National Institute for Earth Physics

Historical monuments (palaces, churches, protected urban areas, so on) are integral part of the national cultural heritage which need activities and measures to protect them. The Banloc complex, located in the west of Romania, encompasses five buildings: the castle, the hunting pavilion, the kitchen, housekeeper house, and the stall. This study involves historical, architectural, structural and physical-chemical investigations with the aim of finding the correct restoration solutions, in good agreement with the Venice Charter, respectively the Nara Charter. Being affected by time and by the seismic sequence (Mw 5.6), the restoration of the building is absolutely necessary for recovering the functionality, structural solution and highlighting the artistic elements. Some analytical techniques, as: X-ray diffraction (XRD), Wavelength Dispersive X-ray fluorescence (WDXRF), optical microscopy (OM), and zoom microscopy, Infrared spectroscopy Fourier Transformed (FTIR), and Raman spectroscopy, have been used in this paper in order to clarify the main aspects about the structure, composition and morphological of the samples collected from Pavilion building. The oxides composition, and the possible organic impurities has been identified by WDXRF amd FTIR/Raman spectra, respectively. The microscopy images revealed the colours, fissures, cracks and even holes due to the time weathering processes, and damages they suffered in time.

Ambient seismic noise has proven to be a particularly effective tool for subsurface imaging in the last decades, with applications ranging from near surface imaging, to crustal or upper mantle tomography. Fundamentally, it relies on the cross correlations of continuous recordings of ground motion data at pairs of seismic stations. Processing steps have become more refined and promising in extracting meaningful signals that can further be used in a range of classic seismology tools. The processing, however, is usually cumbersome, time- and memory-consuming, as it requires years of continuous noise recorded at broadband seismic stations, to obtain high signal-to-noise ratio empirical Green’s functions. In order to ease the research effort, we built a database of ambient seismic noise cross correlations between pairs of broadband seismic stations that operated in Central and Eastern Europe between 1999 and 2020. The database is part of the Eastern European Ambient Seismic Noise (EENSANE) research project hosted by the National Institute of Earth Physics in Romania and will progressively grow as more stations become available, such as those provided by the new AdriaArray network. Based on this database and using state-of-the-art inversion techniques, we developed a series of near-surface and crustal tomography models of the Eastern European craton, the Trans-European Suture Zone and younger accreted terranes from Central Europe. Our integrated models provide both isotropic and azimuthally anisotropic seismic velocities from surface wave dispersion and attenuation parameters from the decay envelopes of Rayleigh waves. Using horizontal-to-vertical ratios of ambient noise, we also recovered the fundamental frequency of resonance and near surface shear wave velocity models beneath stations located across the Carpathian Orogen. Based on cross-correlation functions, we were also able to retrieve the seismic wavefield and peak ground displacement amplitudes from past earthquakes decades after their occurrence, offering a chance for improving seismic hazard and risk models in seismically vulnerable and developing regions of Europe. Our multidisciplinary results prove the versatility of ambient noise uses and the importance of the EENSANE database for a wide range of seismological imaging and hazard applications.

We analyze the usefulness of the foreshocks in forecasting seismic mainshocks. The analysis is based on possible correlations which may exist between foreshocks and mainshocks. Such correlations are expressed by a previously established time-magnitude relationship, which indicates the presence of an abrupt magnitude-decreasing sequence of correlated foreshocks in the proximity of a mainshock. By fitting this formula, we are able to derive the occurrence time of a possible mainshock. Also, we can estimate the magnitude of the mainshock, providing we know the parameters of the background seismicity of the seismic region. We report here on the application of this procedure to three Vrancea (Romania) mainshocks, the l’Aquila (Italy), Yangbi (Yunnan, China) and Izmit (Turkey) earthquakes. The limitations of the procedure are discussed. Also, a discussion is included regarding the so-called temporal variability of the Gutenberg-Richter parameter in the proximity of a mainshock, as resulting from time-magnitude and time-time correlations.

