INTERNATIONAL JOURNAL OF CLIMATOLOGY

Int. J. Climatol. 30: 44–57 (2010)

Published online 12 March 2009 in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/joc.1869

Microclimate modelling of street tree species effects within

the varied urban morphology in the Mediterranean city
of Tel Aviv, Israel
Limor Shashua-Bar,a * Oded Potchter,b Arieh Bitan,c Dalia Boltanskyd and Yaron Yaakove
aFaculty of Architecture and Town Planning, Technion-Israel Institute of Technology, Haifa 32000, Israel
bUnit for Applied Climatology and Environmental Aspects, Department of Geography and the Human Environment, Tel Aviv University, Tel
Aviv 39040, Israel
c Unit for Applied Climatology and Environmental Aspects, Department of Geography and the Human Environment, Tel Aviv University, Tel Aviv
39040, Israel
d Dalia Boltansky, Boltansky Architects Ltd, Tel-Aviv, Israel
e Unit for Applied Climatology and Environmental Aspects, Department of Geography and the Human Environment, Tel Aviv University, Tel Aviv
39040, Israel

ABSTRACT: Microclimate formation and its significance in urban planning was examined through two components that
predominate in affecting the city’s climate: built-up morphology and urban shade trees.
The methodological approach focus is on a generalization procedure for quantifying the thermal effect of any studied
situation through parameterization of the vegetated variables and the built-up forms. The analysis is integrative, using
empirical climatic data followed by an analytical study for generalization and sensitivity analysis using an integrative
model, the Green CTTC model.
Three urban tree species predominant in the Tel Aviv gardens and streets, with different canopy characteristics, and three
levels of building densities were analysed to determine their thermal effect on an urban street microclimate. The variables
were parameterized according to six basic cooling attributes for the studied tree species in urban gardens in Tel Aviv, and
according to three geometric built-up parameters for the studied urban street.
The integrative modelling approach of considering all changes simultaneously was illustrated on an urban boulevard in
Tel Aviv. The analysis demonstrates the shortcomings of piecemeal modelling and the merits of the integrative approach.
The study indicates the importance of urban trees in alleviating the heat island effect in a hot and humid summer. The
tree cooling effect was found to be strongly related to the built form geometry. In all the studied cases, the thermal effect
of the tree was found to depend mainly on its canopy coverage level and planting density in the urban street and little on
other species characteristics.
The methodology of analysis presented in this paper can be applied to develop an operational tool in assessing for an
urban open space the integrative thermal effects of different tree species, and of the varied urban morphology and the
interaction between them. Copyright  2009 Royal Meteorological Society
KEY WORDS urban shade trees; Mediterranean climate; vegetation cooling effect; microclimate modelling; Green CTTC
model
Received 12 October 2007; Revised 11 January 2009; Accepted 12 January 2009

1. Introduction which is the climate in the streets between the buildings

Design strategies for sustainable passive cooling in urban and beneath the roof level (Oke, 1987), as these spaces
planning for mitigating heat island intensity are becoming cover more than two thirds of the urban area of a
important in terms of energy savings and improved modern city (Shashua-Bar and Hoffman, 2004). Among
human thermal comfort (Steemers, 2003; Yezioro et al., the variables to be modelled are the vegetation coverage
2006; Grimmond, 2007). This mitigation is important (Saito et al., 1990/1; Grimmond et al., 1996; Jonsson,
from the scientific aspect by means of reducing the 2004), the thermophysical properties of the built-up
urban warming and heat stress, and from the policy elements (Rosenfeld et al., 1995; Akbari et al., 2003), the
aspect by reducing costs of energy consumption. The geometry of the studied open space (Oke, 1988; Swaid
focus of this study is on the microclimate modelling and Hoffman, 1990/1; Mills, 1997) and the surrounding
of open spaces within the Urban Canopy Layer (UCL) building forms (Meir et al., 1995; Ratti et al., 2003).
The quantitative analysis followed in this paper is
through the use of parameterization of the vegetated
* Correspondence to: Limor Shashua-Bar, Faculty of Architecture and
Town Planning, Technion-Israel Institute of Technology, Haifa 32000, variables and the built-up forms. Three urban tree species
Israel. E-mail: [email protected] are studied in open spaces; they predominate in the

Copyright  2009 Royal Meteorological Society

 MICROCLIMATE MODELLING OF STREET TREE SPECIES 45

Tel Aviv gardens and streets and differ in their canopy load received from surface surroundings (Kjelgren and
characteristics and coverage levels. The effects of these Montague, 1998; Kuttler et al., 2007).
trees in summer can then be applied to urban sites In the context of large green areas as compared to
with various built-up configurations. The typology of built-up surroundings, the impact of PCI of well-treed
the studied building forms is parameterized through a parks can reach a maximum of 4–5 ° C during day time,
generic model for urban streets. The interaction thermal while at night the PCI is small and reaches up to
effects of vegetation with the urban building density are 1 ° C (Sham, 1991; Saito et al., 1990/1; Ca et al., 1998;
estimated through simulations using the analytical Green Potchter et al., 1999). It has also been shown that the
CTTC (Cluster Thermal Time Constant) model (Shashua- park cooling effect extends beyond the park. The range
Bar and Hoffman, 2002) for predicting the diurnal air effect was found to depend mainly on the park size and on
temperature variations in a given UCL site. the distance of the built-up surroundings from the park
border (Honjo and Takakura, 1990/1; Jauregui, 1990/1;
1.1. Background Spronken-Smith, 1994; Ca et al., 1998; Upmanis et al.,
1998).
Design strategies for sustainable urban planning are In microscale cases, the study of vegetation effects on
important in passive cooling and in finding sustainable the UCL microclimate is quite recent. Besides the prop-
energy sources for mitigating the urban heat island inten- erties of the vegetation species, three other main factors
sity and heat stress. A potentially powerful means for affect the microclimate of an open space: (1) The geom-
passive cooling in the urban context is vegetation, par- etry of the open space obtained as the aspect of the sur-
ticularly shade trees. One of the strategies for urban rounding buildings’ height to the width of the street (Oke,
heat island mitigation is the provision of urban parks 1987), (2) The thermophysical properties are the ther-
in order to enhance evaporation (Grimmond, 2007). In mal inertia of the built materials within the urban open
a study conducted in Colombo, Sri Lanka (Emmanuel space, and (3) The anthropogenic heat release, mainly
and Johansson, 2006), shading is proposed as the main the transportation and fuel consumption for domestic use
strategy for reducing air and radiant temperatures, espe- in the flanking buildings of the open spaces. These fac-
cially achieved by deep canyons, covered walkways and tors were found to affect the vegetation effect and thus
shade trees. The measurement results showed a maxi- introduce an interaction effect (Shashua-Bar et al., 2006).
mum difference of 7 ° C, while temperature differences In addition to the climatic variables, the effect of the
between sunlit and shaded urban surfaces reached up to above three factors and of vegetation depends on vari-
20 ° C. Recent studies demonstrate the importance of mod- ous variables which are under the control of the designer
eling the UCL in urban canyon streets; It is shown that (herein named as control variables). The overall effect
the interaction between buildings material, their location of the various control variables within an urban space
and geometry play an important role in the street heat can be significant in hot climatic regions in terms of
exchange which influence its climate conditions (Erell energy savings and improved human comfort. In tem-
and Williamson, 2006; Offerele et al., 2007; Emmanuel perate and hot regions, the maximum thermal effects in
et al., 2007) and on thermal comfort (Al-Toudert and summertime of these variables on the UCL microclimate
Mayer 2006, Pearlmutter et al., 2007) within the street. is of the order of 3–4 ° C for vegetation in urban parks
The cooling effect of urban parks creates a Park Cool (e.g. Oke, 1989; Taha et al., 1991; Potchter et al., 1999),
Island (PCI), and is most pronounced in urban envi- about 1 ° C for wall albedo modification, (Shashua-Bar
ronments which have high urban heat island intensi- and Hoffman, 2004) and up to 7 ° C for differences of
ties (Spronken-Smith and Oke, 1998). Studies in hot building forms (Shashua-Bar et al., 2006). These ther-
humid cities indicate that urban trees in parks and streets mal impacts are significant, taking into account that
can generate cool islands leading to a large PCI effect the average daily range of the summer air temperature
(e.g. Bernatzky, 1982; McPherson and Dougherty, 1989; in temperate climates, as found in the Mediterranean
Oke, 1989; Dimoudi and Nikolopoulou, 2003). Besides region, are of the order of 6–10 ° C (Bitan and Rubin,
the positive effect of the green areas on the urban cli- 1994). According to Akbari et al. (2001), urban trees in
mate, several empirical results (e.g. Jauregui, 1990/1; streets can have a substantial cooling effect on the urban
Grimmond et al., 1996; Potchter et al., 2006) indicate air temperature, and can reduce cooling energy demand
that in some cases, during the daytime hours, the veg- by 20%.
etated areas can be warmer than the surrounding built-up
environment, leading to unpleasant microclimatic con- 1.2. Objectives
ditions. It appears that the cooling effect of vegetation The study examines the factors affecting the microclimate
may differ from site to site, depending on the vegeta- and its significance in urban planning through two
tion type and coverage, the surface temperatures, and the components that dominantly affect the city’s climate:
built-up surroundings. This interrelation between vegeta- built-up density and urban shade trees. The study seeks to
tion coverage and its surroundings is most noticeable in develop an operational tool for assessing the integrative
the context of an urban open space, such as the urban thermal effects of different tree species and the site’s
street. As opposed to the urban park, the trees’ effect is built-up elements, and the interaction between them in
local and is greatly influenced by the increased energy an urban open space.

