Study on heat island mitigation effect of large-scale greenery using ...
Study on heat island mitigation effect of large-scale greenery using numerical simulation Hom Bahadur Rijal1, Ryozo Ooka1, Hong Huang2, Takeaki Katsuki1 and Byoungchull Oh1 1 2
The University of Tokyo, Tokyo, Japan Tsinghua University, Beijing, China
Corresponding email: [email protected]
SUMMARY The effect of large-scale greenery is expected to mitigate urban heat islands. However, most research is measurement-based, with limited research using numerical simulations. In order to investigate the effect of thermal mitigation by large-scale greenery, a tree model was built into the computational fluid dynamics (CFD) code. The air temperature leeward of an urban block with large-scale greenery is approximately 1.5°C lower than without such greenery. This result showed that large-scale greenery provides an effective way to mitigate urban heat islands.
INTRODUCTION Due to artificial urbanization, urban heat islands have become a serious problem. Greenery is expected to be an effective countermeasure, and much research is available regarding green roofs, green walls, street trees, parks, and woodland, etc. In large-scale greenery especially, the effect of cool air seeping in and the extensive mass as a cool spot are expected to expand in large urban areas. Table 1 shows the thermal mitigation effect of large-scale greenery in existing research [1-20]. The surface and air temperatures in or around large-scale greenery are lower than those of urban areas, and the air temperature of the urban area is reduced by the influence of such greenery. Moreover, greenery has valuable potential for the environment, such as absorbing carbon dioxide and reducing carbon dioxide emissions thanks to a 5 to 15% cut in air-conditioning usage [9, 15, 16]. In addition, atmospheric purification is found in areas of large-scale greenery, and thus NO and NO2 concentrations are lower than urban areas. However, most research is measurement-based, and only limited research based on numerical simulations. In terms of the temperature-reduction effect of greenery, there is much research at ground level (pedestrian area), and only a few attempts to research the vertical direction. If the temperature-reduction effect of large-scale greenery is confirmed in the vertical direction, it is possible to introduce cool outdoor air to buildings by opening windows in the upper part of the building, and thus it may be possible to cut cooling costs. In summer, wind velocity in urban areas is reduced by large-scale greenery, and thus this negative effect needs to be thoroughly investigated. In this research, a tree model is built into the CFD code for a real urban block, and the temperature-reduction effect in the horizontal (pedestrian level) and vertical directions by large-scale greenery are clarified quantitatively. The reduction by large-scale greenery of the wind velocity in urban blocks is also investigated. If large-scale greenery can effectively mitigate urban heat islands, existing green areas can be preserved from the viewpoint of the thermal environment, and this can be applied to new urban planning.
1
Table 1. Comparison of thermal mitigation effect of large-scale greenery Reference
Country
Location
Method
Season
[1]
Japan
Sendai
M
Summer
[2]
Japan
Tokyo
M
Summer
[3]
Japan
Tokyo
M
Summer
[4]
Japan
Tokyo
M
Summer
Air temperature of green areas is 0.5~1.5°C lower than surrounding urban areas. Maximum and average air temperature of the park is 0.8°C and 0.5°C lower than the park entrance. Maximum air temperature of the Meiji Shrine forest is 4°C lower than the surrounding urban block. Air temperature of the park is 1.2~2.1°C lower than the urban block.
[5]
Japan
Tokyo
M
Summer
Air temperature of the park is 1.5~2.5°C lower than the urban block.
[6]
Japan
Tokyo
M
Summer
Air temperature of the Shinjuku Gyoen (park) is 1~3°C lower than the urban area.
[7]
Japan
Tokyo
M
Summer
[8]
Japan
Tokyo
S
Summer
[9]
Japan
Tokyo
S
Summer
Air temperature of large-scale greenery is 1~2°C lower than the urban area. When the large green tracts and the preparation of permeable pavements are introduced, the daily maximum and average air temperature of Tokyo City is 0.4°C and 0.2°C lower. When the green area is 60%, the maximum air temperature is 1.0°C lower.
[10]
Japan
Zushi
S
Summer
When trees are planted, the maximum air temperature in the urban area is 1.6°C lower.