The lithosphere‐asthenosphere boundary and mid‐lithospheric discontinuities are primary attributes of the upper mantle. The Pannonian region is an extensional sedimentary basin enclosed by collisional orogens. Here, we estimate the negative phase depth of S‐to‐P receiver functions to image the lithospheric thickness and other discontinuities with high resolution, based on the recent dense seismological broadband networks. The lithosphere‐asthenosphere boundary is relatively shallow (<90 km) in the Pannonian Basin system, and deeper (∼90–140 km) in the surrounding orogens, where average surface heat flow values are higher (120 mW/m²) and lower (50–70 mW/m²), respectively. The 1D and 2D common conversion point migration with 3D velocity model provide comparable but different resolution images beneath the wider region of the Pannonian Basin. We obtained deeper values in the Western (∼120 km) and Southern‐Carpathians orogens (∼135 km). Furthermore, we provide new information on the lithospheric thickness and its seismic properties in the eastern part of the study region (e.g., Apuseni Mountains (∼95 km), Eastern‐Carpathians (∼120 km), Moesian Platform (∼90 km) and Transylvanian Basin (∼85 km). The shallower negative phase depth can be interpreted as the lithosphere‐asthenosphere boundary beneath the Pannonian Basin system in agreement with its high heat flow values. In contrast, the deeper negative phase depth estimates in the colder surroundings can be interpreted as intra‐ or mid‐lithospheric discontinuities, when compared with local seismic tomography models. In this region, the correlation with heat flow implies that the observed negative phase depth is of thermo‐chemical or rheological nature.

The analysis of the relationship between radon and seismicity was previously carried out in the seismic zone of Vrancea (Romania), positioning the measuring stations on tectonic faults. This article analyzed the evolution of radon under conditions of deep and surface seismicity and the presence of mud volcanoes, as well as fires caused by gasses emanating from the ground. The monitoring area was extended to the Black Sea and the area of the Făgăraș-Câmpulung fault, where a special radon detection system was established and proposed for patenting. The case study was the impact of the earthquakes in Turkey (7.8 R and 7.5 R on 6 February 2023) on the seismically active areas in Romania in terms of gas emissions (radon, CO2). The main analysis methods for radon (we also included CO2) were applied to integrated time series and the use of anomaly detection algorithms. Data analysis showed that the effects of global warming led to variations in seasonal gas emissions compared to previous years. This made it difficult to analyze the data and correlate it with seismicity. Several of the cases presented require more in-depth analysis to determine the cause of the unusually high radon levels. The primary purpose of establishing the monitoring network is to use the gas emissions as seismic precursors, but the measurements are affected by the conditions under which the monitoring is conducted. In some cases, we are dealing with the effects of pollution, and in other cases, more extensive studies are required. One solution we plan to use is to expand the measurement points to locate the source of the anomalies and use weather data to determine the impact of global warming on the measurements. The main conclusions related to the development of a radon monitoring network and, in general, to the emission of gasses in earthquake-prone areas relate to the importance of the choice of equipment, monitoring location, and installation method.

Understanding the seismo–ionospheric coupling mechanism requires a quiet geomagnetic condition, as this represents an ideal situation to detect abnormal variations in the geomagnetic field. In reality, continuous interactions between solar wind and Earth’s magnetosphere create many fluctuations in the geomagnetic field that are more related to sun–magnetosphere interactions than to seismotectonic causes. A triaxial magnetometer was installed at the Muntele Rosu Observatory near the Vrancea seismic zone in 1996 to measure the local magnetic field. Since 2002, the data have become more consistent, allowing for the representation of long time series. Since then, variations have been observed on the eastern component (By) of the magnetic field, which sometimes overlaps with significant earthquakes. Previous studies have shown that high decreases in amplitude recorded on the By component of the magnetic field measured at Muntele Rosu have been accompanied by higher seismicity, while small decreases have been accompanied by lower seismic energy release. This research analyzes the geomagnetic data collected between September 2002 and May 2008 from two geomagnetic observatories, one located in the proximity of the Vrancea seismic zone and another one situated 120 km away. For each geomagnetic anomaly identified, the daily seismic energy released was plotted logarithmically, along with seismicity and Kp indices. Additionally, the daily seismic energy released was also plotted logarithmically for all earthquakes with Mw ≥3. To identify variations in the By component, datasets recorded at Muntele Rosu (MLR) were compared with those recorded at Surlari National Geomagnetic Observatory (SUA), to discriminate between global magnetic variations associated with solar activity and possible seismo–electromagnetic variations. The standard deviation (SDBy) was calculated for each anomaly recorded on the By component of the magnetic field and compared with the cumulative seismic energy release. To determine if this type of variation was present in other components of the magnetic field, the following ratios were calculated for all data recorded at Muntele Rosu: Bz/Bx, Bz/By, and Bz/BH. The size of the anomalies resulting from the standard deviation measured on the By component (SDBy) partially validates the relationship between the size of the anomalies and the seismic energy release during the anomaly. The relationship between the released seismic energy and the anomaly magnitude is vaguely respected, but these variations seem to follow two patterns. One pattern is described by smooth decreases, and the other pattern involves decreases where the By component varies significantly over short periods, generating decreases/increases in steps. It was noticed that seismic activity is greater for the second pattern. Additionally, using standard deviation measured on the magnetic field represents a great tool to discriminate external magnetic field variations from local, possibly seismo–magnetic variations.

