Imaging of R3 profile of Chicxulub offshore seismic data using prestack split-stip Fourier migration in the plane wave domain

Date

2008

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Sixty-five million years ago, a bolide approximately 10 km in diameter traveling over 20 km/sec collided with earth in the Yucatan Peninsula leaving behind the wide multi-ring Chicxulub crater. Two-dimensional (2D) marine seismic reflection data were acquired in 1996 and 2005 to image the crustal deformation. Radial line R3, a 100 km seismic reflection profile, was processed using a conventional seismic data processing flow (McDonald, MS Thesis, 2006). In this study, line R3 is processed using a different scheme using prestack split-step Fourier migration in the plane wave domain. This new seismic imaging of the R3 data collapses the scattered waves, moves the temporal reflection events to their true structural position in depth and increases the signal to noise ratio. The field shot gathers are contaminated with low-frequency guided waves due to the shallow water column and the hard water bottom offshore Yucatan as well as the Scholte waves propagating along the seafloor interface. A 2D normal derivative operator was applied to remove this coherent noise for NMO corrected data. This multichannel filtering approach attempts to reveal the horizontal or nearly-horizontal reflections while non-horizontal evenets such as groundroll and Scholte waves are attenuated. Before migration of the reflections, the seismic shot gathers were mapped from the offset-time (X-T) domain to the vertical delay time, τ, and the horizontal ray parameter, p, or simply τ-p domain. In the τ-p domain, predictive deconvolution often works better since multiples are periodic and hence easier to remove and this usually gives better results than applying the deconvolution in the original offset-time (X-T) domain. Moreover, groundroll and Scholte waves are mapped to points in the τ-p domain and there can then be readily excluded for the imaging, improving the signal to noise ratio of the final depth section. For depth migration, a good velocity model is required to image the data to the correct position and depth. Thus, an optimized velocity model was used for prestack plane wave migration. Prestack depth migration was applied directly on the transformed τ-p gathers that are sorted into constant ray parameter sections. Each plane wave component, i.e. constant p value, was imaged separately and prestack-migrated common-image-gathers (CIGs) are collected. They are in the depth and ray parameter z-p domain, at each shot position. The migrated and stacked results are obtained by stacking a selected range or all the traces in each CIG to generate the final image. Residual depth versus p "moveout" is then used to refine the interval velocity of the depth section. The result of this new processing is an improved image in depth of the crater which is important to understanding the actual structural geometry of this large impact event. The improved image can give a greater confidence in both the geologic structure and the velocity model than time migration since the events are now in their true spatial position

Description

Citation