[11] [12]
Japan Japan
Nagoya Kyoto
S M/S
[13]
Japan
Kumamoto
M
Summer Summer Spring / Summer
Air temperature of the urban area is approx. 1.0°C lower during day and night. Air temperature around trees is 0.9°C lower than urban areas. Maximum air temperature within the green area is 3°C lower than surrounding areas. (The surface temperature of the green area is 15~20°C lower than surrounding areas.) Air temperature of the tree canopy is 1.5~2.0°C lower than surrounding areas. (The surface temperature of the tree canopy is 7.8°C lower than concrete.) Air temperature of the park is 0.4~1.7°C lower than urban blocks.
[14]
Singapore
[15]
Singapore
Bishan & Serangoon CWP
M/S
Summer
[16]
Singapore
NUS
M/S
Summer
[17]
Israel
Tel-Aviv
M
Summer
[18]
Germany
Hanover
M
Autumn
[19]
USA
M
Summer
Air temperature (Surface temperature) (°C)
Air temperature of the park is 3~4°C lower than urban blocks. Maximum air temperature within the green area is 4°C lower than outside the green area. The air temperature of urban area is 0.5~2.9°C lower than the reference point. (The maximum surface temperature of the green area is 20°C lower than artificial ground.) When the green area is 65%, the air temperature of urban areas is approx. 2.0°C lower.
St. Louis S Summer Mexico Rainy / [20] Mexico M Air temperature of the park is 1.0~4.0°C lower than urban blocks. City Dry M/S: Measurement/Simulation, CWP: Central & western parts, NUS: National University of Singapore
INVESTIGATION OF THE TREE MODEL The tree has multiple influences on the thermal environment in summer. It effectively reduces solar and long-wave radiation, and generates evaporative latent heat, and thus it is expected to mitigate the thermal environment of urban areas. However, trees produce moisture and reduce wind velocity, and thus it is necessary to investigate these influences. For the purpose of widespread use of the tree model developed by the authors, it is built using the radiation analysis software “RADX” [21] and the CFD software “Star CD”. Heat transfer analysis of the tree (1) Radiation of the tree In modeling the radiation reduction effect of the tree crown, absorption of the radiation flux is used which considers the transmission distance and the leaf area density to radiation flux of the tree crown, and the effective transmittance was calculated from the absorption coefficient. (1) τ eff = exp ( −kax ) where, τ eff is the effective transmittance, k is the absorption coefficient [-], a is the leaf area density [1/m], and
x
is the transmission distance [m].
(2) Fluid resistance of the tree In order to consider the fluid dynamics resistance of the tree, Equation 2 is added to the equation of wind velocity motion. The turbulence energy and dissipation factor were also considered at the same time. 2 (2) −η Cd a ( x1 , x2 , x3 ) ui uj where
is the resistance coefficient of the tree crown (=0.2) [-], a ( x1 , x2 , x3 ) is the leaf area density of the tree crown [m2/m3], and η is the ratio of green coverage (the proportion of the horizontal area covered by the tree out of the horizontal area of mesh including tree) [-]. Cd
2
oC
Top Wind Wind
North
Tree 樹木 crown
East
West South Z
地表園 Ground Y
Bottom
X
Figure 1. The simulation model
Figure 2. The surface temperature of the tree crown
Out of consideration for the sensible heat and latent heat generated by the tree, the latent and sensible heat fluxes that flow from the heat balance equation of the tree crown as obtained by the radiation calculation were build into the generation term of the transport equation of the temperature and the absolute humidity. The details are as per Yoshida et al. [22]. Simple analysis example of the tree (1) The simulation model
N Tree crown C
o
Figure 3. Ground surface temperature
The simulation model is shown in Fig. 1. The computational domain is 10 m × 10 m × 10 m, and the tree size is 3.33 m × 3.33 m × 3.33 m. The inflow wind velocity is 1 m/s, and the inflow air temperature is 31.6°C. The evaluation time is 3 p.m. The tree was assumed to be a cube floating in the air, and the influence of the trunk tree was disregarded in this study. (2) Results The surface temperature distribution of the tree is shown in Fig. 2. Due to the effect of solar radiation at 15:00, the surface temperature at the west side and top of the tree are higher than the south side. The surface temperatures of the east side and north side of the tree are lowest. The surface temperature of the bottom of the tree is high due to the influence of reflection from the ground. The ground surface temperature is more than 10°C lower in the shade of the tree (Fig. 3). This result is similar to existing research [23]. The vertical temperature distribution is low in the tree (Fig. 4). Because of the southwest direction of the sun at 15:00, the surface temperature of the northeast ground is low. The tree becomes an obstacle to the wind, and thus the wind velocity decreased on the inner and leeward side of the tree (Fig. 5). The vertical absolute humidity distribution is high in the centre of the tree (Fig. 6).
