Wavenumber-domain joint inversion method for magnetic anomaly and its gradient vector data

Abstract

The invention discloses a wavenumber-domain joint inversion method for magnetic anomaly and its gradient vector data, relating to the field of geophysical exploration data processing technology. The invention comprises acquiring magnetic field data disturbed by ocean currents and the precise instantaneous spatial position of each sensor; regularizing the data and calculating the spatial offset and confidence weight of each grid point; constructing a novel wavenumber domain joint inversion kernel function directly embedding the position offset; using the kernel function for forward modeling to generate predicted data and compare with observed values; performing iterative solution based on the kernel function to update the subsurface model until convergence conditions are met; and outputting the final magnetization intensity model compensated for position error. By directly incorporating the spatial position offset information of data acquisition points into the construction of the inversion kernel function, the invention realizes in-situ compensation for errors caused by position offset within the wavenumber domain inversion framework, thereby improving the accuracy and reliability of inversion results under non-ideal observation conditions.