Available online at www.sciencedirect.

com

ScienceDirect
Russian Geology and Geophysics 56 (2015) 1791–1800
www.elsevier.com/locate/rgg

Investigation of density contrasts and geologic structures of hot springs

in the Markazi Province of Iran using the gravity method
J. Nouraliee a, S. Porkhial b , M. Mohammadzadeh-Moghaddam c,*, S. Mirzaei c,
D. Ebrahimi a, M.R. Rahmani d
a
Niroo Research Institute (NRI), Western End of Dadman Blvd., Shahrake Qods, Tehran, 1468617151, Iran
b
Islamic Azad University, Karaj Branch, Imam Ali Complex, Moazen Blvd, Karaj, POB 31485-313, Tehran, Iran
c
Research Institute of Applied Sciences, Shahid Beheshti University, Daneshjoo Blvd., Tehran, 196151171, Iran
d
Renewable Energy Organization of Iran (SUNA), Energy Affairs Deputy Building, Western End of Dadman Blvd.,
Shahrake-Qods, Tehran, 1468611387, Iran
Received 26 December 2014; accepted 28 May 2015

Abstract

In 2012, the renewable energy organization of Iran (SUNA) performed a gravity survey around hot springs of the Mahallat geothermal
field in the Markazi Province of Central Iran, as part of the explorations and developments of geothermal energy investigation program in
the region. The Mahallat region has the greatest geothermal field in Iran. This work presents interpretation results of various gravity maps
and a calculated 3D inversion model. The residual gravity map shows three negative gravity anomaly zones (A1, A2, and A3) associated with
the geothermal reservoirs in the region. The horizontal gradient maps reveal a complex fault system. In order to attain more information about
the Khorhe geothermal reservoir, a 3D density contrast model was calculated using the Li–Oldenburg method. The attained 3D model provides
an in-depth image of the evolution, showing the density contrast and the A1 zone having a high potential for the geothermal reservoir in the
region. The results also show that the rocks which exist between 1000 and 3000 meters under the Earth’s surface in the A1 zone are the most
suitable aquifers for utilization of geothermal energy.
© 2015, V.S. Sobolev IGM, Siberian Branch of the RAS. Published by Elsevier B.V. All rights reserved.

Keywords: gravity; geothermal system; Bouguer anomaly; Euler; inversion; Mahallat

Introduction al., 2013; Salem et al., 2005; Schiavone and Loddo, 2007).
Therefore, gravity method is one of the most economic
In geothermal exploration, gravity method generally is used geophysical methods for modeling of a geothermal system.
to delineate subsurface structures that control the geothermal The Mahallat geothermal field of the Markazi Province in
system. Basically, a given geothermal reservoir and its fluid Iran is a popular tourist destination due to the occurrence of
content cause density differences between geothermal reser- hot springs and one of the richest geothermal fields in Iran.
voir and the surrounding rocks. The basis of gravity method According to the available scientific data about Abgarm
is density contrast in the rocks. Gravity studies in different geothermal region, it is found that in the NW of Mahallat city
regions of the world have yielded useful results for geothermal has favorable conditions for the exploitation of geothermal
exploration, such as: investigation of topography of basement energy. The local heat flow and the temperature maxima in
in a geothermal field (Salem et al., 2005; Soengkono, 2011), the Mahallat geothermal field originate from strong convec-
magma chambers and intrusive body related to the heat source tional heat transport mainly in the igneous basement.
of the geothermal system (Represas et al., 2013) and deline- In the current study area, several studies have been
ation of faults and fractured zones corresponding to the conducted by the renewable energy organization of Iran
reservoir of geothermal system (Abiye and Haile, 2008; (SUNA) and other research institutes. Using a ground mag-
Gottsmann et al., 2008; Montesinos et al., 2003; Represas et netic study Oskooi at al. (accepted for publication) and
Mirzaei et al. (2013) showed that heat source of the Mahallat
geothermal system originates from a hot igneous intrusive
* Corresponding author. mass which exists at the depth of more than 1.5 km at the
E-mail address: [email protected] (M.Mohammadzadeh-Moghaddam) east of the Mahallat hot springs. Based on geological and
1068-7971/$ - see front matter D 201 5, V.S. So bolev IGM, Siberian Branch of the RAS. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.rgg.2015.11.011
1792 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800

