The application of satellite technology in the study of geodynamic movements in the wider Zagreb area

Abstract

Recognizing the fact that the region around the capital belongs seismically in one of the most vulnerable areas in the Republic of Croatia, the interdisciplinary geodynamic research was initiated which lasts for 18 years now. Since the establishment of the geodynamic network, ten series of precise GPS measurements on specially stabilized points of the geodynamic network were conducted in order to determine tectonic movements and related seismic activities in the wider Zagreb area. From all conducted measurements in the period from 1997 to 2015, the original geodetic model of tectonic movements was created. Following the analysis of geodetic and geological data, a unique interdisciplinary surface layers movement model of the research area was determined. During 2015, PSInSAR (Persistent Scatterer InSAR) technology was included in the research and was combined with the GPS measurements. The paper describes the combination of these two methods for the period 2004 ÷ 2009 which resulted in denser and more reliable movement model in the urban part of the research area, especially in its height component, which is very important for the safety of buildings.

Full text

B. Pribičević i dr. Primjena satelitskih tehnologija u istraživanju geodinamičkih pomaka na širem zagrebačkom području
Tehnički vjesnik 24, 2(2017), 503-512 503
ISSN 1330-3651 (Print), ISSN 1848-6339 (Online)
DOI: 10.17559/TV-20160817013320
THE APPLICATION OF SATELLITE TECHNOLOGY IN THE STUDY OF GEODYNAMIC
MOVEMENTS IN THE WIDER ZAGREB AREA
Boško Pribičević, Almin Đapo, Marin Govorčin
Original scientific paper
Recognizing the fact that the region around the capital belongs seismically in one of the most vulnerable areas in the Republic of Croatia, the
interdisciplinary geodynamic research was initiated which lasts for 18 years now. Since the establishment of the geodynamic network, ten series of precise
GPS measurements on specially stabilized points of the geodynamic network were conducted in order to determine tectonic movements and related
seismic activities in the wider Zagreb area. From all conducted measurements in the period from 1997 to 2015, the original geodetic model of tectonic
movements was created. Following the analysis of geodetic and geological data, a unique interdisciplinary surface layers movement model of the research
area was determined. During 2015, PSInSAR (Persistent Scatterer InSAR) technology was included in the research and was combined with the GPS
measurements. The paper describes the combination of these two methods for the period 2004 ÷ 2009 which resulted in denser and more reliable
movement model in the urban part of the research area, especially in its height component, which is very important for the safety of buildings.
Keywords: geodynamics; GPS measurements; PSInSAR; tectonic movement; Zagreb
Primjena satelitskih tehnologija u istraživanju geodinamičkih pomaka na širem zagrebačkom području
Izvorni znanstveni članak
Uvažavajući činjenicu da šire zagrebačko područje spada seizmički u jedno od najugroženijih područja u Republici Hrvatskoj, započeta su predmetna
interdisciplinarna geodinamička istraživanja koja traju punih 18 godina. Od uspostave geodinamičke mreže provedeno je deset serija preciznih GPS
mjerenja na specijalno stabiliziranim točkama predmetne mreže u svrhu određivanja pomaka odnosno tektonskih pokreta i srodnih seizmičkih aktivnosti
na širem zagrebačkom području. Iz svih provedenih mjerenja od 1997. do 2015. godine stvoren je originalni geodetski model tektonskih pomaka.
Provedenom analizom geodetskih i geoloških podataka, kao rezultat je određen jedinstveni interdisciplinarni model gibanja pripovršinskih slojeva na
području istraživanja.Tijekom 2015. godine u istraživanje je uključena PSInSAR (Persistent Scatterer InSAR) tehnologija koja je kombinirana sa GPS
mjerenjima. U radu je prikazana kombinacija ovih dviju metoda za period 2004. ÷ 2009. te je rezultirala puno gušćim i još vjerodostojnijim modelom
pomaka u izgrađenom dijelu područja istraživanja, i to posebice u njegovoj visinskoj komponenti, što je posebno važno za sigurnost građevina.
Ključne riječi: geodinamika; GPS mjerenja; PSInSAR; tektonika; Zagreb
1 Introduction
This paper is the result of integrating all of the
collected data from eighteen years’ research of
interdisciplinary scientific team which focused on the
monitoring and study of geodynamic processes of the
wider area of the City of Zagreb. During the planning of
the included areas of research the size and spatial
extension of the City of Zagreb and other interdisciplinary
aspects were taken into account so that the total area
which respective geodynamic studies included is about
800 km2.
The main objective of the underlying geodynamic
research is to determine the actual geodynamic
movements with very high accuracy over a long period of
time. In order to represent geodynamic events in the area
of research in the best way, Geodynamic Network was
designed and established in 1997 with an initial total of 43
specially stabilized points [1, 2].
GPS technology in the last twenty years has
established itself as a major geodetic contribution to the
geodynamic research [3], and the establishment of the
respective geodynamic network, in 1997, opened the
space for development of the geodynamic project to
monitor tectonic movements in the Zagreb mountain
Medvednica and boundary areas of several important
tectonic units that meet in the wider Zagreb area [4, 5].
On one side are the branches of the southeast Alps, on the
other side the Dinarides, and on the third side is the
Pannonian Basin. In the research area numerous important
faults stand out along which tectonic activity determines
the number and intensity of earthquakes [6, 7]. GPS
measurement campaigns on Geodynamic network were
conducted ten times in total in the period from 1997 to
2015, which represent a set of measured data, which is
used for the determination of geodynamic movements in
the network [8].
