A. Cruise Report: S04I A.1. Highlights WHP Cruise Summary Information WOCE line designation S04I Expedition designation (ExpoCode) 320696_3 Chief scientists and affiliation Thomas Whitworth III and James H. Swift* Ship RVIB NATHANIEL B. PALMER Cruise dates 1996.MAY.03 - 1996.JUL.04 Ports of call Cape Town, S. Africa - Hobart, Australia Number of stations 108 full-depth CTD stations 58ƒ 0.23' S Geographic boundaries 20ƒ 0.34' E 120ƒ 0.08' E 65ƒ 41.97' S Floats and drifters deployed 17 ALACE floats Moorings deployed or recovered 9 self-reporting current meter moorings *Thomas Whitworth III James H. Swift Department of Oceanography Scripps Institution of Oceanography Texas A&M University University of California, San Diego College Station, TX, 77843 La Jolla, CA, 92093 twhitworth@tamu.edu jswift@ucsd.edu Authors S. Rutz, D. Chipman, M. Mensch, F. Delahoyd, D. Breger, R. Key, E. Peltola, T. Whitworth WOCE Hydrographic Program Line S04I was conducted on the RVIB NATHANIEL B. PALMER on voyage S229 from 3 May to 4 July, 1996. The voyage began in Cape Town, Republic of South Africa, and ended in Hobart, Australia. Co-Chief Scientists for the cruise were Thomas Whitworth III/TAMU and James H. Swift/SIO. WHP leg S04I was a cooperative effort among the PIs listed in Table 1. The members of the scientific party are listed in Table 2. Table 1. Principal Investigators for WOCE S04I Component Principal Investigator Institution ----------------- ---------------------- ----------- CTD/Hydrography J. Swift SIO CFCs W. Smethie/M. Warner LDEO/UW Tritium, 3He, 18O P. Schlosser LDEO CO2 T. Takahashi LDEO Alkalinity F. Millero Miami 14C R. Key Princeton current meters W. Nowlin/T. Whitworth TAMU Transmissometer W. Gardner TAMU LADCP E. Firing/P. Hacker UH ALACE floats R. Davis SIO Table 2. Participants on WOCE S04I Participant Affiliation Responsibility ---------------- ----------- ----------------------------------------------- Isabelle Ansorge UCT CTD console, sampling, salinities Dee Breger LDEO Tritium, 3He, 18O Christie Campbell ASA deck ops, sampling, hazardous waste, salinities Kent Chen ASA sampling, oxygen, NBP computer David Chipman LDEO CO2 Scott Colburn ASA PDR, sampling, NBP ET Craig Hallman ODF deck ops, sampling, oxygen Steve Covey UW CFCs Frank Delahoyde ODF ODF systems, data q.c. Bob Key PU 14C, gadfly Leonard Lopez ODF deck ops, oxygen Guy Mathieu LDEO CFCs Carl Mattson ODF TIC, deck supv., ET Rod McCabe ASA sampling, NBP computer Manfred Mensch LDEO CFCs Stacey Morgan ODF nutrients Jim Noyes SIO CTD console, sampling Alex Orsi UW CTD console, sampling, analysis Ron Patrick ODF deck supv., bottle q.c. Esa Peltola UM Alkalinity Erik Quiroz TAMU nutrients Blaine Reynolds ASA PDR, rosette prep., NBP ET Stephany Rubin LDEO CO2 Steve Rutz TAMU watch leader., CTD console, sampling, ADCP Buzz Scott ASA deck ops., salinities, MT Colm Sweeney LDEO CO2 Jim Swift SIO watch leader., CTD console, sampling, analysis Mark Talkovic ASA deck ops., salinities, MT Tom Whitworth TAMU indirection, analysis Kevin Wood ASA deck ops., sampling, MPC ASA: Antarctic Support Associates UM: University of Miami (RSMAS) 61 Inverness Dr. East, Suite 300 4600 Rickenbacker Cswy. Englewood, CO 80112 Miami, FL 33149 SIO: Scripps Instit. of Oceanog. LDEO: Lamont-Doherty Earth Observatory Univ. of Calif.-San Diego Columbia University La Jolla, CA Palisades, NY 10964 TAMU: Texas A&M University UW: University of Washington Dept. of Oceanography School of Oceanography College Sta. TX 77843 Seattle, WA 98195 ODF: Ocean Data Facility UCT: University of Cape Town Scripps Instit. of Oceanog. Department of Oceanography 9500 Gilman Dr. Rondebosch, Cape Town La Jolla, CA 92093 South Africa PU: Princeton University Dept. of Geosciences 207 Guyot Hall Princeton, N.J. 08544 A2. Scientific Program Summary Narrative The cruise constituted the Indian Ocean portion of WOCE line S04, a meridional circumnavigation of Antarctica at a nominal latitude of 60ƒS. This segment covered the longitudes 20ƒE to 120ƒE. After departure from Cape Town, a bottom-tracking course was set to provide about 8 hours of depths along the 200-m isobath to calculate the offset between the ship's gyro and the underway Acoustic Doppler Current Profiler (ADCP). Upon reaching deep water, the CTD wire was lowered to 5500 m wire out to tension the wire on the winch, and subsequently, two test CTD casts were made to choreograph the procedures for launching and recovering the package in the unfamiliar setting of the Palmer's Baltic Room. At 0330 on 7 May, the Palmer turned back toward South Africa to seek medical attention for a crew member. The ship was diverted to the naval base at Simonstown where fuel was available, and the morning and afternoon of May 10 were spent getting the crewman treated and refueling. Bottom-tracking for the ADCP calibration was repeated into and out of Simonstown. On Sunday, 12 May, a third CTD test cast was being recovered when a sudden wave lifted the rosette out of the water and then dropped it. The wire parted at the sheave, and the entire package was lost. The only piece of equipment without back-up was the lowered ADCP unit belonging to the University of Hawaii. Subsequent days were spent preparing a second rosette unit, considering alternative launch and recovery procedures and defining guidelines for the sea state in which CTD operations could be conducted on the Palmer. Because the Palmer reacted differently from UNOLS vessels we were accustomed to, the planned cruise track was modified to lie in, or closer to the ice where swell would be less of a problem. Station 1 was occupied at 58ƒS, 20ƒE and the first station line was run southeast to Gunnerus Ridge, about 50 miles south of the ice edge. Station positions for the cruise are shown in Fig. 1. During the transit to station 1 and continuing to 58ƒS, 17 ALACE floats were launched. Details are provided in Table 3. Stations across the Enderby Abyssal Plain trended east-northeast from 66ƒS at 33ƒE, to 61ƒS at 83ƒE on the Kerguelen Plateau. A line of stations (35-42) was made north from the 500-m isobath on the continental slope at 53ƒE, and three self-reporting current meters were deployed along the slope. Details of the current meter deployments are given in Table 4. A line of stations (65- 72) extending east from the crest of the Kerguelen Plateau was made at about 59ƒ S, and three more current meters were placed in the boundary current on the eastern flank of the Plateau. On June 8, after station 72, science operations were suspended for seven days when the Palmer was diverted to Mirnyi Station in the Davis Sea to deliver emergency food supplies. On June 14, the Palmer left Mirnyi and began a line of stations (73-86) from the shelf break of the Davis Sea to Kerguelen Plateau. One current meter was placed near the 3000-m isobath north of the Antarctic Continental Slope, and two were deployed at the southern end of Kerguelen Plateau. The zonal line of stations at a nominal latitude of 62ƒS was resumed at 90ƒE. Ice conditions, fuel and time considerations necessitated 45-mile station separation for most of the final 22 stations, which terminated with station 108 at 120ƒE. Summary Information 108 full-depth CTD stations were made exclusive of test stations at the beginning of the cruise and a dedicated CFC archive-sample cast at the end of the cruise. Nine self-reporting current meters and 17 ALACE floats were deployed Table 3. ALACE deployments on WOCE S04I Ser# Type Lat Long Time/Date Ser# Type Lat Long Time/Date ---- ---- -------- -------- --------- ---- ---- -------- -------- ---------- 628 T 37-58.3S 20-16.1E 2056Z 5/4 641 Std 47-59.4S 19-12.0E 2221Z 5/13 346 Std 38-59.1S 21-46.4E 0610Z 5/5 642 Std 49-32.6S 19-16.5E 0700Z 5/14 629 T 39-59.1S 21-46.1E 1546Z 5/5 643 Std 50-59.8S 19-20.7E 1458Z 5/14 559 Std 40-59.4S 21-42.2E 2118Z 5/5 607 CTD 51-58.7S 19-23.8E 2045Z 5/14 634 Std 41-58.8S 21-38.5E 0228Z 5/6 644 Std 52-59.8S 19-26.4E 0224Z 5/15 566 Std 43-29.9S 21-32.9E 1040Z 5/6 608 CTD 54-30.0S 19-31.7E 1035Z 5/15 604 CTD 44-59.9S 21-27.3E 2042Z 5/6 645 Std 55-59.4S 19-47.2E 1855Z 5/15 640 Std 45-59.9S 19-06.5E 1118Z 5/13 609 CTD 57-30.0S 19-58.6E 0331Z 5/16 605 CTD 47-00.0S 19-09.7E 1651Z 5/13 *T = temperature, Std = Standard, CTD = cond/temp/depth Table 4. Self-reporting current meter deployments CMß serial# time date latitude longitude depth MAB* --- ------- ----- --------- --------- --------- ------ --- A 26935 0732Z 28 May 96 65-22.8 S 53-14.2 E 1740 m 50 B 26939 1125Z 28 May 96 65-14.0 S 53-06.5 E 1900 m 50 C 26936 1904Z 28 May 96 65-07.0 S 52-59.6 E 2260 m 100 D 26937 1706Z 6 June 96 59-42.9 S 84049.9 E 2020 m 50 E 26941 2209Z 6 June 96 59-40.6 S 85-07.9 E 3080 m 50 F 26943 2231Z 6 June 96 59-40.3 S 85-10.9 E 4225 m 50 G 26944 1331Z 16 June 96 64-03.8 S 92-21.5 E 3275 m 100 H 26938 0614Z 19 June 96 63-00.0 S 85-00.0 E 3058 m 50 I 26940 1212Z 19 June 96 62-59.6 S 84-31.5 E 2740 m 50 ß letters correspond to Fig. 1 * MAB = meters above bottom World Ocean Circulation Experiment Southern Indian Ocean S4I R/V Nathaniel B. Palmer NBP96-3 3 May - 4 July 1996 Cape Town, South Africa - Hobart, Tasmania, Australia Expocode: 320696_3 Co-Chief Scientists: Dr. Thomas Whitworth (Texas A&M University) Dr. James H. Swift (Scripps Institution of Oceanography) S4I Cruise Track Oceanographic Data Facility (ODF) Final Cruise Report 1 August 2003 Data Submitted by: Oceanographic Data Facility Scripps Institution of Oceanography La Jolla, CA 92093-0214 http://odf.ucsd.edu DESCRIPTION OF MEASUREMENT TECHNIQUES AND CALIBRATIONS 1. Basic Hydrography Program The basic hydrography program consisted of salinity, dissolved oxygen and nutrient (nitrite, nitrate, phosphate and silicate) measurements made from bottles taken on CTD/rosette casts, plus pressure, temperature, salinity and dissolved oxygen from CTD profiles. 109 CTD/rosette casts were made at 108 stations, usually to within 5-15 meters of the bottom. Station 2 cast 1 was aborted at the surface because of signal failure at 322m on the down- cast; it is not otherwise mentioned in this release or documentation. Water was found inside the CTD case; after repairs, station 2 cast 2 was successfully accomplished. 17 ALACE floats were deployed during the transit from Cape Town to station 1. 9 expendable current meters were deployed following 8 stations along the cruise. The R/V Nathaniel B. Palmer departed from Cape Town, South Africa on May 3, 1996. One test cast was accomplished on May 6; on May 7, the ship turned back toward South Africa to seek medical attention for a crew member. The ship docked at the naval base at Simonstown on May 10, departing later the same day to resume the expedition. 2 more test casts were done during the transit. During the recovery of the second of these casts, a rogue wave lifted the rosette out of the water and then dropped it. The wire parted at the sheave, and the rosette package was lost. A backup rosette was prepared and used for the remainder of the cruise. 108 CTD/Rosette stations were occupied between May 16 and June 27 along the nominal S4I line (60 deg.S), between 58-66 deg.S latitude and 20-120 deg.E longitude. An additional line (stations 35-42) was made northward from the 500m isobath on the continental slope at 53 deg.E back to the main track. There was a 6.5-day (June 8-14) diversion from the track after station 72 to deliver emergency food supplies to Mirnyy Station in the Davis Sea. After Mirnyy, an extra line (stations 73-86) was run northward, then westward, from the shelf break of the Davis Sea back toward the S4I line. The cruise ended in Hobart, Tasmania, Australia on July 4, 1996. 3655 bottles were tripped resulting in 3651 usable bottles. Any problems encountered during data acquisition or processing are described later in this document. The resulting data set met and in many cases exceeded WHP specifications. The distribution of samples is illustrated in Figures 1.0, 1.1 and 1.2. Figure 1.0 S4I sample distribution, stas 1-34. Figure 1.1 S4I sample distribution, stas 35-72. Figure 1.2 S4I sample distribution, stas 73-108. 2. Water Sampling Package Hydrographic casts were performed with a rosette system consisting of a 36-bottle rosette frame (ODF), a General Oceanics (GO) 36-place pylon (Model 2216) and 36 10-liter PVC bottles (ODF). Underwater electronic components consisted of an ODF-modified NBIS Mark III CTD (ODF #3) and associated sensors, SeaTech transmissometer (TAMU) and Benthos pinger (Model 2216). The CTD was mounted horizontally along the bottom of the rosette frame, with the transmissometer, a SensorMedics dissolved oxygen sensor and an FSI secondary PRT sensor deployed next to the CTD. The pinger was monitored during a cast with a precision depth recorder (PDR) in the ship's laboratory. The rosette system was suspended from a three- conductor 0.322" electro-mechanical cable. Power to the CTD and pylon was provided through the sea cable from the ship. Separate conductors were used for the CTD and pylon signals. The transmissometer, dissolved oxygen and secondary temperature were interfaced with the CTD, and their data were incorporated into the CTD data stream. Deep Sea Reversing Thermometers (DSRTs) were used occasionally on this leg to monitor for CTD pressure or temperature drift. Three rosette test casts were performed prior to station 1: 998 (6 May), 997 (11 May) and 996 (12 May). During retrieval on the third test cast (996), a wave caught the rosette, and the wire jumped the sheave and broke. The rosette, bottles and all associated electronics were lost. The only instruments that did not have backup units were the UH LADCP and an ODF altimeter. A spare altimeter was used during stations 1-5 and 8-10, but was removed for the rest of the cruise; it never worked properly and was identified as the source of the degraded signal seen during the up-cast for station 5. The deck watch prepared the rosette approximately 45 minutes prior to each cast. All valves, vents and lanyards were checked for proper orientation. The bottles were cocked and all hardware and connections rechecked. Time, position and bottom depth were logged by the console operator at arrival on station. The rosette system was deployed from the Palmer's main deck out of the starboard-side Baltic Room, a protected rosette room and winch shed with an external door and an extension boom. The deployment door to the Baltic Room was opened after the ship had finished positioning, which sometimes entailed clearing a hole in the ice. Deployment was assisted by tag lines threaded through rings on the rosette for stabilization. Each rosette cast was lowered to within 5-15 meters of the bottom, unless the bottom return from the pinger was extremely poor. As noted