NAME
mctshift - module to perform a multi-component time-shift,
as for a layer-stripping operation on split-shear reflection
data. mctshift accepts four input traces (xx,xy,yx,yy), and
applies a user-specified time advance of dt: In the reflec-
tion case: to the yy trace (by default), and of dt/2 to the
xy and yx traces; In the VSP case: to both the xy and the yy
traces (by default).
SYNOPSIS
mctshift [ -Nroot_ntap ] [ -Oroot_otap ] [ -Pxsd_ptap ] [
-PFpro_fast ] [ -PSpro_slow ] [ -swheader_mnemonic ] [
-rsirs ] [ -reire ] [ -nsins ] [ -neine ] [ -sindex ] [
-dtms ] [ -V ] [ -? ] [ -h ]
DESCRIPTION
mctshift
This routine is needed to complement
multisource/multireceiver rotation of shear reflection data,
to resolve the effects of azimuthal anisotropy, in the case
where the principal directions of such anisotropy vary with
depth. In such a case, it is not sufficient to rotate
independently at each time-step (as with routine ackr),
since the shear-waves will split (and re-split while upcom-
ing) at every horizon where the anisotropy changes its
orientation. This appears to be a common circumstance in
the sedimentary crust; Winterstein and Meadows (Geoph, 56,
1331, 1991) show that where they have used the VSP version
of the present algorithm, they do find that the direction
varies, on a coarse-layer basis.
Hence, the rotation (eg using routine rottnsr) must be fol-
lowed by an analysis (currently done visually) to determine
at which reflection time the optimum angle of reflection no
longer adequately fulfills the Alford criterion (ie elimina-
tion of signal on the off-diagonal (mismatched) traces).
This determines the time corresponding to the bottom of a
coarse layer of uniform anisotropic direction. Then, appli-
cation of routine mctshift performs a layer-stripping
action, replacing the split-shear displacements at this
layer-bottom by a single wave, polarized in the fast direc-
tion. The operation is then repeated for the next layer. See
details in APR report F93-G-17 (Section III).
Strictly speaking, the algorithm should be applied only to
traces resulting from a single raypath; a stacked trace may
be an acceptable noise-reduced surrogate for a vertical-ray
trace under certain (poorly-understood) conditions.
In the reflection case, the layer-stripping action consists
of advancing the slow diagonal trace by the user-supplied
(2-way) delay dt, and the off-diagonal traces by the 1-way
delay dt/2 (since for each of these, the reflection from the
next layer down has passed through the upper layer once (eg
down) as a fast wave, and once as a slow wave. The static
shift applied, if trace-variable, is output to the trace
header mnemonic TVPT20 and to an xgraph file
root_name_Xgraph. This file may be plotted using xgraph to
show the time delay vs cdp data.
In the VSP case (using "-VSP"), we implement Winterstein's
original algorithm: advancing both xy and yy (by default) by
the 1-way delay dt (but see below).
If the slow trace is NOT the 22 trace, you must specify it
using "-s11" (the ONLY legitimate alternative to the default
[22] is 11). There is NO provision at present for handling
the case where 11 becomes slower than 22 somewhere along the
line. It is EXPECTED that the user has handled this problem
during rotation [using the trace-variant capabilities of
rottnsr]. If you specify "-s11" and "-VSP", then mctshift
will shift the xx and yx traces.
In order to eliminate the need for you to input manually all
four input and four output filenames, a MULTICOMPONENT NAM-
ING CONVENTION is used (see below). Hence, the input files
are identified on the command line (after "-N") by the root
of their filenames only, as with a matrix. (BUT see IKP
exception, below!)
MULTICOMPONENT NAMING CONVENTION
All multicomponent files are stored as SEPARATE COMPONENTS
in SIS format, and conform to all conventions established
for single- component seismic data (including headers, his-
torical line headers, etc.). The tie between the various
components is established in their names, which end in
'.ij', like subscripts on a matrix. The subscripts i and j
are taken as integers (1,2, or 3) following the normal alge-
braic convention (also adapted in the proposed multi- com-
ponent data standards of the SEG) as follows:
* The FIRST subscript (i) refers to the orientation of the
source, since the source action occurs prior to the recep-
tion.
* The SECOND subscript (j) refers to the orientation of the
receiver.
* The indices refer to a RIGHT-HANDED coordinate system,
which may be either:
# line-oriented (if only a single line is under considera-
tion), or
# map-consistent (if multiple line-orientations are
considered, then most multicomponent situations require that
a single coordinate system be used throughout, in order to
avoid error).
As long as one is consistent, some flexibility is avail-
able, but it is easy to err; the best advice is to just fol-
low the "most natural" choices defined below. Use your
creativity somewhere else.
* The POLARITY of the signals must conform to the coordinate
system, so that a positive trace value on any component
implies motion in the positive direction of that axis. With
multicomponent data, it is much easier to screw up the
polarities than with single-component data (roughly 3**2 = 9
times easier!); take care (in both acquisition and process-
ing)!
* The numerical identifiers are:
1 = x-axis (horizontal) If line-oriented, this is most
naturally the INLINE direction (since we are accustomed to
drawing x-z cross-sections); the polarity must be the same
on both sides of the source. If map-consistent, this is
most naturally EAST.
