© daizuoxin/AdobeStock
Applications are now open for three in-person ECMWF training courses taking place in Reading in January 2027.
Covering data assimilation, satellite observations and the use of ECMWF forecast products, the courses combine expert-led lectures with practical sessions to deepen participants' understanding of numerical weather prediction and forecasting.
Training and education are a key part of ECMWF’s work, helping Member and Co-operating States and the wider community build expertise in weather forecasting and make the most effective use of ECMWF products and services.
Data assimilation & Machine Learning
11–15 January
This five-day course provides an overview of data assimilation methods in numerical weather prediction, including the assimilation of observations and the practical implementation of assimilation techniques. It also explores the growing role of machine learning within data assimilation workflows.
Participants will learn about core assimilation methods, uncertainty estimation, error modelling and the use of observations, including satellite data. The programme also examines emerging AI applications, such as model error correction, hybrid modelling and generative AI techniques, through a combination of lectures, discussions and hands-on exercises.
The course is intended for participants with a strong background in meteorology and mathematics, with some experience in numerical weather prediction beneficial.
Application deadline: 27 September 2026
Full details and how to apply on the Data assimilation & Machine Learning course webpage.
EUMETSAT/ECMWF NWP-SAF satellite data assimilation
18–22 January
Sponsored by EUMETSAT, this five-day course provides a comprehensive overview of how meteorological satellite observations are used in modern numerical weather prediction.
Through lectures and practical sessions, participants will explore the principles of satellite data assimilation, observation operators and radiative transfer, alongside the use of data from a wide range of passive and active satellite instruments. The programme also includes an introduction to machine learning applications using satellite observations for weather prediction.
The course is aimed primarily at early-career scientists, including PhD students with relevant experience, as well as others with a suitable background in satellite data assimilation.
Application deadline: 27 September 2026
Full details and how to apply on the EUMETSAT/ECMWF NWP-SAF satellite data assimilation course webpage.
Use and interpretation of ECMWF products
25–28 January
Designed for operational weather forecasters and others who work directly with ECMWF forecast products, this four-day course focuses on understanding, interpreting and making effective use of ECMWF forecast products.
Participants will explore key elements of the forecasting system, analyse real forecast case studies and gain experience using applications such as ecCharts.
Topics include forecast verification, model physics, machine learning models for operational forecasting, and products such as the Extreme Forecast Index (EFI), Shift of Tails (SoT) and ecPoint.
To help participants prepare for the in-person training, a series of pre-course activities is required.
Application deadline: 27 September 2026
Full details and how to apply on the Use and interpretation of ECMWF products course webpage.
Further ECMWF training resources
Explore the ECMWF training catalogue for courses, tutorials and information on upcoming online and in-person training opportunities.