ELLIS Unit Jena – kick off

ELLIS Unit Jena Kick-off

The ELLIS Kick-Off Event, held on January 12, 2023, marked an exciting beginning for a new chapter in European AI research, with a specific focus on Earth. Bringing together leading scientists, industry experts, and stakeholders from various fields, the event showcased the vision and goals of the ELLIS initiative. Participants engaged in insightful discussions, networking opportunities, and collaborative brainstorming sessions, all aimed at fostering innovation and advancing the field of machine learning in relation to Earth System Sciences. The kick-off not only highlighted the commitment to excellence in AI research but also emphasized the importance of interdisciplinary collaboration for tackling complex environmental challenges. This event set the stage for ELLIS to become a pivotal player in shaping the future of artificial intelligence in Europe, particularly in addressing critical issues related to our planet’s systems.

PROGRAM

8:45 – 9:15 | Arrival at Rosensäle Jena (Fürstengraben 27, 07743 Jena)

Moderator
Prof. Dr. Alexander Brenning | Friedrich Schiller University Jena, ELLIS Unit Jena Member

9:15 – 9:20 | Welcome notes
Dr. Conrad Philipp
ELLIS Unit Jena Coordinator

9:20 – 9:30 | Welcome address
Prof. Dr. Markus Reichstein
Director of the Max Planck Institute for Biogeochemistry, ELLIS Unit Jena Director
Welcome & Introduction, or: Why are we here?

9:30 – 9:50 | Welcome speeches
Dr. Katja Böhler
Thuringia’s State Secretary for Research, Innovation and Economic Development

Prof. Dr. Walter Rosenthal
President of the Friedrich Schiller University Jena

Peter Höhne
General Manager of the Chamber of Commerce and Industry East Thuringia

9:50 – 10:00 | Webinar launch at Unites Nations platform AI4Good
Dr. Reinhard Scholl | Unites Nations platform AI4Good at the International Telecommunication Union
Accelerating the United Nations Sustainable Development Goals

10:00 – 10:15 | Coffee Break

Moderator
Prof. Dr. Markus Reichstein | Director of the Max Planck Institute for Biogeochemistry, ELLIS Unit Jena Director

10:15 – 11:00 | Keynote speech
Prof. Dr. Bernhard Schölkopf | Director of the Max Planck Institute for Intelligent Systems, Tübingen

11:00 – 11:15 | Research direction
Prof. Dr. Joachim Denzler | Friedrich Schiller University Jena, ELLIS Unit Jena Director
Causal Inference as Driver for Understanding Spatio-Temporal Earth and Environmental Systems

11:15 – 11:45 | Future thoughts
Dr. Shijie Jiang | Helmholtz Centre for Environmental Research (UFZ), Leipzig
Knowledge Integration for Spatiotemporal Environmental Modeling

11:45 – 12:45 | Lunch break and networking

12:45 – 14:15 | Scientific talks session 1

Moderator
Dr. Nuno Carvalhais | Max Planck Institute for Biogeochemistry, ELLIS Unit Jena Member

Prof. Dr. Jonas Peters | University of Copenhagen
How causality can benefit from the ELLIS unit Jena

Prof. Dr. Ribana Roscher | Research Center Jülich
Generating the Unseen and Explaining the Seen

Prof. Dr. Jürgen Gall | University of Bonn
From Efficient 3D CNNs and Transformers to Applications in Agriculture and Wildfire Forecasting

14:15 – 15:00 | Break and Poster session

Optimization for Machine Learning
Matthias Mitterreiter and Joachim Giesen

Why Capsule Neural Networks Do Not Scale: Challenging the Dynamic Parse-Tree Assumption
Matthias Mitterreiter, Marcel Koch, Joachim Giesen and Sören Laue

Rapid landslide hazard assessment using active and transfer learning
Zhihao Wang and Alexander Brenning

Geospatial Machine Learning for Earth Surface Processes and Natural Hazards
Prof. Dr. Alexander Brenning

Causal Discovery in Non-linear Time-series using Model Invariance via Knockoff Interventions
Wasim Ahmad, Maha Shadaydeh and Joachim Denzler

