Why we need to get smart about data to be better stewards: Making smarter virtual observatories

For some time, the term cyberinfrastructure has been used to describe the collected set of computer-based software and services utilized to support and advance science, engineering, and education. Over the past decade there has been an evolution of that cyberinfrastructure aimed at enabling more robust and virtual, distributed scientific research set in a context of an interdisciplinary virtual observatory, i.e. using cyberinfrastructure to enhance data stewardship. Early efforts, in for example, the Virtual Solar-Terrestrial Observatory (VSTO) applied the then emergent semantic web data frameworks to provide production access to observational datasets from solar-terrestrial physics. That observatory provides virtual access to highly distributed and heterogeneous datasets in a common way, i.e. they appear to be organized, stored, accessed and used as if they were local or in other words, not exposing the underlying organization of the data holdings. More recently, an even larger science community, oriented around carbon in Earth has implemented a Deep Carbon Virtual Observatory (DCVO). The DCVO integrates (rather than develops) an existing set of application cyberinfrastructure, i.e. data and information stores, catalogues, collaboration and network tools, and application services. In this paper we indicate the development of the embedded semantic technology within the DCVO and indicate future directions.