In the published version of the article, the caption of Figure 3, page 6, incorrectly identifies two of the scales as Fahrenheit scales.
To establish a robust reference framework for future seismic hazard evaluations, we introduce the first deep shear wave velocity (Vs) model of the Venice Lagoon (Italy), developed through seismic noise interferometry. A seismometer was deployed in Venice’s historic center, and its continuous noise recordings were cross‑correlated with data from the sole permanent station located on Lido Island. Using seismic noise processing methods and joint inversion of Rayleigh wave group and phase velocity dispersion curves, we obtained a detailed deep Vs profile of the Venice Lagoon subsurface, extending down to the engineering bedrock and the base of the Pliocene sequence. Our model shows strong consistency with geological information from deep boreholes and seismic reflection surveys. Additionally, the determined engineering bedrock depth corresponds well with existing models of the deep alluvial Po Plain from earlier research. This study delivers the first comprehensive deep Vs reference model for Venice, establishing an essential baseline for future seismic response assessments and risk mitigation initiatives designed to safeguard the city’s priceless cultural heritage.
Geological CO2 storage in sedimentary basins associated with depleted hydrocarbon reservoirs represents a relevant strategy for climate change mitigation. In Colombia, the Valle Medio del Magdalena Basin, with a long history of hydrocarbon exploration and production, has been identified as a potential target region, particularly the Campo Colorado field. An additional factor supporting the selection of Campo Colorado is its proximity to the Barrancabermeja refinery, a major CO2 emission source. However, baseline seismic studies are required to assess its feasibility from a seismic risk management perspective. Given that the Valle Medio del Magdalena is tectonically active, a regional seismic characterization was conducted using the Colombian National Seismological Network catalog. The analysis focused on Campo Colorado within a 100 × 100 km polygon and included the Bucaramanga Seismic Nest as a comparative reference. Results indicate that progressive improvements in the seismic network enhanced catalog completeness, while the estimated b‑value for cortical seismicity in Campo Colorado, slightly higher than that estimated for the Bucaramanga Seismic Nest. Higher probabilities of occurrence for cortical events with magnitudes below 5.0 are observed in the study area. The estimated fractal dimension (Dc = 1.45) indicates a moderately clustered spatial distribution of seismicity, consistent with the regional structural framework and providing complementary information for the characterization of seismicity patterns. No significant temporal anomalies were identified, indicating that current seismicity is predominantly related to natural tectonic processes. Strengthening the local seismic network is recommended to improve baseline characterization, support real‑time monitoring, and facilitate future geological CO2 storage initiatives in the region.
The Earth’s ionosphere is a layer of ions and electrons embedded in the neutral atmosphere and extending from about 50 km to 1000 km of altitude. This layer of plasma is a fundamental component of the near-Earth space environment and plays a crucial role in the propagation of radio waves and satellite signals. Its retarding and refractive effects directly influence a wide range of scientific and technological applications, including satellite navigation, telecommunications, remote sensing, over-the-horizon radar, and space-based observations. As our society becomes increasingly dependent on space-based infrastructure and services, the need for an accurate and reliable representation of ionospheric conditions becomes ever more important. Recognizing this need, the Committee on Space Research (COSPAR) and the International Union of Radio Science (URSI) established the International Reference Ionosphere (IRI) as the empirical standard model of the ionosphere. Based on a broad body of ground-based and space-based observations, IRI has become the internationally accepted reference for ionospheric specification, being widely used in science, engineering, and education (Bilitza et al., 2022). Its continuous development over the decades has relied on the integration of new datasets, the testing of improved modelling approaches, and the systematic validation of outputs against observations collected under different geophysical conditions and over diverse regions of the globe.This Special Issue provides an updated overview of current efforts to assess and further improve the IRI model. The contributions collected in this issue address key aspects of IRI development, including the proposal of new formulations for specific ionospheric parameters and regions, the evaluation of model performance using new ground-based and space-based datasets, the assimilation of real-time or retrospective observations, and examples of applications that highlight the practical relevance of IRI for both research and operations. The twelve contributions