


     GP4-97-KA
     NOAA Ship Ka'imimoana
     Kwajalein - Honolulu, HI
     June 8 - July 3, 1997

     Chief Scientist: Mr. Andy Shepherd
     Survey Department: CST Dennis Sweeney
     CTD Personnel: D. Sweeney, M. O'Haleck
     Final Processing: K. McTaggart


     ACQUISITION: 

     Thirty-three CTD profiles were collected on this cruise.  A deep test cast
     was made enroute to the first station along 165E.  Then seventeen CTD 
     casts were made along 165E from 8N to 8S.  Fifteen profiles were collected 
     along 180W from 5S to 8N.  The majority of CTD casts were to 1000 m; 
     one was to 500 m, one to 1300 m, and three were deep (>3000 db).

     The ship's Sea-Bird 9plus CTD s/n 09P10493-0405 measuring pressure (s/n 
     61183), temperature (s/n 1708, 2027), and conductivity (s/n 1536, 1537)
     (PMC8.CON same as PMC7.CON) was used throughout the cruise.  

     The CTD was mounted in a custom 24-bottle frame with a Sea-Bird rosette 
     sampler (s/n 88).  The CTD data stream was passed through a Sea-Bird 
     11plus deck unit (s/n 392) with factory settings.  An analog signal was 
     recorded onto the audio portion of VCR tape as a backup.  Digitized data 
     were sent to a Dell 4100 personal computer equipped with Sea-Bird's 
     SEASOFT acquisition software (version 4.216) where calibrated data were 
     displayed in graphical form in real-time, as well as stored in raw form 
     onto the hard disk.  Backups of the raw data were made on QIC-80 1/4" 
     cartridge tapes and returned to PMEL for post-cruise processing.

     Stations 1-4 were originally acquired using the wrong configuration 
     file (PMC6.CON).  These casts were reprocessed through the Seasoft 
     modules with the correct conductivity calibration coefficients 
     (SPECIAL>CON) at the lab post-cruise before calibrations.  There were
     no raw data files for KA40241, and the VCR recording was errorneously
     incomplete.  So there is no cast KA40241 in the final data set.  
     Final data for station 30 were truncated at 460 db on the downcast
     due to noisy data in all channels as a result of cable damage.
     

     SALINITIES:

     For calibration purposes, bottle samples were usually taken at 6 depths 
     on 1000 m or shallower casts and at 10 depths during deep casts.  Two 
     bottle samples were taken at the deepest depth.  Duplicate samples were 
     analyzed on a subsequent day from the rest.  Salinity analysis was 
     performed using Guildline Autosal 8400B salinometer s/n 61.383 (last 
     calibrated at NRCC February 13, 1996).  IAPSO standard seawater batch 
     #P130 was used for all casts.  NRCC calibrations were not applied to 
     this data set, only a drift-during-run linear interpolation correction 
     in ship program DISAL.  Standard operating temperature was 27 degrees 
     Celsius.
 

     POST-CRUISE CONDUCTIVITY CALIBRATIONS:

     GP497S.CAL, not including duplicate salts, was created at sea using batch
     routine MAKECAL.  Anomalous differences between CTD and bottle salinities
     were scrutinized.  Station 30 bottle data were removed from the .CAL file
     because CTD values were bad.  Samples from bottle number 5 tripped at 200 
     db were often compromised.  These bottle salinities are flagged as bad and
     questionable.

     Final pressure and temperature calibrations were pre-cruise.  Conductivity
     fit coefficients were determined for all casts as one group, stations 1-29.
     These were best determined using a third order station-dependent fitting 
     routine, CALCOS3, written by Greg Johnson in MATLAB.  NOTE: conductivity
     calibrations were redone on January 6, 1999.  See below.  

    			number of points used   171
	    		total number of points  197
	  		% of points used in fit  86.8
	  		fit standard deviation  0.001163
	  		fit bias -0.0074280895 mS/cm
	  		min fit slope 1.0004942
	  		max fit slope 1.0005485 

     Slope and bias values were applied to CTD data using PMEL Fortran program
     GP497_EPIC; and to bottle files using CALMSTR4.


     FINAL PROCESSING:

     The following are the standard SEASOFT processing modules used to
     reduce Sea-Bird CTD data:

     DATCNV converts raw data to engineering units and creates a bottle
     file if a Sea-Bird rosette sampler was used.  (MARKSCAN creates a 
     bottle file if a General Oceanics rosette was used.)