oC
Tree crown
Figure 4. Vertical air temperature distribution
Tree crown
m/s
Figure 5. Vertical wind velocity distribution
Tree crown kg/kg’
INVESTIGATION OF LARGE-SCALE GREENERY Investigated area The investigated area is Minami Aoyama, Tokyo. The area is composed of detached houses, high-rise apartment buildings, office buildings and commercial buildings, etc. Trees are planted around the high-rise apartments, and create a comfortable open outdoor space. When we enter an alley 3
Figure 6. Vertical absolute humidity distribution
Table 2. Meteorological conditions Day and time Wind direction Wind velocity Air temperature Latitude Longitude
23 July (15:00) Southwest 3 m/s (74.6 m above ground level) 31.6°C 35.68°N 139.79°E
Table 3. Radiation conditions Building
from the main street, the residential area is extended. Most buildings are made of concrete, with only a few wooden houses. In a few office buildings, the air conditioning is set at ground level, and the heat is discharged in the pedestrian space. Parks are scattered in the targeted area, with the Aoyama graveyard on the west side, and Jingu Gaien and Akasaka Imperial Palace on the north side. In this research, the ‘Aoyama graveyard’ was the model for ‘large-scale greenery’.
Wall Solar reflectance Solar transmittance Thickness Conductance Indoor convection heat transfer coefficient Indoor air temperature Long wavelength emissivity Ground Solar reflectance Long wavelength emissivity
Concrete 0.1 0 0.22 m 1.64 W/(mK) 2
Ground
4.64 W/(m K) Calculation method 26°C In this research, coupled simulation of radiation 0.9 and convection was used [24]. The date and Asphalt 0.1 meteorological conditions are shown in Table 2. 0.9 rd The analysis date is 23 July and the investigated 26°C (0.5 m Underground temperature below ground time is 15:00; i.e. a fine summer’s day. The level) radiation calculation conditions are shown in Table 3. The radiation calculation was Table 4. CFD Conditions (Huang et al. 2005) conducted in an unsteady condition, Standard k-ε model (inclusion of buoyancy and the pre-calculation time is 15 hours. Turbulent model effect) Discretization Finite volume method The surface temperature, which was Algorithm SIMPLE method obtained from the radiation calculation, Differencing scheme First-order upwind scheme Side No slip is used as a boundary condition in the Sky Slip CFD simulation. Wall Generalized logarithmic law The CFD calculation conditions are U = U0(Z/Z0)0.25, Z0=18.0 m, U0=1.0 m/s Z 0: Representative height (met station) (m) shown in Table 4. The turbulent model Inlet (Power law) U0: Wind velocity at Z0 (m/s) is a standard k-ε model. The Z: Height from the ground level (m) U: Wind velocity at Z (m/s) dimensions of the computational k=1.5(IU)2, I=0.1 Turbulent kinetic energy domain are 1,591 m (x) × 1,668 m (y) × (k) I: Turbulent intensity 600 m (z) with a mesh size of 354,070 Turbulent kinetic energy ε=Cμk1.5/l, Cμ=0.09 dissipation rate (ε) for the large-scale greenery model. In l=4(Cμk)0.5Z00.25Z0.75/U0 Turbulence length scale (l) order to investigate the effect of such greenery, CFD simulations were performed with and without large-scale greenery. The area of large-scale greenery is assumed to be 195,255 m2. The height of the tree trunk is 4 m above ground and the height of the tree crown is 15 m.
