Abstract
This master thesis in Geophysics is about detecting and attenuating seismic interference noise. The thesis is written in collaboration with the Marine R&D Department at CGG and Department of Geosciences at the University of Oslo. The main focus of this work was to detect and attenuate seismic interference from 2D marine seismic data measured by three-component (3C) streamers. The multicomponent measurement consist of pressure (P) and particle velocity component Vy and Vz. During the course of the project it was found that the two velocity sensor components Vy and Vz could not be used to characterize and detect the seismic interference (SI). The reason is that they are very sensitive to streamer born noise such as transversal vibration noise. Such noise will be almost perpendicular to the direction of the propagating wave in the water column. Thus the particles do not move along with the wave, but oscillate up and down around their equilibrium position as the wave pass by. Therefore, in this thesis we had to limit the analyses to pressure data only. Unlike velocity sensors, hydrophones can be made insensitive to vibration by design. Therefore, the development of multicomponent streamers that measure particle velocity is more challenging than design of conventional streamers that measure pressure only. The attenuation of seismic interference in this work was carried out using a τ-p to common-p method which after such transformation and sorting, makes the SI appear with a random character which can be efficiently removed. In the analysis, data before and after removal of SI were compared at three different processing steps: pre-stack in the form of source gather, stacked section and post-stack migrated sections. At the level of stacking, the SI corrected dataset performed slightly better. However, after migration, both datasets performed equally well. This result is not surprising, since the data considered here were dominated by SI with a moveout opposite that of the recorded data. In cases where the two moveouts are fairly similar, the attenuation of SI is believed to play a more important role.