Interoperability of Argo Essential Climate Variables

Argo is an international observation array of approximately 4 000 autonomous profiling floats measuring oceanic Essential Climate Variables (ECV) consisting of physical (pressure, temperature and salinity) and biogeochemical (dissolved oxygen, pH, nitrate, chlorophyll-a, downwelling irradiance and suspended particles) variables, from 2000-meter depth (or from 6000-meter depth for the deep floats) to the surface every 10 days, all over the ocean. More than 3 millions of vertical profiles have been collected in 25 years. 

The Argo array is unique as it samples the global ocean, even in regions or seasons when vessels cannot operate, and depths that satellite sensors cannot probe. Argo is tightly connected to other observation arrays, in calibration or cross-validation efforts, such as with the accurate measurements performed onboard research cruises, essential for Argo to achieve the accuracy required for climate studies, or satellite data. 

Argo contributes to monitor and understand climate change for several key climate change phenomena, including increase of the ocean heat content (and sea level rise), deoxygenation phenomenon, ocean acidification and carbon cycle. Because of its importance in science studies, including carbon cycle, the presentation will focus on the interoperability of the Argo dissolved oxygen data.

To facilitate science studies and support for public policies based on ECVs, FAIR share of both data and metadata is essential. The Argo international program has been continuously improving the findability, accessibility, interoperability and reusability of its dataset since its inception. Recently, Argo has increased its interoperability by exposing its vocabulary on the web, specifically on the NERC Vocabulary Server applying the I-ADOPT framework and thus facilitating the mapping of Argo vocabulary with the vocabulary of other research infrastructures. Argo metadata and data access services have also been improved to match evolving users’ needs. They have been made more accessible by being exposed under federative platforms, such as the ENVRI-Hub, currently developed under the ENVRI-Hub NEXT EU project, or on Galaxy Europe for a biogeochemical calibration collaborative workflow that has been developed under the FAIR-Ease EU project. Interoperability improvement activities are also currently undertaken within the AMRIT EU project for European marine research infrastructures datasets, including oxygen.  

FAIRness challenges faced by the Argo program are multiple, for instance the necessity to simplify the dataset for a given audience, which is done through the development of products (e.g. easyOneArgo) or the necessity to increase the interoperability of the measures’ conditions, including methods and uncertainties. 

Indeed, uncertainties are key information to climate change analyses: foreseeing that the sea level will rise by 10 meters +/- 5 cm has not the same meaning as foreseeing that the sea level will rise by 10 meters +/- 10 meters for a decision-maker. Argo uncertainties have the same dimension as the dataset itself (i.e., an error value is associated with each observation point), which means that uncertainties are to be considered as data. For the interoperability of essential climate variables, in science studies in general, and in predictive studies in particular, the FAIR share of uncertainties associated with ECVs is crucial.