                                   If symbols do not display correctly change 
                                   your browser character encoding to unicode 



CRUISE REPORT: AAIW05
(Updated FEB 2012)



HIGHLIGHTS

                        CRUISE SUMMARY INFORMATION

          WOCE Section Designation  AAIW05
Expedition designation (ExpoCodes)  316N20050821
                  Chief Scientists  Dr. Teresa Chereskin
                             Dates  21 August 2005 - 6 October 2005
                              Ship  R/V KNORR
                     Ports of call  Puenta Arenas, Chile - Puerto Monif, Chile
                                    
                                                 45° 18.9' S
             Geographic Boundaries  103° 5.33' W             72° 40.95' W
                                                 62° 0.56' S

                          Stations  136
      Floats and drifters deployed  13 ARGO floats, 20 surface drifters deployed
    Moorings deployed or recovered  0
                                    

                                Chief Scientist:
           Dr. Teresa Chereskin • Scripps Institution of Oceanography
    University of California, San Diego • 9500 Gilman Drive • Mail Code 0230 
                            La Jolla, CA 92093-0230
                         email: tchereskin(at)ucsd.edu 














                            Preliminary Cruise Report

                                 5 October 2005



                               Data Submitted by:



             Shipboard Technical Support/Oceanographic Data Facility

               Kristin M. Sanborn, Teresa Kacena, Dan G. Schuller

             Shipboard Technical Support/Shipboard Electronics Group

                     Carl Mattson, Scott Hiller, Bob Green

                 Shipboard Technical Support/Computing Resources

                                 Frank Delahoyde

                       Scripps Institution of Oceanography

                            La Jolla, Ca. 92093-0214



SUMMARY


A hydrographic survey consisting of LADCP/CTD/rosette sections, underway 
shipboard ADCP, XCTD profiling, float and drifter deployments in the 
southeast Pacific was carried out August to October 2005. The RN Knorr 
departed Punta Arenas, Chile on 21 August 2005. A total of 135 
LADCP/CTD/rosette stations were occupied, 399 XCTDs were launched, 13 ARGO 
floats and 20 surface drifters were deployed from 23 August - 5 October. 
Water samples (up to 24), LADCP, and CTD data were collected on each cast in 
most cases to within 10 meters of the bottom. Salinity, dissolved oxygen and 
nutrient samples were analyzed for up to 24 water samples from each cast of 
the principal LADCP/CTD/rosette program. Water samples were also measured 
for C02 and CFCs, and underway surface pCO2, N20, temperature, conductivity, 
oxygen, and meteorological measurements were made. The cruise ended in 
Puerto Montt, Chile on 6 October 2005.







Introduction



Antarctic Intermediate Water (AAIW) is a low salinity water mass that fills 
most of the southern hemisphere and the tropical oceans at about 800 to 1000 
m depth. As the densest of the circumpolar Subantarctic Mode Waters (SAMW), 
AAIW is formed as a thick, outcropping mixed layer in the southeastern 
Pacific just north of the Subantarctic Front (SAF). SAMW and AAIW formation 
have a major impact on the oceanic sink for anthropogenic C02, whose largest 
uncertainty is at intermediate depths. The goal of Knorr cruise 182-07 was 
to characterize the wintertime AAIW formation processes. A followon summer 
hydrographic survey of the AAIW outcropping region and the fronts that bound 
it is scheduled for January to March 2006.



A sea-going science team gathered from three oceanographic institutions 
participated on the cruise. Several other science programs were supported 
with no dedicated cruise participant. The science team and their 
responsibilities are listed below.



Personnel


Scientific Personnel AAIW 2005  

Duties               Name               Affiliation    email
-------------------  -----------------  -------------  ------------------------
Chief Scientist      Teresa Chereskin   UCSD/SIO       tchereskin@ucsd.edu
ET/Deck/Salinity/02  Carl Mattson       UCSD/SIO/STS   cmattson@ucsd.edu
ET/Deck/Salinity/02  Scott Hiller       UCSD/SI O/STS  scott@odf.ucsd.edu
ET/DeckO2            Bob Green          UCSD/SI O/STS  bobg @odf.ucsd.edu
CTD/Data             Frank Delahoyde    UCSD/SI O/STS  fdelahoyde@ucsd.edu
Bottle Data          Kristin Sanborn    UCSD/SIO/STS   ksanborn@ucsd.edu
Nutrients/02/Deck    Dan Schuller       UCSD/SIO/STS   dan@odf.ucsd.edu
Nutrients/02/Deck    Teresa Kacena      UCSD/SI O/STS  teresa@odf.ucsd.edu
CTD/LADCP/XCTD       Sharon Escher      UCSD/SIO       sescher@ucsd.edu
C02                  Justine Afghan     UCSD/SIO       jafghan@ucsd.edu
C02                  Jeffrey Skacel     UCSD/SIO       jafghan@ucsd.edu
DIC                  George Anderson    UCSD/SIO       ganderson@ucsd.edu
DIC                  Brendan Carter     UCSD/SIO       brcarter@ucsd.edu
CTD/ADCP/XCTD        James Holte        UCSD/SIO       jholte@ucsd.edu
CTD/ADCP/XCTD        Yueng-Djern Lenn   UCSD/SIO       ylenn@ucsd.edu
CFC                  Jim Happell        RSMAS          jhappell@rsmas.miami.edu
CFC                  Kim Van Scoy       RSMAS          fleece@eritter.net
PCO2, N20            Mauricio Gallegos  U. Concepcion  mauricio@profc.udec.cl
PCO2, N20            Victor Villagran   U. Concepcion  victor@profc.udec.cl
CTD watchstander     Eduardo Navarro    U. Concepcion  eduardo@dgeo.udec.cl
SSSG Tech            Robert Laird       WHOI           sssg@knorr.whoi.edu
SSSG Tech            Sacha Wichers      WHOI           sssg@knorr.whoi.edu




PRINCIPAL PROGRAMS


Principal Programs of AAIW 2005

Analysis                 Institution   Principal          Investigator
-----------------------  ------------  -----------------  ---------------------
CTDO/S/02/Nutrients      UCSD/SIO      Lynne Talley       ltalley@ucsd.edu
                                       James H. Swift     jswift@ucsd.edu
Transmissometer          TAMU          Wilf Gardner       wgardner@tamu.edu
C02-Alkalinity           UCSD/SIO      Andrew Dickson     adickson@ucsd.edu
C02-DIC                  UCSD/SIO      Andrew Dickson     adickson@ucsd.edu
CFCs                     RSMAS-UMiami  Rana Fine          rline@rsmas.miami.edu
ADCP/LADCP               UCSD/SIO      Teresa Chereskin   tchereskin@ucsd.edu
ARGO Floats              UCSD/SIO      Dean Roemmich      droemmich@ucsd.edu
CTD/XCTD/satellite data  UC            Samuel Hormazabal  sam@profc.udec.cl
Underway pCO2            UC            Osvaldo Ulloa      oulloa@profc.udec.cl
                                       Samuel Hormazabal  lfarias@profc.udec.dk
pCO2 drifter             MBARI         Francisco Chavez   chfr@mbari.org




CRUISE NARRATIVE

The Knorr departed Punta Arenas, Chile on 21 August 2005 at 0900 local. 
Almost immediately we hove to for repairs on the starboard steering, which 
was not functioning when we left the dock. The repairs took about 8 hours 
during which time the various groups finished testing equipment and tying 
down gear. We also took the opportunity to deploy a shallow test cast in the 
Straits of Magellan. The other serious problem that was discovered in Punta 
Arenas was that the wire on both winches had only one good conductor. We 
used the starboard one for the duration of the survey. The condition of the 
wire influenced our choice of a 24-place over a 36-place rosette and dictated 
conservative wire speeds for this cruise. Another major factor influencing 
wire speed was wire tension, especially during times when we had several 
sets of large swells coming from multiple directions.

During AAIW, CTD stations at roughly 50 km spacing were supplemented by XCTD 
sampling every 15 to 20 km. Generally, three XCTDs were launched between CTD 
stations. Additionally, two intensive surveys were carried out in regions of 
deep mixed layers, steaming a diamond pattern centered on the main AAIW 
track, with dense XCTD sampling throughout and CTD stations at the corners. 
The first intensive survey began after CTD station 9, triggered by deep 
mixed layers (400 m) observed at stations 6 and 7. We turned back and 
steamed a diamond pattern centered on station 7, with CTD stations 8, 10, 6, 
and 11 located at the corners. The second intensive survey was triggered by 
crossing the Subantarctic Front (station 14). We turned back to survey the 
deep mixed layers north of the front. We again steamed a diamond pattern, 
centered on station 13, with CTD stations 14, 15, 12, and 16 at the corners. 
Surface drifters were deployed at the corners of the intensive surveys, 
accounting for 8 of our 20 deployments.

In total we made 6 crossings of the Subantarctic Front (SAF) and 2 crossings 
of the Polar Front (PF). Microwave SST images made for our region by Lynne 
Talley and downloaded from the internet were very helpful in tracking the 
fronts. Based on a 1980 cruise by McCartney, we anticipated that the first 
pair of SAF crossings along 77W and 79W, nearest to Drake Passage, would 
have the deepest mixed layers. In fact, we found equally deep mixed layers 
on the second pair of SAF crossings, located near/on the 89W meridian. The 
89W PF crossing was our furthest south, to 62S.

The third pair of SAF crossings was our furthest west, meant to measure the 
SAMW upstream condition. ARGO floats supplied by Dean Roemmich (SlO) were 
deployed at predetermined sites, with the first deployments made along this 
section of our track. It was also along this portion of our survey that we 
encountered our heaviest seas and winds. The steaming speed fell below 9kts 
and our wire speed was often limited to 20 m/s for the upper 1500 m. From 
station 62 to 69, our station spacing was increased from 50 km to 100 km in 
order to keep on schedule, and the density of XCTDs was increased to 
compensate. However, strong and gusty winds (above 50 kts) were often the 
cause of XCTD cast failure along the westward line from station 62 to 
station 70. Finally, at station 70, conditions were deemed unsafe for 
deployment and we hove to for 24 hours until winds and seas abated. Sea 
cable reterminations were required after several of these stations to remove 
wire kinks caused by snap loading of the wire by ship roll/heave.

Our weather improved from stations 77 to 92, but from 92 to 100 we again 
encountered high winds and swell as we continued east towards the Chilean 
coast. However, we maintained 50 km station spacing, closer at the coast. 
From the coast, we transited back to pick up our line northward along 89W. 
The 89W line repeats the 1980 McCartney cruise line, as does our final 
eastward segment to the coast along 45S. On our final eastward segment we 
replaced 3 stations with XCTD casts, because of time constraints.

Science operations halted at 1000 local on 5 October 2005 to begin the 26 
hour steam to Puerto Montt, which required making rendezvous with pilot 
boats at two locations for our final transit through Chilean coastal waters.

The science parties and the officers and crew of the Knorr are to be 
commended for their hard work and careful measurements. A CDROM of 
preliminary data obtained within the Chilean EEZ was produced and given to 
the Chilean observer/participating scientist, Eduardo Navarro.



DESCRIPTION OF MEASUREMENT TECHNIQUES


1. CTDLHYDROGRAPHIC MEASUREMENTS PROGRAM

The basic CTD/hydrographic measurements consisted of salinity, dissolved 
oxygen and nutrient measurements made from water samples taken on 
CTD/rosette casts, plus pressure, temperature, salinity, dissolved oxygen 
and transmissometer from CTD profiles. A total of 136 CTD/rosette casts were 
made usually to within 10 meters of the bottom. No major problems were 
encountered during the operation. The distribution of samples is illustrated 
in figure 1.0 - 1.8.


