Enabling HPC Earthquake and Tsunami Simulation Data Mining - an NFDI4Earth Lighthouse Use Case Project
Wednesday, June 24, 2026 3:45 PM to 5:15 PM · 1 hr. 30 min. (Europe/Berlin)
Foyer D-G - 2nd Floor
Project Poster
Development of HPC SkillsDigital Twins and MLGeosciencesStorage Technologies and Architectures
Information
Poster is on display.
Earthquakes and tsunamis are among the most devastating natural hazards, posing unpredictable hazards to densely populated regions worldwide. Physics-based simulations of large devastating earthquakes (and their associated tsunami generation) using high-performance computing (HPC)
now allow for the first time a rapid understanding of earthquake processes, and their tsunamigenic
potential permitting unprecedented insight into the dynamics of fault systems, and their ground
shaking and inundation hazards. Pioneer simulation campaigns for hazard assessment, learning of
surrogate models or parameter inference create massive amounts of simulation data, and are
often produced and exploited for a single purpose only, thus the concrete problem and challenge
lies in the lack of standardized and accessible frameworks for handling, curating, reusing and mining
this valuable resource. Our objectives are 1) to (further) develop and promote international
standards for simulation (meta-)data formats for large-scale earthquake/tsunami simulations, 2) to
develop tools to access, combine and postprocess (visualize) simulation and observational datasets;
and 3) to provide derived, labelled products (e.g., shake maps, tsunami inundation maps) for
downstream AI data mining and use in scientific and civil protection use cases. The result will be
unified, FAIR simulation data formats and a set of integrated geo-referenced case studies (giant
subduction zone earthquakes, induced seismicity in cities) adding the perspective of a simulation
data lake to NFDI4Earth infrastructure, including the OneStop4All and KnowledgeHub platforms,
leveraging international collaborations of the PIs. Target users include Earth System scientists,
research software engineers, as well as professionals in relevant sectors (e.g., geological surveys,
reinsurance industries, city planners). Through this LUC, we aim to advance the Research Data
Management workflows for physics-based natural hazard assessment, a field of considerable
societal relevance, making world-leading HPC simulation results accessible to a broad Earth system
community.
Contributors:
Earthquakes and tsunamis are among the most devastating natural hazards, posing unpredictable hazards to densely populated regions worldwide. Physics-based simulations of large devastating earthquakes (and their associated tsunami generation) using high-performance computing (HPC)
now allow for the first time a rapid understanding of earthquake processes, and their tsunamigenic
potential permitting unprecedented insight into the dynamics of fault systems, and their ground
shaking and inundation hazards. Pioneer simulation campaigns for hazard assessment, learning of
surrogate models or parameter inference create massive amounts of simulation data, and are
often produced and exploited for a single purpose only, thus the concrete problem and challenge
lies in the lack of standardized and accessible frameworks for handling, curating, reusing and mining
this valuable resource. Our objectives are 1) to (further) develop and promote international
standards for simulation (meta-)data formats for large-scale earthquake/tsunami simulations, 2) to
develop tools to access, combine and postprocess (visualize) simulation and observational datasets;
and 3) to provide derived, labelled products (e.g., shake maps, tsunami inundation maps) for
downstream AI data mining and use in scientific and civil protection use cases. The result will be
unified, FAIR simulation data formats and a set of integrated geo-referenced case studies (giant
subduction zone earthquakes, induced seismicity in cities) adding the perspective of a simulation
data lake to NFDI4Earth infrastructure, including the OneStop4All and KnowledgeHub platforms,
leveraging international collaborations of the PIs. Target users include Earth System scientists,
research software engineers, as well as professionals in relevant sectors (e.g., geological surveys,
reinsurance industries, city planners). Through this LUC, we aim to advance the Research Data
Management workflows for physics-based natural hazard assessment, a field of considerable
societal relevance, making world-leading HPC simulation results accessible to a broad Earth system
community.
Contributors:
Format
on-demandon-site
