固定翼时间域航空电磁数据拟二维整体反演  

Quasi 2D holistic inversion of fixed-wing time-domain electromagnetic data

作　　者：王琦[1] 张琼[2] 于生宝[2] 王凌群 朱凯光[2] 

机构地区：[1]吉林大学通信工程学院,长春130026 [2]吉林大学仪器科学与电气工程学院,长春130026

出　　处：《地球物理学进展》2016年第3期1173-1180,共8页Progress in Geophysics

基　　金：国家自然科学基金项目(41274076);国家高技术研究发展计划(863)项目(3013AA063904);国家重大科研装备研制项目(ZDYZ 2012_1_03)联合资助

摘　　要：航空电磁探测数据量大,二维、三维反演算法复杂、计算速度慢,通常采用一维反演,利用层状模型拼接描述地下复杂结构,但航空电磁数据信噪比低,容易引起一维反演结果横向连续性差等问题.本文针对上述问题,基于一维反演算法,通过整合测线观测数据,建立了测线数据整体的目标函数,并根据Tikhonov正则化反演理论,引入包含空间粗糙度和先验信息的模型参数约束项,确定了拟二维整体反演的目标函数,推导了反演迭代方程组,利用超松弛共轭梯度算法,求得由于整条测线整体反演所致的大型稀疏矩阵的极小化解,实现了对整条测线数据同时反演的固定翼航空电磁数据的拟二维整体反演算法.在反演迭代过程中,正则化因子采用线性搜索自适应迭代的方法自动选取,提高了反演结果的稳定性.对比分析了仿真数据的一维反演与拟二维整体反演结果,得出拟二维整体反演算法横向连续性较好,对高导覆盖层下的导体分辨率优于一维反演,同时受高斯噪声的影响较小.最后,将直升机飞行实测噪声加入仿真数据中,拟二维整体反演结果平均相对误差较一维反演结果降低了31.6%,进一步验证了拟二维整体反演算法的有效性.Since the two-dimensional（ 2D） and three-dimensional（ 3D） algorithm are complexity and time-consuming for inversion of densely sampled airborne electromagnetic data. The one-dimensional（ 1D） inversion is usually used to describe the complex underground structure. Due to the limited of the low signal-to-noise ratio,the airborne electromagnetic data often results in poor lateral continuity of 1D inversion. To solve the problems mentioned above,based on one-dimensional inversion algorithm we have constructed the data misfit of survey line by integrating observation data,and introduced the model parameters spatial roughness and prior information according to Tikhonov regularization inversion theory. Then the iterative inversion equations are derived. Solutions of large sparse matrix of whole survey line data is obtained by using ultra relaxation conjugate gradient algorithm. Quasi 2D holistic inversion of fixedwing airborne electromagnetic data is realized. In the inversion iteration,regularization factor is selected by a linear search method automatically. The quasi 2D holistic inversion results are compared with single-site Marquart inversion for the synthetic data. The results indicate that the quasi 2D inversion algorithm have better lateral continuity and less affected by Gaussian noise. To further demonstrate the effectiveness,the observed noise is added to the simulation data. The inversion results show that average relative error of quasi 2D holistic inversion is reduced by 20. 6%.

关 键 词：航空电磁 拟二维 整体反演 正则化 

分 类 号：P631[天文地球—地质矿产勘探]