Florin Pavel

Fore-Arc and Back-Arc Ground Motion Prediction Model for Vrancea Intermediate Depth Seismic Source

A next generation ground motion model for the prediction of spectral accelerations both in the fore-arc and back-arc regions of the Carpathians Mountains is developed in this research for the Vrancea intermediate depth seismic source in Romania. This ground motion prediction equation (GMPE) is an updated version of the model given in Vacareanu et al. [2014] and is applicable in both the fore-arc and the back-arc regions. The strong ground motion database from which the prediction model is derived consists of over 700 triaxial accelerograms from Vrancea subcrustal seismic events, as well as from other intermediate-depth earthquakes produced in other seismically active regions in the world. The applicability of this ground motion prediction model in both the fore-arc and the back-arc region is tested using the analysis of residuals. Moreover, the appropriateness of this GMPE for soil classes B and C defined in EN 1998-1, as well as for average soil conditions is investigated. All results suggest that this model is an improvement of the previous versions of ground motion prediction equations for Vrancea intermediate-depth seismic source and its use in both the fore-arc and the back-arc regions make it a reliable candidate for more accurate seismic hazard studies of Romania.

An Updated Probabilistic Seismic Hazard Assessment for Romania and Comparison with the Approach and Outcomes of the SHARE Project

The probabilistic seismic hazard analysis for Romania is revisited within the framework of the BIGSEES national research project (http://infp.infp.ro/bigsees/default.htm) financed by the Romanian Ministry of Education and Scientific Research in the period 2012–2016. The scope of this project is to provide a refined description of the seismic action for Romanian sites according to the requirements of Eurocode 8. To this aim, the seismicity of all the sources influencing the Romanian territory is updated based on new data acquired in recent years. The ground-motion models used in the analysis, as well as their corresponding weights, are selected based on the results from several recent papers also published within the framework of the BIGSEES project. The seismic hazard analysis for Romania performed in this study are based on the traditional Cornell-McGuire approach. Finally, the results are discussed and compared with the values obtained in the recently completed SHARE research project. The BIGSEES and SHARE results are not directly comparable since the considered soil conditions are different—actual soil classes for BIGSEES and rock for SHARE. Nevertheless, the analyses of the seismic hazard results for 200 sites in Romania reveal considerable differences between the seismic hazard levels obtained in the present study and the SHARE results and point out the need for further analyses and thorough discussions related to the two seismic hazard models, especially in the light of a possible future harmonized hazard map for Europe.

Kappa and regional attenuation for Vrancea (Romania) earthquakes

In this short paper, we investigate ground motion recordings from nine intermediate-depth Vrancea (Romania) earthquakes with Mw ≥ 5.2 which occurred between 1986 and 2013. From these recordings, the high-frequency spectral decay parameter (kappa) is computed for 57 seismic stations in Romania. The relation between kappa and several parameters (event, source-to-site distance, soil class, geographical region) is evaluated through inversion techniques. The results show a very distinct influence of the earthquake magnitude and of the geographical position on the kappa values. Subsequently, a conventional frequency-dependent Q model of the form Q(f) = 100*f1.20 is derived from the geometric spreading functions. The proposed Q model and the site specific kappa values represent the basis for future stochastic simulations of ground motions generated by the Vrancea subcrustal seismic source.

Spectral Characteristics of Strong Ground Motions from Intermediate‐Depth Vrancea Seismic Source

Abstract This study focuses on the investigation of spectral characteristics of ground motions recorded from intermediate‐depth Vrancea earthquakes using Newmark–Hall‐type response spectrum. The database consists of more than 700 horizontal components recorded during 10 Vrancea earthquakes. In the first step of the analysis, the component‐to‐component variability is assessed and the corresponding influence of various parameters (earthquake magnitude, peak ground acceleration, soil class, position of the seismic station, or instrument type) is evaluated. Subsequently, the correlation between the spectral acceleration and the values of the peak ground motion parameters (acceleration, velocity, and displacement) is investigated. The analyses reveal an important influence of the magnitude and a significant influence of the soil class (in the case of large magnitude earthquakes) on the results. These findings are validated through the use of the analysis of variance (ANOVA) method. The computed response amplification factors for the Newmark–Hall type spectrum are similar with those from other studies in the literature. Finally, a check of three control periods such as T B , T E , and T F of the Newmark–Hall‐type spectrum yields results somewhat different from the ones given in the Romanian seismic design code P100‐1 or in Eurocode 8.

Investigation on the Variability of Simulated and Observed Ground Motions for Bucharest Area

ABSTRACT In this research, we evaluate the variability of ground motions from Vrancea intermediate-depth earthquakes for Bucharest area. In the first step of the analysis, the variability of ground motions recorded during four intermediate-depth Vrancea seismic events is assessed. Subsequently, the variability of ground motions simulated using the stochastic method is evaluated. The soil conditions in Bucharest are taken into consideration using the horizontal-to-vertical spectral ratios computed from the ground motions recorded during Vrancea seismic events with magnitudes MW ≤ 6.0. The stochastic simulations are performed for earthquakes with magnitudes in the range MW = 5.5–7.5 and four focal depths (90, 110, 130, and 150 km), while the epicentral positions take into account the instrumental seismicity of the 20th century. The results show similar mean normalized acceleration response spectra for the 13 analyzed sites in Bucharest area, as well as a significant influence of the earthquake magnitude on the spectral shapes. A comparison of the simulated ground motions with the amplitudes obtained from a ground motion model derived for the Vrancea subcrustal seismic source shows much larger spectral amplitudes for the former category. In addition, the largest part of the ground motion variability can be attributed to earthquake magnitude.

