S2S-JEPA: Predicting the Predictable at Subseasonal-to-Seasonal Timescales
Not on the map. Why?
Abstract (arXiv)
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.
Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 26 of its 46 references that are in the map. First posted on arXiv 2 Oct 2026.
Builds on
26 of 46 references.
References · 26
- 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
- 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
- 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
- 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 3,045
- 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,333
- 2023Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecturecited by 1,039
- 2025V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planningcited by 809
- 2022FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operatorscited by 776
- 2022Message Passing Neural PDE Solverscited by 386
- 2023ClimaX: A foundation model for weather and climatecited by 337
- 2026LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixelscited by 273
- 2026VLA-JEPA: Enhancing Vision-Language-Action Model with Latent World Modelcited by 118
14 more
- 2023Scaling transformer neural networks for skillful and reliable medium-range weather forecastingcited by 110
- 2021Sub-seasonal forecasting with a large ensemble of deep-learning weather prediction modelscited by 79
- 2024ChaosBench: A Multi-Channel, Physics-Based Benchmark for Subseasonal-to-Seasonal Climate Predictioncited by 43
- 2026Hierarchical Planning with Latent World Modelscited by 43
- 2024Prithvi WxC: Foundation Model for Weather and Climatecited by 31
- 2022Adaptive Bias Correction for Improved Subseasonal Forecastingcited by 19
- 2025OmniCast: A Masked Latent Diffusion Model for Weather Forecasting Across Time Scalescited by 12
- 2025Towards mechanistic understanding in a data-driven weather model: internal activations reveal interpretable physical featurescited by 12
- 2025Utilizing Strategic Pre-training to Reduce Overfitting: Baguan -- A Pre-trained Weather Forecasting Modelcited by 7
- 2025CondensNet: Enabling stable long-term climate simulations via hybrid deep learning models with adaptive physical constraintscited by 6
- 2026Can AI Weather Models Predict Beyond Two Weeks? A Quantitative Benchmark and Analysis of Long Rolloutscited by 4
- 2025XAI4Extremes: An interpretable machine learning framework for understanding extreme-weather precursors under climate changecited by 3
- 2025TianQuan-S2S: A Subseasonal-to-Seasonal Global Weather Model via Incorporate Climatology Statecited by 3
- 2026Tracing the Unlabeled Storm: Cross-Variable Transfer in a Lagrangian Atmospheric JEPA Frameworkcited by 1
Built on by
Not cited yet.