Depth migration of monostatic and bistatic georadar data

Robert James Ferguson, Matthew J. Yedlin, Christian Pichot, Jean-Yves Dauvignac, Nicolas Fortino, Stephane Gaffet

A 200 trace subset of a larger survey conducted in the "aniti-blast" tunnel of the Inter-Disciplinary Underground Science & Technology Laboratory are compared for image quality. These data are unique in that two distinct acquisition geometries are acquired simultaneously. The first, and we will call it "bistatic", is the conventional georadar acquisition where the transmitting and recording antennae are separated by 65 cm. The second we will call "monostatic". Monostatic acquisition is unique in that the transmitting and recording antennae are exactly co-located - they are the same physical antenna, and this is a recent technical development. Monostatic acquisition reproduces exactly the geometry of the well known "exploding reflector model" of seismic imaging and therefore, zero-offset migration (ZOM) of the data is not an approximation but a legitimate imaging approach. In particular, the image of the near-surface (1 m or so) should be precisely imaged (given an exact velocity model) - bistatic data require prestack depth migration (PSDM) to achieve equal precision. PSDM, of course, is much more expensive and time consuming than ZOM and so monostatic acquisition is very desirable. Here we demonstrate a number of important differences.