by Ben Johnson, Senior GIS Consultant
Spatial Vision has worked on a number of projects where we have co-ordinated the collection of data from a number of private sector and government agencies.
In some cases, the organisations we are working with have a limited understanding of GIS, let alone the ANZLIC page 0 metadata requirements. One of the benefits of producing metadata in these situations is that the organisation is forced to take some ownership and responsibility for the quality of the data they are providing.
This article discusses the approaches taken to gathering metadata, particularly in relation to ESRI technology, and how these can be harnessed in small to medium sized projects.
ANZLIC Metadata with ArcCatalog
One of these projects has been the Maribyrnong defence site project, where data has been gathered from a number of different project consultants to support the disposal of land at the old Maribyrnong Defence Site in Western Melbourne. As part of this project, Spatial Vision utilised the ANZLIC metadata entry tool developed by ESRI Australia to store metadata associated with each dataset.
With the introduction of ArcGIS, ESRI introduced ArcCatalog. ArcCatalog allows you to store and edit metadata along with the spatial data in XML format. The metadata travels with the data when you copy and paste the data from one location to another. In addition, dataset derived information such as the dataset extent can be automatically calculated. The ANZLIC metadata entry tool has been developed by ESRI Australia in Visual Basic to allow for the editing of ANZLIC compliant metadata within ArcCatalog.
The main disadvantage of this approach is that you need to know where the data is to locate the metadata. It may also be an issue when, for practical purposes, you need to maintain a number of versions of the same dataset. In this instance, insuring that the current version of metadata applies to the current dataset version can be problematic.
In another project, the National Oceans Office (NOO) has selected Spatial Vision to produce a National Marine Atlas showing the non-fisheries uses of Australia’s marine jurisdiction. As part of the data collection process, Spatial Vision is gathering metadata from a number of government agencies in an Access database. At the final stage of the project, data will be exported into the ‘Neptune’ data directory maintained by NOO.
This is clearly a different approach, with metadata being centrally located and having no physical link between the spatial datasets they describe. One of the benefits of centrally locating metadata is that it can immediately be searched. This is critical when there is poor knowledge of the data, or the location of the data is not known. In this project we have been able to use information within the metadata, notably custodial information, to construct disclaimers that are used on the maps.
Metadata and ArcIMS
To counter the data-centric view of metadata that is inherent with ArcCatalog, ESRI allows ArcCatalog users to publish their metadata to a metadata service that is hosted by ArcIMS. ArcSDE stores and indexes the associated metadata documents in XML format within the database platform on which ArcSDE resides (SQL Server/Oracle/DB2 etc). A customised HTML/JSP client interface provided by ESRI allows you to search for metadata within the metadata service. In addition, ArcCatalog users can search for datasets in a metadata service on the basis of attribute details or coverage area. This approach is particularly suited to larger projects and organisations that are supported by an investment in corporate infrastructure, ArcIMS and ArcSDE. Further details on the implementation of a metadata server using this technology can be obtained from the ESRI website. For small to medium size projects, the infrastructure requirements of a metadata service, ArcSDE and ArcIMS, make this approach impractical.
As part of the Maribyrnong project, the project consultants were provided with a CD containing an ArcReader application and all the project data that Spatial Vision had collected. (ArcReader is a freely available application that provides a cut-down version of the ArcMap interface) The challenge was to ‘harvest’ all the metadata that was stored with each dataset and make it available via a central interface. To do this, a tool was written to run in ArcMap (using ArcObjects) that would interrogate each dataset and export its metadata into HTML format. Data could just as readily be exported to XML format also, if required. A controlling HTML page was created with a link to the HTML page for each dataset.
The advantage of this approach was that links could be created to each of the metadata records with the same names as stored in the ArcReader application, making the metadata easily accessible via a simple logical interface. Finally, all the data, the metadata and the ArcReader application were packaged up into a CD that was distributed to all the project consultants.
This approach gives users the advantage of working with ArcCatalog’s tools for maintaining and editing metadata, and an approach for gathering and collecting metadata so that is more readily accessible.
or further information on the ArcMap metadata exporter tool, please contact:Ben Johnson
Ph: (03) 9691 3000