already, no altimeter data were available to assist with bottom approaches after station 5. Bottles on the rosette were each identified with a unique serial number. Usually these numbers corresponded to the pylon tripping sequence, 1-36, where the first (deepest) bottle tripped was bottle #1. Bottle #8 had repeated drain valve leakage problems and was replaced with bottle #37 (stations 13-25 and 35-47), Ocean Instrument Tech. (OIT) test bottle #61 (stations 26-34) and Antarctic Support Associates (ASA) test bottle #63 (stations 48-108). Bottle #4 was missing (apparently imploded) after station 77, and was replaced with bottle #39 for stations 78-108. GO test bottle #62 replaced bottles #10 (stations 26-28) and #6 (stations 82-83). Averages of CTD data corresponding to the time of bottle closure were associated with the bottle data during a cast. Pressure, depth, temperature, salinity and density were immediately available to facilitate examination and quality control of the bottle data as the sampling and laboratory analyses progressed. Recovering the package at the end of deployment was essentially the reverse of the launching with the additional use of air-tuggers for added stabilization. The rosette was placed onto the Baltic Room deck, then the deployment door was closed prior to sampling. The bottles and rosette were examined before samples were taken, and any unusual situations or circumstances were noted on the sample log for the cast. Seawater froze on rosette bottles several times during recovery, but quickly thawed in the Baltic Room. There was never any evidence of water freezing in the bottles or spigots. Routine CTD maintenance included soaking the conductivity and CTD O2 sensors in distilled water between casts to maintain sensor stability. Beginning at station 20, the distilled water was replaced by salt water ~1 hour prior to deployment to reduce the possibility of sensors freezing before entering the water. This preventive measure was not totally successful, and freezing did occur during deployment on some casts. When freezing was detected by the console operator, the rosette was lowered to 30-80 meters to thaw the sensors, then raised back to the surface. Rosette maintenance was performed on a regular basis. O-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. The transmissometer windows were cleaned prior to deployment approximately every 20 casts. The air readings were noted in the TAMU transmissometer log book after each cleaning. Transmissometer data were monitored for potential problems during every cast, but were not processed by ODF beyond initial block averaging. The starboard-side Baltic Room Markey winch was used throughout the cruise. Only one sea cable retermination was necessary, prior to station 57. 3. Underwater Electronics Packages CTD data were collected with a modified NBIS Mark III CTD (ODF #3). This instrument provided pressure, temperature, conductivity and dissolved O2 channels, and additionally measured a second temperature with an FSI oceantemperature module (OTM) as a calibration check. An FSI ocean pressure module (OPM) was substituted in place of the secondary temperature OTM for four casts. Other data channels included elapsed-time, several power supply voltages and transmissometer. The instrument supplied a 15-byte NBIS-format data stream at a data rate of 25 Hz. Modifications to the instrument included revised pressure and dissolved O2 sensor mountings; ODF-designed sensor interfaces for O2, FSI-OTM PRT and transmissometer; implementation of 8-bit and 16-bit multiplexer channels; an elapsed-time channel; instrument ID in the polarity byte and power supply voltages channels. Table 3.0 summarizes the serial numbers of instruments and sensors used during S4I. Table 3.0 S4I Instrument/Sensor Serial Numbers | ODF | SensorMedics | SeaTech Station(s) | CTD+ | Model 147737 | Transmissometer | ID# | Oxygen Sensor | (TAMU) -----------|------|-----------------|---------------- 1-31,39-61 | 3a | | -----------|------| | 32-38 | 3b | | -----------|------| | 62-100 | 3c | | -----------|------| 5-02-22 | 151D 101-102 | 3d | | -----------|------| | 103-106 | 3e | | -----------|------| | 107-108 | 3f | | -----------|------|-----------------|---------------- + See table below for ODF CTD serial numbers ODF CTD #3 sensor serial numbers: NBIS | Pressure | Temperature | Conductivity MKIIIB | Paine Model | PRT1 | PRT2/(PRS2) | CTD | 211-35-440-05 | Rosemount | FSI | NBIS Model (ODF-ID#) | strain gage/0-8850psi | Model 171BJ | OTM/(OPM) | 09035-00151 ----------|-----------------------|-------------|-------------|------------- 3a | | | | E55 ----------| | | OTM/1320T |------------- 3b | | | | P42 ----------| | |-------------|------------- 3c | | | OTM/1322T | ----------| 77011 | 14373 |-------------| 3d | | | OTM/1321T | ----------| | |-------------| O17 3e | | | (OPM/1326P) | ----------| | |-------------| 3f | | | OTM/1320T | The CTD pressure sensor mounting had been modified to reduce the dynamic thermal effects on pressure. The sensor was attached to a section of coiled, oil-filled stainless-steel tubing that was connected to the end-cap pressure port. The transducer was also insulated. The NBIS temperature compensation circuit on the pressure interface was disabled; all thermal response characteristics were modeled and corrected in software. The O2 sensor was deployed in a pressure-compensated holder assembly mounted separately on the rosette frame and connected to the CTD by an underwater cable. The O2 sensor interface was designed and built by ODF using an off-the-shelf 12-bit A/D converter. The transmissometer interface was a similar design. Although the secondary temperature sensor was located within 6 inches of the CTD conductivity sensor, it was not sufficiently close to calculate coherent salinities. It was used as a secondary temperature calibration reference rather than as a redundant sensor, with the intent of eliminating the need for mercury or electronic DSRTs as calibration checks. Three secondary temperature sensors were interchanged during S4I. The General Oceanics (GO) 1016 36-place pylon was used in conjunction with an ODF-built deck unit and external power supply instead of a GO pylon deck unit. This combination provided generally reliable operation and positive confirmation. The pylon emitted a confirmation message containing its current notion of bottle trip position, which could be useful in sorting out mis-trips. The acquisition software averaged CTD data corresponding to the rosette trip as soon as the trip was initiated until the trip confirmed, typically 3.5+/-1 seconds on S4I. There were 13 random bad trip confirmations during S4I; 12 of these were noticed in a timely manner by the console operator and re-tripped successfully. 3 odd trip confirmations resulted in open bottles at the surface. There were 255 other odd trip confirmations, most of which were duplicates of valid confirmations or in place of normal confirmations. 2 casts (stas 78 and 79) were re-started mid-up-cast because of pylon communication or confirmation problems. 2 casts (stas 40 and 79) had trip confirmations that were off by 1 level on many or all bottles. 2 other casts (stas 52 and 56) confirmed normally but returned to the surface with the first two bottles tripped at unknown depths, and the rest 2 trip levels deeper than expected. The bottles for these casts were matched up to the correct CTD trip depths after the casts, by comparison of CTD and bottle data water properties. Bad or odd confirmations that affected bottle trips are documented in Appendix D. 4. Navigation and Bathymetry Data Acquisition Navigation data were acquired from the ship's Ashtech GPS receiver via the network, which reported full P-code position information. Data were logged automatically at one-minute intervals by one of the Sun SPARCstations. Underway bathymetry was logged manually from the 12 kHz Raytheon/EPC PDR at five-minute intervals (or when possible in the ice), then corrected according to Carter [Cart80] and merged with the navigation data to provide a time-series of underway position, course, speed and bathymetry data. These data were used for all station positions, PDR depths and bathymetry on vertical sections. 5. CTD Data Acquisition, Processing and Control System The CTD data acquisition, processing and control system consisted of a Sun SPARCstation LX computer workstation, ODF-built CTD and pylon deck units, CTD and pylon power supplies, and a VCR recorder for real-time analog backup recording of the sea cable signal. The Sun system consisted of a color display with trackball and keyboard (the CTD console), 18 RS-232 ports, 2.5 GB disk and 8mm cartridge tape. Two other Sun SPARCstation LX systems were networked to the data acquisition system, as well as to the rest of the networked computers aboard the Palmer. These systems were available for real-time CTD data display and provided for hydrographic data management and backup. Two HP 1200C color inkjet printers provided hardcopy capability from any of the workstations. The CTD FSK signal was demodulated and converted to a 9600 baud RS-232C binary data stream by the CTD deck unit. This data stream was fed to the Sun SPARCstation. The pylon deck unit was connected to the Sun LX through a bi-directional 300 baud serial line, allowing bottle trips to be initiated and confirmed by the data acquisition software. A bitmapped color display provided interactive graphical display and control of the CTD rosette sampling system, including real-time raw and processed CTD data, navigation, winch and rosette trip displays. The CTD data acquisition, processing and control system was prepared by the console watch a few minutes before each deployment. A console operations log was maintained for each deployment, containing a record of every attempt to trip a bottle as well as any pertinent comments. Most CTD console control functions, including starting the data acquisition, were initiated by pointing and clicking a trackball cursor on the display at icons representing functions to perform. The system then presented the operator with short dialog prompts with automatically generated choices that could either be accepted as defaults or overridden. The operator was instructed to turn on the CTD and pylon power supplies, then to examine a real-time CTD data display on the screen for stable voltages from theunderwater unit. Once this was accomplished, the data acquisition and processing were begun and a time and position were automatically logged for the beginning of the cast. A backup analog recording of the CTD signal on a VCR tape was started at the same time as the data acquisition. A rosette trip display and pylon control window popped up, giving visual confirmation that the pylon was initializing properly. Various plots and displays were initiated. When all was ready, the console operator informed the deck watch by radio. Once the deck watch had deployed the rosette, it was immediately lowered without pausing at the sea surface. The deck watch informed the console operator that the rosette was on its way down (also confirmed by the computer displays). If the console operator noticed that sensors were frozen on entry, the package was stopped at 30-80 meters, then raised to just below the surface to allow the sensors to thaw. The console operator or deck watch leader then provided the winch operator with a target depth (wire-out) and maximum lowering rate, normally 60-70 meters/minute for this package. The package built up to the maximum rate during the first few hundred meters, then optimally continued at a steady rate without any stops during the down-cast. The console operator examined the processed CTD data during descent via interactive plot windows on the display, which could also be run at other workstations on the network. Additionally, the operator decided where to trip bottles on the up-cast, noting this on the console log. The PDR was monitored to insure the bottom depth was known at all times. The deck watch leader assisted the console operator by monitoring the rosette's distance to the bottom using the difference between the rosette's pinger signal and its bottom reflection displayed on the PDR. No altimeter was available to assist with bottom approaches. The winch speed was usually slowed to ~30 meters/minute during the final approach. The winch and PDR displays allowed the watch leader to refine the target depth relayed to the winch operator and safely approach to within 5-15 meters of the bottom. Bottles were closed on the up-cast by pointing the console trackball cursor at a graphic firing control and clicking a button. The data acquisition system responded with the CTD rosette trip data and a pylon confirmation message in a window. A bad or suspicious confirmation signal typically resulted in the console operator repositioning the pylon trip arm via software, then re-tripping the bottle, until a good confirmation was received. All tripping attempts were noted on the console log. The console operator then instructed the winch operator to bring the rosette up to the next bottle depth. The console operator was also responsible for generating the sample log for the cast. After the last bottle was tripped, the console operator directed the deck watch to bring the rosette on deck. It was sometimes necessary to close the surface bottles "on the fly" due to a risk of slack wire at higher sea states. Once the rosette was on deck, the console operator terminated the data acquisition and turned off the CTD, pylon and VCR recording. The VCR tape was filed. Usually the console operator also brought the sample log to the rosette room and served as the sample cop. 6. CTD Data Processing ODF CTD processing software consists of over 30 programs running under the Unix operating system. The initial CTD processing program (ctdba) is used either in real-time or with existing raw data sets to: o Convert raw CTD scans into scaled engineering units, and assign the data to logical channels o Filter various channels according to specified filtering criteria o Apply sensor- or instrument-specific response-correction models o Provide periodic averages of the channels corresponding to the output time-series interval o Store the output time-series in a CTD-independent format Once the CTD data are reduced to a standard-format time-series, they can be manipulated in various ways. Channels can be additionally filtered. The time-series can be split up into shorter time-series or pasted together to form longer time-series. A time-series can be transformed into a pressure- series, or into a larger-interval time-series. The pressure calibration corrections are applied during