It's better for the users to access a unique source of data (in this case, we have two for reliability/redundancy). A central website will provide an extensive set of tools to query, retrieve, plot and compare the profiling float data dynamically. The choice of which data server to access could be determined by its proximity to the user while attempting to alleviate the load put on either server.
Each of the two Global servers receive data directly from every National Centres with the latest version of their float profiles, trajectory data and float metadata. Both servers should be updated simultaneously in order to ensure consistency between the two datasets. Each file is the responsibility of a single Dac (i.e. the data provider) who guarantees the quality and integrity of the data. A security system is set up on each server to ensure that only the appropriate Dac is given the necessary privileges to create/modify a NetCDF file for a given profiling float.
The two GDacs provide an FTP and a WWW access to the Argo data. The FTP server, suitable for data retrieval by a script/program, will provide users data organized: