You are on page 1of 9

Urban Forestry & Urban Greening 37 (2019) 56–64

Contents lists available at ScienceDirect

Urban Forestry & Urban Greening


journal homepage: www.elsevier.com/locate/ufug

Reprint of: Review on urban tree modelling in CFD simulations: T


Aerodynamic, deposition and thermal effects☆
Riccardo Buccolieria, , Jose-Luis Santiagob, Esther Rivasb, Beatriz Sáanchezb

a
Dipartimento di Scienze e Tecnologie Biologiche ed Ambientali, University of Salento, S.P. 6 Lecce-Monteroni, 73100 Lecce, Italy
b
Environment Department, Research Center for Energy, Environment and Technology (CIEMAT), Spain

ARTICLE INFO ABSTRACT

Keywords: This paper reviews current parameterizations developed and implemented within Computational Fluid
Aerodynamic Dynamics models for the study of the effects linking vegetation, mainly trees, to urban air quality and thermal
CFD modelling conditions. In the literature, passive mitigation via deposition is parametrized as a volumetric sink term in the
Deposition and thermal effects transport equation of pollutants, while a volumetric source term is used for particle resuspension. The aero-
Trees
dynamics effects are modelled via source and sink terms of momentum, turbulent kinetic energy and turbulent
Urban air quality
Vegetation parameterizations
dissipation rate. A volumetric cooling power is finally considered to account for the thermal (transpirational
cooling) effects of vegetation. The most recent applications are also summarized with a focus on the relative
importance of both aerodynamic and deposition effects, together with recent studies evaluating thermal effects.
Those studies have shown that the aerodynamic effects of trees are stronger than the positive effects of de-
position, however locally the pollutant concentration increases or decreases depending on the complex inter-
relation between local factors such as vegetation type and density, meteorological conditions, street geometry,
pollutant characteristics and emission rates. Unlike aerodynamic and deposition effects on pollutant dispersion
which were also found in street far from trees, the thermal effects were in general locally restricted to the close
vicinity of the vegetation and to the street canyon itself. Future requirements in CFD modelling include more in
depth investigation of resuspension and thermal effects, as well as of the VOCs emissions and chemical reactions.
The overall objective of this review is to provide the scientific community with a comprehensive summary on the
current parameterizations of urban vegetation in CFD modelling and constitutes the starting point for the de-
velopment of new parametrizations in CFD as well as in mesoscale models.

1. Background Bosch (2015), who reviewed benefits and challenges associated with
implementing the compact city approach by looking at the effects of
The urban population in 2015 accounted for 54% (4 billion) of the urban densification and compact city development on urban green
total global population and it is expected to increase to 60% (4.9 billion space and its planning. Their review underlines that provision of urban
people) of world population by 2030 (UN-Habitat, 2016). People often green space in compact city environments and during densification
breathe air that does not meet the legislation standards making urban processes is still a major challenge. Loss of private urban green space
air quality one of the most important environmental challenges and the rarely is counterbalanced by provision of more public green space.
largest environmental health risk in Europe (EEA, 2015). Together with Several ways are identified to deal with these challenges, e.g. how loss
legislative actions to improve air quality, several scientific works have of green space quantity can be offset by increased green space quality. A
been conducted on the potential of urban vegetation as mitigation tool systematic review of epidemiological studies on health benefits of green
for air quality improvement, carbon sequestration, micro-climate reg- spaces in the living environment has been recently provided by Van den
ulation, noise reduction, rainwater drainage, airborne pollution mod- Berg et al. (2015), who showed a strong evidence for significant posi-
ification (Salmond et al., 2016; Livesley et al., 2016). Psycho-logical tive associations between the quantity of green space and perceived
and recreational values were also reported by Haaland and van den mental health and all-cause mortality, and a moderate evidence for an


This article is part of a special issue entitled “Green Infrastructure: Nature Based Solutions for sustainable and resilient cities” published at the journal Urban
Forestry & Urban Greening 37C.

Corresponding author.
E-mail address: riccardo.buccolieri@unisalento.it (R. Buccolieri).

https://doi.org/10.1016/j.ufug.2018.07.004

Available online 24 July 2018


1618-8667/ © 2018 Elsevier GmbH. All rights reserved.
R. Buccolieri et al. Urban Forestry & Urban Greening 37 (2019) 56–64

association with perceived general health. 2. Brief overview of regulating (dis)services of urban street trees
When dealing with air quality at local scale, urban vegetation,
especially trees, as well as other obstacles and barriers, have been de- The recent papers by Salmond et al. (2016) and Livesley et al.
monstrated to influence flow patterns and thus the levels of pollutant (2016) have systematically reviewed works analysing the role of street
concentrations (see recent reviews by Gallagher et al., 2015; Janhäll, vegetation, mainly, trees in provision of regulating (dis)services which
2015; Grote et al., 2016; Abhijith et al., 2017). Some green infra- are briefly summarized here:
structures as vegetation barriers (e.g. hedges) can potentially improve
urban air quality acting as a barrier between traffic emissions and po- - microclimate regulation: increasing vegetation in urban areas may
pulation. In addition, pollutant concentration is mitigated by means of lead to reduced ambient and surface temperatures and increased
deposition on these green infrastructures. However, the impact of street evapotranspiration, precipitation interception and reduced runoff.
trees is more complex and reduce or increase the concentration de- This is an effective option for mitigating urban heat and adapting to
pending on the case. Results have highlighted the complexity and the climate changes caused by regional-scale changes in land use and
need for a careful design of vegetation structures to optimize benefits global scale changes in atmospheric composition (Gill et al., 2007).
and reduce the potential for unintended consequences. A systematic Local reductions in temperature may change the rate of chemical
evaluation of the impact of vegetation health, height, species, density, reactions within the atmosphere, leading to reduced concentrations
distance from the road, or leaf area index (ratio of leaf area to ground of other pollutants such as ozone;
area, LAI) on the processes determining pollutant dispersion, deposition - noise attenuation: little is known about the specific value of street
and atmospheric chemistry has been recently provided by Abhijith et al. trees in reducing noise pollution in street canyons, although there is
(2017), who attempted to review the inter-relations between local certain evidence that trees can attenuate traffic noise roadside of
factors such as vegetation type and density, climatic conditions, street open busy streets (Kalansuriya et al., 2009);
geometry, pollutant characteristics and emission rates. There remains - emission of biogenic volatile compounds: street trees directly emit
however considerable controversy as to whether increased vegetation gases precursors to the formation of secondary pollutants such as
within urban areas has the potential to provide a positive or negative ozone. Trees emit biogenic volatile organic compounds (bVOCs) as a
contribution to local scale air quality. reaction to stress in their environment, such as high light intensities
In reviewing past works on such topics, it is worth noting that and/or temperatures or low water availability (Seinfeld and Pandis,
aerodynamic and deposition effects have been separately investigated 2006; Leung et al., 2011). In the presence of NOx and sunlight, VOC
in simple geometries, and only few recent studies have brought them contribute to ozone and particulate formation, which may accu-
together to enable an appreciation of their relative influence and how mulate locally when ventilation is limited (Calfapietra et al., 2013);
they interact in real scenarios. Further, very limited studies have re- - pollen release: exposure to allergenic pollen produced in the flowers
cently accounted for resuspension and thermal effects of vegetation on of trees is associated with a range of health effects. The timing of the
street temperatures, which, in turn, affect flow pattern and pollutant release varies depending on the tree species and environmental
distribution. conditions (Cariñanos et al., 2016). Dispersion of tree pollen is de-
In this context, this paper focuses on the parametrizations of the pendent on a number of environmental factors, including local
effects of urban vegetation, trees in particular, on flow and pollutant meteorological conditions;
dispersion in urban areas by reviewing the most recent Computational - absorption and deposition of pollutants: street trees have the potential
Fluid Dynamics (CFD) microscale simulations. The overall objective is to regulate air quality by absorbing pollutants and increasing pol-
to provide the scientific community with a comprehensive overview on lutant deposition. Trees increase both the surface roughness
the current parameterizations of aerodynamic, deposition, resuspension (slowing air flow thus enhancing deposition and absorption pollu-
and thermal effects of trees in CFD models (Fig. 1), as well as insights on tant removal processes) and the area of the ground surface that at-
the relative importance of aerodynamic and deposition effects on flow mospheric pollutants come into contact with (acting as biological
and pollutant concentration. This review could constitute a systematic filters, enhanced by the properties of their surfaces). Trees absorb
summary for scientists interested in modelling vegetation in their stu- CO2 and gaseous pollutants such as O3, NO2, SO2 primarily by up-
dies and the starting point for the development of new parametrizations take via leaf stomata or surface, and accumulate airborne particu-
in CFD models as well as in mesoscale models, which is an open lates (by interception, impaction or sedimentation) more effectively
question for the scientific community. than other urban surfaces (Escobedo and Nowak, 2008; Janhäll,
2015).

