磁梯度张量异常反演算法与算例

发布时间：2018-07-02 09:52

本文选题：磁法勘探 + 梯度张量　；参考：《中国地质大学(北京)》2015年硕士论文

【摘要】：磁梯度张量方法是磁法勘探的重要发展方向,磁梯度张量是利用磁场在各个方向上的分量的导数来进行勘探解释的一种方法。随着超导磁力仪以及多磁通门测量装置的发展,直接测量磁梯度张量成为可能,而磁梯度张量相比于普通的磁法勘探来说,可以提供更多的信息。此外磁梯度张量也存在着更多的优点,它本身是一个位场,具备对称性与无源性,可以利用延拓等位场处理中常见的方法对磁梯度张量数据进行处理,而且受到地磁场方向的影响相对较小。梯度张量测量能提高磁法测量的可解释性,尤其是异常形态被剩磁或者低纬度磁化扭曲的地区。磁梯度张量的反演,相较于传统磁异常反演来说,能够又更好的分辨率,而且磁梯度张量具有多个分量,当我们观测出这些分量之后,可以利用加权的方式改造目标函数,进行联合反演。联合反演理论上包含了更多的信息,因此也会对异常体的构造、形态、位置等因素有着更好的刻画。在进行了磁梯度张量的网格单元正演拟合之后,采用适当的反演方法进行反演,就可以得到地下空间内的异常分布状况。这里我们采用的反演方法是预条件共轭梯度法,预条件共轭梯度法和一般的共轭梯度法相比,可以提高反演的稳定性,提高运算的迭代效率,是目前地球物理位场反演所采用的主要算法之一。此外我们根据磁梯度张量的性质,加入了多种约束因子,主要包括深度约束因子,稀疏约束因子等,进一步提高的反演的稳定性与分辨率。磁梯度张量具有九个分量,其中有五个独立分量,通过实验我们发现,每个分量的单独反演均可以起到一定的效果,但因为各个分量的核函数有所差异,反演效果各异,也不利于地质解释工作。将多个分量联合反演之后,得到的效果更加理想,组合模型上对深部异常也有着更高的分辨率。这些结果也证明了反演的有效性。由于条件有限,论文亦存在很多不足,并在文章最后提出了结论与建议,希望能够对继续从事这一方面工作的人有所启发。
[Abstract]:Magnetic gradient Zhang Liang method is an important development direction of magnetic exploration. Magnetic gradient Zhang Liang is a method to interpret exploration by using the derivative of magnetic field components in each direction. With the development of superconducting magnetometers and multi-flux gate measuring devices, it is possible to measure magnetic gradient Zhang Liang directly, and magnetic gradient Zhang Liang can provide more information than conventional magnetic exploration. In addition, magnetic gradient Zhang Liang has many advantages. It is a potential field, which has symmetry and passivity. It can be used to process the magnetic gradient Zhang Liang data by using the common methods of continuation allelic field processing. Moreover, it is less affected by the direction of geomagnetic field. Gradient Zhang Liang measurements can improve the interpretability of magnetic measurements, especially in areas where anomalous forms are distorted by remanence or magnetization at low latitudes. The inversion of magnetic gradient Zhang Liang has better resolution than traditional magnetic anomaly inversion, and the magnetic gradient Zhang Liang has many components. When we observe these components, we can transform the objective function in a weighted way. Joint inversion is carried out. Joint inversion theory contains more information, so it also has a better description of the structure, shape, location and other factors of anomalous bodies. After the forward fitting of the grid element of the magnetic gradient Zhang Liang is carried out, the abnormal distribution in the underground space can be obtained by using the appropriate inversion method. The inversion method used here is the preconditioned conjugate gradient method, which can improve the stability of inversion and the iterative efficiency compared with the general conjugate gradient method. It is one of the main algorithms used in geophysical potential field inversion. In addition, according to the properties of magnetic gradient Zhang Liang, we add a variety of constraint factors, including depth constraint factor, sparse constraint factor and so on, to further improve the stability and resolution of the inversion. The magnetic gradient Zhang Liang has nine components, of which there are five independent components. Through experiments, we find that the individual inversion of each component can play a certain effect, but because of the different kernel function of each component, the inversion effect is different. Nor is it conducive to geological interpretation. The combined inversion results are more ideal and the combined model has higher resolution for deep anomalies. Because of the limited conditions, there are many shortcomings in the paper, and at the end of the article, the conclusion and suggestion are put forward, hoping to inspire the people who continue to work in this field.
【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.2

【参考文献】