Beruflich Dokumente
Kultur Dokumente
Department of Landscape Architecture, Faculty of Design & Architecture, Universiti Putra Malaysia (UPM), 43400 Serdang, Selangor, Malaysia
Welsh School of Architecture, Cardiff University, Bute Building, King Edward VII Avenue, CF10 3NB, United Kingdom
c
Solar Energy Research Institute (SERI), Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 30 April 2012
Received in revised form
9 July 2012
Accepted 25 July 2012
This study focuses on the optimum cooling effect of trees with ground materials modication in mitigating the urban heat island (UHI) and the benets towards building energy performance in tropical
climate. The modication focused on both physical properties e i.e. tree canopy density and quantity;
and the albedo values of ground materials. Two phases of methodology were developed and applied
using eld measurement and computer simulation. This study measured the average monthly UHI
intensity found to be 2.6 C. In mitigating its impact, higher levels of tree canopy density (LAI 9.7)
coupled with cool materials (albedo of 0.8) produced the largest urban air temperature reduction.
Simulations predicted an average air temperature reduction of 2.7 C when compared with the current
condition. Further, both modications were found to produce a potential building cooling load reduction
of up to 29%. In fact, the optimum improvement of both outdoor and indoor environment was inuenced
by three major physical factors, namely, larger tree quantity, higher canopy density and cool materials.
Thus, it is suggested that appropriate guidelines, inuencing implementation of these improvements
could be implemented in order to mitigate the UHI effect in tropical climate.
2012 Elsevier Ltd. All rights reserved.
Keywords:
Urban heat island mitigation
Tree canopy density and quantity
Cool pavement
Outdoor and indoor air temperature
Optimum cooling effect
Building energy savings
1. Introduction
The main causes of the urban heat island phenomenon are
recognized to be the consequences of increased urbanization and
abrupt changes in the outdoor environment. These temperature
rises in the urban environment are caused by the changes of street
surface materials and reduction of green areas [1]. The variety of
urban grids and buildings generate a wide range of different streets,
squares, courts and open spaces that further modify local climate
into urban microclimates [2]. In addition, the introduction of
concrete buildings has crowded out vegetation and trees [3].
Consequently, heat islands contribute to human discomfort, health
problems, higher energy bills and increased pollution [4]. In
temperate climate, the maximum urban heat island effect can be
found only during the summer season. However, in the tropics it
can be found during hot dry season and anytime in a year due to the
high exposure to solar radiation throughout the year [5,6]. This
246
247
Fig. 1. Twelve location points for eld measurement with different landscape environments and one weather station location point. Source: Authors screenshot of Google Earth
from August 25, 2008.
Lumpur (Lat. 2 550 N, Long. 101420 E). The climate is hot and humid
with an annual average temperature of 26.1 C (with an average
maximum of 27.5 C and minimum 25 C), annual average relative
humidity of 62.6%, an average of 6.1 h of sunlight per day and
4.39 kW h m2 of annually average solar irradiation. Generally, the
wind is light and variable with speeds ranging from 0 to 7.5 m/s, but
wind movement is still weak and sometimes can be static.
3.2. Field measurement programme
The eld measurement programme was designed to measure air
temperature (Ta) and relative humidity (RH%) using Tiny Tag TGP4500 Ultra and TGU-4500 Plus portable data logger (with the
same operating range 25 to 85 C/0e100% RH) at twelve (12)
selected locations. This equipment was used together with additional instruments such as Stevenson screens to provide solar cover.
Ten (10) numbers Tiny Tag data loggers contained in Stevenson
screens were installed on lighting posts, with a further two no. on
tree trunks, all at about 3 m height above the ground. The whole
area was divided into three major environmental conditions such
as green areas, open hard surface areas and mixed environment
areas with division of spots areas (Table 1). Data based on the study
of satellite images, data from Perbadanan Putrajaya and eld
observation were compared with data obtained from a weather
station located on the roof top of the Perbadanan Putrajaya
building. The weather station records ve weather parameters: air
temperature, relative humidity, wind velocity, wind direction and
solar radiation, in order to evaluate the current environment.