Figure 1.0: Sample distribution, stations 1-21
Figure 1.1: Sample distribution, stations 21-33.
Figure 1.2: Sample distribution, stations 33-44.
Figure 1.3: Sample distribution, stations 44-62.
Figure 1.4: Sample distribution, stations 62-70.
Figure 1.5: Sample distribution, stations 70-77.
Figure 1.6: Sample distribution, stations 77-109.
Figure 1.7: Sample distribution, stations 110-118.
Figure 1.8: Sample distribution, stations 118-136.


1.1. Water Sampling Package

LADCP/CTD/rosette casts were performed with a package consisting of a 
24-bottle rosette frame (ODF), a 24-place pylon (SBE32) and 24 10-liter 
Bullister bottles (ODF). Underwater electronic components consisted of a 
Sea-Bird Electronics (SBE) 9plus CTD (ODF #796) with dual pumps, dual 
temperature (SBE3pIus), dual conductivity (SBE4), dissolved oxygen (5BE43) 
and transmissometer (Wetlabs C-Star); an SBE35RT Digital Reversing 
Thermometer, an RDI LADCP (Broadband 150khz) and a Simrad altimeter.

The CTD was mounted vertically in an SBE CTD frame attached to the bottom 
center of the rosette frame. The SBE4 conductivity and SBE3p/us temperature 
sensors and their respective pumps were mounted vertically as recommended by 
SBE. Pump exhausts were attached to inside corners of the CTD cage and 
directed downward. The entire cage assembly was then mounted on the bottom 
ring of the rosette frame, offset from center to accommodate the pylon, and 
also secured to frame struts at the top. The SBE35RT temperature sensor was 
mounted vertically and equidistant between the T1 and T2 intakes. The 
transmissometer was mounted horizontally along the rosette frame adjacent to 
the CTD. The altimeter was mounted on the outside of the bottom frame ring. 
The LADCP was vertically mounted inside the bottle rings on the opposite 
side of the frame from the CTD. The locations of bottles 16, 17 and 18 were 
adjusted to accommodate the LADCP.

The rosette system was suspended from a UNOLS-standard three-conductor 0.322" 
electro-mechanical sea cable. The RN Knorr's starboard-side Markey winch was 
used for all casts. It was discovered at the beginning of the cruise that 
the two sea cables on board had only a single functional conductor each. Sea 
cable reterminations were made prior to casts 22/1, 70/1, 71/1, 73/1 and 
110/1. Cast 75/1 was aborted at 212m on the downcast due to sea conditions.

The deck watch prepared the rosette 10-20 minutes prior to each cast. The 
bottles were cocked and all valves, vents and lanyards were checked for 
proper orientation. Once stopped on station, the LADCP was turned on and the 
rosette moved into position under the starboard-side squirt boom using an 
airpowered cart and tracks. The CTD was powered-up and the data acquisition 
system in the main lab started when directed by the deck watch leader. Tag 
lines were threaded through the rosette frame, and syringes were removed 
from the CTD intake ports. The winch operator was directed by the deck watch 
leader to raise the package, the boom and rosette were extended outboard and 
the package quickly lowered into the water. The tag lines were removed and 
the package was lowered to 10 meters, by which time the sensor pumps had 
turned on. The winch operator was then directed to bring the package back to 
the surface (0 winch wireout) and to begin descent. The entry procedure was 
frequently modified as dictated by weather and sea conditions and for many 
casts no attempt was made to return close to the surface prior to descent.

Each rosette cast was usually lowered to within 10 meters of the bottom, 
using the altimeter to determine a safe distance.

On the up cast the winch operator was directed to stop at each bottle trip 
depth. The CTD console operator waited 30 seconds before tripping a bottle 
to insure the package wake had dissipated and the bottles were flushed, then 
an additional 10 seconds after receiving the trip confirmation to allow the 
SBE35RT temperature sensor time to make a measurement. The winch operator 
was then directed to proceed to the next bottle stop.

Sea conditions were sufficiently poor toward the end of several casts that 
no stops were made shallower than 200m. In these cases, the rosette was 
hauled at a constant rate (20m/min) and the remaining bottles closed 
"on-the-fly". These bottles have a quality code of "4" (did not trip 
correctly) associated with them and are well-documented.

Standard sampling depths were used throughout AAIW 2005 , depending on the 
overall water depth (table 1.1.0). These standard depths were staggered 
every other station.

Recovering the package at the end of the deployment was essentially the 
reverse of launching, with the additional use of poles and snap-hooks to 
attach tag lines, and air-tuggers on the tag lines for added safety and 
stability. The rosette was moved into the forward hangar for sampling. The 
bottles and rosette were examined before samples were taken, and anything 
unusual noted on the sample log.

Each bottle on the rosette had a unique serial number. This bottle 
identification was maintained independently of the bottle position on the 
rosette, which was used for sample identification. Six bottles were replaced 
on this cruise, and various parts of bottles were occasionally changed or 
repaired.

Routine CTD maintenance included soaking the conductivity and DO sensors in 
fresh water between casts to maintain sensor stability. Rosette maintenance 
was performed on a regular basis. 0-rings were changed as necessary and 
bottle maintenance was performed each day to insure proper closure and 
sealing. Valves were inspected for leaks and repaired or replaced as needed.


Table 1.1.0: AAIW 2005 water sampling guidelines.

(1)       top bottle within sight of the surface  
(2)       bottom bottle within 10 meters of bottom  
(3)       0-about 500 meters: spacing no greater than 50-60 meters  
(4)       500-800 meters: spacing no greater than 100 meters  
(5)       800-1200 meters: spacing no greater than 200 meters  
(6)       1200-2000 meters: spacing no greater than 350 meters  
(7)       2000-bottom: spacing no greater than 500 meters  
(8)       bottom of SAMW: resolve the property break with one  
          bottle above and one below if the layer is obvious (within  
          50 meters of break)  
(9)       AAIW if obvious salinity minimum (north of SAF): try to hit  
          the minimum  
(General) Stagger the sampling so that sample depths are not  
          exactly the same from one to the next.  



1.2.  Underwater Electronics Packages

CTD data were collected with a SBE9plus CTD (ODF #769). This instrument 
provided pressure, dual temperature (SBE3), dual conductivity (SBE4), 
dissolved oxygen (5BE43), transmissometer (Wetlabs SeaStar) and altimeter 
(Simrad 807 or 1007) channels. The CTD supplied a standard SBE-format data 
stream at a data rate of 24 frames/second (fps).


Table 1.2.0: AAIW 2005 Rosette   Underwater Electronics.

Sea-Bird 5BE32 24-place Carousel Water Sampler  S/N 3223219-0320
Sea-Bird SBE35RT Digital Reversing Thermometer  S/N 35-0034
Sea-Bird SBE9plus CTD                           S/N 09P39801-0796
Paroscientific Digiquartz Pressure Sensor       S/N 98627
Sea-Bird SBE3pIus Temperature Sensor            S/N 03P-4486 (Primary)
Sea-Bird SBE3pIus Temperature Sensor            S/N 03P-4476 (Secondary)
Sea-Bird SBE4C Conductivity Sensor              S/N 04-3023 (Primary)
Sea-Bird SBE4C Conductivity Sensor              S/N 04-3002 (Secondary, 1/1-72/1)      
Sea-Bird SBE4C Conductivity Sensor              S/N 04-2319 (Secondary, 73/1-136/1)
Sea-Bird 5BE43 DO Sensor                        S/N 43-0872 (1/1-32/1, 35/1) 
Sea-Bird 5BE43 DO Sensor                        S/N 43-0185 (33/1-34/1, 36/1, 37/1-136/1)
Wetlabs Sea-Star Transmissometer                S/N 327DR (owned by TAMU)
Roll and Pitch Sensor                           S/N SBE13-475
Accelerometer Sensor                            S/N SBE13-471
Simrad 807 Altimeter                            S/N 9711090
Simrad 1007 Altimeter                           S/N 20174
RDI Broadband 150khz LADCP                      S/N 1394
LADCP Battery Pack  


The CTD was outfitted with dual pumps. Primary temperature, conductivity and 
dissolved oxygen were plumbed on one pump circuit and secondary temperature 
and conductivity on the other. The sensors were deployed vertically. The 
primary temperature and conductivity sensors (T1 #03P-4486 and C1 #04-3023) 
were used for reported CTD temperatures and conductivities on all casts. The 
secondary temperature and conductivity sensors were used for calibration 
checks.

The SBE9plus CTD and SBE35RT temperature sensor were both connected to the 
5BE32 24-place pylon providing for single-conductor sea cable operation. The 
sea cable armor was used for ground (return). Power to the SBE9plus CTD (and 
sensors), 5BE32 pylon, SBE35RT and Simrad 807 altimeter was provided through 
the sea cable from the SBE1 1 plus deck unit in the main lab.


1.3.  Navigation and Bathymetry Data Acquisition

Navigation data were acquired at 1-second intervals from the ship's C-Nav GPS 
receiver by one of the Linux workstations beginning August 21. Data from the 
ship's Knudsen 320B/R Echosounder (12 KHz transducer) were also acquired and 
merged with the navigation. The Knudsen bathymetry data were noisy and 
subject to washing out when the seas were choppy or the ship's bow thruster 
engaged.

Bathymetric data from the ship's multibeam echosounder system (Seabeam 2000) 
were also logged and archived independently.


1.4.  CTD Data Acquisition and Rosette Operation

The CTD data acquisition system consisted of an SBE-1 1 plus (V2) deck unit 
and three networked generic PC workstations running Fedora Core Linux. Each 
PC workstation was configured with a color graphics display, keyboard, 
trackball and DVD+RW drives. One of the three systems also had 8 additional 
RS-232 ports via a Comtrol Rocketport PCI serial controller. The systems 
were connected through a 100BaseTX ethernet switch, which was also 
connected to the ship's network. These systems were available for realtime 
operational and CTD data displays, and provided for CTD and hydrographic 
data management and backup.

One of the workstations was designated the CTD console and was connected to 
the CTD deck unit via RS-232. The CTD console provided an interlace and 
operational displays for controlling and monitoring a CTD deployment and 
closing bottles on the rosette.

CTD deployments were initiated by the console watch after the ship had 
stopped on station. The watch maintained a console operations log containing 
a description of each deployment, a record of every attempt to close a 
bottle and any pertinent comments. The deployment and acquisition software 
presented a short dialog instructing the operator to turn on the deck unit, 
examine the on screen CTD data displays and to notify the deck watch that 
this was accomplished.

Once the deck watch had deployed the rosette, the winch operator would begin 
the descent. When permitted by sea conditions, the rosette was lowered to 10 
meters, raised back to the surface then lowered for the descent. This 
procedure was adopted to allow the immersion-activated sensor pumps time to 
start and flush the sensors.

Profiling rates were frequently dictated by sea conditions but never exceeded 
60m/minute.

The progress of the deployment and CTD data quality were monitored through 
interactive graphics and operational displays. Bottle trip locations were 
decided and transcribed on the console and sample logs. The sample log would 
later be used as an inventory of samples drawn from bottles.