Since in the wider Zagreb area within the underlying
research, a full range of geological measurements was
conducted too, a scientific comparison of the two models
was realized by independent methods: geodetic model,
based on the precise GPS measurements and geological
model, based on long-term geological research, were
performed.
Therefore, scientifically-based testing and analysis of
correlation between the geodetic and geological models is
carried out in order to create a unique interdisciplinary
movement model in the wider Zagreb area [8].
Data from GPS measurements conducted on the
points of the geodynamic network in the period from
1997 to 2015, have been processed in the scientific
software GAMIT/GLOBK, which is designed specifically
for processing of GPS measurements in geodynamic
networks. GAMIT/GLOBK calculates the velocities on
the points by applying modern methods of Kalman
filtering [11].
The result are the velocity models on the points of the
geodynamic network for the periods between the
individual campaigns, and cumulatively for the whole
period from 1997 to 2015. Also the interpolation of the
velocity values obtained for the entire period of research
by IDW (Inverse Distance Weighting) method was

The application of satellite technology in the study of geodynamic movements in the wider Zagreb area B. Pribičević et al.
504 Technical Gazette 24, 2(2017), 503-512
performed with the inclusion of the Zagreb area fault
model. In the end the original geodetic model of velocity
field for the whole research area was created.
The research was extended by including the SAR
(Synthetic Aperture Radar) technology. The reasons for
the use of this technology in the field of geodynamic
research are tremendous technological advances in the
collection, as well as data processing and large coverage
of the observations. Namely, it is the application of radar
technology from space to obtain a digital representation of
the configuration of the ground on which the periodic
observations can detect changes with millimeter accuracy.
As part of the research, 40 satellite (Envisat) radar images
for the period 2004÷2009 were processed. During the
processing of images, interferometric method PSInSAR
(Persistent scatterer Interferometry) was applied which
resulted with the denser and more credible model of the
movements in the urban areas of research, in particular in
its height component, which is especially important for
the safety of buildings. It's worth noting that the SAR
observations were combined with the GPS measurements,
and the reason for defining the aforementioned
observation period (2004÷2009) was in high
representation of the GPS campaigns in that period (2004,
2006, 2007, 2008 and 2009). The GPS data enable precise
determination of horizontal and vertical movements on
discrete points while the SAR data provide precise
vertical movements over large areas with a temporal
resolution of one month. By combining these two
methods the accurate and detailed perception of
geodynamic activities in the area of research with a focus
on urban areas is acquired.
2 Conducted GPS measurements, processing and
results
Research on the Geodynamic Network of the City of
Zagreb began in 1997 and so far during the eighteen
years, ten GPS campaigns were carried out in order to
determine the geodynamic movements on the network
points.
Figure 1 Geodynamic network with marked major faults and epicenters
of earthquakes.
A total of ten GPS campaigns was conducted: 1997,
2001, 2003, 2004, 2005, 2006, 2007, 2008, 2009 and
2015. Only seven out of ten of those measurement
campaigns cover all of the points of the Network, as
follows: 1997, 2001, 2004, 2006, 2008, 2009 and 2015
(41 point). Two of the campaigns were carried out with
the aim of surveying on densification points 2005 and
2007 (11 and 21 points) [1, 8]. Tab. 1 shows all conducted
campaigns including the number of sessions, points
included and receivers used. Fig. 1 shows the respective
network with the most important faults and epicenters of
earthquakes plotted.
All conducted GPS measurements on the
Geodynamic Network of the City of Zagreb were
processed in the same way, by using the latest version of
the scientific GPS software GAMIT/GLOBK ver. 10.6
which runs under the Linux operating system.
GAMIT/GLOBK uses Kalman filter to determine the
velocities from the discrete GPS campaigns. This
software gives multipoint solutions, therefore, all the
point velocities are processed at once and the covariance
matrix is common to the entire network [11].
All GAMIT parameters are set for regional or local
campaign. For the reference of the geodynamic network
two points were selected: the city's permanent GPS
stations, CAOP 1038, because it is the most stable point
in the City of Zagreb and in addition the city permanent
GPS station, and point ZZFP 1039. Referent points also
had the possibility of relative movement, but have proven
to be very stable. Shifts of all other points in the network
refer to the vector between the two points mentioned. In
this way, the possibility of overestimating the movement
of all points in the network due to possible movement of a
unique origin was avoided.
Table 1 All conducted GP campaigns with number of sessions, points
included and used receivers
Campaign Date
Nr.
sesion
Nr.
Pts
Nr.
GPS
Zagreb 1997
27.10.1997 - 29.10.1997
2
43
27
Zagreb 2001
25.06.2001 - 28.06.2001
3
40
16
Zagreb 2003
22.06.2003 - 23.06.2003
1
13
13
Zagreb 2004
17.06.2004 - 20.06.2004
3
39
13
Zagreb 2005
10.09.2005 - 11.09.2005
1
11
11
Zagreb 2006
22.06.2006 - 25.06.2006
3
41
15
Zagreb 2007
13.07.2007 - 15.07.2007
2
21
13
Zagreb 2008
10.06.2008 - 13.06.2008
3
41
15
Zagreb 2009
11.06.2009 - 14.06.2009
3
41
15
Zagreb 2015
11.06.2015 - 14.06.2015
3
41
15
Point velocities were calculated using GLOBK
modules, where the velocity field was calculated using a
method of Kalman filtering [9÷15]. Velocities are
expressed in millimeters per year (mm/yr). Also the
magnitude of the total spatial vector was calculated in all
three dimensions.