Output data from the weather station was required as a basic
meteorology setting in simulating the current climate condition
Name of areas
1
2
3
Wawasan Park
Mahkamah
Perbadanan
4
5
Putra Square
P3P Road
Rakyat Square
Open Space
8
9
P2N Road
Perdana Boulevard
10
11
12
13
248
trees were not evaluated, indeed, the indirect effect was the major
concern. In the preliminary simulations, both the average air and
surface temperature were underestimated. Thus, some of the
input data from the weather station had to be calibrated in order
to achieve a better correlation with the measured results. The
results showed that air temperature and surface temperature was
inuenced by wind speed. Based on the model interpretations,
this may be due to the high wind speed causing high volume of
cool energy, simultaneously cooling the surrounding air. Thus, this
shows that diurnal amplitude tends to increase with decreases in
wind speed. This is similar to adjustment problems experienced in
Spangenbergs [23] study where it became necessary to nd
a better correlation with measured results due to the wind speed
input. It is noted that the measured wind speed varied strongly in
direction and speed throughout the whole day. Therefore, in order
to adjust the air and surface temperature curve, the average mean
wind speed measured was changed from 2.1 m/s to 1.1 m/s. The
initial temperature was increased by 2 C in order to obtain
a better correlation between measured and computed temperature. Considering the complexity of the three dimensional environment, ENVI-met was found to represent the trends of air and
ground surface temperature well, the 24 h average error was
found to be at 0.1 Ce0.8 C for air temperature and from 0.1 C to
1.5 C for ground surface temperature, respectively. In fact, similar
24 h air temperature curves were achieved with the calculated
correlation coefcient, R2, between measured and computed for
measurement point ranging from 0.860 to 0.975 for both
parameters (Table 2). Thus, it has been shown that the measured
and computed ENVI-met was correlated and dependable in presenting the current air and ground surface temperature of Persiaran Perdana.
3.3.2. Development of proposed scenarios
Besides the current condition, three scenarios were designed in
order to compare and predict the optimum cooling potential made
by vegetation and ground material modications. Thus, the basic
input congurations as presented in Table 3 are used throughout all
of the simulations to enable conrmation that the cause of modication was genuinely made by vegetation and ground materials.
The scenarios were as follows:
(A) Current condition;
Table 2
Correlation coefcients and average error between the measured and the computed
air and ground surface temperatures for each location in 24 h period.
Environment
Location points
1. Wawasan Park
2. Mahkamah
3. Perbadanan
4. Putra Square
5. P3P Road
6. Rakyat Square
7. Open space
8. P2N Road
9. Perdana Boulevard
10. Wisma Tani
Entrance
11. Wisma Tanis
Pedestrian Area
12. Tunku Abdul
Rahman Road
Air temperature
Ground surface
temperature
Correlation
coefcient (R2)
Average
error ( C)
Correlation
coefcient (R2)
Average
error ( C)
0.956
0.975
0.972
0.970
0.958
0.970
0.965
0.923
0.959
0.933
0.8
0.5
0.2
0.1
0.2
0.2
0.6
0.1
0.2
0.3
0.919
0.860
0.972
0.973
0.973
0.969
0.952
0.972
0.971
0.972
0.7
0.5
0.1
1.0
0.6
0.0
0.5
0.9
0.9
0.5
0.905
0.1
0.969
0.6
0.957
0.5
0.970
1.5
Current medium
density tree
Low
density
tree
High
density
tree
2.4
0.3e0.4
0.9
0.3e0.4
9.7
0.3e0.4
9.7
0.8
Tree quantities
Current
Add tree
Increase (%)
1. Wawasan Park
2. Mahkamah
3. Perbadanan
4. Putra Square
5. P3P Road
6. Rakyat Square
7. Open space
8. P2N Road
9. Perdana Boulevard
10. Wisma Tani
Entrance
11. Wisma Tanis
Pedestrian Area
12. Tunku Abdul
Rahman Road
13. Other Areas
90
28
36
20
0
64
4
28
0
0
0
0
0
30
124
22
288
36
4
0
0
0
0
60%
100%
35%
98%
56%
100%
0
10
10
30
75%
4810
4502
94%
Total
5,100
5,036
99%
249
average albedos of 0.2 and 0.4 respectively. This is due to the large
open area in which asphalt materials are used for parking spaces.
