GPR technologies

[GP-05] Time Reversal Imaging based Subsurface Velocity Estimation

*SWARNA LAXMI PANDA¹, SUBRATA MAITI¹, UPENDRA KUMAR SAHOO¹ (1. NATIONAL INSTITUTE OF TECHNOLOGY ROURKELA (India))

Ground Penetrating Radar (GPR) is the most reliable tool for nondestructive subsurface imaging. The most crucial parameter in ground-penetrating radar (GPR) survey is the estimation of electromagnetic wave velocity of the subsurface medium. Velocity estimates are the integral part for estimation of many physical properties of the medium. Furthermore, it is very much essential for subsurface imaging to pin point the buried object location. Therefore, accurate estimation has a great impact because a slight deviation may leads to wrong interpretation. Without any prior information, subsurface velocity estimation is a challenging task. Various techniques are well documented in literature for this purpose. Some approaches based on the ray- approximation of wave equation which uses only the travel time information. Since the ray-approximation based techniques are noise sensitive, it makes difficulty in getting the arrival time. On the other hand, Full Waveform Inversion (FWI) uses entire wave information to get back the subsurface properties. But this inversion is highly nonlinear and requires the information about the source wavelet. Time reversal imaging (TRI) uses a basic physical concept based on the reciprocity feature of wave equation which incorporates the time reversing and back propagating the reversed recorded wave through the same media. By effective autocorrelation, this will create a focusing at exact source location regardless of the medium complexity. A novel TRI based subsurface velocity estimation is proposed which overcame some of the drawbacks of existing travel time tomography and FWI techniques. The proposed method is validated on synthetic data (using GPRMAX data) as well as on measurement data.