Interdisciplinary Institute in Artificial Intelligence

The goal of SBI4C is to develop machine learning methods to better understand and model physical phenomena, with a specific focus on climate sciences and inverse problems that arise in this context. Simulation-based inference (SBI) enables the estimation of nonlinear model parameters from observational data, which are useful, for example, when tuning the parameterizations of atmospheric and oceanic models. However, current SBI methodology has been limited to small-scale models, and our project seeks to extend this framework. We will address two main challenges: developing new SBI algorithms that account for the computational cost of simulations and proposing strategies to generate data using efficient emulators. We expect methodological and software contributions, applications to climate models with experiments on supercomputers, and technology transfers to industrial partners.