Time Coded Impulse Seismic Technique
LACS test week 14 - 2006

The Low-frequency Acoustic Source (LACS) manufactured by Naxys is used in the verification of Time Coded Impulse Seismic Technique (TCIST), currently being developed by the The University of Bergen, in collaboration with Statoil and Naxys. The TCIST development is part of a Ph.D. project that is described here.

Click to see photos from sea trial 5 April 2006.

1. Test procedure

Click to download test procedure (PDF, 600 kB)

Click to download test procedure.

2. Test equipment from Dept. of Earth Science

Item

Description

Specifications etc.

Mini-streamer, two channels

Manufacturer: Mark Products Inc., Houston, USA
Mfr date: 7/3/1980
Configuration:
- Channel 1 (front): 21 ea hydrophones, 51 cm spacing
- Channel 2: 8 hydrophones, 1.75 m spacing
Report from 2005 calibration.

Seismic recorder

Mfr: Geometrics
Model: GEODE
S/N: 3699
Specifications

Interface PC soundcard O/P and GEODE trigger I/P.

Input: Signal from sound card speaker output.
Assumes negative pulse, approx. to -1 V.
Adjust sound card outoul level if needed.
Trigger level can be adjusted by pot. meter.
Will trigger on negative edge.
Generates output pulse 5 ms duration.
Output is closure from opto-coupler.
Keep in mind that this is rapid prototyping!

Click to enlarge

Trigger interface schematics. Click to enlarge. Download PDF version.

GPS connected to GEODE data collecting laptop.

Mfr: Garmin
Model: GPS-35 HVS
S/N: 71998362
Specifications
Manual: Technical Specification, Rev. E, Mar, 2000 Download (PDF, 139 kB)

GPS used for differential correction

Mfr: Garmin
Model: GPS-35 HVS
S/N: 71951015
Specifications
Manual: Technical Specification, Rev. E, Mar, 2000 Download (PDF, 139 kB)

Used a Garmin GPS-35. Unfortunately, logging of GPS stopped after a short while, for unknown reasons. Will be investigated further.

3. Test sequence overview (5 April 2006 - B.A.)

Test #	Description	Setup Note: Streamer offset measured between vessel stern and center of first hydrophone group in streamer. LACS positioned approx. 5 m forward from stern, on port side.	FFID range (called fldr in trace headers)e	Recording parameters	Raw nav data files
1	LACS far field signature test Sequence 1 test Sequence 2 test Sequence 1 and 2 test	LACS at 5m depth Far-field hydrophone at 85-90 m Sensitivity=-195 dB rel V/uPA Ch. 4 in recording Near-field hydrophone at 7 m Ch. 3 in recording Depth according to map: 200-285 m	FFID1000-1195 FFID1196-1205 FFID1206-1216 FFID1217-1228	1000 Hz Channel assignment: #1 = Mini-streamer first hydrophone section #2 = Mini-streamer second hydrophone section #3 = Near-field hydrophone #4 = Far-field hydrophone	Here
2	LACS as seismic source Sequence type 1 LEG1	Vessel speed 2.4-2.8 kt LACS at approx. 5m Streamer offset 60 m	FFID1229-1287 (logging towards starting point of LEG1) FFID1288-1440 (LEG1)	1000 Hz Recording time:12 s Sequence interval: 16 s Channel assignment: As per test #1	-"-
3	LACS as seismic source Sequence type 1 and 2 LEG2-a1	Vessel speed 2.4-2.8 kt LACS at approx. 5m Streamer offset 60 m	FFID1441-1451 (logging towards starting point of LEG2-a FFID 1452-1491 (LEG2-a) LACS misfires	1000 Hz Recording time:12 sec Sequence interval: 16 s Channel assignment: As per test #1	-"-
4	LACS as seismic source Sequence type 1 and 2 LEG2-a2	Vessel speed 2.4-2.8 kt LACS at approx. 5m Streamer offset 60 m	FFID1492-1545(LEG2-a2) LACS misfires	1000 Hz Recording time: 12 s Sequence interval: 16 s Channel assignment: As per test #1	-"-
4	LACS as seismic source Fixed firing rate:0.476 Hz. LEG2-b	Vessel speed 2.4-2.8 kt LACS at approx. 5m Streamer offset 60 m	FFID1546-1811	1000 Hz Recording time: 0.8 s Every second pulse recorde Channel assignment: As per test #1	-"-
5	LACS as seismic source Sequence type 1 and 2 LEG3	Vessel speed 2.4-2.8 kt LACS at approx. 5m Streamer offset 60 m	FFID1812-1968	1000 Hz Recording time:12 s Sequence interval: 16 s Channel assignment: As per test #1	-"-
5	Noise measurement	Vessel speed 2.4-2.8 kt LACS shut off Streamer offset: 60 m	FFID1969-1978	1000 Hz Recording time:12 s Channel assignment: As per test #1	-"-

4. Nav data

4.1 Processing

Raw navdata is here.