However, there are also public areas, i.e. squares and open spaces
designed with higher albedo pavements such as white/grey granite
and polished white granite with albedos of 0.65 and 0.8 respectively. Furthermore, for building properties the average albedo for
walls is 0.3 and the average albedo for a roof is 0.5 [26].
3.4. HTB2 software
In further analyzing on the prediction of the indirect effect of
building energy savings based on the current, and three different
environmental scenarios modication resulting from ENVI-met,
the detailed building heat transfer simulation programme HTB2
was used. HTB2 [27] is a dynamic building energy simulation model
that can account for complex, time varying climatic and occupation
conditions in the prediction of cooling loads and indoor environmental conditions in a building [28e30]. Preferences come when
the model can predict the thermal performance of a building when
it is subject to the inuence of outdoor temperature, solar gain,
shading, ventilation and inltration. Thus, one of the strengths that
was applied for this study is the ability of the user of this model to
easily alter or replace all input components, for example meteorological conditions, ventilation, building construction and shading
properties in order to assess building energy performances. Based
on this exibility, comparison between different outdoor microclimate conditions that inuenced indoor environment of buildings
can be predicted and simulated using this model. Besides, it also
allows the user to dene complex heating and cooling schedules,
occupancy, lighting, and internal load intensity patterns, and to
vary control settings during run-time, to mimic realistic occupation
conditions [30]. HTB2 has been shown to be capable of providing
predictions that correlate with measurements in buildings [30,31].
3.4.1. Simulation development
In order to predict the indirect effect of building energy savings
based on the current, and three different environmental scenarios
modication resulting from ENVI-met, the layout of the Industrial
Design building, which was formed and selected as a basis for this
energy savings study. The building was selected due to its
simplicity and yet extreme outdoor environmental condition, with
fully hard, low albedo, pavement, that allows signicant exploration of the impact of different external environments through
simulation. This single-storey building has a simple layout with
major facades (entrance and second entrance) facing west and east,
with two other facades with windows facing north and south. The
oor comprises an open plan studio of 624 m2 sized area. Input data
for HTB2 was prepared according to the characteristics of typical
Malaysian construction, which is henceforth referred to as the
baseline model, namely: outdoor wall slabs with a painted 20 mm
thick plastered layer on the outdoors, a 110 mm brick wall layer, and
20 mm thick painted plastered layer on the indoors (Fig. 3); 12
windows with 6 mm single glazing; a solid ground oor with two
layer properties of 100 mm concrete and 1600 mm of earth and
nally a at roof composed of 100 mm of concrete, a 65 mm cavity
and 9 mm plasterboard indoors. This construction information
formed the basis for every building simulation in this study, with
the meteorological input varied for each scenario from ENVI-met
simulation output.
The additional information needed and gathered in the HTB2
input le was services data consisting of occupancy, lighting, small
power, heating and ventilation. The whole area sized at 624 m2 is
occupied by average of 28 people during ofce hours from 09:00 to
17:00 h in weekdays. These occupancy patterns were assumed to be
applicable on all days throughout the measurement period. These
250
Fig. 2. Three dimensional model outputs in four different conditions of environment (A) vegetation coverage and (B) mean LAI for environmental condition A to D.
details enabled the HTB2 software to calculate heat gains from the
users of the building. The light was assumed to be turned on from
09:00 to 12:30, turned off during the lunch hour and then turned
on again from 14:00 to 17:00 (Table 4). This information is used by
HTB2 to determine the internal heat gains from lighting the
building. This table also summarizes the assumed proles of energy
loads in the building caused by small computers. Thus, the information would enable the HTB2 software to calculate the internal
heat gains from these devices. The air conditioners were installed in
the building. The air conditioners in the building served the whole
areas from 08:00 to 17:00 h. Full time running of the air conditioning was chosen in order to determine those energy savings that
can be obtained from the inuence of the outdoor environment
during ofce hours and was set at 25 C. When the building was air
conditioned with windows shut, the assumed inltration rate for
this room was 0.5 ach. Additionally, the diary le was used for
Fig. 3. The basic layout of selected building (left) and the wall slabs layer properties (right).
