A. Cruise Narrative A.1 Highlights A.1.a WOCE designation P18S P18N A.1.b EXPOCODE P18S: 31DSCG94/2 P18N: 31DSCG94/3 A.1.c P18S: Chief Scientist Bruce Taft NOAA/PMEL NOAA Building 3 Bin C15700 7600 Sand Point Way NE Seattle WA 98115-0070 Phone: 206-526-4146 Fax: 206-526-6744 Internet: taft@noaapmel.gov Co-Chief Scientist John Bullister NOAA/PMEL NOAA Building 3 Bin C15700 7600 Sand Point Way NE Seattle WA 98115-0070 Phone: 206-526-6741 Fax: 206-526-6744 Internet: bullister@noaapmel.gov P18N Chief Scientist Gregory Johnson NOAA/PMEL NOAA Building 3 7600 Sand Point Way NE Seattle WA 98115-0070 Phone: 206-526-6806 Fax: 206-526-6744 Internet:gjohnson@noaapmel.gov Co-Chief Scientist Richard Feely NOAA/PMEL NOAA Building 3 7600 Sand Point Way NE Seattle WA 98115-0070 A.1.d Ship R/V Discoverer A.1.e Ports of Call P18S: Punta Arenas to Easter Island P18N: Easter Island to San Diego A.1.f Cruise dates P18S: Feb 22 - March 2 1994 P18N: March 27 - April 3 1994 A.2 Cruise Summary Information A.2.a Geographic boundaries 23 N 110 W 103 W 67 S A.2.b Stations Occupied A total of 185 full water column CTD/water sample stations were made along the section from 67 S 103 W to 23 N 110 W. Of these, 158 stations were made using a 36-position, 10-liter bottle frame with a lowered Acoustic Doppler Current Profiler (ADCP) and a transmissometer. The other 27 stations were made using a 24-position, four-liter bottle frame deployed primarily during heavy weather. A Sea-Bird Electronics 911plus CTD was mounted in each frame. In addition to a set of temperature and conductivity sensors resident on each CTD, a single set of mobile temperature, conductivity, and dissolved oxygen sensors was used at every station for quality control. Water samples were collected at every station for analyses of salt, dissolved oxygen and dissolved nutrients (i.e., silicate, nitrate, nitrite and phosphate). Samples were drawn at selected locations for analysis of CFC-11, CFC-12, dissolved inorganic carbon (DIC), total alkalinity, pH, pCO2, 3He, tritium, dissolved organic carbon, carbon isotopes, oxygen isotopes, and other variables. Daily shallow casts were made for assessment of various biological parameters, including productivity. A total of 25 ALACE (Autonomous Lagrangian Circulation Explorer) floats were deployed during the cruise. Nineteen XCTDs were successfully launched between CTD/O2 stations from 1 - 9.5 N. Underway measurements included ADCP data, meteorological variables, bottom depth, pH, pCO2, atmospheric CFCs, nitrate, and chlorophyll. A.2.c Floats and drifters deployed ALACE Floats were launched at 25 locations listed in Table 1. Twelve ALACE floats were released on Leg 2 and thirteen on Leg 3. Table 1: Time and location of ALACE float deployments Date Time Latitude Longitude ----------------------------------------------------------------- 022494 0756 55 50.17 S 80 22.34 W 022494 1636 56 39.64 S 81 46.87 W 022494 2130 57 30.02 S 83 17.12 W 022594 0228 58 19.87 S 84 45.79 W 022594 0725 59 09.26 S 86 18.96 W 022594 1210 59 59.90 S 87 51.50 W 030894 1025 55 10.40 S 103 01.09 W 031094 2028 49 49.28 S 103 00.10 W 031394 0637 44 58.99 S 103 00.25 W 031594 0117 40 00.99 S 103 00.55 W 031894 1200 35 00.40 S 103 00.74 W 032094 0739 30 00.15 S 103 01.53 W 032994 1341 25 00.24 S 103 00.05 W 033194 2011 20 29.51 S 102 59.98 W 040494 0005 14 59.70 S 103 00.01 W 040694 1917 9 59.76 S 103 00.70 W 040994 1441 6 09.09 S 108 38.61 W 041094 2307 3 59.28 S 110 19.78 W 041294 1838 1 20.27 S 110 19.94 W 041494 1443 1 00.38 N 110 19.96 W 041694 1431 3 59.69 N 110 19.93 W 041794 1731 5 59.90 N 110 20.30 W 041994 1956 10 00.78 S 110 00.19 W 042194 1819 14 29.77 S 110 00.03 W 042394 2246 18 59.93 S 109 59.80 W ------------------------------------------------------------------ A.2.d Moorings deployed or recovered A.3 Principal Investigators Table 2: List of Principal Investigators Measurement Principal Investigator Institution --------------------------------------------------------------------- CTD/O2 B. Taft, G. Johnson PMEL Chlorofluorocarbons (CFCs) J. Bullister PMEL C-14 (AMS radiocarbon), C-13 P. Quay UW Nutrients K. Krogsland UW Dissolved Oxygen J. Bullister PMEL Helium/tritium W. Jenkins WHOI CO2 (alkalinity) F. Millero UM Total CO2 (coulometry), pCO2 R. Feely PMEL pH R. Byrne USF ADCP P. Hacker UH ALACE floats R. Davis SIO Underway stmospheric/surface halocarbons, nitrous oxide J. Butler CMDL Productivity F. Chavez MBARI --------------------------------------------------------------------- NOAA/PMEL National Oceanic and Atmospheric Adminstration Pacific Marine Environmental Laboratory USF University of South Florida MBARI Monterey Bay Aquarium Research Institute SIO Scripps Institution of Oceanography UM University of Miami UW University of Washington UH University of Hawaii WHOI Woods Hole Oceanographic Institution CMDL NOAA Climate Modelling and Diagnostics Laboratory --------------------------------------------------------------------- A.4 Scientific Programme and Methods The long term objective of the Climate and Global Change Program is to provide reliable predictions of climate change and associated regional implications on time scales ranging from seasons to a century or more. In support of NOAA's Climate Program, PMEL scientists have been measuring the growing burden of greenhouse gases in the Pacific Ocean and the overlying atmosphere since 1980. The NOAA Office of Global Programs (OGP) sponsored Ocean Tracers and Hydrography Program and Ocean-Atmosphere Carbon Exchange Study (OACES) studies ocean circulation, mixing processes, and the rate at which CO2 and chlorofluorocarbons (CFCs) are taken up and released by the oceans. Work on this cruise was cooperative with the World Ocean Circulation Experiment (WOCE) and the U.S. Joint Global Ocean Flux Study (JGOFS). The research was designed to (1) describe water properties and relate them to circulation processes throughout the water column in the eastern Pacific Ocean; (2) determine the sources and sinks of carbon dioxide along 103-110W; (3) study the invasion of CFCs in the ocean; and (4) provide a high quality set of baseline measurements for the continuing evaluation of changes in ocean content of dissolved gasses, water properties, and circulation. This section fills a gap in the