EFECTOS NATURACIÓN
-
Upload
ana-garcia-garcia -
Category
Documents
-
view
224 -
download
0
Transcript of EFECTOS NATURACIÓN
-
7/30/2019 EFECTOS NATURACIN
1/14
Building and Environment 43 (2008) 480493
Temperature decreases in an urban canyon due to green walls and green
roofs in diverse climates
Eleftheria Alexandria, Phil Jonesb,
aMantzakou 2-6, 114 73 Athens, GreecebWelsh School of Architecture, Cardiff University, King Edward VII Avenue, Cardiff CF10 3NB, UK
Received 23 February 2006; received in revised form 24 July 2006; accepted 31 October 2006
Abstract
This paper discusses the thermal effect of covering the building envelope with vegetation on the microclimate in the built environment,
for various climates and urban canyon geometries. A two-dimensional, prognostic, micro scale model has been used, developed for the
purposes of this study. The climatic characteristics of nine cities, three urban canyon geometries, two canyon orientations and two wind
directions are examined. The thermal effect of green roofs and green walls on the built environment is examined in both inside the canyon
and at roof level. The effects of this temperature decrease on outdoors thermal comfort and energy savings are examined. Conclusions
are drawn on whether plants on the building envelope can be used to tackle the heat island effect, depending on all these parameters
taken into consideration.
r 2006 Elsevier Ltd. All rights reserved.
Keywords: Green roofs; Green walls; Urban canyon; Plants; Built environment
1. Introduction
Since the beginning of human existence man has clearly
intended to alter his microclimate, to a more human-
friendly one, protecting himself from extreme climatic
conditions. Even from the first evidence of Neolithic houses
and settlements, it is obvious that they were not sited in a
purely natural environment, but in a part of nature
transformed according to a human plan [1]. With the
evolution of human societies, settlements were trans-
formed, evolved into villages, towns or cities, developed
or faded away, according to the geographical, economic,
social and cultural transformations taking place through-out time. With the Industrial Revolution, urban spaces
expanded dramatically, much faster and with much more
significant changes than in their previous evolutionary
periods. The large areas modern cities occupy, their
structure, materials and the general lack of vegetation
cannot but have altered the climatic characteristics of
urban spaces.
These changes have a direct effect on the local climate ofurban spaces, especially the central parts of the city,
causing a significant rise of the urban temperature and
other alterations, known as the heat island effect. This may
cause serious local climatic unpleasant conditions and even
imperil human health, especially for cities in climates with a
distinctively hot season [2,3]. The moderation of extreme
heat in the local environment of such climates could mean
not only their sustainability, but also the potential of
occupying them without the morbidity and mortality risks
caused by excessive heat [4,5].
On prima facie evidence, the general lack of vegetation in
existing cities is one of the factors affecting the formationof raised urban temperatures. In most urban spaces,
appreciable amounts of vegetation exist mostly concen-
trated in parks or recreational spaces. Although parks
manage to lower temperatures within their vicinity [69],
they are incapable of thermally affecting the concentrated
built spaces where people live, work and spend most of
their urban lives. By placing vegetation within the built
space of the urban fabric, raised urban temperatures can
decrease within the human habitats themselves and not
only in the detached spaces of parks. Urban surfaces which
ARTICLE IN PRESS
www.elsevier.com/locate/buildenv
0360-1323/$- see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.buildenv.2006.10.055
Corresponding author.
http://www.elsevier.com/locate/buildenvhttp://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.buildenv.2006.10.055http://localhost/var/www/apps/conversion/tmp/scratch_3/dx.doi.org/10.1016/j.buildenv.2006.10.055http://www.elsevier.com/locate/buildenv -
7/30/2019 EFECTOS NATURACIN
2/14
-
7/30/2019 EFECTOS NATURACIN
3/14
specific humidity (in kg/kg), c1 is the isobaric specific heat
of component 1 (moisture) of the mixture (in J/kg K) and c2the isobaric specific heat of component 2 (air) of the
mixture (in J/kg K). According to Eckert and Drake [12],
the effect of thermal diffusion on heat transfer (fourth term
of Eq. (1)) can be neglected in normal engineering mass-
transfer processes. However, they point out [12] that itcontributes essentially when temperature gradients are
extremely large. In the description of the thermal ex-
changes in the built environment, where temperature
gradients in the boundary layer of surfaces exposed to
direct solar radiation are relatively large, the expression of
thermal diffusion is essential for the accurate description of
the phenomenon.
When air velocity is considered, eddy diffusion is much
stronger than molecular diffusion (conduction) in the air in
the atmosphere away from the boundary layer of the
surface. Despite the fact that molecular diffusion always
takes place in the air, it is omitted from both heat and mass
transfer in the air at these levels, as it is 104105 smaller
than eddy diffusion [13]. The effect of vapour gradients
onto temperature in the air nodes well above the ground is
expressed through eddy diffusion coefficients [14,15]. Heat
and mass transfer in the two-dimensional model of the
binary airwater vapour mixture thus becomes:
q
qt u
q
qx w
q
qz
T
q
qzKHz
qT
qz
q
qxKHx
qT
qx
,
(3)
q
qt u
q
qx w
q
qz q q
qzKEz
qq
qz q
qxKEx
qq
qx , (4)
where KHz and KEz are the eddy diffusion coefficient of
energy and water vapour, respectively, in the vertical axis
and KHx and KEx are the respective diffusion coefficients in
x-axis. The expression of these diffusion coefficients is
given by the MoninObukhov similarity theory [14,15].
The water vapour gradients are taken into consideration in
the calculation of the eddy diffusion coefficients of energy.
Regarding solid materials, they are considered as a
system, consisting of a capillary-porous building material
in the medium of wet air and in a region of positive
temperatures (no ice). The equations describing the one-
dimensional heat and mass transfer can be expressedby [11]
dT
dt ac
q2T
qz2
l
cc
qq
qt, (5)
dq
dt am
q2q
qz2, (6)
where ac is the building material thermal diffusion
coefficient (in m2/s), cc is the building material specific
heat capacity (in J/kg K), e is the evaporation number of
the building material, and am is the diffusion coefficient of
moisture in the building material (in m2/s).
Regarding plants, they are considered to be a layer
consisting of canopy leaves and the air among them.
Equations describing heat and mass transfer in the air are
the ones given by Eqs. (1) and (2), while heat transfer in the
leaf is given by
rcpdT
dt Fn C lE, (7)
where r is the density of the leaf tissue (in kg/m3), cp is
the specific heat capacity of the leaf tissue (in J/kg K), Tis the leaf surface temperature (in K), Fn is the net heat
gain from radiation (in W/m2), C is the net sensible heat
loss (in W/m2) and lE is the net latent heat loss (in W/m2).
