CEAREX Hydrography Data


Data Description
Data Acquisition 
Data Processing
Data Organization 
Data Format 

The CEAREX hydrography data include conductivity-temperature-depth (CTD) data collected during two cruises and from two camps, and data obtained during the Seasonal Ice Zone Experiment (SIZEX) and the Eurasian Basin Experiment (EUBEX). The hydrography data files also include data taken from the Fram Strait 11-Year CTD Data Set, bottle data obtained from a National Oceanographic Data Center data base, and ship data acquired from a merged data set of 300 Norwegian and Greenland Seas stations.

Information in this document has been derived from documentation files on NSIDC's CD-ROM titled Eastern Arctic Ice, Ocean and Atmosphere Data, Volume 1: CEAREX-1, version 1.0, 8/91.

Data Description:

Described here are the SIZEX data, the Fram Strait data, the bottle and ship data, and the data processed at the Lamont-Doherty Geological Observatory.


The Lamont data were collected during the POLARBJORN drift phase. Three data files contain 208 calibrated, decimated CTD hydrographic stations. Time is in GMT; all latitudes are in degrees north; all longitudes are in degrees east; temperature is in degrees Celsius; pressure is in decibars; salinity is in parts per thousand (PSS78 units); conductivity is in millimohs.


The SIZEX data include remote sensing, oceanographical, ocean acoustical, meteorological, and sea ice data.

Fram Strait

The Fram Strait data consists of all readily available hydrographic information obtained from salinity-temperature-depth (STD) or conductivity-temperature-depth (CTD) profiling instruments north of 76 degrees north and within the region of Fram Strait. Data from 4,114 stations comprise the Fram Strait data file. The data consists mostly of spring, summer and fall measurements. A table in the CD-ROM documentation file identifies the 26 experiments undertaken between 1977 and 1987 whose data have been collected into the Fram Strait data file.

Bottle and Ship

The bottle data represents 1549 stations obtained through a search of a copy of the NOAA/National Oceanographic Data Center (NODC) data base. No quality control measures were applied to the data after selection. The original investigators' and/or institutions' quality control measures are documented at NODC. Examination of the data shows very few stations with the quality, resolution, or range of parameters expected from modern observations.

The ship data files comprise a special merged data set of Norwegian Sea and Greenland Sea stations from 1980 through 1984, having relatively good data quality. Data are in five files. All station times are missing from the data files.

Data Acquisition and Equipment:

Described here are the data acquisition methods and equipment for the Lamont, Acoustics Camp (A-Camp), Oceanography Camp (O-Camp), EUBEX, SIZEX (including the Seasoar Program), and the bottle and ship data.


The Lamont data were collected during the first three legs of CEAREX (PB1, PB2, and PB3) aboard the ship POLARBJORN:

     LEG   Dates of Stations      Chief Scientist
     PB1   10/17/88 - 12/14/88    R. Pritchard
     PB1   12/15/89 - 01/08/89    J. Ardai
     PB2   01/14/89 - 02/01/89    E. D'Asaro
     PB3   02/09/89 - 03/01/89    O. Johannessen

Bottom depths for the hydrographic stations are from the ship-mounted echo sounder. The values are in uncorrected meters and should be used as estimates. When the echo sounder was inoperable the following alternative methods were used:

     Leg     Stations             Source of Depth Data
     PB1   1, 2               General Bathymetric Chart of the Oceans 
     PB2   5-7                (GEBCO), 5th edition (corrected meters)
     PB1   60-104             Ship's charts and winch wire counter
     PB3   10-11, 52, 45-60   ADCP

Raw CTD data were collected while the CTD unit was lowered (downtrace) and raised (uptrace). The data values presented here are from the downtrace except occasionally when the downtrace values were contaminated and unusable; in these cases the uptrace values were processed. At the following stations, uptrace data were used at 12 PB1 stations, 17 PB2 stations, and 14 PB3 stations. Uptrace data tend to be lower quality than downtrace data.

The navigation available when this data set was assembled was the original navigation collected on the POLARBJORN by ERIM. Logging mistakes and gross navigational errors in station positions have been corrected using this raw navigation and the CTD log sheets. All station locations should be re-evaluated if the navigation is processed further.

