Numerical modeling of the conductive heat transfer in western Anatolia

15 Pages·2004·0.88 MB·English

JOURNAL OF THE BALKAN GEOPHYSICAL SOCIETY, Vol. 6, No. 1, February 2003, p. 1 – 15, 5 figs.

Numerical modeling of the conductive heat transfer
in western Anatolia

G. Göktürkler *, M. Salk * and C. Sari *

* Dokuz Eylul University, Department of Geophysics, Tinaztepe Campus, 35160 Buca-Izmir, TURKEY

Abstract: Using the available data, the crustal temperatures were modeled along two profiles
in the western Anatolia. Modeling was achieved by solving a two-dimensional steady-state heat
conduction equation in a heterogeneous medium by successive overrelaxation method of finite
differences. The Bouguer gravity data were used to construct the crustal structures along the
profiles. The average crustal thickness was taken as 30 km. An exponential decrease with depth
in heat production was assumed. The temperatures in the grabens and regions under them are
obtained as relatively higher than those in the surrounding regions. The thickness of the
seismogenic zone and depth to the Curie isotherm in the region were estimated as 10 and 15 km
respectively. Also, the temperature at the bottom of the crust was calculated as 1075-1100ºC
approximately. Sedimentary fills in the grabens affect the temperature distribution in the upper
crust.

INTRODUCTION 1995), we have carried out a two-
dimensional modeling in the region using
Temperature distribution within the available data to define temperature
earth's crust has important effects on the distributions along several cross-sections.
active tectonics and seismicity. Even The main driving force of active
though it is one of the most important tectonics in Anatolia is the lithospheric
parameters in the earth, our knowledge of regime connected with the continental
the thermal state within the crust is poor. collision in the Alpine-Himalayan zone.
Temperature measurements are usually Interaction among the Eurasian, African
restricted at shallow depths and are under and Arabian plates controls the active
the influence of fluid motions, topography tectonics in the western Anatolia. Until the
and local geology. Therefore, some beginning of the late Miocene the region
mathematical models are widely used to was under a north-south shortening, and
estimate temperature change with depth this yielded a thick crust (> 50 km) (Le
(Mayhew, 1982). Based on some Pichon and Angellier, 1981; Sengor, 1982;
assumptions, numerical modeling of steady- Jackson and McKenzie, 1988; Yilmaz,
state conductive heat transfer is one of the 1989). Since the late middle Miocene the
methods, which is widely preferred by region has experienced a north-south
researchers, to define temperature extension. This caused the thick and
distribution both horizontally and vertically partially melted crust (lower parts only) to
in the crust and upper mantle (Sams and be stretched. Thus, this generated a thin and
Thomas-Betts, 1988; Clauser and Villinger, brittle crust (~30 km) (Ezen, 1991;
1990; Baumann and Rybach, 1991; Swift, Saunders et al., 1998). There are several
1991; Cermak et al., 1991; Kukkonen, and types of models for the extension
Joeleht, 1996; Jokinen and Kukkonen, mechanism in the region. These are tectonic
1999). Defining temperature distribution escape (Dewey and Sengör, 1979; Sengör
may reveal the characteristics of thermal 1979, 1980, 1987; Sengör et al., 1985),
regime in the crust. Since the thermal back-arc spreading (Le Pichon and
regime in western Anatolia is anomalous
(Hurtig et al., 1992; IlkiSik, 1995; Tezcan,
1
2 Göktürkler et al.

FIG. 1. Major tectonic features and the profiles along which temperature fields were modeled
(faults are drawn by B. Rojay from the Landsat images)
(Modified from SimSek and Güleç, 1994).

Angelier, 1979; Meulenkamp et al., 1988) trending major grabens with northeast-
and orogenic collapse (Dewey, 1988; southwest trending secondary (cross-
Seyitoglu and Scott, 1996; Seyitoglu et al., cutting) grabens (Koçyigit, et al., 1999)
1992). (Fig. 1). Western Anatolia is one of the
The present geomorphology is most rapidly deformed continental regions
characterized by a series of east-west in the earth, and the present widely spread
Numerical modeling of the conductive heat transfer 3

