Study on Ground Motions in Southwest Bulgaria based on in-Situ
and Satellite Data
Mila Atanasova-Zlatareva
1a
, Hristo Nikolov
2b
and Nikolay Dimitrov
1c
1
Department of Geodesy, National Institute of Geophysics Geodesy and Geography,
Bulgarian Academy of Sciences, Sofia, Bulgaria
2
Institute of Space Research and Technology, Bulgarian Academy of Sciences, Sofia, Bulgaria
Keywords: Ground Movements, SAR Data, GNSS, Crustal Deformation.
Abstract: In the last decades data from satellites are being used more frequently to study the ground movements. This
fact is evidenced by the increased number of research papers and projects using freely provided data by
space agencies such ESA (European Space Agency) and JAXA (Japan Aerospace Exploration Agency) and
increased revisiting time of the new instruments on-board satellites. Other reason for this increase are the
latest developments in processing methods such as PSI (Persistent Scatterer Interferometry) and even
increasing number of cloud processing options provided by universities and research centres. Nevertheless
the information obtained by this manner has some drawbacks for example moderate spatial resolution. This
is why in-situ data from precise GNSS (Global Navigation Satellite System) measurements are essential. In
this study the authors used both kinds of data to study one of the regions of Bulgaria which is recognized to
be highly prone to seismic and geological hazards namely the Southwest region. For this research two
sources of data have been used – SAR (Synthetic Aperture Radar) data from Sentinel-1 mission of ESA and
in-situ acquired contemporary and older GPS (Global Positioning System) data. As a result of SAR data
processing produced were interferometric images from ascending and descending orbits to decrease the
effect of the mountainous topography, while the results from the GNSS measurements were used for
verification.
a
https://orcid.org/0000-0002-3105-3266
b
https://orcid.org/0000-0001-5764-1499
c
https://orcid.org/0000-0002-5369-5921
1 INTRODUCTION
The main objective of this research is monitoring of
the ongoing geodynamic processes by
complementary use of SAR and GNSS data. GNSS
data from permanent and local geodetic networks are
used for validation of the SAR derived information
concerning the study area. The study will give
reliable data for ongoing risky geo-processes for the
region of the Southwest Bulgaria.
Geodynamic processes and seismic activity are
considered to be the prime driver of horizontal and
vertical movements of the Earth's crust in the Balkan
Peninsula. One proven method for continuous
monitoring of ground deformations is the use of data
from active radar remote sensing. These data are the
basis for the creation of interferometric images
(IFIs) for quantitatively assessment the registered
ground movements of the Earth’s surface within a
fixed time interval. For this research a set of IFIs
was created for the areas surrounding the city of
Sofia.
2 PREVIOUS GPS
MEASUREMENTS, RESULTS
AND ANALYSIS
In 1996 a GPS geodynamic network for long-term
monitoring of the crustal movements is established
in the region within a joint project with the
Massachusetts Institute of Technology (Kotzev et
al., 2006). All points are stabilised with metal bolts
Atanasova-Zlatareva, M., Nikolov, H. and Dimitrov, N.
Study on Ground Motions in Southwest Bulgaria based on in-Situ and Satellite Data.
DOI: 10.5220/0010503101570164
In Proceedings of the 7th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2021), pages 157-164
ISBN: 978-989-758-503-6
Copyright
c
2021 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
157
in rocks after geological study. The points have been
placed on the main structural blocks in the area - the
Vitosha, Plana and Lozenska mountains. The
network includes the first order point of Cherni Vrah
peak and the Geodesic Observatory of the NIGGG
“Plana”, thus covering the main tectonic structures
in the area.
A complete measurement of the entire network
with processing and analysis of the results has so far
been performed only in two epochs 1997 and 2000
(Kotzev et al, 2001). Data from GPS measurements
conducted on the points of the geodynamic network
have been processed in the scientific software
GAMIT/GLOBK, which is designed specifically for
processing of GNSS measurements in geodynamic
networks, calculates the velocities on the points by
applying modern methods of Kalman filtering.
During the period 2001-2019, the network was
expanded with the stabilization of new points and
GPS campaigns were periodically conducted to
determine crustal movements in the area. (Georgiev
et. al, 2011). The horizontal crustal movement and
seismicity have been studied by GPS analysis in
numerous studies (McClusky et al., 2000, Georgiev
et al., 2013, Kotzev et al, 2008).
