authors identify the following indicators at the neighborhood scale: density, land-use
mix, street linkages, the relationship between scales of roads and buildings, and
beauty and attractiveness. Table 4.1 is organized with reference to such studies
(Lawrence and Low 1990; Asami 2001; Handy et al. 2002; Li et al. 2005); it omits
the indicators that refer to a subjective assessment, such as beauty and attractiveness,
and limits the objective indicators showing physical features. Further, the classification in this chapter focuses on the neighborhood scale, and the indicators for the
scale’s features can be organized into four categories: (1) land use, (2) shape of urban
tissue, (3) density, and (4) proportion of buildings by type. Subordinate to each
category is a group of specific indicators that indicate the category’s status. This
study selected the following four indicators that can be visually identified from the
aforementioned categories:
1. Land use: ratio of the area of green space and farmland to unit area
This indicator identifies the proportions of trees, paddy fields, and other green
spaces within and around the settlement.
2. Shape of urban tissue: block shape and building layout
This indicator identifies whether the block shape and the arrangement of
buildings are regular or not.
3. Density: building density
This indicator identifies the building density (the number of buildings in unit
area) in the residential area.
4. Percentage of buildings by type: building height
This indicator identifies the average height of buildings in the residential area.
Earlier classifications were generally limited to (1) and (3), on which one could
easily obtain existing data. However, the inclusion of (2) and (4), which can be
inspected visually, enables a more detailed classification.
4.2.3 Classification Unit
Two main methods can be followed to visually classify residential areas based on
the indicators described Sect. 4.2.2. One is to extract the residential areas having the
same characteristics and map them as vector data. The other method is to divide the
city into meshes of a specific size and classify each mesh according to the indicators.
Since residential areas vary in size and shape, the first method is more accurate than
the second. However, since megacities have vast residential areas, tracing all these
areas requires significant effort and is not practical. Therefore, this study adopted the
second method of classification.
In this approach, the mesh size determines the deviation of the actual condition
from the classification. This is because the mesh size determines the smallest unit of
classification, and it is assumed that one mesh encompasses one residential type.
Therefore, first, this study examined the mesh in two scales, 500 m  500 m and
250 m  250 m. For 500 m  500 m, it was difficult to consider one mesh as a
homogeneous residential area in many cases. In contrast, the number of such cases
4 Diversity and Historical Continuity of the Residential Landscape of a. . .
49
mix, street linkages, the relationship between scales of roads and buildings, and
beauty and attractiveness. Table 4.1 is organized with reference to such studies
(Lawrence and Low 1990; Asami 2001; Handy et al. 2002; Li et al. 2005); it omits
the indicators that refer to a subjective assessment, such as beauty and attractiveness,
and limits the objective indicators showing physical features. Further, the classification in this chapter focuses on the neighborhood scale, and the indicators for the
scale’s features can be organized into four categories: (1) land use, (2) shape of urban
tissue, (3) density, and (4) proportion of buildings by type. Subordinate to each
category is a group of specific indicators that indicate the category’s status. This
study selected the following four indicators that can be visually identified from the
aforementioned categories:
1. Land use: ratio of the area of green space and farmland to unit area
This indicator identifies the proportions of trees, paddy fields, and other green
spaces within and around the settlement.
2. Shape of urban tissue: block shape and building layout
This indicator identifies whether the block shape and the arrangement of
buildings are regular or not.
3. Density: building density
This indicator identifies the building density (the number of buildings in unit
area) in the residential area.
4. Percentage of buildings by type: building height
This indicator identifies the average height of buildings in the residential area.
Earlier classifications were generally limited to (1) and (3), on which one could
easily obtain existing data. However, the inclusion of (2) and (4), which can be
inspected visually, enables a more detailed classification.
4.2.3 Classification Unit
Two main methods can be followed to visually classify residential areas based on
the indicators described Sect. 4.2.2. One is to extract the residential areas having the
same characteristics and map them as vector data. The other method is to divide the
city into meshes of a specific size and classify each mesh according to the indicators.
Since residential areas vary in size and shape, the first method is more accurate than
the second. However, since megacities have vast residential areas, tracing all these
areas requires significant effort and is not practical. Therefore, this study adopted the
second method of classification.
In this approach, the mesh size determines the deviation of the actual condition
from the classification. This is because the mesh size determines the smallest unit of
classification, and it is assumed that one mesh encompasses one residential type.
Therefore, first, this study examined the mesh in two scales, 500 m  500 m and
250 m  250 m. For 500 m  500 m, it was difficult to consider one mesh as a
homogeneous residential area in many cases. In contrast, the number of such cases
4 Diversity and Historical Continuity of the Residential Landscape of a. . .
49