WIND: Weather Inverse Diffusion for Zero-Shot Atmospheric Modeling

2026-02-03

AI Summary

Current weather and climate modeling relies heavily on numerous task-specific models, lacking a unified foundational architecture. This work proposes the first fine-tuning-free unified atmospheric foundation model: an unconditional video diffusion model pretrained via self-supervised video reconstruction, which casts diverse atmospheric modeling tasks as inverse problems and enables zero-shot multitask inference through posterior sampling. The approach matches or surpasses specialized models in probabilistic forecasting, spatiotemporal downscaling, sparse reconstruction, and conservation-law-constrained tasks, while efficiently generating physically consistent counterfactual extreme weather scenarios—such as those under global warming—thereby significantly enhancing model generalizability and scientific utility.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Neural Architectures and Foundation Models
• Natural Language Processing: (Large) Language Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Large language models for search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Deep learning has revolutionized weather and climate modeling, yet the current landscape remains fragmented: highly specialized models are typically trained individually for distinct tasks. To unify this landscape, we introduce WIND, a single pre-trained foundation model capable of replacing specialized baselines across a vast array of tasks. Crucially, in contrast to previous atmospheric foundation models, we achieve this without any task-specific fine-tuning. To learn a robust, task-agnostic prior of the atmosphere, we pre-train WIND with a self-supervised video reconstruction objective, utilizing an unconditional video diffusion model to iteratively reconstruct atmospheric dynamics from a noisy state. At inference, we frame diverse domain-specific problems strictly as inverse problems and solve them via posterior sampling. This unified approach allows us to tackle highly relevant weather and climate problems, including probabilistic forecasting, spatial and temporal downscaling, sparse reconstruction and enforcing conservation laws purely with our pre-trained model. We further demonstrate the model's capacity to generate physically consistent counterfactual storylines of extreme weather events under global warming scenarios. By combining generative video modeling with inverse problem solving, WIND offers a computationally efficient paradigm shift in AI-based atmospheric modeling.

Problem

• atmospheric modeling
• foundation model
• zero-shot learning
• weather forecasting
• climate modeling

Innovation

• foundation model
• inverse problem
• diffusion model
• zero-shot learning
• atmospheric modeling