Tab. 2 provides the statistical absolute values of
cumulative annual velocity of network points for the
period 2004÷2009. To obtain a cumulative model of
velocities of points of the observed area for the entire
period from 2004 to 2009, in further processing of the
data it was necessary to combine the resulting solution of
each individual campaign.
The first two columns show the velocity as a
component of latitude and longitude (v
ϕ
and vλ) the third

B. Pribičević i dr. Primjena satelitskih tehnologija u istraživanju geodinamičkih pomaka na širem zagrebačkom području
Tehnički vjesnik 24, 2(2017), 503-512 505
column is the resultant of the first and second component
(horizontal velocity - vhor), the last column is the vertical
velocity component (vver).
Figure 2 Cumulative velocities on the network points for the period
from 2004 to 2009
Fig. 2 shows the cumulative solution of the point
velicities of the Network points obtained for the period
observed from 2004 to 2009.
Absolute velocity values for the cumulative solution
in horizontal direction are 5,4 mm/yr and in vertical 21,1
mm/yr. Computed cumulative solution is used for creating
the geodetic model of tectonic movements and in the
combination with geologic data for the creation of the
original movement model of the Earth‘s crust surface on
the wider Zagreb area.
Figure 3 Geodetic velocity model computed using IDW interpolation
with inclusion of the foult model of the wider Zagreb area
Table 2 Statistical overview of the absolute values of the annual
velocities of the points for the period from 1997 to 2009
vφ (mm/yr)
vλ (mm/yr)
vhor (mm/yr)
vver (mm/yr)
Min
0,0
0,1
0,4
0,0
Max
3,5
4,9
5,4
21,1
Avg
1,1
1,9
2,4
4,0
3 Synthetic Aperture Radar Interferometry (InSAR)
Synthetic Aperture Radar Interferometry (InSAR),
also referred to as SAR (Synthetic Aperture Radar)
interferometry is one of the radar based techniques used
in geodesy, as a part of remote sensing techniques. The
InSAR technique exploits two or more SAR images to
measure the signal phase changes, or interference over
time. Results are interferograms, which can be used for
monitoring different geophysical and surface processes
occurring on the Earth’s surface, such as earthquakes,
vulcanic eruptions, landslide, uprising and subsidence
events, glacial movements and etc. Moreover, with this
technique, it is possible to achieve a determination of
ground surface deformations with sub-centimeter
accuracy, over an area of interest in period of a few days
to several years.
Figure 4 SAR acquisition geometry [16]
Synthetic Aperture Radar is an active radar imaging
system usually located on the moving platforms
(airplanes, satellites), which are flying over and mapping
the areas of interest. The SAR systems are side-looking
imaging systems and the geometry of the data acquisition
is shown in Fig. 4. This imaging system represents one of
the active remote sensing techniques, where the SAR
antenna emits and receives the backscattered radar signals
(electromagnetic radiation of the objects)(Fig. 5).
Figure 5 Emitting and receiving backscattered electromagnetic radiation
[17]
The most common SAR platforms are satellites
equipped with SAR sensors orbiting the Earth on a near-
polar orbit at an altitude ranging from 500 to 800 km
above the Earth’s surface [18]. Satellite missions
equipped with SAR sensors and their main characteristics
are shown in Tab. 3.

The application of satellite technology in the study of geodynamic movements in the wider Zagreb area B. Pribičević et al.
506 Technical Gazette 24, 2(2017), 503-512
Table 3 Satellite missions equipped with SAR sensors
Satellite
Band and signal
wavelength (cm)
Max spatial
resolution (m)
Polarisation
Max. temporal
resolution (days)
State, yr. of launch
Sentinel-1
C-band – 5,6
5
HH, VV, HV, VH
6
Europe 2014
ALOS-2
L-band – 23,5
3
HH, VV, HV, VH
14
Japan 2014
CosmoSkymed
X-band – 3
1
HH, VV, HV, VH
16
Italy 2007
TerraSAR-X
X-band – 3
1
HH, VV, HV, VH
11
Germany, 2007
Radarsar- 2
C-band – 5,6
3
HH, VV, HV, VH
24
Canada 2007
ALOS
L-band – 23,5
10
HH, VV, HV, VH
46
Japan 2006
Envisat
C-band – 5,6
23
HH, VV, HV, VH
35
Europe 2002
Radarsat
C-band – 5,6
8
HH
24
Canada 1995
ERS-2
C-band – 5,6
23
VV
35
Europe 1995
ERS-1
C-band – 5,6
23
VV
35
Europe 1991
The angle between satellites orbit and true North-
South varies slightly depending on the satellite but, in
general lies in the range of 10°. Furthermore, satellites are
orbiting around the Earth in two directions; from the
North to the South Pole (descending orbit) and back, from
the South to the North Pole (ascending orbit) (Fig. 6)
[18].
Figure 6 Ascending and descending satellite orbits [19]
Data acquired by SAR acquisitions are first stored on
the satellite memory, but the memory is limited, thus the
gathered spatial data must be transmitted to numerous
control stations around the Globe for further data
processing and analysis. The gathered data, radar images,
contain two important pieces of information concerning
backscattered radar signals: amplitudes and phase values.
Taking into account that the radar signal’s wavelength
and phase are in a direct correlation, any change of the
signal’s wavelength is also expressed as a change of its
phase.
Figure 7 The relationship between a change of electromagnetic signal’s
phase and ground movement [18]
The SAR interferometry technique measures the
phase changes during the exposition time and exploits this
to determine the surface ground deformations. When
some particular point of the Earth’s topography is
moving, the distance between that point on ground and
satellite is changing, which will also take an effect on the
measured phase value of the returned signal. The
relationship between that ground point movement and
corresponding shift in signal phase between two SAR
signals acquired over the same area is depicted in Fig. 7.
The change in signal phase (Δφ) can be expressed
with the following simple equation:
,Δ
π4
Δ
α
λ
ϕ
+= R
(1)
where λ is the wavelength, ΔR is the displacement and α
is a phase shift due to different atmospheric conditions at
the time of the two radar acquisitions. Consequently, any
radar target along the satellite line of sight creates a phase
shift in the radar signal that can be detected by comparing
the phase values of two SAR images acquired at different
times [18].
Digital representation of all surface changes derived
from the difference of the phase values corresponding to a
certain area is an interferogram. Determination of true
displacement values from interferometric phases is not a
trial task, as interferometric phase values are a blend of
different signal contributors.
Interferometric phase is impacted by four
contributors: topographic distortions arising from slightly
different viewing angles of the two satellite passes (t),
atmospheric effects (α) arising from the wavelength
distortion that occurs when signals enter and leave a
moisture-bearing layer, any range displacement of the
radar target ΔR, and noise; range being the distance
between the sensor and the target. Therefore, the equation
1 can be extended to more precise form [18]:
. Δ
π4
ΔnoisetR +++=
α
λ
ϕ
(2)
Process of removing the topographic effect on
interferometric phase values is Differential InSAR
technique (DInSAR). DInSAR uses a Digital Elevation
Models (DEMs) of the area of interest to remove the
topographic effects. After the phase contribution due to
the local topography is accurately compensated, in

B. Pribičević i dr. Primjena satelitskih tehnologija u istraživanju geodinamičkih pomaka na širem zagrebačkom području
Tehnički vjesnik 24, 2(2017), 503-512 507
interferometric phase values are left atmospheric, noise
(negligible) and displacements contributors to phase shift.
Following the realization that atmospheric effects on
signal phase values were significant, a method emerged in
the late 1990’s that sought to mitigate this effect by
‘averaging’ data within multiple interferograms. This
process was referred to as Interferogram Stacking. [16]
The two most used multi-stacking techniques are Small
Baseline Subset (SBAS) and Persistent Scatter (Ps)
technique.
3.1 PsInSAR
Persistent Scatterer Interferometry (PSI) is the
collective term used within the InSAR community to
distinguish between single interferogram DInSAR and the
second generation of InSAR technologies. The PS
technique first appeared in 1999, when the base algorithm
was developed at the Politecnico di Milano (Polimi).
PsINSAR is an advanced interferometry technique based
on the phase shift determination on the permanent
scatterers (PSs) of the electromagnetic radiation located
on the Earth’s surface.
Permanent scatterers are defined as a radar targets,
within a resolution cell of radar image, which display
stable amplitude properties and coherent signal phase
throughout all of the images within a data stack. Objects
that make good PS are varied and can be natural or man-
made. Among the natural forms are: rock outcrops, hard
un-vegetated earth surfaces and boulders. Among the
man-made objects are: buildings, street lights,
transmission towers, bridge parapets, above-ground
pipelines, appurtenances on dams and roof structures and
any rectilinear structures that can create a dihedral signal
reflection back to the satellite [18].
The main aim of PS technique is to overcome the
errors introduced into signal phase values by atmospheric
artifacts. That can be accomplished with the identification
of PS targets on multiple radar images on which the
atmospheric correction procedure can be performed.
Minimum required scenes for successful utilization of PS
technique are 15 radar images. There is also possibility to
perform the technique with smaller radar image stack
(especially if the AOI is a mainly urban area) with a good
coherence between images but in general it is not
recommended. The precision of the technique is better
with larger radar data stack as the reliability of the
technique‘s results depends on the number of images used
in process. Having removed the atmospheric artifacts with
this technique, the displacements on the aforementioned
natural and man-made objects can be determined with the
accuracy of a few millimeters per year.
3.1 The application of InSAR technique on the wider
Zagreb area
Satellite Radar Interferometry is implemented in the
research to obtain a detail insight into on-going surface
deformations occurring in the wider Zagreb area and to
enhance previous research findings gathered through the
aforementioned GPS geodynamic research. In order to
successfully utilize InSAR technique, it is important to
know its advantages and disadvantages, topography
features and expected surface deformations of the area of
interest (AOI). Therefore, it is necessary to have a good
preparation phase before the data processing starts.
The application of InSAR technique on the wider
Zagreb area (Fig. 8) can be divided into three phases of its
realization. Preparation phase is the first phase and it
represents a process of finding the right radar images by
searching the online databases (archives) and a detail
analysis of topography features of the AOI. Second phase
is the main phase and it is comprised of two steps: an
acquiring of the adequate radar images from institutions
responsible for InSAR data dissemination and the
processing of the acquired data. Final phase is an analysis
of obtained results and a creation of deformation maps.
Figure 8 Area of the research (the wider Zagreb area)
Radar images can be obtained from the institutions
such as European Space Agency (ESA), German
Aerospace Center (DLR), Canadian Space Agency and
others, by ordering archived data or by requesting future
data acquisitions of AOI. The archived data are only
available for certain areas investigated in the past as SAR
data acquisitions mainly depend on user demands. Also,
there are certain requirements that SAR observations need
to fulfill for interferometric applications such as
perpendicular and temporal baseline restrictions, which
are important for data processing and thereby selection of
right radar images. Number of adequate radar images for
interferometric processing available over the wider
Zagreb area are: 21 images of ERS1, 148 images of
ERS2, 59 images of Envisat-ASAR, 260 images of
Radarsat 2 and 69 images of Sentinel 1.
Taking into account the good temporal coverage of
GPS campaigns on the Geodynamic GPS Network of the
City of Zagreb for the period of 2004 to 2009, it was
decided to use that period for an investigation of InSAR
usability over the wider Zagreb area. Moreover, GPS
campaigns were conducted once per every year which
ensured good temporal coverage of ongoing ground
displacements over the AOI and also mitigated the
additional space for interpolation. That enabled the
possibility to combine the two independent geodetic
techniques, InSAR and GPS for the same research period.
Therefore, after finding all available radar data for an
interferometric processing over the AOI, it was decided to

The application of satellite technology in the study of geodynamic movements in the wider Zagreb area B. Pribičević et al.
508 Technical Gazette 24, 2(2017), 503-512
use Enivisat satellite mission and its 40 radar images over
the Zagreb area for the aforementioned time span.
Both orbit directions are also considered to ensure the
maximum spatial coverage of the area by bypassing the
shadowing effect of Medevednica Mountain in the data
acquisitions. Therefore, the acquired data for the
realization of the research were: 23 Enivsat ASAR scenes
from the track 265 (descending orbit) and 17 Enivsat
ASAR scenes from the track 272 (ascending track).
Furthermore, Envisat satellite mission ended in 2012, so
in the interest of further InSAR investigation over the area
the use of some other satellite missions like
CosmoSkyMed, Radarsat-2 or Sentinel 1 should be
considered.
First phase of the InSAR application ended with a
detailed analysis of the data availability and, afterwards,
by choosing the right radar images for interferometric
processing. The next step was an establishment of the
cooperation with the European Space Agency (ESA) in
order to acquire the necessary data. ESA recognized the
importance of this kind of research in this region by
providing the 40 Enivsat ASAR scenes for the purpose of
the research.
3.3 Radar data processing
Radar data processing was conducted with several
scientific software packages developed on different
universities. Each software is used in a different phase of
data processing due to the inability to conduct the data
processing in only one software from the beginning to the
end. Moreover, they are scientific software solutions that
enable users to adapt or upgrade their source code to their
needs.
First step of data processing is focusing the radar
images to obtain the single look complex images (SLC).
Single look Complex data retains the phase and amplitude
information of the original SAR data. They are also
corrected for satellite reception errors and include a
latitude/longitude positional information. In addition,
SLC data retains the optimum resolution and is suitable
for interferometric processing [20]. Usually that step can
be done in the scientific package ROI-PAC (Repeat Orbit
Interferometry Package) developed at the Jet Propulsion
Laboratory (JPL) on the Caltech (California Institute of
Technology) university, but in the research that step is
skipped as ESA directly provided SLC radar data.
Furthermore, a process of interferogram creation is
conducted in the scientific software package called
DORIS (Delft object-oriented radar interferometric
software) developed at the TU Delft in Netherland. In the
process, DORIS also uses additional software packages,
for modeling the path of satellite orbit from precise orbit
information (Getorb, TU Delft) and for phase unwrapping
process (SNAPHU, Stanford). Phase unwrapping is the
reconstruction process of the phase difference in radar
data by adding the correct integer multiple of 2π to the
interferometric fringes [21]. Final phase of processing is
conducted in the scientific software StaMPS/MTI
(Stanford method for persistant Scatterers/Multi Temporal
InSAR) developed at the Stanford University and later on
upgraded at the universities: Iceland, Delft and Leeds.
StaMPS connects all previous processing phases by
taking the final products from the aforementioned
software packages and applying the PS algorithm on the
unwrapped interferograms. Result is the deformation map
that shows the ground displacements in Line-of-Sight
direction (LOS).
Other software packages used in the data processing
are: ADORE (TU Delft) used for a graphical visualization
of relationship between master and slave radar images,
SNAP (Sentinel Application Platform) developed by ESA
used in the research for obtaining the information about
signal polarization, wavelength, date and time of
acquisitions, image resolution and other information
related to radar images, and Matlab software used as a
platform for StaMPS code and for visualization of final
products. Process of radar data processing is shown in
Fig. 9.
Figure 9 Process of radar data processing [22]
It is important to have the right folder hierarchy
according to the source code in software packages.
Therfore, before the data processing explained in the
aforementioned software packages, the necessary step is
to arrange the radar data by grouping them according to
the acquisition track (265 and 272) and putting the groups
and other relevant data (precise orbit, DEM) in a separate
folders. Afterwards, it is essential to select the right
master images for the interferometric processing in both
groups. Master image is the referent image in the radar
data stack, where all other images are called slaves and
they are co-registered to the master. That step is carried
out with an assistance of software packages SNAP and
ADORE. The co-registration step is fundamental in
interferogram generation, as it ensures that each ground
target contributes to the same (range, azimuth) pixel in

B. Pribičević i dr. Primjena satelitskih tehnologija u istraživanju geodinamičkih pomaka na širem zagrebačkom području
Tehnički vjesnik 24, 2(2017), 503-512 509
both the master and the slave image [21]. Co-registration
is performed with DORIS in two steps; coarse co-
registration and fine co-registration. Coarse co-
registration is based on the orbits of slaves and master,
and it is computed with an accuracy of about 30 pixels.