seismicity is an indicator of this According to these measurements, the
deformation (McKenzie, 1978; Alptekin et eastern Mediterranean Sea has relatively
al., 1990). However, there is a close low heat flow values mainly in the southern
association between the present day part. Also, low heat flow is observed in the
deformation and the rate of convergence Black Sea. However, high heat flow values
and subduction along the Hellenic-Cyprus are obtained after correcting original
Arc (Royden, 1993a,b). observations for the fast sedimentation
High heat flow and seismicity, intensive (Erickson, 1970). In the Aegean Sea three
faulting and volcanism are the main high heat flow anomalies connected with
characteristics of the region. Tezcan (1995) tectonic zones are observed. The highest
mentions that the heat flow values over heat flow (>120 mWm-2) is related to the
Anatolia are much higher than those over subduction zone (Fytikas, 1980).
the Meditterranean Sea and Black Sea. A The overall heat flow values over
similar result was also reported by IlkiSik Turkey are generally much higher than
(1995). He calculated a mean heat flow those over the Mediterranean Sea and Black
value as 107±45 mWm-2, which is Sea (Tezcan, 1995), although a poor heat
approximately 50% higher than the world flow measurement coverage is available in
average. Also, the surface temperatures of Turkey (Tezcan, 1995; IlkiSik, 1989). The
hot springs in the western Anatolia are first heat flow map for Turkey was
generally higher than those in the eastern published by Tezcan (1979) using the
Anatolia (Erentöz and Tenek, 1968). measurements in the wells drilled for
Moreover, there is substantial thermal geothermal energy. This map has been
activity in the area revealed by numerous revised by including temperatures measured
hot springs, fumaroles, hydrothermal in 204 oil and coal wells mostly drilled in
alterations and recent mineralizations. the south-eastern Turkey and Thrace
Major geothermal fields are usually (Tezcan and Turgay, 1991) (Fig. 2). They
associated with the horst-graben systems calculated heat flow using geothermal
and active-subactive volcanism. For gradients in exploration wells and assuming
instance, the Kizildere and Pamukkale a constant thermal conductivity because no
geothermal fields exist in the Büyük thermal conductivity measurements were
Menderes graben. Especially, major performed in the wells. The temperature at
geothermal systems exist where the north- the bottom of the hole and the one at 1m
south and east-west trending horst-graben below the surface were employed to
systems intersect each other. calculate the geothermal gradient. Since
A two-dimensional steady-state most of the wells have been drilled in
conductive heat flow modeling was sedimentary units, a typical value of 2.1
performed along some profiles in the Wm-1K-1 has been assigned as the thermal
western Anatolia (Fig. 1) to estimate conductivity in all calculations.
temperature distribution in the crust. Based According to the map in Figure 2, the
on the modeling results, the temperatures highest heat flow values (>140 mWm-2) are
in the grabens and regions under them are observed in the western Anatolia. It is
obtained as relatively higher than those in obvious that there is a close association
the surrounding regions. The temperature at between the heat flow anomalies and horst-
the bottom of the crust was calculated to be graben systems. Central Anatolia, on the
approximately 1075-1100ºC. Also, the other hand, is characterized by values
grabens affect the temperature distribution between 70 and 100 mWm-2. There are two
in the upper crust. heat flow closures with the values greater
than 100 mWm-2. The shapes of the
HEAT FLOW DATA anomalies imply that they are very likely
related to the young volcanism (Koçak,
Unlike western Anatolia, a large number 1990). The eastern part of the map (not
of heat flow measurements is available in shown in Fig. 2) does not indicate any
Europe and the surrounding seas. anomalies. The values range from 70-100
4 Göktürkler et al.

FIG. 2. Heat flow map of western Anatolia (from Tezcan and Turgay, 1991), calculated using
geothermal gradients in exploration wells and assuming a constant thermal conductivity
(k = 2.1 Wm-1 K-1 )

mWm-2. Since there is no enough sampling Büyük Menderes, Küçük Menderes and
in the region, this part of the map may not Gediz grabens). Also, there is a close
represent the actual thermal state of the correlation between high heat flow (> 100
crust in the eastern Anatolia. Due to the mWm-2) and Tertiary and younger
available temperature data collected from volcanism.
the wells drilled for oil exploration, the Recently, MTA (The Directorate of
southeastern part of the map (not shown in Mineral Research and Exploration) and
Fig. 2) is also representative of the thermal TÜBITAK (The Scientific and Technical
regime in the area. The values vary between Research Council of Turkey) together with
50 to 90 mWm-2 approximately, and there some universities carried out some heat
are some closed contours in the region. flow measurements in the western Anatolia
Tezcan (1995) considers them to be using shallow wells drilled for water
associated with buried topography. explorations (which are not in use). This
IlkiSik (1995) determined a regional study gives similar results like previous
heat flow pattern in the western Anatolia ones. Based on these measurements, an
using silica geothermometers from 187 average heat flow is 103.68 mWm-2 in the
thermal springs. He calculated the mean region (IlkiSik et al., 1996).
value of the heat flow as 107– 45 mWm-2. Following Rybach and Buntebarth
This value is about 50-60% higher than the (1982), IlkiSik (1995) evaluated an average
world average. IlkiSik's (1995) heat flow heat production rate from radioactive decay
pattern given in semi-quantitative nature for the upper crust as 3.73 m Wm-3 by U, Th
indicates that there is an association and K concentrations in the volcanic and
between high heat flow and the grabens (the plutonic rocks of the western Anatolia
Numerical modeling of the conductive heat transfer 5

