Geodatabases and Relationships

A Geodatabase is a relational database that contains both vector and raster tables along with other GIS data and the various 'relationships' between the data. Its hierarchical structure uses "primary" and "secondary" keys as a way of relating various feature classes and attribute tables. These relationships can be one-to-one, one-to-many, or many-to-many depending on how various variables are related. Essentially, the geodatabase allows one to define relationships between various geographical objects, and provides rules for maintaining these relationships.

 

Geodatabases and Data Modeling

Data Modeling is the process of defining GIS data in terms of its inherent relationships and characteristics. This requires definition of the general scope and business requirements of a database (Lo, C.P. and Yeung, K.W. 2002). This modeling approach is supported by ArcGIS during Geodatabase creation.

The data model is first entered as Unified Modeling Language (UML) into Microsoft Visio with all geographical object relationships being defined. Programs known as Computer Aided Software Engineering (CASE) tools (incorporated to ArcCatalog) help to generate a geodatabase schema from the UML model by developing XMI (XML Metadata Interchange) files. XMI or Microsoft Repository files are intermediate formats for CASE Tools in creating the geodatabase schema. XMI format creates personal geodatabases while Microsoft Repository creates personal or multiuser geodatabases.

The XMI file format is preferred because it is faster and preferred to the Microsoft Repository. The XMI file is imported through the CASE Tools schema wizard, which creates empty feature class layers, attribute class tables, relationship class tables, and the associated relationships originally established in the UML data model. ArcCatalog now contains an empty geodatabase. Feature layers are keyed in as ESRI Feature Classes in Visio while attribute tables are keyed in as ESRI Object Classes. The relationships between layers, as well as the attribute column names and length are mapped. After exporting the UML model to XMI or Microsoft Repository database file, semantic checks are performed to finalize the geodatabase.

Personal and Multiuser Geodatabases

A Geodatabase data model can be either personal or multiuser (enterprise). Personal Geodatabases are stored as Microsoft Access database files and allow many readers, but with a single editor. Multiuser Geodatabases, also known as ArcSDE Geodatabases, are used with relational database management systems (RDMS) such as Microsoft SQL Server or Oracle, which can be read and edited by many users (hence multiuser). The product provided to the District is a Personal Geodatabase, but can be readily converted to Multiuser Geodatabase by a series of simple steps.
Note: Since District of Pitt Meadows uses ArcGIS and Microsoft Sequel (SQL) Server, the selection of the database format is limited to that of an ArcGIS Geodatabase.

Data Standardization
One of the members of OD contacted a majority of GVRD municipalities to try to determine any existing municipal standards for data or data models. It was determined that essentially none exist, and that each municipality typically develops their own standards originating from historic data organization and drafting standards/naming conventions. For now, universal naming and structuring of municipal data is not a reality.

In the past few years, however, the Integrated Cadastral Initiative (ICI) was launched as an attempt to standardize data collection, attribute names, metadata, and geodatabase structure. The ICI is defined as 'a partnership among local governments, utility companies, and provincial agencies for establishing, maintaining, and distributing the ICI database'. The ICI database contains province-wide data for the integrated parcel fabric and related parcel links (i.e. utilities), and permits easier data sharing between participating groups. Until now, of Lower Mainland municipalities, only Pitt Meadows and Surrey have joined the ICI family.

Data standardization was a high priority for OD. In an attempt to standardize the data, OD followed the ICI naming attribute convention (data name, data type and data size), and more generally the "Integrated Cadastral Fabric (ICF) Standards and Specifications" standards, wherever possible. Metadata standards were based on the ArcCatalog FGDC/ESRI metadata structure, currently used by the Victoria Capital Regional District under the guidance of the Integrated Cadastral Initiative Society (ICIS).