reduction of the data to time-series. Temperature, conductivity and oxygen corrections to the series are maintained in separate files and are applied whenever the data are accessed. ODF data acquisition software acquired and processed the CTD data in real- time, providing calibrated, processed data for interactive plotting and reporting during a cast. The 25 Hz data from the CTD were filtered, response-corrected and averaged to a 2 Hz (0.5-second) time-series. Sensor correction and calibration models were applied to pressure, temperature, conductivity and O2. Rosette trip data were extracted from this time- series in response to trip initiation and confirmation signals. The calibrated 2 Hz time-series data, as well as the 25 Hz raw data, were stored on disk and were available in real-time for reporting and graphical display. At the end of the cast, various consistency and calibration checks were performed, and a 2-db pressure-series of the down-cast was generated and subsequently used for reports and plots. CTD plots generated automatically at the completion of deployment were checked daily for potential problems. The two PRT temperature sensors were inter-calibrated and checked for sensor drift. The CTD conductivity sensor was monitored by comparing CTD values to check-sample conductivities, and by deep theta-salinity comparisons between down- and up-casts as well as adjacent stations. The CTD O2 sensor was calibrated to check-sample data. Two casts (stations 30 and 31) exhibited an unacceptable level of primary PRT temperature noise which was traced to a water leak in the sensor turret. The secondary PRT temperature was used in these cases. CTD salinity for these casts is noisier than usual because of the greater distance of the secondary PRT from the conductivity sensor, and because of potential noise induced on the conductivity sensor by the flooded turret. There was a high level of conductivity drift during stations 32-38, which used a new and apparently defective conductivity sensor, and during stations 55-61, just before the original conductivity sensor was replaced with yet another new sensor. Since down- and up-cast conductivities were very different for these casts, it was necessary to use the up-casts for these stations, where bottle-CTD differences could be used to determine pressure-dependent conductivity corrections for each cast individually. A few casts exhibited conductivity offsets due to biological or particulate artifacts. Some casts were subject to noise in the data stream caused by sea cable or slip-ring problems, or by moisture in the interconnect cables between the CTD and external sensors (i.e. O2). Intermittent noisy data were filtered out of the 2 Hz data using a spike-removal filter. A least- squares polynomial of specified order was fit to fixed-length segments of data. Points exceeding a specified multiple of the residual standard deviation were replaced by the polynomial value. Density inversions can be induced in high-gradient regions by ship- generated vertical motion of the rosette. Detailed examination of the raw data shows significant mixing occurring in these areas because of "ship roll". In order to minimize density inversions, a ship-roll filter was applied to all casts during pressure-sequencing to disallow pressure reversals. The first few seconds of in-water data were excluded from the pressure-series data, since the sensors were still adjusting to the going- in-water transition. Pressure intervals with no time-series data can optionally be filled by double-quadratic interpolation/extrapolation. Most pressure intervals missing/filled during this leg were within the top 0-4 db, caused by chopping off going-in-water transition data during pressure-sequencing. However, there were a number of casts where temperature or conductivity sensors froze in transit from the deck into the water. Ideally, these were noticed by the console operator, and the casts were returned to near- surface water and restarted after thawing. However, a number of casts with freezing problems were not noticed. At the request of one of the co-chief scientists, down-cast data were extrapolated from the "thaw" point back to the surface whenever there was a clear, stable mixed layer. The resulting data were compared to original down-cast data from the un-frozen sensor, up-cast data from the same cast and density profiles. When the down-cast CTD data have excessive noise, gaps or offsets, the up- cast data are used instead. This also applied to frozen-sensor casts where down-casts could not be extrapolated without distortion, or where sensors remained frozen below the mixed layer. CTD data from down- and up-casts are not mixed together in the pressure-series data because they do not represent identical water columns (due to ship movement, wire angles, etc.). The up-casts used for final S4I CTD data are indicated in Appendix C. There is an inherent problem in the internal digitizing circuitry of the NBIS Mark III CTD when the sign bit for temperature flips. Raw temperature can shift 1-2 millidegrees as values cross between positive and negative, a problem usually avoided by offsetting the raw PRT readings by ~1.5 deg.C. The conductivity channel also can shift by 0.001-0.002 mS/cm as raw data values change between 32768/32767, where all the bits flip at once. This is typically not a problem in shallow to intermediate depths because such a small shift becomes negligible in higher gradient areas. There were a number of casts colder than -1.5 deg.C, where raw temperature values crossed the 0 deg.C threshold. All transitions falling in lower- gradient areas were shallower than 480 db and showed no density inversions. All raw conductivity values were lower than 32768 and unaffected by this problem. Appendix C contains a table of CTD casts requiring special attention. S4I CTD-related comments, problems and solutions are documented in detail. 7. CTD Laboratory Calibration Procedures Pre-cruise laboratory calibrations of CTD pressure and temperature sensors were used to generate tables of corrections applied by the CTD data acquisition and processing software at sea. These laboratory calibrations were also performed post-cruise. Pressure and temperature calibrations were performed on CTD #3 at the ODF Calibration Facility in La Jolla. Pre-cruise calibrations were done in March 1996, and post-cruise calibrations were done in July 1996. The CTD pressure transducer was calibrated in a temperature-controlled water bath to a Ruska Model 2400 Piston Gage pressure reference. Calibration data were measured pre-/post-cruise at -1.89/-1.10 deg.C to a maximum loading pressure of 6080 db, and 10.08/30.34 deg.C to 1190 db. An additional pressure calibration was done post-cruise at 4.07 deg.C to 6080 db. Figures 7.0 and 7.1 summarize the CTD #3 laboratory pressure calibrations performed in March and July 1996. Figure 7.0 Pressure calibration for ODF CTD #3, March 1996. Figure 7.1 Pressure calibration for ODF CTD #3, July 1996. Additionally, dynamic thermal-response step tests were conducted on the pressure transducer to calibrate dynamic thermal effects. These results were combined with the static temperature calibrations to optimally correct the CTD pressure. CTD PRT temperatures were calibrated to an NBIS ATB-1250 resistance bridge and Rosemount standard PRT in a temperature-controlled bath. The primary and secondary CTD temperatures were offset by ~1.5 and ~2 deg.C to avoid the 0-point discontinuity inherent in the internal digitizing circuitry. Standard and CTD temperatures were measured pre-cruise for the primary PRT at 7 different bath temperatures between -1.9 and 10.1 deg.C. The primary and secondary PRT #FSI-1320T were both calibrated post-cruise at more than a dozen bath temperatures between -1.9 and 30.3 deg.C. Figures 7.2 and 7.3 summarize the laboratory calibrations performed on the CTD #3 primary PRT during March and July 1996. Figure 7.4 shows the laboratory calibration performed on the CTD #3 secondary PRT (FSI-1320T only) during July 1996. Figure 7.2 Primary PRT Temperature Calibration for ODF CTD #3, March 1996. Figure 7.3 Primary PRT Temperature Calibration for ODF CTD #3, July 1996. Figure 7.4 Secondary PRT (FSI-1320T) Temperature Calibration for ODF CTD #3, July 1996. These laboratory temperature calibrations were referenced to an ITS-90 standard. Temperatures were converted to the IPTS-68 standard during processing in order to calculate other parameters, including salinity and density, which are currently defined in terms of that standard only. Final calibrated CTD temperatures are reported using the ITS-90 standard. 8. CTD Calibration Procedures ODF CTD #3 had recently been acquired by ODF and did not have an extensive calibration history. A redundant PRT sensor was used as a temperature calibration check while at sea. CTD conductivity and dissolved O2 were calibrated to in situ check samples collected during each rosette cast. Final pressure, temperature, conductivity and oxygen corrections were determined during post-cruise processing. 8.1. CTD #3 Pressure There was a pre- to post-cruise shift in the loading curves (increasing pressure) of less than -0.5 db in the top 2000 db, gradually shifting to a maximum +0.5 db at the maximum pressure in the cold-bath laboratory calibrations for pressure. The unloading curves were similar in the top 1000 db, and shifted a fairly consistent +0.5 db in the post-cruise. The intermediate-temperature (10/4 deg.C pre-/post-cruise) pressure calibrations were less easily compared, since they differed by 6 deg.C and were done to different maximum pressures. For easier comparison, the deep extrapolation of the pre-cruise 10 deg.C calibration was used. The loading curves were within +/-0.5 db of each other in the top 3500 db, with the post-cruise shifting by a maximum -0.5 db at 6080 db. The unloading curves crossed around 4500 db, with the post-cruise calibration showing a maximum +0.8 db at 1100 db, then closing in again to within +/-0.2 db near the surface. The 4 deg.C calibration (post-cruise) would typically be twice as close as the 10 deg.C calibration (pre-cruise) to the -1 deg.C calibrations, if there were no shift in CTD pressure. However, the difference between the cold and intermediate calibrations at maximum pressure became twice as large instead (0.6 db in 12 deg.C pre-cruise vs 1.3 db in 5 deg.C post- cruise). The differences between the calibrations were still less than 1 db at any calibration temperature or pressure, a relatively insignificant amount. A test comparing the results of using one calibration or the other showed less than +/-0.3 db differences in maximum pressures for each cast deeper than 1500 db, and 0.3 to 0.9 db differences in casts shallower than 1500 db. The pre-cruise calibration data, plus the dynamic thermal- response correction, were applied to S4I CTD #3 pressure data to generate final pressures. Down-cast surface pressures were automatically adjusted to 0 db as the CTD entered the water; any difference between this value and the calibration value was automatically adjusted during the top 50 decibars. Residual pressure offsets at the end of each up-cast (the difference between the last corrected pressure in-water and 0 db) averaged 0.9 db, indicating no significant problems with the final pressure corrections. The entire pre- to post-cruise laboratory calibration shift for the pressure sensor on CTD #3 was less than one-half the magnitude of the WOCE accuracy specification of 3 db. Final adjusted S4I CTD pressures should be well within the desired standards. An FSI-OPM/pressure module (1326P) was substituted for the secondary PRT during stations 103 through 106 as a test of the OPM. These secondary pressure data were neither processed nor calibrated. 8.2. CTD #3 Temperature Three different FSI-OTM/PRT sensors (S/N 1320T, 1322T, 1321T) were deployed as a second temperature channel (PRT2) and compared with the primary PRTchannel (PRT1) on all casts except stations 103-106 to monitor for drift. The response times of the primary and secondary PRT sensors were matched, then preliminary corrected temperatures were compared for a series of standard depths from each CTD down-cast. OTM-1320T was used for stations 1-61 and 107-108, OTM-1322T was used for stations 62-100, and OTM-1321T was used for stations 101-102 only. Since no OTM was attached during the pre-cruise calibration, a simple offset of -2.0 was used to correct PRT2 for comparison to PRT1 data, a correction within 0.0025 deg.C of calibration checks of all 3 OTMs in November 1996. The differences between the CTD #3 primary PRT and all 3 OTM sensors remained a fairly stable +/-0.0005 deg.C for pressures deeper than 1500 db. A stable conductivity correction also indicated no shift in the primary PRT. Figure 8.2.0 summarizes the comparison between the primary and secondary PRT temperatures. Figure 8.2.0 S4I comparison of CTD #3 primary vs. secondary PRT temperatures, pressure > 1500 db (no Sta.031). The primary temperature sensor laboratory calibrations indicated a -0.0015 deg.C shift at -1.5 to 6 deg.C, with no slope change, from pre- to post- cruise. Figure 8.2.1 shows the pre-/post-cruise PRT1 calibrations plotted together, using only uncorrected PRT1 values above 0 deg.C. Figure 8.2.1 WOCE96-S4I Primary temperature (PRT1) correction for ODF CTD #3, March + July'96 calibs, rawPRT1 > 0 deg.C only. The post-cruise PRT1 calibration measured more temperature points and was more consistent, so it was offset by +0.00075 deg.C (half of the pre- to post-cruise change) and applied to S4I temperature data. Figure 8.2.2 shows the offset post-cruise temperature calibration used to correct CTD #3 PRT1 data. Figure 8.2.2 WOCE96-S4I Primary temperature (PRT1) correction for ODF CTD #3, July'96 calib. +0.00075 deg.C. Two casts (stations 30 and 31) had problems with PRT1 readings, caused by a flooded sensor turret; the problem was repaired before station 32. It was necessary to use PRT2 for the primary temperature data on these two casts, despite the expected noisier salinity caused by the distance between PRT2 and the conductivity sensor. The post-cruise secondary temperature sensor laboratory calibration showed a fairly constant -1.9997 deg.C offset between -1.1 and 9 deg.C, covering the full range of temperatures seen on these two casts. This offset was applied to correct the PRT2 temperature data for stations 30 and 31. Figure 8.2.3 shows the post-cruise temperature calibration data used to correct CTD #3 PRT2 data. Figure 8.2.3 WOCE96-S4I Secondary temperature (PRT2) correction for ODF CTD #3, July'96 calib., rawPRT2 from 0.5 to 10.5 deg.C only. The pre- to post-cruise laboratory calibration shift for the primary temperature sensor on CTD #3 was less than the magnitude of the WOCE accuracy standard of 0.002 deg.C for the temperature range of the S4I line. Since the difference between the two calibrations was essentially split and applied to the data, S4I CTD temperatures should be within the WOCE accuracy specifications. PRT2 data compared well to PRT1 data throughout the cruise, and should also be within the same accuracy range as PRT1. The exception to these accuracy figures would be where uncorrected CTD temperatures cross between positive and negative values: the discontinuity described in the "CTD Data Processing" section may offset colder data. This error may be as much as +0.0025 deg.C for corrected CTD temperatures below ~-1.49 deg.C, an amount apparent in the figures for PRT1 Temperature Calibrations seen in the previous section. Fortunately, all such temperatures on S4I are shallower than 480 db and fall in areas where the temperature gradient is larger than the error, so it is not readily detectable. 