The LMG data acquisition systems continuously log data from a suite of instrumentation throughout the cruise. This document describes the format of that data and its location on the distribution CDs. It also contains important information which may affect how this data is processed such as instrument failures or other known problems with acquisition.

The data collected during this cruise is distributed on a CD-ROM written in ISO9660 level-1 format. This data format has very strict requirements on filenames and organization. However, it is readable by virtually every computing platform.

All of the data has been archived with the Unix "tar" command and/or compressed using Unix "gzip" compression. Tar files have a ".tar" extension and Gzipped files have a ".gz" extension. Tools are available on all platforms for uncompressing and de-archiving these formats. On Macintosh, Stuffit Expander with DropStuff will open a tar archive and uncompress gzipped and Unix compressed files. For Windows9X, WinZip, a shareware utility included on this CD (remember, it is shareware) will open these files.

In some cases to adhere to the ISO9660 format the .tar extension was removed. When we tarred the files then gzip the tar archive the name of the file became File.tar.gz. This name does not follow the 8.3 naming convention of the ISO9660 format. So the file was renamed without .tar making the file name File.gz. On Windows and Mac Platforms Winzip and Stuffit Expander handles this just fine. When they expand the File.gz the expanded file becomes File.tar, which both software packages can handle. On Unix platform gunzip expands File.gz but it does not append the .tar extension. So you may not recognize the file as a tar archive, but OS does recognize it as a tar archive. If you use the file command it will return saying it is a tar file. The below tar command will un-archive the file just fine.

It is often useful to know exactly how an archive was produced when expanding its contents. Tar files were created using the following commands:

To create a list of the files in the archive:

To extract the files from the archive:

G-zipped files will have a ".gz" extension on the filename. These files can be decompressed after de-archiving, using:

Disk1/Adcp/

The ADCP DAS data files are named pingdata.xxx (xxx representing a file number). Note that these extensions do NOT represent Julian day numbers. Please refer to the file’s creation date. The ADCP DAS computer creates a new pingdata file when the current one reaches a size of 320K. The ping files logged on Gentoo, the Linux processing computer, however are created new each day.

Some ADCP data is also transmitted to RVDAS. East and North vectors for ship’s speed relative to the reference layer and ship’s heading are archived in the navigational data section of RVDAS.

There was one problem that occurred with the ADCP. When we left Palmer Station at the end of the cruise, we started the ADCP DAS collection again. For some unknown reason it started writing to pingdata.023 instead of pingdata.028. We were unsure if it was overwriting the file or just appending. So we moved the pingdata files with extensions 000-027 and started over with a new set of pingdata files beginning with pingdata.000. The first set of data is in the ADCP/LEG1/ directory and the second set is in the ADCP/LEG2 directory.

Disk1/Cal/

The tar files in the Cal directory contain images of calibration sheets for each of the following systems: Sound Velocity Probe(SVP_CALS.TAR), Meteorological System(MET_CALS.TAR), Underway System(UW_CALS.TAR), and CTD(CTD_CALS.TAR).

Disk1/CTD/

The ctd data was collected and processed on a windows 98 computer, using Seasave Win32 – Version 5.25a and SBE Data Processing – Version 5.21.

See the Readme.txt file for cast notes.

In the Setup directory there is a Config file in html, excel, and text form which contain information of which sensors where used and what freq or volt the where connected to. The file also contains a table with the vertical distance in meters from the pressure port that each sensor was mount. The distances are positive as pressure increases. The sensor connections are also shown in the two diagrams of the fish’s end caps: Freq.jpg and Volt.jpg. Also in this directory are jpeg images of the CTD rosette. They also show how the Rothera seacat was mounted, which was done for one cast for calibration. We used three different con files through out the cruise, which are in this directory. Lmg0301.con was used for cast 1 to 16, event 007 to 122. The oxygen sensor failed so we had to replace it, so lmg0301a.con was created with updated cals for this new sensor. It was used for cast 17 to 60, event 137 to 525. We then decided to add the SPAR to the CTD data set, so a third con file was created with the new cals added, lmg0301b.con. It was used for the remaining casts 61 to 93, event 534 to 729.

This directory contains the Pscript.tar archive, which contains the batch file and psu files that we used for post processing the data. The data was processed with the standard seabird processing method. This is just a preliminary processing which was done to verify that the sensors were functioning properly during the cruise. The raw data should be re-processed using the pre and post cruise sensor calibrations.

