Premium
A sigmoid stabilizing function for fast sparse 3D inversion of magnetic data
Author(s) -
Rezaie Mohammad
Publication year - 2020
Publication title -
near surface geophysics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.639
H-Index - 39
eISSN - 1873-0604
pISSN - 1569-4445
DOI - 10.1002/nsg.12080
Subject(s) - sigmoid function , tikhonov regularization , inversion (geology) , inverse problem , conjugate gradient method , inverse , algorithm , synthetic data , parametric statistics , function (biology) , error function , computer science , geology , mathematics , artificial intelligence , mathematical analysis , geometry , evolutionary biology , artificial neural network , biology , paleontology , statistics , structural basin
An interesting geological objective of quantitative interpretation of magnetic data is to find inverse models which can determine sharp geological interfaces below the surface. The stabilizing function in the Tikhonov parametric functional governs sparseness constraint in the recovered model. This paper introduces a novel stabilizer based on a sigmoid function which can provide non‐smooth models in the inversion of magnetic data efficiently. An inversion algorithm is developed based on the reweighted regularized conjugate gradient to get the solution of the inverse problem using this stabilizing function. The performance of the proposed algorithm is checked on two synthetic data sets and real aeromagnetic data from McFaulds Lake in Ontario, Canada, in comparison with the results of the minimum support stabilizing function. The inverse problem converges to the solution faster when the sigmoid stabilizing function is used instead of the minimum support stabilizing function.