URBAN HEAT ISLAND AND MITIGATION ...

In: Advances in Environmental Research. Vol. 54 ISBN: 978-1-53610-667-1 Editor: Justin A. Daniels © 2017 Nova Science Publishers, Inc.

Chapter 6

URBAN HEAT ISLAND AND MITIGATION SOLUTIONS EVALUATION IN COLD CLIMATES: A CASE OF MONTREAL Yupeng Wang1,2, and Hashem Akbari2 1

School of Human Settlement and Civil Engineering, Xi’an Jiaotong University, Xi’an, China 2 Department of Building, Civil and Environmental Engineering, Concordia University, Montreal, Canada

ABSTRACT The increasing awareness of the urban heat island (UHI) effect has raised the attention for monitoring and evaluating the outdoor thermal comfort in cities worldwide. However, most of the UHI studies are carried out in hot climates. Urban climate change in cold-climate cities of Canada is also becoming an important consideration for global climate moderation, energy consumption, citizen safety and wellbeing. Research carried out in eight Canadian cities compared the annual hot days (with a temperature high of 30°C) between 1961 and 1990 to the forecasted average after 2020 will be increased from 10 to 22 days. Projected temperature increases in Canada are even more dramatic than in the southern latitudes. Currently, in Canada, some municipalities have discussed UHI mitigation at strategy level, carried out literature review, revealed current status of UHI impacts in Canadian cities on energy

Email: [email protected].

Complimentary Contributor Copy

144

Yupeng Wang and Hashem Akbari consumption, citizen well-being and health. However, there is little discussion about detailed technical guidelines for UHI mitigation application. They need policy-relevant data and analysis in these efforts. The current data mostly focuses on the consideration of greenhouse gas emissions levels rather than providing specific adaptation strategies on the urban scale. In this chapter, the current policies for UHI mitigation and adaptation in Canada are discussed. In particular, the effect of street vegetation planting, albedo and urban canopy characteristics in Montreal is discussed and demonstrated by using numerical simulation. ENVI-met (a three-dimensional computer model which analyzes micro-scale thermal interactions within urban environments) is used to evaluate the urban typology in typical micro-scaled urban areas, in order to indicate the detailed types of urban form. Simulated environmental conditions including air temperature (Ta), human weighted mean radiant temperature (MRTh-w), wind speed, and physiologically equivalent temperature (PET) at the community scale is carried out for demonstrating the effects from the current urban development on urban thermal environment. These simulation comparisons demonstrate the effects on each environmental factor for a typical summer day and provide hints for mitigation of the urban heat island (UHI) and design of new urban development. The effectiveness of each UHI mitigation strategy is evaluated for providing guidelines for policy development.

1. INTRODUCTION The urban heat island (UHI) effect is a phenomenon whereby a metropolis is usually significantly warmer than its rural surroundings. This occurs because 1) urban surfaces are typically darker than those of surrounding areas, 2) there is less vegetation in urban areas, and 3) buildings and street surface materials, which have high heat capacities, store heat during the day and release heat slowly at night (Rao, 2012; Oke, 1988). The adverse energy and environmental effects of UHIs, and methods to alleviate them, has become a major research topic in sustainability programs. Decreasing the energy consumption of buildings is also an important topic in environmental engineering. Most of the UHI studies are carried out in hot climates. Urban climate change in cold-climate cities of Canada is also becoming an important consideration for global climate moderation, energy consumption, citizen safety and wellbeing. Research carried out in eight Canadian cities compared the number of annual hot days (with a temperature high of 30°C) between


Urban Heat Island and Mitigation Solutions Evaluation …

145

1961 and 1990 to the forecasted average after 2020 will be increased from 10 to 22 days. Projected temperature increases in Canada are even more dramatic than in the southern latitudes. Currently, in Canada, some municipalities have discussed UHI mitigation at strategy level, carried out literature review, revealed current status of UHI impacts in Canadian cities on energy consumption, citizen well-being and health. However, there is little discussion about detailed technical guidelines for UHI mitigation application. They need policy-relevant data and analysis in these efforts. However, the current data mostly focuses on the consideration of greenhouse gas emissions levels rather than providing specific adaptation strategies on the urban scale. Studies of air-conditioning energy use in hot climates have shown that the air-conditioner load in residential buildings can be particularly high in the nighttime. As an example, in a study of residential air-conditioner operation on a summer day in nine Chinese cities, the peak time was observed between 6 pm and 11 pm (Yoshino et al., 2006). In a study of two Chinese office buildings, the highest energy consumption was observed in the daytime, as the operating schedule of offices is normally between 7 am to 6 pm (Pan et al., 2008). A similar result was also observed in an American campus, which showed the highest energy demand between to 6 am to 5 pm, and the peak hours were 3 pm and 4 pm (Agarwal et al., 2009). In cold climates, airconditioning energy use in residential buildings may not be very significant for most of the summer. However, the few days in the summer that require residential AC use at night can significantly impact utility peak demand. A report from Hydro Quebec states that energy demand is likely to peak at two times of the day: 7-9 am and 4-8 pm (Hydro Quebec, 2014). The first peak period roughly aligns with the beginning of a typical work day and the second peak period may be when people return home after work. Therefore, heating or cooling energy consumption peak reduction should be carried out by focusing on the evening peaks. Meanwhile, UHI mitigation techniques provide different contributions in daytime and nighttime. Urban open space obtains more solar radiation and leads to higher air temperature in the daytime, it releases more heat to the sky and leads to lower air temperature in the nighttime (Wang et al. 2014). Environmental policy implications should consider urban energy distribution and the features of mitigation techniques, and create the optimized guidelines. These policies could address urban development density, which is influenced by urban roughness, and potentially affect the urban vegetation styles. These three dimensions of urban fabric are discussed in this chapter. In addition, this chapter takes into account the effect of UHI on energy use, urban air quality,


146

Yupeng Wang and Hashem Akbari

and urban thermal comfort. The evaluation approach in urban thermal comfort includes wind speed, Ta, MRTh-w, and PET. This chapter focuses on summer UHI mitigation options. Urban vegetation planning, urban albedo, urban building density and urban canopy control are discussed, based on simulation comparisons in the city of Montreal. The ultimate goal is to provide a framework for developing effective UHI mitigation policies.

2. OVERVIEW OF MODELS AND TOOLS 2.1. Urban Environment Simulation We used ENVI-met simulation model (a three-dimensional computer model that analyzes micro-scale thermal interactions within urban environments) (Bruse, 1999, 2015) to simulate the environmental conditions in the 4 districts in Montreal, Canada. ENVI-met is designed to simulate the surface-plant-air interactions in urban environments. It has a typical spatial resolution of 0.5m to 10m, and a temporal resolution of 10 seconds. A simulation is typically carried out for at least 6 hours (usually for 24-48 hours). The optimal time to start a simulation is at night or sunrise, so that the simulation can follow the atmospheric processes. ENVI-met requires an area input file which defines the 3-dimensional geometry of the target area. This includes buildings, vegetation, soils and receptors. A configuration file, which defines the initialization input, is also required. The data flow is shown in Figure 3 summaries the general interaction between input data files and output data (Chow and Brazel, 2012). The MRT calculation in ENVI-met is taking into account the radiation that absorbed by a standing human body using projection factor (fp). This presents the radiation absorbed by human skin that affected by the height of sun. Therefore, at the mid-noon, when the sun is high above the head, the projection factor becomes to be the minimum. The value of MRT (human weighted) shows the sum of all short-wave and long-wave radiation fluxes absorbed by the human body that affects his energy balance. Indicator of MRTh-w is more accurate than using to evaluate the thermal comfort (Peng et al., 2011). ENVI-met calculation of MRTh-w has been validated in previous studies (Ali-Toudert and Mayer, 2006; Bruse, 1999). The surrounding environment affects MRTh-w: building surfaces, ground surface and the sky.



147

For the MRTh-w calculations at street level, half of the long-wave radiation fluxes are assumed to originate from the upper hemisphere, and the other half from the ground and building surfaces. For these simulations, the geometry of urban street canyons in four areas was identified using satellite pictures and Google StreetView image from Google Map, the information include building layout, building height, vegetation placement and vegetation height. Area input files were built by ENVI-met; we input satellite pictures into the editing files, and defined the ground, vegetation, building facade and building layout by cubic grids. It is common to use geometric resolution of the order of 3 meters in urban studies (Steinicke et al., 2009), and the pixel of several digital data base of earth’s surface (Taubenböck et al., 2009; Herold et al., 2001). Meanwhile, average floor height of urban building is common to be 3 meters, while doing urban simulation (Chandurkar and Pajgade, 2013; Huang et al., 2012). In this analysis, the cubic grids with the size of 27m3 (3×3×3m) is used to present the standard unit for urban environmental evaluation. Soil characters are indicated by asphalt road, pavement, and loamy soil, observed via Google Maps in these four study areas, and modeled by the setting of natural soils and seal materials that stored in ENVI-met database. In ENVI-met, the basic structure of the receptors files is the same for the atmosphere, the soil and surface, but showing the status at selected points inside the model. Receptors are installed at the center of each road segment and the intersections at the center area of each model, and used the average amount of all the receptors in this areas. Nine receptors for each district were defined for SVF calculation (Figure 2.1). All the receptors were placed in the middle of streets. However, with the spatial limitation of the simulation model, a deviation will emerge around the edge of simulation model. This deviation will also affect the UHI simulation results with ENVI-met. Therefore, we made efforts to set all the receptors close to the center of each area in order to reduce the grid edge effect.

2.2. Simulation Validation ENVI-met simulations are affected by the boundaries of the studied areas. In our summary data analysis, we have mostly taken data from the interior part of the study domain, minimizing the boundary effects. Figure 2.2 shows the correlativity between simulation results at YUL airport area in Montreal, and


Sources Database (SOURCES.DAT)

Soils Database (SOILS.DAT)

Soil Profiles (PROFILS.DAT)

Plant Database (PLANTS.DAT)

Global Databases and Settings

ENVI-met Basedata

ENVI-met Model V3.1

Simulation

Output Data

Data link to BOTworld

Receptor 1D- Output 䞉Time Series Files 䞉Snapshot Files

1D-Inflow Profile

Main 3D Output Files: ..AT...䠖Atmosphere ..FX...䠖Surface & Fluxes ..SO...䠖Soil


Figure 2.1. Data flow in ENVI-met V3.1 (http://www.envi-met.com/).

Additional databases

Simulatiion specific databases 䞉Plants 䞉Sources adding to global databases

Simulation Files

definees 䞉Buildings 䞉Plants 䞉Soils 䞉Sources 䞉Receptors

Area Input File (.IN)

Main Configuration File (.CF)

Input data


149

Figure 2.2. The correlativity between simulation results and measured results. (Location information: 45°28'00.000" N, 73°45'00.000" W; Montreal city).

measured data obtained from the government of Canada website, located at about 5-10km from these four study areas. In the comparison, the data from 5 am on 22nd July to 5 am on 23rd July of 2012 is used. The elevation height of the measurement was 36m. ENVI-met is found to represent well the trends of Ta and relative humidity, which shows the high R2 of 0.73 (Ta), 0.63 (relative humidity). Numerically, simulated values of Ta is about 2 to 3°C underestimated in the daytime (10 am to 3 pm), and about 4°C underestimated at night. Meanwhile, the relative humidity is overestimated for most of the simulation period. This discrepancy could be explained by inaccuracies in simulations input for surface materials, soil, and vegetation conditions. ENVImet simulation was set for cloud-free sky conditions. Without regards to actual cloud cover in the model can lead to lower temperature compared to clouded sky conditions, because the energy distribution is shifted towards the diffuse component (Watson and Johnson, 1987). Meanwhile, the formulation for longwave flux divergence does not take into account the effect of horizontal longwave fluxes, or the effects of vegetation, only indirectly through air and surface temperatures. The maximal Ta underestimation is 4°C, observed at the evening (6pm to 8pm), and the largest overestimation of relative humidity is


150


about 20%, occurred at the evening (6pm). The large differences could be explained by the large time steps used for updating the sun position and the radiative fluxes from the sky (set at 10 minutes). Without regarding the results at the evening (6pm to 8pm); R2 of simulation and measurement data increases to 0.78 (Ta), 0.70 (relative humidity).

