Valentina Magni
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Valentina Magni has a total of 101 co-authors. Here are the top co-authors by number of shared publications:
- Jeroen van Hunen
Durham University
15 shared publications - Finn Løvholt
Norwegian Geotechnical Institute, Offshore Energy Department, Sandakerveien 140, 0484 Oslo, Norway
8 shared publications - Pierre Bouilhol
Université de Lorraine, CNRS, CRPG, Nancy, France
7 shared publications - Steven John Gibbons
Norwegian Geotechnical Institute Oslo Norway
7 shared publications - Manuela Volpe
Istituto Nazionale di Geofisica e Vulcanologia Roma Italy
5 shared publications - Fabrizio Romano
Istituto Nazionale di Geofisica e Vulcanologia Roma Italy
4 shared publications
Latest Publications
- Open Access
- 22 September 2026
- Published by Elsevier in Quaternary Science Advances
- Vol. 24
Propagation of the Storegga tsunami in the southeastern North Sea
- 22 September 2026
- Published by Elsevier in Quaternary Science Advances
- Vol. 24
Open AccessPropagation of the Storegga tsunami in the southeastern North Sea
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
Probabilistic tsunami forecasting for earthquakes at global scale
Probabilistic Tsunami Forecasting (PTF) provides a rapid estimation of tsunami hazard intensity probabilities at given forecast points when a potentially tsunamigenic earthquake has occurred. According to predefined rules provided by the decision makers, PTF can also convert these values into uncertainty-informed alert levels that can be used in operational tsunami early warning or post-event actions for risk reduction (for example, evacuation). The PTF workflow is planned to become operational at the Italian Tsunami Warning Center (CAT-INGV) with a specific setting for delivering early warning messages in the Mediterranean area. Indeed, the PTF implemented for the CAT-INGV relies on the long-term hazard model NEAMTHM18 and on a large database of precomputed tsunami scenarios.Here we present the first prototype of the PTF extension at global scale, where the ensemble of seismic scenarios is defined from scratch using real-time data (hypocenter and magnitude of the event) and moment tensor solutions provided by an ad hoc integrated tool and external agencies in quasi real time. Each source parameter is discretized within a given range of values around the provided solution and the corresponding uncertainties are assigned through weight distributions of these parametrizations. For each scenario, the initial sea floor displacement is computed based on a standard uniform rupture model. The corresponding tsunami impact is estimated using on-the-fly numerical simulations, requiring dedicated HPC resources.This global-scale version of the PTF is here presented through the hindcast of two major megathrust events in the Pacific Ocean; the 2010 Mw 8.8 Maule, Chile and the 2011 Mw 9.1, Tohoku, Japan earthquakes and tsunamis.This work was partially funded by the DT-GEO project (A Digital Twin for GEOphysical extremes, https://dtgeo.eu/) through the European Union’s Horizon Europe research and innovation programme under grant agreement nº 101058129.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessProbabilistic tsunami forecasting for earthquakes at global scale
Probabilistic Tsunami Forecasting (PTF) provides a rapid estimation of tsunami hazard intensity probabilities at given forecast points when a potentially tsunamigenic earthquake has occurred. According to predefined rules provided by the decision makers, PTF can also convert these values into uncertainty-informed alert levels that can be used in operational tsunami early warning or post-event actions for risk reduction (for example, evacuation). The PTF workflow is planned to become operational at the Italian Tsunami Warning Center (CAT-INGV) with a specific setting for delivering early warning messages in the Mediterranean area. Indeed, the PTF implemented for the CAT-INGV relies on the long-term hazard model NEAMTHM18 and on a large database of precomputed tsunami scenarios.Here we present the first prototype of the PTF extension at global scale, where the ensemble of seismic scenarios is defined from scratch using real-time data (hypocenter and magnitude of the event) and moment tensor solutions provided by an ad hoc integrated tool and external agencies in quasi real time. Each source parameter is discretized within a given range of values around the provided solution and the corresponding uncertainties are assigned through weight distributions of these parametrizations. For each scenario, the initial sea floor displacement is computed based on a standard uniform rupture model. The corresponding tsunami impact is estimated using on-the-fly numerical simulations, requiring dedicated HPC resources.This global-scale version of the PTF is here presented through the hindcast of two major megathrust events in the Pacific Ocean; the 2010 Mw 8.8 Maule, Chile and the 2011 Mw 9.1, Tohoku, Japan earthquakes and tsunamis.This work was partially funded by the DT-GEO project (A Digital Twin for GEOphysical extremes, https://dtgeo.eu/) through the European Union’s Horizon Europe research and innovation programme under grant agreement nº 101058129.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
