DOI QR코드

DOI QR Code

Surface Deformation Measurement of the 2020 Mw 6.4 Petrinja, Croatia Earthquake Using Sentinel-1 SAR Data

  • Received : 2021.02.03
  • Accepted : 2021.02.15
  • Published : 2021.02.26

Abstract

By the end of December 2020, an earthquake with Mw about 6.4 hit Sisak-Moslavina County, Croatia. The town of Petrinja was the most affected region with major power outage and many buildings collapsed. The damage also affected neighbor countries such as Bosnia and Herzegovina and Slovenia. As a light of this devastating event, a deformation map due to this earthquake could be generated by using remote sensing imagery from Sentinel-1 SAR data. InSAR could be used as deformation map but still affected with noise factor that could problematize the exact deformation value for further research. Thus in this study, 17 SAR data from Sentinel-1 satellite is used in order to generate the multi-temporal interferometry utilize Stanford Method for Persistent Scatterers (StaMPS). Mean deformation map that has been compensated from error factors such as atmospheric, topographic, temporal, and baseline errors are generated. Okada model then applied to the mean deformation result to generate the modeled earthquake, resulting the deformation is mostly dominated by strike-slip with 3 meter deformation as right lateral strike-slip. The Okada sources are having 11.63 km in length, 2.45 km in width, and 5.46 km in depth with the dip angle are about 84.47° and strike angle are about 142.88° from the north direction. The results from this modeling can be used as learning material to understand the seismic activity in the latest 2020 Petrinja, Croatia Earthquake.

Keywords

1.Introduction

Earthquakes are one of the types of natural disasters that are considered as one of the most destructive event on earth surfaces. It usually causes deformation on the earth’s surface. Over 3400 earthquake happened around the world during 1980-1990 and 2016 with 1260 of them are continental events with magnitude around Mw5.5 (Storchak et al., 2015). With the number of earthquakes that occur, it is important to observe and mitigate this disaster and measure its deformations in order to determine how massive the damage can be.

Nowadays, there are a lot of instrument that could detect the surface deformation due to earthquake event. One of them is Global Navigation Satellite Systems (GNSS). This instrument could get the coordinate point position with a high degree of accuracy that could measure surface deformation during earthquake event. It is already been applied in some areas such as, the 2015 Mw7.8 Nepal earthquake (Geng et al., 2016), the 2016 Mw6.6 Vettore, Italy earthquake (Wilkinson et al., 2017), and the 2011 Mw9.0 Tohoku-Oki, Japan earthquake (Tobita, 2016). However, the areas affected by the earthquake have a wide coverage making the observation using GNSS are not efficient. The deformation information generated from GNSS observations is limited to the point of view. Thus, space-borne remote sensing techniques by using synthetic aperture radar (SAR) could become an alternative solution. Interferometric synthetic aperture radar (InSAR) is one of the techniques of remote sensing technology that can be used to detect surface deformation caused by earthquakes (Xu, Sandwell, and Smith-Konter, 2020). InSAR are also could be used in various areas such as volcano (Achmad et al., 2020) and detecting land subsidence in cities (Fadhillah, Achmad, and Lee, 2020; Hakim et al., 2020).

At 29 December 2020, an earthquake with Mw6.4 occurred and hit central Croatia. The epicenter of the earthquake is located around 5 km southwest of the town of Petrinja. This earthquake devastated the town with a lot of building damaged, at least seven people dead, and many were wounded. The earthquake is also felt in the neighbor country such as Slovenia and Bosnia and Herzegovina.

