基于交叉梯度的重磁和大地电磁三维联合反演研究

发布时间：2018-04-17 06:33

本文选题：三维反演 + 交叉梯度　；参考：《中国地质大学(北京)》2015年硕士论文

【摘要】：基于交叉梯度理论,本文实现了重力、磁法与大地电磁的三维联合反演算法。首次将交叉梯度引入到三种参数的三维联合反演中,使三种物探方法在纵、横向上分辨率实现互补。分别对大地电磁、重磁三维反演做了基础性的研究工作:在大地电磁方面,研究了数据空间三维反演方法,有效降低了三维反演所需要的内存,提高了计算效率。在重磁方面,研究了重磁正则化约束反演方法,并加入深度加权和对数障碍项,使模型扰动量更加稳定,物性范围更加合理,收敛速度更快。对于大地电磁数据空间三维反演和重磁三维正则化反演,本文都分别使用多个理论模型合成数据进行试算,并对单方法反演结果做了具体分析。本文用上述三维联合反演算法对理论模型合成数据进行了反演试算,理论模型计算的结果表明,使用交叉梯度耦合电阻率、密度和磁化率的联合反演,不依赖于岩石物性之间的定量关系,能够使三种参数在反演迭代过程中在结构上相互约束,可获得更可靠的反演结果,明显优于传统的单一反演结果,说明了本方法的有效性和普遍适用性。
[Abstract]:Based on the theory of cross gradient, this paper realized the 3D inversion algorithm of gravity, magnetic and electromagnetic. Three dimensional joint inversion for the first time will cross the gradient is introduced to the three parameters, the three kinds of geophysical methods in the longitudinal, horizontal resolution to achieve complementarity. Of magnetotelluric, 3D inversion of gravity and magnetic research has been done on the basis of the study of the magnetotelluric data, 3D inversion method, effectively reduces the memory requirement of 3D inversion, the computational efficiency is improved. In the aspect of gravity and magnetic, gravity and magnetic regularization inversion method, adding depth weighted and logarithmic barrier, the perturbation model is more stable, physical property a more reasonable range, faster convergence speed. For the three-dimensional magnetotelluric inversion of gravity and magnetic data space and 3D regularized inversion, this paper used several theoretical models of synthetic data test, and the single method The inversion results made a concrete analysis. This paper used the three-dimensional joint inversion algorithm on synthetic data of theoretical models for the inversion calculation, the theoretical model calculation results show that the use of cross coupled resistivity gradient, joint inversion density and magnetic susceptibility, does not depend on the quantitative relationship between petrophysical, can make three kinds of parameters in the iterative inversion in the process of structure of mutual constraints, can obtain more reliable inversion results, the inversion result is better than the traditional single, illustrate the effectiveness and general applicability of the method.

【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

【相似文献】