GENAI-X project

AI Generalizability in Non-stationary Environmental Regimes:
The Case of Hydro-climatic Extremes 

Project Acronym: GENAI-X
Project Duration: April 2026 - March 2031
Funding Volume: €6,000,000

Summary

GENAI-X addresses a fundamental AI challenge: achieving robust model generalizability in non-stationary environmental systems, where conditions vary across space and evolve unpredictably. Focusing on hydro-climatic extremes (floods, landslides, droughts and late-frost events) and their impacts, we advance AI methods that adapt to shifting data patterns and uncertainties. We integrate hybrid modeling, causal modeling, equation discovery, dimension reduction and uncertainty quantification to improve environmental understanding and prediction. Beyond theory, we develop AI-driven tools for assessing risks, interpreting environmental changes, and supporting decisions. Strengthening the ELLIS (European Laboratory for Learning and Intelligent Systems) Unit Jena, GENAI-X positions FSU as a leader in environmental AI.

The GENAI-X project originated from the ELLIS Unit JenaExternal link, reflecting its strong foundation in AI and environmental research. It aims to foster interdisciplinary collaboration and establish the basis for high-impact, integrative follow-up projects centered around this hub within the European Laboratory for Learning and Intelligent Systems (ELLIS)External link — a pan-European network of excellence founded in 2018 to advance fundamental research, technological innovation, and societal benefit.
A core strategic goal of GENAI-X is to further strengthen the ELLIS Unit Jena as a leading interdisciplinary research center and position Friedrich Schiller University Jena (FSU) at the forefront of AI-driven environmental science.

Scientific Motivation:
Conventional machine learning assumes stable data distributions. Yet in real-world environmental systems, data is often sparse, uncertain, and shaped by changing climatic drivers. This non-stationarity undermines model reliability and calls for AI systems that can adapt to shifting baselines and novel conditions.

Research Goals:
GENAI-X advances cutting-edge AI methodologies that enhance adaptability, resilience, and predictive performance in dynamic Earth and environmental systems. The project integrates five complementary approaches: (AI1) Hybrid modeling (AI2) Causal modeling (AI3) Equation discovery (AI4) Dimension reduction (AI5) Uncertainty quantification.

Impact and Application:
Beyond theoretical development, GENAI-X creates AI-driven tools for environmental monitoring, hazard assessment, and decision support that bridge scientific research and real-world application. These innovative tools aim to improve predictions of hydro-climatic risks, such as floods, droughts, and landslides, and to guide evidence-based policy and management strategies. By integrating satellite observations, in-situ sensor data, and large-scale simulations, the project enables a deeper understanding of dynamic Earth system processes and their socio-economic impacts. The resulting tools support sustainable resource management, biodiversity protection, and ecosystem resilience. In addition, GENAI-X enhances data-driven decision-making through transparent, explainable, and transferable AI models. The project strengthens interdisciplinary collaboration across computer science, geosciences, and environmental research, contributing to global efforts in climate adaptation, disaster risk reduction, and the long-term preservation of natural systems essential for human well-being.

Broader Relevance:
The results of GENAI-X are of importance for achieving social and environmental benefits by developing fundamental AI solutions and implementing them in the domain sciences in the field of hydro-climatic extremes. Beyond the domain-specific AI, multiple ongoing or planned research projects in the domain sciences will directly or indirectly benefit from the proposed advances in environmental AI, including collaboration with government stakeholders.