发表机构
University of Colorado Denver; National Snow and Ice Data Center (NSIDC), CIRES, University of Colorado Boulder（科罗拉多大学丹佛分校; 科罗拉多大学博尔德分校环境科学研究合作研究所（CIRES）国家冰雪数据中心（NSIDC））
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文量化海冰发育阶段制图中的多标注者标注不确定性与模型不确定性，发现软监督提升两者相关性，且在冰缘处相关性最强，蒙特卡洛dropout校准最佳。
AI 中文摘要
海冰发育阶段（SoD）描述海冰的年龄及相关厚度，为航行和业务化冰监测提供重要信息。SoD标签来源于业务化冰图，由训练有素的分析人员解读卫星观测结果，并将标准阶段代码分配给具有相似冰况的区域。这些代码通常表示兼容冰厚度的范围，而非精确的物理值。深度学习方法能够自动化SoD制图，并通常采用业务化冰图作为训练参考标签。然而，这些标注并不精确，因为冰图解读依赖于分析人员的判断以及当时可用的观测数据，因此不同的冰服务可能对相同条件分配不同的SoD标签。我们将这种独立生成的专业标注之间的差异称为多标注者标签不确定性；将标注合并为单一确定性目标会丢弃这种差异。第二个不确定性来源源于学习模型本身。在本文中，我们量化了这两种来源：来自独立冰服务图之间不一致的标注不确定性，以及来自学习预测模型的模型不确定性。然后，我们通过测试模型不确定性在冰服务不一致处是否更高来评估它们之间的关系。我们观察到，纳入来自多个标注者信息的监督可以改善这种对应关系，其中软监督实现了最高总体相关性0.256。在冰缘附近，这种关系显著增强，模型预测不确定性紧密追踪多标注者不一致性，在0-10公里范围内达到0.704的相关性。在不确定性估计方法中，蒙特卡洛dropout提供了校准最佳的置信度估计，期望校准误差为0.050。
英文摘要
Sea-ice stage of development (SoD) describes the age and associated thickness of sea ice and provides important information for navigation, and operational ice monitoring. SoD labels are obtained from operational ice charts, where trained analysts interpret satellite observations and assign standardized stage codes to regions with similar ice conditions. These codes often represent ranges of compatible ice thicknesses rather than exact physical values. Deep-learning methods can automate SoD mapping and commonly adopt operational ice charts as reference labels for training. These annotations are not exact, however; this is because chart interpretation relies on analyst judgement and on the observations available at the time, so different ice services may assign different SoD labels to the same conditions. We term this variation across independently produced expert annotations multi-annotator label uncertainty; collapsing the annotations into a single deterministic target discards this variation. A second source of uncertainty originates in the learned model itself. In this paper, we quantify both sources: annotation uncertainty from disagreement among independent ice-service charts and model uncertainty from the learned predictive models. We then evaluate their relationship by testing whether model uncertainty is higher where ice services disagree. We observe that supervision incorporating information from multiple annotators can improve this correspondence, with soft supervision achieving the highest overall correlation of 0.256. The relationship becomes substantially stronger near the ice edge, where model predictive uncertainty closely tracks multi-annotator disagreement, reaching a correlation of 0.704 within 0--10 km. Among the uncertainty-estimation approaches, Monte Carlo dropout provides the best-calibrated confidence estimates, with an expected calibration error of 0.050.
