"Dataset for Physics-Informed Conditional Diffusion Inpainting of Gridded Three-Component Magnetic Fields"

"This dataset supports the study \u201cPhysics-Informed Conditional Diffusion Inpainting of Gridded Three-Component Magnetic Fields.\u201d It contains only synthetic planar three-component magnetic-field data generated by forward modeling and does not include real survey data. The samples are produced from subsurface susceptibility distributions under controlled conditions and record the three magnetic components on the observation plane. The dataset is intended for reconstruction and inpainting research on planar magnetic-field observations with missing regions. It includes magnetic-field samples, masks, and related experimental files, and may be used for magnetic-field reconstruction, geophysical data recovery, physics-informed learning, and data-driven inversion studies. It can also support benchmarking and comparative evaluation of related methods."