Marine Controlled Source Electromagnetic Methods1-D Forward Modeling And OCCAM Inversion Research

In the year of1982, Chave and Cox got the solutions of the electric dipole source infrequency domain and the controlled source electromagnetic methods has been studied.Marine controlled source electromagnetic(CSEM)surveying has been in commercial usefor predrill reservoir appraisal and hydrocarbon exploration for10years. In thisdissertation, we examine the sensitivity of the CSEM method to such layers withforward and Occam’s inversion modeling in one dimension. A deep-towed transmitterclose to the seafloor injects a current of several hundred amps into the seawater froman electric dipole, creating magnetic and electric fields that propagate diffusively intothe seafloor. The electromagnetic waves generated from the source propagate into theseafloor and interact with the high resistance layer in there.Electric dipole receiversrecord the seafloor electric fields at various ranges from the transmitter. Because theelectromagnetic waves spread along the conductive layer would get greaterattenuation than spread along the high resistance layer, compared with the structurewhich lack of high resistance layer, the emergence of the reservoir will strengthen thereceive signal. So comparing the observed data and the prediction model or inversionusing the CSEM data, the geophysicists can recognize the resistivity anomalies andspeculate the geological informations.Based on existing ground electromagnetic method, the main research work andachievements are as follows:(1) In the beginning of this dissertation，we studied the basic principle of marineCSEM through collecting and analyzing a large number of informations about marineMT and marine CSEM.The history and the research status of marine CSEM soundingis reviewed, and we lookforward to the future.(2) According to Maxwell’s equations and the basic theory of electromagneticfields, especially the ideal from Nabighian’s method of deriving formula of electric dipole, we deduced the expressions for the electromagnetic induction fields producedby horizontal current source in the conducting ocean overlying a one-dimensionalearth.(3) Research on Marine CSEM1D modeling. First, we verified the feasibilityand effectiveness of the CSEM by using a series of common models.Through thesimulation analysis, the electric and the magnetic field components are used tocompare the sensitivity to the high resistance layer. The geometry of generator andreceiver is studied, and the result shows that radial dipole geometr are more sensitiveto the presence of the thin resistive layer than the azimuthal fields. The differenttransmit frequency will effect the response results.The air wave is one of the mostimportant factors in CSEM, in the dissertation this would be discussed. In addition,we also discussed the sea depth and the offset distance which would effect theresponse results. And a conclusion will be given that CSEM methods is not completelT-equivalent, particularly at higher frequencies.(4) Research on Marine CSEM1D Occam’s inversion. Based on the1D forwardmodeling, this dissertation gives a brief summarizes of the geophysics inversionmethods and focus on the theory of Occam’s inversion. The model studies here useOccam’s inversion method which solves the regularized problem by searching for thesmoothest model that fits the data. The utility of this method is that it generallyproduces smooth peaks in the model that correspond to features that are wellconstrained by the data, whereas features not constrained by the data will be smoothedover or entirely absent in the model.A brief reviewe of the Occam’s inversionapproach is given in this paper. In this dissertation Synthetic data studies indicate thatjointly inverting frequencies of0.1and1.0Hz offers better resolution than invertingeither frequency alone. An inline horizontal electric dipole is found to provide betterresolution than either broadside or vertical electric dipoles. Separate inversions ofelectric and magnetic fields perform equally well at recovering the reservoir. Smoothinversion for a multiple resistive layer model detects the presence of all resistivelayers, and shallow thin resistive layers do not impact the ability to image deeperresistive layers. The accuracy of the inverted models is improved substantially byincluding the boundary depths of resistive layers as a priori structure in the inversion.