As for air quality at local scale, as already mentioned in the


Introduction, the atmosphere-urban vegetation interaction is complex
and it is needed a careful design of green infrastructure to optimize air
quality improvements and reduce unintended consequences.

3. Brief overview of experimental studies for CFD validation


purposes

Both field and wind tunnel investigations have looked at the effects
of urban vegetation, mainly trees, hedges and other vegetation barriers,
on flow and pollutant concentrations in urban areas. While most field
experiments were performed in real scenarios, wind tunnel experiments
mainly considered idealized cases of isolated trees or located within
simple street canyons. For a comprehensive review see Abhijith et al.
(2017). Here a brief overview of wind tunnel studies which are used for
CFD validation purposes is provided.
In full-scale field tests flow and concentrations are measured under
Fig. 1. Effects of street trees parametrized in CFD models and discussed in the real atmospheric conditions. They have the advantage that the full
present review. complexity of a real problem is considered even though they are limited

57
R. Buccolieri et al. Urban Forestry & Urban Greening 37 (2019) 56–64

by low spatial resolution and uncontrollable meteorological conditions. are not considered. For example, for particle matter, the filtering ca-
As for vegetation, the main advantage is that the “real” effect has been pacity of vegetation by particle deposition on leaves has an optimum for
considered, but it is not likely to distinguish between aerodynamic, a specific permeability, and then findings related to the permeability
deposition, resuspension and thermal effects on concentration levels. that improves air quality cannot be directly transferred to particle
This makes field tests not suitable for evaluating the relative con- concentrations (Gromke et al., 2016).
tribution of those effects, even though they are important to validate
modelling simulations under real conditions. 4. CFD modelling of urban vegetation
In wind tunnel experiments boundary conditions can be chosen
based on the problem and the stationary flow conditions can be Complementary to experimental studies, flow and pollutant dis-
maintained throughout the experiments. However, they also suffer from persion models constitute a relatively economic approach for assessing
the limited set of spatial points. In addition to high costs (i.e. personnel, urban air quality as well as temporal and spatial variations. They give
instrumentation and others), the set-up is time-consuming and requires insights into the physical and chemical processes that govern the dis-
adherence to similarity criteria. As for vegetation, the aerodynamic persion of atmospheric pollutants. At the micro-scale, the CFD tech-
effects, which have been proven in field tests to dominate over the nique, even though computational expensive, is the preferred way of
deposition in street canyon, have been investigated without the influ- investigation and the most suitable for studies of various physical flow
ence of deposition, resuspension or thermal effects and this makes wind and dispersion processes in complex geometries such as cities. CFD
tunnel data suitable to validate modelling simulations under controlled offers some advantages compared to other methods since provide re-
conditions. sults of flow features at every point in space simultaneously and do not
Some wind-tunnel experiments were performed in real vegetation suffer from potentially incompatible similarity because simulations can
canopies. Brunet et al. (1994) analysed flow profiles in and above a be conducted at full scale.
continuous forest and Raupach et al. (1987) studied the flow at the edge In this context, the main objective is to review some of the most
of a forest performing wind-tunnel experiments in different forest- recent CFD modelling studies which have developed and proposed
clearing configurations. These experiments were used to test and tune parameterizations of aerodynamic and deposition (including re-
some CFD models (see Section 4.1). Few experiments were conducted suspension) effects of urban vegetation, mainly trees, and looked at
using real trees in wind tunnel to investigate the particle removal ef- their effects on flow and pollutant concentrations in urban areas.
ficiency, confirming that trees can improve air quality by capturing Recent studies which have investigated the modelling of resuspension
particles in their foliage (Rasanen et al., 2013; Huang et al., 2013; Lin and thermal effects of vegetation are also considered. The importance
and Khlystov, 2012). A systematic evaluation of the aerodynamic ef- of thermal effects is twofold. First, the impact of thermal effects on flow
fects of street trees and hedgerows has been performed in other wind within the urban canopy induces changes in the pollutant distribution.
tunnel tests in idealized scenarios. Two different sets of wind tunnel And second, urban vegetation is widely used to improve thermal
experiments have been performed at the University of Karlsruhe, La- comfort but its effects on air quality are not being jointly investigated
boratory of Building- and Environmental Aerodynamics. They are re- for planning purposes.
ported here since they have constituted the basis for validation of most
of the CFD modelling studies performed until now. 4.1. Parameterization of aerodynamic effects
The first wind tunnel dataset is CODASC (COncentration DAta of
Street Canyons – www.windforschung.de/CODASC.htm) and comprises The representation of vegetation in the CFD modelling is crucial to
pollutant concentrations in street canyons of street-width-to-building- capture the its effects on wind flow in urban areas. Salim et al. (2015)
height ratio W/H = 1 (narrow) and W/H = 2 (broad) and a street- investigated the aerodynamic effects of the inclusion of trees with dif-
length-to-building-height ratio of L/H = 10 with avenues of trees of ferent approaches: a) an implicit approach where trees were included in
different tree-stand densities ρts and crown porosities PVol subjected to the surface parameterization as a roughness length and b) an explicit
various wind directions (Gromke and Ruck, 2012). Trees were not real approach where trees were considered as porous media and additional
but realized using custom-made lattice cages. A filament/fibre-like terms are added to the momentum and turbulence equations. Results
synthetic wadding material was used to fill the cells, whose purpose showed that only explicit approach could be used to simulate wind flow
was to facilitate a uniform distribution of the wadding material in urban areas. In this sense, most of CFD studies have considered ve-
throughout the entire length of the lattice cage. The strongest effects of getation as a porous medium, which is modelled by the addition of a
avenues of trees on traffic pollutant dispersion were observed for ob- momentum source (sink) term to the standard fluid flow equations. The
lique wind directions for which also the largest concentrations at the source term is composed of two parts: a viscous loss term (Darcy) and
canyon walls were found. Thus, the prevailing assumption, which at- an inertial loss term. Some studies (e.g. Buccolieri et al., 2009, 2011;
tributes the most harmful dispersion conditions to a perpendicular Jeanjean et al., 2015, 2016, 2017) used a pressure loss coefficient λ
wind, does not hold for street canyons with avenues of trees. [m−1] as inertial loss term based on measurements of porous material
The second wind tunnel dataset collects concentration data ob- in wind tunnel. This coefficient is the ratio between the static pressure
tained from the dispersion of gaseous traffic pollutants in an isolated difference between the windward and leeward side of the porous ma-
street canyon of W/H = 2 and L/H = 10 with roadside hedges (Gromke terial ΔPstatic [Pa] and the dynamic pressure divided by streamwise
et al., 2016). Hedgerows were realized as trees in the CODASC case. depth of the material l [m] (Balczó et al., 2009):
Various hedge configurations of differing height, permeability and
Pstatic
longitudinal segmentation (continuous over street length L or dis- = [m 1]
((1/2) u2) l (1)
continuous with clearings) were investigated under a perpendicular and
parallel wind. Two arrangements were examined: two eccentric This coefficient was applied to simulate wind tunnel experiment of
hedgerows sidewise of the main traffic lanes and one central hedgerow CODASC database (see previous section). In this approach, turbulence
between the main traffic lanes. This study indicates that continuous in the medium is treated as though the porous medium has no effect on
hedgerows can be employed as an effective passive control measure for the turbulence generation or dissipation rates.
air quality in urban street canyons. Other recent CFD works about urban vegetation (e.g. Katul et al.,
Findings obtained in the two wind tunnel experiments described 2004; Mochida et al., 2008; Santiago et al., 2013; Amorim et al., 2013;
above are obviously restricted to the aerodynamic effects of gaseous Krayenhoff et al., 2015; Salim et al., 2015; Vranckx et al., 2015;
pollutants because the effect to absorb gases and filter out particle Gromke and Blocken, 2015a; Santiago et al., 2017a; Moradpour et al.,
matter (deposition effect), as well as resuspension and thermal effects, 2017), following Green (1992) and Liu et al. (1996), parameterized the

