A novel algorithm for 3D multi-level gravity interpretation: application to northwestern Iran

We present a novel algorithm, termed the "Forward Calculation of the Adaptive Mesh Grid" designed to unveil density structures by modeling Bouguer gravity data with minimal prior constraints. The fundamental concept of this algorithm is the innovative use of analytical signals. Upward continuation acts as a low-pass filter, decomposing the gravity signal into multi-level frequency bands across a broad range of wavelengths. The amplitude of analytical signals in each band is analyzed using radially averaged power spectrum moving windows to estimate the dimensions of a corresponding subsurface mesh. A forward modeling approach then computes the 3D density distribution for each frequency band. Finally, stacking the results from short to long wavelengths generates a 3D density model. To assess the efficacy of the algorithm, we conducted several synthetic tests, demonstrating its ability to recover anomalous features. Subsequently, we applied the algorithm to a gravity data set in northwestern Iran. Through comparative analysis with a velocity cross-section from seismic tomography, we evaluate the algorithm's effectiveness. Our findings confirm that the uppermost mantle of the Arabian plate in the Zagros collision zone and South Caspian Basin exhibits higher density compared to the lower-density uppermost mantle beneath Central Iran. We uncover lithospheric features, including a crustal-scale low-density shear zone between the underthrusting Arabian plate and the overriding Central Iran, as well as a thin, high-velocity lithospheric mantle of the Arabian plate underthrusting beneath Central Iran. These features, detected through high-resolution seismic modeling as well, demonstrate the capability of method to uncover lithospheric-scale structures.