Artificial Intelligence (AI) and Machine Learning (ML) methods are becoming increasingly important in both science and society. In climate science - where complex biophysical and societal processes interact across diverse temporal and spatial scales, and datasets are often large, heterogenous and incomplete - AI and ML methods offer new powerful solutions.
Join us for the next AI for Climate Science Seminar!
After a short summer break, we are back in September with a guest talk by Dingling Yao from the Institute of Science and Technology Austria.
Dingling Yao is an ELLIS PhD candidate at the Institute of Science and Technology Austria, advised by Francesco Locatello, and currently a visiting researcher at Caltech with Anima Anandkumar. Her research lies at the intersection of causal representation learning, scientific machine learning, and climate science. She studies how AI systems can move beyond predictive shortcuts to recover meaningful physical structure and answer reliable what-if questions from observational data. She is a 2025 Google PhD Fellow and a 2022 Amazon–MPI Science Hub Fellow.
For online participation, a registration is necessary.
Title:
Beyond Prediction: Answering Climate What-If Questions from Observations
Abstract:
Machine learning models can now forecast complex scientific systems with remarkable accuracy. But science often asks a harder question than what happens next?: What would happen if the system were different? Answering such what-if questions from observational data is fundamentally challenging because machine learning models are inherently designed to match patterns in their training data and, by default, need not generalize when asked about unseen scenarios. Dingling will illustrate this challenge through the Perception-Physics Paradox. Using tropical cyclones as a case study, she will show that vision foundation models can predict physical quantities from satellite imagery while failing to represent the underlying physical state—especially for intense storms, precisely where scientific reliability matters most. She will then discuss how combining causal representation learning with dynamical-system identification can help recover scientifically meaningful latent parameters from observed trajectories, enabling downstream causal analyses of real-world sea-surface temperatures. Finally, she will extend this perspective to high-dimensional gridded climate emulators through ReRoute, an mechanism-guided approach for answering unseen CO₂ what-if questions, without requiring controlled perturbation experiments.
Together, these works point toward a path beyond pattern matching: building scientifically grounded AI systems that can use observational data not only to predict the world as it is, but also to reason reliably about how it could be.
The monthly AI seminar at IIASA features global experts in the field of AI and ML who will showcase the newest methodological advancements and applications in the field. Through a series of invited talks, the seminar showcases cutting edge research with the aim of strengthening AI and ML expertise at IIASA and to foster external collaborations. Additionally, it serves as an institute-wide platform for discussions and knowledge exchange across programs and working groups to spark new ideas and innovations.
As an initiative from the ECE/ ICI Theme on Extreme Weather and Climate Dynamics, this seminar is designed for both experts already integrating AI and ML into their workflows and those eager to expand their knowledge in these fields.
Upcoming Events
Royal Belgian institute of Natural Sciences, Rue Vautier 29, 1000 Bruxelles, Belgium
The Final Event of NaturaConnect: Pathways to a Trans-European Nature Network
Chulalongkorn University, Thailand