58
R. Buccolieri et al. Urban Forestry & Urban Greening 37 (2019) 56–64

Table 1
Summary of additional term added to the momentum equations for modelling aerodynamic effects of vegetation and values of sectional drag coefficient used in CFD
simulations.
Additional Term [Pa m−1] Sectional drag coefficient References
Cd

Sui = LADCd Uui 0.1–0.3 Katul et al. (2004) and Vranckx et al. (2015)
0.2 Santiago et al. (2013), (2017a,b), Amorim et al. (2013), Krayenhoff et al. (2015), Salim et al. (2015), Gromke and
Blocken (2015a), Gromke et al. (2015) and Moradpour et al. (2017)
0.25 Jeanjean et al. (2015), (2017)Jeanjean et al., 2015Jeanjean et al. (2015), (2017)
0.3 Liu et al. (1996)

form drag (inertial term) at grid cell scale due to the interactions of 2
2
3 3
foliage and atmosphere in terms of leaf area density LAD [m2 m−3] of = Cµ0.5 + [ ]
d p
k (5)
the vegetation and added a sink of momentum term (Sui ) to each mo-
mentum equations (one for each velocity component, ui) as: 2
2 Cµ0.5 2 3
C 4 (=C 5) = (C 2 C 1) [ ]
Sui = LADCd Uui [Pa m 1] (2) k
6 (6)
where ρ is the air density [kg m−3], ui is the appropriate wind velocity
where it is assumed Cε4 = Cε5, α = 0.05 (Dalpé and Masson, 2009) and
component [m s−1], U is the wind speed [m s−1] and Cd is the sectional
(Cμ, σk, σε, Cε 1, Cε2) = (0.09, 1, 1.3, 1.44, 1.92). As summary, βp is
drag for vegetation (dimensionless). The sectional drag is a constant
usually assumed equal to 1 and the values of βd, Cε4 and Cε5 range
related to aerodynamic features of vegetation (see Section 4.3 for values
between 4–6.5, 0.9–2 and 0.9–1.8, respectively. However, the optimal
usually employed). Higher foliage and wind speed induce higher sink of
values of these parameters depend on the study case. Mochida et al.
momentum. Please note that this is the same approach as using the
(2008) optimized the values of Cε4 and Cε5 by comparing numerical
pressure loss coefficient λ = LADCd.
results with field measurements. Recently, Krayenhoff et al. (2015)
In addition, the vegetation modifies the mean flow motion into
carried out analysis about the optimisation of Cε5 finding it is case
wake turbulence, but the smaller length scale compared to turbulence
dependent. This parameter in particular controls the destruction of
generated by the shear induces a fast dissipation. Then, this process is
dissipation due to the foliage, i.e. a low Cε5 yields a higher dissipation
usually parametrized as source and sink terms of turbulent kinetic en-
rate and small turbulent kinetic energy in the canopy (and elsewhere
ergy (k) and turbulent dissipation rate (ε) (Amorim et al., 2013;
depending on vertical transport). They used two wind-tunnel data sets:
Santiago et al., 2013; Gromke and Blocken 2015a; Krayenhoff et al.,
a) a “continuous forest” configuration (Brunet et al., 1994) and b)
2015; Santiago et al., 2017a,b,c; Moradpour et al., 2017; Hong et al.,
several “forest-clearing” configurations (Raupach et al., 1987). Vertical
2018) as follows:
profile of mean velocity, turbulent kinetic energy and Reynolds stress
Sk = LADCd ( p U 3 d Uk ) [kg m 1 s 3] were analysed. In the first case, better results were found for Cε5 = 1
(3)
and in the second one for Cε5 = 1.10.These datasets have been used to
test several other RANS and LES model implementation: the k-l model
S = LADCd C 4 ( p kU
3 C5 dU ) [kg m 1 s 4]
(4) of Wilson and Flesch (1999), the k-ε model of Foudhil et al. (2005), and
the LES models of Yang et al. (2006) and Dupont and Brunet (2008)
where βp is the fraction of mean kinetic energy converted into turbulent where a different set of parameters were used by each model. In this
kinetic energy by means of drag and takes a value between 0 and 1; βd is way, a wide range of values for Cε5 (0.6–1.8), as well as for other
the dimensionless coefficient for the short-circuiting of the turbulence parameters involved in the parametrization of the impacts of foliage on
cascade; Cε4 and Cε5 are model constants. Several values of these k and ε, were employed in different studies (Green, 1992; Liu et al.,
parameters could be found in the literature. βp = 1 and βd = 4 have 1996; Foudhil et al., 2005; Mochida et al., 2008; Endalew et al., 2009;
been used by Amorim et al. (2013) and Salim et al. (2015). Green Dalpe and Masson, 2009; Rosenfeld et al., 2010) and there is no evi-
(1992) defined Cε4 = 1.5 and Cε5 = 1.5 and these values were applied dence that one of these set of parameters works better than other in a
to urban vegetation modelling by Amorim et al. (2013) and Hong et al. general case.
(2018). Other studies as Santiago et al. (2013), Gromke and Blocken A summary of the additional terms added to the momentum and
(2015a), Moradpour et al. (2017) or Santiago et al. (2017a) computed turbulence equations for modelling aerodynamic effects is shown in
βd, Cε4 and Cε5 values based on analytical formulation of Sanz (2003). Tables 1–3. The tables also report on some of the reference studies
Santiago et al. (2013) used βp = 1 and computed the other values (di- which employed those parameterizations.
mensionless) as follows: Finally, it is worth mentioning that there is a lack of studies

Table 2
Summary of additional terms added to the turbulent kinetic energy equation for modelling aerodynamic effects of vegetation and values of constants used in CFD
simulations.
Additional Term [kg m−1 s−3] Constant values References

0 Buccolieri et al. (2011) and Jeanjean et al. (2015), (2017)Jeanjean et al., 2015Jeanjean et al. (2015), (2017)
Sk = LADCd ( p U3 d Uk )
βp = 1 Green (1992), Liu et al. (1996), Katul et al. (2004), Mochida et al. (2008), Amorim et al. (2013) and Salim et al. (2015)
βd = 4
βp = 1 Hong et al. (2018)
βd = 3
βp = 1 Santiago et al. (2013), (2017a,b) and Krayenhoff et al. (2015) using βd = 6.5; Katul et al. (2004), Gromke and Blocken
(2015a), Gromke et al. (2015) and Moradpour et al. (2017) using βd = 5.1
2
d = Cµ0.5 ()
2 3
p +
3
k

59
R. Buccolieri et al. Urban Forestry & Urban Greening 37 (2019) 56–64

Table 3
Summary of additional term added to the turbulent dissipation rate equation for modelling aerodynamic effects of vegetation and values of constants used in CFD
simulations.
Additional Term [kg m−1 s−4] Constant values References