09:00e12:30
12:30e02:00
02:00e05:00
34 nos 65 W standard
uorescent light
22 nos computers
(150 W)
2210
0
2210
3300
3300
3300
251
Fig. 4. Air temperatures at four different conditions; current condition (A) and modied environments (B) to (D) were compared at 15:00 in 3 m height.
252
yellow spots such as Putra Square and Rakyat Square also had
higher temperatures. This condition is believed to be due to the lack
of greenery, lower ground albedo materials and higher building
density within that area. On the other hand, there are greenery
areas such as Wawasan Park that produced a noticeable cooling
effect and kept the area cooler during the peak hours. However, the
effect was not enough to provide the optimum cooling to the entire
area of Persiaran Perdana, especially at the focal point where most
activities occur during the daytime.
Meanwhile, after modications were made to produce condition
(B), the magenta spots reduced, turning from yellowish to greenish
in colour, indicating that the air temperature has been reduced, as
more vegetation has been added at the centre of boulevard. The
modied areas such as P3P Road, Perdana Boulevard area and along
Rakyat Square show an improvement in air temperature reduction,
where these improvements are believed to be due to the increase of
trees quantities from 60% to 100%. However, the effect of cooling is
reduced due to the use of loose canopy density trees in this
scenario. Indeed, the actual reduction achieved is believed to be
due to the increase in number of trees rather than from an increase
in tree canopy density.
In condition (C), the results are similar, where the greenish
colour spotted in condition (B) becomes more bluish with fewer
magenta spots, indicating that the areas have become cooler than
the previous scenarios (B), this time due to increased tree densities
(LAI). Thus, it is suggested that higher canopy density trees create
a remarkable modication reducing air temperature. This condition
reduced the ground surface temperature that inuenced the
evapotranspiration rate and consequently reduced the air
Fig. 5. Average (above) and maximum (below) air temperatures for 12 location points from condition A to D.
condition (B), (C) and (D), respectively. On one hand, the results
show that almost 60% of average temperature differences were
achieved through the increased in tree canopy densities, loose
(condition B) to high canopy density trees (condition C). While, it
also noted that the differences in changing to cool materials was
less than 10% (condition C to D). These results revealed that within
the whole 24 h periods, the highest reduction of air temperature
were originally obtained from high canopy density and large tree
quantities while, a lesser effect was found from the changes of
ground surface material properties. Thus, it sufces to say here that
modication of the trees physical aspect offers larger cooling
potential than is available through ground material modications.
It can be seen in Fig. 6 where the variation of air temperature was
inuence by the variation of tree canopy density and quantities. The
location with higher LAI values and numbers of trees offers a better
reduction. Alternatively, it can be seen in P3P Road location where
the increase of quantities of high canopy density trees (LAI 9.7),
from 0 to 124 numbers additional trees, created a remarkable
reduction of 2.5 C. However, it was found only 2 C differences
when the albedo of the ground surface was changed from 0.3 to 0.8
albedo of ground material. On the contrary, it is believed that
modication of ground materials could enhance more on the
optimum cooling effect of vegetation due to low absorption of solar
radiation and high ground surface moisture contents during the
daytime. In fact, both combinations can provide an optimum effect
in comparison to current conditions with the range of average air
temperature was found to be from 27.6 C to 30.3 C. This makes
a maximum average difference with magnitude up to 2.7 C.
As the maximum air temperature is the main concern in minimizing the effect of UHI, the pattern is very obvious and there is
signicant difference in every condition, (B), (C) and (D), from the
baseline A (Fig. 5). By having a large amount of tree quantities the
largest air temperature differences can be found in each of the test
location point, besides, after modication each critical location was
found to have improved and to compare favourably with those
locations originally dened as green spaces (i.e. Wawasan Park,
Mahkamah, Perbadanan). Condition (D) shows the lowest
maximum air temperature, where the highest maximum differences were found in Open Space, with a magnitude of 2.7 C
compared to the baseline condition (A). This is due to the effectiveness of the cooling processes of trees and ground surface during
peak hours. During these hours, it is believed that all cooling
processes were taken into consideration such as radiation ltration,
253
Fig. 6. Air temperature reduction inuence by variation of tree canopy density, quantity and types of ground material (albedo).
