Joint Inversion Of Total Field Anomaly And Its Gradients

Magnetics is a common geophysical technique used to image subsurface structure though variations of magnetic properties. Three-dimensional inversion has been used successfully to achieve quantitative interpretation of magnetic data. With the development of the magnetic observations, many kinds of magnetic data from multiple observe altitudes can provide more information, which can also be used in 3D inversion to image the source more precisely.The magnetic data studied in the thesis are total field anomaly and its gradients, and the target of the research is the joint inversion of the magnetic data. Considering the situations when strong remanent magnetization is present, the converted magnetic data like the total gradient and the amplitude of the anomalous total magnetic, which translated from total field anomaly and its gradients, also were used in our study.The inversion of total field anomaly and its gradients is a linear process, while the inversion of total gradient or the amplitude of the anomalous total magnetic belong to the nonlinear calculation. Here a general form of the inversion is used to solve both the linear and nonlinear problems. And the inversion methodology with technique details like the selection of best regularization parameter, bound constraints using logarithmic conversion and using preconditioned conjugate gradient method to solve the linear problems, are introduced in detail. Then I studied the relationship between the type of the source and model objective function, and proposed an preliminary strategy of the best model objective function based on the way of subsurface division and the type of sources.The joint inversion of total field anomaly and its gradients can be achieved through two points, one is increase the information about the anomaly data, and another is the increase of the model structure. Usually the total field anomaly and its gradients can be added in the data misfit function simultaneously to increase the information about the source. With the presence of strong remanent magnetization, the combination of total gradient and the amplitude of the anomalous magnetic field are also be presented in the simultaneous inversion. While the increase of model structure generally realized by utilizing the information about the source and incorporating them into the model objective function to decrease the non uniqueness of the inversion problem and let it more stable. The source edges extracted from the probability tomography results of gradient anomalies and the curie point depth estimated from the results of spectral analysis are used to self-constrain the inversion result. Two-dimentional and three-dimentional synthetic tests and four real data examples are used to illustrate the joint inversion methods’ usefulness.