Mixed-Phase Wavelet Estimation for Designature in Marine Seismic Data Processing

De-signature is a key step in marine seismic data processing, aimed at removing the source signature to enhance temporal resolution. This study compares two de-signature approaches that differ in how the source wavelet is obtained. The first approach is deterministic and uses a modeled far-field source wavelet derived from the air-gun array configuration. The second approach is data-driven and extracts a mixed-phase source wavelet from the data using higher-order statistics.

The performance of both methods is evaluated on a deep-water dataset. The comparison focuses on their ability to achieve zero-phase alignment across all the frequency band up to 250 Hz. Results indicate that the data-driven approach provides better bubble attenuation and achieves improved zero-phasing up to approximately 64 Hz. At higher frequencies, both methods perform comparably