A Neil Brown IIIB CTD underwater unit (serial number 01-2276-01) was used to collect the Lamont data. The instrument was calibrated before the start of CEAREX on March 22, 1988 by the Naval Oceanographic Office. A total of 208 stations were collected between October 10, 1988 and March 1, 1989 (PB1 through PB3). No post-cruise processing by the Naval Oceanographic Office was possible because of damage to the conductivity cell at the start of PB4. A post-cruise calibration of pressure only was run for the instrument on August 22, 1989 at Lamont-Doherty Geological Observatory.

Continuous cruise calibration of the conductivity cell was possible using Autosal salinity measurements from water samples collected by Niskin water bottles. The conductivity data were calibrated separately for each leg of the cruise. The reversing thermometers provided to calibrate the CTD thermometer were of poor quality and their results were not used in the temperature calibration. A total of 234 bottles were available for use with the 208 stations. The lack of bottle data has caused the calibration errors to be larger than is desirable. Data values within approximately 5 meters of the surface are to be used with caution. The upper water layer was often contaminated by the ship. In some cases the surface values have been extrapolated to allow all stations to start at zero meters.

Please refer to the lists of calibration results and equations contained in the CD-ROM documentation file.


The CTD data were acquired from a drifting ice camp (A-Camp) and from a helicopter that was also based out of the ice camps. Muench et al. (1991) describe the helicopter CTD data. The data from the A-Camp (cruise designator ICE) and the helicopter (cruise designator CX2) were obtained using SeaCat portable, self-contained CTD systems. Some salinity spiking was encountered with the SeaCat systems in regions of strong vertical temperature gradient, and this was most pronounced in portions of the A-Camp data (cruise ICE), where lowering rates were maintained at 0.5 m/s (a slow rate) in an effort to minimize the problem. Each of the CTD systems was calibrated prior to and following the field program.

Two SeaCats used at the A-Camp (ICE) were intercalibrated through simultaneous casts during the program, though loss of one of the two instruments precluded such intercalibration later in the program. The helicopter-borne SeaCat (CX2) systems were intercalibrated periodically at the O-Camp. Except where severe spiking was present, accuracy of the observations is +/- 0.01 degree C and 0.02 psu.


The O-Camp daily cast data were collected from the Oceanography Camp via a "yoyo" CTD system. The yoyo CTD system was cycled nearly continuously down to 400 meters. Typically a single cast was made each day to 600 meters. Near the very end, the continuous casts went deeper than 400 meters.

Data were also obtained from a helicopter (cruise designator CX3) using a SeaBird model SBE/11 CTD system, which had a pumped conductivity cell and was not subject to salinity spiking. The CX3 data constitute a small subset of the total volume of CTD data obtained from O-Camp, and is integrated with that data on the CEAREX CD-ROM.


EUBEX data were collected from 35 CTD stations near Spitsbergen between March 8 and April 17, 1981. The surveys were conducted using aircraft and from the sea ice surface, and extended down to 1 km where the water was deep enough to allow this. A table in the CD-ROM documentation file provides location information for each of the stations.

Three CTD casts taken off Greenland are included in a separate EUBEX data file. For more complete information on the EUBEX CTD data, please refer to Lewis and Perkin (1983) and Perkin and Lewis (1984).


The SIZEX and related MIZEX data were collected using various observational platforms such as ice-strengthened ships, open ocean ships, drifting buoys, bottom-moored buoys, helicopters, and satellites. SIZEX data were collected in the Barents Sea and the Greenland Sea in February and March 1989.

On February 8, the POLARBJORN departed Tromso heading towards the Barents Sea. Between Fugloya and Bjornoya, 12 CTD stations were obtained. The ice edge east of Bjornoya was encountered on February 11 when one acoustic buoy was deployed in ice-covered area at 75 degrees 04 minutes north, 23 degrees 00 minutes east. The next day one current meter rig and a second acoustic buoy were deployed in open water at 74 degrees 57 minutes north, 29 degrees 15 minutes east. Then the POLARBJORN headed northeastward towards the Hopen area.