3. CASE STUDIES Urban planning is covering many factors such as urban vegetation, urban materials, urban typology that are affect to the urban environment. And also, these factors affect each other at the same time. For evaluating the effects from each urban factors, 3 case studies are carried out to compare and discuss.

3.1. Urban Typology 3D geometrical configuration plays an important role in managing longwave radiation heat loss. Outgoing long-wave radiation loss depends on the urban design, because only a small part of the sky is seen from the ground surface caused by narrow streets and tall buildings creating deep canyons. The sky view factor (SVF) can be a representative indicator for urban building density and layout. SVF is the ratio of the radiation received (or emitted) by a planar surface to the radiation emitted (or received) by the entire hemispheric environment (Watson and Johnson, 1987). Videlicet, SVF affects urban radiation exchange and urban microclimate. Numerous studies have studied the effect of SVF on UHI (Gulvas et al., 2006; Thorsson et al., 2007; Gal et al., 2009; Oke, 1988; Unger, 2004, 2009; Bottyan and Unger, 2003). Many studies are related to thermal comfort in urban environment (Brode, 2012; Mayer and Hoppe, 1987; Lin and Matzarakis, 2008; Matzarakis, 2001; Matzarakis and Nastos, 2011). The importance of research field of urban thermal comfort with using simulation was indicated in many studies (Bruse, 2007, 2009; Huttner et al., 2009). Most studies have focused on UHI intensity (the comparison between urban and rural areas) but not radiation transfer between various urban surface geometries inside urban areas. Here, an urban development policy is provided for SVF control, and the effects on urban geometry change and urban thermal environment is evaluated. The micro-scale urban environment simulation system ENVI-met is used to



151

compare the results under various policies of controlling floor area ratio, building set back, and land coverage ratio, between specified urban geometry changes. The simulations were carried out in the central area (300m by 300m in size) at Montreal. The Ta and MRTh-w at 1.8m height above the ground are compared and evaluated. These technical analyses contribute to environmental policy development for UHI mitigation.

3.1.1. Methods For the discussion, 6 micro urban canopy models are built and simulated for evaluating the environmental effects (air temperature, and mean radiant temperature) in varied building set back. The ENVI-met inputs of the models are presented in Table 3.1.1, and the properties are presented in Table 3.1.2. Table 3.1.1. Area input file to the ENVI-met for comparison of different set back distances

For these simulations, the geometry of an urban street canyon in Montreal city was identified using satellite images and street maps from Google Map. Area input files were built by ENVI-met; we input satellite images into the editing files, and defined the ground, vegetation, building facade and building layout by cubic grids of 27 m3 (3 m × 3 m × 3 m). However, the spatial limitation of the simulation model caused a deviation that emerged around the


152


edge of the model, which affected the UHI simulation results with ENVI-met. This leads to an omission of the effects from surrounding buildings outside of the selected area, and an overestimation of the effects from the environment (air temperature, wind speed, humidity) outside of the city. Table 3.1.2. Land use and building height properties in each model area Base Land Coverage (%) Green Coverage (%) Pavement Coverage (%) Building height (m) Side walk width (m)

39.4

Set back 1 34.9

Set back 2 28.8

Base + Green 39.4

Set back 1 + Green 34.9

Set Back 2 + Green 28.8

0

0

0

6.1

6.4

14.1

60.6

65.1

71.2

54.5

58.7

57.1

15

18

21

15

18

21

3

6

9

3

6

9

The urban environmental simulation model “RayMan” provides the assessment of MRT (Matzarakis and Rutz, 2005). We used Canadian historical weather data in simulation at ENVI-met and RayMan (Environment Canada website, 2015) to evaluate the thermal comfort for human. The simulation results of MRT from ENVI-met and RayMan are compared and discussed. For the discussion, 6 micro urban canopy models are built and simulated for the discussion about the environmental effects (air temperature, and mean radiant temperature) with varied building set back. The simulations ran for 30 hours for a typical summer day, starting at 10 pm (about six hours before sunrise). The details of the initialization input data for simulation is shown in Table 5. The selected area in Montreal was a 300 m × 300 m, high-density residential district next to the city’s main commercial core and a university. There are many high-rise residential buildings of more than 15 floors. Most of the ground surfaces are asphalt road. The image of the selected area is shown in Figure 3.1.1. To compare the environmental difference between varied building types, we constructed two models of the selected area with different building height and set back, as shown in Table 3.1.3. The “Concordia” model



153

shows the original conditions of the urban setting. The “Height Control” model shows the building style modification to unificate the building height of all of the buildings in this selected area with remain the same foot print and the same global floor area (building volume) of this whole area. The “Set Back” model show is enlarged the width of the sidewalk for 3m. The details of input data is shown in Table 3.1.3. The land use and building height properties in each area are measured from the area input file that showing in the Table 3.1.4.

Figure 3.1.1. Selected area for simulation and comparison (Concordia area in Montreal, Canada).

Table 3.1.3. Area input file to the ENVI-met for compare the effects from “height control” and “set back”


154

Yupeng Wang and Hashem Akbari Table 3.1.4. Land use and building height properties in Concordia model areas

Land Coverage (%) Green Coverage (%) Asphalt Coverage (%) Average Building height (m) Median Building height (m)

Concordia 46.3 3.4 50.4 30 12

Height Control 46.3 3.4 50.4 30 30

Set Back 40.2 3.4 50.4 30 16

Table 3.1.5. Details of the initialization input parameters for the simulation Category Starting Time Wind Speed Wind Direction Air Temperature Specific Humidity in 2500m Relative Humidity in 2m Building Interior Temperature Mean Heat Transmission of Walls Mean Heat Transmission of Roofs Mean Walls Albedo Mean Roofs Albedo

21st, July, 2012 22:00 3.6 m/s South West 298.15K (25Ԩ) 7 g/kg 51% 299.15 K (26Ԩ) 1.94 W/m2K 6 W/m2K 20% 30%

3.1.2. Results Comparing the “Set Back 1” and “Set Back 2” to the “Base” model in Figure 3.1.2, increasing the building set back could provide wider walking space, increase the urban openness (without vegetation), as well as provide more opportunity to plant vegetation in the urban spaces (refer to Table 3.1.2), without affecting the global floor area (building volume). In the models with vegetation, more vegetation are placed in the open spaces, and bigger set back leaded to a lower sky view factor, because of the bigger amount of trees.



