Dataset underlying the publication: "Efficient deep learning surrogate method for predicting the transport of particle patches in coastal environments"

The data provided in this repository can be used to run the surrogate and optimal prediction experiments described in the manuscript "Efficient deep learning surrogate method for predicting the transport of particle patches in coastal environments". This paper introduces a revolutionary tool for forecasting the spread of tracers or pollutants in our oceans. We have developed a unique surrogate modeling method that combines the power of deep learning with physical oceanographic understanding. This translates to accurate forecasts that achieve at least two orders of magnitude faster than traditional systems – once the deep learning model is trained. In our paper, the experiment "surrogate prediction" is used to assess the performance of our current deep learning approach, whereas the experiment "optimal prediction" shows what can be achieved if a perfect deep learning prediction is obtained. A small sample of the data is also stored in the GitHub repository (https://github.com/JeancarloFU/paper_Efficient_Deep_Learning_Surrogate_Method_For_Lagrangian_Transport). Here, scripts, and notebooks (based on Python v3.8) used to run the surrogate and optimal prediction experiments described in the manuscript are archived.

本仓库提供的数据可用于复现研究论文《面向海岸环境中粒子团输运预测的高效深度学习替代方法》中所述的替代预测（surrogate prediction）与最优预测（optimal prediction）实验。该论文提出了一款用于预测海洋中示踪剂或污染物扩散的变革性工具。我们研发了一种独特的替代建模方法，将深度学习的优势与物理海洋学原理相结合。在深度学习模型完成训练后，该方法可实现相较于传统系统至少快两个数量级的精准预测。在本论文中，替代预测实验用于评估当前深度学习方法的性能，而最优预测实验则展示了在获得完美深度学习预测结果时所能达成的效果。本GitHub仓库（https://github.com/JeancarloFU/paper_Efficient_Deep_Learning_Surrogate_Method_For_Lagrangian_Transport）中还存储了该数据集的少量样本。本仓库中存档了用于复现该论文所述替代预测与最优预测实验的脚本与Jupyter Notebook（基于Python v3.8版本）。