Noise in Marine Seismic Data

Abstract

Marine seismic is a well established method to search for subsurface hydrocarbon deposits. However, the method is often limited by various sources of noise, of which flow and swell noise are the dominating types. This study takes advantage of 3-D direct numerical simulations of fluid flow combined with real life, and full scale measurements of flow and swell noise acquired on purpose built seismic streamer cables in the ocean, to study the mechanisms responsible for flow noise generation. The combined knowledge obtained by the simulations and the measurements are then put to use in order to come up with practical methods to reduce noise in seismic data. Two different paths are followed: The first is in the form of a software de-noising algorithm developed and implemented as a module in a commercial seismic processing software package. It works in the frequency domain by statistically comparing neighboring traces, and attenuates amplitudes that are found to be abnormal. The module is in daily use, and has successfully been applied to attenuate various types of noise found in both land, and marine seismic data. The second path followed to reduce the amount of noise in seismic data is to use so-called superhydrophobic surfaces. This is in the form of a coating material that can be applied to seismic streamers to reduce both drag and flow noise. The flow noise reduction capabilities of superhydrophobic surfaces is a new discovery, which holds great promise.