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
46 L. SHASHUA-BAR ET AL.

2. Methodology geometry, thermophysical properties, anthropogenic fac-

2.1. Methodological approach tors, and the vegetation in a given urban open space. In
this study, two main analysed control variables are the
The methodological approach in this study is focused on urban space geometry and urban trees and their interre-
generalization. The generalization procedure for quantify- lationship effects on microclimate. This interrelationship
ing the thermal effect of any studied situation is achieved effect is analysed through modelling using parameteriza-
here through parameterization of the variables according tion of the main attributes of each of the two impacts:
to basic measurable attributes. This process is applied for for the urban space geometry, three built-up ratios, and
two main control variables namely, the built-up density for the urban trees, six attributes that capture the thermal
and the urban shade trees. The tree thermal characteristics effect of tree cooling.
are parameterized by attributes namely, the tree cov- The paper’s sections are as follows: Section 2 describes
erage, canopy thermal characteristics including canopy the studied sites and the measurement setup. Section 3
solar absorptivity, canopy solar transmissivity, canopy reports on the experimental study and the empirical find-
thermal time constant, and the tree cooling coefficient ings. The analytical study is described in Section 4, which
due to evapotranspiration, and to changes in the tree’s includes a short description of the model formulation,
heat storage. The built forms are parameterized by three its suitability for simulation purposes, and its validity
geometric relations (aspect ratio, building depth ratio, in predicting air temperature variations for the studied
and spacing ratio) representing the various building form sites. The section also discusses the estimated value of
configurations. the tree’s cooling effect and its dependence on the urban
The control variables studied in this paper in mod- built-up density. Section 5 illustrates the shortcomings
elling the UCL microclimate are multivariate in nature. of piecemeal modelling and argues for the merits of the
The vegetation variable covers many tree species, each integrative approach.
having different cooling potential due to its properties.
Likewise, the built forms cover different types of open
2.1. Sites and observations
spaces in urban streets, pavilions, and courtyard hous-
ings. The multiplicity of the factors calls for developing The study considers the case of a hot humid climate
a generalization procedure in order to reach generality in summer. It was conducted in the city of Tel Aviv
and meaningful quantitative results. Thus, online anal- in Israel (32° 06
N 34° 47
E), situated on a plain along
yses for each individual type are not only cumbersome, the east coast of the Mediterranean Sea. The city has a
but will hinder decision making in identifying a preferred subtropical climate, Csa according to Koeppen classifica-
choice among possible alternatives. tion (Potchter and Saaroni, 1998), its climatic conditions
Besides parameterization, the analysis is integrative. are a result of the combined effect of the geographical
It starts by an empirical study to describe the actual location, the influences of the Mediterranean Sea, and of
situation followed by an analytical study for generaliza- urbanization. During the summer season, the city suffers
tion and sensitivity analysis. The quantification of ther- from hot and humid climatic conditions: The daily max-
mal effects for each attribute and built form aspect, the imum temperature is around 29.0–30.0 ° C on average,
total thermal effect of the major variables of vegetation with minimum relative humidity around 61%. The daily
and building density, as well as the thermal interaction minimum temperature is around 22.2 ° C, with average
between them are estimated using a well-validated ana- relative humidity around 83%. Heat stress in summer is
lytical model, the Green CTTC model. noticeable during most of the daytime hours. The wind
Figure 1 shows the modelling procedure followed in regime is influenced by the Mediterranean Sea breeze
this study in quantifying the effects of the system’s con- which blows during the day from the west and northwest,
trol variables. The figure shows a set of variables affect- whereas during the night, it is affected by a light land
ing the microclimate and to which the designer can assign breeze, which blows from the east and southeast (Bitan
appropriate values. The level of these variables can be and Rubin, 1994). Significant heat island intensities were
controlled; hence, they are named in the paper as control recorded in Tel Aviv at midday and nighttime during the
variables. The control variables consist of urban space summer (Bitan et al., 1992) and winter (Saaroni et al.,

Modeling
Set of Control Variables affecting the microclimate
Procedure

Urban Space Urban Space Interrelationship Urban

Features Geometry effects Trees

Parameterization Geometric ratios Tree thermal characteristics

Figure 1. The urban open space microclimate modelling procedure.

