Improve Float Technology

The present float technology has a long history of development and it can be considered as a well proven component of the Euro-argo infrastructure. The EA-RI will pursue its involvement in R&D on float technology, in particular in light of specific European requirements (e.g. operations in marginal seas). Examples of such developments concern improved capabilities and performance (communications, handling, deeper measurements) and additional or improved sensors (e.g. bio-geochemical and bio-optical sensors).

1. Lithium Batteries

Provor uses Lithium batteries which has high capacity, high current capabilities, low auto-discharge, high reliability, high price (~5% of total float cost). They are designed to supply the float to perform up to 250 cycles at 2000m, continuously pumping during profile and transmitting 110 CTD points to the satellite. Higher capacities batteries exists (18 A.h instead of 13) and are available in the same volume, but with lower current discharge. These technology could be used for lower depth operation (eg coastal floats) or coupled to "super capacitors" to absorb peak current needed by engine (increasing price).On the other hand, alkaline batteries could lead to reduce costs and simplify shipping rules (but reducing also performances).

Apex floats delivered from the manufacturer at present are only supplied with Mn-Alkali batteries. These are designed to provide sufficient energy for the floats to perform a 4-year (140 cycle) mission to 2000m. But increasingly float operators have opted to fit lithium batteries to increase float life and thus provide a greater margin of error. Starting in 2010 WRC will supply floats fitted with lithum batteries.


2. PROVOR and ARVOR Communications
2.1 - ARVOR-i (iridium satellite transmission)

The standard Arvor (CTD measurements for Argo) uses Argos2 transmission. For marginal seas, an iridium model has been developed. The deployment of Argo in marginal seas has specific requirements. In particular, it is critical to reduce the transmission time at the surface to reduce the risk of thefts in these highly trafficked seas, and to have better estimates of subsurface currents with a given reduced cycling period (e.g. 5 days).

The ARVOR-i model includes a small modem and the iridium/ GPS bi-band antenna has replaced the Argos one.

All of the standard Arvor specifications have been maintained, and some specific features have been added as follows :

- The transmitting algorithm has been modified and data has been gathered in 200 bytes SBD (short burst data) messages. The software drivers of the iridium modem and the GPS receiver have been improved (compared to previous one used on ProvBio).

- The technical message contains more information about the behavior of the float and includes the last CTD raw data before stopping the CTD pump at the end of the rising profile. This is useful for the knowledge of the surface properties.

- The parameters of the mission (period cycle, parking depth, profile depth, …) can be remotely modified by the downlink capability of the satellite transmission.

- In order to get a lower volume / weight antenna, and a product easier to manufacture, a new patch antenna has been designed (as an alternative to the current model) and tested for 2000m depth operation.





Two Arvor-i have been ordered by Ifremer and delivered by our industrial partner NKE. Software design and tests have been done by the department of technology at Ifremer. Environmental tests have been done in the pressure tanks at Ifremer, including real mission operation with permanent control of the weight of the float.

Arvor-i sea trials
  • The first Arvor-i (wmo 6900794) has been deployed by Coriolis (France) in the Mediterranean Sea (south of Cyprus) in december 2009, 4th. It was programmed to make one cycle per day from 700m depth. 254 cycles has been done up to 2010/08/16
  • The second float (wmo 1900848) was launched in Adriatic sea by OGS team in February 2010 :

This float is cycling in a shallow area, proving its good capability to manage very frequent grounding behavior. Several remote commands have been sent to the float in order to modify the cycling period and the profile depth. Initially, the float was programmed to cycle every 5 days, drifting at 350m and profiling from700m depth. Every 10 cycles it was programmed to dive at 1100m depth.

At the end of the year, the float has drift northward, moving away from the south Adriatic Basin. In December, in order to reverse this drift, the parking and starting profile depths have been reprogrammed to respectively 150m and 200m.

This float is cycling in an area where depth changes very often from shallow to deep scheme. The float “follows” the ground and samples all the water column, proving its good capability to manage very frequent grounding behavior. At the end of 2011, this float has done 134 cycles. At most of cycles, the float has grounded.

  • A third float (wmo 6900800) was bought by OGS and deployed in the Ligurian sea in July 2010 :

Several remote commands have been sent to the float in order to modify the cycling period and the profile depth. Particularly, the float has been recovered 2 times and redeployed thanks to a new specific remote command allowing the float to stay at surface at the end of a cycle.

The performances of the float :

These floats are able to better sample the water column. They have been programmed to transmit 240 PTS (pressure, temperature, salinity) triplets between 2000m depth and surface, instead of 100 triplets for standard Argo float. Particularly, the upper area is sampled every 2m. When the float has reached the surface, only 3 minutes are needed to transmit a high sampled data profile (every 2 dbars). The total time at surface is approximatively 30 minutes, including the time to increase the buoyancy for good satellite transmission, and afterwards to reduce buoyancy to start a new cycle. This performance has to be compared to more than 8 hours with Argos2. The objective to reduce the time at surface has been reached.

2.2 - ARVOR with Argos3 satellite transmission

In order to have an alternative way to improve communication performances on floats, assessment
and integration of the new Argos-3 system has been done.

Argos system represents 6 non-stationary polar-orbit satellites, with 1 active Argos3-mode (MetopA) available today with 2 ways communication. In order to have an alternative way to improve communication performances on floats, assessment of the new PMT (Platform messaging transmitter, supplied by CLS-Argos) has been done.