Result is the coarse offset within a few pixels explained
with a simple equation:
),()()( ms p,loffsetp,lPp,lP +=
(3)
Where Ps is a slave image, Pm is a master image, l an p
are image coordinates (line, pixel). The obtained offset
values are then used as initial parameters for the fine co-
registration, which is estimated by computing the
correlation of the magnitude images for shifts at pixel
level [23]. Fine co-registered images are then used as an
input for interferogram generation. Generation of the
interferogram requires the pixel-to-pixel computation of
Hermitian product of two co-registered images shown in
Eq. (4) [21]
,
*
S
MI ×=
(4)
where M and S refer to the master and slave images.
Afterwards, on each generated interferogram the phase
unwrapping process is performed and the results are used
as an input for StaMPS processing. The final step of data
processing is an application of PS algorithm on the data
stack through which the atmospheric and orbital errors are
eliminated from the data. Obtained results are the velocity
maps (mm per year) in LOS direction. Furthermore, LOS
displacements are displacement vectors regarding to the
satellite imaging geometry, hence, they are showing the
movement towards or away from the satellite under a
certain inclination angle. Therefore, LOS displacement
values consist of both horizontal and vertical surface
displacements. Number of PS points representing the
surface velocity in the AOI per year are: 156898 points
from track 265 and 138086 points from track 272.
Figure 10 The velocity map of the track 265
Velocity maps are shown in Figs. 10 and 12, and
related maps of standard deviations in Figs. 11 and 13
with the color bar scale from −10 to 10 mm per year.
Furthermore, satellite orbit direction is depicted with the
black arrow and LOS direction with the red arrow in the
following figures.
Positive velocity values (blue color on the color bar)
shown in figures are showing the movement towards the
satellite. The area of 200 meters around the asterisk has
been considered to be without surface deformations and
as such it is taken as the reference point.
Figure 11 Track 265's map of the standard deviation of the velocity
model
Figure 12 The velocity map of the track 272
The detail analysis of obtained PS results has shown
that the track 265 gave more reliable results than the track
272, which is confirmed by the maps of standard
deviations (Figs. 11 and 13) where the maximum standard
deviation of track 265 result is 3,9 mm per year and
standard deviation of track 272 result is 8 mm per year.
Reasons can be found in the following, in the data
processing of the track 265 more images were used than
in track 272 and there were better weather conditions
during the SAR acquisitions. The several SAR

The application of satellite technology in the study of geodynamic movements in the wider Zagreb area B. Pribičević et al.
510 Technical Gazette 24, 2(2017), 503-512
acquisitions of the track 272 were accompanied with rainy
weather which is reflected in the obtained results.
Obtained results revealed that the average range of
surface velocities in the wider Zagreb area is from –2 to 2
mm per year (142230 of 156898 PS points, 90 %) (Fig.
14) on the velocity map of the track 265 and from −4 to 4
mm per year (129 799 of 138 086 pS points, 94 %) on the
velocity map of the track 272 (Fig. 15).
Figure 13 Track 272‘s map of the standard deviation of the velocity
model
Figure 14 Velocity value distribution on PS points of tracks 265
Figure 15 Velocity value distribution on PS points of tracks 272
Detail statistics of the obtained results are shown in
Tab. 4. Final results after elimination of gross errors in the
results are shown in Tab. 5.
4 The Combination of GPS and InSAR results
The combination of GPS and InSAR results has
provided a more detail insight in ongoing ground surface
deformations in the wider Zagreb area, for the period
2004÷2009. Data gathered through GPS campaigns has
ensured the precise determination of horizontal and
vertical ground displacements on the discrete points,
whereas InSAR provided information of surface
displacements on more than 100.000 points but in the
LOS direction. GPS data has ensured the referent (more
accurate) velocity model of the AOI and InSAR provided
better coverage of the area, especially in urban parts of
the area. Therefore, with the combination of these two
techniques, the more accurate and detailed investigation
of ongoing geodynamic processes in the area is enabled.
The result of the combination of the two independent
geodetic techniques for the determination of ground
displacements in the AOI is depicted in Fig. 15.
In the combination used there is only InSAR result
from the track 265 because it is more reliable than the
result from the track 272. Due to the enormous amount of
points, the result is shown in Fig. 16 as a surface map.
Furthermore, it can be concluded that the InSAR results
provided a very dense spatial information about the
surface displacements in the urban parts of the area,
whereas the highly vegetated areas (Medvednica
Mountain) are exposed to interpolation due to the lack of
data.
Table 4 Detail statistics of the obtained PS results
Statistical parameters
T265
(velocity mm/year)
T265
(std of velocity mm/year)
T272
(velocity mm/year)
T272
(std of velocity mm/year)
Min
−15,24
0,21
−18,67
0,38
1st Qu
−0,65
0,60
−0,45
1,53
Median
0,01
0,76
0,73
1,98
Mean
1,07
0,83
0,75
2,24
3rd Qu.