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
 MICROCLIMATE MODELLING OF STREET TREE SPECIES 47

2000). The results indicate maximum warming up to 5 ° C simultaneously in the middle of the boulevard and the
from air temperature measurements between the city cen- street canyon on a traffic island. Air temperature and
tre and its surrounding areas indicating the existence of relative humidity were measured at 1.5 m above ground,
an urban heat island. From remotely acquired thermal and wind velocity and direction at 2 m height. Figure 2
readings, differences up to 10 ° C in surface temperature shows the map of the studied sites’ locations.
were recorded indicating the existence of an surface heat The instruments included Campbell HMP45C-type
island due to differences in urban texture, vegetated areas, temperature and relative humidity sensors. The sensors
and anthropogenic activities. were installed inside a Gill Multi-Plate 41 002 unaspirated
Table I lists the data of measurements and the tree radiation shield. The temperature and relative humidity
coverage level for each site. Three tree species with accuracy of the sensors is ±0.2 ° C and ±2%, respectively.
different canopy characteristics and canopy coverage Wind velocity and direction were measured using Young
levels were studied: The Ficus retusa tree (common name: 05 103 anemographs. Sensor accuracy was ±0.3 m/s for
Indian Laurel Fig) with a highly dense and broad canopy the velocity, and 3° for the direction. Readings were
(about 14 m in diameter), the Tipuana tipu tree (common taken every second with the resulting data averaged
name: tipu tree) with a moderate size canopy (about 10 m and stored every five minutes using a Campbell 21X
in diameter), and the Phoenix Dactylifera tree (common data logger. For wind direction, the mode direction was
name: Date Palm) with a sparse and narrow canopy stored. All the instruments were calibrated, checked,
(about 6 m in diameter). These three species of shade and compared before and after the experiment under
trees are dominant in the Tel Aviv gardens and streets. the same conditions. According to specifications of the
Some basic characteristics of the three tree species are screen manufacturer (Young Co, 1998), the screen may
shown in Table II. overestimate the air temperatures during the daytime
All three species were studied from measurements hours (up to 1.5 ° C under stable calm conditions, and
in situ in different urban gardens. The Ficus tree was less than 0.7 ° C and 0.4 ° C with winds of 2 and 3 m/s,
studied in Tagore Garden in the Ramat-Aviv area. The respectively). In order to avoid erroneous conclusions
Tipuana tipu was studied in the Reading Garden in the stemming from the screen effect, temperatures of stations
Ramat-Aviv area. The third tree species, the Date Palm exposed to different wind velocities were compared to
tree (planted over grass), was studied in a garden situated those of a station sheltered from the wind. Temperature
in the Tel Aviv University area. The distance between any differences were less than 0.2 ° C, indicating that the error
two gardens is less than 1 km. effect due to the small ventilation in the exposed screens
In addition to the garden sites, an urban boule- in the park was negligible. It should also be noted that
vard at the centre of Tel Aviv city was also stud- during most hours during the day, wind velocity was
ied, namely, Chen Boulevard, with Ficus trees planted between 1.5 and 3 m/s, so that the screens had natural
along it. The measurements were conducted in the ventilation.
boulevard and in an urban street, Iben Gvirol Street In addition, mobile instruments were also used for
parallel to it, which has no trees. Both sites are evaluating some of the site parameters, such as the solar
north–south oriented, having similar building density, absorptivity and canopy solar transmissivity factors, and
with an aspect ratio of 0.4, and both carry heavy vehicular the tree coverage levels. The experimental study for all
traffic.
the sites concerns the summer situation. Besides the
onsite measurements, climatic diurnal data corresponding
2.3. Measurement setup to the measurement days, on solar radiation, wind veloc-
Campbell automatic meteorological stations were set up ity, ground surface and air temperatures, and humidity,
in the centre of each urban garden, representing the three were obtained from a meteorological station at Beit-
studied tree species. In the case of the urban streets, Dagan, located in a rural environment at a distance of
Campbell automatic meteorological stations were set up about 7 km southeast of Tel Aviv. The data were used

Table I. The studied sites.

Site Date of measurements Site area (m2 ) Bordering buildings height (m) Tree Species Tree coverage (%)

Urban Garden
Reading Garden 31/7–1/08/2001 35 000 12 Tipuana tipu 70
Tagore Garden 15–16/08/2005 60 000 6–12 Ficus retusa 90
Tel Aviv University 15–16/08/2005 5000 6–12 Date Palm∗ 50
Urban Boulevard
Chen Boulevard 03–06/06/1997 3000∗∗ 12 Ficus retusa 70
Iben Gvirol Street 03–06/06/1997 3000∗∗ 12 – –

∗ The trees are planted over grass.

∗∗ According to width of 30 m and measured length of 100 m.

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
 48

Table II. Basic characteristics of the studied tree species.

Botanical name Ficus retusa Tipuana tipu Phoenix dactylifera

Family Moraceae Fabaceae Arecaceae (Palmae)

Common name Indian Laurel Fig tipu tree Date Palm

Copyright  2009 Royal Meteorological Society

Height at maturity 10–15 m 6–15 m 10–20 m
Canopy width Vertically growing branches, 15–20 m in Up to 10–15 m. Broad silhouette with flattened Large feather palm, 6 m diameter.
diameter. Suited for formal shearing. May be crown. Usually wider than higher.
pruned at any time of year to desired shade and
size.
Trunk Central smooth trunk, with growth of air roots Central trunk branching. Needs support as young Slender trunk, 30–40 cm in diameter. Trunk
which can develop into sub-branches. tree especially in windy areas. patterned with base of old leafstalks. Suckers from
base, natural state is clump of several trunks.
Leaves at top only.
Foliage Dense foliage. Leaves are clear lustrous green, Leaves divided into 11–21 oblong; 3 cm long. Grey-green waxy leaves. 3 m long. Leaflet stiff
similar in size. Light green leaflets. and sharply pointed at top.
L. SHASHUA-BAR ET AL.

DOI: 10.1002/joc
Int. J. Climatol. 30: 44–57 (2010)
 MICROCLIMATE MODELLING OF STREET TREE SPECIES 49

D1

D2

KM
A- Tagor Garden, B- TA campus garden, C- Reading Garden, D- Urban Boulevard
(where: D1- Chen Blvd, D2-Iben-Gvirol street)

Figure 2. Map of the studied sites locations. This figure is available in colour online at www.interscience.wiley.com/ijoc

for comparison purposes, such as the air temperature and Table III. Daily air temperature values (° C) measured at the
for running the analytical model. studied sites, summer data, Tel Aviv area, Israel.