By using InSAR, the surface deformation that formed during the earthquake event could be detected and measured. But the result based on one pair of InSAR could still including some error factors such as topographic error, orbital error, atmospheric effect factor error, and some other noise factor error (Lee et al., 2012). Thus, with the abundance of current SAR satellite data available these days, to overcome that problem in InSAR generation for surface deformation analysis, new advanced algorithm using a lot of InSAR data are also developed; a multi-temporal interferogram framework algorithm. This method will generate a high coherent target that could overcome most of the difficulties in unwrapping phases from interferometry process, Permanent/Persistent Scatterers Interferometry (PSI) is one of the method developed (Ferretti et al., 2001; Hakim et al., 2020; Hooper et al., 2007; Nur, et al., 2020). This process also generate the time- series deformation information in the region where surface deformation occurred and provide much better information for surface deformation purpose.

Therefore, in this study we applied Stanford Method for Persistent Scatterer (StaMPS) algorithm (Hooper et al, 2007; Hooper et al., 2012) in the latest Mw6.4 earthquake event occurred in Petrinja, Croatia in December 2020 that also affected neighbor country such as Slovenia and Bosnia and Herzegovina. The mean deformation map that has been generated from the StaMPS algorithm then were processed by Okada model to generate the best-fitted model in order to understand the deformation that occurred in this earthquake event.

2. Study Area

At 29 December 2020, an earthquake with Mw 6.4 occurred and hit central Croatia. The InSAR images showing the surface deformation is shown in Fig. 1.

OGCSBN_2021_v37n1_139_f0001.png 이미지

Fig. 1. Petrinja Town location (a) in Croatia (marked with red box) (b) by Sentinel-2 imagery at 20 August 2020 (c) aftermath of the 2020 Mw 6.4 Earthquake in Petrinja (marked with red box in Fig. 1(b)) (Ministry of Internal Affairs of Republic of Croatia, 2020).

The epicenter of the earthquake is located around 5 km southwest of Petrinja as seen in Fig. 1(b). Petrinja is a town in central Croatia that located near Sisak in the historic region of Banovina at 45°26′35.3″N and 16°16′39.4″E. The geological map showing that the epicenter of the earthquake at 29 December 2020 are close with two fault underground (Pokupsko Fault and Petrinja Fault) of the broader area of Sisak, Petrinja, and Glina town (Pikija, 1987). It is dominated by terrace sediments with silt, sand gravel on Petrinja area while nearby the epicenter is mostly sand, gravel, clay, sandstone, coal, and conglomerates. Earthquakes are common event in Croatia as in its region (Circum- Adriatic region) are mostly resulted from the movement of the African tectonic plate that move towards the relatively stable Eurasian plate about 160 million years ago (Ustaszewski et al., 2014).

According to Croatian Geological Survey (HGI- CGS) (Croatian Geological Survey, 2021), most earthquakes that occurred in Croatia, including the wider epicentral area of Pokupsko-Petrinja-Sisak is driven by the continuous movement of the Adriatic lithospheric microplate (Adria) to the north. In the upper parts of earth’s crust, the strain occurred at the contact of Dinarides and Pannonian Basin that could reactivate individual fault when the strain reaches a critical level. It will make a sudden movement of blocks of the crust measuring from several hundred to thousands of cubic kilometers that released a massive amount of energy, making the earthquakes occurred.

3. Methods

To generate surface deformation of earthquake that occurred in Petrinja, Croatia. We used data from Sentinel-1 SAR data (C-band) with 5.5 cm wavelength provided by the European Space Agency (ESA). It acquired data from September 2020 to December 2020 in ascending track. We use only four month acquisition with 6 days of revisit time with total of 17 datasets. This amount of datasets are still acceptable due to that the minimum images needed to perform a C-band persistent scatterer analysis is about 15-20 data (Crosetto et al., 2016). In addition, the selection of using 16 interferograms is also in order to minimize the temporal decorrelation that resulted from the interferometric process from single master date scene. In earthquake event, the surface deformation is mostly occurred once and rarely showing surface deformation before the earthquake event. Thus, 16 interferogram is enough for analysis in this study. It is also enough to be processed in StaMPS in this study in order to generate the result faster but still could give more deformation information from the time-series interferometry point of view. The data that used are shown in Table 1 below.