Radiative heat exchanges between the canyon surfaces
have been described analytically, according to the radiative
heat transfer theory [16] and not with the use of a
combined convection and radiative heat transfer coeffi-
cient. Thus the radiative heat exchanges between surfaces
with different emissivities in closed enclosures is expressed
by
qii
XNj1
1
j 1
Fijqj H0i
XNj1
FijEbi Ebj for i 1; 2; . . . ; N, 8
where qi is the radiation emitted from the surface i(W/m2),
ei is the isurfaces emmisivity and ej is the emmisivity of the
rest of the surfaces, Fij is the view factor of surface i
towards surface j, H0i is any external radiation arriving at
surface i, and Ebi equals to sTi4, where s is the
StefanBoltzmann constant (5.67 108
W/m2
K4
) and Tiis the temperature of the ith surface.
Climatic characteristics, such as air temperature, relative
humidity and wind speed, are set as the boundary nodes of
the model, placed 10 m above the upper part of roofs.
These climatic characteristics, as well as solar radiation
derive from meteorological data from METEONORM
[17]. Solar radiation is input onto the surfaces, according to
their orientation, inclination and shading pattern. The
shading pattern, determined by the canyon geometry and
the geographic latitude, was calculated with the software
ECOTECT [18], where the same canyon geometries, as the
ones described below were input, for the different latitudes
and longitudes examined. Air velocities in the vicinity of
the canyon were calculated with the CFD code WinAir4
[19]. WinAir4 is an in-house code, which uses the fixed
viscosity models, with a variation on simple solution
scheme. The canyon geometries in the CFD model are
the same as for the heat and mass transfer model. As the
CFD model is three-dimensional, the length of the canyon
was 40 m and the rest of the canyon dimensions (height,
width) varied, according to the canyon geometry, as
discussed below. The CFD code mesh was also the same
as for the two-dimensional heat and mass transfer model,
as described in Fig. 1. The grid is not uniform; near the
building and road surfaces the grid is 0.30 m, while two
ARTICLE IN PRESS
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493482
-
7/30/2019 EFECTOS NATURACIN
4/14
nodes away from the surface, the grid varies according to
the canyon geometry, from 1 to 3 m for the canyon width,
and from 1 to 2 m for the canyon height. For the length, it
is constantly set at 2 m. At the boundary surface are set
10 m away from the roof of the buildings, 10 m away from
the windward surface and 100m away from the leeward
surface when the building height is 5 m, and 10, 20 and200 m, respectively, when the building height is 10 m. The
input climatic data of wind speed, air temperature and
relative humidity are input as the boundary conditions for
each hour of the diurnal profile. The heat gains on
buildings and the street are calculated according to the
solar radiation absorbed by the surface, which depends on
its orientation, and shading pattern, the latter having been
defined by [18]. The output air velocities from the CFD
code in the middle of the canyon are input at the respective
nodes of the two-dimensional heat and mass transfer
model.
Four types of vegetation covering the building envelope
are examined for each canyon geometry:
(a) a base case, where no green is placed in and around the
canyon, referred to as the no-green case,
(b) the green-roofs case, where both roofs are covered
with vegetation (ground-covering grasses)
(c) the green-walls case, where both walls inside the
canyon are covered with vegetation (ivies) and
(d) the green-all case, where both roofs and walls are
covered with vegetation.
Three types of canyon geometries are examined, accord-
ing to the wind flow developed in each:
(a) a canyon with height (H) 10m and width (W) 5 m ,
referred to as H10W5 canyon, where, according to
Santamouris [6], skimming flow is developed, with very
low air velocities, and sun shaded,
(b) a canyon with 5 m height and 10 m width, referred to as
H5W10 canyon, where wake interference flow is
developed, with bigger air velocities and more exposed
surfaces to direct solar radiation and
(c) the H5W15 canyon, with 5 m height and 15 m width,
where isolated roughness flow is developed, with much
larger air velocities, and greater exposure to solar
radiation.
The canyons are examined with two orientations:
(a) one where the canyons axis was parallel to the
EastWest axis (referred to as the EW canyon) and
(b) one where the canyons axis was parallel to the
NorthSouth axis (referred to as the NS canyon).
Two directions of wind flow are considered:
(a) one aligned to (referred as x) and
(b) one parallel to the canyons axis (referred as y).
Buildings are made of concrete, and the street is covered
with asphalt. A summary of the hydrothermal properties of
the materials and vegetation considered in the canyons is
made in Table 1. All these cases are examined for nine cities
in nine different types of climates, where cities and
evapotranspiring vegetation can be found. Based on
Koeppens climatic classification [20], the nine citiesstudied, and the climatic type in which they belong, are
summarised in Table 2. All cases are examined for a typical
day of their hottest month. Their climatic data have
derived from hourly data from METEONORM [17]. The
effect of vegetation on the urban texture of each city is
examined for its hottest month. For Athens, Hong Kong,
London, Montre al, Moscow and Riyadh, July is chosen as
the hottest month, while for Mumbai May is used, for
Beijing June, and for Braslia September. The typical day
ARTICLE IN PRESS
Table 1
Hydrothermal properties of plants, soil, building materials (concrete) andstreet materials
Characteristic Concrete Asphalt Soil Plants
Specific thermal capacity
(MJ/m3K)
1.60 2.00 1.15 2.60
Thermal conductivity (W/mK) 1.70 1.30
Vapour diffusivity (106 m2/s) 0.55 1.58
Ratio of vapour diffusion
coefficient to total moisture
diffusion coefficient
0.20 0.10
Emissivity 0.94 0.81 0.94 0.94
Albedo 0.23 0.10 0.23 0.30
Hydraulic conductivity
(10
4
m/s)
0.01
Moisture potential, when soil is
saturated (cm)
49.0
Maximum volumetric water
content (m3/m3)
0.492
Coefficient b 10.40
Convective heat resistance
(s/m)
200
Resistance expressing the plant
type (s/m)
100
Canopy extinction coefficient 1.4
Level of soil moisture below
which permanent wilting of the
plant occurs (m3/m3)
0.25
Table 2
Table of cities studied
City Climate Location
London, UK Temperate 51.32N, 0
Montreal, Canada Subarctic 45.31N, 73.34W
Moscow, Russia Continental cool summer 55.45N, 37.37E
Athens, Greece Mediterranean 37.59N, 23.43E
Beijing, China Steppe 39.48N, 116.23E
Riyadh, Saudi Arabia Desert 24.38N, 46.43E
Hong Kong, China Humid subtropical 22.16N, 114.12E
Mumbai, India Rain forest 18.54N, 72.5E
Braslia, Brazil Savanna 15.48S, 47.54W
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493 483
-
7/30/2019 EFECTOS NATURACIN
5/14
of the hottest month is a 24-h profile whose climatic data
are averaged for each hour of the month examined, to
make a diurnal typical climatic profile for the month
studied. The maximum, minimum, average and daytime
average quantities of temperature, relative humidity, wind
speed and solar radiation on a horizontal plane are
presented in Table 3. In Figs. 25 the diurnal profile ofair temperature, relative humidity, solar radiation on a
horizontal plane and wind speed are presented for the
typical day of the hottest month of all nine cities.
