Pre-eruptive ground deformation of Azerbaijan mud volcanoes detected through satellite radar interferometry (DInSAR)

Abstract

Mud volcanism is a process that leads to the extrusion of subsurface mud, fragments of country rocks, saline waters and gases. This mechanism is typically linked to hydrocarbon traps, and the extrusion of this material builds up a variety of conical edifices with a similar morphology to those of magmatic volcanoes, though smaller in size. The Differential Interferometry Synthetic Aperture Radar (DInSAR) technique has been used to investigate the ground deformation related to the activity of the mud volcanoes of Azerbaijan. The analysis of a set of wrapped and unwrapped interferograms, selected according to their coherence, allowed the detection of significant superficial deformation related to the activity of four mud volcanoes. The ground displacement patterns observed during the period spanning from October 2003 to November 2005 are dominated by uplift, which reach a cumulative value of up to 20 and 10 cm at the Ayaz-Akhtarma and Khara Zira Island mud volcanoes, respectively. However, some sectors of the mud volcano edifices are affected by subsidence, which might correspond to deflation zones that coexist with the inflation zones characterized by the dominant uplift. Important deformation events, caused by fluid pressure and volume variations, have been observed both (1) in connection with main eruptive events in the form of pre-eruptive uplift, and (2) in the form of short-lived deformation pulses that interrupt a period of quiescence. Both deformation patterns show important similarities to those identified in some magmatic systems. The pre-eruptive uplift has been observed in many magmatic volcanoes as a consequence of magma intrusion or hydrothermal fluid injection. Moreover, discrete short-duration pulses of deformation are also experienced by magmatic volcanoes and are repeated over time as multiple inflation and deflation events.

Full text

Pre-eruptive ground deformation of Azerbaijan mud volcanoes detected
through satellite radar interferometry (DInSAR)
Benedetta Antonielli
a,b,
⁎, Oriol Monserrat
c
,MarcoBonini
d
,GaiaRighini
e
,FedericoSani
a
, Guido Luzi
c
,
Akper A. Feyzullayev
f
, Chingiz S. Aliyev
f
a
Dipartimento di Scienze della Terra, Università di Firenze, via G. La Pira n.4, 50121, Firenze, Italy
b
Tuscan EarthScience PhD Program, Earth Science Department, University of Pisa, Via S. Maria 53, 56126 Pisa, Italy
c
Centre Tecnològic de Telecomunicacions de Catalunya (CTTC), Av. Carl Friedrich Gauss, 7, Castelldefels, Spain
d
CNR, Consiglio Nazionale delle Ricerche, Istituto di Geoscienze e Georisorse, via G. La Pira 4, I-50121 Firenze, Italy
e
ENEA, Italian National Agency for NewTechnologies, Energy and Sustainable Economic Development, via Martiri di Monte Sole, 4, 40129, Bologna, Italy
f
Geology Institute of the Azerbaijan National Academy of Sciences, H. Javid pr., 29 A, Baku AZ1143, Azerbaijan
abstractarticle info
Article history:
Received 14 April 2014
Received in revised form 19 September 2014
Accepted 3 October 2014
Available online 15 October 2014
Keywords:
DInSAR
Mud volcanoes
Volcano deformation
Pre-eruptive uplift
Deformation pulses
Azerbaijan
Mud volcanism is a process that leads to the extrusion of subsurface mud, fragments of country rocks, saline waters
and gases. This mechanism is typically linked to hydrocarbon traps, and the extrusion of this material builds up a
variety of conical edifices with a similar morphology to those of magmatic volcanoes, though smaller in size. The
Differential Interferometry Synthetic Aperture Radar (DInSAR) technique has been used to investigate the ground
deformation related to the activity of the mud volcanoes of Azerbaijan. The analysis of a set of wrapped and
unwrapped interferograms, selected according to their coherence, allowed the detection of significant
superficial deformation related to the activity of four mud volcanoes. The ground displacement patterns
observed during the period spanning from October 2003 to November 2005 are dominated by uplift,
which reach a cumulative value of up to 20 and 10 cm at the Ayaz–Akhtarmaand Khara-Zira Island mud volcanoes,
respectively. However, some sectors of the mud volcano edifices are affected by subsidence, which might
correspond to deflationzones that coexist with the inflation zones characterized by the dominant uplift. Important
deformation events, caused by fluid pressure and volume variations, have been observed both (1) in connection
with maineruptive events in the form of pre-eruptive uplift, and (2) in the form of short-lived deformation pulses
that interrupt a period of quiescence. Both deformation patterns show important similarities to those identified in
some magmatic systems. The pre-eruptive uplift has been observed in many magmatic volcanoes as a
consequence of magma intrusion or hydrothermal fluid injection. Moreover, discrete short-duration pulses of
deformation are also experienced by magmatic volcanoes and are repeated over time as multiple inflation and
deflation events.
© 2014 Elsevier B.V. All rights reserved.
1. Introduction and aims of the work
Mud volcanism is a well-known phenomenon that drives the
extrusion of fluids and solid material originated from deeply buried
sediments, such as saline waters, gases (mostly methane), mud and
clasts of country rock. This mechanism is typically linked to hydrocarbon
traps (Higgins and Saunders, 1974) and builds up a variety of features.
Conical extrusive edifices are the most typical feature and may vary in
size from centimeters to a few hundred meters in height and several
kilometers long. Mud volcanoes are generally smaller than magmatic
volcanoes, although in some cases their size is comparable to that of
the large magm atic edifices. In particular, in the Marianas subduction
zone, mud volcanoes may be 2 km high and have a diameter of 25 km
(Fryer et al, 1999).
Mud volcanoes are usually located in fold-and-thrust belts and
accretionary prisms, and develop at convergent plate margins where
the active tectonic shortening affects the sediments by increasing
stresses and temperatures leading to the maturation of organic matter
(Brown, 1990; Higgins and Saunders, 1974; Kopf, 2002). During most
of their lifetime, mud volcanoes show a background activity of quiet to
vigorous expulsion of fluids and mud breccias. However, such quiescent
activity may be occasionally interrupted by paroxysmal events which
violently release large mud flows and flaming eruptions caused by the
self-ignition of the methane.
The eastern Greater Caucasus in Azerbaijan hosts the highest density
of mud volcanoes in the world (Jakubov et al., 1971; Guliyev and
Feizullayev, 1997;Figs. 1 and 3a). Some mud volcanoes may be up
to 400 m tall, and 4–5 km long, with a dimension and morphological
Tectonophysics 637 (2014) 163–177
⁎Corresponding author.
E-mail address: benedetta.antonielli@unifi.it (B. Antonielli).
http://dx.doi.org/10.1016/j.tecto.2014.10.005
0040-1951/© 2014 Elsevier B.V. All rights reserved.
Contents lists available at ScienceDirect
Tectonophysics
journal homepage: www.elsevier.com/locate/tecto