eastern Pacific between WOCE Hydrographic Programme (WHP) meridional sections P19 (along 90W) and P17 (along 135W). The southern end of this section intersects WHP S4, an E-W section along 67S occupied in 1992. During the transit (leg 1) from Seattle, Washington to Punta Arenas, Chile, a test station was occupied in the Puget Sound to evaluate the CTD/rosette system. This profile was not processed and is not included in this data report. In response to significant volcanic activity detected by the VENTS monitoring system at the East Blanco Depression (44 12N, 129 42W), 6 stations were occupied in this area during leg 1. The NOAA/PMEL VENTS program focuses research on determining the oceanic impacts and consequences of submarine hydrothermal venting. This event was particularly interesting as the area is a pull-apart basin in a transform zone, possibly the site of early ridge formation. Occupation of WOCE section P18 began with station 10 of leg 2, after two test casts were completed enroute to 67S, 103W from Punta Arenas, Chile. Seventy-eight full water column hydrographic stations were occupied east of the Pacific Rise along 103W from 67S to 27S. Stations were spaced at 30 nm intervals except from 58 30S to 48S where spacing was increased to 40 nm intervals to make up time lost from bad weather and winch level wind problems. Features sampled during leg 2 included the Polar and Subantarctic Fronts of the Antarctic Circumpolar Current, the Subtropical Front, the Subantarctic Mode Water, the Antarctic Intermediate Water, the Circumpolar Deep Water spreading to the northern reaches of the Southeast Pacific Basin, and currents along the Sala y Gomez Fracture Zone. During leg 3 stations continued northward along 103W to 10S at 30 nm intervals. The section turned northwestward from 10S to 5S with 40 nm station spacing to cross the East Pacific Rise in a perpendicular fashion. The 30 nm spacing was resumed from 5S to 3S northward along 110 20W. From 3S to 3N stations were occupied every 20 nm along the same longitude. From 3N to 22 30N stations were occupied at 30 nm intervals, except from 12N to 16N, where the spacing was again increased to 40 nm to make up for time lost to winch level wind problems. A gradual shift in the longitude from 110 20W to 110W was made between 8N and 10N. North of 22 30N station spacing was reduced to as little as 3 nm over the rapidly shoaling bathymetry approaching Cabo San Lucas. The line was completed in 200 m of water at 22 51N, 110W. During leg 3, 107 full water column hydrographic stations were occupied sampling the deep waters of the Bauer Basin, currents associated with the flanks of the East Pacific Rise, tropical water masses and currents over the full water column, the northern mid-depth helium-3 plume, and the oxygen depleted layer of the tropical Eastern Pacific. Full water column CTD/O2 profiles were collected at all stations. Lowered Acoustic Doppler Current Profiler (ADCP) measurements were also collected on most casts. In addition, underway salinity, temperature, and CO2 measurements were taken along the cruise track. Shallow productivity casts were made daily, ALACE floats were launched at predetermined locations, and XCTDs were successfully dropped in a high-resolution survey from 1N to 9.5N. Water samples were analyzed for a suite of anthropogenic and natural tracers including salinity, dissolved oxygen, inorganic nutrients, CFCs, pCO2, total CO2, pH, total alkalinity, helium, tritium, C-13, C-14, O-18, dissolved organic carbon, and dissolved organic nitrogen. Samples were collected from productivity casts for chlorophyll and primary productivity. Leg 1 (Seattle, Washington to Punta Arenas, Chile) This leg was a transit leg with a test station occupied in the Puget Sound to evaluate the CTD/rosette system. This profile was not processed and is not included in this data report. In response to significant volcanic activity detected by the VENTS monitoring system at the East Blanco Depression (44 12N, 129 42W), 6 stations were occupied in this area during leg 1. Leg 2 (Punta Arenas - Easter Island). This leg consisted of 78 stations along 103 W; the first station on the WOCE Line P18 (#10) was occupied at 67 00 S 103 00 W on 26 February 1994 and the final station at 26 00 S 103° 00 W on 23 March 1994. Except for 10 degrees of latitude span (58 30 S - 48 30 S), the station spacing was 30 miles. The station spacing was increased to 40 miles in the above mentioned latitudinal band because of time lost to heavy weather and slower than normal retreival rates of the CTD package due to problems with the winch level wind. All CTD stations were full depth (nominally 10 m above the bottom). Two CTD/rosette packages were used: a 24 position 4 l bottle rosette (21 stations) and a 36 position 10 l bottle rosette (57 stations). The choice between the two systems was usually dictated by the severity of the weather. On stations where the large rosette was used, a LADCP was attached to the rosette frame which reduced the number of bottle positions from 36 to 33. Shallow (200 m) productivity bottle casts with light transmission profiles were made at 23 stations. Twelve ALACE floats were released at predetermined locations along the section and on the transit to the first station. Leg 3 (Easter Island - San Diego). A similar observational program was carried out on this leg (107 stations) with the following changes from the nominal 30-mile station spacing. Stations were occupied at 40 mile intervals along a dog-leg section across the East Pacific Rise from 10 S 103 W to 5 S 110 20 W. Thirty-mile spacing was resumed between 5 S and 3 S and then reduced to 20 miles between 3 S and 3 N. From 3 N to 22 30 N stations were occupied at 30 mile intervals except between 12 N and 16 N, where spacing was again relaxed to 40 miles. Between 8 N and 10 N a gradual shift in longitude from 110 20 W to 110 00 W was made. As the ship approached Cabo San Lucas, at the end of the section, spacing was reduced to as little as 3 miles over the steeply shoaling bathymetry. Only on six stations, during reterminations of the CTD cable, was the 24 bottle rosette used. Discussion: The basic goals of the cruise were accomplished. All casts were made to the bottom. Station spacing only occasionally