Report on the Argo Project for JCOMM-II

DBCP-XXI/Doc. 2.4.1(1), p. 1

Report by the Argo Project for JCOMM-II

(Submitted by Argo Director, John Gould)

ACTION PROPOSED

The panel will be invited to to comment, and particularly make decisions or recommendations, as appropriate on the following topics:

(a) Note and comment on the information contained in this document;

(b) Take into account the contents of the report when discussing relevant agenda items;

______

Report on the Argo Project for JCOMM-II

Submitted by the Argo Director

Background

Argo is a pilot programme of GOOS and GCOS using autonomous profiling floats to collect temperature and salinity profiles from the upper 2000m of the world’s ice-free oceans and velocity data primarily from the ocean interior (mostly around 1000m) and from the sea surface. Argo is co-sponsored by the WCRP’s Climate Variability and Predictability project (CLIVAR) and by the Global Ocean Data Assimilation experiment (GODAE). Argo is an integral part of GOOS and GCOS but is specifically designed to complement the satellite radar altimetry data from JASON.

The Argo technology is based on that developed during the WCRP’s World Ocean Circulation Experiment (WOCE) in the 1990s.

The project is overseen by the Argo Steering Team (AST) and the Argo Data Management Team (ADMT). Day to day project supervision is given by a part-time Argo Director (Dr W John Gould, presently funded by the USA) and by the Argo Technical Co-ordinator (Mathieu Belbéoch), based in JCOMMOPS in Toulouse and funded by contributions from Australia, Canada, France, UK and USA.

Array growth

The target is to establish and maintain a 3°x3° array that will total approximately 3000 floats delivering data from each float at 10 day intervals (ca 100,000 CTD profiles per year).

The first floats were deployed in 1999 and since then the array has grown steadily to reach its present 2000 floats. (Figure 1).

Figure 1 Growth of the Argo array

Floats are being deployed at a rate of around 800 per year. This is the rate that needs to be maintained if the array is to reach and maintain its target density (assuming floats achieve their designed lifetimes of 4 years). The target array should be complete by early 2007. The floats are considered expendable since the cost of recovery far outweighs the float’s value (ca $15,000).

Initially there was a distinct northern hemisphere bias to the Argo array but this has been addressed through targeted deployments in the southern hemisphere basins and the Southern Ocean. (Figure 2)

Figure 2 The Argo array in August 2005 (top) and August 2003 (bottom) showing the increased coverage in the Southern hemisphere oceans.

Floats are deployed from research and merchant ships and can also be deployed from aircraft. Many countries have assisted with float deployments.

Technology and reliability issues

The array is made up of three float designs. Two are commercially manufactured (Apex, Webb Research Corporation, USA; PROVOR, Martec, France) and one produced by Scripps Institution of Oceanography, USA. All are essentially similar in their functionality and comprise respectively 64, 10 and 25% of the present array with the remainder made up of a small number of experimental float designs. The Argo philosophy has been to safeguard the array by using multiple float suppliers. Almost all floats have salinity sensors supplied by SeaBird Electronics.

Communication and float positioning is carried out by Système Argos with a small number of experimental floats using Low Earth Orbit systems such as Iridium and Orbcomm coupled with GPS positioning. These provide improved accuracy, greater data bandwidth and the possibility of two-way communication. The present Argos floats transmit data from of order 100 depth levels together with a small amount of diagnostic data. Float have to remain at the surface for 6-12 hours to ensure all data are downloaded via Argos.

The technical challenge of Argo is considerable since floats are required to work autonomously and provide high quality data for periods of order 4 years. Considerable progress has been made in extending float lifetime. This is illustrated in Figure 3.

Figure 3 Number of floats “dying” in various age classes.

This shows the steady increase in the most common age of float death and the increase in the number of floats living beyond the target 48 months.

Almost all floats fail at depth when the battery voltage falls to a level where it cannot propel the float to the surface. A small number that have been recovered at sea or by being washed ashore have given valuable information on float and sensor performance.

Data delivery and quality

Argo data are available through two streams. Real-time data are available on the Global Telecommunication System and via the internet and Live Access Servers (LAS) from Global Data Assembly Centres (GDACs) in the USA (US GODAE Server) and France (Coriolis). (Figure 4)

Figure 4. The Argo data flow structure

Since 2002 when monitoring began, between 80 and 90% of Argo profile data were available to users within 24 hours of a profile being made. These data are subject to similar integrity and quality checks as the real-time XBT data stream.

The salinity sensors drift with time due to biological fouling and to physical deformation caused by repeated cycling over as much as 200 atmospheres. However the floats are capable of delivering data that approach the quality of ship-based observations. The required validation and correction of salinity data is carried out by comparing float data with ship-based climatological data and by comparing data from old floats with that from nearby new floats.

Standardised procedures for delayed-mode quality control were agreed at a workshop in April 2005 and the backlog of data needing these corrections should be reduced by early 2006. Delayed-mode quality control is carried out on data in a 12 month sliding window, meaning that Delayed Mode data cannot be available within 6 months of the measurement being made. The delayed-mode QC is carried out by Argo scientists and by regional oceanographic specialists.