The raw.gz file is a tar archive file that has been compressed with gzip, for more information on this see the above Introduction section. This archive contains the raw file collect at each CTD cast, which is represented by a set of four files containing a bottle-firing file (.bl), a configuration file (.con), a data file (.dat) and a header file (.hdr). Casts are named with the following g301EEE.ext, where g is for the LMG, 301 is the cruise 0301, EEE is the event number, and ext is the extention (bl, con, dat, hdr). For example; the raw files associated with the event 007 on this cruise are: g0301007.bl, g0301007.con, g0301007.dat, g0301007.hdr. The raw data files(*.dat) are binary files.

The process.gz file is a tar archive file that has been compressed with gzip. For more information on this see the above Introduction section. This archive contains the processed data files for each CTD cast, the processing method used is briefly described in the above section CTD/Pscripts/. Also see the above section CTD/Data/raw for a description of the file naming convention used. Each processed cast is represented here by a set of ten files:

The graph.gz file is a tar archive file that has been compressed with gzip, for more information on this see the above Introduction section. This archive contains three postscript files for each cast, which are plots of the processed CTD data. The graphs were generated with a CShell script written by Fred Stuart, which uses the GMT, General Mapping Tool, software package. The file naming convention is TTT.LMG0301.EEE.ps where TTT represents the graph type( Pri, Sec, Other) and EEE is the event number for the cast. The Pri graph type is a plot of the primary sensors, the Sec graph type is a plot of the secondary sensors, and the Other graph type is a plot of all the other sensor on the CTD. These files were use to compare the data from cast to cast to make sure that all the sensors were working properly.

The data in this directory was collected using the Palmer Station seacat from a zodiac. The cast were done on the Foraging Grid near Palmer Station, see the readme.txt in this directory for more information on the Foraging grid location. The subdirectories here are the same as describe in the above CTD sections, except for the naming convention see the readme.txt file to correspond the file name to an event number, latitude, and longitude.

disk1/SALT/

Salts were collected during the southbound Drake crossing of LMG03-01 as per the standard routine established by Janet Sprintall, using the ship’s underway system. Additional samples were taken during the cruise to calibrate the CTD. Approximately six cruise samples were collected once per day from the rosette niskin bottles. Most water was sampled from the mixed regions (typically the bottom several bottles, though we did have a few surface mixed layers).

Samples were analyzed every couple of days, allowing room temperature to be reached and for multiple samples to be run at the same time for conservation of standard seawater and technician sanity.

The salinity was determined from use of equation 1 and the AutoSal’s "2 x conductivity ratios". Readings were taken until three replicates within +/-0.00005 were attained. The salinity was then determined from the mean of these three replicates (as shown in the electronic data sheets).

Equation 1. a₀ = 0.0080, b₀ = 0.000; a₁ = -0.1692, b₁ = -0.0056; a₂ = 25.3851, b₂ = -0.0066;

For calibration/processing, the salts spreadsheet includes the salinity values from primary and secondary sensor on the CTD, and each of these values subtracted from the salinity determined via the AutoSal. The final column also lists the values of the two CTD salinities subtracted from each other.

All cruise salts from the CTD were collected and analyzed by the Marine Science Technician, Jordan Watson. The machine appeared to function well. The Guildline AutoSal protocol suggests that during the DI rinses in between samples, a conductivity reading of less than 0.00050 be attained before continuing to the next sample. However, after the first several runs, I switched this number to 0.00020 as my personal guideline and seemed to achieve greater precision with a fewer number of required replicates, at the expense of extra DI rinses instead of extra sample rinses. I believe that this took extra time, but achieved better results.

disk1/MAP/

PostScript cruise track plots have been produced for this cruise, in two sizes lter8in.ps and lter18in.ps. The lter8in.ps is 8.5"x11" and the lter18in.ps is 18"x24". The lter8in.ps has been converted to a jpeg format also. Additionally, the cruise track file track.asc is also included and contains the latitude and longitude in one-minute intervals.

disk1/PUV-GUV/

The GUV files are in both binary and ASCII format they have been zipped using pkzip and called GUV_LOG.ZIP and GUV_CSV.ZIP respectively.

The GUV log files were converted to CSV (Comma Separated Value) format using the Biospherical Instruments’ PUVLOG program.

The columnar format of the CSV file is listed below:

disk1/Isobars/isobar.tar

The isobar.tar file contains GIF image file. These file are an analysis of mean sea level pressure from the National Center for Environmental Prediction's Medium Range Forecast Model. They are updated every 6 hours. Naming the convention is as follows yyjjj.hh.gif where yy is the year, jjj is the day number, and hh is the hour.

disk1/Report/

Copies of this report in MS Word, HTML, and text formats.

disk1/Weekly/

Copies of weekly activity reports from the Chief Scientist, Robin Ross.

