AlphaEarth 是由 Google DeepMind 与 Google Earth Engine 团队于 2025 年发布的一个全球地理空间嵌入数据集，相关论文成果为「AlphaEarth Foundations: An embedding field model for accurate and efficient global mapping from sparse label data」，旨在把多源遥感和地理数据压缩为统一、可复用的时空嵌入，在标注稀缺条件下更高效地完成制图与监测。 该数据集由 2017–2024 的年度全球嵌入图层组成，每个 10 米像素对应 64 维单位长度向量，覆盖全球陆地与浅水区，当前公开版本标注为 AEFv2.1 。数据集全球覆盖、统一坐标系，对云、条带与缺测值具有鲁棒性，支持 空间或时间维度的线性聚合，可替代传统光谱指数或手工特征。适用于土地覆盖与作物分类、生态与风险监测、变化检测、回归建模及大范围制图等场景。

Citation

@misc{brown2025alphaearthfoundationsembeddingfield, title={AlphaEarth Foundations: An embedding field model for accurate and efficient global mapping from sparse label data}, author={Christopher F. Brown and Michal R. Kazmierski and Valerie J. Pasquarella and William J. Rucklidge and Masha Samsikova and Chenhui Zhang and Evan Shelhamer and Estefania Lahera and Olivia Wiles and Simon Ilyushchenko and Noel Gorelick and Lihui Lydia Zhang and Sophia Alj and Emily Schechter and Sean Askay and Oliver Guinan and Rebecca Moore and Alexis Boukouvalas and Pushmeet Kohli}, year={2025}, eprint={2507.22291}, archivePrefix={arXiv}, primaryClass={cs.CV} }