Table 1. Acquisition date of Sentinel-1 used in this study. Master date are marked with bold color

OGCSBN_2021_v37n1_139_t0001.png 이미지

1) Interferometric SAR generation

Two-pass InSAR approach is applied to SAR datasets. Those images are being co-registered at sub-pixel accuracy to form the interferometric pair. The co-registered images cropped to generate the image that contains only the study area to shorten the time required for further process. The topographic phases are subtracted from Shuttle Radar Topography Mission (SRTM) Global one arcsec with a 30 meter resolution to generate Differential Interferometry (DInSAR) phase where only the deformation phase remains. Generation of DInSAR in this process is during earthquake event and also reduces the temporal decorrelation as seen in Fig. 2 below.

OGCSBN_2021_v37n1_139_f0002.png 이미지

Fig. 2. Baseline graph of StaMPS process in this study.

2) StaMPS algorithm

The interferometry that already been generated then through a process in the StaMPS algorithm as seen in Fig. 3 below.

OGCSBN_2021_v37n1_139_f0003.png 이미지

Fig. 3. Flowchart of StaMPS processing (modified from (Sousa et al., 2011)).

This method is having advantages in natural terrains because it is assumed that the deformation and the interferometric phases are spatially correlated. These algorithms also do not rely on the assumption about ground deformation’s temporal nature. A total of 16 interferograms with master date at 24 December 2020 are generated as seen in Fig. 2 above. We select master image based on four terms, dependent on temporal baseline,spatial perpendicular baseline,Doppler centroid baseline and thermal noise (Hooper, 2006). From that term, the image on 24 December 2020 was chosen as master images because this date could maximize the sum correlation of all the interferogram. All of the interferogram data generated then loaded into the StaMPS module. StaMPS algorithm will select the persistent scatterer by using phase characteristics suitable for finding the low amplitude natural target with phase stability as seen in equation 1 below (Hooper, 2008; Hooper et al., 2012; Hooper et al., 2004).

\(\mathrm{D}_{\mathrm{A}}=\frac{\sigma_{\mathrm{A}}}{\mu_{\mathrm{A}}}\)       (1)

With DA indicated the amplitude dispersion value, while σA and μA are standard deviation and mean of amplitude values, respectively. In this algorithm, points with DA of less than 0.4 are selected as persistent scatterer point candidates(PSC). StaMPS then will reselect the PS candidates by employing an iterative model where coherence of the PSCs are calculated using other nearby PSCand points with low coherence values are rejected (Osmanoğlu et al., 2016).

After the PS selected, then PS weeding process is started by dropping the PS that too close to each other based on their proximity and the smoothness of their deformation in time. The remaining selected PS then corrected for spatially-uncorrelated DEM error. After the DEM error correction is done, the refined PS point phases are being unwrapped. The unwrapping process is running on the module after all refined PS pointfrom several iteration was completed. StaMPS is using 3-D unwrapping method that consists of estimating the probability density function in space dimension after temporal unwrapping, interpolating the point to a regular grid, and then define the cost-maps to optimize spatial unwrapping by SNAPHU algorithm (Chen and Zebker, 2002; Hooper, 2010). After the unwrapping, last step is estimating the total DEM error from look angle SAR data as well as master atmosphere and orbital error phases before the deformation time series and mean deformation velocity is estimated over the analyzed persistent scatterer.

3) Okada Modeling

The mean deformation velocity that been generated from StaMPS algorithm then were modeled by using Okada equation to calculate an analytical solution for surface deformation due to shear and tensile faults in an elastic half space (Okada, 1985). Okada model is widely used to simulate ground deformation by some local perturbation such astectonic faults because of the earthquake. It is also being used as model equation for magmatic intrusion from volcanic dykes.

