Recording of recoveries and redeployments

[delphinedobler]

This subject may be related to issue OneArgo/ArgoVocabs#26.

The number of recovered and/or redployed floats is increasing. There is no formal mechanism to follow tracks of them.

Keeping track of these recoveries helps in computing expected life time of floats, and in checking consistencies in sensor serial number duplicates.

There is a need, at least at the European level, to record:

• if floats were recovered
• if recovered floats were refitted and what kind of refit was performed (battery changed, card changed, sensor changed or sensor removed/added, nothing was changed, etc.)
• if recovered floats were redeployed, with which WMO ?
• in case of multiple redeployments, the order of redeployment ?
• if a float is a redeployed float: from which WMO it stems ?

Up to now, some of the needed information is spread here and there, part of them can be found in OceanOPS, other in the Coriolis DB, some in the "float ending cause" field in the meta file. An internal listing was also created at Euro-Argo, which gather available information from these sources and from analyses of duplicates regarding float sn, sensor sn and/or ptt.

The questions are the followings:

• is there also a need at the international level to keep track of such records in a formal way ?
• where and with which formal semantics should we record these information ?