geochemical studies, it was revealed that the hydrothermal down to the deep levels. Some important alterations have also
fluids cycle in the area was controlled mainly by N–S faults been documented for the area.
(Porkhial et al., 2013). Using a Magnetotelluric (MT) survey Some major faults have been identified in the region.
along the 8 km-long profile over the Mahallat hot springs, Generally, they have two main directions, NW–SE and
Oskooi et al. (2013) and Oskooi and Darijani (2013) it is NE–SW. The main geological feature present in the current
determined that Mahallat geothermal system included a cap study area is the segment of the NNE–SSW Mahallat–Abgarm
rock (from 100 to 600 m), a reservoir (from 500 to 2000 m) fault (MAf) zone (Nouraliee and Shahhosseini, 2012). This
and a source (from 1000 m to the deepest reachable point). structure stretches for 50 km, from southeast of Mahallat city
However, these results can’t illustrate the lateral extent of the to the north of Abgarm. All of the Abgarm hot springs is
structures of the Mahallat geothermal system, because their directly connected to this fault zone.
interpretation was two-dimensional, and the amount of data
was insufficient. Geothermal surface manifestations
Gravity survey which is presented in this paper is the most
important geophysical study conducted in the Mahallat geo- The area of the Abgarm region is composed of young
thermal region that allows us to generate a comprehensive volcanic rocks, hydrothermally altered zones, wide travertine
overview of the area. The only other performed geophysical outcrops and intrusive bodies as well. All of the hot springs
and geological surveys are too local and incomplete to be of are genetically related to tectonic activities and concentrated
any use for this purpose. In the current study, a total of 380 along the MAf zone. Due to the hot spring activity, a lot of
gravity observations along several regular profiles were re- travertine deposits have been accumulated in the Mahallat
corded, covering nearly 200 km2. region. Travertine is a chemically-precipitated continental
It should be mentioned that so far no wells have been limestone formed around seepages, hot springs and along
drilled in the Mahallat geothermal region. Therefore all the streams and rivers. It consists of calcite or aragonite, of low
data about this region are based on the surface studies. to moderate inter-crystalline porosity and often high mouldic
or framework porosity within a vadose or occasionally shallow
phreatic environment. Precipitation results primarily through
Geologic setting transfer (evasion or invasion) of carbon dioxide from or to a
groundwater source leading to calcium carbonate super-satu-
In the Markazi Province of Iran (Fig. 1), the geologic ration, with nucleation/crystal growth occurring upon a sub-
structures are characterized by a dextral rotational movement, merged surface (Pentecost, 2005). Travertine deposits usually
which has formed by northward under-thrusting of the Arabian formed around hot springs can be a useful sign of geothermal
plate beneath Central Iran (McKenzie, 1972). The Cenozoic activity in a region, either active or fossil one. Irregular fault
geologic history and the stratigraphy of this region are movements can be responsible for spring migration leading to
complex, due to its tectonic framework. As a result, the the deposition of a range of travertines at different levels
stratigraphic record of the area is made up of units with (Soligo et al., 2002).
different structural characteristics. During the Eocene, igneous Before gravity studies, geological and geochemical studies
activities took place in this area, causing accumulation of were conducted in the area. In geological investigations, a very
volcanic rocks over the Mesozoic and Paleozoic sediments. accurate 1:25,000 scale geological map of the Mahallat
Subsequently, these rocks were thermally metamorphosed by geothermal area was generated (Nouraliee and Shahhosseini,
an Early Miocene monzonitic batholith, elongated in a NW– 2012). All faults and lithological units were studied in detail.
SE direction. Moreover, hydrothermally altered areas were recognized by
Being the most important thermal water sources of the area, Aster satellite images. They were also visited and sampled in
Abgarm hot springs are located 15 km northeast of Mahallat order to perform mineralogical analysis.
city in the Markazi Province. They are considered as a small Based on the Aster satellite images studies, checking
part of the Central Iran zone, which is located in the outcrops and petrography studies, several altered zones were
Urumieh–Dokhtar volcanic belt. The geologic setting of this found in the study area. The most important of alteration types
area and the distribution of hot springs are presented in Fig. 1. include argillitization-sericitization and kaolinitization-aluniti-
The study area is characterized by sedimentary deposits from zation. Argillitization-sericitization alteration is detected in a
the Permian to the Quaternary, with the existence of magmatic close relationship with hydrothermal fluid activities. This type
and volcanic outcrops. Notable formations cropped out in the of alteration is representative of high-temperature geothermal
study area include the Shemshak Formation (shale and systems. Previous studies suggest that the heat source of hot
sandstone), Cretaceous Orbitolina bearing limestones, marly springs is due to the cooling processes of the molten magma
limestones of the Qom Formation and volcanic rocks (grano- (Oskooi et al., accepted for publication). Hot springs are
diorite, tuff and lava). The hot springs (depicted as orange charged by meteoric waters mixed with magmatic fluids
hexagons in the geological map in Fig. 1) caused deposition (Beitollahi, 1996; Oskooi and Darijan, 2013). The hydrology
of travertine sediments with considerable thickness of about and geothermometry studies (Rezaie et al., 2009) showed the
500 m. In this area, faults and fracture zones played an average temperature of the hot springs to be about 46 °C and
important role in the circulation of water from the surface their pH to range from acidic to neutral. The water of hot
 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800 1793

Fig. 1. Study area in the Markazi Province of Iran, plate tectonic structure and large scale geological map adopted from 1:25,000 geological map published by NRI
(Nouraliee and Shahhosseini, 2012). Legend of geological map: 1, trace and fan (Quaternary); 2, travertine (Quaternary); 3, Limestone with fossile fragments
(Tertiary); 4, marl (Oligocene); 5, ignimbrite and rhyolite (Eocene); 6, green tuff and tuffite (Eocene); 7, granite and granodiorite (Eocene); 8, conglomerate and
sandstone (Eocene); 9, sandy conglomerate (Eocene); 10, sandstone and siltstone (Cretaceous); 11, limestone and sandstone (Jurassic); 12, shale and sandstone
(Jurassic); 13, dolomitic limestone (Jurassic); 14, limestone and dolomite (Permian–Jurassic); 15, shale (Permian); 16, dolomite (Cambrian); 17, Main Mahallat–Ab-
garm fault (MAf); 18, fault; 19, hot spring.