In this study, the Okada model is carried out with a two-step approach (Wright et al.,2003); a nonlinear optimization based on the Levemberg-Marquardt algorithm (Marquardt, 1963) and also linear inversion with Monte Carlo restarts to avoid local minima (Wright et al., 1999). The inversion for the fault parameters from surface deformation data can be solved by computing a forward model of the surface deformation repeatedly as it is a non-linear inverse problem, then adjusting the fault parameters until the misfit between modeled and the observed deformation is decreased (Weston et al., 2012). This problem can be solved with a non-linear optimization algorithm such as the Levemberg-Marquardt algorithm to vary the source parameters systematically to find the modeled result with the best fit to the observed data. The modeled deformation that been generated, then being converted into a satellite line of sight.We choose to use 100 time iteration and then select the best-fitted model from the lowest value of the root mean square error (RMSE) from 100 simulations explained in equation 2 as follow (Pawluszek-Filipiak and Borkowski, 2020):

\(\mathrm{RMSE}=\sqrt{\frac{\sum_{i=0}^{n}\left(f_{o}-f_{m}\right)^{2}}{n}}\)       (2)

with fo is the result observed from StaMPS, fm is modeled result, and n is the number of processed data. This method has been widely used to evaluate the prediction with observed results in classification purpose (Dodangeh et al. 2020).

4. Results

About 1,001,124 PS point is collected from StaMPS algorithm. The unwrapped result afterselecting1,001,124 PS point are shown in Fig. 4 below. This result shown that before the earthquake happened (before 24 December 2020), no significant deformation occurred.

OGCSBN_2021_v37n1_139_f0004.png 이미지

Fig. 4. Unwrapped InSAR of available SAR data processed in StaMPS. Master images date is 24 December 2020. All images are overlaid with SAR mean amplitude images.

The generated unwrapped InSAR then being estimated over all the analyzed persistent scatterer, generating the mean deformation velocity maps as well as the Okada model of the 2020 Mw 6.4 Petrinja, Croatia Earthquake was also successfully generated by using two-step approaches, a nonlinear and linear optimization as shown in Fig. 5 below.

OGCSBN_2021_v37n1_139_f0005.png 이미지

Fig. 5. Result of line-of-sight projected deformation maps from (a) observed InSAR pair generated from 20201224 and 20201230 pair data (b) observed data from StaMPS result (c) modeled result (d) residual result.

It is shown that the mean deformation velocity occurred during the 2020 Petrinja Earthquake are about 30 radians or equal with 13.1 cm per 4 months. On the western part of the earthquake are showing red color scale which means that the deformation detected from SAR data are about 13 cm toward to the satellite LOS direction (uplift) while on the eastern part are showing deformation about 12.3 cm away to the satellite LOS direction (subsidence).

Meanwhile, the time series deformation graph are shown that no significant deformation detected during pre-seismic event as seen in Fig. 6 below.

OGCSBN_2021_v37n1_139_f0006.png 이미지

Fig. 6. Time-series deformation graph of the study area.

The closest GPS station was located about 70 km east of the epicenter (station BJEL) in Blejova city (Bruyninx et al., 2019). This station is beyond SAR data image subsets generated in this study. Therefore apart to maximize the modeling processing and also due to the location is quite far away from the epicenter, we choose another point location for analysis. P1 and P2 location were chosen on the basis of the area showing the largest mean deformation velocity compared to other areas as seen in Fig. 5(b). The P1 point shows that after the earthquake, the deformation are about 15.3 cm as uplift direction while in P2 point is showing deformation about 11.4 cm as subsidence direction.

The best-fitted model was selected among the 100 times simulations shown in Fig. 5(c). We select the best-fitted model from the lowest value of covariance. The residual result is generated by subtracting the original data from StaMPS with modeled results as seen in Fig. 5(d). The source parameters generated from Okada model were shown in Table 2 below.

