Time-lapse Tomographic Inversion Using a Gaussian Parameterization of the Velocity Changes

Changes in acoustic velocity due to production of hydrocarbons may occur in reservoir or overburden layers. These changes may be detected as time shifts in time lapse seismic data. To map these velocity changes we propose using a least squares tomographic inversion algorithm utilizing pre stack time shifts as input. The velocity changes are modeled as Gaussian functions in two dimensions, thus limiting the amount of parameters to six per Gaussian. Ray tracing is used for the forward modeling. The inversion is carried out using a iterative re-weighted least squares algorithm, and several synthetic data experiments have successfully been conducted on both simple cases with one Gaussian anomaly and more complex cases with and without noise added to the data.