0 Buccolieri et al. (2011) and Jeanjean et al. (2015), (2017)Jeanjean et al.,


2015Jeanjean et al. (2015), (2017)
Cε4 = 1.5 Green (1992), Katul et al. (2004), Amorim et al. (2013) and Hong et al. (2018)
S = LADCd C ( 4 pkU
3 C5 dU ) Cε5 = 1.5
Cε4 = 1.5 Liu et al. (1996)
Cε5 = 0.6
Optimized values of Cε4 and Cε5 Mochida et al. (2008)
(Cε4 = 1.5–2; Cε5 = 1.5–1.8)
Cµ0.5
2 Santiago et al. (2013), (2017a,b) and Krayenhoff et al. (2015) using
C4= k
2
6 ()
2 3
(C 2 C 1) Cε4 = Cε5 = 1.26; Katul et al. (2004), Gromke and Blocken (2015a), Gromke et al.
(2015) and Moradpour et al. (2017) using Cε4 = Cε5 = 0.9
Cε4 = Cε5

comparing models with vegetation parameterizations using only the In real world, the leaves are aligned in different orientation to the wind
term of sink of momentum with parameterizations using also turbulent flows which is not included in the equation. Which thus can be seen as
kinetic energy and dissipation sink/source terms. However, we can an average sink term of the trees on particulate over the whole tree
analyse the results from validation exercises carried out by different crown area.
CFD models using CODASC wind-tunnel experiments. Gromke et al. Trees also affect particulate matter in the atmosphere through re-
(2008) modelled the vegetation only taking into account the sink of suspension of particles captured on the plant surface (Nowak et al.,
momentum and applying a RANS model obtained a general good 2013; Chen et al., 2017). Nowak et al. (2013) estimated a PM2.5 re-
agreement with concentration measured. Improved results were ob- suspension from leaves to range between 4.5% and 12% for wind
tained by Moonen et al. (2013) which use the same sink of momentum speeds between 3–9 m/s, respectively. The particulate resuspension is
but with a LES model. Santiago et al. (2017a) and Vranckx et al. (2015) usually neglected in CFD simulations. Recently, Hong et al. (2018)
used Eqs. (2)–(4) to model vegetation and they did not obtained much parameterized it in CFD simulations as a volumetric source term in the
better fit of experimental concentration. This comparison is limited transport equation of pollutants (Bell and Treshow, 2003):
because only concentrations are considered, and in these conditions,
there is no evidence to establish the differences between parametriza- Sr = LADVr Csink (x , y, z ) [kg m 3 s 1]
(9)
tion performances and more studies would be necessary. −1
where Vr is the particle resuspension velocity from plant foliage [m s ]
and Csink is the particle concentration deposited on plant foliage
4.2. Parameterizations of deposition and resuspension effects [kg m−3].
A summary of additional terms added to the pollutant dispersion
In addition to the aerodynamic effects, the presence of vegetation equation for modelling deposition and resuspension effects of vegeta-
removes pollutants from air by means of deposition on the leaves. tion and values of deposition velocities used in CFD simulations is
Traditionally, this effect has been parametrized by means of a down- shown in Table 4. The table also reports on some of the reference stu-
ward flux as follows: dies which employed those parameterizations.
2 s 1]
Fdepo = Vd (z ref ) C (z ref ) [kg m (7)
4.3. Values of Vd, Cd and LAD
where C(zref) is the mean concentration of pollutant at zref [kg m−3] and
Vd(zref) is the deposition velocity at zref [m s−1]. This approach is used The values of deposition velocity Vd depend on the type of vegeta-
in mesoscale modelling. However, since the resolution at the microscale tion and pollutant. Many discrepancies between published values are
is much finer and resolves the vegetation into several computational found (Nowak et al., 2006, 2013, 2018; Janhäll, 2015). Deposition
cells, urban vegetation has been modelled in CFD simulations as a vo- velocities for vegetated surfaces are usually less than 1 cm s−1 for some
lumetric sink term in the transport equation of pollutants. This term is gases to several cm s−1 for particles. For example, Litschke and Kuttler
proportional to LAD, Vd and pollutant concentration C(x, y, z) within (2008) in their review of urban particle deposition found that the
each cell (Vos et al., 2013; Vranckx et al., 2015; Santiago et al., average published value for PM10 deposition velocity is around
2017a,b; Jeanjean et al., 2016, 2017): 1 cm s−1, however current in-situ measurements indicate deposition
velocities higher than this value and for PM1 above 10 cm s−1. For
Sd = LADVd C (x , y, z ) [kg m 3 s 1]
(8)
PM2.5 values of 0.02 cm s−1 (lower end) (Peters and Eiden, 1992),
Please note that Eq. (8) works for homogenous vegetation surfaces. 0.64 cm s−1 (intermediate) (Pugh et al., 2012) and 30 cm s−1 (higher

Table 4
Summary of additional terms added to the pollutant dispersion equation for modelling deposition and resuspension effects of vegetation and values of deposition
velocities used in CFD simulations.
Additional Term [kg m−3 s−1] Deposition velocity Vd[cm s−1] References

Sd = LADVd C (x , y, z ) 0.2–1 (for PM10) Vos et al. (2013)


0.5–5 (for PM10) Vranckx et al. (2015)
0.64 (for PM2.5) Jeanjean et al. (2016) and Jeanjean et al. (2017)
0.25–10 (Sensitivity analysis) Santiago et al. (2017a)
0.5–3 (for NOx) Santiago et al. (2017b)
4.58 (for PM2.5) Hong et al. (2018)
Sr = LADVr Csink (x , y, z ) Resuspension velocity Vr [cm s−1] Concentration deposited on plant Csink [kg m−3] Hong et al. (2018)

