Stale Thoughts and Broken Links
Old entries from my weblog on geophysics and the energy industry.
Last week I attended an all-day meeting at UH for the DOE National Lab
projects titled "Testing Advanced Computational Tools for 3D Seismic
Imaging" and "Next Generation Seismic Modeling and Imaging." These are
the current continuations of the Imaging-SEG Project.
Over the last five years, I received some off-and-on funding from this
program through contracts with Rice University. Here are my notes from
- Leigh House of Los Alamos has been relocated to D.C., and is not working on
these projects this year. Peter Roberts of LANL has taken over Leigh's
- Prolific researcher Biondo Biondi presented details of his wave equation
prestack depth migration algorithm, which he refered to as Fourier finite
difference plus interpolation (FFDPI). He also presented some prestack imaging
results of a well-known North Sea salt body. I missed Biondo's presentations at SEG
last month, so I was happy I got to see these talks.
- Peter Roberts gave a presentation on "Prestack
split-step Fourier migration of the SEG physical model." I've worked a
lot with split-step Fourier migration in the past, so I was excited to see this.
I believe that Peter was presenting work that was mostly done by Lian-Jie Huang
and Michael Fehler, also from Los Alamos. I checked the SEG site to see if they
presented a paper at the meeting last month. Instead, I found this paper
on the Fourier finite difference method!
- AGL at UH is still very
active in these projects, through physical modeling and the
development of new wave equation migration methods. Anning Hou of AGL presented
results using two algorithms for pseudo spectral prestack depth migration.
- UH professor Arthur Weglein gave a brief review of his SEG presentation on
imaging and inversion without velocities. Arthur has not been involved with the DOE
projects in the past.
- Shawn Larsen of Lawrence Livermore, the author of the E3D finite difference
elastic modeling code, showed examples of recent elastic modeling runs using
our extended SEG salt model. Shawn is currently running his code at 85
Gflops; he expects to have access to a 12 Tflop machine in the near future.
- From the audience, Bee Bedner of ADS responded to Shawn's talk by asking
something like, "Now that we can generated all this data, what are we going
to do with it?" Leon Thomsen of BP suggested that we should be considering
high frequency models with subseismic details and lithologic effects for
reservoir characterization studies.
- At the end of the meeting, Bob Wiley announced that he has been asked by the SEG
research committee to chair a committee to design a "benchmark" model
for a next generation modeling effort. He is looking for volunteers to work
on this new project.
I'm still very interested in this project, but I haven't formed an opinion on what
direction it should take next. If I have any coherent thoughts in the next month,
I'll post them on my homepage.