Table 2. Source Parameter of Okada model from StaMPS result

OGCSBN_2021_v37n1_139_t0002.png 이미지

The predicted resultfromOkadamodel is shows that the deformation is mostly dominated by strike-slip with 3 m deformation. The geometry of the fault predicted by Okada is about 11.63 km of length and 2.45 km of width with depth of about 5.46 km beneath the surface. The strike direction and dip direction from the predicted model are showing 142.88° and 84.47°, respectively.

5. Discussion

The deformation detected by the StaMPS method shows that there is no significant value at points 1 and 2 before the earthquake occurred as seen in Fig. 6. Although the InSARresult of the 2020 Mw 6.4 Petrinja, Croatia earthquake is not the first time result generated, there have been some improvement presented in this study by accommodate the time-series interferometry as well as improve the modeling result by using mean deformation result value from StaMPS. The StaMPS algorithm has succeeded in showing that PS can be found in the area around the earthquake area even though the area is quite dominated by forest area. This can occur due to the use of Sentinel-1 data with a span of 6 days which causes the number of coherent points to be generated due to the low temporal decorrelation. Apart from the low temporal decorrelation factor, the PS point selection technique used by the StaMPS algorithm by looking for stable phase characteristics in space and time regardless of their amplitude (Hooper et al., 2004) also makes StaMPS can be used to measure deformation in non-urban areas (Osmanoğlu et al.,. 2016). Thus, even though the points are called persistent scatterers, the term only agrees in regards to the phase characteristics and not in the process to find the points. By using this method, other error factors such as atmospheric effect, baseline error, temporal decoration error, topographic error(DEM error) can be suppressed (Fig. 5(a), Fig. 5(b)); this also can be seen from the low deformation range value before the earthquake occurred in Fig. 6 above.

The best fitted model result using Okada equation is showing that the source parameter in Petrinja Earthquake is mostly dominated by strike-slip with 3 meter deformation generated. There are also some other slip deformation such as dip-slip with 6.8 cm and opening with 26.8 cm generated. The residual phase value is generated after subtracting the observed data from StaMPS mean deformation velocity with the modeled result. It is shows that the remaining phases are having maximum value of about 5 radians or equal with 2 cm deformation. In addition, the RMSE calculation results shows that the value obtained in this modeling process are about 1.4 radians or equal with 0.62 cm. The use mean deformation result is showing improvement in the model generated. The mean deformation result that include all interferogram generated (pre-seismic and co-seismic event) from StaMPS algorithm (Fig. 5(b)) could suppress the error factor generated (i.e. atmospheric error, topographic error, and orbital error) compared by using the coseismic event only (Fig. 5(a)). This residual result value is suppressed much better from the acceptable as phase error factor generated in one InSAR pair process that are ranging about 5 cm generated from atmospheric effect, topographic error, orbital error, and temporal decorrelation error (Lee et al., 2012).

In comparison with preliminary GPS station result analysis, the GPS displacements are showing very small value at the horizontal components about 1 mm, except for station that located to the NE of the epicenter where a motion of ~2.1 mm towards the south was detected. In other word, this motion is in agreement with rupture kinematics(Ganas et al., 2021).This small deformation value detected by GPS data due to the location of GPS station ranging for about 108 209 km from earthquake epicenter.The GPS data are also could not get the evidence for any pre-seismic deformation, which is showing agreement with our time-series interferometry analysis. The model generated from two-step approach is also shows an agreement with preliminary report from Croatian Geological Survey (HGI-CGS) by compiling information from geological maps, numerous field data published in the media, and available seismological data that there is an evident that there are an intersection of longitudinal and transverse fault to the strike in the area (Croatian Geological Survey, 2021).Those faults systems consist of multiple faults with strike-slip movement are Petrinja fault and Pokupsko fault. Further research and field research data and information are still in process to be compiled by the Croatian Geological Survey.

