VSR Seminar with two talks

1st talk: Understanding coupled reactive transport in fractured crystalline rocks across scales

Speakers:
Guido Deissmann, Institute of Fusion Energy and Nuclear Waste Management (IFN-2), Forschungszentrum Jülich GmbH, Jülich, Germany
Paolo Trinchero, Amphos21 Consulting, Barcelona, Spain

Abstract:

The fate of solutes and contaminants in the subsurface is controlled by various strongly coupled thermal-hydraulic-mechanical-chemical-biological (THMC/B) processes. In the last decades “reactive transport modelling” (RTM) has emerged as a versatile simulation tool integrating groundwater flow and geo(bio)chemical processes to quantitatively describe coupled THMC processes, for example, in the context of mining operations, environmental remediation or safety assessments for deep geological repositories for radioactive wastes. Irrespective of the approaches or the combination of approaches being used in RTM, the systems are formulated as a set of highly non-linear algebraic equations that must be solved by means of iterative numerical schemes.

Various countries (e.g. Sweden, Finland, Canada, and Germany) consider building geological disposal facilities for radioactive wastes in crystalline “granitic” rocks. For demonstrating the safety of a repository system over time scales of up to 1 million years, a fundamental understanding and quantification of the long-term evolution of the repository system and the migration behaviour of radionuclides potentially released from the disposed wastes are essential. However, coupled reactive transport simulations in fractured and (micro)porous media are computationally highly demanding, for example, due to the significantly different timescales of various hydrogeological and geochemical processes and inherent heterogeneities of the systems on various scales. Moreover, on the repository scale, simulations of large spatial domains and extended time scales are required.

Here, we discuss the application of reactive transport simulations using high performance computing to analyse effects of heterogeneities in crystalline rocks on processes affecting subsurface radionuclide migration, to address transport/retention of radionuclides in crystalline rocks on various scales, and to develop upscaling methodologies. These simulations were performed on the supercomputers JUQUEEN and JURECA at the Jülich Supercomputing Centre, using the massively parallel reactive transport code PFLOTRAN. The results of these simulations contribute to a refined understanding of the effects of system heterogeneities on key processes controlling the geochemical evolution and radionuclide migration in the subsurface. Moreover, they serve as a starting point for the development of approaches to represent system heterogeneity on multiple scales and for upscaling strategies for larger-scale reactive transport models, thus providing a more realistic view on the evolution of repository systems.

2nd talk: Towards a digital twin of the cortical network at cellular resolution

Speaker: Markus Diesmann, Institute for Advanced Simulation (IAS-6), Computational and Systems Neuroscience, Forschungszentrum Jülich GmbH, Jülich, Germany

Abstract:

Computational neuroscience is entering a new era. This originates from the convergence of two developments: First, biological knowledge has expanded, enabling the construction of anatomically detailed models of one or multiple brain areas. Second, simulation has firmly established itself in neuroscience as a third pillar alongside experiment and theory. A conceptual separation has been achieved between concrete network models and generic simulation engines. Neuroscientists can now work with digital twins of certain brain structures to test their ideas on brain functions and probe the validity of approximations required for analytical approaches.

However, the use of this capability also requires a change in mindset. Computational neuroscience seems stuck at a certain level of model complexity for the last decade not only because anatomical data were missing or because of a lack of simulation technology. The fascination of the field with minimal models leads to explanations for individual mechanisms, but the reduction to the bare equations required provides researchers with few contact points to build on these works and construct larger systems with a wider explanatory scope. In addition, creating large-scale models goes beyond the period of an individual PhD project. The change of perspective required is to view digital twins as research platforms and scientific software as infrastructure.

As a concrete example, the presentation discusses how the universality of mammalian brain structures motivates the construction of large-scale models [1,2,3] and demonstrates how digital workflows on infrastructures like EBRAINS1 help to reproduce results and increase the confidence in such models.

A digital twin promotes neuroscientific investigations but can also serve as a benchmark for technology. The talk shows how a model of the cortical microcircuit has become a de facto standard for neuromorphic computing [4] and has sparked a constructive race in the community for ever larger computation speed and lower energy consumption.

[1] Jiang H-J., Qi G., Duarte R., Feldmeyer D., van Albada SJ. (2024) A Layered Microcircuit Model of Somatosensory Cortex with Three Interneuron Types and Cell-Type-Specific Short-Term Plasticity. Cerebral Cortex 34 (9), bhae378. DOI: 10.1093/cercor/bhae378

[2] Pronold J., van Meegen A., Vollenbröker H., Shimoura R.O., Senden M., Hilgetag C.C., Bakker R., van Albada, S.J. (2024) Multi-Scale Spiking Network Model of Human Cerebral Cortex. Cerebral Cortex. DOI: 10.1093/cercor/bhae409

[3] Senk J., Hagen E., van Albada S.J., Diesmann M. (2024) Reconciliation of weak pairwise spike-train correlations and highly coherent local field potentials across space. Cerebral Cortex. 34(10). DOI: 10.1093/cercor/bhae405

[4] Kurth AC., Senk J.,Terhorst D.,Finnerty J., Diesmann M. (2022) Sub-realtime simulation of a neuronal network of natural density.