The seismic activity in the Himalayan region results from the ongoing collision between the Indian and Eurasian tectonic plates. The Himalayan thrust zone, comprising critical fault zones like the Main Central Thrust (MCT), Main Boundary Thrust (MBT), and Main Frontal Fault (MFF), is highly seismically active, leading to numerous moderate-to-high magnitude earthquakes annually. Nepal, situated in one of the world's most seismically active continental collision orogenic belts, has experience numerous devastating earthquakes in its history. Earthquakes can result in wide-ranging devastation that includes other, cascading natural disasters, including avalanches, landslides and glacial lake outburst floods (GLOFs). The present study aims to identify ground surface deformations associated with the 2023 Nepal earthquake sequence using C-band radar interferometry from the ESA Sentinel-1A/B synthetic aperture radar (SAR) datasets. The earthquake sequence includes a mainshock (M_w 5.7 triggered at a hypocentre depth of 32.6 km) on November 3, 2023, followed by an aftershock (M_w 5.3 triggered at a hypocentre depth of 10 km) on November 6, 2023. The mainshock's influence radius, determined using shake maps from the USGS earthquake catalog, is 57 km. Because the aftershock's influence radius of 51 km which is smaller than that of the mainshock, we use that as the study radius. Differential interferometric SAR (DInSAR) is employed for this region, utilizing co-seismic single-look complex (SLC) datasets acquired on October 25 and November 6, 2023, respectively. The DInSAR analysis reveals a maximum reliable (regions with coherence ≥ 0.4) and atmospherically-corrected ground deformation of -79 mm. The most significant ground deformations are observed around the Sisne Himal glacial region and the mountain slopes of the Ragda region. Other areas with ground deformations are identified over the mountain slopes of the Guthi Chaur region. The topographic slope of these regions is ≥35° except for Sisne Himal glacial region as observed through ALOS PALSAR high-resolution terrain corrected digital elevation model (DEM) at 12.5m ground resolution. Analysis of pre- and co-seismic coherence images revealed decreased co-seismic coherence in certain locations within the influence radius. These areas are further investigated for soil liquefaction/cyclic mobility using the Temporal Difference Liquefaction Index (TDLI) using Landsat-8 and -9 datasets of October 27 and November 4, 2023 respectively. TDLI detects changes in soil moisture content after an earthquake event. The observed ground deformations indicate potential earthquake-induced slope failures including some of the locations with liquefaction/cyclic mobility susceptibility. This emphasises the importance of monitoring such vulnerable areas for enhanced seismic risk assessment and disaster preparedness.

How to cite: Mondal, S. K., Bharti, R., and Tiampo, K.: Seismic Deformations due to 2023 Nepal Earthquake Sequence using Satellite Remote Sensing Techniques, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17792, https://doi.org/10.5194/egusphere-egu24-17792, 2024.

Determining the arrival time of P waves is crucial in the context of earthquake early warning. Accuracy and calculating time play an important role in the earthquake parameter prediction. By using the initial power of the P-wave (IPP), STA/LTA, and artificial neural network, the arrival time of the P wave has been determined. 4,482 earthquake signals which occurred between 2021 to 2023 in the western of Java obtained by accelerometer were carefully investigated. The arrival time calculated by using Initial of P-wave (IPP) then compared with traditional STA/LTA picking and artificial neural network (ANN). Interestingly, IPP method can reduce background noise better than the other implying very clear P-wave to be proceed. It is found that IPP method exhibit less deviation, higher accuracy, and more precise suggesting candidate to apply as earthquake early warning system. This research contributes the reliability of seismic monitoring, thereby enhancing our ability to provide timely and accurate earthquake predictions for enhanced public safety.

How to cite: Kadnan, K., Djuhana, D., Handoko, D., and Florida Riama, N.: Automatic Determination Of P-Wave Arrival Time Using The Initial Power Of The P-Wave: A Case Study In Western Java, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7445, https://doi.org/10.5194/egusphere-egu24-7445, 2024.

Regarding foreshocks, it is believed that the percentage of the mainshock events with observable foreshocks is very low. But a recent study suggests that 15%–43% of large mainshocks have at least one foreshock. That is to say, if we can effectively identify foreshocks, it can help us carry out short-term and imminent earthquake forecasting, which is of great significance to the local people.

To identify the foreshock, the first step is to understand the characteristics. For example, the low b-value of the foreshock sequence, the cumulative number of foreshock sequences satisfying the anti-Omori law,  the foreshock sequences having similar focal mechanism solutions, and having the characteristics of migration towards the mainshock. Among them, the low b-value of foreshock sequence is widely recognized as a typical feature of foreshocks. However, this method is mainly based on the earthquake catalogs, and requires the number of earthquakes to meet the calculation conditions. Actually, it is difficult to obtain the b-value results in a short time after the earthquake.

In this study, we selected the foreshock event waveform recorded by the station closest to the earthquake epicenter and obtained the envelope function of the waveform. Each peak of the envelope can be regarded as an earthquake event. The amplitude and time of the function peak correspond to the magnitude and time of the earthquakes. Similar to the G-R relationship, the β-value, corresponding relationship between magnitude and peak number, is preliminarily obtained.

For all the M≥6.0 earthquakes with foreshocks in Chinese Mainland since 2010, we calculate the β-values of foreshocks and mainshocks using waveform envelope method. Most β-values of the foreshocks and mainshocks are calculated except the Yushu mainshock for its unsatisfactory data quality. In addition, we also apply this method to the M≥5.0 earthquakes occurred in Chinese mainland since 2022. The results indicate that the β-values of foreshocks are all lower than its mainshocks. If the β-value is less than 0.7, the earthquake can be considered as a foreshock. The β-value could also provide us with some hints about the magnitude and time interval between the foreshock and mainshock.

How to cite: Han, Y., Zang, Y., Xie, M., and Meng, L.: Research on the foreshock characteristics of earthquakes in Chinese Mainland using waveform envelope method, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3831, https://doi.org/10.5194/egusphere-egu24-3831, 2024.

The near field condition in seismic events is characterized by its immediate proximity to the seismic source, and is widely proven that ground motion near a causative fault (Near field) can differ significantly from typical ground motion observed at greater distances (far field).

Features of near-fault ground motion are high vertical accelerations and the occurrence of high-amplitude, long-duration (2–5s) pulses observed in velocity–time and displacement–time histories aligned with the fault's normal direction. These features and the other effects linked to this condition are critical factors in causing potential damage to structures as the seismic motion in the near-field can subject structures to seismic demands that differ from the design criteria, primarily in terms of intensity and the nature of ground motion.

As the seismic hazard quantifies the ground motion expected at a given site, understanding and predicting near-field effects are vital for seismic hazard assessment, structural design, and risk mitigation in the areas where near-field conditions occur.

This study aims to investigate the near-field effects in seismic events by employing two-dimension numerical simulations carried out with FLAC 2D Finite Difference Code, to reproduce the features observed during a real earthquake occurred.

The selected area is the Norcia plain, one of the intermountain basins widely present in Central Italy—a context of significant interest due to its association with high seismic hazard and high exposure in urban agglomerations. Several active seismic stations have recorded the last important seismic sequence (Central Italy 2017-2018) in particular the third and largest event on 30^th October (6.5 Mw), whose epicenter was located close to Norcia (4 km). The validity of near-field conditions for this event has already been established by previous studies.

Other scientific studies have been carried out in this direction in similar geological contexts with other software and the advancing here proposed is performing simulations using a non-horizontal interface geometry to apply the seismic input, with both horizontal and vertical components. The simulations consider a geological and tectonic model with few variables changing to provide a comprehensive understanding of how results may be affected by the knowledge of the geological and geotechnical setting, gained from basic studies.