springs is saturated by calcite minerals, hinting that the materials mainly originated from magma bodies or other
geothermal reservoir is probably formed in limestone. high-temperature rocks in the earth. The density of infiltrated
A useful way for characterizing of various types of waters was subsequently reduced by the increase in their
geothermal waters is to use chloride-sulphate-bicarbonate temperature. Such a phenomenon occurs in the depth and in
ternary diagram described by Giggenbach (1991). Based on the proximity of cooling magma and/or hot rocks. Moreover,
the chemical data, it was found that the Mahallat hot springs capturing volatile materials can reinforce this process. As a
are slightly acidic and sulphatic (Nouraliee and Ebrahimi, consequence, infiltrated waters become more buoyant and rise
2012). Based on the geothermometery studies, the Mahallat up along the faults. During rising and reacting with surround-
geothermal reservoir approximate temperature was estimated ing carbonate rocks, their temperature decreases, and more
to vary from 88 to 194 °C (Nouraliee and Ebrahimi, 2012). volatiles dissolve in the waters. So, the solubility of the hot
Rezaei et al. (2009) presented a conceptual model of hot waters increases, and therefore, deep karst networks are
springs formation in the targeted area. According to their formed in hosting carbonate sequences. Likewise, mixing of
model, rainfalls infiltrate into the earth along the transverse hot waters with shallow cool waters can control the dissolution
faults, fractures and permeable rocks. The infiltrated waters in carbonate intervals. Rising of waters to upper (shallower)
become more soluble when mixed with vapors and other levels causes depletion in hydrostatic pressure. All of these
volatiles (i.e., H2O and CO2) due to high pressure. Such have resulted in outgassing (e.g., CO2) and formation of
1794 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800

Fig. 2. The topographic map of the study area. The black dots represent the Fig. 3. The complete Bouguer anomaly map of the study area.
gravity stations.

super-saturated waters with respect to the carbonate calcium coordinates ~461,000 m), there is a prominent high-value
(CaCO3). Subsequent precipitation of dissolved carbonate anomaly coinciding with the location of travertine and young
calcium leads to the formation of travertine that we can see traces. However, these rocks are characterized by low density
around the hot springs. and cannot be associated with this gravity anomaly, thus it
seems that there is an igneous intrusive body in the subsurface
of the locality (Mirzaei et al., 2013). From the south towards
Gravity survey Abgarm village, along the MAf zone, there is a low-value
anomaly (–129 to –134 mGal), probably related to the
Data acquisition and reductions fractured zone in the subsurface, as its lithological boundary
clearly follows the strong gravity gradient. In addition, there
Gravity survey was carried out in Mahallat geothermal field is another prominent negative anomaly east of Khorhe. These
by the Renewable Energy Organization of Iran (SUNA) in negative anomaly areas are considered as probable reservoirs
2012. The measurements were made with the Scintrex CG5 of the geothermal system in the region.
gravimeter at 380 stations on an area of approximately
200 km2. Figure 2 shows distribution of the gravity stations Regional / residual separation
in the region. The spacing between the stations varies from
200 m to 450 m, depending on the accessibility of the terrain. The complete Bouguer anomaly contains superposition of
Mahallat gravimetric base station (National Cartographic regional and residual anomalies. The regional Bouguer gravity
Center of Iran—NCC) was used to correct relative values to anomaly is resulting from the presence of deeper and broader
absolute gravity. The 3D station coordinates were obtained structures. The residual Bouguer gravity anomaly is resulting
using trigonometric calculus from topographic survey obser- from the presence of smaller and more superficial ones. In
vations (planimetry and altimetry) performed from geodetic this work, to highlight the gravity anomalies related to the
marks. The accuracy of these determinations is about 0.03 m. interesting source, a regional-residual separation of the com-
International Association of Geodesy 1967 formula and the plete Bouguer anomaly was performed. Correct estimation and
average crustal density of 2.67 g/cm3 were used for calculating removal of the regional anomaly from the observed field yields
free air and Bouguer corrections. Topographic corrections the residual anomaly produced by the target sources. There
were reduced using the “Hammer” method over the 1:25,000 are at least four approaches to doing so (Li and Oldenburg,
topographic maps. 1998a), from more empirical to the graphical ones (see, e.g.,
Figure 3 shows the complete Bouguer anomaly map of the Telford et al., 1990). In the current study, there is a regional
study area. In the northern portion of the map (northeast of trend from the east and the north of the area towards the west.
Khorhe village), there is a high-value anomaly (–110 to Thus the regional field is estimated by least-squares fitting a
–102 mGal) coinciding with the high-density igneous rocks. second-order polynomial to the observed gravity field, so that
Also, in the eastern portion of the map (from UTM (X) this regional trend would not absorb the local gravity anoma-
 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800 1795

