Need for robust spatial data management solutions for SDI implementation
Chandra Kavuri
Manager
Intergraph consulting India ltd.,
Email: chandra.kavuri@intergraph.com
Spatial Data Infrastructures (SDIs) provide a basis for spatial data discovery, evaluation and application. This includes sharing of Geographic data, associated metadata, and services through clearinghouse mechanisms. GSDI (Global Spatial Data Infrastructure) and all NSDIs (National Spatial Data Infrastructure) of different nations have a common goal (i.e) share Geospatial data across national boundaries. Such sharing needs common standards and interoperable practices meeting the needs of all nations. The efforts of OGC (Open Geospatial Consortium) and ISO-TC 211 lead towards implementation of these goals.
This paper focuses on some practical aspects and implementation challenges in spatial metadata management for global discovery of Geospatial resources. Spatial metadata is the data describing the Geospatial resources in terms of its content, quality, condition and other characteristics. This data helps in structured searching of geospatial resources from published clearing houses for the user’s context.
The first challenge is that of creating metadata and assuring its currency. Most metadata creation tools consider each metadata record to be an independent entity. This assumption leads to difficulties in keeping data consistent. A complex model associating the metadata entities using the concepts of hierarchy and inheritance is needed to manage the data consistency. Such models reduce redundancy and improve reusability. The paper discusses some such models.
The second challenge is that of associating the metadata to a spatial entity. The normal notion is to support metadata to a “data set” which is realized as a publishable map. However what constitutes a “dataset” is implementation specific. In a typical scenario, metadata can be associated to a dataset series, or to each spatial entity which constitutes the dataset series, up to the level of a feature entity. In such cases, the spatial entity model and the metadata model have to be associated in a complex way so that spatial entities of lower granularity inherit metadata from their parents of higher granularity. The paper discusses a typical spatial entity model and its metadata model relating the granularity of spatial data and metadata inheritance.
The third challenge is that of synchronizing the metadata to the changes in properties of the spatial entities. This needs close coupling of spatial entities and metadata. When either of the entities are relocated or modified, content must be synchronized and links have to be re-established to this effect. This paper discusses some aspects when the spatial data and metadata are geographically distributed.
The fourth challenge is that of supporting the relationship of metadata instances that use different standards or profiles with spatial data entities. Several standards and profiles currently exist, and some are in standardization process. It is not essential that the spatial warehouses need to have metadata of same standard. Each standard (profile) has unique way of representing the information. Organizing the metadata from multiple standards and associating to spatial entities is another important aspect in implementation. This paper discusses approaches which utilize the relational database and XML database concepts for supporting metadata of multiple standards.
The final challenge is the provision of catalog services for publishing resource descriptions in such a way as to make them accessible to prospective users (publish), discovering resources based on certain criteria (discover), and then interacting with the resource provider to access desired resources (bind). Standardization through the OGC catalog services specification and related profiles is the direction for SDI implementation. The paper discusses different protocol bindings and implementation aspects specifically CSW (Catalog services for Web) and the profiles under standardization process (ISO19115/19119, ebRIM and FGDC).