3D joint inversion of gravity and magnetic data based on structure similarity of node physical property model
S.L. Li, N.P. Wang, X.H. Meng and J. Wang
Event name: EAGE-HAGI 1st Asia Pacific Meeting on Near Surface Geoscience and Engineering
Session: Poster session
Publication date: 09 April 2018
Info: Extended abstract, PDF ( 888.17Kb )
Price: € 20
The node physical property model with good sparseness is the difference form of its corresponding block physical property model. It can not only eliminate computational redundancy and improve the efficiency of forward computation, but also can effectively recover simple geology models. The developed gravity-magnetic simultaneous inversion method introduces the structure similarity constraint between two node physical property models to make the inversion results have the characteristics of structural consistency. The Cauchy norm constraint can be used to get a sparse solution. Within the conventional inversion framework, this strategy does not need to apply nonlinear functions such as physical property transformation function to overcome the problem of nonlinear enhancement, and the inversion results are not affected by the initial value. The model tests show that the inversion method can effectively recover simple geology models. The boundary of the recovered anomalies is clear and the location is close to the real position. The block physical property values of the recovered anomalies are also closer to their true values. Compared with the inversion using only one geophysical method data, the joint inversion improves the vertical resolution of anomalies to a certain extent, and inhibits the generation of partial interference anomalies.