6. Conclusion

The earthquake that occurred on December 2020 in Petrinja, Croatia had a massive impact not only for the Croatian citizen, but also the neighborhood country such as Slovenia, and Bosnia and Herzegovina. This work has shown that PSI-StaMPS showing better information regarding deformation occurred in the 2020 Mw 6.4 Petrinja, Croatia Earthquake with the mean deformation velocity are about 30 radians or equal with 13.1 cm per 4 months of acquisition. The time-series graph is also showing no significant deformation occurred before the co-seismic event at selected point (Point 1 and Point 2). The predicted result from Okada model is shows that the deformation is mostly dominated by strike-slip with 3 meter deformation as right lateral strike-slip. The Okada sources are having 11.63 km in length, 2.45 km in width, and 5.46 km in depth with the dip angle are about 84.47° and strike angle are about 142.88° from the north direction. This result is in an agreement with preliminary report from Croatian Geological Survey (HGI-CGS). Thus, further research and field data collection in order to understand more about this earthquake in order to help authorities and other decision-maker in this region for risk assessment and mitigation plan.

Acknowledgements

This research was supported by a grant from the National Research Foundation of Korea (No. 2019R1 A2C1085686), which is funded by the Government of Korea.

References

  1. Achmad, A.R., S. Lee, S. Park, J. Eom, and C.-W. Lee, 2020. Estimating the potential risk of the Mt. Baekdu Volcano using a synthetic interferogram and the LAHARZ inundation zone, Geosciences Journal, 24(6): 755-768, https://doi.org/10.1007/s12303-020-0032-9
  2. Bruyninx, C., J. Legrand, A. Fabian, and E. Pottiaux, 2019. GNSS metadata and data validation in the EUREF Permanent Network, GPS Solutions, 23(4): 106, https://doi.org/10.1007/s10291-019-0880-9
  3. Chen, C.W. and H. Zebker, 2002. Phase unwrapping for large SAR interferograms: Statistical segmentation and generalized network models, IEEE Transactions on Geoscience and Remote Sensing,40(8):1709-1719,https://doi.org/10.1109/TGRS.2002.802453
  4. Croatian Geological Survey, 2021. Press release of the Croatian Geological Survey (HGI-CGS), Hrvatski geoloski institut, https://www.hgicgs.hr/press-release-of-the-croatian-geologicalsurvey-hgi-cgs/, Accessed on Jan. 25, 2021.
  5. Crosetto, M., O. Monserrat, M. Cuevas-Gonzalez, N. Devanthery, and B. Crippa, 2016. Persistent Scatterer Interferometry: A review, ISPRS Journal of Photogrammetry and Remote Sensing, 115: 78-89, https://doi.org/10.1016/j.isprsjprs.2015.10.011
  6. Dodangeh, E., M. Panahi, F. Rezaie, S. Lee, D. TienBui, C.-W. Lee, and B. Pradhan, 2020. Novel hybrid intelligence models for flood-susceptibility prediction: Meta optimization of the GMDH and SVR models with the genetic algorithm and harmony search, Journal of Hydrology, 590: 125423, https://doi.org/10.1016/j.jhydrol.2020.125423
  7. Fadhillah, M.F., A.R. Achmad, and C.-W. Lee, 2020. Integration of InSAR Time-Series Data and GIS to Assess Land Subsidence along Subway Lines in the Seoul Metropolitan Area, South Korea, Remote Sensing, 12(21): 3505, https://doi.org/10.3390/rs12213505
  8. Ferretti, A., C. Prati, and F. Rocca, 2001. Permanent scatterers in SAR interferometry, IEEE Transactions on Geoscience and Remote Sensing, 39(1): 8-20, https://doi.org/10.1109/36.898661
