干旱区生态环境知识资源中心

Application of the finite-difference reverse time migration (FD-RTM) to ground penetrating radar (GPR) data may generate high-resolution imaging for complex geologic bodies. However, in conditions of complex topography, due to topographic roughness it is very difficult to get clear images of near-surface internal structure with the traditional approach of "static correction + flat datum migration" and RTM using Cartesian coordinate. In this paper, we developed a staggered grid finite difference forward modeling algorithm using Maxwell's equations with curved coordinates. transformed from coordinates, based The wave equation in TM mode is Cartesian coordinates to curved on irregular grid subdivision. By comparing with RTM section of the synthetic data in Cartesian coordinates and curved coordinates, we demonstrated the advantage of RTM imaging in curved coordinates. After validation, we also applied this approach to the imaging of the internal structure of a sand dune using GPR field data acquired in the Badain Jaran desert, China.