SeasonCast: A Masked Latent Diffusion Model for Skillful and Scalable Subseasonal-to-Seasonal Prediction

Accurate weather prediction on the subseasonal-to-seasonal (S2S) scale is critical for anticipating and mitigating the impacts of climate change. However, existing data-driven methods struggle beyond the medium-range timescale due to error accumulation in their autoregressive approach. In this work, we propose SeasonCast, a scalable and skillful probabilistic model for S2S prediction. SeasonCast consists of two components, a VAE model that encodes raw weather data into a continuous, lower-dimensional latent space, and a diffusion-based transformer model that generates a sequence of future latent tokens given the initial conditioning tokens. During training, we mask random future tokens and train the transformer to estimate their distribution given conditioning and visible tokens using a per-token diffusion head. During inference, the transformer generates the full sequence of future tokens by iteratively unmasking random subsets of tokens. This joint sampling across space and time mitigates compounding errors from autoregressive approaches. The low-dimensional latent space enables modeling long sequences of future latent states, allowing the transformer to learn weather dynamics beyond initial conditions. SeasonCast performs competitively with leading probabilistic methods at the medium-range timescale while being $10\times$ to $20\times$ faster, and achieves state-of-the-art performance at the subseasonal-to-seasonal scale across accuracy, physics-based, and probabilistic metrics.