Propagation of the Storegga tsunami in the south eastern North Sea
Numerical simulations of the 8150 BP Storegga slide event in the Norwegian Sea need to be consistent both with landslide runout and with observations of tsunami run-up. Here we focus on the southern North Sea, South East of Dogger Bank, and the shores of Denmark and Germany. The slide volume and dynamics need to generate a realistic initial tsunami wave and the simulation of the propagation and inundation of the tsunami would need the correct bathymetry of the North Sea 8150 years ago. From both Denmark and Germany there have been reports of deposits claimed to be associated with the Storegga tsunami. We simulate the slide using a recent cohesive landslide model, that has demonstrated success in predicting both the extent of runout deposits and tsunami run-up heights farther north, and we run suites of tsunami simulations to predict tsunami surface elevations, velocities, and arrival times. Appreciating the uncertainty associated with the paleobathymetry, we parametrize a continuum of bathymetric models between the present day topobathymetry and two different paleobathymetric models and perform simulations for multiple candidate bathymetric models. For many models in which Dogger Bank is completely submerged, the shallow water in this part of the North Sea still represents a significant barrier to the tsunami propagation. We compare tsunami metrics between the alternative topobathymetries systematically and conclude that the tsunami surface elevations and velocities in the southeastern North Sea are probably too small to have caused erosion and significant deposition of tsunami debris along the coasts of Denmark and Germany. All numerical simulation results performed for this study are openly archived on the Geo-INQUIRE Simulation Data Lake.We acknowledge Geo-INQUIRE, funded by the European Commission under project number 101058518 within the HORIZON-INFRA-2021-SERV-01 call, DT-GEO, funded by Horizon Europe under Grant Agreement No 101058129, and ChEESE-2P, funded by the European Union and the European High Performance Computing Joint Undertaking (JU) together with Spain, Italy, Iceland, Germany, Norway, France, Finland, and Croatia under grant agreement No 101093038.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessPropagation of the Storegga tsunami in the south eastern North Sea
Numerical simulations of the 8150 BP Storegga slide event in the Norwegian Sea need to be consistent both with landslide runout and with observations of tsunami run-up. Here we focus on the southern North Sea, South East of Dogger Bank, and the shores of Denmark and Germany. The slide volume and dynamics need to generate a realistic initial tsunami wave and the simulation of the propagation and inundation of the tsunami would need the correct bathymetry of the North Sea 8150 years ago. From both Denmark and Germany there have been reports of deposits claimed to be associated with the Storegga tsunami. We simulate the slide using a recent cohesive landslide model, that has demonstrated success in predicting both the extent of runout deposits and tsunami run-up heights farther north, and we run suites of tsunami simulations to predict tsunami surface elevations, velocities, and arrival times. Appreciating the uncertainty associated with the paleobathymetry, we parametrize a continuum of bathymetric models between the present day topobathymetry and two different paleobathymetric models and perform simulations for multiple candidate bathymetric models. For many models in which Dogger Bank is completely submerged, the shallow water in this part of the North Sea still represents a significant barrier to the tsunami propagation. We compare tsunami metrics between the alternative topobathymetries systematically and conclude that the tsunami surface elevations and velocities in the southeastern North Sea are probably too small to have caused erosion and significant deposition of tsunami debris along the coasts of Denmark and Germany. All numerical simulation results performed for this study are openly archived on the Geo-INQUIRE Simulation Data Lake.We acknowledge Geo-INQUIRE, funded by the European Commission under project number 101058518 within the HORIZON-INFRA-2021-SERV-01 call, DT-GEO, funded by Horizon Europe under Grant Agreement No 101058129, and ChEESE-2P, funded by the European Union and the European High Performance Computing Joint Undertaking (JU) together with Spain, Italy, Iceland, Germany, Norway, France, Finland, and Croatia under grant agreement No 101093038.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
