地面磁共振与瞬变电磁橫向约束联合反演方法研究

Joint and laterally constrained inversion of surface MRS and TEM data

大地电阻率分布信息是影响磁共振地下水探测反演结果准确性的重要因素.在众多电磁法勘探技术中,瞬变电磁法具有高分辨率、高效率和大探测深度等优势,能准确探测地下几百米范围内的电阻率分布信息.因此磁共振与瞬变电磁联合解释方法具有重要意义.然而,利用单一测点拼接的磁共振与瞬变电磁联合解释方法进行模拟二维反演时存在解释结果不唯一,容易出现错误异常体等问题,尤其在复杂地质情况下,同一测线上相邻测点探测结果连续性差,解释结果偏离实际.基于此,本文提出磁共振与瞬变电磁横向约束联合反演方法(Laterally Constrained Inversion,简称LCI),重点引入外推积分法(quadrature with extrapolation,简称QWE),解决了传统正演过程中基于直接数值积分方法引起的求解效率低的问题,保证了联合反演方法的顺利实施,进而以相邻测点地下结构应具备连续性为依据,引入横向约束反演思想,通过在联合反演目标函数中加入相邻测点间各模型参数约束矩阵,提高磁共振解释结果准确性,加强探测剖面地质结构和含水模型连续性.经过理论模型证实,本文提出的LCI方法能有效提高传统一维反演结果的稳定性和唯一性.最后,对安徽黄山野外实际探测数据进行横向约束联合反演,验证了磁共振与瞬变电磁LCI联合反演方法的实用性.本文的研究成果将为磁共振与瞬变电磁空间约束联合反演奠定基础.

The earth resistivity distribution information is an important factor influencing the accuracy of the inversion of MRS data for groundwater detection. In the many electromagnetic exploration methods, TEM has the advantages of high resolution, high efficiency and great exploration depth, which can detect the underground resistivity distribution accurately within the scope of a few hundred meters. Therefore, joint inversion of MRS and TEM data is of great significance. However, there are such problems as non-uniqueness and wrong location of abnormal bodies when we use the single-point splicing way in the joint interpretation of MRS and TEM to perform two-dimensional inversion. Especially in the complex geological conditions, adjacent point detection results of the same measuring line have poor continuity. The inversion results thus deviate from real geological situations. In order to ensure the inversion method implemented smoothly, we propose a method of joint and laterally constrained inversion of MRS and TEM data （Laterally Constrained Inversion, LCI）, which introduces the QWE method （quadrature with extrapolation, QWE） to solve the problem of low efficiency caused by traditional forward modeling based on the direct numerical integration method. Based on that underground structure of adjacent measuring points should have continuity, we introduce the idea of laterally constrained inversion. Joining the constraint matrix of parameters between adjacent points in the joint inversion objective function permits to improve the accuracy of MRS. At the same time, it allows us to strengthen the continuity of the geological structure and water content model of the detection section. A series of tests on the theoretical model show that the LCI method can effectively increase the uniqueness and stability of traditional 1D inversion results. Finally, we invert the actual detection data from Huangshan in Anhui by the method of joint and laterally constrained approach. The results are consistent with the basic geological data, which verifies the practicability of the algorithm. The research results in this paper will provide a basis for the joint inversion of MRS and other electromagnetic data. At the same time they can lay the foundation for the spatially constrained joint inversion of MRS and TEM data.