The region of South Bulgaria, especially
Southwest Bulgaria and Rhodope Mountains and
Northern Greece is an active tectonic and
seismotectonic area in the South Balkans with
proved recent active tectonic structures and crustal
motions. Based on these alleged boundaries,
geological structures and faults, (Atanasova M.,
2014) formed the contours of blocks structuring of
this research area. The GPS velocity fields are
estimated and they are used to estimate the block
rotation motions. The region of Southwest (SW)
Bulgaria has the most pronounced tectonic and
seismotectonic activity on the whole Bulgarian
territory (Shanov et al. 2001). The investigated area
exhibits diverse relief structures, which are subjected
to horizontal and vertical movements of various
intensity. Southwest Bulgaria falls within a zone of
contemporary extension of the Earth’s crust with
complex interaction between horizontal and vertical
movements of the geological structures (Zagorchev,
2001).
Figure 1: A network for GNSS measurements consisting of 30 points.
GISTAM 2021 - 7th International Conference on Geographical Information Systems Theory, Applications and Management
158
GPS measurements of eight points in a region of
central western Bulgaria were carried out by
specialists from the Department of Geodesy at
National Institute of Geophysics, Geodesy and
Geography (NIGGG, Bulgarian Academy of
Science) in several periods between 1997 and 2006
is used in the estimation. Those points have been
stabilized on the main structural blocks in the area.
The results show that the general tendency of
movement of the points in the region of Central
West Bulgaria is in the south direction with respect
to the stable Eurasia, which is in line with the
extensive movement of southern Bulgaria and
northern Greece The existence of the recent tectonic
activity in the area is well justified (Dimitrov N.,
2019).
The area of Sofia is a moderately active
geodynamic area. However, it is exposed at a
significant geological hazard, based on the presence
of numerous active faults, seismicity, landslides,
rockfalls, etc. The results produced as output from
this research will contribute to the assessment of
natural risks and seismic hazards in the study area
and will have a positive impact for the sustainable
development of the region.
2.1 GNSS Network
At the end of 2019 the project "Monitoring of
geodynamic processes in the region of Sofia" was
launched by the Department of Geodesy, the
National Institute of Geophysics, Geodesy and
Geography at the Bulgarian Academy of Sciences,
funded by the Bulgarian National Science Fund
(Dimitrov et. al., 2020).
This project allowed for a new comprehensive
measurement of the geodynamic network (Fig. 1).
The geodetic GNSS measurements were performed
by four teams of specialists from the Department of
Geodesy of the National Institute of Geophysics,
Geodesy and Geography in the period from
13.07.2020 to 31.08.2020. CHC N71 receivers with
CHCA220GR antennas and V100 with HITV100
antennas were used. Point PLA1, is continuous
operating reference station for many years. To study
the modern movements of the Earth's crust, two-day
GNSS measurements of a network of 30 points were
performed. The measurements and analysis of the
results of the GNSS campaign in 2020 were
processed. The results of the processing and analysis
of the GNSS 2020 campaign show very good quality
(Dimitrov et al. 2020) and will be used for joint
processing with other, significant number of
previous and future GNSS measurements to obtain a
model of intraplate movements of the Earth's crust in
the study area.
3 SYNTHETIC APERTURE
RADAR INTERFEROMETRY
(InSAR)
Lately the interferometric approach was widely used
in geodetic studies. This method uses
interferometrically processed data from ground or
satellite based synthetic aperture radars. The aim of
the processing is to obtain the change of the phase
signal present in two radar images acquired at
different dates. Based on this difference information
on diverse type’s geophysical phenomena
earthquakes, landslides and rockfalls etc. can be
monitored.
It should be noted that the SAR instrument is an
active radar imaging system which usually is
mounted on flying platforms in order to map the
Earth’s’ surface. In this instrument the main element
is the SAR antenna which emits and receives back
the radar signals (Pribičević et al, 2017).
The acquired data form a radar image which is
composed by two different components of the
electromagnetic wave – the amplitude and the phase.
It should be taken into account that the wavelength
and the phase of radar signal are mutually correlated.
