Abstract Glaciers with termini at sea level may calve glacial ice features that can pose threats to offshore installations in the Barents Sea, especially in the central and northern part of this sea. It is therefore of great importance to estimate the annual iceberg encounter frequencies to select robust concepts for offshore field development, to design offshore structures and possibly to plan ice management operations. These encounter frequencies are often estimated using numerical models. Regardless of the model, considerable uncertainties often exist in the input data of icebergs at the sources, i.e., the annual number of icebergs released at the source and their size characteristics. The aim of this work is to reduce these uncertainties by utilizing state-of-the-art satellite remote sensing data and a complementary numerical model of iceberg drift and deterioration. Iceberg length and width distributions derived using Sentinel-2 optical imagery are presented at the major iceberg sources in the Barents Sea, which are Franz Josef Land, the eastern side of Svalbard and Novaya Zemlya. Over 22,000 icebergs were manually identified, with the largest observed iceberg being approximately 1 km long, originating from Franz Josef Land. Furthermore, a methodology is proposed to estimate the annual number of icebergs released into the Barents Sea by comparing the model results against Copernicus iceberg density data derived from the satellite synthetic aperture radar system onboard Sentinel-1. The importance of satellite remote sensing data cannot be understated because it is undoubtedly the best way to calibrate and validate the results of numerical iceberg drift models. Finally, with the calibrated model and the derived iceberg size, a map of the Barents Sea with updated annual iceberg encounter frequencies is presented.

中文翻译：

---

使用 Sentinel-1 和 Sentinel-2 估计主要冰川源的释放速率和尺寸特征，增强了巴伦支海的冰山漂移模型

摘要 在海平面有终点的冰川可能会崩解冰川冰特征，这可能对巴伦支海的海上设施构成威胁，尤其是在该海的中部和北部。因此，估计每年的冰山遭遇频率以选择用于海上油田开发的稳健概念、设计海上结构以及可能规划冰管理操作非常重要。这些遭遇频率通常使用数值模型进行估计。无论采用何种模型，源头冰山的输入数据往往存在相当大的不确定性，即源头每年释放的冰山数量及其大小特征。这项工作的目的是通过利用最先进的卫星遥感数据和冰山漂移和恶化的补充数值模型来减少这些不确定性。使用 Sentinel-2 光学图像得出的冰山长度和宽度分布显示在巴伦支海的主要冰山来源，即弗朗兹约瑟夫地、斯瓦尔巴群岛和新地岛的东侧。人工识别了超过 22,000 座冰山，观察到的最大冰山长约 1 公里，起源于弗朗兹约瑟夫地。此外，通过将模型结果与来自 Sentinel-1 上的卫星合成孔径雷达系统的哥白尼冰山密度数据进行比较，提出了一种估计每年释放到巴伦支海的冰山数量的方法。卫星遥感数据的重要性不容小觑，因为它无疑是校准和验证数值冰山漂移模型结果的最佳方式。最后，使用校准模型和导出的冰山大小，显示了具有更新的年度冰山遭遇频率的巴伦支海地图。