Technological advancements and the appearance of low-cost Raspberry Shake seismographs have enabled the development of citizen science seismic networks in many areas worldwide. These networks can help reduce seismic risk and increase citizens’ understanding of seismology and earthquakes. Such a network exists in Bucharest, one of the cities in Europe that are struck and affected by strong Vrancea earthquakes. The paper aims to show that data from such networks can be used in both outreach programs and research studies. There are presented, for the first time, seismic observations collected over two years beginning in the summer of 2020 in the Bucharest area based on the low-cost seismometers from the citizen science Raspberry Shake network. A significant number of earthquakes from the Vrancea region were recorded by the Bucharest Raspberry Shake Seismic Network (BRSSN). Some of them were felt by Bucharest inhabitants. The National Institute for Earth Physics in Magurele (Romania) organizes educational events that promote geosciences among the population and presents the tools at its disposal for a better understanding of earthquakes and their effects, contributing this way to the development of the concept of citizen science. Citizens are the first witnesses to seismic events and the citizen science seismic network provides them with the first direct information about the event via web apps available for any internet-connected device. Their involvement as non-professional participants helps in providing data for scientists via questionnaire forms to improve scientific research for earthquake assessment. Since citizen seismometers are installed in urban areas, an analysis of the ambient seismic noise (ASN) was performed in addition to the analysis of recorded seismic events. The analysis indicates that the level of seismic noise is mainly controlled by human activities. At the same time, for one citizen seismometer installed in a school in Bucharest, the results show patterns of noise variations due to students’ activity.

The seismicity of Romania is mostly represented by earthquakes produced by the Vrancea seismic source with intermediate depth events (3 shocks/century with magnitude MW greater than 7.0). The seismic activity in Romania also includes crustal earthquakes. The crustal seismicity is more scattered and moderate compared to the intermediate-depth one. A stable and automatic method has been implemented in the real-time data acquisition and processing system ANTELOPE to estimate the seismic moment, the moment magnitude and the corner frequency of events recorded by the velocity sensors, using spectral analysis applied to S waves. The main goals are the independent estimation of the seismic moment and the common characterization for all events recorded by the National Seismic Network. The main target of this paper is represented by the fast estimation of moment magnitude MW and ground motion parameters that are derived using Gallo et al. (Bull Earthquake Eng 12:185–202, 2014) methodology and their validation with other magnitude determination algorithms existing at the National Institute for Earth Physics (NIEP). To test this new methodology, we have analyzed 331 seismic events, most of them being automaticaly located, and afterwards added a new, manually processed solution for events with ML ≥ 4.5, to obtain a larger interval of magnitudes.

In September 2022, two destructive earthquakes of moment magnitude (Mw) 6.6 (foreshock) and 7.1 (mainshock) occurred in Taitung County, south-eastern Taiwan. To understand their complex rupture processes, we analysed these earthquakes using the Potency Density Tensor Inversion method, which can stably estimate the rupture propagation process, including fault geometry, without overfitting the data. The analyses revealed that the major rupture of the foreshock propagated towards shallow depth, in a south–southwest direction, following an initial rupture that propagated towards the deeper part of the fault. The mainshock, with its epicentre on the north–northeast side of that of the foreshock, consists of two distinct episodes. During the first episode (0–10 s), the initial rupture propagated north–northeast, through a deep path, followed by the main rupture that propagated bilaterally in a north–northeast and south–southwest direction. The second rupture episode (10–16 s) started near the hypocentre of the mainshock, and the rupture propagated towards the shallow side of the fault. The results suggest that the stress concentration from both the foreshock and mainshock’s first rupture episode may have caused the second rupture episode in the high fracture surface energy area between the foreshock and the first rupture episode of the mainshock. The irregular rupture process of the foreshock and mainshock may reflect the heterogeneity of stress and structure in the source region.