was increased to 40 miles from the nominal WOCE interval of 30 miles. There were no significant gaps in sampling any of the variables. Preliminary analysis of the Seabird CTD measurements and bottle data indicate that they will meet the WOCE standards. A.5 Major Problems and Goals not Achieved A.6 Other Incidents of Note A.7 List of Cruise Participants A list of cruise participants is found in Table 3. TABLE 3: Cruise Participants --------------------------------------------------------------------- Institutions: NOAA Pacific Marine Environmental Laboratory (PMEL) NOAA Atlantic Oceanographic and Meterological Laboratory (AOML) NOAA Climate Modelling and Diagnostics Laboratory (CMDL) Monterey Bay Aquarium Research Institute (MBARI) Scripps Institution of Oceanography (SIO) University of Hawaii (UH) University of Washington (UW) University of Miami (UM) University of South Florida (USF) Woods Hole Oceanographic Institution (WHOI) --------------------------------------------------------------------- Name Responsibility Leg 1 Leg 2 Leg 3 Andrew Dickson, SIO Chief Scientist x Bruce Taft, PMEL Chief Scientist x John Bullister, PMEL Co-Chief Scientist x Gregory Johnson, PMEL Chief Scientist x x Richard Feely, PMEL Co-Chief Scientist x Kristene McTaggart, PMEL CTD x x x Nordeen Larson, Sea-Bird CTD x x Gregg Thomas, AOML salinity x x David Wisegarver, PMEL CFC x x C.J. Beegle, UW CFC x Kirk Hargreaves, PMEL oxygen/CFC x x David Jones, PMEL oxygen x Katherine Krogslund, UW nutrients x x Calvin Mordy, PMEL nutrients x x Kerry Jones, PMEL TCO2 x x x Thomas Lantry, AOML TCO2 x x Marilyn Roberts, PMEL TCO2 x Dana Greeley, PMEL pCO2 x x Catherine Cosca, PMEL pCO2 x Matthew Steckley, AOML pCO2 x Sonya Olivella, UM alkalinity x J. Zhang, UM alkalinity x Bernardo Vargas, UM alkalinity x Essa Peltola, UM alkalinity x Michael De Alessi, UM alkalinity x Mary Roche, UM alkalinity x Robert Byrne, USF pH x Renate Bernstein, USF pH x Huining Zhang, USF pH x x Sean McElligott, USF pH x Frederick Stengard, USF pH x Craig Huhta, UH ADCP x Claude Lumpkin, UH ADCP x Joshua Curtice, WHOI helium, tritium x Scott Birdwhistell, WHOI helium, tritium x Dennis Hansell, BBSR DOC x Rhonda Kelly, BBSR DOC x James Green, UW C-13, C-14 x Elizabeth Houzel, UW C-13, C-14 x Kurt Buck, MBARI productivity x x Thomas Hayden, USC productivity x Gregory Morris, USC productivity x Raphael Kudela, MBARI productivity x Dante Gutierrez-Besa, SHOA Chilean Observer x Diego Lopez-Veneroni, Texas A&M Mexican Observer x Humberto Perez-Ortiz, Direccion General de Oceanografia Naval, Secretaria de Marina Mexican Observer x --------------------------------------------------------------------- B.1 Navigation and bathymetry SeaBeam multibeam sonar was used continuously for bathymetry during both legs. Navigation was by means of the Global Positioning System (GPS). B.2 Acoustic Doppler Current Profiler (ADCP) Shipboard ADCP measurements, along with global position system (GPS) data, were collected continuously along the track to measure the velocity profile in the upper 500 m. B.3 Thermosalinograph and underway dissolved oxygen, etc A thermosalinograph was operated continuously on both legs. pCO2 and pH were measured while underway together with photosynthetically active radiation, nitrate and chlorophyll concentrations. B.4 XBT and XCTD Nineteen XCTDs were dropped along 110 20 W between 1 10 N and 9 45 N at locations halfway between successive CTD stations on Leg 3. Times and positions of each deployment are shown in Table 4. TABLE 4: Deployment times and locations for XCTD casts Date Time Latitude Longitude -------------------------------------------------------- 012994 0355 44 12.97 N 129 37.08 W 030294 1916 62 27.85 S 102 58.45 W 030394 0941 61 25.90 S 102 58.90 W 031094 0556 51 09.50 S 103 00.60 W 041494 1540 1 10.01 N 110 19.87 W 041494 2208 1 30.10 N 110 19.60 W 041594 0340 1 50.30 N 110 19.70 W 041594 0933 2 10.10 N 110 20.00 W 041594 1455 2 30.00 N 110 19.80 W 041594 2116 2 50.00 N 110 19.90 W 041694 0250 3 15.00 N 110 19.90 W 041694 0942 3 45.00 N 110 19.40 W 041694 1546 4 15.00 N 110 19.80 W 041694 2313 4 45.00 N 110 20.00 W 041794 0536 5 16.28 N 110 19.77 W 041794 1227 5 45.00 N 110 20.00 W 041794 1845 6 15.03 N 110 20.46 W 041894 0038 6 45.00 N 110 20.60 W 041894 0659 7 15.00 N 110 20.61 W 041894 1307 7 45.00 N 110 19.90 W 041894 2011 8 15.00 N 110 17.74 W 041894 0159 8 45.10 N 110 12.50 W 041994 0822 9 15.00 N 110 07.60 W -------------------------------------------------------- B.5 Meteorological observations B.6 Atmospheric chemistry 3/8" O.D. Dekaron air sampling lines (reinforced plastic tubing) was run from the CFC van to the bow and stern and air was analyzed continuously for: CFC-11 CFC-12 CFC-113 Carbon tetrachloride Methyl chloroform C. Hydrographic Measurements C.1. CTD/O2 Measurements and Calibrations (K.E. McTaggart, G.C. Johnson, and B.A. Taft) C.1.1. STANDARDS AND PRE-CRUISE CALIBRATIONS The CTD system is a real time data system with the CTD data from a Sea-Bird Electronics, Inc. (SBE) 9plus underwater unit transmitted via a conducting cable to the SBE 11plus deck unit. The serial data from the underwater unit is sent to the deck unit in RS-232 NRZ format using a 34560 Hz carrier-modulated differential-phase-shift-keying (DPSK) telemetry link. The deck unit decodes the serial data and sends it to a personal computer for display and storage in a disk file using Sea-Bird SEASOFT software. The SBE 911plus system transmits data from primary and auxiliary sensors in the form of binary number equivalents of the frequency or voltage outputs from those sensors. The calculations required to convert from raw data to engineering units of the parameters being measured are performed by software, either in real-time, or after the data has been stored in a disk file. The SBE 911plus system is electrically and mechanically compatible with standard, unmodified rosette water samplers made by General Oceanics (GO), including the 1016 36-position sampler. An optional modem and rosette interface allows the 911plus system to control the operation of the rosette directly, and without interrupting the data from the CTD, eliminating the need for a rosette deck unit. The SBE 9plus underwater unit uses Sea-Bird's standard modular temperature (SBE 