Estimation of swell heights and reflection coefficients in rough-sea from dual-sensor towed streamer data

Subsurface imaging from conventional streamer acquisition is flawed by the sea surface level variation and sea surface reflection coefficient fluctuation. Based on a novel dual-sensor towed streamer acquisition system, a new tool for imaging the sea-surfa ce from decomposed wavefields will be developed and implemented in PGS' main processing system. This new tool is designed to quantify the rough-sea effects on conventional and dual-sensor streamer data and to estimate reliable correction values needed in subsequent data processing and subsurface imaging. The dual sensor towed streamer, developed at PGS, senses simultaneously on hydrophones the pressure field and on motion sensors the vertical particle velocity field, at the same spatial position. The acq uired wavefield may be separated at the receiver level into the up-going/down-going pressure fields and up-going/down-going vertical velocity fields. A pertinent fact is that the up- and down-going wavefields of either sensor can be used to image the se a-surface and obtain the air-water reflection coefficients above the streamer. Based on this principle, Orji Okwudili Chuks developed in his Master Thesis (2009) a method of sea-surface imaging. The method is based on stepwise extrapolating the decomposed (up- and down-going) wavfields from the acquisition level towards the sea-surface and applying an adequate imaging condition at each extrapolation step. In his PhD-project, Orji will work on a continuation and extension of this activity. He will aim to achieve a higher degree of accuracy by optimizing the imaging condition in the current development and by generalizing the underlying physical model. In contrast to previous work, the imaging tool will be generalized in this PhD activity to handle near s urface layers of rapidly changing acoustic properties overlaying a smooth background model. An extended imaging condition will be developed in order to achieve optimal statistical image stability.