included in this Special Issue are organized into three main thematic sections: TEC observations and comparisons and evaluation of IRI; New inputs for IRI; and Representation of irregularities. Together, these papers provide a broad and timely picture of the diversity of present-day IRI-related research and of the directions along which further model development is expected to proceed. The first section, TEC observations and comparisons and evaluation of IRI, focuses on model assessment against observations and on methodologies that improve the derivation of reliable ionospheric quantities from measurements. Mondal et al. (2025) compare total electron content (TEC) predictions from IRI-2016 and IRI-2020 over Indian near-equatorial and equatorial ionization anomaly regions, providing a detailed evaluation of the NeQuick and COR2 topside options during low solar activity conditions. Adero et al. (2026) investigate low-latitude topside ion composition using complementary COSMIC-2 and ICON observations and compared the measured O+ and H+ fractions with IRI-2020 predictions, showing that the model reproduces the large-scale climatological structure while still exhibiting systematic discrepancies in the transition region and around dawn and dusk. Kenpankho et al. (2026) focus on GNSS receiver-bias modelling for near-real-time TEC monitoring over low-latitude Thailand, demonstrating the importance of local bias correction for improved TEC estimation. Keokhumcheng et al. (2026) examine TEC fluctuations and their implications for GPS signal delay over Thailand during the ascending phase of Solar Cycle 25, emphasizing the practical relevance of ionospheric variability for satellite navigation systems. The second section, New Inputs for IRI, presents contributions that offer new observational benchmarks, reconstruction methods, and modelling tools of direct relevance to future IRI developments. Moses et al. (2026) provide a comparison of sporadic-E (Es) characteristics derived from coincident COSMIC-2 radio occultation and digisonde observations, contributing important information for future Es occurrence and intensity modelling. Yenen et al. (2026) apply the IONOLAB-Fusion computerized ionospheric tomography framework over Africa and Türkiye, demonstrating that advanced four-dimensional reconstructions can improve regional electron density specification even in areas with sparse data coverage. Gulyaeva and Shubin (2026) explore the use of different two-dimensional climatological models within the IRI-Plas framework, showing the potential of alternative foF2 and hmF2 formulations to improve the representation of electron density profiles and TEC. Erdem Kocak and Arikan (2026) present IONOLAB-RAY, a flexible three-dimensional ray-tracing tool able to synthesize both virtual and real-height ionograms, thus linking empirical ionospheric modelling to radio-wave propagation studies and ionogram interpretation. Wang et al. (2026) introduce a global foF2 prediction model based on COSMIC-1 and COSMIC-2 radio occultation data and an interpretable XGBoost approach. The model reproduces key ionospheric structures and, when validated against independent GRACE and GIRO observations, outperforms IRI-2020 and other reference models. By combining predictive skill with SHAP-based interpretability, this study offers an interesting new data-driven contribution to future improvements in foF2 modelling. The third section, Representation of irregularities, is about ionospheric variability under disturbed or highly structured conditions. Pansong and Kenpankho (2026) compare the ionospheric effects of two matched intense equinoctial geomagnetic storms during Solar Cycles 24 and 25, documenting marked differences in TEC, foF2, and hmF2 responses across low latitudes. Lukianova (2026) addresses the high-latitude F-region ionosphere with a regional numerical model and corresponding comparisons with Swarm satellite observations, showing how large-scale structures such as the tongue of ionization and the polar hole can be reproduced by physics-based modelling, thereby complementing the smoother morphology of empirical climatologies. Hegy and Abdelrahman (2026) evaluate neural-network forecasting of geomagnetic indices during Solar Cycle 25, a topic of relevance for operational space-weather prediction and, more broadly, for the specification of disturbed ionospheric conditions.Taken together, the papers in this Special Issue show how the IRI development continues to rely on the close interaction between empirical modelling, new observational datasets, advanced reconstruction techniques, and studies of disturbed ionospheric behaviour. The Special Issue brings together contributions that range from the validation of IRI outputs to the introduction of new tools and inputs that may support future model improvements. We warmly thank all authors for their valuable contributions. We are equally grateful to the reviewers, whose expertise and constructive comments helped to improve the quality of the papers collected in this issue. Finally, we sincerely thank the editorial team of Annals of Geophysics for their invaluable support and assistance throughout the preparation of this Special Issue.