The effect of magma poor and magma rich rifted margins on continental collision dynamics
Rifted margins form when continents rift apart and are commonly characterized by a thinned transition zone between the continental crust and the oceanic crust. This transition zone can display a wide range of characteristics, which primarily depend on the regional tectonic evolution. The velocity and duration of the rifting process as well as the geodynamic setting influence the properties and geometry of the margins, which are often grouped into two main categories: magma-poor and magma-rich. Magma-rich margins are characterized by large input of mafic melt, while magma-poor margins are characterized by much less magma production during the rifting process, resulting in variations in geometry and rheology of rifted margins worldwide.Using the finite elements code Citcom, we show how different types of rifted margins can influence the dynamics of continental collision, focusing on the time and depth of slab break-off after collision and the fate of margin material. We compared these models as a function of various parameters (e.g., margin length, density, and viscosity), in order to understand how the architecture of a passive margin affects the dynamics of continental collision.We find that rifted margins have a noticeable impact on subduction dynamics, as we observe large variability in slab break-off times and depths. In particular, the presence of a rifted margin can delay slab break-off to up to 60 Myr after the onset of collision. Our results show that a large portion of the weak crust of magma-poor margins is likely to detach from the subducting plate and accrete to the upper plate, while the dense and strong mafic and ultramafic component of magma-rich margins causes most of the margin to subduct and be lost into the mantle, leaving only a small fraction of transitional and oceanic crust at the surface. Therefore, the volume of accreted material is much larger when the margin is magma-poor than magma-rich, which is consistent with geological observations that fossil magma-poor rifted margins are preserved in many mountain ranges, whereas remnants of magma-rich rifted margins are scarce.Importantly, our results show that rifted margin type controls the architecture of the subsequent collisional phase of the Wilson cycle.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessThe effect of magma poor and magma rich rifted margins on continental collision dynamics
Rifted margins form when continents rift apart and are commonly characterized by a thinned transition zone between the continental crust and the oceanic crust. This transition zone can display a wide range of characteristics, which primarily depend on the regional tectonic evolution. The velocity and duration of the rifting process as well as the geodynamic setting influence the properties and geometry of the margins, which are often grouped into two main categories: magma-poor and magma-rich. Magma-rich margins are characterized by large input of mafic melt, while magma-poor margins are characterized by much less magma production during the rifting process, resulting in variations in geometry and rheology of rifted margins worldwide.Using the finite elements code Citcom, we show how different types of rifted margins can influence the dynamics of continental collision, focusing on the time and depth of slab break-off after collision and the fate of margin material. We compared these models as a function of various parameters (e.g., margin length, density, and viscosity), in order to understand how the architecture of a passive margin affects the dynamics of continental collision.We find that rifted margins have a noticeable impact on subduction dynamics, as we observe large variability in slab break-off times and depths. In particular, the presence of a rifted margin can delay slab break-off to up to 60 Myr after the onset of collision. Our results show that a large portion of the weak crust of magma-poor margins is likely to detach from the subducting plate and accrete to the upper plate, while the dense and strong mafic and ultramafic component of magma-rich margins causes most of the margin to subduct and be lost into the mantle, leaving only a small fraction of transitional and oceanic crust at the surface. Therefore, the volume of accreted material is much larger when the margin is magma-poor than magma-rich, which is consistent with geological observations that fossil magma-poor rifted margins are preserved in many mountain ranges, whereas remnants of magma-rich rifted margins are scarce.Importantly, our results show that rifted margin type controls the architecture of the subsequent collisional phase of the Wilson cycle.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