8.3. CTD #3 Conductivity The corrected CTD rosette trip pressure and temperature were used with the bottle salinity to calculate a bottle conductivity. Differences between the bottle and CTD conductivities were then used to derive a conductivity correction. This correction is normally linear for the 3-cm conductivity cell used in the Mark III CTD, but CTD #3 sensors required pressure- dependent conductivity corrections as well. Three different CTD conductivity sensors were used during S4I; all three sensors were essentially new at the start of S4I. o #E55 was used on stations 1-31. It was replaced because the sensor turret leaked during stations 30-31. o #P42 was used on stations 32-38. It was replaced because of nonlinear sensitivity and lack of stability. o #E55 was again used on stations 39-61. This sensor became extremely noisy during stations 56-58. The sensor was cleaned with RBS prior to station 59, which caused a shift in the offset while significantly reducing the noise level. The sensor was replaced because of nonlinear sensitivity and lack of stability. o #O17 was used on stations 62-108. It was fairly stable, with a small shift after the 6.5-day break in station work to deliver supplies to Mirnyy. Conductivity differences above and below the thermocline were fit to CTD conductivity for each conductivity sensor to determine conductivity slopes. Stations 1-31, 39-55 and 56-61 were treated separately for sensor #E55, and stations 62-72 and 73-108 were grouped separately for sensor #O17. Figures 8.3.0.0-8.3.0.5 show the data used to determine preliminary conductivity slopes. Figure 8.3.0.0 CTD #3 prelim. conductivity slopes for WOCE96-S4I, stations 1-31 (C-sensor #E55). Figure 8.3.0.1 CTD #3 prelim. conductivity slopes for WOCE96-S4I, stations 32-38 (C-sensor #P42). Figure 8.3.0.2 CTD #3 prelim. conductivity slopes for WOCE96-S4I, stations 39-55 (C-sensor #E55). Figure 8.3.0.3 CTD #3 prelim. conductivity slopes for WOCE96-S4I, stations 56-61 (C-sensor #E55). Figure 8.3.0.4 CTD #3 prelim. conductivity slopes for WOCE96-S4I, stations 62-72 (C-sensor #O17). Figure 8.3.0.5 CTD #3 prelim. conductivity slopes for WOCE96-S4I, stations 73-108 (C-sensor #O17). These preliminary conductivity differences were fit to conductivity, with outlying values (4,2 standard deviations) rejected. Shallower stations were omitted from all groups; only stations 56-58 were used to determine slopes for stations 56-61 because of the offset caused by cleaning the sensor prior to station 59. Conductivity slopes were calculated from the first-order fits. The slopes calculated for stations 1-31 and 39-55 were averaged, as were the slopes for stations 62-72 and 73-108. Preliminary slopes were then applied to each S4I cast. Once the conductivity slopes were applied, residual CTD conductivity offset values were calculated for each cast using bottle conductivities deeper than 1400 db for stations 1-31, 39-55 and 62-108. More restricted pressure ranges were used to determine preliminary offsets for casts with unstable conductivity sensors, while pressure-dependent conductivity corrections were pending: only 0-70 db differences were used for stations 32-38, and 2300-2800 db for stations 56-61. Figure 8.3.1 illustrates the S4I preliminary conductivity offset residual values. Figure 8.3.1 S4I CTD #3 preliminary conductivity offsets by station number. Casts were grouped together based on drift and/or known CTD conductivity shifts or problems to determine average offsets. This also smoothed the effect of any cast-to-cast bottle salinity variation, typically on the order of +/-0.001 PSU. Some casts were omitted from the fits because there were no bottle differences within the specified pressure ranges used, or because of known CTD shifts relative to nearby casts. Smoothed offsets were applied to each cast except stations 32-38 and 56-61, which had individual offsets applied because of sensor instabilities. Some offsets were then manually adjusted to account for discontinuous shifts in the conductivity transducer response or bottle salinities, or to maintain deep theta-salinity consistency from cast to cast. After applying preliminary conductivity slopes and offsets to each cast, residual CTD conductivity differences above and below the thermocline were fit to CTD pressure for each sensor. Stations 1-31 + 39-55 conductivity differences varied +/-0.002 mS/cm and warranted a second-order correction as a function of pressure. Stations 62-108 needed a linear correction as a function of pressure to pull in the 0.001 mS/cm differences at intermediate pressures. Stations 32-38 and 56-61 required individual second-order corrections (linear for shallow station 35) as a function of pressure to pull in much larger residual differences. Figures 8.3.2.0-8.3.2.3 show the residual conductivity differences used for determining these corrections. Figure 8.3.2.0 CTD #3 residual conductivity vs. pressure for WOCE96-S4I, stas 1-31 + 39-55 (C-sensor #E55). Figure 8.3.2.1 CTD #3 residual conductivity vs. pressure for WOCE96-S4I, stas 32-38 (C-sensor #P42). Figure 8.3.2.2 CTD #3 residual conductivity vs. pressure for WOCE96-S4I, stas 56-61 (C-sensor #E55). Figure 8.3.2.3 CTD #3 residual conductivity vs. pressure for WOCE96-S4I, stas 62-108 (C-sensor #O17). After applying the pressure-dependent corrections to conductivity, conductivity slopes were re-examined for any leftover dependence on conductivity. Two groups needed minor adjustments to conductivity slopes as a function of conductivity. Figures 8.3.3.0 and 8.3.3.1 show the residual corrections calculated for stations 55-61 and stations 62-108. Figure 8.3.3.0 CTD #3 adjustments to conductivity slopes for WOCE96-S4I, stas 55-61 (C-sensor #E55). Figure 8.3.3.1 CTD #3 adjustments to conductivity slopes for WOCE96-S4I, stas 62-108 (C-sensor #O17). The final S4I pressure-dependent coefficients and conductivity-dependent slopes are summarized in Figures 8.3.4 and 8.3.5. Figure 8.3.6 summarizes the final conductivity offsets (combined conductivity- and pressure- dependent corrections) by station number. Figure 8.3.4 S4I CTD #3 pressure-dependent correction coefficients by station number. Figure 8.3.5 S4I CTD #3 conductivity-dependent slope corrections by station number. Figure 8.3.6 S4I CTD #3 combined conductivity offsets by station number. S4I temperature and conductivity correction coefficients are also tabulated in Appendix A. Summary of Residual Salinity Differences Figures 8.3.7, 8.3.8, 8.3.9 and 8.3.10 summarize the S4I residual differences between bottle and CTD salinities after applying the conductivity corrections. Only CTD and bottle salinities with final quality code 2 (acceptable) were used to generate these figures and statistics. Residual differences exceeding +/-0.025 PSU are included in the calculations for averages and standard deviations, even though they arenot plotted. Figure 8.3.7 S4I Salinity residual differences vs pressure (after correction). Figure 8.3.8 S4I Salinity residual differences vs station # (after correction). Figure 8.3.9 S4I Deep salinity residual differences vs station # (after correction). The CTD conductivity calibration represents a best estimate of the conductivity field throughout the water column. 3-sigma from the mean residual in Figures 8.3.8 and 8.3.9, or +/-0.0059 PSU for all salinities and +/-0.0015 PSU for deep salinities, represents the limit of repeatability of the bottle salinities (Autosal, rosette, operators and samplers). This limit agrees with station overlays of deep theta-salinity. Within most casts (a single salinometer run), the precision of bottle salinities and CTD salinities appears to be better than 0.001 PSU. Final calibrated CTD data from WOCE96-S4I and various cruises were compared at their closest stations. Non-S4I WOCE data were extracted from http://whpo.ucsd.edu in March 2003. A table of the comparisons follows: Table 8.3.10 S4I Compared To Historical Data S4I |Crs.ID/ |Crs. |IAPSO SSW |Distance |Avg. Salinity Diffc. (Crs-S4I) Sta.No. |Sta.No. |Date |Batch No. |Apart(nm) |(PSU at Deepest 1 deg.C Theta) --------|----------------|-------|----------|----------|------------------------------------- 1 |WOCE-S4A/20 |Mar.96 |P-127 |68 |0 to +0.001 (vs. S4A btls - |(06AQANTXIII_4) | | | |CTD salinity quality-coded 4) --------|----------------|-------|----------|----------|------------------------------------- 63,64 |WOCE-I8S/76 |Dec.94 |P-124 |13,27 |+0.001 92 |WOCE-I9S/92 |Jan.95 |P-124 |9 |+0.001 |(316N145_5) | | | | --------|----------------|-------|----------|----------|------------------------------------- 105 |WOCE-S3+S4/17 |Jan.95 |P-123 |9 |+0.003 to +0.0035 106 |WOCE-S3+S4/18 |Jan.95 |P-123 |12 |+0.0005 to +0.002 107 |WOCE-S3+S4/3-4 |Dec.94 |P-123 |14 |+0.006 to +0.007 * 108 |WOCE-S3+S4/2 |Dec.94 |P-123 |0.5 |+0.001 to +0.002 108 |WOCE-S3+S4/19 |Jan.95 |P-123 |0.1 | 0 (0.5+ deg.C Theta) to +0.004 (deep) |(09AR9404_1) | | | | --------|----------------|-------|----------|----------|------------------------------------- 11 |WOCE-S4/12345 |Feb.93 |P-120 |46 |-0.002 (below -0.04 deg.C Theta) | | | | |-0.0005 (-0.04 to 0.45 deg.C Theta) 86 |WOCE-S4/12351 |Feb.93 |P-120 |21 |-0.004 (S4I 300m shallower than S4) |(74DI200_1) | | | | --------|----------------|-------|----------|----------|------------------------------------- 48,49 |GEOSECS/430 |Feb.78 |P-61 |201,188 |+0.004/+0.003 85 |GEOSECS/431 |Feb.78 |P-61 |76 |+/-0.002 above 0.2 deg.C Theta | | | | |(incomparable deeper) 88,89 |GEOSECS/430 |Feb.78 |P-61 |180,178 |+0.002 * these S3+S4 casts were +0.003 to +0.0045 PSU compared to nearby casts on the same cruise IAPSO Standard Seawater batch corrections are similar for S4I (P-125) and most of the standards used for the other cruises listed in the chart: at most, -0.0004 PSU in salinity. The P-123/P-125 batch difference may account for up to a +0.001 PSU difference between S3+S4/S4I salinity data [Culk98]. S4A stations 3-4 are probably not good for comparison, since they are offset from nearby casts on the same cruise. S4I stations 105-108 all agree within +/-0.0005 PSU. IAPSO batch corrections would bring the GEOSECS data about 0.001 PSU closer to S4I [Mant87] [Culk98]. 8.4. CTD Dissolved Oxygen SensorMedics oxygen sensors have a finite shelf life, so new sensors are usually employed at the start of a cruise. A single, new O2 sensor was used throughout S4I. The pressure-related response problems observed during WOCE95-I10 were not apparent during this leg. The oxygen sensor from this cruise was used again 8 months later, during WOCE97-ICM3 at 20S. The extremely cold temperatures during S4I apparently caused problems with the CTD O2 fits, since no fitting problems occurred for this same sensor on ICM3. Either the surface mixed layer fit the bottle oxygen data, causing arelatively shapeless deeper fit; or the deeper data fit the bottle-defined structure well at the expense of surface fits. Since freezing problems at the surface were observed with temperature and conductivity sensors, it is likely that the oxygen sensor was also affected. Most surface oxygen fits were sacrificed in order to define sub-thermocline CTD O2 structure; these poorly fit areas are documented in Appendix C, and the data are quality- coded 3 or 4. There are a number of problems with the response characteristics of the SensorMedics O2 sensor used in the NBIS Mark III CTD, the major ones being a secondary thermal response and a sensitivity to profiling velocity. Stopping the rosette for as little as half a minute, or slowing down for a bottom approach, can cause shifts in the CTD O2 profile as oxygen becomes depleted in water near the sensor. Such shifts could usually be corrected by offsetting the raw oxygen data from the stop or slow-down area until some time after the sensor has been moving again, occasionally until the bottom of the cast. All offset sections, winch stops or slow-downs that affected CTD oxygen data are documented in Appendix C. Because of these same stop/slow-down problems, up-cast CTD O2 data cannot be optimally calibrated to O2 check samples. Instead, down-cast CTD O2 data are derived by matching the up-cast rosette trips along isopycnal surfaces. When down-casts were deemed to be unusable (see Appendix C), up- cast CTD O2 data were processed despite the signal drop-offs typically seen at bottle stops. The differences between CTD O2 data modeled from these derived values and check samples are then minimized using a non-linear least-squares fitting procedure. Figures 8.4.0 and 8.4.1 show the residual differences between the corrected CTD O2 and the bottle O2 (ml/l) for each station. Only CTD and bottle oxygens with final quality code 2 (acceptable) were used to generate these figures and statistics. Residual differences exceeding +/-0.5 ml/l are included in the calculations for averages and standard deviations, even though they are not plotted. Figure 8.4.0 S4I O2 residual differences vs station # (after correction). Figure 8.4.1 S4I Deep O2 residual differences vs station # (after correction). The standard deviations of 0.044 ml/l for all oxygens and 0.015 ml/l for deep oxygens are only intended as indicators of how well the up-cast bottle and pressure-series (mostly down-cast) CTD O2 values match up. ODF makes no claims regarding the precision or accuracy of CTD dissolved O2 data. The general form of the ODF O2 conversion equation follows Brown and Morrison [Brow78] and Millard [Mill82], [Owen85]. ODF does not use a digitized O2 sensor temperature to model the secondary thermal response but instead models membrane and sensor temperatures by low-pass filtering the PRT temperature. In situ pressure and temperature are filtered to match the sensor response. Time-constants for the pressure response Taup, and two temperature responses TauTs and TauTf are fitting parameters. The Oc gradient, dOc/dt, is approximated by low-pass filtering 1st-order Oc differences. This gradient term attempts to correct for reduction of species other than O2 at the cathode. The time-constant for this filter, Tauog, is a fitting parameter. Oxygen partial-pressure is then calculated: Opp=[c1*Oc+c2]*fsat(S,T,P)*e**(c3*Pl+c4*Tf+c5*Ts+c6*dOc/dt) (8.4.0) where: Opp = Dissolved O2 partial-pressure in atmospheres (atm); Oc = Sensor current (uamps); fsat(S,T,P) = O2 saturation partial-pressure at S,T,P (atm); S = Salinity at O2 response-time (PSUs); T = Temperature at O2 response-time (deg.C); P = Pressure at O2 response-time (decibars); Pl = Low-pass filtered pressure (decibars); Tf = Fast low-pass filtered temperature (deg.C); Ts = Slow low-pass filtered temperature (deg.C); dOc/dt = Sensor current gradient (uamps/secs). S4I CTD O2 correction coefficients (c1 through c6) are tabulated in Appendix B. 9. Bottle Sampling At the end of each rosette deployment, water samples were drawn from the bottles in the following order: o CFCs; o 3He; o O2; o PCO2; o Total CO2; o AMS 14C; o Nutrients; o Salinity; o 18O/16O; o Tritium; o Alkalinity. Since some properties were not sampled on every cast, the actual sample- drawing sequence was modified as necessary. The correspondence between individual sample containers and the rosette bottle 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 to their respective laboratories for analysis. Oxygen, nutrients and salinity analyses were performed on computer-assisted (PC) analytical equipment networked to Sun SPARCstations for centralized data analysis. The analysts for each specific property were responsible for insuring that their results were updated into the cruise database. 