Pure and Applied Geophysics

Sandeep, ; Devi, Sonia; Kumar, Parveen; Monika, ; Kumar, Rohtash

doi: 10.1007/s00024-022-02947-8pmid: N/A

A devastating earthquake (Mw 7.4) struck near Izmit city, northwestern Turkey, on 17 August 1999. The relatively large size of this earthquake and site amplification conditions caused severe damage in the Marmara region as well as the city of Istanbul. The disastrous outcome of this large earthquake requires careful analysis of the seismic hazard including local site effects in this region. Therefore, the present work addresses the issue of local site effects and modelling of strong ground motion for this earthquake. We have estimated the soil effects using the horizontal-to-vertical (H/V) ratio. Furthermore, high-frequency records are simulated using a modified semi-empirical technique (MSET) by including estimated site effects. This modification will provide more reliable and precise simulated strong motion records. In the present study, strong ground motions are simulated at nine seismic stations in the epicentral range of 40–99 km. These stations were selected based on their recorded data quality. To validate the technique, we have compared the synthetic records with the observed records in terms of root mean square error (RMSE). This comparison includes time records, pseudo-acceleration spectra, mean period and predominant period. This comparison shows MSET has successfully simulated the 1999 Izmit earthquake.

Thornley, John; Douglas, John; Dutta, Utpal; Yang, Zhaohui

doi: 10.1007/s00024-022-02945-wpmid: N/A

Anchorage, Alaska, is located in one of the most active tectonic settings in the world. The city and region were significantly impacted by the MW 9.2 Great Alaska Earthquake in 1964, and they were recently shaken by a MW 7.1 event in 2018. The city was developed in an area underlain by complex soil deposits of varied geological origins and stiffnesses, with the deposits’ thicknesses increasing east to west. Situated at the edge of the North American Plate, with the actively subducting Pacific Plate below, Anchorage is susceptible to both intraslab and interface earthquakes, along with crustal earthquakes. Strong-motion stations were installed across the city in an attempt to capture the variability in site response. Several previous studies have been performed to evaluate that variability but have not included larger magnitude events and have not benefited from the current density of instrumentation. The work presented here provides background information on the geology and tectonic setting of Anchorage and presents details related to the dataset and methods used to perform the site-response analysis. This study has collected strong-motion recordings from 35 surface stations across Anchorage for 95 events spanning from 2004 to 2019, including the MW 7.1 Anchorage Earthquake in 2018. The more than 1700 three-component recordings from those 95 events with moment magnitudes ranging from 4.5 to 7.1 were used to evaluate site response variability across the city. Using the Generalized Inversion Technique and a reference rock site, spectral amplifications were calculated and analyzed for frequencies between 0.25 and 10 Hz for each strong-motion station. The study results were used to develop contour maps at 1 Hz and 5 Hz, using logarithmic-band averages, to describe the variability of spectral amplifications at these two frequencies of interest. The results were also compared to geologic conditions across Anchorage, and the overlaying of different soil deposits can be seen to have an impact on the spectral amplification at the sites. The results of this study provide improvements on past microzonation studies and, using sensitivity analyses, offer support for the use of small and moderate earthquakes to evaluate spectral amplifications.

Huded, Praveen M.; Dash, Suresh R.

doi: 10.1007/s00024-021-02929-2pmid: N/A

The Indian peninsular region has witnessed moderate to large earthquakes in the recent past. The present work demonstrates probabilistic seismic hazard assessment at the bedrock level using six ground-motion prediction equations with a logic tree approach considering Odisha state as a case study. Regional rupture characteristics are established for the tectonic features of the study region. Uniform hazard response spectra are computed for district places. Seismic hazard results are disaggregated for Bhubaneswar, the capital city of Odisha. The results have shown the distribution of seismicity in the state is not in sync with the seismic zone map provided by the Indian code.

Κaviris, George; Zymvragakis, Angelos; Bonatis, Pavlos; Sakkas, Georgios; Kouskouna, Vasiliki; Voulgaris, Nicholas

