ELLIS Winter School 2027

AI & ML in Earth System Science

March 1-5, 2027 | Friedrich-Schiller-University Jena (Germany).

The ELLIS Winter School on “AI & ML in Earth System Science: Exploring current and future challenges in leveraging AI and ML for improving our understanding and modelling of the Earth system” is a 5-day hands-on program hosted by AI4PEX and ELIAS Horizon Europe projects and the ELLIS Unit Jena, in cooperation with MeDiTwin and EXPECT. The school offers a combination of theoretical lectures, hands-on training, and real-world applications, bringing together experienced leaders and scientists with early career researchers to explore and discuss the latest advances in AI and ML for solving classic, new and long-standing challenges in Earth System science.

Logos

Follow us on social media & share our posts!

|||AI4PEX|ELLIS Unit Jena|ELIAS|MediTwin|EXPECT

Event Details

Jena City View
  • When: 1.- 5. March 2027
  • Location: Friedrich Schiller University Jena - Rosensäle (map)
  • Format: in-person (hands-on)
  • Working Language: English
  • Participant Fees: 400 €

Key Pillars

EARTH SYSTEM DOMAINS

Investigate physical, chemical, and biological processes driving land, ocean, atmosphere, and biosphere. Focus on critical interactions and feedbacks defining Earth system behaviour across vast spatial and temporal scales.
ARTIFICIAL INTELLIGENCE & MACHINE LEARNING

Learn to analyse Earth observations and enhance system models using cutting-edge techniques. From foundation models and generative AI to stochastic modelling and agentic AI, adapt methods for specific scientific tasks.
TRUSTWORTHY AI

Assess robustness, reliability, physically meaningfulness and identify and identify limitations in AI/ML data-driven approaches. Focus on explainability, responsible use and the ethical implications of applying AI in Earth system science.
OPEN CHALLENGES

Tackle open scientific and methodological challenges, through open discussions and collaborative group challenges, identify unresolved questions, promising research directions and opportunities for interdisciplinary discovery.

Keynotes by Experts

Keynotes will feature leading scientists and practitioners sharing cutting-edge insights and forward-looking perspectives on Earth system science and machine learning/AI, setting the stage for the school's themes and inspiring participants with new ideas and directions.

Zhijing Jin

Zhijing Jin
 University of Toronto

Ethics │ Causality │ Responsible AI

Tapio Schneider

Tapio Schneider
 Caltech

Earth System Modeling │ Coupling

Konstantin Klemmer

Konstantin Klemmer
 ETH Zurich

Foundation Models │ Land │ Biosphere

Peter Dueben

Peter Dueben
 ECMWF

Foundation Models │ Coupling

Lectures

Lectures Introducing Theory and Basics
The lectures will provide foundational knowledge and theoretical underpinnings of specific topics, ensuring participants have a solid understanding of the concepts before diving into hands-on activities. They will cover essential principles and methods in an accessible and engaging manner.

Christian Reimers

Christian Reimers
 Max Planck Institute for Biogeochemistry

Land │ Biosphere │ generative AI

Shijie Jiang

Shijie Jiang
 Max Planck Institute for Biogeochemistry

Climate Extremes │ explainable AI

Zero2Hero & Challenges

Zero2Heros (Guided Hands-on Sessions)
Zero2Heros are structured, guided hands-on sessions where participants learn a specific technique or method through practical exercises, with the support of tutors. These sessions aim to rapidly build skills and confidence in applying new tools and approaches to real-world problems.

Challenges (Group Projects)
Challenges are collaborative, problem-based activities where participants work in groups to address a real-world or simulated problem. These sessions encourage teamwork, creativity, and critical thinking, allowing participants to apply learned concepts to solve complex tasks and develop solutions under guidance from experienced mentors.

Programme

WinterSchoolProgramme
Understand Earth System Processes & Interactions

Explore the processes, interactions, and feedbacks across land, ocean, atmosphere, and biosphere to understand the complexity of the Earth system.
Apply AI & ML to Earth System Models

Discover and experiment with AI and ML methods to improve Earth system understanding and modelling, from data-driven approaches to emerging techniques.
Understand and Explain Models

Learn how to make AI models more transparent, physics-informed, and causally grounded, while considering uncertainty, limitations, and ethical principles for the responsible use of AI in Earth system science.
Collaborate and Exchange Ideas

Connect with experts in AI and Earth system science, exchange perspectives with researchers from different backgrounds, and develop new ideas through hands-on challenges and open discussions.

Registration

Registration is open until 15. November 2026, 23:00 CET . To register use our online form. Participant number is limited to 100.
Notifications of acceptance will be sent no later than by 14. December 2026.

For registration you need the following: