author/Co-Presenter
Chris Stewart


ECMWF is constantly evolving their training offerings to provide the meteorological community with information to understand, access, and use ECMWF forecast products and services.

Machine Learning in Weather & Climate MOOC

Due to the rise of machine learning in meteorology and climate, in January 2023, ECMWF launched its first ever Massive Open Online Course (MOOC) on Machine Learning in Weather & Climate (MLWC). Through collaboration with iFAB and the University of Luxembourg Competence Centre (ULCC) a three Tier MOOC was developed with input from experts in the field. The MOOC featured a mixture of learning activities from videos and podcasts to forums, e-learning and interactive Jupyter notebooks.

The MOOC ran live from 09/01/2023 to 30/04/2023 and contained 40 hours of learning content across three Tiers:

  1. Machine Learning in weather & climate
  2. Concepts of Machine Learning
  3. Practical Machine Learning applications in weather and climate

Key figures:

  • 9026 participants
  • 6021 certificates delivered (participants recieved a certificate for each Tier they completed)
  • 159 countries were represented by participants
  • 4.3/5 overall satisfaction rate
  • 60 experts contributed to the MOOC

Final report for the Machine Learning in Weather & Climate MOOC: https://etrp.wmo.int/pluginfile.php/67303/mod_page/content/7/Final%20report%20for%20the%20ECMWF%20Machine%20Learning%20in%20Weather%20and%20Climate%20MOOC.pdf

The content of the MOOC is still available on the ECMWF Learning Platform (learning.ecmwf.int) however certificates for completion can no longer be recieved, they were only available if the MOOC was participated in during it's 'live' dates.

Lessons learned:

  • The undertaking of creating a MOOC is a massive task requiring substantial resource and effort
  • The partnership with iFAB and ULCC was very positive and helped to facilitate the MOOC development
  • This MOOC was probably more like three MOOCs in one due to it's size
  • Based on feedback Jupyter notebook exercises could have been improved

Future plans:

  • Due to the pace of Machine Learning the content of the MOOC will need to be reviewed frequently
  • ECMWF is looking at developing a series of mini-MOOCs in the future

New Learning Management System

In July 2023, ECMWF moved to a new Learning Management System (LMS) utilising the Moodle Platform developed by University of Luxembourg Competence Centre (ULCC) after the success of the MOOC. The new LMS (learning.ecmwf.int) combines Forecasting, Research, Software, Data and Computing and Copernicus Climate Change Services (C3S) e-learning modules into one place. The aim is for the LMS to grow and eventually encompass all ECMWF learning resources. Moodle was favoured for the new LMS as it will allow for interoperability with LMSs of other organisations with which ECMWF collaborates.

ECMWF Learning Platform: https://learning.ecmwf.int/









Modifié le: mardi 7 novembre 2023, 05:16