Seismic Microzonation studies of Level 3 (SM3) require a detailed and spatially homogeneous characterization of subsurface conditions, typically obtained by integrating geognostic and geophysical datasets. However, the distribution of existing investigations (boreholes (S), Multichannel Analysis of Surface Waves (MASW), Down‑Hole test (DH), Horizontal to Vertical Spectral Ratio (HVSR)) is often irregular and clustered and, consequently, insufficient to represent the geological and geotechnical variability of the entire study area. This is particularly true in those geological contexts, such as volcanic settings, characterized by strong lateral and vertical heterogeneities. This paper presents a reproducible GIS‑based approach for designing an optimal investigation plan for SM3 through the construction of a regular network of Control Points (CPs). The methodology is entirely implemented using open‑source GIS tools and consists of: (i) generating two regular grids (size: 500 m and 1000 m), (ii) extracting and cleaning grid centroids by removing those located outside SM3 areas, (iii) integrating additional CPs where necessary, and (iv) associating each CP with pre‑existing investigations within a significance‑based distance threshold and within the same SM1 stable or instable zone (stab/instab). The method is tested in the Etnean area (Sicily, Southern Italy), where complex volcanic architectures strongly influence the local seismic response. Results indicate that the CP network allows for a rapid identification of data gaps and supports a rational design of new investigations, ensuring homogeneous spatial coverage and improving the reliability of the SM3 subsurface model. This GIS‑based framework provides a transparent and fully reproducible workflow that can be applied to any SM3 municipality studies at national scale.
The Toba volcanic caldera represents a highly heterogeneous volcanic system where seismic wave propagation is strongly influenced by complex subsurface structures. This study presents an integrated analysis of resonance and wavefield evolution using the Finite Difference Time Domain (FDTD) method combined with observed seismic data. A physics‑based subsurface model derived from the CRUST 1.0 dataset incorporates spatial variations in P‑wave velocity (Vp), S‑wave velocity (Vs), and density (𝜌). Frequency‑domain analysis reveals a dominant resonance band at 1.7‑1.9 Hz, in close agreement with theoretical predictions from a layered medium, with deviations below 10%, highlighting the primary control of near‑surface sediment thickness and shear‑wave velocity. Time‑domain simulations further exhibit prolonged coda waves, particularly in the central caldera, indicating efficient energy trapping and sustained multiple reflections. Cumulative energy analysis confirms that seismic energy is not rapidly attenuated but redistributed over time within the subsurface structure. Importantly, this study demonstrates that the observed wavefield complexity is governed by the coupled effects of structural resonance and vertical reverberation rather than true lateral scattering, aligning with the limitations of 1D horizontal layer extensions. By integrating frequency and time‑domain perspectives within a unified FDTD framework, this work provides new insights into seismic wave behavior in volcanic basins and offers a robust basis for improving seismic hazard assessment in the Toba caldera region.
Structured metadata defined in a machine-readable format are the foundation of FAIR data and open research. To describe rock deformation laboratory experiments, standardized metadata formats are currently missing. This limits the annotation of metadata describing the experiments, as well as the data and metadata produced during the experiments. To address this issue, we propose a structured metadata model, a configuration for OpenBIS data platform, and a web interface to annotate the metadata of experiments and data files. We developed a structured metadata model based on our user experience on performing rock deformation experiments. Then, we fully configured OpenBIS data platform using a docker container stack. Finally, we modified an existing Streamlit application allowing users to annotate metadata according to our structure, upload and annotate data files automatically. This approach allows us to annotate data and metadata produced in rock deformation laboratory experiments in a systematic and structured way to foster the adoption of FAIR principles and open science across rock deformation laboratories and hopefully across geosciences laboratories in general.
