Multi-Channel Seismic Reflection System used from
1977-1980 by Scripps Institution of Oceanography
From Paul Henkart's memory after 20 years!
DATA ACQUISITION
---- -----------
Scripps used an array of 3 Bolt airguns (20, 120, and 40 cubic
inches), operating at 1800 - 2000 psi, and firing simultaneously
every 15 seconds (150ft at 6kts). The signal was received by a Seismic
Engineering Company 24 channel hydrophone array, digitized by a Geospace
1590 Digital Field System (binary gain amplifier), and recorded on a
Bell and Howell serial 7 track tape at 16,600 bpi. Twelve seconds of data
were recorded at .001 or .002 second sample interval. The high density
tapes permitted approximately 28 minutes of data to recorded per track.
resulting in almost three hours of data on one tape.
A Hewlett Packard MX21 computer was used in comjunction with the
high density tapes to produce a real-time four trace stack or a single
trace display on a Versatec Matrix ( 18 in.) raster plotter. The
internal HP clock was used to identify the seismic shots rather than
the Geospace shot number.
The hydrophone streamer was an Exxon research streamer, consisting
of some 150ft. long sections and some 300ft. sections. Initially, in
1977, the streamer had six 150ft. and eighteen 300ft. sections. When
shooting at a 300ft. interval, this resulted in 21 fold subsurface
points alternating with 3 fold data. The 300ft. sections had nonlinear
phone spacing (these diagrams have been lost).
Several of the 150ft. sections were changed to 300ft sections
by adding a 150ft dead section (before?) the live section. The
ranges listed below are as if a 300ft section had been used rather
than 150ft dead section and a 150ft live section (the true range
should be to the center of the live section).
The 150ft. section was a 100ft live section, with linearly
spaced hydrophone, followed by a 50ft stretch section (the live
section being closest to the ship).
There were two types of 300ft sections. Groups 1-9 (1 being the
furthest from the ship) were made up as 200ft of "tapered" phone
spacing. Presumably there was 50ft dead on either side.
Groups 10-20 were made up with a 100ft linear phone spacing section,
presumably between two 50ft dead section. This 200ft section was
then followed by a 100ft stretch section.
The airguns were normally towed at a depth of 10 - 15 feet while
the streamer was set for a depth of 30 feet.
The Geospace usually had a low cut filter of 12Hz with 120db/octave
slope. The high cut filter was always open and a 60Hz. notch filter
was always on. The anti-aliasing filter was set for 125Hz. on 2ms data
and 250Hz. on .001 second data.
INITIAL PROCESSING
------- ----------
The first step in processing at Scripps was to copy the high density
7 track tapes to normal 9 trace tapes. The data were then gained in
order to remove the binary gain applied by the Geospace. These multiplexed
data were demultiplexed and decimated to 4 milliseconds, without
anti-alias filters.
The 9-track tapes had a density of 1600BPI and held ~60MB, thus
a deep water delay was introduced while copying from the high density
tapes to 9-track. SIOSEIS plots default to plotting the data at
the deep water delay first (i.e. the plots do not zero pad from
time zero to the delay). SIOSEIS plots contain a gap whenever the
deep water delay changes.
The preprocessing also included applying a .2 second time shift to
account for a recording delay. This recording delay prevented very
shallow data to be recorded.
The data traces were renumbered to match the industry standard of
numbering the long range trace as the first ( trace 1 is the first in
the water when deploying the streamer).
Shot numbers were created during preprocessing also. Dummy or dead
shots were created based on time and assuming a 15 second rep. rate.
Sixteen shots (four minutes of shooting) were lost every three hours
when the 7-track tape had to be changed.
Duplicate shot numbers are possible. Don't know if a duplicate
shot number means a identical or repeated record or whether the
it's just the shot counter that got reset.
Occasionally an entire 7-track tape track was unreadable causing
a lose of 28 minutes of data. The high density tapes (HDDR?) had been
used by JPL prior to Scripps acquiring them for free.
9-TRACK TAPE RECOVERY
------- ---- --------
Unreadable traces (those with a parity) are dropped. They are not
zeroed or marked. Therefore, a shot may not have 24 traces. No
modifications were made to the headers or data.
NAVIGATION
----------
Scripps' navigation prior to GPS was based on transit satellite
and lots of editing and smoothing using gravity and gyro. Generally,
the navigation is sampled every minute of time.