Understanding the effects of plant diversity on soil temperature stability
Gideon Stein, Yuanyuan Huang, Maha Shadaydeh, Anne Ebeling, Nico Eisenhauer and Joachim Denzler

Fine-grained Recognition and Continuous Learning for Biodiversity Monitoring
Paul Bodesheim, Daphne Auer, Julia Böhlke, Dimitri Korsch, Matthias Körschens and Joachim Denzler

Investigating Neural Network Training on a Feature Level using Conditional Independence
Niklas Penzel, Christian Reimers, Paul Bodesheim and Joachim Denzler

Hybrid Modelling of Land-Atmosphere Fluxes: Inferring Leaf Stomatal Regulation of Evapotranspiration and Photosynthesis
Reda ElGhawi, Basil Kraft, Christian Reimers, Markus Reichstein, Marco Körner, Pierre Gentine and Alexander J. Winkler

Forest Disturbances Classification
Franziska Müller, Laura Häbold, Vitus Benson, Markus Reichstein and Ana Bastos

Necessary and sufficient graphical conditions for optimal adjustment sets in causal graphical models with hidden variables
Jakob Runge

Conditional Independence Testing with Heteroskedastic Data and Applications to Causal Discovery
Wiebke Günther, Urmi Ninad, Jonas Wahl and Jakob Runge

Time Series Causal Link Estimation under Hidden Confounding using Knockoff Interventions
Violeta-Teodora Trifunov, Maha Shadaydeh and Joachim Denzler

An objective estimate of water stress (- going beyond PDSI)
Wenli Zhao, Alexander J. Winkler, Markus Reichstein, Rene Orth and Pierre Gentine

Data-driven modelling of canopy greenness dynamics accounting for meteorological memory effects on vegetation phenology
Guohua Liu, Alexander J. Winkler, Christian Reimers and Mirco Migliavacca

Learning to forecast vegetation greenness at fine resolution over Africa with ConvLSTMs
Claire Robin, Christian Requena-Mesa, Vitus Benson, Lazaro Alonso, Nuno Carvalhais and Markus Reichstein

15:00 – 16:00 | Scientific talks session 2

Moderator
Dr. Paul Bodesheim | Friedrich Schiller University Jena

Dr. Georg Martius | Max Planck Institute for Intelligent Systems, Tübingen
Model-based Reinforcement Learning

Prof. Dr. Gustau Camps-Valls | University of Valencia
Gaussianizing the Earth

16:00 – 16:30 | Break and Poster session

16:30 – 18:00 | Flashlight talks

Moderator
Dr. Ana Bastos | Max Planck Institute for Biogeochemistry, ELLIS Unit Jena Member

Prof. Dr. Christian Wirth | University of Leipzig
AI for understanding biodiversity-climate feedbacks – a wish list

Dr. Fernando Iglesias-Suarez | German Aerospace Center (DLR)
Causally-informed deep learning to improve climate models and projections

Dr. Ankit Kariryaa | University of Copenhagen
Deep learning for continental scale mapping of trees and biomass

Dr. Teja Kattenborn | University of Leipzig
Tracking plant properties from local to global scales with plant photographs

Konrad Heidler | Technical University Munich
AI for cold regions

Prof. Dr. Hanna Meyer | University of Münster
Machine learning-based global maps of ecological variables and the challenge of assessing them

Francesco Martinuzzi | University of Leipzig
Learning Biosphere Response to Climate Drivers Using Echo State Observers

Prof. Dr. Richard Turner | University of Cambridge
Neural Process for Environmental Research

18:15 – 19:00 | Capstone talk

Moderator
Prof. Dr. Markus Reichstein | Director of the Max Planck Institute for Biogeochemistry, ELLIS Unit Jena Director

Prof. Dr. Yoshua Bengio | Université de Montréal
Generative Models and Experimental Design in the Fight Against Climate Change

19:00 | Closing remarks
Prof. Dr. Markus Reichstein
Director of the Max Planck Institute for Biogeochemistry, ELLIS Unit Jena Director

19:30 | Beverages & snacks at exhibition „I AM A.I. – explaining artificial intelligence“
Campus Ernst-Abbe-Platz, Foyer, Carl-Zeiß-Str. 3, 07743 Jena