Korean Journal of Remote Sensing (대한원격탐사학회지)

Korean Society of Remote Sensing (대한원격탐사학회)
권호 표지
DOI QR코드

DOI QR Code

The Estimated Source of 2017 Pohang Earthquake Using Surface Deformation Modeling Based on Multi-Frequency InSAR Data

  • Received : 2021.01.29
  • Accepted : 2021.02.17
  • Published : 2021.02.26
Download PDF
⟨ Previous Next ⟩

Abstract

An earthquake occurred on 17 November 2017 in Pohang, South Korea with a strength of 5.4 Mw. This is the second strongest earthquake recorded by local authorities since the equipment was first installed. In order to improve understanding of earthquakes and surface deformation, many studies have been conducted according to these phenomena. In this research, we will estimate the surface deformation using the Okada model equation. The SAR images of three satellites with different wavelengths (ALOS-2, Cosmo SkyMed and Sentinel-1) were used to produce the interferogram pairs. The interferogram is used as a reference for surface deformation changes by using Okada to determine the source of surface deformation that occurs during an earthquake. The Non-linear optimization (Levemberg-Marquadrt algorithm) and Monte Carlo restart was applied to optimize the fault parameter on modeling process. Based on the modeling results of each satellite data, the fault geometry is ~6 km length, ~2 km width and ~5 km depth. The root mean square error values in the surface deformation model results for Sentinel, CSK and ALOS are 0.37 cm, 0.79 cm and 1.47 cm, respectively. Furthermore, the results of this modeling can be used as learning material in understanding about seismic activity to minimize the impacts that arise in the future.

Keywords

1.Introduction

Earthquakes are a natural phenomenon that often occurs on earth ranging from a small scale to a larger scale. An earthquake with a large force can cause large losses then affect damage to infrastructure and environmental changes (Syifa et al., 2019). In general, earthquakes can be caused by several factors such as volcanic and tectonic activity, earthquakes are disasters that cannot be predicted until now. Many scientists have developed methods to study the mechanisms of an earthquake that occurs so that it can be used as a reference to reduce the impact in the future (H.-S. Kim et al., 2018). The occurrence of earthquakes is driven by pressure in the earth’s crust, and the quakes themselves can change the stress field as they relieve and redistribute the pressure that has built up during interseismic periods. Modeling of stress changes caused by large earthquakes has shown that these changes can affect the occurrence of subsequent earthquakes (Hardebeck and Okada, 2018).

The Korean peninsula belongs to the interior of the Eurasian Plate, where seismic activity is not high, and about 400 km from the plate boundary where four plates interact with the Philippine Sea Plate, the Pacific Plate, and the North American Plate (Sun et al., 2012). Therefore, the frequency of seismic activity is smaller than neighborhood countries such as Japan and China. However, there is a possibility of a strong earthquake, and the recent Gyeongju earthquake in 2016 (Mw= 5.8) and Pohang earthquake in 2017 (Mw= 5.4) caused a loss of economic and human damage in Korea (Choi et al., 2019; Uchide and Song, 2018). In order to prepare for such earthquake damage, it is important to study the characteristics of surface deformation caused by earthquakes. However, since observation records were reported only for earthquakes that occurred after 1978 by the Korean Meteorological Administration, there are very few records of earthquake ground motion to evaluate the characteristics of ground motion. Previous studies have studied to determine the causes that can induce the 2017 Pohang earthquake using InSAR data and modeling using Columb stress modeling (Grigoli et al., 2018). On the other hand, the estimated deformation from this earthquake was carried out using InSAR data from Sentinel-1 to estimate static slip distribution on the fault plane that was caused by the Pohang earthquake by inverting the surface deformation InSAR data (Song and Lee, 2019). Other studies have also been conducted to find factors that induce this earthquake such as drilling activity and geothermal condition around the Pohang (Kim et al., 2018; Park et al., 2017; Woo et al., 2019), as well as the potential impact of liquefaction after the earthquake (Gihm et al., 2018).