This study presents the Obsidian Sieve, an Excel-based tool designed to provide a simple, rapid, and flexible first-step provenance assessment for Central Mediterranean obsidians using only major and minor elements. Its development was grounded in a systematic review of provenance studies that report quantitative concentrations of major and-minor oxides/elements, including chlorine (Cl), still relatively uncommon in the literature despite its high discriminant potential. Seven bivariate systems traditionally employed in obsidian characterization were re-evaluated through a unified statistical workflow combining Kernel Density Estimation, Mahalanobis distance, 97.5% confidence ellipses, and quantitative measures of cluster overlap. This comparative analysis demonstrates that the Na2O vs Cl diagram provides the clearest and most stable separation among the four main Italian sources, yielding the highest Discriminant Power Index and uniquely displaying non-overlapping confidence fields. These results form the statistical basis for defining conservative Na2O-Cl windows that underpin the Obsidian Sieve. The tool assigns archaeological specimens to Monte Arci, Palmarola, Pantelleria, or Lipari with high specificity, while flagging atypical or ambiguous cases as Unknown, thus directing research towards further analysis with more advanced geochemical techniques. Because the Sieve is transparent, editable, and easily adaptable, users can extend its capabilities by incorporating additional published datasets, refining source windows, or testing alternative discriminant bivariate plots. In this way, the Obsidian Sieve provides not only a practical screening instrument but also a flexible framework for cumulative improvement as new analytical data become available.
The submerged Roman districts of the ripa puteolana present along almost 2 km of coastline between the center of the port of Puteoli (now Pozzuoli, Italy) and the Portus Julius constitute an underwater archaeological area of extraordinary importance. The headquarters of merchants and peregrini from every corner of the Mediterranean in the vicus Lartidianus, and the endless rows of horrea (warehouses) at the service of the fleets of Rome in the vicus Annianus have been only scarcely explored in the past, due to the heavy and impactful presence of the industries that formost of the twentieth century characterized the west coast of modern Pozzuoli. With the project "Tra terra e mare. Studi e ricerche nelle aree costiere dei Campi Flegrei / Between land and sea. Studies and research in the coastal areas of the Phlegraean Fields", born in 2021 from an agreement between the SABAP for the metropolitan area of Naples and the University of Campania Luigi Vanvitelli-with the collaboration, for the submerged areas, of the Scuola Superiore Meridionale, a systematic documentation program for the ripa puteolana has finally been launched. The underwater research conducted in 2021, 2022 and 2023 led to the reconstruction of the entire ancient waterfront, on the basis of an aerial photogrammetric survey calibrated with direct dives, to the identification of submerged horrea, granaria (granaries) and quays in the vicus Lartidianus and vicus Annianus, to the localization of a previously unknown peninsula, home to administrative premises of the harbour, and above all to the discovery of a submerged temple of the Nabataeans in Puteoli, unique outside of Nabataea. Together with these archaeological results, the collaboration established with the INGV enabled the collection of important new data on the relative sea level change in the bradyseismic area of Puteoli during the period between the 1st and 4th centuries AD, when an impressive effort was made to rebuild the breakwaters in response to the relative increase in sea level.
Geophysical prospecting was carried out with GPR (Ground Penetrating Radar) at the orchestra floor of the Ancient Theatre in Catania (Italy) to reconstruct the lithological pattern of the substrate and to identify cavities and/or channels used in the past for the drainage of rainwater or groundwater. On the basis of the data acquired, the existence of pre-existing north-south oriented drainage channels below the orchestra floor can be confirmed, now partly blocked by sediments that have accumulated over the centuries. In particular, interpretation of radar sections indicated the occurrence, in the center of the orchestra, of a masonryvault one meter deep, supported laterally by jambs. It has been interpreted as a channel that should have been used to drain groundwater, since it appears connected towards the south with another channel, still visible below the stage, almost completely filled with debris. Another channel probably occurs at the same depth in the eastern sector of the orchestra.