The Vrancea slab, Romania, is a subducted remnant of the Tethyan lithosphere characterized by a significant intermediate-depth seismicity (60–170 km). A recent study showed a correlation between this seismicity and major dehydration reactions, involving serpentine minerals up to 130 km depth, and high-pressure hydrated talc deeper. Here we investigate the potential link between the triggering mechanisms and the retrieved focal mechanisms of 940 earthquakes, which allows interpreting the depth distribution of the stress field. We observe a switch from horizontal compression to vertical extension between 100 and 130 km depth, where the Clapeyron slope of serpentine dehydration is negative. The negative volume change within dehydrating serpentinized faults, expected mostly sub-horizontal in the verticalized slab, could well explain the vertical extension recorded by the intermediate-depth seismicity. This apparent slab pull is accompanied with a rotation of the main compressive stress, which could favour slab detachments in active subduction zones.

The parameters influencing the sea level measured with ultrasonic devices that are analyzed in this paper are the air temperature, atmospheric pressure and wind speed. As these variations are independent to each other and to the sea level, they can be removed from the measured sea level by applying a filtering algorithm based on independent component analysis (FastICA), adapted and improved for this application. The sound speed increases with temperature, so an internal temperature sensor is required to compensate for the sound-speed variation. Though this may improve the measurement accuracy, it is not enough to achieve the best results because there is a discrepancy between the internal sensor and the actual environment temperature. For high accuracy measurements, an external temperature sensor is required. In our case, we imported temperature datasets from a weather station, along with other datasets regarding atmospheric pressure and wind speed. The use of these external datasets, along with an algorithm based on principal component analysis (PCA) for error removal and the filtering algorithm based on FastICA for environmental phenomena extraction, allows us to achieve more accurate values for the Black Sea level in Constanta (2017–2020), independent of external influences.

Seismic velocity is the geophysical property that has a key role in characterizing dynamic processes and the state of the stress around the faults, providing valuable information regarding the change in the tectonic regime. The stress in the crust is an important indicator of the possible occurrence of a major earthquake, and the variation of seismic velocities, in time, can provide a clearer picture on the tectonic processes taking place in the region. In the crust, velocities change before, during, and after earthquakes through several mechanisms related to fault deformations, pore pressure, stress changes, and recovery processes. In this study, we investigate the possible correlation between the changes of seismic velocities (Vp/Vs) in time and the occurrence of moderate size crustal and intermediate depth earthquakes from the Vrancea region. Our findings show that there are no significant variations in Vp/Vs for the intermediate depth earthquakes, while crustal events have decreased seismic activity prior to the main earthquake and no high Vp/Vs anomalies. Our results indicate key aspects, and such analyses should be carried out in real-time to continuously explore any unusual pattern pointed out by the seismic velocity changes. Vp/Vs and their standard errors can also be used to describe seismic activity patterns that shape the tectonic evolution of the area.

The goal of this paper is to describe cluster analyses of the focal mechanism solutions estimated from local and teleseismic measurements and stress inversions to support the recent and previously published studies in the investigated region, that is the Vrancea seismic zone. We have applied different established clustering methods—e.g. hierarchical density-based clustering for applications with noise (HDBSCAN) and agglomerative hierarchical analysis—to the geographical coordinates, focal depths and parameters of the focal mechanism solutions of the gathered seismic events. We have attempted to develop a fully automated algorithm for the classification of earthquakes and to support the further investigation of stress inversions. This algorithm does not call for the setting of hyper-parameters by the users, thus the contribution from any bias of the user can be reduced significantly and the time required to carry out the clustering can also be decreased. In most cases, the resulting stress tensors are in close agreement with those we found in the literature.

We designed a convolutional neural network application to detect seismic precursors in geomagnetic field records. Earthquakes are among the most destructive natural hazards on Earth, yet their short-term forecasting has not been achieved. Stress loading in dry rocks can generate electric currents that cause short-term changes to the geomagnetic field, yielding theoretically detectable pre-earthquake electromagnetic emissions. We propose a CNN model that scans windows of geomagnetic data streams and self-updates using nearby earthquakes as labels, under strict detectability criteria. We show how this model can be applied in three key seismotectonic settings, where geomagnetic observatories are optimally located in high-seismicity-rate epicentral areas. CNNs require large datasets to be able to accurately label seismic precursors, so we expect the model to improve as more data become available with time. At present, there is no synthetic data generator for this kind of application, so artificial data augmentation is not yet possible. However, this deep learning model serves to illustrate its potential usage in earthquake forecasting in a systematic and unbiased way. Our method can be prospectively applied to any kind of three-component dataset that may be physically connected to seismogenic processes at a given depth.