Navdata is not merged with seismic data (into SEG-Y trace headers). Instead, GPS info associated with each recording is placed in separate file, with GPS info in alternating lines; e.g.:

$GPGGA,091614,6028.0945,N,00514.6330,E,1,10,1.2, - Received at 09:16:49.21 for File 1166
 FFID    1166 (Stack  1, Shot Loc: 0 Meters) 09:16:14.00 04/05/2006   66 KBytes SAVED IN 1166.SGY
$GPGGA,091619,6028.0945,N,00514.6326,E,1,10,1.2, - Received at 09:16:54.31 for File 1167
 FFID    1167 (Stack  1, Shot Loc: 0 Meters) 09:16:19.00 04/05/2006   62 KBytes SAVED IN 1166.SGY

We need to extract Recording Number - called Shotpoint = SP from now - and time, latitude (lat) , longitude (lon) from these files, and also convert lon/lat to UTM Easting/Northing (these fields are highlighted).

Conversion between lat/lon and UTM: Download this Python library http://www.pygps.org/LatLongUTMconversion-1.1.tar.gz.
Unpack it, and run "python setup.py install" as root. In order to build Python modules you must first install 'python-devel'. In SUSE: Yast -> Install and remove software -> Select filter: "Package Groups" -> Development -> Languages -> Python. There you will find "python-devel".

This python script extracts & converts the needed parameters. Input nav data file is given as command line parameter. Output to stdout, so just redirect to file to save output.
To process all *.log files in a directory write the following. Processed files with have the extension ".extract" added to the filename.

for FILE in `ls *.log`; do echo $FILE; ../software/navdata-processing.py $FILE > $FILE.extract; done

Then merge all processed nav data files into one file called "navdata.txt", also sorting by SP, by typing:

cat *.extract | sort > navdata.txt

4.2 Plotting

Have used Generic Mapping Tools (GMT) to make shotpoint maps.

Here is example map in png format (click to enlarge).

Example SP map (click to enlarge).

Example scripts:

Bash script to make map above
Script to plot far-field signature test position
Trial with UTM grid markers included (1x1 km grid cell size, must be improved). Must first generate UTM grid points by this Python script.
PNG result is here.

GMT generates PostScript output. Convert to other formats:

PNG, JPG: Use "convert", which is part of ImageMagick. Output filename extension determines output format.
The -density option determines size (in pixels) of output image.

convert -rotate 90 -trim  -antialias -density 120x120 SP-plot-with-UTM-grid.ps SP-plot-with-UTM-grid.png

PDF: Use "ps2pdf", which is part of GMT:

ps2pdf SP-plot-with-UTM-grid.ps SP-plot-with-UTM-grid.pdf

5. Seismic data

Data is in SEG-Y format.

Here's an overview of SEG-Yformat
The SEG-Y format specification from SEG:

We'll process data in Seismic Unix (SU). To see all options to a command, just write the command name itself.

5.1 Importing SEG-Y data to Seismic Unix

In the following we will use the far-field LACS signature test. Data is in 1166.sgy:

segyread tape=1166.sgy endian=0 conv=1 | segyclean > 1166.su

Convert all SEG-Y files in a directory:

for FILE in `ls *.sgy`; do echo $FILE; segyread file=$FILE conv=1 endian=0 | segyclean | >  ${FILE%.*}.su; done

${FILE%.*} is a so-called "string operation" in Bash; it strips off existing file extension ".sgy", so we can replace it by ".su".

Parameters:

conv=1 : According to SEG-Y standard, floating point samples should be in IBM format. Modern PCs use IEEE 754 floating point format, so setting conv=1 will convert samples from IBM to IEEE floating point format (here is some background information on problems caused by mix-up of these formats).
endian=0: PCs are little-endian machines, so this option should be set to zero.