Results and discussions Surface temperature distribution The distribution of the surface temperature, which was obtained from the radiation calculation at 15:00, is shown in Fig. 7. Due to the influence of solar radiation, the variation in surface temperatures of the buildings or ground is high in shaded and non-shaded areas. The sun is in a southwest direction at 15:00, and thus the surface temperature of the west-facing walls is higher than the south-facing walls. The surface temperature of the roof rises to 56°C due to the effect of direct solar radiation. The surface temperature is about 50°C on the west-facing walls, and about 38°C on the south-facing walls. The surface temperature of the ground is about 54°C in areas exposed to the sun, and about 42°C in shaded areas. The surface temperature of large-scale greenery is about 28°C due to the effect of evaporation. The result is similar to existing research that the surface of the green area is 15 to 20°C lower than the surrounding area (Table 1, Fig. 8). 4
Air temperature (1) Horizontal air temperature distribution The horizontal air temperature distribution at 1.5 m above ground level is shown in Fig. 9. The air temperature difference between with and without greenery is shown in Fig. 10. The mean and standard deviation are shown in Table 5. The air temperature is high around the area where the surface temperature is also high. The air temperature of the middle part of the greenery is 4.0°C lower than if without greenery. The air temperature of the leeward side of the urban area of large-scale greenery is about 1.5°C lower than without greenery, and thus large-scale greenery can be considered effective as urban heat island mitigation. These results are similar to existing research (Table 1). The cool air of the greenery area flows to the inner part of the residential area, and the Table 5. Mean and standard deviation (SD) temperature reduction is as much as near to Air temperature (°C) Height Mesh the greenery area (Fig. 10). The effect of (m) number Without greenery (T) With greenery (Tg) the air temperature reduction from largeMean SD Mean SD 1.5 8,801 32.5 0.4 31.2 1.2 scale greenery extends to about 500 m into 5 8,801 32.4 0.4 31.2 1.2 the urban blocks, which is in line with 10 8,801 32.3 0.4 31.3 1.0 existing research of 80 to 1,000 m (Table 6). 20 8,801 32.2 0.4 31.4 0.6 Due to poor ventilation, heat more Table 6. Range of air temperature reduction easily builds up in the densest spaces effect by large-scale greenery in summer (alleys, narrow spaces) and thus the air Place Method Range (m) Reference temperature is highest in such locales. Tokyo Existing research (M) 100 to 1000 [5] Conversely, the ventilation is good in Tokyo Existing research (M) 80 to 250 [6] Japanese city Existing research (S) 300 to 400 [25] open spaces (wide streets, open land) Singapore Existing research (S) Approx. 350 [15] and thus the air temperature is Israel Existing research (M) 100 [17] comparatively lower there. Tokyo Present (S) 500 Present study M: Measurement, S: Simulation 55
oC
Temperature (o C)
50 45 40 35
Urban area (Concrete surface temp.) Urban area (Air temp. above 1.5 m GL) Green area (Air temp. above 1.5 m GL) Green area (Ground surface temp.)
30 25 20 7:00 11:00 15:00 19:00 23:00 3:00 7:00 Time (-)
Figure 8. Measured temperatures from existing research [2]
(a) Without greenery (b) With greenery Figure 7. Surface temperature distribution
N
N
oC
oC
A
A’
A
A’
(a) Without greenery (b) With greenery Figure 9. Air temperature distribution (GL +1.5 m)
5
Figure 10. Air temperature (difference between with and without greenery)
(2) Vertical air temperature distribution The vertical temperature distribution decreases with increased height (Fig. 11). This result shows that an unstable layer is formed near ground level. In the absence of greenery, the air temperature is high across the entire calculated area. On the other hand, when an area of greenery exists, the air C temperature remains low up to building height. Green area The vertical temperature distribution of the entire calculated area is shown (a) Without greenery (b) With greenery in Fig. 12. With largeFigure 11. Vertical air temperature distribution (A-A’ section) scale greenery, the air 31.6 temperature is as low as 250 T = -0.233T + 38.8 Without green (n=21, R = 0.87, p
The University of Tokyo, Tokyo, Japan Tsinghua University, Beijing, China
Corresponding email: [email protected]
SUMMARY The effect of large-scale greenery is expected to mitigate urban heat islands. However, most research is measurement-based, with limited research using numerical simulations. In order to investigate the effect of thermal mitigation by large-scale greenery, a tree model was built into the computational fluid dynamics (CFD) code. The air temperature leeward of an urban block with large-scale greenery is approximately 1.5°C lower than without such greenery. This result showed that large-scale greenery provides an effective way to mitigate urban heat islands.