For use the interferometric method two radar images
(often named interferometric pair - IFP) must be
processed in conjunction and in the resulting image
the phase difference of the backscattered signal
between both measurements is placed. This
difference is directly related with the changes (if
any) that have occurred on the Earth’s surface in the
time interval between the two acquisitions. It also
has to be accounted that the displacements registered
by this manner are in the line-of-sight (LOS) of the
antenna and can’t be directly interpreted as motion
in horizontal or vertical plane. In order to do this
additional calculations are needed.
3.1 DInSAR Technique in Measuring
the Ground Motions in SW
Bulgaria
DInSAR (Differential Interferometric Synthetic
Aperture Radar) approach used to produce the
results in this study is composed of following stages.
The first one is the data preparatory stage at which
based on the expertise of the authors from the
available online archives with SAR data maintained
Study on Ground Motions in Southwest Bulgaria based on in-Situ and Satellite Data
159
Figure 2: Area of study covered by the SAR images of ESA’s mission Sentinel-1 ascending and descending orbits tracks
102, 7, 29 and 80.
by different institutions the needed ones have been
obtained. In this process the perpendicular baselines
between all possible IFPs were calculated and the
most suitable ones for further processing at next
stage were selected. Further auxiliary information
(digital maps and other data) concerning the
topography of the studied region was collected too.
This information was used to minimize the
shadowing effect of the Vitosha, Rila Mountains in
the DInSAR processing carried out on the next
processing stage which results in layover and
foreshortening as well. Other consideration that was
taken into account to mitigate the mentioned effects
was to process data from ascending and descending
orbits and co-registers them. At the end of this stage
we ended up with 48 Sentinel 1A/B scenes for the
period 2015-2020 (see Fig. 2).
3.2 DInSAR Processing Procedure
In the second stage of this research implemented
was the procedure for DInSAR processing of every
IFP and it is considered to be the most important to
obtain the final results.
At this stage only data in Single look Complex
(SLC) format were processed. In this format both
data for the amplitude and phase signals are
registered. Additionally those data free of platform
and orbital errors, have good positional information
for latitude/longitude, and at the same time provide
spatial resolution which is relevant for the objectives
of this research. For creation of a single IFP the SAR
data were processed by SNAP software provided by
ESA into which a well-established methodology for
DInSAR is easily realized (see Fig.3).
In this study a set of 32 IFPs was created for the
time period between April 2015 and November
2020. It has to be clarified that the produced IFPs
cover a four-month interval which essential to
minimize the negative effects of radar signal
attenuation and decrease of coherence caused by the
vegetation present in the studied areas. Other
constrain that was accounted at the beginning of
processing was to maximize the modelled coherence
for each IFP. During the creation of the
interferometric image of importance is to select the
DEM that will provide the best possible spatial
resolution of the final product (Nikolov et. al. 2017).
GISTAM 2021 - 7th International Conference on Geographical Information Systems Theory, Applications and Management
160
Since no better DEM was made available at the
moment of processing the 1arcsec SRTM raster was
used. To guarantee the high quality of the final IFPs
only precise orbital data for the co-registration of the
initial SLC images were used. This step is of
importance for single interferogram formation since
at it the registered amplitude and phase signals from
a single ground element in both range and azimuth
directions from both SCL images are matched.
Next step in the processing was to improve the
quality by high-pass filtering and multilooking to
produce square ground elements in the IFI. Since for
this research we narrowed the area of study it was
necessary to produce a smaller polygon by
subsetting the IFI. After this procedure the IFI can
be geocoded and exported to other software (e.g.
QGIS, Google Earth) for additional visualization and
analysis. The colours present in this IFI indicate the
degree of change of the phase signal. In case of
single smooth colour there is no phase change i.e. no
surface deformations are detected. In the areas
where “salt and paper” effect is visible it can be
concluded that the phase is highly decorrelated due
to low coherence in the single ground element. Only
for those areas in the IFI where fringes are present it
can be concluded that ground deformations were
reliable registered from phase signal change in LOS.
To obtain surface displacements in LOS in
metric units one more procedure is needed which is
known as phase unwrapping. In fact this is a process
of reconstruction of the phase difference by adding
the correct integer multiple of to the
interferometric fringes (Veci L. 2016). This
procedure for phase unwrapping can be performed
by several software products, but in this study we
used Snaphu module (Snaphu, 2020).
To study the ongoing geodynamic processes in
the region of SW Bulgaria created was a local
repository with SAR data from Sentinel-1 mission.
