Destination Earth ML Modelling Retreat

Day 

Time 

Session 

Wednesday (26 March) 

14:00 – 14:30 

Participants arrive 

14:30 – 15:30 

Icebreaker activities in small groups 

15:30 – 16:00 

Coffee break 

16:00 – 17:30 

Walking Dialogues: AI and the Future of Earth System Modelling 

17:30 – 18:30 

Scene setting talks   

• Irina Sandu - Destination Earth  

• Jorn Kristensen – Why are we here?  

• Matthew Chantry - ML Overview  

• Sara Hahner - Earth system ML modelling in Destination Earth 

19:00 

Dinner 

Thursday (27 March) 

09:00 – 10:15 

Oral session 1: Probabilistic Predictions and Weather Extremes  

Chair Mariana Clare  

• Data-driven modelling at MET Norway and in DE330 - Thomas Nipen (20 min) 

• Leveraging deep learning for probabilistic regional forecasting - Laure Raynaud (20 min) 

• AI activity at DWD and by E-AI - Roland Potthast (20 min) 

• 15 min discussion led by Mariana Clare  

10:15 – 10:45 

Coffee break 

10:45 – 12:00 

Walking Dialogues: AI and the Future of Earth System Modelling  

12:00 – 13:30 

Lunch 

13:30 – 15:00 

Oral session 2: Earth system modelling across Europe  

Chair Nina Raoult 

• MLWP modelling activities at MeteoSwiss - Oli Fuhrer (15 min) 

• Multi-domain dynamical graph training and other progress at KNMI towards a data-driven weather model - Sophie Buurman (15min) 

• Catching the Wind: RMI’s experience with anemoi-LAM - Michiel Van Ginderachter (15min) 

• Status and outlook on data-driven limited area forecasting at DMI - Simon Kamuk Christiansen (15min) 

• 30 min discussion  
  Led by Nina Raoult 

15:30 – 16:00 

Coffee break 

16:00 – 17:00 

Breakout groups 1: Challenges towards high resolution  

Led by Christian Lessig 
 

17:15 – 18:15 

Shaping the Future of AI Earth System Modeling in Europe – A DestinE Fireside Chat 

Chair: Florian Pappenberger 

Panel: Grazyna Piesiewicz, Roar Skålin, Roland Potthast, Mark Liniger, Laure Raynaud 

19:00 

Dinner 

Friday (28 March) 

09:00 – 10:30 

Oral session 3: Machine Learning (ML) for Climate Applications  

Chair Francisco Doblas-Reyes 

• Early-Stage Development and Implementation of a Climate Emulator for DestinE  - Fernando Iglesias-Suarez & González-Abad (15min) 

• The Met Office AI4Climate programme - Doug McNeall & Kate Halladay (15 min) 

• From physics-based to AI-based storyline simulations in DestinE - Nikolay Koldunov (15min) 

• Exploring the use of AI and Machine Learning in the Climate DT - Amanda Duarte (15 min) 

• 30 min discussion led by Francisco Doblas-Reyes

10:30 – 11:00 

Coffee break 

11:00 – 12:00 

Breakout groups 2: Challenges in AI Climate modelling 

Led by Peter Dueben 

12:00 – 13:00 

Summary from breakout groups, conclusions, and next steps 

13:00 

Lunch