Source Effects on the Spectral Characteristics of Strong Ground Motions Recorded in Bucharest Area During Vrancea Earthquakes of 1986 and 1990

The August 30, 1986 (MW  = 7.1) and May 30, 1990 (MW  = 6.9) Vrancea intermediate-depth earthquakes, despite their almost similar magnitudes, have produced very different spectral contents as shown by the strong ground motions recorded in Bucharest and its surroundings. The differences can be attributed to different epicentral distances and to different values of the stress drop. The characteristics of the seismic ground motions recorded in Bucharest area in the 1986 and 1990 seismic events are discussed in the context of (a) the source characteristics of the two earthquakes and (b) the local soil conditions in Bucharest. Furthermore, an attempt is made to determine the soil factors S defined in EN 1998-1 for the Bucharest area, based on the strong ground motion dataset recorded during the two seismic events.

Site Dependent Seismic Hazard Assessment for Bucharest Based on Stochastic Simulations

The stochastic method has been and is still one of the most versatile tools for simulating earthquake ground motions. This method has been used in the past for deriving ground motion models or for evaluating seismic hazard. In this study, an attempt is made to evaluate the seismic hazard at INCERC site in Bucharest from Vrancea intermediate-depth earthquake using the stochastic method of simulating ground motions applied in conjunction with a stochastic catalogue. The simulated seismic catalogue which contains only intermediate-depth earthquakes originating in the Vrancea subcrustal seismic source with magnitudes M W ≥ 5.5, has a time length of 2500 years and it is based on the seismicity parameters derived from the ROMPLUS catalogue of the National Institute of Earth Physics. For each individual earthquake, the ground motion levels at INCERC station are assessed using stochastic finite-fault simulations. Subsequently, the seismic hazard curves for the INCERC site are obtained and compared with the ones computed through classical Cornell-McGuire probabilistic seismic hazard assessment (PSHA). In addition, the influence of the earthquake magnitude and source-to-site distance on the frequency contents of the simulated ground motions is investigated. Finally, the site-dependent displacement response spectrum constructed based on the simulated ground motions puts into evidence large displacement demands for long period structures.

Investigation of the variability of strong ground motions from Vrancea earthquakes

A systematic investigation of the applicability of several ground motion prediction models for Vrancea intermediate-depth seismic source is conducted in this research. Two ground motion prediction models recommended by previous evaluations (Vacareanu et al. in Bull Earthq Eng 11(6):1867–1884, 2013a; Pavel et al. in Earthq Struct 6(1):1–18, 2014), as well as two new state-of-the-art ground motion prediction equations (Vacareanu et al. in J Earthq Eng, 2013b; Earthq Struct 6(2):141–161, 2014) are tested using an increased strong ground motion database consisting of 150 recordings from Vrancea subcrustal earthquakes. The evaluation is performed by using several goodness-of-fit parameters from the literature. Moreover, the applicability of the single-station sigma method is also investigated by using the same strong ground motion database recorded in 30 seismic stations from southern and eastern Romania. The influence of the soil conditions on the numerical results obtained in this study is investigated and discussed using the results provided by the analysis of variance method. The impact of the single-station standard deviation on the levels of seismic hazard is also assessed in this study, and the results show, in the analyzed cases, significant reductions of the hazard levels.

Bi-normalized Response Spectra for Various Frequency Content Ground Motions

This article focuses on bi-normalized response spectra for recorded strong ground motions with various frequency contents. The bi-normalized response spectrum introduced in Xu and Xie [2004] is obtained by dividing the spectral response by the corresponding peak parameter and by dividing the vibration period of the SDOF by a frequency content-related parameter. More than 100 seismic motions recorded in 17 countries in the past 70 years are analyzed. The ground motions are classified into several categories according to: (a) the peak ground acceleration PGA and (b) the control period TC of the response spectra or the mean period, TM. The analyses performed on bi-normalized response spectra aim at highlighting the differences between the various categories of strong ground motions. The influence of the peak ground acceleration and of the frequency content on the bi-normalized response spectra is emphasized. This article also compares the computed spectra with the elastic response spectra given in EN 1998-1. It can be concluded that the bi-normalized response spectra could be used in seismic design codes and that the design spectra from EN 1998-1 should be re-shaped for the near-field region (due to the limited database).

Impact of Long-Period Ground Motions on Structural Design: A Case Study for Bucharest, Romania

In this chapter, the soil conditions in the Bucharest area are evaluated using several approaches based on both observed and simulated ground motion recordings. The analysis of the recorded ground motions shows the fact that significant long-period spectral ordinates are encountered for large magnitude Vrancea intermediate-depth seismic events. A soil class C according to EN 1998-1/2004 can be assigned for Bucharest based on the shear-wave velocities of the existing boreholes in this area. However, the much deeper Quaternary sediments from the Bucharest area can generate significant long-period spectral ordinates largely exceeding the ones for a typical soil class C site. The nonlinear ground response analysis performed for INCERC site in the eastern part of Bucharest revealed ground motions with a frequency content similar to the one of the ground motion recorded during the Vrancea 1977 earthquake. Besides the peak in amplitude corresponding to a spectral period of around 1.5 s, some additional analyses have shown that another peak at around 5 s can be inferred from the available data.