17.01.2013 Views

SDI Convergence - Global Spatial Data Infrastructure Association

SDI Convergence - Global Spatial Data Infrastructure Association

SDI Convergence - Global Spatial Data Infrastructure Association

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

data and grid computing infrastructures with associated metadata to deliver on-line<br />

spatial data processing and modelling.<br />

Figure 1 organizes metadata into several simple functional classes and assists in illustrating<br />

the conceptual relationships between spatial data, spatial data metadata and the<br />

different forms of model metadata.<br />

Figure 1: Functional classes of metadata.<br />

This approach to describing metadata is slightly unorthodox because it extends the<br />

common paradigm that uses metadata largely for purposes of search and discovery.<br />

The term metadata, purportedly first used in 1969 (Howe, 1996) is defined as “structured<br />

information that describes, explains, locates or otherwise makes it easier to retrieve,<br />

use or manage an information resource” (NISO, 2004). A more commonly used<br />

definition is that metadata is “data about data”. Here we use ‘model metadata’ as a<br />

term or descriptor meaning ‘data about models’. This is not to be confused with the<br />

terms meta-modelling and meta-models as these describe a contemporary approach to<br />

model selection, construction and assembly (Keller and Dungen, 1999; Bridewell et al.,<br />

2006).<br />

The NISO (2004) metadata report states “Metadata is key to ensuring that resources<br />

will survive and continue to be accessible into the future”. In Agriculture and Natural<br />

Resource Management (NRM) this is undoubtedly true for metadata about data and<br />

information resources as these are often primary sources of knowledge describing the<br />

state and nature of environmental systems and are important for comparative analysis<br />

regardless of their age. However, this is probably less so for model metadata as models<br />

represent the interpolative, inferential or processing systems that are used in research.<br />

These methodologies are generally subject to continual improvement and consequently<br />

it could be argued that model metadata has a greater potential to age and is<br />

most useful when it is more current. Although there are metadata standards for data<br />

and information (ISO 19115) similar standards are only beginning to emerge for model<br />

metadata.<br />

The functional levels shown in Figure 1 are closely related to three main types of metadata:<br />

descriptive metadata, structural metadata and administrative metadata (NISO<br />

2004). At level 1 the metadata is primarily used as a resource for discovery and identi-<br />

139

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!