10. Bottle Data Processing Bottle data processing began with sample drawing, and continued until the data were considered to be final. One of the most important pieces of information, the sample log sheet, was filled out during the drawing of the many different samples. It was useful both as a sample inventory and as a guide for the technicians in carrying out their analyses. Any problems observed with the rosette before or during the sample drawing were noted on this form, including indications of bottle leaks, out-of-order drawing, etc. Oxygen draw temperatures recorded on this form were at times the first indicator of rosette bottle-tripping problems. Additional clues regarding bottle tripping or leak problems were found by individual analysts as the samples were analyzed and the resulting data were processed and checked. The next stage of processing was accomplished after the individual parameter files were merged into a common station file, along with CTD- derived parameters (pressure, temperature, conductivity, etc.). The rosette cast and bottle numbers were the primary identification for all ODF-analyzed samples taken from the bottle, and were used to merge the analytical results with the CTD data associated with the bottle. At this stage, bottle tripping problems were usually resolved, sometimes resulting in changes to the pressure, temperature and other CTD properties associated with the bottle. All CTD information from each bottle trip (confirmed or not) was retained in a file, so resolving bottle tripping problems consisted of correlating CTD trip data with the rosette bottles. Diagnostic comments from the sample log, and notes from analysts and/or bottle data processors were entered into a computer file associated with each station (the "quality" file) as part of the quality control procedure. Sample data from bottles suspected of leaking were checked to see if the properties were consistent with the profile for the cast, with adjacent stations, and, where applicable, with the CTD data. Various property- property plots and vertical sections were examined for both consistency within a cast and consistency with adjacent stations by data processors, who advised analysts of possible errors or irregularities. The analysts reviewed and sometimes revised their data as additional calibration or diagnostic results became available. Based on the outcome of investigations of the various comments in the quality files, WHP water sample codes were selected to indicate the reliability of the individual parameters affected by the comments. WHP bottle codes were assigned where evidence showed the entire bottle was affected, as in the case of a leak, or a bottle trip at other than the intended depth. WHP water bottle quality codes were assigned as defined in the WOCE Operations Manual [Joyc94] with the following additional interpretations: 2 | No problems noted. 3 | Leaking. An air leak large enough to produce an | observable effect on a sample is identified by a code of | 3 on the bottle and a code of 4 on the oxygen. (Small | air leaks may have no observable effect, or may only | affect gas samples.) 4 | Did not trip correctly. Bottles tripped at other than | the intended depth were assigned a code of 4. There may | be no problems with the associated water sample data. 5 | Not reported. No water sample data reported. This is a | representative level derived from the CTD data for | reporting purposes. The sample number should be in the | range of 80-99. 9 | The samples were not drawn from this bottle. WHP water sample quality flags were assigned using the following criteria: 1 | The sample for this measurement was drawn from the water | bottle, but the results of the analysis were not (yet) | received. 2 | Acceptable measurement. 3 | Questionable measurement. The data did not fit the | station profile or adjacent station comparisons (or | possibly CTD data comparisons). No notes from the | analyst indicated a problem. The data could be | acceptable, but are open to interpretation. 4 | Bad measurement. The data did not fit the station | profile, adjacent stations or CTD data. There were | analytical notes indicating a problem, but data values | were reported. Sampling and analytical errors were also | coded as 4. 5 | Not reported. There should always be a reason | associated with a code of 5, usually that the sample was | lost, contaminated or rendered unusable. 9 | The sample for this measurement was not drawn. WHP water sample quality flags were assigned to the CTDSAL (CTD salinity) parameter as follows: 2 | Acceptable measurement. 3 | Questionable measurement. The data did not fit the | bottle data, or there was a CTD conductivity calibration | shift during the up-cast. 4 | Bad measurement. The CTD up-cast data were determined | to be unusable for calculating a salinity. 7 | Despiked. The CTD data have been filtered to eliminate | a spike or offset. WHP water sample quality flags were assigned to the CTDO (CTD O2) parameter as follows: 1 | Not calibrated. Data are uncalibrated. 2 | Acceptable measurement. 3 | Questionable measurement. 4 | Bad measurement. The CTD data were determined to be | unusable for calculating a dissolved oxygen | concentration. 5 | Not reported. The CTD data could not be reported, | typically when CTD salinity is coded 3 or 4. 7 | Despiked. The CTD data have been filtered to eliminate | a spike or offset. 9 | Not sampled. No operational CTD O2 sensor was present | on this cast. Note that CTDO values were derived from the down-cast pressure-series CTD data, except for 18 stations where up-casts were processed because of conductivity problems on the down-casts. CTD data were matched to the up- cast bottle data along isopycnal surfaces. If the CTD salinity is footnoted as bad or questionable, the CTD O2 is not reported. Table 10.0 shows the number of samples drawn and the number of times each WHP sample quality flag was assigned for each basic hydrographic property: Table 10.0 Frequency of WHP quality flag assignments for S4I. Rosette Samples Stations 001-108 ---------------------------------------------------------------------------- Reported WHP Quality Codes Levels 1 2 3 4 5 7 9 ----------||---------|------------------------------------------------------ Bottle || 3655 | 0 3542 1 108 0 0 4 CTD Salt || 3655 | 0 3493 0 36 0 126 0 CTD Oxy || 3619 | 0 2909 111 599 36 0 0 Salinity || 3604 | 0 3521 24 59 9 0 42 Oxygen || 3630 | 0 3568 33 29 7 0 18 Silicate || 3640 | 0 3638 1 1 0 0 15 Nitrate || 3640 | 0 3639 0 1 0 0 15 Nitrite || 3640 | 0 3639 0 1 0 0 15 Phosphate || 3640 | 0 3602 3 35 0 0 15 Additionally, all WHP water bottle/sample quality code comments are presented in Appendix D. 11. Pressure and Temperatures All pressures and temperatures for the bottle data tabulations on the rosette casts were obtained by averaging CTD data for a brief interval at the time the bottle was closed on the rosette, then correcting the data based on CTD laboratory calibrations. The temperatures are reported using the International Temperature Scale of 1990. 12. Salinity Analysis Equipment and Techniques Two Guildline Autosal Model 8400A salinometers were available for measuring salinities. The salinometers were modified by ODF and contained interfaces for computer-aided measurement. Autosal #57-396 was a backup unit but was not used on this expedition. Autosal #55-654 was used to measure salinity on all stations. Its water bath temperature was set and maintained at 24 deg.C for all runs except stations 32-39, where the bath temperature was set at 21 deg.C. The salinity analyses were performed when samples had equilibrated to laboratory temperature, within 7-28 hours after collection. The salinometer was standardized for each group of analyses (typically one cast, usually 36 samples) using two fresh vials of standard seawater per group. A computer (PC) prompted the analyst for control functions such as changing sample, flushing, or switching to "read" mode. At the correct time, the computer acquired conductivity ratio measurements, and logged results. The sample conductivity was redetermined until readings met software criteria for consistency. Measurements were then averaged for a final result. Unstable readings were encountered during analysis of the first 5 samples from station 42. The Autosal flow cell was cleaned, and sample analysis was resumed ~10 hours later without further problems. Sampling and Data Processing 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 collecting each sample, inserts were inspected for proper fit and loose inserts were 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 one run at the end as an unknown was applied linearly to the data to account for any drift. The data were added to the cruise database. 3604 salinity measurements were made and 233 vials of standard water were used. The estimated accuracy of bottle salinities run at sea is usually better than 0.002 PSU relative to the particular standard seawater batch used. Laboratory Temperature The temperature stability in the salinometer laboratory was fair, ranging from 18.7 to 25.8 deg.C and drifting an average of 0.5 deg.C during a run of samples. The laboratory temperature was between -4 and +2 deg.C of the Autosal bath temperature during all sample runs. Standards IAPSO Standard Seawater (SSW) Batch P-125 was used to standardize the salinometers. 13. 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. ODF used a whole-bottle modified-Winkler titration following the technique of Carpenter [Carp65] with modifications by Culberson et al. [Culb91], but with higher concentrations of potassium iodate standard (approximately 0.012N) and thiosulfate solution (50 gm/l). Carbon disulfide was added to the thiosulfate as a preservative. Standard solutions prepared from pre- weighed potassium iodate crystals were run at the beginning of each session of analyses, which typically included from 1 to 3 stations. Nine standards were made up during the cruise and compared to assure that the results were reproducible, and to preclude the possibility of a weighing or dilution error. Reagent/distilled water blanks were determined, to account for presence of oxidizing or reducing materials. Sampling and Data Processing Samples were collected for dissolved oxygen analyses soon after the rosette sampler was brought on board, and after samples for CFCs and helium were drawn. Using a Tygon drawing tube, nominal 125ml volume-calibrated iodineflasks were rinsed twice with minimal agitation, then filled and allowed to overflow for at least 3 flask volumes. The sample draw temperature was measured with a small platinum resistance thermometer embedded in the drawing tube. Reagents were added to fix the oxygen before stoppering. The flasks were shaken twice to assure thorough dispersion of the precipitate, once immediately after drawing, and then again after about 20 minutes. The samples were analyzed within 1-9 hours of collection (18 hours for station 1 only), and then the data were merged into the cruise database. Thiosulfate normalities were calculated from each standardization and corrected to 20 deg.C. The 20 deg.C normalities and the blanks were plotted versus time and were reviewed for possible problems. New thiosulfate normalities were recalculated after the blanks had been smoothed as a function of time, if warranted. These normalities were then smoothed, and the oxygen data were recalculated. Oxygens were converted from milliliters per liter to micromoles per kilogram using the in situ temperature. Sample temperatures were measured at the time the samples were drawn from the rosette bottle. These temperatures were useful in indicating whether or not a bottle tripped properly. 3630 oxygen measurements were made, with no major problems encountered during the analyses. Volumetric Calibration Oxygen flask volumes were determined gravimetrically with degassed deionized water to determine flask volumes at ODF's chemistry laboratory. This is done once before using flasks for the first time and periodically thereafter when a suspect bottle 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 Potassium iodate standards, nominally 0.44 gram, were pre-weighed in ODF's chemistry laboratory to +/-0.0001 grams. The exact normality was calculated at sea after the volumetric flask volume and dilution temperature were known. Potassium iodate was obtained from Johnson Matthey Chemical Co. and was reported by the supplier to be >99.4% pure. All other reagents were "reagent grade" and were tested for levels of oxidizing and reducing impurities prior to use. 14. 