Contiguously Settled Areas: Boundaries and User Documentation

Skip to text

Text begins

1  Summary

This product contains contiguously settled area (CSA) boundaries for a subset of Canadian population centres^Note  for 2010 and 2020.

The CSA boundaries are derived from land cover data and represent the geographic extent of settled areas based on their physical footprint on the landscape. The boundaries can be used for reference, mapping and spatial analysis of settled areas and urban ecosystems.

The CSA boundaries are created and maintained under the umbrella of the Census of Environment,^Note  and will support Statistics Canada’s ecosystem accounting efforts following the System of Environmental-Economic Accounting — Ecosystem Accounting (SEEA EA) framework.^Note  The primary purpose of the CSA boundaries is to represent the extent of the contiguous urban footprint and allow for the measurement of settled area expansion over time. They will also contribute to the urban and industrial ecosystem class^Note  in Statistics Canada’s ecosystem accounting efforts. The boundaries will be used for other types of urban ecosystem analyses, including measures of urban ecosystem condition and services. The CSA boundaries may also be used for urban thematic accounts or to inform where users of ecosystem services are located.

Moving forward, CSA boundaries will be updated on an ongoing basis.

CSA boundary files and documentation (ZIP file)

2  User documentation

2.1  Background

Settled areas represent the areas where people live and work and are part of the intensive land-use biome class of the International Union for the Conservation of Nature’s Global Ecosystem Typology, which is the proposed reference classification for classifying ecosystems following the SEEA EA. While they account for a small percentage of the overall land area in Canada, settled areas represent the places where natural ecosystems are the most modified and where the demand for ecosystem services is most concentrated.

The expansion of settled areas has implications for the surrounding environment and ecosystems. While the permanent conversion of land to settled areas provides benefits to humans in terms of providing places to live and work, there are environmental trade-offs, including the loss or fragmentation of forests and the loss of high-quality agricultural land. Habitat loss, often caused by urbanization, is the most common threat facing species-at-risk (Ventor et al., 2006; McCune et al., 2013).

To measure urban expansion and understand how the physical form of the settled area is changing, it is first necessary to delineate the footprint of settled areas. Because administrative and statistical boundaries representing settled areas do not necessarily represent urban morphology or change in morphology over time (European Environment Agency, 2011), they are inappropriate for some types of analysis such as measuring settled area expansion.

Census metropolitan areas, census subdivisions and population centres are some examples of statistical geographies representing metropolitan areas, cities, towns, or densely populated areas in Canada. These boundaries, while valuable for many purposes, often overbound the physically settled area; that is, they include other land covers and uses, such as forest and agriculture, that occur on the periphery of the main settled area. Conversely, administrative and statistical boundaries may also underbound the settled area, by not capturing settled areas that are contiguous with the core urban area. Further, their relative size and the proportion of settled area within these boundaries can vary from region to region, making comparisons complicated and possibly misleading.

An example of overbounding and underbounding of a population centre is shown in Figure 1, while a comparison of the proportion of settled and non-settled areas in a selection of census metropolitan areas or census agglomerations is shown in Table 1. These issues make it problematic to use administrative and statistical boundaries to measure settlement form and to conduct some types of analysis on urban ecosystems.

The contiguously settled area boundaries aim to represent the morphology of settled areas as their physical footprint on the landscape. Data derived from Earth observation (EO) are useful for delineating settled area footprints because they are detailed, comprehensive, up-to-date, and not subject to the limitations associated with administrative or statistical boundaries (Wang et al., 2017; Neptis Foundation, 2010). CSA boundaries use EO-derived land cover data to delineate settled areas morphologically, including contiguous development outside the urban core while excluding natural and semi-natural land covers on the periphery of the main settled area. An example of a CSA is shown in Figure 1.

Since the CSA approach creates representations of settled areas that are based on morphology rather than on administrative or statistical spatial definitions, it is relatively consistent over space and allows for comparisons across regions. Similar approaches for measuring growth in settled areas have been used by others including national statistical offices, the European Environment Agency and non-governmental organizations. The United Kingdom’s Office for National Statistics, for example, uses “Built Up Areas” created by the Ordnance Survey as a geographical area for which census data are tabulated and to define the urban extent for natural capital accounts of urban areas (Ordnance Survey, 2023; Office for National Statistics, 2019). The European Environment Agency has defined “Urban Morphological Zones” based on built-up land cover, from which urban growth is measured over time (European Environment Agency, 2011), and the Neptis Foundation has used land cover to create an urban land base to assess urban growth patterns in Toronto, Calgary and Vancouver (Neptis Foundation, 2010).

