The research project “Reduced Complexity Models”, funded in the framework of Helmholtz Incubator Initiative, is a joint interdisciplinary effort for an integrated approach addressing three ubiquitous core issues in modern model- and data-based science across any research field: (1) uncertainty quantification of complex models, (2) development of fast surrogate (emulator) models and (3) the identification of key parameters and dependencies. With the explosion of complexity of available data and computer models, mastering these three aspects is essential for knowledge extraction and simulation-based decision making, irrespective of the field of application.

The project brings together already existing expertise distributed at the different Helmholtz Centers and aims at developing transferable data science methods and softwarein a unified manner, which will beneficial to any Helmholtz Centre and other users from science and industry. A selection of relevant and strategic use cases from the diverse research fields will serve as demonstration for the developed approaches. 

Related activities at the Fluid Systems Modelling Section mainly focus on the second key issue, the development of fast surrogate models, in particular to speedup reactive transport simulations, which deal with fluids, chemically interacting with host rocks along their migration in the geological subsurface. The idea is to substitute computationally-intensive and fully-coupled geochemical simulations by statistical surrogates in coupled reactive transport simulations. This is expected to enhance both, their computational speed and numerical stability, finally enabling large¬scale assessments of uncertainties and risks as well as optimization of efficiency and sustainability in geological subsurface utilization.

Contact:

Dr. Marco De Lucia