Virtual Access to the next generation of software as a service

WP5 provides Virtual and Transnational Access (VA and TA) to more than 20 compound installations, formed by almost 40 sub-installations by 7 RIs belonging to the EPOS-ERIC and the ChEESE Centre of Excellence for Exascale in Solid Earth, and to two of the largest European High-Performance Computing (HPC) centres (CINECA, BSC).

Virtual Access is provided to the Simulation Data Lake (SDL) and to 10 Software services from four different infrastructures:

Simulation Data Lake

Hosting Institutions:

Consorzio Interuniversitario (CINECA), Italy

Helmholtz Zentrum Potsdam Deutsches GeoForschungsZentrum (GFZ), Gemany

Istituto Nazionale di Geofisica e Vulcanologia (INGV), Italy

Consejo Superior de Investigaciones Científicas (CSIC), Spain

Stiftelsen Norges Geotekniske Institutt (NGI), Norway

Ludwig-Maximilians-Universität München (LMU), Germany

Barcelona Supercomputing Center - Centro Nacional de Supercomputación (BSC CNS), Spain

Universidad de Málaga (UMA), Spain

URLs:

Description:

The Simulation Data Lake (SDL) provided by CINECA (SDL@CINECA infrastructure) is engineered to ensure the enduring preservation, accessibility, and insightful exploration of results from numerical simulations and in-silico experiments. Committed to the FAIR principles (Findable, Accessible, Interoperable, and Reusable), this innovative service not only guarantees the reproducibility of analyses but also opens avenues for pioneering research while nurturing collaboration across diverse geoscience domains.

SDL@CINECA comprises multiple installations. At the core lies the actual storage server at CINECA, serving as the Virtual Access (VA) provider. Complementing this, the sub-installations refer to the effort that different institutions will dedicate to curating existing simulation datasets and integrating new ones from various Virtual Access (VA) and Task Areas (TAs) within the Geo-INQUIRE WP5 framework int the SDL.

CINECA is actively developing a state-of-the-art SDL hosted on its cloud infrastructure, envisioning it as a strategic asset within the geo-science domains. Upholding FAIR principles, the future SDL@CINECA will efficiently manage vast and intricate simulation datasets, fostering advancements in geoscience research.

In preparation for the full release of SDL@CINECA, a temporary initial VA service is now available. This temporary solution relies on the EUDAT Collaborative Data Infrastructure and its B2* service ecosystem. To enhance data sharing and dissemination efforts, a dedicated Geo-INQUIRE B2SHARE community has been established. Additionally, seamless integration with B2ACCESS, B2HANDLE, and B2FIND services enhances the SDL@CINECA infrastructure's capabilities in data access, management, and discovery, respectively.

Target community/Users:

Researchers, developers of geoscience-specific software, and those responsible for Geo-INQUIRE VA and/or TA services are the expected users of the SDL@CINECA installation.

Estimated users/year:

1.000 to 10.000

Community Standards:

Specific standards for the SDL will be defined and implemented during Geo-INQUIRE.

ChEESE CoE

The EU Center of Excellence for Exascale in Solid Earth (ChEESE) develops exascale transition capabilities in the domain of Solid Earth. The first implementation phase of the project (ChEESE-1P; 2018-2022) addressed scientific and technical computational
challenges in seismology, tsunami science, volcanology, and magnetohydrodynamics, in order to understand the phenomena, anticipate the impact of natural disasters, and contribute to risk management. The project initiated the optimisation of 10  community flagship codes for exascale systems and implemented 12 Pilot Demonstrators that combine the flagship codes with dedicated workflows in order to address the underlying capability and capacity computational challenges. (For more information, see “The EU Center of Excellence for Exascale in Solid Earth (ChEESE): Implementation, results, and roadmap for the second phase”, Arnau Folch et. al.) ChEESE is continued as ChEESE-2P (2023-2026).