3. Discussion and analysis
3.1. Direct cooling effects
Air and surface temperatures lower significantly in all
climates examined, when walls and roofs are covered with
vegetation, as can be observed in Figs. 710. The heat
fluxes on the vegetated and on the non-vegetated surfacesare very different. As can be observed in Fig. 6 for a green
roof and for a concrete roof in Montre al, the 24-h profile of
ARTICLE IN PRESS
T
able3
M
aximum,minimum,averageandday-timeaveragevaluesofclimaticcharacteristics(totalsolarradiationonahorizontalplane,air
temperature,relativehumidity,windspeed)forthe24-hprofileof
th
etypicaldayofthehottestmonthofeach
city
C
ity
Totalsolarradiationonahorizontalplane
Airtem
perature
Relativehu
midity
Windspeed
Max
Min
Average
Daytime
average
Max
Min
Average
Daytime
average
Max
Min
Average
Day-time
average
Max
Min
Average
Daytime
average
A
thens,July
809.0
0.0
298.4
457.5
30.1
22.3
25.7
27.0
57.4
37.3
50.2
45.9
6.4
3.5
4.8
5.2
B
eijing,June
589.2
0.0
216.4
331.9
29.1
19.6
24.2
25.6
75.0
47.1
62.7
57.9
4.3
3.3
3.8
4.0
B
raslia,September
726.0
0.0
228.4
437.8
28.0
20.4
23.5
25.4
65.7
44.0
58.6
52.1
4.7
3.2
3.8
4.0
H
ongKong,July
704.3
0.0
235.8
417.2
33.1
26.7
29.5
30.9
85.1
62.9
77.1
70.9
4.9
3.0
3.8
4.2
L
ondon,
July
497.3
0.0
199.3
286.5
19.7
14.5
17.2
18.0
77.3
63.5
74.3
70.5
4.4
3.9
4.2
4.2
M
ontreal,July
692.2
0.0
261.9
376.5
25.9
17.1
21.2
22.3
77.7
49.7
66.1
62.1
5.5
3.3
4.3
4.6
M
oscow,
July
527.8
0.0
220.2
297.9
20.9
13.9
17.3
18.0
86.3
60.3
74.7
71.6
5.2
3.5
4.3
4.5
M
umbai,
May
846.9
0.0
283.8
502.1
33.5
25.3
28.9
30.8
79.1
53.1
69.5
62.4
4.6
2.3
3.3
3.8
R
iyadh,
July
850.0
0.0
284.2
466.9
42.8
31.2
36.5
38.3
45.8
23.5
36.2
32.4
5.9
3.8
4.8
5.2
0
5
10
15
20
25
30
35
40
45
1 9 11 13 15 17 19 21 23
Athens Beijing Brasilia
Hong Kong London Montreal
Moscow Mumbai Riyadh
Air Temperature
Time (Hours)
Temperature(C)
3 5 7
Fig. 3. Twenty four-hour profile of air temperature for the hottest month
of each city, which is input at the boundary nodes of the heat and mass
transfer model.
0
1 11 13 15 17 19 21 23
Athens
Beijing
Brasilia
Hong Kong
London
Montreal
Moscow
Mumbai
Riyadh
Total Solar Radiation on a Horizontal Plane
SolarRadiation(W/m2)
900
800
700
600
500
400
300
200
100
3 5 7 9
Time (Hours)
Fig. 2. Twenty four-hour profile of the total solar radiation on a
horizontal plane, for the hottest month of each city, which is input on the
unshaded, horizontal surfaces of the heat and mass transfer model.
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493484
-
7/30/2019 EFECTOS NATURACIN
6/14
the convective, conductive, evaporative and radiative heat
fluxes on the green and the concrete roofs differ
significantly. The convective heat flux density at the
external surface of the concrete roof is much larger than
the convective heat flux density in the upper part of the green
roofs canopy. For the concrete roof its 24-h profile ranges
from 345.1 to 128.6 W/m2, while for the green roofs upper
surface, it only ranges from 51.3 to 99.9W/m2. The
convective heat exchanges between the grass foliage and
the air are milder than those between the solid concreteroof and the air. The total radiative heat flux density (both
short and long wave radiation) on the external surface of
the concrete roof is also larger than that on the upper part
of the canopy layer. It ranges from 158.2 to 355.1 W/m2
on the concrete roof and from 38.8 to 229.5 W/m2 on the
green roof. Due to the redistribution of radiation within
the vegetated layer, the total radiative heat exchanges are
smaller on the vegetated surface, when compared with the
concrete roof. As can be observed in Eq. (7), the conductive
heat component is omitted in the relationship governing
heat transfer in plants as too small [1315,21,22], while it is
an important factor in the heat transfer of a concrete roof,
with a range from 444.5 to 154.5 W/m2 on the external
part of the roof. Nonetheless, the greatest differences are
observed at the evaporative heat fluxes, which range from
46.3 to 170.6 W/m2 for the concrete roof and from
593.2 to 26.4 W/m2 for the green roof. As the
evaporative heat transfer on the green roof acts constantly
as a heat sink and the radiative energy absorbed by the
green roof is smaller than that absorbed by the concrete
roof, the energy fluxes on a green surface can only offer
lower surface and air temperatures, when compared to
those produced by concrete surfaces.
Because of these energy distributions, canyon air
temperature lowers the most when both walls and roofsare covered with vegetation in all climates examined. This
can be explained by the fact that when roofs are covered
with vegetation, air masses enter the canyon much cooler,
from the vegetated roofs. On the other hand, when only
walls are covered with vegetation, air masses enter the
canyon heated by the plain roofs, which absorb the quite
ARTICLE IN PRESS
0
10
20
30
40
50
60
70
80
90
1 11 13 15 17 19 21 23
RelativeH
umidity(%)
Athens Beijing Brasilia
Hong Kong London Montreal
Moscow Mumbai Riyadh
100
Time (Hours)
Relative Humidity
3 5 7 9
Fig. 4. Twenty four-hour profile of relative humidity for the hottest
month of each city, which is input at the boundary nodes of the heat and
mass transfer model.
0
1
2
3
4
5
6
7
1 11 13 15 17 19 21 23
Athens Beijing Brasilia
Hong Kong London Montreal
Moscow Mumbai Riyadh
Wind Speed
WindSpeed(m/s)
3 5 7 9
Time (Hours)
Fig. 5. Twenty four-hour profile of wind speed for the hottest month of
each city, which is input at the boundary nodes of the heat and mass
transfer model.