Site Time (hour)

3. Empirical findings 6 : 00 12 : 00 15 : 00 24 : 00

Table III summarizes the sites’ air temperature values for Urban Garden
times 6 : 00, 12 : 00, 15 : 00 and 24 : 00 for each urban Reading Garden 24.4 30.8 30.1 25.2
wooded sites (where 15 : 00 is equal to 14 : 10 solar time). (Tipuana tipu trees)
Table III provides for each site the corresponding values Meteorological station 24.4 30.1 31.1 24.6
at the Beit-Dagan Meteorological station (located in a (31/7–1/08/2001)
rural area). Note that the difference between the site and Tagore Garden (Ficus 24.0 29.3 29.4 26.0
the meteorological station values is not a correct measure retusa trees)
of the trees’ cooling effect, since it includes other thermal Meteorological station 23.6 32.3 32.3 24.5
effects due to differences in wind speed, humidity, ground (15–16/08/2005)
absorptivity, and urban factors. Tel Aviv University 24.9 29.8 30.2 26.3
The diurnal patterns of air temperature, wind speed, (Date Palm trees∗ )
and humidity are shown in Figure 3 for the three Meteorological station 23.6 32.3 32.3 24.5
urban gardens (Reading garden, Tagore garden and (15–16/08/2005)
the Tel Aviv University) and for the urban boule- Urban Boulevard
vard (Chen Boulevard) and urban street (Iben Gvirol Chen Boulevard 19.5 25.5 24.6 21.2
(Ficus retusa trees)
Street).
Iben Gvirol Street 21.5 27.3 27.7 23.4
The air temperature patterns in Table III indicate sig-
Meteorological station 19.4 27.4 26.7 19.4
nificant differences between the urban wooded sites and (03–06/06/1997)
the rural meteorological station at Beit-Dagan, and also
∗ The trees are planted over grass.
Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
50 L. SHASHUA-BAR ET AL.

Urban Gardens
33 6 33 6
Vapour Pressure [mmHg]

Vapour Pressure [mmHg]

31 31
5 5

Wind Speed [m/sec]

Wind Speed [m/sec]

Air Temperature [C]

Air Temperature [C]

29 29
27 4 27 4
25 25
23 3 23 3
21 21
19 2 19 2
17 1 17 1
15 15
13 0 13 0
4 8 12 16 20 24 4 4 8 12 16 20 24 4
Time [hr] Time [hr]
Reading garden (Typu Tipuana trees) Tagor garden (Ficus Retusa trees)

33 6
Vapour Pressure [mmHg]

31
5

Wind Speed [m/sec]

Air Temperature [C]

29
27 4
25
23 3
21
19 2
17 1
15
13 0
4 8 12 16 20 24 4
Time [hr]
TA university (Date Palm trees)

Urban Boulevard
33 6 33 6
31 31
Vapour Pressure [mmHg]

Vapour Pressure [mmHg]

Wind Speed [m/sec]

5 5
Wind Speed [m/sec]
Air Temperature [C]

Air Temperature [C]

29 29
27 4 27 4
25 25
23 3 23 3
21 21
19 2 19 2
17 1 17 1
15 15
13 0 13 0
4 8 12 16 20 24 4 4 8 12 16 20 24 4
Time [hr] Time [hr]
Chen Boulevard (Ficus Retusa trees) Iben-Gvirol Street

Air temperature at site Vapour pressure at site Wind speed at site

Figure 3. Daily data at the studied sites. Summer data, Tel Aviv area, Israel.

noticeable differences among the sites. The variability The quantitative thermal impact on the air temperature
among the hourly air temperature values is due to dif- variability for each variable is discussed in the analytical
ferent days of measurements and also to different levels analysis.
of the variables affecting the site’s microclimate. Large
differences between the two urban streets were recorded,
due to differences in the traffic load and to the fact that 4. The analytical study
Chen Boulevard includes dense broad trees while Iben As shown in previous studies (Shashua-Bar and Hoffman
Gvirol has none. 2004; Shashua-Bar et al., 2006), the cooling effect does

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
 MICROCLIMATE MODELLING OF STREET TREE SPECIES 51

not depend solely on the tree attributes, but also on the climatic variables and the effect of control variables
the site’s specific conditions such as its geometry, the but refers to thermal radiation exchange.
building density of its surrounding buildings, its wind The term,
TAHR (t), is the anthropogenic effect to air
regime, etc. Hence, an interrelationship among the effects temperature. It includes the effects of man-made nature
of all the variables affecting the site’s microclimate. such as heat release owing to transportation and fuel
The effect of each variable and its interaction with the consumption for domestic use. It is estimated in Wm−2
urban texture can be studied through simulations using and is added to the absorbed solar radiation at time t.
the analytical Green CTTC model. The thermal effect
of different trees and urban forms is incorporated in 4.2. Parameterization
the analytical model through parameterization of their The basic characteristics of the three tree species in
characteristic attributes. Table II provide little information as regards the cooling
attributes of the trees, as these basic characteristics
4.1. The Green CTTC model–outline are specific for each species and state of the tree
The Green CTTC model for estimating the diurnal urban development, and thus, are too generally stated for
air temperature pattern is an extension of the CTTC analytic purposes. In this study, the attributes affecting
model (Swaid and Hoffman, 1990/1). The predicted air the cooling ability are common for any tree species. Six
temperature of the site is calculated through the contribu- such attributes are chosen as input in the analytical model:
tion of the heat received from external sources, mainly the
net solar radiation, anthropogenic heat release, and vege- • Tree coverage: The tree coverage is the difference
tation effects. It comprises design variables and parame- between the values of two attributes: sunny spots under
ters related directly to the physical structure and proper- the canopy area and partial tree-shaded area at noon.
ties of the built-up complex (mainly, the thermophysical It is a measure of the canopy size net of sunny spots,
properties of the surfaces and the open space geometry) relative to the site’s ground area.
and to the density and canopy characteristics of shade • Canopy solar absorptivity: A function of the leaf
trees. The levels of these factors can be controlled by the properties, the canopy structure, the canopy solar
designer, hence, as such, the model is suitable for mod- transmittance, and the soil solar reflection (Monteith
elling purposes. Specifically, it can enable the designer and Unsworth, 1995, p 88).
to predict the thermal effect of any change in the level of • Canopy solar transmissivity: Average canopy solar
one or more of its control variables. The model does not transmittance under the tree canopy.
take into consideration advection effects, thus, its pre- • CTTC: Thermal time constant of the tree canopy.
dictions are valid only for calm days, and therefore, not Its reciprocal determines the heating rate of the tree
valid on rainy days, stormy days or advection of hot dry canopy; thus, the higher the time constant, the lower
winds. It should be noticed that the effect of UHI is more the heating rate.
pronounced in calm conditions. • Cooling coefficient of tree: That part of solar energy
The formulation of the Green CTTC model is given diverted from heating the site to changing of the tree’s
in (Shashua-Bar and Hoffman, 2002; Shashua-Bar et al., heat storage and evapotranspiration.
2006). For the convenience of the reader the predicting
The levels and parameters for the three studied tree
equation for the cluster’s air temperature at time (t) is
species are listed in Table IV. The parameter values of the
given here:
trees’ attributes were estimated by the authors based on
measurements conducted by the Department of Geogra-
T (t) = T0 +
TSOLAR (t) −
TNLWR (t)
phy and Human Environment at Tel Aviv University. The
+
TAHR (t) (1) estimation of the two parameter values – thermal time
constant and cooling coefficient – follows the methods
T (t), is the predicted air temperature at time (t). T0 , expressed in (Shashua-Bar and Hoffman, 2002).
is the regional base (or background) temperature that Table IV indicates differences in the levels and in the
was found to be equal to the mean daily air temperature parameter values of the six attributes as estimated in the
measured at a rural meteorological station, representative studied sites, which follows the canopy characteristics
of the non-built-up area near the site in question. of the three tree species. The significant differences

TSOLAR (t) is a step function of the penetrated short- occur in the tree coverage (net of sunny spots), canopy
wave radiation, using the reciprocal of the CTTC param- solar transmissivity, and the cooling coefficient. The
eter as the attenuating factor. It includes the vegetation two attributes, the canopy absorptivity and the canopy
effect, expressed as a reduction in the incident solar radi- thermal time constant, as average estimates for a tree
ation which is diverted by the trees for the purpose of canopy, range from about 0.6 to 0.7, and 8 to 9 hours,
evapotranspiration and for changes in the plant’s heat respectively, showing small differences among the three
storage. tree species.
The term,
TNLWR (t), is the contribution of the net The Ficus tree has a large and dense canopy which
outgoing longwave radiation exchange flux-to-air cool- is expressed in a small area of sunny spots (5–10%)
ing. Like the absorbed solar radiation term, it includes and a small amount of solar transmittance under its

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
52 L. SHASHUA-BAR ET AL.