lies. When the regional field was calculated, it was subtracted

from the complete Bouguer field to calculate the residual field.
Figure 4 shows the calculated regional and residual fields.
Due to the existence of significant fractures and faults in
a geothermal reservoir, low density zones were formed.
Therefore, in a geothermal exploration program, generally but
not in all cases, negative gravity anomalies are likely places
associated with the geothermal reservoir. In the current study
area, three negative gravity anomalies (A1, A2, and A3 in
Fig. 4) are found in the residual gravity map. Two of them
(A1 and A2) are situated along the MAf zone of the area.
They constitute a particular interest, as they are related to the
main fault where hydrothermal systems are known to exist.
A1 anomaly is the most intense negative anomaly in the area
and reflects the lower density sedimentary deposits where they
are fractured by the main fault. A1 covers a broader area of
the survey and consists of many thermal surface manifestations
such as hot springs, various alterations and thick travertine
layers (Oskooi et al., accepted for publication). This particular
site has been subjected to several geophysical surveys, namely
magnetic and magnetotelluric (e.g., Oskooi and Darijani, 2013;
Oskooi et al., 2013 and accepted for publication), thus it is a Fig. 4. The residual gravity anomaly map of the study area after separation of the
second-order polynomial fitting from the complete Bouguer anomaly. The
matter for this study.
calculated regional gravity map is depicted on the bottom left corner. A1, A2,
A2 anomaly is located along the MAf zone of the area and and A3 indicate the negative gravity anomalies analyzed in the text.
creates a less intense gravity anomaly. A2 seems to be a
fractured zone due to MAf zone and can be a candidate for
a geothermal reservoir in the area but there are no significant
tinuities perpendicular to the direction of gradient and detect-
geothermal surface manifestations in this zone. The previous
ing more clearly the margins of the structures (Nabighian,
studies (Oskooi et al., 2013 and accepted for publication) and
1984).
the geological information show that this zone is not fed well
We used the GRIDGRAD Geosoft extension (available in
by the surface fluids.
the Geosoft Oasis Montaj 7.0.1 software) to calculate the
A3 anomaly is located in western Khorhe, where there are
horizontal gradient in a specified direction. A 3 × 3 point
significant geothermal manifestations. A small and less intense
convolution filter was applied to the gridded data to produce
gravity anomaly was detected in A3 zone. It is believed that
the gradient results. It is notable that the gradient data contain
this anomaly is related to the limestone and sandstone rocks,
an inherent zero reference and should not be reduced to
as a good correlation can be observed between the shape of
residual anomaly data. Therefore in the current study, we used
gravity anomaly and the outcrop of these rocks.
the Bouguer anomaly field for applying the horizontal gradient
Directional horizontal gradients filter.
The direction of the main fault (MAf) in the study area is
Geothermal fields in many sedimentary basins are distrib- approximately N–S. Therefore, two maps of the first-order
uted along the fault-controlled linear trends, and faults detec- horizontal gradients were calculated, one along the direction
tion is often used effectively for target-area selection in E-W, perpendicular to the direction of the main fault to
geothermal exploration. Also, almost all of geothermal reser- enhance the structures almost parallel to the main fault
voirs in various geologic settings are commonly associated (Fig. 5A), and another along the direction N-S, to reveal the
with the fluid conducting faults. In this section, application of structures almost perpendicular to the main fault (Fig. 5B). In
gravitational gradients of the study area is reviewed. It shows Fig. 5, the black lines represent the known faults in the
that directional horizontal gradients are well-suited for inves- geological map and the white dashed lines are discontinuities,
tigation of structures, delineation of the faults and their identified using the horizontal gradient map. These maps
orientations. The first-order horizontal gradient of the gravi- clearly show the importance of the main lineament, the MAf
tational field in a given direction emphasizes the short zone that controls the evolution of the sedimentary basins.
wavelength spatial components of gravity anomalies and In the central part of the map, in Fig. 5A, the gradients are
attenuates the long wavelength spatial components associated stronger and the main fault is recognized as strong gradient
with regional trends and large scale anomalies along that with a NNE–SSW orientation. In the west of this confirmed
direction. This filter enhances the anomalies arising from near main fault, there seems to be another main fault, located
source over those arising from deeper or more distant sources. almost along that. This lineament doesn’t seem to be reflected
The gradient will reach a minimum/maximum in regions at the surface faults but the high values of the anomalies
where the density contrast is higher, thus highlighting discon- suggest that they could have expression in depth and they may
1796 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800

Fig. 5. Directional horizontal gradients of the Bouguer anomaly field along the directions E–W (A) and N–S (B). The black lines are known faults in geological map
and the white dashed lines are discontinuities, identified using the horizontal gradient map.

represent important discontinuities in lithology. Also, in the regional field. By specifying the SI, Euler’s equation is solved
map in Fig. 5B, more than 10 main faults (white dashed lines) using the linear least-squares method (Reid et al., 1990).
have been detected using the first-order horizontal gradient. Thompson (1982) revealed different values of structural
In Fig. 5, both proved faults by the geological map (black index for the geological geometry. The structural index
solid lines) and the faults identified by the horizontal gradients depends on the source geometry and its range is between 0
show the mechanism of the complex faults in the Mahallat to 2 for the gravity sources (and 0 to 3 for the magnetic
geothermal field. These faults play an important role in sources). Stavrev and Reid (2007) show that the structural
hydrothermal circulations which are acting as preferential index depends on the type and physical parameters of the
paths for the hydrothermal fluids circulating in the area. potential field and provides an excellent overview of the
parameters of Euler’s homogeneity equation in general and
Euler depth estimation the structural index in particular. It is notable that a zero index
which implies the potential field is constant, regardless of the
In order to estimate the location of the gravity anomaly distance from the anomaly source. These solutions are physi-
sources, we applied the standard method of Euler deconvolu- cally impossible for real data, and a zero index indicates a
tion to the data. The Euler method uses three orthogonal physical limit which can only be approached as the so-called
gradients of the gravitational field quantity. A 3D form of “infinite” dimensions of the real source increase. In this case,
Euler’s equation can be defined (Reid et al., 1990) as, in practice, an index of 0.5 can often be used to obtain reliable
results when an index of zero would indicate otherwise.
∂g ∂g ∂g ∂g ∂g ∂g
x +y +z + NT = x0 + y0 + z0 + NB, (1) In order to obtain reliable results in Euler method, the
∂x ∂y ∂z ∂x ∂y ∂z
structural index and the window size must be carefully
where ∂g/∂x, ∂g/∂y and ∂g/∂z are the derivatives of the total selected. The window size determines the area in grid cells
field versus x, y, and z axes, respectively. x0, y0, and z0 are used to apply Euler deconvolution. All the points within the
the coordinates of a gravitational source, N indicates the window are used to solve Euler’s equation to locate the source.
structural index (SI) and relates with the rate of change of a It should be large enough to incorporate the entire anomaly
potential field with the distance, and B is the local background to be interpreted and small enough to avoid significant effects
representing the “regional” field within a sliding window with from adjacent or multiple sources (Bournas et al., 2003). In
adjustable size. case of anomalies caused by close sources considered, in-
Equation 1 can be solved in a window centered on a given cluded in a given window, the uncertainty of the solution will
grid point to find the unknown source point (x0, y0, z0) and increase.
the regional field (Reid et al., 1990). This approach introduces In order to apply Euler method to gravity data of Mahallat
a nonlinear relationship between the SI and the unknown region, we used the package of Geosoft Oasis Montaj®
 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800 1797

software. After preparing of data grid, we applied gravitational

field derivatives in three directions of x, y, and z which are
necessary for the Euler deconvolution process. Then, structural
index and window sizes were selected appropriately. To attain
the most reliable structural index, different values were tested.
Since we had no idea about the geometry of the gravity
anomaly source, we tested structural indices from 0.5 (for
dike) to 2 (for sphere) by steps of 0.1. Finally, an appropriate
structural index was estimated to be between 1.2 and 1.4.
Structural indices in this range suggested the best clustering
and regularity in the solutions. Therefore, we considered a
value of 1.3 for structural index as an average. Then Euler
method was applied to residual data by moving different
window sizes. The best results were attained by a window size
of 200 m × 200 m, while the structural index was fixed to be
equal to 1.3.
Figure 6 shows Euler depth estimation solutions on the
residual gravity map using the structural index of 1.3. In this
figure, Euler solutions on the residual gravity map show
different depths for different anomalies. The estimated depth
for both positive and negative anomaly sources in the current Fig. 6. Euler depth estimation solutions on the residual gravity map.
area lies in the range between 500 m and 2000 m. The depth
values estimated in the negative zones (A1, A2, and A3 in 2
Fig. 6) range from about 500 m to 2000 m. Based on the Euler ⎧⎡ ∂w (z) ⎤ ⎫
+ αx ∫ wx ⎨⎢ ⎥ (ρ − ρ0)⎬ dv
depth estimation, estimated depth for the negative anomaly v ⎩⎣ ∂x ⎦ ⎭
source in zone A1 starts from about 500 m in the northern
+ αy ∫ wy ⎧⎨[∂w (z) / ∂x] (ρ − ρ0)⎫⎬ dv
2
part of this zone, while it is located at about 1500 m in the
v ⎩ ⎭
southern part of zone A1 along the MAf zone. Also, in zone
+ αz ∫ wz ⎧⎨[∂w (z) / ∂x] (ρ − ρ0)⎫⎬ dv
A2, estimated depths of the negative anomaly sources range 2
(3)
from about 1000 m to 2000 m and for zone A3 from 750 m v ⎩ ⎭

to 1000 m. The estimated depths in these zones can be

where m is gravimetric model element, ρ0 the reference model,
interpreted as the depth of the geothermal reservoirs in the
wS, wx, wy, and wz weighting functions, αS, αx, αy, and αz
Mahallat geothermal region.
coefficients which affect the relative importance of different
components in the objective function and w(z) a generalized
3D inversion
depth weighting function. This function has flexibility in
constructing a variety of models. The aim of the objective
The three-dimensional density contrast model was calcu-
function is to counteract the geometrical decay of the sensi-
lated with the UBC-GIF Grav3D software with the algorithms tivity with the distance from the observation location, so that
of Li and Oldenburg (1998b), which often provided interesting the recovered density is not concentrated near the observation
results (e.g., Kalateh and Kahoo, 2013; Louro and Mantovani, locations. In the next step, the data misfit (Ød) between the
2012; Oldenburg and Pratt, 2007). This algorithm starts from observed data and the predicted data is calculated, using Eq. 4:
Eq. 2:
Ød = ||wd (dmod − dobs|| 2 (4)
d = Gρ (2)
where d is the vector of real extracted in the survey, G is the where wd is a diagonal matrix in which the ith element is the
standard deviation of the ith datum, dmod the predicted density,
sensitivity matrix and ρ is the density vectors of the tri-or-
and dobs the observed data. The inversion objective is to
thogonal mesh to be created for the inversion.
minimize the difference between both the objective function
The inverse problem can be formulated as an optimization
and the data misfit:
problem where an objective function of the model is mini-
mized, subject to the constraints in Eq. 2. The objective Ø = Ød + μØm, (5)
function of the density model in Eq. 3 is minimized under in which μ is a regularization parameter that controls the
determined constraints in order to reproduce the data inside relative importance of the model norm and the data misfit.
an error tolerance. The described methodology provides a basic structure for
Øm = αS ∫ wS w2 (z) (ρ − ρ0)2 dv
solving 3D gravimetric inversion. More solutions of Grav3D
v are available in Li and Oldenburg (1998b).
1798 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800

the high density rocks are marked with the dark color and the
low density rocks are shown with the light color.
Figure 7 shows the slices of the density contrast model at
five different depths (500, 1000, 2000, 3000, and 4000 m). At
about 500 m depth (Fig. 7), calculated model shows several
small anomalies. In zones of A1 and A3, these anomalies are
very significant but in zone A2 they are not clear. At the depth
of about 1000 m (Fig. 7), in zones A1 and A2, more distinct
negative density contrasts were created, and these contrasts
also can be seen at the depth of more than 2000 m. From the
depth of around 3000 m and more, the density contrasts have
been smoother, as shown in Fig. 7.
Sedimentary rocks are usually less dense than igneous and
metamorphic rocks (see Table 1). A wide range of density of
sedimentary rocks is observed, which is primarily due to
variation in porosity (Telford et al., 1990). The nature of the
pore fluids also affects the bulk density. All types of rocks,
especially the sedimentary ones, are influenced by age,
previous history and depth of burial. In general, although some
previous histories of rocks increase density, such as compac-
tion by depth and time, some other ones decrease the density
of rocks, such as pores and faults. Under normal conditions,
the density contrast between adjacent sedimentary formations
in the field is seldom greater than 250 kg/m3, except for a
fractured zone.
In the current study, the MAf has produced a fractured zone
in limestone rocks (possible reservoir rock in the study area).
According to geological and hydrochemical studies, it is
Fig. 7. First 4000 m of the 3D density contrast model in the Mahallat geothermal
field calculated using the Li–Oldenburg (1998b) method. considered that carbonate rocks host the geothermal reservoir
of Mahallat region. These rocks are formed of limestone (with
density of 2700 kg/m3) surrounded by mostly dolomites (with
density of 2550 kg/m3). Therefore, the density contrast of less
In this approach, the 3D subsurface volume of the ground than –150 kg/m3 can be considered for reconstruction of the
is discretized using a mesh file, the parameters of which are reservoir model during 3D inversion. In other words, based
defined by the user. The mesh should be designed in on the table of rock densities, we considered the areas with
accordance with the vastness of the area and the desired density contrast of less than –150 kg/m3 as suitable places for
resolution. The size of the cells within the mesh determines containing of geothermal reservoirs in the study area. In Fig. 8,
the resolution of the model. The cell size depends on the scale the recovered model is depicted in a perspective view at cutoff
of the study area and the mean distance between the gravity lower than –150 kg/m3. Regarding this figure, the main fault
measurements. In the current inversion, a cell size of 100 me- of the area is responsible for a negative density contrast of
ters was set for the mesh, which was considered an appropriate rocks. As expected, based on the gravitational field maps, zone
compromise between the model discrimination and the model A1, which corresponds to this fault, shows an area of
resolution. considerable negative density contrast at all of the depths. This
The inversion method requires specifying the upper (maxi- area has a more continuous extension to the depths of 1000
mum) and lower (minimum) bounding values for density
contrast of each cell at the beginning of the process. Table 1
shows the main rock densities, which constitute the structure Table 1. Average densities for the main rock types in the study area (after
of the Mahallat geothermal system. In the current study area, Telford et al., 1990)
based on the rock density table, the upper and lower bounding Rock type Average density (kg/m3)
values of density contrast were set to –500 kg/m3 and
500 kg/m3, respectively. Also, other density ranges were Limestone 2700
tested, but the above-mentioned selected range was the one Dolomite 2550
that provided a model which showed the best agreement with Sandstone 2350
the available geological data. Shale 2530
Inversion of the residual gravity data using Grav3D Granite 2690
software (Li and Oldenburg, 1998b) resulted in a density
Granodiorite 2730
contrast which is shown in Figs. 8 and 9. In these Figures,
 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800 1799

Fig. 8. The 3D model of the contrast density lower than 150 kg/m3 of the first 5000 m in the Mahallat geothermal field using the Li–Oldenburg (1998b) method.

to 3000 m, which suits the anomaly. Thus, it seems that the gives an image of the evolution of the density contrast at
geothermal reservoir has higher quality at this depth. The different depths. The most significant aspect is the evolution
activity of MAf zone, together with its tributaries, has resulted in depth of the negative anomalies in A1 and A3 zones. The
in the formation of many fractures and joints in the rocks. As 3D recovered model shows a large negative density contrast
a result, an appropriate space for circulation of hydrothermal at the depths between 1000 m and around 3000 m in A1 zone,
fluids was prepared in the rocks of zone A1 at depths between which can result from high fractured rocks and the activity of
1000 and 3000 m, so indicating of an important geothermal the main fault, allowing for comfortable hydrothermal fluid
reservoir in this zone. These results are supported by geother- circulation consequently. At the same point, A3 zone is partly
mal surface manifestations in this area (A1 zone), such as hot considerable to allow a negative density contrast and as
springs and intense hydrothermally altered areas. consequence is suitable for a geothermal reservoir. The results
The other lesser noticeable area of negative density anom- of the Euler depth estimation, the 3D inversion and the
aly is zone A3, located in the western part of Khorhe village. geothermal surface manifestations in A1 zone, such as hot
This zone is as small as zone A1 (Figs. 7 and 8). Although, springs and intense hydrothermally altered rocks, and lack of
the best place for formation of geothermal reservoir in this these manifestations in A3 zone indicate high potential for
zone is at the depth between 1000 and 2000 m, where the geothermal reservoir forming in A1 zone along the main fault,
negative density contrast is considerable (see Figs. 7 and 8), at the depths between 1000 m and around 3000 m. Also, the
the lack of significant geothermal surface manifestations found results of the Euler depth estimation and the 3D recovered
in this zone indicates that this zone cannot be a suitable model of density contrast show the thickness of this identified
candidate for a possible geothermal reservoir in the region or geothermal reservoir to increase from Abgarm village toward
maybe it is a hidden or blind geothermal system. No doubts, SSW along the main faults.
more scientific data would be needed to prove this theory. This work also revealed the importance of the gravity
survey and its data interpretation to ensure better under-
Conclusion standing of the relationship between the geothermal reservoirs
and the negative density contrasts.
Mahallat geothermal region is one of the most interesting Acknowledgments. The authors wish to thank the Renew-
localities for geothermal exploration in Iran. This paper able Energy Initiative Council of Vice-Presidency of Scientific
presents the interpretation of the gravity data over this area, and Technology and Niroo Research Institute (NRI) for
around the Mahallat–Abgarm fault (MAf). Analysis of the providing gravity data for this study. The authors would also
residual gravity field shows three negative anomaly zones (A1, like to thank Hamzeh Mehrabi and Amir Hossein Enayati for
A2, and A3) which can be associated with the geothermal their useful geological comments.
reservoirs in the area. The results of the horizontal gradient
filters show the most of the observed anomaly to be associated
with faults and geological contacts. The main structure, the References
MAf zone, has a strong expression in the direction of
Abiye, T.A., Haile, T., 2008. Geophysical exploration of the Boku geothermal
NNE-SSW, and most of the geothermal surface manifestations
area, Central Ethiopian Rift. Geothermics 37 (6), 586–596.
in the region exist along this fault. It seems that the most Beitollahi, A., 1996. Travertine formation and the origin of the high natural
important negative anomaly zone in the residual gravity map radioactivity in the region of Mahallat hot springs. The M.Sc. Thesis.
(A1) is due to this fault. The 3D inversion of gravity data Islamic Azad University of Tehran, Iran.
1800 J. Nouraliee et al. / Russian Geology and Geophysics 56 (2015) 1791–1800

Bournas, N., Galdeano, A., Hamoudi, M., Baker, H., 2003. Interpretation of Oskooi, B., Darijani, M., 2013. 2D inversion of magnetotelluric data from
the aeromagnetic map of Eastern Hoggar (Algeria) using the Euler Mahallat geothermal field in Iran using finite element approach. Arabian
deconvolution, analytic signal and local wavenumber methods. J. Afr. J. Geosci., 2013, doi:10.1007/s12517-013-893-6.
Earth Sci. 37 (3), 191–205, doi:http://dx.doi.org/10.1016/j.jafrearsci.2002. Oskooi, B., Darijani, M., Mirzaie, M., 2013. Investigation of electrical
12.001. resistivity and geological structure on the hot springs in Markazi province
Giggenbach, W.F., 1991. Chemical techniques in geothermal exploration, in: of Iran using magnetotelluric method. Bollettino di Geofisica ed Applicata
D‘Amore, F. (coordinator), Applications of geochemistry in geothermal 54, 245–256.
reservoir development. UNITAR/UNDP Publication, Rome, pp. 119–142. Oskooi, B., Mirzaei, M., Mohammadi, B., Moghaddam, M.M., Ghadimi, F.
Gottsmann, J., Camacho, A.G., Martí, J.,Wooller, L., Fernández, J., García, A., Integrated interpretation of the magnetotelluric and magnetic data from
Rymer, H., 2008. Shallow structure beneath the Central Volcanic Complex Mahallat geothermal area, Iran. Studia Geophysica et Geodaetica (accepted
of Tenerife from new gravity data: implications for its evolution and recent for publication).
reactivation. Phys. Earth Planet. Inter. 168, 212–230. Pentecost, A., 2005. Travertine. Springer-Verlag, Berline.
Kalateh, A.N, Kahoo, A.R., 2013. Estimation of 3D density distribution of Porkhial, S., Nouraliee, J. Rahmani, M., Ebrahimi D., 2013 Resource
chromite deposits using gravity data. J. Mining Environ. 4, 97–104. assessment of Vartun geothermal region, central Iran. J. Tethys 1, 29–40.
Li, Y., Oldenburg, D.W., 1998a. Separation of regional and residual magnetic Reid, A.B., Allsop, J.M., Granser, H., Millet, A.J., Somerton, I.W., 1990.
field data. Geophysics 63, 431–439. Magnetic interpretation in three dimensions using Euler deconvolution.
Li, Y., Oldenburg, D.W., 1998b. 3D inversion of gravity data. Geophysics Geophysics 55, 80–91.
63, 109–119. Represas, P., Santos, F.A., Ribeiro, J., 2013. Interpretation of gravity data to
Louro, V.H.A., Mantovani, M.S.M., 2012. 3D inversion and modeling of delineate structural features connected to low-temperature geothermal
magnetic and gravimetric data characterizing the geophysical anomaly resources at Northeastern Portugal. J. Appl. Geophys. 92, 30–38.
source in Pratinha I in the southeast of Brazil. J. Appl. Geophys. 80, Rezaie, M., Ghorbani, M., Bomeri, M., 2009. The hydrogeology and
110–120. geothermology of the Mahallat hot springs, in: 1st National Conference
McKenzie, D.S., 1972. Active tectonics of the Mediterranean region. Geophys. on Hydrogeology, Behbehan, Iran, Extended Abstract, 4 pp.
J. R. Astron. Soc. 30, 109–185. Salem, A., Furuya, S., Aboud, E., Elawadi, E., Jotaki, H., Ushijima, K., 2005.
Mirzaei, M., Moghaddam, M.M., Oskooi, B., Ghadimi, F., Jazayeri, S., 2013. Subsurface Structural Mapping Using Gravity Data of Hohi Geothermal
Processing and interpretation of ground magnetic data corresponding to Area, Central Kyushu, Japan, in: Proc. World Geothermal Congress.
geothermal resources using Euler and AN-EUL method, north-east of Antalya, Turkey.
Mahallat. J. Earth Space Phys. 39, 83–96. Schiavone, D., Loddo, M., 2007. 3-D density model of Mt. Etna Volcano
Montesinos, F.G., Camacho, A.G., Nunes, J.C., Oliveira, C.S., Vieira, R., (Southern Italy). J. Volcanol. Geotherm. Res. 164, 161–175.
2003. A 3-D gravity model for a volcanic crater in Terceira Island Soengkono, S., 2011. Deep Interpretation of Gravity and Airborne Magnetic
(Azores). Geophys. J. Int. 154, 393–406. Data over the Central Taupo Volcanic Zone, in: New Zealand Geothermal
Nabighian, M.N., 1984. Toward a three-dimensional automatic interpretation Workshop Proc. (21–23 November 2011). Auckland, New Zealand.
of potential field data via generalized Hilbert transforms: fundamental Soligo, M., Tuccimei, P., Barberi, R., Delitala, M.C., Miccadei, E., Tad-
relations. Geophysics 49, 780–786. deuci, A., 2002. U/Th dating of freshwater travertine from Middle Velino
Nouraliee, J., Ebrahimi, D., 2012. Geochemical Studies in Mahallt Geothermal Valley (Central Italy): paleoclimatic and geological implications. Palaeo-
Region, Internal Report [in Persian]. Niroo Research Institute (NRI), geogr. Palaeoclim. Palaeoecol. 184, 147–161.
Tehran, Iran. Stavrev, P., Reid, A., 2007. Degrees of homogeneity of potential fields and
Nouraliee, J., Shahhosseini, A., 2012. Geological map of Mahallat geothermal structural indices of Euler deconvolution. Geophysics 72 (1), L1–L12, doi:
region, Scale: 1:25,000. Niroo Research Institute (NRI), Tehran, Iran. 10.1190/1.2400010.
Oldenburg, D.W., Pratt D.A., 2007. Geophysical inversion for mineral Telford, W.M., Geldart, L.P., Sheriff, R.E., 1990. Applied Geophysics,
exploration—A decade of progress in theory and practice, in: Milkereit, B. second ed. Cambridge University Press, Cambridge.
(Ed.), Proc. Exploration 07: Fifth Decennial Int. Conf. Mineral Exploration Thompson, D.T., 1982. EULDPH a new technique for making computer-as-
(9–12 September 2007). Toronto, Canada, pp. 61–95. sisted depth estimates from magnetic data. Geophysics. 47, 31–37.

Editorial responsibility: A.D. Duchkov