characteristics similar to those of magmatic volcanoes.There is a rather
complete record of mud volcanoes eruptions since 1810 (Mellors et al.,
2007). In particular, the second edition of the catalogue provides the
chronology and a brief description of 387 eruptions occurred between
1810 and 2007 at 93 mud volcanoes located, both onshore and offshore,
in Azerbaijan (Aliyev et al., 2009).
Satellite based Synthetic Aperture Radar Interferometry (InSAR)
techniques have been commonly used to monitor and investigate the
ground deformation connected to the eruptive phases of magmatic
volcanoes. In the early 90s, the launch of satellites carrying SAR sensors
onboard purposely built for interferometric applications, opened new
opportunities for mapping and monitoring ground deformations
produced by different processes at active magmatic volcanoes,
particularly: (i) uplift before eruption and co-eruptive subsidence
(Amelung et al., 2000; Jay et al., 2014), (ii) pulses of inflation and
deflation on volcanic systems (Biggs et al., 2009, 2011; Brunori et al.,
2013; Chang et al., 2007; Pagli et al., 2006; Pritchard and Simons, 2004),
(iii) subsidence associated with magma phase change (Caricchi et al.,
2014), (iv) linked hydrothermal and magmatic systems (Wicks et al.,
1998), (v) crustal rifting episodes (Pagli et al., 2012; Sigmundsson et al.,
1997; Wright et al., 2012), (vi) downslope movements on the volcano
flanks (Ebmeier et al., 2010), and (vi) lava thicknesses and extrusion
rate (Ebmeier et al., 2012). InSAR techniques have also been employed
to explore the ground deformation associated with the LUSI mud volcano
in Indonesia (Abidin et al., 2009; Aoki and Sidiq, 2014; Fukushima et al.,
2009; Rudolph et al., 2013).
Using similar techniques, we have carried out research on ground
deformation related to the activity of the mud volcanoes of Azerbaijan.
For this purpose, the deformation of mud volcanic systems has been
analyzed using the differential interferometry (DInSAR) technique
applied to ENVISAT data. A few studies have been carried out using
the InSAR technique to analyze the mud volcanoes of Azerbaijan.
Hommels et al. (2003) investigated the dynamics of the largest mud
volcanoes (Touragai, Great and Lesser Kjanizadag) and found prelimi-
nary indications of deformation in the dataset analyzed. Mellors et al.
(2005) focused on the analysis of the Absheron Peninsula and the
Lokbatan mud volcano, but they did not observe any large-scale
movement (N10 cm line-of-sight) during the analyzed period. In
the following sections we describe the concepts and tools that drive
the assessment of mud volcanism and the geological framework of
the area (Section 2), the key steps of the DInSAR data processing
(Section 3), and the results for each case study (Section 4). Finally, the
principal strengths and weaknesses are discussed together with the
follow-up and recommendations for future work (Sections 5 and 6).
2. Mud volcanism
2.1. Mud volcano processes and terminology
Several theories have been proposed in the literature regarding the
mechanisms that might control the development ofmud volcanism. In par-
ticular, this process has often been related to subsurface intrusive mud or
shale diapirism (Fig. 2a; Brown, 1990; Morley and Guerin, 1996; Kopf,
2002). Mud diapirs are subsurface fluid-rich overpressured muddy masses
that are driven upward in response to their buoyancy resulting from the
bulk density contrast with respect to the denser surrounding overburden
(Brown, 1990). The overpressure is produced by the organogenic activity
and the subsequent gas production at depth (e.g., Higgins and Saunders,
1974). The expansion and degassing of the methane dissolved in the mud
may further increase both the overpressure and buoyancy of the rising dia-
pir (Brown, 1990). Other models suggest that mud volcanism is sourced
from mud–water–gas mixes rising up through intricate systems of con-
duits and pipes and networks of anastomosing fault-controlled planar
pathways exploiting deeper fluid-rich source layers (Cooper, 2001;
Davies et al., 2007; Dimitrov, 2002; Fowler et al., 2000; Mazzini et al.,
2009; Morley, 2003; Planke et al., 2003; Roberts et al., 2010). Fluid res-
ervoirs may occur at various depths and are likely to contain zones with
dense networks of mud-filled fractures.
Mud volcanoes are thus closely associated with petroleum systems,
and the development of overpressures in source rocks is a necessary
condition to trigger mud volcanism (Dimitrov, 2002). For this reason,
mud volcanoes are often located at anticlines where sealing layers in
the fold core may efficiently trap the rising hydrocarbon fluids and readily
built-up overpressures (Bonini, 2012). Tectonic stress provides an impor-
tant source of overpressure, as indicated by the widespread occurrence
of mud volcanoes in many active compressional belts worldwide
(e.g., Kopf, 2002). The state of stress is also inferred to control distribu-
tion, geometry and shape of mud volcano features (Bonini, 2012).
‘Mud volcano’is a generic term that indicates the various morphologic
features associated with the extrusion of subsurface material. As soon as
it reaches the topographic surface, the extruded mud gives rise to the
typical conical edifices. Conventional nomenclature subdivides the small
sub-conical extrusive edifices in gryphons (≤3 m high, Fig. 2b) and
mud cones (N3 m high, Fig. 2c), while the term mud volcano should be
restricted to the edifices reaching some tens of meters in height (say
≥50 m) or a few kilometers across (Fig. 2d). The largest mud volcanoes
are the gigantic submarine serpentinite mud volcanoes (conical
seamounts) of the Marianas forearc, which may reach 25 km in diameter
and exceed 2 km in height. Mud volcanoes and magmatic volcanoes
display very similar morphologic features and, for this reason, many
terms used for mud volcanism are borrowed from the terminology of
magmatic features. For instance ‘crater’is used to indicate the sub-
Fig. 1. (a) Simplified structural sketch map of the Greater Caucasus–eastern Caspian Basin
(modified from Jackson et al., 2002). (b) Regional cross-section through the Absheron Sill
and the South Caspian Basin (vertical exaggeration 2×; simplified from Stewart and Davies,
2006).
164 B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

circular collapsed areas topping the extrusive edifices, ‘caldera’indicates
the depressions formed as a consequence of the removal of subsurface
material, and ‘vent’refers to openings through which fluids are released.
Other manifestations are mud pools (or ‘salses’) characterized by
bubbling gas centers.
2.2. Geologic setting of Azerbaijan mud volcanoes
Mud volcanoes of Azerbaijan are typically associated with hydrocarbon
traps in thrust anticlines (Jakubov et al., 1971; Guliyev and Feizullayev,
1997; Reynolds et al., 1998; Fowler et al., 2000; Cooper, 2001; Planke et al.,
2003; Stewart and Davies, 2006; Bonini, 2012;Fig. 2a). In particular, mud
volcanoes typically pierce the crest of the fold anticlines occurring in on-
shore —Eastern Kura basin (Low Kura depression) and southeast ending
of the Great Caucasus (Gobustan area, Absheron Peninsula) and adjoin
offshore —Baku and Absheron archipelagoes (Guliyev and Feizullayev,
1997; Jakubov et al., 1971). In the South Caspian Sea, mud volcanoes ac-
tivity is documented since 3.5 Ma (Yusifov and Rabinowitz, 2004), there-
by in close connection with the development of onshore and offshore
folds that probably began to form during Early–Late Pliocene times
(e.g., Devlin et al., 1999; Yusifov and Rabinowitz, 2004).
These folds are generally considered to be active structures resulting
from the folding of both the fluvial–deltaic sands of the latest Miocene–
early Pliocene ‘Productive Series’and the middle–late Miocene Diatom
Suite. The folds are detached in the underlying Oligocene–Miocene
overpressured shales of the ‘Maykop Series’(Devlin et al., 1999; Jackson
et al., 2002; Soto et al., 2011).TheMaykopSeriesisa200–1200 m-thick
(up to 3000 m offshore) regionally continuous package of fine-grained
clastic sediments with high content of organic material that provides
the main regional source rock for hydrocarbons and for the mud volcano
systems (Abrams and Narimanov, 1997; Hudson et al., 2008; Inan et al.,
1997). The overlying marls and shales of the Diatom Suite (Fig. 2a)
represent another important source layer of hydrocarbon. The
Productive Series affords the majority of hydrocarbon reservoirs
and consist of 5–7kmthick,fluvial–deltaic sand bodies, interbedded
with mudstones deposited during short-lived lacustrine events
(Allen et al., 2002; Fowler et al., 2000; Guliyev and Feizullayev,
1997; Reynolds et al., 1998; Soto et al., 2011; Stewart and Davies,
2006). Sedimentation of the Productive Series was followed by the
deposition of the late Pliocene marine mudstones of the Akchagyl Series,
and afterward by the Absheron Series and younger units (Fig. 2a), which
were deposited in non-marine, commonly brackish conditions.
3. The DInSAR data and processing
The dynamic analysis of surfacedeformation at volcanoes is usually
carried out by means of observations obtained during its seismic or
eruptive activities. In the case of the mud volcanoes of Azerbaijan,
detailed information exists only for specific cases. In order to widen
this information,we have applied the satellite based DInSAR technique.
Fig. 2. (a) Conceptual setting of Azerbaijan mud volcanism, which typically localizes at tight buckle folds detached from the overpressured Maykop shales (slightly m odified from Bonini et al.,
2013). Hydrocarbon gases, oil (mostly produced in the organic-rich Maykop Series), saline waters and mud are inferred to migrate into the fold core and collect in reservoirs within the
Productive Series. This mud–fluid mix may eventually ascend toward the topographic surface along structurally-controlled pathways (i.e., outer-arc normal faults, faults and joints associated
with the folding, etc.), giving rise toextrusive edifices constructed by mud breccias. Examples of onshore mud volcano edifices in eastern Azerbaijan: gryphon s and mud cones at the (b) Kichik
Maraza (June 2013) and (c) Pirekyushkul mud volcano fields (June 2013). (d) Lateral view (looking west) of the ~400 m-tall Kyanizadag mud volcano (June 2013).
165B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

