Patent attributes
A method for reconstructing seismic data signals of poor quality to improve the signal-to-noise ratio of the data for display and analysis in connection with the selection of drilling sites for recovery of hydrocarbons. The method includes providing a signal model by applying a Karhunen-Loeve transform to selected input seismic data collected for the target zone, to form a co-variance matrix from the dot products of all pairs of input data. Eigenvalues and eigenvectors for the matrix are computed, and the most significant eigenvectors are inversely transformed to provide a coherent estimate of the signal. The input data is combined with the model data based on the determination that the model data lacks continuity, wherein the good quality signal-to-noise ratio data experiences little change and discontinuous data is enhanced by a contribution of the signal estimate data. The reconstructed seismic data of the target zone can be displayed for analysis.
