       Bottle and Pumpcast Data from the 1988 Black Sea Expedition

                  MBARI Technical Report No. 90-3
                       Gernot E. Friederich
                        Louis A. Codispoti
                        Carole M. Sakamoto
             Monterey Bay Aquarium Research Institute
                        160 Central Avenue
                     Pacific Grove, CA 93950
                           408-647-3700


     Introduction
     
          This report contains inorganic nutrient chemistry, sulfide and
     oxygen data collected during cruises 2 through 5 of the 1988 Black Sea
     Oceanographic Expedition aboard the R/V Knorr. Continuous nutrient and
     sulfide data were obtained in the upper 375 m using a pumped profiling
     system. Discrete samples were collected from rosette-CTD casts. The
     corresponding physical oceanographic data have been presented by White
     et al. (1989). Although all of the data reported has been edited at
     least twice, errors may remain. We encourage queries and plan to
     distribute updates on electronic media if there are any non-trivial
     changes.
     
     
     Analysts
     
       Cruise 2: G.E. Friederich             14 May  - 28 May  1988
       Cruise 3: G.E. Friederich              3 June - 16 June 1988
       Cruise 4: L.A. Codispoti and          21 June -  8 July 1988
                G.E. Friederich
       Cruise 5: L.A. Codispoti and          13 July - 29 July 1988
                J. Christensen
     
     
     Pump Casts
     
          The profiling pump system was based on a prototype described by
     Friederich and Codispoti (1987). A hydraulic winch was used to deploy
     400 m of cable to which the pump and a Seabird SBE-9/11 CTD were
     attached. The winch was designed to give smooth lowering speeds from
     near 0 to 20 m per min. Tests during very calm conditions indicated
     that lowering speeds less then 6 m per min did not resolve any
     additional water column structure in the region of the highest
     gradients. Consequently lowering speeds of 6-10 m per min were used
     for the profiles presented in this report.
          The center of the pump cable was a continuous nylon hose with an
     internal diameter of 6 mm. The hose was surrounded by a Kevlar
     strength member; electrical conductors for data and power transmission
     formed the next layer. A Dacron braid formed the outer jacket.
          The pump was a stainless steel and graphite positive displacement
     vane pump, coupled to a submersible deep well pump motor. The flow
     rate of about 4 l per min  resulted in a travel time through the pump
     tubing of about 3 minutes. Delay times for the entire system including
     the time for each chemical analysis were derived for each cast. Most
     delay times were obtained by stopping the pump/CTD system for a short
     time during a profile and then matching the plateau generated in the
     chemical profile with that of the CTD pressure readings. A few delay
     times were derived by injecting enriched water into the pump intake
     while it was held in a tank on deck.
          Data was recorded to disk from all chemistry channels and from
     the CTD every three seconds during each cast. All pump data was
     averaged in 1 decibar (db) bins. The complete 1 db data files are
     available on disk. For the purpose of this report data are given in 2
     db increments between the surface and 150 db and in 5 db increments
     below 150 db.
     
     Inorganic Nutrients
     
          Nutrient analysis of both the pump profiles and the bottle
     samples were made using a computer controlled (Hewlett-Packard Series
     85) Alpkem Rapid Flow Analysis (RFA) system. Ammonium, nitrite,
     phosphate and dissolved silicon measurements were made using slight
     modifications of the methods described by Whitledge et al.(1981).
     Nitrate was measured using a slight modification of the method
     supplied by Alpkem which is based on the work of Patton (1982).
          In order to eliminate the interference of sulfide in the analysis
     of phosphate, dissolved silicon and ammonium; bottle samples from the
     anoxic zone were stripped with nitrogen for about 20 min. The amount
     of potassium antimony tartrate in the phosphate analysis was doubled
     in order to reduce sulfide interference. Nitrate levels decreased to
     undetectable levels just above the sulfide bearing waters and it was
     therefore not necessary to measure nitrate in the sulfidic waters.
     Nitrate analysis were occasionally performed on stripped samples from
     sulfide bearing waters near the oxic/anoxic interface to confirm the
     absence of nitrate in sulfidic water.
          During pump casts, the only precautions taken for sulfide
     interference were disconnecting the nitrate channel at about the depth
     were sulfide first appeared and the doubled potassium antimony
     tartrate concentration used in the phosphate analysis.
     Intercomparisons between pump and bottle data suggested that other
     precautions were not necessary for the sulfide concentrations
     encountered during pump casts.
     
     Sulfide
          
          Continuous sulfide determinations were made from the pump stream
     using a continuous flow adaptation of the method developed by Cline
     (1969). Samples from the rosette casts, however, could not be analyzed
     with this method since considerable loss of sulfide occurred in the
     interval between sampling and analysis (< 15 min.).
          
     Oxygen
          
          Dissolved oxygen concentrations in bottle samples were measured
     using the Chesapeake Bay version of the Winkler titration (Carpenter,
     1965). During some pump casts the low concentration colorimetric
     method of Broenkow and Cline (1969) was employed at concentrations
     less than 25 M on discrete samples collected during the pump cast.
     When collecting these samples the pump was held at selected depths for
     about 5 minutes  to ensure complete flushing  of the pump tube.
          
     CTD
          
          The CTD system on the pump package was a Seabird SBE-9/11
     identical to the one used on the rosette system. During cruise 4 and 5
     a Sea Tech transmissometer was added to the pump package. Each CTD
     data record taken during the pump casts consisted of an average of six
     scans taken over a period of 0.25 seconds.
          
