This online course introduces how machine learning (ML) is transforming weather and Earth system modelling within the Destination Earth (DestinE) initiative. As part of a new three-course learning pathway developed by ECMWF, the training shows how ML methods complement physical modelling to support next-generation Earth digital twins.

Participants will explore the foundations of machine learning for Earth System Sciences (ESS), understand how AI supports DestinE digital twins and emerging simulation systems, and examine ethical and regulatory aspects alongside future skills required in this rapidly evolving domain.

The course combines conceptual understanding with applied examples, preparing learners to engage with modern ML-driven Earth system workflows.

Main topics

The course will cover the following themes:

• Core ML components and terminology applied to Earth Systems modelling
• DestinE initiative and how ML is embedded in its datasets to enhance data assimilation, emulation, and predictive workflows
• Modern AI weather models and recent technological advances
• Key regulatory frameworks and ethical considerations for using AI
• Future technological and societal transitions linked to the use of AI for the Earth Systems modelling
• Perspectives on the use of ML in Earth Systems modeling from research, operations, industry, and policy

Target audience

This course is part of a series designed for:

• Researchers and practitioners in meteorology and climate science.
• Technical specialists from meteorological services and climate centres.
• Academic and industry professionals working with Earth Systems data.

This first course, Foundations and New Frontiers, specifically targets a broad audience of high-level information users and policy/decision-makers from both European public and private sectors, academia, and industry. This course will also be of interest to a technical audience of non-ML meteorologists who want to be onboarded into the world of ML.

Requirements

Participants should have a background in meteorology, climate systems, or Earth Systems Sciences. Participants should also have familiarity with statistics and data interpretation. Some prior knowledge of ML concepts and some coding experience is helpful, but not essential.

Register now!

Register for this course through our eLearning portal.

All lectures are given in English and will remain online and available to anyone who would like to take them as self-study, according to their own pace and schedule.

Coming up soon...

This course is the first of a three-part series of courses on ML and Destination Earth.

1. Foundations, Context and New Frontiers

Launch date: March 16th 2026

Learn about the role of ML in DestinE and its Digital Twins, how ML complements traditional physics-based models, and how emerging AI prediction systems such as AIFS and hybrid approaches are reshaping weather and climate modelling.

2. Architectures, Data, and Prediction. 

Expected launch date: June 2026

A comprehensive, hands-on exploration of modern AI-based prediction systems used in Earth System Science. You will work with real ML workflows, including data pipelines, neural architectures, model training, benchmarking, and the Anemoi ecosystem used within DestinE. 

3. Applications and New Directions. 

Expected launch date: Q3/Q4 2026 (TBD)

Cutting-edge applications of machine learning in weather and climate science. Topics include AI weather prediction models at scale, downscaling, fire and flood prediction, anomaly detection, hybrid and foundation models, ML-based data assimilation, and end-to-end observation-driven approaches.