3) and conductivity (SBE 4) sensors which are mounted with a single clamp and "L" bracket to the lower end cap. The conductivity cell entrance is co-planar with the tip of the temperature sensor's protective steel sheath. The pressure sensor is mounted inside the underwater unit main housing and is ported to outside pressure through the oil-filled plastic capillary tube seen protruding from the main housing bottom end cap. A compact, modular unit consisting of a centrifugal pump head and a brushless DC ball bearing motor contained in an aluminum underwater housing pump flushes water through sensor tubing at a constant rate independent of the CTD's motion. This improves dynamic performance. Motor speed and pumping rate (3000 rpm) remain nearly constant over the entire input voltage range of 12-18 volts DC. The SBE 11plus deck unit is a rack-mountable interface which supplies DC power to the underwater unit, decodes the serial data stream, formats the data under microprocessor control, and passes the data to a companion computer. It provides access to the modem channel and control of the rosette interface. Output data is in RS-232 (serial) format. C.1.1.a. Conductivity The flow-through conductivity sensing element is a glass tube (cell) with three platinum electrodes. The resistance measured between the center electrode and end electrode pair is determined by the cell geometry and the specific conductance of the fluid within the cell, and controls the output frequency of a Wien Bridge circuit. The sensor has a frequency output of approximately 3 to 12 kHz corresponding to conductivity from 0 to 7 S/m (0 to 70 mmho/cm). The SBE 4 has a typical accuracy/stability of +/- 0.0003 S/m/month; resolution of 0.00004 S/m at 24 samples per second; and 6800 meter anodized aluminum housing depth rating. Pre-cruise sensor calibrations were performed at Sea-Bird Electronics, Inc. in Bellevue, Washington. The following coefficients were entered into SEASOFT using software module SEASON: S/N 1177 September 22, 1993 S/N 1247 January 21, 1994 a = 2.28847772e-05 a = 1.76162580e-05 b = 5.58250114e-01 b = 5.50791410e-01 c = -4.14341657e+00 c = -4.07804361e+00 d = -9.59251789e-05 d = -9.32262258e-06 m = 4.1 m = 4.2 Conductivity calibration certificates show an equation containing the appropriate pressure-dependent correction term to account for the effect of hydrostatic loading (pressure) on the conductivity cell: C (S/m) = (af^m + bf^2 + c + dt) / [10 (1 - 9.57e-8 p)] where a, b, c, d, and m are the calibration coefficients above, f is the instrument frequency (kHz), t is the water temperature (C), and p is the water pressure (decibars). SEASOFT automatically implements this equation. C.1.1.b. Temperature The temperature sensing element is a glass-coated thermistor bead, pressure-protected by a stainless steel tube. The sensor output frequency ranges from approximately 5 to 13 kHz corresponding to temperature from -5 to 35 degrees Celsius. The output frequency is inversly proportional to the square root of the thermistor resistance which controls the output of a patented Wien Bridge circuit. The thermistor resistance is exponentially related to temperature. The SBE 3 thermometer has a typical accuracy/stability of +/- 0.004 C per year; and resolution of 0.0003 C at 24 samples per second. The SBE 3 thermometer has a fast response time of 70 milliseconds. It's anodized aluminum housing provides a depth rating of 6800 meters. Pre-cruise sensor calibrations were performed at Sea-Bird Electronics, Inc. in Bellevue, Washington. The following coefficients were entered into SEASOFT using software module SEASON: S/N 1455 January 13, 1994 S/N 1461 February 11, 1994 a = 3.68103063e-03 a = 3.68110418e-03 b = 6.03073078e-04 b = 6.00486851e-04 c = 1.51707342e-05 c = 1.48701147e-05 d = 2.20648879e-06 d = 1.99797919e-06 f0 = 6228.23 f0 = 6212.56 Temperature (IPTS-68) is computed according to T (deg C) = 1/{a+b[ln(f0/f)]+c[ln^2(f0/f)]+d[ln^3(f0/f)]}-273.15 where a, b, c, d, and f0 are the calibration coefficients above and f is the instrument frequency (kHz). SEASOFT automatically implements this equation. C.1.1.c. Pressure The Paroscientific series 4000 Digiquartz high pressure transducer uses a quartz crystal resonator whose frequency of oscillation varies with pressure induced stress measuring changes in pressure as small as 0.01 parts per million with an absolute range of 0 to 10,000 psia (0 to 6885 decibars). Also, a quartz crystal temperature signal is used to compensate for a wide range of temperature changes. Repeatability, hysteresis, and pressure conformance are 0.005% FS. The nominal pressure frequency (0 to full scale) is 34 to 38 kHz. The nominal temperature frequency is 172 kHz + 50 ppm/degree Celsius. Pre-cruise sensor calibrations were performed at Sea-Bird Electronics, Inc. in Bellevue, Washington. The following coefficients were entered into SEASOFT using software module SEASON: S/N 53960 August 4, 1993 S/N 53586 October 29, 1993 c1 = -43150.48 c1 = -39204.51 c2 = 4.54280e-01 c2 = 6.23456e-01 c3 = 1.34438e-02 c3 = 1.35057e-02 d1 = 0.037952 d1 = 0.038943 d2 = 0.0 d2 = 0.0 t1 = 30.34230 t1 = 30.46303 t2 = -1.80938e-04 t2 = -9.018862e-05 t3 = 4.61615e-06 t3 = 4.52889e-06 t4 = 2.08422e-09 t4 = 3.30959e-09 t5 = 0.0 t5 = 0.0 Pressure coefficients are first formulated into c = c1 + c2*U + c3*U^2 d = d1 + d2*U t0 = t1 + t2*U + t3*U^2 + t4*U^3 + t5*U^4 where U is temperature in degrees Celsius. Then pressure is computed according to P (psia) = c * [1 - (t0^2/t^2)] * {1 - d[1 - (t0^2/t^2)]} where t is pressure period (microsec). SEASOFT automatically implements this equation. C.1.1.d. Oxygen The SBE 13 dissolved oxygen sensor uses a Beckman polarographic element to provide in-situ measurements at depths up to 6800 meters. This auxiliary sensor is also included in the path of pumped sea water. Oxygen sensors determine the dissolved oxygen concentration by counting the number of oxygen molecules per second (flux) that diffuse through a membrane. By knowing the flux of oxygen and the geometry of the diffusion path the concentration of oxygen can be computed. The permeability of the membrane to oxygen is a function of temperature and ambient pressure. The interface electronics outputs voltages proportional to membrane current (oxygen current) and membrane temperature (oxygen temperature). Oxygen temperature is used for internal temperature compensation. Computation of dissolved oxygen in engineering units is done in the software. The range for dissolved oxygen is 0 to 15 ml/l; accuracy is 0.1 ml/l; resolution is 0.01 ml/l. Response times are 2 seconds at 25 degrees Celsius and 5 seconds at 0 degrees Celsius. The following oxygen calibrations were entered into SEASOFT using SEACON: S/N 130309 September 7, 1993 m = 2.4544 e-7 b = -4.6633 e-10 k = 8.9224 c = -6.9788 The use of these constants in linear equations of the form I = mV + b and T = kV + c will yield sensor membrane current and temperature (with a maximum error of about 0.5 degrees Celsius) as a function of sensor output voltage. These scaled values of oxygen current and oxygen temperature were carried through the SEASOFT processing stream unaltered. C.1.2. DATA ACQUISITION CTD measurements were made using one of two Seabird 9plus CTDs each equipped with a fixed pumped temperature-conductivity (TC) sensor pair. A mobile pumped TC pair with dissolved oxygen sensor was mounted on whichever CTD was in use so that dual TC measurements and dissolved oxygen measurements were always collected. The TC pairs were monitored for calibration drift and shifts by examining the differences between the two pairs on each CTD and comparing CTD salinities with bottle salinity measurements. PMEL's Sea-Bird 9plus CTD/O2 S/N 09P8431-0315 (sampling rate 24 Hz) was mounted in a 36-position frame and employed as the primary package. Auxiliary sensors included a lowered ADCP, Metrox load cell, Benthos altimeter, and SeaTech transmissometer. Water samples were collected using a General Oceanics 36-bottle rosette and 10-liter Nisken bottles. The primary package was used for the majority of 194 casts. PMEL's Sea-Bird 9plus CTD/O2 S/N 329053-0209 (sampling rate 24 Hz) was mounted in a 24-position frame and employed as the backup package. Auxiliary sensors included a Metrox load cell and Benthos altimeter. Water samples were collected using a Sea-Bird 24-bottle rosette, and 4-liter Niskin bottles. There were 29 bad weather stations made using the smaller backup package. The package entered the water from the stern of the ship and was held 5-20 m beneath the surface for one minute in order to activate the pump and attach tag lines for package recovery. Under ideal conditions the package was lowered at a rate of 30 m/min to 50 m, 45 m/min to 200 m, and 60 m/min to depth. Ship roll often caused substantial variation about these mean lowering rates, especially at southern ocean stations. Load cell values were monitored in real-time during each cast. The position of the package relative to the bottom was monitored on the ship's Precision Depth Recorder (PDR). A bottom depth was estimated from bathymetric charts and the PDR ran during the bottom 1000 m of the cast. Fig. 2 shows the depths of bottle closures during the upcast. Upon completion of the cast, sensors were flushed with deionized water and stored with a dilute Triton-X solution in the plumbing. Niskin bottles were sampled for salinity, dissolved oxygen, inorganic nutrients, CFCs, total CO2, pCO2, pH, C-13, C-14, O-18, helium, tritium, total alkalinity, dissolved organic carbon, and dissolved organic nitrogen. Sample protocols conformed to those specified by the WOCE Hydrographic Programme. A Sea-Bird 11plus deck unit received the data signal from the CTD. The analog data stream was recorded onto video cassette tape as a backup. Digitized data were forwarded to a 286-AT personal computer equipped with SEASOFT acquisition and processing software version 4.201. Temperature, salinity, and oxygen profiles were displayed in real-time. Raw data files were transferred to a 486 personal computer using Laplink version 3 and backed up onto 1/4" cartridge tapes using a Microsolutions Backpack QIC-80 external tape drive. C.1.2.a. Data Acquisition Problems During leg 2, station spacing increased to 40 nm between 58.5S and 48S owing to a delay in departure from Punta Arenas, delays owing to winch problems for some casts, and bad weather. About 36 hours were lost waiting for the weather to moderate at 58S. Other problems included poor level winding of the winch resulting in non-uniform lays on the drum and high tension crossing and snapping of the cable, compromised chemistry samples owing to contamination from the ship's stack output, and difficulties associated with doing CTDs from the stern of the ship in heavy to moderate seas at high latitudes. Stations 8 and 9 test casts were very noisy. Modulo errors persisted through cast 14. Station 11 cast 1 did not sample the upper 800 meters and so a second cast was performed at this station for these bottles. Station 11 cast 2 CTD data was not processed. Station 111 stopcocks and vents were left open therefore no samples were collected. At station 120, upcast water sampling was skipped from 800 to 400 db while a fishing vessel cleared it's net out of the water. Prior to station 123, the cable was reterminated after cutting off 2500 m of cable to get below bad wraps. At station 131 the package sat on the bottom for several minutes. The upcast CTD data were bad. Uptrace pressures were matched to downtrace pressures for bottle sample CTD data. Station 160 had increasing modulo errors during the downcast and was aborted. Water was found in the ground wire at the termination. No samples collected at station 160. There was no sample from station 190 bottle 11 owing to a stuck lanyard. C.1.2.b. Salinity Analyses Bottle salinity analyses were performed in a temperature-controlled van using two Guildline Model 8400A inductive autosalinometers standardized with IAPSO Standard Seawater batch P114. The autosalinometer in use was standardized before each run and either at the end of each run or after no more than 48 samples. The drift between standardizations was monitored and the individual samples were corrected for that drift by linear interpolation. Duplicate samples taken from the deepest bottle on each cast were analyzed on a subsequent day. Bottle salinities were compared with preliminary CTD salinities to aid in identification of leaking bottles as well as to monitor the CTD conductivity cells' performance