5.2 Viewing trace header information

First check which trace header parameter fields are in use, and their min/max values, by the surange command:

olem@linux:~/www/Naxys/test-week-14-2006/data> surange < 1166.su
180 traces:
 fldr=(1166,1195)  tracf=(1,6)  trid=1 nvs=1 nhs=1
 scalco=1 gx=(0,5)  counit=1 ns=2500 dt=1000
 year=6 day=95 hour=9 minute=(16,18)  sec=(4,59)
 timbas=1

Here's a list of all trace header parameters. In this particular file the fldr (=SP) goes from 1166 to 1195, and tracf from 1 to 6 (most likely 6 channels were recorded by mistake - should only be 4 - but it's easy to discard ch. 5 and 6, see later).

Use suxedit to investigate individual trace headers, and also "plot" each sample:

olem@linux:~/www/Naxys/test-week-14-2006/data> suxedit < 1166.su

suxedit: ! examine only (no header editing from STDIN)

180 traces in input file
 fldr=1195 tracf=6 trid=1 nvs=1 nhs=1 scalco=1
 gx=5 counit=1 ns=2500 dt=1000 year=6 day=95
 hour=9 minute=18 sec=39 timbas=1
> p
    1 -1.3893e+00                     --------|
    2 -2.0331e+00                 ------------|
    3  1.5508e+00                             |++++++++
    4  4.5063e+00                             |+++++++++++++++++++++++++
    5  2.1204e+00                             |+++++++++++
    6  2.2197e+00                             |++++++++++++
    7  3.6537e-01                             |++
    8 -2.5042e+00              ---------------|
    9  1.0019e+00                             |+++++
   10  3.3873e+00                             |+++++++++++++++++++
   11  1.7798e+00                             |++++++++++
   12 -7.7628e-01                        -----|
   13 -5.1471e+00-----------------------------|
   14 -2.3701e+00               --------------|
   15  4.5092e-01                             |++
   16 -3.1452e+00           ------------------|
   17 -1.2180e+00                      -------|
   18 -2.1906e+00                -------------|
   19 -4.8839e+00 ----------------------------|
>

5.3 Selecting specific traces from a SU data file

The "suwin" command selects specific traces from a data file, based on criteria that is entered on the command line. All examples refers to file 1166.su mentioned above.


suwind key=fldr min=1195 max=1195 < 1166.su | suxwigb
suwind key=fldr min=1195 max=1195 < 1166.su | suwind key=tracf min=4 max=4 | suxwigb
suwind key=fldr min=1195 max=1195 < 1166.su | suwind key=tracf min=4 max=4 |\
   sufilter f=5,10,200,400 | suxwigb
suwind key=tracf min=4 max=4 < 1166.su > 1166-farfield.su
suwind key=tracf min=4 max=4 < 1166.su | sufilter f=5,10,200,400 | suxgraph style=normal
suwind key=fldr min=1166 max=1166 < 1166.su |suwind key=tracf min=4 max=4 |\
 sustrip outpar=/dev/null | b2a n1=1 outpar=/dev/null > 1166-SP1166-ch4.txt
suwind key=fldr min=1166 max=1166 < 1166.su |suwind key=tracf min=3 max=4 |\
 sustrip outpar=/dev/null | b2a n1=2 outpar=/dev/null > 1166-SP1166-ch3+4.txt
suwind key=fldr min=1166 max=1166 < 1166.su | suwind key=tracf min=4 max=4 |\
 sugabor | suximage
suwind key=fldr min=1166 max=1166 < 1166.su | suwind key=tracf min=4 max=4 |\
 suspecfx | suxgraph style=normal
This is a Python script. Download from here.

NOTE: For diagram annotations, grids, axis text and headings in Seismic Unix plots - see examples below.

6. Data preview

We will just show some results from the LACS far-field signature test (single shot). All these plots made with Seismic Unix.

Thumbnail image - click to enlarge	Script	Comments
Fig. 6.1: Far-field signature, FLDR 1166-1195, BP filter: 10,20,200,400	./software/plot-ff-test.sh	Some jitter in both starttime and amplityde. Compare to similar test in 2004 with LACS, and with airgun. NOTE: Time axis grids differ in size! Also signal polarity must be inverted. To do: Verify that recording system does not introduce any jitter of significance.
Fig. 6.2: Far-field hydrophone, Gabor spectrogram, FLDR=1166	./software/plot-gabor-ff-test.sh
Fig. 6.3: Detail of far-field hydrophone, Gabor spectrogram, FLDR=1166	./software/plot-gabor-ff-test.sh

Time Coded Impulse Seismic Technique LACS test week 14 - 2006