A Landslide Probabilistic Tsunami Hazard Analysis for the Åkerneset and Hegguraksla rockslides (Norway)
Landslide tsunami hazard analysis is characterized by substantial uncertainty, particularly in estimating exceedance probabilities for tsunami heights. As a result, no standardized hazard methodology exists for landslide tsunamis, and most studies rely on deterministic or scenario-based approaches, while Probabilistic Tsunami Hazard Analysis (PTHA) methods are seldom applied. The limited use of probabilistic frameworks is largely due to scarce observational data on past landslide tsunamis and the increased modelling complexity required to represent landslide sources, especially subaerial failures producing impact tsunamis. To address these challenges, we apply a Landslide Probabilistic Tsunami Hazard Analysis (LPTHA) framework previously developed for Norwegian fjord environments, focusing here on the potentialÅkerneset and Hegguraksla rockslides in western Norway. The LPTHA combines landslide occurrence rates derived from slope stability assessments with an event-tree formulation describing uncertainty in key landslide kinematic parameters controlling tsunami generation, such as impact velocity, impact frontal area, and runout distance. Quantification of epistemic uncertainty requires large ensembles of simulations spanning wide parameter ranges, which necessitates computationally efficient modelling approaches. In this study, tsunami generation is represented using a "rounded block” landslide source model, coupled with a linear dispersive wave propagation model and a non-linear shallow-water model for nearshore propagation and inundation. The resulting LPTHA provides probabilistic estimates of tsunami and run-up heights within the fjord system, explicitly accounting for uncertainty in landslide dynamics and tsunami response. The analysis highlights the sensitivity of hazard estimates to landslide source parameterization and demonstrates the applicability of LPTHA as a systematic framework for probabilistic tsunami hazard assessment in fjord settings affected by large unstable rock slopes. Finally, the methodology enables a presentation of run-up heights with corresponding probabilities as required by the Norwegian Planning and Building Act.
- 14 March 2026
- Published by Copernicus GmbH
Open AccessA Landslide Probabilistic Tsunami Hazard Analysis for the Åkerneset and Hegguraksla rockslides (Norway)
Landslide tsunami hazard analysis is characterized by substantial uncertainty, particularly in estimating exceedance probabilities for tsunami heights. As a result, no standardized hazard methodology exists for landslide tsunamis, and most studies rely on deterministic or scenario-based approaches, while Probabilistic Tsunami Hazard Analysis (PTHA) methods are seldom applied. The limited use of probabilistic frameworks is largely due to scarce observational data on past landslide tsunamis and the increased modelling complexity required to represent landslide sources, especially subaerial failures producing impact tsunamis. To address these challenges, we apply a Landslide Probabilistic Tsunami Hazard Analysis (LPTHA) framework previously developed for Norwegian fjord environments, focusing here on the potentialÅkerneset and Hegguraksla rockslides in western Norway. The LPTHA combines landslide occurrence rates derived from slope stability assessments with an event-tree formulation describing uncertainty in key landslide kinematic parameters controlling tsunami generation, such as impact velocity, impact frontal area, and runout distance. Quantification of epistemic uncertainty requires large ensembles of simulations spanning wide parameter ranges, which necessitates computationally efficient modelling approaches. In this study, tsunami generation is represented using a "rounded block” landslide source model, coupled with a linear dispersive wave propagation model and a non-linear shallow-water model for nearshore propagation and inundation. The resulting LPTHA provides probabilistic estimates of tsunami and run-up heights within the fjord system, explicitly accounting for uncertainty in landslide dynamics and tsunami response. The analysis highlights the sensitivity of hazard estimates to landslide source parameterization and demonstrates the applicability of LPTHA as a systematic framework for probabilistic tsunami hazard assessment in fjord settings affected by large unstable rock slopes. Finally, the methodology enables a presentation of run-up heights with corresponding probabilities as required by the Norwegian Planning and Building Act.
