Ocean Carbon and Biogeochemistry Data System

SOFeX 2002 Polar Star cruise
Data processing notes for Seabird Seacat CTD profile data
from Transect 1 and BC Bottle Casts

2 May 2007: Prepared for OCB data system by Terry McKee (PO Dept., Woods Hole Oceanographic Institution)

Original data (ascii .cnv and hex .hex files) and Seabird calibration files were copied to DMO server from Ken Buesseler's post-cruise CD-ROM and then to processing computer.

Comments lines from original Seabird data files
Bottle Casts (BC) 3-6
* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_bottlecastnumber3.hex
** kevlar line bottle cast
** Bottle cast number 3, 2_17
* cast 5 02/17 03:32:12 samples 11382 to 14610 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_bottlecastnumber3.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_bottlecastnumber4.hex
** Bottle cast number four
* cast 0 02/17 23:15:10 samples 0 to 3227 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_bottlecastnumber4.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_bottlecastnumber5.hex
** Bottle cast number 5
* cast 0 02/18 04:29:15 samples 0 to 2759 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_bottlecastnumber5.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_bottlecastnumber6.hex
** Bottle Cast number six
* cast 0 02/18 22:49:58 samples 0 to 4594 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_bottlecastnumber6.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

Transect 1 casts A-H
* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castA.hex
** kevlar line bottle cast
** Transect One, Cast "A"
* mode = PROFILE ncasts = 6
* cast 0 02/16 03:40:30 samples 0 to 2504 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castA.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castE.hex
** kevlar line bottle cast
** Transect One, Cast "E"
* mode = PROFILE ncasts = 6
* cast 1 02/16 06:38:02 samples 2505 to 5244 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castE.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castF.hex
** kevlar line bottle cast
** Transect One, Cast "F"
* cast 2 02/16 08:00:56 samples 5245 to 7283 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castF.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castG.hex
** kevlar line bottle cast
** Transect One, Cast "G"
* cast 3 02/16 09:49:54 samples 7284 to 9256 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castG.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

* FileName = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castH.hex
** kevlar line bottle cast
** Transect One, Cast "H"
* cast 4 02/16 11:31:19 samples 9257 to 11381 sample rate = 1 scan every 0.5 seconds stop = switch off
# datcnv_in = C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_transect1_castH.hex C:\Program Files\Sea-Bird\SOFEX\SBE19_sofex_with_seapoint.con

A header check indicated that the variables were:
timeJ   time   t090C   c0S/m   density00   depSM   sal00   flSP   prSM   flag

To derive parameters like salinity, potential temperature, and density, .con files were copied from the post-cruise CD-ROM (dmoserv1)
Seabird processing sequence beginning with program Datcnv was used.
Pressure and fluorescence were included.

Procedure followed was as outlined in Seabird document on processing SBE19.

• Datcnv;
• AlignCTD - advanced T by 0.5 (standard for SBE19)(did not advance C - having no good reason to (i.e., could not observe spikes in S at high T gradients)
• Wildedit
• Celltm -- alpha .042 beta 10.0 (standard for SBE19)
• Filter -- using A with time constant 0.05 for all variables except pressure, for which I used B with time constant 0.10 (standard for SBE19)
• Loopedit
• Binavg to 1db
• Derive - deriving Salinity; Potential Temperature; and 3 densities.
• Split to uptrace and downtrace.

Example Bottle Cast CTD cast header record from a final processed file:
Bottle Cast CTD cast header

Example Transect 1 CTD cast header record from a final processed file:
Transect 1 CTD cast header

Checkplot revealed surface had some bad data averaged in. Solution was to remove scans before cast descended, then reprocess to _X directory

Truncations done in datcnv
bottlecast3 - scan 110
bottlecast4 - scan 210
bottlecast5 - scan 140
bottlecast6 - scan 85
transectA - scan 315
transectE - scan 310
transectF - scan 485
transectG - scan 340
transectH - scan 422

Doing the truncation made the 1db averaged surface values correct.
Overplotted Salinity from original .cnv, final 1db to check that there were no errors introduced during processing.
Compared downtrace, uptrace, and bottle salinities.
Trace is spikey, but this can be explained by the fact that the cast was stopped frequently to attach
water bottles, which stirs up the water and confuses the conductivity sensor. Except at the surface, the trace compares reasonably well with the bottle data.

The plot below explains the offset between the finished data and the bottle salts,
using SBE19-5 as an example. The magenta is up and downtraces of raw files (datcnv from hex.)

Since there was so much scatter, the bottles line up with some point on the line.

Differences were as follows:

[saldiff(use) bdepth(use)]
For station 3: stdv = .0335
-0.0311 20.0000
-0.0740 40.0000
-0.0778 60.0000
-0.0118 80.0000
-0.0012 100.0000
-0.0110 120.0000

For station 4: stdv = .0557
0.0276 20.0000
-0.0058 40.0000
-0.0687 60.0000
-0.1343 80.0000
-0.0603 100.0000
-0.0500 120.0000

For station 5: stdv = 0.1160
-0.0068 20.0000
0.0041 40.0000
0.2922 60.0000
0.1253 80.0000
0.1961 100.0000
0.1747 120.0000

For station 6: stdv = 0.0815
-0.0042 40.0000
0.1110 120.0000

Differences are not due to drift, but possibly to the conditions under which the casts were taken -- regularly halting the cast to attach sample bottles to the kevlar line during the downtrace, magnified by the additional affect of ship heave during that process. Plots of temperature and conductivity did not indicate any lag in the sensor response.

Process of converting from Seabird 1db-averaged to OCB format was similar to that used for the CTD profiles. Seacat header is slightly different, so conversion program had to be modified to meet its specifications. Fluorometer readings indicated that fluorometer data were questionable:

dSBE19_3.cnv:# span 9 = 0.0000e+00, 0.0000e+00 a definite ND
dSBE19_4.cnv:# span 9 = 9.1475e-02, 2.4448e+00 could be real
dSBE19_5.cnv:# span 9 = 9.6198e-02, 3.7718e+00 could be real
dSBE19_6.cnv:# span 9 = 8.7419e-02, 6.5368e+00 could be real
dSBE19_A.cnv:# span 9 = 6.1050e-02, 6.1050e-02 probably garbage
dSBE19_E.cnv:# span 9 = 6.1050e-02, 6.1050e-02 " "
dSBE19_F.cnv:# span 9 = 6.1050e-02, 6.1050e-02 " "
dSBE19_G.cnv:# span 9 = 6.1050e-02, 6.1050e-02 " "
dSBE19_H.cnv:# span 9 = 6.1050e-02, 6.1050e-02 " "

Contact: Ken Buesseler (WHOI)

  Go back to your previous page