Track synthesis, materials and catalysis as it is published.
Search papers, labs, and topics across Lattice.
S2S-JEPA adapts the Joint-Embedding Predictive Architecture (JEPA) paradigm to subseasonal-to-seasonal (S2S) weather forecasting by focusing on predicting only the slowly varying, predictable components in latent space, rather than fine-scale details. This approach addresses the "predictability desert" challenge where existing AI weather models degrade beyond two weeks. S2S-JEPA demonstrates comparable skill to the ECMWF physics-based ensemble and outperforms it on key metrics at weeks 5-6.
Predicting the predictable in latent space unlocks S2S weather forecasting skill beyond current AI model limits, matching and exceeding physics-based ensembles at critical 5-6 week horizons.
The subseasonal-to-seasonal (S2S) timescale, roughly from two weeks to two months ahead, is a critical forecast window for sectors such as agriculture, energy, and water management. Yet, it is widely known as the `predictability desert'. Recent AI weather models excel up to two weeks ahead but deteriorate beyond, largely because they are trained to predict fine-scale details that are neither predictable nor essential at S2S timescales. We argue that a more physically grounded objective is to forecast only the slowly varying components that remain predictable. Computer vision reached the same conclusion with the Joint-Embedding Predictive Architecture (JEPA), which predicts in latent space, discarding unpredictable details. In this work, we introduce S2S-JEPA, which brings the JEPA paradigm to S2S forecasting. It is tailored to this task through design elements from state-of-the-art AI weather models. S2S-JEPA achieves comparable skill to the gold-standard ECMWF physics-based ensemble and surpasses it on multiple metrics at weeks 5 to 6.