Search

During the European Commission-funded project EUROVOLC, our team has created an interactive webpage to collect citizen-science observations of volcanic events, and to visualise, map and download previously collected data.

Recent eruptive activity at Stromboli is mostly characterized by persistent, Strombolian explosive activity from summit craters. Occasionally, so-called major explosions occur, and more rarely paroxysms take place, such as on 3 July and 28 August 2019. We analyze monitoring data recorded between 2016 and 2022 to identify and characterize patterns in multi-parameter time series (considering in particular seismic, geochemical, meteorological, and sea-level gauges), with special interest in analyzing patterns observed before and during the occurrence of major explosions and/or paroxysmal eruptions. In practice, (i) we implement algorithms to automatically identify families of seismic events using waveform features; (ii) we analyze the effect of oceanic microtremor on continuous and discrete amplitude-based measurements; (iii) we study the temporal and size distribution of event occurrences, and use this information to assess likely trends in eruptive behavior. Moreover, (iv) we use noise cross-correlations and auto-correlations to compute seismic velocity variations of the shallow crust. Multi-parametric measurements provide interesting insights in the temporal evolution of the eruptive activity at Stromboli; for example, correlated changes in the pattern at which events occur in time, patterns in the distribution of extreme, large events, and evidence of a decrease in seismic velocity, seem to be phenomena occurring before paroxysmal eruptions. Detailed analyses of the produced time series, including also pattern recognition techniques, are required for better revealing likely patterns and for better understanding and interpretation of observations.

Accurate assessment of rate and state friction parameters is essential for producing realistic rupture scenarios and, in turn, for seismic hazard analysis. Those parameters can be directly measured in the laboratory, with experimental apparati that reproduce fault conditions in nature. Alternatively, indirect estimates (i.e., inversion) of rate and state parameters are based on postseismic slip evolution studies and numerical modeling. Both direct and indirect approaches require a series of assumptions that might bias the results.Here we take advantage of a downscaled analog model reproducing experimentally megathrust earthquakes. The analog model shares many characteristics of real subduction zones, although being intentionally oversimplified with respect to nature. This allows reducing the number of potential sources of bias (e.g., fault geometry and asperity size). We perform five analog models with a single, rectangular asperity of different lengths embedded in a nearly velocity neutral matrix. We focus on two different physical conditions, namely the along-strike asperity length and the asperity to neighboring segments length ratio, and study systematically how they tune the model seismic behavior. Then, by coupling quasi-dynamic numerical models with the simulated annealing algorithm, we retrieve rate and state parameters that allow reproducing both the recurrence time, rupture duration and slip amplitude of the analog model, in ensemble. We identify a tradeoff between (a-b) of the asperity and (a-b) of neighboring creeping segments, with multiple combinations that allow mimicking the analog model behavior and variability. We also identify a negative correlation between (a-b) of the asperity and asperity size, with Dc remaining relatively constant within the investigated asperity size range. When estimating (a-b), poorly constrained properties of neighboring segments are responsible for uncertainties in the order of per mille. Roughly one order of magnitude larger uncertainties derive from asperity size. Those results provide a first order assessment of the variability that rate and state friction estimates retrieved for nature conditions might have when used as constraint to model fault slip behavior.

During explosive eruptions a large amount of tephra is dispersed and deposited on the ground with the potential to cause a variety of damage and disruption to residential buildings and infrastructure, including road networks. The quantification of the tephra ground load is, therefore, of significant interest to reduce environmental and socioeconomic impact, and for managing crisis situations during volcanic eruptions. Tephra dispersal and deposition is a function of multiple factors, including mass eruption rate (MER), degree of magma fragmentation, vent geometry, top plume height (HTP), particle size distribution (PSD) and wind velocity and pattern.In this work we quantify the tephra load deposited on the road network of the eastern flank of Mt. Etna, in Italy, during the sequence of lava fountains occurred between February 2021 and 2022. In particular we analyse those events generating volcanic plumes mostly dispersed towards the east-southeast direction and focus our study on the main road networks of some municipalities which are found in this section of Mt. Etna as Milo, Santa Venerina, Fleri.We applied the volcanic ash radar retrieval (VARR) approach to a large dataset of short-lasting and intense lava fountains detected by the X-band weather radar, located at about 32 km from the Etna summit, to retrieve the eruption source parameters. When the radar data were unavailable, we analysed images of the SEVIRI satellite and of the visible calibrated camera images of the Istituto Nazionale di Geofisica e Vulcanologia, Osservatorio Etneo.Two numerical models (TEPHRA2 and FALL3D) were used to simulate tephra ground accumulation. The model calibration was performed using data collected during an eruptive event in 2021. Tephra load was calculated for areas of particular interest as e.g., building roofs, infrastructure and road networks, requiring clean-up.We compute a cumulative in time of deposited tephra on some locations of the road network obtaining values ranging from 40-140 kg/m2, and from 110-480 kg/m2 in function of model considered and selected location.As a result, we produce fast estimates of total tephra deposited on specific infrastructures (e.g., roads) during sequences of eruptive events; such information can be a valuable input for quick planning and management of the short-term tephra fall hazard.

Central volcanoes can produce a wide spectrum of volcanic eruptions, with different magmatic compositions, styles, sizes, and recurrence periods. On volcanic islands, the eruptive record of central volcanoes is often incomplete, due to the small subaerial area, irregular topography, and high erosion rates typical of these islands, generating large uncertainties about the past eruptive activity and making the estimation of eruptive parameters, recurrence times and probabilities of future eruptions very challenging.São Miguel Island (Azores archipelago) is one of these cases, where most eruptions of the three active central volcanoes (Sete Cidades, Fogo, and Furnas) are undated or poorly reconstructed. Based on the known stratigraphy, Sete Cidades volcano erupted at least 36 times in the last 15 ky, producing 24 trachytic events, almost all explosive, and 12 basaltic flank eruptions; Fogo volcano erupted at least 21 times also in the last 15 ky, producing 16 trachytic explosive eruptions and 5 basaltic flank eruptions; while Furnas volcano erupted at least 22 times over the last 17 ky, all trachytic explosive events.Here, we model eruptive event times based on the generation of synthetic catalogues that follow the known stratigraphic sequence and include the uncertainty of eruption ages. The completeness of the eruptive records of each volcano was assessed by plotting the cumulative number of eruptions in time and identifying breaks in slope, which may indicate changes in the recording rate of events, as well as possible changes in the eruptive behaviour. In the parts of catalogues after the first break-in-slope we also checked the stationarity to identify the portions of the catalogues that could be modelled by renewal models. Fitting the stationary portion of data with several renewal models allowed to identify which statistical model best describes how eruptive events occur in time. This study presents a statistical analysis where data uncertainties are accounted for to model eruptive inter-event times, estimate recurrence periods and probabilities of future eruptions. This approach is crucial for a more robust long-term assessment of volcanic hazard, providing important clues to forecast the eruptive behaviour of central volcanoes, even in cases of low-activity systems or volcanic islands, where eruptive catalogues are frequently incomplete. In the case of São Miguel this approach allowed to estimate recurrence periods and the probability of a future event for each of the three active central volcanoes of the island.

In the last decades, the enormous potential for direct geothermal heat from aquifers attracted special attention, particularly toward those thermal springs indicating areas in which exploitation of geothermal energy might be economically feasible for indirect uses such as electrical power production. The availability of geochemical data besides the location of thermal spring areas assumes particular importance, especially in the first stages of a geothermal exploration program. In this work, we present a digitised format of the literature review of Gerald Ashley Waring on the geothermal springs of the world. This unprecedented dataset contains geographical coordinates (from georeferentiation) of ~6,000 geothermal spring areas, including complementary data such as temperatures, flow rates, total dissolved solids content (TDS, expressed in ppm), and quantitative chemical analysis of major elements (only for a few hundred sites). Using temperature and flow rate, we derive the heat discharged from 1483 thermal spring areas (between ~10-5 and ~103 MW, with a median value of ~0.5 MW and ~8300 MW in total). We complement this information with geological data sets currently available in the literature and analyse them using statistical and geospatial tools and a supervised machine-learning algorithm. We show that terrestrial heat flow, topography, volcanism, and extensional tectonic play a key role in the occurrence of thermal waters around the globe. These results can also be beneficial to address the geothermal interest towards specific and less studied areas and significantly drive the first steps of the geothermal surveys and detailed investigations. Finally, this data set in electronic format will be beneficial for future research on the spatial distribution of thermalism at a small scale and the variation of temperature and flow rate of several thermal springs in the last decades in certain regions.

The creation of hazard maps relative to volcanic phenomena requires taking into account the intrinsic complexity and variability of eruptions. Here we present an example of how HPC can allow producing a high resolution multi-source probabilistic hazard assessment due to tephra fallout over a domain covering Southern Italy. The three active volcanoes in the Neapolitan area, Somma-Vesuvius, Campi Flegrei and Ischia, were considered as volcanic sources. For each one, we explored three explosive size classes (Small, Medium and Large) for Somma Vesuvius and Campi Flegrei, and one explosive size class (Large) for Ischia. For each size class, we performed 1500 numerical simulations of ash dispersion (total of 10500) using the Fall3D (V8.0) model over a computational domain covering Southern Italy with a 0.03° ⨉ 0.03° (~3 km ⨉ 3 km) resolution. Within each size class, the eruptive parameters have been randomly sampled from well-suited probability distributions and with different meteorological conditions, obtained by randomly sampling a day between 1990 and 2020 and retrieving the relative data from the ECMWF ERA5 database. This allowed exploring the intra-class variability and to quantify aleatoric uncertainty. The results of these simulations have been post-processed with a statistical approach by assigning a weight to each eruption (based on its eruption magnitude) and the annual eruption rate of each size class. For the case of Campi Flegrei, the variability in the eruptive vent position has also been explored by constructing a grid of possible vent locations with different spatial probability. By merging the results obtained for each source and size class we produced a portfolio of hazard maps showing the expected mean annual frequency of overcoming selected thresholds in ground tephra load. A disaggregation analysis has also been performed in order to understand which particular source and/or size class had the greater impact on a particular area. The completion of this work, considering both numerical simulations and the statistical elaboration of the results has required a total of more than 5000 core hours and the processing of more than 2TB of data, an effort that wouldn’t have been possible without the access to high level HPC resources.

Nowadays, tephra fallout hazard is based on coupling the physical modeling of the tephra dispersion processes with a probabilistic analysis that takes into account the natural variability of the volcanic phenomena in terms of eruption probability, eruption sizes, vent position and meteorological conditions. In this framework, we present a prototypal methodology to carry out a multi-volcano long-term tephra fallout hazard assessment in Southern Italy from the active Neapolitan volcanoes: Somma-Vesuvius, Campi Flegrei, and Ischia. FALL3D model (v.8.0) has been used to run thousands of numerical simulations (1,500 per eruption size class), considering the ECMWF ERA5 meteorological dataset over the last 30 years. The output in terms of tephra ground load has been processed within a new workflow for large-scale, high-resolution volcanic hazard assessment, in order to quantify the mean annual frequency with which the tephra load at the ground exceeds given critical thresholds at a target site within a 50-years exposure time, and the relative epistemic uncertainty. This work provides, for the first time, a multi-volcano probabilistic hazard analysis for tephra fallout, fully comparable with those used for seismic phenomena and other natural disasters in which multiple sources are integrated together, and it accounts for potential changes in regimes of each single considered volcano. This allows us to discuss also how the full information can be traced back to provide specific information about the prevalence of different volcanoes and eruptive style in the different target areas, based on hazard disaggregation. The methodology is applicable to any other volcanic areas or over different exposure times.

The EU Center of Excellence for Exascale in Solid Earth (ChEESE) develops exascale transition capabilities in the domain of Solid Earth, an area of geophysics rich in computational challenges embracing different approaches to exascale (capability, capacity, and urgent computing). The first implementation phase of the project (ChEESE-1P; 2018–2022) addressed scientific and technical computational challenges in seismology, tsunami science, volcanology, and magnetohydrodynamics, in order to understand the phenomena, anticipate the impact of natural disasters, and contribute to risk management. The project initiated the optimisation of 10 community flagship codes for the upcoming exascale systems and implemented 12 Pilot Demonstrators that combine the flagship codes with dedicated workflows in order to address the underlying capability and capacity computational challenges. Pilot Demonstrators reaching more mature Technology Readiness Levels (TRLs) were further enabled in operational service environments on critical aspects of geohazards such as long-term and short-term probabilistic hazard assessment, urgent computing, and early warning and probabilistic forecasting. Partnership and service co-design with members of the project Industry and User Board (IUB) leveraged the uptake of results across multiple research institutions, academia, industry, and public governance bodies (e.g. civil protection agencies). This article summarises the implementation strategy and the results from ChEESE-1P, outlining also the underpinning concepts and the roadmap for the on-going second project implementation phase (ChEESE-2P; 2023–2026)., Future Generation Computer Systems, 146, ISSN:0167-739X, ISSN:1872-7115

Volcanic activity typically switches between high-activity states with many eruptions and low-activity states with few or no eruptions. We present a simple two-regime physics-informed statistical model that allows interpreting temporal modulations in eruptive activity. The model enhances comprehension and comparison of different volcanic systems and enables homogeneous integration into multivolcano hazard assessments that account for potential changes in volcanic regimes. The model satisfactorily fits the eruptive history of the three active volcanoes in the Neapolitan area, Italy (Mt. Vesuvius, Campi Flegrei, and Ischia) which encompass a wide range of volcanic behaviors. We find that these volcanoes have appreciably different processes for triggering and ending high-activity periods connected to different dominant volcanic processes controlling their eruptive activity, with different characteristic times and activity rates (expressed as number of eruptions per time interval). Presently, all three volcanoes are judged to be in a low-activity state, with decreasing probability of eruptions for Mt. Vesuvius, Ischia, and Campi Flegrei, respectively.

