Electromagnetic Wave Propagation In The Atmospheric Duct And The Key Techniques Of Inversion

In this dissertation, the key techniques related to electromagnetic wave propagation and inverse problem in the marine atmospheric duct are systematically investigated. Emphasis is put on studying the electromagnetic scattering model of rough sea surface and the forward propagation model of atmospheric duct, the inverse problem related to the typical atmospheric duct (evaporation duct and surface duct) is examined using the learning algorithm. The GPS signal scattering properties, propagation properties and the inversion of atmospheric duct using GPS are also studied. The main works are as follows:Firstly, the rough (sea) surface electromagnetic scattering theory is investigated, which includes the Small Slope Approximation (SSA) and modified Kirchhoff Approximation. The scattering coefficient of rough (sea) surface by the SSA is derived in the polar coordinate. Based on the Gaussian dielectric rough surface and rough sea surface, the numerical results are compared with the experimental data and those by Kirchhoff Approximation. In addition, the effects of incident wave frequency and wind speed on the electromagnetic scattering properties of rough sea surface are also investigated. To study the electromagnetic scattering of large incidence and extract the scattering field information, the electromagnetic scattering from rough sea surface is analyzed using the modified Kirchhoff Approximation.Secondly, the parabolic equation for atmospheric duct propagation and the derivation of the common used narrow-angle and wide-angle parabolic equation are investigated, and the key steps in solving parabolic equation are presented. To validate the accuracy of the parabolic equation for atmospheric duct propagation, the numerical simulations are compared with the experimental data, the results of Mode theory and the AREPS model. The Improved Discrete Mixed Fourier Transform (IDMFT) is used to analyze the influence of the sea surface roughness on the electromagnetic wave propagation in atmospheric duct.Thirdly, the Eikonal equation satisfied by the ray propagation is derived based on the parabolic equation and Maxwell's equations. A convenient Ray Tracing Approach is used to intuitively simulate the over-the-horizon propagation path in typical ducting environment (evaporation duct and surface duct). The ray tracing results are compared with the propagation loss distribution map of parabolic equation model rigorously solved with IDMFT and the AREPS. The relation between the antenna position and the formation of the atmospheric duct is discussed. Fourthly, according to the complex nonlinear relation among the propagation loss, the refractivity profile and its influence factor, the RBF Neural Network (RBFNN) and Least Square Support Vector Machine (LSSVM) are used to predict the propagation loss and refractivity profile in the atmospheric duct, respectively. To reduce the number of sampling and reflect the generality, the Latin hypercube sampling method is adopted to sample refractivity parameter, and the statistical analysis for the inversion results of surface duct are presented.Fifthly, the concept of cut-off wavelength is used to analyze the feasibility of GPS signal propagation in the atmospheric duct. The polarization and scattering properties of GPS scattering signal are discussed using the modified Kirchhoff Approximation, and the influence of wind speed and incident angle on the polarizability and bistatic scattering coefficient are investigated. In addition, the initial field of GPS, such as the initial field of plane wave and the GPS scattering field proposed in this work are investigated. The parabolic equation model is used to analyze the GPS signal propagation properties in the evaporation duct, and the influence of evaporation duct height and the elevation angle of GPS on the propagation properties are discussed in detail. Finally, the evaporation duct refractive index profile is inverted by the GPS scattering signal.