The ability to discover and access historical datasets is a significant challenge for many agencies charged with managing spatially defined marine areas. Unless data have been previously aggregated by (for example) method it is difficult to discover historic datasets as many of these have yet to be made available through the existing national data infrastructure. This need led to the development of the Australian Region MArine Data Aggregation (ARMADA) tool through the NESP Marine Biodiversity Hub partners. ARMADA provides a comprehensive snapshot in time of data available through individual research agency’s geoservers, providing government agencies, industry and other stakeholders a more complete listing and access biological and physical data within Australia’s national marine estate. While this prototype tool was developed to meet a specific user case (Australian Government Department of the Environment and Energy (DOEE) data identification for all biological samples that fall within the boundaries of an irregular shaped polygon, such as the boundaries of a Commonwealth Marine Reserve, Key Ecological Feature or Biologically Important Area) it has clear user benefits for any management or research agencies in identifying the spatial extent of historical research.
In developing this tool, the primary aims have been to meet a specific user need and to demonstrate/test the utility of a spatial data aggregation tool. While it appears there is a clear end user need, this tool does require ongoing support and development, and ideally will be incorporated into national data infrastructure such as the AODN network. Differences in software architecture and the challenge of getting research agencies to publish their data in compatible formats may limit immediate adoption by AODN, however, a clear recommendation from this work is that the Hub partners and AODN/IMOS work towards a single source for accessing this marine data in the future.
Currently, a significant limiting factor in data discovery and aggregation is that existing data served by WFS through the AODN Portal lack harmonisation, as each provider is free to use any data structure and convention that can be delivered by their map server. While AODN has developed a single controlled ‘parameter’ vocabulary intended to be used for all Australian marine metadata, ARMADA overcame this limitation via a manual process of creating shared vocabularies across similar datasets (thus facilitating aggregation of the data itself), however, this was a significant, and time-consuming technical challenge. We recommend that such a challenge be overcome and automated by development of a national standard for data served via web feature services (WFS), and ongoing discussions and development of controlled vocabularies to structure WFS and annotate metadata records supplied to AODN.
Overall, the ARMADA development process has demonstrated pathway to aggregate data from many research agencies across a defined set of data types/gear types within defined spatial boundaries. It highlights a clear need in the existing national data infrastructure, that ideally would be filled via ongoing development within the existing AODN infrastructure.