Probabilistic volcanic hazard assessment is a standard methodology based on running a deterministic hazard quantification tool multiple times to explore the full range of uncertainty in the input parameters and boundary conditions, in order to probabilistically quantify the variability of outputs accounting for such uncertainties. Nowadays, different volcanic hazards are quantified by means of this approach. Among these, volcanic gas emission is particularly relevant given the threat posed to human health if concentrations and exposure times exceed certain thresholds. There are different types of gas emissions but two main scenarios can be recognized: hot buoyant gas emissions from fumaroles and the ground and dense gas emissions feeding density currents that can occur, e.g., in limnic eruptions. Simulation tools are available to model the evolution of critical gas concentrations over an area of interest. Moreover, in order to perform probabilistic hazard assessments of volcanic gases, simulations should account for the natural variability associated to aspects such as seasonal and daily wind conditions, localized or diffuse source locations, and gas fluxes. Here we present VIGIL (automatized probabilistic VolcanIc Gas dIspersion modeLling), a new Python tool designed for managing the entire simulation workflow involved in single and probabilistic applications of gas dispersion modelling. VIGIL is able to manage the whole process from meteorological data processing, needed to run gas dispersion in both the dilute and dense gas flow scenarios, to the post processing of models' outputs. Two application examples are presented to show some of the modelling capabilities offered by VIGIL., Fabio Dioguardi (FD) has been supported from UK National Capability funding (BGS Innovation Flexible Fund). This funding supported the code preparation, testing and publishing. Fabio Dioguardi (FD), Silvia Massaro (SM), Laura Sandri (LS), Jacopo Selva (JS), Giovanni Macedonio (GM) and Antonio Costa (AC) have also been supported by the European Union’s Horizon 2020 project EUROVOLC (grant agreement no 731070). This funding supported the initial idea and the interaction among the co‑authors. We thank John A. Stevenson (British Geological Survey) for his technical support in the development of VIGIL and its repository. This work is published with permission of the Executive Director of British Geological Survey (UKRI)

The effects of volcanic hazards can be quantified by applying new methods to provide support for rational decision-making. Mt Etna is one the most active volcanoes in the world, producing both effusive and explosive eruptions together with a very intense seismic activity, which significantly affect the territory and human society. We present the preliminary results obtained in the framework of the PANACEA project (INGV’s project “Pianeta Dinamico”, funded by the Italian Ministero dell’Università e la Ricerca, MUR) regarding the multi-hazard assessment around Mt Etna. These include: (i) the production of an updated spatio-temporal probability map of vent opening at Etna, using a procedure exploiting different Kernel functions (e.g. the exponential, Cauchy, and Gaussian functions), and testing volcanic deformation patterns to explore possible dynamic, structural conditioning on the vent opening process; (ii) the identification of a set of cascading effects scenarios that account for volcanic phenomena (i.e., volcanic unrest, seismicity, volcanic explosions, volcanic effusive events, lava flows, tephra/ballistic fall, and PDC), as well as other external hazards potentially linked in such chains (e.g., flooding, forest fires, etc.); and (iii) the identification of scenarios involving systemic impacts (e.g., impacts on the functionality or connectivity of networks).

Volcanic eruptions are amongst the most jeopardizing natural events due to their potential impacts on life, assets, and environment. In particular, atmospheric dispersal of volcanic tephra and aerosols during the explosive eruptions poses a serious threat to life and has significant consequences for infrastructures and global aviation safety. The volcanic island of Jan Mayen, located in the North Atlantic under trans-continental air traffic routes, is considered the northernmost active volcanic area in the world, with at least five eruptive periods recorded during the last 200 years. However, quantitative hazard assessments on the possible consequences for air traffic of a future ash-forming eruption are nonexistent. This study presents the first comprehensive long-term volcanic hazard assessment for Jan Mayen volcanic island in terms of ash dispersal and airborne tephra concentration at different flight levels. In order to delve in the characterization and modelling of that potential impact, a probabilistic approach based on merging a large number of numerical simulations is adopted, varying the volcano’s Eruptive Source Parameters (ESPs) and meteorological scenario. Each ESP value is randomly sampled following a continuous Probability Density Function (PDF) defined from the Jan Mayen geological record. Over 20 years of climatic data are considered in order to explore the natural variability associated with meteorological conditions and used to run thousands of simulations of the ash dispersal model FALL3D on a 2 km-resolution grid. The simulated scenarios are combined to produce probability maps of airborne ash concentration, arrival time and persistence at different flight levels in the atmosphere. The resulting maps represent an aid to civil protection, decision makers and aviation stakeholders in assessing and preventing the potential impact from a future eruption at Jan Mayen.

International audience; Despite the growing spatiotemporal density of geophysical observations at subduction zones, predicting the timing and size of future earthquakes remains a challenge. Here we simulate multiple seismic cycles in a laboratory-scale subduction zone. The model creates both partial and full margin ruptures, simulating magnitude M w 6.2-8.3 earthquakes with a coefficient of variation in recurrence intervals of 0.5, similar to real subduction zones. We show that the common procedure of estimating the next earthquake size from slip-deficit is unreliable. On the contrary, machine learning predicts well the timing and size of laboratory earthquakes by reconstructing and properly interpreting the spatiotemporally complex loading history of the system. These results promise substantial progress in real earthquake forecasting, as they suggest that the complex motion recorded by geodesists at subduction zones might be diagnostic of earthquake imminence. Plain Language Summary Large and devastating subduction earthquakes, such as the 2011 magnitude 9.0 Tohoku-oki earthquake (Japan), are currently considered unpredictable. Scientists lack a long enough seismic catalog that is necessary for drawing statistical insights and developing predictions. For this reason, we simulate tens of earthquakes using a small-scale experimental replica of a subduction zone. We show that machine learning (a group of algorithms that make predictions based on the "information" acquired in past "experience") can predict when, where, and how big the next experimental earthquake will be. The "information" in our study is provided by the slow deformation accumulating in the analog tectonic plates during the periods in between earthquakes. Since such slow deformation is also measured by means of space geodesy along real subduction zones, there is the possibility that, in the future, variations of this machine learning approach can predict the timing and size of natural subduction earthquakes.