given by Ercan et al. (1983). Assuming the et al. (1998) investigated the crustal
characteristic depth (h) to be 13-15 km, structure of western Turkey using
r
IlkiSik (1995) also suggested the crust’s teleseismic receiver functions. They used
contribution to the surface heat flow and the the data at two locations, Kula and USak.
reduced heat flow as 52 and 55 mWm-2 They estimated the crustal thicknesses at
respectively, based on the linear these locations as ~30 and ~34 km,
relationship between the near surface heat respectively. While the crustal thickness
production and surface heat flow. increases from Kula to USak, the effects of
As to the magnitude of the mean heat the extension in the region decrease. Makris
flow value in the western Anatolia, both (1978) used topography, gravity and limited
Tezcan and Turgay (1991) and IlkiSik seismic refraction data in the Aegean Sea
(1995) calculated very high main values, and estimated a crustal thickness ranging
about 120 and 107 mWm-2, respectively. from about 22 km in the central Aegean Sea
They both used the measurements taken to over 40 km in the Anatolian Plateau.
only in geothermal systems, and this has Mindevalli and Mitchell (1989) give an
likely affected the magnitudes because average crustal thickness of about 40 km
hydrothermal circulation plays an important which is thinner on the west coast (~34 km)
role in the geothermal systems. This using fundamental-mode Rayleigh and
circulation might produce high near surface Love wave group velocities for western
heat flow, which may not represent the Turkey. Also Ezen (1991), based on
deeper heat flow and main values of the Rayleigh wave dispersion, reported similar
region. result (~30-32 km thick) for the crust in the
These studies reveal that the western western Anatolia.
Anatolia has anomalous heat flow Gravity data can indicate the broad
compared to the rest of the country. Since features of the crust and upper mantle
the heat flow data are insufficient, to structure. Therefore, the Bouguer gravity
explain the cause of such high heat flow is map of Turkey published by MTA (1979)
problematic. A generally accepted on the scale of 1/500000 were used to
hypothesis is that there is a crustal thinning model crustal structures along the Profiles I
in western Anatolia as a result of the and II (Fig. 1). The profiles were sampled
extensional tectonic regime, and this also using a 5-km sampling interval.
formed a quite shallow asthenosphere under Appropriate forward modeling was used for
western Anatolia. Therefore, the constructing crustal models (Talwani et al.,
asthenosphere should be the main source of 1959). The method is based on an
anomalous heat flow in the region evaluation of gravity anomalies caused by
(Alptekin et al., 1990; IlkiSik, 1995). On bodies of arbitrary shapes. The method
the other hand, Tezcan (1995) suggests an performs a double integration analytically
association between high heat flow and a single integration numerically. The
anomalies and metamorphic massifs (the body whose anomaly has to be evaluated is
Menderes, Kazdag, KirSehir massifs, etc.). first represented by contours. Each contour
He considers high content of radioactive is then replaced by a polygonal lamina. By
elements in massifs as the source of the making the number of sides of this polygon
high heat flow. larger, the contour boundary can be
represented as accurately as desired. The
CRUSTAL STRUCTURE numerical integration is then carried out
with respect to Z axis (for details, see
Modeling temperature distribution Talwani, 1965).
within the crust requires the knowledge of Generally, rocks making up the upper
crustal structure, thermal properties of the parts of the crust tend to be felsic in
geological units and some boundary composition and have densities between 2.6
conditions. Even though the crustal and 2.8 g/cm3. Rocks in the lower part of
structure in western Anatolia is of great the crust are more mafic in composition and
interest, there are no data from explosion their densities range from 2.8-3.0 g/cm3.
seismology that can be employed to Mantle is mainly composed of ultramafic
construct a reliable crustal structure rocks with the density of 3.2 g/cm3
(Alptekin et al., 1990). Recently, Saunders approximately (Dobrin and Savit, 1988).
6 Göktürkler et al.

