Istituto Nazionale di Geofisica e Vulcanologia

Validation for computational modelling of volcanic ballistic projectiles (VBPs) has often been accomplished by replication of field-surveyed VBP distributions. However, a range of ejection parameters can result in the same impact location, with differing impact energies and therefore hazard implications. Ballistic hazard models must accurately represent the whole trajectory to better estimate ballistic energy and interpret eruption dynamics. We examine model performance by directly measuring Strombolian VBP ejection parameters and trajectories from high-speed video and comparing with modelled trajectories using the same parameters. We illustrate the effects of velocity and size on drag. We synthesize and investigate different methods by which the Reynolds number ( Re ) - drag coefficient (C D ) relationship is incorporated into models and how this affects modelling results. Our computational model allows for application of different existing Re -C D methods and direct comparison to tracked trajectories on the metrics of impact distance and peak height. Even with validated ejection parameters and appropriate Re -C D relationships, the model rarely accurately reproduces both height and distance within a 5% threshold. The most common scenario is model overestimation of both metrics. The gas jet is the largest factor contributing to this discrepancy. Significant slowing of VBPs in excess of gravity is observed while bombs are rising (and within the gas jet radius) but not while falling. Slowing in the gas jet is potentially caused by friction with ash and lapilli, surface temperature effects, or other fluid dynamics effects. • Drag affects small, fast volcanic bombs more. • Models often overestimate the height and impact distance of bomb paths. • The gas jet above the vent slows bombs down. • Slowdown might be due to friction with ash, denser gases, or the hot bomb surface.

Abstract. High levels of bromine monoxide (BrO) observed in volcanic plumes indicate significant catalytic destruction of tropospheric ozone (O3) at local to regional scales. The underlying chemical mechanisms are still incompletely understood and the quantification of O3 concentrations and their distribution in volcanic plumes remain a major challenge. Common atmospheric O3 measurement techniques (UV absorption spectroscopy and electrochemical sensors) suffer from strong interferences, especially from sulphur dioxide (SO2), which is low in the atmospheric background but a main constituent of volcanic plumes (ppmv levels). This problem can be circumvented by using chemiluminescence (CL) O3 monitors, which have negligible interference with SO2 and other trace gases commonly found in volcanic plumes. However, volcanic plume measurements with modern CL O3 monitors are impractical because they are heavy and bulky. Here we report on the development and application of a lightweight version of a CL O3 instrument (l.5 kg, shoebox size) that can be mounted to a commercially available drone. Besides measurements of vertical O3 profiles over several hundred metres, we present drone-based CL O3 measurements in the volcanic plume of Mount Etna in Italy. Within 3 km of the emitting craters we measured an anti-correlation between SO2 and O3 concentrations, corresponding to ozone reductions by up to 60 % in the volcanic plume with respect to the surrounding atmosphere.

The comprehension of the complex dynamics of degassing is critical for volcano monitoring and assessing volcanic hazards. In this study, we apply visibility graph analysis (VGA) to a decadal, high-resolution time series of daily soil CO2 flux recorded by a standardized monitoring network at Mt. Etna volcano (Italy). By mapping these time series into complex networks, we demonstrate that the connectivity degree distributions follow a power law described by the exponent γ, which reveals a self-similar behavior of gas emissions. We introduce the γ-deviation, namely the variation of the scaling exponent from its long-term site-specific baseline, as a novel proxy for degassing efficiency. The long-term baseline is interpreted as a site-specific measure of flux efficiency, while its variations are attributed to other factors, such as fluctuations in the sources or changes in the efficiency of fluids transport pathways. Our results identify a transition from a period of discordance across the monitoring sites (pre-2016) to a phase of network-wide concordance (after 2016). The striking correlation between topological γ-deviations and the established normalized network signal (Φnorm) validates the methodology, suggesting that VGA is able to capture the same underlying magmatic drivers. This study establishes VGA as a robust and reliable tool for medium- and long-term monitoring, potentially capable of identifying the occurrence of large-scale magmatic processes and refining the characterization of fluid transport dynamics in active volcanic systems.