Campi Flegrei is an active volcano located in one of the most densely inhabited areas in Europe and under high-traffic air routes. There, the Vesuvius Observatory’s surveillance system, which continuously monitors volcanic seismicity, soil deformations and gas emissions, highlights some variations in the state of the volcanic activity. It is well known that fragmented magma injected into the atmosphere during an explosive volcanic eruption poses a threat to human lives and air-traffic. For this reason, powerful tools and computational resources to generate extensive and high-resolution hazard maps taking into account a wide spectrum of events, including those of low probability but high impact, are important to provide decision makers with quality information to develop short- and long- term emergency plans. To this end, in the framework of the Center of Excellence for Exascale in Solid Earth (ChEESE), we show the potential of HPC in Probabilistic Volcanic Hazard Assessment. On the one hand, using the ChEESE's flagship Fall3D numerical code and taking advance of the PRACE-awarded resources at CEA/TGCC-HPC facility in France, we perform thousands of simulations of tephra deposition and airborne ash concentration at different flight levels exploring the natural variability and uncertainty on the eruptive conditions on a 3D-grid covering a 2 km-resolution 2000 km x 2000 km computational domain. On the other hand, we create short- and long-term workflows, by updating current Bayesian-Event-Tree-Analysis-based prototype tools, to make them capable of analyze the large amount of information generated by the Fall3D simulations that finally gives rise to the hazard maps for Campi Flegrei.

The emission of volcanic gases can occur both during volcanic eruptions and in quiescent stages of the volcanic activity. This process can affect the air quality in the areas downwind; in fact, many gas species can be a threat to human health and even life at concentrations and doses above species-specific thresholds. Gas emissions can be of different types, the two main categories being dilute passive degassing and heavy gas flow. The former occurs when the gas concentration is low and/or temperature is high, hence its density is lower than the atmospheric density; the latter takes place when the gas density is higher than the atmosphere and the gas accumulates onto the ground and may flow as a gravity current more or less affected by the wind. Examples of the first and second types of emissions are fumaroles and limnic explosions, respectively.Numerical modelling is one of the approaches used to quantify the hazard related to these processes. Ideally, for hazard quantification purposes numerous simulations originating from varying the most important input parameters (e.g. wind field, emission rates, etc.) in their range of uncertainty should be carried out. The whole process of gas dispersion modelling is time consuming, since it starts with the assessment of the wind field with an ad-hoc meteorological model, proceeds with the actual gas dispersion simulation and concludes with the post-processing stage. In order to simplify the whole workflow with the final aim to manage numerous simultaneous simulations for hazard assessment applications, we created APVGDM (Automatic Probabilistic Volcanic Gas Dispersion Modelling), a simulation tool made of a collection of Python scripts. APVGDM is interfaced with two dispersion models that can be selected by the user depending on the application of interest: a dilute (DISGAS) and a dense gas (TWODEE) dispersion model. The post-processing script is capable of building Empirical Cumulative Distribution Functions (ECDF) of the gas concentrations combining the outputs of multiple simulations; the ECDF can be interrogated by the user to produce outputs at the desired exceedance probability. Here we present APVGDM and some application examples showing the wide range of options that the tool offers.

In recent decades, reliable computational models have significantly advanced, and now represent a valuable tool to make quantitative and testable predictions, supporting gas dispersal forecasting and hazard assessments for public safety. In this study, we carried out a number of tests aimed to validate the modelling of gas dispersal at La Soufrière de Guadeloupe volcano (Lesser Antilles), which has shown quasi-permanent degassing of low-temperature hydrothermal nature since its last magmatic eruption in 1530 AD. In particular, we focused on the distribution of CO2 and H2S discharged from the three main present-day fumarolic sources at the summit, using the MultiGAS measurements of continuous gas concentrations collected during March-April 2017. We implemented the open-source Eulerian code DISGAS-2.0 for passive gas dispersion coupled with the mass consistent Diagnostic Wind Model (DWM), using wind measurements and atmospheric stability information from a local meteorological station and the ECMWF-ERA5 reanalysis data. We found that model outputs are highly dependent on the resolution of the topographic data, which affect mainly the reliability of DWM meteorological fields, especially on and around the steep dome. Our results satisfactory reproduce the observed data, indicating the potential usefulness of DISGAS-2.0 as a tool for quantifying gas hazard and reproducing the fumarolic degassing and at La Soufrière de Guadeloupe.

A statistical analysis of explosive eruptive events can give important clues on the behavior of a volcano for both the time- and size-domains, producing crucial information for hazards assessment. In this paper, we analyze in these domains an up-to-date catalog of eruptive events at Galeras volcano, collating data from the Colombian Geological Survey and from the Smithsonian Institution. The dataset appears to be complete, stationary and consisting of independent events since 1820, for events of magnitude ≥2.6. In the time-domain, Inter-Event Times are fitted by various renewal models to describe the observed repose times. On the basis of the Akaike Information Criterion, the preferred model is the Lognormal, with a characteristic time scale of ∼1.6 years. However, a tendency for the events to cluster in time into “eruptive cycles” is observed. Therefore, we perform a cluster analysis, to objectively identify clusters of events: we find three plausible partitions into 6, 8 and 11 clusters of events with magnitude≥2.6the 6-cluster partition being the preferred. The Inter-Event Times between cluster onsets (inter-cluster) and between events belonging to the same cluster (intra-cluster) are also modeled by renewal models. For inter-cluster data, the preferred model is the Brownian Passage Time, describing a periodical occurrence (mean return time ∼36 years) perturbed by a Gaussian noise. For the intra-cluster explosions, the preferred model is the Lognormal, with a characteristic time scale of ∼0.9 years. In the size-domain, we analyze only single events, due to the low number of clusters. Considering two independent parts of the catalog, we cannot reject the null hypothesis of the erupted mass being described by a power law, implying no characteristic eruption size. Finally, looking for time- and size-predictability, we find a significant inverse linear relationship between the logarithm of the erupted mass during a cycle and the time to the subsequent one. These results suggest that, presently, Galeras is still in the eruption cycle started in 2007; a new eruptive cycle may be expected in a few decades, unless the present cluster resumes to activity with magnitude ≥2.6.