This study could have important implications to suggest an updating of the seismic code and the general approach in the seismic design of structures located in the near field domain, for a careful and reliable assessment of seismic risk.

How to cite: Caciolli, M. C., Giallini, S., Pagliaroli, A., Barchi, M. R., De Franco, R., Fiorentino, G., Mancini, M., and Moscatelli, M.: “Evaluation of the near field effects trough numerical modeling: the case of Norcia (Italy)”, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17952, https://doi.org/10.5194/egusphere-egu24-17952, 2024.

The phenomenon of liquefaction is nowadays sufficiently understood in terms of phenomenology and predisposing conditions. However, a better assessment of liquefaction risk is necessary to mitigate its effects and guide land-use planning choices, particularly in the context of post-earthquake reconstruction.

This evidence comes from some recent events (e.g., New Zealand, 2010-2011; Emilia-Romagna 2012; Palu, 2018), in which liquefaction induced effects were, in some instances, considerably more severe than expected. This is the case of Terre del Reno (Emilia-Romagna region, Italy) which experienced significant liquefaction phenomena during the 2012 Emilia-Romagna earthquake sequence, characterized by two main events: Mw 6.1 and 5.9. In this area sand eruptions, settlements, lateral spreading, and ground fractures were observed, resulting in extensive and irregularly distributed damage to structures and infrastructure.

This study deals with the evaluation of liquefaction susceptibility and development of liquefaction hazard map in complex stratigraphic condition through an integrated method and multilevel approach. The study area is characterized by complex geologic conditions and abrupt slope changes, typical of riverbank-channel systems.

The analysis of liquefaction potential was conducted using simplified semi-empirical methods. The safety factor against this phenomenon was estimated at different depths, relying on  soil properties obtained from penetrometric tests and seismic input. In addition to the calculation of liquefaction potential, the study also addressed the phenomenon of lateral spreading due to liquefaction. To date, the delimitation and representation of area prone to lateral spreading is not yet ruled by guidelines for the mitigation of liquefaction risks. Therefore, this study employed an empirical methodology based on original criteria and procedures to establish the perimeter of such areas.

The cross-analysis between the prediction of indicators of liquefaction potential and the evidence of damage found following the May 20, 2012 earthquake (Mw 6.1) showed a clear correlation between slope and damage frequency, suggesting the possibility of applying an empirical method to define the probability of lateral spreading occurrence.

How to cite: Giallini, S., Fortunato, C., Sirianni, P., Baris, A., Caciolli, M. C., Fabozzi, S., Gaudiosi, I., Mancini, M., Martelli, L., Modoni, G., Moscatelli, M., Paolella, L., Simionato, M., Spacagna, R. L., Stigliano, F., Tentori, D., and Varone, C.: Assessment of liquefaction susceptibility under complex stratigraphic conditions: the case of Terre del Reno (FE), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13175, https://doi.org/10.5194/egusphere-egu24-13175, 2024.

The geometry and soil conditions at a site can strongly influence the ground shaking induced by an earthquake and an important part of seismic hazard assessment is therefore to characterize such site effects. At stiff sites and for weak ground motions, the soil response behaves linearly, however, for larger ground motions and softer sites the site response becomes nonlinear and thus more challenging to assess. Because recordings of such strong shaking are rare, nonlinear soil parameters are often defined from laboratory measurements and numerical simulations. However, the increase of available ground motion data, in particular from sites with recordings from both weak and strong motions, has increased the possibility of deriving nonlinear site parameters directly from empirical data. In this study, we take advantage of the comprehensive KiK-net network in Japan, consisting of stations with recording instruments at both surface and at depth. We use the surface-to-depth ratios of each event recorded by each station, to systematically capture the empirical effects of nonlinear soil response on the local response. Station-specific parameters for the degree of nonlinearity and PGA thresholds for the onset of nonlinear behaviour are then derived and the statistical correlation between these parameters and a selection of geotechnical and geological indicators are investigated. Our results show that - although finding site parameters suitable for predicting nonlinear site effects remains challenging as the nonlinear soil-behaviour is largely site-specific - proxies describing the depth to bedrock, sediment thickness and characterization of the shallowest part of the soil layer display a certain degree of potential for prediction.