The combination of altimeter distance, CTD depth, winch wire-out and 
echo-sounder depth provided reliable, precise control of package distance 
from the bottom and allowed routine approaches to within 10 meters.

Bottles were closed on the up cast by operating an on-screen control. The 
winch operator was given a target wire-out for the bottle stop, proceeded to 
that depth and stopped. Bottles were tripped at least 30 seconds after 
stopping to allow the rosette wake to dissipate and the bottles to flush. 
The winch operator was instructed to proceed to the next bottle stop at 
least 10 seconds after closing bottles to allow the SBE35RT calibration 
temperature sensor time to make a measurement.

After the last bottle was tripped, the console watch directed the deck watch 
to bring the rosette on deck. Once on deck, the console watch terminated the 
data acquisition, turned off the deck unit and assisted with rosette 
sampling.


1.5.  CTD Data Processing

The shipboard CTD data acquisition was the first stage in shipboard 
processing. The raw CTD data were converted to engineering units, 
filtered, response-corrected, calibrated and decimated to a more manageable 
0.5 second time-series. The laboratory calibrations for pressure, 
temperature and conductivity were applied at this time. The 0.5 second 
time-series data were used for real-time graphics during deployments, and 
were the source for CTD pressure and temperature associated with each 
rosette bottle. Both the raw 24hz data and the 0.5 second time-series were 
stored for subsequent processing steps.

At the completion of a deployment a series of processing steps were 
performed automatically. The 0.5 second time-series data were checked for 
consistency, clean sensor response and calibration shifts. A 2 decibar 
pressure-series was then generated from the up cast. The up cast data were 
selected because of missing near-surface down cast data in many of the 
deployments due to sea conditions. Both the 2 decibar pressure-series and 
0.5 second time-series data were then made available for downloading, 
plotting and reporting on the shipboard cruise website.

CTD data were routinely examined for sensor problems, calibration shifts 
and deployment or operational problems. The primary and secondary 
temperature sensors (SBE 3) were compared to each other and to the SBE35 
temperature sensor. CTD conductivity sensors (SBE 4) were compared and 
calibrated by examining differences between CTD and check-sample 
conductivity values. The CTD dissolved oxygen sensor data were calibrated 
to check-sample data. Additional deep TS and theta-02 comparisons were made 
between down and up casts as well as with adjacent deployments. Vertical 
sections were made of the various properties derived from sensor data and 
checked for consistency.

A total of 136 casts were made (including 1 aborted cast). The 24-place 
10-liter rosette and CTD #796 were used on all casts.


1.6.  CTD Sensor Laboratory Calibrations

Laboratory calibrations of the CTD pressure, temperature, conductivity, 
dissolved oxygen and the SBE35RT Digital Reversing Thermometer sensors 
were performed prior to AAIW 2005. The calibration dates are listed in 
table 1.6.0.


Table 1.6.0: AAIW 2005 CTD sensor laboratory calibrations.  

                                              Calibration  Calibration 
Sensor                              S/N       Date         Facility
----------------------------------  --------  -----------  -----------
Paroscientific Digiquartz Pressure  98627      7-July-05   SIOIODF
Sea-Bird SBE3pIus T1 Temperature    03P-4486   7-July-05   SIO/ODF
Sea-Bird SBE3pIus T2 Temperature    03P-4476   7-July-05   SIO/ODF
Sea-Bird SBE4C Cl Conductivity      04-3023   14-July-05   SBE
Sea-Bird SBE4C C2 Conductivity      04-3002   14-July-05   SBE
Sea-Bird SBE4C C2 Conductivity      04-2319   04-March-05  SBE
Sea-Bird 5BE43 Dissolved Oxygen     43-872       N/A       SBE
Sea-Bird 5BE43 Dissolved Oxygen     43-848       N/A       SBE
Sea-Bird SBE35RT Digital            35-0034   18-May-05    SIO/ODF
  Reversing Thermometer


1.7.  CTD Shipboard Calibration Procedures

CTD #796 was used for all AAIW 2005 casts. The CTD was deployed with all 
sensors and pumps aligned vertically, as recommended by SBE. The primary 
temperature and conductivity sensors (T1 & C1) were used for all reported 
CTD data on all casts. The secondary temperature and conductivity sensors 
(T2 & C2) served as calibration checks for the primary sensors. The 
SBE35RT Digital Reversing Thermometer (S/N 35-0034) served as an 
independent calibration check. In-situ salinity and dissolved O2 check 
samples collected during each cast were used to calibrate the conductivity 
and dissolved O2 sensors.

1.7.1.  CTD Pressure

The Paroscientific Digiquartz pressure transducer (S/N 98627) was 
calibrated in July 2005 at the SIO/STS Calibration Facility. Calibration 
coefficients derived from the calibration were applied to raw pressures 
during each cast. Residual pressure offsets (the difference between the 
first and last submerged pressures) were examined to check for calibration 
shifts. All were <0.5db, and the sensor exhibited <0.2 db offset shift 
over the period of use. No additional adjustments were made to the 
calculated pressures.

1.7.2.  CTD Temperature

A single primary temperature sensor (SBE 3, S/N 03P-4486) and secondary 
temperature sensor (SBE 3, S/N 03P-4476) served the entire cruise. 
Calibration coefficients derived from the pre-cruise calibrations were 
applied to raw primary and secondary temperatures during each cast.

Two independent metrics of calibration accuracy were examined. The primary 
and secondary temperatures were compared at each rosette trip, and the 
SBE35RT temperatures were compared to primary and secondary temperatures at 
each rosette trip.

Calibration accuracy was first examined by tabulating Ti-T2 over a range of 
pressures (bottle trip locations) for each cast. These comparisons are 
summarized in figure 1.7.2.0.


Figure 1.7.2.0: T1-T2 by station, p>2000db.


Although there appears to be a slight (<0.0003°C) drift between the sensors 
over the cruise, it is less than than the calibration accuracy. The 95% 
confidence limit for the mean differences is < 0.0008°C.

The SBE35RT Digital Reversing Thermometer is an internally-recording 
temperature sensor that operates independently of the CTD. It is triggered 
by the SBE32 pylon in response to a bottle trip. According to the 
Manufacturer's specifications the typical stability is 0.001°C/year. The 
differences between the SBE35RT and T1 (primary CTD temperature) are 
summarized in figure 1.7.2.1, and between the SBE35RT and T2 (secondary CTD 
temperature) in figure 1.7.2.2.


Figure 1.7.2.1: SBE35RT-T1 by station, p>2000db.
Figure 1.7.2.1: SBE35RT-T2 by station, p>2000db.


The SBE35RT used on AAIW 2005 was calibrated in May 2005 at which time it 
was reported to have < -0.0002°C correction over the entire operating range. 
The sensor was used for a 40-day cruise prior to AAIW 2005 during which it 
exhibited a -0.00082 °C offset relative to the CTD sensors. Evidently the 
SBE35RT began to drift significantly on 67/1. Examining casts 1/1-66/1, the 
mean differences are -0.0017988°C for SBE35RT-Ti and -0.0020826°C for 
SBE35RT-T2. Since T1 and T2 had been calibrated more recently (July 2005) 
than the SBE35RT, had not been used prior to AAIW 2005 since calibration and 
had a mean calibrated difference of -0.00023°C the SBE35RT differences were 
only used to check for calibration shifts. No additional corrections were 
applied to either T1 or T2 temperatures.

Post-cruise calibrations for all the temperature sensors are pending.

1.7.3. CTD Conductivity

A single primary conductivity sensor (SBE 4, S/N 04-3023) and two secondary 
conductivity sensors (SBE 4, S/N 04-3002 1/1-72/1, S/N 04-2319 73/1-136/1) 
served the entire cruise. Conductivity sensor calibration coefficients 
derived from the pre-cruise calibrations were applied to raw primary and 
secondary conductivities.

Comparisons between the primary and secondary sensors and between each of 
the sensors to check sample conductivities (conductivity calculated from 
bottle salinities) were used to derive conductivity corrections. None of the 
sensors showed any appreciable conductivity slope. The second C2 sensor used 
(04-2319) showed a slight (9.65e-8m5/cm/db) pressure slope. C1 was determined 
to have a slight drift amounting to a +0.0021 mS/cm offset change over the 
cruise. This drift correction was actually applied in 5 separate groupings 
as determined by secondary sensor and bottle conductivity differences. C2 
#3002 had a constant offset of -0.0012m5/cm relative to corrected C1. C2 
#2319 had a constant offset of -0.0001 2m5/cm for casts 73/1-109/1, and 
+0.00066m5/cm for 110/1-136/1 relative to corrected C1.

The comparison of the primary and secondary conductivity sensors by station, 
after applying shipboard corrections, is summarized in figure 1.7.3.0.


Figure 1.7.3.0: Cl and C2 conductivity differences by cast, p>2000db.


Salinity residuals after applying shipboard T1/C1 corrections are summarized 
in figures 1.7.3.1 through 1.7.3.3.

Figure 1.7.3.1: salinity residuals by pressure, all pressures.
Figure 1.7.3.2: salinity residuals by cast, all pressures.
Figure 1.7.3.3: salinity residuals by cast, p>2000db.


Figure 1.7.3.3 represents an estimate of the deep salinity accuracy on AAIW 
2005. The 95% confidence limit is ±0.0018 PSU relative to the bottle salts.

1.7.4.  CTD Dissolved Oxygen

Two SBE43 dissolved O2 (DO) sensor were used during this cruise: S/N 43-0872 
(1/1-32/1, 34/1-35/1) and 43-0848 (33/1, 36/1-136/1). The sensor was plumbed 
into the primary T1 /C1 pump circuit after C1. Sensor 0872 was replaced 
prior to 33/1 because of diminishing sensor response. Problems with the 
sensor cable on 33/1 rendered the DO data unusable, and 0872 was returned to 
service for two more casts.

The DO sensors were calibrated to dissolved O2 check samples at bottle stops 
by calculating CTD dissolved O2 then minimizing the residuals using a 
non-linear least-squares fitting procedure. The fitting procedure determined 
the calibration coefficients for the sensor model conversion equation, and 
was accomplished in stages. The time constants for the exponential terms in 
the model were first determined for each sensor. These time constants are 
sensor-specific but applicable to an entire cruise. Next, casts were fit 
individually to check sample data. The resulting calibration coefficients 
were then smoothed and held constant during a refit to determine sensor 
slope and offset.

Standard and blank values for bottle oxygen data were smoothed and the 
bottle oxygen recalculated prior to the final fitting of CTD oxygen.

The residuals are shown in figures 1.7.4.0-1.7.4.2.


Figure 1.7.4.0: O2 residuals by station number, all pressures.
Figure 1.7.4.1: O2 residuals by pressure, all pressures.
Figure 1.7.4.2: O2 residuals by station number, p>l000db.


The standard deviations of 0.97 uM/kg for all oxygens and 0.74 uM/kg for 
deep oxygens are only presented as general indicators of goodness of fit. 
ODF makes no claims regarding the precision or accuracy of CTD dissolved O2 
data.

