大型复杂介质模型三维快速正演是瞬变电磁数据精细化解释的关键。本项目针对城市地下空间结构复杂、电性反差大和多尺度特征，通过结合多尺度杂交混合理论和杂交间断有限元法建立一套瞬变电磁三维快速正演算法。通过多尺度网格剖分精细刻画地下复杂电性结构。构建独立的微观局部问题和宏观全局问题，实现瞬变电磁正演问题尺度分离。基于局部杂交问题构造满足瞬变电磁扩散规律的多尺度基函数，捕捉宏观单元内非均匀地电信息，实现尺度提升。针对瞬变电磁大尺度时间特征，开发基于局部时间步长的多尺度时间离散策略，捕捉精细电性结构引起的电磁扩散特征。采用杂交间断有限元法并行求解包含强电性反差和多尺度结构的微观局部问题，快速获得多尺度基函数和场源影响。最后在粗网格上求解宏观全局问题，实现瞬变电磁三维快速正演。本项目的研究成果对提高大型复杂介质模型三维瞬变电磁数据精细化解释水平，促进城市地下空间精细探测技术发展具有积极的推动作用。

The fast 3D forward modeling algorithm for large complex models is the key to the fine interpretation for transient electromagnetic data. In this project, we establish an efficient 3D transient electromagnetic forward modeling algorithm for the urban underground space with complex geological structure, high electrical contrast and multiscale characteristics by combining the multiscale hybrid-mixed theory and hybridized discontinuous finite element method. Multiscale meshing is used to describe the complex underground electrical structure. By constructing the independent micro-local problems and macro-global problem, we achieve the scale separation of transient electromagnetic forward problem. Based on the local hybridization problems, the multiscale basis functions are constructed to capture non-uniform geoelectric information in macro cells. To deal with the large-scale time of transient electromagnetic method, we develop a multiscale temporal discretization strategy based on local time step to capture the electromagnetic diffusion feature caused by micro electrical structures. The hybridized discontinuous finite element method is used to solve the micro-local problems with strong electrical contrast and multiscale structures in parallel. Then the multiscale basis functions and the fields produced by current source are obtained quickly. Finally, the macro-global problem is solved on the coarse mesh to realize the 3D fast forward modeling for transient electromagnetic method. The research results of this project have a positive effect on improving the level of fine interpretation of 3D transient electromagnetic data on large complex models, and promoting the development of fine detection technology for urban underground space.