Figure 3.1.2. Sky view factor in Discussion Models.

155

Figure 3.1.3. Sky view factor in Concordia Models.

In Figure 3.1.3, after height control, all of the buildings are changed to the height of averaged building (30m) of this whole area. Both high-rise buildings and low-rise buildings are modified to the same height. This is an extreme example of building height control, the median SVF in the Height Control model is about 5% lower than that in the Concordia model. This demonstrated that with the same building volume, higher building could provide more opened urban canopy than that of lower buildings. With increasing the building setback for 3m, the median SVF of the whole area is increased about 2% than that in the original condition. Figure 3.1.4 and 3.1.5 show the simulation results from ENVI-met. Compared to the base model, urban air temperature (Ta) in the models with more building set back is obviously lower in the day time, about 0.3°C Ta reduction in the “Set Back 2” models. Comparing the Ta in “Set Back 1” model and “Set Back 2” model, with 3m more building set back from the road, Ta reduction could be increase more than 0.1°C.


156


Figure 3.1.4. Comparison of average air temperature at human height (1.8 m above ground) in the 6 models.

Figure 3.1.5. Comparison of reduced air temperature compare to the base model with building set back.

Figure 3.1.6 shows the simulation results of MRTh-w with ENVI-met and RayMan. The curves during the day-time are showing that the MRTh-w with effects from surrounding buildings, because the surrounding buildings shadowed down the surfaces from solar radiation (Chen et al., 2014). The affection from surrounding buildings on MRTh-w calculation in RayMan is higher than that calculated in ENVI-met. During the day (before 7pm), the MRTh-w in “Back 1” and “Back 2” models are higher than that in the base



157

model. This is because the higher SVF with larger building set back could receive more solar radiation at the human height level. However, during the night after 7pm, MRTh-w in the base model is slightly higher than that in “Back 1” and “Back 2” models. This is because the stocked heat from the solar during the day in the deeper urban canopy is more difficult to be released than that in an opened urban space during the night.

Figure 3.1.6. Comparison of average mean radiant temperature at human height (1.8m above ground).

Figure 3.1.7. Comparison of average Ta at human height (1.8m above ground) in Concordia area.


158


Figure 3.1.8. Comparison of average MRTh-w at human height (1.8m above ground) in Concordia area.

In the Concordia area, Ta in the daytime is reduced by nearly 2°C with 3m building set back in the whole area, and the contribution is continued to the night. However, in the “height control” model, Ta is up to 0.8°C higher than that in the original Concordia model. Therefore, with the same building volume (global floor area) slime and high building (“set back” model) could reduce urban heat island in the summer, compare to low-rise buildings. Looking at the MRTh-w in Figure 3.1.8, height control could reduce the solar energy storage at human height level, because it provides more shadow area in the day. The MRTh-w in “set back” model is higher than that in the “height control” model in the morning, but lower than the other two models after 12 am. The opened urban canopy with building set back could also help to release the heat during night. The MRTh-w after sunset in “set back” model is 0.3 to 1°C lower than that in the other two models.

3.1.3. Discussion SVF control could be achieved by several building development indexes such as floor area ratio (the proportion of total building floor area and the site area, which is related to building volume and building facade design); building set back (the distance between the buildings and the roadside, which is related to the minimum openness among buildings); and building coverage ratio (the proportion of built area and the site area, which is related to the amount of



159

urban open space and the potential for urban vegetation). The combinations of these building development indexes could create a variation of urban canopy, and the details are presented in Figure 3.1.9. A study on urban density control in residential areas was carried out and discussed widely for reasonable urban functional planning and economic development, and the transportation cost (Mills, 2005). UHI mitigation-related environmental issues should also be considered during the urban development process. Urban development is a constant and long-term issue that goes along with social development and economic growth. The oversight of this long-term consideration should be avoided by providing long-term urban development policies. Currently, urban planning control is implemented in most countries in the world. The SVF control discussed in this section is an important factor is urban development plans. Here, we demonstrated the importance of SVF control policy in residential areas. This is an additional consideration for sustainable urban development in the long term. Further analysis on the relationship between urban SVF control and urban density control should be carried out in future research.

Figure 3.1.9. The details of building development indexes combinations and the effects.

3.1.5. Summery This analysis is based on the simulation. Simulation program ENVI-met and RayMan are used to evaluate the urban environment and building design in simplified models and the central area of Montreal city. It is demonstrated


160


that with the same building volume (global floor area), the slim and tall buildings could contribute to UHI mitigation. This urban development control could be achieved by establishing related policies or guidelines to formulate the building set back in the planning stage. The urban environmental optimization by regulating building set back could be varied in different locations of the city, because of the building density and the global floor area amount. Therefore, it’s demonstrated that the UHI mitigation and environmental optimization by regulating the building set back creates better thermal environment in the city center, but it still needs further investigation in the other locations of the city. Overall, the technical analyses presented here provide hints for the policy makers to develop environmental policies for UHI mitigation in the city center of Montreal.

3.2. Urban Albedo Asphalt and concrete constitute as much as 40% of Canadian urban surface area (Natural Resources Canada, 2009; Gui et al., 2007). On average, in Toronto, asphalt area is 16% and concrete area is almost 14%, to which should be added the roof top areas (Krayenhoff et al., 2003); obviously these average values are much higher in downtown. Asphalt and concrete have low albedo, with values as low as 0.1 on average for asphalt and 0.3 or 0.4 for concrete (Asaedaa and Ca, 2000). As a strategy to mitigate UHI, surface materials with high albedo and emissivity have been proposed worldwide since they remain cooler when exposed to solar energy (Akbari et al., 2001; Akbari and Konopacki, 2004; Synnefa et al., 2007; Pisello and Cotana, 2014).