In most instances there is a time associated with the data. The
time came from the HP21MX internal clock and was inserted into the
SEG-Y header during demux.
PROJECT SCHIZOPHREANIA
------- --------------
The scientists call the cruise by a different project name
than the technicians. The scientist used a name descriptive of
the area of the project and the technicians used a cruise-leg
number.
The ranges listed below came from various notes, but it is
unknown what the measurement points were. It's suspected that
the it's the distance from the transom or the MCS reel to the
closest group and NOT between the guns and the closest group. It
might be just the length of tow leader that was used.
INDP11WT (Indopac Leg 11, R/V Thomas Washington) was Joe Curray's
Andaman Sea cruise.
The streamer had six 150 foot sections, closest to the ship, and
18 300ft. sections further out. The nominal offset is "530ft or
480ft" but when it changed is not known.
Thus the ranges are: 24 450, 18 1350, 17 1650, 1 6450.
INDP12WT (Indopac Leg 12, R/V Thomas Washington) was the
Sunda Arc cruise.
The streamer had six 150 foot sections, closest to the ship, and
18 300ft. sections further out. The nominal offset is 450ft.
Thus the ranges are: 24 450, 18 1350, 17 1650, 1 6450.
GUAY01WT (Guaymas Leg 1, R/V Thomas Washington) was Joe Curray
and Greg Moore's ODP Guaymus cruise.
The streamer had six 150 foot sections, closest to the ship, and
18 300ft. sections further out. The nominal offset is 450ft.
A 150ft dead section was inserted between groups 18 and 19 at
some unknown time.
Thus the ranges are: 24 450, 18 1350, 17 1650, 1 6450.
GUAY02WT (Guaymas Leg 2, R/V Thomas Washington) was Joe Curray
and Greg Moore's ODP Guaymus cruise.
The streamer had five 150 foot sections, closest to the ship, and
19 300ft. sections further out. The nominal offset is 450ft.
Thus the ranges are: 24 450, 19 1200, 18 1500, 1 6600.
MARA09WT (Marianna Leg 09, R/V Thomas Washington) was
The streamer had four 150 foot sections, closest to the ship, and
20 300ft. sections further out. The nominal offset is 450ft.
Thus the ranges are: 24 450, 20 1050, 19 1350, 1 6750
MARA10WT (Marianna Leg 10, R/V Thomas Washington) was
The streamer had four 150 foot sections, closest to the ship, and
20 300ft. sections further out. The nominal offset is 450ft.
Thus the ranges are: 24 450, 20 1050, 19 1350, 1 6750
Guaymus was the first MCS data read at Scripps. The data
were also needed in a hurry in order to meet some DSDP
deadlines.
The original Bell & Howell tapes were written on an HP21mx
computer and were transcribed to 9 track on shore using the
HP21mx and an IBM 1800 with 9 track drives (located in
Shipboard Computer Group's area in the basement of Ritter
Hall). The transcription alone took four times real-time.
The 9 track tapes were taken to the Prime Computer in the
basement of IGPP for demultiplexing.
The year was added during the demultiplex/reformat stage.
I suspect files:
guay02_0205-0078-067-0601z.segy
guay02_0206-0078-067-0704z.segy
guay02_0207-0078-067-0807z.segy
guay02_0208-0078-067-0911z.segy
are named and dated incorrectly because:
1) day 67 is March 8 and the ship was in port by then.
2) The tape index says Guay-02WT Line 10
3) Guay02 Line 11, 0616z 03Mar78 to 0815z 03Mar78 is missing
(but is on film, so check the processed data against it!)
SCS cruises read from 9-track tape:
RAMA05WT
RAMA06WT
RAMA12WT
ARIA03WT
MISSING / UNREADABLE TAPES
------- - ---------- -----
There was a small percentage of tapes that were missing, or
unreadable. A few of the unreadable tapes were mashed and crinkled
beyond repair, we suspect a couple others were erased when they were
sent in for cleaning.
Missing Tapes:
CRGN01WT 44
SMNT01WT 01
MRTN06WT 01,03,04
MRTN04WT 13
ARIA01WT 01,02
ARIA03WT 132
INDP11WT 350-360
INDP12WT 23
MARA09WT 139
MARA10WT 180,181,182
Unreadable Tapes:
ARIA01WT 04,40,44
ARIA03WT 102