Regularized equidimension inversion of gravity data

Abstract

The inversion of gravity data is to recover a 3-D distribution of density contrast based on the gravity observation data. In reality, we can merely utilize the limited and discrete gravity data known only on the surface in the inversion. That causes the inherent non-uniqueness and low depth resolution of the gravity data inversion results. In this paper, we developed a new method of the regularized equidimension inversion (REI), which achieved the inversion of the 3-D gravity anomaly model with 3-D gravity anomaly data. We combined the BG theory and Tikhonov regularization method, developed the regularized inversion of the potential field data. The character of equidimension inversion is to inverse the subsurface model utilizing the 3-D anomaly data. The 3-D anomaly is including the anomaly data of several height planes besides the 2-D anomaly data of the surface.

Steady and higher accuracy inversion results can be obtained by utilizing the regularization even the observation data with errors. The accuracy of the 3-D REI of gravity data is better than the accuracy of the current inversion which is utilizing the singular gravity anomaly data on the surface, and this new application is a powerful way for the improving of the depth resolution. The accuracy of the inversion and the improvement of the depth resolution are better as the increasing of the number of data layers on the height because the data in different heights contain the information of the subsurface anomaly orebody in different depth.