发表机构
Politecnico di Milano; Food and Agriculture Organization of the United Nations（米兰理工大学; 联合国粮食及农业组织）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究评估AlphaEarth基础嵌入在缅因州耕地制图中的表现，证明轻量分类器无需微调即可达93.7%准确率，且时间迁移稳定，并通过独立人工验证优于CDL，支持低计算量区域制图方案。
AI 中文摘要
地理空间基础模型提供了卫星影像的可复用表示，支持以有限的任务特定建模进行下游制图。我们评估了年度AlphaEarth嵌入是否支持美国缅因州的二元耕地与非耕地制图，使用了192个空间分离的图块以及源自美国农业部耕地数据层（CDL）的标签。在不微调基础模型的情况下，一个轻量级分类器在留出图块上达到了93.7%的总体准确率和90.8%的平衡准确率。逻辑回归与梯度提升集成模型的差距在0.3个百分点以内，而使用类别质心但不拟合参数的最近类质心规则达到了90.2%的准确率。一个包含60,000个标记像素的平衡样本与全部860万像素的池相比，差距在1.3个百分点以内；由于像素存在空间自相关，这一结果涉及的是像素样本效率，而非60,000个独立标注点。在同类区域迁移实验中，在某一年训练的分类器在2018年至2023年间保持准确。针对一个连续的2023年区块中385个随机采样点，采用盲法、两位解译员共识作为参照，AlphaEarth加随机森林地图的一致性为95.3%（κ=0.82），而CDL为91.7%（κ=0.72；精确双侧McNemar检验p=0.0161）。这一局部结果与CDL标签噪声的部分平滑一致，但并未确立对参考产品进行全州范围的修正。在相同点上，与微调后的TerraMind分割模型的差异不具有统计学显著性（95.3%对93.5%；p=0.14），且该实验并非计算成本的控制比较。这些结果支持将冻结的地理空间嵌入作为区域农田制图的低计算量候选方案，但需注意其局限性：仅基于单一州的研究、30米分辨率的训练参考以及单区块的人工验证。
英文摘要
Geospatial foundation models provide reusable representations of satellite imagery that support downstream mapping with limited task-specific modelling. We evaluate whether annual AlphaEarth embeddings support binary cultivated-versus-non-cultivated mapping in Maine, USA, using 192 spatially separated patches and labels derived from the USDA Cropland Data Layer (CDL). Without fine-tuning the foundation model, a lightweight classifier reaches 93.7% overall accuracy and 90.8% balanced accuracy on held-out patches. Logistic regression is within 0.3 percentage points of a gradient-boosted ensemble, while a nearest-class-centroid rule, which uses class centroids but fits no parameters, reaches 90.2%. A balanced sample of 60,000 labelled pixels is within 1.3 percentage points of the full pool of 8.6 million pixels; because pixels are spatially autocorrelated, this result concerns pixel-sample efficiency rather than 60,000 independent annotation sites. In a same-region transfer experiment, classifiers trained in one year remain accurate across 2018 to 2023. Against a blind, two-interpreter consensus at 385 randomly sampled points in one contiguous 2023 block, the AlphaEarth-plus-random-forest map agrees at 95.3% ($κ=0.82$), compared with 91.7% for the CDL ($κ=0.72$; exact two-sided McNemar $p=0.0161$). This local result is consistent with partial smoothing of CDL label noise, but it does not establish statewide correction of the reference product. On the same points, the difference from a fine-tuned TerraMind segmentation model is not statistically significant (95.3% versus 93.5%; $p=0.14$), and the experiment is not a controlled comparison of computational cost. These results support frozen geospatial embeddings as a low-compute candidate for regional cropland mapping, subject to the limits of a single-state study, a 30 m-derived training reference, and a one-block human validation.
Comments23 pages, 10 figures. Code: https://github.com/Black-Lights/alphaearth-cropland-maine