For coastal areas located close to offshore seismic sources, earthquake risk is inherently multi-hazard: intense ground shaking and tsunami inundation can occur in rapid succession, with very short lead times for protective actions. In these near-field settings, traditional Tsunami Early Warning Systems (TEWS), which rely on seismic source parameters available several minutes after origin time, may provide alerts that are only marginally earlier than tsunami arrival, limiting their effectiveness.In this study, we evaluate the integration of rapid earthquake magnitude and location estimates from QuakeUp, an impact-based Earthquake Early Warning System (EEWS), into a tsunami early warning workflow. QuakeUp processes real-time seismic observations to issue initial earthquake alerts within a few seconds, based on fast estimates of magnitude, location, and potential damage zone, supporting immediate risk mitigation for ground shaking. These estimates are then simultaneously used to initialize Probabilistic Tsunami Forecasting (PTF), enabling a coordinated multi-hazard warning strategy in which earthquake and tsunami risks are addressed within a unified, time-critical framework.We present a real test case to quantify the earliness and accuracy of EEWS-derived source parameters and assess their impact on tsunami forecasting. A hindcast of the 30 October 2020 Mw 7.0 Aegean Sea earthquake, whose tsunami reached nearby coastlines in approximately 10 minutes, shows that the EEWS delivers stable and accurate hypocenter and magnitude estimates about 40 seconds after origin time. These estimates are comparable to those provided by the operational Early-Est system, which become available only after several minutes in the Mediterranean region. When used to initialize a PTF procedure, the EEWS-based source characterization yields coastal runup estimates in reasonable agreement with observations, despite the substantially reduced warning latency.To further assess the robustness of the results obtained in the real case, we analyze a second scenario based on 150 simulated seismogram datasets for earthquakes in the Messina Strait (Southern Italy), a region characterized by high exposure and extremely short tsunami travel times. For events with moment magnitudes between 6.0 and 7.0, the analysis confirms that the integrated EEWS–TEWS workflow can provide reliable source estimates within one minute, supporting its applicability to near-field tsunami early warning. This study provides the first demonstration of a combined EEWS–TEWS approach for near-coastal tsunamigenic events, highlighting its potential for dual risk mitigation within a multi-hazard early warning perspective. Future work will focus on testing performance in an operational setting, including the effects of network latencies and configuration-dependent efficiency. In addition, the EEWS technique employed, by providing a preliminary mapping of the most damaged zone, offers a promising perspective for extracting early constraints on seismic source geometry. This information could further reduce uncertainty in near-field tsunami inundation forecasts.

The Bedretto Underground Laboratory for Geosciences and Geoenergies (BULGG, ETH Zurich) operates a dense, multi-scale seismic monitoring framework for on-fault observation of induced seismicity during underground stimulation experiments. The system integrates permanent seismic stations (500 to 10k samples per second, sps), high-rate experimental deployments (up to 200k sps), and distributed acoustic sensing (DAS; up to 4000 sps), providing high spatial and temporal data coverage at the reservoir and local scale. This contribution documents real-time seismic monitoring during underground experiments and the generation of associated post-processed earthquake catalogues.The combined instrumentation is used for monitoring of hydraulic stimulation and geothermal project experiments, including the Fault Activation and Earthquake Rupture (FEAR) ERC project and Geothermal Test-Bed (GTB) activities. Permanent monitoring provides manually reviewed earthquake locations and magnitudes in a continuous reference earthquake catalogue, continuously updated since 2021. Experimental deployments enable near real-time event detection, characterization, and Traffic Light System (TLS) alerting, including borehole DAS observations since 2024.For the FEAR and GTB experiments, additional post-processed catalogues provide optimal yet conventional event parameters based on the final instrumentation configuration, optimized phase picking and hypocenter locations. The resulting catalogues provide consistent absolute locations and high-resolution double-difference relocations based on fully repicked waveform data.Event magnitudes are estimated using multiple complementary approaches. The backbone national network provides a Swiss-specific local magnitude (Mlhc) using station corrections, while the permanent BedrettoLab network applies conventional local magnitude (MLc) from Wood–Anderson amplitudes. Experimental catalogues include relative moment magnitudes derived from acoustic emission amplitude ratios (MwA) calibrated against collocated accelerometers. Ongoing developments include spectral moment magnitudes and DAS strain-rate-based magnitude estimates to improve characterization of small-magnitude events.The BedrettoLab monitoring framework, implemented in SeisComP, addresses challenges such as electromagnetic noise, tunnel operations, and heterogeneous sensor coverage while ensuring continuous acquisition, archiving, and FDSN-compliant data access. Beyond Bedretto, this on-fault, densely instrumented approach demonstrates how integrated DAS, permanent, and experimental monitoring can substantially improve event detection, location accuracy, and magnitude estimation, providing a transferable framework for induced seismicity monitoring, operational decision-making, and risk mitigation in geothermal and underground engineering applications.