FIG. 3.. Crustal structure used in the thermal modeling. Gravity data (A) and the model (B)
along the Profile I (top); gravity data (A) and the model (B) along the Profile II (bottom).
Numerical modeling of the conductive heat transfer 7

Therefore, considering the geology and
rock distribution in the region, we assigned METHOD OF THERMAL MODELING
the density values of 2.45, 2.65, 2.9, 3.2
g/cm3 for the sedimentary fill in the To calculate temperature field in the
grabens, upper and lower crust and upper crust a two-dimensional steady-state heat
mantle, respectively during the forward conduction equation was numerically
modeling of the gravity data. solved using successive overrelaxation
Figure 3 shows the gravity data and method of finite-differences. The two-
crustal models along the profiles. The dimensional steady-state heat conduction
length of the Profile I is approximately 200 equation is given as the following:
km. It crosses the Büyük Menderes, Küçük
Menderes and Gediz grabens. Low gravity ¶ (cid:230) ¶ T (cid:246) ¶ (cid:230) ¶ T (cid:246)
values exist at the grabens. The length of (cid:231) k (cid:247) + (cid:231) k (cid:247) + A= 0 (1)
¶ xŁ ¶ x ł ¶ zŁ ¶ z ł
the Profile II is about 350 km. This profile

includes the Simav-Gördes, Demirci
where k(x,z) is thermal conductivity; T(x,z)
subgrabens, the Gediz, Büyük Menderes
is temperature field; A(x,z) is heat
grabens and Mugla-Yatagan subgrabens.
production; x and z are coordinates
Again, the grabens are characterized by low
(Cermák et al., 1991).
gravity values. On the other hand, there is a
Following Kukkonen and Jõeleht
gradual increase in the Bouguer gravity
(1996), we took into account the
anomalies toward the west along the Profile
temperature dependence of thermal
I indicating a relative thinning of the crust
conductivity (Wm-1K-1), including the
in this direction whereas no such thing
effects of both lattice conductivity and
could be observed along the Profile II
radiative heat transfer as:
running in the north-south direction. This

fits very well with the regional setting of
the Aegean domain where there is a back- k = ( ko )+b(T + 273.15K)3 (2)
arc uplifting behind the Hellenic arc. +
1 aT
According to geological data, grabens in

the western Anatolia have asymmetric Here, k is the thermal conductivity (Wm –1
forms (Yilmaz et al., 2000; Sözbilir, 2001). K-1) in othe surface conditions (20(cid:176) C), T is
Our forward modeling results also indicate
the temperature ((cid:176) C), a (K-1) and b (Wm-1K-
such a feature. In our models, the depths of
4) are the experimental parameters. For a
the grabens range from 2.5 to 3.5 km; they
the value of 0.0015 was assigned
are asymmetric; the average thickness of
approximating the lattice conductivity of
the crust is about 30 km (Makris, 1978;
common rock types (Zoth and Haenel,
Mindevalli and Mitchel, 1989; Ezen, 1991;
1988) for temperatures lower than 1000-
Saunders et al., 1998). Paton (1992) also
(cid:176)
1200 C. The radiative heat transfer
studied Bouguer gravity data over the
becomes important in the lower crust and
Büyük Menderes and Gediz grabens and
(cid:176)
mantle at temperatures beyond 800 C. The
estimated an average depth to the basement
b value was assumed as 1(cid:215)10-10 (Schatz and
as ~1.5-2 km and Gürer et al. (2002)
Simmons, 1972) for ultrabasic mantle
suggest a varying sediment thickness in the
material. The variation of thermal
range of 1-3.8 km along the Gediz graben
conductivity with pressure has been ignored
based on geoelectrical data (mainly
since its effect on the conductivity is small
magnetotelluric data). Since there are no
compare to temperature's effect. Also, how
deep seismic studies in the study area, the
thermal conductivity changes with pressure
knowledge of the sediment thickness in the
is not well documented (Sams and Thomas-
grabens is not satisfactory. Even though
Betts, 1988).
there are wells drilled for geothermal
For heat production, we assumed an
explorations they do not give enough
exponential decrease with depth for
information for sediment thickness because
radioactivity in the crust because it is one of
they are located close to the borders of the
the available models (step and linear
grabens.
models) satisfying the linear heat flow
8 Göktürkler et al.