The general form of the ODF O2 conversion equation for Clark cells follows 
Brown and Morrison [Brow78] and Millard [Mill82], [Owen85]. ODF models 
membrane and sensor temperatures with lagged CTD temperatures and a lagged 
thermal gradient. In-situ pressure and temperature are filtered to match 
the sensor response. Time-constants for the pressure response Tp, two 
temperature responses TTs and rTf, and thermal gradient response rdT are 
fitting parameters. The thermal gradient term is derived by lowpass 
filtering the difference between the fast response (Tf) and slow response 
(Ta) temperatures. This term is SBE43-specific and corrects a non-linearity 
introduced by analog thermal compensation in the sensor. The 0 gradient, 
dOc/dt, is approximated by low-pass filtering 1st-order 0, differences. This 
gradient term attempts to correct for reduction of species other than O2 at 
the sensor cathode. The timeconstant for this filter, Tag, is a fitting 
parameter. Dissolved O2 concentration is then calculated:


                                                    dOc
                                  (c3Pl+c4Tf+c5T+c6 --- +c7dT)
O2ml/l [C1 Oc + C2]• fsat(S,T,P) e                  dt             (1.7.4.0)


where:

O2ml/l      = Dissolved O2 concentration in ml/l;
Oc          = Sensor current (amps);
fsat(S,T,P )= O2 saturation concentration at S,T,P (ml/l);
S           = Salinity at O2 response-time (PSUs);
T           = Temperature at O2 response-time (°C);
P           = Pressure at O2 response-time (decibars);
Pl          = Low-pass filtered pressure (decibars);
Tf          = Fast low-pass filtered temperature (°C);
T           = Slow low-pass filtered temperature (°C);
dOc 
---         = Sensor current gradient (µamps/secs);
dt
dT          = low-pass filtered thermal gradient (T1 - TS).


1.8. Bottle Sampling

At the end of each rosette deployment water samples were drawn from the 
bottles in the following order:

  • CFCs
  • O2
  • Dissolved Inorganic Carbon (DIC)
  • Total Alkalinity
  • Nutrients
  • Salinity
  • Nitrous Oxide

The correspondence between individual sample containers and the rosette 
bottle position (1-24) from which the sample was drawn was recorded on the 
sample log for the cast. This log also included any comments or anomalous 
conditions noted about the rosette and bottles. One member of the sampling 
team was designated the sample cop, whose sole responsibility was to 
maintain this log and insure that sampling progressed in the proper drawing 
order.

Normal sampling practice included opening the drain valve and then the air 
vent on the bottle, indicating an air leak if water escaped. This 
observation together with other diagnostic comments (e.g., "lanyard caught 
in lid", "valve left open") that might later prove useful in determining 
sample integrity were routinely noted on the sample log. Drawing oxygen 
samples also involved taking the sample draw temperature from the bottle. 
The temperature was noted on the sample log and was sometimes useful in 
determining leaking or mis-tripped bottles.

Once individual samples had been drawn and properly prepared, they were 
distributed for analysis. Oxygen, nutrient and salinity analyses were 
performed on computer-assisted (PC) analytical equipment networked to the 
data processing computer for centralized data management.


1.9. Bottle Data Processing

Water samples collected and properties analyzed shipboard were managed 
centrally in a relational database (PostgreSoL-8.0.3) run on one of the 
Linux workstations. A web service (OpenAcs-5.1.5 and AOLServer-4.0.10) 
front-end provided ship-wide access to CTD and water sample data. Web-based 
facilities included on-demand arbitrary property-property plots and vertical 
sections as well as data uploads and downloads.

The Sample Log (and any diagnostic comments) was entered into the database 
once sampling was completed. Quality flags associated with sampled 
properties were set to indicate that the property had been sampled, and 
sample container identifications were noted where applicable (e.g., oxygen 
flask number). Each Sample Log was also scanned and made available as a 
JPEG file on the website.

Analytical results were provided on a regular basis by the various 
analytical groups and incorporated into the database. These results 
included a quality code associated with each measured value and followed 
the coding scheme developed for the World Ocean Circulation Experiment 
(WOCE) Hydrographic Programme (WHP) [Joyc94].

Sea conditions were sufficiently poor at the end of several deployments 
that no bottle stops were made shallower than 200m. In these cases, the 
rosette was hauled at a constant rate (20m/min) and the remaining bottles 
closed "on-the-fly". These bottles have a quality code of "4" (did not trip 
correctly) associated with them and are well-documented.

Various consistency checks and detailed examination of the data continued 
throughout the cruise. The individual sample comments are included in 
Appendix A.


1.10.  Salinity Analysis

Equipment and Techniques

Two Guildline Autosal Model 8400A salinometers (S/N 57-526 & S/N 53-503), 
located in the 01 lab, were used for all salinity measurements. The 
salinometers were modified by ODF to contain an interface for computer-aided 
measurement. The water bath temperatures were set and maintained at a value 
near the laboratory air temperature. They were set to 21°C for stations 1-92 
and 118-124 analyses, then switched to 24°C for stations 92-117 and 121-134.

The salinity analyses were performed after samples had equilibrated to 
laboratory temperature, usually within 8-26 hours after collection. The 
salinometers were standardized for each group of analyses (usually 1-2 
casts, up to -48 samples) using at least two fresh vials of standard 
seawater per group. Salinometer measurements were made by computer, where 
the analyst was prompted by software to change samples and flush.

Sampling and Data Processing

3114 salinity measurements were made and approximately 280 vials of standard 
water (SSW) were used.

Salinity samples were drawn into 200 ml Kimax high-alumina borosilicate 
bottles, which were rinsed three times with sample prior to filling. The 
bottles were sealed with custom-made plastic insert thimbles and Nalgene 
screw caps. This assembly provides very low container dissolution and 
sample evaporation. Prior to sample collection, inserts were inspected for 
proper fit and loose inserts replaced to insure an airtight seal. The draw 
time and equilibration time were logged for all casts. Laboratory 
temperatures were logged at the beginning and end of each run.

PSS-78 salinity [UNES81] was calculated for each sample from the measured 
conductivity ratios. The difference (if any) between the initial vial of 
standard water and the next one run as an unknown was applied as a linear 
function of elapsed run time to the data. The corrected salinity data were 
then incorporated into the cruise database. Temperature control was 
somewhat problematic and a few runs were rendered unusable for calibration 
purposes because of a lack of temperature stability. The estimated accuracy 
of bottle salinities run at sea is usually better than ±0.002 PSU relative 
to the particular standard seawater batch used. The 95% confidence limit 
for residual differences between the bottle salinities and calibrated CTD 
salinity relative to SSW batch P-145 was ±0.0037 PSU for all salinities, and 
±0.0028 PSU for salinities deeper than 1000db.

Laboratory Temperature

The temperature in the salinometer laboratory varied from 17.0 to 24.0°C, 
during the cruise. The air temperature during any particular run varied from 
-7 to +4.5°C.

Standards

IAPSO Standard Seawater (SSW) Batch P-145 was used to standardize for 
stations 1-122 salinity measurements and IAP5O Standard Seawater Batch P144 
was used to standardize for stations 123-134.


1.11.  Oxygen Analysis

Equipment and Techniques

Dissolved oxygen analyses were performed with an ODF-designed automated 
oxygen titrator using photometric end-point detection based on the 
absorption of 365nm wavelength ultra-violet light. The titration of the 
samples and the data logging were controlled by PC software. Thiosulfate was 
dispensed by a Dosimat 665 buret driver fitted with a 1.0 ml buret. ODE used 
a whole-bottle modified-Winkler titration following the technique of 
Carpenter [Carp65] with modifications by Culberson et al. [Culb9l], but with 
higher concentrations of potassium iodate standard (~0.012N) and thiosulfate 
solution (~55 gm/I). Pre-made liquid potassium iodate standards were run once 
a day approximately every 4 stations, unless changes were made to system or 
reagents. Reagent/distilled water blanks were determined every day or more 
often if a change in reagents required it to account for presence of 
oxidizing or reducing agents. The auto-titrator performed well.


Sampling and Data Processing

3126 oxygen measurements were made. Samples were collected for dissolved 
oxygen analyses soon after the rosette was brought on board. Using a Tygon 
and silicone drawing tube, nominal 125m1 volume-calibrated iodine flasks were 
rinsed 3 times with minimal agitation, then filled and allowed to overflow 
for at least 3 flask volumes. The sample drawing temperatures were measured 
with a small platinum resistance thermometer embedded in the drawing tube. 
These temperatures were used to calculate uM/kg concentrations, and as a 
diagnostic check of bottle integrity. Reagents were added to fix the oxygen 
before stoppering. The flasks were shaken twice (10-12 inversions) to assure 
thorough dispersion of the precipitate, once immediately after drawing, and 
then again after about 20 minutes.

The samples were analyzed within 1-12 hours of collection, and the data 
incorporated into the cruise database.

Thiosulfate normalities were calculated from each standardization and 
corrected to 20°C. The 20°C normalities and the blanks were plotted versus 
time and were reviewed for possible problems.

The blanks and thiosulfate normalities for each batch of thiosulfate were 
smoothed (linear fits) in three groups during the cruise and the oxygen 
values recalculated.

A noisy endpoint was occasionally acquired during the analyses, usually due 
to small waterbath contaminations. These endpoints were checked and 
recalculated using STS/ODF designed software.

Volumetric Calibration

Oxygen flask volumes were determined gravimetrically with degassed deionized 
water to determine flask volumes at STS/ODF's chemistry laboratory. This is 
done once before using flasks for the first time and periodically thereafter 
when a suspect volume is detected. The volumetric flasks used in preparing 
standards were volume-calibrated by the same method, as was the 10 ml 
Dosimat buret used to dispense standard iodate solution.

Standards

Liquid potassium iodate standards were prepared in 6 liter batches and 
bottled in sterile glass bottles at STS/ODF's chemistry laboratory prior to 
the expedition. The normality of the liquid standard was determined at ODF 
by calculation from weight. Two standard batches were used during AAIW 2005

Potassium iodate was obtained from Acros Chemical Co. and was reported by 
the supplier to be 98% pure. The second standard was supplied by Alfa Aesar 
and has a reported purity of 99.4-100.4%. Tests at ODF indicate no 
difference between these 2 batches. All other reagents were "reagent grade" 
and were tested for levels of oxidizing and reducing impurities prior to 
use.


1.12.  Nutrient Analysis

Equipment and Techniques

Nutrient analyses (phosphate, silicate, nitrate and nitrite) were performed 
on an ODF-modified 4-channel Technicon AutoAnalyzer II, generally within one 
to two hour after sample collection. Occasionally samples were refrigerated 
up to 4 hours at ~4°C. All samples were brought to room temperature prior to 
analysis.

The methods used are described by Gordon et al. [Gord92]. The analog outputs 
from each of the four colorimeter channels were digitized and logged 
automatically by computer (PC) at 2-second intervals.

Silicate was analyzed using the technique of Armstrong et al. [Arms67]. An 
acidic solution of ammonium molybdate was added to a seawater sample to 
produce silicomolybdic acid which was then reduced to silicomolybdous acid 
(a blue compound) following the addition of stannous chloride. Tartaric acid 
was also added to impede P04 color development. The sample was passed 
through a 15mm flowcell and the absorbence measured at 660nm.

A modification of the Armstrong et al. [Arms67] procedure was used for the 
analysis of nitrate and nitrite. For the nitrate analysis, the seawater 
sample was passed through a cadmium reduction column where nitrate was 
quantitatively reduced to nitrite. Sulfanilamide was introduced to the 
sample stream followed by N-(1-naphthyl)ethylenediamine dihydrochloride 
which coupled to form a red azo dye. The stream was then passed through a 
15mm flowcell and the absorbence measured at 540nm. The same technique was 
employed for nitrite analysis, except the cadmium column was bypassed, and a 
50mm flowcell was used for measurement.