This study aims to estimate the source of surface deformation due to the 2017 Pohang earthquake in Pohang, South Korea. By using InSAR data from three satellites with different wavelengths, namely ALOS-2 (L-band), Cosmo Sky-Med (X-band), and Sentinel-1 (C-band) during the time the earthquake occurred. The SAR dataset is then used to create an interferogram which will later be applied as a deformation reference in the modeling process. Estimation of source surface deformation was carried out using Okada modeling calculations. This method has been widely applied in deformation modeling studies, especially in the study of earthquakes and volcanoes (Bonforte et al., 2019). The results of this research can be used as a reference in determining the mechanism of an earthquake. Furthermore, it can be a mitigation reference for stakeholders to make decisions in the future.

2. Method

1) Study Area

This study focuses on the Pohang area in the eastern part of the Korean peninsula. Pohang has located at positions 35°50′07″ N and 36°16′34″ N and longitude 128°05′31″ East Longitude and 128°34′57″ E. The area of the study is 1127 km2with a 516, 471 population. The northeast region consists of coastal hills which mostly consist of Tertiary Semi-Consolidated sediments. This area is called the ‘Pohang Basin’. The northwest area is steep mountains consisting of volcanic rock and granite. Most rainfall is from July to September. The average annual rainfall between 1973 and 2009 was 1119 mm. The geology of the bedrock of the study area consists mainly of limestone, granite, and volcanic sedimentary rocks; Tertiary granite, volcanic, and sedimentary rocks; and Quaternary basalt (Lee et al., 2012). In the study area, Cretaceous volcanic and granite rocks are scattered mainly in the west, Tertiary sediments in the east, and Tertiary volcanic and granite rocks in the south. The study area can be shown in Fig. 1.

OGCSBN_2021_v37n1_57_f0001.png 이미지

Fig.1. The location of the study area (a) Map of South Korea, (b) Map of Pohang city from Sentinel-2 imagery data.

2) Interferometric SAR processing

The Differential Interferogram SAR (DInSAR) technique is the most widely-used method for measuring surface deformation, with a two-pass interferogram approach applied to the SAR datasets. Prior to interferogram generation, the SAR data from each satellite was applied on coregistered process (Achmad et al., 2020). In order to the co-registration process, the alignments of two SAR images will be processed in subpixel accuracy for the accurate determination of phase and noise reduction, then to form the interferometric pairs. Those SAR images will be resampling of the slave images to match with the master image. On completion of the co-registration process, the coregistered images will be cropped to focus on the study area, then the interferogram generation process can be started. While the interferogram is processed, the topography InSAR is generated from the process. As for the topography reference, we use Shuttle Radar Topography Mission (SRTM) Global one arcsec with a 30-meter resolution for Sentinel-1 (Farr et al., 2007). Meanwhile, for ALOS-2 and Cosmo SkyMed dataset we use the WorldDEM digital elevation model with a resolution of 12 m. After the InSAR images were subtracted, the topographic phase was removed and the DInSAR phase was generated where only contains the deformation phase. In order to increase the smoothness of the interferogram, the phase is filtered by adaptive filtering. Phase unwrapping procedure applied to DInSAR phase to generate the unwrapped-DInSAR. For the Pohang co-seismic data, the interferogram is generated from SAR data mentioned in Table 1 below.

