Using GPR and seismic reflection measurements to characterize buried objects: large-scale simulations

Using ground-penetrating radar (GPR) and seismic reflection measurements simultaneously can give a much better characterization of buried objects than a stand-alone measurement since electromagnetic and seismic measurements provide complementary information of the buried objects and the surrounding environment. However, successful interpretation and processing of these measurements in large-scale problems rely on fast simulations. In this work the authors use the pseudospectral time-domain (PSTD) method newly developed by the author together with the finite-difference time-domain (FDTD) method to perform large-scale simulations of ground penetrating radar (GPR) and seismic reflection measurements. The PSTD method uses the fast Fourier transform (FFT) together with the perfectly matched layer (PML) to solve the partial-differential equations. It requires only two cells per wavelength regardless of the problem size. For multidimensional problems, the PSTD method is 4^D-32^D times more efficient than the conventional FDTD method. Hence, the PSTD algorithm is ideal for large-scale problems. The FDTD method, on the other hand, is used to model structures with fine details below 1/8 wavelength. Both the FDTD and PSTD algorithms are developed for conductive and viscous media, and thus can be used to model realistic losses in the subsurface media