doi: 10.1007/s00024-022-02950-zpmid: N/A

Messinia is located in SW Peloponnese (Greece), in the vicinity of the Hellenic Arc which is one of the most seismically active areas of Europe. The arc is dominated by reverse faulting, whereas normal faults are mapped onshore, mainly striking N–S. Large earthquakes have occurred in the study area, both in the historical and instrumental periods. In the present study, a Probabilistic Seismic Hazard Assessment (PSHA) is applied to estimate the Peak Ground Acceleration (PGA), Peak Ground Velocity (PGV), Peak Ground Rotational Acceleration (PGRA) and Peak Ground Rotational Velocity (PGRV) for the broader Messinia region. PGRA and PGRV are not often examined in detail in the literature, even though they are useful for the evaluation of the possible damages in structures. The widely used approach proposed by Cornell and McGuire is implemented taking into account: (a) the seismotectonic model proposed by Seismic Hazard Harmonization in Europe (SHARE), (b) an earthquake catalogue for the instrumental period and (c) Ground Motion Prediction Equations (GMPEs) proposed for the Greek territory. The computational grid spacing for Messinia was set to 1 km, in order to accurately calculate the hazard parameters and to reduce the bias of the results through interpolation processes. For PGA and PGV, a logic tree approach is considered, where every branch is a hybrid version of each GMPE considering the percentage of the type (normal or non-normal) of focal mechanisms for all zones of the seismotectonic model. The results show an increase of the hazard values in the NW part of the study area, where the Greek Building Code provides lower PGA values. In addition, hazard curves in terms of PGA for multiple probabilities of exceedance in 50 years are determined for five major towns of Messinia, i.e. Kalamata, Messini, Filiatra, Kyparissia and Pylos. Kyparissia has the higher hazard in all probabilities of exceedance, which is in full agreement with the PGA and PGV results. Pylos and Filiatra have intermediate to high hazard, whereas Kalamata and Messini show intermediate hazard values. Furthermore, the Uniform Hazard Spectrum (UHS) in terms of Spectral Acceleration for the same towns is computed. The latter provides essential information about the design parameters. Lastly, the results are compared to the seismic histories of the five major towns in terms of European Macroseismic Scale (EMS98) intensity, plotted for the last 200 years.

Provost, Ludmila; Antonucci, Andrea; Rovida, Andrea; Scotti, Oona

doi: 10.1007/s00024-021-02943-4pmid: N/A

We investigate the differences in seismicity rate estimates from two historical earthquake catalogues obtained with two methodologies (Boxer and QUake-MD) calibrated on a common dataset of macroseismic intensities and calibration events. The two methodologies were then applied to a test data set of historical earthquakes covering the France, Italy and Switzerland Alpine region. Differences between the resulting magnitude estimates and instrumental magnitudes show a standard deviation of 0.4 for both methodologies, with a mean residual of 0.01 for Boxer and − 0.04 for Quake-MD. A systematic difference in magnitude estimates between the two methodologies that correlates with the depth estimated by Quake-MD has been observed. This is attributed to the difference in the treatment of the depth parameter between Boxer and QUake-MD. Nevertheless, differences in magnitude estimates between the two methodologies show a mean residual of 0.006 and a standard deviation of 0.35 resulting in seismicity rates that are not significantly different considering the associated uncertainties. Such results made us believe that the European community could gain in the reduction of epistemic uncertainties associated with the estimate of historical earthquake parameters by agreeing on a common macroseismic and calibration dataset across borders. These efforts should be strongly encouraged. On the other hand, we show that even in the ideal conditions of this benchmark (same calibration events and same macroseismic intensity dataset), methodological differences can lead to systematic differences in magnitude estimates. It is therefore paramount to explore different methodologies for a more realistic quantification of the epistemic uncertainties in estimates of maximum magnitudes and seismic activity rates.

Bayik, Caglar; Gurbuz, Gokhan; Abdikan, Saygin; Gormus, Kurtulus Sedar; Kutoglu, Senol Hakan

doi: 10.1007/s00024-022-02944-xpmid: N/A

The East Anatolian Fault Zone in Turkey produced a major earthquake (Mw 6.8) on 24 January 2020. In this study, we analyze the displacements of Global Navigation Satellite Systems (GNSS) stations. We also use the interferometric synthetic aperture radar (InSAR) technique to determine deformation caused by the earthquake. For this analysis, interferograms were produced with Sentinel-1 coseismic pairs. The coseismic displacement is about 60 cm as a maximum in the line-of-sight (LOS) direction from ascending and descending tracks. The displacements are modeled with uniform slip modeling to understand the fault mechanism and to estimate the source parameters of the earthquake. In addition, horizontal and vertical displacements are calculated using the decomposition of LOS solutions. According to the InSAR results, the fault length measured 37 km, and the magnitude of the earthquake was calculated as 6.6, not 6.8. The InSAR results were inverted into the elastic half-space model and estimated the earthquake source parameters as follows: depth 17 km, strike 242°, dip 80°, rake 1°.

Korrat, I. M.; Lethy, Ahmed; ElGabry, M. N.; Hussein, H. M.; Othman, Adel S.

