Objective: To standardize methodologies of ARGO data centers and to document methods in an ARGO Data Management Handbook

Chair: Sylvie Pouliquen, Claudia Schmid
Co-chair: Bob Molinari

Format: Volunteer discussion leaders will review the particular issue, lead a discussion and summarize conclusions. Meeting attendees are expected to contribute ideas on issues before and during the discussion to the leaders.

 3 October

0900: Local greetings
0915: Summary of meeting objectives, Chair and co-chair
0930: Data products: Fabienne Gaillard

 Issues:
> examples from participants
> regional versus global

1045  Break

1115: Discussion of data formats: Leader, Gilbert Maudire

Issues:

> meta-file i) standardize definitions
                  ii) develop look-up tables

 > header data i) meta-file items to be included
                        ii) other data (e.g., positions, time, etc.)
                        iii) unique profile id

 > data i) flags, errors
            ii) two types of real-time data; (1) GTS (i.e., salinity only) and
                post-GTS, pre delayed mode qc (i.e., S and conductivity)
            iii) differences between real and delayed mode

1215: Lunch

1400 : Discussion of data formats (continued)

1530 Break

1600: Quality control and flags: Leader, Bob Keeley

 Issues:

> real time tests and flags
> delayed mode tests and flags
> degree of standardization
> compulsory versus optional tests

1700: End first day
=======================================

4 October

0900: Quality control and flags(continued): Leader, Bob Keeley

1030 Break

1100: Data transmission and access: Etienne Charpentier, Sylvie Pouliquen

 Issues:
> GTS formats (including resolution)
> ftp/www/e-mail

1400: Data Exchange between Centers: Sylvie Pouliquen

 Issues:
> formats (ASCII, netCDF)
> meta-files
> Real time versus delayed mode data
> Mirroring strategy
> how and where are original data stored?

1530: Break

1600: Trajectory data: Claudia Schmid

 Issues:
> quality control
> accepted real-time position

1730: End of second day
====================================

5 October

 0900: Data Archiving : Sylvie Pouliquen

 Issues:
> Real time versus delayed mode (replace real-time with delayed mode profiles?)
> Multiple delayed mode profiles resulting from several qc iterations of the same profile (e.g., PI,    basin-wide, re-analyses, etc.)

1030: Break

1100: Remaining issues