We are pleased to announce the upcoming International Training School “AI 4 Seismology”, scheduled for May 5-8, 2025. This training program aims to explore the synergies between Artificial Intelligence (AI) and Seismology, and provide a comprehensive platform for knowledge sharing and skills development.

Registration

The participation fee includes all sessions and workshops as well as lunch, snacks and the social evening. Accommodation and travel can unfortunately not be covered by the registration fee and require self-organization.

Program

The program covers a wide range of topics at the intersection of AI and seismology:

Theoretical Foundations

• Principles of Seismology and Instrumentation
• Fundamentals and Advanced Sessions of Artificial Intelligence and Machine Learning, including Transformers, Physics Informed NNs, Graph NNs, Probabilistic Geophysical Inversion and Unsupervised Learning

Advanced AI Applications in Seismology

• Earthquake Seismology, Phase Picker Algorithms, First-Motion Polarities
• UMAP
• Denoising
• Environmental Signals

Data Management and Computational Techniques

• Scientific Software Development and Management
• Open Source Methodologies
• High-Performance Computing (HPC) in Seismological Research
• Big Data Analytics for Seismic Data

Schedule

	Time	Subject	Trainer
Monday – 05.05.	14:00-15:30	Fundamentals in Earthquake / Environmental Seismology & Instrumentation	Dr. Sven Schippkus
		30-min break
	16:00-17:30	Basics in ML (supervised, unsupervised, deep learning)	Dr. Robert Haase
	17:30-18:00	Technical HPC Set-Up
	18:00+	Icebreaker
Tuesday – 06.05.	08:30-09:15	Theory I: Transformers for Time Series ML	Matthias Täschner
	09:15-10:00	Theory I – Tutorial	Matthias Täschner
		30-min break
	10:30-11:15	Theory II: Physics Informed NNs	Prof. Dr. Tarje Nissen-Meyer
	11:15-12:00	Theory II – Tutorial	Prof. Dr. Tarje Nissen-Meyer
		2-hour lunch break
	14:00-14:45	Theory III: Graph NNs	Dr. Martijn van den Ende
	14:45-15:30	Theory IV: Advanced methods in probabilistic geophysical inversion	Prof. Dr. Andrew Curtis (online)
		30-min break
	16:00-16:45	Theory V: Unsupervised Learning	Dr. René Steinmann
	16:45-17:30	Theory V – Tutorial	Dr. René Steinmann
	17:30+	Poster Session + Drinks
Wednesday – 07.05.	08:30-10:00	Applications I + II: Earthquake Seismology, Phase picking, First-Motion Polarities	Prof. Dr. Marine Denolle
		30-min break
	10:30-11:15	Applications III: UMAP	Dr. Maria Tsekhmistrenko
	11:15-12:00	Applications IV: Denoising	Dr. Han Xiao
		2-hour lunch break
	14:00-14:45	Applications V: Neural Operators	Prof. Dr. Jonathan Bedford
	14:45-15:30	Applications VI: Environmental Signals	Dr. Josefine Umlauft
		30-min break
	16:00-18:30	Optional Tutorial (“Bring your own data”)
	19:00+	Scientific Networking Event at Moritzbastei
Thursday – 08.05.	08:30-10:00	Scaling Workflows to Big Data	Elias Werner, Dr. Siavash Ghiasvand, Dr. Christoph Lehmann
	10:30-12:00	Scaling Workflows to Big Data – Tutorial	Elias Werner, Dr. Siavash Ghiasvand, Dr. Christoph Lehmann
		2-hour lunch break
	14:00-15:30	Data & Software Management, Open Source	Dr. Robert Haase

Methodological Approach

The training school will use a multifaceted approach to ensure comprehensive learning:

• Lectures: Delivered by experts, providing in-depth theoretical knowledge
• Practical Sessions: Hands-on workshops to apply theoretical concepts
• Poster Presentations: Opportunity for participants to present their research
• Scientific Networking Event: Facilitating collaboration and knowledge exchange

Target Audience

This training school is designed for graduate students, postdoctoral researchers, and professionals in the fields of seismology, geophysics, computer science, and related disciplines who wish to enhance their expertise in AI applications for seismological research.

Confirmed Trainers

Prof. Dr. Tarje Nissen-Meyer (University of Exeter, UK)
Prof. Dr. Marine Denolle (University of Washington)
Dr. René Steinmann (Deutsches GeoForschungsZentrum GFZ Potsdam, Germany)
Dr. Maria Tsekhmistrenko (University College London, UK)
Prof. Dr. Andrew Curtis (online) (University of Edinburgh, Scotland)
Dr. Han Xiao (Deutsches GeoForschungsZentrum GFZ Potsdam, Germany)
Matthias Täschner (ScaDS.AI Dresden/Leipzig)
Dr. Sven Schippkus (Universität Hamburg)
Prof. Dr. Jonathan Bedford (Ruhr Universität Bochum)
Elias Werner (Technische Universität Dresden)
Dr. Siavash Ghiasvand (Technische Universität Dresden)
Dr. Christoph Lehmann (Technische Universität Dresden)

Organizers

Dr. Josefine Umlauft (ScaDS.AI Dresden/Leipzig, Germany)
Dr. Jannes Münchmeyer (ISTerre Grenoble, France)
Dr. Martijn van den Ende (Université Côte d’Azur, France)
Dr. Marius Isken (Deutsches GeoForschungsZentrum GFZ Potsdam, Germany)
Dr. Robert Haase (ScaDS.AI Dresden/Leipzig, Germany)

Cooperation