This study investigates the Rocca di Mezzo area in the central Apennines (Italy) to evaluate the presence of a hypothesized active and capable normal fault underneath a school edifice. A multimethod approach was applied, including geologic field data, borehole data analysis, paleoseismological trenching, stratigraphic analysis. Field observations indicate the absence of structural features affecting the Meso‑Cenozoic bedrock associated to extensional faulting in the sector where the supposed fault should be located. Only structural evidence of reverse faulting related to an inactive compressive tectonic phase was found. Stratigraphic data also indicate that the Meso‑Cenozoic bedrock dips gradually westward, buried by Quaternary sediments in the area where the fault was hypothesized, so that the lateral contact between the bedrock and the Quaternary sediments is just related to the stratigraphic setting. Trench excavations across the southern sector of the suspected fault trace revealed lacustrine sequences, with interbedded paleosols, undisturbed by any fault planes. Radiocarbon and archaeological dating confirm the Holocene age of these lake deposits, consistent with other nearby lacustrine sequences, indicating that multiple sectors of the plateau hosted small lakes during the Holocene whose oscillations have been controlled by karstic‑related processes. This oscillation has determined phases of deposition and erosion that have strongly conditioned the evolution of this part of the plateau. The present‑day geomorphic characteristics of the area, including ponds and numerous dolines, suggest that karstic drainage and local erosion controlled the apparent deepening of the bedrock. The deepening is not therefore caused by the activity of the presumed fault, but it is just caused by local erosion led by karstic processes. These results demonstrate that indirect data or few hints of supposed fault activity can lead to misinterpretation of the presence of active faults. Misinterpretation can be solely ruled out by integrating multiple lines of geological evidence framed in a comprehensive neotectonic analysis, which allows resembling the whole geological evolution of an area over long time spans.
Accurate and rapid localization of microseismic events in single‑well downhole data is highly significant for microseismic monitoring. Although traditional methods, such as diffraction stacking and grid search, achieve high localization accuracy, they are computationally expensive, limiting their applicability for real‑time microseismic monitoring. To address this limitation, we propose an attention‑enhanced fully convolutional neural network (FCN‑CBAM) for efficient single‑well microseismic event localization. The method ingests three‑component waveform data and outputs three one‑dimensional Gaussian distributions representing the probability of the source location along the X, Y, and Z axes. The model is trained using 11,500 theoretical samples generated using the geometry and velocity model of the field data. Compared to traditional grid search methods, which require picking arrival time, calculating back‑azimuth, and locating the 3D seismic source, the FCN‑CBAM model can predict the event locations for field data within seconds, with prediction accuracy comparable to traditional methods. Furthermore, velocity perturbation and signal‑to‑noise ratio tests are performed to demonstrate the robustness and efficiency of our method.
The magnetic field of the Earth’s crust provides essential information about the geological features of a region, its origin, and the spatial and temporal characteristics of the total magnetic field.Numerical estimation and consideration of the contribution of the crustal field are a necessary step in constructing accurate models of the secular variation of the main magnetic field of the Earth based on observatory data. To accurately determine the corrections for the crustal field at the locations of magnetic observatories, we apply an algorithm to process the original observatory data, which allows us to define all three orthogonal components of the crustal magnetic field vector. To assess the reliability of the obtained values, a modified standard deviation 𝜎′ is proposed, which depends on the data series length and involves statistical indicators from long‑term observation series of the benchmark observatories. As a result, updated corrections were obtained for 118 observatories using data over 1998‑2025. For 65 observatories, such data are published for the first time. For the remaining observatories, the results were compared with previously acquired values. For some components and observatories, the deviations exceed 3𝜎′, which confirms the need for regular updating the contribution of the Earth’s crustal field at magnetic observatories. The results obtained also made it possible to assess the accuracy of existing models of the crustal field.
This study presents finite‑fault rupture models for four destructive earthquakes in Türkiye: the 1966 Varto (MW = 6.7), 1967 Mudurnu Valley (MW = 7.3), 1971 Bingöl (MW = 6.7), and 1975 Lice (MW = 6.6) events. Using a teleseismic waveform inversion technique, we derive the spatial distribution of slip, rupture geometry, and seismic moment for each earthquake. Our models reveal that the ruptures are characterized by distinct asperities and are significantly influenced by geometric fault complexities such as bends and step‑overs. The 1966 Varto earthquake involved oblique faulting with three asperities. The 1967 Mudurnu Valley rupture is dominated by a large asperity, with propagation affected by a 20‑degree fault bend. The 1971 Bingöl earthquake model supports the segmentation of the East Anatolian Fault Zone at the Göynük Bend and suggests further sub‑segmentation of the Ilıca Fault Segment at a geometric discontinuity corresponding to the northeastern edge of a large asperity. The 1975 Lice earthquake model shows rupture arrest at a western fault step‑over. These results underscore the critical role of structural discontinuities in controlling rupture initiation, propagation, and termination, providing essential insights for seismic hazard assessment in these tectonically active regions.