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 a few hours after sample collection. Occasionally samples were refrigerated up to a maximum of 8 hours at 2-6 deg.C. All samples were brought to room temperature prior to analysis. The methods used are described by Gordon et al. [Gord93]. The analog outputs from each of the four 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 PO4 color development. The sample was passed through a 15mm flowcell and the absorbance 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 absorbance 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 deg.C to enhance color development, then passed through a 50mm flowcell and the absorbance measured at 820nm. Sampling and Data Processing Nutrient samples were drawn into 45 ml polypropylene, screw-capped "oak- ridge type" centrifuge tubes. The tubes were cleaned with 10% HCl and rinsed with sample twice before filling. Standardizations were performed at the beginning and end of each group of analyses (typically one cast, usually 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 6-7 different standard concentrations were analyzed periodically to determine any deviation from linearity as a function of concentration for each nutrient analysis. A correction for non-linearity was applied to the final nutrient concentrations when necessary. After each group of samples was analyzed, the raw data file was processed to produce another file of response factors, baseline values, and absorbances. Computer-produced absorbance readings were checked for accuracy against values taken from a strip chart recording. The data were then added to the cruise database. 3640 nutrient samples were analyzed. No major problems were encountered with the measurements. The pump tubing was changed four times, and deep seawater was run as a substandard check. The temperature stability of the laboratory used for the analyses was good, ranging from 20 to 24 deg.C. 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 an assumed laboratory temperature of 25 deg.C. Standards The silicate primary standard (Na2SiF6) was obtained from Aesar and was reported by the suppliers to be >98% pure. The nitrite (NaNO2) primary standard was obtained from GFS and was reported by the suppliers to be >97% pure. Primary standards for nitrate (KNO3) and phosphate (KH2PO4) were obtained from Johnson Matthey Chemical Co., and the supplier reported purities for each of 99.999%. B. Underway Measurements B1. Navigation and Bathymetry Navigation data were acquired from the ship's Ashtech GPS receiver via the network. They were logged automatically at one-minute intervals by one of the Sun Sparcstations. Underway bathymetry was logged manually from the ship's 12 kHz Raytheon/EPC PDR at five-minute intervals (or when possible in the ice), then merged with the navigation data to provide a time-series of underway position, course, speed and bathymetry data. These data were used for all station positions, PDR depths, and for bathymetry on vertical sections (Carter, 1980). Depth data on the transit from Cape Town to station 1, and from station 108 to Hobart were not logged. Data on station were not logged. B2. Meteorological Observations Five-minute average meteorological data are routinely recorded by the Palmer. Data recorded consist of time, position, air temperature, relative humidity, wet-bulb temperature, PSR, PIR, barometric pressure, and wind speed and direction. Data were recorded continuously from Cape Town to Hobart. Significant data gaps (longer than 20 minutes, but less than 32 minutes in all cases) occurred on 19 and 27 May, and 1,2,4,7,20 and 24 June. B3. Hull-mounted Acoustic Doppler Current Profiler (S. Rutz) Ocean velocity observations were taken using a hull-mounted Acoustic Doppler Current Profiler (ADCP) system and GPS navigation data. Data were recorded from May 3, 1995 to July 4, 1996 between Capetown, South Africa and Hobart, Australia, along the nominal latitude of 62ƒS from 20ƒE to 120ƒE with two transects across the Antarctic continental slope. The purpose of the observations was to document the upper ocean horizontal velocity structure along the cruise track. The observations provide absolute velocity estimates including the ageostrophic component of the flow. Fig. 2 shows the cruise track and the near-surface currents measured by the ADCP. The hull-mounted ADCP is part of the ship's equipment aboard the Palmer. The ADCP is a 150 kHz unit manufactured by RD Instruments. The instrument pings about once per second, and for most of the cruise the data were stored as 100- second averages or ensembles. The user-exit program, ue4, receives and stores the ADCP data along with both the P-code navigation data from a Trimble receiver and the positions from an Ashtech gps receiver. The ship gyro provided heading information for vector averaging the ADCP data over the 100- second ensembles. The user-exit program calculates and stores the heading offset based on the difference between the heading determination from the Ashtech receiver and from the ship gyro. The ADCP transducer is mounted in a glycol bath at a depth of about 7 meters below the sea surface. As setup parameters, a blanking interval of 16 meters, a vertical pulse length of 16 meters, a vertical bin size of 8 meters, and 60 bins were used. A 300- second sampling interval was used at the beginning of the cruise and the interval was decreased to 100-seconds shortly after entering pack ice to increase the amount of usable data (cruising through ice severely limited the percent of good return pings). 100-seconds was the sampling interval for the remainder of the cruise. Bottom tracking was activated during the shallow water transits near South Africa, Antarctica, and Tasmania. For the processing of the ADCP data aboard ship, a rotation amplitude of 0.97, a rotation angle of -1.65 degrees (added to the gyro minus gps heading), and a time filter width of one hour were used. Final editing and calibration of the ADCP data has not yet been done. For example, some spikes due to pinging off the CTD wire or rosette on station are still present in the data. A set of preliminary plots was generated during the cruise. The plots display velocity vectors averaged over several depth intervals, and over one hour in time. The velocity was measured from a depth of 23 meters to a depth of about 500 meters. During the first few weeks of the cruise, the ADCP hung a half-dozen times for unknown reasons. Several measures were taken to prevent this (e.g., the keyboard was locked) or to minimize its effect (e.g., a "watch dog" program was installed that would reboot the PC if it hung for more than about five minutes). These measures were mostly successful though the ADCP did hang one more time for unexplained reasons late in the cruise. A Trimble P-code receiver was used for navigation. The data from the receiver was stored once per second for the entire cruise. The Ashtech receiver uses a four antennae array to measure position and attitude. The heading estimate was used with the ship gyro to provide a heading correction for the ADCP ensembles. The Ashtech data was stored by the ADCP user-exit program along with the ADCP data. The Ashtech receiver at times (especially after it had been reset) could not lock onto enough satellites to determine the ship's heading. This was remedied by temporarily disabling certain satellites that were low on the horizon so that the Ashtech would not waste its time in a futile attempt to lock onto them. Also, the ship gyro input to the ADCP hung about two dozen times during the cruise for intervals ranging from several minutes to hours. The hangs were mostly due to the ship's data acquisition system (DAS) crashing. An attempt to feed the ship gyro directly to the ADCP, bypassing the DAS, was unsuccessful and had some unintended consequences (i.e., the auto-pilot went berserk). B4. Atmospheric Chemistry (D. Chipman and M. Mensch) Air samples for analysis were drawn from a single inlet located just forward of the ship's bridge through a continuous run of 3/8 inch diameter Dekoron tubing. A KNF Neuberger pump with a teflon-covered rubber diaphragm was use to pressurize the air for distribution to the CO2 analysis system in the Hydro Lab and the CFC analysis system in the Dry Lab. A vent line with a needle valve from a tee fitting at the CFC system provided backpressure for the line while allowing it to be continuously flushed with fresh atmospheric air. CO2 Analysis The LDEO underway pCO2 analysis system was used to determine the concentration of CO2 in dried atmospheric air. At intervals of approximately one half hour, air from the atmospheric sampling line was allowed to flow through a countercurrent-flow permeation gas dryer and then through the cell of the Licor infrared gas analyzer for three minutes at a flow rate of 25-35 ml/min. The sample flow was stopped for 20 seconds prior to reading the analyzer output, to allow time for the pressure to vent to the atmospheric value and for the sample to come to cell temperature. Immediately following each atmospheric sample, the instrument was calibrated using a set of four compressed air-CO2 mixtures (which have CO2 concentrations traceable to the WMO scale of C.D. Keeling); a second- order polynomial response curve was fitted to the instrumental signals given by these gases and used to calculate the concentration of CO2 in the sample. Atmospheric measurements were made whenever the pCO2 analysis system was operating, which was essentially continuously in open water and periodically (usually at stations only) when operating in the ice. Because of the problem of contamination with stack gas when the relative wind was from behind the ship, only those analyses made when the ship's meteorological monitoring system indicated relative winds from ahead were considered valid and retained. CFC Analysis Marine air samples for CFC analysis were taken from the bow air line immediately in front of the T-fitting leading to the vent line. The air was dried by flowing through magnesium perchlorate and then analyzed in exactly the same way standard gases were measured (Section C7). Marine air was analyzed whenever the necessary time was available and the relative wind direction was from the bow. The results will provide information about the current atmospheric CFC levels and will allow the calculation of the CFC saturation levels in the surface water. B5. Thermosalinograph and underway pCO2 (D. Chipman) The Palmer is fitted with two separate uncontaminated seawater lines- a 1-inch line of stainless steel and PVC, which supplies the thermosalinograph and other instruments in the Hydro Lab, and a 2-inch stainless steel line which provides water to the Aquarium Lab. Both have inlets located at a depth of approximately 6.7 meters, well aft of the bow to reduce the entrainment of air and ice during icebreaking operations. Due to a failure of the pump on the thermosalinograph line about one third of the way through the cruise, the thermosalinograph and underway pCO2 equilibrator were replumbed to be supplied water from the larger seawater line. Both lines were plagued with blockages due to ice entrainment during operations in the heavy ice, especially when snow-covered, and in general uncontaminated seawater was only available when operating in open water or unconsolidated floes, or when on station within the ice. The ship is fitted with a Seabird Model SBE-21 thermosalinograph, located in the Hydro Lab, operated and maintained by ASA personnel. The unit is provided with a remote temperature sensor located near the inlet of the smaller uncontaminated seawater line, to provide an approximate sea surface temperature. Data are logged continuously during operation by the ship's RTDAS. Due to the failure of the pump on the thermosalinograph line, the thermosalinograph received water from the other seawater line during most of the cruise, and the remote temperature was thus unavailable. A very approximate underway surface temperature during the later part of the cruise was calculated using an offset from the thermosalinograph temperature, calibrated against CTD mixed-layer temperatures during station work and against bucket thermometer temperatures during the transit from the last station to Hobart. Although the seawater line in the Hydro Lab is provided with a vortex-type debubbler, it is plumbed in parallel with the thermosalinograph, and water for the latter instrument is not routinely debubbled. Near the beginning of the cruise it became obvious that the very high noise level on the thermosalinograph salinity channel was caused by entrained air in the seawater line and the unit was replumbed to receive water from the outlet of the debubbler, which reduced the noise appreciably. Underway pCO2 Underway measurements of the surface seawater pCO2 were made using a shower-type seawater-air equilibrator similar to that originally designed by Takahashi (Broecker and Takahashi, 1966). Seawater from the same uncontaminated pumped water line which supplies the ship's thermosalinograph was used as a source for the CO2 equilibrator. The equilibrator was located downstream of a vortex debubbler to remove air entrained with the water. Air was continuously recirculated through the headspace of the equilibrator by means of a small air pump, and aliquots of this air were removed for analysis using a Licor infrared analyzer built into a fully automated analysis system. Sample gases were dried by means of a countercurrent-flow permeation gas dryer immediately prior to analysis. After eight samples of equilibrated air were analyzed, a single sample of atmospheric air pumped from a sampling point just ahead of the ship's bridge was similarly dried and analyzed. This was followed by calibration of the instrument using four air-CO2 mixtures (150 to 450 ppm range) which are traceable to the WMO calibration scale of C. D. Keeling of SIO. Barometric pressure (essentially the same as the pressure of equilibration) was measured at the time of each analysis by means of an AIR electronic barometer, and the temperature of equilibration was measured at the same time by means of a platinum resistance thermometer within the equilibrator, calibrated against a NIST-traceable mercury thermometer. The entire cycle of eight equilibrated air samples, one atmospheric air sample, and four calibration gases required approximately one half hour, and was repeated continuously. Measurements were made whenever the ship was in open water outside the territorial waters of the Republic of South Africa or Australia, and to a limited extent while operating within the ice (due to the clogging of the seawater lines during ice operations). C. Hydrographic Measurements C1. Water Sampling Package (Rosette and CTD) (F. Delahoyd) Hydrographic (rosette) casts were performed with a 36-place 10-liter rosette system consisting of a 36-bottle rosette frame (ODF), a 36-place pylon (General Oceanics 1016) and 