Sparse Joint Inversion of Gravity and Magnetic Data Using Group <i>L</i>₀-Norm Model Constraint (opens in new tab)

Owing to the nonconvexity of the $L_{0}$ -norm and the associated optimization difficulty, existing sparse joint inversion methods typically construct objective functions using approximating functions of the $L_{0}$ -norm. Although such approaches can produce focused multiphysics models, they often lose the capability of directly measuring and explicitly controlling the volume of the reconstructed models. In fact, the $L_{0}$ -norm provides a direct measure of the model volume. If the $L_{0}$...