2.2  Data sources and methodology

Contiguously settled area boundaries were produced for population centres with a 2021 population greater than 5,000 people or for those that fell within a census metropolitan area or census agglomeration (Statistics Canada, 2022a).

The CSA boundaries are derived from Agriculture and Agri-Food Canada’s (AAFC) 2010 and 2020 Semi-decadal Land Use Time Series data products (AAFC, 2021). These data have a spatial resolution of 30 metres and are based on high quality spatial datasets, many of which are derived from Earth observation. For each CSA year, the land use product for the corresponding year was used as the base data (i.e., the 2010 Semi-decadal Land Use 2010 product was used as the base data for the 2010 CSA boundaries). Because the Semi-decadal Land Use Time Series does not have coverage for areas north of 60° North latitude, CSA boundaries were not created for the population centres of Iqaluit, Whitehorse and Yellowknife although they met the criteria of having a population greater than 5,000 people.

Settled areas were identified by extracting pixels from the Semi-decadal Land Use Time Series that were classified as one of the following classes: settlement, high reflectance settlement, settlement forest, roads, vegetated settlement or very high reflectance settlement. Examples of land covers and uses included in these classes include residential, commercial, industrial or transportation areas; bare pavement; vegetated areas associated with settlements, including parks, lawns, and golf courses; and settlement areas mostly covered by tree canopy. A more detailed description of the settlement classes is provided in Appendix 1.

The extracted settlement pixels were collapsed into a single settlement class, which was then converted to vector format. Shared boundaries were dissolved and polygons less then 3,600 square metres (four pixels) were removed from further analysis.^Note 

The 2021 population centre (Statistics Canada, 2022b) served as the starting point for identifying the core area of the CSA. For each population centre, the settled area intersecting that population centre was selected and edited to create a CSA boundary according to a set of delineation guidelines, explained below. Visuals supporting the guidelines are provided in the appendices.

Delineation guidelines for contiguously settled area boundaries

1. Settled areas that are not contiguous with the core settled area are generally removed. However, in the following situations, non-contiguous areas are included in the CSA:
   1. A settled area that is separated into multiple parts because of the physical geography of the area, such as a settled area that is divided by a river (see Appendix 2) or located along the shores of a body of water.
   2. Settled areas that are non-contiguous but that are located within the population centre boundaries and have a population of at least 1,000 people.^Note 
   3. If a population centre has a compound name, which occurs when a population centre is made up of two or more principal census subdivisions (e.g., St. Catharines–Niagara Falls), the non-contiguous settled areas of each principal census subdivision are retained.
   4. Other exceptions, such as relatively large settled area polygons that are separated from the core settled area by a very short distance (e.g., a width of one or two pixels), are made on a case-by-case basis. Ancillary datasets (including population centre boundaries, population counts, roads, water bodies and Google Earth imagery) are used to support these decisions.
2. The CSA boundaries are truncated when the settled area is reduced to a width of three or fewer pixels (see Appendix 3). Areas narrower than this width are considered as transportation connectors between settled areas and are excluded from the CSA.
   1. Before applying the three-pixel rule, natural and semi-natural land covers that are less than 10 hectares and surrounded by settled areas are filled in.^Note  This is to account for small fragments of natural and semi-natural land such as those formed by highway interchanges. These areas are retained in the CSA even if the settled area that encloses them is narrower than three pixels in width (see guideline 3).
   2. Some exceptions to this guideline, such as when a relatively large area would have been included in the 2020 CSA but not the 2010 CSA because of a difference in classification of a small number of pixels, are made on a case-by-case basis. Supplementary datasets, such as census population counts and high-resolution imagery, are used to support these decisions.
3. Natural and semi-natural land covers enclosed by settled areas are included within the CSA boundary. As a result, the CSA includes land covers other than built-up areas, such as forest, water, cropland and grassland, when these areas are predominantly surrounded by settled areas at least three pixels in width (see Appendix 4).
4. If the CSAs of two or more population centres are attached (i.e., the settlement corridor linking them is wider than three pixels), one CSA is created for the entire area (Appendix 5). These areas are called “agglomerated CSAs” and are flagged in the attribute table.

2.3  Using the contiguously settled area boundaries to measure the expansion of settled areas

Urban growth can occur primarily in two ways: densification and expansion (Mohajeri et al., 2015). Densification can refer to new built-up areas within the existing boundaries of a settlement or the increasing density of population or buildings within a settlement, while expansion refers to the addition of new settled areas at the margins of the present boundaries of a settlement (Mohajeri et al., 2015). Using the CSAs as the boundaries of a settlement at one time point, it is possible to measure urban expansion beyond the boundaries over time.

CSA boundaries can be used to measure settled area expansion in two ways: growth of the CSA area over time, which is indicative of the growth of the settled area as an ecosystem, and growth in land cover converted to built-up in the area surrounding the CSA.

• Growth in the CSA can be measured by subtracting the area of the CSA at an initial time point from the area of the corresponding CSA at a later point. It is important to note that increases in the area of a CSA for a given settled area over time do not necessarily imply a change in land cover for the entire area of change. They may also be because of natural and semi-natural land cover classes becoming enclosed by settled area growth or areas meeting the CSA delineation guidelines that did not meet the rules at an earlier time point (e.g., a settled area becoming contiguous with the core settled area that was previously non-contiguous).
• The CSA can also be used as the boundary from which to measure land cover conversions to built-up (defined as one of the following classes: settlement, high reflectance settlement, settlement forest, roads, vegetated settlement or very high reflectance settlement from the AAFC’s Semi-decadal Land Use Time Series). Buffers can be created around each CSA and growth areas calculated within these buffers by identifying new areas classified as built-up.

The two measures of settled area expansion—growth in the CSA and growth in the built-up area within a buffer around the CSA—are available from Table 38-10-0163-01.^Note 

2.4  Attribute descriptions

A list of attributes of the CSA boundary files and their descriptions is shown in Table 2.

CSA unique identifiers (CSAUID) are unique four-digit codes that are assigned sequentially based on alphabetical order of the associated population centre name upon creation of a CSA. These codes remain consistent between years. If a CSA becomes agglomerated with another CSA over time, the CSAUID of the population centre with the larger population is retained, while that of the other is retired.

2.5  Notes to users

1. The CSA boundaries identified in this product do not reflect all settled areas in Canada. CSAs were created for 540 of 1,026 population centres, and do not include settled areas that are non-contiguous with these areas.
2. Increases in the area of a CSA for a given settled area over time do not necessarily imply a change in land cover for the entire area of change. Increases to the CSA area over time involve the conversion of land cover to built-up area at the periphery of a CSA but may also result from two other factors. First, natural and semi-natural land cover classes (such as forest, water, cropland or grassland) that were on the periphery of the CSA can become enclosed by settled area growth over time. Second, an existing built-up area that was not initially contiguous may meet the CSA delineation guidelines at a later time point. For example, a settled area connected to the core CSA by a narrow transportation corridor at an earlier time point may become connected by a wider corridor of more than three pixels at a later point, becoming part of the CSA.
3. Over time, non-agglomerated CSAs may become agglomerated CSAs as settled areas increase and become merged with adjacent settled areas.  

2.6  Maintenance

Moving forward, CSA boundaries will be maintained on an ongoing basis.

2.7  Spatial reference system

The projection information is as follows:

Projected Coordinate System: NAD 1983 Albers
Projection: Albers
False Easting: 6200000.00000000
False Northing: 3000000.00000000
Central Meridian: -91.86666667
Standard Parallel 1: 49.00000000
Standard Parallel 2: 77.00000000
Latitude of Origin: 63.39067500
Linear Unit: Meter

Geographic Coordinate System: GCS North American 1983
Datum: D North American 1983
Prime Meridian: Greenwich

Angular Unit: Degree

2.8  Limitations

The CSA boundaries are based on AAFC’s Semi-decadal Land Use Time Series, which is derived from EO data. Uncertainty is inherent in EO-derived data and some level of misclassification of data is expected although new technologies and methods have resulted in improved data quality over time. Uncertainty in EO-derived data arises from a variety of factors, including scale dependency, classification errors and limitations, and positional and temporal accuracy. Because of these limitations, caution should be used when considering change of CSA boundaries over time, as change in land cover datasets at different time points can result from both real change and classification errors. Users are encouraged to examine the data for their area of interest to better understand its fitness for purpose. Because the Semi-decadal Land Use Time Series data on which the CSA boundaries are based are created from the integration of multiple spatial datasets representing different temporal periods, the CSA boundary files should be considered as circa the year they are named.

3  References

Agriculture and Agri-Food Canada (AAFC). (2021). AAFC Semi-decadal land use time series. [Data set]. Retrieved October 1, 2021 from https://open.canada.ca/data/dataset/fa84a70f-03ad-4946-b0f8-a3b481dd5248

European Environment Agency. (2011, February 16). Analysing and managing urban growth. Retrieved April 12, 2023 from https://www.eea.europa.eu/articles/analysing-and-managing-urban-growth/#urban-morphological-zones

McCune, J. L., Harrower, W. L., Avery-Gomm, S., Brogan, J. M., Csergo, A., Davidson, L. N. K., Garani, A., Halpin, L. R., Lipsen, L. P. J., Lee, C., Nelson, J. C., Prugh, L. R., Stinson, C. M., Whitney, C. K. & Whitton, J.  (2013). Threats to Canadian species at risk: An analysis of finalized recovery strategies, Biological Conservation, 166, 254-265. Retrieved April 20, 2023 from https://doi.org/10.1016/j.biocon.2013.07.006

Mohajeri, N. Gudmundsson, A., & Scartezzini, J. L. (2015). Expansion and densification of cities: Linking urban form to urban ecology. [Conference paper, Future Buildings & Districts Sustainability from Nano to Urban Scale]. Retrieved May 5, 2023 from https://www.researchgate.net/publication/280244337_Expansion_and_densification_of_cities_linking_urban_form_to_urban_ecology

Neptis Foundation. (2010). Growing cities: Comparing urban growth and regional growth policies in Calgary, Toronto, and Vancouver. Retrieved March 31, 2023 from https://neptis.org/sites/default/files/growing_cities/growing_cities_report_interior_20100721_print.pdf

Office for National Statistics. (2019). UK natural capital: Urban accounts. Retrieved March 31, 2023 from https://www.ons.gov.uk/economy/environmentalaccounts/bulletins/uknaturalcapital/urbanaccounts#extent-of-urban-space-in-the-uk

Ordnance Survey. (2023). OS Open built up areas. Retrieved March 16, 2023 from https://www.ordnancesurvey.co.uk/products/os-open-built-up-areas

Statistics Canada. (2022a). Dictionary, Census of Population, 2021 – Census metropolitan area (CMA) and census agglomeration (CA), Catalogue no. 98-301-X. Retrieved June 6, 2023 from https://www12.statcan.gc.ca/census-recensement/2021/ref/dict/az/Definition-eng.cfm?ID=geo009

Statistics Canada. (2022b). 2021 Census Boundary files, Catalogue no. 92-166-X. Retrieved June 6, 2023 from https://www12.statcan.gc.ca/census-recensement/2021/geo/sip-pis/boundary-limites/index2021-eng.cfm?year=21

Ventor, O., Brodeur, N. N., Nemiroff, L., Belland, B., Dolinsek, I .J., & Grant, J. W. A. (2006). Threats to endangered species in Canada, BioScience, 56(11), 903-910. Retrieved April 20, 2023 from https://doi.org/10.1641/0006-3568(2006)56[903:TTESIC]2.0.CO;2

Wang, P., Huang, C., Brown de Colstoun, E. C., Tilton, J. C. & Tan, B. (2017). Documentation for the global human built-up and settlement extent (HBASE) dataset from Landsat. NASA Socioeconomic Data and Applications Centre. Retrieved March 31, 2023 from https://sedac.ciesin.columbia.edu/data/set/ulandsat-hbase-v1/docs

4  Appendices