Phosphate was analyzed using a modification of the Bernhardt and Wilhelms 
[Bern67] technique. An acidic solution of ammonium molybdate was added to 
the sample to produce phosphomolybdic acid, then reduced to phosphomolybdous 
acid (a blue compound) following the addition of dihydrazine sulfate. The 
reaction product was heated to ~55°C to enhance color development, then 
passed through a 50mm flowcell and the absorbence measured at 820nm.

Explicit corrections for carryover in nutrient analyses are not made. In a 
typical AutoAnalyzer system, sample to sample carryover is 1-2% of the 
concentration difference between samples. This effect is minimized by 
running samples in order of increasing depth such that concentration 
differences between samples are minimized. The initial surface samples could 
be run twice or a low nutrient sea water sample run ahead of the surface 
sample since these samples generally follow standard peaks.

Sampling and Data Processing

3126 nutrient samples were analyzed.

Nutrient samples were drawn into 45 ml polypropylene, screw-capped "oak-ridge 
type" centrifuge tubes. The tubes were cleaned with 10% HCI and rinsed with 
sample 2-3 times before filling. Standardizations were performed at the 
beginning and end of each group of analyses (typically one cast, up to 36 
samples) with an intermediate concentration mixed nutrient standard prepared 
prior to each run from a secondary standard in a low-nutrient seawater 
matrix. The secondary standards were prepared aboard ship by dilution from 
primary standard solutions. Dry standards were pre-weighed at the laboratory 
at ODF, and transported to the vessel for dilution to the primary standard. 
Sets of 7 different standard concentrations were analyzed periodically to 
determine any deviation from linearity as a function of absorbence for each 
nutrient analysis. A correction for non-linearity was applied to the final 
nutrient concentrations when necessary. A correction for the difference in 
refractive indices of pure distilled water and seawater was periodically 
determined and applied where necessary. In addition, a "deep seawater" high 
nutrient concentration check sample was run with each station as an 
additional check on data quality. The pump tubing was changed 3 times.

After each group of samples was analyzed, the raw data file was processed to 
produce another file of response factors, baseline values, and absorbences. 
Computer-produced absorbence readings were checked for accuracy against 
values taken from a strip chart recording. The data were then added to the 
cruise database.

Nutrients, reported in micromoles per kilogram, were converted from 
micromoles per liter by dividing by sample density calculated at 1 atm 
pressure (0 db), in situ salinity, and a per-analysis measured laboratory 
temperature.

Some stations showed small yet significant concentrations of N02 deeper than 
expected (i.e. ~0.01 uM below the thermocline). These stations were carefully 
reviewed and included in the final data report. It should be noted, however, 
that 0.01 uM is at the detection limit of the autoanalyzer system.

Standards

Primary standards for silicate (Na2S1F6) and nitrite (NaNO2) were obtained 
from Johnson Matthey Chemical Co.; the supplier reported purities of >98% and 
97%, respectively. Primary standards for nitrate (KN03) and phosphate 
(KH2P04) were obtained from Fisher Chemical Co.; the supplier reported 
purities of 99.999% and 99.999%, respectively. The efficiency of the cadmium 
column used for nitrate was monitored throughout the cruise and ranged from 
99-100%.

No major problems were encountered with the measurements. The temperature of 
the laboratory used for the analyses ranged from 21.6°C to 25.8°C, but was 
relatively constant during any one station (±1.5°C).




References

Arms67.
Armstrong, F. A. J., Stearns, C. R., and Strickland, J. D. H., "The 
measurement of upwelling and subsequent biological processes by means of the 
Technicon Autoanalyzer and associated equipment," Deep-Sea Research, 14, pp. 
381-389 (1967).

Bern67.
Bernhardt, H. and Wilhelms, A., "The continuous determination of low level 
iron, soluble phosphate and total phosphate with the AutoAnalyzer," 
Technicon Symposia, I, pp. 385-389 (1967).

Brow78.
Brown, N. L. and Morrison, G. K., "WHOI/Brown conductivity, temperature and 
depth microprofiler," Technical Report No. 78-23, Woods Hole Oceanographic 
Institution (1978).

Carp65.
Carpenter, J. H., "The Chesapeake Bay Institute technique for the Winkler 
dissolved oxygen method," Limnology and Oceanography, 10, pp. 141-143 
(1965).

Culb9l.
Culberson, C. H., Knapp, G., Stalcup, M., Williams, R. T., and Zemlyak, F., 
"A comparison of methods for the determination of dissolved oxygen in 
seawater," Report WHPO 91-2, WOCE Hydrographic Programme Office (Aug 1991).

Gord92.
Gordon, L. I., Jennings, J. C., Jr., Ross, A. A., and Krest, J. M., "A 
suggested Protocol for Continuous Flow Automated Analysis of Seawater 
Nutrients in the WOCE Hydrographic Program and the Joint Global Ocean Fluxes 
Study," Grp. Tech Rpt 92-1, OSU College of Oceanography Descr. Chem Oc. 
(1992).

Joyc94.
Joyce, T., ed. and Corry, C., ed., "Requirements for WOCE Hydrographic 
Programme Data Reporting," Report WHPO 90-1, WOCE Report No. 67/91, pp. 
52-55, WOCE Hydrographic Programme Office, Woods Hole, MA, USA (May 1994, 
Rev. 2). UNPUBLISHED MANUSCRIPT.

Mill82.
Millard, R. C., Jr., "CTD calibration and data processing techniques at WHOI 
using the practical salinity scale," Proc. Int. STD Conference and Workshop, 
p. 19, Mar. Tech. Soc., La Jolla, Ca. (1982).

Owen85.
Owens, W. B. and Millard, R. C., Jr., "A new algorithm for CTD oxygen 
calibration," Journ. of Am. Meteorological Soc., 15, p. 621 (1985).

UNES81.
UNESCO, "Background papers and supporting data on the Practical Salinity 
Scale, 1978," UNESCO Technical Papers in Marine Science, No. 37, p. 144 
(1981).




                   APPENDIX A: INDIVIDUAL SAMPLE COMMENTS


APPENDIX A

This appendix contains remarks for samples and bottles having a quality code 
of other than "2" (no problem noted).


Individual Sample Comments


CFC-ll, CFC-12, CFC-113, and CCl4
Analysts: Jim Happell and Kim VanScoy


Sample Collection

All samples were collected from depth using 10 liter Niskin bottles. None of 
the Niskin bottles used showed a CFC contamination throughout the cruise. All 
bottles in use remained inside the CTD hanger between casts. All spare 
bottles were stored on a spare rosette under a tarp, sitting on the Maindeck.

CFC sampling was conducted first at each station, according to WOCE protocol. 
This avoids contamination by air introduced at the top of the Niskin bottle 
as water was being removed. A water sample was collected directly from the 
Niskin bottle petcock using a 100 ml ground glass syringe which was fitted 
with a three-way stopcock that allowed flushing without removing the syringe 
from the petcock. Syringes were flushed several times and great care was 
taken to avoid contamination by air bubbles. One duplicate sample was taken 
on each station from random niskin bottles. Air samples, pumped into the 
system using an Air Cadet pump, were run about every 2 - 4 days from a 
Dekoron air intake hose mounted high on the foremast.


Equipment and technique

Chlorofluorocarbons CFC-11, CFC-12, and CFC-113, and CCL4 were measured on 134 
stations for a total of 3,199 samples. Halocarbon analyses were performed on 
a gas chromatograph (GC) equipped with an electron capture detector (ECD). 
Samples were introduced into the GC-EDC via a purge and dual trap system. The 
samples were purged with nitrogen and the compounds of interest were trapped 
on a main Porapack N trap held at ~-15°C with a Vortec Tube cooler. After the 
sample had been purged and trapped for several minutes at high flow, the gas 
stream was stripped of any water vapor via a magnesium perchlorate trap prior 
to transfer to the main trap. The main trap was isolated and heated by direct 
resistance to 140°C. The desorbed contents of the main trap were back-flushed 
and transferred, with helium gas, over a short period of time, to a small 
volume focus trap in order to improve chromatographic peak shape. The focus 
trap was also Porapak N and is held at ~-15°C with a Vortec Tube cooler. The 
focus trap was flash heated by direct resistance to 155 °C to release the 
compounds of interest onto the analytical pre-column. The pre-column was the 
first 5 meters of a 60 m Gaspro capillary column with the main column 
consisting of the remaining 55 meters. The analytical precolumn was held 
in-line with the main analytical column for the first 1.5 minutes of the 
chromatographic run. After 1.5 minutes, all of the compounds of interest were 
on the main column and the pre-column was switched out of line and 
back-flushed with a relatively high flow of nitrogen gas. This prevented 
later eluting compounds from building up on the analytical column, eventually 
eluting and causing the detector baseline signal to increase.

The syringes were stored in a flow-through seawater bath and analyzed within 
8-12 hours after collection. Bath temperature was recorded continuously for 
use in calculating the mass of water analyzed. Every 12 to 13 measurements 
were followed by a purge blank and a standard, gas7.175m1. The surface 
sample was held after measurement and was sent through the process in order 
to "restrip" it to determine the efficiency of the purging process.


Calibration

A gas phase standard, S39, was used for calibration. The concentrations of 
the CFCs and Cd4 in this standard are reported on the SlO 1998 absolute 
calibration scale. A calibration curve was run every 3-5 days. Estimated 
accuracy is ±2%. Precision for CFC-12, CFC-1 1, CFC-1 13 and Cd4 was less 
than 1%. Estimated limit of detection is 0.010 pM/kg for CFC-12 and CFC-1 
13, and 0.005 pMlkg for CFC-1 1 and Cd4.


Technical Problems

In large part, sample collection and measurement were very successful. The 
integration of the computer software with the GC-EDC system hardware made the 
procedure almost completely automated. Two stations, 38 and 39, were skipped 
to try and determine why instrument sensitivity dropped when changing the 
MgC1O4 drying trap. During the first part of the cruise several changes of 
the drying trap resulted in sensitivity changes, while several did not. We 
were not able to determine the cause of the sensitivity drop at this time. 
During a drying trap change on station 99, the flow controller for the ECD 
makeup gas flow was mistakenly turned down when it was meant to turn down 
the purge gas flow. Lowering the makeup gas flow increased the sensitivity 
and it was therefore considered likely that the earlier sensitivity loss 
occurred when the makeup flow controller knob was bumped when turning down 
the purge gas flow during drying trap changes. The makeup flow was set to 60 
ml/min and care was taken not to bump the controller throughout the 
remainder of the cruise. The sensitivity changes did not seem to affect the 
limit of detection, because instrument blanks and noise decreased when the 
sensitivity decreased.



AAIW 2005: ANALYST'S REPORT FOR TOTAL ALKALINITY AND DISSOLVED 


Inorganic Carbon Measurements.

From August 2lst to October 6th 2005, the RIV Knorr sailed on a NSF-funded 
oceanographic cruise off the coast of Southern Chile to collect physical and 
chemical data. Stations were occupied between ~45°S and ~62°S and ~75°W and 
~105°W (see map in pdf version for station locations): the domain of the Polar 
and Subantarctic Fronts and the region where Antarctic Intermediate Water (AAIW) 
formation is suspected to occur. This report details the corrections that 
were made to the dissolved inorganic carbon (CT) and total alkalinity (AT) 
data and an assessment of these measurements' uncertainties. The symbol 
"∆∆∆" preceeds any paragraph that details an adjustment made to the data.

