Geoscience Reference
In-Depth Information
example, the strategic plan for the US Integrated Earth Observation System, the US
contribution to Global Earth Observation System of Systems (GEOSS), calls for the
implementation of GEOSS services within a web-enabled, component-based archi-
tecture in its overall data management strategy so that the value of Earth observations
data and information resources is maximized (Hood, 2005; IWGEO, 2005). Likewise,
the Integrated Ocean Observing System and the NOAA Group on Earth Observations
Integrated Data Environment (GEOIDE) are both planning to use a SOA/web services
approach for providing data services to their respective communities.
Another computing model that is beginning to transform how resources are ap-
plied to solve complex scientifi c problems is grid computing, a term that originated in
the 1990s as a metaphor for making distributed computer power as easy to use as an
electrical power grid. While many defi nitions of grid computing exist (Wikipedia: The
Free Encyclopedia, 2006), the most defi nitive and widely used one is by Foster and
Kesselman (1997). According to them, the grid refers to “an infrastructure that enables
the integrated, collaborative use of high-end [and distributed] computers, networks,
databases, and scientifi c instruments owned and managed by multiple organizations.”
Grid applications often involve large amounts of data and/or computing and often
require secure resource sharing across organizational boundaries. Grid computing and
the science enabled by it, eScience, are two major trends in distributed computing.
A key advantage of grid computing over historical distributed computing systems is
that the grid concept permits the virtualization of computing resources such that end-
users have the illusion of using a single source of “computing power” without know-
ing the actual location where their computations are performed. The use of digital
certifi cates to access systems on behalf of a user and third-party fi le transfer between
grid nodes authenticated via certifi cates are specifi c examples of how grid technology
enables resource allocation and virtualization. Grid Services, which implement web
services in a grid architecture, are in still in their infancy, although several proof-of
concept test beds have been deployed in a number of disciplines, including Earth and
atmospheric sciences, high energy physics, and biomedical informatics. Although the
distinction between traditional web services and grid services is subtle, grid services,
ideally, should enable virtualization for building and running applications that span
organizations and share resources and infrastructure in a seamless way. Gannon et al.
(2005) and Foster (2002) provide the distinguishing characteristics of a grid service
and specify what is needed for a web service to qualify as a grid service.
DATA SERVICE ATTRIBUTES
As articulated by Cornillon (2003), the ultimate objective of a data system or service
is to provide requested data to the user or user's application (e.g., analysis or visual-
ization tool) in a transparent, consistent, readily useable form. The users do not care
as much about the technology behind those systems or services, but do about trans-
parency and usability. The key to achieving Cornillon's two objectives is through
interoperability of components, systems and services, via the use of standards.
In the opinion of this author, an ideal data service should have the following
attributes:
