Document Type : Research Paper
Authors
1 Petroleum Engineering and Geophysics Laboratory, School of Mining Engineering, College of Engineering, University of Tehran, Iran.
2 Petroleum Engineering and Geophysics Laboratory, School of Mining Engineering, College of Engineering University of Tehran, Iran
3 Institute of Geophysics, University of Tehran, Iran.
Abstract
This study develops a unified framework for full-tensor gravity-gradient and full-tensor magnetic forward modelling and inversion using a mixed-norm regularization approach for subsurface physical property reconstruction. By utilizing all components of both gravity and magnetic gradient tensors, the forward modelling stage captures the full directional sensitivity of the potential fields, providing a more informative representation of subsurface structures than conventional approaches. The inversion is formulated as a constrained optimization problem and solved using a projected Gauss–Newton conjugate gradient method combined with an iterative reweighted least squares scheme. Mixed-norm ensures a balance between compactness and smoothness, enabling recovery of sharp interfaces and smooth background variations. Sensitivity-based weighting and physical property bounds enhance stability and robustness. The results demonstrate reliable reconstruction of density and susceptibility models with good agreement between observed and predicted full-tensor responses. Overall, the study highlights the effectiveness of full-tensor gravity and magnetic inversion for improved subsurface imaging.
Keywords
)