Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 36
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

A machine learning estimator trained on synthetic data for real-time earthquake ground-shaking predictions in Southern California

  • Monterrubio-Velasco, Marisol
  • Callaghan, Scott
  • Modesto, David
  • Carrasco, Jose Carlos
  • Badía, Rosa M.
  • Vázquez-Novoa, Fernando
  • Pienkowska, Marta
  • de la Puente, Josep
  • Pallarés-Font de Mora, Pablo|||0000-0002-0120-3251
  • Quintana-Ortí, Enrique S.|||0000-0002-5454-165X
[EN] After large-magnitude earthquakes, a crucial task for impact assessment is to rapidly and accurately estimate the ground shaking in the affected region. To satisfy real-time constraints, intensity measures are traditionally evaluated with empirical Ground Motion Models that can drastically limit the accuracy of the estimated values. As an alternative, here we present Machine Learning strategies trained on physics-based simulations that require similar evaluation times. We trained and validated the proposed Machine Learning-based Estimator for ground shaking maps with one of the largest existing datasets (<100M simulated seismograms) from CyberShake developed by the Southern California Earthquake Center covering the Los Angeles basin. For a well-tailored synthetic database, our predictions outperform empirical Ground Motion Models provided that the events considered are compatible with the training data. Using the proposed strategy we show significant error reductions not only for synthetic, but also for five real historical earthquakes, relative to empirical Ground Motion Models., This work has been funded by the European Commission's Horizon 2020 Framework program and the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 955558 and by MCIN/AEI/10.13039/501100011033 and the European Union NextGenerationEU/PRTR (PCI2021-121957), project eFlows4HPC. This research has been supported by the European High-Performance Computing Joint Undertaking (JU) as well as Spain, Italy, Iceland, Germany, Norway, France, Finland, and Croatia under grant agreement no. 101093038, (ChEESE-CoE) The authors acknowledge the Center for Advanced Research Computing (CARC) at the University of Southern California for providing computing resources that have contributed to the research results reported within this publication. URL: https://carc.usc.edu. This research used resources of the Oak Ridge Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC05-00OR22725, and the National Center for Supercomputer Applications under the Blue Waters Sustained-Petascale Computing Project. SCEC is funded by USGS Cooperative Agreement G17AC00047 and NSF Cooperative Agreement EAR-1600087. The authors also thank Dr. Kevin Milner, who provided the ASK-14 estimations needed for the comparisons presented in this paper. The authors thank Dr. Arnau Folch who reviewed the draft of the paper, improving it with his comments and suggestions.
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

11th EGU Galileo Conference: Solid Earth and Geohazards in the Exascale Era Consensual Document

  • Folch, Arnau
  • Bhihe, Cedric
  • Caviedes-Vouillième, Daniel
  • de la Puente, Josep
  • Esposti Ongaro, Tomaso
  • Garg, Deepak
  • Gibbons, Steven J.
  • Kaus, Boris
  • Monterrubio, Marisol
  • Räss, Ludovic
  • Reis, Claudia
  • Scaini, Chiara
  • Srivastava, Nishtha
  • Vilarrasa, Víctor
  • Zwinger, Thomas
The 11th Galileo Conference in Barcelona (May 23-26, 2023) addressed Exascale computing challenges in geosciences. With 78 participants from 15 countries, it focused on European-based research but welcomed contributions from worldwide institutions. The conference had four sessions covering HPC applications, data workflows, computational geosciences, and EuroHPC infrastructures. It featured keynote presentations, poster sessions, and breakout sessions, including Master Classes for 22 Early Career Scientists supported by EGU. This document represents the consensus among participants, capturing outcomes from breakout sessions and acknowledging diverse opinions and approaches., The 11th Galileo Conference of the European Geosciences Union (EGU) focused on "Solid Earth and Geohazards in the Exascale Era." This abstract presents the main outcomes and conclusions from the conference breakout sessions, which aimed to provide recommendations for the future of solid earth research. The discussions highlighted the challenges and opportunities associated with high-performance computing (HPC) in solid earth sciences. The key findings include the need for collaboration between computer scientists and solid earth domain-specific scientists, the importance of portability software layers for different hardware architectures, the adoption of programming models for easier development and deployment of applications, the necessity of HPC training at all career stages, the improvement of accessibility and authentication mechanisms for European machines, and the readiness of urgent computing services for natural catastrophes. The conference also emphasized the significance of sustainable funding, software engineering best practices, and the development of modular and interoperable codes and workflows. Overall, the conference provided insights into the current status of computational solid earth research and offered recommendations for future advancements in the field., European Geosciences Union (EGU), the EuroHPC Center of Excellence for Exascale in Solid Earth (ChEESE) under Grant Agreement No 101093038 (https://cheese2.eu), and the European Union's Next Generation/PRTR Program through grant PCI2022-134973-2., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Reconstructing tephra fall deposits via ensemble-based data assimilation techniques

  • Mingari, Leonardo
  • Costa, A.
  • Macedonio, Giovanni
  • Folch, Arnau
In recent years, there has been a growing interest in ensemble approaches for modelling the atmospheric transport of volcanic aerosol, ash, and lapilli (tephra). The development of such techniques enables the exploration of novel methods for incorporating real observations into tephra dispersal models. However, traditional data assimilation algorithms, including ensemble Kalman filter (EnKF) methods, can yield suboptimal state estimates for positive-definite variables such as those related to volcanic aerosols and tephra deposits. This study proposes two new ensemble-based data assimilation techniques for semi-positive-definite variables with highly skewed uncertainty distributions, including aerosol concentrations and tephra deposit mass loading: the Gaussian with non-negative constraints (GNC) and gamma inverse-gamma (GIG) methods. The proposed methods are applied to reconstruct the tephra fallout deposit resulting from the 2015 Calbuco eruption using an ensemble of 256 runs performed with the FALL3D dispersal model. An assessment of the methodologies is conducted considering two independent datasets of deposit thickness measurements: an assimilation dataset and a validation dataset. Different evaluation metrics (e.g. RMSE, MBE, and SMAPE) are computed for the validation dataset, and the results are compared to two references: the ensemble prior mean and the EnKF analysis. Results show that the assimilation leads to a significant improvement over the first-guess results obtained from the simple ensemble forecast. The evidence from this study suggests that the GNC method was the most skilful approach and represents a promising alternative for assimilation of volcanic fallout data. The spatial distributions of the tephra fallout deposit thickness and volume according to the GNC analysis are in good agreement with estimations based on field measurements and isopach maps reported in previous studies. On the other hand, although it is an interesting approach, the GIG method failed to improve the EnKF analysis., This work has been partially funded by the H2020 Center of Excellence for Exascale in Solid Earth (ChEESE) (grant no. 823844) and by the European Union's Horizon Europe Research and Innovation Programme (DT-GEO (grant no. 101058129)). The research leading to these results has received funding from EuroHPC (ChEESE-2P (grant no. 101093038))., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Reconstructing tephra fall deposits via ensemble-based data assimilation techniques

  • Mingari, Leonardo
  • Costa, A.
  • Macedonio, Giovanni
  • Folch, Arnau
In recent years, there has been a growing interest in ensemble approaches for modelling the atmospheric transport of volcanic aerosol, ash, and lapilli (tephra). The development of such techniques enables the exploration of novel methods for incorporating real observations into tephra dispersal models. However, traditional data assimilation algorithms, including ensemble Kalman filter (EnKF) methods, can yield suboptimal state estimates for positive-definite variables such as those related to volcanic aerosols and tephra deposits. This study proposes two new ensemble-based data assimilation techniques for semi-positive-definite variables with highly skewed uncertainty distributions, including aerosol concentrations and tephra deposit mass loading: the Gaussian with non-negative constraints (GNC) and gamma inverse-gamma (GIG) methods. The proposed methods are applied to reconstruct the tephra fallout deposit resulting from the 2015 Calbuco eruption using an ensemble of 256 runs performed with the FALL3D dispersal model. An assessment of the methodologies is conducted considering two independent datasets of deposit thickness measurements: an assimilation dataset and a validation dataset. Different evaluation metrics (e.g. RMSE, MBE, and SMAPE) are computed for the validation dataset, and the results are compared to two references: the ensemble prior mean and the EnKF analysis. Results show that the assimilation leads to a significant improvement over the first-guess results obtained from the simple ensemble forecast. The evidence from this study suggests that the GNC method was the most skilful approach and represents a promising alternative for assimilation of volcanic fallout data. The spatial distributions of the tephra fallout deposit thickness and volume according to the GNC analysis are in good agreement with estimations based on field measurements and isopach maps reported in previous studies. On the other hand, although it is an interesting approach, the GIG method failed to improve the EnKF analysis., This work has been partially funded by the H2020 Center of Excellence for Exascale in Solid Earth (ChEESE) (grant no. 823844) and by the European Union's Horizon Europe Research and Innovation Programme (DT-GEO (grant no. 101058129)). The research leading to these results has received funding from EuroHPC (ChEESE-2P (grant no. 101093038))., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D1.1 Project Management and Quality Guidelines

  • García Miraz, María José
  • Folch, Arnau
  • Beltrán Torres, Barbara
  • Montellier, Vadim
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This document lays out the Project Management and Quality Guidelines for ChEESE- 2P project. The guidelines aim to describe the processes that will ensure the desired quality of all tasks and deliverables are achieved. It describes the project´s structure and governing bodies, the tools to promote a successful development and the mechanisms to avoid potential risks. Thus, this is a report that supports the demonstration deliverable D1.1., Project funded by EuroHPC under the grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D1.2 First Data Management Plan (DMP)

  • Folch, Arnau
  • Esposti-Ongaro, Tomaso
  • Barsotti, Sara
  • Vinciarelli, Valerio
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., Project funded by EuroHPC under the grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D2.1 First code audit report

  • Gracia, Jose
  • Lanucara, Piero
  • Bader, Michael
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., Project funded by EuroHPC under the grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D3.1 Selected set of mini-apps and its requirements

  • Masini, A.
  • Raffin, Erwan
  • Hamitou, Okba
  • García, M.
  • Mantovani, F.
  • Amati, Giorgio
  • Pascual, H.
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This document provides information related to ChEESE-2P WP3, including outlines of other projects involved in the co-design phase and the selected set of mini-applications that will be used, together with their software and platform requirements., Project funded by EuroHPC under the grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D7.1 Project Website

  • Vedia, Varvara
  • Gregorio, Rose
  • Zamora, Natalia
  • Filosa, Silvia
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., The purpose of this document is to elaborate on the design decisions made for the ChEESE official website, which has been established to provide information and increase accessibility to both partners and the general audience. This report serves as a supplement to the demonstration deliverable D7.1, Project funded by EuroHPC under the grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D7.2 Communication and Dissemination Plan

  • Vedia, Varvara
  • Gregorio, Rose
  • Pourhiet, Laetitia Le
  • Denamiel, Clea
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This Communication and Dissemination Plan (D7.2) outlines the strategy and implementation plan for CHEESE's communication and dissemination efforts. This document will serve as a framework for organising these activities and assessing their impact on the project., Project funded by EuroHPC under the grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D7.3 Collaboration plan with complementary projects

  • García Miraz, María José
  • Folch, Arnau
  • Beltrán Torres, Barbara
  • Lanucara, Piero
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This deliverable outlines the collaboration with other Centres of Excellence for HPC applications, NCCs, industry as well other EU funded activities., Project funded by EuroHPC under the grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D7.4 Training Plan

