The increasing complexity and resource intensity of state-of-the-art Earth System Science (ESS), from Earth System Model simulations to large-scale data processing and analysis, require advanced approaches to orchestrate, automate, document and reproduce scientific processes. Scientific workflows provide a fundamental means to manage this complexity and to enable scalable and reproducible research across increasingly complex computational and data infrastructures.

Researchers in ESS are faced with heterogeneous computing architectures, federated research infrastructures, distributed and increasingly large datasets and restrictive data-movement policies. Workflows thus have to operate across systems while making efficient use of available computational and data resources. Portability, interoperability and provenance therefore become essential characteristics of scientific workflows rather than optional additions.

Advances in artificial intelligence (AI) introduce an additional dimension. AI can support the composition, optimization and execution of scientific workflows. Emerging agentic approaches may increasingly enable workflows to adapt autonomously to data, computational environments and scientific tasks. Such capabilities also raise new questions: How can scientific intent, assumptions, dependencies and execution environments be represented in AI generated workflows? How can AI-assisted and agentic workflows remain transparent, verifiable and reproducible? How can their decisions and adaptations be captured through appropriate provenance and metadata?

This session explores advances in scientific workflow concepts, technologies and applications for ESS, ranging from established workflow management approaches to emerging AI-assisted and agentic systems. We particularly welcome contributions addressing workflow orchestration and portability, scalable HPC and data-intensive workflows, distributed, federated and hybrid infrastructures, data locality and efficient resource utilisation, workflow FAIRnes, and AI-assisted generation, adaptation, optimization and execution of scientific workflows.

We encourage contributions presenting methods, tools, lessons learned and concrete use cases from ESS, including European and international research infrastructures and initiatives. Contributions demonstrating both mature solutions and emerging approaches that explore emerging capabilities of scientific workflow technologies are particularly welcome.