60
R. Buccolieri et al. Urban Forestry & Urban Greening 37 (2019) 56–64

end) (White and Turner, 1970) are found. In addition, Beckett et al. thermal effects should be appropriate parameterized. On one hand trees
(2000), Freer-Smith et al. (2005) and Nowak et al. (2013) mentioned shade buildings and ground and on the other hand they provide tran-
deposition velocities well above 1 cm s−1 depending on wind speed. spirational cooling.
However, these values are rather exceptional and deposition velocities In general, thermal CFD simulations are not common in the litera-
of 1 cm s−1 or less are more frequent. See Table 4 for a summary of ture (Sini et al., 1996; Santiago et al., 2014; Nazarian and Kleissl, 2016)
values of deposition velocity Vd used in CFD simulations. and few studies including vegetation have been performed (Gromke
Recently Santiago et al. (2017a) analysed the differences between et al., 2015; Moradpour et al., 2017). The transpirational cooling of
deposition velocities from mesoscale and microscale point of view. vegetation has recently been included in thermal CFD simulations to
They showed as in urban configurations the deposition velocity from evaluate the impact on temperature. In addition, Moradpour et al.
mesoscale point of view is not similar to microscale deposition velocity (2017) also analysed and the dispersion including photochemistry re-
due to the emissions are below the vegetation canopy. Considering a actions, and how can this modify the flow field and thus the final
downward flux of pollutant over a forest, the mesoscale deposition concentration levels within the streets. Concerning photochemistry re-
velocity can be calculated from concentration at a reference height actions, the photostationary steady state defined by three reaction
above the canopy, taking into account an aerodynamic resistance in system of NO, NO2 and O3 is used in the CFD model as in the previous
addition to microscale deposition velocity. However, in urban config- studies of Baker et al. (2004) and Sanchez et al. (2016).
urations the scenario is different. The traffic emissions are located In order to account for the effect of vegetation on air temperature,
below urban vegetation canopy and the resistance model, traditionally Gromke et al. (2015) employed a volumetric cooling power Pc [W m−3]
used at mesoscale for computing deposition velocity, cannot be applied per unit volume vegetation as a function of LAD. This method can be
as in a forest configuration. seen as an integral and rather coarse scale approach where the tran-
As for the leaf drag coefficient Cd, values of for forest range between spirational cooling effect is allocated to a volume containing vegetation,
0.15 and 0.3 (Li et al., 1985; Massman 1987), however, usually, the contrary to a microscale approach where the latent heat flux at leaves
drag coefficient is set 0.2 in CFD modelling of urban vegetation to re- via stomata is considered (Defraeye et al., 2013).
flect an average value instead of a species specific value (Gromke and The basic principle is that when air flows through vegetation it gets
Blocken, 2015a). See Table 1 for a summary of values of leaf drag cooled by transpiration mainly from the leaf surfaces. The heat H (J)
coefficients Cd used in CFD simulations. required to change the temperature of an object can be calculated ac-
Values of LAD depend on the kind of vegetation and for deciduous cording to (Incropera et al., 2011):
trees also on the season considered (Di Sabatino et al., 2015). Lalic and
Mihailovic (2004) showed values of distribution of LAD for different
H = cp m T [J] (10)
species with an average of approximately to 1 m2 m−3. The values used with cp the specific heat capacity [J K−1 kg−1], m the mass of the object
in CFD modelling ranges from 0.1 (Santiago et al., 2017a) to 4 m2 m−3 [Kg] and ΔT the change in temperature [K]. Since the volumetric
(Vranckx et al., 2015) with an average value in the literature about cooling power Pc is understood as the transfer of heat per time t [s] and
1 m2 m−3, and, in general, LAD is considered constant with height. In volume V [m3], Eq. (10) can be rearranged to:
CFD modelling, in addition to tree species or season of the year, it
should also be taken into account that it is modelled a zone where the H 1 1 1
= PC = cP m T [W] T = PC V [K]
vegetation is located but the individual trees are not usually considered. V V m cp (11)
For this reason, the equivalent LAD of this zone should be less than the with Ḣ the heat transfer rate [W] and ṁ the mass flow rate [kg s−1].
LAD for individual trees because in the vegetation zone we have to The right part of Eq. (11) states that the change in temperature ΔT when
consider also the space between trees. A summary of values of LAD used air flows through vegetation is proportional to the cooling power Pc and
in CFD simulations is shown in Table 5 which also reports on some of the volume of the vegetation V and inversely proportional to the air
the reference studies. mass flow ṁ through the vegetation volume and the specific heat ca-
pacity cp of air. Eq. (11) is based on the simplified assumption that the
4.4. Parameterization of thermal (cooling) effects heat/mass transfer at the leaf surface is not a function of the flow re-
gime and the corresponding heat transfer coefficient. The same model
Urban vegetation plays an important role on thermal comfort inside has been recently applied by Moradpour et al. (2017). Gromke et al.
the city, and, with this purpose, it is planning an increase of green in- (2015) and Moradpour et al. (2017) found for a volumetric cooling
frastructure in many of them as a measure to mitigate high outdoor air power of 250 W m−3 per unit of LAD the best agreement with experi-
temperatures. However, the impact of vegetation on air quality jointly mental air temperature in different validation studies (Table 6).
with its thermal effects has been poorly investigated. In this way, the
CFD technique is useful tool to address this problem, but vegetation
5. The contribution of street trees effects in idealized and real
scenarios
Table 5
Summary of leaf area density (LAD) values used in the CFD simulations.
5.1. Aerodynamic effects vs deposition effects
Leaf Area Density LAD [m2 m−3] References
The modification of airflow within the streets has been widely
0.7 (Trees) Vos et al. (2013)
2–5 (Hedges) studied in idealized and complex scenarios. A large number of micro-
1 Amorim et al. (2013) scale simulations were based on configuration of CODASC database (see
1 Gromke and Blocken (2015a) Section 3) in order to evaluate modelling results against wind-tunnel
0.55–2 Gromke et al. (2015)
data. Gromke et al. (2008) firstly applied a Reynolds-averaged Navier-
1.6–4 Vranckx et al. (2015)
1.6 Jeanjean et al. (2015), (2016)
Jeanjean et al., 2015Jeanjean Table 6
et al. (2015), (2016) Summary of additional term added to energy equation for modelling tran-
0.1–0.5 Santiago et al. (2017a,b) spirational cooling of vegetation used in CFD simulations.
1–1.6 Jeanjean et al. (2017)
0.5–2 Moradpour et al. (2017) Additional Term [W m−3] References
2.3 Hong et al. (2018)
Pc = 250LAD Gromke et al. (2015) and Moradpour et al. (2017)

61
R. Buccolieri et al. Urban Forestry & Urban Greening 37 (2019) 56–64

Stokes (RANS) model with two different turbulent closure: k-ε and effects of vegetation were shown to be beneficial for air quality.
Reynolds Stress Model (RSM) to a two dimensional street canyon with a In a neighbourhood of Pamplona (Spain), Santiago et al. (2017b)
line of trees. They found a better prediction of RSM and a dependence analysed the effects of trees with different LAD and the inclusion of new
of the performance with turbulent Schmidt number (Sct). Sct relates trees in a street. They observed as the aerodynamic effect on spatial
diffusivity with turbulent viscosity, increasing the diffusivity as Sct average pollutant concentration was moderate as LAD changed from
decreases. The commonly value used range from 0.7 to 0.9 but they 0.1 to 0.5 m2 m−3 (an increase of 5.4%) and this increase was greater