01 7 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
HeatFlux
Density(W/m2)
Conv, rf-gr
Evap, rf-gr
Rad, rf-gr
Cond, rf-con
Conv, rf-con
Evap, rf-con
Rad, rf-con
400
300
200
100
-100
-200
-300
-400
-500
-600
Time (Hours)
2 3 4 5 6 8
Fig. 6. Convective (Conv), evaporative (Evap), long and short-wave
radiative (Rad) and conductive (Cond) heat fluxes on a concrete roof (rf-
con) and on a green roof (rf-gr) in Montre al. -12
-10
-8
-6
-4
-2
0
4 12 16 20 24
TemperatureDecrease(C)
Athens
Beijing
Brasilia
HongKong
London
Montreal
Moscow
Mumbai
Riyadh
Time (Hours)
Decrease of canyon air temperature,
green-all case
8
Fig. 7. Air canyon temperature decrease in the EW, H5W10 canyon, with
parallel wind flow, when both roofs and walls are covered with vegetation,
for all climates examined.
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493 485
-
7/30/2019 EFECTOS NATURACIN
7/14
great amounts of summer insolation. For this reason, in the
instance of Hong Kong, canyon air temperature decrease
reaches 8.4 1C maximum and 6.9 1C daytime average in the
green-all case (Fig. 7), while for the green-wall case these
numbers become only 3.9 and 2.5 1C, respectively (Fig. 8).
In general, the temperature decrease is quite significant
for both surface and air temperatures both inside thecanyon and at roof level. Regarding surface temperature
decrease of the south-oriented wall, it reaches from 18.7 1C
maximum and 14.3 1C daytime average for Riyadh to 9.8
and 5.6 1C, respectively, for Moscow (Fig. 9). Roof surface
temperatures lower even more, due to the greatest amounts
of solar radiation horizontal surfaces receive in summer;
the greatest day-time average temperature decrease is noted
for Riyadh (12.8 1C) and the greatest maximum for
Mumbai (26.1 1C), while the smallest decreases are noted
for Moscow and London (Fig. 10). Moscow reaches the
smallest daytime average surface temperature decrease
(9.1 1C), while London the smallest maximum (19.3 1C). In
the subject of air temperature decrease inside the canyon
for the green-all case, it reaches its peak for Riyadh
(11.3 1C maximum and 9.1 1C daytime average), while its
smallest decreases are noted in Moscow (3.6 and 3.0 1C,
respectively) (Fig. 7). For the green-wall case, air tempera-ture decrease reaches its maximum again for Riyadh
(5.1 1C maximum and 3.4 1C daytime average) and its
lowest decreases for Moscow (2.6 and 1.7 1C, respectively)
(Fig. 8).
3.2. Indirect radiative cooling effects
On prima facie evidence, the air inside a canyon with
vegetated walls is reduced due to the evapotranspirational
rate from plants and the lower surface temperatures of
vegetated surfaces. The latter are responsible not only for
lowering the air temperature but also for lowering surface
temperatures of surfaces not covered with vegetation.
As the radiative heat exchanges between the urban
canyon surfaces have been modelled analytically, a
decrease is observed in the asphalt surface temperature
when walls are covered with vegetation. In Fig. 11 the
decrease of the asphalt surface temperature is presented for
the H5W10 canyon for all the climates examined. As can be
observed, the greatest decreases occur for hot and with
high solar radiation Riyadh, with a maximum decrease of
2.0 1C and a daytime average 1.3 1C. The lowest surface
asphalt temperature decreases take place in much colder
and with lower insolation Moscow (maximum 0.9 1C,
daytime average 0.6 1C). As the air temperature near the
ARTICLE IN PRESS
-6
-5
-4
-3
-2
-1
0
4 12 16 20 24Athens
Beijing
Brasilia
HongKong
London
Montreal
Moscow
Mumbai
Riyadh
Decrease of canyon air temperature,
green-wall case
TemperatureDecrease(C) 8
Time (Hours)
Fig. 8. Air canyon temperature decrease in the EW, H5W10 canyon, with
parallel wind flow, when only walls are covered with vegetation, for all
climates examined.
-10
-5
0
4 12 16 20 24
TemperatureDecrease(C) Athens
Beijing
Brasilia
HongKong
London
Montreal
Moscow
Mumbai
Riyadh
Decrease of surface temperature of
south-oriented wall
Time (Hours)
-15
-20
8
Fig. 9. Surface temperature decrease of the south-oriented wall, when
covered with vegetation, in the EW, H5W10 canyon, for all climates
examined.
-5
5
4 12 16 20 24
TemperatureDecrease(C) Athens
BeijingBrasiliaHongKongLondonMontreal
MoscowMumbaiRiyadh
Time (Hours)
Decrease of roof surface temperature
-35
-25
-15
8
Fig. 10. Roof surface temperature decrease when covered with vegetation,
for all climates examined.
4 12 16 20 24
TemperatureDecrease(C)
Athens
Beijing
Brasilia
HongKong
London
Montreal
Moscow
Mumbai
Riyadh
Decrease of asphalt surface temperature
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
Time (Hours)
8
Fig. 11. Asphalt temperature decrease when walls are vegetated in the
H5W10 canyon for all the nine climates examined.
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493486
-
7/30/2019 EFECTOS NATURACIN
8/14
ground is primarily affected by the horizontal surface
temperature [14,15,21], the effect of vertical surface
temperatures is not as crucial as that of the horizontal
surface. This radiative cooling of the street asphalt by green
walls, reaching up to 2.0 1C in hot Riyadh has an
additional effect on lowering air temperatures, apart from
evapotranspirational and convective cooling effects.
3.3. Climatic characteristics
It can be said with certainty that, the hotter and drier a
climate is, the more important the effect of green walls and
green roofs on mitigating urban temperatures is (Figs. 7
and 8). As can be observed in Figs. 3 and 4 and Table 3,
Riyadh is the hottest and most arid of all the cases
examined, with urban temperatures reaching 42.81C
maximum and 31.2 1C minimum, with a daytime average
of 38.3 1C, while relative humidity spans from 45.8% to
23.5%, with a daytime average of only 32.4%. These
extreme climatic inputs, benefit the most from green walls
and green roofs and, as has been mentioned in Section 3.1,
the green-all case reaches temperature decreases of the
magnitude of 11.31C maximum and 9.1 1C daytime
average and the green-walls case 5.1 and 3.41C, respec-
tively. Much more humid Mumbai (Fig. 4) reaches smaller
decreases, of the magnitude of 6.6 1C daytime average and
8.0 1C maximum for the green-all case and 2.7 and 4.4 1C,
respectively, for the green-walls case. The colder climates
of London, Moscow and Montre al benefit the least,
reaching daytime average decreases from 1.7 to 2.11C
and maxima from 2.6 to 3.2 1C for the green-walls case and
from 3.0 to 3.8 1C and from 3.6 to 4.5 1C, respectively, forthe green-all case.