Table IV. Levels and parameter values of the six attributes of the studied tree species. Summer data, Tel Aviv area, Israel.

Site Levels Parameters

Sunny spots Tree coverage Canopy solar Canopy solar Canopy time(1) Cooling coeff.
(%) (%) absorptivity transmissivity constant of tree

Tagore Garden 5–10 90 0.60 0.05 9.0 0.60

(Ficus retusa)
Reading Garden 15–20 50 0.60 0.12 7.9 0.50
(Tipuana tipu)
Tel Aviv University 30–40 50(2) 0.70 0.15 7.6 0.70(3)
(Date Palm)
Chen Boulevard 5–10 70 0.60 0.04 8.0 0.40
(Ficus retusa)

(1) Thermal time constant for ground = 8 h, for walls = 6 h.

(2) The grass coverage level (under the Date Palm trees) was 0.80
(3) Cooling coefficient for Date Palm = 0.40, for grass = 0.30

canopy (an average of 5% transmittance). Its massive without trees, was found to be 0.3, leading to a cooling
canopy produces relatively large tree coverage in any coefficient of 0.4 for the Date Palm trees.
open space, and accordingly, leads to a high cooling The urban site chosen for a typological study is a
coefficient of 0.6 in the Tagore Garden due to high heat typical residential street 27 m wide, commonly found
storage and evaporative energy which reduce the solar in the centre of Tel Aviv. The urban space factor
energy received by the open space by about 60%. Due considered in this study is characterized by its built-up
to changes in the soil properties and the presence of geometry. For a residential street, following the generic
surrounding built surfaces, the cooling coefficient was model for urban streets (Shashua-Bar et al., 2006), the
found to be lower in the case of the Chen Boulevard built-up geometry is parameterized in this study by three
(0.4). geometric relations: the aspect ratio H /W (height to
The Tipuana tipu tree is characterized by a moderate width), the built-up depth ratio D/L2 (building depth,
canopy having more sunny spots (15–20%) and higher D, to the frontal length, L2), and the spacing ratio,
L1/L2 (distance between adjacent buildings L1 to the
solar transmissivity (average of 12%) compared to the
frontal length, L2). For all the studied configurations,
Ficus tree. Its cooling coefficient was found to be 0.5 in
W = 27 m, L1 = 10 m, L2 = 20 m and D = 25 m.
the Reading Garden, diverting 50% of the received solar
For the studied street of 27 m width with two rows
energy from heating the site.
planted, the maximum tree coverage (net of sunny spots)
The Date Palm tree characteristics are a sparse and for the Ficus retusa tree is found to be up to 90% of
narrow canopy, with a considerable percentage of sunny the ground area; for the Tipuana tipu tree, 70%; and
spots (30–40%) and high solar transmissivity (an average for the Date Palm tree, 25% only. These three coverage
of 15%). As opposed to the Ficus and tipu trees, the levels were considered for analysis in this study. Figure 4
Date Palm trees can be planted short distances apart in demonstrates the three geometric ratios of the studied
an open space, reaching to a maximum canopy coverage street and the planting scheme of the maximum coverage
of 50% in the Tel Aviv University site. The Date Palm of each of the tree species.
trees in the studied site were planted over grass. This
combination led to the highest cooling coefficient of 0.7. 4.3. Simulating the air temperature patterns of the sites
The cooling coefficient of the same grass species, located The air temperature patterns of the sites were simulated
in the vicinity of the Date Palm site, but fully exposed and using the Green CTTC model. The purpose was to test

Generic model of urban Ficus Retusa Tree Tipuana Tipu Tree Date Palm Tree
street 90% coverage 70% coverage 25% coverage
STREET

L1
L2 H

D W

Figure 4. The studied street with maximum coverage of the studied tree species.

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
 MICROCLIMATE MODELLING OF STREET TREE SPECIES 53

Table V. Daily air temperature values measured at the studied Table VI. Simulated trees cooling effects at the studied sites,
sites vs simulated (° C). (° C).

Site Time (hour) Tree species Tree Time (hour)

