Google DeepMind / WeatherNext

This repo contains the code for WeatherNext 2 (WN2), the global, medium-range atmospheric and cyclone forecasting model developed by Google DeepMind and Google Research. It also contains code for prior generation models GraphCast and GenCast.

该仓库包含了由 Google DeepMind 和 Google Research 开发的全球中程大气及气旋预报模型 WeatherNext 2 (WN2) 的代码。它同时也包含了前代模型 GraphCast 和 GenCast 的代码。

Accessing Forecast Data Feeds

If you are interested in directly accessing daily data feeds of WN2 model outputs rather than running the model yourself, we provide them across multiple platforms:

• Google Cloud (including Earth Engine, BigQuery, and Vertex AI).
• WeatherLab (including cyclone tracks).
• OpenMeteo (including an API and interactive builder).

访问预报数据源

如果您希望直接获取 WN2 模型的每日预报数据源，而不是自行运行模型，我们通过多个平台提供这些数据：

• Google Cloud（包括 Earth Engine、BigQuery 和 Vertex AI）。
• WeatherLab（包括气旋路径）。
• OpenMeteo（包括 API 和交互式构建器）。

Learn More

• Model Guide & Documentation: Google Developers WeatherNext Guide
• WeatherNext Cyclones Paper: Operational tropical cyclone forecasting with AI
• FGN/WN2 Technical Report: Skillful joint probabilistic weather forecasting from marginals
• WeatherNext 2 Blog Post: WeatherNext 2: Our most advanced weather forecasting model
• WeatherNext Cyclones Blog Post: WeatherNext: AI model achieves breakthrough in forecasting cyclones

了解更多

• 模型指南与文档： Google Developers WeatherNext 指南
• WeatherNext 气旋论文： 利用人工智能进行业务化热带气旋预报
• FGN/WN2 技术报告： 基于边缘分布的高效联合概率天气预报
• WeatherNext 2 博客文章： WeatherNext 2：我们最先进的天气预报模型
• WeatherNext 气旋博客文章： WeatherNext：AI 模型在气旋预报领域取得突破

Older Models

This repository serves as the primary home for the WeatherNext family models. Alongside WN2, this repository also hosts the code and documentation for our legacy and specialized models:

• WeatherNext Graph: Deterministic medium-range weather forecasting using graph neural networks. Published as GraphCast.
• WeatherNext Gen: Diffusion-based ensemble forecasting for medium-range weather. Published as GenCast.

旧版模型

本仓库是 WeatherNext 系列模型的主要存储库。除了 WN2，该仓库还托管了我们遗留模型和专用模型的代码与文档：

• WeatherNext Graph： 使用图神经网络进行确定性中程天气预报。即已发布的 GraphCast。
• WeatherNext Gen： 基于扩散模型的中程天气集合预报。即已发布的 GenCast。

Provided Pretrained Models

This repository provides code to run the different versions of WeatherNext 2 and WeatherNext Cyclones. The only difference between them is that WN2 can also predict 100m wind. In particular, WN2 also forecasts cyclones with the exact same algorithm as WN Cyclones. Their weights are different due to independent training runs.

提供的预训练模型

本仓库提供了运行不同版本 WeatherNext 2 和 WeatherNext Cyclones 的代码。它们之间唯一的区别在于 WN2 还可以预测 100 米风速。特别地，WN2 使用与 WN Cyclones 完全相同的算法来预报气旋。由于训练过程独立，它们的权重参数有所不同。

WeatherNext 2

• WeatherNext2_<2025 (Used Operationally): 0.25° resolution (~30km). Fine-tuned on ECMWF HRES data and designed to be initialized directly from operational HRES initial conditions rather than ERA5 reanalysis. Trained on data through 2024. Corresponding weights files: WeatherNext2_<2025_model{1,2,3,4}.npz.