254
Fig. 7. Summary on the comparison of outdoor and indoor air temperature modication in six selected location points in Persiaran Perdana.
Condition D
Ta mean out 0:0091 Q 0:8216 R2 0:96
(1)
Ta mean in 0:0077 Q 0:6217 R2 0:91
(2)
Condition C
Ta mean out 0:0091 Q 0:783 R2 0:74
(3)
Ta mean in 0:0078 Q 0:5707 R2 0:89
(4)
Condition B
Ta mean out 0:0078 Q 0:4374 R2 0:93
(5)
Ta mean in 0:0059 Q 0:318 R2 0:91
(6)
The slope of the lines indicates the average outdoor and indoor
air temperature to tree quantity increases. The slope value in
condition (D) is corresponding to a 0.01 C average outdoor air
temperature reduction per single high canopy density trees added
to the site, starting from 0.8 C reduction in 4 trees. Meanwhile,
condition (C) and (B) shows average air temperature reduction of
0.01 C, starting from 0.8 to 0.4 C in 4 trees, respectively. This
shows the condition (C) and (D) provide highest reduction in
average outdoor air temperature due to the modication. This
correlation also conrmed the effect of cooling from high canopy
density trees on outdoor air temperature modication. On the
other hand, the effect of outdoor modication provides a signicant
reduction in indoor air temperature. It can be observed that the
slope value corresponds to a 0.01 C average indoor air temperature
reduction per single trees added to the site, starting from 0.6 C
Fig. 8. Correlation between average (left) and maximum (right) outdoor and indoor air temperature and tree quantities in six selected location points.
Condition D
Ta max out 0:006 Q 1:3659 R2 0:92
(7)
Ta max in 0:006 Q 1:092 R2 0:88
(8)
255
Condition C
Ta mean out 0:006 Q 0:877 R2 0:92
(9)
Ta mean in 0:007 Q 0:7022 R2 0:88
(10)
Condition B
Ta mean out 0:0084 Q 0:3738 R2 0:95
(11)
Ta mean in 0:0061 Q 0:3301 R2 0:93
(12)
Fig. 9. Summary on the comparison of total cooling load based on each modication in six selected location points in Persiaran Perdana.
256
Fig. 10. Correlation between total cooling load percentage reduction and tree quantities in six selected location points.
Condition D TclY 0:1239 Q 10:682 R2 0:85
(13)
Condition C TclY 0:1312 Q 8:985 R2 0:90
(14)
Condition B TclY 0:1053 Q 4:3182 R2 0:68
(15)
The slope of the tted line indicates the total cooling load
percentage reduction to tree quantity increases. The slope values
within the range of 0.1 corresponding to a 0.12%, 0.13% and 0.11%
total cooling load reduction per single high canopy density trees
added to the site for condition (D), (C) and (B), starting from
a minimum reduction of 10.7%, 9.0% and 4.3% due to the planting of
4 trees, respectively. It appears that the building cooling load
reduces with the increase of tree quantities. In fact, the highest
building load reduction can be found in condition (D) where
application of high canopy density trees and cool material was
proven effective in improving building cooling load. Finally, it can
be concluded that modication of tree canopy density, quantity and
cool materials signicantly benet building energy savings in Persiaran Perdana.
5. Discussion
The ndings show that tree quantities and their canopy densities are among the paramount factors in providing signicant
257
[6] Tran H, Uchihama D, Ochi S, Yasuoka Y. Assessment with satellite data of the
urban heat island effects in Asian mega cities. Int J Appl Earth Obs 2006;8:
34e48.
[7] Shashua-Bar L, Hoffman ME. Vegetation as a climatic component in the design
of an urban street: an empirical model for predicting the cooling effect of
urban green areas with trees. Energ Buildings 2000;31:221e35.
[8] Akbari H, Pomerantz M, Taha H. Cool surfaces and shade trees to reduce
energy use and improve air quality in urban areas. Sol Energy 2001;70:
295e310.