Table 1. The information of SAR dataset that used for generating Interferogram

OGCSBN_2021_v37n1_57_t0001.png 이미지

3) Surface deformation model

After the co-seismic data from three different satellite SAR data was generated, we try to estimate the surface deformation of the co-seismic interferogram images. The surface deformation was computed using the Okada method, which has widely used to estimate the model of fault plane in the study of the earthquake. This could be used as ground deformation simulation by assuming a homogeneous half-space. This is often done using the “Okada model”, which is derived from a Green’s function solution to the elastic half-space problem. Uniform displacement of the solid over a finite rectangular patch specified using the parameters defined by Okada, when inserted in a homogeneous elastic half-space a distance depth below the free surface, leads to a steady-state solution in which the free surface is deformed. In this research the Okada model was carried out using a two-step approach (De Novellis et al., 2019); a nonlinear optimization based on the Levember-Marquardt algorithm to define the rupture mechanism and fault geometry (Marquardt, 1963), then a linear inversion to get the opening distribution with Monte Carlo restart to avoid local minima (Wright et al., 1999). The inversion for the geometry fault parameter in the surface deformation data can be solved by repeatedly calculating the forward model of surface deformation as it is a nonlinear inverse problem, and then adjust the fault parameters until the difference between the modeling and the observed deformation is reduced. Non-linear optimization algorithms (such as the Levemberg- Marquardt algorithm) can be used to solve this problem by systematically changing the source parameters to find the most suitable modeling results for the observed data. The calculated result is then converted by projecting the 3D displacement vector into the satellite line of sight. After that, the model was applied to the earthquake source of Pohang in 2017. The root mean square error (RMSE) was carried out for the validation purpose, lower RMSE shows that modeling result more acceptable (Yastika et al., 2019). RMSE method was carried out by subtracting the prediction (model) value and observation value, the equation that used to calculate as follow (Pawluszek-Filipiak and Borkowski, 2020):

\(R M S E=\sqrt{\frac{\sum_{i=0}^{n}\left(f_{o}-f_{p}\right)^{2}}{n}}\)       (1)

Where fo is the result observed from DInSAR and fp is prediction value. This method has been widely used to evaluate the prediction with observed results in terms of classification purpose (Dodangeh et al., 2020).

3. Result

1) InSAR result

ALOS (L-band), Cosmo SkyMed (X-band), and Sentinel-1 C-band SAR (Synthetic Aperture Radar) were exploited to describe the surface deformation accompanying the Pohang earthquake. The SAR data pairs used are between the earthquakes that occurred, with different time intervals. The interferogram of SAR is produced by applying two-pass DInSAR (Differential Interferometric Interferometry) processing to each SAR data.

Based on the results of the DInSAR interferogram, it shows changes in the complex surface deformation and width at the epicenter of the earthquake. According to the magnitude of the earthquake that is felt, has an impact on the condition of the buildings and surfaces around the earthquake epicenter area. The affected area is estimated to be 10 km (north-south) and 6.5 km (east- west), where the conditions of decline and uplift occur in separate areas (Yun et al., 2018).

The use of satellite data with different wavelengths indicates a difference in the deformation that occurred in the earthquake in Pohang shown in Fig. 2. Based on Sentinel-1 satellite data using C-band waves, the deformation that occurs during an earthquake around the epicenter is -4-5 cm as shown in Fig. 2(a). Meanwhile, the Cosmo SkyMed data taken on 12 November 2017 (d) and 20 November 2017 as shown in Fig. 2(b) shows the location of deformations that are not much different. There is a big difference in the deformation that occurs where the results of the Cosmo SkyMed data show the deformation of about -4 to 6 cm around Pohang. In addition, for the ALOS-2 data, we used a pair of interferograms on 16 August 2016 and 18 November 2017 showing a deformation change of ~ 2.5 cm in Fig. 2(c). In general, the use of Cosmo SkyMed data using X-band shows a higher resolution than the Sentinel-1 and ALOS-2 satellites. The deformation pattern and the location of the surface changes shown by the satellite show similarities based on the DInSAR results.

OGCSBN_2021_v37n1_57_f0002.png 이미지

Fig. 2. Unwrapped phase DInSAR of co-seismic in Pohang from (a) Sentinel-1, (b) Cosmo SkyMed, (c) ALOS-2, and (d) LOS deformation over profile A-A’ from each satellite (above) with elevation over profile A-A′(below).

In addition, Fig. 2(d) shows the surface deformation profile from different satellite with elevation along A-A′. Based on the deformation profile obtained, the highest deformation change was calculated at a distance of 2 km from point A as indicated by the results from Sentinel-1 and CSK about 3.7 cm and 4.3 cm, respectively. As for ALOS-2, the deformation profile has increased up to 1.6 cm, but not as significant as the other two satellites. Besides, the three satellites show a decrease in deformation value after reaching the highest point with different values. This deformation profile value is still influenced by several errors during the interferogram formation process or temporal decorrelation.

