Technical Contributions

My novel contributions in this dissertation are the following:
$\bullet$ In iterative multisource simulations with marine streamer data, the mismatch between the limited number of live hydrophones in the survey and the pervasive number of live traces generated by modeling is recognized as a form of crosstalk. This crosstalk noise is, for the first time, completely removed by the encoding and decoding scheme of frequency selection.
$\bullet$ This encoding scheme is applied to multisource lsm, regularized by a Cauchy norm in the presence of noise. Relative to conventional migration I show with empirical simulations that multisource lsm with frequency encoding provides a $8\times$ speedup. I also show that if starting from csg in the time domain, then my iterative method requires preprocessing that incurs $2\times$ the I/O cost of conventional migration. Afterwards my iterative method only requires a minimal fraction, say $1/12$ , of the I/O cost of conventional migration.
$\bullet$ While the preceding application uses split-step migration, formulated in the frequency and wavenumber domains, for fwi I adopt the fdtd method, which is preferred by many as the method of choice for 3D fwi. This presents new challenges to the frequency selection scheme, such as spectral leakage, a problem mitigated by doubling the simulation time and retaining only the second half of the simulated records. I show that multisource FWI with frequency encoding achieves a speedup of about $4\times$ and $8\times$ for synthetic and field data compared to the conventional method.
$\bullet$ The resolution limits of various constituent wavepaths pertaining to fwi is comprehensively studied. For the first time, I show that the gap of missing reconstructed intermediate wavenumbers is partially filled by capitalizing on diffractor and interbed multiples.
$\bullet$ This dissertation has resulted in one journal paper (Huang and Schuster, 2012) published in Geophysical Prospecting. A second paper (Huang et al., 2013) is submitted to Geophysical Journal International, and a third paper based on this dissertation will soon be submitted for publication. Its associated expanded abstract (Huang and Schuster, 2013) was presented at the EAGE London meeting in June 2013, and was recognized as one of the outstanding papers delivered at the this international meeting. A special letter was sent to me in July 2013 inviting me to publish this work in Geophysical Prospecting, the leading applied geophysics journal in Europe. These publications are listed in Appendix H.