Step Change in 3D Two-Component OBC Seismic Imaging in the Arabian Gulf

Abstract

The common industry approach to the processing of 3D, two-component (2C), Ocean Bottom Cable (OBC) data includes summing the hydrophone and geophone sensors at an early stage of the processing sequence. In this paper, we demonstrate that a step change in seismic imaging quality has been achieved on a shallow water Arabian Gulf dataset via completely separate processing of the hydrophone and geophone data through CMP stack. The results of this work are compared with those obtained using more conventional approachs. The primary utilization of the vertical geophone sensor in a conventional two-component OBC processing flow is in the mitigation of receiver-side reverberation. The theory behind this is approach is well documented (Barr, F., 1997). However, the noise generated in a variable-depth, shallow-water environment often contains both source- and receiver-side reverberation modes that are not sufficiently reduced by conventional hydrophone-geophone summation techniques. The underlying physics of surface-wave ("mud roll") and water column trapped-wave modes detected by geophone and hydrophone sensors are quite different and do not necessarily conform to the assumptions in current processing methods. Geophone motion in particular may be quite complex (Norris et al, 2006). The effectiveness of sensor (2C) summation techniques is highly dependant upon extracting accurate scaling parameters from the data and proper spectral matching of the hydrophone and geophone records. Residual noise in the data may affect both the scaling/matching processes as well as 2C summation itself. It is commonly observed that raw vertical geophone OBC data contain more noise than the hydrophone. Inadequate noise attenuation before 2C summation often leads to "over-weighting" of the hydrophone data or, worse, discarding the geophone data entirely. Previous work (Shatilo, A., et al., 2004) demonstrated that improved imaging can be obtained via separate application of adaptive noise-suppression techniques (after Kim, 1998; Anderson et al., 2006) prior to sensor summation. The fundamental principle in this study is that the highest signal/noise ratio is generally achieved in the latest stages of the processing sequence (in this case post CMP stack), which appears to be the optimal stage for sensor summation. This not only allows for an optimal combined image but also for creation of separate images of each component. The separate OBC hydrophone and geophone images may be used in a workflow that allows one to distinguish between residual noise and primaries. The results of this study are consistent with previous work on four-component OBC data (Stewart, J., et. al., 2004). Dataset The study area is characterized by shallow water, varying from less than 7 meters to over 25 meters. Living reefs are scattered throughout, and water depth, seabed reflectivity and geophone coupling vary significantly, as is typical in these environments. The noise modes can be characterized by:"trapped-mode" water-borne energy;"mixed-mode" water-bottom-interface surface waves ("mudroll");Short-period "reverberation" multiples, with periodicity varying with water depth;Secondary-source diffractions from reefs and production facilities;Ambient "random" noise.