Tools have been designed to understand and evaluate Argos3 performances. The aim was to find an appropriate methodology for using Argos3 on floats.

Phase 1: understanding and evaluation of the Argos-3 system (2009)

An evaluation software has been designed to do the trials. The aim was to test random protocol (Argos2 mode), interactive mode (Argos3 low data rate mode) or "pseudo-track" mode (based on satellite pass predictions). Assessment has been done on transmission performance (error rates, influence of satellite elevation and orientation, power balance, scheduling strategy for the float to surface, using pass prediction tables), and downlink communication, determination of the strategy to be implemented on profiling floats. The results for the transmission of one profile during one satellite pass in less than 15mn was successful, instead of 6 to 8 hours with Argos2.

Phase 2: integration on profiling floats (2009/2010)

The most suitable algorithm was determined for the implementation in a profiling float. The specifications of the software have been written, using UML description. Finally, the software has been written on the embedded processor target, and partially tested in many transmission. There is a design of a new double band and a pressure resistant antenna. We use an environment simulator for running complete mission test

In other respects, the issue of the double band antenna has been resolved and the new design has been successfully tested in IFREMERs' pressure tank. Two floats have been assembled and tested in pressure tank and in seawater pool. Then, satellite transmission has been tested at sea, near IFREMER institute, in order to get results in real environmental conditions (sea water, buoyancy, satellite visibility,…) in October 2010. Some trouble on the electronic PCB of the float (Eeprom default), and a default on one antenna (RF wire solder) have been detected and corrected.

Arvor-Argos3 deployments :

Now, Argos 3rd generation transmission is being embedded on Arvor. It is intended to use the interactive mode capability (low data rate) of the MetopA satellite, using its prediction pass tables to make a « rendez-vous » at surface. Argos2 standard communication is maintained in case of interactive mode failure.

Two Arvor floats (6900947- see figure - and 6900952) equipped with Argos-3 communication had been deployed at the start 2011 in the Mediterranean Sea by OGS team. They cycled every 3 days at 350m, and make profile from 700m to the surface. 77 profiles for the float 6900947, and 98 profiles for the 6900952, are registered since 2011/12/12. At most of cycles, the float 6900947 has grounded near the coast.



3. Ice-sensing algorithm for arctic/subarctic region

In its current form, the Argo system is restricted to oceanic regions that are ice free year-round, as the floats need to surface regularly to be localized and to transmit the data. During recent years, the ice-resilience of Argo floats for the Southern Ocean has been considerably increased by the development of an ice-sensing algorithm (ISA). This algorithm is based on a relation between the subsurface-temperature and the probability of sea ice at the surface. Within the Euro-Argo project the Antarctic ISA was adjusted for Arctic/Subarctic regions. For this purpose, temperature profiles collected during several Polarstern cruises and from Ice-Tethered Profilers (ITP) have been analyzed in conjunction with the corresponding sea ice situation as obtained from satellite observations (AMSR-E data using the ARTIST sea ice algorithm).

The near surface temperatures show a promising temperature/ice – correlation, albeit less pronounced compared to that of the Weddell Sea. Unlike the Southern Ocean, large areas of the Arctic Ocean remain ice covered even in summer, which might result in Arctic floats being trapped under the sea-ice for several years. Thus, a major development objective was to try ascending to the surface to submit data even in small ice free areas. For this reason, previously used safety margins were abandoned in this first version of Arctic-ISA, which was defined as to abort ascent, if the median of the temperature measurements between 10 and 20db is less than ‑1.3°C. Both, the averaging interval and the temperature are adjustable and can be easily optimized and adapted to different areas.

Arctic floats will ascend without avail more frequently than Antarctic floats, i.e. will have physical contact with sea-ice more frequently. To provide additional protection against harsh contacts, a synthetic cage is added to protect the CTD and the lower part of the antenna. To increase the reliability of the Arctic-ISA, the profile-resolution within the near surface layer (upper 50dbar) has been increased to 1dbar.

Simultaneously, the NEMO-floats have been fitted with an Iridium modem coupled with a GPS receiver and a high pressure antenna. Using Iridium satellite communication has been drastically reduced the surface time.

For more informations on NEMO Floats, click here.

Deployments :

After the 2009 technical failure, two more NEMO floats have been equipped with the new Arctic ISA and were deployed into the Sub-Arctic Ocean in July 2010 in close collaboration with the Euro-Argo partner “Institute of Oceanology Polish Academy of Science”. Both floats are working, i.e. 19 profiles with temperature, salinity and oxygen data have been carried out by each float by the 5th January 2011. The evaluation of the performance of these floats will continue throughout their live time.


Implications for the Argo Array

Several Float types have been prepared for alternative communication routes (different from classical Argos 2); among these floats are the widespread Webb floats, Nemo (tested at the moment), and the above described Arvor floats. Surface times could be drastically reduced to just a few minutes which would be of great importance for regions with heavy traffic (Med Sea) or in partly ice covered regions. Bi-directionality would also offer the opportunity to alter the schedule of the floats, e.g. from high frequency profiling to normal schedules. Second, bidirectional communication offers the opportunity of altering the Floats schedule on demand – profiling depths and profiling frequency both are issues.
This work is ongoing and requires time beyond the Euro Argo preparatory phase for final evaluation. Work regarding the use of oxygen sensors will continue within science driven programs and (slow) evolution should be expected.