ChEESE-2P is a European Center for Excellence in Research that focuses on modeling geophysical processes such as earthquakes, volcanic eruptions, and tsunamis exploiting Europe's most powerful supercomputers. In ChEESE-2P, three key areas related to supercomputing applications within geophysical processes are addressed: i) models that calculate complex geohazard processes in three physical dimensions, ii) urgent computing for acute situations such as early warning or for monitoring ongoing events, and iii) conducting a large number of calculations to quantify uncertainty. The objective in ChEESE-2P is to enable the models to exploit the most powerful supercomputers we have available, so-called Exascale machines that perform at least one billion trillion floating-point operations (FLOP) per second, to solve challenges that we cannot solve today.
The Norwegian part of the project focuses tsunami hazards, and in particular, two new pilot applications: i) a global model that quantifies the probability of tsunami run-up heights generated by earthquakes along all coastlines in the world, and ii) a multiphysics model that connects the different components, coupling possible chain reactions involved in tsunami generation and propagation for complex events involving several tsunami sources. The various activities that have been completed so far are described below for each of the two pilots:
The global tsunami model addresses the modeling of various earthquake scenarios with different probabilities for all areas of the world that have the potential to generate large earthquakes (earthquakes above magnitude 7). In ChEESE-2P, we have started work on breaking down the various zones into subcomponents that are used as so-called unit sources. These have been further combined to model earthquake scenarios with probabilities. Other activities have included the analysis and setup of computational grids for all the world's oceans, as well as the development of a new dataset that is used as input for calculating tsunami run-up heights globally. The work done so far has laid the groundwork for setting up large-scale tsunami calculations globally later in the project. The work has been carried out in close collaboration with the research institute INGV in Italy.
The multiphysics model is used to model events that involve multiple phenomena occurring simultaneously and involving chain reactions between phenomena that have not previously been embedded in tsunami models. An example of such an event is the global tsunami from the Tonga eruption in 2022. The Tonga event involved the collapse of the volcano, landslide dynamics, and a massive acoustic pressure wave that was recorded worldwide. The focus in the project so far has been on first developing models for these individual components. The first of these components connects landslide dynamics with local tsunami propagation. The second of these components is the connection of the acoustic pressure field with global tsunami propagation. NGI has started the development of a workflow model that links the local model with the global one, and has explored how this should be set up on relevant high-performance computing infrastructure in Europe. The work has been carried out in close collaboration with several different research institutes, including Institut Rüder Boskovic in Croatia, University of Malaga in Spain, Technical University Munich and Ludwig Maximilian University in Germany, and INGV.
This project aims at implementing the second phase of the Center of Excellence for Exascale in Solid Earth (ChEESE-2P). Solid Earth (SE) sciences address fundamental problems in understanding the formation, composition, and the dynamics of the geosphere from its deep interior to the surface. This encompasses also the study of geohazardous phenomena that originate in the Earth’s interior but that manifest at the interface with the atmosphere, the hydrosphere, and the biosphere, causing a variety of natural hazards and geophysical extremes across all spatial and temporal scales. SE is extremely rich in computational challenges, requiring petascale and exascale infrastructures both to address fundamental scientific questions and to anticipate, mitigate, and manage the occurrence of geohazards and their impacts.
The part of this project with Norwegian participation concerns adressing supercomputing challenges related to tsunami hazards. To this end, two novel computational pilot demonstrators within tsunami science will be established
i) A global probabilistic tsunami hazard and uncertainty quantification development that allow
users to carry out local probabilistic tusnami hazard analysis at high resolution at any location worldwide.
ii) Complex multi-source tsunami modeling combining several multiphysics processes from earthquakes, landslides, volcanoes, and acoustic pressure and atmospheric forcing to tsunami generation and propagation in a single workflow that allows coupling of different flagship codes. Present tsunami prediction capabilities in use today use a single source representation. It is hence necessary to develop a coupled approach where all the different physics can be tied together, including complex earthquake rupture, landslides, volcanic explosions, acoustic pressure and explosions, as well as atmospheric tsunami sources.