  9. Ganas, A.,P. Elias, S. Valkaniotis, V. Tsironi, I. Karasante, and P. Briole, 2021. Petrinja earthquake moved crust 10 feet, https://doi.org/10.32858/temblor.156, Accessed on Feb. 8, 2021.
  10. Geng, T., X. Xie, R. Fang, X. Su, Q. Zhao, G. Liu, H. Li, C. Shi, and J. Liu, 2016. Real-time capture of seismic waves using high-rate multi-GNSS observations: Application to the 2015 Mw 7.8 Nepal earthquake, Geophysical Research Letters, 43(1): 161-167, https://doi.org/10.1002/2015GL067044
  11. Hakim, W.L., A.R. Achmad, J. Eom, and C.-W. Lee, 2020.Land Subsidence Measurement of Jakarta Coastal Area Using Time Series Interferometry with Sentinel-1 SAR Data, Journal of Coastal Research, 102(sp1): 75-81, https://doi.org/10.2112/SI102-010.1
  12. Hakim, W.L., A.R. Achmad, and C.-W. Lee, 2020. Land Subsidence Susceptibility Mapping in Jakarta Using Functional and Meta-Ensemble Machine Learning Algorithm Based on Time-Series InSAR Data, Remote Sensing, 12(21): 3627, https://doi.org/10.3390/rs12213627
  13. Hooper, A.J., 2006. Persistent Scatterer Radar Interferometry for Crustal Deformation Studies and Modeling of Volcanic Deformation, Stanford University, Stanford, CA, USA.
  14. Hooper,A.J., P. Segall, and H. Zebker, 2007. Persistent scatterer interferometric synthetic aperture radar for crustal deformation analysis, with application to Volcán Alcedo, Galapagos, Journal of Geophysical Research: Solid Earth, 112(7): B07407, https://doi.org/10.1029/2006JB004763
  15. Hooper, A.J., 2008. A multi-temporal InSAR method incorporating both persistent scatterer and small baseline approaches, Geophysical Research Letters, 35(16):L16302, https://doi.org/10.1029/2008GL034654
  16. Hooper, A.J., 2010. A statistical-cost approach to unwrapping the phase of InSAR time series, http://radar.tudelft.nl/-ahooper/Hooper_FRINGE_2009.pdf, Accessed on Dec. 16, 2020.
  17. Hooper, A.J., D. Bekaert, K. Spaans, and M. Arikan, 2012. Recent advances in SAR interferometry time series analysis for measuring crustal deformation, Tectonophysics, 514-517: 1-13, https://doi.org/10.1016/J.TECTO.2011.10.013
  18. Hooper, A.J., H. Zebker, P. Segall, and B. Kampes, 2004. A new method for measuring deformation on volcanoes and other natural terrains using InSAR persistent scatterers, Geophysical Research Letters, 31(23):L23611, https://doi.org/10.1029/2004GL021737
  19. Lee, C.-W., Z. Lu, and H.S. Jung, 2012. Simulation of time-series surface deformation to validate a multi-interferogram InSAR processing technique, International Journal of Remote Sensing, 33(22): 7075-7087, https://doi.org/10.1080/01431161.2012.700137
  20. Marquardt, D.W., 1963. An Algorithm for Least Squares Estimation of Nonlinear Parameters, Journal of the Society for Industrial and Applied Mathematics, 11(2): 431-441, https://doi.org/10.1137/0111030
  21. Ministry of Internal Affairs of Republic of Croatia, 2020. Earthquake near Petrinja, https://civilnazastita.gov.hr/vijesti/potres-kod-petrinje-3357/3357,Accessed on Jan. 25, 2021 (In Croatian).
  22. Nur, A.S., A.R. Achmad, and C.-W. Lee, 2020. Land Subsidence Measurement in Reclaimed Coastal Land: Noksan UsingC-Band Sentinel-1 Radar Interferometry, Journal of Coastal Research, 102(sp1): 218-223, https://doi.org/10.2112/SI102-027.1