Prepared by Brendan Carter: brcarter@ucsd.edu



CT: ESTIMATED UNCERTAINTY: 1.7 µmol kg-1
Analzyed by Justine Afghan Brendan Carter and Jeffrey SkaceL


Sample Density and Measurement Technique

Dissolved inorganic carbon (CT) analyses were made by coulometric titration 
with a SOMMA measurement apparatus and a UIC coulometer as described in SOP 
2 of the Department of Energy's Carbon Dioxide System in Sea Water Version 2 
(DOE handbook).1 Due to the length of time required to analyze a sample 
(~25min) many stations were not measured for CT at every depth for which 
water was collected. The pattern of every other station being sampled only 
for the shallowest half of the depths was adopted, though not rigorously 
adhered to when timing permitted otherwise. A total of 2,935 CT analyses 
were made including:

1. 189 measurements of Certified Reference Materials (CRM Batch 71, CT 
   2032.85 ± 0.40 µmol kg-1).

2. 196 sets of duplicate measurements made on separate seawater samples 
   collected from the same Niskin.

3. 86 measurements made on four separate batches of a reference material 
   that were prepared at sea.

4. 25 rejected measurements of samples with known errors.

5. 2439 measurements of sample CT that are presented with quality control 
   codes.



Data Quality Control and Applied Corrections


CRMs:

Measurements on bottles of CRMs were used to calibrate the SOMMA system and 
the coulometer. A calibration factor (calfactor), expressed as the number of 
counts registered by the coulometer per µmol of carbon titrated, was 
obtained from each of the CRM measurements. These calfactors were used for 
all samples measured until another CRM was analyzed and a new calfactor was 
obtained. Measurements of CRMs were made at the beginning, middle, and end 
of a cell's life.

A control chart for the calfactor values was prepared (Figure 1) as 
described in SOP 22 of the DOE handbook. The mean and standard deviation 
for all data are 4806.1 ± 4.9 counts/µmol of carbon. A number of changes 
were made to the system circa the 3rd of September including switching 
coulometers, and separate calfactor control charts were prepared for both 
the pre and post September 3rd CRM measurements (Figures 2 and 3). These 
subsets of the data can be shown to be statistically different from one 
another with greater than 99.9% confidence. For the pre-September 3rd data, 
the calfactor average and standard deviation are 4809.4 ± 3.5 counts/µmol. 
For the post-September 3rd data, after excluding one known outlier, the 
average and standard deviation are 4804.6 ± 3.9 counts/µmol.


Duplicates:

While the CRM data provides a measure of the long-term reproducibility of the 
system, shorterterm reproducibility (sample to sample within a given cell) 
was estimated from measurements of duplicate samples (i.e. pairs of bottles 
filled with water from the same Niskin from a rosette cast). Duplicate 
measurements were made at the beginning, the midpoint, and end of a cast, and 
two duplicates with water from the same Niskin were never run consecutively.

A control chart was prepared using the duplicate data as described in SOP 22 
of the DOE handbook (Fiure 4). The average non-absolute difference (2nd - 1St 
in the order of analysis) was 0.38 µmol kg and the standard deviation was 1.2 
µmol kg-1. The average value is statistically distinguishable from zero, where 
zero is the ideal result that would indicate all sources of error were random 
or independent of both time and the amount of carbon titrated within a given 
cell. It is thought that this error is related to the chemical evolution of 
the cell solutions with repeated titrations. However, since the exact nature 
of these cells' chemistry is not well-understood, no correction was made to 
account for this observation.


Uncertified Reference Materials:

The uncertified reference materials collected at sea were the only inter-cell 
checks on reproducibility that were not used for calibration. The first 
measurement of such a reference material occurred on the 13th of September, 
and the practice was continued the rest of the cruise. Most often one 
uncertified reference material would be measured at the beginning of a cell 
and another at the end. The standard deviation for these reference material 
measurements is 1.6 µmol a precision lower than the duplicate data. 
Subsequent tests have shown that the plastic bottles used to hold the 
reference materials may not form a proper seal, and this could easily account 
for some of the increase in variability.


Samples:

With a gap in the data below 2000 meters at every other station and no 
samples collected from station 93 (which shared a location with station 58), 
there is no meaningful comparison of samples from a quality control 
perspective.


Mercuric Chloride Dilution Correction:

∆∆∆ A ~120 µL volume of 50% saturated mercuric chloride was added to the 
~285 mL samples prior to measurement. An upward adjustment of the final 
reported sample CT values by a factor of 1.0004 was made to account for this 
dilution.


Measurement Uncertainty

The uncertainty for this measurement is a combination of the reproducibility 
of the system (± 1.6 µmol kg-1, estimated from the deviation of post September 
3rd calfactor data converted to µmol kg-1 using the certificate CT for batch 
71 CRMs) and the uncertainty in the certificate value to which the system was 
calibrated (± 0.40 µmol kg-1). Adding these in quadrature yields a total 
uncertainty of ± 1.7 µmol kg-1.



AT: ESTIMATED UNCERTAINTY: 1.3 µmo1 kg-1
Analyzed by George Anderson Justine Afghan and Brendan Carter.


Sample Density and Measurement Technique

Total Alkalinity (AT) analyses were made using a two-stage, potentiometric, 
open-cell titration by coulometrically analyzed hydrochloric acid. The 
equivalence point was evaluated from titration points in the pH region 3.0 
to 3.5 using a non-linear least squares procedure that corrects for 
reactions with sulfate and fluoride ions. A limited number of the Niskin 
samples that were not analyzed for CT due to time constraints were similarly 
not analyzed for AT to allow time for system diagnostics. A total of 3,837 
AT analyses were made including:


1. 396 measurements of Certified Reference Materials (CRM Batch 71, 
   alkalinity 2254.50 ± 0.71 µmol kg-1).

2. 251 sets of duplicate measurements made on separate seawater samples 
   collected from the same Niskin.

3. 172 repeat measurements made to verify values, to check outliers, as 
   system diagnostics, or when the system either was operated incorrectly 
   or behaved abnormally.

4. 3018 values of samples AT that are presented with quality control codes.


Data Quality Control and Applied Corrections


CRMs:

Measurements of bottles of CRMs were used to assess the system's 
reproducibility and accuracy at the beginning and end of each cast.

A control chart for the CRM AT values (Figure 5) was prepared as described 
in SOP 22 of the DOE handbook. The mean and standard deviation for all CRM 
analyses are 2056.3 ± 3.0 µmol kg-1.

The control chart allows us to see that the series does not meet the control 
criteria early in the cruise. Using the of August as a cutoff, it is 
statistically more than 99.9% likely that the data plotted before and after 
this date are distinct series. The two most likely causes for the change 
over time of the CRM results are a change in a calibrated syringe's volume 
(used for extracting the portion of seawater sample to be titrated) or a 
gradual change in the concentration of the acid used during titration. 
Gravimetric analyses of volumes of seawater collected from the syringe 
before and after this cutoff suggest that the syringe volume was constant.

∆∆∆ To determine whether a shift in the acid concentration could have 
produced the observed effect, a model was created to estimate how the 
analytical result of an AT measurement of a CRM would vary when measured 
using acid of changing concentration. The model assumed that the acid 
reservoir, initially filled with HC1 of one concentration, was repeatedly 
refilled by 200 mL portions of acid of another concentration every time the 
reservoir's acid volume decreased to 850 mLs. The actual dates of refilling 
were recorded by the operators, and these records were used to time the 
dilutions in the model. The modeled value expected for a measurement of a CRM 
using this acid of evolving concentration changed in a stepwise fashion that 
tracked an exponential (each step coinciding with a dilution) from 2254.25 
µmol kg-1 (an initial best guess) towards 2257.25 µmol kg-1 (a best guess for 
the concentration given infinite dilutions). A chart was prepared that 
displays this evolution of the modeled CRM value as the acid concentration 
changed with successive dilutions (Figure 6). Since this seemed the more 
plausible cause of this change, and since the curve fit the observed changes 
reasonably well, the final values for the CRMs and samples have been 
recalculated to the values that would be obtained by the titration if the 
acid were of constant (initial best-guess) concentration. This correction was 
accomplished by multiplying the measured AT by the ratio of the certificate 
CRM AT over the modeled average CRM AT expected for that time period. The CRM 
and sample data were then multiplied by a constant factor of 1.000118 to 
align the mean corrected value with the certified value.

The majority of the outliers on the control chart had lower AT than the mean 
for all CRMs measured at sea. Subsequent laboratory experimentation has 
reproduced this trend and shown that the degree of deviation from the 
certified AT is correlated with the amount of time that has elapsed since the 
CRM was sampled for CT or AT. This problem was unique to the CRMs, as only 
they were analyzed for more than one quantity: CT and AT. Data points that 
were subject to this deviation were labeled "suspect" and were excluded from 
any further calculation of standard deviation. This was done because the 
lower deviation obtained by excluding them would more accurately reflect the 
deviation of the collected station samples, as these samples were measured 
for only one quantity. The criterion for being labeled as "suspect" was that 
there had to be two consistent measurements of the same CRM bottle, the 
average AT for which was more than two standard deviations below the 
inclusive mean. Prior to eliminating these data points the skew for the 
dataset was -3.4 µmol kg-1. Afterwards it was -0.57 µmol kg-1. The standard 
deviation excluding suspect points, but without making the correction for the 
acid evolution, was 1.3 µmol kg-1.

A final control chart was created following correction for the acid, 
exclusion of the suspect datapoints, and alignment of the mean corrected CRM 
value with the certified value (Figure 7).  The final average and standard 
deviation for the measurement of the CRM's alkalinities are 2254.5 ± 1.1 
µmolkg-1.


Duplicates:

An estimate of the short term reproducibility of the alkalinity measurements 
made on this cruise was obtained from the duplicates. There were two pairs of 
duplicates measured per station: one drawn from the shallowest and one drawn 
from the deepest Niskin. One of each of the sets of duplicates was measured 
as the first station sample of a cast and the companion duplicates were 
measured midway through and as the last. From these measurements a duplicate 
control chart was prepared (Figure 8) as described in SOP 22 of the DOE 
handbook. The standard deviation of duplicate measurements was 0.84 µmol kg-1, 
and the average non-absolute difference between all pairs of duplicate 
measurements (2nd minus 1st) was statistically indistinguishable from zero.


Samples:

Two sets of stations were near enough to be useful for comparison: stations 
and 15 and 27, and 9 and 28. Additionally, station 58 was reoccupied later in 
the cruise as station 93. Plots were prepared of the measured AT vs. the 
potential density (relative to 4000 decibars) for each of these station pairs 
(Figures 9, 10, and 11). The plots agree with one another to the degree 
expected by measurements made several days or weeks apart, and, as expected, 
they agree most closely in the deep water where the effects of being near, 
and possibly straddling, the Southern Ocean's fronts would be minimal. Since 
sampling below 1000 meters was done with fairly large vertical spacing and 
there was vertical heterogeneity at these depths, no attempt was made to 
interpolate between points to allow for direct comparison of values.


Mercuric Chloride Dilution Correction:

∆∆∆ A 57 µL volume of 100% saturated mercuric chloride was added to the ~285 
mL samples prior to measurement. An upward adjustment of the final reported 
sample AT values by a factor of 1.0002 was made to account for this dilution.