36 10-liter PVC bottles (ODF). Underwater electronic components consisted of an ODF-modified NBIS Mark III CTD and associated sensors, SeaTech transmissometer and 12 kHz Benthos pinger. The CTD was mounted horizontally along the bottom of the rosette frame, with the transmissometer, dissolved oxygen and secondary PRT sensors deployed alongside. The Benthos pinger was monitored during a cast with a precision depth recorder (PDR) in the ship's laboratory. The rosette system was suspended from a three-conductor electro-mechanical (EM) cable. Power to the CTD and pylon was provided through the cable from the ship. Separate conductors were used for the CTD and pylon signals. CTD data were collected with a modified NBIS Mark III CTD (ODF CTD #3). This instrument provided pressure, temperature, conductivity and dissolved O2 channels, and additionally measured a second temperature (FSI temperature sensor) as a calibration check. Other data channels included elapsed-time, an altimeter, several power supply voltages and a transmissometer. The instrument supplied a standard 15-byte NBIS-format data stream at a data rate of 25 fps. Modifications to the instrument included revised pressure and dissolved O2 sensor mountings; ODF-designed sensor interfaces for O2, FSI PRT and the SeaTech transmissometer; implementation of 8-bit and 16-bit multiplexer channels; an elapsed-time channel; instrument id in the polarity byte and power supply voltages channels. The CTD pressure sensor mounting had been modified to reduce the dynamic thermal effects on pressure. The sensor was attached to a 20 cm length of coiled, oil- filled stainless-steel tubing threaded into the end-cap pressure port. The transducer was also wrapped in foam rubber. The NBIS temperature compensation circuit on the pressure interface was disabled; all thermal response characteristics were modeled and corrected in software. The O2 sensor was deployed in a pressure-compensated holder assembly mounted separately on the rosette frame and connected to the CTD by an underwater cable. The O2 sensor interface was designed and built by ODF using an off-the-shelf 12- bit A/D converter. The transmissometer interface was a similar design. Although the secondary temperature sensor was located within 1 meter of the CTD conductivity sensor, it was not sufficiently close to calculate coherent salinities. It was used as a secondary temperature calibration reference rather than as a redundant sensor, with the intent of eliminating the use of mercury or electronic DSRTs as calibration checks. Three secondary temperature sensors were used during the cruise. Standard CTD maintenance procedures included soaking the conductivity and O2 sensors in salt water between casts to maintain sensor stability and reduce the possibility of sensors freezing before entry. The General Oceanics pylon deck unit was not used. Instead, an ODF-built deck unit and external power supply were employed. The pylon emits a confirmation message containing its current notion of bottle trip position, an invaluable aid in sorting out mis-trips. The General Oceanics 1016 36-place pylon provided generally reliable operation, with mis- or non-confirmations occurring on 21 of 108 casts. Water properties from the bottles were used to positively identify the closing depth from four of these casts. The rosette system was deployed from the Palmer's Baltic room (a protected rosette room and winch shed with an external door and an extension boom) on the starboard side. Sampling took place in the deployment area after the rosette was secured. The deck watch prepared the rosette 45 minutes prior to each cast. All valves, vents and lanyards were checked for proper orientation. The bottles were cocked and all hardware and connections rechecked. Upon arrival on station, time, position and bottom depth were logged and the deployment begun. The deployment door to the Baltic room was opened after the ship had finished positioning. Positioning sometimes entailed clearing a hole in the ice. Two stabilizing tag lines were threaded through rings on the frame. CTD sensor covers were removed and the pinger turned on. Once the CTD acquisition and control system in the ship's forward laboratory had been initiated by the console operator and the CTD and pylon had passed their diagnostics, the watch leader would verify with the bridge that deployment could begin. The winch operator would raise the package and extend the boom out through the Baltic room door and over the side of the ship. The package was then quickly lowered into the water, the tag lines removed and the console and winch operators notified by radio of the target depth (wire- out). The CTD conductivity and temperature sensors were soaked in seawater between casts to minimize the problem of sensors freezing before entry. This preventive measure was not totally successful, and freezing did occur during deployment on some casts. In these cases, the rosette was lowered to 30 meters to thaw the sensors, then raised back to the surface. During each cast, the rosette was lowered to within approximately 10 meters of the bottom. Averages of CTD data corresponding to the time of bottle closure were associated with the bottle data during a cast. Pressure, depth, temperature, salinity and density were immediately available to facilitate examination and quality control of the bottle data as the sampling and laboratory analyses progressed. Recovering the package at the end of deployment was essentially the reverse of the beginning. Two tag lines terminating in large snap hooks were manipulated on poles by the deck watch to snag recovery rings on the rosette frame. The package was then lifted out of the water under tension from the tag lines, the boom retracted, and the rosette lowered onto the Baltic room deck. Sensor covers were replaced and the pinger turned off. The rosette was then secured and the Baltic room deployment door closed before allowing others in the room for sampling. A detailed examination of the bottles and rosette would occur before samples were taken, and any extraordinary situations or circumstances noted on the sample log for the cast. Seawater froze on rosette bottles several times during recovery, but quickly thawed in the rosette room and there was never any evidence of water freezing in the bottles or spigots. Rosette maintenance was performed on a regular basis. O- rings were changed as necessary and bottle maintenance performed each day to insure proper closure and sealing. Valves were inspected for leaks and repaired or replaced. The starboard side Baltic room Markey winch was used throughout the cruise. Only one seacable retermination was necessary, prior to cast 057/01. C2. CTD Measurements (F. Delahoyd) The CTD data acquisition, processing and control system consisted of a Sun SPARCstation LX computer workstation, ODF-built CTD and pylon deck units, CTD and pylon power supplies and a VCR recorder for real-time analog backup recording of the seacable signal. The Sun system consisted of a color display with trackball and keyboard (the CTD console), 18 RS-232 ports, 2.5 GB disk and 8-mm cartridge tape. Two other Sun LX systems were networked to the data acquisition system, as well as to the rest of the networked computers aboard the Palmer. These systems were available for real- time CTD data display as well as providing hydrographic data management and backup. Two HP 1200C color inkjet printers provided hardcopy from any of the workstations. The CTD FSK signal was demodulated and converted to a 9600 baud RS-232C binary data stream by the CTD deck unit. This data stream was fed to the Sun SPARCstation. The pylon deck unit was also connected to the Sun through a bi- directional 300 baud serial line, allowing rosette trips to be initiated and confirmed through the data acquisition software. A bitmapped color display provided interactive graphical display and control of the CTD rosette sampling system, including real-time raw and processed data displays, navigation, winch and rosette trip displays. The CTD data acquisition, processing and control system was prepared by the console watch a few minutes before a deployment. A console operations log was maintained for each deployment, containing a record of every attempt to trip a bottle as well as any pertinent comments. Most CTD console control functions, including starting the data acquisition, were performed by pointing and clicking a trackball cursor on the display at pictures representing functions to perform. The system would then present the operator with a short dialog prompting with automatically-generated choices that could either be accepted as default or overridden. The operator was instructed to turn on the CTD and pylon power supplies, then to examine a real-time CTD data display on the screen for stable voltages from the underwater unit. Once this was accomplished, the data acquisition and processing was begun and a time and position automatically associated with the beginning of the cast. The backup analog recording of the CTD signal on a VCR tape was started. A rosette trip display and pylon control window popped up, giving visual confirmation that the pylon was initializing properly. Various plots and displays were initiated. When all was ready, the console operator informed the deck watch by radio. Once the deck watch had deployed the rosette it was immediately lowered without pausing at the sea surface. If the console operator noticed that sensors were frozen on entry, the package would be stopped at 30 meters, then raised to just below the surface after the sensors had thawed. The deck watch leader provided the winch operator with a target depth (wire-out) and lowering rate (normally 60-70 meters/minute for this package). The console operator would examine the processed CTD data during descent via interactive plot windows on the display, which could also be initiated from other workstations on the network. Additionally, the operator would decide where to trip bottles on the up-cast, noting this on the console log. The PDR was monitored to insure the bottom depth was known at all times. The rosette distance above the bottom was monitored by the deck watch leader using the distance between the rosette pinger signal and its bottom reflection displayed on the PDR. The winch and PDR displays allowed the watch leader to refine the target depth relayed to the winch operator and safely approach to about 10 meters above the bottom. Bottles would be closed on the up cast by pointing the console trackball cursor at a graphic firing control and clicking a button. The data acquisition system would respond with the CTD rosette trip data and a pylon confirmation message in a window. All tripping attempts were noted on the console log. The console operator would then direct the winch operator to the next bottle stop. The console operator was also responsible for generating the sample log for the cast. After the last bottle was tripped, the console operator would inform the deck watch and the rosette would be brought on deck. It was sometimes necessary to close the surface bottles on-the-fly due to a risk of slack wire at higher sea states. Once the rosette was on deck, the console operator would terminate the data acquisition and turn off the CTD, pylon and VCR recording. The VCR tape was filed. Frequently the console operator would also bring the sample log to the rosette room and serve as the sample cop. CTD Data Processing ODF CTD processing software consists of approximately 35 programs running under the Unix operating system. The initial CTD processing program (ctdba) is used either in realtime or with existing raw data sets to: o Convert raw CTD scans into scaled engineering units, and assign the data to logical channels; o Filter data channels according to specified filtering criteria; o Apply sensor or instrument-specific response-correction models; o Provide periodic averages of the channels corresponding to the output time-series interval; and o Store the output time-series in a CTD-independent format. Once the CTD data are reduced to a standard-format time-series, they can be manipulated in a number of various ways. Channels can be additionally filtered. The time-series can be split into shorter time-series or pasted to form longer time-series. A time-series can be transformed into a pressure-series, or a different interval time-series. Calibration corrections to the series are maintained in separate files and are applied whenever the data are accessed. ODF data acquisition software acquired and processed the CTD data in real-time, providing calibrated, processed data for interactive plotting and reporting during a cast. The 25 Hz data from the CTD were filtered, response-corrected and averaged to a 2 Hz time-series. Sensor correction and calibration models were applied to pressure, temperature, conductivity and O2. Rosette trip data were extracted from this time-series in response to trip initiation and confirmation signals. The calibrated 2 Hz time-series data were stored on disk (as were the 25 Hz raw data) and were available in real-time for reporting and graphical display. At the end of the cast, various consistency and calibration checks were performed, and a 2.0 db pressure-series of the down-cast was generated and subsequently used for reports and plots. CTD plots generated automatically at the completion of deployment were checked daily for potential problems. The two PRT temperature sensors were compared for sensor drift. The CTD conductivity sensor was monitored by comparing CTD values to check-sample conductivities and by deep TS comparisons with adjacent stations. The CTD dissolved O2 sensor was calibrated to check-sample data. Note that all CTDOXY values were derived from the down cast data, matched to the upcast along isopycnal surfaces. WHP water sample quality flags were assigned to the CTDSAL (CTD salinity) parameter as follows: 2 Acceptable measurement. 3 Questionable measurement. The data did not fit the bottle data, or there was a CTD conductivity calibration shift during the cast. 4 Bad measurement. The CTD data were determined to be unusable for calculating a salinity. 8 The CTD salinity was derived from the CTD down cast, matched on an isopycnal surface. WHP water sample quality flags were assigned to the CTDOXY (CTD oxygen) parameter as follows: 2 Acceptable measurement. 4 Bad measurement. The CTD data were determined to be unusable for calculating a dissolved oxygen concentration. 5 Not reported. The CTD data could not be reported. 