Giuseppe Davide Chiappetta

Understanding how fault strength evolves through time remains a central challenge in seismology, particularly in complex plate-boundary regions characterized by structural segmentation and heterogeneous coupling. We investigate whether ground-motion variability from small earthquakes can track mechanical evolution of active faults in the southeastern Alps (northeast Italy—W Slovenia)—a region that hosted destructive earthquakes such as the 1511 (∼Mw 6.8) and the 1976 Mw 6.5 Friuli events. We analyze ∼115,000 near-fault peak ground acceleration (PGA) recordings from 6,219 earthquakes (0.5≤Mw≤4.2; 2016–2025) using a mixed-effects ground-motion model that isolates the between-event term (δBe), representing source-related variability independent of magnitude, distance, and site effects. δBe exhibits clear spatial clustering and correlates positively with independently estimated stress-drop residuals, confirming that ground-motion variability reflects rupture efficiency and relative fault strength. Mapping the temporal evolution of a normalized cumulative stress index derived from δBe reveals persistent east–west mechanical segmentation and multi-year trends. Notably, the epicentral area of the 1976 Friuli earthquake exhibits a progressive post-2022 increase in relative stress drop. Our results demonstrate that small-earthquake ground-motion variability provides a model-independent proxy for tracking spatial and temporal changes in fault mechanics. In the southeastern Alps, mechanically distinct fault domains evolve differently through time, suggesting heterogeneous stress loading and redistribution across adjacent segments of the plate boundary.

Distributed acoustic sensing (DAS) is rapidly emerging as a powerful complement to conventional seismic monitoring, enabling dense spatial sampling over tens of kilometers using existing fiber-optic telecommunication infrastructures. In this study, we present the design, implementation, and initial performance of a regional-scale DAS seismic monitoring network integrated within the operational framework of the Seismological Research Center of OGS in north-eastern Italy. The system is built upon dark fiber segments of a public telecommunication network and interrogated by multiple DAS units, covering several hundred kilometers of optical fiber and producing tens of thousands of virtual sensing channels. We describe the end-to-end workflow developed to support continuous DAS acquisition in an operational environment, including data conversion from proprietary formats to standard seismological archives, channel aggregation strategies, metadata handling, and real-time streaming through SeedLink and International Federation of Digital Seismograph Networks services. Attention is given to practical constraints related to data volume, latency, channel naming, and integration with existing seismic monitoring infrastructures. Using representative examples, we illustrate the capability of the network to record teleseismic and local seismic signals, to characterize spatially variable seismic noise conditions along fiber routes, and to contribute to earthquake location when combined with conventional seismic stations. These case studies highlight both the opportunities and current limitations of distributed acoustic sensing (DAS) for regional seismic monitoring, including issues related to channel localization, directional sensitivity, and effective sensing range. The presented network represents one of the most extensive operational DAS deployments for seismic monitoring in Europe to date. Our results demonstrate the feasibility of integrating large-scale DAS systems into routine monitoring activities and provide a reference framework for future developments toward multisensor, fiber-optic-based seismic observatories.

On 1 August 2024, a magnitude Mw 5.0 earthquake struck Calabria, Italy. Seismograms reveal a clear double-rupture source consisting of an initial mb 4.4 event followed 1.1 s later by a stronger rupture. The source kinematics computed from the P-wave first motion polarity for the first rupture and from waveform modeling for the entire event provide markedly different results. The first rupture exhibited normal-fault kinematics with a nearly vertical plane, whereas the low-frequency radiation of the whole earthquake is consistent with strike-slip motion. Nevertheless, both focal mechanisms share a common feature: one nodal plane is nearly vertical, oriented northeast–southwest, and dips to the southeast. The location of more than 400 aftershocks delineates a diffuse cloud of hypocenters over a broad seismogenic volume with a pronounced northeast–southwest elongation, coincident with the orientation of one nodal plane of the mainshock moment tensor solution. Spectral analysis of P waves indicates that the stronger rupture of the mainshock occurred on the northeast–southwest fault with right-lateral slip and a clear directivity effect. Focal mechanisms of 17 aftershocks in the magnitude range 2.0≤ML≤3.1 show highly variable solutions, even among events located close to one another, reflecting all kinematic styles. The mainshock was followed by fewer aftershocks than predicted by the Gutenberg and Richter law, distributed across a broad volume that is probably highly fractured and affected by a complex stress field.