Between February 13 and February 15, ten Argos buoys were deployed on ice floes in an area approximately 100 km by 100 km. Five of these buoys were toroids with current meter strings. The deployment area west of Hopen was selected because the expected drift of the array was southwestward towards Bjornoya where the array could be recovered later. In this period the ship had to plow through fairly heavy ice, much of it was multiyear up to four or five meters thick, and she could move only a few miles per day. Weather conditions were good with four to six hours of daylight, and the helicopter could be used almost every day for ice reconnaissance and buoy deployment.

Between February 17 and February 24, the ship drifted southwestward with the pack ice, or moved only slightly, while the different groups collected data. Every time the ship was towed to a floe, radar and in situ measurements of snow and ice were made, sessions of ADCP data and wave data were obtained, and vertical and horizontal acoustic arrays were deployed. Ice photographs were obtained using helicopter, and other in situ measurements for SAR calibration were carefully coordinated with the SAR flights. Sonobuoy deployments were also coordinated with the Norwegian P3 flights from Andoya on February 18 and February 27. One iceberg, which was approximately 200 m by 100 m and grounded at 50 m depth, was visited for in situ measurements. Meteorological and remote sensing data were collected regularly throughout the experiment.

From February 24 through February 27, all the toroids were recovered, partly by use of helicopter since the ship could make only slow progress in the pack ice. Three small Argos buoys were left in the area to continue monitoring the ice drift. On February 27, wave/acoustic studies had first priority with one dedicated SAR flight, deployment of a vertical acoustic array, and a wave buoy near open water.

Barents Sea operations were very successful. In ice-covered areas, the weather was very good most of the time. All important instruments were recovered, and a lot of interesting data were collected. The POLARBJORN arrived in Tromso on March 4. After repair in a shipyard in Harstad and a cargo outhaul to Longyearbyen, the POLARBJORN was ready for the Greenland Sea operations. On March 11, she left Longyearbyen and headed southwestward to occupy deep CTDs in the acoustic tomography array in the Greenland Basin in cooperation with the HAKON MOSBY. Between March 13 and March 16, only four deep CTD casts were obtained in this area because there were problems with both the winch and the CTD sonde. The HAKON MOSBY had to be called for technical assistance before the program could be continued.

Between March 17 and March 28, the weather conditions were quiet and operations were carried out in the Boreas Basin at about 78 degrees north. The CTD system functioned normally after the repair and 70 CTD stations were obtained in this period. Only two toroid buoys with current meters and four other Argos buoys were deployed. The study of eddies, deep convection and chimneys using the CTD and water samples was emphasized. Therefore less time was spent on buoy deployment and more on CTD work compared to the Barents Sea. Meteorological and remote sensing data were collected regularly, and snow/ice measurements and acoustic data were obtained in between the CTD casts. A dedicated acoustic experiment was carried out on March 27 and March 28. Seven SAR flights over the area by NADC P-3 were completed in this period.

From March 28 through April 1, two strong storms passed the area and weather conditions changed from moderate to rough. Of highest priority in this period was the recovery of the last toroid, which had five current meters and one thermistor chain. It had been deployed from a multiyear floe at about 78 degrees 25 minutes north, and during a seven day period it drifted southward with the East Greenland current. Fortunately, the toroid survived the storms and was successfully recovered at 76 degrees 30 minutes north on March 31. After one more deep CTD cast in the Greenland Basin, the POLARBJORN headed back to Tromso where she arrived on April 2.

Two of the drifting Argos buoys were caught by eddies and circulated in the Boreas Basin for about three weeks after the ship left the area. One buoy drifted south to about 73 degrees north where data transmission stopped on May 18.

The Greenland Sea leg was successfully completed and much interesting data were obtained. Ice conditions, ocean conditions and weather are different in the Greenland Sea compared to the Barents Sea. Therefore the data sets from these two areas complement each other.