WeatherNext 2

• WeatherNext2_<2025（业务化使用）： 0.25° 分辨率（约 30 公里）。在 ECMWF HRES 数据上进行了微调，旨在直接从业务化 HRES 初始条件而非 ERA5 再分析数据进行初始化。使用截至 2024 年的数据进行训练。对应的权重文件：WeatherNext2_<2025_model{1,2,3,4}.npz。

WeatherNext Cyclones - models which reproduce the results in paper

• WeatherNextCyclones_<2025 (Used Operationally): 0.25° resolution. The model that ran live during the 2025 Atlantic hurricane season, publicly referred to as FNV3 (NHC’s postprocessed version was called GDMI). Trained on data through 2024. Corresponding weights files: WeatherNextCyclones_<2025_model{1,2,3,4}.npz.
• WeatherNextCyclones_<2024: 0.25° resolution. Reproduces results from the paper on 2024. Trained on data through 2023. Corresponding weights files: WeatherNextCyclones_<2024_model{1,2,3,4}.npz.
• WeatherNextCyclones_<2023: 0.25° resolution. Reproduces results from the paper on 2023. Trained on data through 2022. Corresponding weights files: WeatherNextCyclones_<2023_model{1,2,3,4}.npz.

WeatherNext Cyclones - 复现论文结果的模型

• WeatherNextCyclones_<2025（业务化使用）： 0.25° 分辨率。该模型在 2025 年大西洋飓风季节期间实时运行，公开称为 FNV3（NHC 的后处理版本称为 GDMI）。使用截至 2024 年的数据进行训练。对应的权重文件：WeatherNextCyclones_<2025_model{1,2,3,4}.npz。
• WeatherNextCyclones_<2024： 0.25° 分辨率。复现了论文中 2024 年的结果。使用截至 2023 年的数据进行训练。对应的权重文件：WeatherNextCyclones_<2024_model{1,2,3,4}.npz。
• WeatherNextCyclones_<2023： 0.25° 分辨率。复现了论文中 2023 年的结果。使用截至 2022 年的数据进行训练。对应的权重文件：WeatherNextCyclones_<2023_model{1,2,3,4}.npz。

WeatherNext Cyclones Mini

• WeatherNextCyclones_Mini_<2024: 1° resolution. A lightweight version suitable for lower memory and compute constraints. Not expected to match the performance of the larger versions. Trained on data through 2023. Corresponding weights file: WeatherNextCyclones_Mini_<2024.npz.
• WeatherNextCyclones_Mini_<2023: As above, but only trained on data through 2022. Corresponding weights file: WeatherNextCyclones_Mini_<2023.npz.

WeatherNext Cyclones Mini

• WeatherNextCyclones_Mini_<2024： 1° 分辨率。轻量级版本，适用于内存和计算资源受限的环境。性能预计无法达到大型版本水平。使用截至 2023 年的数据进行训练。对应的权重文件：WeatherNextCyclones_Mini_<2024.npz。
• WeatherNextCyclones_Mini_<2023： 同上，但仅使用截至 2022 年的数据进行训练。对应的权重文件：WeatherNextCyclones_Mini_<2023.npz。

Quick Start Guide

The easiest way to get started with WeatherNext 2 is by running our interactive Colab Notebook. In general, we recommend running WeatherNext 2 on TPU where possible. If choosing to run on GPU, the attention implementation must be switched. The non-Mini models require H100 for sufficient VRAM. The Mini models should manage inference on a P100.

快速入门指南

开始使用 WeatherNext 2 最简单的方法是运行我们的交互式 Colab Notebook。通常情况下，我们建议尽可能在 TPU 上运行 WeatherNext 2。如果选择在 GPU 上运行，则必须切换注意力机制的实现方式。非 Mini 版本模型需要 H100 以获得足够的显存，而 Mini 版本模型在 P100 上即可完成推理。

Setup Installation

[!NOTE] This is research code provided as-is for the purpose of running and experimenting with the published models. There are no guarantees of API stability and future updates may introduce breaking changes without notice. We recommend pinning to a specific release.

安装设置

[!NOTE] 这是按原样提供的研究代码，旨在用于运行和实验已发布的模型。我们不保证 API 的稳定性，未来的更新可能会在不另行通知的情况下引入破坏性变更。我们建议锁定特定版本。