3.4. Roof versus canyon
Temperatures at roof level decrease more than inside the
canyon, when the building envelope is covered with
vegetation. This is because the roof, being more exposed
to the much larger amounts of summer solar radiation on
the horizontal plane, raises its temperatures even more
when plain, low albedo-building materials are exposed to
direct solar gains. However, the canyon, due to its
geometry, is generally more shaded, not reaching the peak
temperatures roof surfaces do. By covering the roof with a
vegetated medium, which regulates its temperature so as
not exceed some crucial levels, roof temperatures decrease
more than the temperature inside the canyon, when both
roofs and walls are covered with vegetation. For all the
nine climates examined, the maximum temperature de-
crease at the air layer 1 m above the roof reaches from
26.0 1C for Riyadh to 15.51C for London and daytime
average temperatures from 12.8 1C for Riyadh to 5.8 1C for
Moscow (Fig. 12). The air inside the canyon reaches lower
decreases; for the green-all case the air temperature
decrease reaches a maximum from 11.3 1C for Riyadh to
3.6 1C for Moscow and daytime average from 9.1 1C for
Riyadh to 3.0 1C for Moscow (Fig. 7). However, tempera-
tures at roof level start falling after 12:00, while for the
more stable conditions inside the canyon, temperatures due
to vegetation on walls start decreasing from early in the
morning, as can be observed by comparing Figs. 7 and 12.
3.5. Canyon orientation
Canyon orientation determines the shading pattern on
both the horizontal and the vertical parts of the canyon
geometry. It determines the amount of insolation received,
especially for the vertical planes, depending on their
orientation. During the summer months examined, the
amount of irradiation received on vertical planes is much
smaller than the horizontal one, for all orientations. Thus,
the orientation, despite the fact that it plays an importantrole in temperature distributions in and around the canyon,
it does not affect temperature decreases so significantly
when vegetation covers its vertical surfaces and roofs. The
magnitude of the effect strongly depends on the geographic
latitude. The examples of Hong Kong (22.16N) and Athens
(37.59N) are discussed below.
For all the climates examined, it has been observed that
the amount and geometry of vegetation is more important
than the canyons orientation. In the instance of Hong
Kong, solar radiation in all vertical orientations is not so
high, reaching a maximum of only 185 W/m2 for the
south orientation and 427 W/m2 for the west orientation
(Fig. 13). The green-walls case of the EW and NS oriented
H5W10 canyons result in 2.4 and 2.0 1C daytime average
temperature decrease, respectively, inside the canyon, with
an only 0.4 1C difference between the two orientations. For
the maximum, this difference becomes 0.7 1C (3.81C
maximum temperature decrease for the EW oriented
canyon and 3.11C for NS one). For the green-all case,
the differences between the two orientations become even
smaller, reaching 0.2 1C for the daytime average (tempera-
ture decrease being 6.8 1C for EW and 6.6 1C for NS), and
0.0 1C for the maximum (maximum temperature decrease
being 8.5 1C for both EW and NS orientation). It can be
observed in Fig. 14 that the amount and geometry of
ARTICLE IN PRESS
-30
-25
-20
-15
-10
-5
0
5
4 12 16 20 24
Temperatu
reDecrease(C)
AthensBeijingBrasiliaHongKongLondonMontreal
MoscowMumbaiRiyadh
-35
8
Decrease of air temperature 1m above the roof
Time (Hours)
Fig. 12. Air temperature decrease 1 m above the roof, for all climates
examined.
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493 487
-
7/30/2019 EFECTOS NATURACIN
9/14
vegetation mostly affect the temperature decrease and not
so much the canyons orientation. The green-all case in the
EW-oriented canyon results in 5.4 1C higher for the
daytime average and 4.7 1C higher for the maximum
temperature decrease than the green-walls case. For the
NS orientation, these differences become 4.6 and 4.4 1C,
respectively.
However, in the example of Athens, the amount of
irradiation received in the east and west oriented vertical
planes is much larger than on the south and north
orientations and proportionally larger to those of Hong
Kongs. The maximum solar radiation received by the
south-oriented vertical plane is 374.0 W/m2, while for the
east-oriented plane reaches the magnitude of 616.7 W/m2
(Fig. 15). This has a direct effect on the way the canyon
orientation affects temperature decreases due to vegetated
surfaces. For the green-all case, the difference between the
temperature decrease of the air inside the canyon remains
small, of the magnitude of 0.1 1C for the daytime average
(temperature decrease being 5.6 1C for EW and 5.5 1C for
NS), and 0.2 1C for the maximum (temperature decrease
being 6.6 1C for EW and 6.8 1C for NS orientation). For
the green-walls case these differences become larger,
reaching 0.81C for the daytime average (temperature
decrease being 3.01C for EW and 2.2 1C for NS) and
1.2 1C for the maximum (temperature decrease being 4.5 1C
for EW and 3.3 1C for NS). Yet again, the difference
between the two amounts of vegetation (green-all and
green-walls) is more crucial than the difference between the
decreases of different orientations. The difference between
the temperature decrease of the green-all and the green-
walls cases of the EW oriented canyon reaches 2.6 1C for
the daytime average and 2.1 1C for the maximum. For the
NS orientation, these differences become larger, 3.2 and
3.4 1C, respectively (Fig. 16).
In general, it can be concluded that the orientation may
play a countable role in temperature decreases due to
vegetation, only when the amounts of solar radiation
received by the vertical planes differ significantly. Yet
again, concerning temperature decreases, the amount of
vegetation placed on buildings is more crucial than the
orientation of the canyon, with the green-all case, when
ARTICLE IN PRESS
0
100
200
300
400
500
600
700
800
SolarRadiation(W/m2)
Global rad, horiz.
Globalrad, south
Global rad, north
Global rad, east
Global rad, west
1.00
Time (Hours)
4.00 7.00 10.00 13.00 16.00 19.00 22.00
Fig. 13. Hong Kong global solar radiation on horizontal and vertical
planes of east, west, south and north orientation in July.
0
5
10
15
20
25
30
35
12 15 18 21
Temperature(C)
-9
-8
-7
-6
-5
-4
-3
-2
-1
0 Ta,can[no gr]EW-HgKg
Ta,can [no gr]NS-HgKg
Ta,can[gr a]EW-HgKg
Ta,can[gr a]NS-HgKg
Ta,can[gr w]EW-HgKg
Ta,can[gr w]NS-HgKg
DTa,can[gr a]EW-HgKg
DTa,can[gr a]NS-HgKg
DTa,can[gr w]EW-HgKg
DTa,can[gr w]NS-HgKg
Time (Hours)
TemperatureDecrease(C)
Fig. 14. Hong Kong temperature distributions and decreases inside the
canyon for no-green [no-gr], green-all [gr-a], green-walls [gr-w], for EW
and NS oriented H5W10 canyon.
0
100
200
300
400
500
600
700
800
900
1 10 13 16 19 22
SolarRadiation(W/m2)
Global rad, horiz.
Global rad, south
Global rad, north
Global rad, east
Global rad, west
4 7
Time (Hours)
Fig. 15. Athens global solar radiation on horizontal and vertical planes of
east, west, south and north orientation in July.