2) Surface deformation modeling

In order to describe the fault geometry as the source of the 2017 Pohang earthquake, we used fault slip modeling using the Okada model. The search for the fault geometry of the Pohang earthquake was carried out on the interferogram pair of three satellites with different wavelengths. Modeling with this method is used to get an overview of the deformation conditions that are in good agreement with the InSAR results. Based on the results of the modeling that has been carried out, we can produce the closest deformation signal, this indication is obtained by looking at the interferogram residue that appears clear result. The interferogram residual data is obtained by reducing the value of the real interferogram with the modeling results. Modeling result depends on the parameters that exist in the Okada equation to achieve a good result. After completion on modelling process, the best-fit parameters for each SAR dataset interferogram can be seen in Table 2 below:

Table 2. Best-fit parameters that used for modeling in each satellite data

OGCSBN_2021_v37n1_57_t0002.png 이미지

Modeling results show that the synthetic interferogram calculation has good agreement with the observations using DInSAR. The residual images on each interferogram are quite clean, although there is still a little deformation in the southeast area from the epicenter of the earthquake. Based on Table 2, the best parameter for the deformation model using Sentinel-1 data is where the fault plane is 6.39 km length, 2.36 km wide, and 5.17 km deep with a dip of 64.97 degrees and strikes of 229.65 degrees. This earthquake had a strike-slip component of -306 mm and was equipped with a dip-slip component of 214 mm. Meanwhile modeling with Cosmo SkyMed data, the best fit parameter for this deformation model is a fault plane with a strike of about 225.58 degrees, a length of 6.28 km with a width of 2.44 km, and a depth of 5.25 km. As well as fault dips of 63.87 degrees to the north-east. The strike-slip component is -343.62 mm and the dip-slip component is 228.69 mm. The best parameter in modeling using the ALOS-2 satellite is the fault plane which has a length of 6.08 km, a width of 2.2 km, and is at a depth of 5.09 km. Then, the dip component is 61 degrees from north to east and the strike component is 231 degrees. And to complement the rest of the fault parameter, we get a strike-slip value of -329.15 mm and a dip-slip of 197.88 mm.

The modeling result was carried out by applying the Okada model to each interferogram from satellite as shown in Fig. 3 which has the observations from InSAR (Fig. 3(a), (b), and (c)), modeled (Fig. 3(d), (e), and (f)), and residual (Fig. 3(g), (h), and (i)). Based on the results of surface deformation modelling on Fig. 3(d), e and f, we can estimate the projection of fault plane on the surface for each SAR dataset which marked with rectangle. In addition, the location of fault trace on the surface is marked with dashed line by assuming the fault extended to the surface with the same dip angle from fault parameter modeling result. Also, these results may indicate unmapped fault system activity and could be a preliminary study in the context of seismic mapping in Pohang area.

OGCSBN_2021_v37n1_57_f0003.png 이미지

Fig. 3. Input of the unwrapped DInSAR from (a) Sentinel-1 (b) Cosmo SkyMed (c) ALOS-2; Result of Okada model from (d) Sentinel-1 (e) Cosmo SkyMed (f) ALOS-2 with rectangle solid line that represents fault plane and dashed line are fault traces which are an extension of fault plane on the surface; Residual image from modelling result of (g) Sentinel-1 (h) Cosmo SkyMed (i) ALOS-2.

For validation purposes, the co-seismic deformations of the unwrapped phase from three satellites are compared with those simulated using the best-fit fault geometry parameters, listed in Table 2. The RMSE values between the observed and modeled deformations were 0.37, 0.79, and 1.47 cm for Sentinel, CSK and ALOS, respectively. The RMSE shows the difference between the observed and modeled surface deformations. The amount of residual deformation is considerably smaller than the deformation modeled on the three satellites. Seismic deformation measurement performance needs to be improved by minimizing the noise source.

4. Discussion

Deformation modeling in the 2017 Pohang earthquake with Okada models from multi-frequency satellite data has been carried out. This deformation model can adequately describe how the deformation occurred in the earthquake. Using interferogram pairs of SAR data from three satellites, it shows a picture of the different deformations.

