AIMET-NL 2026
AI for Weather and Climate
Explainability, Multimodal Learning, Mechanistic Interpretability, and Operational Forecasting Applications
AIMET-NL Meeting / Workshop
Date: 19 June 2026
Location: Radboud University, Nijmegen, The Netherlands
Time: 10:00 - 16:00, followed by drinks and networking
Register here
Registration deadline: Monday 15 June 2026 - Fee: €20 (free seats available for SIKS members) - See details below.
About the Workshop
On 19 June 2026 a 1-day workshop will be organized at Radboud University, Nijmegen, by AIMET-NL, the Dutch research network on AI for Weather & Climate. The topic of this meeting is Explainability, Multimodal Learning, Mechanistic Interpretability, and Operational Forecasting Applications.
The use of AI methods for weather and climate applications is developing very rapidly. Only a few years ago, fully data-based weather forecasts made with AI/ML methods seemed like a remote possibility. Nowadays weather forecasting systems based on AI have reached a level of predictive skill that is competitive in many respects with state-of-the-art physics-based weather forecasting models. More broadly, researchers are quickly tapping into the potential of AI for advancing modelling and analysis of weather and climate.
Reflecting these developments, the AIMET-NL network was initiated in 2025 by a group of researchers based in the Netherlands. This edition at Radboud University will bring together the research community around explainability, mechanistic interpretability, multimodal learning, and transformer architectures for Earth system applications, with a complementary focus on the operational use of weather forecasts in the energy sector. The workshop aims to stimulate interaction between meteorologists, climate scientists, AI/ML researchers, and industry partners, and to foster collaborations within the Dutch AI for weather and climate community.
We kindly ask you to confirm your attendance by registering no later than Monday 15 June 2026. The registration fee for the workshop is 20 euros. We also note that the number of participants is limited due to room capacity constraints, so early registration is recommended.
The tentative schedule for the workshop is below.
This edition is organised in cooperation with SIKS, the Dutch Research School for Information and Knowledge Systems.
Organizers: Yuliya Shapovalova & Charlotte Cambier van Nooten
Confirmed Speakers
- Tom Heskes (Radboud University)
- Ana Lucic (University of Amsterdam)
- Yuliya Shapovalova (Radboud University)
- Wouter Tromp and Simon Veldkamp (TenneT)
- Ment Reeze and Jessica Loke (Alliander)
- Representative of Beyond Weather (TBA)
Preliminary Program
| Time | Session |
|---|---|
| 09:00 | Welcome & Coffee |
| 09:40 | Opening |
| 09:45 | Yuliya Shapovalova (RU) - Multimodal Learning and Gradient Starvation |
| 10:20 | Wouter Tromp & Simon Veldkamp (TenneT), Ment Reeze & Jessica Loke (Alliander) - Operational Challenges of Using Weather Forecasts |
| 10:55 | Coffee break |
| 11:20 | Panel discussion - Operational challenges of producing and using weather forecasts (KNMI, Alliander, TenneT) |
| 12:00 | Lunch |
| 13:00 | Tom Heskes (RU) - Explainability and Causal Shapley Values |
| 14:05 | Ana Lucic (UvA) - Mechanistic Interpretability for Scientific Machine Learning |
| 14:40 | Coffee break |
| 15:10 | Community discussion, posters, demos |
| 16:00 | Drinks, networking |
The program will be updated as additional details are confirmed.
Registration
Please confirm your attendance by registering no later than Monday 15 June 2026.
Registration link: Register here
Registration fee: €20
Due to the cooperation between SIKS and the local organisation of AIMET-NL, there will be a fixed number of free seats available for SIKS members (PhD students and staff). Applications will be honoured on a first-come-first-serve basis.
The number of participants is limited due to room capacity, so early registration is recommended. Participants interested in presenting a poster or demo are encouraged to indicate this during registration.
After registration, you will receive a separate email from Radboud University with a payment link (PayByLink). Available payment methods include credit card, PayPal, and iDEAL.
Practical Information
- Dietary requirements: Please indicate any dietary requirements during registration. Note that the venue catering cannot accommodate halal or kosher diets.
- Posters and demos: If you would like to present, indicate this on the registration form.
Contact
For questions, please contact: yuliya.shapovalova@ru.nl