  • Baghetti, María Paz
  • Vedia, Varvara
  • Folch, Arnau
  • Lanucara, Piero
  • García Miraz, María José
  • Kaus, Boris
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This document introduces the training plan of ChEESE-2P project. It outlines the key aspects of the training program, including the target audiences, topics, and expected learning outcomes. Additionally, the plan includes a tentative timetable for the delivery of the courses, ensuring a structured approach to their implementation., Project funded by EuroHPC under the grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Numerical simulations reveal the dynamics of the most intense eruption of Hunga Tonga in January 2022

  • Suzuki, Yujiro J.
  • Martínez Montesinos, Beatriz
  • Costa, Antonio
A violent undersea explosive eruption occurred at Hunga Tonga–Hunga Ha’apai volcano on 15 January 2022, generating an eruption cloud more intense than any previously observed. We performed numerical simulations of eruption cloud dynamics using a 3D fluid-dynamic model and an ensemble-based tephra dispersal inversion model to reconstruct the eruption’s climactic phase and compare it with available observations. Our results reveal that during this phase, 190–1500 Tg of seawater interacted with magma, producing a mass flow rate of the eruptive magmatic mixture of 3.2–6.3 × 10⁹ kg s−1, which is several times more intense than the 1991 Pinatubo eruption. Moreover, we show that the eruption cloud, which injected approximately 1 Tg of volcanic ash and 0.1 Tg of seawater into the mesosphere, was in a state of thermal disequilibrium with the surrounding environment. The eruption injected 0.3–11 Tg sulfur dioxide into the atmosphere. These results suggest that a substantial amount of magmatic material, water vapor, and sulfur dioxide was injected into the stratosphere and mesosphere during this eruption, which could have a significant impact on the global climate several years after the eruption. Our work also shows the importance of high-resolution simulations in capturing the complex dynamics of eruption plumes generated by undersea volcanic eruptions, leading to more accurate predictions of eruption impacts., This study was supported by JSPS KAKENHI (21K21353), MEXT’s “Integrated Program for Next Generation Volcanic Research and Human Resource Development” (J005391), and Earthquake Research Institute, The University of Tokyo (JURP 2022-S-C104). AC and BM acknwoledge the Centre of Excellence for Exascale in SolidEarth (ChEESE; EuroHPC Grant Agreement No. 101093038) and the project “A Digital Twin for GEO-physical extremes” (DT-GEO; Horizon Europe Grant Agreement No.101058129), Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Eruption Source Parameters in Volcanic Plume Modeling: Advances, Challenges, and Future Directions

  • Costa, A.
  • Folch, Arnau
  • Suzuki, Yujiro J.
  • Mingari, Leonardo
  • Macedonio, Giovanni
Accurately predicting the atmospheric dispersion of volcanic ash and gases is crucial for both scientific understanding and hazard mitigation. Estimating Eruption Source Parameters (ESP), such as mass eruption rate, plume height, duration, and particle size distribution and properties, remains challenging due to the complex nature of volcanic processes and measurement uncertainties. This review examines recent advancements in assessing ESPs from the perspective of modeling, evaluating current limitations and potential future improvements. We overview how current models quantify crucial ESPs, either directly from observations or indirectly from proxies or modeling strategies, and address the inherent uncertainties. We identify critical areas needing further research and emphasize the importance of developing more robust and widely applicable methodologies. Finally, we propose innovative strategies to enhance ESPs estimations, ultimately improving volcanic hazard assessments and our understanding of eruption dynamics and their atmospheric interactions., This work has been partially funded by the Centre of Excellence for Exascale in Solid Earth - Second Phase (ChEESE-2P; EuroHPC Grant Agreement No. 101093038) and the project “A Digital Twin for GEO-physical extremes” (DT-GEO; Horizon Europe Grant Agreement No. 101058129). A.C., G.M. and L.M. thank the Istituto Nazionale di Geofisica e Vulcanologia, Italy, grant Progetto INGV Pianeta Dinamico (code CUP D53J19000170001) funded by Italian Ministry MIUR (“Fondo Finalizzato al rilancio degli investimenti delle amministrazioni centrali dello Stato e allo sviluppo del Paese”, legge 145/2018). Y.J.S. was supported by MEXT's “Integrated Program for Next Generation Volcanic Research and Human Resource Development” (J005391). A.C. acknowledges the UK NERC-funded projects Radar-supported Next-Generation Forecasting of Volcanic Ash Hazard (R4AsH) and Volcanic Plumes: Science to Services (V-PLUS)., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D5.3 Demonstrator of PD4

  • Gibbons, Steven J.
  • Bader, Michael
  • Denamiel, Clea
  • Castro Díaz, Manuel J.
  • Gabriel, Alice-Agnes
  • González del Pino, Alejandro
  • Lorito, Stefano
  • Løvholt, Finn
  • Macías Sánchez, Jorge
  • Romano, Fabrizio
  • Ulrich, Thomas
  • Wille, Mario
  • Monteiller, Vadim
  • Fournier, Alexandre
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This pilot demonstrator (PD) combines multiphysics processes from earthquakes, landslides, volcanoes, and atmospheric forcing to tsunami generation, propagation, and inundation in a joint framework for coupling different flagship codes and solving compound problems where several physical processes and phenomena occur together. This exploratory PD contains multiple workflows within the framework covering different sequences of events. Specifically, we consider three separate workflows: tsunami generation from complex earthquake rupture (coupling the flagship codes SeisSol and HySEA), tsunami generation from landslide (coupling ExaHyPE and HySEA), and global tsunami propagation and inundation driven by atmospheric forcing from a volcanic explosion (featuring Meteo-HySEA)., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D2.2 First report on improving code performance at exascale