The results of this study indicate a good agreement with previous research in modeling the 2017 Pohang earthquake deformation (Grigoli et al., 2018; Song and Lee, 2019). However, there are still some discrepancies where there are parameters that are not used in modeling this time. To get better insight, the addition of these parameters can be used in further research. This can better describe the deformation mechanism associated with the fault. Also, the process of making an interferogram can affect the results of this deformation modeling, several factors can be influence to this result. One of the causes of the differences is the noise in the interferogram generation. Noise or errors in the interferogram usually appear due to several things such as orbital errors, atmospheric errors and unwrapping errors (Kim et al., 2017). To overcome the presence of this error, we can use time series deformation analysis such as PSI and SBAS which have been widely used (Osmanoğlu et al., 2016).

Apart from these error factors, the type of frequency used on the satellite sensor can affect the measurement of the deformation value on this study. The Cosmo Sky-Med satellite which uses the X sensor has a better level of spatial resolution than the C band sensor owned by Sentinel-1. One of the advantages of using X-band compared to C-band satellites is the resolution and detail of the recorded buildings (Lu et al., 2019). Besides, the use of X-band satellites is better for monitoring in urban areas which provide many kinds of man-made object such as building, road and other infrastructures. However, this does not imply that the use of C-band satellites is not suitable for monitoring deformation especially earthquakes. An increase in satellite resolution and a shorter temporal cycle can improve the reliability of measurements made by C- band satellites such as the Sentinel-1. With a short interval revisiting cycle, the Sentinel-1 satellite more reliable to monitoring deformation caused by the earthquake, it can be early detection of the deformation. Also, the measurements made using the L-band satellite (ALOS-2) show a smaller value than the other two satellites. This occurs because the temporal decorrelation is quite wide that affect some condition of the data. Even so, L-band data are sensitive to change in surface conditions and low coherence affect the measurement result. Besides, several variables can influence the coherence of the L-band satellite such as snow coverage and vegetation condition of the study area (Wempen and McCarter, 2017).

For the RMSE calculation results show that the smallest RMSE values are obtained from the deformation model using Sentinel-1 (0.37 cm), while for CSK and ALOS-2 the values are 0.79 cm and 1.47 cm, respectively. This can be caused by several things, such as the period in which SAR data is collected by satellites where the Sentinel has a shorter visit cycle. Meanwhile, for ALOS-2, the higher RMSE value could be caused by the temporal change that is too far compared to other satellites in this deformation monitoring. Even so, the resulting RMSE value is not too large from the observed deformation, this can be related to the sensor used ALOS-2 (L-band) which has deeper penetration capability so that the observed deformation is not too constrained. Modeling complex surface deformations is a difficult challenge and requires a lot of improvisation to get the best model. The use of an optimization algorithm can be used to optimize setting the initial fault parameter. This algorithm can reduce the shortcomings in determining the initial fault parameters and improved modeling result.

5. Conclusions

The earthquake that occurred on November 17, 2017, in Pohang, South Korea had an impact on damage and losses to local life. The disaster was a major disaster caused by the largest earthquake ever recorded in modern Korea today. Seeing the massive impact this had caused researchers to study the deformation that occurred during the incident. In this research, we tried to estimate the surface deformation using the Okada modeling approach where the modeling is widely used to study deformation mechanisms. Deformation modeling was constructed using interferogram SAR data as a reference. We used three satellites with different wavelengths to generate co-seismic deformation interferogram of the earthquake. Based on the modeling results, we get the best-fit parameters that describe the fault mechanism from a geometric perspective. From the point of view of the fault geometry, we estimate the source of the earthquake has a width of ~2 km with a length of ~6 km leading from northeast to southwest and with a depth of ~5 km. For validation purposes, we calculated the RMSE of each model showing the values 0.37, 0.79, and 1.47 cm for Sentinel-1, Cosmo SkyMed, and ALOS-2, respectively. By looking at this value, it implies that this model is compatible with the observed surface deformation determined by InSAR data.