INTRODUCTION Due to artificial urbanization, urban heat islands have become a serious problem. Greenery is expected to be an effective countermeasure, and much research is available regarding green roofs, green walls, street trees, parks, and woodland, etc. In large-scale greenery especially, the effect of cool air seeping in and the extensive mass as a cool spot are expected to expand in large urban areas. Table 1 shows the thermal mitigation effect of large-scale greenery in existing research [1-20]. The surface and air temperatures in or around large-scale greenery are lower than those of urban areas, and the air temperature of the urban area is reduced by the influence of such greenery. Moreover, greenery has valuable potential for the environment, such as absorbing carbon dioxide and reducing carbon dioxide emissions thanks to a 5 to 15% cut in air-conditioning usage [9, 15, 16]. In addition, atmospheric purification is found in areas of large-scale greenery, and thus NO and NO2 concentrations are lower than urban areas. However, most research is measurement-based, and only limited research based on numerical simulations. In terms of the temperature-reduction effect of greenery, there is much research at ground level (pedestrian area), and only a few attempts to research the vertical direction. If the temperature-reduction effect of large-scale greenery is confirmed in the vertical direction, it is possible to introduce cool outdoor air to buildings by opening windows in the upper part of the building, and thus it may be possible to cut cooling costs. In summer, wind velocity in urban areas is reduced by large-scale greenery, and thus this negative effect needs to be thoroughly investigated. In this research, a tree model is built into the CFD code for a real urban block, and the temperature-reduction effect in the horizontal (pedestrian level) and vertical directions by large-scale greenery are clarified quantitatively. The reduction by large-scale greenery of the wind velocity in urban blocks is also investigated. If large-scale greenery can effectively mitigate urban heat islands, existing green areas can be preserved from the viewpoint of the thermal environment, and this can be applied to new urban planning.
1
Table 1. Comparison of thermal mitigation effect of large-scale greenery Reference
Country
Location
Method
Season
[1]
Japan
Sendai
M
Summer
[2]
Japan
Tokyo
M
Summer
[3]
Japan
Tokyo
M
Summer
[4]
Japan
Tokyo
M
Summer
Air temperature of green areas is 0.5~1.5°C lower than surrounding urban areas. Maximum and average air temperature of the park is 0.8°C and 0.5°C lower than the park entrance. Maximum air temperature of the Meiji Shrine forest is 4°C lower than the surrounding urban block. Air temperature of the park is 1.2~2.1°C lower than the urban block.
[5]
Japan
Tokyo
M
Summer
Air temperature of the park is 1.5~2.5°C lower than the urban block.
[6]
Japan
Tokyo
M
Summer
Air temperature of the Shinjuku Gyoen (park) is 1~3°C lower than the urban area.
[7]
Japan
Tokyo
M
Summer
[8]
Japan
Tokyo
S
Summer
[9]
Japan
Tokyo
S
Summer
Air temperature of large-scale greenery is 1~2°C lower than the urban area. When the large green tracts and the preparation of permeable pavements are introduced, the daily maximum and average air temperature of Tokyo City is 0.4°C and 0.2°C lower. When the green area is 60%, the maximum air temperature is 1.0°C lower.
[10]
Japan
Zushi
S
Summer
When trees are planted, the maximum air temperature in the urban area is 1.6°C lower.