doi: 10.1007/s00024-022-02953-wpmid: N/A

Source parameters calculated from displacement spectra of both P and S waves are used to discriminate between earthquakes and quarry blasts in three regions of Egypt during the 2009–2015 period. We use vertical component seismograms from 440 earthquakes and 450 quarry explosions with MD 1.5 to 3.3 to calculate source parameters, including scalar moments and corner frequencies. The Mo(P,S) vs. fc(P,S) and P- to S-wave corner frequency fc(P)/fc(S) ratios are used to distinguish quarry blasts from earthquakes. A comparison of Mo(P,S) vs. fc(P,S) for both earthquakes and explosions in Egypt demonstrates that explosions had significantly lower corner frequencies than earthquakes, particularly for S-wave displacement spectra. In contrast to the Northern and Central regions, the Southern Egyptian region provides a perfect separation of corner frequencies of earthquakes and explosions for both P- and S-waves. The empirically derived average ratio of fc(P)/fc(S) for earthquakes is 1.28, 1.26 and 1.26 in the Northern, Central and Southern Egyptian regions, respectively. For explosions, average fc(P)/fc(S) ratios are 1.89, 1.86 and 2.0 in the three Egyptian regions, respectively. According to these findings, the average ratio of fc(P) to fc(S) for explosions is higher than those for earthquakes, implying that the differences in ratios enhance the ability of the fc(P) vs. fc(S) approach to discriminate between earthquakes and explosions. Based on the average fc(P)/fc(S) ratios vs. Mw in the whole of Egypt, the observed fc(P)/fc(S) discrimination threshold value for separating quarry explosions from earthquakes is 1.51–1.52.

Radziminovich, Natalia A.

doi: 10.1007/s00024-022-02952-xpmid: N/A

Earthquake depth frequency distribution is often used as a constraining factor in assessing lithospheric strength. The accuracy of hypocenter locations is therefore a key factor in these types of analyses. This paper presents a summary of earthquake depth distributions in the Baikal rift system reported in various previous studies. Datasets on both background seismicity and aftershock sequences were analyzed by taking into account hypocenter depth uncertainties. The results show that the most seismically active part of the Earth's crust is in a depth range of 10–25 km. The lower cutoff depth varies for different datasets from 20 to 40 km. The latter value corresponds to the Kichera sequence and coincides with the Moho depth, but other datasets show that the lower 8–18 km of the crust is aseismic. The South Baikal basin is distinguished by the shallower cutoff depth (20–24 km) compared to the datasets outside of it. The peak depths of the earthquake depth frequency distributions vary from 8 to 23 km. Given the depth uncertainties, only the Kumora sequence appears to have double peaks at 9–10 km and 16–17 km. Variations in the lower limit of earthquake distribution and depth of the main peak may imply a spatial change in the depth of the brittle–ductile transition and consequently in the thermomechanical properties of the crust.

Bhukta, Kuntal; Paul, Ajay; Khan, Prosanta K.

doi: 10.1007/s00024-021-02935-4pmid: N/A

An experiment on SKS and SKKS splitting is carried out using teleseismic earthquakes recorded at 10 broadband seismic stations deployed in parts of the NW Himalaya. A total of 47 reliable splitting parameters is estimated using rotation correlation and transverse component minimization methods. A large variation of delay time (δt) of 0.3 to 1.6 s and polarization direction (ϕ) of 42° and 98° of the fast wave is obtained. A strong component of fast axes along the strike of the Delhi–Hardwar Ridge (DHR) near the Himalayan foothills is found to be correlated with the structurally controlled strain-induced shearing. The structural fabrics were likely frozen because of induction of past plate tectonics. The orientations of the fast axes surrounding the Main Central Thrust (MCT) are correlated with the foliation and folding caused by strike-orthogonal compression as well as underthrusting of the Indian plate beneath the Himalaya. A low value of 0.3 s delay time is proposed to be caused by multi-layered anisotropy. A two-layered anisotropy model developed for the stations Adibadri, Gaurikund and Chakrata reveals that the directions of fast axes in the shallow and deeper levels are parallel to the strike of the DHR and to the absolute plate motion (APM) vector of the Indian plate, respectively. We thus propose that the anisotropy observed in the NW Himalaya is a combined effect of the fossil anisotropy preserved in the DHR, the mantle flow and the foliation plane arisen due to collision between the Indian and the Eurasian plates.

Cao, Jing-Jie; Yao, Gang; da Silva, Nuno V.

doi: 10.1007/s00024-021-02936-3pmid: N/A

Seismic data interpolation is an essential tool for providing complete seismic data when field data are incomplete due to the influence of obstacles, topography and acquisition cost. Among the existing interpolation methods, sparsity inversion-based methods are commonly used for noisy data interpolation. These methods assume that seismic data can be sparsely expressed in a transformed domain, and a sparse inversion should be solved to obtain sparse coefficients. The L1 norm is often chosen as the regular operator since it is a convex function and can measure sparsity of solutions. Nonetheless, nonconvex regularization represented by the L1/2 norm has better numerical properties than those of the L1 norm regularization, since the L1/2 norm is a closer expression of sparsity. Based on the idea of nonconvex regularization, a novel nonconvex regularization model was developed to realize seismic interpolation. The L1 norm and a nonconvex, hyperbolic tangent-based function were combined as the regularization constraint, and a revised proximal method was proposed to efficiently solve the inversion model. A revised Newton direction of the nonconvex term was used to ensure that the proposed method gave stable results. Synthetic and field data examples demonstrate that the proposed technique is especially useful for interpolating missing traces in recorded seismic data sets, and it is robust even when data are contaminated with noise.