coverage
6 : 00 12 : 00 15 : 00 24 : 00 (%) 6 : 00 12 : 00 15 : 00 24 : 00

Urban Garden Ficus retusa Tree 90 −2.1 −2.5 −3.3 −2.4

Reading Garden measured 24.4 30.1 30.1 25.2 (Tagore Garden)
(Tipuana tipu trees) simulated 24.5 30.2 30.0 24.6 Tipuana tipu Tree 50 −1.0 −2.0 −2.8 −1.8
Tagore Garden measured 24.0 29.3 29.4 26.0 (Reading Garden)
(Ficus retusa trees) simulated 24.7 29.2 29.1 26.6 Date Palm Tree 50 −1.7 −1.9 −2.0 −2.1
(Tel Aviv Univ.)
Tel Aviv University measured 24.9 29.8 30.2 26.3
Ficus retusa Tree 70 −0.9 −3.0 −3.8 −1.7
(Date Palm trees) simulated 24.2 29.4 30.1 25.9
(Chen Boulevard)
Urban Boulevard
Chen Boulevard measured 19.5 25.5 24.6 21.2
(Ficus retusa trees) simulated 20.3 24.6 23.7 19.8
4.5. The impact of urban geometry on the cooling
Iben Gvirol Street measured 21.5 27.3 27.7 23.4 effect of trees
(without trees) simulated 20.9 26.2 25.8 21.2
The impact of urban geometry on the cooling effect
of trees was studied on a typical urban street 27 m
the validity of the model and the validity of the parameter wide (wall to wall). Among the three geometric ratios
estimates. In all the validity tests, anthropogenic heat was characterizing the generic built form, the aspect ratio
not taken into account. (height to width ratio, H /W ) was found to have the
The results of the simulations are shown in Table V for largest impact. To illustrate the urban impact, only the
the times 6 : 00, 12 : 00, 15 : 00, and 24 : 00. The diurnal aspect ratio impact was considered in this study.
patterns are shown in Figure 5 for the wooded sites and Three levels of aspect ratios were considered in the
for the urban street, Iben Gvirol, with no trees. The simulations: 0.6 (= 16 m/27 m), 0.8 (= 22 m/27 m),
patterns of the three park gardens show a very close fit, and 1.0 (= 28 m/27 m) corresponding, for this kind of
indicating a high level of validity. street, to buildings with 4, 6, and 8 floors, respectively.
In the two urban streets, the deviations are systemati- The other two geometric ratios, the depth and spacing
cally positive and large along the whole range of the day ratios, for the 27-m-wide street were held constant
pattern. The deviations are about an average of 1 ° C for at the levels of 1.25 and 0.5 respectively in all the
Chen Boulevard, and 1.5 ° C for Iben Gvirol Street indi- simulations. The simulations were run for a common
cate the anthropogenic effect of vehicular traffic in these coverage level of the three tree species, as determined
two streets. The estimate for the Chen Boulevard during by the smallest maximum coverage level allowed for
the noon hours is about 2000 passenger cars per hour, each tree species. In the 27-m-wide street, taking into
and that of Iben Gvirol about 3000, consistent with the account the canopy diameter and the planting distance
deviation values of air temperatures. The simulated esti- of the trees, the maximum coverage level (net of sunny
mates for the traffic load in the rush hours are 1.5 ° C for spots) was found to be 90% for the Ficus trees, 70%
the Chen Boulevard and 2.5 ° C for the Iben Gvirol Street. for the tipu trees, and 40% for the Date Palm trees
planted in 3 rows, and only 25% for those in 2 rows.
The maximum common coverage level allowed for
4.4. Estimating the cooling effect of the trees
comparison purposes is then 40%, as restricted by the
Estimating the cooling effect as the difference in air canopy size relative to the street’s width. The Date Palm
temperature between a park garden with trees and a trees considered in the simulations are not planted over
nearby site without trees, yields, in most cases, mislead- grass. Their cooling coefficient is estimated to be about
ing results. As mentioned earlier, urban factors play an 0.4.
important role in the UCL air formation. In addition, the Table VII shows the cooling effects (at 15 : 00 hours)
presence of trees affects the levels of some of the cli- for the three tree species, all calculated at the same
matic variables such as wind velocity and humidity. In coverage of 40%, net of sunny spots, which is the
addition, the albedo of trees and of the ground and walls maximum allowed coverage for the Date Palm trees in
are not the same. a street of 27 m wide (planted in three rows). In this
In this study, the cooling effect of the trees is estimated manner, the figures in Table VII can be compared to
as the difference between the site’s simulated values with the cooling ability of the other attributes taken together.
trees to those without trees. Table VI shows the cooling For all the three tree species, the effect of increasing
effect of the trees at times 6 : 00, 12 : 00, 15 : 00, and the aspect ratio (in this study, by increasing the wall
24 : 00. The differences among the sites are due mainly height of the open space) results in reducing the tree
to relatively large differences in their net tree coverage cooling effect. Going across the rows, from 4 to 6
levels. building floors, the cooling effect is reduced for all

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
54 L. SHASHUA-BAR ET AL.

Urban Gardens

32 32

Air Temperature [C]

Air Temperature [C]

30 30
28 28
26 26
24 24
22 22
20 20
4 8 12 16 20 24 4 4 8 12 16 20 24 4
Time [hr] Time [hr]
Reading Garden, Typu Tipuana trees Tagor Garden, Ficus Retusa trees
31.7-1.8/1997 15-16/8/2005

32
Air Temperature [C]

30
28
26
24
22
20
4 8 12 16 20 24 4
Time [hr]
TA University garden, Data Palm trees
15-16/8/2005

Urban Boulevard

30 30
Air Temperature [C]

Air Temperature [C]

28 28
26 26
24 24
22 22
20 20
18 18
4 8 12 16 20 24 4 4 8 12 16 20 24 4
Time [hr] Time [hr]
Chen Boulevard, Ficus Resuta trees Iben gvirol street
3-6/6/1997 3-6/6/1997

Measured values at studied site Simulated values at studied site

Figure 5. Diurnal air temperature measured at the studied sites vs simulated values (° C).

tree species by 14% (1.32/1.52); going from 4 to 8 planted with the trees compared to the street without
building floors, the cooling effect is reduced by 22% trees. The potential cooling effect for each tree species
(1.19/1.52). Going down the columns in the Table, we is that calculated at its maximum coverage level of 90,
find the effect of the three tree species: percentage-wise, 70, and 25% for the Ficus, tipu, and Date Palm trees
the Ficus has 67% (1.52/0.91) more potential than the (all planted in 2 rows), respectively. In the simulations,
Date Palm tree, and only 30% more than the tipu tree, the cooling potential of the trees reaches its maximum
and this is found to be so for all the three studied levels at 15 : 00 hours. In contrast to the first group of figures
of H /W . in Table VII, where the coverage level is held constant
The potential of the cooling effect in summer of at 40%, the second group of figures in Table show the
the three urban tree species in the studied street at practical side of selecting the tree species. In this respect,
15 : 00 hours, are shown also in Table VII; the diurnal the Ficus and the tipu trees are found to be effective as
pattern is shown in Figure 6. The trees’ cooling effect is cooling agents in urban streets, while the Date Palm tree
estimated by the difference between the studied street is found to be relatively ineffective.

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
 MICROCLIMATE MODELLING OF STREET TREE SPECIES 55

Table VII. The simulated cooling effect of the three tree species in an urban street (15 : 00 hours), July data, Coastal Mediterranean
region, case study: an urban street, 27 m width.

Term of simulation Species Tree Trees potential cooling effect at 15 : 00 [° C]

coverage
[%] H/W = 0.6 H/W = 0.8 H/W = 1.0

At similar tree coverage Ficus Retusa Tree 40 −1.5 −1.3 −1.2

Tipuana Tipu Tree 40 −1.2 −1.0 −0.9
Date Palm Tree 40 −0.9 −0.8 −0.7
At maximum tree coverage Ficus Retusa Tree 90 −3.4 −3.0 −2.7
Tipuana Tipu Tree 70 −2.1 −1.8 −1.6
Date Palm Tree 25 −0.5 −0.4 −0.4

H H H

W W W
H/W = 0.6 H/W = 0.8 H/W = 1.0

33 33 33
32 32 32
Air Temperature [C]

Air Temperature [C]

Air Temperature [C]

31 31 31
30 30 30
29 29 29
28 28 28
27 27 27
26 26 26
25 25 25
24 24 24
4 8 12 16 20 24 4 4 8 12 16 20 24 4 4 8 12 16 20 24 4
Time [hr] Time [hr] Time [hr]

Max Min Max Min Max Min

90% 90% 90%
3.4 °C 0.8 °C 3.0 °C 0.7 °C 2.7 °C 0.6 °C
Ficus Ficus Ficus
70% 70% 70%
2.1 °C 0.5 °C 1.8 °C 0.4 °C 1.6 °C 0.4 °C
Typu Typu Typu
25% 25% 25%
0.5 °C 0.1 °C 0.4 °C 0.1 °C 0.4 °C 0.1 °C
Palm Palm Palm

Street without trees Street with 25% Palm trees

Street with 70% Typu trees Street with 90% Ficus trees

Figure 6. The simulated diurnal cooling effect potential of the three tree species in an urban street, July data, Coastal Mediterranean region.