However, several constraints are evaluated in making this model, such as orbital errors and atmospheric disturbances. This shortcoming can be overcome by the use of multi-temporal InSAR methods such as PSI and SBAS. Besides, the optimization algorithm can be used to determine the initial parameters to reduce errors. Also, this research has achieved its goal of estimating the deformation caused by the 5.4 Mwearthquake in 2017. This can be a reference in the learning of deformation, surface deformation modeling, and SAR application in some future natural phenomena.

Acknowledgments

This work has supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2019R1A2C1085686).

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. Bonforte, A., F. Guglielmino, and G. Puglisi, 2019. Large dyke intrusion and small eruption: The December 24, 2018 Mt. Etna eruption imaged by Sentinel-1 data, TerraNova, 31(4): 405-412.
  3. Choi, J.H., K. Ko, Y.-S. Gihm, , C.-S. Cho, H. Lee, S.-G. Song, E.-S. Bang, H.-J. Lee, H.-K. Bae, S.-W. Kim, S.-J. Choi, S.-S. Lee, and S.-R. Lee, 2019. Surface deformations and rupture processes associated with the 2017 Mw 5.4 Pohang, Korea, earthquake, Bulletin of the Seismological Society of America, 109(2): 756-769. https://doi.org/10.1785/0120180167
  4. De Novellis, V., S. Atzori, C. De Luca, M. Manzo, E. Valerio, M. Bonano, C. Cardaci, R. Castaldo, D. DiBucci, M. Manunta, G. Onorato, S. Pepe, G. Solaro, P. Tizzani, I. Zinno, M. Neri, R. Lanari, and F. Casu, 2019. DInSAR Analysis and Analytical Modeling of Mount Etna Displacements: The December 2018 Volcano-Tectonic Crisis, Geophysical Research Letters, 46(11): 5817-5827. https://doi.org/10.1029/2019GL082467
  5. Dodangeh, E., M. Panahi, F. Rezaie, S. Lee, D. Tien Bui, 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
  6. Farr, T.G., P.A. Rosen, E. Caro, R. Crippen, R. Duren, S. Hensley, M. Kobrick, M. Paller, E. Rodriguez, L. Roth, D. Seal, S. Shaffer, J. Shimada, Umland, M. Werner, M. Oskin, D. Burbank, and D.E. Alsdorf, 2007. The shuttle radar topography mission, Reviews of Geophysics, 45(2): 1-33.
  7. Gihm, Y.S., S.-W. Kim, K. Ko, J.-H. Choi, H. Bae, P.-S. Hong, Y. Lee, H. Lee, K.J. Lee, S.J. Choi, J.-C. Kim, M.-S. Choi, and S.-R. Lee, 2018. Paleoseismological implications of liquefaction-induced structures caused by the 2017 Pohang Earthquake, Geosciences Journal,22(6):871-880. https://doi.org/10.1007/s12303-018-0051-y
  8. Grigoli, F., S. Cesca, A.P. Rinaldi, A. Manconi, J.A. Lopez-Comino, J.F. Clinton, R. Westaway, C. Cauzzi, T. Dahm, and S. Wiemer, 2018. The November 2017 Mw 5.5 Pohang earthquake: A possible case of induced seismicity in South Korea, Science, 360(6392): 1003-1006. https://doi.org/10.1126/science.aat2010
  9. Hardebeck, J.L. and T. Okada, 2018. Temporal Stress Changes Caused by Earthquakes: A Review, Journal of Geophysical Research: Solid Earth, 123(2): 1350-1365. https://doi.org/10.1002/2017JB014617
  10. Kim, H.-S.,C.-G. Sun, and H.-I.Cho, 2018. Geospatial Assessment of the Post-Earthquake Hazard of the 2017 Pohang Earthquake Considering Seismic Site Effects, ISPRS International Journal of Geo-Information, 7(9): 375. https://doi.org/10.3390/ijgi7090375
  11. Kim, J.-R., S.-Y. Lin, H.-W. Yun, Y.-L. Tsai, H.-J. Seo, S. Hong, and Y. Choi, 2017. Investigation of Potential Volcanic Risk from Mt. Baekdu by DInSAR Time Series Analysis and Atmospheric Correction, Remote Sensing, 9(2): 138. https://doi.org/10.3390/rs9020138