3.2.1. Methods The way of building simulation model and the resolution of the model is the same as that showed in section 3.1.1. The simulations ran for 30 hours for a typical summer day, starting at 10 pm (about six hours before sunrise). The weather data was obtained from the “Weather Spark” database (Vector Magic, Inc., 2014). However, the spatial limitation of the simulation model caused a deviation that emerged around the edge of the model, which affected the UHI simulation results with ENVI-met.



161

Figure 3.2.1. Selected area for simulation and comparison (Concordia area in Montreal, Canada).

Table 3.2.1. Details of ground surface materials

Asphalt road Granite pavement

-3 -1 6 -1 -1 Albedo Heat Capacity [Jm K ]*10 Heat Conductivity [Wm K ] 0.2 2.251 0.90 0.4 2.345 4.61

Table 3.2.2. Details of the initialization input parameters for the simulation Category Starting Time Wind Speed Wind Direction Temperature Specific Humidity in 2500m Relative Humidity in 2m Building Interior Temperature

21st, July, 2012 22:00 pm 3.6 m/s South West 298.15 K (25 Ԩ) 7 g/kg 51% 299.15 K (26Ԩ)

Mean Heat Transmission of Walls

1.94 W/m2K

Mean Heat Transmission of Roofs Mean Walls Albedo Mean Roofs Albedo

6 W/m2K 20% 30%

The image of the selected area is shown in Figure 3.2.1, and the details of land use and input data are shown in Tables 3.2.1, and 3.2.2. To compare the environmental difference between varied vegetation types, we constructed two


162


models of the selected area with different vegetation types, as shown in Figure 2.

3.2.2. Results Ground surface materials properties affected community environment and the details are shown in Figure 3.2.2. During the day-time, with a higher albedo, granite pavement reflected more solar irradiation than asphalt. Therefore, the ground surface temperature and Ta of granite was lower than for asphalt. The surface temperature difference between using granite and asphalt was 6.3°C at 11 am. The surface temperature of granite was about 6.5°C higher than the Ta, and the surface temperature of asphalt was about 10.7°C higher than the Ta. Between 8 am and 5 pm, compared to using asphalt, the MRTh-w above granite pavement was about 3 to 3.3°C and Physiological Equivalent Temperature (PET) was about 3 to 4°C lower. While granite surfaces have higher albedo and could thus contribute to urban heat mitigation during the day, it has a higher heat capacity than that of asphalt, and it has greater heat absorption. At night after 7 pm, the surface temperature of granite becomes higher than that of asphalt. This is because the granite pavement starts to release heat after sunset. At night, the Ta associated with granite pavement was about 0.4-0.7°C, MRTh-w was about 0.1-0.2°C, and PET was about 0.1-0.5°C higher than asphalt. These findings show that ground surface materials with high heat capacity can intensify the UHI phenomenon at night. 3.2.3. Discussion High-albedo pavement can reduce the UHI effect during the day. It has long been known that the temperature of a pavement affects its performance (Bahadur, 2009). Reflectivity of pavements is also a safety factor in visibility at night and in wet weather, affecting the demand for electric street lighting. Therefore, high albedo could benefit the local environment at daytime and nighttime. However, we also show that the high heat capacity of material increases the Ta at night. It is therefore ideal to increase urban albedo while reducing the heat capacity of urban surface materials. The initial cost of more reflective pavements may be higher than the other pavement types, but it may be offset by lifetime savings related to energy and smog and by a substantially longer lifetime.



163

Figure 3.2.2. Comparison of environmental conditions at 1.5 m above the ground on a typical summer day (from 5 am, 22 July, to 2 am the next day) with two ground surface materials.

3.2.4. Summery Here, we focused on the dynamic change of energy demand and various urban function characteristics of urban albedo in a summer day in Montreal. We showed that urban albedo can reduce the UHI effect in daytime. Investigating the effects of various pavement properties can assist officials to design effective policies that take adaption of urban land use into consideration.

3.3. Street Tree Planting Another important aspect of urban areas is that the fraction of the ground covered by trees and other vegetation is smaller and contains less biomass than in nonurban areas. The absence of vegetation impacts the UHI in several ways. Vegetation, and in particular trees, intercept solar energy, and their shade reduces the temperature of surfaces below while increasing the latent heat


164


exchange for the evapotranspiration process (Oke, 1988; Bonan, 2000; Bass et al., 2003; Shashua et al., 2009). Vegetation have shown to cool the surrounding environment by reflecting more solar radiation (higher albedo compared to common pavements), by absorbing and accumulating less heat, while the evapotranspiration contributes to cool the environment (Sailor, 2011; Millward and Sabir, 2011; Ng et al., 2012). Air temperature reductions due to vegetation of 2˚C in Tokyo, 5˚C in Singapore and 8˚C in Athens were reported (Papadopoulos, 2001).

3.3.1. Methods As shown in Figure 3.3.1, the selected area in Montreal is a high-density residential area, next to the city’s main commercial area and a university. The 300×300m2 area models is built for microclimate in simulation, and the simulation domains are defined by 3 × 3 × 3m3 grids. The detailed land use and building height in this area is shown in Table 3.3.1.

Figure 3.3.1. Images of the selected area: 1) aerial satellite view pictures, 2) CAD map include floor numbers for simulation inputting, and 3) input file images for ENVI-met simulations, based on the aerial satellite view pictures.