- Open Access
- 14 March 2026
- Published by Copernicus GmbH
Multi-Source Tsunami Hazard Assessment for Digital Twin Workflows
Recent advancements in the Digital Twin Component for Tsunamis, developed within the EU-funded DT-GEO project, are transforming rapid hazard assessment from static pre-computed databases to dynamic, data-informed workflows. In this presentation, a novel workflow for Probabilistic Tsunami Forecasting (PTF) due to earthquake-triggered landslides is presented through a site demonstrator for the Mediterranean Sea motived by the 1908 Messina Strait earthquake and tsunami. A key innovation is the integration of earthquake-triggered submarine landslides and the application of AI driven inundation emulators for rapid prediction linked to earthquake workflows and related shakemaps. In addition, we showcase possible use of the workflow for new geophysical settings for a submarine slope in Southwest India. These synergies between digital twin architectures and machine learning provide a robust framework for anticipatory action and disaster risk management at both regional and global scales. This work was partially funded by the EU DT-GEO project (A Digital Twin for GEOphysical extremes, https://dtgeo.eu/) through the European Union’s Horizon Europe research and innovation programme under grant agreement nº 101058129 and PCTWIN project, jointly funded by the Natural Environment Research Council (NERC), UKRI and the Ministry of Earth Sciences (MoES), Government of India (Grant: NE/Z503496/1).
- 14 March 2026
- Published by Copernicus GmbH
Open AccessMulti-Source Tsunami Hazard Assessment for Digital Twin Workflows
Recent advancements in the Digital Twin Component for Tsunamis, developed within the EU-funded DT-GEO project, are transforming rapid hazard assessment from static pre-computed databases to dynamic, data-informed workflows. In this presentation, a novel workflow for Probabilistic Tsunami Forecasting (PTF) due to earthquake-triggered landslides is presented through a site demonstrator for the Mediterranean Sea motived by the 1908 Messina Strait earthquake and tsunami. A key innovation is the integration of earthquake-triggered submarine landslides and the application of AI driven inundation emulators for rapid prediction linked to earthquake workflows and related shakemaps. In addition, we showcase possible use of the workflow for new geophysical settings for a submarine slope in Southwest India. These synergies between digital twin architectures and machine learning provide a robust framework for anticipatory action and disaster risk management at both regional and global scales. This work was partially funded by the EU DT-GEO project (A Digital Twin for GEOphysical extremes, https://dtgeo.eu/) through the European Union’s Horizon Europe research and innovation programme under grant agreement nº 101058129 and PCTWIN project, jointly funded by the Natural Environment Research Council (NERC), UKRI and the Ministry of Earth Sciences (MoES), Government of India (Grant: NE/Z503496/1).
- Open Access
- 13 March 2026
- Published by Copernicus GmbH
Probabilistic Tsunami Hazard Assessment for the Rabat-Salé Coastline ,Morocco.
The Atlantic coast of Morocco is exposed to potentially damaging tsunami events generated by offshore seismic sources. Although several studies have investigated tsunami hazard along the Moroccan coastline, most have relied on deterministic approaches and remain limited in their ability to quantify uncertainty.In this study, we perform a Probabilistic Tsunami Hazard Assessment (PTHA) for the Rabat–Salé coastal region based on thousands of high-resolution tsunami simulations. The methodology follows a three-step workflow: (1) hazard disaggregation and scenario selection, (2) high-resolution tsunami modeling using the Tsunami-HySEA model, and (3) hazard aggregation. High-resolution topo-bathymetric datasets provided by the Moroccan Ministry of Equipment and Water are incorporated to ensure accurate simulation of wave propagation and inundation processes. The results include local hazard curves and probabilistic inundation maps that provide quantitative estimates of tsunami hazard for risk-informed coastal planning and decision-making. This work represents one of the first local-scale PTHA implementations along the Moroccan Atlantic coast and demonstrates the added value of combining advanced numerical modeling with detailed national geospatial datasets for improved coastal risk management. This work was carried out under the Geo-INQUIRE project, funded by the European Commission under project number 101058518 within the HORIZON-INFRA-2021-SERV-01 call.
- 13 March 2026
- Published by Copernicus GmbH
Open AccessProbabilistic Tsunami Hazard Assessment for the Rabat-Salé Coastline ,Morocco.