  23. Okada, Y., 1985. Surface deformation due to shear and tensile faults in a half-space, Bulletin of the Seismological Society of America, 75(4): 1135-1154. https://doi.org/10.1785/BSSA0750041135
  24. Osmanoglu, B., F. Sunar, S. Wdowinski, and E. Cabral-Cano, 2016. Time series analysis of InSAR data: Methods and trends, ISPRS Journal of Photogrammetry and Remote Sensing, 115: 90-102, https://doi.org/10.1016/j.isprsjprs.2015.10.003
  25. Pawluszek-Filipiak, K. and A. Borkowski, 2020. Integration of DInSAR and SBAS Techniques to Determine Mining-Related Deformations Using Sentinel-1 Data: The Case Study of Rydultowy Mine in Poland, Remote Sensing, 12(2): 242, https://doi.org/10.3390/rs12020242
  26. Pikija, M., 1987. Basic geological map of SFRY 1:100000, List Sisak L-33-93. Croatian Geological Survey (HGI-CGS), Belgrade, HRV (in Croatian).
  27. Sousa, J.J., A.J. Hooper, R.F. Hanssen, L.C. Bastos, and A.M. Ruiz, 2011. Persistent Scatterer InSAR: A comparison of methodologies based on a model of temporal deformation vs. spatial correlation selection criteria, Remote Sensing of Environment, 115(10): 2652-2663, https://doi.org/10.1016/j.rse.2011.05.021
  28. Storchak, D.A., D. Di Giacomo, E.R. Engdahl, J. Harris, I. Bondár, W.H.K. Lee, P. Bormann, and A. Villasenor, 2015.The ISC-GEM Global Instrumental Earthquake Catalogue (1900-2009): Introduction, Physics of the Earth and Planetary Interiors, 239: 48-63, https://doi.org/10.1016/j.pepi.2014.06.009
  29. Tobita, M., 2016.Combined logarithmic and exponential function model for fitting postseismic GNSS time series after 2011 Tohoku-Oki earthquake, Earth, Planets and Space, 68(1): 41, https://doi.org/10.1186/s40623-016-0422-4
  30. Ustaszewski, K., M. Herak, B. Tomljenovie, D. Herak, and S. Matej, 2014. Neotectonics of the Dinarides-Pannonian Basin transition and possible earthquake sources in the Banja Luka epicentral area, Journal of Geodynamics, 82: 52-68, https://doi.org/10.1016/j.jog.2014.04.006
  31. Weston, J., A.M.G. Ferreira, and G.J. Funning, 2012. Systematic comparisons of earthquake source models determined using InSAR and seismic data, Tectonophysics, 532-535: 61-81, https://doi.org/10.1016/j.tecto.2012.02.001
  32. Wilkinson, M.W., K.J.W. McCaffrey, R.R. Jones, G.P. Roberts, R.E. Holdsworth, L.C. Gregory, Walters, R.J. Wedmore, L., H. Goodall, and F. Iezzi, 2017. Near-field fault slip of the 2016 Vettore Mw 6.6 earthquake (CentralItaly) measured using low-cost GNSS, Scientific Reports, 7(1): 1-7, https://doi.org/10.1038/s41598-017-04917-w
  33. Wright, T.J., B.E. Parsons, J.A. Jackson, M. Haynes, E.J. Fielding, P.C. England, and P.J. Clarke, 1999. Source parameters of the 1 October 1995 Dinar (Turkey) earthquake from SAR interferometry and seismic bodywave modelling, Earth and Planetary Science Letters, 172(1-2): 23-37, https://doi.org/10.1016/S0012-821X(99)00186-7
  34. Wright, T.J., Z. Lu, and C. Wicks, 2003. Source model for the Mw 6.7, 23 October 2002, Nenana Mountain Earthquake (Alaska) from InSAR, Geophysical Research Letters, 30(18): 1974, https://doi.org/10.1029/2003GL018014
  35. Xu, X., D.T. Sandwell, and B. Smith-Konter, 2020. Coseismic displacements and surface fractures from sentinel-1 InSAR: 2019 Ridgecrest earthquakes, Seismological Research Letters, 91(4): 1979-1985, https://doi.org/10.1785/0220190275