Reducing the impact of volcanic eruptions on society is a major challenge of volcanology. Although science is one basic component of risk reduction, the achievement of this goal requires competencies that go beyond natural sciences. Nowadays, the importance of non-scientific factors in reducing volcanic risks is often and dangerously overlooked, possibly leading to decision making that cannot be rationally justified. In this paper we explore the basic components of an ideal decision-making process, identifying the roles and responsibilities of the different partners/tasks that are involved. In particular, we advocate the use of the hazard/risk separation principle, which can help discern unambiguously the role of volcanology (and more in general of science) in the whole risk-reduction process. Although this distinction may be of low relevance in some real cases—for example, when the costs of mitigation actions are low and the likelihood of eruption is high—it becomes of paramount importance when dealing with high uncertainty on the eruption onset or size and expensive mitigation actions, such as the evacuation of a large city. Volcanologists can play different roles in the decision-making process, but they have to be aware that this demands competencies that go beyond being a good volcano scientist. The final intent of this paper is to encourage constructive cooperation between volcanologists and public-policy makers keeping separated their own tasks as defined by their roles and their competences, with the intent of establishing fully transparent decision-making protocols well before volcanic crises. These protocols can be very helpful to audit the decision-making process at any time, and they may be an excellent communication tool for the interested stakeholders, including society.

Jan Mayen Island (Norway), located in the North Atlantic, is considered the world’s northernmost active subaerial volcano, with at least five eruptive periods recorded during the last 200 years. Explosive activity of the volcano may seriously affects the nearby important air traffic routes. However, no quantitative studies on the possible impact of a new explosive volcanic eruption on the air traffic have been conducted. In this work, we statistically characterise the spatial and temporal distribution of airborne volcanic ash cloud and its persistence at different flight levels. Since current operational forecast products do not always meet the requirements of the aviation sector and related stakeholders (using coarse time and space scales, with outputs on a 40 km horizontal resolution grid and 6 hour time averages), and they neglect epistemic/aleatory uncertainties in quantitative forecasts on real time, we propose hourly high resolution hazard maps over a 3D-grid covering a 2 km-resolution spatial domain 2000 km x 2000 km wide. We present the use of high-performance computing (HPC) to overcome the computational limitations associated with unbiased long-term probabilistic volcanic hazard assessment (PVHA) .Considering a continuum of possible combinations of Eruptive Source Parameters (ESP) to assess and quantify the uncertainty, and the natural variability associated with wind fields over 20 years of data, from 1999 to 2019, we run thousands of analytical solutions (numerical simulations) using the most recent version of the FALL3D model. As a result, the first comprehensive long-term PVHA for Jan Mayen volcanic island is presented.

The Center of Excellence for Exascale in Solid Earth (ChEESE; https://cheese-coe.eu) is promoting the use of upcoming Exascale and extreme performance computing capabilities in the area of Solid Earth by harnessing institutions in charge of operational monitoring networks, tier-0 supercomputing centers, academia, hardware developers and third parties from SMEs, Industry and public-governance. The scientific challenging ambition is to prepare 10 European open-source flagship codes to solve Exascale problems on computational seismology, magnetohydrodynamics, physical volcanology, tsunamis, and data analysis. Preparation to Exascale is considering code inter-kernel aspects of simulation workflows like data management and sharing following the FAIR principles, I/O, post-process and visualization. The project is articulated around 12 Pilot Demonstrators (PDs) in which flagship codes are used for near real-time seismic simulations and full-wave inversion, ensemble-based volcanic ash dispersal forecasts, faster than real-time tsunami simulations and physics-based hazard assessments for earthquakes, volcanoes and tsunamis. This is a first step towards enabling of operational e-services requiring of extreme HPC on urgent computing, early warning forecast of geohazards, hazard assessment and data analytics. Additionally, and in collaboration with the European Plate Observing System (EPOS), ChEESE will promote and facilitate the integration of HPC services to widen the access to codes and fostering transfer of know-how to Solid Earth user communities. In this regard, the project aims at acting as a hub to foster HPC across the Solid Earth Community and related stakeholders and to provide specialized training on services and capacity building measures.

In the last decade, probabilistic volcanic hazard assessment (PVHA) has become one of the most rapidly developing topics in volcanology. PVHA relies on a number of simulation tools, which have been catalogued within H2020 EUROVOLC project.Here we apply two of these tools that will concur to a probabilistic multi-hazard assessment for volcanic phenomena at La Soufrière de Guadeloupe, as reported in reference scenarios elaborated by OVSG-IPGP and communicated to the authorities. In the last 9 kyr the activity at La Soufrière is characterized by recurrent effusive to explosive activity, sector collapses and intense fumarolic emissions. Based on literature data, we focus particularly on the most likely explosive (phreatic or hydrothermal) scenario and explore the hazard posed by gas dispersal and ballistics impact, which have never been the focus of PVHA. We also set up a preliminary spatial map for phreatic vent opening, a baseline for the PVHA here presented.

Despite the growing spatio-temporal density of geophysical observations, our understanding of the megathrust earthquake cycle continues to be limited by a series of factors, in particular the short observation time compared to mega-earthquake recurrence and the partial spatial coverage of geodetic data. Here, we attempt to compensate for these natural limitations by simulating dozens of seismic cycles in a laboratory-scale analogue model of subduction. The model creates analog earthquakes of magnitude Mw 6.2–8.3, with a coefficient of variation in recurrence intervals of 0.5, similar to real subduction megathrusts. Using a digital image correlation technique, we measure coseismic and interseismic deformation – this is akin to having a dense continuous geodetic network homogeneously distributed over the whole margin. We show how, by deciphering the spatially and temporally complex surface deformation history, machine learning can predict the timing and size of analog earthquakes. Then, we investigate data characteristics that maximize the performance of a machine learning binary classifier predicting slip-events imminence. We show how this framing can be used for designing an efficient geodetic network, and defining the minimum space-time coverage requirements for analog earthquake prediction. Converting the laboratory scale to the natural scale, we found that a 70-85 km wide coastal swath gives the most important information on slip imminence and that model performance is mainly influenced by the alarm duration, with density of stations and record length playing a secondary role. Under optimal monitoring conditions, about ten seismic cycles long record is enough to predict alarm periods in good agreement with those observed.