This technique has allowed us to look back in time and to document the
deformation processes of four volcanoes during the period 2003 to
2005.
3.1. The DInSAR data analysis
DInSAR is a radar based technique that exploits the information
contained in the phase of at least two SAR images acquired over
the same area, at different times, to infer terrain deformation measure-
ments (Hanssen, 2001). The process of measuring deformation with
DInSAR is not always straightforward because the deformation contri-
bution, φ
Mov
, is not the only factor affecting the phase: (i) the effect of
the different positions of the sensor at each acquisition time named
topographic component, φ
Topo
, which can be removed during the
differential interferogram calculation by using a digital elevation
model (DEM) of the area; (ii) the effect of atmospheric propagation,
φ
Atm
, and; (iii) finally a noise term, φ
Noise
, that is the sum of the instru-
mental noise and the contribution of the physical changes of each
measured object which can affect radar response. Moreover, the DInSAR
phase ambiguity, given that the phase is modulus 2π, has to be solved
(phase wrapping).
Hence, measuring deformations with DInSAR, i.e. to obtain φ
Mov
,
requires solving the following equation:
ΔφInt ¼φTopo þφMov þφAtm þφNoise

þ2kπ:ð1Þ
Different approaches to solve quantitatively Eq. (1) are described
in the literature (Berardino et al., 2002; Biescas et al., 2007; Biggs et al.,
2007; Ferretti et al., 2001; Hooper et al., 2004). However, these
approaches were discarded for this work because all of them are
based in processing large data stacks (a minimum of approximately
15 images are required). In this work, a simpler approach is based on
the simultaneous analysis of small sets of interferograms where
the contribution of the non-deformation components of the phase is
assumed to be negligible. The first step of the applied procedure is to
fully screen each interferogram in order to detect phase spatial variation
located in mud volcano areas. Once a phase spatial variation is detected,
the significance of the contribution of each one of the non-deformation
components is evaluated by using a pairwise logic criterion (Massonnet
and Feigl, 1998)asbriefly described here:
–φ
Topo
: This component is linearly related to the perpendicular baseline
oftheinterferogram.Hence,givenaspecific area, if the phase variation
has opposite gradient with opposite perpendicular baselines, then the
observed pattern is mainly due to φ
Topo
:otherwise,theφ
Topo
can be
neglected. At this point, it is worth emphasizing that this analysis is
adequate for two reasons: first, the contribution of the φ
Topo
is
expected to be negligible since the studied areas are mostly flat,
and secondly, the perpendicular baselines of the interferogram
network are mostly small (b216 m).
–φ
Atm
: Let's assume that we observe a phase variation in an inter-
ferogram Δφ
21
=φ
2
−φ
1
and that the phase variation is mainly
due to the atmospheric contribution of φ
1
.Then,thesamephase
variation should appear in an interferogram Δφ
1k
=φ
1
−φ
k
but
with opposite sign. Therefore, analyzing different combinations of
φ
1
and φ
2
makes possible the discrimination of a significant φ
Atm
contribution.
–φ
Noise
: This contribution is evaluated for each interferogram by
means of the coherence. The coherence threshold has been chosen
iteratively for each interferogram and eachmud volcano separately.
The final coherence threshold is based on a trade-off between
the level of noise and the spatial point density, in order to obtain a
reliable phase unwrapping. It is worth noting that the reliability of
the unwrapped phase is qualitatively evaluated comparing the
unwrapped interferogram with its corresponding wrapped version.
In particular, for the Ayaz–Akhtarma mud volcano the selected
coherence threshold is around 0.1 for all the used interferograms
(see caption of Fig. 5). Regarding the other mud volcanoes, the
selected coherence threshold is around 0.3, 0.2 and 0.25, for the
Khara-Zira, the Akhtarma–Pashaly, and the Byandovan, respectively
(see captions of Figs. 8, 10 and 11).
Phase unwrapping is applied to the interferograms that show
negligible non-deformation components. In this step, the absolute
cycles of the phase are estimated, which leads to a correct interpretation
of the measured deformation. The phase unwrapping is an ill posed
problem. The success of this step strongly depends on both the
magnitude of the deformation and the density of coherent points.
For this reason, only those interferograms with a trade-off between
coherence and spatial density have been used. The analysis of interfero-
grams led us to select a maximum temporal baseline of 105 days. Longer
temporal baselines show poor correlation in most of the cases. The phase
unwrapping have been performed using the minimum cost flow
algorithm described in Costantini (1998).
The last step consisted of interpolating the unwrapped interferograms
to obtain a continuous deformation field. The interpolation method used
was a distance weighted bilinear interpolation. It is worth noting that the
interpolated points are sparse and, in the worst case (interferogram 11),
correspond to 15% of the total number of points. Hence, there are not
expected interpolation artifacts.
To conclude, it is important to underline that the final deformation
maps are represented in the Line-of-Sight (LOS) direction, i.e. the
measured deformation at one point is the projection of the actual
deformation along the satellite's LOS.
3.2. The used SAR dataset
A dataset of 9 ENVISAT descending images (track 235–frame 2799;
Fig. 3c) spanning from October 2003 to November 2005 have been
processed in this study. Radar images have been provided by the
European Space Agency(ESA) in the frameworkof the ESA Cat-1 Project
13866 titled: “Assessing the relationships between tectonics and mud
volcanism by integrating DInSAR analysis and seismic data in active
tectonic areas”.
An area of 100 km
2
, in which the majority of the Azerbaijan mud
volcanoes (about 300) are contained, was surveyed (Fig. 3a). All the
available interferometric pairs were processed by using software
developed at the Centre Tecnològic de Telecomunicacions de Catalunya
(CTTC). The Shuttle Radar Topography Mission (SRTM) Digital Terrain
Model (DTM), with a spatial resolution of 90 m, was used to estimate
the differential interferograms. Finally, a subset of 8 interferograms
which showed a good phase unwrappingand covered the entire period
of observations was selected for the analysis. Table 1 shows the main
characteristics of the selected interferograms. It is worth noting that,
with the exception of interferogram no. 19, all the interferograms
have temporal baselines shorter than 105 days, assuring working with
interferograms with high coherence, which ease the phase unwrapping
process. At the same time, the analyzed dataset was heterogeneous
enough to provide a robust discrimination of both atmosphere and
topographic errors.
4. Results
DInSAR observations have allowed us to detect significant
deformation at four volcanic edifices in the Absheron Peninsula, in
the Baku Archipelago and in the Gobustan region, namely the Ayaz–
Akhtarma, Khara-Zira Island, Akhtarma–Pashaly and Byandovan (or
Bandovan) mud volcanoes (Fig. 3a, b). After a general screening of the
used ENVISAT frame, we focused our study on these four mud volcanoes
because the displacement signals are the most evident and the deforma-
tion at these mud volcanoes clearly stands out.
166 B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

As no dedicated monitoring networks exist, the mud volcanic
activity in the region remains poorly documented, and the associ-
ated hazard cannot be forecasted. A geological–structural field
survey was carried out in selected areas of Azerbaijan during
June 2013 to collect morphological and structural data on the
volcanoes according to the preliminary results obtained from the
Fig. 3. (a) Mainmud volcano fields aroundthe Greater Caucasusfront and Absheron area(adapted from Jakubovet al., 1971; Jackson et al.,2002; Bonini and Mazzarini, 2010;Bonini et al.,
2013); thesefeatures are superposed on a shaded digital terrainmodel (U.S. Geological Survey, available from:http://www.gisweb.ciat.cgiar.org/sig/90m_data_tropics.htm). (b) Location
of the four studied mud volcanoes. White boxes indicate the location of wrapped interferograms at Ayaz–Akhtarma (1 September 2005/10 November 2005), Khara-Zira Island (14 April
2005/28 July2005), Akhtarma–Pashaly (25 November 2004/30 December2004), and Byandovan (25 November 2004/30 December 2004).Each fringe represents 2.8 cm of motion in the
satellite Line-of-Sight (LOS). (c) Location of the area (red box) covered by the ENVISAT descending image (track 235).
167B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