Polar motion (PM) describes the motion of the Earth’s rotation axis as it wanders across the Earth’s crust and is essential for space geodesy, navigation, and precise geophysical measurements. Real‑time PM data are not directly observable, and existing prediction models, such as those in IERS Bulletin A, typically provide forecasts limited to one year. This study proposes a deep learning framework based on a Long Short‑Term Memory (LSTM) network augmented with a multi‑head attention mechanism for long‑term PM prediction. The model jointly predicts PMX and PMY components using differenced input sequences to capture nonlinear temporal dependencies and multi‑frequency periodic behaviors. Experiments on the IERS C04 dataset demonstrate that the proposed model substantially outperforms classical linear predictors. Specifically, the proposed model reduces the total prediction error by approximately 61% at a 600‑day horizon and 39% at 1100 days compared with a conventional baseline, achieving a mean absolute error of 16 mas at 600 days and 32 mas at 1100 days. This demonstrates the effectiveness of hybrid LSTM‑attention architectures in capturing long‑range temporal dynamics in geophysical time series and their potential for extended Earth rotation forecasting.
We investigate aftershock centroid moment tensor solutions along the Pazarcık Mw7.7 and Elbistan Mw7.6 February 6, 2023 rupture zones during the time period February 2, 2023 – November 25, 2024. Six spatial clusters of aftershock activity are analyzed specifically the Çardak Cluster (ÇC), Doğanşehir Cluster (DC), Sürgü Cluster (SC), Erkenek Cluster (EC), Pazarcık Cluster (PC) and Amanos Cluster (AC) named according to the segmentation models of the Pazarcık‑Elbistan rupture zone. Based on an aftershock hypocenter catalogue restricted to events with horizontal and vertical errors <2 km, we determine fault plane solutions for 178 events. Centroid moment tensor solutions with moment magnitudes between 3.7 and 7.7 are computed by applying a waveform inversion method on data from the Kandilli Observatory and Earthquake Research Institute (KOERI) and the Disaster and Emergency Management Presidency Earthquake Department (AFAD) broadband seismic networks. The high number of moment tensors at the Pazarcık‑Elbistan rupture zone could be determined only thanks to the local seismic network and allows us to resolve the local deformation pattern with unprecedented precision. Moment tensors along the Çardak, Doğanşehir, Sürgü, Erkenek and Pazarcık Clusters allow us to identify dominantly sinistral strike‑slip mechanisms with normal faulting components on NE‑SW trending fault planes. Moreover, moment tensors in the Amanos Cluster predominantly exhibit NE‑SW extensional normal faulting but also a substantial component of strike‑slip faulting. North and southeast of the Çardak Cluster we observe a high variance in stress field orientation correlated with lower Coulomb stress values. While the Pazarcık Cluster reflects a predominant sinistral strike‑slip regime with normal faulting components, the Çardak Cluster further to the north that also hosted the forthcoming Elbistan Mw7.6 mainshock 9 hours after the Pazarcık Mw7.7 earthquake represents pure left‑lateral strike‑slip faulting. Stress tensor inversions of the aftershock focal mechanisms show rotations of the local stresses following the Pazarcık and Elbistan mainshocks. In the Pazarcık‑Elbistan mainshock area, the maximum horizontal compressive stress axis is horizontally rotated clockwise by ~65° with respect to the coseismic and long‑term regional stress field. Towards the northern end of the rupture where the Elbistan mainshock area is located, stress orientations are rotated clockwise steadily. We conclude that the Pazarcık earthquake caused significant stress partitioning along the rupture. The direction of stress rotation is related to the orientation of the individual fault segments like the Çardak Fault along the EAFZ.