relation (Lachenbruck, 1970). Also it is the corresponding thermal conductivities in
model which represents well the actual surface conditions (k).
o
situation in the continents (Condie, 1976).
This model is defined as follows: Table 1

(cid:230) - (cid:246) Units k
( ) z o
A z = A exp(cid:231)(cid:231) (cid:247)(cid:247) (3) (Wm-1 K-1) Reference
0 Ł h ł

r
Sedimentary fill in grabens 2.1
where A(z) (m Wm-3) is the heat production (Tezcan and Turgay, 1991)
at depth z; A is the near-surface heat Upper crust 3.0
o
production (m Wm-3); h is the characteristic (Cermák et al., 1991)
r
depth (km), and z is the depth (km). For A Lower crust 2.0
o
we assigned the value of 3.73 m Wm-3 (Cermák et al., 1991)
(IlkiSik, 1995). h is taken as 12 km for the Upper Mantle 2.8
r
western Anatolia considering the average assumed
thickness of the upper crustal layer
(granitic-gneissic part of the crust) in the
models. The depth z ranges from 0-35 km. RESULTS AND DISCUSSION
As to the heat generated by radioactive
decay, no heat production in the grabens Since we assign a constant heat flow
was considered, i.e. the heat production value to the lower boundary of the models
equals to zero everywhere in the grabens. and the topographies of the model
Solution of equation (1) is subject to the interfaces change smoothly, relatively flat
following boundary conditions: isotherms are obtained. Thus, we preferred
to plot temperatures at different depths
• T(x, z=0) = 18ºC, which is considered versus distance to reveal spatial temperature
as the mean annual surface temperature changes. Figure 4 and Figure 5 show the
(Tezcan, 1992), calculated temperatures along the Profiles I
• No-flux boundary conditions are and II, respectively. These figures indicate
that temperature distributions in the upper
applied to the vertical boundaries of the
and lower crust are remarkably different.
(cid:230) ¶ T (cid:246)
models, i.e. (cid:231) (cid:247) = 0 at x=0 and The distributions in the upper crust exhibit
Ł ¶ x ł undulating variations, whereas the ones in
x=L, L is the length of the model, the lower crust show some smooth spatial
• A constant vertical heat flow is changes.
assigned along the lower boundaries of Figure 4 shows the temperatures at the
the models. Considering the linear depths of 1, 5, 10, 15 and 30 km along the
relationship between the surface heat Profile I. Considering the depth range of the
flow and near-surface heat production, grabens in the models, Figure 4a displays
the reduced heat flow is the heat the temperature change along a depth level
coming from the mantle and lower crust cross-cutting the Büyük Menderes, Küçük
(Condie, 1976). Therefore, integrating Menderes and Gediz grabens. According to
the heat production function A(z) from this plot, temperatures outside the grabens
(cid:176)
h to 35 km yields the lower crust’s are approximately 55 C while they are
r
contribution. Subtracting this from the about 65(cid:176) C inside the grabens. In other
reduced heat flow should give the words, temperatures at 1 km in the grabens
contribution of mantle, which can be are about 10(cid:176) C higher than those in the
used as lower boundary condition; and same depth in the surrounding region. At
this is about 30 mWm-2. the depth of 5 km (Fig. 4b), we see a
similar situation even though the grabens
Solving equation (1) requires a certain do not extent down to this depth. The
crustal model and thermal conductivity regions under the Büyük Menderes and
values for the geological units (Cermák et Gediz grabens are about 20(cid:176) C, and the
al., 1991). Table (1) shows the units and region under the Küçük Menderes graben is
Numerical modeling of the conductive heat transfer 9

about 10(cid:176) C hotter than the rest of region. relatively high thermal conductivity (3.0
This can be explained considering the Wm-1K-1). Because of this, the grabens
thermal conductivities of the various units. behave like thermal insulators, which do
Because the grabens are filled by clayey not let heat flow transfer easily through the
sediments (Tezcan, 1995) whose thermal surface, and this causes the grabens and the
conductivities are as low as 2.1 Wm-1K-1, regions under them to become relatively
they cannot conduct the heat from below as hotter compared to the surrounding regions.
fast as the surrounding region which has
SW NE
a h=1 km
B. Menderes K. Menderes Gediz
b h=5 km
)
C
(
E

R
U
T c
A h=10 km

R
E
P
M
E
T
h=15 km
d
e
h=30 km
D I S T A N C E (km)

FIG. 4. Modelled temperatures at various depths along the Profile I. Plots show temperature
change with distance at the depths of 1, 5, 10, 15, 30 km.
10 Göktürkler et al.