5. The integrative approach in modelling correctly formulating the system’s mechanism. This
approach is illustrated on the Chen Boulevard site. The
In Section 4.5, the impact of changing one control changes considered from Chen Boulevard’s base-levels
variable, the building density (or number of floors), on are:
the cooling effect of trees is illustrated. It was found that
at any level of tree coverage, the cooling effect is not
constant, but depends on the level of the other variables in
the system. The conclusion is that piecemeal modelling, (a) Raising the building height of Chen Boulevard’s open
usually the result reached by empirical analysis, would space from 4 floors, the base level, to 6 floors.
lead to incorrect conclusions as a result of ignoring the (b) Modifying the albedo of walls from 0.5, the base
interrelationship and the interaction effects among the level, to 0.7, that is, a lighter colour for the walls.
system control variables. (c) Reducing Chen Boulevard’s tree coverage from the
The integrative approach of considering all changes base level of 70 to 40%, for better ventilation but a
simultaneously can be applied using an analytical model warmer microclimate.

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
56 L. SHASHUA-BAR ET AL.

The results of the simulations of the piecemeal mod- findings were applied to a typical residential street, 27 m
elling and of the integrative approach are shown in wide, with buildings of 4, 6, and 8 floors.
Table VIII, in degrees centigrade, relative to the base Modelling effects and sensitivity analysis were studied
case (as in a, b, c above) of Chen Boulevard, for the by the piecemeal procedure (one variable modification at
times 6 : 00, 12 : 00, 15 : 00, and 24 : 00. In Table VIII, a time) and by the integrative procedure (a simultaneous
the piecemeal modelling is expressed as the sum of the change of all variables). The difference between the two
separate effects of each modified level (a + b + c), while sets of estimates indicates the interaction effect.
the integrative modelling is the total effect of the three Other analytical studies that simulated the vegetation
modified levels simultaneously, and the interaction effect effect within the urban tissue, found different results.
is the difference between these two estimates. According to Emmanuel et al. (2007), in a tropical
The figures in Table VIII indicate that for any hour climate, the effect of vegetation is less significant than
of the day, the total thermal effect reached by piecemeal the effect of the urban geometry and surfaces albedo,
modelling (row A) is smaller than that reached by the while Dimoudi and Nikolopoulou (2003) pointed out
integrative solution (row B). The bias is cumulative as that in a Mediterranean climate, vegetation can play an
the day goes on: from 16% at 6 : 00, 57% at 15 : 00, to important role depending on the urban geometry. This
a maximum of 129% at night. The bias is the result of study contributes to the significance of tree species as a
the interaction effects resulting from the interrelationship factor that must be considered in estimating their cooling
among the system’s control variables. The larger the effect within the urban varied morphology.
change of the variables’ levels, the greater the bias. The main findings are:

6. Summary and conclusions • The cooling effect of the trees in summer in a hot
humid climate is significant and can reach up to 3–4
The study examines the mechanism of microclimate degrees of cooling, which is about 50% of the air
and its significance in urban planning through two temperature rise from sunrise to noon hours.
components that dominantly affect the city’s microcli- • The potential cooling effect of the tree was found
mate – built-up density and urban shade trees. Both com- to depend mostly on its canopy coverage level and
ponents are multivariate in nature, hence, regularity is planting density in the urban street, and little on other
sought throughout the analysis in order to enable the species characteristics. This should be considered when
designer to draw general conclusions and guidelines. The selecting tree species for climatic sustainable planning.
problem is resolved through parameterization of the vari- • The Date Palm tree in a residential street (27 m wide)
ables – for vegetation, six basic cooling attributes are was found to be ineffective as a cooling agent.
used, and for the building density, three built-up geo- • A tree’s thermal effect is strongly related to the
metric ratios were found to fully cover most situations. urban street geometry. For any tree coverage level, the
Three tree species with different canopy characteristics cooling effect is not constant: the deeper the open space
and canopy coverage levels were studied: Ficus retusa, (high H /W ), the smaller the tree cooling effect.
Tipuana tipu, and Date Palm. These three species of shade • In urban streets with deep open spaces, the effect of
trees are dominant in Tel Aviv gardens and streets. The trees on the related microclimate is small. In such
cases, trees with small canopy size such as the canopy
Table VIII. Piecemeal vs integrative modelling, (° C). Case of the Date Palm, are more suitable than large ones,
study: Chen Boulevard. especially if the deep open space is that of a narrow
street width rather than of high flanking walls. Such
Simulated effects Time (hour) trees are mostly needed along sidewalks for the benefit
6 : 00 12 : 00 15 : 00 24 : 00 of pedestrians, mainly for shading and thermal comfort
in the noon hours.
Effect of building density 4∗ −0.08 −1.38 −1.60 −0.37 • In the case of urban streets with shallow open spaces,
floors to 6 floors a tree’s potential cooling effect can be enhanced by
Albedo modification from −0.18 −0.41 −0.73 −0.34 using broad-leaf trees (such as the Ficus and Tipuana
0.5∗ to 0.7
tipu tree species) in minimum planting intervals, so
Tree coverage reduction −0.05 +1.00 +1.31 +0.54
from 70%∗ to 40%
that canopies will overlap at maturity.
• Significant interaction thermal effects were found to
Sum of modeling effects:
exist between the vegetation effects and the built-up
Piece-meal effect (A) −0.31 −0.79 −1.02 −0.17
effects. Piecemeal modelling of the variables led to
Integrative effect (B) −0.36 −1.15 −1.60 −0.39
Interaction effect (B-A) −0.05 −0.36 −0.58 −0.22 thermal effects with a significant bias as compared to
Interaction effect in [%] 16 46 57 129 the integrative modelling. The integrative solution is
(B/A - 1) × 100 reached through the use of an analytical model. In
this study, the Green CTTC model is found to yield
∗ Values of the base case satisfactory results.