[11] [12]
Japan Japan
Nagoya Kyoto
S M/S
[13]
Japan
Kumamoto
M
Summer Summer Spring / Summer
Air temperature of the urban area is approx. 1.0°C lower during day and night. Air temperature around trees is 0.9°C lower than urban areas. Maximum air temperature within the green area is 3°C lower than surrounding areas. (The surface temperature of the green area is 15~20°C lower than surrounding areas.) Air temperature of the tree canopy is 1.5~2.0°C lower than surrounding areas. (The surface temperature of the tree canopy is 7.8°C lower than concrete.) Air temperature of the park is 0.4~1.7°C lower than urban blocks.
[14]
Singapore
[15]
Singapore
Bishan & Serangoon CWP
M/S
Summer
[16]
Singapore
NUS
M/S
Summer
[17]
Israel
Tel-Aviv
M
Summer
[18]
Germany
Hanover
M
Autumn
[19]
USA
M
Summer
Air temperature (Surface temperature) (°C)
Air temperature of the park is 3~4°C lower than urban blocks. Maximum air temperature within the green area is 4°C lower than outside the green area. The air temperature of urban area is 0.5~2.9°C lower than the reference point. (The maximum surface temperature of the green area is 20°C lower than artificial ground.) When the green area is 65%, the air temperature of urban areas is approx. 2.0°C lower.
St. Louis S Summer Mexico Rainy / [20] Mexico M Air temperature of the park is 1.0~4.0°C lower than urban blocks. City Dry M/S: Measurement/Simulation, CWP: Central & western parts, NUS: National University of Singapore
INVESTIGATION OF THE TREE MODEL The tree has multiple influences on the thermal environment in summer. It effectively reduces solar and long-wave radiation, and generates evaporative latent heat, and thus it is expected to mitigate the thermal environment of urban areas. However, trees produce moisture and reduce wind velocity, and thus it is necessary to investigate these influences. For the purpose of widespread use of the tree model developed by the authors, it is built using the radiation analysis software “RADX” [21] and the CFD software “Star CD”. Heat transfer analysis of the tree (1) Radiation of the tree In modeling the radiation reduction effect of the tree crown, absorption of the radiation flux is used which considers the transmission distance and the leaf area density to radiation flux of the tree crown, and the effective transmittance was calculated from the absorption coefficient. (1) τ eff = exp ( −kax ) where, τ eff is the effective transmittance, k is the absorption coefficient [-], a is the leaf area density [1/m], and
x
is the transmission distance [m].
(2) Fluid resistance of the tree In order to consider the fluid dynamics resistance of the tree, Equation 2 is added to the equation of wind velocity motion. The turbulence energy and dissipation factor were also considered at the same time. 2 (2) −η Cd a ( x1 , x2 , x3 ) ui uj where
is the resistance coefficient of the tree crown (=0.2) [-], a ( x1 , x2 , x3 ) is the leaf area density of the tree crown [m2/m3], and η is the ratio of green coverage (the proportion of the horizontal area covered by the tree out of the horizontal area of mesh including tree) [-]. Cd
2
oC
Top Wind Wind
North
Tree 樹木 crown
East
West South Z
地表園 Ground Y
Bottom
X
Figure 1. The simulation model
Figure 2. The surface temperature of the tree crown
Out of consideration for the sensible heat and latent heat generated by the tree, the latent and sensible heat fluxes that flow from the heat balance equation of the tree crown as obtained by the radiation calculation were build into the generation term of the transport equation of the temperature and the absolute humidity. The details are as per Yoshida et al. [22]. Simple analysis example of the tree (1) The simulation model
N Tree crown C
o
Figure 3. Ground surface temperature
The simulation model is shown in Fig. 1. The computational domain is 10 m × 10 m × 10 m, and the tree size is 3.33 m × 3.33 m × 3.33 m. The inflow wind velocity is 1 m/s, and the inflow air temperature is 31.6°C. The evaluation time is 3 p.m. The tree was assumed to be a cube floating in the air, and the influence of the trunk tree was disregarded in this study. (2) Results The surface temperature distribution of the tree is shown in Fig. 2. Due to the effect of solar radiation at 15:00, the surface temperature at the west side and top of the tree are higher than the south side. The surface temperatures of the east side and north side of the tree are lowest. The surface temperature of the bottom of the tree is high due to the influence of reflection from the ground. The ground surface temperature is more than 10°C lower in the shade of the tree (Fig. 3). This result is similar to existing research [23]. The vertical temperature distribution is low in the tree (Fig. 4). Because of the southwest direction of the sun at 15:00, the surface temperature of the northeast ground is low. The tree becomes an obstacle to the wind, and thus the wind velocity decreased on the inner and leeward side of the tree (Fig. 5). The vertical absolute humidity distribution is high in the centre of the tree (Fig. 6).
