Improved Interface Inversion Based On Constrained Varying Density And Its Application

Calculation of density interface using inversion of gravity data is of great significance in research of regional geological structure and mineral resources exploration. Existing inversion methods focus on the constrained varying density interface inversion algorithm, however, less considering accuracy of the constraint data and lateral density changes of targeted areas, thus affecting reliability of final inversion results. To solve this problem, based on the forward modeling and inversion algorithm of varying density interface and actual requirements, this paper proposes a density variable interface inversion method under the constraint of known depth information. It includes three aspects: algorithm of varying density constrained inversion, verification of known constraint information and partition of the density model. The algorithm provides an iterative inversion for density interface with known depth constraint information; verification of constraint information evaluates the accuracy of the constraint and reduces its systematic error; and horizontal density partitioning is introduced in the inversion to deal with density interface models of different structure backgrounds, and to improve the credibility of the results. The proposed method is applied to estimation of the Moho depth in the South China Sea. More reliable depth is obtained with known constraints and a partition density model, which proves effectiveness of the method suggested.