Distributed Acoustic Sensing (DAS) is rapidly emerging as a powerful complement to conventional seismic networks, enabling dense and continuous ground-motion monitoring by exploiting existing fiber-optic telecommunication infrastructures. In this contribution we present the design, implementation, and first results of a regional-scale DAS seismic monitoring network deployed in Friuli Venezia Giulia (NE Italy), developed through a close collaboration between the regional authorities, the public ICT provider INSIEL, and the scientific community.The DAS infrastructure is an integral part of the SMINO network, the multi-sensor seismic and geodetic monitoring system for north-eastern Italy operated by OGS, and is fully integrated within its operational framework. The network is based on five strategically distributed interrogation points across the region, collectively instrumenting nearly 250 km of telecommunication fibers belonging to the regional backbone. This configuration provides several tens of thousands of virtual sensing channels with meter-scale spatial sampling, covering urban areas, transportation corridors, and heterogeneous geological settings. To our knowledge, this represents one of the most extensive—and likely the largest—operational regional DAS seismic monitoring networks currently deployed in Europe.We describe the criteria adopted for fiber selection, network geometry, and acquisition parameters, with particular emphasis on signal optimization, performance, and long-term operational robustness. Examples of recordings from local and regional earthquakes demonstrate the capability of the system to capture coherent seismic wavefields, clear phase arrivals, and spatial variability of ground motion. The network also systematically records anthropogenic signals, which are discussed both as a challenge for data quality control and as an opportunity for multi-purpose environmental monitoring.Finally, we address the integration of DAS data into existing seismic and civil protection workflows, including real-time data streaming, event detection, and rapid situational awareness. The Friuli Venezia Giulia experience demonstrates the maturity and added value of fiber-optic sensing for operational seismology at regional scale.

In this work, we present a new and improved version of the NExt STrOng Related Earthquake (NESTORE) software, originally released as NESTOREv1.0 and publicly available as a MATLAB-based package. The original version of NESTORE was specifically designed to forecast the occurrence of strong aftershocks in the first few hours following a mainshock, providing a valuable tool for short-term seismic hazard assessment. The newly developed version introduces several methodological and computational improvements aimed at increasing the robustness and reliability of the forecasting framework. Among the main upgrades is the integration of the REPENESE (RElevant features PErcentage class weighting NEighborhood detection SElection) algorithm, an advanced outlier detection method explicitly designed to handle class imbalance and skewed datasets, which are characteristic of the seismicity features we used. This integration enables a more effective identification and treatment of anomalous events, thereby improving classifier performance. In addition, the new version implements a k-fold cross-validation strategy to estimate model performance. This approach allows a more stable and unbiased evaluation of predictive capabilities compared to single-split validation methods, especially with limited or heterogeneous data. Overall, the combination of these enhancements results in a more flexible, accurate, and reliable tool for the analysis of earthquake clusters and the early forecasting of strong aftershocks. Funded within the RETURN Extended Partnership and received funding from the European Union Next-Generation EU (National Recovery and Resilience Plan—NRRP, Mission 4, Component 2, Investment 1.3—D.D. 1243 2/8/2022, PE0000005) and by the grant “Progetto INGV Pianeta Dinamico: NEar real-tiME results of Physical and StatIstical Seismology for earthquakes observations, modelling and forecasting (NEMESIS)” - code CUP D53J19000170001 - funded by Italian Ministry MIUR (“Fondo Finalizzato al rilancio degli investimenti delle amministrazioni centrali dello Stato e allo sviluppo del Paese”, legge 145/2018).