In the Montreal city biodiversity report in 2013, the most common trees in Montreal are indicated (Ville de Montreal, 2013). For understanding the characteristics of each tree type, in order to analyse the effects in the next step, a research of the tree size and lifetime is carried out. The results are shown in Table 3.3.3 (Arbor Day Foundation, 2014; North Dakota Tree Information Center, 2014; Kentucky’s Champion Trees, 2014). Using the average tree height and crown diameter, all common trees are classified to tree types (10 m height, 9m crown diameter; 15m height, 12m crown diameter; 20m height, 12 m crown diameter). Two tree planting patterns which are with space (A1; B1;



165

C1) and without space (A2; B2; C2) are indicated. In order to compare the variation of tree types and tree planting spaces, six proposed models for each road side tree planting patterns are simulated (Figure 2). The parameters for trees simulation are explained in Table 3.3.4, and the tree size and planting space between trees are explained in Table 3.3.5. Table 3.3.1. Land use and building height

Category Land cover (%) Grass (%) Tree (%) Vegetation cover (%) Asphalt (%) Average building height (m) Median building height (m)

Value 46.3 1.5 1.9 3.4 50.4 28 12

Table 3.3.2. Initialization input properties parameters for simulation

Starting time Total simulation time Wind speed Wind direction Temperature Relative humidity in 2m Building interior temperature

21th July, 2013 21:00 pm 30 hours 3.6 m/s Southwest 298.15K (25Ԩ) 51% 299.15K (26Ԩ)

Table 3.3.3. Detailed information of trees that most common in Montreal

Elm Ash

Siberian elm Northern red ash Basswood linden Linden Little leaf linden Silver maple Maple Norway maple Red maple Honey-locust Hackberry

Height (m) Crown (m) Average Lifetime (Year) Range Average Range Average Height&Crown (m) 8㹼15 11.5 6㹼12 9.0 60㹼150 10㹼9 11㹼20 15.5 9㹼12 11.5 30㹼50 15㹼12 15㹼21 18.0 9㹼15 12.0 1000 20㹼12 18㹼21 19.5 9㹼15 12.0 100㹼 20㹼12 12㹼20 16.0 9㹼15 12.0 80㹼130 15㹼12 15㹼23 19.0 10㹼15 12.5 60㹼250 20㹼12 12㹼18 15.0 9㹼12 10.5 80㹼150 15㹼12 9㹼15 12.0 9㹼12 10.5 120㹼150 10㹼9 12㹼18 15.0 7㹼14 10.5 300㹼400 15㹼12


Type A1; A2 B1; B2 C1; C2 C1; C2 B1; B2 C1; C2 B1; B2 A1; A2 B1; B2

166

Yupeng Wang and Hashem Akbari Table 3.3.4. Detailed parameters for trees in different patterns

Type A1; A2 B1; B2 C1; C2

Stomata Resistance Height Root Depth Leaf Area Density Root Area Density (sec/cm) (m) (m) (m²/m³) (m²/m³) 400 10 2 0.473 0.1 400 15 2 0.935 0.1 400 20 2 0.935 0.1

Table 3.3.5. Tree size and planting space in six simulation patterns Type Height (m) Crown (m) A1 10 9 B1 15 12 With space C1 20 12 A2 10 9 B2 15 12 No space C2 20 12

Figure 3.3.2. Images of proposed six tree planting patterns in Concordia area.



167

The averaged SVF of each simulation patterns at human height level (1.8 m) are output from ENVI-met, and compared in Figure 3.3.3. The SVF in base pattern is the highest in seven patterns. This is because the less tree cover leads to the urban openness. The averaged SVF in C2 is the lowest, because planting high trees without space between each other creates a deeper urban canopy at the human height level.

Figure 3.3.3. Average sky view factor in the proposed six tree planting patterns and in the base pattern.

3.3.2. Results Looking at the one day air temperature (Ta) change in Figure 3.3.4, the Ta in the base pattern is higher than the other patterns. With adding road side trees in the six proposed patterns, the average Ta difference after sun rise is reduced from 1.2°C (in pattern A1) to 3.3°C (in pattern C2). The Ta difference by planting trees could be observed remarkably between 11 am to 16 pm. The Ta different between A1 and B1 is almost 2.0°C in the mid-day, and the Ta different between B1 and C1 is under 0.3°C. After 0:00 am, the Ta different between A1 and B1 is almost 0.8°C, and that between B1 and C1 is under 0.1°C. This is to say, trees with larger crown diameter could provide higher contribution on reducing UHI in day-time and in night-time. The average Ta in A2 is 0.6°C lower than that in A1, average Ta in B2 is 0.7°C lower than that in B1, and the average Ta in C2 is 0.7°C lower than C1. Planting trees without space between the tree crowns could reduce Ta; and the maximum Ta


168


difference could be observed in the mid-day, reach up to 1.3°C between B1 and B2. The average daily MRTh-w in the base pattern is 2.9°C (A1) to 10.2°C (C2) higher than the proposed patterns. The largest differences could be observed between 9:00 am to 15:00 pm. Planting big trees without space between the tree crowns in pattern C2, street environment would mostly be shaded by trees, and the solar radiation could be all blocked from reaching the ground. The highest MRTh-w reduction between C2 and the base pattern could reach up to 40°C at 10 am.

Figure 3.3.4. Diurnal air temperature and MRTh-w change at human height level (1.8 m) on a typical summer day, from 3 am 22 nd July to 23rd the next day.



169

The different schemes of tree cover lead to the variation of SVF. The effect of tree cover on the outdoor thermal environment is shown in Figure 3.3.5. In mid-day, the R2 between SVF and Ta is 0.45 (ᬅ). This indicated that more tree cover could help to reduce the outdoor air temperature. The MRTh-w distribution is separated by the shaded areas and unshaded areas (ᬆ). In the shaded areas, the R2 between SVF and MRTh-w is 0.44; and in the unshaded areas it is not showing a high relation between tree cover and MRTh-w, the R2 is only 0.08. In mid-night, the R2 between SVF and Ta is 0.64 (ᬇ), the Ta reduction is about 3°C and that between SVF and MRTh-w is 0.69 (ᬈ). The Ta distribution range (the different between the highest spot and the lowest spot) in mid-day is about 7°C, and the range in mid-night is about 3°C. Tree shading could reduce the heat storage during the day-time, and this reduction leads to the lower Ta in night-time.

(1)

(2) Figure 3.3.5. (Continued).


170


(3)

(4) Figure 3.3.5. The effect of vegetation SVF on air temperature and MRTh-w at human height level (1.8 m) in mid-day (12:00 pm, 22nd July) and in mid-night (1:00 am, 23rd July). R2 = coefficient of determination.