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc
 MICROCLIMATE MODELLING OF STREET TREE SPECIES 57

Acknowledgements urban street canyon. Boundary-Layer Meteorology 122: 273–292,

DOI 10.1007/s10546-006-9099-8.
The authors wish to thank to the Israel Ministry of Oke TR. 1987. Boundary Layer Climates. Methuen Press: London,
252–302.
Environmental Protection for financing this research Oke TR. 1988. Street design and urban canopy layer climate. Energy
(Grant no. 4-060). and Buildings 13: 103–111.
Oke TR. 1989. The micrometeorology of the urban forest. Journal of
Philosophical Transactions of the Royal Society of London B 324:
References 335–349.
Pearlmutter D, Berliner P, Shaviv E. 2007. Integrated modelling of
Akbari H, Pomerantz M, Taha H. 2001. Cool surfaces and shaded trees
pedestrian energy exchange and thermal comfort in urban street
to reduce energy use and improve air quality in urban areas. Solar
canyons. Building and Environment 42: 2396–2409.
Energy 70: 295–310.
Potchter O, Choen P, Bitan A. 2006. Climatic behavior of various
Ali-Toudert F, Mayer H. 2006. Numerical Study on the effects of
urban parks during hot and humid summer: Mediterranean city of Tel
aspect ratio and orientation on an urban street canyon on outdoor
Aviv, Israel. International Journal of Climatology 26: 1695–1711,
thermal comfort in hot and dry climate. Building and Environment
DOI 10.1002/joc.1330.
41: 94–108.
Potchter O, Saaroni H. 1998. An examination of the maps of climatic
Bernatzky A. 1982. The contribution of trees and green spaces to a
regions of Israel, according to the Koppen classification. Studies in
town climate. Energy and Buildings 5: 1–10.
the Geography of Israel 15: 179–194, (in Hebrew).
Bitan A, Noy L, Turk R. 1992. The Impact of the Sea Shore on the
Potchter O, Yaakov Y, Bitan A. 1999. Daily and seasonal climatic
Urban Climate of Tel Aviv. Scientific reports. Wissenschaftliche
behavior of a small urban park in a Mediterranean climate: A case
Berichte des Instituts für Meteorlogie und Klimaforschung der
study of Gan-Meir Park, Tel-Aviv, Israel. In Proceedings of the
Universität Karlsruhe Nr. 16, 147–160.
15 th International Congress of Biometeorological and International
Bitan A, Rubin S. 1994. Climatic Atlas of Israel for Physical and
Conference on Urban Climatology, Sydney.
Environmental Planning and Design. Ramot Publishing Company,
Ratti C, Raydan D, Steemers K. 2003. Building form and environmen-
Tel-Aviv University: Tel-Aviv, Israel.
tal performance: archetypes, analysis and an arid climate. Energy and
Ca VT, Asaeda T, Abu EM. 1998. Reduction in air conditioning energy
Buildings 35: 49–59.
caused by a nearby park. Energy and Buildings 29: 83–92.
Rosenfeld AH, Akbari H, Bretz S, Fishman BL, Kurn DM, Sailor D,
Dimoudi A, Nikolopoulou M. 2003. Vegetation in the urban
Taha H. 1995. Mitigation of urban heat islands: materials, utility
environment: microclimate analysis and benefits. Energy and
programs, updates. Energy and Buildings 22: 255–265.
Buildings 35: 69–76.
Saaroni H, Ben-Dor E, Bitan A, Potchter O. 2000. Spatial Distribution
Emmanuel R, Johansson E. 2006. Influence of urban morphology and
and microscale characteristics of the urban heat island in Tel-Aviv,
sea breeze on hot humid microclimate: the case of Colombo, Sri
Israel. Landscape and Urban Planning 48: 1–18.
Lanka. Climate Research 30: 189–200.
Saito I, Ishihara O, Katayama T. 1990/91. Study of the effect of green
Emmanuel R, Rosenlund H, Johansson E. 2007. Urban shading - a
areas on the thermal environment in an urban area. Energy and
design option for the tropics? A study in Colombo, Sri Lanka.
Buildings 15–16: 443–446.
International Journal of Climatology 27(14): 1995–2004, DOI:
Sham S. 1991. Urban climatology in Malaysia: a review. Energy and
10.1002/joc.1609.
Buildings 15–16: 105–117.
Erell E, Williamson T. 2006. Simulating air temperature in an urban
Shashua-Bar L, Hoffman ME. 2002. The Green CTTC model for
street canyon in all weather conditions using measured data
predicting the air temperature in small urban wooded sites. Building
from a reference meteorological station. International Journal of
and Environment 37: 1279–1288.
Climatology 26: 1671–1694.
Shashua-Bar L, Hoffman ME. 2004. Quantitative evaluation of passive
Grimmond CSB. 2007. Urbanization and global environmental change:
cooling of the UCL microclimate in hot regions in summer – case
local effects of urban warming. Geographical Journal 173: 83–88,
study: urban streets and courtyards with trees. Building and
DOI: 10.1111/j.1475–4959.2007.232 3.x.
Environment 39: 1087–1099.
Grimmond CSB, Souch C, Hubble MD. 1996. Influence of tree cover
Shashua-Bar L, Hoffman ME, Tzamir Y. 2006. Integrated ther-
on summertime surface energy balance fluxes, San Gabriel Valley,
mal effects of generic built forms and vegetation on the
Los Angeles. Climate Research 6: 45–57.
UCL microclimate. Building and Environment 41: 343–354,
Honjo T, Takakura T. 1990/91. Simulation of thermal effects of urban
DOI:10.1016/joc.buildenv.2005.01.032.
green areas on their surrounding areas. Energy and Buildings 15–16:
Spronken-Smith RA. 1994. Energetics and cooling in urban parks.
443–446.
Unpublished PhD Thesis, The University of British Columbia,
Jauregui E. 1990/91. Influence of a large urban park on temperature
Vancouver.
and convective precipitation in a tropical city. Energy and Buildings
Spronken-Smith RA, Oke TR. 1998. The thermal regime of urban
15–16: 457–463.
parks in two cities with different summer climates. International
Jonsson P. 2004. Vegetation as an urban climate control in the
Journal of Remote Sensing 19: 2085–2104.
subtropical city of Gaborone, Botswana. International Journal of
Steemers K. 2003. Energy and the city: density, buildings and transport.
Climatology 24: 1307–1322, DOI: 10.1002/joc.
Energy and Buildings 35: 3–14.
Kjelgren R, Montague T. 1998. Urban tree transpiration over turf and
Swaid H, Hoffman ME. 1990/91. Climatic impacts of urban design
asphalt surfaces. Atmospheric Environment 32: 35–41.
features for high and mid-latitude cities. Energy and Buildings 14:
Kuttler W, Weber S, Schonnefeld J, Hesselschwerdt A. 2007. Urban/
325–336.
rural atmospheric water vapour pressure differences and urban
Taha H, Akbari H, Rosenfeld A. 1991. Heat island and oasis effects
moisture excess in Krefeld, Germany. International Journal of
of vegetative canopies: micro-meteorological field-measurements.
Climatology 27: 2005–2015.
Theoretical and Applied Climatology 44: 123–138.
McPherson EG, Dougherty E. 1989. Selecting trees for shading in the
Upmanis H, Eliasson I, Lindqvist S. 1998. The influence of green areas
Southwest. Journal of Arboriculture 15: 35–42.
on nocturnal temperatures in a high latitude city (Göteborg, Sweden).
Meir I, Pearlmutter D, Etzion Y. 1995. On the microclimatic behavior
International Journal of Climatology 18: 681–700.
of two semi-attached courtyards in hot dry region. Building and
Young Co RM. 1998. Technical Data on Model 41002 Gill Multi-Plate
Environment 30: 563–572.
Shield, Traverse City, MI.
Mills G. 1997. The radiative effects of building groups on single
Yezioro A, Capeluto IG, Shaviv E. 2006. Design guidelines for
structures. Energy and Buildings 25: 51–61.
appropriate insolation of urban squares. Renewable Energy 31:
Monteith JL, Unsworth MH. 1995. Principles of Environmental
1011–1023.
Physics, 2nd edn. St. Martin Press Inc: New York.
Offerle B, Eliasson I, Grimmond CSB, Holmer B. 2007. The influence
of surface heating on air temperature, wind and turbulence in an

Copyright  2009 Royal Meteorological Society Int. J. Climatol. 30: 44–57 (2010)
DOI: 10.1002/joc