Massive 3D seismic data compression and inversion with hierarchical Tucker
Title | Massive 3D seismic data compression and inversion with hierarchical Tucker |
Publication Type | Conference |
Year of Publication | 2017 |
Authors | Yiming Zhang, Curt Da Silva, Rajiv Kumar, Felix J. Herrmann |
Conference Name | SEG Technical Program Expanded Abstracts |
Month | 09 |
Keywords | 3D, FWI, SEG, tensor algebra |
Abstract | Modern-day oil and gas exploration, especially in areas of complex geology such as fault belts and sub-salt areas, is an increasingly expensive and risky endeavour. Typically long-offset and dense sampling seismic data are required for subsequent shot based processing procedures, e.g. wave-equation based inversion (WEI) and surface-related multiple elimination (SRME). However, these strict requirements result in an exponential growth in data volume size and prohibitive demands on computational resources, given the multidimensional nature of the data volumes. Moreover the physical constraints and cost limitations impose restrictions on acquiring fully sampled data. In this work, we propose to invert our large-scale data from a set of subsampled measurements, resulting in an estimate of the true volume in a compressed low-rank tensor format. Rather than expanding the data to its fully-sampled form for later downstream processes, we demonstrate how to use this compressed data directly via on-the-fly common shot or receiver gathers extraction. The combination of massive compression and fast on demand data reconstruction of 3D shot or receiver gathers leads to a substantial reduction in memory costs but with minimal effects on results in the subsequent processing procedures. We demonstrate the effective implementation of our proposed framework on full-waveform inversion on a 3D seismic synthetic data set generated from a Overthrust model. |
Notes | (SEG, Houston) |
URL | https://slim.gatech.edu/Publications/Public/Conferences/SEG/2017/zhang2017SEGmsd/zhang2017SEGmsd.html |
DOI | 10.1190/segam2017-17742951.1 |
Presentation | |
Citation Key | zhang2017SEGmsd |