Deep Learning and Foundation Models for Weather Prediction: A Survey

TMLR Paper4075 Authors

Abstract: Conventional numerical weather prediction models are computationally expensive, while deep learning approaches often offer faster, and sometimes more accurate predictions. However, challenges do persist. This paper presents a comprehensive survey of recent deep learning and foundation models for weather prediction. We propose a taxonomy to classify existing models based on their training paradigms: deterministic predictive learning, probabilistic generative learning, and pre-training and fine-tuning. For each paradigm, we delve into the underlying model architectures, address major challenges, and offer key insights. Furthermore, we explore related applications of these methods and provide a curated summary of open-source code repositories and widely used datasets, aiming to bridge research advancements with practical implementations. Finally, we propose potential future research directions in this fast-growing field. The related sources are anonymously available at https://anonymous.4open.science/r/Survey.

Submission Length: Long submission (more than 12 pages of main content)

Assigned Action Editor: ~Binhang_Yuan1

Submission Number: 4075