DInSAR analysis. The results for each mud volcano are described
below.
4.1. Ayaz–Akhtarma mud volcano
The Ayaz–Akhtarma mud volcano is a large edifice with elliptical
base (major axis about 2700 m) characterized by a wide flat top surface
that might be a filled caldera (Fig. 4a–b). Current fluid expulsion occurs
at several gryphons and cones clustering in a circular area (about 430 m
in diameter) at the center of the crater (Fig. 4a–b). The major eruptions
in the last 20 years occurred in 2001, 2005, 2006 and 2007, even though
the precise date of these eruptive events is unknown (Aliyev et al.,
2009).
Significant surface ground displacement within the mud volcano
edifice is registered in the interferograms from October 2003 and
November 2005 (Fig. 5), in connection with the strong eruptive activity
of 2005, 2006 and 2007. Specifically, the wrapped interferograms
(column on the left of Fig. 5) show a very complex pattern of concentric
Table 1
Interferograms selected according to theircoherence,spanning from October 2003 to November 2005.
Interferogram no. Date of the master Date of the slave Perpendicular baseline (m) Temporal baseline (days)
11 02/10/2003 15/01/2004 −213 105
19 15/01/2004 25/11/2004 62 315
27 25/11/2004 30/12/2004 −212 35
1 30/12/2004 10/03/2005 −30 70
40 10/03/2005 14/04/2005 118 35
48 14/04/2005 28/07/2005 −160 105
51 28/07/2005 01/09/2005 −319 35
8 01/09/2005 10/11/2005 106 70
Fig. 4. Ayaz–Akhtarma mud volcano. (a) Panoramic view looking south. (b) Google Earth image (March 2004). (c) Fracture observed during the field survey carried out in June 2013.
(d) Contour plot (Fisher distribution;equal area, lower hemisphere) of poles to fracture and fault segments. (e–f) Fault segments showing dominant vertical displacement (maximum
vertical throw ~1 m). (g) En-echelon fractures pointing to some dextral strike–slip component of displacement. Images in (b) and (c) are extracted from Google Earth®; http://earth.
google.it/download-earth.html.
168 B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

fringes in the eastern part of the mud volcano. The unwrapped interfer-
ograms (central column of Fig. 5) indicate that the deformation in the
mud volcano shows a similar trend in all the analyzed interferograms.
Although at different rates, the western part of mud volcano has
moved away from the satellite and the eastern sector toward the satellite.
This implies that, assuming a purely vertical deformation, the eastern part
of the volcanic edifices has lifted and the western sector has subsided.
This deformation pattern is shown in Fig. 6, where the cumulative LOS
displacements for the two orthogonal cross-sections (displayed at the
bottom-right corner of Fig. 5) are represented. The figures of the
right column of Fig. 5 represent the sum of the contributions of the
interferograms. In particular, the measured ground uplift (LOS
displacement) reached up to 20 cm in the period October 2003–
November 2005. On the hypothesis that deformation has not stopped,
the final cumulative ground uplift is probably underestimated because
it has not been possible to analyze the interferograms with low coher-
ence values, specifically January–November 2004 and April–July 2005
interferograms (white panels in Fig. 5). The uplift has indeed increased
in rate (up to 6 cmin 70 days) between July and November 2005 and it
is located in a specific semicircular zone showing a larger diameter in
comparison to the previous interferograms (Fig. 5). This observation
suggests that the area of highest uplift probably corresponds to the
center of inflation. The western area of the mud volcano is affected by
a smaller subsidence which decreases in the last interferograms.
Volumetric deformation rates of the order of those measured in this
case mightbe associated with some surfacebrittle faulting/fracturing. A
geological–structural field survey was carried out in June 2013 in order
to investigate this possibility. Thefield survey allowed the identification
of a main fault/fracture 600 m long and in a direction trending about
N42°E (Fig. 4c–d). The north-eastern part of this brittle element is
Fig. 5. Wrapped(figures on the leftcolumn), unwrapped(figures on the centralcolumn) and cumulative (figureson the right column) interferogramsof the Ayaz–Akhtarmamud volcano.
Unwrappedand relative cumulative interferogramswith low coherence are notreported (white figures). The meancoherence thresholdfor all the used interferograms is 0.1 ± 0.05. The
detectedLine-of-Sight (LOS) ground displacementconsists of a maximummotion toward the satellite (i.e. uplift,blue colors and negative sign in thechromatic scale, asthe sensor-target
distancedecreases) in the eastern part of the mudvolcano, and a maximumdisplacement awayfrom the satellite (i.e.subsidence, red color and positivesign in the chromaticscales, as the
sensor-target distance increases) in the western part of the volcano. The bottom right figure reports the traces of the cross-sections of Fig. 6.
169B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

characterized by a normal vertical throw varying between 25 cm and
1m(Fig. 4e–f). The remaining south-western tract consists of en-
echelon fractures, which suggest a right strike–slip component of
displacement (Fig. 4g). GPS reference points were collected and
used to locate the fault/fracture on the geocoded radar digital data
and on Google Earth images. This fault/fracture surveyed in 2013
canalsobeobservedinGoogleEarthimagesdatingbackto2004
(Fig. 4c), thereby suggesting that this brittle system has been active
at least since 2004.
4.2. Khara-Zira Island
The Khara-Zira Island lies south-southwest of the Baku Archipelago
and is one of the islands generated by mud volcanoes in the Caspian
Sea (Mellors et al., 2007; Aliyev et al., 2009;Fig. 3a, b). The Khara-Zira
Island is elliptical in shape (with a maximum length of ~2.4 km) and
is topped by a rather flat surface (Fig. 7a). According to the catalogue
of mud volcano eruptions, a major paroxysmal phase occurred on the
20th of November 2006 (Aliyev et al., 2009). The catalogue describes
this event as an explosion with inflammation of gas with a 50 to
200 m high fire column. A large amount of mud breccias was outburst
and the altitude of the island increased significantly. The eruption
zone can be easily identified on satellite images slightly after the
eruption date (Fig. 7a). Unfortunately, the time span covered by
the interferograms showing good coherence (October 2003 to
November 2005) can only be used to observe the initial stages of
pre-eruptive deformation.
The wrapped interferograms show concentric, elliptical fringes
located on the most active part of the mud volcano. All the interferograms
show a relative ground uplift affecting most the mud volcano surface
(Fig. 8,figures on the left and central columns). The cumulative ground
uplift exceeds 10 cm in the two years analyzed (Fig. 9). At the
north-western part of the island, the displacement displays opposite
sign in two interferogram pairs, namely nos. 19–27, and nos. 51–08
(Fig. 8). The two interferogram pairs share a common master or
slave date, which suggests that the change in the sign of the displacement
might be the result of peculiar atmospheric conditions on two dates (25
November 2004 and 01 September 2005 for the first and second pair,
respectively). Such rapid signal changes are likely caused by local
atmospheric disturbances, and thus, may not represent actual ground
displacement.
4.3. Akhtarma–Pashaly mud volcano
The Akhtarma–Pashaly is one of the largest mud volcanoes in
Azerbaijan and is located over the crest of one of the thrust anticlines
bordering the active south-eastern margin of the Great Caucasus. It
has a roughly elliptical shape and is topped by a wide flat area with
a major axis about 2.2 km long (Fig. 7b). The Akhtarma–Pashaly mud
volcano had an intense period of activity in the eighties, with two
eruptionsin 1982 and one in 1986. Since then, we have noinformation
of any relevant volcanic activity until 01 April 2013, when the volcano
expelled a massive amount of mud and breccias that covered the entire
perimeter of the edifice. During the field survey carried out in June
2013, a few months after the eruption, the surface of the volcano
appeared as a large expanse of mud breccias, exceeding 1 m in
thickness in some areas. The mud breccias covered part of the pre-
existing morphological features, including the mud cones and gryphons
that erected from the top of the mud volcano (Fig. 7c).
The Akhtarma–Pashaly mud volcano appears to undergo a period of
quiescence in the time span covered by the analyzed interferograms
that was interrupted by an isolated intense deformation event between
November and December 2004.This deformation event was anticipated
by a phase of slightuplift during the previous 10 months. Unfortunately,
the interferograms covering the period before the deformation event
show low coherence. The strong deformation event (interferogram no.
27, Fig. 10) is revealed by fringes with complex patterns. A wide area
in the central part of the volcano was affected by a strong subsidence
(6 cm of LOS displacement in about one month), while a portion at its
southern border experienced uplift (4.5 cm of LOS displacement in
about one month). The deformation reduces drastically during the
next 70 days until it vanishes (interferogram no. 01; Fig. 10). On
the basis of the available eruption catalogue (Aliyev et al., 2009), the
observed deformation would only represent a transitory pulse during
the quiescence phase thatlasted about 27 years, and it does not coincide
with an eruption (which is indeed not found on the catalogue list).
4.4. Byandovan mud volcano
The Byandovan mud volcano is an extrusive edifice with a diameter
of approximately 1.5km and 55 m high relative to the adjacent Caspian
Sea level (altitude of −29 m). The so-called “Byandovan Mountain”
Fig. 6. Observed cumulative Line-of-Sight (LOS) displacements along cross-sections of the Ayaz–Akhtarma mud volcano. The traces A–A′and B–B′are indicated in Fig. 5 (bottom right figure).
170 B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

