A novel constrained approach for improving the precision of downward continuation of potential field data

The resolution of potential field data generally decreases with the observation height, which brings adverse effects for various transformations. To address this issue, the method of downward continuation has been proposed to estimate the field measured at a lower elevation, even closer to the causative sources. Unfortunately, the downward continuation operator is similar to that of the high-pass filter which also amplifies the high-frequency components of the data. Therefore, the computation of downward continuation is an ill-posed problem that is very unstable. The useful information contained in the original data would be masked by the amplified noise only after downward continuation with three to five data intervals.

To further improve the accuracy of the downward continuation of potential field data, we formulate a constrained scheme following the idea of the constrained inversion of geophysical data. The integrated computational scheme consists of the truncated Taylor series expansion, the principal component analysis, the iterative downward continuation method and the incorporation of prior information. We present the detailed calculation procedure of the constrained scheme and tested its accuracy on synthetic and real field data. Numerous model tests demonstrate that downward continuation using the constrained strategy can yield more precise results compared to other downward continuation methods without constraints and is relatively insensitive to noise even for downward continuation over a large distance. The proposed scheme is also applied to real magnetic data collected within the Dapai polymetallic deposit from the Fujian province in South China. This practical application also indicates the superiority of the presented scheme.