Argo Ocean Temperature and Salinity Profiles

 The following was contributed by Josh Willis, Jet Propulsion Laboratory, January, 2014:

The Argo array of profiling floats is a globally distributed network of autonomous, profiling robots that observe temperature and salinity over the top 2000 m (roughly the top half) of the oceans.  They are distributed globally, with approximately 1 float every 3 degrees of latitude and longitude, providing a new profile once every 10 days.

 The Argo array provides the most accurate and comprehensive means of observing global temperature and salinity change in the world’s oceans. Although salinity sensors in the Argo floats can be subject to biases that vary slowly over the course of months and years, the Argo data system is carefully constructed so data centers now provide corrected salinities, where these drifts are estimated and corrected for.  These corrected data are referred to as delayed-mode data and are available within one year of collection of any individual profile.  Data that have not had the salinity correction are referred to as real-time data, and are available within 24 hours of collection.  In practice, these data are accurate enough for many scientific and operational purposes.

 Data from the Argo array now provide the most accurate, ongoing observations of global ocean heat uptake, which is a direct result of human-caused climate change. Nevertheless, they do have some limitations. The Argo floats do not sample at very fine resolution in the ocean and are note capable of resolving small features such as ocean eddies.  At present, they also do not sample the deep ocean, and in some (usually high-latitude) regions it has been suggested that human-caused warming of the ocean penetrates all the way to the sea floor.  Finally, Argo provides only a few observations in regions with seasonal sea ice.  Specially designed floats have been deployed in a few regions that can avoid seasonal sea ice, but the Arctic Ocean and many parts of the Southern Ocean are not sampled by Argo.  Finally, because most Argo floats profile to a depth of 2000 m, they cannot sample the shallow marginal seas that are important for many coastal areas.

It is often suggested that because Argo floats drift freely in the oceans, they do not provide good uniform sampling. However, there is very little evidence for floats clumping, or experiencing significant divergence anywhere in the global oceans except for the equatorial Pacific. Most floats spend about 9 days drifting at 1000 m depth, and 1 day at the surface transmitting data.  In most regions, this duty cycle results in a rather random pattern of float trajectory.  In the Equatorial Pacific, however, the strong surface divergence tends to push floats off the equator making it difficult to sample.  In recent years, floats that spend at little as 1 hour at the surface have been used to populate this region and the resulting sampling along the equator has been much improved.##