SFINCS-LSG dataset, model files, python environment and scripts

Scripts supporting the manuscript Eilander, D., de Goede, R., Leijnse, T., & Fraehr, N. (n.d.). Stress-testing a physics-guided surrogate model for compound flood dynamics across hydrodynamic regimes. This study evaluates a hybrid physics guided SFINCS–LSG surrogate model, which combines low resolution SFINCS simulations with Empirical Orthogonal Function (EOF) decomposition and Sparse Gaussian Process learning to emulate high resolution flood dynamics. Two contrasting case studies, Charleston, South Carolina, USA, and Brisbane, Australia, are used to assess model performance under diverse compound flood conditions. See readme.md file for a detailed description, folder stucture and getting started.本数据集配套支撑论文：Eilander, D., de Goede, R., Leijnse, T., & Fraehr, N. (未刊稿) 《针对不同水动力工况下复合洪水动力学的物理引导替代模型压力测试》 本研究评估了一款混合物理引导的SFINCS–LSG替代模型，该模型将低分辨率SFINCS模拟与经验正交函数（Empirical Orthogonal Function, EOF）分解及稀疏高斯过程（Sparse Gaussian Process）学习相结合，以重现高分辨率洪水动力学过程。研究选取美国南卡罗来纳州查尔斯顿与澳大利亚布里斯班两个对比鲜明的案例区域，用以评估该模型在多样复合洪水工况下的性能。 详细的数据集说明、文件夹结构及快速上手指南，请参阅readme.md文件。