     Data Quality
          
          CTD data listed with the bottle casts was taken from the upcast
     of the rosette-CTD system. These data have not been corrected for the
     drift of the CTD as given by White et al. (1988). The above omission
     should cause errors smaller than 0.0025 Deg. C in temperature and
     0.005 ppt. in salinity
          During these cruises leaking bottles and out of sequence tripping
     of bottles on the rosette were encountered. We hope to have eliminated
     most of the data points plagued by these problems.
          The CTD system coupled to the pump was calibrated before the
     cruise but it was not possible to perform a post cruise calibration.
     Comparison with the rosette-CTD system and occasional salinity
     determinations of pumped water on an Autosal salinometer indicated
     that the data are reliable to at least .01 deg. C in temperature and
     .01 ppt. in salinity.
          Due to the large range of phosphate, dissolved silicon, ammonium
     and sulfide concentrations, the detection limit for these methods was
     larger than for more typical marine environments. Ammonium data in the
     oxygenated zone are unlikely to have any meaning since concentrations
     were of the same magnitude as the noise. An additional problem that
     degraded data quality was the interference of sulfide in the analysis
     of phosphate and to a lesser extent in the analysis of dissolved
     silicon and ammonium. The extra manipulation necessary to remove
     excess sulfide from the samples prior to analysis increased the risk
     of contamination and sample degradation. Another problem that effected
     some of the dissolved silicon data was the poor control of laboratory
     temperature which could vary as much as 15 deg. C over the course of a
     few hours.
          At all times at least two independent nutrient standards were
     maintained and intercomparisons of the various standards used during
     these cruises assured the absence of any systematic calibration
     errors. In addition to the  full calibration curves, most analytical
     runs were accompanied by single point standards that could be used to
     correct drift due to such factors as temperature. In a few instances
     dissolved silicon data was normalized using deep samples from the same
     location as internal standards. Sulfide standards were titrated with
     thiosulfate and simultaneously analyzed in the RFA system.
          The estimates for accuracy given in the table below are a best
     estimate based on replicate standards, replicate deep samples,
     comparisons on isopycnal surfaces and our knowledge of various
     analytical problems encountered during these cruises. Richard Mortlock
     provided us with results from his high precision phosphate analysis to
     help in our assessment of data quality. These comparisons suggest:
     
                              Detection      High Range
     Analysis  Range (M)     Limit (M)     Error (+/- M)
     ----------------------------------------------------------------------
     Ammonium  0 - 100           0.5             2
     Nitrate   0 - 10            0.05            0.2
     Nitrite   0 - 5             0.02            0.1
     Phosphate 0 - 10            0.05            0.2
     Silicon   0 - 350           2              10
     Sulfide   0 - 200           1              10
     
     At intermediate values the expected error should be between  the
     detection limit and the high range error. Due to the method of
     assessment the high range error cannot be expressed as a statistical
     quantity, but it should be roughly equivalent to one standard
     deviation.
          Due to the configuration of the pumping system during cruise 5 a
     systematic offset appears to have been introduced to the ammonium
     profiles during cruise 5. The pump profiles on this disc have been             corrected for this offset.

     Acknowledgements
          
          This research was supported by National Science Foundation grant
     OCE-8614400. Financial support was also provided by the Monterey Bay
     Aquarium Research Institute. We thank the crew of the R/V Knorr for
     their assistance during the field work and their help in the repair of
     mechanical problems. C. Goyet assisted with some of the initial
     colorimetric oxygen determinations. A.Gough's secretarial assistance
     is also much appreciated.
          
     References
     
     Carpenter, J.H. (1965) The Chesapeake Bay Institute technique for the
          Winkler dissolved oxygen method. Limnol. Oceanogr., 10, 141-143.
          
     Broenkow, W. W. and J.D. Cline (1969) Colorimetric determination of
          dissolved oxygen at low concentrations. Limnol. Oceanogr., 14,
          450-454.
          
     Cline, J.D. (1969) Spectrophotometric determination of hydrogen
          sulfide in natural waters. Limnol. Oceanogr., 14, 454-458.
          
     Friederich, G. E. and L. A. Codispoti (1987) An analysis of continuous
          vertical nutrient profiles taken during a cold-anomaly off Peru.
          Deep-Sea Research 34, 1049-1065.
          
     Patton, C. J., Doctoral Dissertation, Michigan State University, 1982,
          pp. 87-121
          
     White, G., M. Relander, J. Postal and J. W. Murray (1989) Hydrographic
          data from the 1988 Black Sea Oceanographic Expedition. University
          of Washington School of Oceanography Special Report No. 109.
          
     Whitledge, T. E., S. C. Malloy, C. J. Patton and C. O. Wirick (1981)
          Automated nutrient analysis in seawater. Brookhaven National Lab
          Report 51398, 216 pp.


                                Table of Units

 Hydrocast           Sequential cast number used by White et al.(1989)
 Pumpcast            Sequential pumpcast number
 Latitude            In degrees and decimal degrees
 Longitude           In degrees and decimal degrees
 Date                Greenwich Julian day and decimal day
 Depth               Pressure in decibars
 Temperature         In degrees Celsius
 Salinity            In practical salinity units
 Sigma-t             In kg per cubic meter
 Light transmission  Per cent transmission at 660 nm
 Oxygen              In milliliters of oxygen per liter at NTP
                     or micro molar (ml O2/l = .0224 x M)
 Low Conc. Oxygen    Micro molar
 Phosphate           Micro molar reactive phosphorous
 Dissolved Silicon   Micro molar
 Nitrate             Micro molar (corrected for nitrite)
 Nitrite             Micro molar
 Ammonium            Micro molar
 Sulfide             Micro molar
 Null values         -999 (any 0 in %transmission also indicates no data) 