Hosting Institution:

Ludwig-Maximilians-Universität München (LMU), Germany

URLs:

Description:

SeisSol is an earthquake and wave simulation open-source community code. SeisSol solves the seismic wave propagation problem (elastic, viscoelastic, poroelastic) linked to earthquake dynamic rupture in complex 3D models. SeisSol uses Discontinuous Galerkin discretization which is high-order accurate in space and time and local time-stepping on unstructured adaptive tetrahedral meshes. Scalable performance at Petascale has been demonstrated up to several thousands of nodes (on several  supercompers, e.g., Cori, SuperMUC, Hazel Hen, Shaheen, Frontera). As part of the ChEESE/ChEESE-2P Centre of Excellence, SeisSol has been ported and is being optimized for GPU-based supercomputers.

ExaHyPE is an open source simulation engine to solve hyperbolic PDE systems, as stemming from conservation laws. It is built on top of dynamically adaptive Cartesian meshes and offers support for Finite Volume and Discontinuous Galerkin discretizations. ExaHyPE is written in a way that most computer science aspects as well as most of the numerics are hidden away from the user: Users plug in user functions for their PDE formulation (such as flux functions and eigenvalues) into the engine and then delegate all further work to ExaHyPE. A concrete model for seismic wave propagation and dynamic rupture problems has been developed within ChEESE. The model is based on high-order Discontinuous Galerkin discretization and works on octree-structured Cartesian meshes.

Through the VA we offer several different options:

Target community/Users:

Researchers of any level either from Seismology or interdisciplinary fields (e.g. Tsunami, Geodynamics, Engineering); Students.

Estimated users/year:

Less than 100

Community Standards:

HDF5, NetCDF, Standard Rupture Format (SRF, https://seissol.readthedocs.io/en/latest/standard-rupture-format.html?#standard-rupture-format)

Hosting Institution:

Consejo Superior de Investigaciones Científicas (CSIC), Spain

URLs:

Description:

FALL3D is a parallel (MPI+OpenACC) Eulerian model for the transport and deposition of volcanic tephra developed and maintained at CSIC in collaboration with scientists of the INGV (Italy). The model solves the advection-diffusion sedimentation  equation using a Finite Volumes explicit scheme and can be used both to reproduce past events and to forecast on-going events.

The model can run in forecast mode or, if combined with a workflow (TA-541-2), to generate scenarios for Probabilistic Volcanic Hazard Assessment (PVHA).

VA to FALL3D atmospheric dispersal code can be presently accessed from a gitlab repository. During Geo-INQUIRE, in synergy with the ChEESE CoE, it will be provided using containerized applications and micro services, and it will be made available to users of EuroHPC via automated deployment, documentation, testing and quality checks.

Target community/Users:

Civil protection, aviation sector

Estimated users/year:

Less than 100

Hosting Institution:

Istituto Nazionale di Geofisica e Vulcanologia (INGV), Italy

URLs:

Description:

OpenFOAM is an open-source framework for developing computational fluid dynamics applications. It is composed by C++ libraries and solvers based on the Finite Volume method, and utilities for development and analysis.

The Geo-INQUIRE Ch-OpenFOAM is a suite of OpenFOAM-based solvers and libraries for volcanological and geophysical applications, optimized in the framework of ChEESE for HPC usage.

VA to ChEESE volcano dynamics simulation codes is made  available through download, documentation, and user support for OpenPDAC and MagmaFOAM open-source codes.

OpenPDAC is a multiphase Eulerian-Eulerian-Lagrangian solver built upon the multiphaseEulerFoam native solver, customized for compressible, turbulent volcanic gas-particle mixtures. It comes with automatic workflows for High Performance Computing simulations of explosive volcanic eruptions over complex 3D topographies. MagmaFOAM is a modular framework for the simulation of magmatic systems based on OpenFOAM. It incorporates models for solving the dynamics of multiphase,  multicomponent magmatic systems and several benchmarks. MagmaFOAM sets a robust framework for in-house and community model development, testing and comparison.

Potentially, new OpenFOAM-based applications will be made available during the course of the project.

Target community/Users:

Researchers (of any level), Students (graduate)

Estimated users/year:

Less than 100

Hosting Institution:

Universidad de Málaga (UMA), Spain

URLs:

Description:

Ch-HySEA is the numerical model of the HySEA family specifically designed for earthquake and landslide generated tsunami simulations. It combines robustness, reliability and good accuracy in a model based on a multi-GPU faster than real time (FTRT) implementation. Ch-HySEA model implements the generation, propagation, and coastal inundation.

Ch-HySEA VA service provides VA to code for simulating earthquake and landslide generated tsunamis at global, regional or local levels.

The output is a suite of NetCDF files with all the data corresponding to the numerical simulations. Several Python/bash codes are provided as tools for pre or post-processing data.

Target community/Users:

Academic and Research community, Hazard Practitioners. Results can be used as input for warning systems in this framework

Estimated users/year:

Less than 100

Community Standards:

NetCDF

EPOS TCS Tsunami

The candidate Thematic Core Service (cTCS) Tsunami coordinates within the EPOS nfrastructure the provision of various tsunami-related services organised within four thematic pillars: (1) Support to Tsunami Service Providers, (2) Tsunami Data, (3) Numerical Models, and (4) Hazard and Risk Products.

Hosting Institution:

Stiftelsen Norges Geotekniske Institutt (NGI), Norway

URLs:

Description:

This service equips scientists with an advanced tool for simulating landslide dynamics in both subaerial and subaqueous conditions, both as a tool towards simulating the landslide dynamics and its run-out distance, but also for generating tsunamis. The  depth-averaged model, named BingClaw, is primarily tailored for taking into account the behaviour of clay-rich materials during landslide motion for both onshore and offshore applications.

The service computes the motion of the landslide as a function of time. It uses a finite volume scheme to solve the equations of motion, which is depth averaged for the Herschel-Bulkley fluid. It considers the complex terrain deflection, as well as the  remoulding of the material, as well as the fluid resistance, and includes the transition between the fluid phases in air and water.

The BingClaw numerical model is presently provided in a repository for local usage. Some pre- and post-processing tools will also be provided at a later stage. The model will also be containerised later to simplify deployment.

Target community/Users:

Researchers and engineers. The program license is limited to research use.

Estimated users/year:

Less than 100

Community Standards:

Most input and out formats mimic closely ESRII ASCII.

Hosting Institution:

Stiftelsen Norges Geotekniske Institutt (NGI), Norway

URLs:

Description:

This service combines tools and data for estimating tsunami maximum inundation heights based on precomputed global amplification factors, and for computing new amplification factors. The amplification factors consider both the local bathymetry and  the wave characteristics extracted from the user's tsunami simulations at predefined offshore points along the 50 m depth contour.

The first part of the service available at present is based on a lookup table with global amplification factors connected at predefined points (points of interest – POI) distributed globally along the 50 m depth contour. Based on the amplification factor for the specified wave characteristics the estimation of the maximum inundation height on nearby land is then found by multiplying the amplification factor with the height of the wave at the POIs.

The service is presently being updated for i) extracting wave information from the users own simulations, and ii) to allow users to make their own (new) amplification factors on arbitrary profiles. The tool for extracting the wave characteristics will take as  input the users own tsunami simulations (externally to this service) to define the shape of the leading wave, wave period and wave height at each POI. These parameters are needed as input for extracting the amplification factor.

In future versions of the service in Geo-INQUIRE, the users can also compute their own amplification factors for e.g. due to improved bathymetric profiles or higher sampling rate of (own) POIs, this part of the service is presently being updated.

Target community/Users:

All users working with tsunami hazard risk on a regional or global scale.

Estimated users/year:

Less than 100

Hosting institutions:

Helmholtz Zentrum Potsdam Deutsches GeoForschungsZentrum (GFZ), Germany

Istituto Nazionale di Geofisica e Vulcanologia (INGV), Italy

URL:

  • Available by October 2023.

Description:

The “GaussToolbox” will provide instant tsunami propagation modelling based on the technique of pre-computed Greens’ functions. The service, in its first implementation will be presented as a suite of source codes and sample data for the  Mediterranean region. The package will include documentation and a cook-book to facilitate quick start with the package. On-demand training will be provided as a part of this VA. The first version of the service will be available starting from October 2023.

Target community/users:

tsunami modellers; hazard and risk analysts; early warning centres

Estimated users/year:

Less than 100

Community Standards:

netCDF

Hosting Institutions:

Institut Français de Recherche pour l'Exploitation de la Mer (IFREMER), France

Universitá degli Studi di Napoli Federico II (UNINA), Italy

URLs:

Geo-INQUIRE virtual training course:

Generating earthquake slip on complex geological faults: combining EFSM20 and k223d/ANTI-FASc

Description:

IFREMER:

Code for generating fractal slip distributions on non-planar faults.

This service provides a programme that can generate self similar slip distributions on non-planar surfaces. The programme calculates slip distributions with a k^-2 spectra by using the composite source model method, meaning no Fourier transforms are required. This allows the use of irregular unstructured mesh to describe the fault surface.

The service involves downloading the source code which can then be compiled using the provided makefile and fortran compiler (has been tested on gfortran and ifort to date). An example code is also provided for the generation of a simple planar fault mesh. The service is being updated to enhance interoperability with other services (e.g. tsunami simulation codes and kinematic / dynamic modelling), for example including the output of ground deformation due to the slip based on a half space Earth. Within Geo-Inquire, Jupyter notebooks will be developed as cookbooks for users demonstrating how to use the service, for example on how to generate a mesh from Slab 2.0 and use within the programme.

UNINA:

The service basically grounds on the software ANTI-FASc. This software is used to compute optimized ensembles of stochastic slip distributions, based on k-2 paradigm. The slip distributions can be defined on complex geometries, such as non-planar surfaces, and hence are suitable for modelling seismic source scenarios on realistic models of subducting plates. The output of the software can be promptly used as initial conditions for tsunami scenario simulators such as Tsunami-HySEA. The  computed scenario ensembles take into account a broad range of seismic source complexities (e.g. variable rigidity, geometries and stress drop) and are being currently efficiently used for several applications such as S-PTHA and PTF Tsunami EW.

In the current version, the service allows to download the software ANTI-FASc along with a training documentation that allows to run some test-case examples and to learn how to modify the input parameters to run all the possible applications. It  provides the stand-alone executables and the source codes of the various modules of the software ANTI-FASc along with a brief manual of using and some test-case examples. In the framework of Geo-Inquire activities the output will be adapted to the  standard format used by tsunami simulation services and a more detailed documentation will be provided

Target community/Users:

Tsunami researchers, tsunami modelers, also with application to hazard and early warning

Estimated users/year:

100 to 1.000

Community Standards:

VTK format, GeoJSON

EPOS TCS Volcanology

The Volcano Observations (VO) Thematic Core Service (TCS) provides long-term access to the Volcano Observatories and Research Institutions data and products. VO TCS aims at implementing facilities allowing easy access to volcanological data and  interoperable services. Virtual Access to web-based volcanological software, is also offered through the interface. Applications include volatile partitioning in magmas; kinematics of pyroclastic density currents; long-term probability estimation of volcanic hazard phenomena.

Hosting Institution:

Istituto Nazionale di Geofisica e Vulcanologia (INGV), Italy

URLs:

Description:

The Volcano Dynamics Computational Centre (VDCC) is an infrastructure providing Virtual Access to volcanological software through an interactive web framework. It allows the user to run volcanological applications online, but it is not easy to maintain and to expand because it requires a significant workload to implement new software in the web framework.

