Application of the signal processing to a short-offset seismic data in the Hupo basin, offshore Korea

Abstract

In this study, seismic data were acquired using Tamhae2 R/V to identify the subsurface fault structures in the Hupo basin. The seismic data were generated by the air-gun source (30 cu. in.). The source distance is 12.5 m, and the receiver distance is 6.25 m. The number of channels is 32. The offset range of the seismic data is 50 to 250 m. The data processing for short-offset seismic data is mainly applied with simple processing techniques such as frequency filter, trace editing, and velocity analysis in consideration of cost efficiency. However, these simple data processing techniques cannot accurately image complex subsurface structures because it is difficult to remove severe noise and water-bottom (WB) multiples effectively. Therefore, in order to accurately identify the geological structures, it is necessary to apply high-resolution signal processing techniques that can remove severe random noise and WB multiples included in raw seismic data. Severe noise was removed by applying data processing techniques such as a low-cut filter, trace editing, swell noise attenuation, and random noise attenuation. In addition, predictive deconvolution, SRME, and Radon filter were applied to effectively attenuate WB multiples that cause difficulties in geological interpretation. Finally, pre-stack Kirchhoff time migration was applied to more accurately image the subsurface structures. From the data processing results, we confirmed that the high-resolution signal processing techniques applied in this study greatly improved the signal-to-noise ratio of seismic data and effectively eliminated WB multiples.