Research On Inversion Of 2D MCSEM Anisotropic Media

The marine controlled source electromagnetic method(MCSEM)usually transmits low frequency(0.1Hz-10Hz)electromagnetic wave signal from an electric dipole source to the receiving array on the seabed.The receiver measures and records the value of the corresponding electromagnetic field components.The electrical structure of the seabed within a range of tens of meters to several kilometers can be determined by analyzing the amplitude and phase of the signal with the change of the source distance.Since the MCSEM is very sensitive to the resistivity of the reservoir,when the oil and gas is contained,the region exhibits a high resistivity anomaly relative to the surrounding surroundings.Due to this feature,the measurement of the resistivity value has attracted attention.However,the conductivity characteristics of subsurface media in nature often show anisotropic.Therefore,the study of anisotropic is of great significance for the cognitive underground electrical structure.Based on the previous research results,this thesis deeply studies the related technologies of MCSEM in 2D forward and inversion problem.The major contents are summarized as follows:(1)In two-dimensional MCSEM adaptive finite element algorithm,in order to eliminate the problem of source singularity,the superposition principle is used to decompose the electromagnetic field into primary and secondary fields.The primary field is the background field generated in one-dimensional uniform half-space or layered medium,which can be directly solved by the analytical method.The solution of second-field partial differential equation needs to be discrete by the finite element method.In the finite element solution process,an unstructured triangular element mesh is used to simulate a complex electrical boundary structure.The posterior error estimation method is used to evaluate the accuracy of finite element,and the grid elements with large errors are gradually refined to ensure the rationality of the global mesh refinement,so as to obtain a higher precision numerical solution while saving computing resources.(2)In the inversion of two-dimensional MCSEM,in order to improve the computational efficiency and reduce the memory requirement,the sensitivity matrixon the inversion domain is solved by the partition parallelism.The measurement area is firstly divided according to the size of the “sensitive area” to reduce the calculation scale of the inversion domain sensitivity matrix.After completing the electromagnetic data inversion of each measurement area,the total conductivity distribution in the inversion area is obtained.Thereby improving the efficiency of inversion calculation and reducing the time required during the inversion.(3)In the process of marine controlled source electromagnetic method inversion,if the background information of the geophysical field about a certain area is known,it can be added as constraint information to the inversion,but the background information is often not easy to obtain directly.In order to set a more reasonable initial model in inversion,the inversion region is divided into multiple horizontal layers according to the seismic profile,each layer resistivity value is inverted as a separate parameter,and then the inversion result is used as an initial model for OCCAM constrained inversion until convergence.The above research results show that the adaptive finite element method can simulate complex geological bodies and improve the accuracy of the numerical solution.The introduction of partitioned parallelism can increase computational efficiency and reduce the time required for inversion.Due to the large amount of inversion calculation and high memory demand in the process of MCSEM practical data interpretation,it is necessary to introduce the partition parallel method into the practical data processing.However,by analyzing the electrical structure of the area,it is found that the resistivity layers in the area are close to each other,and the range of resistivity value changes is small.It is difficult to obtain better results without adding any additional information.In order to get a good inversion result,the practical data is processed by the OCCAM inversion method based on background model constraints.The results show the effectiveness of the method.