Disk1/QC_PLOTS/

Postscript files of data stored each day on RVDAS for quality control analysis during the cruise. There are 3 types of files, named metXXX.ps, navXXX.ps, and oceanXXX.ps, where XXX is represents the Julian day. Met files are a summary of the data from the meteorological instruments, Nav files are a summary of navigational data, and Ocean files are a summary of the underway seawater and bathymetry data.

disk1/XBT-XCTD/SBCross/XBT

During the Southbound crossing Expendable Bathythermographic (XBT) probes were used to obtain water column temperature profiles. These XBT were launched from the auto-launcher off the port aft quarter of the ship. The data files from these launches are included here in the SFILES.ZIP. The NAV.ZIP file contains the navigation files. The LOG.ZIP contains log files generated by the auto-xbt software. The DAT.ZIP file contains the configuration file used and generated by the auto-xbt software. The handwritten logs take during the transect were scanned in and saved as .jpg files (XBTlog1.jpg, XBTlog2.jpg, XBTlog3.jpg). For more information contact Glenn Pezzoli, project coordinator, at gpezzoli@ucsd.edu.

disk1/XBT-XCTD/SBCross/XCTD

During the Southbound crossing, Expendable Conductivity Temperature Depth (XCTD) probes were used to obtain water column conductivity and temperature profiles. These XCTD were the analog type, and where manually launch from tube 1 of auto-launcher. The data files from these launches are included both in binary (RDF.ZIP) and ascii (EDF.ZIP) format. The logsheet (XCTDLog) has been saved as three different formats excel spreadsheet, HTML, and text file.

disk1/XBT-XCTD/LTER

During the cruise, at various stations, Expendable Conductivity Temperature Depth (XCTD) probes were used to obtain water column conductivity and temperature profiles. Analog and digital XCTD were used. The analog probe can goto depth of 1850m and the digital to 1100m, so the analog probes were use at stations deeper than 1100 meters. The XCTDData.gz file contains both the binary and ASCII file for each digital XCTD drop. The XCTDDeep.gz file contains both the binary and ASCII file for each analog XCTD drop. The log sheets were scanned in as jpegs as XCTDLog#.jpg where # is a integer corresponding to the log sheet page number.

disk1/JGOF/

Additionally, 3 separate QC plots are generated daily by the ET using the JGOFS data set. These plots include TSG and Bathymetry data, meteorological data, and navigation data. The files are called ocean<julian_day>.ps, met<julian_day>.ps, and nav<julian_day>.ps respectively.

disk2/rvdas/

RVDAS (Research Vessel Data Acquisition System) was developed at Lamont-Doherty Earth Observatory of Columbia University and has been used on the R/V Maurice Ewing for several years. It was adapted for use on the Nathaniel B. Palmer and her sister ship, the R/V Laurence M. Gould.

Below you will find detailed information on the data included. Be sure to read the "Significant Acquisition Events" section below for important information about data acquisition during this cruise.

RVDAS data is divided into two broad categories, Underway and Navigation. The groups are abbreviated "uw" and "nav". Thus, these two tar files, lmguw.tar and lmgnav.tar exist under the top-level rvdas directory. The instruments are broken down as shown. Each data file is g-zipped to save space on the distribution. Not all data types are collected everyday or on every cruise.

RVDAS data files are named following the convention: LMG[FileID].dDDD.

• The FileID is a 4-character code representing the system being logged, for example: lmet (for meteorology)
• DDD is the Julian day of the data collection

Data is received by the RVDAS system via RS-232 serial connections. The data files that comprise the rvdas data set are described below. A time tag is added to each line of data received and the data is written to disk.

Where, YY: two-digit year, DDD: Julian Day, HH: 2 digit hours, MM: 2 digit minutes SS.SSS: seconds. All times are UTC.

The delimiters used to separate fields in the raw data files are usually spaces and commas, but other delimiters are used (::, =, @) and occasionally there is no delimiter. Care should be taken when reprocessing the data that the fields separations are clearly understood. An example data

99+099:00:18:19.775 hf,305.2,lf,304.3

02+314:23:59:50.067 01.2 047 028 01.3 063 042 0988.8 001.7 084 -000.2192 0000.9358 0025.5875

02+310:23:57:30.200 8542 -1.2580 34.1740 -1.2030 3.435 0.349 27.361205

02+319:23:59:13.748 2002319.99851 7154.27 26.49 1033.6 325.79 6.74 329.3 53.76 0 Equil

00+348:01:59:52.128 1539.40

00+019:23:59:59.099 $PUHAW,UVH,-1.48,-0.51,250.6