Reminders

1
Reminders

• I will be away from 18 Oct. 13 Nov.
• Two classes during that time
• Thursday 21 Oct., 1230-230, A208 Ryan Faerber
  demonstrating some examples of seismic processing
• Friday 22 Oct., 200-400, Bevill 3O Denise
  Hills discussing ground penetrating radar.
• I will be sending 2 homework assignments from
  sea.
• These will be due in my mailbox in Bevill (see
  Betty Fagen if you do not know where this is) on
  Friday Nov. 12.
• I will send weekly reports from sea on what I am
  doing.
• Use this time to work on your projects
• Do not leave this until the last moment.
• I will have complete e-mail access during this
  time, so e-mail me if you have any questions.

2
ProMAX

• Log into the PC
• Click on your dtterm window
• Type
• su l promax
• Enter password
• Setenv DISPLAY computername0.0
• geo545
• promax
• You will see a list of areas named GEO369
  yourname
• Click on the appropriate area
• Click on Example Line

3
The Data

• Today we will be covering some basics of
  multi-channel seismic processing. We will be
  using a survey carried out in 2001.
• 23 July August 1, 2001.
• Gulf of Corinth, Greece.
• R/V Maurice Ewing.
• 240 Channels, 6 km streamer.
• Tow speed 4.5 knots.
• Shot distance 50 m.
• 20 airgun source array totaling 8445 cu.in.
• Birds spaced at 300 m, compass birds at 600 m.
• GPS on tail buoy.
• UKOOA navigation.
• Seismic data recorded in SEG-D format.
• Record length 16 seconds.

4
Streamer Navigation
The UKOOA navigation data has already been loaded
but we should check it to make sure that it is
OK. Click on Database with the left hand mouse
button (LHMB) (only once). In the new window,
make sure the TRC tab is selected. With the LHMB
click on View 2D Matrix. A new window comes
up. In the Horizontal window, with the LHMB,
select CDP_X In the Vertical window, with the
LHMB, select CDP_Y In the Color window, with the
LHMB, select CHN In the Histogram window, with
the LHMB, select CHN with the LHMB, click on OK
5

• This plot shows the location of all the channels
  on the streamer while the line was being shot.
• Red represents the channel closest to the ship
  (240), blue represents the channel furthest from
  the ship (1).
• Note that the streamer is not directly behind the
  ship. This phenomenon is known as feathering.

From Dobrin and Savit, 1988
6
Close the Database and DBTools windows. With the
LHMB, click on the 01 Trace Display flow Click
on Disk Data Input with the middle mouse button
(MMB).
This is a flow designed to display a single shot
gather. The primary sort of field file ID number
for the purpose of this survey, this is
equivalent to the shot number. The secondary sort
is the channel number. This number is given in
such a way that decreasing channel number
(increasing offset) is to the right on the
upcoming screen. Note in the flow that Automatic
Gain Control is commented out.
7
Click on Trace Display with the MMB.
Various display options are given here. Click on
Execute with the LHMB
8
The Trace Display window shows a single shot
gather in pairs, comment on some of the things
that you can see in this plot and the possible
causes. After a few minutes, report your findings
to me.
9
Refractions
Seafloor
Reflected noise from surrounding seafloor slopes
Wave slap as streamer is in turn
10
Close the Trace Display window. In the flow
window, click on Automatic Gain Control with the
RHMB button. This uncomments the process. Click
on Execute with the LHMB
Comment on the differences. Click on the top
arrow icon this will take you to the next shot
gather.
11
Filtering
The data obviously need filtering. We are going
to apply a 4-10-70-80 bandpass filter to the data.
4 10 70 80

• Below 4 Hz, no data is passed.
• From 4-10 Hz the amount of data passed is ramped
  from 0 to 100.
• From 10-70 Hz, all data is passed.
• From 70-80 Hz, the amount of data passed is
  ramped from 100 to 0.
• Above 80 Hz, no data is passed.

Go back to the list of flows. With the LHMB click
on 02 Filter Click on Bandpass Filter with the
MMB
12

• This specifies that we are going to apply a
  single Ormsby zero phase bandpass filter with the
  parameters specified. Click on Execute with the
  LHMB.
• How does the filtered version compare to the
  unfiltered version (you can display both at the
  same time by commenting out Bandpass Filter and
  re-running the flow.
• Try different filter parameters.

