Perpustakaan Digital ITB

Ghost reflections in marine seismic data occur due to wave reflections from the sea surface that are recorded again by the receivers. The interference between the primary wave and ghost reflections causes energy attenuation at certain frequencies, thereby limiting the effective bandwidth of the data. This study focuses on receiver deghosting using three synthetic data scenarios: a constant-depth streamer, a variable-depth streamer with known receiver depths, and an unknown variable-depth streamer. In the first scenario, deghosting is performed using a global operator in the F-K domain. In the second scenario, deghosting is performed trace-by-trace in the F-X domain. In the third scenario, the deghosting operator is constructed based on optimum notch frequencies obtained through auto-picking without using receiver-depth information.The results show that receiver ghost characteristics are influenced by offset, propagation angle, and receiver depth. The application of receiver deghosting in all three scenarios produces data that closely approximate the synthetic primary data. The deghosting results of Experiment 1 yield an RMS residual of 0.00112, a relative RMS error of 1.67%, and a correlation coefficient of 0.99987. Experiment 2 yields an RMS residual of 0.00083, a relative RMS error of 0.88%, and a correlation coefficient of 0.99997. In Experiment 3, the notch-frequency auto-picking method yields an RMS residual of 0.00105, a relative RMS error of 1.1%, and a correlation coefficient of 0.99989. These results indicate that notch frequency can be used as an alternative parameter for receiver deghosting when streamer-depth information is unavailable. The notch-frequency auto-picking method is subsequently applied to two shot gathers of real marine seismic data. Evaluation in the time and frequency domains shows that the method can be applied to real data and is capable of reducing the effects of receiver ghosts and recovering energy around the notch frequencies and at low frequencies