[Seisunix] DIP MOVE OUT

[AKI]

Hello every one !

I am stuck with few questions, will be thankful if community can answer it !

1.What is DMO

Simple defination

1. DMO is a DIP move out program which we use in addition with NMO if there are high dipping reflectors as NMO fails in case of higher dips. DMO is generally applied in following order before migration.

DMO + NMO +STACK

Correct me if i am wrong..

2. There are many DMO modules

SUDMOFK - DMO via F-K domain (log-stretch) method for common-offset gathers

SUDMOTX - DMO via T-X domain (Kirchhoff) method for common-offset gathers

SUDMOVZ - DMO for V(Z) media for common-offset gathers

SUTIHALEDMP - Based on Hale paper

so, which one is usually apply..

3. Also.  DMO theory is based on constant RMS velocity. So do these methods take constant velocity  (RMS) acorss the whole section , As its difficut to calculate interval velocity in case of dipping reflectors

Akshay

[Karl Schleicher]

Akshay,

DMO or dip moveout is just the correction you need so that in constant velocity

nmo + dmo + stack + poststack migration = prestack migration + stack

The typcal sequence is 

 nmo with a simple velocity field - perhaps velocity functions every 2 miles of 3 km.

dmo on offset gathers

inverse nmo 

velocity analsys 

nmo 

stack 

post stack migration

I think constant velocity dmo is pretty reasonable.  You might get some lift from dz dmo if you have very steep dip and a simple velocity function with depth.  Probably better to move to prestack migration if you need to extra accuracy.

I think I saw artifacts in the SU dmo programs a year ago, but I did not try to track them down of check parameters.

Karl

---

Date: Sat, 18 Feb 2012 01:16:05 -0700
From: gulati...@gmail.com
To: seis...@mailman.mines.edu
Subject: [Seisunix] DIP MOVE OUT

_______________________________________________ seisunix mailing list seis...@mailman.mines.edu https://mailman.mines.edu/mailman/listinfo/seisunix Unsubscribe: seisunix-u...@mailman.mines.edu

[gulati...@gmail.com]

Just want to be sure I am on right path..when we apply dmo we assume constant velocity both in vertical and lateral direction . And that contsnt velocity is rms velocity ..

Sent wirelessly from my BlackBerry device on the Bell network.
Envoyé sans fil par mon terminal mobile BlackBerry sur le réseau de Bell.

---

From: Karl Schleicher <k_schl...@hotmail.com>

Date: Sat, 18 Feb 2012 15:41:43 +0000

To: Akshay Gulatia<gulati...@gmail.com>; seismic unix<seis...@mailman.mines.edu>

Subject: RE: [Seisunix] DIP MOVE OUT

[Robert Hardy]

Akshay,

You generally follow Karls flow...the velocity applied at the NMO stage could be constant (in which case Vrms=Vint) but more normally is a simple smooth stacking velocity field picked as far as possible from events without strong dips. Each of the su algorithms constucts a DMO (partial prestack migration) operator in a constant or V(z) or VTI media and applies that to the NMO corrected data.

In a more (e.g. last 15 years) modern processing flow it would be more common to skip DMO and just iteratively apply full prestack time migration (e.g. using module suktmig2d) refining the velocity field at each iteration.

A DMO flow might still be useful if for example you wanted to improve the imaging of diffractions in the final stack i.e. you didn't want post-stack migration. Other than that you might save some CPU cycles but any modern PC should be able to cope with PSTM.

Regards

Rob

________________________________________
From: seisunix...@mailman.mines.edu [seisunix...@mailman.mines.edu] On Behalf Of gulati...@gmail.com [gulati...@gmail.com]
Sent: 18 February 2012 21:39
To: Karl Schleicher; seismic unix
Subject: Re: [Seisunix] DIP MOVE OUT

[John Stockwell]

> Hello every one !
>
>I am stuck with few questions, will be thankful if community can answer it !
>
>1.What is DMO
>
>Simple defination
>
>1. DMO is a DIP move out program which we use in addition with NMO if there
>are high dipping reflectors as NMO fails in case of higher dips. DMO is
>generally applied in following order before migration.
>
>DMO + NMO +STACK
>
>Correct me if i am wrong..

Dip Moveout is a partial migration to zero offset of seismic data. The
NMO process assumes that the world consists of a single horizontal
reflector with a constant velocity medium. You could think of NMO
as a really simple model of prestack time migration. Dips, however
are not really honored by NMO, so another process was created to
to partially correct for the failure of NMO to properly correct for
the moveout of dipping reflectors. This is called "DMO (dip moveout)".

To get an idea of how such a formula would be created consider the
situation where you perform a prestack time migration, and then remodel
the data back to a zero-offset section. As stated, this would be a
cascade of a bunch of integrals and would not be computationally
practical. However, thanks to the magic of high frequency asymptotics,
a number of the integrals can be done by the method of stationary phase,
to produce a formula that looks like a migration formula, but with a
different stretching factor.

NMO followed by DMO or DMO followed by NMO are possibilities.

>
>
>2. There are many DMO modules
>
>SUDMOFK - DMO via F-K domain (log-stretch) method for common-offset gathers
>
>
>SUDMOTX - DMO via T-X domain (Kirchhoff) method for common-offset gathers
>
>SUDMOVZ - DMO for V(Z) media for common-offset gathers
>
>SUTIHALEDMP - Based on Hale paper
>
>
>so, which one is usually apply..

It depends on what you are doing. We use sudmofk and sudmotk most of
the time for our seismic processing lab. The TI one is for when you
are explicity studying problems where you can estimate anisotropy
parameters.

>
>
>3. Also.� DMO theory is based on constant RMS velocity. So do these methods
>take constant velocity� (RMS) acorss the whole section , As its difficut to
>calculate interval velocity in case of dipping reflectors

The NMO combined with DMO process is a partial migration of the data
to zero offset. If you do this sort of thing, then you are going to
stack the data, and then do a poststack migration.

It is also possible to use NMO+DMO as a velocity analysis tool, though
nothing like that has been implemented in SU as yet.

>Akshay

-John

John Stockwell | jo...@dix.Mines.EDU
Center for Wave Phenomena (The Home of Seismic Un*x)
Colorado School of Mines
Golden, CO 80401 | http://www.cwp.mines.edu/cwpcodes
voice: (303) 273-3049

Our book:
Norman Bleistein, Jack K. Cohen, John W. Stockwell Jr., [2001],
Mathematics of multidimensional seismic imaging, migration, and inversion,
(Interdisciplinary Applied Mathematics, V. 13.), Springer-Verlag, New York.