Forward Modeling, Continuation And Data Interpretation Of Magnetic Gradient Tensor

In recent years, the measurement of magnetic gradient tensor is graduallybecoming a popular method of magnetic detection for geophysical exploration. Aseries of magnetic gradient tensor systems have been carried out in American，German，Australia and so on. Besides, many surveys have been undertaken with theinstrumentation outdoors. Magnetic gradient tensor provides unique advancescompared with conventional magnetic surveys. For these reasons, it is desirable to paymore attention to the method of data processing and interpretation during the researchof the device. We just attempt to improve magnetic gradient tensor techniques, so it isvery desirable now to discuss the relevant theory problems.In order to model magnetic field, an understanding of the mathematical laws thatgovern these fields is vital. Therefore forward modeling is a fundamental step formagnetic data interpretation. This thesis introduces mathematical modeling theory ofmagnetic survey firstly, especially theories for the magnetic gradient tensor. Thenbasic field sources are applied in the process of forward modeling, showing exampleswhere appropriate. Characteristics of tensor gradient components and derivedquantities are discussed then. Numerous specific advantages of magnetic gradienttensor are listed at the end of this part. Some difficulties and disadvantage are alsodiscussed. This part of the research is the basis for this thesis.The process of transforming a data set so that it appears that it has beenmeasured at a different height is called continuation. Continuation is one of theimportant methods for magnetic data processing. Upward or downward continuationis adopted due to different conditions, so that information about the causative bodycan be obtained. Because magnetic gradient tensors meet the continuation conditions,it can be applied in the data processing. This thesis firstly discusses upwardcontinuation of the tensor data, which suppresses shallow features and highlightsregional anomalies due to deep targets. In the process of downward continuation, an iteration method has been used to resolve closely spaced causative bodies. Thismethod has better effect than many other downward continuation methods. In the end,we analyze the accuracy of both continuation methods, so as to make sure that theycan be applied in real data processing.The majority of potential field inversion routines assume that only onecomponent of the field is being used. Gradient tensor data, however, have fivecomponents that can be used for inversion. The particularity of magnetic gradienttensor allows numerous data interpretation methods available. The thesis mainlydiscusses two techniques for data interpretation of magnetic gradient tensor. Above all,three eigenvalues of the magnetic gradient tensor are used to estimate thedimensionality of the target. Also models created by COMSOL Multiphysics which isa kind of software that based on finite element principle are used to confirm thereliability of the method. Then I discuss the technique of analytic signal which can beapplied to magnetic gradient tensor. Three analytic signals in different directions areintroduced, which enhance the edges of the causative bodies. The magnitude ofmagnetic vector components is also important in the process of data interpretation.Point dipole and line of dipoles are used to test the determination of locations anddepths of the analytic signals. Eventually，the precision of the method in obtainingtarget locations and depths have been discussed.