0,72
0,97
1,85
2,71
Max
13,05
3,91
23,98
9,91

B. Pribičević i dr. Primjena satelitskih tehnologija u istraživanju geodinamičkih pomaka na širem zagrebačkom području
Tehnički vjesnik 24, 2(2017), 503-512 511
Table 5 Final PS results after elimination of gross errors in the results
Track id. (velocity map)
Number of PS
points
Min
(mm/year)
1st Qu.
(mm/year)
Median
(mm/year)
Mean
(mm/year)
3rd Qu.
(mm/year)
Max
(mm/year)
T265 (max std of 1
mm/per year)
121490 −6,82 −0,65 −0,03 −0,01 0,60 9,42
T272 ( max std of 2,5
mm/per year)
95530 −18,67 −0,48 0,55 0,39 1,35 14,24
Figure 16 Combination of InSAR and GPS results of ground surface deformations in the wider Zagreb area
5 Conclusion
The geodynamic research in the wider Zagreb area
started in 1997 with the project "Geodynamic GPS
Network of the City of Zagreb" and, since then, 10 GPS
campaigns have been conducted on the network points
with the aim to determine the geodynamically induced
ground deformations in the area. The research has
revealed so far that the average velocity values on the
network points are 1,72 mm per year in horizontal
direction and 5,17 mm per year in vertical direction.
The next logical step in the research was to enhance
the aforementioned research with InSAR (Interferometric
Synthetic Aperture Radar) technique in order to obtain a
more detail insight of ongoing ground deformations. That
was enabled by European Space Agency (ESA) in 2015,
when they provided 40 Enivsat ASAR radar images. The
radar data was processed with an advanced multi-stacking
Persistent Scatteres (PS) technique in StaMPS scientific
software package. PS algorithm has shown to be
extremely efficient in urban parts of the research area,
where a highly dense and reliable surface velocity model
is obtained. The model represents surface displacements
mainly in the vertical direction, which can be useful for
monitoring stability of buildings in the city.
Final deliverables from the InSAR data processing
were two surface velocity maps showing the average
surface displacements per year, in Line-of-Sight direction.
Obtained surface displacements are determined on the
156898 PS points from track 265 and 138086 PS points
from track 272. After a detailed analysis of the surface
velocity maps it is evident that the ranges of surface
displacements from the InSAR data are: –2 to 2 mm per
year on the velocity map of track 265 and from –4 to 4
mm per year on the velocity map of track 272.
The utilization of InSAR data in the research was
conducted for the period from to 2004 to 2009, due to the
good temporal coverage of GPS campaigns. That ensured
the possibility to combine these two independent geodetic
techniques on a reliable level. GPS data ensured the more
accurate velocity model of the area determined from 40
discrete network points, whereas InSAR data provided a
better coverage of the area, especially in urban parts of
area. The combination of these two techniques gave a
more accurate and detailed insight into ongoing surface
and geodynamic processes in the wider Zagreb area.
Acknowledgments
We would like to thank the European Space Agency
(ESA) for their support and for the data supplied for this
research. We would also like to acknowledge the
contribution of the universities where the software
packages used in the research were developed.
Furthermore, we appreciate the free accessibility of the
aforementioned software packages.

The application of satellite technology in the study of geodynamic movements in the wider Zagreb area B. Pribičević et al.
512 Technical Gazette 24, 2(2017), 503-512
7 References
[1] Pribičević, B.; Medak, D.; Đapo, A. Densification of the
Zagreb Geodynamic Network in the area of northeast
Medvednica. // Geodetski list. 61(84), 4(2007), pp. 247-
258.
[2] Pribičevic, B.; Đapo, A.; Medak, D. Geodetic-Geologic
Research on Wider Zagreb area based on Geodynamic
Network of the City of Zagreb. // Geodetski list. 65(88),
(2011), pp. 1-19.
[3] Pinter, N; Grenerczy, G.; Webber, J.; Stein, S.; Medak, D.
(Eds.) The Adria Microplate: GPS Geodesy. // Tectonics
and Hazards, Vol. 61. Veszprem, Hungary: Springer,
(2004).
[4] van Gelder, I. E.; Matenco, L.; Willingshofer, E.;
Tomljenović, B.; Andriessen, P. A. M.; Ducea, M. N.;
Beniest, A.; Gruic, A. The tectonic evolution of a critical
segment of the Dinarides-Alps connection: Kinematic and
geochronological inferences from the Medvednica
Mountains, NE Croatia. // Tectonics. 34, (2015). DOI:
10.1002/2015tc003937
[5] Matoš, B.; Tomljenović, B.; Trenc, N. Identification of
tectonically active areas using DEM: a quantitative
morphometric analysis of Mt. Medvednica, NW Croatia. //
Geological quarterly. 58, 1(2014), pp. 51-70. DOI:
10.7306/gq.1130
[6] Herak, D.; Herak, M.; Tomljenović, B. Seismicity and
earthquake focal mechanisms in North-Western Croatia. //
Tectonophysics. 485, (2009), pp. 212-220. DOI:
10.1016/j.tecto.2008.12.005
[7] Herak, M.; Allegretti, I.; Herak, D.; Ivančić, I.; Kuk, K.;
Marie, K.; Markušić, S.; Sović, I. Seismic hazard maps of
Croatia. // Geophysical Challenges of the 21st century
Zagreb, (2011), (poster) Zagreb, Croatia.
[8] Đapo, A.; Pribicević, B.; Medak, D.; Prelogović, E.
Correlation between Geodetic and Geological Models in the
Geodynamic Network of the City of Zagreb. // Reports on
Geodesy. 86, (2009), pp. 115-122.