found that in these cases a reduction of Sct to 0.6 or 0.3, increases the than the reduction due to deposition (2.8% for a deposition velocity of
diffusion and induces a better agreement with experiments. In this 0.01 m s−1). This was also analysed for a wide range of deposition ve-
sense, Vranckx et al. (2015) observed that the optimum Sct ranges from locities and, in all cases, the NOx concentration was clearly higher
0.3 to 1.0 depending on the case, and Santiago et al. (2017a) and (1–8%) for LAD = 0.5 m2 m−3. However, local effects could be stronger
Gromke and Blocken (2015a) found the best agreement for their si- (increase and reductions > 100%), similar to what found by Gromke
mulations using Sct = 0.3 and 0.5, respectively. Large Eddy Simulations and Blocken (2015b) for an idealized neighbourhood considering only
(LES) have also carried out to model CODASC scenarios (Salim et al., aerodynamics effects of trees. In addition, Santiago et al. (2017b)
2011; Moonen et al., 2013). Better quantitative agreement with the showed as the inclusion of vegetation in only one street can change the
experimental data was obtained. However, this improvement in the flow fields and, although the variation of spatial average concentration
performance increases the computational cost at least one order of in the whole neighbourhood was minor, the local differences of pollu-
magnitude in comparison with RANS models. tant concentration could be greater than 100%.
There are only few studies which have looked at the relative con-
tribution of the aerodynamic and deposition effects, while there are no 5.2. Thermal (cooling) effects
studies also evaluating the relative contribution of resuspension effects.
Overall, those studies indicate that both the reduction of ventilation Thermal CFD studies which attempted to implement thermal
induced by street trees and the pollutant removal capacity of vegetation (cooling) effects of vegetation were those by Yoshida et al. (2006),
should be taken into account. Vos et al. (2013) simulated several con- Gromke et al. (2015) and Moradpour et al. (2017). The first analysed
figurations of urban vegetation founding locally aerodynamic effects of outdoor thermal comfort by coupled simulation of convection and ra-
roadside vegetation on pollutant concentration are much stronger than diation combined with a tree canopy model considering aerodynamic
the deposition effects. Therefore, they lead to an increase of con- and thermal effects. The second studied the transpirational cooling ef-
centration rather than to mitigate air quality problems. They proposed fects in urban street canyons of the J.P. van Muijlwijkstraat in the Dutch
to avoid dense rows of trees in street canyons with busy traffic and only city of Arnhem, in conjunction with aerodynamic effects, with the aim
used vegetation with a limited impact on flow fields (e.g. isolated trees, of assessing their effectiveness to mitigate high outdoor air tempera-
low hedges). Vranckx et al. (2015) modelled annual average con- tures in heatwaves. The second also included transport equations for
centration of elemental carbon (EC) and PM10 in a street canyon taking nitric oxide (NO), nitrogen dioxide (NO2), and ozone (O3) with the
into account the presence of trees and their dynamical and deposition photochemistry reactions, and analysed the influence of temperature
effects on air quality. They also found the effect of the reduction of reduction by trees on the formation of O3 in idealized regular arrays of
ventilation is greater than the positive effect of deposition. However, buildings.
the increase of concentration obtained is smaller than in earlier works The case study of Gromke et al. (2015) showed that locally applied
as Vos et al. (2013). Annual average concentrations of EC increased vegetative measures at the street and building scale can contribute to a
about 8% (range of 1%–13%) by the presence of vegetation, and in the reduction of air temperatures in urban street canyons during hot
case of PM10, this increase is even less, about 1.4% varying from 0.2% summer days by transpirational cooling. The intensity of cooling in
to 2.6%. In order to analyse the scenarios where aerodynamic effects terms of air temperature reductions and their spatial extent differed
are stronger than deposition (and vice versa), Santiago et al. (2017a) distinctly between the various vegetative measures. The avenue-trees
carried out a series of simulations over an idealized building config- provided a marked transpirational cooling within the street canyon
uration with different vegetation designs (several positions, LADs and both in terms of air temperature reductions and spatial extent. And the
deposition velocities). The height of vegetation and the magnitude of combination of avenue trees together with facade greening is most
microscale deposition velocity were observed in this configuration as promising. Different from aerodynamic and deposition effects on pol-
the key parameters that determine which of the two effects (deposition lutant dispersion which were also found in street far from trees, the
or aerodynamic) prevails. In general, the vegetation above the building transpirational cooling effect was in general locally restricted to the
roof height increases the concentration within the street in comparison close vicinity of the vegetation and to the street canyon itself. The
with the no vegetation case, even taking into account pollutant de- modifications in the flow fields would induce in modifications in the
position. Then, street trees below building roof height with a high de- pollutant distributions in the cases where dispersion is considered.
position velocity were proposed to mitigate air quality problems in In this sense, Moradpour et al. (2017) studied not only the impact
streets. In addition, they observed as, in some cases, the spatial average on temperature but also on NO, NO2 and O3 dispersion considering
pollutant concentration at pedestrian level was reduced by the presence photochemistry reactions. In addition to thermal effects, the aero-
of vegetation but the maximum at this level was increased. dynamic effects are considered modelling vegetation as a momentum
This shows as the analysis of the street trees effects, and in general sink. For similar configurations to that used by Gromke and Blocken
of vegetation, on air quality is complex even in simple configurations. (2015b), Moradpour et al. (2017) also found an increase of primary
The effects of trees including deposition in several real cities (Leisceter pollutants concentrations (NO, NO2) as vegetation is considered (note
and London, UK and Pamplona, Spain) were modelled in different that photochemistry is not considered by Gromke and Blocken
studies (Jeanjean et al., 2016, 2017; Santiago et al., 2017b). These (2015b)). This fact is due to ventilation efficiency of NO and NO2 is
studies concluded that the aerodynamic effect of trees prevails over weakened revealing that the dispersion of nitrogen oxides was influ-
deposition. In Marylebone neighbourhood (London), Jeanjean et al. enced by the flow patterns. In the presence of trees, flow is reversed,
(2017) found an increase of 7% for typical meteorological conditions at and it enhances as leaf area density increases; additionally, vegetation
the monitoring site, and only an additional reduction via deposition of creates downward and vortex flows which resulted in the accumulation
2% in PM2.5 concentration. They showed that changes in PM2.5 via of the cool air in the lower layer of the canyons. Nevertheless, chemical
deposition on trees are less important than the aerodynamic effects in reactions were significant for the dispersion of O3 and ventilation ef-
terms of magnitude in this street. However, these results were depen- ficiency of O3 improved in LAD = 0.5 and 1.0 cases. Aspect ratios and
dent on local meteorology and for parallel winds in street canyon the leaf area densities were also found to interact with each other;