Outdoor thermal environment affects pedestrian thermal comfort, as well as building energy consumption. In a city centre like Concordia area which is developed by high-rise buildings, the vertical outdoor thermal environment should be considered, in order to reduce the indoor energy consumption at higher floors. Figure 3.3.6 shows the Ta distribution at various heights. In the six proposed vegetation patterns, with increasing tree size and the amount of tree planting, at the tree height level (20m above ground), the Ta is reduced up to 4°C. At higher heights over 60m above ground (approximate 20 floors), the observed Ta in pattern C2 is about 2°C lower compare to that in the base pattern. This is to say, road side tree planting could provide UHI mitigation to



171

create better pedestrian thermal environment, and also reduce temperature at level which could relate to indoor energy reduction in the city centre areas.

Figure 3.3.6. The image of air temperature distribution in the cross section of simulation areas in the mid-day (12:00 pm, 22nd July).


172


3.3.3. Discussion It is widely reported that roadside tree planting could efficiently cool communities and reduce energy use in summer (Akbari, 2001). It is necessary to encourage vegetation planting to the appropriate areas through relevant policy. Planting trees reduces the UHI effect during the day, because the shading from trees reduces solar irradiance. Therefore, planting trees in areas that are mostly used in the daytime, such as business and factory districts, could provide a more significant UHI mitigation effect. Urban development policy for these areas should include guidelines about roadside tree planting and vegetation placement. Further analysis on implementation scenarios and financial considerations of planting and maintenance should be carried out in future research. 3.3.4. Summery In this section, the size of street trees and the effect of tree planting patterns are simulated and discussed. Increasing tree crown diameter could provide significant Ta reduction than increasing tree height, because of the bigger shading areas could reduce solar energy absorption in summer. Planting trees without space between the tree crowns could maximize to the environmental effect. Urban tree cover also reduces the UHI effect in nighttime, and the correlation (R2) between tree cover (SVF) and urban Ta is about 0.64 at summer mid-night. The effect of UHI mitigation by planting trees is not only at the tree height level, but also at higher elevation which could contribute on building energy consumption reduction of high-rise buildings in city central areas.

CONCLUSION Using numerical simulation, the effect of street vegetation planting, albedo and urban canopy characteristics on urban thermal environment in Montreal are discussed and demonstrated. These simulation comparisons demonstrate the effects on each environmental factor for a typical summer day and provide hints for mitigation of the urban heat island (UHI) and design of new urban development. The effectiveness of each UHI mitigation strategy are evaluated, in order to providing hints for policy guidelines development.



173

ACKNOWLEDGMENTS This research was funded by a Discovery grant from the Natural Resources and Engineering Council of Canada (NSERC) and partially funded by an NSERC Postdoctoral Fellowships Program. The authors wish to express appreciate to Professor Michael Bruse (University of Mainz, Germany) for providing the advanced free environmental simulation program (ENVI-met), and Professor Andreas Matzarakis (Alberts-Ludwigs-University Freiburg, Germany) for providing the advanced free model “RayMan.”

REFERENCES Agarwal Y, Weng T, Gupta R K. The energy dashboard: improving the visibility of energy consumption at a campus-wide scale[C]//Proceedings of the First ACM Workshop on Embedded Sensing Systems for EnergyEfficiency in Buildings. ACM, 2009: 55-60. Akbari H, Pomerantz M, Taha H. 2001. Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas. Solar energy, 70(3): 295-310. Akbari H, Konopacki S. 2004. Energy Effect of Heat island reduction strategies in Toronto Canada. Energy, 29, 191-210. Ali-Toudert F, Mayer H. Numerical study on the effects of aspect ratio and orientation of an urban street canyon on outdoor thermal comfort in hot and dry climate [J]. Building and Environment, 2006, 41(2): 94-108. Arbor Day Foundation. http://www.arborday.org/trees/ (accessed November 2014). Asaedaa T, Ca VT. 2000. Characteristics of permeable pavement during hot summer weather and impact on the thermal environment. Building and Environment, 35:363-375. Bahadur K C K. 2009, Mapping soil erosion susceptibility using remote sensing and GIS: a case of the Upper Nam Wa Watershed, Nan Province, Thailand [J]. Environmental geology, 57(3): 695-705. Bass B, Krayenhoff ES, Martilli A, Stull RB, Auld H. 2003. The impact of green roofs on Toronto’s urban heat island. Proceedings of the First North American Green Roof Conference: Greening Rooftops for Sustainable Communities, 292–304.


174


Bonan G B. 2000. The microclimates of a suburban Colorado (USA) landscape and implications for planning and design. Landscape and Urban Planning, 49(3): 97-114. Bottyan Z, Unger J. A multiple linear statistical model for estimating the mean maximum urban heat island [J]. Theoretical and Applied climatology, 2003, 75(3-4): 233-243. Brown M.J, Grimmond C.S.B. and Ratti C. Comparison of methodologies for computing sky view factor in urban environment. Internal Report Los Alamos National Laboratory, 2001; Los Alamos, NM, LA-UR-01-4107. Bröde P, Krüger E L, Rossi F A, et al., Predicting urban outdoor thermal comfort by the Universal Thermal Climate Index UTCI—a case study in Southern Brazil [J]. International journal of biometeorology, 2012, 56(3): 471-480. Bruse. M. ENVI-met website: www.envimet.com (2015). Bruse. M. The influences of local environmental design on microclimate – development of a prognostic numerical model ENVI-met for the simulation of Wind, temperature and humidity distribution in urban structures. Ph.D. Thesis, Germany: University of Bochum (1999). Chandurkar P P, Pajgade D P S. Seismic analysis of RCC building with and without shear wall [J]. International Journal of Modern Engineering Research, 2013, 3(3): 1805-1810. Chow W T L, Brazel A J. Assessing xeriscaping as a sustainable heat island mitigation approach for a desert city [J]. Building and Environment, 2012, 47: 170-181. Gal T, Lindberg F, Unger J. Computing continuous sky view factors using 3D urban raster and vector databases: comparison and application to urban climate [J]. Theoretical and applied climatology, 2009, 95(1-2): 111-123. Gui J, Phelan PE, Kaloush KE, Golden JS. 2007. Impact of pavement thermophysical properties on surface temperatures. Journal of Materials in Civil Engineering, 1:683-690. Gulyás Á, Unger J, Matzarakis A. Assessment of the microclimatic and human comfort conditions in a complex urban environment: modelling and measurements [J]. Building and Environment, 2006, 41(12): 1713-1722. Herold M, Menz G, Clarke K C. Remote sensing and urban growth modelsdemands and perspectives[C]//Symposium on remote sensing of urban areas. 2001, 35. Huang L, Cai H, Cui X, et al., Layered design and simulation of TD-SCDMA system based on BBU+ RRU[C]//Automatic Control and Artificial