One of the aims of EUROVOLC is to raise awareness and share data by exploiting existing tools for hazard and risk. Here we present the ongoing effort within EUROVOLC WP12 to create an online tool to collect information from people witnessing volcanic events at European or other volcanoes.In the recent past, building on the experience from earthquakes, and from the trans-national effects of Eyjafallajökull eruption, European research groups have built tools (e.g. questionnaires or apps) for facilitating the collection of data by citizens. These efforts are presently fragmented and sparse across Europe (and across the world).As the first step we have conducted a reconnaissance survey of existing citizen science tools in volcanology (from operational and research projects), available for download through EUROVOLC websiteOne of the aims of EUROVOLC is to raise awareness and share data by exploiting existing tools for hazard and risk. Here we present the ongoing effort within EUROVOLC WP12 to create an online tool to collect information from people witnessing volcanic events at European or other volcanoes.In the recent past, building on the experience from earthquakes, and from the trans-national effects of Eyjafallajökull eruption, European research groups have built tools (e.g. questionnaires or apps) for facilitating the collection of data by citizens. These efforts are presently fragmented and sparse across Europe (and across the world).As the first step we have conducted a reconnaissance survey of existing citizen science tools in volcanology (from operational and research projects), available for download through EUROVOLC website.The new EUROVOLC tool will:- access and collate data collected by several pre-existing tools. These tools currently include ‘myVolcano’ by British Geological Survey; sulphur dioxide and ash recording tools by Iceland Met Office; Osservatorio Vesuviano web questionnaire & Tefranet by INGV-Catania. These tools were selected based on whether their data can be ‘pulled’ in real-time;- allow additional tools to be incorporated as they become available;- allow recording of new data by the users;- allow visualizing on a map the data in which the users are interested in, that can be selected by region/country, by recording time, or by observed phenomenon;- allow downloading the data in which the users are interested inIn this way, the users of EUROVOLC tool will have access to observations collected by the multiple tools available across EUROPE through a single access point.The EUROVOLC tool will become available in July 2020.

A methodology for a comprehensive probabilistic tsunami hazard analysis is presented for the major sources of tsunamis (seismic events, landslides, and volcanic activity) and preliminarily applied in the Gulf of Naples (Italy). The methodology uses both a modular procedure to evaluate the tsunami hazard and a Bayesian analysis to include the historical information of the past tsunami events. In the urn:x-wiley:jgrc:media:jgrc23818:jgrc23818-math-0001 the submarine earthquakes and the submarine mass failures are initially identified in a gridded domain and defined by a set of parameters, producing the sea floor deformations and the corresponding initial tsunami waves. Differently volcanic tsunamis generate sea surface waves caused by pyroclastic density currents from Somma‐Vesuvius. In the urn:x-wiley:jgrc:media:jgrc23818:jgrc23818-math-0002 the tsunami waves are simulated and propagated in the deep sea by a numerical model that solves the shallow water equations. In the urn:x-wiley:jgrc:media:jgrc23818:jgrc23818-math-0003 the tsunami wave heights are estimated at the coast using the urn:x-wiley:jgrc:media:jgrc23818:jgrc23818-math-0004's amplification law. The selected tsunami intensity is the wave height. In the urn:x-wiley:jgrc:media:jgrc23818:jgrc23818-math-0005 the probabilistic tsunami analysis computes the long‐term comprehensive Bayesian probabilistic tsunami hazard analysis. In the prior analysis the probabilities from the scenarios in which the tsunami parameter overcomes the selected threshold levels are combined with the spatial, temporal, and frequency‐size probabilities of occurrence of the tsunamigenic sources. The urn:x-wiley:jgrc:media:jgrc23818:jgrc23818-math-0006 probability density functions are integrated with the urn:x-wiley:jgrc:media:jgrc23818:jgrc23818-math-0007 derived from the historical information based on past tsunami data. The urn:x-wiley:jgrc:media:jgrc23818:jgrc23818-math-0008 probability density functions are evaluated to produce the hazard curves in selected sites of the Gulf of Naples. Plain Language Summary Probabilistic analyses are essential to estimate the natural hazards caused by infrequent and devastating events and to elaborate risk assessments aiming to mitigate and reduce the impact of the natural disasters on society. Probabilistic tsunami hazard analyses use procedures that trace and weight the different tsunami sources (submarine earthquakes, aerial/submarine slides, volcanic activity, meteorological events, and asteroid impacts) with varying probability of occurrence. The scope of the present methodology is the reduction of possible biases and underestimations that arise by focusing on a single tunamigenic source. The multisource probabilistic tsunami hazard analysis is applied to a real case study, the Gulf of Naples (Italy), where relevant threats due to natural events exist in a high densely populated district. The probabilistic hazard procedure takes into account multiple tsunamigenic sources in the region and provides a first‐order prioritization of the different sources in a long‐term comprehensive analysis. The methodology is based on a Bayesian approach that merges computational hazard quantification (based on source‐tsunami simulations) and past data, appropriately including in the quantification the epistemic uncertainty. For the first time a probabilistic analysis of the tsunami hazard in the region is presented taking into consideration multiple tsunamigenic sources.

Subduction zones are monitored using space geodesy with increasing resolution, with the aim of better capturing the deformation accompanying the seismic cycle. Here, we investigate data characteristics that maximize the performance of a machine learning binary classifier predicting slip‐event imminence. We overcome the scarcity of recorded instances from real subduction zones using data from a seismotectonic analog model monitored with a spatially dense, continuously recording onshore geodetic network. We show that a 70–85 km‐wide coastal swath recording interseismic deformation gives the most important information on slip imminence. Prediction performances are mainly influenced by the alarm duration (amount of time that we consider an event as imminent), with density of stations and record length playing a secondary role. The techniques developed in this study are most likely applicable in regions of slow earthquakes, where stick‐slip‐like failures occur at time intervals of months to years., Plain Language Summary: Machine learning, a group of algorithms that produce predictions based on past “experience,” has been successfully used to predict various aspects of the earthquake process, including slip imminence. The accuracy of those algorithms depends on a variety of data characteristics, for example, the amount of data used for building the “experience” of the model. We focus on this point using a scaled representation of a seismic subduction zone and a monitoring technique similar to Global Navigation Satellite System. We identify the most useful surface regions to be monitored and the parameter that most strongly influences prediction accuracy for the timing of upcoming laboratory earthquakes. The routine implemented in this study could be used to predict the onset and extent of slow earthquakes., Key Points: We investigate the performances of a binary classifier predicting slip‐event imminence in analog models of megathrust seismic cycling. A 70–85 km‐wide coastal swath is the region producing the most important information for the imminence classification. Length of time that we consider an event imminent plays a primary role in tuning the performances of a binary classifier predicting the imminence of analog earthquakes., DAAD‐Prime

International audience; The observed maximum magnitude of subduction megathrust earthquakes is highly variable worldwide. One key question is which conditions, if any, favor the occurrence of giant earthquakes (Mw ≥ 8.5). Here we carry out a multivariate statistical study in order to investigate the factors affecting the maximum magnitude of subduction megathrust earthquakes. We find that the trench-parallel extent of subduction zones and the thickness of trench sediments provide the largest discriminating capability between subduction zones that have experienced giant earthquakes and those having significantly lower maximum magnitude. Monte Carlo simulations show that the observed spatial distribution of giant earthquakes cannot be explained by pure chance to a statistically significant level. We suggest that the combination of a long subduction zone with thick trench sediments likely promotes a great lateral rupture propagation, characteristic of almost all giant earthquakes.