How to cite: Loviknes, K., Bergamo, P., Cotton, F., and Fäh, D.: Assessment of the correlation between site-specific nonlinear soil behavior for Japanese KiK-net stations and geological, geotechnical parameters , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16964, https://doi.org/10.5194/egusphere-egu24-16964, 2024.

A spectral decomposition approach is applied to separate the site amplification, path attenuation, and source parameters of earthquakes that occurred in Madagascar between 2011 and 2013. Concerning source parameters, the stress drop is derived from the source spectra, fitting the Brune model. Our findings indicate an increase in stress drop with magnitude, ranging from 0.001 to 0.1 MPa. Additionally, the results unveil a consistent attenuation curve diminishing with distance, characterized by a swift decay at higher frequencies below the 1/R^2 decay function. The frequency-dependent behavior of the estimated quality factor for S waves suggests rapid damping and slow dissipation.Regarding site amplification, a robust agreement is noted between resonance frequency obtained through GIT and the H/V spectral ratio. Furthermore, we explore the correlation between site amplification and geological characteristics such as sediment depth and lithology.

How to cite: Ravoson, M. and Razafindrakoto, H. N. T.: Application of Spectral Decomposition approach to examine the attenuation, source parameters, and site effects in Madagascar, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19444, https://doi.org/10.5194/egusphere-egu24-19444, 2024.

Seismic hazard maps deriving from probabilistic seismic hazard analysis (PSHA) collect the intensities, in terms of one ground motion intensity measure (𝐼𝑀), that, at each site taken individually, have the same probability of being exceeded in a time interval or, equivalently, exceedance return period. In the case of Italy, there are three authoritative nationwide PSHA studies that can be currently considered of interest. Given the return period, they provide hazard maps that can differ even significantly in some areas of the country. This contribution pertains to the assessment of the fractional area of Italian territory where 𝐼𝑀 values from hazard maps have been exceeded, at least once, due to seventy-one historical mainshocks that occurred in the country from 1117 to 1968. Ground shaking data for such events were derived from a recently developed large database of ShakeMap inferred from macroseismic intensity data. Such database is not complete, with the Italian catalogue (Catalogo Parametrico dei Terremoti Italiani; CPTI) counting more than two thousand mainshocks in that time interval, yet it is, to date, the highest level of information on shaking data due to historical events. For each hazard model, the exceedance area was quantified considering hazard maps with four return periods, that is, 50yr, 475yr, 975yr and 2475yr, and three 𝐼𝑀𝑠, that is, peak ground acceleration and pseudo-spectral acceleration associated to a vibration period of 0.3s and 1s. It was found that, based on the available regional shaking estimates for historical earthquakes in Italy, the fraction of the country exposed to at least one exceedance, in almost one thousand years, is comparable, given return period and 𝐼𝑀, for all the hazard models, despite their apparent differences. Such comparability was also found when considering instrumental, rather than historical, earthquakes that occurred in Italy in a continuously monitored time interval spanning twelve years. In this case, the exceedance area was quantified considering ShakeMap data for nineteen mainshocks that occurred from 2008 to 2019 according to CPTI, and therefore the dataset can be deemed complete. Thus, the fraction of the country possibly subjected to exceedance of 𝐼𝑀 values from hazard maps according to ShakeMap estimates was also compared with its expected value from PSHA, something that depends only the return period the maps refer to. It was found that, for each return period, the estimated fractional exceedance area in the available twelve years is one order of magnitude lower (or slightly less) than the expected value according to all PSHA studies.

How to cite: Cito, P., Vitale, A., and Iervolino, I.: Exceedance of probabilistic seismic hazard maps in Italy, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-21465, https://doi.org/10.5194/egusphere-egu24-21465, 2024.

The distribution of earthquakes in time and space is clustered and may exhibit a non-stationary behaviour. The impacts of non-stationarity are further amplified when the observation window is short compared to the timescales of the underlying tectonic process, such as in regions of low-seismicity. This can preclude a robust statistical analysis for PSHA models, which commonly assume stationary Poisson models. We investigate the performance of forecasts for PSHA, such as smoothed-seismicity models (SSM), with respect to the available training data. We design bootstrap experiments for multiple pairs of consecutive training/forecast windows of a catalogue to: (i) analyse the lowest available amount of training data for which SSM performs spatially better than the least-informative Uniform Rate Zone (URZ) model; (ii) characterise the temporal variability of rates in terms of their over-dispersion and non-stationarity. The results show rate variability up to 10 times higher than predicted by Poisson forecasts, and demonstrate the impact of non-stationarity when assuming a constant mean rate derived during a training period for forecasting purposes. Analytical distributions are used to describe rate variability, which are conditioned on the information available from a training period. Furthermore, we devise a data-driven method based on strain-rate maps to spatially delineate URZs, under the assumption that the strain-rates field is related to the time scales of earthquake occurrence and interaction. For each URZ, a rate temporal distribution is inferred from the training events within it. We provide forecasts for the update of the New Zealand Seismic Hazard Model that have increased rates by up to 10 times higher in extensive low-seismicity regions compared to optimised SSMs. The new forecasts are implemented as negative-binomial distributions in the hazard integral. For a 10% exceedance probability in 50 years, the use of URZ with rate variability descriptions increases the expected PGA by up to 0.16 g in low seismicity regions (e.g. Auckland, Dunedin) compared to SSM. Our results highlight the relevance, as well as the feasibility, of incorporating analytical formulations of seismicity that go beyond the inadequate stationary Poisson description of seismicity.

How to cite: Iturrieta, P., Gerstenberger, M., Rollins, C., Van Dissen, R., Wang, T., and Schorlemmer, D.: Accounting for earthquake rates’ variability through Uniform Rate Zone forecasts in the 2022 Aotearoa New Zealand Seismic Hazard Model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20762, https://doi.org/10.5194/egusphere-egu24-20762, 2024.

How to cite: Sadeghi-Bagherabadi, A., Fülöp, L., and Korja, A.: The Comparison of the 2020 European and the Finnish Seismic Hazard Models at Two Nuclear Power Plant Sites in Finland , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15663, https://doi.org/10.5194/egusphere-egu24-15663, 2024.

Ground motion models (GMMs) play a pivotal role in both deterministic and probabilistic seismic hazard assessments, which are essential for identifying the seismic safety of nuclear power plants. In regions with abundant seismic data, especially strong earthquake records, GMMs could be empirically derived. However, in areas like South Korea with scarce strong earthquake records, development of empirical GMMs is impractical, leading to the utilization of alternative methods such as stochastic simulations. There have been a few GMMs developed in South Korea, all of which relied on stochastically simulated motions. In this study, GMMs are developed for rock sites in South Korea using the hybrid empirical method (HEM) suggested by Campbell (2003). Western United States (WUS) is selected as a host region and five Next Generation Attenuation (NGA)-West2 GMMs are used as GMMs of the host region. The seismological parameters employed in the simulation, including effective point source distance, source and path duration, and path attenuation, duly encompass the findings of recent studies. The high-frequency spectral attenuation parameters, kappa, utilized as site attenuation parameters in ground motion simulations for the target region, are estimated in this study. It is primarily estimated using the classical method proposed by Anderson and Hough (1984). Additionally, the estimation process considers the standardized procedure and the recommended lower bound magnitude decisions put forth by Ktenidou et al. (2013) and Van Houtte et al. (2014), respectively. Since the shear wave velocity for bedrock is considered to be 760 m/s in South Korea, the site amplification functions have been applied with reference to this velocity for both the host and target regions. The adjustment factors obtained from simulated ground motions in both the host and target regions are applied to adjust NGA-West 2 Ground Motion Models (GMMs). Derived GMMs are for magnitudes from 5.0 to 7.5 and rupture distances from 10 to 500 km. Median GMMs are provided with aleatory standard deviations. Predictive GMMs are compared with observed ground motions from the available earthquake records for moment magnitudes 5.0 and 5.5. The notable advantages of the GMMs developed in this study are as follows: Distinct from previous researches utilizing stochastic methods, the implementation of HEM served to complement the limitations inherent in stochastic approaches such as lack of near-source ground motion characteristics. Defining the sites where GMMs are employed at Vs₃₀ = 760m/s enables the derivation of seismic motions applicable to rock layers having Vs₃₀ of 760m/s. Since aleatory standard deviations are quantitatively defined, they can serve as the sigma parameter within GMMs in Probabilistic Seismic Hazard Analysis (PSHA).