0
5
10
15
20
25
30
35
40
12 15 18 21
-8
-7
-6
-5
-4
-3
-2
-1
0 Ta,can[no gr]Ath-EW
Ta,can[no gr]Ath-NS
Ta,can[gr a]Ath-EW
Ta,can[gr a]Ath-NS
Ta,can[gr w]Ath-EW
Ta,can[gr w]Ath-NS
DTa,can[gr a]Ath-EW
DTa,can[gr a]Ath-NS
DTa,can[gr w]Ath-EW
DTa,can[gr w]Ath-NS
Temperature(C)
TemperatureDecrease(C)
Time (Hours)
Fig. 16. Athens temperature distributions and decreases inside the canyon
for no-green [no-gr], green-all [gr-a], green-walls [gr-w], for EW- and NS-
oriented H5W10 canyon.
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493488
-
7/30/2019 EFECTOS NATURACIN
10/14
both roofs and walls are covered with vegetation, leading
to much larger temperature decreases.
3.6. Canyon geometry
For all geometries and cases examined, it can be
concluded that the wider a canyon is the smaller the effectof green roofs and green walls on its temperature decreases.
For wider canyons, temperatures inside the canyon are
dominated by the proportionally larger street surface and
the fact that it is more exposed to direct solar radiation. In
Riyadh, temperature decreases in the wide H5W15 canyon
are of the magnitude of 1.2 1C for the daytime average and
1.7 1C for the maximum of the green-wall case and 7.3 and
9.3 1C, respectively, for the green-all case. For the narrower
H10W5 canyon these decreases reach 6.3 1C daytime
average and 9.1 1C maximum for the green-walls case and
8.9 and 12.3 1C, respectively, for the green-all case (Figs. 17
and 18). It can be observed in Fig. 18 that the wider a
canyon is the smaller the effect of green walls is on its
temperatures. Nevertheless, when the combination of green
roofs and green walls is implemented (green-all case),
temperature decreases rise significantly. In contrast, for the
narrow H10W5 canyon, whose walls are proportionally
more dominant than the street, the green-walls case has a
significant effect on lowering urban temperatures. The
combination of both green roofs and green walls does not
lead to such significant further decreases, as is the case in
the wider H5W10 and H5W15 canyons. The differences
between the temperature decreases of the green-all and the
green-walls case reaches an average of 5.7 and 6.1 1C,
respectively, for the H5W10 and the H5W15 canyons,while for the H10W5 canyon it is only 2.6 1C. For the
maximum, the discrepancy between the two wider and the
narrower canyon is even larger, reaching 6.2, 7.6 and
3.2 1C, respectively.
3.7. Wind direction
Wind direction affects temperature decreases inside the
canyon even less than orientation. Although it is a
significant factor for temperature distributions, for the
decreases due to vegetation it is the vegetation itself that
plays the most important role. The differences betweentemperature decreases in the same canyons for different
wind directions are insignificant. In Fig. 19 temperature
distributions and temperature decreases for the EW-
oriented H5W10 canyon in Mumbai are presented, for
both parallel (y) and perpendicular (x) to the canyons
axis wind directions. It can be observed, that for the low air
velocities inside the canyon, temperature distributions are
not so different for the two wind directions as they were for
the canyon orientations (Figs. 14 and 16). Temperature
differences between the temperature decreases of the two
wind directions become quite insignificant, reaching a
0.1 1C difference for both green-walls and green-all cases. It
can also be observed that for both wind directions the most
important factor for temperature decreases is the amount
of vegetation, with the green-all case reaching temperature
decreases 3.7 1C higher than the green-walls case.
For the wider H5W15 canyon, with its much larger air
velocities, temperature decreases are similar for the two
wind directions (Fig. 20). The differences between the
temperature decreases of the two wind directions reach a
maximum of 0.3 1C, with 0.2 1C daytime average for the
ARTICLE IN PRESS
-5
0
7 11 15 19
TemperatureDecrease(C)
DTa [gr a]H5W10
DTa [gr a]H10W5
DTa [gr a]H5W15
Air temperature decrease for different canyon
geometries in Riyadh for the green-all case
Time (Hours)
-15
-10
Fig. 17. Air temperature decrease during the day in the H5W10, H10W5
and H5W15 canyon, for the green-all case, Riyadh.
-8
-6
-4
-2
0
7 11 15 19
TemperatureDecrease(C)
DTa[gr w]H5W10
DTa[gr w]H10W5
DTa[gr w]H5W15
Time (Hours)
-10
Air temperature decrease for different canyon
geometries in Riyadh for the green-wall case
Fig. 18. Air temperature decrease during the day in the H5W10, H10W5
and H5W15 canyon, for the green-wall case, Riyadh.
0
5
10
15
20
25
30
35
40
12 15 18 21
Temperature(C)
-9
-8
-7
-6
-5
-4
-3
-2
-1
0 Ta,can[no gr]Mumb-EW-x
Ta,can[no gr]-Mumb-EW-y
Ta,can[gr a]Mumb-EW-x
Ta,can[gr a]-Mumb-EW-y
Ta,can[gr w]Mumb-EW-x
Ta,can[gr w]-Mumb-EW-y
DTa,can[gr a]Mumb-EW-x
DTa,can[gr a]-Mumb-EW-y
DTa,can[gr w]Mumb-EW-x
DTa,can[gr w]-Mumb-EW-y
TemperatureDecrease(
C)
Time (Hours)
Fig. 19. Mumbai temperature distributions and decreases inside the
canyon for no-green [no-gr], green-all [gr-a], green-walls [gr-w], for a
parallel (y) and perpendicular (x) to the canyons axis wind direction in
the EW-oriented H5W10 canyon.
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493 489
-
7/30/2019 EFECTOS NATURACIN
11/14
green-wall case and 0.4 and 0.3 1C, respectively, for the
green-all case. Again, the amount of vegetation has a
stronger effect than the wind flow direction, even in the
wider canyon.
It can thus be concluded, that for the generally low air
velocities inside the urban canyons [6], the effect of wind
direction is not so strong on temperature decreases due to
vegetated roofs and walls, as is the amount and geometry
of vegetation itself.
4. Thermal comfort
In order to assess the thermal comfort improvements in
outdoors spaces when walls and roofs are covered with
vegetation, the physiological equivalent temperature (PET)
is used, its expression deriving from Ref. [23] and its
relationship with thermal sense from Ref. [24]. The results
for the EW-oriented H5W10 green-all and no-green cases
are presented here, for Moscow, Athens and Riyadh in
Figs. 2123. Emphasis is given on thermal comfort, not
only inside the canyon (symbolised with EW in the graphs),
but also at the roof level (symbolised with rf).
It can be observed in Fig. 21, that for the much milder
summer of Moscow, the greening of the building envelope
does not lead to such major improvements of the outdoors
thermal comfort. PET ranging from slightly warm and
comfortable levels on the roof and inside the canyon for
the no-green case, lowers to cooler levels, from comfortable
to slightly cool, during daytime, when roofs and walls are
covered with vegetation (green-all case). Although moving
from slightly warm to comfortable might not be so
spectacular, it could prove to be beneficial for the thermal
comfort and well being of populations used to cooler
climatic conditions.