The ultimate goal of volcanology is forecasting eruptions. This task is particularly challenging at calderas, where unrest is frequent, affects wider areas and its evidence is often masked by the activity of hydrothermal systems. A recent study has compiled a database on caldera unrest, derived from seismicity, geodetic, gravity and geochemical monitoring data at calderas worldwide, from 1988 to 2014. Here we exploit this database, searching for the most recurring features of unrest and, in turn, its possible dynamics. In particular, we focus on: a) the duration of unrest at calderas; b) recurring patterns in unrest; c) unrest episodes culminating in eruptions, including time- or size-predictability and a multivariate regression analysis. Our analysis indicates that pre-eruptive unrest is shorter than non-eruptive unrest, particularly with open or semi-open calderas, calderas with mafic or mixed composition of past eruptive products, or unrest driven by mafic magma; conversely, lack of data on pre-eruptive unrest driven by felsic magma and/or at felsic or plugged calderas prevents an analysis of these specific subsets. In addition, 72% of pre-eruptive unrest lasts

Since 1978 water chemistry of the Laguna Caliente crater lake has been used to monitor volcanic activity at Poas, Costa Rica, making it arguably the best studied hyper-acidic crater lake on Earth. During these 39 years, three phases of unrest occurred, manifested through frequent phreatic eruptions, with each a duration of several years to over a decade (1978–1980, 1986–1996, 2006–2016). We here present a novel technique to deal with the long time series of the chemical composition of water of Laguna Caliente, independent on previous deterministic research and resulting conceptual models. Common patterns of chemical parameters in relation with phreatic eruptive activity for the period 1978–September 2014 are sought, applying the objective statistical method of Pattern Recognition. This resulted in the definition of the strongest precursory signals and their respective thresholds. Numerical outcomes often confirm findings based on geochemical models (e.g. SO4, SO4/Cl and pH are strong monitoring parameters). However, some surprising parameters (opposite behavior of Mg/Cl ratios, decreases in Ca and Mg concentrations, increasing Al/Mg ratios) still need a geochemical explanation and should be a focus for future research strategies. The obtained parameters and thresholds were retrospectively applied for the “test period” of the Pattern Recognition method (November 2014–February 2016). This test provided hints that suggested that eruptive activity at Poas was not yet over, despite apparent quiescence in early 2016. Indeed, after new phreatic eruptions since May 2016, the 2006–2016 phreatic eruptive cycle culminated in phreatomagmatic activity in April 2017. We conclude that evaluating time series of chemical composition of crater lakes framed in the Pattern Recognition method can be a useful monitoring approach. Moreover, increased sampling frequency can provide more details and more adequate prediction of phreatic activity at Poas. Comparing Laguna Caliente with other two well monitored acidic crater lakes (Ruapehu Crater Lake, New Zealand and Yugama, Kusatsu-Shirane, Japan) Poas results unique in many ways and undoubtedly the most active crater lake of the three during the past four decades.

Quantitatively assessing long-term volcanic risk can be challenging due to the many variables associated with volcanic hazard and vulnerability. This study presents a structured first-order approach for considering variables in hazard and vulnerability analyses, such as eruption style and cyclic fragility, in order to quantitatively estimate risk. Probabilistic volcanic hazard data derived from advection–diffusion–sedimentation tephra fall model TEPHRA2 and probabilistic volcanic hazard analysis tool BET_VH (Bayesian Event Tree for Volcanic Hazards) are combined with fragility functions and seasonal vulnerability coefficients for agricultural production to calculate volcanic risk indices which represent the likelihood of damage or loss to farm production over a given time frame. The resulting dataset allows for approximations of quantitative risk over a continuous range of ash thickness thresholds, at multiple levels of uncertainty, and in the context of fluctuating hazard and vulnerability environments (e.g., seasonal wind patterns and crop phases). We illustrate this approach through a case study which evaluates the risk of incurring 90% damage to agricultural production at dairy and fruit farms in the Bay of Plenty region of New Zealand (BoP) due to ashfall from a Plinian eruption phase at the large local caldera volcano, the Okataina Volcanic Centre (OVC). Consideration of seasonal wind profiles, seasonal fluctuations in fruit and dairy farm vulnerability, multiple possible OVC eruption styles, different possible OVC vent locations, and a continuous distribution of ash thickness and damage thresholds enables a multi-dimensional analysis that aims to reflect the natural complexity and interdependencies associated with volcanic risk. A risk uncertainty matrix is introduced as a conceptual scheme to help guide evaluation and communication of the results of such quantitative risk analyses by showing how different types of uncertainty can yield “maximum”, “average”, or “minimum” estimates of risk. Results of this case study indicate that BoP fruit farms are at higher risk of experiencing damage and production loss from OVC ashfall than dairy farms, and farms to the east of the OVC are typically at higher risk than farms to the north of the OVC. Forecasts based on the annual maximum estimate of risk for fruit farms show a regional average of 2.3% probability (greater than 1 in 50 likelihood) of experiencing 90% damage from a basaltic or rhyolitic Plinian eruption from anywhere within the OVC over a period of 100 years. Seasonal-level analyses revealed that the risk of experiencing losses due to OVC ashfall at fruit farms is cyclic and fluctuates with time of year and harvest season, with the highest risk experienced during peak harvest season (15 October–14 April) when crop vulnerability is high and westerly winds dominate in the BoP.

