3D joint inversion of gravity and magnetic data based on structure similarity of node physical property model

Author(s):  
S.L. Li ◽  
N.P. Wang ◽  
X.H. Meng ◽  
J. Wang
Geophysics ◽  
2006 ◽  
Vol 71 (3) ◽  
pp. L35-L42 ◽  
Author(s):  
Mark Pilkington

Gravity and magnetic data are inverted jointly in terms of a model consisting of an interface separating two layers having a constant density and magnetization contrast. A damped least-squares inversion is used to determine the topography of the interface. The inversion requires knowledge of the physical property contrasts across the interface and its average depth. Since the relationship between model parameters and data is weakly nonlinear, a constant damped least-squares inverse is used during the iterative solution search. The effect of this inverse is closely related to a downward continuation of the field to the average interface depth. The method is used to map the base of the Sept-Iles mafic intrusion, Quebec, Canada, and the shape of the central uplift at the Chicxulub impact crater, Yucatan, Mexico. At Sept-Iles, the intrusion reaches a thickness of [Formula: see text], coincident with the maximum gravity anomaly, south of the intrusion center. At Chicxulub, the top of the central uplift is modeled to be [Formula: see text] deep and has a single peak form.


2020 ◽  
Vol 224 (2) ◽  
pp. 1344-1359
Author(s):  
Zhengwei Xu ◽  
Guangui Zou ◽  
Qianqian Wei ◽  
Junqi Tian ◽  
Hemin Yuan

SUMMARY This paper develops a minimum-support focusing stabilizer to perform a joint inversion of the vertical components of gravity and magnetic data using fuzzy c-means clustering (FCM) with the regularized Newton method in a space of weighted parameters. Not only does this joint inversion technology arrive at the conditionally well-posed traditional potential field inversion, but it also increases the structural correlation between multiple inverted models. The FCM and the focusing stabilizer make it possible to balance the convergence of the data space (D) and the model space (M), guiding multimodal geophysical parameters toward assigned petrophysical values, which makes the results more stable and realistic. Two model studies are presented to illustrate the method, a simple synthetic model with two rectangular bodies in a homogenous background and a realistic model of the Volcanogenic Massive Sulfide (VMS) deposits in northeastern New Brunswick, Canada. These models demonstrate that the new focusing joint inversion algorithm produces better images than traditional methods because the FCM function uses the structural correlation of density contrast and magnetic susceptibility as constraints.


2019 ◽  
Author(s):  
Zhengwei Xu* ◽  
Guangui Zou ◽  
Jiang Wang ◽  
Junqi Tian ◽  
Yue Mao ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document