The EPOS Computational Infrastructure for Volcanology (CIV) is a software catalogue made available by EPOS. The service offers a description and some documentation about volcanological software, and a catalogue of metadata accessible through the EPOS portal, but does not offer the possibility for online execution of the applications.

The Geo-INQUIRE VO-CIV@INGV service being implemented aims to combine (and to overcome the limits of) the EUROVOLC VDCC VA infrastructure and the EPOS CIV, to provide APIs to make volcanological software portable and accessible as a Web Processing Service (WPS).

The VO-CIV infrastructure is designed to ease the implementation of Pilot Demonstrator Applications during the Geo-INQUIRE project, in a sustainable perspective.

Target community/Users:

Researchers (of any level); Undergraduate and graduate students; Volcano observatories and decision makers.

Estimated users/year:

Less than 100

GFZ Pyrocko

Pyrocko is an open source seismology toolbox and library, written in the Python programming language. It can be utilised flexibly for a variety of geophysical tasks, like seismological data processing and analysis, modelling of InSAR, GPS data and dynamic waveforms, or for seismic source characterization.

Hosting Institution:

Helmholtz Zentrum Potsdam Deutsches GeoForschungsZentrum (GFZ), Germany

URLs:

Description:

Pyrocko is a powerful seismology toolbox and library written in Python, with an active and growing community of users and contributors. It can be utilized flexibly for a variety of geophysical tasks, like seismological data processing and analysis,  modelling of InSAR, GPS data and dynamic waveforms, or for seismic source characterization.

To improve the user experience and facilitate the installation and usage of Pyrocko, implementing improving the continuous integration/continuous delivery (CI/CD) pipeline for Linux, Apple MacOS and Microsoft Window can be very beneficial. By  automating the build, testing, and deployment processes, it can ensure that new features and bug fixes are promptly and reliably delivered to users, and can also reduce the time and effort required for maintaining the software. The CD pipelines will publish Pyrocko on PyPI, users can easily install it using tools such as `pip`, withouthaving to manually download and install the software or its dependencies. These workflows ensure that each new release is tested and packaged consistently across different platforms and Python versions. This can improve the user experience by providing a more streamlined and reliable installation process, and can also increase the visibility and accessibility of Pyrocko to a wider community of users.

Implementing an online service that calculates synthetic waveforms and surface displacements via an API can be a valuable addition to the existing online Green's function databases. This service can provide users with a more streamlined and user-friendly experience by allowing them to easily access the calculations they need without having to download and install large databases. Users can simply send a request to the API with the necessary input parameters, and the API can respond with the  calculated synthetic waveforms or surface displacements. This approach can also be useful for researchers who require custom modeling or calculations that are not available in the pre-calculated Green's function databases. By leveraging the power of Pyrocko and its Python-based API, this service can offer a flexible and customizable platform for seismic modeling and analysis. Moreover, by implementing security and access controls, this service can ensure that the calculations and results are only accessible to authorized users, and that the service remains scalable and reliable over time.

To improve the user experience and facilitate the usage of the Pyrocko toolbox, we plan to generate online documentation and training material. These materials will be provided as interactive Jupyter notebooks, which will allow to learn Pyrocko and  tools from the ecosystem in a self-paced, hands-on manner. Additionally, we will also provide video tutorials that complement the notebooks, providing an overview of the various features and use cases of the toolbox. These materials will be designed  to be accessible and engaging, catering to both novice and experienced users, and will cover a wide range of topics, from basic usage to advanced topics like waveform inversion and source characterization. By providing a range of learning materials  that cater to different learning styles and preferences, we aim to make Pyrocko more accessible and user-friendly.

Target community/Users:

Seismologists, Geodesy, Earthquake modelers

Estimated users/year:

100 to 1.000

Community Standards:

MiniSeed and various waveform data formats, StationXML, QuakeML, FDSNWS and many others