62
R. Buccolieri et al. Urban Forestry & Urban Greening 37 (2019) 56–64

consequently, the optimum LAD was different for each aspect ratio. The analysis of the literature has also pointed out some shortages
Furthermore, vegetation as compared to tree-free environment, mostly and future perspectives:
led to a better chemical equilibrium. This research illustrated an ap-
proach to urban planner in order to design urban vegetation for im- - resuspension and thermal effects of vegetation have been considered
proving air quality and outdoor thermal environment. An interesting in few studies and there is the need for more works to gain con-
task for future study is to evaluate various conditions of air temperature fidence on the accuracy of their parametrizations in CFD modelling;
and solar radiation. - the different orientation of leaves to the wind flow and moving
Studies above which considered thermal effects of vegetation, in crown may be taken into consideration in the parameterizations of
conjunction with aerodynamic effects, and thus the influence of buoy- aerodynamic and deposition effects;
ancy on flow and pollutant dispersion, are limited. It is not possible to - emissions of VOCs from vegetation and chemical reactions may be
achieve any conclusion on the relative importance of thermal effects on sink or source of NOx and O3 and thus strongly affect the pollutant
air quality with respect to aerodynamic and deposition effects. concentrations in urban areas. Such effects are usually neglected in
However, this may represent one of the next step the scientific com- CFD modelling;
munity is called to deal with. - wind tunnel studies employed for validation purposes refer to
idealized scenarios under specific meteorological and conditions
6. Discussion and conclusions and for limited combinations of different types of vegetation (e.g.
trees, hedges) and LAD. There is the need for systematic and com-
6.1. The right tree in the right street prehensive filed and wind tunnel experiments of aerodynamic, de-
position, including resuspension, and thermal effects of different
The microscale studies on the role of vegetation, mainly trees, re- types of trees and vegetation in general. This would allow to better
viewed here suggest that urban vegetation cannot be used as a general evaluate which is the dominant effect for the different scenarios and
mitigation measure of air quality problems. From a local point of view, provide useful suggestion for planning purposes.
the aerodynamic effects of trees are stronger than the positive effects of
deposition and, this, usually, induces a reduction of ventilation in the Acknowledgements
streets and an average increase of pollutant concentration. However,
locally the pollutant concentration increases or decreases due to the JLS, ER and BS acknowledge funding from the projects: LIFE+
presence of trees depending on the studied zone and meteorological RESPIRA (LIFE13 ENV/ES/000417) funded by EU and TECNAIRE-CM
conditions. This paper is in line with previous reviews by Gallagher (S2013/MAE-2972) funded by The Regional Government of Madrid.
et al. (2015), Janhäll (2015) and Abhijith et al. (2017) which provide a
detailed overview on the impacts of green infrastructure on dispersion References
in the urban environment suggesting that urban planners need to ac-
count for the impact of trees and vegetation on pollutant dispersion. Abhijith, K.V., Kumar, P., Gallagher, J., McNabola, A., Baldauf, R., Pilla, F., Broderick, B.,
Di Sabatino, S., Pulvirenti, B., 2017. Air pollution abatement performances of green
One should be aware that any measure needs to be analysed for each
infrastructure in open road and built-up street canyon environments—a review.
particular case in conjunction with several other factors, i.e. pre- Atmos. Environ. 162, 71–86.
dominant meteorology, building morphology, types of trees, location of Amorim, J.H., Rodrigues, V., Tavares, R., Valente, J., Borrego, C., 2013. CFD modelling of
the aerodynamic effect of trees on urban air pollution dispersion. Sci. Total Environ.
pollutant sources (e.g. traffic). 461, 541–551.
In exploring the opportunity of planting trees and vegetation, it also Baker, J., Walker, H.L., Cai, X., 2004. A study of the dispersion and transport of reactive
should be taken in mind that they produce social benefits (e.g. making pollutants in and above street canyons—a large eddy simulation. Atmos. Environ. 38,
6883–6892.
urban environment more pleasant), economic benefits (e.g. increasing Balczó, M., Gromke, C., Ruck, B., 2009. Numerical modeling of flow and pollutant dis-
property value), health benefits (psycho-logical values), visual and aes- persion in street canyons with tree planting. Meteorol. Z. 18 (2), 197–206.
Beckett, K.P., Freer-Smith, P.H., Taylor, G., 2000. Particulate pollution capture by urban
thetic benefits, micro-climate regulation, storing carbon, reducing energy trees: effect of species and wind speed. Global Change Biol. 6, 995–1103.
use, etc. (Roy et al., 2012; Haaland and van den Bosch, 2015; Van den Bell, J.N.B., Treshow, M., 2003. Air Pollution and Plant Life, 2nd ed. John Willey & Sons
Berg et al., 2015). In addition, the vegetation could be used to improve Ltd., Chichester, UK.
Brunet, Y., Finnigan, J.J., Raupach, M.R., 1994. A wind tunnel study of air flow in waving
wind comfort of pedestrian reducing wind speed in zones where high wheat: single-point velocity statistics. Bound. Layer Meteorol. 70, 95–132.
speeds are usually presented. However, if the vegetation is planted in Buccolieri, R., Gromke, C., Di Sabatino, S., Ruck, B., 2009. Aerodynamic effects of trees on
urban areas for mitigating traffic-related pollution, benefits should be pollutant concentration in street canyons. Sci. Total Environ. 407, 5247–5256.
Buccolieri, R., Salim, S.M., Leo, L.S., Di Sabatino, S., Chan, A., Ielpo, P., de Gennaro, G.,
counterbalanced with the impact on flow and pollutant dispersion. “The Gromke, C., 2011. Analysis of local scale tree-atmosphere interaction on pollutant
more trees the better” approach is clearly inadequate in urban areas, while concentration in idealized street canyons and application to a real urban junction.
Atmos. Environ. 45, 1702–1713.
the most appropriate approach is “the right tree in the right street”. Failure Calfapietra, C., Fares, S., Manes, F., Morani, A., Sgrigna, G., Loreto, F., 2013. Role of
to account for the functional mechanisms may not achieve the desired biogenic volatile organic compounds (BVOC) emitted by urban trees on ozone con-
outcomes and may actually increases local pollution concentrations, with centration in cities: a review. Environ. Pollut. 183, 71–80.
Cariñanos, P., Adinolfi, C., Díaz de la Guardia, C., De Linares, C., Casares-Porcel, M.,
concomitant negative effects on human health. In addition, urban vege- 2016. Characterization of allergen emission sources in urban areas. J. Environ. Qual.
tation is usually planted to mitigate high outdoor air temperature, but its 45, 244–252.
Chen, L., Liu, C., Zhang, L., Zou, R., Zhang, Z., 2017. Variation in tree species ability to
impact on air quality is not jointly analysed. In these scenarios, both ve-
capture and retain airborne fine particulate matter (PM2.5). Sci. Rep. 7, 3206.
getation effects (i.e. on air quality and thermal comfort) should be jointly Dalpé, B., Masson, C., 2009. Numerical simulation of wind flow near a forest edge. J.
investigated in order to obtain improvements on thermal comfort without Wind Eng. Ind. Aerodyn. 97, 228–241.
Defraeye, T., Verboven, P., Derome, D., Carmeliet, J., Nicolai, B., 2013. Stomatal tran-
a deterioration of air quality of the area. spiration and droplet evaporation on leaf surfaces by a microscale modelling ap-
proach. Int. J. Heat Mass Transfer 65, 180–191.
6.2. Future requirements in experiments and CFD modelling Di Sabatino, S., Buccolieri, R., Pappaccogli, G., Leo, L.S., 2015. The effects of trees on
micrometeorology in a real street canyon: consequences for local air quality. Int. J.
Environ. Pollut. 58, 100–111.
The developed parameterizations of trees (and urban vegetation in Dupont, S., Brunet, Y., 2008. Edge flow and canopy structure: a large-eddy simulation
study. Bound. Layer Meteorol. 126, 51–71.
general) are appropriate for taking into account deposition and the EEA, European Environment Agency, 2015. Air Quality in Europe—2015 Report. EEA
aerodynamic effects on air quality in urban areas. In fact, validation Report No 5/2015. EEA, Copenhagen, Denmark. Available online at: http://www.
studies, mostly performed using wind tunnel experiments, have shown eea.europa.eu//publications/air-quality-in-europe-2015.
Endalew, A.M., Hertog, M., Gebrehiwot, M.G., Baelmans, M., Ramon, H., Nicolai, B.M.,
a good performance of CFD modelling provided that the appropriate Verboven, P., 2009. Modelling airflow within model plant canopies using an in-
turbulence closures and tuning of some parameters are employed. tegrated approach. Comput. Electron. Agric. 66, 9–24.