  • Lanucara, Piero
  • Amati, Giorgio
  • Rose, Martin
  • Hamitou, Okba
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D4.1 Initial report on workflow setups and workflow auditing

  • Niethammer, Christoph
  • Song, Bin
  • Raffin, Erwan
  • Cordrie, Louise
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023-2026., This deliverable introduces the methodology to collect and present details for the simulation workflows in ChEESE-2P. We report the details for workflows around pilot demonstrators and derived simulation cases available up to the current point. Further, initial results from workflow analysis and optimization activities are presented., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D4.2 Plan for development and deployment tool usage for workflows and software

  • Niethammer, Christoph
  • Piero Lanucara
  • Pascual, Heribert
  • Folch, Arnau
  • Vedia, Varvara
  • Brogi, Federico
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023-2026., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable D4.3: Intermediate report on workflow setups and optimization

  • Niethammer, Christoph
  • Lanucara, Piero
  • Rosati, Mario
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023-2026., This deliverable gives an update on the status and work implementing the simulation workflows in ChEESE-2P. We report the details for workflows around pilot demonstrators and derived simulation cases available up to the current point. Further, initial results from workflow analysis and optimization activities are presented., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D5.2 Demonstrator of PD1

  • Bader, Michael
  • Monterrubio, Marisol
  • Monteiller, Vadim
  • Halldorsson, Benedikt
  • Gabriel, Alice-Agnes
  • Bhihe, Cedric
  • Blanco Prieto, Rut
  • Cubuk Sabuncu, Yesim
  • Schliwa, Nico
  • Ulrich, Thomas
  • Abril, Claudia
  • Jourdon, Anthony
  • Kurapati, Vikas
  • Schneller, David
  • Kaus, Boris
  • Denamiel, Clea
  • Kaus, Boris
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., In this report we report on prototypes and test cases for the Pilot Demonstrator PD1 in the ChEESE-2P centre of excellence, which is concerned with extreme-scale mod-elling of seismic hazards. We provide 5 prototypes considering seismic hazards from underwater explosion, seismic cycle simulations to better assess long-term seismic risk, augmented earthquake (dynamic rupture) simulations coupling to geodynamic simulation or tsunami generation, and augmenting probabilistic seismic hazard as-sessment by AI/machine learning approaches or reduced order models. Test cases demonstrate preparedness of the workflows for executing extreme-scale simulation cases, and cover several areas of high seismic risk in Europe, such as the South Iceland Seismic Zone, the Eastern Anatolian Fault Zone or the Hellenic Arc Subduc-tion Zone., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D5.4 Demonstrator of PD6 : Multiphase 3D volcanic explosion modeling

  • Michieli Vitturi, Mattia de
  • Esposti-Ongaro, Tomaso
  • Pardini, Federica
  • Neri, Augusto
  • Brogi, Federico
  • Zamora, Natalia
  • Mingari, Leonardo
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., The main objective of this PD is to demonstrate the capability of a workflow based on the OpenPDAC flagship code to develop multiscale simulation (from sub-meter scale for the initial explosion phase to kilometre scale for the flow and ballistic impacted area) of phreatic eruption processes, accounting for complex, 3D multiphase flow interactions occurring at the microscopic to macroscopic levels. The modular workflow is ready to exploit exascale supercomputers for Urgent Computing (SC6.1) and Probabilistic Volcanic Hazard Assessment (SC6.2). Pilot application and first results to the cases of Vulcano and Campi Flegrei (Italy) are presented, and the potential scientific and societal impact of the results highlighted., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D5.6 PD8 Geodynamics to geohazards

  • Ibragimov, Iskander
  • Kaus, Boris
  • Jourdon, Anthony
  • Pourhiet, Laetitia Le
  • Bader, Michael
  • Morales, Tomas
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., We report on prototypes and test cases for Pilot Demonstrator PD8 in the ChEESE-2P centre of excellence which focuses on simulating processes that operate on geodynamic timescales and their links to a seismic cycle timescale. We provide 2 prototypes, the first of which addresses the ongoing motion of the eastern flank of the Etna volcano. The new workflow shows how to generate model setups that take a range of datasets into account. The second prototype demonstrates a new workflow that links discrete and smooth faults from geodynamic simulations that can serve as an input for seismic rupture models. Test cases demonstrate the ability of the workflows to simulate the Etna flank instability and to perform seismic rupture simulations on strike slip faults generated in geodynamic simulations., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D5.5 The Earth’s dynamo model

  • Fournier, Alexandre
  • Frezat, Hugo
  • Schaeffer, Nathanaël
  • Moguilny, Geneviève
  • Costa, Antonio
  • Ulrich, Thomas
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This deliverable presents advances in modeling Earth’s core dynamo through neural network (NN)-based subgrid-scale (SGS) closure models, aiming to reduce computational costs versus direct numerical simulations (DNS). Prototype 1 demonstrates NN-SGS modeling for quasi-geostrophic (QG) turbulence in an annular domain, employing spectral methods (Fourier-Chebyshev) and online learning —a training approach where NNs dynamically adapt to SGS terms during simulation runtime via adjoint optimization, ensuring stability and accuracy. The framework uses Galerkin coarse-graining for boundary preservation and and leverages JAX, a Python library for high-performance machine learning and automatic differentiation, to enable differentiable programming and efficient optimization, achieving a 5% error in Reynolds number compared to DNS. The NN-SGS model accurately captures zonal energy ratios, enstrophy, and dissipation scales, enabling a 6 times runtime reduction (30.5 vs. 181 minutes) on NVIDIA A100 GPUs. Prototype 2 introduces a JAX implementation of the xshells code for rotating spherical shell convection, verified against benchmarks, though current differentiable spherical harmonic transforms remain 4 times slower than the high-performance SHTns library used in the flagship xshells code. Benchmark results for QG turbulence and thermal convection confirm the viability of NN-SGS and JAX-xshells, with kinetic energy convergence matching references. Future work targets SGS extensions to magnetoconvection and dynamo regimes, alongside GPU optimization for SHTns. Public repositories ensure reproducibility, though 40GB VRAM GPUs are required for the QG turbulence demonstrator. The study highlights a path toward tractable geodynamo simulations by combining spectral methods, automatic differentiation, and data-driven SGS closure models, despite current limitations in JAX-SHTns performance., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D6.1. Simulation Cases and Trials definition

