Improving databases in scientific research

These databases play a role in the improvement of:

• the automation of marine observation, by handling data reception from onshore/offshore transmissions, their collation and pre-processing,
• the standardisation of acquired and banked information, making data collected independently comparable and thus allowing consistent baseline conditions to be established (long time series, climatologies, etc.),
• data quality, given the validation and labelling processes associated with the banking procedure. Standardisation and quality control include an ever increasing number of complex algorithms applied routinely (obtaining geophysical quantities, geostatistical analyses, etc.) and implemented by the data banking teams.

More integrated scientific approaches

This fuller integration is conceived in terms of:

• geographical areas (region, continental shelf, basin)
• time scales (multi-scale approaches, long term variability analysis)
• interdisciplinarity (habitats, ecosystemic approach).

These approaches require the constitution of large sets of consistent data, of established quality and with a sufficient level of processing to allow direct use by multidisciplinary teams (elimination of observation artefacts).

Diversifying data acquisition sources

For a long time, oceanographic cruises and satellites were the main sources of marine data acquisition. Yet thanks to technological advances, information production methods have greatly diversified: floats, fixed stations, autonomous vehicles transmitting data in real time, etc. Certain instruments now generate new types of data such as high frequency data or images. Furthermore, new observation data producers such as NGOs, professionals and the general public also supplement databases.

Strengthening inter-organisation cooperation

The aim is to facilitate the aggregation of data from different sources and in different areas, to coordinate initiatives and to adopt a unified distribution policy.

Scientific projects address complex issues as a whole: ecosystemic approach, modelling, study of the variability of the marine environment, economic aspects. These new approaches simultaneously require:

• access to long time series so as to establish baseline conditions and determine evolutions,
• rapid data availability for modelling and forecasting (real time aspects),
• the compilation of datasets covering broad thematic, temporal and spatial ranges, while remaining homogeneous and consistent, so as to characterise the entire phenomenon studied.

Reinforcing the interoperability of information systems

Because an interdisciplinary approach is required for so-called "ecosystemic" or "intregrated management" studies, information systems become tools which:

• provide access to datasets with a broad spatial, temporal or thematic range,
• allow communication between multidisciplinary teams,
• distribute the information produced.

This interoperability should apply in technical terms (communication protocols, exchange formats, component availability) as well as in thematic terms: harmonisation of repositories and descriptions (metadata), qualification procedures. Promoting interoperability should result in the creation of networks of data centres, capable of archiving and distributing consistent, aggregated data.