How to cite: Park, S.: Ground Motion Models for rock sites in South Korea, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3265, https://doi.org/10.5194/egusphere-egu24-3265, 2024.

Ground Motion Models (GMM) are an essential component in seismic hazard analysis. In recent years, Machine Learning (ML) methods have made significant progress in advancing GMM research. However, their effectiveness is hindered by certain limitations. Firstly, some ML methods, such as neural networks, lack interpretability, which poses challenges for researchers and engineers. Secondly, these methods may have limited extrapolation capabilities and may struggle to accurately capture data distribution characteristics when predicting scenarios beyond the scope of the training data. To address these shortcomings, we propose a physics-informed symbolic learning (PISL) approach for constructing GMMs. Symbolic learning extracts an interpretable expression from data using sparse regression. To tackle sparse data issues in the near-field of large earthquakes, we have introduced a near-field saturation factor into the model. This factor contains physics information and guides the model to accurately capture near-field saturation properties even with limited data. We used the NGA-West2 database to develop GMMs for Peak Ground Acceleration (PGA), Peak Ground Velocity (PGV), and Pseudo-Spectral Acceleration (PSA) within the period range of 0.01s-10s. The study conducted comparative analyses using M_w-scaling, R_JB-scaling, and V_S30-scaling cases with empirical regression models. Model stability was assessed through residual calculations and standard deviation analysis. The results indicate that our model predictions are comparable to those of classical empirical models, with no significant differences in residual distribution and standard deviations. The extrapolation ability of the symbolic learning approach was validated using seismic events outside the training data, including the Wenchuan earthquake and Turkey earthquake. In conclusion, the method described above integrates physical knowledge and data-driven approaches, simplifying the equations while achieving similar results. Symbolic learning methods then provide a new perspective for the application of ML in engineering seismology.

How to cite: Liu, X., Chen, S., Fu, L., Li, X., and Cotton, F.: Ground motion prediction using Physics-Informed Symbolic Learning (PISL), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9728, https://doi.org/10.5194/egusphere-egu24-9728, 2024.

Ground Motion Prediction Equations (GMPEs) are semi-empirical relationships commonly used to model ground motion intensity measures, such as peak ground accelerations (PGA) and velocity (PGV) and pseudo-spectral amplitudes (SA) at a specific site, conditional to earthquake parameters such as magnitude, source-site distance, and local site amplification effects. They are used for several seismological and earthquake engineering applications, such as probabilistic seismic hazard and rapid response (ShakeMap) analyses.

In the last decade, the very densely populated volcanic area of Campi Flegrei in Southern Italy, has experienced an intense seismic activity, related to the inner-caldera resurgency and ground uplift, with more than eight-thousand recorded events. During the last two years, both the uplift rate and the seismic activity accelerated, leading to the occurrence of about forty events with duration magnitude between 2.5 and 4.2 whose shaking has been well perceived by the population. Some of these earthquakes showed ground motion intensity (i.e., spectral pseudo-acceleration, SA), leading to non-negligible seismic actions on structures at specific natural vibration periods. Nevertheless, even structures located at less than few km from the source did not sustain significant damage.