For much hotter Athens (Fig. 22) and Riyadh (Fig. 23),
the improvements of outdoors thermal comfort are more
dramatic. For both climates, the bare concrete roof reaches
the very hot level in the afternoon. When covered with
vegetation, the sensation warm is reached only for 4 h in
Athens and 5 in Riyadh. Most of the daytime, the exposed
to direct solar radiation roof reaches the slightly warm
and comfortable zone for both cities. For inside the
ARTICLE IN PRESS
0
510
15
20
25
30
35
40
45
12 15 18 21
-8
-7-6
-5
-4
-3
-2
-1
0
1 Ta,can[no gr]MumbH5W15-EW-x
Ta,can[no gr]-MumbH5W15-EW-y
Ta,can[gr a]MumbH5W15-EW-x
Ta,can[gr a]-MumbH5W15-EW-y
Ta,can[gr w]MumbH5W15-EW-x
Ta,can[gr w]-MumbH5W15-EW-y
DTa,can[gr a]MumbH5W15-EW-x
DTa,can[gr a]-MumbH5W15-EW-yTemperatureDecrease(C)
T
emperature(C)
Time (Hours)
Fig. 20. Mumbai temperature distributions and decreases inside the
canyon for no-green [no-gr], green-all [gr-a], green-walls [gr-w], for a
parallel (y) and perpendicular (x) to the canyons axis wind direction in
the EW-oriented H5W15 canyon.
Fig. 21. PET for the EW-oriented H5W10 canyon, for the no-green [no
gr] and green-all [gr-a] cases, inside the canyon (EW) and on the roof (rf),for Moscow.
Fig. 22. PET for the EW-oriented H5W10 canyon, for the no-green [no
gr] and green-all [gr-a] cases, inside the canyon (EW) and on the roof (rf),
for Athens.
Fig. 23. PET for the EW-oriented H5W10 canyon, for the no-green [no
gr] and green-all [gr-a] cases, inside the canyon (EW) and on the roof (rf),
for Riyadh.
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493490
-
7/30/2019 EFECTOS NATURACIN
12/14
canyon, the thermal sensation improves even more, from
hot, in both cases to slightly warm and comforta-
ble, reaching even slightly cool for both cases in the
early morning and late evening hours and even cool for
Athens, in the early morning hours.
In general, green roofs and walls can improve outdoors
thermal conditions not only at street level, but also at rooflevel, turning these empty urban spaces into potentially
usable ones, in the form of superterrestrial gardens. By
covering roofs and walls with vegetation, thermal comfort
in the built environment can improve significantly, not only
for hot climates, but for cooler ones, in which populations
are acclimatised to lower temperatures.
5. Energy savings from green walls and roofs
Apart from creating outdoor conditions, which are more
human-friendly, from a thermal point of view, green
roofs and green walls can also prove beneficial for indoorthermal conditions. In addition to the fact that they add a
further insulation layer to the buildings fabric, they can
decrease cooling load demands inside the building quite
significantly due to the microclimatic modifications dis-
cussed in this paper.
In a simplified steady-state analysis, without taking into
consideration internal thermal gains, heat gains/losses (qE)
from the buildings fabric with an average U-value U, an
indoors temperature Tin and an outdoors temperature Toutare given by the relationship:
qE UTout Tin. (9)
For the no-green base case [no gr], the cooling load for
the non-vegetated canyon is given by the relationship:
qEno gr U Tno gr Tin
, (10)
where T[no gr] is the averaged air temperature inside the
canyon when no vegetation is placed either on walls or on
roofs. For the green-all case, with an average air
temperature inside the canyon T[gr a], heat gains are1:
qEgr a U Tgr a Tin
. (11)
Thus, the decrease in the cooling load, when both walls
and roofs are covered with green is given by
DqEgr a qno gr qgr a
qno gr) ,
DqEgr a Tno gr Tgr a
Tno gr Tin. 12
For T[no gr]6Tin, T[no gr]4Tin and T[gr a]4Tin.
Similarly, for the green-walls case, if T[gr w] is the
average air temperature inside the canyon with the green
walls, the cooling load decrease becomes:
DqEgr w Tno gr Tgr w
Tno gr Tin. (13)
For T[no gr]6Tin, T[no gr]4Tin and T[gr w]4Tin.
Considering an indoor limit temperature for cooling of
23 1C for all climates studied, the cooling load decreases
due to green-all and green-walls cases are given as a
daytime average in Fig. 24 and for an hourly basis in
Fig. 25.
As can be observed in Fig. 24, the largest cooling load
decreases in all climates examined, occur for the green-all
case. For the geometries examined for Braslia and Hong
Kong, the cooling load decreases for the green-all casereach 100%; no cooling load is needed after covering roofs
and walls with vegetation, while in both cities cooling load
is needed in the afternoon and early evening hours for the
no-green case (Fig. 25c and d). London and Moscow are
not affected, regarding cooling loads, as no cooling load is
needed for the typical day examined, even before vegeta-
tion was placed around the canyon. Riyadh experiences a
quite high cooling load decrease, of the magnitude of 90%,
as does Montre al (85%) for the green-all case, lowering
their total hours of cooling demand from 12 to 5 and from
8 to 4, respectively (Fig. 25e and g). For the green-all case,
Mumbai reaches a 72% decrease, lowering its cooling
energy demand from 11 h to 6 (Fig. 25f), while for Athens
and Beijing the decrease is 66% and 64%, respectively,
lowering their energy demand by 4 and 3 h, respectively
(Fig. 25a and b).
For the green-walls cases, cooling load decreases are less
dramatic. The largest one is noted for Braslia (68%), with
6 h decrease in cooling demand (Fig. 25d). It is followed by
a 66% and 2 h decrease for Hong Kong (Fig. 25e), 52%
and 2 h for Montre al (Fig. 25e), 43% and 2 h for Athens
(Fig. 25a), 37% and 2 h for Beijing (Fig. 25b), 37% and 3 h
for Riyadh (Fig. 25(g) and 35% and 3 h for Mumbai
(Fig. 25f). It can be noted that the differences between the
green-all and green-walls cooling loads are smaller for
ARTICLE IN PRESS
Fig. 24. Average cooling load decreases (%), with a 231C indoors
temperature, for the green-all and green-walls cases of all the climates
examined.
1Despite the fact that the U-value is altered, when vegetation is placed
on the buildings fabric, leading to further cooling load decreases, this is
not taken into consideration here, as the aim is to directly compare
between the effects of the microclimatic alterations on the buildings
cooling load, without it being affected by alterations to the fabric.
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493 491
-
7/30/2019 EFECTOS NATURACIN
13/14
humid climates (of the magnitude of 3237%) and greater
for arid climates (53% for Riyadh), due to the different
humidity concentrations in the two climatic groups.
