Inversion of wave parameters to perform spectral recomposition

The seismic attributes associated with frequencies are important to characterize stratigraphic sequences and hydrocarbon reservoirs. The spectral recomposition is a technique to extract wave parameters as frequencies and amplitudes of the components of a seismic spectrum. The spectral recomposition to recover the frequencies and its associated amplitudes (weights) was performed as an inverse problem by an optimization criterion. The analyses of sensibility, unicity and stability were also performed. With the frequencies and amplitudes recovered, the spectral recomposition showed to be an efficient way to reach a better characterization of a stratigraphic model. With the sensibility analysis, the frequencies which are more difficult to be recovered were determined. The complexity of the objective function was studied analyzing the unicity and the stability of the function.