Uncertainty Estimate

The uncertainty of these AT values is due to a combination of the 
uncertainties in the short term reproducibility of the system (± 1.1 µmol kg-1 
from CRM analyses) and in the certificate CRM concentration with which the 
technique was calibrated (± 0.71 µmol kg-1). Adding these two errors in 
quadrature yields an estimated total uncertainty of ± 1.3 µmol kg-1. More 
discussion of the uncertainty of the system employed can be found in Dickson 
et al., 2003.2

The variance of the CRM measurements was not found to be significantly 
affected by errors in the best-guess initial acid concentration used during 
the acid-reservoir concentration adjustment of the data. However, if an 
application for this dataset considers primarily data collected early in the 
cruise (before August 30th) then adopting a greater uncertainty may be 
appropriate given the larger relative affect errors in the best-initial-guess 
concentration would have for these points.



References.

1) DOE (1994) Handbook of methods for the analysis of the various parameters 
   of the carbon dioxide system in sea water; version 2, A. G. Dickson & C. 
   Goyet, eds., ORNL/CDIAC-74

2) Dickson, A. G., Afghan, J.D., Anderson, G.C., 2003. Reference Materials 
   for Oceanic CO2 analysis: a Method for the Certification of Total 
   Alkalinity. Marine Chemistry 80, 185-197



                     AAIW: PRELIMINARY XCTD CRUISE REPORT



       ANTARCTIC INTERMEDIATE WATER FORMATION IN THE SOUTHEAST PACIFIC

                         21 August to 6 October 2005

                           by Teresa K. Chereskin



1  AAIW XCTDs

During the Antarctic Intermediate Water (AAIW) winter cruise, dense underway 
profiling of upper ocean temperature and salinity was carried out with 
expendable conductivity-temperature-depth probes (XCTDs). The sampling at 15 
to 20 km spacing supplemented the full-depth CTD stations that were spaced 
at approximately 50 kin. Generally, three XCTDs were launched between CTD 
stations. Additionally, two intensive surveys were carried out in regions of 
deep mixed layers, steaming a diamond pattern centered on the main AAIW 
track, with dense XCTD sampling throughout and CTD stations at the corners. 
The first intensive survey began after CTD station 9 and ended with station 
12 (Stations 8, 10, 6, and 11 are the corners). The second intensive survey 
began after CTD station 14 and ended with station 17 (Stations 14, 15, 12, 
and 16 are the corners). Surface drifters were deployed at the corners of 
the two diamond patterned intensive surveys.


1.1  Instrumentation

The XCTDs were digital TSK probes purchased from Sippican (Sippican, Inc.) 
and manufactured by TSK (The Tsurumi-Seiki Co.). The computer, deck unit, 
and launcher were supplied as standard ship's equipment on the Knorr. The 
deck unit was the Sippican MK-21 model.


1.2  Data acquisition

Data acquisition was on a pc computer with the Windows 2000 Professional 
operating system. (Minimum computer requirements for the Sippican software 
are a P3/700 with 64 MB of RAM, with W2000 or XP). Two copies of the data 
files were made; one on the pc hard disk and the second on either a 
networked drive or in a backup directory on the pc. The Sippican software 
versions were WinMK21 v2.1.2, MK21COEF v2.3.1, and MK21AL v2.5.1. The XCTD 
computer and the Sippican MK-21 deck unit were located in the computer rack 
in the main lab. The hand launcher and XCTDs were kept in the aft hangar, 
and the launches were staged from the hanger.


1.3  Launch Procedure

XCTD launching was a two person effort because the weather deck on the Knorr 
was secured while underway during most of the cruise, thus requiring two 
persons on deck and radio communication to the bridge. XCTD launch times 
were determined from the ETA time and range to station from the main ODF 
AAIW webpage. The bridge was notified via radio. One person opened the"New 
Launch" window of the MK-21 software while the second person went aft to 
load a new probe in the hand launcher. The software cycles through "Testing 
Probe", "Prepare to Launch", and "Launch Probe". If it is successful in 
reading the probe's EPROM, it will usually get through to the "Launch Probe" 
window. At this point both persons, in work vests and equipped with a 
handheld radio, would go out to launch the probe. There were four launch 
locations, and the choice was dictated by wind and seas. A permanent launch 
tube was located on the port side, just aft of the hangar. A second launch 
tube was tried in various locations; it was usually located on the rail on 
the port side of the fantail. The third location was the starboard side of 
the fantail, and a fourth was the starboard side immediately across from the 
aft hangar. The fall rate is approximately 200 m/min, and a cast typically 
took 5 mins. If a launch tube was used, both persons would come back inside 
and monitor the launch on the computer. If the probe was hand-launched, one 
person would watch from the hangar where they could also view the computer 
screen if they moved to the doorway into the main lab. The spent canister 
was retrieved after the launch. The data file was inspected and serial 
number (SN), time, latitude, and longitude were recorded to logsheets and 
reported to the bridge.


1.4  Problems encountered

We launched 399 probes and had a total of 342 good casts (defined as casts 
to depths of at least 800 m) with an overall success rate of 86%. The main 
reasons for XCTD failures were 1) XCTD wire contacting the ship, usually due 
to wind, 2) XCTD launch not recognized by the software (despite the fact 
that the EPROM was read ok and the SN displayed correctly in the "Launch 
Probe" window), 3) a false splash (the software starts recording when the 
probe is in air, but the computer operator has gone aft to launch and does 
not know to abort the cast), 4) loading a probe too early while the software 
is running (operator error - we did not know that the batteries in the probe 
run down in about 15 mm if the program is active). The table at the end of 
this section summarizes the XCTD performance. Also note that some casts did 
not profile to maximum depth, usually for unknown reasons.

Part of our launch procedure initially was to wet the end of the probe with 
Jet-Dry, to improve adhesion and thus decrease spiking near the surface. 
This technique is used for analog XCTDs, but it may have been the cause of 
some of our early false splash failures with the digital probes. After 
communication with Sippican, we stopped using Jet-Dry; however, we still 
encountered false launches and other failures. The system was examined for 
wiring defects such as ground loops. The launcher cable appeared to be wired 
correctly, after testing with an ohm meter between the launcher and the 
Sippican connector box at the computer. However, the computer chassis, 
monitor, and Sippican deck unit were all independently grounded, and we were 
advised by Justine Afghan and Glenn Pezzoli that floating all but one ground 
was considered critical for the setup of the XCTD systems that they install 
on the high resolution XBT container ships. We floated the computer and deck 
unit ground, using a single ground in the connector box. However the 
monitor, the GPS, and other computers in the rack all had independent 
grounds which we were unable to change, and these could be a potential 
problem. Our success rate improved dramatically (from 75% to 90% or 95%) 
after we floated the ground on the computer chassis and deck unit, until the 
last week of the cruise when we had a string of failures (mostly undetected 
launches). A loose ground wire was found on the connector box between the 
computer and the launcher and reconnected, but that did not solve the 
problem. The source of the failure turned out to be a broken wire in the 
stress-relief section of the launcher cable, located where it entered the 
hand launcher. Other things that were tried at this time were 1) rebooting 
the computer, power cycling the deck unit, and restarting the software 
before every cast, 2) installing the software on a new computer outside of 
the rack, 3) turning off virus detection software, 4) disconnecting from 
GPS, and 5) disconnecting from the network. We had a Sippican digital test 
canister that did not detect the wiring problem, possibly because the wire 
ends still had intermittent contact. Continued testing by the Knorr's SSSG 
technician (Robbie Laird) indicated that the broken wire only had to make 
contact for an instant for the test canister to initiate a successful test 
cast. It is unknown whether the same is true for an actual XCTD cast.


                           XCTD Cast Statistics
----------------------------------------------------------------------------
             No. of casts                  Drop quality
             ------------  -------------------------------------------------
Used probes      342       Good cast to depth > 800 m
                 10        Good cast to depth < 800 m
                 16        Bad or truncated cast due to wind or wire on deck
                 17        Bad cast; failure to recognize launch
                 9         Bad cast; reason unknown
                 5         Bad cast; false splash
-----------------------------------------------------------------------------
Unused probes    9  
-----------------------------------------------------------------------------
Total            408       34 cases of 12



1.5  Data processing

The Sippican automated processing was the only processing that was applied. 
Two files exist for each cast: RDF (binary, raw) and EDF (ascii, edited by 
the Sippican autoprocessing). An example of a succession of good temperature 
casts to 1000 m is shown in Fig. 1.


Figure 1: Sample of XCTD temperature profiles from casts 168 to 172



1.6 Recommendations

After diagnosing our final problem (broken wire in the launcher cable), we 
stopped the reboot/power cycle/restart routine, and we put the XCTD computer 
back on the network. We kept the virus software turned off, and we did not 
connect to the GPS. All of these things could potentially cause problems, 
but it is our assessment that they were probably not an issue in the Knorr 
setup. It is recommended to bring our own computer, cable, and launcher with 
a stable and tested version of the Sippican software installed. Then the 
science party has control with respect to: network, external inputs (GPS), 
grounding, and virus software. As long as time is accurate, the system does 
not need to be networked and does not need real-time GPS input.

Other recommendations for the summer cruise entail additional items to log 
for each cast that are helpful in assessing the quality of the cast and in 
diagnosing problems.

We suggest:

• Log whether spiking occurs in the profile and the maximum depth of the 
  profile.
• Log the serial no. prior to deployment so that all probes get logged, even 
  failures.
• Track the box number as well as serial no., in case the probe is part of a 
  bad batch that needs to be reported to Sippican for a refund.
• Check whether the probes in a single box are all in sequence, again for 
  tracking purposes.



1.7 Acknowledgements

Thanks to Justine Afghan and Glenn Pezzoli for advice during the AAIW winter 
cruise.





                   AAIW: PRELIMINARY ADCPILADCP CRUISE REPORT



         ANTARCTIC INTERMEDIATE WATER FORMATION IN THE SOUTHEAST PACIFIC

                          21 August to 6 October 2005

                             by Teresa K. Chereskin





1  INTRODUCTION

During the Antarctic Intermediate Water (AAIW) winter cruise, direct 
velocity measurements were made by the Chereskin lab group of Scripps 
Institution of Oceanography (SlO) from hull mounted shipboard acoustic 
Doppler current profilers (SADCPs) and from a Lowered Acoustic Doppler 
Current Profiler (LADCP).



2  SHIPBOARD ADCPS


2.1  Instrumentation

Data was recorded from two shipboard ADCPs: an Ocean Surveyor 75 kHz phased 
array (OS75) and an RD Instruments 150 kHz narrowband ADCP (NB 150).

The OS75 is standard ship's equipment on R/V Knorr. The OS75 ADCP transducer 
was mounted in an instrument well located near the center line of the ship 
and below the laundry room. The well is open to the sea, and the transducer 
is located at approximately 5 m depth, with beam 3 oriented 45 deg to 
starboard.

The NB 150 is an obsolete instrument, no longer supported by the 
manufacturer, that was installed by WHOI on request from the P1 specifically 
for the AAIW cruise in order to profile currents at higher resolution and at 
shallower depths than the OS75. The NB 150 ADCP transducer was mounted in an 
instrument well located below the lower lab at frame 85, about 8 feet 
starboard of the center line. The well is open to the sea, and the 
transducer is located at approximately 5 m depth, with beam 3 oriented 45 
deg to starboard. The NB 150 that was installed in Miami for AAIW failed 
prior to the ship's arrival in Punta Arenas, Chile. A second complete system 
was sent via air freight. Although the system had checked out satisfactorily 
at WHOI, it reported error messages after installation on Knorr. In actual 
use, the problem was very low signal on beam 2 (unsuitable for a 4-beam 
velocity solution). We collected NB 150 data with the intention of 
implementing a 3-beam solution.


