GEPH316 Field Data: Gathers

GEPH316 Seismic Data Processing

Field Data Workshop 2000

15 Gathers


Previous: F-K filtering	Up: Contents	Next: Velocity analysis

Types of gather

A gather is a collection of seismic traces made according to some trace header parameter. It is defined as a side-by-side display of seismic traces which have some acquisition parameter in common.
Useful forms of gathers include:
- A common-shotpoint gather: This is the collection of all the data from one shot. This gather sequence is how the data is usually shot and is the representation in shot section files S05 and S06 that have been used so far.
- A common-receiver gather: This would be the complete set of data that was received at one geophone. It would be useful for studying problems arising from one particular geophone, such as incorrect damping or noise.
- A common-offset or common-range gather: This displays the succession of data traces which have the same offset distance between the shot point and geophone. It would use one trace per shot section and is useful give a quick look at the variations in layering along a line.
- A common-midpoint CMP or common-depthpoint CDP gather: This displays all the data for the same reflection midpoint, often after correction for normal moveout and statics. It is the essential gather for NMO correction and stacking. This CDP gather will appear similar to a common-shotpoint gather, but should have less variability as all the traces refer to a common midpoint.

Common-shotpoint gather

The seismic unix command used to extract a particular set of seismic traces is the windowing command suwind. For example, it is used to extract single seismic shots from the complete seismic line. Type the command to inspect its parameters:

suwind

Repeat the process, previously performed in the viewing section, to create the shotpoint gather S05 for the shot at 05 metres.
Each seismic trace has various parameters in the header of the trace, so we may use the shot x position parameter sx to select the desired data. The values of the desired sx value lie between the identical minimum and maximum values of 05.
Run the following command to make the common-shotpoint gather file s05, containing just the traces for the shot at 05m, and then display the gather.

suwind < line.velfilter > s05 key=sx min=05 max=05
suxwigb < s05 title='Shot gather sx=05' windowtitle='Shot gather sx=05' label1='Time (ms)' label2='Trace numbers' &

This shotpoint gather S05 should look like a filtered version of the raw data that appears on the screen of the Geometrics ES1225 as it is being collected.

Common-receiver gather

We may use the geophone x position parameter gx to select the desired data. In order to get a full set of arrivals at a geophone, we may choose the detector at the distance of 36m, which received arrivals from the far shot at sx=00m to the near shot at sx=22m.
Run the following command sequence to make the common-receiver gather file g36, containing just the traces for the geophone at 36m, and then display the gather.

suwind < line.velfilter > g36 key=gx min=36 max=36
suxwigb < g36 title='Geophone gather gx=36' windowtitle='Geophone gather gx=36' label1='Time (ms)' label2='Trace numbers = (sx+2)/2' &

The appearance of the data seems to be back to front, because the longest offset from shot sx=00m to the geophone gx=36m gives the first trace and the shortest offset from shot sx=22m gives the last trace.
This is because the traces are displayed in the order that they occur in the whole line, in the sequence of shooting.
However, because of reciprocity, it could be regarded as identical to the result of a reversed shot at sx=36m and a line of geophones running from gx=22m back to 00m at the left end of a reversed line. Think through this carefully until you understand it!
Compare the display with the copy of the reversed shot at 36m in the handout. The pictures should look very similar, apart from different offset distance to the geophones.

Sorting traces

To change the sequence of traces into the usual appearance, we need to reverse the order of the traces.
The seismic unix command susort allows the sequence of a set of traces to be rearranged. Type the command to inspect its parameters:

susort

The easiest sorting parameter here is the offset distance between the shot and the geophone, which the seismic unix set header word command sushw has previously calculated as offset=gx-sx.
Sort the traces according to increasing offset distance and display the result using:

susort < g36 > g36.sort offset
suxwigb < g36.sort title='Resorted geophone gather gx=36' windowtitle='Resorted geophone gather gx=36' label1='Time (ms)' label2='Trace numbers = (offset-12)/2' &

This now looks like a normal shot section, but remember that it has been produced by multiple shots and just one geophone. Any abnormalities which occur on all the traces will have been caused by the geophone, as different shots, temporal noise and recorder channels were involved in creating each separate trace.
Remove the unwanted graphics windows and data files.

zap xwigb
rm g36.*

Common-offset gather

The most interesting type of gather is a common-offset gather. It gives a crude simple approximation to a fully stacked reflection line.
The process can be viewed as a synthetic seismic experiment which maintains a fixed separation between the source and receiver as they are moved as a pair along the seismic line. This is the actual technique commonly used in Ground Probing Radar experiments.
Keeping the offset separation rigidly fixed means that any changes in the appearance of the traces are due to real geological variations along the line.
The key decision is to choose the optimum offset distance between the shot and receiver at which the effects of deeper reflecting interfaces may best be observed, with minimum interference from other arrivals.
Inspect the printouts for the sections in the handout. At what distances are possible reflection events most clearly visible?
Although the reflections are not obvious anywhere, the following comments may be made.
- After the F-K velocity filtering, any reflection events will be much clearer.
- The near traces 1-3 will have near vertical reflections. The first might be at about the intercept time of the refracted arrival.
- The middle 3 traces 4-6 have a later leading refraction event with reflection events following close behind.
- The final 6 traces 7-12 have wide-angle asymptotic reflections. The leading refraction event should be quite late.
It is difficult to decide which are the best offset distances, so a selection can be tested.
Use the following commands to select and plot the common-offset gathers with offsets of 14m, 20m, 34m and 36m:

suwind < line.velfilter > offset14 key=offset min=14 max=14
suwind < line.velfilter > offset20 key=offset min=20 max=20
suwind < line.velfilter > offset34 key=offset min=34 max=34
suwind < line.velfilter > offset36 key=offset min=36 max=36
suxwigb < offset14 title='Offset gather at 14m' windowtitle='Offset gather at 14m' label1='Time (ms)' label2='Trace numbers' &
suxwigb < offset20 title='Offset gather at 20m' windowtitle='Offset gather at 20m' label1='Time (ms)' label2='Trace numbers' &
suxwigb < offset34 title='Offset gather at 34m' windowtitle='Offset gather at 34m' label1='Time (ms)' label2='Trace numbers' &
suxwigb < offset36 title='Offset gather at 36m' windowtitle='Offset gather at 36m' label1='Time (ms)' label2='Trace numbers' &

Inspect the common-offset plots, especially for 34 and 36m.
- Do they show an undulating refraction event at a time of about 0.03s?
- Do they show any undulating reflection events at a time of about 0.04s?
- Do they show any deeper undulating reflection events?
- Do they show any deeper air wave events?
- Do they show any deeper surface wave events?
- Do the undulating reflection events behave similarly on several plots, with slightly later arrival times on the larger offset plots?
- Which of the plots displays the reflection events most clearly?
Use the following line to try the suxmovie presentation of the data.

suxmovie < offset36 title='Offset gather at 36m' windowtitle='Offset gather at 36m' label1='Time (ms)' label2='Trace numbers' &

Inspect portions of the xmovie plot. Does the xmovie presentation show more subtle reflection features?
Use the following lines if you want sample print-outs of the offset gathered sections.

supswigp < offset36 title='Offset gather at 36m'  label1='Time (ms)' label2='Trace numbers' | lp -d mono
supsmovie < offset36 title='Offset gather at 36m' label1='Time (ms)' label2='Trace numbers'  | lp -d mono

Remove the unwanted graphics windows.

zap xmovie
zap xwigb

Common-midpoint gather

Common-midpoint gathers CMP (also called common-depth point CDP) are required for velocity analyses, Normal Moveout Correction NMO and for stacking.
The whole line can be sorted so that it is in order of increasing CDP points and also so that the traces are in the order of increasing offset for each CDP point.
Sort the complete line with the two stages and then inspect the headers using sugethw.

susort < line.velfilter cdp offset > line.sorted
sugethw < line.sorted key=sx,gx,offset,cdp | more

Now plot the complete line.

suxmovie < line.sorted title='Line sorted into CDP gathers' windowtitle='Line sorted into CDP gathers' label1='Time (ms)' label2='Trace numbers' &

Use the mouse to expand portions of the line to inspect it.
In the centre of the line, the traces should be in groups of 12 which look like normal seismic shot sections. However, they are CDP gather sections with traces all relating to the same CDP position. The traces have offsets which increase by 2m each time. For each trace, the shot position sx moves 1m to the left and the geophone postion gx moves 1m to the right.
At the ends of the line, the traces should form increasingly smaller groups. For example at the left hand end of the line, the first CDP point will be at cdp=07m with the shot at sx=00m and the geophone at gx=14m. There will only be one trace giving information on this CDP point. This is called single-fold data.
As you move further into the line, the number of traces for each CDP point will increase by one each time until the full 12-fold data is reached in the central region of the line.
Similarly at the right hand end of the line the CDP gathers gradually decrease from 12-fold back to single-fold at the end of the line. Therefore, the quality of the final processed and stacked data will not be as good at the end of the lines as it is in the centre.
Use the following to pick out the typical CDP gather at cdp=20m, plot it and inspect it.

suwind < line.sorted > cdp20 key=cdp min=20 max=20
suxwigb < cdp20 title='CDP gather at 20m' windowtitle='CDP gather at 20m' label1='Time (ms)' label2='Trace numbers' &

Although this looks like a normal shot gather, it is a CDP gather with the shots and geophones spaces symmetrically around the single central CDP point.
Use the following line if you want a print-out of the CDP gathered section.

supswigp < cdp20 title='CDP gather at 20m' label1='Time (ms)' label2='Trace numbers' | lp -d mono

Remove the unwanted graphical windows and files.

zap xmovie
zap xwigb
rm g*
rm off*
rm c*
rm line.velfilter
ls


Previous: F-K filtering	Up: Contents	Next: Velocity analysis

Go to <--->