SIZEX CTD and Seasoar Program

In the Barents Sea, a total of 55 CTD stations were completed. The water masses in the shallow area between Bjornoya and Hopen (40 m to 60 m depth) are dominated by homogeneous cold polar water with temperature around -1.8 degrees C and salinity of 34.7 parts per thousand. In the western part of the experiment area, some intrusion of warm and saltier Atlantic water below 50 m was observed. In the Greenland Sea, 75 CTD stations were obtained, 13 of which were deeper than 2000 m. The deep CTD casts were made in cooperation with the HAKON MOSBY to study deep convection and bottom water formation. The shallow casts (500 m) were made to map eddies and upper ocean chimneys in the Boreas Basin. (Johannessen et al. 1991).

Bottle and Ship

The bottle data were obtained by searching a copy of the NOAA/National Oceanographic Data Center (NODC) data base, June 1990 version, held by Joe Reid. The search was performed using the following coincident criteria:
< >maximum observed (sampling) depths greater than or equal to 200 meters
latitudes above 80 degrees north, or latitudes between 70 degrees north and 80 degrees north
with longitudes between 120 degrees west and 180 degrees or 180 and 100 east
both temperature and salinity data at the same station
salinity values reported to at least two decimal places

the YMER 1980 Fram Strait and northern Barents Sea slope expedition
the KNORR 1981 expedition for the North Atlantic Study of the Transient Tracers in the Ocean(TTO-NAS)
the HUDSON 1982 winter expedition
the METEOR 1982 spring expedition
the POLARSTERN 1984 post-MIZEX expedition in and near Fram Strait

Data Processing:

Due to the predominance of interleaving of the various water masses within this region, virtually all of the profiles displayed density inversions and spiking over a rather wide range of amplitudes (1 m through 10 m). Since these characteristics were undesirable for the intended use of the data, a variable-knot cubic-spline smoothing algorithm was used to remove all density inversions and spikes from the data while still preserving structure having vertical amplitudes greater than 20 meters. This algorithm consecutively fit a series of cubic splines with continuous first and second derivatives over the entire profile. More splines were used at the beginning of the profile to insure better fit of the thermal layering. Subsequent verification procedures were incorporated into the processing to insure a closeness of fit to the original profile and to insure that no inversions were present. Additionally, the mixed layer was removed from the smoothing process since it was important that the data maintain the original observational values as well as prevent the modification (smearing) of the base of the mixed layer. Data were subsampled every 5 meters and then truncated at a maximum depth of 800 m to produce the final data. The Lamont CTD data were processed at Lamont-Doherty Geological Observatory. All data were filtered with a median filter using a window of five scans after which the correction coefficients were applied for each sensor and a 0.25-second lag in temperature/pressure was applied to account for response time. The final data set was then averaged into 1 m bins.

Data Organization:

The hydrography data are presented in 49 data files in nine subdirectories on the CEAREX CD-ROM, one each for:

Lamont data

Bio-physical Cruise data

Acoustics Camp data

Oceanography O-Camp daily cast data

EUBEX data

SIZEX data

Fram Strait data

Bottle data

Ship data

Data Format:

At the Scripps Institution of Oceanography, Oceanographic Data Facility, all CTD files were converted to the "S87" standard CTD data format. The S87 format data files were then provided to the National Snow and Ice Data Center (NSIDC), converted to fixed-length records, and written to magnetic tape for CD-ROM mastering. Each record contains 64 characters; the end of each data record is padded to 64 characters with blanks, and the 64th character is a "newline."

The S87 data format was developed at the Lamont-Doherty Geological Observatory as a standard format for ASCII station data. The main parts of the S87 format file are:

< >the header line(s)optional identification line(s) describing the datathe data lines

The header line contains all the pertinent station information needed to identify the station. When the file contains data for more than one station, the header line may be repeated within a single file.

0        1         2         3         4         5         6     { column
1234567890123456789012345678901234567890123456789012345678901234 { counter

        T         = data type (C: ctd, B: bottle, A: axbt, X: xbt)
        PP        = NODC platform code
        CC        = NODC country code of the platform
        SSSS      = station number
        CC        = cast number
        SDD.DDDD  = latitude in decimal degrees
        SDDD.DDDD = longitude in decimal degrees
        YY/MM/DD  = date (including "/")
        JUL       = year-day for year of collection (also called "Julian Day")
        HH:MM     = time (including ":")
        CRUISE_ID = optional cruise identifier.

