The Canada Census is collected by Statistics Canada every 5 years. It contains a plethora of socio-economic data about the Canadian population and is useful across many disciplines. In Geographic Information Science (GIS), census data is particularly useful because of its inherent integration with a spatial component. This spatial component, that is stored with the census attribute data is often used to conduct spatial analysis. A hierarchical structure summary of all Canada Census data can be seen in Figure 1.1. The following eight levels of aggregation are present at the spatial level:

1. Enumeration Areas (EA) 2. Census Tracts (CT) 3. Federal Electoral Districts (FED) 4. Consolidated Census Subdivisions (CCS) 5. Census Subdivisions (CSD) 6. Census Divisions (CD) 7. Census Metropolitan Areas (CMA) 8. Provinces (PR) The attribute data is organized into the following categories: 1. Age, Sex and Families 2. Immigration 3. Mother Tongue 4. Labour Market Activities 5. Aboriginal 6. Education and Mobility 7. Families 8. Ethnic Origin 9. Sources of Income

Census data may be purchased from Statistics Canada in a wide variety of media formats. Originally, our group accessed the census data in a proprietary format (.ivt) compatible with Ivation's 20/20 Browser. Prior to the completion of our project, the 1996 census data was primarily accessed by researchers and students of SFU through the Research Data Library. The conversion and extraction of data from Beyond 20/20 for GIS is difficult and time consuming for all but the most experienced users. This browser works to reduce the amount to time GIS users have to perform analysis. This could also work to raise the cost of a given research project. The elimination of this browser from the hands of SFU students and researchers is one of the main goals of our project, which will be covered in more detail in Project Goals.

anchor

It was important for our group to establish a good understanding of each census level in order to maintain all inherent relationships upon creation of the geodatabase. The following definitions apply to all census levels that we worked with throughout the project.

A Federal Electoral District (FED) is any region that elects a member to represent them in the federal House of Commons. There are a total of 295 FEDs in Canada.

Census Divisions are derived from the term applied to areas established by provincial law. They are the intermediate geographic areas between the municipality (Census Subdivision) and the province level. Census Divisions are representative of counties, regional districts, regional municipalities and types of provicially legislated areas.

Census Subdivisions (CSD) are areas termed as municipalities or their equivalent by provincial legislation. Example of CSDs include Indian reserves, Indian settlements and unorganized territories.

Census Metropolitan Areas (CMA) are urban cores or very large urban areas combined with nearby urban or rural settlements which have a strong dependency both socially and economically on the core. An example of this is the Vancouver CMA, which reaches all the way to Maple Ridge. The requirement for this division is that it must include an urban area with a population of at least 100,000 in the last census.

Census Agglomeration Areas (CA) are derived in a similar fashion to CMAs, but are only required to meet an urban core size of 10,000. Within our database, the only Census Agglomerations included are those that exceeded an urban core population of 50,000 at the time of the last census. This is the minimum population requirement for CAs to be subdivided into Census Tracts.

Census Tracts (CT) are small geographic areas that represent urban or rural communities, derived from CMAs or CAs. All CMAs are subdivided into CTs, however only the CAs with a population greater than 50,000 are divided into CTs.

Enumeration Areas (EA) are the smallest geographic census unit. These areas cover the entire country and are defined by the area that each census representative canvasses. The maximum number of homes and EA encompasses is usually 440 in urban areas and 125 in rural regions.

anchor

Many different pieces of software were used for the construction and delivery of our geodatabase. The following is an outline to each software package used and its applied functionality. The specific implementation of this software towards our project goals is covered in the methodology section.

ArcGIS is a structured system of GIS software. It offers a wide variety of options in terms of interoperability and practical functionality that benefit our project. ArcSDE, ArcCatalog, ArcMap, ArcToolbox and ArcIMS are the five main components of ArcGIS which are discussed below on an individual basis. For the construction of our geodatabase, conversion tools within ArcToolbox and ArcCatalog were of particular help. Figure 1.2 outlines the integrated nature of the ArcGIS package. ArcCatalog's personal geodatabase is a major improvement from the database management systems available in previous software packages. This fact alone led to the obvious choice of implementing the ArcGIS software. To learn more about this software, click the logo in Figure 1.2.

The Beyond 20/20 browser is proprietary software that gives the user the ability to view download our statistical data. The 1996 Canada Census data is disseminated in this format. Figure 1.3 provides a screenshot of a 20/20 browser window. 20/20 has the ability to store tables titles up to 255 characters and up to 64,000 dimension (fields). These capabilities are useful when long descriptive names are used for fields, which is often the case with census data. Labour Market Activities, for example contains 540 fields within the Canada Census data. As mentioned earlier, we have come to conclusions about 20/20's lack of user-friendliness. This is covered in Project Limitations.

ArcCatalog is used for accessing and managing data. Folders and databases are loaded into ArcCatalog where they are registered for structured access. A browse function is available in this program that allows for the previewing of maps and data. In addition, Metadata is created and edited in ArcCatalog. This component of ArcGIS was the base for the entire project, allowing for the creation and management of our geodatabase. The details pertaining to the creation of our geodatabase can be found in Methodology.

ArcMap is used to visualize important trends within a given dataset. Data relationships are often built using this component of ArcGIS. The user interface of ArcCatalog can be built and customized through the use of Visual Basic for Applications (VBA). ArcToolbox is used for the geoprocessing and conversion of data. Using this component of ArcGIS, topology can be created and maintained, and map sheets can be clipped and joined. The alteration of co-ordinate reference systems can also be accomplished using this software component. Arc Macro Language (AML) scripts can be written within ArcToolbox to customize this process.

Arc Internet Map Server (IMS) was developed by the Environmental Systems Research Institute (ESRI) to allow GIS functionality and spatial processes to migrate to a web-based platform. The diverse uses of ESRI's ArcIMS include real-time analysis and exchange of data, with great flexibility and opportunity for full customization. Figure 1.4 gives an example of how ArcIMS provides GIS on the web in the form of a "viewer." ArcIMS was introduced as an "extension" of our original project goals. It provides a gateway to our census data through visualization and web flexibility. ArcIMS functionality for the 1996 Canada Census is detailed in Project Goals.

ArcObjects is a new platform existing within the ArcGIS package. It helps to provide customization with a full range of functionality to any software developer. ArcObjects was developed using the Microsoft Component Object Model (COM) client-server technology.

This is a new model that was developed for ArcGIS, which works to incorporate spatial data into a given database. Regular relational databases only stores data in attribute form, so this is a great advantage to have. The geodatabase works similar to the coverage model in ArcInfo, yet it also topologically integrates the feature classes. This model provides support for complex networks and relationships between feature classes and other object-oriented features. There are two main types of geodatabase - personal and multi-user. The personal geodatabase is a collection of tables and coverages in ArcCatalog. The multi-user geodatabase is also known as an ArcSDE database.Arc Spatial Database Extension (SDE) connections provide an open interface to a relational database, compatible with various database platforms such as Oracle and Microsoft SQL Server. The ArcSDE connection also facilitates ease of downloading and multiple user access through ArcIMS and works to enhance the performance of a given database.