Using the above described method produced was a
set of IFIs at intervals of 4, 8 and 12 months and
when specific ground motion event had to be
studied. The validation of the information produced
from the IFIs was done using the data from the
GNSS network (see sec. 2 and Fig. 4). (Larsen Y. et
al., 2020)
The third and a final phase comprises the
analysis of the information produced from DInSAR
processing and is based on the produced deformation
maps in metric units for the corresponding period. It
needs to be underlined first that the registered
surface displacements are in the LOS of the antenna
and provide information if the points move toward
Figure 3: Radar data processing.
or away from it, and second they are relative to a
point with known deformation. One more practical
consideration when interpreting LOS deformations
is that the real three dimensional ground motion
(North, East, Up) is an estimation and should made
carefully.
Study on Ground Motions in Southwest Bulgaria based on in-Situ and Satellite Data
161
Figure 4: Sample IFI created from two images dated Nov 12 2016 (master) and Apr 5 2017 (slave).
On Fig. 5 an example of the occurred ground
displacements in LOS is presented. Those were
calculated from the unwrapped phase signal which is
color-coded. It can be noted that for the period of 4
months for 2016-2017 the calculated displacements
are in the range between 2 (uplift) and -25
(subsidence) cm. The resulting IFIs reveal that the
registered deformations are concentrated in some
areas exhibiting non-uniform pattern. From them a
map of the concentration of deformations of the
Earth's crust was created as well.
4 COMBINED USE OF GNSS AND
InSAR RESULTS
The combination of GPS/GNSS and DInSAR results
from the project "Monitoring of geodynamic
processes in the region of Sofia" will provide a more
detail insight in about the ongoing ground surface
deformations for the period 2015÷2023. The
information gathered through several GNSS
campaigns has ensured the precise determination of
horizontal and vertical ground displacements on the
discrete geodetic points while the information
produced after DInSAR reflects surface
GISTAM 2021 - 7th International Conference on Geographical Information Systems Theory, Applications and Management
162
Figure 5: Subset of displacement map created from two images: Nov/12/2016 (mast) and Apr/5/2017 (slave).
displacements for thousands of ground targets but
only in LOS direction. The velocity model obtained
for the GNSS points is of high accuracy and for this
reason is considered as referent while from DInSAR
processing much larger coverage was obtained.
Therefore, with the combination of these two
techniques, more accurate and detailed investigation
of the ongoing geodynamic processes in the area is
produced. (Pribičević et al, 2017). The final result of
the combined use of these two independent geodetic
techniques for the determination of the ground
displacements in the studied region of SW Bulgaria
is depicted in Fig. 5.
5 CONCLUSIONS
This research demonstrated the potential and
capability of radar satellite data and DInSAR time-
series technique to investigate and monitor the
ground-surface deformations, and also to measure
their variations in LOS with centimetre accuracy
over time using freely available data and software.
The DInSAR can be regarded as an attractive
technique and operational tool for geological hazard
tasks such as detection and monitoring of ground-
surface deformations. It needs to be underlined that
the surface deformations produced from satellite-
based data require precise calculations and are
complementary to the field measurements and could
direct them, but can’t replace them.
Final products from the DInSAR processing were
surface displacement maps showing the average
surface displacements in LOS direction. One more
advantage of the DInSAR results is that they
provided dense spatial information in non-vegetated
areas only while in the highly vegetated areas where
the coherence is low an interpolation was done. The
combination of GNSS and DInSAR increased the
quality of the information regarding the current
geodynamic processes and surface motions.
Study on Ground Motions in Southwest Bulgaria based on in-Situ and Satellite Data
163
ACKNOWLEDGEMENTS
This paper has been made available with the
financial support provided by National Science
Fund, Bulgaria, call identifier “Competition for
financial support of basic research projects 2019”
Project “Monitoring of geodynamic processes in the
area of Sofia”. Contract No KП-06-H 34/1.
REFERENCES
Atanasova–Zlatareva, M. 2014 Research of the
Horizontal Crustal Motions, Based on GPS Data for
the Territory of Bulgaria and the Balkans (7093)
Proceedings FIG Congress 2014 in Kuala Lumpur,
Malaysia 16-21 June 2014 ISBN 978-87-92853-21-9,
ISSN 2308-3441, p 11
Georgiev, I., D. Dimitrov, P. Briole, E. Botev,
2011Velocity field in Bulgaria and northern Greece
from GPS campaigns spanning 1993-2008.