Following the header line is an optional secondary header line for end-of-cast information. There may also be another line describing important physical characteristics at the station location. This line must begin with the character "&" in the first column. For a list of physical characteristics mnemonics in use when the data set was assembled, please refer to the CD-ROM documentation file.

A sample optional "&" line is as follows:

	&ZZ=4766   TA=-4.2   PA=0990   WS=0.6   WD=122

Comment lines follow this optional "&" line; they may not begin with "&" or "@". Use these comment lines to note the name of the program used to write the data file, the date the file was written, and the name of the programmer.

A column identification line contains mnemonics of at least two unique characters that identify the data types. This line must start with the character "@" in the first column. For a list of data type mnemonics in use when the data set was assembled, please refer to the CD-ROM documentation file.

The following is an example of data that illustrates all parts of the S87 format:

	CPB32   55  1  74.4490   19.5095 89/03/01  60 17:56 PB3
	&ZZ=4766   TA=-4.2   PA=0990   WS=0.6   WD=122
	90/02/01  sohara  program: s87interp -i 1
@PR TE CO SA PT S0 0 -1.877 27.375 34.912 -1.877 28.112 1 -1.877 27.375 34.912 -1.877 28.112 2 -1.877 27.375 34.911 -1.877 28.111 3 -1.877 27.375 34.910 -1.877 28.110 4 -1.877 27.375 34.910 -1.877 28.110 5 -1.877 27.375 34.909 -1.877 28.109 6 -1.877 27.375 34.908 -1.877 28.109 7 -1.878 27.383 34.920 -1.878 28.118 8 -1.876 27.384 34.919 -1.876 28.118 9 -1.873 27.385 34.916 -1.873 28.115

All Fram Strait files are named *.jms, where the "*" represents the name and year of the platform, and the ".jms" extension tags the files as relating to Manley et al. (1991). To avoid duplicate filenames, the name and year of the experiment were used instead in some cases. Please refer to the CD-ROM documentation file for a table of Fram Strait filenames and the associated "Cruise_id" value found in positions 54 through 62 of each file's header records.

The CEAREX hydrography data are also archived in the original format (as provided to Scripps by each investigator) and are available on magnetic media from NSIDC User Services. Please inquire for current distribution format(s) and cost.


Data accuracy varies with experimental program and the types of sensors used, but for the data set as a whole, the accuracy estimates are +/- 0.02 degree C and +/- 0.02 psu for temperature and salinity, respectively. Residual (smoothed - original) standard errors were used as an indication of the quality of fit between the smoothed and original profiles of temperature, salinity and density. Less than 10 percent of the station data had residuals greater than +/- 0.01, but they were still within the limits of the data set reliability (i.e., +/- 0.02).

Errors related to the hydrography CTD data have been identified.

SIZEX data descriptions in this document were extracted from the SIZEX portion of the CD-ROM documentation file, which was taken from Johannessen and Sandven (1989). Lamont data descriptions in this document were derived from the Lamont section of the CD-ROM documentation file, which was taken from O'Hara (1990).

Please refer to the CD-ROM documentation file for the abstract from:

Manley, T. O., R. H. Bourke and K. L. Hunkins. 1991. Near-surface circulation over the Yermak Plateau in northern Fram Strait. Journal of Marine Systems 3(1/2):107-125.


Roger Andersen, University of Washington - O-Camp data
Ola Johannessen, Nansen Remote Sensing Center - SIZEX data
Thomas Manley, Marine Research Corporation - Bio-Physical Cruise data and Fram Strait data Set
James Morison, University of Washington - O-Camp data
Robin Muench, Science Applications International Corporation - A-Camp data
Suzanne O'Hara, Lamont-Doherty Geological Observatory - Lamont data
Ronald Perkin, Institute of Ocean Sciences - EUBEX data
James Swift, Scripps Institution of Oceanography - bottle and ship data

For a complete list of all CEAREX investigators, please refer to the CEAREX Investigator Address List.