Hydro camp 2008 Sub-bottom
Processing
The data was collected with a
Knudsen 3.5 kHz chirp echo sounder and stored in the Knudsen KEB format
(binary format). The web site is:
http://www.knudsenengineering.com/html/company/archives2002.htm
The following are the steps in
the sub-bottom processing, the command lines are shown in italics:
1. Convert Knudsen format to OMG Gloria format using command line:
knud2glo \
-omg -VX
-min16_LF 0 -max16_LF 16000 -min16_HF 0 -max16_HF 16000
-out line1 -in line1.keb
The variables are defined as
follows:
-omg specifies format
-VX version
-min16_LF specifies where to start for low frequency
-min16_HF specifies where to start for high frequency
-max16_LF specifies where to end for low frequency
-max16_HF specifies where to end for high frequency
-out specifies output file name
-in specifies input file name
2. View merged data:
swathed
line1.merged
The following is an example of the image generated:
In this screen you can use the SS (sidescan) function key. The default image that is shown is the high frequency, by using channel button it will switch to low frequency:
To enhance the image use the
log/lin function which uses a logarithmic conversion. This does
not change anything in the file and is strictly for viewing. You
may also look at the bottom pick as selected by the Knudsen for each
frequency. By using the space bar, it allows you to scroll
through.
3. Covert from 16 bit to 8 bit data:
k16to8 \
-log -in
line1.merged.16_data -out line1.merged.ss_data -min16 100 -max16 16000
The variables are defined as
follows:
-log convert using logarithmic
-in input line
-out output line
-min16 minimum
-max16
maximum
At this stage it is a good time to restrict the levels that you are converting. The following images are of linear conversions with different minimum and maximum parameters:
Linear conversion with minimum 0
and maximum 16,000
Linear conversion with minimum 0
and maximum 1000
4. Apply "Course Made Good" to file to determine proper position for transducer:
useCMG
line1.merged
The CMG command uses the GPS
positions to the merged line to give the heading.
5. Merge navigation file to account for offset:
mergeNav
-delay 0.0 -ahead -2.0 -right 0.0 -below 0.0 line1.merged
Will use above heading to apply
correction for the offset of the transducer from the GPS antenna.
6 . Merge multibeam depth and attitude data:
mergeOtherSonarData \
-in
line1.merged -userealTWTT -replace_depth -replace_attitude -draft -0.30
-auto_delay /em3002/*.mergedhttp://start.ubuntu.com/8.04/
This command uses the more
correct depths as collected by the multibeam sonar. The variables
are defined as follows:
-in line1.merged input merged file
-userealTWTT use real two way travel time
-replace_depth replaces the Knudsen bottom pick with the multibeam derived bottom pick
-replace_attitude uses multibeam bottom pick in the sub-bottom merged file
-draft -3.0 indicates difference between the Knudsen and the multibeam
-auto_delay/em3002/*.merged corrects for time offsets between the Knudsen and multibeam systems
7. Make plot of profile
time vs depth:
makeSeisPlot
-mindepth 0 -maxdepth 70 -below -nosam 700 -out kseis/line1.LF -lf -in
line1.merged
The variables are defined as
follows:
-mindepth 0 minimum depth you want to see in your final profile
-maxdepth70 maximum depth you want to see in your final profile
-below indicates that you only want to see the data below the bottom pick
-nosam 700 indicates the number of samples used
-lf indicates the low frequency transducer
-in line1.merged
sub-bottom merged line
8. View plot:
jview -seis
-initstr 255 0 -autoj line1.LF &
9. Create file containing
boundary of line limits:
getBounds
line1.merged
10. Make blank grid for
viewing navigation:
make_blank
-boundsof *.merged
11.
Select Projection Area:
jview
image.r4 -navlook line1.nav -pickrot
The first step is to
identify a 'box' that is oriented along the
direction of the sub-bottom line. This is done using the pickrot
option in jview. As shown in the first image, you can identify
the red survey line. The second image shows the 'bounding' box
being manipulate using
the arrow keys. The third image shows the
'bounding' box encapsulating the survey line.
12. Make plot of profile
projected by position vs depth:
makeSeisPlot
\ -nosam 750 -lf -mindepth 0 maxdepth 100 -map Box.header0 -pixel 10
-fill_map_gap \ -first 0 -last 100000 -out line1.map.LF -in
line1.merged
This command line allows the
profile to be projected on a vertical plane. The variables
are defined as follows:
-nosam 750 indicates the number of samples used
-lf indicates the low frequency transducer
-mindepth 0 maxdepth 100 minimum and maximum depth of final profile
-map Box.header0
-pixel 10
-fill_map_gap attempt to fill any gaps in the data
-first 0
-last 10000
first
and last options enable you to identify the starting and ending ping in
the line
13. View profile plot:
jview -seis
-initstr 255 0 -autoj $4.map$3.LF
14. Convert to gif:
mk_annotated_gif $SUBDIR $LINE_NUMBER
stretchacres -pgm -low 0 -high 255 -invert -in line1.LF -out line1.pgm
convert -flip -rotate 90 line1.pgm line1.gif