地理空间基础模型旨在学习可跨区域与传感器迁移的表征，但对其在具体任务上的评估需依赖大规模、高质量、多模态的基准数据集，以衡量此类模型从数据中提取价值的能力。在洪水制图领域，现有数据集极少在大规模上同时整合双时相合成孔径雷达（SAR）影像与配准的光学影像，致使基础模型在此下游任务中的价值 largely 未被验证。我们提出 GEOID-Flood——一个源自哥白尼应急管理系统（Copernicus Emergency Management Service）响应事件的大规模多模态洪水分割基准数据集，覆盖十年间65个国家的219起灾害事件。该数据集提供逾14,000个图像块，包含配准的灾前与灾后 Sentinel-1 影像（GRD 与 RTC 格式）、灾前 Sentinel-2 影像合成图及数字高程模型（DEM），并附带经人工校验的标签，区分背景、常年水体与洪涝水体。基于该基准，我们在单图像、多时相与多模态协议下评估基础模型与传统编码器的性能。我们报告三项主要发现：基础模型带来一致但有限的性能提升；经微调的光学-SAR 融合方法最能有效识别瞬态洪水；在 GEOID-Flood 上训练的模型相较于在现有数据集上训练的模型，对未见事件具备更优的泛化能力。数据集与代码详见 https://github.com/links-ads/geoid-flood。

Geospatial foundation models aim to learn representations that transfer across regions and sensors, yet evaluating them on specific tasks requires large, high-quality, multi-modal benchmarks that measure how well such models extract value from data. Concerning flood mapping, existing datasets rarely combine bi-temporal SAR and co-registered optical imagery at scale, leaving the value of foundation models for this downstream task largely untested. We introduce GEOID-Flood, a large-scale multi-modal flood segmentation benchmark, derived from Copernicus Emergency Management Service activations, spanning 219 events across 65 countries over ten years. The dataset provides more than 14,000 tiles with co-registered pre- and post-event Sentinel-1, in GRD and RTC format, pre-event Sentinel-2 composite, and DEM, including manually validated labels that separate background from permanent water and flooded water. Using this benchmark, we evaluate foundation models against conventional encoders across single-image, multi-temporal, and multi-modal protocols. We report three main findings: foundation models offer a consistent but modest advantage; optical-SAR fusion with finetuning best resolves transient flooding; and models trained on GEOID-Flood transfer to unseen events better than those trained on existing datasets. Dataset and code available at https://github.com/links-ads/geoid-flood.