Workshop & Hackathon on

Machine Learning for the Earth System 

25-29 August 2025

Bonn, Germany

Workshop

Hackathon

Registration

Programme

Accommodation

Important Note: Please note that information about the workshop for all registered participants has been sent via one the following email addresses mlesm-workshop@cesoc.net or mlesm-registrations@cesoc.net.

If you have not received the email, please check your spam/junk folder or contact us.

All further communication will be sent from these two addresses, so please make sure to mark them as safe (not spam/junk).

Workshop Programme 25 – 27 Aug. 2025

All times are in CEST (UTC+2)

For the abstracts, please click the respective title.

Monday 25 Aug.:

08:30 – 09:00 Registration 

9:00-10:30 NWP

 

11:00 – 11:55 NWP + Evaluation poster session

  • Lightning talks (15 mins)
  • Poster viewing

 

11:55 – 12:30 Evaluation

12:30 – 13:30 Lunch Break

13:30 – 15:00 Earth system components

15:30 – 16:30 Earth system component posters

  • Lightning talks (20 mins)
  • Poster viewing

 

16:30 – 17:30: Evaluation + NWP

19:00 – : Workshop Dinner @ Tuscolo Münsterblick (Gerhard-von-Are-Straße 8, 53111 Bonn)

————————————

 

Tuesday 26 Aug.:

9:00 – 10:30 Observations

 

11:00 – 12:00 Observation + Evaluation posters

  • Lightning talks (20 mins)
  • Poster viewing

 

12:00 – 12:30 Datasets

12:30 – 13:30 Lunch Break

 

13:30 – 15:00 Foundation Models and NWP

 

15:30 – 16:30 Evaluation + Climate Posters

  • Lightning talks (20 mins)
  • Poster viewing

 

16:30 – 17:30: NWP +  Observations

————————————

 

 

Wednesday 27 Aug.:

9:00 – 10:30 Climate

11:00 – 12:00 Observation + Earth system components

12:00 – 12:15 Closing remarks

————————————-

Important Note: Long talk = 20 min , Regular talk = 10 min

 

List of Posters

Posters sessions:

S1: Monday 25 Aug. 11:00 – 11:55                                                   S2: Monday 25 Aug. 15:30 – 16:30

S3: Tuesday 26 Aug. 11:00 – 12:00                                                  S4: Tuesday 26 Aug. 15:30 – 16:30

Poster Session Paper Title Primary Author
S1  Towards ECMWF’s data-driven flood forecasting system Maria Luisa Taccari
S1 Regional ensemble weather forecasting using a stretched-grid approach over Western Europe Serban Vadineanu
S1 Exploring Explainability for Graph-Based Weather Forecasting Models Using Layer-Wise Relevance Propagation Jens Pruschke
S1 Comparison of Distribution Calibration in Machine Learning and Numerical Weather Prediction Models Kornelius Raeth
S1 Fixing the Double Penalty in Data-Driven Weather Forecasting Through a Modified Spherical Harmonic Loss Function Leo Separovic
S1 Transformer-Based Short-term Precipitation Post-Processing: Leveraging Extensive Data and Short Training Minchan Jeong + Cho
S1 Advancing Heatwave Prediction with the Aurora Foundation Model: Insights from India and Beyond Seshagirirao Kolusu
S2 Improved Neural Network Modeling of Plasmasphere Using Arase and RBSP Data Sadaf Shahsavani
S2 Emulating Significant Wave Height Forecasts Using Deep Learning Katherine Haynes
S2 Autoregressive denoising diffusion for predicting trajectories of floating objects in oceans Christian Donner
S2 A machine learning approach to recover cloud microphysical process rates from ICON model simulations with the two-moment microphysics scheme Miriam Simm
S2 Unlocking the Future of Dry Intrusion Outflows with Deep Learning Owain Harris
S2 Sea-Ice Simulation and Ocean-Coupled Processes in a Deep Learning Earth-System Model Dale Durran
S2 Towards high-resolution land surface temperature forecasting for early detection of extreme events Marieke Wesselkamp
S2 Bias-Correcting Arctic ERA5 Surface Air Temperatures using Deep Learning Sabine Scholle
S2 Machine Learning-Based Reservoir Storage Forecasting for Optimal and Climate-Resilient Operation of Multi-reservoir systems in the Blue Nile River Basin. Selam Sahlu
S3 On the verification of weather forecasts for extremes: a statistical review Romain Pic
S3 Leveraging machine learning to advance the assimilation of new types of satellite observations to better constraint the land carbon cycle in the ECMWF Integrated Forecast System (IFS) Sebastien Garrigues
S3 Leveraging NOAA’s Joint Effort for Data assimilation Integration (JEDI) to force FourCastNet with Tomorrow.io’s microwave sounder data Brandon Taylor
S3 Convolutional Neural Network for the Assimilation of Diurnal Satellite Retrievals of Sea Surface Temperature in Ocean Reanalysis and Forecast Systems Matteo Broccoli
S3 Are observations all you need? Peter Lean
S3 Out-of-sample Heat Extremes in Aurora George Jordan
S3 On simulating the 2019 European summer heatwave event using data-driven weather models Prabhakar Namdev
S3 Learning to Faithfully Compress Sentinel-2 Satellite Data using Vector-Quantized Autoencoders Sebastian Hoffmann
S4 HClimRep: A Foundation Model for Capturing the Atmosphere, Ocean, and Sea Ice Interactions Ankit Patnala
S4 Harnessing Machine Learning for Climate Modeling: Towards the Development of a DestinE Climate Emulator Fernando Iglesias-Suarez
S4 A machine learning emulator for paleoclimate: Using neural network to generate precipitation climatology in Southwest Asia Trang Nguyen
S4 Uncertainty-aware Masked Autoencoders for predicting El Niño Southern Oscillation dynamics Jannik Thuemmel
S4 Sensitivity of Deep learning El Nino Southern Oscillation forecasts to model training Jinhao Wu
S4 Hierarchical Graph Diffusion Networks for ERA5 Reconstruction and Long-Term Climate Forecasting Nishant Kumar
S4 Verification of AI models – Adaptation and extension of an operational verification framework as part of the RAINA project Britta Seegebrecht

 

Questions?

📩 Contact us at: mlesm-workshop@cesoc.net

We gratefully acknowledge the support by
CESOC, ECMWF, the TRA Modelling at the University of Bonn, the Forschungszentrum Jülich and the University of Cologne.

 

Supported by