Data science, machine learning (ML), and physics-informed modelling are rapidly transforming geothermal energy systems and subsurface resource engineering. These digital approaches enable the integrated interpretation of heterogeneous data, predictive simulation of complex coupled processes, uncertainty-aware decision making, and end-to-end digital workflows across the full lifecycle of subsurface energy projects.
This session invites contributions advancing data-driven, physics-informed, and hybrid physics–ML methodologies for geothermal, geotechnical, and geoenvironmental applications relevant to energy and subsurface resources. Topics include subsurface and site characterization, reservoir engineering, subsurface flow and transport, induced seismicity, coupled thermo-hydro-mechanical-chemical processes, as well as geotechnical aspects of geothermal infrastructure such as foundations, tunnelling, and slope stability. Contributions employing supervised and unsupervised learning, deep learning, physics-informed neural networks, surrogate and reduced-order models, and inverse modelling approaches are particularly encouraged. We also welcome contributions that integrate laboratory and field experimentation, such as CT imaging, NMR, scattering methods, core- and rock-mechanical testing, and field monitoring, with data science and physics-informed modelling workflows for parameter inference, model calibration, and validation.
A central focus of the session is digital innovation for geothermal energy systems, spanning exploration, development, monitoring, and operation. Relevant topics include geothermal and subsurface databases; data quality control, uncertainty quantification, and metadata standards; integration of multi-source datasets (geological, geophysical, thermal, geochemical, and operational); and AI/ML approaches for resource assessment, reservoir characterisation and modelling, performance forecasting, and operational risk or failure prediction. Contributions related to open and FAIR data infrastructures, semantic technologies, and European and national initiatives (e.g., GeoERA, DESTRESS, HeatStore, GSEU, GeoMAP, MALEG, WärmeGut) are particularly welcome.
The session also welcomes studies integrating ML and data science with monitoring technologies such as IoT sensor networks, remote sensing, and real-time data streams, as well as the development of digital twins for geothermal and subsurface energy systems.
Solicited authors:
Sergey Oladyshkin