(Fig. 7c, d) is the most famous of the three mud volcanoes present in
the Shirvan National Park, a natural reserve established in July 2003.
This mud volcano lies on the NW–SE-trending Kelameddin–Byandovan
anticline (Zeinalov, 2000). The catalogue of recorded mud volcano
eruptions (Aliyev et al., 2009) reports two important eruptions that
occurred in 1932 and 1989. Subsequently, the volcano went through a
phase of quiescence.
Satellite radar imagery spanning from October 2003 to November
2005 shows that the volcano edifice has deformed, although no
significant eruptive events are reported in the catalogue of mud volcano
eruptions. The analyzed interferograms show a complex sequence of
feeble episodes of ground displacement. In particular, the interferograms
reveal a series of short-term pulses that alternate ground deformations
with opposite sign (i.e., uplift and subsidence), localized at the main
extrusive edifice. It is not possible to attribute the change of sign of the
deformation to atmospheric conditions on a particular date, as the
anomaly is found in many interferograms that do not share the master
or slave images. In particular, a main ground displacement event was
identified in an interferogram between November and December 2004
(interferogram no. 27, Fig. 11). The deformation consisted of relative
uplift of almost 3 cm in 35 days at the mud volcano edifice, and
subsidence of almost 2 cm in 35 days along a localized narrow, linear
area located approximately along the coastline (Fig. 11).
5. Discussion
The interferograms allow the detection of ground displacement
related to mud volcano activity for two distinct cases, particularly
(1) during the time span encompassing pre-eruptive phases (Ayaz–
Akhtarma and Khara-Zira Island mud volcanoes), and (2) short-lived
pulses interrupting a long period of quiescence (Akhtarma–Pashaly
and Byandovan mud volcanoes).
Regarding the pre-eruptive activity, the analysis of the interferograms
has allowed observing the deformation phases connected to the 2005
eruptive event of the Ayaz–Akhtarma mud volcano, as well as the activity
up to one year before the eruption of the Khara-Zira Island on 26
November 2006. In both cases, the deformation originates showing
a relative uplift of the main active zone of the mud volcano. It has
long been recognized that mud volcano activity is driven by internal
fluid pressure changes (e.g., Brown, 1990; Davies et al., 2007;
Fig. 7. Morphologicalcharacteristicsof the analyzed mud volcanoes. (a) Optical satellite image of the Khara-Zira Island (March 2007). (b) Opticalsatellite image ofthe Akhtarma–Pashaly
mud volcano (March 2004). (c) Lateral view of the top surface of the Akhtarma–Pashaly mud volcano with gryphons and mud cones (May 2010). (d) Plan- and (e) oblique view of the
Byandovan mud volcano (September 2009). Images are extracted from Google Earth®; http://earth.google.it/download-earth.html.
171B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

Fig. 8. Wrapped (figures on the leftcolumn), unwrapped(figures on the centralcolumn) and cumulative (figureson the right column) interferograms of the Khara-Zira mud volcano. The
mean coherence threshold for all the used interferograms is 0.3 ± 0.05. The Line-of-Sight (LOS) ground displacement infers a maximum motion toward the satellite (i.e. uplift) in the
south-eastern part of the mud volcano (for explanation see Fig. 5 caption). The bottom right figure reports the traces of the cross-sections of Fig. 9.
172 B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

Dimitrov, 2002; Mellors et al., 2007) and thus, the deformation of
both the Ayaz–Akhtarma and the Khara-Zira Island mud volcanoes
is likely linked to an increase of internal pressure, which may also
induce the circulation of pressurized fluids into the system.
The deformation at the Ayaz–Akhtarma mud volcano is characterized
by two adjacent zones of local subsidence and uplift. The uplift phenom-
enon predominates and is located to the east of the area in which
diffuse seepage and discharge of fluids and solid material occur. Uplift
continuously grows in intensity, especially in the last two interferograms
(Fig. 5). Unfortunately, the exact date of the eruption is unknown, and
thus, we cannot directly relate our results to the eruptive event, which
has been dated approximately 2005 by Aliyev et al (2009). In other
cases, the eruption is generally accompanied by a clear signal of subsi-
dence due to the discharge of material and release of gas pressure, as
occur at the LUSI mud volcano (e.g., Abidin et al, 2009; Aoki and Sidiq,
2014; Fukushima et al., 2009; Rudolph et al., 2013) and at magmatic
volcanoes (e.g., Amelung et al., 2000; Hreinsdóttir et al., 2014; Jay
et al., 2014; Sigmundsson et al., 2010). By contrast, a nearly continuous
Fig. 9. Observed cumulative Line-of-Sight (LOS)displacements along cross-sections of the Khara-Zira mud volcano. The traces A–A′and B–B′are indicated in Fig. 8 (bottom right panel).
Fig. 10.Wr apped (figureson the left column)and unwrapped (figureson the central column) interferogramsof the Akhtarma–Pashaly mud volcano. The meancoherence thresholdfor all
the used interferograms is 0.2 ± 0.05. Displacement away from the satellite (i.e. subsidence)is detected in the central–western part of themud volcano. An area that moves toward the
satellite(i.e. uplift) is perceptible in the southern part ofthe mud volcano (see Fig. 5 caption for explanation). Observeddisplacements along cross-sections (figures on theright column).
The traces are indicated on the unwrapped interferograms.
173B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

uplift of the main active zone is observed at the Ayaz–Akhtarma mud
volcano. Two scenarios may explain this behavior: (1) the eruption
was a minor event, or (2) the eruption was associated with subsidence
but it is unrecorded because it occurred after the date of the last inter-
ferogram (10 November 2005). As the eruption catalogue reports that
the eruptive event was characterized by the outburst of a huge amount
of mud breccias (Aliyev et al., 2009), the second hypothesis is what we
believe to be true.
Abidin et al. (2009) identified the simultaneous presence of subsi-
dence and uplift during the eruption of the LUSI mud volcano. In this
Fig. 11.Wr apped (figures on the left column) and unwrapped (figureson the central column)interferogramsof the Byandovanmud volcano (see Fig. 5 caption for explanation).The mean
coherencethreshold for all theused interferogramsis 0.25 ± 0.05. The mud volcano edifice has experienced a complex sequenceof uplift and feeble subsidence. Observed displacements
along cross-sections (figures on the right column). The traces of the cross-sections are indicated on the unwrapped interferograms.
174 B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