Proceedings of 2nd INQUA-IGCP-567 International
Workshop on Active Tectonics, Earthquake Geology,
Archaeology and Engineering, Corinth, Greece, 2011,
54-56.
Georgiev, I., D. Dimitrov, E., Botev., 2013, Crustal
Motion Monitoring in Bulgaria and Surrounding
Regions by Permanent GPS Array. Proceedings of 7th
Congress of Balkan Geophysical Society Tirana,
Albania, 7-10 October 2013, 5 pp.
Dimitrov N., 2019, Deformation analysis in central west
Bulgaria using triangulation and GPS data, Reports on
Geodesy and Geoinformatics, 2019, Vol. 108, pp. 23–
26, DOI: 10.2478/rgg-2019-0009.
Dimitrov N., P. Danchev, I. Georgiev, A. Ivanov 2020.
Geodetic measurements of the geodynamic network in
the area of Sofia, processing and analysis of the
results. Proceeding XXX International symposium on
“Modern technologies, education and professional
practice in geodesy and related fields” 04-06
November 2020, Sofia, Bulgaia, CD, 12, 1-11, 2020,
ISSN: 2367-6051.
Dimitrov N, I. Georgiev, R. Nakov 2020, Monitoring of
geodynamic processes in the area around Sofia.
SGEM, Volume 20, Book 2.2, 20th International
Multidisciplinary Scientific GeoConference, SGEM,
18-24 August, 2020, Albena, Bulgaria, ISSN:1314-
2704, DOI:10.5593/sgem2020/2.2/s09.010, 79-86.
Kotzev, V., R. Nakov, B. C. Burchfiel, R. King, and R.
Reilinger (2001), GPS study of active tectonics in
Bulgaria: Results from 1996 to 1998, J. Geodyn., 31,
189–200.
Kotzev, V., R. Nakov, Tz. Georgiev, B.C. Burchfiel, R.W.
King. 2006, Crustal motion and strain accumulation in
western Bulgaria, Tectonophysics, 2006, 413, 127–
145.
Kotzev, V., R.W. King, B.C. Burchfiel, A. Todosov, B.
Nurce, R. Nakov. 2008, Crustal motion and strain
accumulation in the South Balkan Region Inferred
from GPS
Larsen Y. et al., 2020 European Ground Motion Service:
Service Implementation Plan and Product
Specification Document, Report to CLMS, ver. 1.01
(Jan 2020)
Veci L., 2016 TOPS Interferometry Tutorial,
http://sentinel1.s3.amazonaws.com/docs/S1TBX%20T
OPSAR%20Interferometry%20with%20Sentinel-
1%20 Tutorial.pdf,
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., 2000, Global
Positioning System constraints on plate kinematics
and dynamics in the eastern Mediterranean and
Caucasus. // Journal of Geophysical Research. 105,
doi: 10.1029/1999jb900351
Nikolov H., Atanasova M.. 2017, Influence of different
DEMs on the quality of the InSAR results – case study
over Bankya and Mirovo areas. Active and Passive
Microwave Remote Sensing for Environmental
Monitoring, edited by Claudia Notarnicola, Nazzareno
Pierdicca, Emanuele Santi, 10426, SPIE, 104260M-1-
104260M-8. https://doi.org/10.1117/12.2278393
Pribičević B., A. Dapo, M. Govorčin 2017 The application
of satellite technology in the study of geodynamic
movements in the wider Zagreb area, Technical
Gazette 24, 2(2017), 503-512, ISSN 1330-3651,
(Print), ISSN 1848-6339 (Online), DOI: 10.17559/TV-
20160817013320
Shanov, S., K. Kourtev, G. Nikolov, A. Boykova, B.
Rangelov 2001, Seismotectonic characteristics of the
western periphery of the Rhodope Mountain region.
Geologica Balcanica, 31, 1/2, pp.53-66.
SNAPHU: Statistical-Cost, Network-Flow Algorithm for
Phase Unwrapping https://web.stanford.edu/group/
radar/softwareandlinks/sw/snaphu/
Zagorchev I. 2001, Southwest Bulgaria, Geological
Guidebook, Bulg. Acad. Sci.
GISTAM 2021 - 7th International Conference on Geographical Information Systems Theory, Applications and Management
164