  • Esposti-Ongaro, Tomaso
  • Pardini, Federica
  • Monteiller, Laura Sandri
  • Monterrubio-Velasco, Marisol
  • Vadim Monteiller, Alice Gabriel
  • Monterrubio-Velasco, Marisol
  • Casarotti, Emanuele
  • Lorito, Stefano
  • Romano, Fabrizio
  • Løvolt, Finn
  • Gibbons, Steven J.
  • Sandri, Laura
  • Folch, Arnau
  • Michieli Vitturi, Mattia de
  • Fournier, Alexandre
  • Pourhiet, Laetitia Le
  • Kaus, Boris
  • Jóhannesson, Tómas
  • Zwinger, Thomas
  • Mangeney, Anne
  • Folch, Arnau
  • Bhihe, Cedric
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This deliverable defines the background, methodology, setup and exascale ambition of the Pilot Demonstrators’ production runs (Simulation Cases), including the collaborative effort to secure computational resources on the EuroHPC systems through regular access and extreme scale access calls. We also report about the expected scientific, technological and societal impact of the Simulation Cases. Finally, we define the methodology and framework for running “Trials”, i.e. the execution of Simulation Cases on Operational Environments, for those Pilot Demonstrators aiming to reach the highest Technology Readiness Levels., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D7.5 Preliminary Exploitation Plan

  • Guimarães, Aerton
  • Farnós, Joan
  • Song, Bin
  • Esposti-Ongaro, Tomaso
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., This Preliminary Exploitation Plan (D7.5) details the strategies for exploiting ChEESE-2P project outcomes, including a market analysis to identify key end users and early adopters. It also outlines plans for collaboration with standardisation bodies and aims to drive adoption across science, public administration, and industry sectors., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038, Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Otros (Other)

ChEESE-2P Deliverable: D5.1 New Version Release of Flagship Codes

  • Bader, Michael
  • Monteiller, Vadim
  • Pascual, Heribert
  • Michieli Vitturi, Mattia de
  • ChEESE-2P
Copyright Members of ChEESE collaboration, 2023/2026., Funded by the European Union. This work has received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Uncertainties, complexities and possible forecasting of Volcán de Colima energy emissions (Mexico, years 2013–2015) based on a fractal reconstruction theorem

  • Monterrubio Velasco, Marisol|||0000-0003-0790-1832
  • Lana Pons, Francisco Javier|||0000-0002-3298-9234
  • Arámbula Mendoza, Raúl
The effusive–explosive energy emission process in a volcano is a dynamic and complex physical phenomenon. The importance of quantifying this complexity in terms of the physical and mathematical mechanisms that govern these emissions should be a requirement for deciding to apply a possible forecasting strategy with a sufficient degree of certainty. The complexity of this process is determined in this research by means of the reconstruction theorem and statistical procedures applied to the effusive–explosive volcanic energy emissions corresponding to the activity in the Volcán de Colima (western segment of the Trans-Mexican Volcanic Belt) along the years 2013–2015. The analysis is focused on measuring the degree of persistence or randomness of the series, the degree of predictability of energy emissions, and the quantification of the degree of complexity and “memory loss” of the physical mechanism throughout an episode of volcanic emissions. The results indicate that the analysed time series depict a high degree of persistence and low memory loss, making the mentioned effusive–explosive volcanic emission structure a candidate for successfully applying a forecasting strategy., This research has been supported by the European High-Performance Computing Joint Undertaking (JU) as well as Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement no.101093038, (ChEESE-CoE)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

A machine learning estimator trained on synthetic data for real-time earthquake ground-shaking predictions in Southern California

  • Monterrubio Velasco, Marisol|||0000-0003-0790-1832
  • Callaghan, Scott
  • Modesto Galende, David|||0000-0001-9540-8815
  • Carrasco Jiménez, José Carlos
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
  • Pallarés Font de Mora, Pablo
  • Vázquez Novoa, Fernando
  • Quintana Ortí, Enrique Salvador
  • Pienkowska, Marta
  • de la Puente, Josep
After large-magnitude earthquakes, a crucial task for impact assessment is to rapidly and accurately estimate the ground shaking in the affected region. To satisfy real-time constraints, intensity measures are traditionally evaluated with empirical Ground Motion Models that can drastically limit the accuracy of the estimated values. As an alternative, here we present Machine Learning strategies trained on physics-based simulations that require similar evaluation times. We trained and validated the proposed Machine Learning-based Estimator for ground shaking maps with one of the largest existing datasets (<100M simulated seismograms) from CyberShake developed by the Southern California Earthquake Center covering the Los Angeles basin. For a well-tailored synthetic database, our predictions outperform empirical Ground Motion Models provided that the events considered are compatible with the training data. Using the proposed strategy we show significant error reductions not only for synthetic, but also for five real historical earthquakes, relative to empirical Ground Motion Models., This work has been funded by the European Commission’s Horizon 2020 Framework program and the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 955558 and by MCIN/AEI/ 10.13039/501100011033 and the European Union NextGeneration EU/PRTR(PCI2021-121957), project eFlows4HPC. This research has been supported by the European High-Performance Computing Joint Undertaking (JU) as well as Spain, Italy, Iceland, Germany, Norway, France, Finland, and Croatia under grant agreement no. 101093038, (ChEESE-CoE). The authors acknowledge the Center for Advanced Research Computing (CARC) at the University of Southern California for providing computing resources that have contributed to the research results reported within this publication. URL: https://carc.usc.edu. This research used resources of the Oak Ridge Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC05-00OR22725, and the National Center for Supercomputer Applications under the Blue Waters Sustained-Petascale Computing Project. SCEC is funded by USGS Cooperative Agreement G17AC00047 and NSF Cooperative Agreement EAR1600087., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Multifractal structure and Gutenberg–Richter parameter associated with volcanic emissions of high energy in Colima, Mexico (years 2013–2015)

