I’m confused. If you already accurately know your velocity model, isn’t the problem already solved? Isn’t the whole reason for migrating to find the velocity model of the earth?
yeah, that's a confusing subject. ;-) We run an iterative process ("PSDM velocity model building") whereby we start with a trial velocity, then measure an attribute of "velocity incorrectness". We then update the velocity model (e.g. using a reflection tomography algorithm) to minimize the "velocity incorrectness", re-migrate, and repeat the velocity update. We might update the velocity model 4-8 times before we call it good. And no, in seismic interpretation, the reflection image itself is what we interpret. If our velocity models had infinite spatial and vertical resolution, then we could interpret the velocity for geology, but in reality this is not achievable. If full waveform inversion fulfills its ultimate promise, we could someday interpret off the velocity model directly. But people have been working on FWI for 40 years now
wiki.seg.org/wiki/Reverse_time_migration gives more tutorial (though probably not that good). The Baysal et al. (1983) reference listed on that page is a good reference.
Hello, On the comparison PSTM (I assume is the old Kirchhoff) are you using the same Velocity Model for both?? (PSTM and RTM) ? should not be a more fair comparison do Kirchhoff PSDM vs RTM? One limitation on the RTM is the Frequency (because of computer power) normally you stop at 25 or 35 HZ on RTM while Kirchhoff gives you the full band. Thanks for the video great review.
In this case, we did not run the PSTM. The company (Wave Imaging Technology Inc) was a depth imaging specialty shop, and all we had was one-way WEM and RTM. I also like Kirchhoff PSDM. I wrote my own Kirchhoff PSDM code in 2017 and run it in my current job, working as an in-house processing geophysicist at Aspect Energy in Denver. Thanks to Amazon's compute cloud, I depth-imaged 5500 sq km of 3D data in 2022. 3x as much as we did in a year at Wave Imaging Technology...and I'm just one guy!
You are very welcome. I should probably make a new video (this one is 8 years old!)
6 ปีที่แล้ว +1
@@MorganBrown I really appreciate if could that be possible to get more videos about RTM, actually, I'm doing my master thesis on RTM, I already started since a couple of months, thanks again for the video.
@ Cool, good luck. I gave a lecture at Colorado School of Mines in 2012 where I expanded the RTM discussion to include a section on angle gathers. Maybe I will build that into a video.
"Wave Equation Migration". The "wave equation" was actually a one-way propagator that only goes up or down, compared to a "two-way" propagator that can propagate turning waves. Most people are using RTM these days.
Great job Morgan! Thank you for your presentation.
FThanks alot for simplifing such typical mathemetical thing in animation.Thanks Dr Brown
I’m confused. If you already accurately know your velocity model, isn’t the problem already solved? Isn’t the whole reason for migrating to find the velocity model of the earth?
yeah, that's a confusing subject. ;-) We run an iterative process ("PSDM velocity model building") whereby we start with a trial velocity, then measure an attribute of "velocity incorrectness". We then update the velocity model (e.g. using a reflection tomography algorithm) to minimize the "velocity incorrectness", re-migrate, and repeat the velocity update. We might update the velocity model 4-8 times before we call it good.
And no, in seismic interpretation, the reflection image itself is what we interpret. If our velocity models had infinite spatial and vertical resolution, then we could interpret the velocity for geology, but in reality this is not achievable. If full waveform inversion fulfills its ultimate promise, we could someday interpret off the velocity model directly. But people have been working on FWI for 40 years now
Thank you very much Morgan Brown for your wonderful theory on RTM, could you suggest any literature regarding algorithms related to RTM
wiki.seg.org/wiki/Reverse_time_migration gives more tutorial (though probably not that good). The Baysal et al. (1983) reference listed on that page is a good reference.
Hello, On the comparison PSTM (I assume is the old Kirchhoff) are you using the same Velocity Model for both?? (PSTM and RTM) ? should not be a more fair comparison do Kirchhoff PSDM vs RTM? One limitation on the RTM is the Frequency (because of computer power) normally you stop at 25 or 35 HZ on RTM while Kirchhoff gives you the full band. Thanks for the video great review.
In this case, we did not run the PSTM. The company (Wave Imaging Technology Inc) was a depth imaging specialty shop, and all we had was one-way WEM and RTM. I also like Kirchhoff PSDM. I wrote my own Kirchhoff PSDM code in 2017 and run it in my current job, working as an in-house processing geophysicist at Aspect Energy in Denver. Thanks to Amazon's compute cloud, I depth-imaged 5500 sq km of 3D data in 2022. 3x as much as we did in a year at Wave Imaging Technology...and I'm just one guy!
very good illustration of RTM. Thanks!
My pleasure. Maybe I will update it sometime. This video is getting a bit long in the tooth
Great!. Simplified. Thanks
Thanks a lot
You are very welcome. I should probably make a new video (this one is 8 years old!)
@@MorganBrown I really appreciate if could that be possible to get more videos about RTM, actually, I'm doing my master thesis on RTM, I already started since a couple of months, thanks again for the video.
@ Cool, good luck. I gave a lecture at Colorado School of Mines in 2012 where I expanded the RTM discussion to include a section on angle gathers. Maybe I will build that into a video.
what is wem??
"Wave Equation Migration". The "wave equation" was actually a one-way propagator that only goes up or down, compared to a "two-way" propagator that can propagate turning waves. Most people are using RTM these days.
good
thank you for your explanations and i'm really grateful, i don't want to be rude, but those mouth noises are really annoying.