and drift. The expected precision of the autosalinometer with an accomplished operator is 0.001 pss, with an accuracy of 0.003. To assess the precision of discrete salinity measurements on this cruise, a comparison was made for data from the instances in which two bottles were tripped within 10 dbar of each other at the same station below a depth of 2000 dbar. For the 138 instances in which both bottles of the pair have acceptable salinity measurements, the standard deviation of the differences is 0.0012 pss. This value is very close to the expected precision. Calibrated CTD salinities replace missing bottle salinities in the hydrographic data listing and are indicated by an asterisk. C.1.3. POST-CRUISE CALIBRATIONS Post-cruise sensor calibrations were done at Sea-Bird Electronics, Inc. during May 1994. For stations 2-8, temperature sensor T1455 (with pre-cruise calibration coefficients dated January 1994) and conductivity sensor C1177 (with pre-cruise calibration coefficients dated September 1993) were selected as the best source of data. Post-cruise calibrations showed T1455 had drifted (offset only) by approximately -0.0015; C1177 displayed a change in slope. For stations 9-194, sensor T1461 (with pre-cruise calibration coefficients dated January 1994) and C1247 (with pre-cruise calibration coefficients dated January 1994) were selected for final data reduction since they were used on both packages. Post-cruise calibrations showed T1461 to be drifting (offset only) by approximately -0.006 C. C1247 had drifted (slope and offset) by approximately -0.0009 S/m. At sea monitoring and post-cruise calibration of redundant TC pair T1460/C1180 showed T1460 had jumped by 0.002 C, warranting repair. Redundant TC pair T1072/C748 post-cruise calibration showed T1072 had drifted to an offset of -0.004 C. These TC pairs were not included in the final processing. C.1.3.a. Conductivity SEASOFT module ALIGNCTD was used to align conductivity measurements in time relative to pressure. Measurements can be misaligned due to the inherent time delay of the sensor response, the water water transit time delay in the pumped plumbing line, and the sensors being physically misaligned in depth. Because SBE 3 temperature response is fast (0.06 seconds), it is not necessary to advance temperature relative to pressure. When measurements are properly aligned, salinity spiking and density errors are minimized. For a SBE 9 CTD with ducted TC sensors and a 3000 rpm pump the typical net advance of conductivity relative to temperature is 0.073 seconds. The SBE 11 deck units advanced primary conductivity 0.073 seconds but do not advance secondary conductivity. Therefore when C1177 or C1247 conductivity data came from a secondary sensor channel the alignment was much larger, typically 0.06 seconds versus coming from a primary sensor channel, typically 0.02 seconds. Conductivity slope and bias, along with a pressure fudge term (beta) were computed by a least-squares minimization of CTD and bottle conductivity differences. The function minimized was BC - m * CC - b - beta * CP where BC is bottle conductivity (S/m), CC is pre-cruise calibrated CTD conductivity (S/m), CP is the CTD pressure (dbar), m is the conductivity slope, b is the bias (S/m), and beta is the pressure fudge term (S/m/dbar). The final CTD conductivity (S/m) is m * CC + b + beta * CP The slope term m is a fourth-order polynomial function of station number to allow the entire cruise to be fit at once with a smoothly-varying station- dependent slope correction. For each sensor a series of fits were made, each fit throwing out bottle values for locations having a residual between CTD and bottle conductivities of greater than three standard deviations. This procedure was repeated with the remaining bottle values until no more bottle values were thrown out. For C1177, the slope correction ranged from 1.00014254 to 1.00014262, the bias applied was -3.8e-4, and the beta term was -5.69e-9. Of 5040 bottles, the percentage of bottles retained in the fit was 84.9 with a standard deviation of CTD versus bottle conductivity differences of 1.19e-4 S/m. For C1247, the slope correction ranged from 1.00021478 to 1.00044972, the bias applied was -7.2e-4, and the beta term was -1.29e-8. Of 5797 bottles, the percentage of bottles retained in the fit was 83.4 with a standard deviation of 0.87e-4 S/m. The slope and bias were applied in SEACON. The beta-fudge term was applied after SEASOFT post-processing in PMEL program POSTCAL. CTD-bottle conductivity differences used for the final fits are plotted against cast number to show the stability of the calibrated CTD conductivities relative to the bottle conductivities. The entire set of CTD-bottle conductivity differences are plotted against pressure to show the tight fit below 1000 m and the increasing scatter above 1000 m. C.1.3.b. Temperature In SEACON, adjustments were made to the bias of the thermistors as deviations from the pre-cruise calibrations on a station by station basis. These deviations were obtained from a linear fit of the pre-cruise and post- cruise temperature residuals from the pre-cruise calibration versus time. Deep temperature differences between primary and secondary sensors were less than 0.001 C. Also, a uniform correction for heating of the thermistor owing to viscous effects was applied to the bias in SEACON. This correction was obtained using the formula: error[C] = B * sqrt(nu)*U*U where B=0.692, U=1.02 m/s, and nu=1.7279e-6 m2/s. The value for viscosity nu is that for the peak in the distribution of the temperature and salinity bottle values (te=1.8 C, sa=34.67 pss). Error[C] = 0.9464e-3 C. All the thermistors read high by this amount and were adjusted down accordingly. The adjustment is near the maximum viscous heating for the encountered temperature and salinity range. Thermistors will read about 0.66e-3 C high near the surface in the tropics (te=30 C, sa=34.5 pss) causing an overadjustment of 0.29e-3 C. For deep values (te=0 C, sa=37 pss) where gradients are small, thermistors will read about 0.97e-3 C high and so will be underadjusted by 0.2e-3 C. C.1.3.c. Oxygen In situ oxygen samples collected during CTD profiles are used for post-measurement calibration. SEASOFT bottle files were merged and