  • Monterrubio Velasco, Marisol|||0000-0003-0790-1832
  • Lana Pons, Francisco Javier|||0000-0002-3298-9234
  • Arámbula Mendoza, Raúl
The evolution of multifractal structures in various physical processes, such as climatology, seismology, or volcanology, serves as a crucial tool for detecting changes in corresponding phenomena. In this study, weexplore the evolution of the multifractal structure of volcanic emissions with varying energy levels (observed at Colima, Mexico, during the years 2013–2015) to identify clear indicators of imminent high-energy emissions nearing 8.0×108 J. These indicators manifest through the evolution of six multifractal parameters: the central Hölder exponent (α0); the maximum and minimum Hölder exponents (αmax, αmin); the multifractal amplitude (W =αmax−αmin); the multifractal asymmetry (γ = [αmax −α0]/[α0 −αmin]); and the complexity index (CI), calculated as the sum of the normalized values of α0, W, and γ. Additionally, the results obtained from adapting the Gutenberg–Richter seismic law to volcanic energy emissions, along with the corresponding skewness and standard deviation of the volcanic emission data, further support the findings obtained through multifractal analysis. These results, derived from multifractal structure analysis, adaptation of the Gutenberg–Richter law to volcanic emissions, and basic statistical parameters, hold significant relevance in anticipating potential volcanic episodes of high energy. Such anticipation can be further quantified using an appropriate forecasting algorithm., This work has been supported by the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland, and Croatia under grant agreement no. 101093038 (ChEESE-2P)., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo preliminar (WorkingPaper)

11th EGU Galileo Conference: Solid Earth and Geohazards in the Exascale Era Consensual Document

  • Folch, Arnau|||0000-0002-0677-6366
  • Bhihe, Cedric
  • Caviedes-Vouillième, Daniel
  • de la Puente, Josep
  • Esposti Ongaro, Tomaso
  • Monterrubio, Marisol
The 11th Galileo Conference in Barcelona (May 23-26, 2023) addressed Exascale computing challenges in geosciences. With 78 participants from 15 countries, it focused on European-based research but welcomed contributions from worldwide institutions. The conference had four sessions covering HPC applications, data workflows, computational geosciences, and EuroHPC infrastructures. It featured keynote presentations, poster sessions, and breakout sessions, including Master Classes for 22 Early Career Scientists supported by EGU. This document represents the consensus among participants, capturing outcomes from breakout sessions and acknowledging diverse opinions and approaches., The 11th Galileo Conference of the European Geosciences Union (EGU) focused on "Solid Earth and Geohazards in the Exascale Era." This abstract presents the main outcomes and conclusions from the conference breakout sessions, which aimed to provide recommendations for the future of solid earth research. The discussions highlighted the challenges and opportunities associated with high-performance computing (HPC) in solid earth sciences. The key findings include the need for collaboration between computer scientists and solid earth domain-specific scientists, the importance of portability software layers for different hardware architectures, the adoption of programming models for easier development and deployment of applications, the necessity of HPC training at all career stages, the improvement of accessibility and authentication mechanisms for European machines, and the readiness of urgent computing services for natural catastrophes. The conference also emphasized the significance of sustainable funding, software engineering best practices, and the development of modular and interoperable codes and workflows. Overall, the conference provided insights into the current status of computational solid earth research and offered recommendations for future advancements in the field., European Geosciences Union (EGU), the EuroHPC Center of Excellence for Exascale in Solid Earth (ChEESE) under Grant Agreement No 101093038 (https://cheese2.eu), and the European Union's Next Generation/PRTR Program through grant PCI2022-134973-2., Peer Reviewed, "Article signat per 115 autors/es:" Folch, A., Bhihe, C., Caviedes-Vouillième, D., de la Puente, J., Esposti Ongaro, T., Garg, D, Gibbons, S. J., Kaus, B., Monterrubio, M., Räss, L., Reis, C., Scaini, C., Srivastava, N., Vilarrasa, V., Zwinger, T."
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Model generation architecture for digital twins and deployment system

