Building an e-Infrastructure for Capture Statistics and Species Distribution Modelling: the D4Science Approach

D4Science-II was presented at the recent International Conference on Marine Data and Information Systems (IMDIS), organized by SeaDataNet, Ifremer and IOC/IODE in Paris.

The presentation explained how D4Science already assists data analysts and environmental scientists to produce collaborative products, such as AquaMaps. The presentation showed how activities that in the past required that data and logic are managed by different platforms and applications can now be accessed in a shared environment that enables true collaboration across parties that are widely dispersed and autonomous.

These collaborations are often cross-discipline and require innovative research environments that integrate data, processing and work-flows to produce new knowledge. They also increase the demand for interoperability, and some collaborative products and initiatives are already emerging, and are brought together in D4Science, e.g. :
- The Environmental community, where ESA provides GPOD interoperability, to share on demand geospatial data such as Sea Surface Temperature maps.
- The Biodiversity community, where AquaMaps enables biologists to create species prediction maps, integrating biological and environmental data.
- The Fishery management community, where extraction of information from statistical data sets combined with geospatial and environmental can improve catch estimates.

VREs were first described from the technological perspective, followed by the possibilities to collect, analyze, and organize data and the collaboration features. The VREs now include services for storage, processing and presentation of potentially huge and heterogeneous datasets, but they also allow semi-automated transformations across different data schemas and formats. Only part of the array of standards and interoperability-oriented approaches underlying the D4Science technology could be discussed.

True collaboration requires an environment with the ability to load and archive various sources of data, to harmonize datasets by mapping correspondences, and to query across these datasets.

In the case presented at IMDIS, the generation of AquaMaps using the services provided by D4Science was shown. Here, model-based, large-scale predictions of occurrence of aquatic species are calculated taking into account various environmental variables.

The D4Science pool of standards and technologies will continue to be enriched with state-of-the-art data exchange capabilities. These will include the Statistical Data and Metadata eXchange (SDMX), Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH) and Open Archives Initiative Object Reuse and Exchange (OAI-ORE).

As a consequence, the Virtual Research Environments will bring the power of data processing and storage to users in an easy to understand and intuitive format, with minimal requirements on bandwidth and processing power.

Delivered by Anton Ellenbroek, FAO of the UN, FIE, anton.ellenbroek@fao.org (Italy)