The Atlantic coast of Morocco is exposed to potentially damaging tsunami events generated by offshore seismic sources. Although several studies have investigated tsunami hazard along the Moroccan coastline, most have relied on deterministic approaches and remain limited in their ability to quantify uncertainty.In this study, we perform a Probabilistic Tsunami Hazard Assessment (PTHA) for the Rabat–Salé coastal region based on thousands of high-resolution tsunami simulations. The methodology follows a three-step workflow: (1) hazard disaggregation and scenario selection, (2) high-resolution tsunami modeling using the Tsunami-HySEA model, and (3) hazard aggregation. High-resolution topo-bathymetric datasets provided by the Moroccan Ministry of Equipment and Water are incorporated to ensure accurate simulation of wave propagation and inundation processes. The results include local hazard curves and probabilistic inundation maps that provide quantitative estimates of tsunami hazard for risk-informed coastal planning and decision-making. This work represents one of the first local-scale PTHA implementations along the Moroccan Atlantic coast and demonstrates the added value of combining advanced numerical modeling with detailed national geospatial datasets for improved coastal risk management. This work was carried out under the Geo-INQUIRE project, funded by the European Commission under project number 101058518 within the HORIZON-INFRA-2021-SERV-01 call.
- Open Access
- 13 March 2025
- Published by American Geophysical Union (AGU) in Geochemistry, Geophysics, Geosystems
- Vol. 26 (3)
Hydrous Regions of the Mantle Transition Zone Lie Beneath Areas of Continental Intraplate Volcanism
Great volumes of water are carried downward into the mantle transition zone (MTZ, 410–670 km depth) by subducting slabs. If this water is later drawn upward, the resulting mantle melting may generate continental intraplate volcanism (IPV). Despite water's importance, its amount and spatial distribution within the MTZ, and its impact on IPV, are poorly constrained. Here we use plate tectonic reconstructions to estimate the rates and positions of water injection into the MTZ by subducted slabs during the past 400 Myr. This allows us to construct global maps of heterogeneous MTZ hydration, which we then compare to IPV eruption locations from the past 200 Myr. We detect a statistically significant correlation between wet MTZ regions and IPV locations at the surface, but only if slabs sink faster than 1 cm/yr, water remains stored in the MTZ for periods of 30–100 Myr, and IPV eruptions occur 10–30 Myr later. We find that 42%–68% of continental IPV is underlain by wet MTZ, with greater fractions associated with longer MTZ residence time. Hydrous underpinning of continental IPV was highest during the Jurassic, when more extensive slab interaction with the MTZ hydrated a wider area of the MTZ. Since the Cretaceous, continents have been moving over the wet MTZ, increasing IPV possibilities. MTZ regions near the northern Pacific, southern Africa, and western Europe have remained dry by avoiding wet slabs. We suggest that subducted water shapes global patterns of intraplate volcanism, with hydrous upwellings rising from the MTZ to generate continental IPV above wet MTZ regions.
- 13 March 2025
- Published by American Geophysical Union (AGU) in Geochemistry, Geophysics, Geosystems
- Vol. 26 (3)
Open AccessHydrous Regions of the Mantle Transition Zone Lie Beneath Areas of Continental Intraplate Volcanism
Great volumes of water are carried downward into the mantle transition zone (MTZ, 410–670 km depth) by subducting slabs. If this water is later drawn upward, the resulting mantle melting may generate continental intraplate volcanism (IPV). Despite water's importance, its amount and spatial distribution within the MTZ, and its impact on IPV, are poorly constrained. Here we use plate tectonic reconstructions to estimate the rates and positions of water injection into the MTZ by subducted slabs during the past 400 Myr. This allows us to construct global maps of heterogeneous MTZ hydration, which we then compare to IPV eruption locations from the past 200 Myr. We detect a statistically significant correlation between wet MTZ regions and IPV locations at the surface, but only if slabs sink faster than 1 cm/yr, water remains stored in the MTZ for periods of 30–100 Myr, and IPV eruptions occur 10–30 Myr later. We find that 42%–68% of continental IPV is underlain by wet MTZ, with greater fractions associated with longer MTZ residence time. Hydrous underpinning of continental IPV was highest during the Jurassic, when more extensive slab interaction with the MTZ hydrated a wider area of the MTZ. Since the Cretaceous, continents have been moving over the wet MTZ, increasing IPV possibilities. MTZ regions near the northern Pacific, southern Africa, and western Europe have remained dry by avoiding wet slabs. We suggest that subducted water shapes global patterns of intraplate volcanism, with hydrous upwellings rising from the MTZ to generate continental IPV above wet MTZ regions.