  • Vázquez Novoa, Fernando
  • Conejero Bañón, Francisco Javier
  • Badia Sala, Rosa Maria|||0000-0003-2941-5499
A Digital Twin is a virtual copy or representation of a physical object. Given the same input, both should produce the same output. A bidirectional data exchange between both of them should keep the objects in the same state by automatically reflecting changes from one of the objects to the other. Despite the popularity of the topic, there is a lack of standardization and no widespread adoption of techniques or architectures, resulting in ad-hoc implementations for each use case. This paper proposes a methodology for the generation of digital twins based on the definition of their life cycle, with phases that include training, operation and retraining. To allow efficient management of resources, each phase of the Digital Twin is implemented as a PyCOMPSs workflow. The proposed methodologies and architectures aims to simplify and strengthen the design and development of digital twins. It also allows for the extension of existing ones, the replacement of their components with more advanced ones, and the support of their reuse. We also consider workflow deployment, which is often a difficult task requiring specialized personnel and time. This article focuses on presenting the training phase of a Digital Twin life-cycle and demonstrates its operation by applying it to two real-world use case scenarios in ongoing projects. The use of the proposal led to a simplification of the developed code, allowed a distribution of the computational work among the available resources in a user-agnostic manner and the digital models created faithfully represent the physical object. Additionally, the paper presents the design of the operating life-cycle phase of a Digital Twin., Author Fernando Vázquez is supported by PRE2022-104134 funded by MICIU/AEI /10.13039/501100011033 and by the FSE+. This work has been partially funded by the the projects PID2019- 107255GB, CEX2021-001148-S, and PID2023-147979NB-C21 from the MCIN/AEI and MICIU/AEI /10.13039/501100011033 and by FEDER, UE, and by the Departament de Recerca i Universitats de la Generalitat de Catalunya, research group MPiEDist (2021 SGR 00412), and by the European Commission through the program Horizon 2020 and the European High-Performance Computing Joint Undertaking (JU) with contract 955558 and by the MCIN/AEI/10.13039/501100011033 y European Union NextGenerationEU/PRTR (PCI2021-121957), project eFlows4HPC; as well as by the European Commission under program Horizon Action under contract 101056886 (CAELESTIS). This work has also received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Model generation architecture for digital twins and deployment system

  • Vázquez Novoa, Fernando
  • Conejero Bañón, Francisco Javier
  • Badia Sala, Rosa Maria
A Digital Twin is a virtual copy or representation of a physical object. Given the same input, both should produce the same output. A bidirectional data exchange between both of them should keep the objects in the same state by automatically reflecting changes from one of the objects to the other. Despite the popularity of the topic, there is a lack of standardization and no widespread adoption of techniques or architectures, resulting in ad-hoc implementations for each use case. This paper proposes a methodology for the generation of digital twins based on the definition of their life cycle, with phases that include training, operation and retraining. To allow efficient management of resources, each phase of the Digital Twin is implemented as a PyCOMPSs workflow. The proposed methodologies and architectures aims to simplify and strengthen the design and development of digital twins. It also allows for the extension of existing ones, the replacement of their components with more advanced ones, and the support of their reuse. We also consider workflow deployment, which is often a difficult task requiring specialized personnel and time. This article focuses on presenting the training phase of a Digital Twin life-cycle and demonstrates its operation by applying it to two real-world use case scenarios in ongoing projects. The use of the proposal led to a simplification of the developed code, allowed a distribution of the computational work among the available resources in a user-agnostic manner and the digital models created faithfully represent the physical object. Additionally, the paper presents the design of the operating life-cycle phase of a Digital Twin., Author Fernando Vázquez is supported by PRE2022-104134 funded by MICIU/AEI /10.13039/501100011033 and by the FSE+. This work has been partially funded by the the projects PID2019- 107255GB, CEX2021-001148-S, and PID2023-147979NB-C21 from the MCIN/AEI and MICIU/AEI /10.13039/501100011033 and by FEDER, UE, and by the Departament de Recerca i Universitats de la Generalitat de Catalunya, research group MPiEDist (2021 SGR 00412), and by the European Commission through the program Horizon 2020 and the European High-Performance Computing Joint Undertaking (JU) with contract 955558 and by the MCIN/AEI/10.13039/501100011033 y European Union NextGenerationEU/PRTR (PCI2021-121957), project eFlows4HPC; as well as by the European Commission under program Horizon Action under contract 101056886 (CAELESTIS). This work has also received funding from the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement No 101093038., Peer Reviewed, Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Uncertainties, complexities and possible forecasting of Volcán de Colima energy emissions (Mexico, years 2013–2015) based on a fractal reconstruction theorem

  • Monterrubio Velasco, Marisol
  • Lana Pons, Francisco Javier
  • Arámbula Mendoza, Raúl
The effusive–explosive energy emission process in a volcano is a dynamic and complex physical phenomenon. The importance of quantifying this complexity in terms of the physical and mathematical mechanisms that govern these emissions should be a requirement for deciding to apply a possible forecasting strategy with a sufficient degree of certainty. The complexity of this process is determined in this research by means of the reconstruction theorem and statistical procedures applied to the effusive–explosive volcanic energy emissions corresponding to the activity in the Volcán de Colima (western segment of the Trans-Mexican Volcanic Belt) along the years 2013–2015. The analysis is focused on measuring the degree of persistence or randomness of the series, the degree of predictability of energy emissions, and the quantification of the degree of complexity and “memory loss” of the physical mechanism throughout an episode of volcanic emissions. The results indicate that the analysed time series depict a high degree of persistence and low memory loss, making the mentioned effusive–explosive volcanic emission structure a candidate for successfully applying a forecasting strategy., This research has been supported by the European High-Performance Computing Joint Undertaking (JU) as well as Spain, Italy, Iceland, Germany, Norway, France, Finland and Croatia under grant agreement no.101093038, (ChEESE-CoE)., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo preliminar (WorkingPaper)

11th EGU Galileo Conference: Solid Earth and Geohazards in the Exascale Era Consensual Document