bottle oxygen values flagged as 'good' were appended to the data records. Because the dissolved oxygen sensor has an obvious hysteresis, PMEL program OXDWNP replaced up-profile water sample data with corresponding down-profile CTD/O2 data at common pressure levels. Oxygen saturation values were computed according to Benson and Krause (1984) in units of umol/kg. The algorithm used for converting oxygen sensor current and probe temperature measurements to oxygen as described by Owens and Millard (1985) requires a non-linear least squares regression technique in order to determine the best fit coefficients of the model for oxygen sensor behavior to the water sample observations. WHOI program OXFITMR uses Numerical REcipes (Press et al., 1986) Fortran routines MRQMIN, MRQCOF, GAUSSJ, and COVSRT to perform non-linear least squares regression using Levenberg-Marquardt method. A Fortran subroutine FOXY describes the oxygen model with the derivatives of the model with respect to six coefficients in the following order: oxygen current slope, temperature correction, pressure correction, weight, oxygen current bias, and oxygen current lag. Program OXFITMR reads the data for a group of stations. The time rate of change of oxygen current is computed using a least squares estimate over 15 second intervals. The data are editted to remove spurious points where values are less than zero or greater than 1.2 times the saturation value. The routine varies the six (or fewer) parameters of the model in such a way as to produce the minimum sum of squares in the difference between the calibration oxygens and the computed values. Individual differences between the calibration oxygens and the computed oxygen values (residuals) are then compared with the standard deviation of the residuals. Any residual exceeding an edit factor of 2.8 standard deviations is rejected. A factor of 2.8 will have a 0.5% chance of rejecting a valid oxygen value for a normally distributed set of residuals. The iterative fitting process is continued until none of the data fail the edit criteria. The best fit to the oxygen probe model coefficients is then determined. Coefficents were applied by PMEL program CALOX2W and CTD oxygen was computed using subroutine OXY6W. By plotting the oxygen residuals versus station, appropriate station groupings for further refinements of fitting were obtained by looking for abrupt station to station changes in the residuals. Sometimes it was necessary to fix values of some oxygen algorithm parameters to keep those parameters within a reasonable range. Final coefficients were applied by PMEL program EPSBE94. C.1.4. POST-CRUISE PROCESSING SEASOFT consists of modular menu driven routines for acquisition, display, processing, and archiving of oceanographic data acquired with Sea-Bird equipment and is designed to work with an IBM or compatible personal computer. Raw data is acquired from the instruments and is stored as unmodified data. The conversion module DATCNV uses the instrument configuration and calibration coefficients to create a converted engineering unit data file that is operated on by all SEASOFT post processing modules. Each SEASOFT module that modifies the converted data file adds information to the header of the converted file permitting tracking of how the various oceanographic parameters were obtained. The converted data is stored in either rows and columns of ascii numbers or as a binary data stream with each value stored as a 4 byte binary floating point number. The last data column is a flag field used to mark scans as good or bad. The following are the SEASOFT processing module sequence and specifications used in the reduction of P18 CTD/O2 data. DATCNV converted the raw data to pressure, temperature, conductivity, oxygen current, oxygen temperature, and transmissometer voltage. DATCNV also extracted bottle information where scans were marked with the bottle confirm bit during acquisition. ROSSUM created a summary of the bottle data. Bottle position, date, and time were output as the first two columns. Pressure, temperature, conductivity, oxygen current, oxygen temperature, and transmissometer voltage were averaged over a two-second interval (48 scans). For the primary package, the time interval was from five to three seconds prior to the confirm bit in order to avoid spikes in conductivity and oxygen current owing to minor incompatibilities between the Sea-Bird 911plus CTD system and General Oceanics 1016 rosette. Bottle data from the backup package were averaged from one second prior to the confirm bit to 1 second after the confirm bit in the data stream. WILDEDIT marked extreme outliers in the data files. The first pass of WILDEDIT obtained an accurate estimate of the true standard deviation of the data. The data were read in blocks of 200 scans. Data greater than two standard deviations were flagged. The second pass computed a standard deviation over the same 200 scans excluding the flagged values. Values greater than 16 standard deviations were marked bad. SPLIT removed decreasing pressure records from the data files leaving only the downcast. FILTER performed a low pass filter on pressure with a time constant of 0.15 seconds. In order to produce zero phase (no time shift) the filter first runs forward through the file and then runs backwards through the file. ALIGNCTD aligned conductivity in time relative to pressure to ensure that all calculations were made using measurements from the same parcel of water. Alignment between stations was checked every time the CTD configuration changed between primary and secondary underwater packages or every ten stations, whichever was less. CELLTM used a recursive filter to remove conductivity cell thermal mass effects from the measured conductivity. Typical values were used for thermal anomaly amplitude (alpha=0.03) and the time constant (1/beta=9.0). DERIVE was used to compute fall rate (m/s) with a time window size for fall rate and acceleration of 2.0 seonds. LOOPEDIT marked scans where the CTD was moving less than the minimum velocity of 0.2 m/s or travelling backwards due to ship roll. BINAVG averaged the data into 1 db pressure bins starting at 1 db with no surface bin. The center value of the first bin was set equal to the bin size. The bin