Due to the strong discrepancy between observed and predicted data using literature GMPEs for Campi Flegrei, in this work, ad-hoc GMPEs were calibrated for PGA, PGV and 21 SA at periods T  in the range [0.01s 10s] . Data come from the largest magnitude events (38) occurred in the last two years, and recorded at thirty-four accelerometric and/or velocimetric stations located at epicentral distance R_epi < 40 km. The events were re-located with a probabilistic, non-linear approach and the moment magnitude was computed from displacement spectrum amplitudes. Results indicate that the re-calibrated GMPEs expect larger PGA and PGV very close to the source (R_epi < 5 km) and a higher attenuation at larger distances with respect to the existing attenuation relations for Italian volcanic areas.

The retrieved GMPEs for the Campi Flegrei caldera have been used to map the minimum magnitude of close-by earthquakes expected to exceed code-mandated design (elastic) seismic actions on structures; so-called strong earthquakes. This minimum magnitude is found in the range 4.1-5.1, depending on the ground motion intensity measure, for sites located within or in proximity of the caldera and earthquakes occurring at epicentral distances smaller than 1km.

How to cite: Scala, A., Cito, P., Strumia, C., Scotto di Uccio, F., Festa, G., Iervolino, I., Zollo, A., Convertito, V., Elia, L., Emolo, A., Bobbio, A., and Iaccarino, A. G.: Ground Motion Prediction Equations for the Campi Flegrei volcanic area, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11465, https://doi.org/10.5194/egusphere-egu24-11465, 2024.

The ground motion intensity measures are often obtained using Ground-Motion-Prediction-Equations (GMPEs) or more in general referred to Ground Motion Models (GMMs), which are empirical mathematical equations that relate the ground motion toseismological parameters (e.g., magnitude, source-to-site distance,focal depth and the average shear-wave velocity in the uppermost 30 m, Vs30). GMPEs are worldwide used as a tool for seismic hazard assessment and seismic design, usually derived from past earthquakes records through linear regression and predefined functional forms.

In the last 20 years, the application of artificial intelligence in earth sciences has been significatively applied to solve nonlinear problems that cannot be explained by empirical approaches.

The last seismic events in central Italy, including the earthquakes in L'Aquila (2009) and the Amatrice-Visso-Norcia sequence (2016-2018), have provided a substantial dataset comprising approximately 34,000 waveforms, contributing to the creation of a robust and accurate database(central Italy dataset).

In this work, we leverage the valuable data compiled by their work for the calibration of prediction models based on supervised Machine Learning (ML) to predict and evaluate the ground motion intensity measures and comparison the result with an existing GMPE currently used in Italy (ITA18).

Several ML regression algorithms are systematically examined and validated, the XGBoost algorithm is identified as the optimal choice, offering a balanced performance in terms of error minimization, interpretability, and computational efficiency. The evaluation encompasses the estimation of diverse Intensity Measures (IMs), such as Peak Ground Acceleration (PGA), Peak Ground Velocity (PGV), and Acceleration Response Spectra at 5% damping (SA) across different time periods (e.g., 0.1 second, 1.0 second, and 2.0 seconds). The ML model developed in this research demonstrates high accuracy, exhibiting notable improvements compared to the Italian Ground Motion Prediction Equation (GMPE). These advancements suggest the model's efficacy in enhancing seismic hazard assessment. Moreover, the versatility of this model extends beyond the study area, as it can be applied to various worldwide geological contexts, provided seismic data is available. The outcomes of this work not only contribute to refining local seismic risk evaluations but also offer valuable insights for seismic studies in diverse global regions.

How to cite: Zaragoza Alonzo, A. D., Zambrano, M., Luzi, L., and Tondi, E.: Using Supervised Machine Learning Algorithms for Ground Motion Prediction: A Comparison with the Traditional functional form Approach in Central Italy, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12618, https://doi.org/10.5194/egusphere-egu24-12618, 2024.