Azerbaijan hosts one of the largest global concentrations of mud volcanoes, which are natural geological formations that erupt mud, water, and gases, with methane (CH4) being the dominant component. Here, we present results from the recent METHANE-To-Go Azerbaijan field campaign, which aimed to quantify CH4 emissions from mud volcanoes using in-situ fixed-wing drone measurements.The campaign was carried out mainly in two field phases in May and October 2025, supplemented by additional flights throughout the year, resulting in a total of 36 flight hours. Measurements were performed using a fixed-wing MAS DIHA 350 VTOL UAV equipped with a lightweight, high precision CH4 analyzer, along with wind measurements derived from pitot tube data. A dedicated measurement strategy was developed and optimized specifically for mud volcanoes in the Azerbaijani environment. The approach follows a mass balance framework based on Gauss’s divergence theorem, whereby the drone flies concentric circles around the emission source at multiple altitudes. Fluxes of CH4 are derived from upwind and downwind measurements, enabling the quantification of the total emission enclosed by the cylindrical flight pattern. To validate the applied methodology, five controlled release experiments were conducted, during which CH4 from bottles was released at a known rate. Additionally, two dedicated flights were conducted to calibrate and evaluate the in-situ wind measurements using the onboard pitot tube. In total, 38 measurement flights were conducted at five continuously emitting mud volcanoes, allowing for quantitative emission estimates for these sites. In addition to continuously emitting sources, the method was applied to an explosive eruption of the Otmanbozdagh mud volcano on 11 October 2025, which was surveyed shortly after the event.In parallel to the flight measurements, ground based air samples were collected and analyzed for gas composition and CH4 carbon stable isotopes, supporting source attribution and the discrimination of mud volcano emissions from potential anthropogenic sources.The dataset provides the first high resolution in-situ CH4 observations shortly after a mud volcano eruption, as well as quantitative emissions from active mud volcanoes using mass balance approach, indicating source dependent emission rates on the order of several tens to approximately one hundred kilograms per hour. These results contribute to a more robust assessment of the role of geological CH4 sources in the global climate system and demonstrate the potential of fixed-wing drone platforms for quantitative measurements of greenhouse gas fluxes.

Probabilistic Fault Displacement Hazard Analysis (PFDHA) quantifies the probability and the expected amount of coseismic displacement associated with the activity of Active and Capable Faults (ACFs) at a given site. Common PFDHA approaches distinguish between primary on-fault displacement and distributed off-fault ruptures occurring on secondary faults or fractures and typically rely on empirical scaling relationships calibrated for specific earthquake magnitudes. However, these methods are often calibrated for specific tectonic or kinematic settings and lack readily available computational tools. Moreover, available PFDHA approaches do not commonly include the possibility to investigate the floating rupture mechanism, i.e., the possibility that surface ruptures involve only portions of the full fault trace.To overcome these limitations, we developed FaulTED, a new user-friendly MATLAB-based code for PFDHA that integrates a comprehensive set of published models, including magnitude–frequency distributions, fault scaling relationships, and surface rupture probability functions. The toolkit comprises two main modules: (i) a site-specific hazard curve calculator and (ii) a fault-specific hazard map generator for user-defined return periods. Both modules explicitly account for on-fault and distributed off-fault ruptures and incorporate the floating rupture approach commonly adopted in probabilistic seismic hazard analysis.The modular architecture of FaulTED allows users to flexibly select and compare alternative models through a structured input file, enabling sensitivity analyses and systematic exploration of epistemic uncertainties. FaulTED is designed as a user-oriented platform to support infrastructure planning in regions affected by ACFs.