Argo is also establishing regional data centres (RDACs) that will serve to ensure uniformity of data quality across regions and will work to ensure the up-to-date ship based reference data sets are assembled for regional Argo DMQC.

Argo organization, infrastructure and oversight.

The Argo array has been made up of floats operated by over 20 countries with contributions ranging in size from half the array (USA) to the operation of one or two floats sometimes the gift of donor nations. Any country wishing to contribute to Argo by providing or deploying floats can attend meetings of the Argo Steering Team.

All data are freely available to anyone wishing to use it. The deployment of floats is covered by IOC resolution XX-6 that inter alia

Concludes that concerned coastal states must be informed in advance, through appropriate channels, of all deployments of profiling floats which might drift into waters under their jurisdiction, indicating the exact locations of such deployments,

Instructs the Executive Secretary IOC, in close collaboration with the Secretary-General of WMO and in consultation with the Executive Director of UNEP:

I .to inform all Member States, the IHO, and appropriate UN agencies, including IMO and FAO, of the acceptance of the Argo project by IOC and WMO;

Ii .to inform all Member States how to determine float locations and access float data;

Iii .to consider how all Member States might participate in and benefit from the Argo project, as well as propose options to that end; and

iv. to appeal for international co-operation in making the Argo project a success;

This notification process is the responsibility of the Argo Technical Co-ordinator who also monitors the growth and performance of the array.

Data use

Argo is already the major source of profile data from the open ocean. The real-time data stream presently provides ca 6000 profiles per month compared with around 2000 XBT profiles. Argo data are used by over a dozen operational forecast and analysis centres that have cited the data’s abundance, uniform seasonal and spatial distributions, greater depth penetration and the availability of salinity information as key advantages of Argo compared with other data sources. Many regional and global products are now based on Argo data.

The wide availability of Argo data has also encouraged its use in research applications and a bibliography of publications based on Argo (and earlier profiling float data is maintained on the Argo web site.

Challenges…

The rapid growth of the Argo array is only the first step towards its full implementation and many challenges remain.

Completing and maintaining the array. Presently almost all Argo funding comes from research budgets. These are rarely able to sustain funds beyond a 5 year period and thus many countries face the prospect of having to find new funding mechanisms for Argo. This is a major challenge since, despite the rhetoric applied to issues such as the establishment of the Global Earth Observations System of Systems (GEOSS) few countries have the infrastructure and funding mechanisms needed for long term ocean (or atmospheric/terrestrial) observations. These issues need to be addressed urgently in such a manner that the array can be completed, maintained and assessed while still assuring the effective involvement of the research community needed to maintain data quality and evaluate the array’s design and performance. JCOMM has a role to play in helping Argo make the transition from a pilot programme to a sustained observing system.

Delayed mode-quality control. In order to provide users with the best possible data sets and thus allow maximum exploitation of the investment in the array, the DMQC process needs to be fully implemented and maintained. One key requirement is the timely and open availability of high-quality ship-based CTD data such that reference data sets can be kept up to date thus allowing the Argo data to be used to study temporal change of ocean salinities. The issues will be address at a workshop to be held in Japan in November 2005.

Supporting infrastructure. At present the level of infrastructure (ATC and part time Argo Director) is insufficient to provide the level of support to the project that it needs. Funds and personnel need to be identified to allow a minimum of two full time staff in support of Argo.

Technology. Constant attention to the array’s performance is needed so that potential technical problems with floats or sensors can be rapidly identified, diagnosed, discussed with manufacturers and rectified. This issue will remain a difficult one requiring effective monitoring of the array and the commitment of resources to technical assessment and evaluation. A first workshop on this issue is being held in September 2005.

Temperature and alinity data are presently collected no nearer to the surface than 5m so as to avoid salinity sensor contamination. There are potential benefits from data nearer to the surface particularly for the validation of satellite data and particularly for future salinity missions.

……… and opportunities

New sensors and communication. The Argo array provides a global platform that could be used for other measurements. Trials have been made of several sensors (dissolved oxygen, acoustic wind and rainfall, velocity shear, fluorescence, optical plankton counter). All provide a burden on energy budgets and a potential decrease in float reliability. The AST has agreed that no new sensors will be incorporated in the core Argo array unless the potential benefits can be demonstrated to outweigh these negative impacts on float reliability and life. Instead floats with novel sensors are deployed in parallel with the Argo array.

Floats capable of making profile measurements under seasonal ice cover have been developed but are limited by the need for acoustic tracking in order to determine where the profile measurements have been made.

The introduction of broad bandwidth, two-way communication open up the potential for higher resolution data, less contamination of both subsurface velocities and salinity data following from shorter surface times and condition sampling (commanding more frequent profiles in interesting environments)

Educational outreach. The easy access to Argo data has great potential to inform the public and school children about the state of the ocean and its role in the earth’s climate. This is already being exploited in the Pacific Island States through the SEREAD project but wider use could be developed.

Scientific evaluation.

Scientific evaluation. In March 2006, Argo will hold its 2nd Science workshop in conjunction with a meeting celebrating 15 years of radar altimetry. This will be an excellent opportunity to demonstrate the wide range of operational and research uses of Argo and to highlight the value of the array.

Key URLs

General Argo information http://www.argo.ucsd.edu

Argo Information Centre

Argo portal