In general, green roofs and green walls cool the
microclimate around them, which can lead to quite
important energy savings for cooling, depending on the
climatic type, the amount and position of vegetation on the
building. In cases where little cooling load is needed,
cooling demand can be reduced to zero by covering
building surfaces with vegetation. In other cases, energy
savings can also be significant, varying from 90% to 35%.2
In addition to the energy savings themselves, this could
lead to successful applications of further passive cooling
techniques, especially ones employing ventilation, which
are not easy to implement in the extremely hot urban
conditions, in cases of large heat island densities.
6. Conclusions
From this quantitative research, it has been shown that
there is an important potential of lowering urban
temperatures when the building envelope is covered with
vegetation. Air temperature decreases at roof level can
reach up to 26.0 1C maximum and 12.81C day-time
average (Riyadh), while inside the canyon decreases reach
up to 11.3 1C maximum and 9.1 1C daytime average, again
for hot and arid Riyadh. It can be concluded that the
ARTICLE IN PRESS
Fig. 25. Cooling load decreases (%) for (a) Athens, (b) Beijing, (c) Hong Kong, (d) Braslia, (e) Montre al, (f) Mumbai and (g) Riyadh for green-all and
green-walls cases.
2These percentages can become even greater, when a higher than 23 1C
limit temperature for cooling is considered. In general, inhabitants of hot
climates are accustomed to higher temperatures (in the instance of Greek
regulations, the limit temperature for cooling is set to 26 1C).
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493492
-
7/30/2019 EFECTOS NATURACIN
14/14
hotter and drier a climate is, the greater the effect of
vegetation on urban temperatures. However, it has been
pointed out that also humid climates can benefit from
green surfaces, especially when both walls and roofs are
covered with vegetation, reaching up to 8.4 1C maximum
temperature decrease for humid Hong Kong. Temperature
decrease due to vegetation is primarily affected by thevegetation itself (amount and geometry), more than the
canyon orientation in hot periods. In general, the larger
amounts of solar radiation a surface receives, the larger its
temperature decreases are when it is covered with vegeta-
tion. For the low air velocities inside the canyon, the wind
direction does not have any significant effect on tempera-
ture decreases due to vegetation.
Regarding the urban geometry, the wider a canyon is,
the weaker the effect green roofs and green walls have on
temperature decrease. For all climates examined, green
walls have a stronger effect than green roofs inside the
canyon. Nonetheless, green roofs have a greater effect at
roof level and, consequently, at the urban scale. The
combination of both green roofs and green walls leads to
the highest mitigation of temperatures inside the canyon. If
applied to only one unit block, green roofs and green walls
can create a small area of mitigated temperatures to the
urban heat island effect, as has been shown in this
microclimatic study. If applied to the whole city scale,
they could mitigate raised urban temperatures, and,
especially for hot climates, bring temperatures down to
more human-friendly levels and achieve energy saving
for cooling buildings from 32% to 100%.
Acknowledgements
This research has been funded by the State Scholarship
Foundation of Greece (IKY) from 2001 to 2003. The
authors are extremely grateful to Panagiotis Doussis for his
guidance and contribution to computer modelling.
References
[1] Benevolo L. The history of the city. London: Scholar Press; 1980.
[2] Koppe C, Kovats S, Jendritzky G, Menne B. Health and global
environmental change; heat-waves: risks and responses, series no. 2,
energy, environment and sustainable development. Copenhagen:
World Health Organization; 2004.[3] White KS, et al. Technical summary, climate change 2001: impacts,
adaptation and vulnerability. In: McCarthy JJ, Canziani OF, Leary
NA, Dokken DJ, White KS, editors. Climate change 2001: impacts,
adaptation and vulnerability. Cambridge: United Nations Inter-
governmental Panel on Climate Change (IPCC) Cambridge Uni-
versity Press; 2001. p. 1973.
[4] Daz J, Ballester F, Lopez-Velez R. Impacts on human health. In:
Moreno Rodrguez JM, editor. The preliminary assessment of the
impacts in Spain due to effects of climate change, project ECCE.
Madrid: Ministry of the Environment; 2005.
[5] Davies RE, Knappenberger PC, Novicoff WM, Michaels PJ. Decadalchanges in summer mortality in US cities. International Journal of
Biometeorology 2003;47:16675.
[6] Santamouris M, editor. Energy and climate in the urban built
environment. London: James & James; 2001.
[7] Dimoudi A, Nikolopoulou M. Vegetation in the urban environment:
microclimatic analysis and benefits, PRECis project, the European
Commission, Directorate General XII, Joule III, contract JOR3-
CT97-0192. Centre for Renewable Energy Sources, Pikermion,
2000.
[8] Bruse M, Fleer H. Simulating surfaceplantair interactions inside
urban environments with a three dimensional numerical model.
Environmental Modelling and Software 1998;13:37384.
[9] Giridharam R, Ganasan S, Lau SSY. Daytime urban heat island in
high-rise and high-density residential developments in Hong Kong.
Energy and Buildings 2004;36:52534.[10] Versteeg HK, Malalasekera W. An introduction to computational
fluid dynamics. Harlow, London: Prentice-Hall; 1995.
[11] Luikov AV. Heat and mass transfer in capillary-porous bodies.
Oxford: Pergamon Press; 1961.
[12] Eckert ERG, Drake RM. Heat and mass transfer. New York:
McGraw-Hill Book Company Inc.; 1959.
[13] Noilhan J. Les facteurs physiques du microclimat au voisinage dun
batiment. Etude bibliographique. Centre Scientifique et Tecnique du
Batiment, Etablissement de Nantes, 1979.
[14] Jacobson MZ. Fundamentals of atmospheric modeling. Cambridge:
Cambridge University Press; 1999.
[15] Pielke RA. Mesoscale meteorological modeling. San Diego: Aca-
demic Press; 2002.
[16] Modest MF. Radiative heat transfer. New York: McGraw-Hill
International Editions; 1993.[17] Remund J, Lang R, Kunz S, Meteonorm, version 3.0, Meteotest,
Bern, 1997.
[18] Marsh A. Ecotect, version 5.20(b), square one research PTY LTD,
2003.
[19] Welsh School of Architecture, Cardiff University, WinAir4. Cardiff,
2003.
[20] Critchfield HJ. General climatology. London: Prentice-Hall Inter-
natioal Inc.; 1974.
[21] Geiger R. The climate near the ground. Cambridge, Massachusetts:
Harvard University Press; 1965.
[22] Monteith JL, Unsworth M. Principles of environmental physics.
London, New York: Edward Arnold; 1990.
[23] Matzarakis A, Mayer H, Rutz F. Radiation and thermal comfort.
Sixth Hellenic conference in meteorology, climatology and atmo-
spheric physics, Ioannina 2; 2002. p. 73944.[24] Kuttler W. Stadklima, Teil 2: pha nomene und wirkungen. In: Mo ller
D. (Ed.), Beitragsserie klimaa nderung und klimaschutz, UWSF,
vol. 16(4); 2004. p. 26374.
ARTICLE IN PRESS
E. Alexandri, P. Jones / Building and Environment 43 (2008) 480493 493