2 = y-axis (horizontal) If line-oriented, this is most
naturally the CROSSLINE direction, and (right-handed) care
must be used in the polarity. If map-consistent, this is
most naturally SOUTH, rather than north.
3 = z-axis (vertical) Whether line-oriented or map-
consistent, the SEG convention is to take positive DOWN,
rather than up, as a physicist might prefer. (This conven-
tion, along with +x = East, is what forces +y to be South.)
In some contexts (eg in an offset VSP), the indexing may
correspond to another right-handed coordinate system (eg
with the z-axis aligned with the downgoing ray, and the x-
axis lying in the saggital plane). In such a case, this
coordinate system should be derived from the foregoing by a
proper rotation.
ALGORITHM
mctshift gets both its data and its parameters from command
line arguments. These arguments specify the input, the start
and end records, the start and end traces, the slow index,
the two-way time shift, and the verbose printout flag. See
details of the algortihm in APR report F93-G-17 (Section
III).
IKP processing
If mctshift is run inside IKP, the input and output com-
ponents are connected via the process box, rather than by
the command line argument -N. The connections are specified
in the following manner: inputs 0,3,5,7 are for input com-
ponents in THIS ORDER: 11, 12, 21, 22 (using here the above
MULTICOMPONENT NAMING CONVENTION). You must specify the
FULL identifier for each input; ie NOT just the name-root,
as is done outside IKP (consequently, the names need not
conform to the CONVENTION). Similarly, outputs 1,4,6,8 are
for the output components (in the same order as the input).
Again, you must specify fully the desired destination of
these outputs. Output 9 is for the output xgraph quality
control file which may be plotted using xgraph and plots
time delay versus cdp.
Command line arguments
-N root_ntap [default: none]
Enter the input data set root filename immediately
after typing -N. This input filename-root should
include the complete path name if the files reside in a
different directory. For example, -N/b/mc/test tells
the program to look for files test.11, test.12,
test.21, and test.22 in directory '/b/mc'.
-O root_otap [optional]
Enter the output data set root filename immediately
after typing -O. This output filename-root should
include the complete path name if you want the files to
reside in a different directory. For example,
-O/b/mc/testout tells the program to create files
testout.shms.xx.11, testout.shms.xx.12,
testout.shms.xx.21, and testout.shms.xx.22 in directory
'/b/mc'.
-P xsd_ptap [optional]
Enter the XSD header value filename immediately
after typing -P. This file gives the user the oppor-
tunity of applying a trace variant time shift to the
data. The file must contain two segments. The first
segment should be a reference horizon picked from the
fast shear dataset. The second is the same event
picked on the slow shear dataset. The header mnemonic
DphInd should be detected when generating the header
value file . mctshift will calculate a time shift
based on the difference between these horizons for each
cdp of the output dataset. If the slow shear pick
drifts above the fast shear pick in time, the program
will react by assuming the slow horizon is now the fast
horizon and change the sign on the time shift. The
program uses linear interpolation between picks on the
horizon and extrapolates using the first or last value
of the horizon to cdp's outside the range picked.
-PF promax_fast [optional]
Enter the ProMAX pmx filename picked from the fast
shear section immediately after typing -PF. This file
must have a PKEYVALUE=CDP and a ZKEYVALUE=TIME. It
doesn't matter what the SKEYVALUE is. The same
interpolation/extrapolation rules apply as used with
the -P option above. If this option is used the -PS
entry must also be used.
-PS promax_slow [optional]
Enter the ProMAX pmx filename picked from the slow
shear section immediately after typing -PS. This file
must have a PKEYVALUE=CDP and a ZKEYVALUE=TIME. It
doesn't matter what the SKEYVALUE is. The same
interpolation/extrapolation rules apply as used with
the -P option above. If this option is used the -PF
entry must also be used.
-sw static [Default: TVPT20]
Enter the trace header mnemonic in which to store the
trace static applied in microseconds.
-rs irs [Default: first record]
Enter start record number.
-re ire [Default: last record]
Enter end record number.
-ns ins [Default: first trace]
Enter the starting trace for rotation. trace.
-ne ine [Default: last trace]
Enter the ending trace for rotation.
-s index [Default: 22]
Enter the index of the slow principal component.
-VSP [Default: off]
Enter for VSP mc-time-shifting.
-dt idt [Default: 0.0]
Enter the 2-way time delay (in msec) of the slow com-
ponent, relative to the fast component. You may
specify any accuracy, but only one decimal place will
be repeated in the default filenames of the output
files.
-V Enter the command line argument '-V' to get additional
printout.
-? or -h
Enter the command line argument '-?' or -h to get
online help. The program terminates after the help
screen is printed.
BUGS
None known; please notify the AUTHOR if you find any.
SEE ALSO
rotzs2(1) rotvctr(1) rottnsr(1) rltt(1) wrot(1) grot(1)
AUTHOR
Leon Thomsen (APR x 3920)
COPYRIGHT
copyright 2001, Amoco Production Company
All Rights Reserved
an affiliate of BP America Inc.
Man(1) output converted with
man2html