The rapid growth of multi-scale Earth observations presents unprecedented opportunities for discovery, yet exploiting them demands new approaches for hypothesis generation, predictive understanding, and synthesis across workflows. Recent advances in large language models (LLMs) and agentic AI—systems capable of autonomous planning, reasoning, and action—offer pathways to enable AI-assisted discovery.

This session invites contributions on the responsible use of LLMs and agentic AI in Earth science applications, including carbon, water, ecosystem, and atmospheric processes. We welcome studies on orchestrating heterogeneous data and models; AI-driven hypothesis generation, falsification, and scientific reasoning; human–AI collaborative systems that combine domain expertise with AI exploration; as well as benchmarking, evaluation, and reliability in Earth science applications.

By uniting Earth scientists and AI researchers, this session aims to define a new frontier where scalable, adaptive, and collaborative AI systems not only streamline research workflows but actively contribute to scientific discovery and predictive understanding of Earth system dynamics.

• A - Atmospheric Sciences
• B - Biogeosciences
• GC - Global Environmental Change
• H - Hydrology