Hiedi Zutter

"Evaluation and application of automated methods to measure sky view factors for urban areas."
BSES Senior Research Project 1999.
Original Paper (PDF)

Introduction

An urban area is comprised of many buildings that when in close proximity to each other form a basic urban surface unit called an urban canyon. The urban canyon is thus a fundamental unit comprising the urban canopy layer, which is defined as the layer of air from ground to roof-level in an urban area.

Urban canyons vary in geometry based on the heights, lengths, and spacing of the buildings that define them. The geometric relationships within them can influence the absorption and emission of incoming solar and outgoing longwave radiation within the urban area and can have a significant impact on the energy balance and temperature of an urban area. An urban canyon is comprised of the walls and ground (road, garden, etc.) between two adjacent buildings as well as the canyon-air volume, which has three sides with active surfaces (walls and ground) and three open sides (ends and top) (Oke 1987). Studying urban canyons and their respective geometry is worthwhile because of their potential in explaining the urban heat island. An urban heat island occurs when the air in an urban area (city) is warmer than that of the surrounding rural (countryside) area. According to Oke (1987), the air becomes warmer in the urban area because urban geometry and density of development influence processes such as the trapping of incoming solar and outgoing long-wave radiation. Thus, knowledge of canyon climate with respect to canyon geometry and the influence of meteorology help to aid urban street design. While much research has been done on the climate and effects of a single urban canyon, little to no work has been done to compare different canyon geometry and their effects on variables such as sky view factor (SVF) and temperature (of air and surfaces).

The objective of this study, then, was to intercompare canyon geometry (size, orientation/direction) and SVF to see how it impacts air and surface temperatures. Additionally, H:W relationships have traditionally been used to estimate SVF. A number of other alternative methods are evaluated in this study. Studies into urban canyon geometry (canyon orientation and SVF) and its effects on radiation fluxes, and air and surface temperature differences have thus far concerned themselves primarily with only one particular type of urban canyon geometry. However, when these studies are looked at together we begin to see the changes that different height-to-width (H:W) ratios, SVFs, and orientation of canyons can have.

Sky view factor is the fraction of overlying hemisphere occupied by sky. For example, a SVF of one would be for an open site such as a wide field or the top of a tall building where there are no obstructions blocking a portion of the sky from view (Figure 1a). In urban areas, buildings and vegetation decrease the fraction of sky visible from the ground (Figure 1b). Urban geometry, which is then related to SVF, thus has implications in blocking some of the direct beam of radiation from the sun. According to Oke (1987), urban geometry influences the trapping of incoming solar and outgoing longwave radiation. This allows the urban air to become warmer. By day, a canyon is especially good at absorbing heat because its geometry acts to trap reflected solar radiation and its materials are often good heat stores. At night, the loss of heat from within canyons is slowed by the screening of the sky energy sink by the bordering of buildings (Arnfield 1982). Mills and Arnfield (1993) found that, as street canyons become increasingly narrow, they become increasingly isolated in terms of heat exchange with the atmosphere above.

Additionally, Arnfield and Grimmond (1998), using a north-south canyon orientation in a model, found that a larger fraction of available canyon net radiation is stored when wall height is large in relation to floor width. Essentially, the narrower a canyon becomes the less radiation that is emitted back out into space because of trapping and so a larger fraction of the net canyon radiation is stored. Eliasson (1996) found surface temperature to be statistically correlated to SVF. Sky view factor thus has a potential influence on the surface temperatures of urban canyons. A look at the mobile method of measuring SVF provides an interesting way to look at how SVF changes across an urban area.

Figure 24(below) shows the results for the mobile transect, change in SVF with distance from the start of the transect (along route), in downtown Bloomington, IN, through the campus of Indiana University. This Figure has three main areas of particular interest. The first area is from distance 0 meters to about a distance of 1300 meters along the route. This area is representative of downtown Bloomington area, circling the courthouse. This area is relatively open due to wider main streets which are circling the courthouse area of town, which is more open. Thus, the SVFs here are consistently in the range from about 0.68 to about 0.87, which is fairly high for an urban area. The second area of interest is at the distance of about 1400 meters. Here there is a large dive in the SVFs from around 0.90 down to about 0.35, and then back up to 0.90. This is representative of an alley between a 5-level parking garage and another large building. The third area of interest is the area from about 2000 m through to the end. Here the SVF has a much wider range varying from about 0.85 to about 0.27. This area is representative of the campus of Indiana University, which has a marked increase in the number of trees, especially large trees lining the streets and the land to either side of the streets. These trees in addition to many large buildings cause the SVF to drop as low as 0.27.

Conclusions
Studies of urban canyons have shown that canyon geometry and materials that make them up can influence the air and surface temperatures of urban canyons. By comparing different canyon geometries and their respective air and surface temperatures, this study went beyond most previous studies that investigated only one particular urban canyon geometry. Sky view factor varied depending upon the method used to measure it. The method we deemed as most accurate was the automated analysis of Nikon imagery. Using this method, the mobile transect appears to be a valuable method to see how SVF varies along a particular area (route). SVF (from automated analysis of Nikon imagery) and air temperatures and surface temperatures appear to be related, but when canyon orientations are accounted for, this relationship loses significance for the north-south canyons. The size of the sample (only 10 canyons) may not be adequate to determine the significance of the relationship completely. When SVF (from H:W) is compared to air temperatures, the results are such that air temperature and SVF are not linearly related, which may explain the contradiction in what our results indicate vs. the results from previous studies. Further study is necessary, with a larger sample size to draw more sound conclusions. Certainly, urban canyons influence higher temperatures adding to the temperatures experienced within the city. This can be controlled by controlling the sky views of canyons through better planning when constructing buildings and city blocks, especially through the inclusion of vegetation.