In recent decades, Central Italy has faced several seismic sequences, such as the one of the 2016-2017, started with the Amatrice mainshock (Mw6.2) in August 2016, followed by the Visso (Mw5.9) and Norcia (Mw6.5) earthquakes in October 2016 (hereafter AVN). Given the region's frequent seismic activity and heigh seismic risk, the use of ground-motion simulations becomes crucial for seismic risk assessment and earthquake engineering applications. Ground motion characteristics have been already investigated in the area for the Mw6.2 Amatrice and Mw6.5 Norcia earthquakes, using stochastic and numerical approaches.

Moreover, to predict ground motion from hypothesized events, it is fundamental to define the attenuation characteristics of the area and the relationship with the ground motion models.

In this study, we investigate the variability of seismic wave attenuation in strong-ground motion simulation in the Central Apennines, employing stochastic simulations. Initially, we compute the quality factor Q values for the region, deriving the total attenuation Q as a function of frequency for the 2016-2017 seismic sequence. To visualize the variation of the attenuation, a 2D kernel-based imaging of coda-Q space is applied, confirming an increment in attenuation during the AVN sequence in the fault plane zones.

Then, we incorporate the acquired frequency-dependent Q values as input parameters into the simulations of ground motion. This methodology replicates stochastically strong-ground motion at high frequencies, reproducing horizontal and vertical accelerograms and using as input information from the source (stress drop, rupture time, slip distribution and radiation pattern). The estimations are then correlated and validated against observed peak ground accelerations and spectral accelerations for the Amatrice and Norcia fault planes, followed by a comparison with the ground motion prediction equations used for the region.

How to cite: Gabrielli, S., Akinci, A., Gutierrez, C., Ojeda Vargas, J., Arriola, S., and Ruiz, S.: Variability in Anelastic Attenuation: Examining Temporal and Spatial Influences on Ground Motion Characteristics in Central Italy, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9397, https://doi.org/10.5194/egusphere-egu24-9397, 2024.

Recent significant seismic events, namely the Zagreb M_W5.3 and Petrinja M_W6.4 earthquakes in 2020, have highlighted the critical need for enhanced seismic hazard assessment. To facilitate a more accurate assessment of seismic hazard, it is imperative to refine and adjust input parameters. Ground Motion Prediction Equations (GMPEs) assume a pivotal role in this aspect. However, the accurate allocation of GMPEs for specific regions necessitates an extensive database comprising strong motion (SM) recordings. This proves challenging for areas characterized by moderate seismic activity, such as Croatia. In response to this challenge, we have established the first systematic SM digital database, continually updated to address this gap. While the BSHAP database (Salic et al., 2017) was formally recognized as the initial Croatian strong motion database, it primarily contained analogue waveforms, often falling short of satisfactory quality standards. Our database confines its scope to the geographical boundaries of 41.2°N – 47.7°N and 12.5°E – 20.5°E. Comprising over 150 good-quality recordings from 2020 to the present day, with magnitudes equal to or surpassing 3.5, this database serves as a valuable resource. In this study, we tested various widely used Ground Motion Prediction Equations (GMPEs) to define the most suitable models. The findings of this investigation lay the foundation for further GMPE development tailored to the Croatian SM database, leveraging the Hybrid Empirical Method (HEM).

References:

Salic, R., Sandikkaya, M.A., Milutinovic, Z. et al. BSHAP project strong ground motion database and selection of suitable ground motion models for the Western Balkan Region. Bull Earthquake Eng 15, 1319–1343 (2017). https://doi.org/10.1007/s10518-016-9950-3

How to cite: Lončar, I., Stanko, D., and Markušić, S.: Croatian strong motion database: ground motion prediction equations tests, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9154, https://doi.org/10.5194/egusphere-egu24-9154, 2024.