Gabor Deconvolution of Nearsurface Georadar Data for improved Fault Imaging

Georadar data are even more strongly attenuated in geological media than are seismic data. For example, when compared to seismic, georadar attenuation Q is stronger by a factor of more than 10. Recently, nonstationary deconvolution has been developed in the seismic industry as a process to correct for the amplitude loss and phase dispersion that is caused by Q. One of the most successful deconvolutions is Gabor domain deconvolution where the deconvolution operator is determined from the time - frequency decomposition of individual traces. We apply this deconvolution to strongly attenuated georadar data and we find it provides exceptionally obvious improvement in data quality. As a demonstration, we present a number of data examples from a nearsurface project in Central Italy where a large number of georadar lines were acquired to image the expression in the nearsurface of dormant earthquake faults.