New magnetic inversion scheme including terrain-effect correction

Abstract Magnetic inversion analysis is used to estimate the subsurface magnetization structure from surface magnetic anomalies and plays a key role in geophysical exploration due to its sensitivity to lithological variations. Surface magnetic anomalies include those created by topography, referred to as terrain effects. When estimating the subsurface magnetization structure using magnetic inversion analysis, accurately separating terrain effects from anomalies associated with the target structure is essential, as failing to do so can distort the estimated magnetization structure model. To account for terrain effects, previous studies have estimated the average magnetization of the model domain primarily using correlation-based methods prior to inversion. However, these methods become ineffective when terrain effects and target anomalies are strongly correlated, leading to distortions in the magnetization structure model obtained through inversion. To address this limitation, the current study develops a new method that simultaneously estimates proper terrain effects and the magnetization structure model by solving a basic magnetic field equation explicitly containing a terrain-effect term. Additionally, accurate terrain-effect correction requires a more precise representation of topography, which necessitates finer model domain division. However, increasing the resolution significantly expands the matrix size required for inversion calculations, leading to a substantial rise in computational costs. To migrate this issue, we introduce a new data-space inversion method. The results of the synthetic tests demonstrate that the proposed method effectively separates terrain effects from target anomalies and enhances the accuracy of the estimated magnetization structure model, even in cases where terrain effects and target anomalies exhibit high correlation. Furthermore, we apply this method to aeromagnetic data acquired over the Himekami pluton in northeastern Japan. The results of the proposed method exhibit greater consistency with prior geological information, especially in the mountain body, which produces terrain effects, compared to the predictions of the conventional method. The proposed method could lead to considerable advances in the refinement of geophysical exploration. Graphical Abstract