Subsurface energy and environmental systems like geothermal energy systems, geological carbon storage, underground hydrogen storage, groundwater flow systems, etc., play a vital role for mitigating climate change and achieving net-zero. At the same time, their complex, heterogeneous, multiscale, and uncertain nature presents major challenges for modelling, prediction, monitoring, and decision-making, particularly in multiphysics and multiscale model coupling, data integration, and real-time monitoring and control.
Artificial intelligence (AI) and scientific machine learning (SciML) are creating new opportunities to address these challenges. This session aims to bring together researchers from geoscience, hydrology, subsurface energy, computational science, and AI to discuss recent advances at the interface between these fields. We invite contributions on the development and application of AI and SciML methods for subsurface energy and environmental systems. Topics include, but are not limited to, physics-informed machine learning, neural operators, reduced-order modelling, surrogate and generative modelling, multimodal data fusion, digital twins, explainable and trustworthy AI, large language models, and agentic AI.
The session will provide a platform for interdisciplinary exchange on emerging methods, practical applications, current limitations, and future research directions.

Earth Observation (EO) offers a powerful means of monitoring changes in climate, ecosystems, and human environments at both global and local scales. These observations generate a wide array of climate and environmental variables, and they are delivered as Analysis-Ready Data (ARD). While ARD is globally accessible and scientifically robust, it might lack the specificity and contextual relevance required to effectively address local challenges. To bridge this gap, ARD must be transformed into Action-Ready Information (ARI): tailored data products and insights that support local decision-making and reflect community priorities. This transformation depends on co-creation, a collaborative process involving local communities, scientists, engineers, policymakers, and private sector stakeholders. For example, by integrating satellite EO with locally collected data from ground, water, and airborne platforms, we can enhance data granularity, validate satellite outputs, and generate customized, equitable, and actionable solutions. This session will explore how data can be harnessed to support environmental monitoring, local climate mitigation and adaptation, and sustainable development. It will emphasize the importance of identifying gaps between global datasets and local needs, and present strategies to close these gaps through innovation (e.g. new technologies and open FAIR science), inclusive engagement, and capacity building. Economic and policy dimensions will also be addressed, including the sustainability of community-led initiatives, the role of citizen science, funding mechanisms, and scalable technologies that enhance data utility for local solutions. The practical implementation challenges confronting policymakers when seeking to engage with EO data, particularly in the context of constrained policy capacities, will also be discussed. We invite participants from across/around the EO ecosystem: researchers in both physical and social sciences, community leaders, and stakeholders from policy and business sectors. We do not limit us only to satellite EO. We do consider non-EO observations and data, and their applications. We will share case studies, identify synergies between global and local efforts, and co-create knowledge that informs both local action and global strategies. By synthesizing diverse experiences, this session aims to advance EO as a tool for addressing the interconnected climate and environmental challenges we face locally and globally.