Time reverse modeling of low-frequency microtremors: Application to hydrocarbon reservoir localization

Time reverse modeling is applied to synthetic and real low-frequency microtremors measured at the Earth surface with synchronized seismometers. In contrast to previous time-reverse applications, no single event or first arrival time identification is applied for microtremor localization. Synthetic low-frequency microtremors are numerically generated within a small underground area and modeled with a finite-difference algorithm simulating two-dimensional elastic wave propagation. Time reverse modeling using synthetic microtremors shows that small underground source areas can be accurately located. Different source characteristics which emit mainly P-waves or S-waves vertically influence the localization accuracy. Time reverse modeling is applied to two real microtremor data sets acquired 16 months apart above known oil reservoirs nearby Voitsdorf, Austria, to investigate whether spectral anomalies observed above the reservoirs originate from the reservoirs. Time reverse modeling indicates that low-frequency microtremor signals originate from the reservoir locations and provides a possible method for reservoir localization.