case, the subsidence is caused by the collapse of the overburden due to
the removal of mud from the subsurface, and the uplift aligns with the
Watukosek Fault system, which caused localized vertical movement.
At the Ayaz–Akhtarma, faults also play a fairly important role in the
deformation pattern. Interestingly, the observed fault/fracture system
broadly occurs within the area which separates sectors with different
rates of subsidence. Particularly, this fault/fracture system downfaults
the sector affectedby the highest subsidence (Fig. 12). Thefault/fracture
may represent either (1) a shallow structure originated to accommo-
date the ground displacement of the area with the maximum subsi-
dence rate, or (2) the surface expression of a deeper fault intercepting
the subsurface fluidreservoir.Inthefirst case, the local subsidence
observed at Ayaz–Akhtarma could correspond to a deflation zone, and
may be the response to a redistribution of fluids. In the second case,
subsidence would be directly related to the movement of a pre-
existingnormal fault that has possiblybeen reactivated bythe increased
fluid pressure regime connected with the pre-eruptive phases. In the
latter hypothesis, thesurface fractures/faults would represent secondary
fault splays connected to the deeper main fault.
Similar uplift–subsidence patterns have also been observed in
magmatic volcanoes, where uplift is typically associated with magma
intrusion in the shallow crust or with hydrothermal fluid injection and
circulation. In several large calderas and magmatic volcanoes, such
as Yellowstone and Campi Flegrei, volcanic uplift can also generate
complex, temporally and spatially, varying patterns of ground deforma-
tion, and may be concurrent with the presence of areas of subsidence
(e.g., Chang et al., 2007; Hutnak et al., 2009; Wicks et al., 1998). Apart
from the obvious differences related to the presence of magma and the
larger dimensions of the magmatic systems, the uplift–subsidence
patterns identified at the Ayaz–Akhtarma show some similarities to
those observed in the Yellowstone caldera, where the area affected
by subsidence could correspond to a deflation zone and the brittle
fracturing/faulting accommodates the differential volumetric variations
(Chang et al., 2007). Similarly, Brunori et al. (2013) identified deformation
patterns of the Cerro Blanco/Robledo Caldera during a resting phase peri-
od. In particular, the caldera subsides with decreasing velocity while a
positive velocity field is detected in the northwestern part of the system
outside the caldera.
Regarding the short-lived pulses at the Akhtarma–Pashaly and
Byandovan mud volcanoes, these are characterized by ground deforma-
tion of considerable magnitude although they do not culminate in an
eruption (we have no reports of eruptions around the period analyzed
with the interferograms). In particular, the Akhtarma–Pashaly mud
volcano underwent an isolated intense deformation event that lasted
3 months approximately, while the Byandovan mud volcano experi-
enced a series of short-term deformation pulses that came to an end
within approximately one month. Important similarities have been
found between these cases and a study of the Murono mud volcano, in
Japan, carried out with high-precision leveling (Kusumoto et al.,
2014). The Murono mud volcano was in a state of quiescence, but the
analysis revealed the occurrence of uplift and subsidence events that
reached approximately 26 mm and 14 mm, respectively. Most of the
benchmarks experienced recurrent uplift and subsidence in a relatively
short period of time (6 months) highlighting the occurrenceof episodic
changes in internal overpressure that may transmit the effects up to the
topographic surface. Pulsed fluids moving through long-lived mud
volcano systems have been indeed found to be important in the erup-
tive history as a mechanism of fluid–mud redistribution and expulsion
(Evans et al., 2006). In general, it is likely that the quiescence periods
of mud volcanoes are cluttered with short-duration phases of deforma-
tion, as those mentioned here.
Similar types of surface deformation events as discrete, short-
duration pulses (shorter than one year) are also experienced by
magmatic volcanoes, such as some of the East African Rift where the
pulses are repeated over time as multiple inflation and deflation events
(Biggs et al., 2009, 2011). Multiple pulses of uplift, several of which did
not end in eruptions, occurred, for example, at Sierra Negra in the
Galapagos Islands (Amelung et al., 2000), Eyjafjallajökull in Iceland
(Pedersen and Sigmundsson, 2006; Sigmundsson et al., 2010)and
Cordón Caulle in Chile (Jay et al., 2014). Volcanic deformation preceding
eruptions can indeed occur episodically at magmatic volcanoes and a
peak in deformation rate does not necessarily imply that an eruption
is about to occur within a relatively short time period (a few years or
months). In particular, Pritchard and Simons (2004) evidenced that
important deformation may occur even during phases of quiescence,
and Hutnak et al. (2009) used numerical simulations and inferred that
the circulation of hydrothermal fluids may explain the observation
that the deformation of some calderas did not culminate in magmatic
eruptions.
Apart from evident differences (i.e., the role of temperature,
crystallization and volatiles), magmatic and mud volcanoes share
some processes and properties, including (i) comparable morphologic
and internal structure (Stewart and Davies, 2006), (ii) a similar eruptive
history (Evans et al., 2006), and (iii) a similar seismic energy density
to be triggered off by earthquakes (Wang and Manga, 2010). The
similarities in the time–space evolution of ground deformation
evidencedfor the studied mud volcanoes of Azerbaijan thus strengthen
the notion that similar processes govern both igneous and mud volcano
systems. The present study indicates that satellite radarinterferometry
represents a suitable tool for studying mud volcano activity, and the
results contribute to a wider understanding of the processes driving
ground deformation at mud volcanoes. In the absence of monitoring
systems and detailed historical information about mudvolcanic activity,
satellite-based observation of the superficial deformation could play a
relevant role in assessing the hazards related to mud volcanoes.
Sentinel-1 radar data, as soon as they become available, will lead to
new challenges in the analysis of ground deformation and will enhance
the potential of radar satellite-based technique thanks to improved
acquisition parameters and increased satellite repeat times (~12 days).
Fig. 12. The location of the fault/fracture detected during the field survey at the Ayaz–
Akhtarma mud volcano is shown in an interferogram (superimposed on the Google Earth
image) and in a graph displaying the final cumulative Line-of-Sight (LOS) displacements of
a cross-section. The trace of the cross-section A–A′is indicated in Fig. 5 (bottom right figure).
Structural field analysis of outcrop-scale faults allowed us to infer the fault dip and to identify
the down-faulted block.
175B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