[9] Solecki WD, Rosenzweig C, Parshall L, Pope G, Clark M, Cox J, et al. Mitigation
of the heat island effect in urban New Jersey. Environ Hazard 2005;6(1):
39e49.
[10] Givoni B. Urban design in different climates. Technical Report. WMO; 1989. p.
346.
[11] Brown RD, Gillespie TJ. Microclimate landscape design: creating thermal
comfort and energy efciency. New York: John Wiley & Sons; 1995.
[12] Taha H. Urban climates and heat islands: albedo, evapotranspiration, and
anthropogenic heat. Energ Building 1997;25:99e103.
[13] Dimoudi A, Nikolopoulou M. Vegetation in the urban environment: microclimatic analysis and benets. Energ Building 2003;35:69e76.
[14] Shahidan MF, Shariff MKS, Jones P, Salleh E, Abdullah AM. A comparison of
Mesua ferrea L. and Hura crepitans L. for shade creation and radiation modication in improving thermal comfort. Landscape Urban Plan 2010;97:
168e81.
[15] Fahmy M, Sharples S, Yahiya M. LAI based trees selection for mid latitude
urban developments: a microclimate study in Cairo, Egypt. Build Environ
2010:345e57.
[16] Kotzen B. An investigation of shade under six different tree species of the
Negev desert towards their potential use for enhancing micro-climatic
conditions in landscape architectural development. J Arid Environ 2003;55:
231e74.
[17] Jonckheere I, Fleck S, Nackaerts K, Muys B, Coppin P, Weiss M, et al. Review of
methods for in situ leaf area index determination: part I. Theories, sensors and
hemispherical photography. Agr Forest Meteorol 2004;121:19e35.
[18] Montes F, Pita P, Rubio A, Canellas I. Leaf area index estimation in mountain
even-aged Pinus silvestris L. stands from hemispherical photographs. Agr
Forest Meteorol 2007;145:215e28.
[19] Lalic B, Mihailovic DT. An empirical relation describing leaf area density inside
the forest for environmental modelling. J Appl Meteorol 2004;43:641e5.
[20] Ali-Toudert F, Mayer M. Numerical study on the effects of aspect ratio and
orientation of an urban street canyon on outdoor thermal comfort in hot and
dry climate. Build Environ 2006;41:94e108.
[21] Wong NH, Yu C. Tropical urban heat islands: climate, buildings and greenery.
New York: Taylor and Francis; 2009.
[22] Bruse M. ENVI-met V3.1, a microclimate urban scale model. www.envi-met.
com. retrieved on 28 February, 2008.
[23] Spangernberg J, Shinzato P, Johansson E, Duarte D. Simulation of the inuence
of vegetation on microclimate and thermal comfort in the city of Sao Paulo.
Piracicaba: Rev SBAU; 2008. p. 1e19.
[24] Bourbia PF, Mansouri O. The effect of albedo on urban street microclimate for
semi arid climate constantine. In: world renewable energy congress (WRECX).
Glasgow, Scotland: 2008.
[25] Bruse M, Fleer H. Simulating surface-plant-air interactions inside urban
environments with a three dimensional numerical model. Environ Model
Softw 1998;13:373e84.
[26] Putrajaya P. Putrajaya annual report 2008. Putrajaya, Malaysia: Perbadanan
Putrajaya; 2008.
[27] Alexander DK. HTB2 users manual. United Kingdom: Welsh School of Architecture; 1996.
[28] Lomas KJ, Eppel H, Martin CJ, Bloomeld DP. Empirical validation of building
energy simulation programmes. Energ Building 1997;26:253e73.
[29] Bojic M, Yik F. Cooling energy evaluation for high-rise residential buildings in
Hong Kong. Energ Building 2005;37:345e51.
[30] Li Q, Jones P, Lannon S. Planning sustainable in Chinese cities: dwelling types
as a means to accessing potential improvements in energy efciency. In:
proceedings of the tenth international conference of the international
building performance simulation (IBPSA). Beijing, China: 2007.
[31] Shahidan MF. The potential optimum cooling effect of vegetation with ground
surface physical properties modication in mitigating the urban heat island
effect in Malaysia, Cardiff University. Unpublished thesis. 2011.