Search

Seismic records are composed of many seismic wavelets that are reflected from underground impedance interfaces. Q-value information of the formations is contained in these seismic wavelets; therefore, it motivates us to decompose the seismic signal for obtaining the reflected seismic wavelets. To achieve it, we first decompose the seismic records into a series of weighted Ricker wavelets using a fast inversion-based sparse seismic decomposition algorithm which is characterized by the ability of embodying contributions of arbitrary phase rotation Ricker wavelets during its decomposition procedure. Then, with the guidance of a new-defined position indicating vector, we stack the weighted Ricker wavelets that appear near same time sample to form the final decomposition results. As a quantitative application, we propose a Q-value estimation strategy according to the centroid frequencies of the acquired seismic wavelets. A single trace and Marmousi model synthetic examples are adopted to demonstrate the effectiveness of our method in terms of seismic decomposition and Q-value estimation. In addition, our method is applied to both onshore and offshore field seismic data decomposition, successfully. The following estimated equivalent and interval Q-value attributes are further used for seismic resolution enhancement and carbonate reservoir heterogeneity evaluation, respectively. [ABSTRACT FROM AUTHOR]