NESTORE (Next STrOng Related Earthquake) is a machine learning algorithm for forecasting strong aftershocks during ongoing earthquake clusters. It has already been successfully applied to Italian, Greek and Californian seismicity in the past. A free version of the software in MATLAB (NESTOREv1.0) is available on GitHub. The method is trained on the region under investigation using seismicity characteristics. The obtained region-specific parameters are used to provide the probability, for the ongoing clusters, that the strongest aftershock has a magnitude greater than or equal to that of the mainshock - 1. If this probability is greater than or equal to 0.5, the cluster is labeled as type A, otherwise as type B. The current version of the code is modular and the cluster identification method is based on a window approach, where the size of the spatio-temporal window can be adjusted according to the characteristics of the analyzed region.In this study, we applied NESTORE to the seismicity of Japan using the Japan Meteorological Agency (JMA) catalogue from 1973 to 2022. To account for the highly complex seismicity of the region, we replaced the cluster identification module with software that uses a stochastic declustering approach based on the ETAS model.The analysis is performed in increasing time intervals after the mainshock, starting a few hours later, to simulate the evolution of knowledge over time. The analysis showed a high prevalence of clusters where there are no strong earthquakes later than 3 hours after the mainshock, leading to an imbalance between type A and type B classes.NESTORE was trained with data from 1973 to 2004 and tested from 2005 onwards. The large imbalance in the data was mitigated by carefully analyzing the training set and developing techniques to remove outliers. The cluster type forecasting was correct in 84% of cases. Funded by a grant from the Italian Ministry of Foreign Affairs and International Cooperation and Co-funded within the RETURN Extended Partnership and received funding from the European Union Next-GenerationEU (National Recovery and Resilience Plan - NRRP, Mission 4, Component 2, Investment 1.3 – D.D. 1243 2/8/2022, PE0000005) and by the NEar real-tiME results of Physical and StatIstical Seismology for earthquakes observations, modeling and forecasting (NEMESIS) Project (INGV).

The advanced machine learning algorithm NESTORE (Next STrOng Related Earthquake) was developed to forecast strong aftershocks in earthquake sequences and has been successfully tested in Italy, western Slovenia, Greece, and California. NESTORE calculates the probability of aftershocks reaching or exceeding the magnitude of the main earthquake minus one and classifies clusters as type A or B based on a 0.5 probability threshold. In this study, NESTORE was applied to Japan using data from the Japan Meteorological Agency catalog (1973-2024). Due to Japan's high seismic activity and class imbalance, new algorithms were developed to complement NESTORE. The first is a hybrid cluster identification method using ETAS-based stochastic declustering and deterministic graph-based selection. The second, REPENESE (RElevant features, class imbalance PErcentage, NEighbour detection, SElection), is optimized for detecting outliers in skewed class distributions. A new seismicity feature was proposed, showing good results in forecasting cluster classes in Japan. Trained with data from 1973 to 2004 and tested from 2005 to 2023, the method correctly forecasted 75% of A clusters and 96% of B clusters, achieving a precision of 0.75 and an accuracy of 0.94 six hours after the mainshock. It accurately classified the 2011 Tōhoku event cluster. Near-real-time forecasting was applied to the sequence after the April 17, 2024 M6.6 earthquake in Shikoku, classifying it as a "Type B cluster," with validation expected on October 31, 2024.

Seismic swarms of low magnitude earthquakes occur frequently in the Calabrian Arc. During the last few years, several earthquakes of magnitude up to M_L4.4 occurred both on land and offshore near the coast of Calabria. Some of them were followed by a sequence of tens to hundreds of smaller, well-clustered earthquakes that occurred during the following weeks or months. In other cases, swarms of low-magnitude earthquakes occur without a classical mainshock-aftershock evolution. In this work, we selected swarms that were well recorded by a high number of seismic stations to perform a detailed analysis consisting of the determination of the relative location and focal mechanism for as many earthquakes as possible. In some cases, the relative location allows to recognize the seismogenic fault and to distinguish the fault plane from the auxiliary plane of the focal mechanism solution. In other cases, the relative location unravels a small cloud of events that is not compatible with a unique fault plane, suggesting the occurrence of the swarm in highly fractured seismogenic volume. The relative hypocenter positions allow to estimate the size of the seismogenic volume, which is very small in most of the cases, often less than 1 km³. However, its extension is greater than the size computed for the mainshock rupture in many cases. The most common source mechanism is of normal type, but strike-slip and reverse kinematics are also found, in particular for swarms located offshore and near the coast. The temporal distribution of events does not show any evident migration of the sources, thus suggesting that the driving mechanism is not related with aseismic phenomena like fluid diffusion and stress waves.