  12. Kim, K.H., J.-H. Ree, Y.-H. Kim, S. Kim, S.-Y. Kang, and W. Seo, 2018. Assessing whether the 2017 Mw 5.4 Pohang earthquake in South Koreawas an induced event, Science, 360(6392): 1007-1009. https://doi.org/10.1126/science.aat6081
  13. Lee, S., Y.-S. Kim, and H-J. Oh, 2012. Application of a weights-of-evidence method and GIS to regional groundwater productivity potential mapping, Journal of Environmental Management, 96(1): 91-105. https://doi.org/10.1016/j.jenvman.2011.09.016
  14. Lu, P.,S. Bai, V. Tofani, and N. Casagli, 2019. Landslides detection through optimized hot spot analysis on persistent scatterers and distributed scatterers, ISPRS Journal of Photogrammetry and Remote Sensing, 156: 147-159. https://doi.org/10.1016/j.isprsjprs.2019.08.004
  15. 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
  16. 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
  17. Park, S., L. Xie, K.-I. Kim, S. Kwon, K.-B. Min, J. Choi, W.-S. Yoon, and Y. Song, 2017. First Hydraulic Stimulation in Fractured Geothermal Reservoir in Pohang PX-2 Well, Procedia Engineering, 191: 829-837. https://doi.org/10.1016/j.proeng.2017.05.250
  18. 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
  19. Song, S.-G. and H. Lee, 2019. Static slip model of the 2017 Mw 5.4 Pohang, South Korea, earthquake constrained by the InSAR data, Seismological Research Letters, 90(1): 140-148. https://doi.org/10.1785/0220180156
  20. Sun, C.-G., K. Bang, and W. Cho, 2012. A State-of-the-Practice Review on Amplification of Earthquake Ground Motion Resulting from Surface Topographic Effects, Journal of Korean Society of Hazard Mitigation, 12(3): 123-131 (in Korean with English abstract).. https://doi.org/10.9798/KOSHAM.2012.12.3.123
  21. Syifa, M., P. Kadavi, and C.-W. Lee, 2019. An Artificial Intelligence Application for Post-Earthquake Damage Mapping in Palu, Central Sulawesi, Indonesia, Sensors, 19(3): 542. https://doi.org/10.3390/s19030542
  22. Uchide, T. and S.-G. Song, 2018. Fault Rupture Model of the 2016 Gyeongju, South Korea, Earthquake and Its Implication for the Underground Fault System, Geophysical Research Letters, 45(5): 2257-2264. https://doi.org/10.1002/2017GL076960
  23. Wempen, J.M. and M.K. McCarter, 2017.Comparison of L-band andX-band differential interferometric synthetic aperture radar for mine subsidence monitoring in central Utah, International Journal of Mining Science and Technology, 27(1): 159-163. https://doi.org/10.1016/j.ijmst.2016.11.012
  24. Woo, J.-U., M. Kim, D.-H. Sheen, T.-S. Kang, J. Rhie, F. Grigoli, W.-L. Ellsworth, and D. Giardini, 2019. An In-Depth Seismological Analysis Revealing a Causal Link Between the 2017 Mw 5.5 Pohang Earthquake and EGS Project, Journal of Geophysical Research: Solid Earth, 124(12): 13060-13078. https://doi.org/10.1029/2019jb018368
  25. 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
  26. Yastika, P.E., N. Shimizu, and H.Z. Abidin, 2019. Monitoring of long-term land subsidence from 2003 to 2017 in coastal area of Semarang, Indonesia by SBAS DInSAR analyses using Envisat-ASAR, ALOS-PALSAR, and Sentinel1A SAR data, Advances in Space Research, 63(5): 1719-1736. https://doi.org/10.1016/j.asr.2018.11.008
  27. Yun, H., J. Yu, S.B. Lee, and M.H. Lee, 2018. Three dimensional disaster monitoring of the Pohang earthquake in the Republic of Korea by sentinel1, Proc. of 2018 IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium, Valencia, ESP, Jul. 22-27, pp. 2228-2230.