The 2006 Yogyakarta MW ~6.4 earthquake epicenter was located near the Opak Fault, but slightly farther east, with aftershocks also spreading more toward the east of the fault. The distribution of aftershock did not align precisely with the Opak Fault, raising questions about whether the mainshock originated from the Opak Fault or from another nearby fault. One hypothesis suggests that an eastward dipping fault caused the eastern distribution of the aftershocks. However, the previous studies by Saputra et al. (2021) and Ramdhan et al. (2025a, b) concluded that the MW ~6.4 earthquake had a westward dipping fault, located to the east of the Opak Fault. In this study, we applied a deep learning method to analyze the arrival times of P and S waves, providing more consistent results than previous approaches. This method has not previously been applied to the analysis of the Yogyakarta MW ~6.4 aftershock sequence. We used the arrival times of P and S waves to determine the locations and local magnitudes (ML) of aftershocks. First, we used the grid‑search method to determine the absolute hypocenter locations. Then, we updated the velocity model and relocated the events to better represent the seismic conditions in the region. Finally, we selected events to refine the distribution pattern and understand the tectonic setting around the mainshock. The results show that aftershocks in the eastern part of the study area occurred at greater depths compared to those on the western side. To further understand this pattern, we calculated Coulomb stress changes using the focal mechanism from the Global CMT, which aligns with the USGS catalog. The analysis reveals that the shallower earthquakes on the western side of the fault correlate with areas of positive Coulomb stress change. These findings suggest that the 2006 Yogyakarta earthquake was likely triggered by a westward‑dipping fault associated with the Ngalang Fault.
Stromboli, one of the most active volcanoes in Italy, is characterized by an ordinary eruptive activity consisting of persistent and mild explosions, occasionally interrupted by lava flows and by significant eruptions named major explosions and paroxysms. During such explosive activity, abundant loose pyroclastic material is emplaced on the upper steep slopes of the volcano. Due to intense or prolonged rainfall, this material can be remobilized, thereby generating volcanoclastic flows of different typologies that invade the lower slopes and, in several cases, reach the inhabited coastal areas causing huge damage to infrastructures. The most recent flows occurred on October 18‑20, 2024 and May 15, 2025, when Stromboli island was hit by very intense rainfall concentrated in a few tens of minutes. Consequently, loose tephra deposits were rapidly remobilized, triggering lahars that reached the villages of Stromboli and Ginostra, as well as the access routes to the volcano summit. In this work, for the first time, the grain‑size characterization of these lahar deposits is presented. Based on the sedimentological characteristics and grain‑size distributions the deposits have been classified as hyperconcentrated stream flows and proves to be dominated by coarse to fine ash. Moreover, using high‑resolution Pleiades optical satellite data, one of the depositional fans sampled during fieldwork was characterized. The results of this study provide the first quantitative grain‑size dataset for Stromboli lahars, representing a key constraint for the reconstruction and numerical modeling of these processes in volcanic hazard assessment at Stromboli.
We investigated the source characteristics of the August 22, 2025 (Mw 4.3) and January 19, 2026 (Mw 4.4) San José earthquakes in Costa Rica. Although moderate in magnitude, both events occurred at a shallow depth (~4.0 km) beneath the central urban area of the capital city. They produced strong ground shaking with peak ground accelerations reaching approximately 230 gal. Using small aftershocks as Empirical Green’s functions (EGFs), we estimated the size of the strong‑motion generation areas (SMGAs), rise time, rupture velocity, and associated stress drop for each event. Despite their similar magnitudes, the two earthquakes exhibit notable differences in SMGA size and stress drop, which appear to control the spatial distribution of observed acceleration.
Legacy seismic reflection profiles are a critical subsurface resource, essential for seismotectonic and basin modeling, yet many remain accessible only as raster images. This paper presents WigglePy, an open‑source QGIS plugin that integrates the entire raster‑to‑SEG‑Y conversion workflow – from image calibration and cleaning to geometry assignment – within a single GIS environment. WigglePy provides an interactive interface utilizing fast extraction algorithms combined with a noise attenuation strategy. A synthetic benchmark shows that the implemented extraction strategies (Fast Fourier Bandpass, Savitzky‑Golay morphological reconstruction, and Wiener deconvolution) achieve reconstruction errors comparable to those of matrix‑based inversion methods, with improved computational performance. By integrating iterative quality control, spatial data management, and post‑extraction spectral spike filtering, WigglePy offers an efficient solution for the scientific rescue and revitalization of vintage seismic archives.
