Generative artificial intelligence improves projections of climate extremes
This addresses the challenge of representing climate extremes for climate modeling and risk assessment, though it appears incremental as a hybrid method building on existing techniques.
The paper tackles the problem of projecting climate extremes by introducing FuXi-CMIPAlign, a generative deep learning framework for downscaling CMIP outputs, which improves accuracy, stability, and generalization across emission scenarios.
Climate change is amplifying extreme events, posing escalating risks to biodiversity, human health, and food security. GCMs are essential for projecting future climate, yet their coarse resolution and high computational costs constrain their ability to represent extremes. Here, we introduce FuXi-CMIPAlign, a generative deep learning framework for downscaling CMIP outputs. The model integrates Flow Matching for generative modeling with domain adaptation via MMD loss to align feature distributions between training data and inference data, thereby mitigating input discrepancies and improving accuracy, stability, and generalization across emission scenarios. FuXi-CMIPAlign performs spatial, temporal, and multivariate downscaling, enabling more realistic simulation of compound extremes such as TCs.