[First Hit]

Datapages, Inc.Print this page

Compressive Sampling Techniques: A Methodology for Sub-ice Data Acquisition

Ian Hanlon

SLIM Compressive Sampling: for sub-ice data acquisition Traditional seismic data acquisition relies on towed or seabed arrays arranged in a geometrically ordered fashion over which a regular acoustic source is applied. Maintaining strict geometry is crucial for conventional processing techniques, and is done such that data with the highest fold and the fewest artifacts contribute to the final image product. Compressive Sampling (CS), as demonstrated in many papers (Candes and Tao, 2006; Donoho, 2006a, Mallat, 2009, Herrmann and Lin 2009a, Herrmann, Mansour & Wason 2011 ), throw these geometric rules out of the window and promote the idea of sparsified stochastic acquisition. We describe schema for controlled (i) random source & receiver positioning, (ii) source mixing, (iii) simultaneous shooting, for imaging; and present data interpolation in the curvelet domain, congruent with the needs for Full-waveform Inversion modelling. In terms of conventional data collection, with billions of dollars of in-water investment, there is little incentive for adopting CS acquisition, however, in regions where conventional data collection methods are highly difficult, there is a compelling case to develop CS methods in order to collect data despite the limitations. The ideal candidate for exploration with CS techniques sub ice in the frozen Arctic. The following presentation discusses CS theory, research and results, to date by the SLIM group, and the potential application of CS acquisition sub ice, and the subsequent recovery of data for processing. Examples are shown using synthetics and OBC datasets, where subsets of randomized source and receiver groupings are resorted into super-shots representing possible CS recording patterns, simultaneous Interpolation and De-Noising in the Curvelet domain allows the application of conventional processing and imaging techniques, we can also exploit Curvlet interpolation and frequency domain conversion to produce regularized data for FWI. Finally, we also explore ideas for interpolating shot lines and receiver lines with varying linear, radial and 'wave' offsets, which (potentially) rotate around the survey area producing MAZ gathers, in essence, illuminating the subsurface from a range of directions. (i) jittered sampling (ii) Previous HitairNext Hit Previous HitgunTop and vibroseis sources (iii) supershots

AAPG Search and Discovery Article #90177©3P Arctic, Polar Petroleum Potential Conference & Exhibition, Stavanger, Norway, October 15-18, 2013