[9] Herring, T.; Davis, J.; Shapiro, I. Geodesy by radio
astronomy: the application of Kalman filtering to Very
Long Baseline Interferometry. // J. Geophys. Res. 95,
(1990), pp. 12561-12581. DOI: 10.1029/JB095iB08p12561
[10] Dong, D.; Herring, T.; King, R. Estimating regional
deformation from a combination of space and terrestrial
geodetic data. // Journal of Geodesy. 72, 4(1998), pp. 200-
214. DOI: 10.1007/s001900050161
[11] Herring, T. A.; King, R. W.; Floyd, M. A.; McClusky S. C.
GAMIT - GPS Analysis at MIT, Reference Manual 10.6.
Department of Earth, Atmospheric, and Planetary Sciences
Massachusetts Institute of Technology. 2015.
[12] Reilinger, R.; McClusky, S.; Vernant, P.; Lawrence, S.;
Ergintav, S.; Cakmak, R.; Ozener, H.; Kadirov, F.; Guliev,
I.; Stepanyan, R.; Nadariya, M.; Hahubia, G.; Mahmoud, S.
K.; Ar-Rajehi, S. A.; Paradissis, D.; Al-Aydrus, A.;
Prilepin, M.; Guseva, T.; Evren, E.; Dmitrotsa, A.; Filikov,
S. V.; Gomez, F.; Al-Ghazzi, R.; Gebran K. GPS
Constraints on Continental Deformation in the Africa-
Arabia-Eurasia Continental Collision Zone and implications
for the Dynamics of Plate Interactions. // Journal of
Geophysical Research. 111 (B05411), (2006). DOI:
10.1029/2005jb004051
[13] McClusky, S.; Balassanian, S.; Barka, A.; Demir, C.;
Ergintav, S.; Georgiev, I.; Gurkan, O.; Hamburger, M.;
Hurst, K.; Kahle, H.; Kastens, K.; Kekelidze, G.; King, R.;
Kotzev, V.; Lenk, O.; Mahmoud, S.; Mishin, A.; Nadariya,
M.; Ouzounis, A.; Paradissis, D.; Peter, Y.; Prilepin, M.;
Reilinger, R.; Sanli, I.; Seeger, H.; Tealeb, A.; Toksöz, M.
N.; Veis, G. Global Positioning System constraints on plate
kinematics and dynamics in the eastern Mediterranean and
Caucasus. // Journal of Geophysical Research. 105, (2000).
DOI: 10.1029/1999jb900351
[14] Davies, P.; Blewitt, G. Methodology for global geodetic
time series estimation: A new tool for geodynamics. //
Journ. Geophys. Res. 105, 11(2000), pp. 11083-11100. DOI:
10.1029/2000JB900004
[15] Lavallee, D.; Blewitt, G.; Clarke, P. J.; Nurutdinov, K.;
Holt, W. E.; Kreemer, C.; Meertens, C. M.; Shiver, W. S.;
Stein, S.; Zerbini, S.; Bastos, L.; Kahle, H. G. GPSVEL
Project: Towards a Dense Global GPS Velocity Field. //
Proceedings of the International Association of Geodesy
Scientific Assembly, Budapest, (2001).
[16] Dawson, J. H. Satellite Radar Interferometry with
Application to the Observation of Surface Deformation in
Australia, The Australian National University (PhD thesis),
(10.2008)
[17] SarMAPSARscape, SAR Guidebook. http://
www.sarmap.ch/pdf/SAR-Guidebook.pdf (10.08.2016)
[18] RiskNET, Interferometric Synthetic Aperture Radar, an
Introduction for Users of InSAR Data. http://www.risknet-
alcotra.org/rna/allegati/insar-manual-20101008_468.pdf
(10.08.2016)
[19] U.S. Department of Transportation, Federal Highway
Administration InSAR Applications for Highway
Transportation Projects, Publ. No. FHWA-CFL/TD-06-002,
(04.2006)
[20] National University of Singapore, The Centre for Remote
Imaging, Sensing and Processing: Radarsat products.
https://crisp.nus.edu.sg/rsat/rsat_prod.html (09.08.2016)
[21] European Space Agency, InSAR processing: A practical
approach. http://www.esa.int/esapub/tm/tm19/TM-
19_ptB.pdf (09.08.2016)
[22] Bekaert, D. InSAR time series analysis of the 2006 slow
slip event on the Guerrero Subduction Zone, Mexico, Delft
University of Technology, Faculty of Aerospace
Engineering, Delft Institute of Earth Observation and Space
Systems (master thesis), (26.12.2010)
[23] TU Delft, DORIS user manual and technical documentation
http://doris.tudelft.nl/software/doris_v4.02.pdf (09.08.2016)
Authors’ addresses
Prof. Boško Pribičević, PhD
University of Zagreb, Faculty of Geodesy, Dept. of Geomatics,
Kaciceva 26, 10000 Zagreb, Croatia
E-mail: bpribic@geof.hr
Assist. Prof. Almin Đapo, PhD
University of Zagreb, Faculty of Geodesy, Dept. of Geomatics,
Kaciceva 26, 10000 Zagreb, Croatia
E-mail: adapo@geof.hr
Marin Govorčin, mag. ing. geod. et geoinf.
University of Zagreb, Faculty of Geodesy, Dept. of Geomatics,
Kaciceva 26, 10000 Zagreb, Croatia
E-mail: mgovorcin@geof.hr