An Imaging Algorithm For Gpr Data

The finite difference time domain (FDTD) method is calculated by the explicit<br>finite approximation scheme of Maxwell’s curl equations. Since it allows arbitrary<br>electrical conductivity and permittivity variations within a model, the FDTD method<br>has become one of the powerful forward modeling methods for electromagnetic (EM)<br>phenomena. On the other hand, the reverse time migration is the imaging algorithms,<br>which is performed by inserting the recorded data as boundary conditions at each<br>recorder position in reverse time order. Due to the advantage of simple calculation<br>algorithm and applicability to arbitrary model, both methods are powerful tools for GPR.<br>In the present paper, the synthetic GPR data are calculated by using of the FDTD<br>method with the exploding reflector method. The reverse time migration is formulated<br>using FDTD scheme and applied to model with lossless and lossy media. In lossless<br>media, the forward and reverse time schemes are expressed by the completely same<br>scheme. The proposed method is successfully applied to synthetic data for two models,<br>these are steeply dipping structure and point diffractors model. In lossy media, the<br>forward scheme includes diffusion term, while the reverse time scheme includes<br>divergence term. In such a case, we show the condition that the EM wave velocity is<br>regarded as constant, then this methodology is applicable successively. We also discuss<br>the reverse time migration for the lossy media after the amplitude recovery under the<br>lossless media condition.