minimum and maximum values are the center value +/- half the bin size. Scans with pressures greater than the minimum and less than or equal to the maximum were averaged. Scans were interpolated so that a data record exists every decibar. STRIP removed scan number and fall rate from the data files. TRANS converted the data file format from binary to ascii. Following the SEASOFT processing modules, PMEL program POSTCAL corrected conductivity with respect to pressure using an additional beta term, beta = -1.29e-8 for C1247 beta = -5.69e-8 for C1177 c2(i) = (c1(i)*10) + beta * p(i) computed salinity, s(i) = SAL78(c2(i)/42.914,t1(i),p(i),0) corrected temperature due to instrument calibration error, t2(i) = 1.00008961734348 * t1(i) - 9.924374518041036e-4 and backed out final conductivity values. c3(i) = SAL78(s(i),t2(i),p(i),1) c3(i) = c3(i) * 42.914 Also, POSTCAL interpolated temperature, conductivity, oxygen current, oxygen temperature, and transmissometer voltage where values were bad as flagged by SEASOFT before the above corrections and repeated to the surface the first good record input interactively by the user. PMEL program EPSBE94 followed POSTCAL and computed doxc/dt, calibrated CTD oxygens, and computed ITS-90 temperature, potential temperature, sigma-t, sigma-theta, and dynamic height. EPSBE94 also introduced the WOCE quality flag associated with pressure, temperature, salinity, and CTD oxygen. Quality flag definitions can be found in the WOCE Operations Manual (1994). 1 db data were output in EPIC format (Soreide, 1995). Processed data were despiked and values linearly interpolated. WOCE flags were ammended to reflect these changes. D. Acknowledgments The assistance of the officers, crew, and survey department of the NOAA ship DISCOVERER is gratefully acknowledged. Funds for the CTD/O2 program were provided to PMEL by the Climate and Global Change program under NOAA's Office of Global Programs. E. References Benson, B.B. and D. Krausse Jr., 1984 : The concentration and isotopic fractionation of oxygen dissolved in freshwater and seawater in equilibrium with the atmosphere. Limnology and Oceanography, 29, 620-632. Denbo, D.W., 1992 : PPLUS Graphics, P.O. Box 4, Sequim, WA, 98382. Owens, W.B. and R.C. Millard Jr., 1985 : A new algorithm for CTD oxygen calibration. J. Physical Oceanography, 15, 621-631. Seasoft CTD Aquisition Software Manual, 1994 : Sea-Bird Electronics, Inc., 1808 136th Place NE, Bellevue, Washington, 98005. Soreide, N.N., M.L. Schall, W.H. Zhu, D.W. Denbo and D.C. McClurg, 1995 : EPIC: An oceanographic data management, display and analysis system. Proceedings, 11th International Conference on Interactive Information and Processing Systems for Meteorology, Oceanography, and Hydrology, January 15-20, 1995, Dallas, TX, 316-321. Unesco, 1983. International Oceanographic tables. Unesco Technical Papers in Marine Science, No. 44. Unesco, 1991. Processing of Oceanographic Station Data. Unesco memorgraph By JPOTS editorial panel. WOCE Operations Manual, 1994 : Volume 3: The Observational Programme, Section 3.1: WOCE Hydrographic Programme, Part 3.1.2: Requirements for WHP Data Reporting. WHP Office Report 90-1, WOCE Report No. 67/91, Woods Hole, MA, 02543. F. WHPO Summary Several data files are associated with this report. They are the 31DSCG94_2.sum and 31DSCG94_3.sum, 31DSCG94_2.hyd and 31DSCG94_3.hyd, 31DSCG94_2.csl and 31DSCG94_3.csl and *.wct files. The *.sum file contains a summary of the location, time, type of parameters sampled, and other pertinent information regarding each hydrographic station. The *.hyd file contains the bottle data. The *.wct files are the ctd data for each station. When submitted to the SAC, the *.wct files are zipped into one file called *wct.zip. The *.csl file is a listing of ctd and calculated values at standard levels. The following is a description of how the standard levels and calculated values were derived for the *.csl file: Salinity, Temperature and Pressure: These three values were smoothed from the individual CTD files over the N uniformly increasing pressure levels. using the following binomial filter- t(j) = 0.25ti(j-1) + 0.5ti(j) + 0.25ti(j+1) j=2....N-1 When a pressure level is represented in the *.csl file that is not contained within the ctd values, the value was linearly interpolated to the desired level after applying the binomial filtering. Sigma-theta(SIG-TH:KG/M3), Sigma-2 (SIG-2: KG/M3), and Sigma-4(SIG-4: KG/M3): These values are calculated using the practical salinity scale (PSS-78) and the international equation of state for seawater (EOS-80) as described in the Unesco publication 44 at reference pressures of the surface for SIG-TH; 2000 dbars for Sigma-2; and 4000 dbars for Sigma-4. Gradient Potential Temperature (GRD-PT: C/DB 10-3) is calculated as the least squares slope between two levels, where the standard level is the center of the interval. The interval being the smallest of the two differences between the standard level and the two closest values. The slope is first determined using CTD temperature and then the adiabatic lapse rate is subtracted to obtain the gradient potential temperature. Equations and Fortran routines are described in Unesco publication 44. Gradient Salinity (GRD-S: 1/DB 10-3) is calculated as the least squares slope between two levels, where the standard level is the center of the standard level and the two closes values. Equations and Fortran routines are described in Unesco publication 44. Potential Vorticity (POT-V: 1/ms 10-11) is calculated as the vertical component ignoring contributions due to relative vorticity, i.e. pv=fN2/g, where f is the coriolius parameter, N is the buoyancy frequency (data expressed as radius/sec), and g is the local acceleration of gravity. Buoyancy Frequency (B-V: cph) is calculated using the adiabatic leveling method, Fofonoff (1985) and Millard, Owens and Fofonoff (1990). Equations and Fortran routines are described in Unesco publication 44. Potential Energy (PE: J/M2: 10-5) and Dynamic Height (DYN-HT: M) are calculated by integrating from 0 to the level of interest. Equations and Fortran routines are described in Unesco publication 44. Neutral Density (GAMMA-N: KG/M3) is calculated with the program GAMMA-N (Jackett and McDougall) version 1.3 Nov. 94.