A regularized Wiener–Hopf filter for inverting models with magnetic susceptibility

For a magnetic target, the spatial magnetic signal can be expressed as a convolutional integral over Green's function of an assumed model with susceptibility as its parameter. A filter can be used to obtain the susceptibility by minimizing the mismatch between observed and the computed magnetic anomalies. In this perspective, we report the development of an advanced digital filter, which is efficient and can be used to map rock susceptibility from the acquired magnetic data. To design the new filter, we modified the space‐domain standard Wiener–Hopf filter by imposing two different constraints: (i) the filter energy constraint; and (ii) normalization of the filter coefficients. These constraints make it capable to characterize source bodies from their produced magnetic anomalies. We assume that the magnetic data are produced by induced magnetization only and interpretation can be as good as the subsurface model. Our technique is less sensitive to the data noise, which makes it efficient in enhancing the interpretation model. The modified filter demonstrates its applicability over the synthetic data with additive white Gaussian noise. In order to check the efficacy and adaptivity of this tool in a more realistic perspective, it is also tested on the real magnetic data acquired over a kimberlitic district adjoining to the western margin of the Cuddapah Basin in India to identify the source bodies from the anomalies. Our result shows that the modified Wiener–Hopf filter with the constraint for the magnetic data is more stable and efficient than the standard Wiener–Hopf filter.