Temperatures at 10 km (Fig. 4c) range gradually to the north of this point, reaching
(cid:176) (cid:176)
from 345-360 C and they present variations 500 C at the end of the profile. Based on
which are similar to Figures 4a and 4b. But this, it is concluded that the thickness of the
the effect of the Küçük Menderes graben on magnetic crust along the Profile I should be
the temperatures is not as strong as the ones around 15 km.
of the Büyük Menderes and Gediz grabens. In our crustal models, the average
It seems that the region under the Küçük thickness of the crust in western Anatolia is
Menderes graben is a few degrees warmer about 30 km. Figure 4e depicts the
than the surrounding region. The depth temperatures at the depth of 30 km, they
extent of the graben can explain this, as the can be regarded as the temperatures in the
depth of the Küçük Menderes graben is vicinity of Moho discontinuity, i.e.
quite shallow and it has less sedimentary temperatures at the bottom of the crust.
fill. Thus, it cannot affect the deeper According to this figure, they range
regions as much as the Büyük Menderes approximately from 1075 to 1145(cid:176) C.
and Gediz grabens. Temperatures at the southwest end of the
Usually, parts of the crust down to the model are about 1100(cid:176) C, increasing
– (cid:176)
depths where 300 100 C isotherms exist towards the northeast and reaching around
are considered to be brittle enough to 1145(cid:176) C at the distance of 80 km. After this
nucleate an earthquake (Chen and Molnar, point they start to gradually decrease, and at
1983). Generally, brittle-ductile transition the northeast end of the model temperature
in the crust is assumed to be at the depth is about 1075ºC. We can therefore conclude
(cid:176)
where 350 C isotherm takes place that the temperature at the bottom of the
(Kukkonen and Jõeleth, 1996); and the crust along the Profile I is about 1100ºC.
brittle part of the crust is considered as Figure 5 shows temperatures at depths
seismogenic zone. As to the temperature of 1, 5, 10, 15 and 30 km along the Profile
(cid:176)
range in Figure 4c (from 345 to 360 C), we II. Like the temperatures along the Profile I,
can conclude that the seismogenic zone is there is noticeable difference between the
about 10 km thick along the Profile I. This temperatures in the upper and lower crust.
result is also in accordance with Similar to the Profile I, the distribution in
seismological observations. Jackson and the upper crust shows undulating variations,
McKenzie (1988) suggest a 10-15 km thick whereas the ones in the lower crust
seismogenic zone in the western Anatolia; smoothly change with distance.
and Eyidoðan and Jackson (1985) has Figure 5a displays the temperatures at
similar result considering the focal depths the depth of 1 km, that is, a depth level
of the largest normal faulting earthquakes cross-cutting the Simav-Gördes, Demirci,
which are about 6-10 km in the region, and Gediz, Büyük Menderes and Mugla-
this implies that the seismogenic zone Yatagan grabens. Like the situation along
should be about 10 km thick. the Profile I, the grabens have relatively
A temperature at which a magnetic higher temperatures, about 65ºC inside the
mineral loses its ferromagnetism is called grabens and 55ºC in the surrounding region,
as the Curie temperature. Even though the i.e. the temperatures at 1 km in the grabens
magnitude of the Curie temperature are about 10ºC higher those at the same
depends upon the mineralogy of the rocks, depth in the surrounding region. Because of
it ranges in a narrow zone, which is about the geometry of the Mugla-Yatagan graben
520-560(cid:176) C within the most part of the temperature gradually rises up to 65ºC
continental crust. Also, the Curie isotherm toward the south. Temperatures at 5 km
defines the base of the magnetic crust (Fig. 5b) have similar character but
(Mayhew, 1982). Figure 4d shows increased magnitudes. The highest
(cid:176)
temperatures of this order at 15 km. They temperature (about 210 C) is obtained in
range from 500 to 560°C. They are almost the Simav-Gördes graben. The other
fixed at the value of 560(cid:176) C up to the grabens reach temperatures of 200ºC. This
distance of 130 km from the beggining of shows that the regions under the grabens is
the profile, and they start to decrease approximately 20-25ºC hotter than the
surrounding region, which is about 180ºC.