The Study On Radial Inversion Method Based On Gravity Data

Gravity inversion is a key step in the quantitative interpretation of gravity data using the model of anomalous distribution.The existing gravity inversion methods mainly include physical inversion methods and geometry inversion methods.Compared with physical inversion,the geometry inversion method can clearly describe the boundary of source and does not require a large amount of prior information.The radial inversion method establishes the radial model in the polar coordinate system.Compared with other geometry inversion methods based on Cartesian coordinate system,the inversion parameters are less,which can better reflect the shape of geological body.In China,there are few studies on geometry inversion methods for geological bodies(noninterface inversion).Therefore,this paper is mainly studied on the radial inversion method which has achieved better practical application results abroad.The author hopes to promote the research on geometry inversion and draws the following results:1.The basic principle of solving the two-dimensional and three-dimensional radial inversion method based on the damped Gauss-Newton iterative method is introduced in detail and the specific algorithm flow is given.The similarities and differences of gravity radial inversion methods in different dimensions are compared and analyzed.For the two-dimensional radial inversion method,the convexity constraint and the inequality constraint need to be introduced in each iteration.The relative proximity constraint can make the inversion result close to the equiaxed shape.And the absolute proximity constraint can make the inversion result as close as possible to the prespecified model.The concentration constraint can make the physical properties concentrated in one direction.For geological bodies with obvious orientation,the inversion effect of introducing concentration constraints is the best.For isometric geologic bodies,the inversion result of introducing relatively proximity constraints is best.Compared with two-dimensional,the three-dimensional radial inversion method needs to invert more types of parameters(the coordinates of the origin of the horizontal cross-sections and the depth to the bottom of the geological body),and requires more prior information(the depth to the top of the geological body).2.Aiming at the problem that the optimal regularization factor selection process of traditional gravity radial inversion method is cumbersome and computationally intensive,the original method is improved by introducing balancing priciple as a method of adaptive selection.3.The traditional method and improved method are applied to the laccolith model and the “step” model experiments,which proves the feasibility and advantages of the improved method.4.The improved gravity radial inversion method is applied to the measured profile data of the Rio Grande Gorge Bridge in New Mexico,USA,and the measured grid data of a metal sulfide deposit in Noranda,Quebec,Canada.The improved gravity radial inversion method has achieved good results in real gravity data processing.