Figure 1: Argo national contributions
In 1999, a small group of oceanographers outlined a plan to set up an array of profiling floats to monitor the state of the upper 2 km of the global ocean. The initial objective was to maintain a network of 3,000 units, in ice-free areas, providing both real-time data and higher quality delayed mode data and analyses to underpin a new generation ocean and climate models. The programme was called Argo.
By late 2007, as a result of remarkable international collaboration, information sharing and the solution of complex technological problems, the 3,000 float target was reached and the array has remained above 3,000 floats ever since despite the global economic downturn.
Most importantly all Argo data were, and continue to be, made freely available.
In late 2012 the Argo array rose to 3,500 floats, the largest it has ever been and the average lifetime of the floats has approximately doubled in the past decade greatly increasing the efficiency of the operation. But this optimistic number, resulting as well from a more diverse use of the subsurface float vector, hides the fact that the core Argo array  is operating below 3,000 units, and that the existing funding level will not allow a proper expansion to the global and multi-disciplinary. Presently 28 countries contribute to the annual $25m cost of operating the programme (see Figure 1). Argo is highly cost-effective (~$250 per observation) but most of national contributions have not yet transitioned from a research based to a sustained funding mode.
Argo floats are gathering profiles of temperature and salinity (together with information on subsurface water movement) at the impressive rate of 1 profile approximately every 4 minutes, (360 profiles per day or 11,000 per month) and on 4 November 2012 the array passed the symbolic milestone of collecting its 1 millionth profile . To put this achievement in context, since the start of deep sea oceanography, in the late 19th century, ships have collected just over half a million temperature and salinity profiles to a depth of 1km (see Figure 2) and only 200,000 to 2 km. At the present rate of data collection, Argo will take only 8 years to collect its next million profiles.
In its short life, the Argo data set has become an essential mainstay of climate and ocean modelers complementing information from earth observing satellites but uniquely providing subsurface information giving new insights into changes in the earth's hydrological cycle and opening the possibility of seasonal climate forecasting.
While continuing its core mission of maintaining the global temperature and salinity array, Argo is extending into ice-covered areas and shallower oceans and is adding measurements of ocean biogeochemistry. Argo continues to strive to improve the network's data quality, the effectiveness of data delivery, to improve coverage of data-sparse areas and the reliability and lifetime of the floats and their sensors. A major challenge will be extending measurements to greater depths.
This symbolic 1 million profile milestone is just one of the programme's many achievements and there will be many more, not solely for Argo but for our understanding of the oceans and their role in earth's climate.
 Core Argo: floats fully operating according to Argo standards
 The float that collected the symbolic profile is operated by the Indian Argo program, (Float # 2901287), and participates in understanding monsoon rainfall seasonal patterns. This is a funny coincidence as the Argo Data Management Team will meet next week in India.
Figure 2: Argo vs historical subsurface data
Courtesy H. Freeland, Canada