CESOC kindly invites you to a talk given by Dr. Richard Forbes (ECMWF) with the title:
Predicting the weather at ECMWF: Improving physics-based modelling in an increasingly data-driven world
Date: 24 November 2025
Time: 10:00 (Berlin time)
Location: University of Cologne
Please use the entrance at Höninger Weg 100, 50969 Cologne. Ring the bell labeled “Hörsaal Universität Köln – Öffnungszeit 7–18 Uhr” to enter; the door will open automatically. The session takes place in Lecture Hall 4.001 on the 4th floor.
For online participation, please contact info@cesoc.net
Abstract:
Over the last 50 years at ECMWF and at meteorological services across the world, we have seen remarkable progress in physics-based numerical weather prediction (NWP). More recently, the emergence of data-driven machine learning weather prediction (MLWP) has offered a step increase in forecast skill and is now also providing weather forecasts every day (e.g. AIFS, AICON). Progress is rapid and machine learning will become an increasing part of weather forecast systems going forward, but physics-based models will continue to play a vital role. Whether this role is as a parallel reference system, part of a hybrid-ML system or providing ML training data, it is imperative to continue to push forward the improvement of the physical model.
The accurate representation of subgrid-scale physical processes remains a critical challenge for improving forecasts, and the effective use of observations is crucial to enhance the accuracy of parametrization schemes and reduce systematic errors. One example is the recent EarthCARE satellite mission with its synergistic lidar, radar, and radiometer measurements, which is providing unprecedented insights into clouds, precipitation, aerosol, radiation and their interactions.
This presentation will explore these themes and describe recent progress on the physical global NWP model (IFS) developments at ECMWF, particularly highlighting clouds and convection, and specifically how using observations from the EarthCARE satellite is helping to constrain physical processes, with benefits for the next generation of Earth-system models, and for weather prediction in whatever form it may take.
Bio: https://www.ecmwf.int/en/about/who-we-are/staff-profiles/richard-forbes