oC
Tree crown
Figure 4. Vertical air temperature distribution
Tree crown
m/s
Figure 5. Vertical wind velocity distribution
Tree crown kg/kg’
INVESTIGATION OF LARGE-SCALE GREENERY Investigated area The investigated area is Minami Aoyama, Tokyo. The area is composed of detached houses, high-rise apartment buildings, office buildings and commercial buildings, etc. Trees are planted around the high-rise apartments, and create a comfortable open outdoor space. When we enter an alley 3
Figure 6. Vertical absolute humidity distribution
Table 2. Meteorological conditions Day and time Wind direction Wind velocity Air temperature Latitude Longitude
23 July (15:00) Southwest 3 m/s (74.6 m above ground level) 31.6°C 35.68°N 139.79°E
Table 3. Radiation conditions Building
from the main street, the residential area is extended. Most buildings are made of concrete, with only a few wooden houses. In a few office buildings, the air conditioning is set at ground level, and the heat is discharged in the pedestrian space. Parks are scattered in the targeted area, with the Aoyama graveyard on the west side, and Jingu Gaien and Akasaka Imperial Palace on the north side. In this research, the ‘Aoyama graveyard’ was the model for ‘large-scale greenery’.
Wall Solar reflectance Solar transmittance Thickness Conductance Indoor convection heat transfer coefficient Indoor air temperature Long wavelength emissivity Ground Solar reflectance Long wavelength emissivity
Concrete 0.1 0 0.22 m 1.64 W/(mK) 2
Ground
4.64 W/(m K) Calculation method 26°C In this research, coupled simulation of radiation 0.9 and convection was used [24]. The date and Asphalt 0.1 meteorological conditions are shown in Table 2. 0.9 rd The analysis date is 23 July and the investigated 26°C (0.5 m Underground temperature below ground time is 15:00; i.e. a fine summer’s day. The level) radiation calculation conditions are shown in Table 3. The radiation calculation was Table 4. CFD Conditions (Huang et al. 2005) conducted in an unsteady condition, Standard k-ε model (inclusion of buoyancy and the pre-calculation time is 15 hours. Turbulent model effect) Discretization Finite volume method The surface temperature, which was Algorithm SIMPLE method obtained from the radiation calculation, Differencing scheme First-order upwind scheme Side No slip is used as a boundary condition in the Sky Slip CFD simulation. Wall Generalized logarithmic law The CFD calculation conditions are U = U0(Z/Z0)0.25, Z0=18.0 m, U0=1.0 m/s Z 0: Representative height (met station) (m) shown in Table 4. The turbulent model Inlet (Power law) U0: Wind velocity at Z0 (m/s) is a standard k-ε model. The Z: Height from the ground level (m) U: Wind velocity at Z (m/s) dimensions of the computational k=1.5(IU)2, I=0.1 Turbulent kinetic energy domain are 1,591 m (x) × 1,668 m (y) × (k) I: Turbulent intensity 600 m (z) with a mesh size of 354,070 Turbulent kinetic energy ε=Cμk1.5/l, Cμ=0.09 dissipation rate (ε) for the large-scale greenery model. In l=4(Cμk)0.5Z00.25Z0.75/U0 Turbulence length scale (l) order to investigate the effect of such greenery, CFD simulations were performed with and without large-scale greenery. The area of large-scale greenery is assumed to be 195,255 m2. The height of the tree trunk is 4 m above ground and the height of the tree crown is 15 m.