We report on the first ‘real-time' application of the BET_UNREST (Bayesian Event Tree for Volcanic Unrest) probabilistic model, during a VUELCO Simulation Exercise carried out on the island of Dominica, Lesser Antilles, in May 2015. Dominica has a concentration of nine potentially active volcanic centers and frequent volcanic earthquake swarms at shallow depths, intense geothermal activity and recent phreatic explosions (1997) indicate the region is still active. The exercise scenario was developed in secret by a team of scientists from University of West Indies (Trinidad and Tobago) and University of Auckland (New Zealand). The simulated unrest activity was provided to the exercise's Scientific Team in three ‘phases' through exercise injects comprising processed monitoring data. We applied the newly created BET_UNREST model through its software implementation PyBetUnrest, to estimate the probabilities of having i) unrest of ii) magmatic, hydrothermal or tectonic origin, which may or may not lead to iii) an eruption. The probabilities obtained for each simulated phase raised controversy and intense deliberations among the members of the scientific team. The results were often considered to be ‘too high', and were not included in any of the reports presented to ODM (Office for Disaster Management) revealing interesting crisis communication challenges. We concluded that the PyBetUnrest application itself was successful and brought the tool one step closer to a full implementation. However, as with any newly proposed method it needs more testing, and in order to be able to use it in the future we make a series of recommendations for future applications. This article is protected by copyright. All rights reserved.

Giant earthquakes (Mw≥8.5) usually occur on the boundary between subducting and overriding plates of converging margins, but it is not yet clear which (if any) subduction zones are more prone to produce such a kind of events. Here we show that subduction zones may have different capabilities to produce giant earthquakes. We analyze the frequency-magnitude distribution of the interplate earthquakes at subduction zones that occurred during 1976–2007 and calculate the propensity (defined as the average annual rate) of giant events for about half of the subduction zones. We find that the b value of interplate earthquakes is significantly different among the subduction zones, and out-of-sample giant earthquakes (before 1976 and after 2007) have occurred preferentially in high-propensity areas. Besides the importance for seismic hazard assessment and risk mitigation, our results seem to indicate that a higher seismicity rate does not necessarily imply a higher likelihood to generate giant earthquakes, and the way in which the stress is released at a subduction interface does not change significantly after the occurrence of such events.

Although most of volcanic hazard studies focus on magmatic eruptions, volcanic hazardous events can also occur when no migration of magma can be recognized. Examples are tectonic and hydrothermal unrest that may lead to phreatic eruptions. Recent events (e.g., Ontake eruption on September 2014) have demonstrated that phreatic eruptions are still hard to forecast, despite being potentially very hazardous. For these reasons, it is of paramount importance to identify indicators that define the condition of nonmagmatic unrest, in particular for hydrothermal systems. Often, this type of unrest is driven by movement of fluids, requiring alternative monitoring setups, beyond the classical seismic-geodetic-geochemical architectures. Here we present a new version of the probabilistic BET (Bayesian Event Tree) model, specifically developed to include the forecasting of nonmagmatic unrest and related hazards. The structure of the new event tree differs from the previous schemes by adding a specific branch to detail nonmagmatic unrest outcomes. A further goal of this work consists in providing a user-friendly, open-access, and straightforward tool to handle the probabilistic forecast and visualize the results as possible support during a volcanic crisis. The new event tree and tool are here applied to Kawah Ijen stratovolcano, Indonesia, as exemplificative application. In particular, the tool is set on the basis of monitoring data for the learning period 2000-2010, and is then blindly applied to the test period 2010-2012, during which significant unrest phases occurred.

International audience; The tectono-magmatic relationships along obliquely convergent plate boundaries, where strain partitioning promotes strike-slip structures along the volcanic arc, are poorly known. Here it is unclear if and, in case, how the strike-slip structures control volcanic processes, distribution and size. To better define the possible tectono-magmatic relationships along strike-slip arcs, we merge available information on the case study of Sumatra (Indonesia) with field structural data. The Sumatra arc (entire volcanic belt) consists of 48 active volcanoes. Of these, 46% lie within 10 km from the dextral Great Sumatra Fault (GSF), which carries most horizontal displacement on the overriding plate, whereas 27% lie at >20 km from the GSF. Among the volcanoes at

The metropolitan area of Napoli (∼3 M inhabitants) in southern Italy is located in between two explosive active volcanoes: Somma-Vesuvius and Campi Flegrei. Pyroclastic density currents (PDCs) from these volcanoes may reach the city center, as witnessed by the Late Quaternary stratigraphic record. Here we compute a novel multivolcano Probabilistic Volcanic Hazard Assessment of PDCs, in the next 50 years, by combining the probability of PDC invasion from each volcano (assuming that they erupt independently) over the city of Napoli and its surroundings. We model PDC invasion with the energy cone model accounting for flows of very different (but realistic) mobility and use the Bayesian Event Tree for Volcanic Hazard to incorporate other volcano-specific information such as the probability of eruption or the spatial variability in vent opening probability. Worthy of note, the method provides a complete description of Probabilistic Volcanic Hazard Assessment, that is, it yields percentile maps displaying the epistemic uncertainty associated with our best (aleatory) hazard estimation. Since the probability density functions of the model parameters of the energy cone are unknown, we propose an ensemble of different hazard assessments based on different assumptions on such probability density functions. The ensemble merges two plausible distributions for the collapse height, reflecting a source of epistemic (specifically, parametric) uncertainty. We also apply a novel quantification for a spatially varying epistemic uncertainty associated to PDC simulations. ©2018. American Geophysical Union. All Rights Reserved.

PyBetVH is a completely new, free, open-source and cross-platform software implementation of the Bayesian Event Tree for Volcanic Hazard (BET_VH), a tool for estimating the probability of any magmatic hazardous phenomenon occurring in a selected time frame, accounting for all the uncertainties. New capabilities of this implementation include the ability to calculate hazard curves which describe the distribution of the exceedance probability as a function of intensity (e.g., tephra load) on a grid of points covering the target area. The computed hazard curves are (i) absolute (accounting for the probability of eruption in a given time frame, and for all the possible vent locations and eruptive sizes) and (ii) Bayesian (computed at different percentiles, in order to quantify the epistemic uncertainty). Such curves allow representation of the full information contained in the probabilistic volcanic hazard assessment (PVHA) and are well suited to become a main input to quantitative risk analyses. PyBetVH allows for interactive visualization of both the computed hazard curves, and the corresponding Bayesian hazard/probability maps. PyBetVH is designed to minimize the efforts of end users, making PVHA results accessible to people who may be less experienced in probabilistic methodologies, e.g. decision makers. The broad compatibility of Python language has also allowed PyBetVH to be installed on the VHub cyber-infrastructure, where it can be run online or downloaded at no cost. PyBetVH can be used to assess any type of magmatic hazard from any volcano. Here we illustrate how to perform a PVHA through PyBetVH using the example of analyzing tephra fallout from the Okataina Volcanic Centre (OVC), New Zealand, and highlight the range of outputs that the tool can generate.