  • Folch, Arnau
  • Bhihe, Cedric
  • Caviedes-Vouillième, Daniel
  • de la Puente, Josep
  • Esposti Ongaro, Tomaso
  • Monterrubio, Marisol
The 11th Galileo Conference in Barcelona (May 23-26, 2023) addressed Exascale computing challenges in geosciences. With 78 participants from 15 countries, it focused on European-based research but welcomed contributions from worldwide institutions. The conference had four sessions covering HPC applications, data workflows, computational geosciences, and EuroHPC infrastructures. It featured keynote presentations, poster sessions, and breakout sessions, including Master Classes for 22 Early Career Scientists supported by EGU. This document represents the consensus among participants, capturing outcomes from breakout sessions and acknowledging diverse opinions and approaches., The 11th Galileo Conference of the European Geosciences Union (EGU) focused on "Solid Earth and Geohazards in the Exascale Era." This abstract presents the main outcomes and conclusions from the conference breakout sessions, which aimed to provide recommendations for the future of solid earth research. The discussions highlighted the challenges and opportunities associated with high-performance computing (HPC) in solid earth sciences. The key findings include the need for collaboration between computer scientists and solid earth domain-specific scientists, the importance of portability software layers for different hardware architectures, the adoption of programming models for easier development and deployment of applications, the necessity of HPC training at all career stages, the improvement of accessibility and authentication mechanisms for European machines, and the readiness of urgent computing services for natural catastrophes. The conference also emphasized the significance of sustainable funding, software engineering best practices, and the development of modular and interoperable codes and workflows. Overall, the conference provided insights into the current status of computational solid earth research and offered recommendations for future advancements in the field., European Geosciences Union (EGU), the EuroHPC Center of Excellence for Exascale in Solid Earth (ChEESE) under Grant Agreement No 101093038 (https://cheese2.eu), and the European Union's Next Generation/PRTR Program through grant PCI2022-134973-2., Peer Reviewed, "Article signat per 115 autors/es:" Folch, A., Bhihe, C., Caviedes-Vouillième, D., de la Puente, J., Esposti Ongaro, T., Garg, D, Gibbons, S. J., Kaus, B., Monterrubio, M., Räss, L., Reis, C., Scaini, C., Srivastava, N., Vilarrasa, V., Zwinger, T.", Postprint (author's final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

A machine learning estimator trained on synthetic data for real-time earthquake ground-shaking predictions in Southern California

  • Monterrubio Velasco, Marisol
  • Callaghan, Scott
  • Modesto Galende, David
  • Carrasco Jiménez, José Carlos
  • Badia Sala, Rosa Maria
  • Pallarés Font de Mora, Pablo
  • Vázquez Novoa, Fernando
  • Quintana Ortí, Enrique Salvador
  • Pienkowska, Marta
  • de la Puente, Josep
After large-magnitude earthquakes, a crucial task for impact assessment is to rapidly and accurately estimate the ground shaking in the affected region. To satisfy real-time constraints, intensity measures are traditionally evaluated with empirical Ground Motion Models that can drastically limit the accuracy of the estimated values. As an alternative, here we present Machine Learning strategies trained on physics-based simulations that require similar evaluation times. We trained and validated the proposed Machine Learning-based Estimator for ground shaking maps with one of the largest existing datasets (<100M simulated seismograms) from CyberShake developed by the Southern California Earthquake Center covering the Los Angeles basin. For a well-tailored synthetic database, our predictions outperform empirical Ground Motion Models provided that the events considered are compatible with the training data. Using the proposed strategy we show significant error reductions not only for synthetic, but also for five real historical earthquakes, relative to empirical Ground Motion Models., This work has been funded by the European Commission’s Horizon 2020 Framework program and the European High-Performance Computing Joint Undertaking (JU) under grant agreement No 955558 and by MCIN/AEI/ 10.13039/501100011033 and the European Union NextGeneration EU/PRTR(PCI2021-121957), project eFlows4HPC. This research has been supported by the European High-Performance Computing Joint Undertaking (JU) as well as Spain, Italy, Iceland, Germany, Norway, France, Finland, and Croatia under grant agreement no. 101093038, (ChEESE-CoE). The authors acknowledge the Center for Advanced Research Computing (CARC) at the University of Southern California for providing computing resources that have contributed to the research results reported within this publication. URL: https://carc.usc.edu. This research used resources of the Oak Ridge Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC05-00OR22725, and the National Center for Supercomputer Applications under the Blue Waters Sustained-Petascale Computing Project. SCEC is funded by USGS Cooperative Agreement G17AC00047 and NSF Cooperative Agreement EAR1600087., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Multifractal structure and Gutenberg–Richter parameter associated with volcanic emissions of high energy in Colima, Mexico (years 2013–2015)

  • Monterrubio Velasco, Marisol
  • Lana Pons, Francisco Javier
  • Arámbula Mendoza, Raúl
The evolution of multifractal structures in various physical processes, such as climatology, seismology, or volcanology, serves as a crucial tool for detecting changes in corresponding phenomena. In this study, weexplore the evolution of the multifractal structure of volcanic emissions with varying energy levels (observed at Colima, Mexico, during the years 2013–2015) to identify clear indicators of imminent high-energy emissions nearing 8.0×108 J. These indicators manifest through the evolution of six multifractal parameters: the central Hölder exponent (α0); the maximum and minimum Hölder exponents (αmax, αmin); the multifractal amplitude (W =αmax−αmin); the multifractal asymmetry (γ = [αmax −α0]/[α0 −αmin]); and the complexity index (CI), calculated as the sum of the normalized values of α0, W, and γ. Additionally, the results obtained from adapting the Gutenberg–Richter seismic law to volcanic energy emissions, along with the corresponding skewness and standard deviation of the volcanic emission data, further support the findings obtained through multifractal analysis. These results, derived from multifractal structure analysis, adaptation of the Gutenberg–Richter law to volcanic emissions, and basic statistical parameters, hold significant relevance in anticipating potential volcanic episodes of high energy. Such anticipation can be further quantified using an appropriate forecasting algorithm., This work has been supported by the European High Performance Computing Joint Undertaking (JU) and Spain, Italy, Iceland, Germany, Norway, France, Finland, and Croatia under grant agreement no. 101093038 (ChEESE-2P)., Peer Reviewed, Postprint (published version)