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
A Dynamic HPC Probabilistic Tsunami Forecast Workflow for Real-time Hazard Assessment
The Urgent Tsunami Computing procedures discussed herein are designed to quantify potential hazards resulting from seismically-induced tsunamis following an earthquake, with a temporal scope ranging from minutes to a few hours. The presented workflow employs comprehensive simulations, encompassing the entire tsunami propagation process, while accounting for uncertainties associated with source parameters, tsunamigenesis and wave propagation dynamics. Within the EuroHPC eFlows4HPC project, we present a High-Performance Computing (HPC) workflow tailored for urgent tsunami computation in which the Probabilistic Tsunami Forecast (PTF) code has been restructured and adapted for seamless integration into a PyCOMPSs framework. This framework enables parallel execution of tasks and includes simulations from Tsunami-HySEA numerical model within a unified computational environment. Of particular significance is the workflow's capability to incorporate new datasets, such as focal mechanism data, seismic records, or real-time tsunami observations. This functionality facilitates an "on-the-fly" update of the PTF, ensuring that the forecasting model remains responsive to the latest information. The development of this workflow involves a systematic exploration of diverse scenarios, realistic simulations, and the assimilation of incoming data. The overarching goal is to rigorously diminish uncertainties, thereby producing updated probabilistic forecasts without compromising precision and enhancing risk mitigation efforts far from the seismic source. Improved risk management, achieved by informing decision-making in emergency situations, underscores the importance of this development. We will showcase the technical advancements undertaken to tailor the workflow for HPC environments, spanning from the developers' perspective to that of the end user. Additionally, we will highlight the scientific enhancements implemented to leverage the full potential of HPC capabilities, aiming to significantly reduce result delivery times while concurrently enhancing the accuracy and precision of our forecasts.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessA Dynamic HPC Probabilistic Tsunami Forecast Workflow for Real-time Hazard Assessment
The Urgent Tsunami Computing procedures discussed herein are designed to quantify potential hazards resulting from seismically-induced tsunamis following an earthquake, with a temporal scope ranging from minutes to a few hours. The presented workflow employs comprehensive simulations, encompassing the entire tsunami propagation process, while accounting for uncertainties associated with source parameters, tsunamigenesis and wave propagation dynamics. Within the EuroHPC eFlows4HPC project, we present a High-Performance Computing (HPC) workflow tailored for urgent tsunami computation in which the Probabilistic Tsunami Forecast (PTF) code has been restructured and adapted for seamless integration into a PyCOMPSs framework. This framework enables parallel execution of tasks and includes simulations from Tsunami-HySEA numerical model within a unified computational environment. Of particular significance is the workflow's capability to incorporate new datasets, such as focal mechanism data, seismic records, or real-time tsunami observations. This functionality facilitates an "on-the-fly" update of the PTF, ensuring that the forecasting model remains responsive to the latest information. The development of this workflow involves a systematic exploration of diverse scenarios, realistic simulations, and the assimilation of incoming data. The overarching goal is to rigorously diminish uncertainties, thereby producing updated probabilistic forecasts without compromising precision and enhancing risk mitigation efforts far from the seismic source. Improved risk management, achieved by informing decision-making in emergency situations, underscores the importance of this development. We will showcase the technical advancements undertaken to tailor the workflow for HPC environments, spanning from the developers' perspective to that of the end user. Additionally, we will highlight the scientific enhancements implemented to leverage the full potential of HPC capabilities, aiming to significantly reduce result delivery times while concurrently enhancing the accuracy and precision of our forecasts.
