2000.04.21

Here's an update from November 2000.

This semester at the University of Houston, I've been taking a mathematical physics class with Arco's Art Weglein. He's a fairly intense theoretician, easily one of the best in the seismic industry.

Art is joining the UH geosciences department in August. At UH, he plans to start a research program devoted to developing new imaging and multiple suppression algorithms.

Next Wednesday he's going to give a talk on his plans entitled "A fundamentally new approach to processing seismic data: imaging and inversion at depth without the velocity model." The talk, which has been completely unpublicized, will be in the UH geosciences building at noon.

When I asked him how his theory compares to the recent work out of Delphi, he answered (paraphrasing):

"We're going to do depth migration without velocities. This is beyond what Berkhout is doing. Berkhout doesn't know about this yet. Berkhout does time migration without velocities, then uses the velocities to find the depth. We're going straight to depth in one step."

Normally, I would be suspicious of science that was so counter-intuitive. But I know how rigorous this guy is, so I'm inclined to believe that he must be on to something.

Next Wednesday is also the second day of a two-day SEG symposium on multicomponent seismic, which I am attending. I plan to cut out of the SEG talks and attend Weglein's presentation at lunch time.