Defining the built boundary

The method (summarized in Appendix B) uses Landsat 5 Thematic Mapper imagery and a series of semi-automated processes that classify every pixel in the image as either "urban," "non-urban," or "water." The process semi-automatically detects built surfaces such as buildings, roofs, roads, or parking lots--these features are defined as "urban" in this process.

Our intent was to define the consolidated urban area; therefore, we were not interested in capturing small, isolated developments in the countryside, nor in leaving "holes" in the urban fabric caused by small pockets of green space or undeveloped land. Lawns, parks, or conservation areas were reclassified as "urban" if they were completely surrounded by built land. (A similar step has also been incorporated in the Province's procedure for defining a built boundary.)

Figure 5a shows the urban area data set derived from 2006 imagery. We plotted this 2006 boundary so that we could compare our built boundary to the data set produced for the Province that will be used to measure intensification and monitor progress towards meeting the target rate.