6. Concluding remarks
The analyzed interferometric data provide new evidence on the
deformation patterns of four mud volcanoes in Azerbaijan.The analyzed
interferograms were selected according to their coherence and span the
period from October 2003 to November 2005. The main outcomes of
this study are the following:
1. The observed ground deformation patterns of mud volcanic edifices
are often characterized by the presence of simultaneous uplift and
subsidence areas, which might correspond to inflation and deflation
zones, respectively.
2. Important deformation events, driven by fluid pressure and volume
variations, can happen in connection with main eruptions. Pre-
eruptive deformation consists of marked uplift and occasional
minor subsidence that is probably related to subsurface redistribu-
tion of pressurized fluids. Ground uplift has been detected to mani-
fest up to one year before the eruption.
3. Ground deformation may also occur as short-lived pulses interrupting
a period of quiescence. Discrete pulses of surface deformation could be
repeated through the long eruptive history of mud volcanoes as a
mechanism of subsurface fluid–mud inflation and redistribution, and
may not be directly related to main eruptive events.
4. The results of this interferometric study show that mud and
magmatic volcanoes display some similarities in the behavior of
ground deformation during quiescence and pre-eruptive stages.
Acknowledgments
Satellite radar data are provided by the European Space Agency (ESA
Category 1 Project 13866). Article is developed within the framework of
the scientific cooperation between Azerbaijan (ANAS) and Italy (CNR)
(two-year program 2012–2013 and 2014–2015). We thank Carolina
Pagli and Maria Cuevas for the helpful comments. We also thank three
anonymous journal reviewers for their constructive comments.
References
Abidin, H.Z., Davies, R.J., Kusuma, M.A., Andreas, H., Deguchi, T., 2009. Subsidence and uplift
of Sidoarjo (East Java) due to the eruption of the LUSI mud volcano (2006–present).
Environ. Geol. 57, 833–844. http://dx.doi.org/10.1007/s00254-008-1363-4.
Abrams, M.A., Narimanov, A.A., 1997. Geochemical evaluation of hydrocarbons and their
potential sources in the western South Caspian depression, Republic of Azerbaijan.
Mar. Pet. Geol. 14, 451–468.
Aliyev, A.A., Guliyev, I.S., Rahmanov, R.R., 2009. Catalogue of Recorded of Mud Volcano
Eruptions of Azerbaijan (1810–2007), Second Edition. ‘Nafta-Press’Publishing-House,
Baku (109 pp.).
Allen, M.B., Jones, S., Ismail-Zadeh, A., Simmons, M.D., Anderson, L., 2002. Onset of
subduction as the cause of rapid Pliocene–Quaternary subsidence in the South
Caspian Basin. Geology 30, 775–778.
Amelung, F., Jónsson, S., Zebker, H., Segall, P., 2000. Widespread uplift and ‘trapdoor’
faulting on Galápagos volcanoes observed with radar interferometry. Nature 407,
993–996. http://dx.doi.org/10.1038/35039604.
Aoki, Y., Sidiq, T.P., 2014. Ground deformation associated with the eruption of Lumpur Sidoarjo
mud volcano, east Java, Indonesia. J. Volcanol. Geotherm. Res. 278–279, 96–102.
Berardino, P., Fornaro, G., Lanari, R., Sansosti, E., 2002. A new algorithm for surface
deformation monitoring based on small baseline differential SAR interferograms.
IEEE Trans. Geosci. Remote Sens. 40 (11), 2375–2383.
Biescas, E., Crosetto, M., Agudo, M., Monserrat, O., Crippa, B., 2007. Two radar interferometric
approaches to monitor slow and fast land deformations. J. Surv. Eng. 133, 66–71.
Biggs, J., Wright, T., Lu, Z., Parsons, B., 2007. Multi-interferogram method for measuring
interseismic deformation: Denali Fault, Alaska. Geophys. J. Int. 170, 1165–1179.
http://dx.doi.org/10.1111/j.1365-246X.2007.03415.x.
Biggs, J., Anthony, E., Ebinger, C., 2009. Multiple inflation and deflationeventsatKenyanvol-
canoes, East African Rift. Geology 37, 979–982. http://dx.doi.org/10.1130/G30133A.1.
Biggs, J., Bastow, I.D., Keir, D., Lewi, E., 2011. Pulses of deformation reveal frequently
recurring shallow magmatic activity beneath the Main Ethiopian Rift. Geochem.
Geophys. Geosyst. 12 (9), Q0AB10.
Bonini, M., 2012. Mud volcanoes: indicators of stress orientation and tectonic controls.
EarthSci.Rev.115,121–152.
Bonini, M., Mazzarini, F., 2010. Mud volcanoes as potential indicators of regional stress
and pressurized layer depth. Tectonophysics 494, 32–47.
Bonini, M., Tassi, F., Feyzullayev, A.A., Aliyev, C.S., Capecchiacci, F., Minissale, A., 2013. Deep
gases discharged from mud volcanoes of Azerbaijan: new geochemical evidence. Mar.
Pet. Geol. 43, 450–463.
Brown, K.M., 1990. The nature and hydrological significance of mud diapirism and
diatremes for accretionary systems. J. Geophys. Res. 95, 8969–8982.
Brunori, C.A., Bignami, C., Stramondo, S., Bustos, E., 2013. 20 years of active deformation
on volcano caldera: joint analysis of InSAR and AInSAR techniques. Int. J. Appl.
Earth Obs. Geoinformation 23, 279–287.
Caricchi, L., Biggs, J., Annen, C., Ebmeier, S., 2014. The influence of cooling, crystallization
and re-melting on the interpretation of geodetic signals in volcanic systems. Earth
Planet. Sci. Lett. 388, 166–174.
Chang, W., Smith, R.B., Wicks, C., Farrell, J.M., Puskas, C.M., 2007. Accelerated uplift and
magmatic intrusion of the Yellowstone caldera, 2004 to 2006. Science 318,
952–956. http://dx.doi.org/10.1126/science 1146842.
Cooper, C., 2001. Mud volcanoes of Azerbaijan visualized using 3D seismic depth cubes: the
importance of overpressured fluid and gas instead of non-existent diapirs (abs.).
Proceedings of the European Association of Geoscientists and Engineers, Conference
on ‘Subsurface Sediment Mobilisation’, Gent, Belgium, September 2001, p. 71.
Costantini, M., 1998. A novelphase unwrappingmethod based on network programming.
IEEE Trans. Geosci. Remote Sens. 36 (3), 813–821.
Davies, R.J., Swarbrick, R.E., Evans, R.J., Huuse, M., 2007. Birth of a mud volcano: East Java,
29 May 2006. Geol. Soc. Am. Today 17, 4–9.
Devlin, W., Cogswell, J., Gaskins, G., Isaksen, G., Pitcher, D., Puls, D., Stanley, K., Wall, G.,
1999. South Caspian Basin: young, cool, and full of promise. Geol. Soc. Am. Today 9
(7), 1–9.
Dimitrov, L.I., 2002. Mud volcanoes—the most important pathway for degassing deeply
buried sediments. Earth Sci. Rev. 59, 49–76.
Ebmeier, S.K., Biggs, J., Mather, T.A., Wadge, G., Amelung, F., 2010. Steady downslope
movement on the western flank of Arenal Volcano, Costa Rica. Geochem. Geophys.
Geosyst. (ISSN: 1525-2027) 11 (12). http://dx.doi.org/10.1029/2010GC003263.
Ebmeier, S.K., Biggs, J., Mather, T.A., Elliott, J.R., Wadge, G., Amelung, F., 2012. Measuring
large topographicchange with InSAR: lavathicknesses, extrusion rate and subsidence
rate at Santiaguito Volcano, Guatemala. Earth Planet. Sci. Lett. 335–336, 216–225.
Evans, R.J., Davies, R.J., Stewart, S.A., 2006. Internal structure and eruptive history of a
kilometer-scale mud volcano system, South Caspian Sea. Basin Res. 19, 153–163.
Ferretti, A., Prati, C., Rocca, F., 2001. Permanent scatterers in SAR interferometry. IEEE
Trans. Geosci. Remote Sens. 39 (1), 8–20.
Fowler, S.R., Mildenhall, J., Zalova, S., Riley, G., Elsley, G., Desplanques, A., Guliyev,F., 2000.
Mud volcanoes and structural development on Shah Deniz. J. Pet. Sci. Eng. 28,
189–206.
Fryer, P., Wheat, C.G., Mottl, M.J., 1999. Mariana blueschist mud volcanism: implications
for conditions within the subduction zone. Geology 27, 103–106.
Fukushima, Y., Mori, J., Hashimoto, M., Kano, Y., 2009. Subsidence associated with the LUSI mud
eruption, East Java, investigated by SAR interferometry. Mar. Pet. Geol. 26, 1740–1750.
Guliyev, I.I., Feizullayev, A.A., 1997. All About Mud Volcanoes. Bafta Press, Baku,
Azerbaijan.
Hanssen, R., 2001. Radar Interferometry: Data Interpretation and Error Analysis Vol. 2.
Springer.
Higgins, G., Saunders, J.B., 1974. Mud volcanoes and their nature and origin. In: Jung, P.
(Ed.), Contributions to the Geology and Paleobiology of the Caribbean and Adjacent
Areas. Verh. Aturforsch. Ges. vol. 84(1), pp. 101–152.
Hommels, A., Scholte, K.H., Munoz-Sabater, J., Hanssen, R.F., Van der Meer, F.D.,
Kroonenberg, S.B., Aliyeva, E., Huseynov, D., Guliev, I., 2003. Preliminary Aster and
InSAR imagery combination for mud volcano dynamics, Azerbaijan. International
Geoscience and Remote Sensing Symposium 3, III, pp. 1573–1575.
Hooper, A., Zebker, H., Segall, P., Kampes, B., 2004. A new method for measuring deformation
on volcanoes and other natural terrains using InSAR persistent scatterers. Geophys. Res.
Lett. 31 (23) (December 2004).
Hreinsdóttir, S., Sigmundsson, F., Roberts, M.J., Björnsson, H., Grapenthin, R., Arason,
P., Árnadóttir, T., Hólmjárn, J., Geirsson, H., Bennett, R.A., Gudmundsson, M.T.,
Oddsson, B., Ófeigsson, B.G., Villemin, T., Jónsson, T., Sturkell, E., Höskuldsson, Á.,
Larsen, G., Thordarson, T., Óladóttir, B.A., 2014. Volcanic plume height correlated
with magma pressure change at Grímsvötn Volcano, Iceland. Nat. Geosci. 7. http://
dx.doi.org/10.1038/Ngeo2044.
Hudson, S.M., Johnson, C.L., Efendiyeva, M.A., Rowe, H.D., Feyzullayev, A.A., Aliyev, C.S.,
2008. Stratigraphy and geochemical characterization of the Oligocene–Miocene
Maikop series: implications for the paleogeography of Eastern Azerbaijan.
Tectonophysics 451, 40–55.
Hutnak, M., Hurwitz, S., Ingebritsen, S.E., Hsieh, P.A., 2009. Numerical models of caldera
deformation: effects of multiphase and multicomponent hydrothermal fluid flow. J.
Geophys. Res. 114, B04411. http://dx.doi.org/10.1029/2008JB006151.
Inan,S.,Yaclin,M.N.,Guliev,I.S.,Kuliev,K.,Feizullayev,A.A.,1997.Deep petroleum
occurrences in the Lower Kura Depression, South Caspian Basin, Azerbaijan: an
organic geochemical and basin modelling study. Mar. Pet. Geol. 14, 731–762.
Jackson, J., Priestley, K., Allen, M., Berberian, M., 2002. Active tectonics of the South
Caspian Basin. Geophys. J. Int. 148, 214–245.
Jakubov, A.A., Ali-Zade, A.A., Zeinalov, M.M., 1971. Mud Volcanoes of the Azerbaijan SSR.
Atlas (in Russian), Azerbaijan Academy of Sciences, Baku.
Jay, J., Costa, F., Pritchard, M., Larac, L., Singer, B., Herrin, J., 2014. Locating magma reservoirs
using InSAR and petrology before and during the 2011–2012 Cordón Caulle silicic
eruption. Earth Planet. Sci. Lett. 395, 254–266.
Kopf, A.J., 2002. Significance of mud volcanism. Rev. Geophys. 40 (2), 1005.
Kusumoto, S., Sudo, K., Kawabata, M., Uda, T., Fukuda, Y., 2014. Vertical movement during the
quiescent phase of the Murono mud volcano, Niigata, Japan. Earth, Planets Space 66, 14.
Massonnet, D., Feigl, K.L., 1998. Radar interferometryand its application to changesin the
earth's surface. Rev. Geophys. 36, 441–500.
Mazzini, A., Svensen, H., Planke, S., Guliyev, I., Akhmanov, G.G., Fallik, T., Banks, D., 2009.
When mud volcanoes sleep: insight from seep geochemistry at the Dashgil mud
volcano, Azerbaijan. Mar. Pet. Geol. 26, 1704–1715.
176 B. Antonielli et al. / Tectonophysics 637 (2014) 163–177

Mellors,R.J., Bunyapanasarn, T., Panahi, B., 2005.InSAR analysis of the Absheron Peninsula
and nearby areas, Azerbaijan. Mud Volcanoes, Geodynamics and Seismicity. NATO
Science Series volume 51, pp. 201–209.
Mellors, R.J., Kilb, D., Aliyev, A., Gasanov, A., Yetirmishli, G., 2007. Correlations between
earthquakes and large mud volcano eruptions. J. Geophys. Res. 112, B04304. http://
dx.doi.org/10.1029/2006JB004489.
Morley, C.K., 2003. Outcrop examples of mudstone intrusions from the Jerudong anticline,
Brunei Darussalam and inferences for hydrocarbon reservoirs. In: Van Rensbergen, P.,
Hillis, R.R., Maltman, A.J., Morley, C.K. (Eds.), Subsurface Sediment MobilizationSpecial
Publications 216. Geological Society, London, pp. 381–394.
Morley, C.K., Guerin, G., 1996. Comparison of gravity-driven deformation styles and
behavior associated with mobile shales and salt. Tectonics 15, 1154–1170.
Pagli, C., Sigmundsson, F., Arnadòttira, T., Einarsson, P., Sturkell, E., 2006. Deflation of the
Askja volcanic system: constraints on the deformation source from combined inversion
of satellite radar interferograms and GPS measurements. J. Volcanol. Geotherm. Res.
152, 97–108.
Pagli, C., Wright, T.J., Ebinger, C.J., Yun, S.H., Cann, J.R., Barnie, T., Ayele, A., 2012. Shallow
axial magma chamber at the slow-spreading Erta Ale Ridge. Nat. Geosci. http://dx.
doi.org/10.1038/NGEO1414.
Pedersen,R., Sigmundsson, F., 2006. Temporal development of the 1999intrusive episode
in the Eyjafjallajökull Volcano, Iceland, derived from InSAR images. Bull. Volcanol. 68,
377–393.
Planke, S., Svensen, H., Hovland, M., Banks, D.A., Jamtveit, B., 2003. Mud and fluid migration
in active mud volcanoes in Azerbaijan. Geo-Mar. Lett. 23, 258–268.
Pritchard, M., Simons, M., 2004. Surveying volcanic arcs with satellite radar interferometry:
the Central Andes, Kamchatka, and beyond. GSA Today 14 (8). http://dx.doi.org/10.
1130/1052-5173.
Reynolds, A.D., Simmons, M.D., Bowman, M.B.J., Henton, J., Brayshaw, A.C., Ali-Zade, A.A.,
Guliyev, I.S., Suleymanova, S.F., Ateava, E.Z., Mamedova, D.N., Koshkarly, R.O., 1998.
Implications of outcrop geology for reservoirs in the Neogene productive series; Ap-
sheron Peninsula, Azerbaijan. AAPG Bull. 82, 25e49.
Roberts, K.S., Davies, R.J., Stewart, S.A., 2010. Structure of exhumed mud volcano feeder
complexes, Azerbaijan. Basin Res. 22, 439–451.
Rudolph, M.L., Shirzaei, M., Manga, M., Fukushima, Y., 2013. Evolution and future of the
LUSI mud eruption inferred from ground deformation. Geophys. Res. Lett. 40, 1–4.
http://dx.doi.org/10.1002/Grl.50189.
Sigmundsson, F., Vadon, H., Massonnet, D., 1997. Readjustment of the Krafla spreading
segment to crustal rifling measured by satellite radar interferometry. Geophys. Res.
Lett. 24 (15), 1843–1846.
Sigmundsson, F., Hreinsdóttir, S., Hooper, A., Arnadottir, T., Pedersen, R., Roberts, M.J.,
Oskarsson, N., Auriac, A., Decriem, J., Einarsson, P., Geirsson, H., Hensch, M.,
Ofeigsson, B.G., Sturkell, E., Sveinbjornsson, H., Feigl, K.L., 2010. Intrusion triggering
of the 2010 Eyjafjallajokull explosive eruption. Nature 468 (7322), 426–430. http://
dx.doi.org/10.1038/nature 09558.
Soto, J.I., Santos-Betancor, I., Sánchez Borrego, I., Macellari, C.E., 2011. Shale Diapirism and
Associated Folding History in the South Caspian Basin (offshore Azerbaijan). Search
and Discovery Article No. 30162.
Stewart, S.A., Davies, R.J., 2006. Structure and emplacement of mud volcano systems in
the South Caspian Basin. AAPG Bull. 90, 771–786.
Wang, C.Y., Manga, M., 2010. Hydrologic responses to earthquakes and a general metric.
Geofluids 10, 206–216.
Wicks,C.,Thatcher,W.,Dzurisin,D.,1998.Migrationoffluids beneath Yellowstone caldera
inferred from satellite radar interferometry. Science 282, 458–462. http://dx.doi.org/
10.1126/science.282.5388.458.
Wright, T.J., Sigmundsson, F.,Pagli, C., Belachew, M., Hamling,I.J., Brandsdóttir, B., Keir, D.,
Pedersen, R., Ayele, A., Ebinger, C., Einarsson, P., Lewi, E., Calais, E., 2012. Geophysical
constraints on the dynamics of spreading centers from rifting episodes on land. Nat.
Geosci. http://dx.doi.org/10.1038/NGEO1428 (Review Article).
Yusifov, M., Rabinowitz, P.D., 2004. Classification of mud volcanoes in the South Caspian
Basin, offshore Azerbaijan. Mar. Pet. Geol. 21, 965–975.
Zeinalov, G.A., 2000. Importance of remote-sensing data in structural geologic analysis of
oil- and gas-bearing regions of Azerbaijan. Nat. Resour. Res. 9 (4).
177B. Antonielli et al. / Tectonophysics 637 (2014) 163–177