- Open Access
- 20 January 2025
- Published by Copernicus GmbH
Uncertainty based amplification factors benchmarked by large ensembles of inundation simulations
Tsunami maximum inundation heights (MIH) are normally estimated by modelling the hydrodynamics of the overland flow over a limited area using high resolution long wave models. However, in situations where there one need to simulate inundation over large areas, we demand methods that are less resource intensive. The method of amplification factors represents one such simplified and resource efficient method. It is based on precomputed factors using linear shallow water models modelled over bathymetric transects to estimate the ratio of the offshore maximum surface elevation to that at the shoreline. It has been shown previously that we can get a relatively good estimate of the median value of the MIH at given coastline location using amplification factors. Yet, as this method is associated with considerably higher epistemic uncertainty than e.g. solving the nonlinear shallow water equations (due to the additional simplifications), there is need to include a measure of its uncertainty and bias. The main sources of uncertainty include among others the misfit of the method (towards data or more accurate methods) and local spatial variability of the inundation. Here, we present results from recent advancements of the amplification factor method emphasising a much more elaborate uncertainty analysis than employed previously. To this end, we use a unique set of synthetic data from several hundreds of thousands of massive scale nonlinear shallow simulations as a benchmark for estimating uncertainty concerned the amplification factors. The simulation dataset comprises ensembles of about 50 000 scenarios each for six different sites and includes sensitivity studies related to the Manning friction. When analysing the entire ensembles, we have revised the previous mathematical model for the uncertainty treatment (Glimsdal et al., 2019, PAGEOPH) by normalising the ensemble scenario outputs with the median MIH from each simulation. We further measure the bias of the amplification factor method by measuring its offset towards median MIH output from the shallow water simulations. The model provides input to the probabilistic tsunami hazard (PTHA) map for Italy, and we present related results of the amplification factor uncertainty analysis here. We will also discuss further and advocate the use of the method, both for long term hazard (PTHA) and for rapid post assessment following an event, e.g. through the ARISTOTLE framework. This work is supported by the European Union’s Horizon Europe Research and Innovation Program under grant agreement No 101058129 (DT-GEO, https://dtgeo.eu/). Computational resources were made available through PRACE grant number 2020225386, TsuHazAP.
- 20 January 2025
- Published by Copernicus GmbH
Open AccessUncertainty based amplification factors benchmarked by large ensembles of inundation simulations
Tsunami maximum inundation heights (MIH) are normally estimated by modelling the hydrodynamics of the overland flow over a limited area using high resolution long wave models. However, in situations where there one need to simulate inundation over large areas, we demand methods that are less resource intensive. The method of amplification factors represents one such simplified and resource efficient method. It is based on precomputed factors using linear shallow water models modelled over bathymetric transects to estimate the ratio of the offshore maximum surface elevation to that at the shoreline. It has been shown previously that we can get a relatively good estimate of the median value of the MIH at given coastline location using amplification factors. Yet, as this method is associated with considerably higher epistemic uncertainty than e.g. solving the nonlinear shallow water equations (due to the additional simplifications), there is need to include a measure of its uncertainty and bias. The main sources of uncertainty include among others the misfit of the method (towards data or more accurate methods) and local spatial variability of the inundation. Here, we present results from recent advancements of the amplification factor method emphasising a much more elaborate uncertainty analysis than employed previously. To this end, we use a unique set of synthetic data from several hundreds of thousands of massive scale nonlinear shallow simulations as a benchmark for estimating uncertainty concerned the amplification factors. The simulation dataset comprises ensembles of about 50 000 scenarios each for six different sites and includes sensitivity studies related to the Manning friction. When analysing the entire ensembles, we have revised the previous mathematical model for the uncertainty treatment (Glimsdal et al., 2019, PAGEOPH) by normalising the ensemble scenario outputs with the median MIH from each simulation. We further measure the bias of the amplification factor method by measuring its offset towards median MIH output from the shallow water simulations. The model provides input to the probabilistic tsunami hazard (PTHA) map for Italy, and we present related results of the amplification factor uncertainty analysis here. We will also discuss further and advocate the use of the method, both for long term hazard (PTHA) and for rapid post assessment following an event, e.g. through the ARISTOTLE framework. This work is supported by the European Union’s Horizon Europe Research and Innovation Program under grant agreement No 101058129 (DT-GEO, https://dtgeo.eu/). Computational resources were made available through PRACE grant number 2020225386, TsuHazAP.