     ROSSUM averages the bottle data specified in the DATCNV or MARKSCAN
     output and derives salinity, theta, sigma-t, and sigma-th.  These
     bottle files are transfered to the PMEL VAX where post-cruise 
     calibrations are computed.

     WILDEDIT makes two passes through the data in 100 scan bins.  The
     first pass flags points greater than 2 standard deviations; the
     seond pass removes points greater than 20 standard deviations from
     the mean with the flagged points excluded.

     CELLTM uses a recursive filter to remove conductivity cell thermal
     mass effects from the measured conductivity.  In areas with steep
     temperature gradients the thermal mass correction is on the order
     of 0.005 psu.  In other areas the correction is negligible.  The
     value used for the thermal anomaly amplitude (alpha) is 0.03.  The
     value used for the thermal anomaly time constant (1/beta) is 9.0.
 
     FILTER applies a low pass filter to pressure with a time constant of
     0.15 seconds, and to conductivity with a time constant of 0.03 seconds.
     In order to produce zero phase (no time shift) the filter is first
     run forward through the file and then run backwards through the file.

     LOOPEDIT removes scans associated with pressure slowdowns and
     reversals.  If the CTD velocity is less than 0.25 m/s or the pressure
     is not greater than the previous maximum scan, the scan is omitted.

     BINAVG averages the data into 1 db bins.  Each bin is centered around
     a whole pressure value, e.g. the 1 db bin averages scans where pressure
     is between 0.5 db and 1.5 db.

     DERIVE uses 1 db averaged pressure, temperature, and conductivity to
     compute salinity, theta, sigma-t, sigma-th, and dynamic height.

     SPLIT removes decreasing pressure records and keeps only the downcast
     data.

     TRANS converts the data file from binary to ASCII format.  These 
     data are transfered to the PMEL VAX.

     PMEL program GP497_EPIC applies post-cruise conductivity calibration 
     coefficients, recomputes the derived variables in DERIVE, and converts
     the ASCII data files to EPIC format.  GP497_EPIC skips bad records 
     near the surface (typically the top 5 m) as well as any records containing 
     -9.990e-29, and extrapolates raw data to the surface (0 db) within 10 db.  
     Because the SBE module LOOPEDIT does not handle package slowdowns and 
     reversals well in the thermocline where gradients are large, GP497_EPIC 
     removes raw data records where a sigma-theta inversion is greater than 
     -0.01 kg/m3.  Data are linearly interpolated such that a record exists
     for every 1 db.  When data are extrapolated to the surface, the WOCE
     quality word is '888'; when interpolated over greater than 2 db, the 
     WOCE quality word is '666'.  The WOCE quality word consists of a 1-digit
     flag for pressure, temperature (ITS-90), and salinity.

     PMEL program CALMSTR4 applies post-cruise conductivity calibration
     coefficients and recomputes the derived variables in ROSSUM.  EPICBOMSTR
     converts the ASCII bottle data file into individual cast EPIC data files.
     There are no .BOT files for station 0281 and 0301.

     Final CTD and bottle files were moved to DISK$EPIC1:[HAYES.DATA] and 
     included in the RIM data management tables on July 17, 1997.

********************************************************************************

     January 6, 1999:  While generating the 1997-1998 NOAA CTD data report, it
     was decided to redo the conductivity cals of GP497, splitting them into two
     groups, stations 1-17 and 18-30.  A 5 day transit occurred between these
     two stations and a .002 jump in conductivity residuals was evident.  Calcop3
     was used to determine the best fit for each group:

     stations 1-17:  number of points used   101
		     total number of points  122
		     % of points used in fit 82.79
		     fit standard deviation  0.0008134
		     fit bias                -0.0087830648
		     fit co pressure fudge   4.1647099e-007
		     min fit slope           1.0004982
		     max fit slope           1.0005514

    stations 18-30:  number of points used   64
		     total number of points  75
		     % of points used in fit 85.33
		     fit standard deviation  0.0007616
		     fit bias                -0.012301793
		     fit co pressure fudge   2.2680421e-006
		     min fit slope           1.0006075
		     max fit slope           1.0007091

Programs were modified to apply these new coefficients.  Bad bottles were
flagged in GP497S.CLB and excluded from the EPIC .bot files: station 1 sample
102 and 106, station 2 sample 103, station 10, 11, 15, 17, and 22 all sample
105.  DISK$EPIC1:[HAYES.DATA.GP497.CTD] was updated.  There was no need to
update the RIM/INGRES tables.
***********************************************************************************
