Time-lapse Q-factor tomography by reflected waves’ inversion

Hydrocarbon production modifies several lithological parameters of a reservoir, as P velocity, density and electrical conductivity, which may have a dominant effect. Anelastic absorption provides a weaker time-lapse signal, but its information may become critical when other parameters are not sensitive enough. We present a new method for estimating the spectral centroid of seismic signals, which is the input for the classical frequency-shift inversion. This approach is validated by a numerical simulation of direct, reflected and head waves that are jointly inverted in a time-lapse framework. Two models are compared that differ only for a local Q-factor anomaly, detected by head waves. This approach can contribute significantly to reservoir characterization and monitoring.