Dataset for the publication "Integrating Score-Based Diffusion Models with Machine Learning-Enhanced Localization for Advanced Data Assimilation in Geological Carbon Storage"

This dataset contains 4,847 NetCDF files generated with the Delft Advanced Research Terra Simulator (DARTS). Each file represents a distinct high-resolution reservoir simulation, designed for machine learning research in carbon storage and reservoir engineering. The simulations include pressure, temperature, saturations, flow fields, permeability, porosity, and production variables. This is the dataset used on the publication Integrating Score-Based Diffusion Models with Machine Learning-Enhanced Localization for Advanced Data Assimilation in Geological Carbon Storage, which is still preprint.

本数据集包含4847个由代尔夫特高级研究Terra模拟器（Delft Advanced Research Terra Simulator，DARTS）生成的网络通用数据格式（NetCDF）文件。每个文件对应一项独立的高分辨率油藏模拟任务，专为碳封存与油藏工程领域的机器学习研究设计。该模拟涵盖压力、温度、饱和度、流场、渗透率、孔隙度及生产变量等多类数据。本数据集被用于题为《集成基于分数的扩散模型与机器学习增强定位以实现地质碳封存中的高级数据同化》的研究，该研究目前仍为预印本。