Terrain correction using spatially variable density for analyzing airborne gravity gradient data

Abstract This study demonstrates that applying spatially variable density correction (VDC) to eliminate terrain effects in airborne gravity gradients enables the estimation of surface density distribution and improves the extraction of structural boundaries through a semi-automatic analysis. Because gravity gradients are more susceptible to nearby density anomalies than gravity, they are expected to have a high correspondence with rock-type distribution near the surface. In this study, airborne gravity gradients acquired using FALCON airborne gravity gradiometry system in the Yuzawa–Kurikoma area, Japan, were used to evaluate the validity of VDC. FALCON system measures only two horizontal components of the gradient tensor, which have anisotropic sensitivity to density anomalies. Hence, we derived the vertical gradients from these components and applied VDC. The estimated densities were generally consistent with the rock-type distribution shown in the geological map and with those obtained from actual rock samples. Subsequently, Euler deconvolution, a semiautomatic analysis technique, was conducted on the VDC-corrected data to extract lineaments based on the continuity of the solutions. The extracted structures agreed well with previously identified or inferred faults, and several new lineaments were found along the extensions of these faults. Additionally, deconvolution was applied to the data after terrain correction with a unique assumed density, and the variability and structural characteristics of the solution distributions were compared between the two correction methods based on information entropy. The VDC-corrected data yielded lower entropy than the conventionally corrected data, indicating that VDC can improve the detectability of structural boundaries when using Euler deconvolution.