3.2.5 Developed Impervious Cover

A key objective in this project is a description of the land cover that adequately represents the hydrologic characteristics of the landscape. While the five part classification does a good job at generally describing the landscape, field excursions revealed significant variability in the amount of impervious cover in developed areas. While a singular imperviousness value is assigned to the other four classes, a further description of the variability within the developed class is required.

Field observations, literature review and airphoto interpretation (API) are used to differentiate among the different types of residential land uses in the subwatershed. The field observations provide a qualitative description of existing residential land use types and knowledge of the land use patterns and landscape context within the watershed. In addition, review and analysis of planning documents for the watershed aided in class definition. Based on these observations and information, a preliminary classification of residential land uses is defined for air photo interpretation.

The general approach to manual airphoto interpretation follows the guidelines outlined by Campbell (1983) (see Table 3). Aerial photographs are acquired from the archives of the MTRCA. The photographs were flown in April, 1993 and have a scale of 1: 800. Based on field observations and trial interpretations, five attributes were defined to characterize the form of residential development. These attributes include: houses and buildings; grass; trees; shrubs and other vegetation; paved areas including parking lots, roads, driveways. The size, shape, shadows, texture, tone, patterns, and location are visual attributes used to interpret elements in the photographs.

Initially, five different residential land use classes were identified, differentiated by the decreasing number of houses per unit area , increasing lot size, and decreasing proportion of impervious areas within the land use class. The low density residential development categories are labeled as R1, R2, R3, R4 and Rf, where R1 to R4 ranges from highest to lowest density, and Rf is rural-residential farm.

Fifteen training areas, three sets of the five representative residential land use classes are located in the East Humber subwatershed in the City of Vaughan, Richmond Hill and King City. Figure 28 indicates the location of these sites. These sites were selected to ensure the form of residential development throughout the subwatershed was represented. A template representing a circular area of approximately 8 ha on the air photo overlay the selected residential classes. The template was divided in a grid with a cell size corresponding to 900 m2 (30m x 30m), on the ground. This area corresponds to the spatial resolution of each pixel of the Thematic Mapper (TM) image. Each cell of the template was characterized in quartiles described by one of the five attributes. The areas were summed and percentage impervious was based on the sum of paved areas and housing and building. Gross housing densities were also calculated.

A survey of the subwatershed indicates that the land use is predominately agricultural. In developed areas, low-density residential development , with few commercial and institutional uses (e.g. municipal offices, schools) comprise the remainder of the land use. Large expanses of natural areas exist throughout the subwatershed in the form of conservation areas (Boyd and Kortright Conservation areas), and vegetated valley and stream corridors. Single-detached and semi-detached dwellings are characteristic of the area, varying in lot size and density. Infill development with estate homes was also existent. Despite a singular form of residential development (e.g. single family detached), there appeared to be a distinctiveness in the different forms of low-density residential development (e.g. lot size, building footprint, driveway, width of road within subdivision).

The observations correspond with the description of existing land uses within the watershed as described in municipal and agency planning reports. A summary of nine land uses are given by Marshal Macklin Monaghan (1996) and is listed in Table 4. In developed areas of the northeast subwatershed, the City of Vaughan which comprises Woodbridge, Maple and Kleinburg, approximately 88%, 98% and 96%, respectively are single family detached dwellings (City of Vaughan 1997). Greater differentiation of the of low and medium density development is given in municipal planning documents. For example, King City permits three different densities of classified low-density development and two estate development residential densities (Township of King 1997). In Richmond Hill in the Oak Ridges community, various densities are permitted with a minimum average net density of 10 units per ha along Wilcox Lake, up to a maximum net density of 60 units per hectare in a commercial, medium residential area along King Road (Richmond Hill Planning Department 1995). Table 5 provides examples of the permitted densities for residential development.

Following a grouping of percent imperviousness and housing densities with the initial five residential land uses, the classification scheme was revised by merging two of the highest low density residential classes into a single category, resulting in four residential land use classes for the north-east subwatershed. The resultant categories are labeled as R1, R2, R3 and Rf and are listed in Table 6.

The coincident tables indicate that both map layers need to be used to define existing land use within the subwatershed. A low correspondence exists for all categories with the highest for the agriculture category. The residential land use category is largely explained by the agricultural land cover classification and secondly by the urban residential land cover class. From this analysis and visually comparing the two map layers, a possible reason for the lack of coincidence between these two layers is that the land use map describes designated land uses for existing as well as future land use. Thus, greater expanse of residential areas are assigned to the land use map layer, consistent with future development proposal in the area (City of Vaughan 1997).