDB = L

Substituting in the geometric factor F for the 4-electrode resistivity method gives:
1 1 1 1 1
=
F rAC rCB rAD rDB

1 1 1 1
= +
L L +a
2 2
L +a
2 2 L
1 1
= 2
2
L L2 + a
L2 + a 2 L
= 2
L L +a
2 2

If L = na, this simplifies to

1 2 n2 + 1 n
=
F a n n2 + 1

The apparent resistivity is therefore

V V n n2 + 1
= 2 F= a
I I n +1 n
2

110
This can be simplified further to

V n n2 + 1 n2 + 1 + n
== a 2
I n +1 n n +1 + n
2

V n n +1 n +1 + n
= a
2 2
( )
I (
n2 + 1 n2 )
=
V
I
an n 2 + 1 ( n2 + 1 + n )
For n >> 1, we can write n 2 + 1 n and so
V
= 2 an 3
I

26. Calculate the speed of an electromagnetic wave in (a) basalt (dielectric constant = 12)
and (b) water ( = 80.4).

The speed v of an electromagnetic wave is obtained from Eq. (4.126):


1
v2 = , where 0 is the magnetic field constant and is the permittivity of the
0
medium, which is related to the permittivity 0 of a vacuum by the dielectric constant :
0 = 0
1 1 1 1
Thus v 2 = = = c 2 , where c 2 = and c is the speed of light in a vacuum
0 0 0 0 0
(299,792 km s1).

In applied geophysics the velocities of electromagnetic waves are commonly


given in units of meters/nanosecond (m ns1) or meters/microsecond (m s1); in these
units, c 0.3 m ns1 300 m s1.

(a) In basalt, with = 12,


1 1
v= c= c = 0.289c = 86, 500 km s1 (0.0865 m ns1)
12

(b) In water, with = 80.4,


1 1
v= c= c = 0.1115c = 33, 400 km s1 (0.0334 m ns1)
80.4

111
27. Calculate the skin depths of penetration in (a) granite ( = 5,000 m) and (b) a
pyrrhotite ore-body ( = 5 105 m) for electromagnetic waves in surveys employing
(i) electromagnetic induction ( = 1 kHz) and (ii) ground penetrating radar ( = 100
MHz). Would these methods detect the conducting bodies if they were buried under a
water-saturated soil layer, 3 m thick with resistivity 100 m?

The skin depth for an electromagnetic wave is given by Eq. (4.126):


2
d= =
0 0 f

(a) For the granite body ( = 5,000 m) the skin depth for electromagnetic induction at a
frequency of 1 kHz is:
5000
d= = = 1100 m.
0 f 4 10 7 10 3
2

The skin depth for ground penetrating radar at 100 MHz is:
5000
d= = = 3.6 m
0 f 4 10 7 10 8
2

(b) For the pyrrhotite ore-body ( = 5 105 m) the skin depth for electromagnetic
induction at a frequency of 1 kHz is:
5 10 5
d= = = 0.11 m.
0 f 4 2 10 7 10 3
The skin depth for ground penetrating radar at 100 MHz is:
5 10 5
d= = = 0.36 10 3 m = 0.36 mm
0 f 4 10 10
2 7 8

The penetration depth in wet soil ( = 100 m) is found similarly.


100
For the EM induction method d = = = 160 m .
0 f 4 10 7 10 3
2

100
For the GPR method d = = = 0.5 m .
0 f 4 10 7 10 8
2

The skin depth d is where the signal is attenuated to e1 of its surface value. In fact
the electromagnetic waves penetrate far beyond this depth. At depth 5d the signal still
has 1% (e5) of its surface amplitude. The weakened signal may penetrate further and
induce a detectable response from greater depths.

The 3 m thick layer of wet soil in this exercise would be penetrated and the buried
conductor surely detected by the 1 kHz EM induction signal. It might also be detected
by the 100 MHz GPR signal, if the transmitter is strong.

112

Das könnte Ihnen auch gefallen