13
Common Mid Point (CMP) Gather
In your 01 - Trace Display flow change the sort
to CDP as above. Click on Execute with the LHMB
14
(No Transcript)
15
CMP - Normal Moveout
Refractions
Seafloor
Direct Wave
The shape of the seafloor reflector forms a
hyperbola. We can use this shape to calculate the
value of vrms down to this point. The same
applies for subsequent reflectors below the
seafloor. Note secondary annotation is by offset.
16
Normal Moveout
The graph of travel time of reflected rays
plotted against offset distance is a hyperbola
whose axis of symmetry is the time axis. Normal
moveout (NMO) at an offset distance x is the
difference in travel time ?T between reflected
arrivals at x and at zero offset.
From Kearey, Brooks, and Hill, 2002
Rearranging, we can calculate velocity as a
function of zero-offset reflection time and NMO
17
Velocity Analysis

• Velocity analysis uses NMO to estimate vrms for
  layers in a CMP gather.
• Prior to velocity analysis a supergather is
  typically formed adjacent CMPs are combined to
  increase the resolution of the CMP display
• In this case, I have combined 5 adjacent CMPs.
• Though this has been done for you, you can
  examine the flow (03 Supergather).

18
Semblance window velocity vs. TWT

• Peaks in semblance mark areas where the data will
  stack.
• Beware of seafloor multiples.
• In the semblance window, zoom into the area
  1400-3000 m/s, 1000-3000 ms.

CMP gather window offset vs. TWT
19

• With the LHMB, click on the bottom icon on the
  left hand of the screen.
• Move the mouse cursor around in the semblance
  window. In the CMP gather window you will see a
  travel time curve corresponding to the NMO given
  by the location of the mouse cursor in velocity
  and TWT.
• Try and pick an NMO curve that fits the seafloor
  reflector. Click the LHMB to select that point.
  Continue with the other reflectors to create a
  series of points defining a vrms curve.
• At this point I will come around and help
  individually.
• To apply the NMO velocities, select Gather
  Apply NMO
• Save velocities (File Save Picks)

Seafloor reflector
20
Top Mute
As this survey uses a long streamer, when we
apply NMO the data from the channels at long
offset is excessively stretched at shallow travel
times. This needs to be muted out.
21
Bottom Mute
We can also remove some of the data at small
offsets just above the multiple. This will
attenuate the multiple when we stack the data.
22
Execute the 05 Top and Bottom Mutes flow
Zoom in on the Trace Display plot
23
With the LHMB select Picking Pick Top Mute.
Enter a table name of geo369 top (or select
geo369 top). Click on the pick icon and pick some
points. With the LHMB select Picking Pick
Bottom Mute. Enter a table name of geo369 bottom
(or select geo369 bottom). Click on the pick icon
and pick some points
Top mute
Bottom mute
24
Stacking
Now that we have applied NMO, top, and bottom
mutes to our data, we can stack the data. This
involves summing all of the traces in a CMP to
create one trace.
25
From the list of flows, with the LHMB, select 06
- Stack

• If we were fully processing this line, we may
  have carried out a number of other processes at
  this stage, i.e
• Trace kills, trace statistics
• Trace amplitude corrections
• Radon Filtering
• Dip Moveout
• f-k filteing
• Pre-stack decon, etc, etc.
• Click on Execute with the LHMB. This will create
  a stacked seismic section.

26
Once the stack job has finished, go to the list
of flows and select 07 Display Stack. Execute
We now have a section in which we can start to
see some geology. However, this section needs
migrating note numerous diffractions and
crossing reflectors.
27
Migration
In the simplest case, a dipping reflector plots
in the wrong location, with the wrong dip.
Migration corrects for this.
From Kearey, Brooks, and Hill, 2002
28
Migration
From Kearey, Brooks, and Hill, 2002
In the top examples, the actual structure and how
it would be imaged in an un-migrated seismic
section is shown. An example of our un-migrated
seismic data is shown for comparison.
29
Select the flow 08 - Migration
An important part of migration is the velocity
model. This has to be optimized for the geology
of our region. For this example, we are assuming
a simple linear gradient. Click on Execute with
the LHMB When the job is finished, select the
flow 09 Display Stack. Click on Execute with
the LHMB.
30
Note how much clearer the faults are in the basin
enlarged view on next slide.
31
Migrated
Stacked
32
Looking at this seismic section, can you see any
obvious artifacts??
33
References Used

• Kearey, P., M. Brooks, and I. Hill, An
  Introduction to Geophysical Exploration, 2002.
• Dobrin, M.B. and C.H. Savit, Geophysical
  prospecting, 1988.