175

Intelligence (ACAI 2012), International Conference on. IET, 2012: 12931296. Huttner S, Bruse M, Dostal P, et al., Strategies for mitigating thermal heat stress in Central European cities: The project KLIMES [C]//Proceedings of the 7th International Conference on Urban Climate. 2009, 29. Hydro Quebec website, 2014, http://www.hydroquebec.com/en/ Rao K V. 2012, Characterisation of Urban Heat Islands in one of the Most Urbanised Corridors of India from Space Based Multi-Sensor, SpatioTemporal Data [J]. AGSE 2012–FOSS4G-SEA, 17. Kentucky’s Champion Trees. (2014). Bigtreescommonname. http://forestry. ky.gov/ChampionTrees/Documents/bigtreescommonname.pdf. Krayenhoff ES, Martilli A, Bass B, Stull R. 2003. Mesoscale simulation of urban heat mitigation strategies in Toronto, Canada, Fifth International Conference on Urban Climate, 1-5 September, Lodz, Poland. Lin T P, Matzarakis A. Tourism climate and thermal comfort in Sun Moon Lake, Taiwan [J]. International Journal of Biometeorology, 2008, 52(4): 281-290. Matzarakis A. Climate and bioclimate information for tourism in Greece[C]//Proceedings of the first international workshop on climate, tourism and recreation, ed. by A. Matzarakis and C. R. de Freitas. International Society of Biometeorology, Commission on Climate Tourism and Recreation. 2001: 171-183. Matzarakis A, Nastos P T. Human-biometeorological assessment of heat waves in Athens [J]. Theoretical and applied climatology, 2011, 105(1-2): 99-106. Mayer H, Höppe P. Thermal comfort of man in different urban environments [J]. Theoretical and Applied Climatology, 1987, 38(1): 43-49. Mills E S. 2005, Why do we have urban density controls? [J]. Real Estate Economics, 33(3): 571-585. Millward A A, Sabir S. 2011. Benefits of a forested urban park: What is the value of Allan Gardens to the city of Toronto, Canada? Landscape and urban planning, 100(3): 177-188. Natural Resources Canada, 2009. Ng E, Chen L, Wang Y, Yuan C. 2012. A study on the cooling effects of greening in a high-density city: An experience from Hong Kong. Building and Environment, 47:256–271. North Dakota Tree Information Center. N.D. Tree Handbook. http://www.ag. ndsu.edu/trees/handbook/ndhand-1.htm (accessed November 2014).


176


Oke TR. 1988. Street design and the urban canopy layer climate. Energy and Buildings; 11:103-113. Pan Y, Yin R, Huang Z. 2008, Energy modeling of two office buildings with data center for green building design [J]. Energy and Buildings, 40(7): 1145-1152. Papadopoulos A.M. 2001. The influence of street canyons on the cooling loads and the performance of air conditioning systems. Energy and Buildings, 33: 601-607. Peng S, Piao S, Ciais P, et al., Surface urban heat island across 419 global big cities [J]. Environmental science and technology, 2011, 46(2): 696-703. Pisello AL, Cotana F. 2014. The thermal effect of an innovative cool roof on residential buildings in Italy: Results from two years of continuous monitoring. Energy and Buildings, 69:154-164. Sailor D J. 2011. A review of methods for estimating anthropogenic heat and moisture emissions in the urban environment. International Journal of Climatology, 31(2): 189-199. Shashua-Bar L, Pearlmutter D, Erell E. 2009. The cooling efficiency of urban landscape strategies in a hot dry climate. Landscape and Urban Planning, 92(3): 179-186. Sheppard S R J. 2008, Local climate change visioning: A new process for community planning and outreach using visualization tools [J]. Plan Canada, 48(1): 36. Steinicke F, Bruder G, Hinrichs K, et al., Transitional environments enhance distance perception in immersive virtual reality systems [C]//Proceedings of the 6th Symposium on Applied Perception in Graphics and Visualization. ACM, 2009: 19-26. Synnefa A, Santamouris M, Apostolakis K. 2007. On the development, optical properties and thermal performance of cool colored coatings for the urban environment. Solar Energy, 81(4), 488–497. Taubenböck H, Wegmann M, Roth A, et al., Analysis of urban sprawl at mega city Cairo, Egypt using multisensoral remote sensing data, landscape metrics and gradient analysis [C]//33rd International Symposium on Remote Sensing of Environment (Stresa, Italy: http://elib. dlr. de/58840/1/136_H. Taubenbock. pdf). 2009. Thorsson S, Lindberg F, Eliasson I, et al., Different methods for estimating the mean radiant temperature in an outdoor urban setting [J]. International Journal of Climatology, 2007, 27(14): 1983-1993.



177

Unger J. Intra-urban relationship between surface geometry and urban heat island: review and new approach [J]. Climate research, 2004, 27: 253264. Unger J. Connection between urban heat island and sky view factor approximated by a software tool on a 3D urban database [J]. International Journal of Environment and Pollution, 2009, 36(1): 59-80. Vector Magic, Inc., WeatherSpark website: weatherspark.com; 2014. Ville de Montreal. (2013). Biodiversity Report. Wang Y, Akbari H. 2014. Effect of Sky View Factor on Outdoor Temperature and Comfort in Montreal [J]. Environmental Engineering Science, 31(6): 272-287. Watson I D, Johnson G T. Graphical estimation of sky viewǦfactors in urban environments [J]. Journal of climatology, 1987, 7(2): 193-197. Yoshino H, Yoshino Y, Zhang Q, et al., 2006, Indoor thermal environment and energy saving for urban residential buildings in China[J]. Energy and buildings, 38(11): 1308-1319.
