Fast, Scale-Adaptive, and Uncertainty-Aware Downscaling of Earth System Model Fields with Generative Foundation Models

arXiv:2403.02774v1 Announce Type: cross
Abstract: Accurate and high-resolution Earth system model (ESM) simulations are essential to assess the ecological and socio-economic impacts of anthropogenic climate change, but are computationally too expensive. Recent machine learning approaches have shown promising results in downscaling ESM simulations, outperforming state-of-the-art statistical approaches. However, existing methods require computationally costly retraining for each ESM and extrapolate poorly to climates unseen during training. We address these shortcomings by learning a consistency model (CM) that efficiently and accurately downscales …

abstract art arxiv change climate climate change cs.cv cs.lg earth economic esm fields foundation generative impacts machine machine learning physics.ao-ph physics.geo-ph results scale simulations state statistical type uncertainty