9 Not sampled. No operational dissolved oxygen sensor was present on this cast. C3. Bottle Measurements (F. Delahoyd) At the end of each rosette deployment water samples were drawn from the bottles in the following order: 1: CFCs 4: PCO2 7: Nutrients 10: Tritium 2: 3He 5: Total CO2 8: Salinity 11: Alkalinity 3: O2 6: AMS 14C 9: 18O/16O Note that some properties were subsampled by cast or by station, so the actual sequence of samples drawn was modified accordingly. The correspondence between individual sample containers and the rosette bottle 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 proper drawing order. Normal sampling practice included opening the drain valve before opening 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 to their laboratory for analysis. Oxygen, nutrients and salinity analyses were performed on computer-assisted (PC) analytical equipment networked to Sun SPARCStations for centralized data analysis. The analyst for a specific property was responsible for insuring that their results updated the cruise database Bottle Data Processing The first stage of bottle data processing consisted of verifying and validating individual samples, and checking the sample log (the sample inventory) for consistency. Oxygen flask numbers were verified, as each flask is individually calibrated and significantly affects the calculated O2 concentration. At this stage, bottle tripping problems were usually resolved, sometimes resulting in changes to the pressure, temperature and other CTD data associated with the bottle. The rosette bottle number was the primary identification for all samples taken from the bottle, as well as for the CTD data associated with the bottle. All CTD trips were retained whether confirmed or not so that they could be used to help resolve bottle tripping problems. Diagnostic comments from the sample log were then translated into preliminary WOCE quality codes, together with appropriate comments. Each code indicating a potential problem would be investigated. The next stage of processing would begin after all the samples for a cast had been accounted for. All samples for bottles suspected of leaking were checked to see if the measurements were consistent with the profile for the cast, with adjacent stations and where applicable, with the CTD data. All comments from the analysts were examined and turned into appropriate water sample codes. The third stage of processing will continue until the data are judged "final". Various property-property plots and vertical sections were examined for both consistency within a cast and consistency with adjacent stations. In conjunction with this process the analysts would review (and sometimes revise) their data as additional calibration or diagnostic results became available. Final assignment of a WHP water sample quality code to an anomalous sample value was typically achieved through consensus. WHP water bottle quality flags were assigned with the following additional interpretations: 3 An air leak large enough to produce an observable effect on a sample is identified by a code of 3 on the bottle and a code of 4 on the oxygen. (Small air leaks may have no observable effect, or may only affect gas samples.) 4 Bottles tripped at other than the intended depth were assigned a code of 4. There may be no problems with the associated water sample data. WHP water sample quality flags were assigned using the following criteria: 1 The sample for this measurement was drawn from a bottle, but the results of the analysis were not (yet) received. 2 Acceptable measurement. 3 Questionable measurement. The data did not fit the station profile or adjacent station comparisons (or possibly CTD data comparisons). No notes from the analyst indicated a problem. The data could be correct, but are open to interpretation. 4 Bad measurement. Does not fit the station profile, adjacent stations or CTD data. There were analytical notes indicating a problem, but data values were reported. Sampling and analytical errors were also coded as 4. 5 Not reported. There should always be a reason associated with a code of 5, usually that the sample was lost, contaminated or rendered unusable. 9 The sample for this measurement was not drawn. Table 5 shows the number of samples drawn and the number of times each WHP sample quality flag was assigned for each basic hydrographic property. C4. Salinity Analysis (F. Delahoyd) Salinity samples were drawn into 200 ml Kimax high alumina borosilicate bottles after 3 rinses, and were sealed with custom-made plastic insert thimbles and Nalgene screw caps. This assembly provides very low container dissolution and sample evaporation. When loose inserts were found, they were replaced to ensure an airtight seal. Salinity was determined after a box of samples had equilibrated to laboratory temperature, usually within 8-12 hours of collection. The draw time and equilibration time, as well as per-sample analysis time and temperature were logged. Table 5. S04I Sample Quality Flags Quality Flag WHP ||----------------------------|| Samples Total || 1 | 2 | 3 | 4 | 5 | 9 || --------- ----- ||---|------|----|----|---|---|| Bottle 3640 || 0 | 3632 | 1 | 4 | 0 | 3 || CTD Salt 3640 || 0 | 3557 | 0 | 83 | 0 | 0 || Nitrite 3854 || 0 | 3853 | 0 | 1 | 0 | 0 || Nitrate 3854 || 0 | 3853 | 0 | 1 | 0 | 0 || Oxygen 3674 || 6 | 3623 | 24 | 21 | 0 | 0 || Phosphate 3854 || 0 | 3811 | 4 | 39 | 0 | 0 || Silicate 3854 || 0 | 3852 | 1 | 1 | 0 | 0 || Salinity 3650 || 9 | 3559 | 26 | 56 | 0 | 0 || Two Guildline Autosal Model 8400A salinometers (55-654 and 57-396) located in a temperature-controlled laboratory were used to measure salinities. The salinometers were modified by ODF and contained interfaces for computer-aided measurement. A computer (PC) prompted the analyst for control functions (changing sample, flushing) while it made continuous measurements and logged results. The salinometer cell was flushed until successive readings met software criteria for consistency, then two successive measurements were made and averaged for a final result. The salinometer was standardized for each cast with IAPSO Standard Seawater (SSW) Batch P-125, using at least two fresh vials per cast. The estimated accuracy of bottle salinities run at sea is usually better than 0.002 PSU relative to the particular Standard Seawater batch used. PSS-78 salinity (UNESCO, 1981) was then calculated for each sample from the measured conductivity ratios, and the results merged with the cruise database. 3650 salinity measurements were made and 233 vials of standard water were used. Salinometer 55-654 was used throughout this leg. The temperature stability of the laboratory used to make the measurements was fair, ranging from 18.7ƒ to 25.8ƒC. The salinometer bath temperature was maintained at 24ƒC for all runs except for casts 032/01-039/01 (21ƒC). The salinities were used to calibrate the CTD conductivity sensor. C5. Oxygen Analysis (F. Delahoyd) Samples were collected for dissolved oxygen analyses soon after the rosette sampler was brought on board and after CFC and helium were drawn. Nominal 125 ml volume-calibrated iodine flasks were rinsed twice, then filled via a drawing tube and allowed to overflow for at least 3 flask volumes. The sample temperature was measured with a small platinum resistance thermometer embedded in the drawing tube. Draw temperatures were somewhat useful in detecting possible bad trips even as samples were being drawn. Reagents were added to fix the oxygen before stoppering. The flasks were shaken twice; immediately after drawing, and then again after 20 minutes, to assure thorough dispersion of the MnO(OH)2 precipitate. The samples were analyzed within 4 hours of collection. Dissolved oxygen analyses were performed with an SIO-designed automated oxygen titrator using photometric end-point detection based on the absorption of 365 nm wavelength ultra-violet light. Thiosulfate was dispensed by a Dosimat 665 buret driver fitted with a 1.0 ml buret. ODF uses a whole-bottle modified-Winkler titration following the technique of Carpenter (1965) with modifications by Culberson et al. (1991), but with higher concentrations of potassium iodate standard (approximately 0.012N) and thiosulfate solution (55 gm/l). Standard solutions prepared from pre-weighed potassium iodate crystals were run at the beginning of each session of analyses, which typically included from 1 to 3 stations. Nine standards were made up during the cruise and compared to assure that the results were reproducible, and to preclude the possibility of a weighing error. Reagent/distilled water blanks were determined to account for oxidizing or reducing materials in the reagents. Carbon Disulfide was added to the thiosulfate as a preservative. The auto-titrator generally performed very well. The samples were titrated and the data logged by the PC control software. The data were then used to update the cruise database on the Sun SPARCstations. Blanks, and thiosulfate normalities corrected to 20C, calculated from each standardization, were plotted versus time, and were reviewed for possible problems. New thiosulfate normalities were recalculated after the blanks had been smoothed. These normalities were then smoothed, and the oxygen values recalculated. Oxygens were converted from milliliters per liter to micro-moles per kilogram using the in-situ temperature. Ideally, for whole-bottle titrations, the conversion temperature should be the temperature of the water issuing from the bottle spigot. The sample temperatures were measured at the time the samples were drawn from the bottle, but were not used in the conversion from milliliters per liter to micro-moles per kilogram because the software was not available. Aberrant drawing temperatures provided an additional flag indicating that a bottle may not have tripped properly. Oxygen flasks were calibrated gravimetrically with degassed deionized water (DIW) to determine flask volumes at ODF's chemistry laboratory. This is done once before using flasks for the first time and periodically thereafter when a suspect bottle volume is detected. All volumetric glassware used in preparing standards is calibrated as well as the 10 ml Dosimat buret used to dispense standard Iodate solution. Iodate standards are pre-weighed in ODF's chemistry laboratory to a nominal weight of 0.44xx grams and the exact normality is calculated at sea. Potassium Iodate (KIO3) is obtained from Johnson Matthey Chemical Co. and is reported by the suppliers to be > 99.4% pure. All other reagents are "reagent grade" and are tested for levels of oxidizing and reducing impurities prior to use. 3674 oxygen measurements were made. No major problems were encountered with the analyses. The temperature stability of the laboratory used for the analyses was fair, ranging from 19.3ƒC to 26.8ƒC. The oxygen data were used to calibrate the CTD dissolved O2 sensor. C6. Nutrient Analysis (F. Delahoyd) Nutrient samples were drawn into 45 ml high density polypropylene, narrow mouth, screw-capped centrifuge tubes which were rinsed three times before filling. The tubes were also rinsed with 1.2N HCl after each use. Standardizations were performed at the beginning and end of each group of analyses (one cast, usually 36 samples) with a set of an intermediate concentration standard prepared in low-nutrient seawater for each run from secondary standards. The secondary standards were prepared aboard ship by dilution from dry, pre-weighed primary standards. Sets of 6-7 different concentrations of shipboard standards were analyzed periodically to determine the deviation from linearity as a function of concentration for each nutrient. Nutrient analyses (phosphate, silicate, nitrate and nitrite) were performed on an ODF-modified 4 channel Technicon Auto-Analyzer II, generally within one hour of the cast. Occasionally some samples were refrigerated at 2ƒC to 6ƒC for a maximum of 4 hours. The methods used are described by Gordon et al. (1992). The colorimeter outputs from each of the four channels were digitized and logged automatically by computer (PC), then split into absorbence peaks. Each was manually verified. Silicate is analyzed using the technique of Armstrong et al. (1967). Ammonium molybdate is added to a seawater sample to produce silicomolybdic acid which is then reduced to silicomolybdous acid (a blue compound) following the addition of stannous chloride. Tartaric acid is also added to impede PO4 color development (interference). The sample is passed through a 15 mm flowcell and the absorbence measured at 820nm. ODF's methodology is known to be non-linear at high silicate concentrations (>120 M); a correction for this nonlinearity is applied in ODF's software. Modifications of the Armstrong et al. (1967) techniques for nitrate and nitrite analysis are also used. The seawater sample for nitrate analysis is passed through a cadmium column where the nitrate is reduced to nitrite. Sulfanilamide is introduced, reacting with the nitrite, then N-(1-naphthyl)ethylenediamine dihydrochloride which couples to form a red azo dye. The reaction product is then passed through a 15 mm flowcell and the absorbence measured at 540 nm. The same technique is employed for nitrite analysis, except the cadmium column is not present, and a 50 mm flow-cell is used. Phosphate is analyzed using a modification of the Bernhardt and Wilhelms (1967) technique. Ammonium molybdate is 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 is heated to 55ƒC to enhance color development, then passed through a 50 mm flowcell and the absorbence measured at 820 nm. Nutrients, reported in micromoles per kilogram, were converted from micromoles per liter by dividing by sample density calculated at zero pressure, in-situ salinity, and an assumed laboratory temperature of 25ƒC. Na2SiF6, the silicate primary standard, is obtained from Aesar and is reported by the supplier to be >98% pure. Primary standards for nitrate (KNO3), nitrite (NaNO2), and phosphate (KH2PO4) are obtained from Johnson Matthey Chemical Co. and the supplier reports purities of 99.999%, 97%, and 99.999%, respectively. 3854 nutrient analyses were performed. No major problems were encountered with the measurements. The pump tubing was changed four times, and deep seawater was run as a substandard on each run. The efficiency of the cadmium column used for nitrate was monitored throughout the cruise and ranged from 99.6-100.0%. The temperature stability of the laboratory used for the analyses was good, ranging from 20.0 to 24.0ƒC. C7. Chlorofluorocarbon Analysis (M. Mensch) The CFC analysis on board as well as the sampling in flame-sealed glass ampoules for subsequent on-shore analysis were performed by Guy Mathieu and Manfred Mensch (Lamont-Doherty Earth Observatory of Columbia University, New York, PI Bill Smethie) and by Steve Covey (Universi