2.2  Data acquisition

Single ping ADCP data from both instruments and ancillary navigation streams 
(GPS, gyrocompass, and POS/MV) were collected on a Dell 1-U rack-mounted 
server running the Linux operating system (Mandrake 10.2) using UHDAS, a 
data acquisition and processing software suite written by Eric Firing and 
Jules Hurnmon, University of Hawaii.The data were processed in real-time on 
the Linux server (currents.knorr.whoi.edu) and were recorded in duplicate on 
a pair of internal, mirrored hard disks. Data were copied to Mac G4 laptops 
via a network (Samba) exported filesystem for further processing. The 
primary heading source was the ship's gyrocompass, and heading corrections 
were made using the POS/MV. After applying the heading corrections, the 
overall additional calibration was an amplitude of 1.0 and a phase of 0.0 
degrees. This calibration will be refined in post-processing.


2.3  Sampling parameters

The NB 150 operating parameters used during AAIW were 50 depth bins and an 8 
m blank, range bin, and pulse length. The OS75 ADCP was configured to 
collect data in narrowband mode. The OS75 operating parameters were 70 depth 
bins and a 16 m blank, range bin, and pulse length.


2.4  Data processing

Overall, the quality of the OS75 ADCP and navigation data acquired during 
AAIW was excellent. High precision GPS was available throughout the cruise, 
with an estimated single position fix accuracy of 1 m. The estimated 
accuracy of the POS/MV heading corrections is 0.1°(King and Cooper, 1992). 
The overall error in absolute currents is estimated at 1-2 cm s-1 
(Chereskin and Harris, 1997). The main problems encountered were bubble 
sweepdown when the bow thruster was used to maintain station and during 
rough weather and heavy seas. The maximum profiling range of the OS75 was 
about 850 m, but this depth range was drastically curtailed when bubbles 
were severe.

The NB 150 data were processed using a 3-beam solution. Where the data 
overlap with the OS75, they are of higher resolution. Unlike the OS75, he NB 
150 was not affected by bubbles from the bow thruster. It was negatively 
affected by bubble sweepdown during rough weather and heavy seas. The 
maximum range was about 225 m; typical range was 180 m.


3  LOWERED ADCP


3.1  Instrumentation

The lowered ADCP was Chereskin's 150 kHz RDI Phase 3 broadband ADCP, serial 
number 1394, firmware versions 1.16 (XDC), 5.52 (CPU), 3.22 (RCDR), and C5d3 
(PWRTIM). The LADCP has custom 30° beam angles. It was mounted on the outer 
edge of the CTD rosette, about 1 inch above the bottom of the frame. A 
rechargeable lead acid gel cell battery in an oil-filled plastic case 
(SeaBattery, Ocean Innovations, La Jolla, CA) was mounted in a steel box 
that was hose-clamped to the bottom of the rosette frame.


3.2  Data acquisition

A Mac G4 laptop computer running OSX (Panther 10.3.9) was used to upload an 
LADCP command set prior to each cast, using serial communication and a 
python terminal emulator (rditerm.py). Data acquired during the cast were 
stored internally on a 20 MB EPROM recorder. Data recovery used the terminal 
emulator, a public domain ymodem program (lib), and a shell script to change 
the baud rate (change-baud) once the ymodem transfer was initiated.


3.3  Sampling protocol

Commands were uploaded from a file for deployment. The profiler was 
instructed to sample in a 2 ping burst every 2.6 seconds, with 0 s between 
pings and 1 s between (single-ping) ensembles, resulting in a staggered ping 
cycle of [1 s, 1.6 s]. Other relevant setup parameters were 16x16 m bins, 
16 m blank, 16 m pulse, bandwidth parameter WB 1, water mode 1, and an 
ambiguity velocity of 330 cm s-1 Data were collected in beam coordinates.

The battery pack was recharged after every cast, using an AmRel linear 
programmable power supply. The power supply was set to 57.31 V constant 
voltage and 1.8 A maximum current. Typically, at the end of a cast, the 
power supply was current-limited at the maximum current. The power supply 
switched within about 10 mm to constant voltage as the current level 
dropped. Charging was stopped nominally at 0.6 A in order to minimize the 
chance of overcharging, although the power supply resorts to trickle 
charging as the battery approaches full charge. Since lead acid gel cells 
outgas small amounts of hydrogen gas when overcharged/discharging, it is 
necessary to vent the pressure case. The pressure case was vented every few 
casts. There was a small but noticeable amount of outgassing.


3.4  Data processing

The LADCP provides a full-depth profile of ocean current from a 
self-contained ADCP mounted on the CTD rosette. Using the conventional 
"shear method" for processing (e.g., Fischer and Visbeck, 1993), overlapping 
profiles of vertical shear of horizontal velocity are averaged and gridded, 
to form a full-depth shear profile. The shear profile is integrated 
vertically to obtain the baroclinic velocity and the resulting unknown 
integration constant is the depth-averaged or barotropic velocity. This 
barotropic component is then computed as the sum of the time-averaged, 
measured velocity and the ship drift (minus a small correction, less than 
1 cm S-1 to account for a nonconstant fall rate) (Fischer and Visbeck, 1993; 
Firing, 1998). Errors in the baroclinic profile accumulate as 1√(N) where N 
is the number of samples (Firing and Gordon, 1990). This error translates to 
the lowest baroclinic mode and, for a cast of 2500 m depth, it is about 2.4 
cm s-1(Beal and Bryden, 1999). The barotropic component is inherently more 
accurate, because the errors result from navigational inaccuracies alone. 
These are quite small with P-code GPS, about 1 cm s-1(2 to 4 cm s-1without). 
Comparisons with Pegasus suggest that the LADCP can measure the 
depth-averaged velocity to within 1 cm s-1(Hacker et al., 1996). The rms 
difference between Pegasus and LADCP absolute profiles are within the 
expected oceanic variability, 3-5 cm s-1(Send, 1994), due primarily to high 
frequency internal waves.

In previous experiments the interference layer, which results from the 
previous ping reflecting off the bottom, has caused a large data gap in the 
LADCP profile, causing an uncertain velocity offset (several cm s-1) between 
the parts of the profile on either side of the gap. For this experiment 
bottom velocities were greatly improved by using Chereskin's instrument 
which pings asynchronously, thereby avoiding complete data loss in the 
interference layer. A second problem with data loss arises at the bottom of 
a CTD/LADCP cast, when the package is held 10 m above the sea bed for bottle 
sampling. At this distance the instrument is 'blind' since the blank after 
transmit is order 20 m, and a time gap in the data stream will result in an 
uncertainty in the absolute velocity. We attempted to minimize the stop at 
the bottom of the cast to keep this gap to a minimum.

Initial processing was done with the University of Hawaii CODAS software. 
The method is the traditional shear method outlined in Fischer and Visbeck 
(1993) as implemented by Eric Firing in the UH CODAS LADCP software. CTD 
time series data were available immediately following the cast which 
provided more accurate depth than from integrating LADCP vertical velocity 
as well as calculated sound speed at the transducer. Typically LADCP casts 
were analyzed through to absolute velocity, including CTD data, prior to the 
next station.

During the cruise, the casts were also processed with Martin Visbeck's LADCP 
Matlab processing routines, version 8a. The method (Visbeck, 2002) differs 
from the shear method in that an inverse technique is used which includes 
two additional constraints, the bottom velocity estimate and the average 
shipboard ADCP profile during the cast. In principle, the Firing shear and 
Visbeck inverse methods should agree when no additional constraints are 
included in the inverse, but at the moment the methods have shown 
unexplained differences on some data sets (Brian King, pers. comm.) 
Qualitatively, the absolute currents computed between the 2 methods agreed 
reasonably well. Detailed comparisons will be made in post-processing. 
Preliminary comparisons of shipboard and lowered ADCP data also showed 
fairly good agreement and suggest that the shipboard data will be a useful 
constraint in the inverse method utilized by Visbeck.



Acknowledgements

Thanks are due to the SIO Oceanographic Data Facility for their outstanding 
support on this cruise. Thanks also to Sharon Escher and Yueng-Djern Lenn 
for their diligent watchstanding and processing efforts.




4  REFERENCES

Beal, L.M., and H.L. Bryden, The velocity and vorticity structure of the 
    Agulhas Current at 32°S, J. of Geophys. Res., 104, 5151-5176, 1999.

Chereskin, T.K., and C.L. Harris, Shipboard Acoustic Doppler Current 
    Profiling during the WOCE Indian Ocean Expedition: 110, Scripps 
    Institution of Oceanography Reference Series, SlO-97-14, 137 pp, 1997.

Firing, E.F. and R. Gordon, Deep ocean acoustic Doppler current profiling, 
    Proceedings of the IEEE Fourth International Working Conference on 
    Current Measurements, Clinton, MD, Current Measurement Technology 
    Committee of the Ocean Engineering Society, 192-201, 1990.

Firing, E., Lowered ADCP development and use in WOCE, It. WOCE Newsletter, 
    30, 10-14, 1998.

Firing, E., Erratum, Intl. WOCE Newsletter, 31, 20, 1998.

Fischer, J. and M. Visbeck, Deep velocity profiling with self-contained 
    ADCPs, J. Atmos. and Oceanic Tech., 10, 764-773, 1993.

Hacker, P., E. Firing, W.D. Wilson, and R. Molinari, Direct observations of 
    the current structure east of the Bahamas, Geophys. Res. Let., 23, 
    1127-1130, 1996.

King, B.A. and E.B. Cooper, Comparison of ship's heading determined from 
    an array of GPS antennas with heading from conventional gyrocompass 
    measurements, Deep-Sea Res., 40, 22072216,1993.

Send, U., The accuracy of current profile measurements - effect of tropical 
    and mid-latitude internal waves, J. Geophys. Res., 99, 16229-16236, 1994.

Visbeck, M., Deep velocity profiling using lowered acoustic Doppler current 
    profilers: bottom track and inverse solution, J. Atmos. and Oceanic 
    Tech., 19, 795-807, 2002.



Figure 1: LADCP section across the Subantarctic Front, stations 009 to 020. 
          Upper panel is eastward current (cm/s). Lower panel is northward 
          current (cm/s). Red line on station map indicates location of 
          section.



CCHDO DATA PROCESSING NOTES

Event Date  Person         Date Type         Summary 
----------  -------------  ----------------  ----------------------------------
2009-11-02  Talley, Lynne  SUM/BTL/CTD/DOCS  Data are public
            Q: (Diggs) "Do you mind if we put this AAIW-2005 cruise data 
               online..."
            A: (Talley) "Yes, they should all be on the CCHDO site. There was a 
               second cruise as well, in 2006, which should also be included on 
               the CCHDO site." 

2009-11-03  Diggs, Steve   SUM/BTL/CTD/DOC   Data online 
            Downloaded files from L. Talley's website (see previous history 
            entries). Corrected minor format problems in exchange bottle and 
            WOCE summary files. Updated all expocodes in all files. Made WHP-
            Exchange CTD files from SUM and WOCE CTD files. Level-0 QC and 
            format checked all data files, then placed online. Added PI's PDF 
            documentation to CCHDO site as well.

            Updated all history database tables. 

