README file describing the directory structure of the GTSPP data server 

This file describes the structure of the GTSPP real-time and best copy (BC) 
data.  

1) The ASCII version of the GTSPP real-time data is updated three times a week,
2) The netCDF version of the GTSPP real-time data is updated every Sunday, and 
3) The best copy data are updated on or about the 7th of each month. 

The data are organized as a small set of directory hierarchies. The top level 
of the directory tree contains a README file (this file) and two plain text 
files, gtspp_add.txt and gtspp_rmd.txt, showing the GTSPP database 
identification numbers and names of the individual files are being added or 
removed, respectively.

The sub-trees are:

bestcopy/
  gtspp_add.txt
  gtspp_rmd.txt
  atlantic/
  indian/
  inventory/
  meds_ascii/
  netcdf/
  pacific/
realtime/
  atlantic/
  images/
  indian/
  meds_ascii/
  netcdf/
  odv/
  pacific/
tools/

The following describes the directory structure and file types for the
aformentioned directories:

[realtime,bestcopy] / [atlantic,indian,pacific] / [YYYY] / [MM] /
  These directories contains individual profile (stations) netCDF files for 
  observations from 1990 to present.  The files are divided into subdirectories
  first based on ocean basin, i.e. the atlantic, indian, and pacific, then year
  and month.

bestcopy/gtspp_[add,rmd].txt
  These text files contain 

bestcopy/inventory/
  This directory contains inventory files for each monthly archive. Each
  file lists the call sign of platforms,  the path and the file names of netCDF 
  files, the stations with observations in that month giving the sampling date 
  (year, month, and day), location (latitude and longitude, negative numbers are 
  for southern latitudes and western longitudes), and data min/maxes. Fields with 
  missing min/maxes indicate the data quality is probably bad or bad. 

bestcopy/meds_ascii/
  This directory contains the GTSPP best copy data sets in the MEDS ASCII 
  format.  Each file contains multiple stations based on ocean basin, year, and
  month.  The files are compressed using the unix utility "gzip".  The file
  naming convention is OOYYYYMM.gz, where OO is a two-character ocean code
  (at=atlantic, in=indian, and pa=pacific), YYYY is a four-digit Year, MM is the
  two-digit Month of the year, and ".gz", is a special extension for files
  compressed using the "gzip" unix utility.

bestcopy/netcdf/
  This directory contains the GTSPP best copy data with only TEMP and PSAL data
  records stored in the netCDF format.  The GTSPP netCDF datasets are split into
  three oceans, and are assembled as months of years and compressed, using the
  unix utility "zip". The file naming convention is "gtspp4_OOYYYYMM.zip", where
  OO is a two-character ocean code (at=atlantic, in=indian, and pa=pacific),
  YYYY is a four-digit Year, MM is the two-digit Month of the year, and ".zip", 
  is a special extension for files compressed using the "zip" unix utility.

realtime/images/
  This directory contains station location maps, in GIF format, for realtime
  data.  The file naming convention is "rtm(M)MYY.gif" where MM is the one to
  two-digit month and YY is the two-digit year.

realtime/meds_ascii/
  This directory contains the GTSPP ASCII monthly near real-time data sorted by
  year-month.  The file naming convention is rtm(M)MYY.meds.gz, where YY is a
  two-digit year and (M)M is the one to two-digit month of the year.  Files with
  TXT extensions are the statistics of the corresponding monthly real-time data
  sets.

realtime/netcdf/
  This directory contains the GTSPP weekly near real-time data stored in the
  NetCDF format.  The file naming convention is gtspp4_rt_nc_yyyymmdd-YYYYMMDD.tgz
  where yyyymmdd-YYYYMMDD is the beginning and ending observations dates of a
  given week.

realtime/odv/
  This directory contains the GTSPP weekly near real-time data stored in the
  Ocean Data View file format.  The file naming convention is similar to the
  weekly real-time NetCDF format, gtspp4_rt_odv_yyyymmdd-YYYYMMDD.txt where
  yyyymmdd-YYYYMMDD is the beginning and ending observations dates of a given
  week.