Results and discussions Surface temperature distribution The distribution of the surface temperature, which was obtained from the radiation calculation at 15:00, is shown in Fig. 7. Due to the influence of solar radiation, the variation in surface temperatures of the buildings or ground is high in shaded and non-shaded areas. The sun is in a southwest direction at 15:00, and thus the surface temperature of the west-facing walls is higher than the south-facing walls. The surface temperature of the roof rises to 56°C due to the effect of direct solar radiation. The surface temperature is about 50°C on the west-facing walls, and about 38°C on the south-facing walls. The surface temperature of the ground is about 54°C in areas exposed to the sun, and about 42°C in shaded areas. The surface temperature of large-scale greenery is about 28°C due to the effect of evaporation. The result is similar to existing research that the surface of the green area is 15 to 20°C lower than the surrounding area (Table 1, Fig. 8). 4
Air temperature (1) Horizontal air temperature distribution The horizontal air temperature distribution at 1.5 m above ground level is shown in Fig. 9. The air temperature difference between with and without greenery is shown in Fig. 10. The mean and standard deviation are shown in Table 5. The air temperature is high around the area where the surface temperature is also high. The air temperature of the middle part of the greenery is 4.0°C lower than if without greenery. The air temperature of the leeward side of the urban area of large-scale greenery is about 1.5°C lower than without greenery, and thus large-scale greenery can be considered effective as urban heat island mitigation. These results are similar to existing research (Table 1). The cool air of the greenery area flows to the inner part of the residential area, and the Table 5. Mean and standard deviation (SD) temperature reduction is as much as near to Air temperature (°C) Height Mesh the greenery area (Fig. 10). The effect of (m) number Without greenery (T) With greenery (Tg) the air temperature reduction from largeMean SD Mean SD 1.5 8,801 32.5 0.4 31.2 1.2 scale greenery extends to about 500 m into 5 8,801 32.4 0.4 31.2 1.2 the urban blocks, which is in line with 10 8,801 32.3 0.4 31.3 1.0 existing research of 80 to 1,000 m (Table 6). 20 8,801 32.2 0.4 31.4 0.6 Due to poor ventilation, heat more Table 6. Range of air temperature reduction easily builds up in the densest spaces effect by large-scale greenery in summer (alleys, narrow spaces) and thus the air Place Method Range (m) Reference temperature is highest in such locales. Tokyo Existing research (M) 100 to 1000 [5] Conversely, the ventilation is good in Tokyo Existing research (M) 80 to 250 [6] Japanese city Existing research (S) 300 to 400 [25] open spaces (wide streets, open land) Singapore Existing research (S) Approx. 350 [15] and thus the air temperature is Israel Existing research (M) 100 [17] comparatively lower there. Tokyo Present (S) 500 Present study M: Measurement, S: Simulation 55
oC
Temperature (o C)
50 45 40 35
Urban area (Concrete surface temp.) Urban area (Air temp. above 1.5 m GL) Green area (Air temp. above 1.5 m GL) Green area (Ground surface temp.)
30 25 20 7:00 11:00 15:00 19:00 23:00 3:00 7:00 Time (-)
Figure 8. Measured temperatures from existing research [2]
(a) Without greenery (b) With greenery Figure 7. Surface temperature distribution
N
N
oC
oC
A
A’
A
A’
(a) Without greenery (b) With greenery Figure 9. Air temperature distribution (GL +1.5 m)
5
Figure 10. Air temperature (difference between with and without greenery)
(2) Vertical air temperature distribution The vertical temperature distribution decreases with increased height (Fig. 11). This result shows that an unstable layer is formed near ground level. In the absence of greenery, the air temperature is high across the entire calculated area. On the other hand, when an area of greenery exists, the air C temperature remains low up to building height. Green area The vertical temperature distribution of the entire calculated area is shown (a) Without greenery (b) With greenery in Fig. 12. With largeFigure 11. Vertical air temperature distribution (A-A’ section) scale greenery, the air 31.6 temperature is as low as 250 T = -0.233T + 38.8 Without green (n=21, R = 0.87, p
