发表机构
Amphitrite; Sorbonne Université; The Ocean Cleanup; INRIA（Amphitrite; 索邦大学; 海洋清理组织; 法国国家信息与自动化研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出漂移场网络（DFN），一种结合模拟预训练与拉格朗日平流损失微调的深度神经网络，用于从卫星观测预测海洋表面流场，显著降低粒子轨迹定位误差。
AI 中文摘要
北太平洋副热带环流（NPSG）是漂浮塑料碎片的主要堆积区，这是由盆地尺度的汇聚性海洋环流造成的。该区域有效的清理策略依赖于对拉格朗日粒子漂移的准确预测。在此，我们引入了漂移场网络（DFN），这是一种深度神经网络，可从业务化卫星观测中预测海洋表面流场。DFN采用一种新颖的两阶段策略进行训练，该策略将模拟数据上的预训练与基于平流一致性损失函数的拉格朗日微调相结合。这种物理信息优化直接提高了粒子轨迹预测的准确性。我们将DFN与一个基于物理的业务化预报系统进行了评估，并展示了深度学习在海洋表面流预测方面的潜力。在原位漂流浮标轨迹上，与业务化模型相比，DFN在7天预报后将平均定位误差减少了20公里。此外，使用所提出的平流损失的拉格朗日微调进一步将定位误差减少了10公里，凸显了将拉格朗日约束纳入训练过程的好处。
英文摘要
The North Pacific Subtropical Gyre (NPSG) is a major accumulation zone for floating plastic debris, resulting from basin-scale convergent ocean circulation. Effective cleanup strategies in this region rely on accurate forecasts of Lagrangian particle drift. Here, we introduce Drift Field Net (DFN), a deep neural network that predicts ocean surface flow fields from operational satellite observations. DFN is trained using a novel two-stage strategy that combines pretraining on simulated data with Lagrangian fine-tuning based on an advection-consistent loss function. This physics-informed optimization directly improves the accuracy of particle trajectory predictions. We evaluate DFN against an operational physics-based forecasting system and demonstrate the potential of deep learning for ocean surface flow prediction. On in situ drifter trajectories, DFN reduces the mean positioning error by 20 km after a 7-day forecast compared with the operational model. Furthermore, Lagrangian fine-tuning with the proposed advection loss further reduces the positioning error by 10 km, highlighting the benefits of incorporating Lagrangian constraints into the training process.
CommentsSubmitted to Artificial Intelligence for the Earth Systems
