At Sea Monitoring Tools

At sea monitoring activities are needed to detect anomalies that may impact a fleet of floats. When a fleet start to be large ( a few hundred of floats ) it's impossible to scrutinize each float behaviour to detect such anomalies. It is therefore important to set up systems that raise automatically alarms on changes in float status and provide an easy access to all the technical information that can then be analysed by engineers or scientists.

Here, we present what organisation and tool have been set up at Coriolis data center to monitor the French fleet (more than 700 hundred floats deployed between 2001 and 2010, about 200 active en 2010 ) and how the experience gained at Coriolis could be extended to the Euro-Argo fleet.


An efficient "At sea monitoring " relies on three main components :

• A rigorous record of exhaustive metadata at all steps of a float life that allows
o traceability of the components integrated by the manufacturer,
o information of the integration and acceptance tests done at manufacturer and at reception by the customer
o information on float programming
o information on float deployment environment

• A monitoring system that compute automatically some statistics and detect some events ( float death, float grounding, data transmission rate,… ) and is able to raise alerts automatically

• A person in charge of periodically studying the alerts and coordinating the necessary actions to
solve the alert.

At Coriolis, two levels of monitoring at sea have been defined, having different temporal timescales :

  • A monthly analysis for quick problems detection requiring rapid action from the engineering team, the manufacturer or the deployment coordination team. This analysis of routine requires 1-2 day's work per month by a trained person. It consists of an analysis of the alerts raised automatically that always lead to a monthly report and to a set of actions to solve the problem(s) detected if necessary..
  • An annual analysis to provide an overview of the float behaviour as well as a report on the evolution made during the period. This work requires 1-2 month of full-time work and generates a report for the users and scientific teams involved in Coriolis. This report, is a useful document that can be used by all the actors involved in Argo from manufacturer, to project manager and scientific users. It is also useful to provide an expert feedback at technical workshops organised by the Argo community.
Monitoring System developped at Coriolis for French fleet

In support to the Argo activity Coriolis set up a web tool (click here to see more) that provide statistics , updated on a monthly basis, on pre-defined batch of floats. Presently, the following batches of floats have been defined :

• Argo Fleet
• French Argo per year of deployment
• By platform type
o Provor CTS 2
o Provor CTS 3 whole fleet or per year of deployment
o Provor CTS 3 with Oxygen sensor
o Arvor
o Apex
• With specific instrument or transmission devices
o Oxygen floats Provor and Apex
o Biochemical floats
o Iridium floats: Provor, Apex and Nemo
For each batch, the following information are provided, organized in 4 index :

  1. The “Float Age” and "Float Status" indexes provide general statistics on the fleet behavior and allows a synoptic view of the fleet both in term of age, number of cycle performed or geographical . It is designed for project manager or scientists. In an ideal world we should see a sudden drop in the float cycle distribution corresponding at the fact that the batteries are empty. Unfortunately there are many other reasons for a float to die at sea.
  2. The “Functional Monitoring” index gives information on the main problems identified on float and that engineering team decided to track ( data transmission, length of the profile, anomaly in temperature , salinity or pressure ) as well as the list of float that encountered one of these problems during it's life. Each float is registered with its’ WMO number and when he clicks on this number the user can access Coriolis individual float pages that provide usual access to the float profiles, trajectory and metadata.
  3. Finally the “Technical Monitoring” index allows an in depth examination of each float behaviour
    and is used by the engineering team and the manufacturer to determine eventual cause of an
    strange behaviour or a float death. It separates active from dead floats, and then provide an index
    per platform type. It highlights the number of dead floats since last monthly report.