3D Pre-stack Reverse Time Migration of Ground Penetrating Radar for Subsurface Imaging

Due to the huge computational cost, reverse time migration (RTM) of ground penetrating radar (GPR) is usually implemented in its two-dimensional (2D) form. However, a 2D RTM algorithm can hardly focus the scattered signal from a object buried in a complicated subsurface environment. This paper proposes a 3D RTM algorithm for GPR based on the high-order finite difference time domian (FDTD) method. The 3D FDTD algorithm has a precision of 8 order in space and 2 order in time, and is used in forward and backward extrapolation of electromagnetic fields. The normalized cross-correlation imaging condition is used to yield the pre-stack RTM results and a Laplace filter is used to suppress the low frequency clutter, which is generated during the correlation process. The effectiveness of the proposed 3D RTM algorithm is validated using a 3D numerical model with a subsurface cubic void. Both the top and bottom interfaces, as well as the side walls of the subsurface void can be clearly imaged. In contrast, the 2D RTM algorithm can not accurately refocus the image of the side walls of the cubic target, and the interpretation of target shape is difficult. We conclude that the 3D RTM algorithm can reconstruct a subsurface image with a higher precision, compared with the 2D one.