63
R. Buccolieri et al. Urban Forestry & Urban Greening 37 (2019) 56–64

Escobedo, F.J., Nowak, D.J., 2008. Spatial heterogeneity and air pollution removal by an Moonen, P., Gromke, C., Dorer, V., 2013. Performance assessment of large eddy simu-
urban forest. Landsc. Urban Plann. 90, 102–110. lation (LES) for modelling dispersion in an urban street canyon with tree planting.
Foudhil, H., Brunet, Y., Caltagirone, J.-P., 2005. A k – ε model for atmospheric flow over Atmos. Environ. 75, 66–76.
heterogeneous landscapes. Environ. Fluid Mech. 5, 247–265. Moradpour, M., Afshin, H., Farhanieh, B., 2017. A numerical investigation of reactive air
Freer-Smith, P.H., Beckett, K.P., Taylor, G., 2005. Deposition velocities to Sorbus aria, pollutant dispersion in urban street canyon with tree planting. Atmos. Pollut. Res. 8,
Acer campestre, Populus deltoids X trichocarpa ‘Beaupre’, Pinus nigra and X 253–266.
Cupressocyparis leylandii for coarse, fine and ultra- fine particles in the urban en- Nazarian, N., Kleissl, J., 2016. Realistic solar heating in urban areas: air exchange and
vironment. Environ. Pollut. 133, 157–167. street-canyon ventilation. Build. Environ. 95, 75–93.
Gallagher, J., Baldauf, R., Fuller, C.H., Kumar, P., Gill, L.W., McNabola, A., 2015. Passive Nowak, D.J., Crane, D.E., Stevens, J.C., 2006. Air pollution removal by urban trees and
methods for improving air quality in the built environment: a review of porous and shrubs in the United States. Urban For. Urban Green. 4, 115–123.
solid barriers. Atmos. Environ. 120, 61–70. Nowak, D.J., Hirabayashi, S., Bodine, A., Hoehn, R., 2013. Modeled pm 2.5 removal by
Gill, S.E., Handley, J.F., Ennos, A.R., Pauleit, S., 2007. Adapting cities for climate change: trees in ten us cities and associated health effects. Environ. Pollut. 178, 395–402.
the role of the green infrastructure. Built Environ. 33 (1), 115–133. Nowak, D.J., Hirabayashi, S., Doyle, M., McGovern, M., Pasher, J., 2018. Air pollution
Green, S.R., 1992. Modellilng turbulent air flow in stand of widely-spaced trees. Phoenics removal by urban forests in Canada and its effect on air quality and human health.
J. 5, 294–312. Urban For. Urban Green. 29, 40–48.
Gromke, C., Blocken, B., 2015a. Influence of avenue-trees on air quality at the urban Peters, K., Eiden, R., 1992. Modelling the dry deposition velocity of aerosol particles to a
neighborhood scale. Part I: quality assurance studies and turbulent Schmidt number spruce forest. Atmos. Environ. 26, 2555–2564.
analysis for RANS CFD simulations. Environ. Pollut. 196, 214–223. Pugh, T.A.M., MacKenzie, A.R., Whyatt, J.D., Hewitt, C.N., 2012. Effectiveness of green
Gromke, C., Blocken, B., 2015b. Influence of avenue-trees on air quality at the urban infrastructure for improvement of air quality in urban street canyons. Environ. Sci.
neighborhood scale. Part II: Traffic pollutant concentrations at pedestrian level. Technol. 46, 7692–7699.
Environ. Pollut. 196, 176–184. Rasanen, J.V., Holopainen, T., Joutsensaari, J., Ndam, C., Pasanen, P., Rinnan, A., et al.,
Gromke, C., Ruck, B., 2012. Pollutant concentrations in street canyons of different aspect 2013. Effects of species-specific leaf characteristics and reduced water availability on
ratio with avenues of trees for various wind directions. Bound. Layer Meteorol. 144, fine particle capture efficiency of trees. Environ. Pollut. 183, 64–70.
41–64. Raupach, M.R., Bradley, E.F., Ghadiri, H., 1987. A Wind Tunnel Investigation into
Gromke, C., Buccolieri, R., Di Sabatino, S., Ruck, B., 2008. Dispersion study in a street Aerodynamic Effect of Forest Clearings on the Nesting of Abbott’s Booby on
canyon with tree planting by means of wind tunnel and numerical investigations Christmas Island Internal Report. CSIRO Centre for Environmental Mechanics,
evaluation of CFD data with experimental data. Atmos. Environ. 42, 8640–8650. Canberra 21 pp.
Gromke, C., Blocken, B., Janssen, W., Merema, B., van Hooff, T., Timmermans, H., 2015. Rosenfeld, M., Marom, G., Bitan, A., 2010. Numerical simulation of the airflow across
CFD analysis of transpirational cooling by vegetation: case study for specific me- trees in a windbreak. Bound. Layer Meteorol. 135, 89–107.
teorological conditions during a heat wave in Arnhem, Netherlands. Build. Environ. Roy, S., Byrne, J., Pickering, C., 2012. A systematic quantitative review of urban tree
83, 11–26. benefits, costs, and assessment methods across cities in different climatic zones.
Gromke, C., Jamarkattel, N., Ruck, B., 2016. Influence of roadside hedgerows on air Urban For. Urban Green. 11, 351–363.
quality in urban street canyons. Atmos. Environ. 139, 75–86. Salim, S.M., Chan, A., Cheah, S.C., 2011. Numerical simulation of atmospheric pollutant
Grote, R., Samson, R., Alonso, R., Amorim, J.H., Cariñanos, P., Churkina, G., Fares, S., Le dispersion in tree-lined street canyons: comparison between RANS and LES. J. Build.
Thiec, D., Niinemets, Ü., Mikkelsen, T.N., Paoletti, E., Tiwary, A., Calfapietra, C., Environ. 46, 1735–1746.
2016. Functional traits of urban trees in relation to their air pollution mitigation Salim, M.H., Schlünzen, K.H., Grawe, D., 2015. Including trees in the numerical simu-
potential: a holistic discussion. Front. Ecol. Environ. 14, 543–550. lations of the wind flow in urban areas: should we care? J. Wind Eng. Ind. Aerodyn.
Haaland, C., van den Bosch, C.K., 2015. Challenges and strategies for urban green-space 144, 84–95.
planning in cities undergoing densification: a review. Urban For. Urban Green. 14, Salmond, J.A., Tadaki, M., Vardoulakis, S., Arbuthnott, K., Coutts, A., Demuzere, M.,
760–771. Dirks, K.N., Heaviside, C., Lim, S., Macintyre, H., McInnes, R.N., Wheeler, B.W.,
Hong, B., Qin, H., Lin, B., 2018. Prediction of wind environment and indoor/outdoor 2016. Health and climate related ecosystem services provided by street trees in the
relationships for PM2.5 in different building-tree grouping patterns. Atmosphere urban environment. Environ. Health 15 (1), 36.
9, 39. Sanchez, B., Santiago, J.-L., Martilli, A., Palacios, M., Kirchner, F., 2016. CFD modeling of
Huang, C.W., Lin, M.Y., Khlystov, A., Katul, G., 2013. The effects of leaf area density reactive pollutant dispersion in simplified urban configurations with different che-
variation on the particle collection efficiency in the size range of ultrafine particles mical mechanisms. Atmos. Chem. Phys. 16, 12143–12157.
(UFP). Environ. Sci. Technol. 47, 11607–11615. Santiago, J.L., Martin, F., Martilli, A., 2013. A computational fluid dynamic modeling
Incropera, F.P., Lavine, A.S., DeWitt, D.P., 2011. Fundamentals of Heat and Mass approach to assess the representativeness of urban monitoring stations. Sci. Total
Transfer. John Wiley & Sons Incorporated. Environ. 454–455, 61–72.
Janhäll, S., 2015. Review on urban vegetation and particle air pollution—deposition and Santiago, J.L., Krayenhoff, E.S., Martilli, A., 2014. Flow simulations for simplified urban
dispersion. Atmos. Environ. 105, 130–137. configurations with microscale distributions of surface thermal forcing. Urban Clim.
Jeanjean, A.P., Hinchliffe, G., McMullan, W.A., Monks, P.S., Leigh, R.J., 2015. A CFD 9, 115–133.
study on the effectiveness of trees to disperse road traffic emissions at a city scale. Santiago, J.L., Martilli, A., Martin, F., 2017a. On dry deposition modelling of atmospheric
Atmos. Environ. 120, 1–14. pollutants on vegetation at the microscale: application to the impact of street vege-
Jeanjean, A.P.R., Monks, P.S., Leigh, R.J., 2016. Modelling the effectiveness of urban tation on air quality. Bound. Layer Meteorol. 162, 451–474.
trees and grass on PM2.5 reduction via dispersion and deposition at a city scale. Santiago, J.L., Borge, R., Martin, F., de la Paz, D., Martilli, A., Lumbreras, J., Sanchez, B.,
Atmos. Environ. 147, 1–10. 2017b. Evaluation of a CFD-based approach to estimate pollutant distribution within
Jeanjean, A., Buccolieri, R., Eddy, J., Monks, P., Leigh, R., 2017. Air quality affected by a real urban canopy by means of passive samplers. Sci. Total Environ. 576, 46–58.
trees in real street canyons: the case of Marylebone neighbourhood in central London. Santiago, J.L., Rivas, E., Sanchez, B., Buccolieri, R., Martin, F., 2017c. The impact of
Urban For. Urban Green. 22, 41–53. planting trees on NOx concentrations: the case of the Plaza De La Cruz neighborhood
Kalansuriya, C.M., Pannila, A.S., Sonnadara, D.U., 2009. Effect of roadside vegetation on in Pamplona (Spain). Atmosphere 8, 131.
the reduction of traffic noise levels. In: Proceedings of the Technical Sessions, 25, Sanz, C., 2003. A note on k −ε modeling of vegetation canopy air-flows. Bound. Layer
1–6. Institute of Physics, Sri Lanka, 2009. . Available online at: http://www. Meteorol. 108, 191–197.
academia.edu/2273952/Effect_of_roadside_vegetation_on_reduction_of_traffic_noise_ Seinfeld, J.H., Pandis, S.N., 2006. Atmospheric Chemistry and Physics: From Air Pollution
levels. to Climate Change. Wiley-Interscience ISBN-13: 978-0471720188 ISBN-10:
Katul, G.G., Mahrt, L., Poggi, D., Sanz, C., 2004. One-and two-equation models for canopy 0471720186.
turbulence. Bound. Layer Meteorol. 113, 81–109. Sini, J.-F., Anquetin, S., Mestayer, P., 1996. Pollutant dispersion and thermal effects in
Krayenhoff, E.S., Santiago, J.L., Martilli, A., Christen, A., Oke, T.R., 2015. urban street canyons. Atmos. Environ. 30, 2659–2677.
Parametrization of drag and turbulence for urban neighbourhoods with trees. Bound. UN-Habitat, 2016. Urbanization and Development: Emerging Futures. World Cities
Layer Meteorol. 156, 157–189. Report 2016. United Nations Human Settlements Programme (UN-Habitat), Nairobi,
Lalic, B., Mihailovic, D.T., 2004. An empirical relation describing leaf-area density inside Kenya. Available online at http://wcr.unhabitat.org/main-report.
the forest for environmental modeling. J. Appl. Meteorol. 43, 641–645. Van den Berg, M., Wendel-Vos, W., van Poppel, M., Kemper, H., van Mechelen, W., Maas,
Leung, D.Y.C., Tsui, J.K.Y., Chen, F., Yip, W.K., Vrijmoed, L.L.P., Liu, C.H., 2011. Effects J., 2015. Health benefits of green spaces in the living environment: a systematic
of urban vegetation on urban air quality. Landsc. Res. 36, 173–188. review of epidemiological studies. Urban For. Urban Green. 14, 806–816.
Li, Z.J., Miller, D.R., Lin, J.D., 1985. A first-order closure scheme to describe counter- Vos, P.E., Maiheu, B., Vankerkom, J., Janssen, S., 2013. Improving local air quality in
gradient momentum transport in plant canopies. Bound. Layer Meteorol. 33, 77–83. cities: to tree or not to tree? Environ. Pollut. 183, 113–122.
Lin, M.-Y., Khlystov, A., 2012. Investigation of ultrafine particle deposition to vegetation Vranckx, S., Vos, P., Maiheu, B., Janssen, S., 2015. Impact of trees on pollutant dispersion
branches in a wind tunnel. Aerosol Sci. Technol. 46, 465–472. in street canyons: a numerical study of the annual average effects in Antwerp,
Litschke, T., Kuttler, W., 2008. On the reduction of urban particle concentration by ve- Belgium. Sci. Total Environ. 532, 474–483.
getation—a review. Meteorol. Z. 17 (3), 229–240. White, E.J., Turner, F., 1970. A method of estimating income of nutrients in a catch of
Liu, J., Chen, J.M., Black, T.A., Novak, M.D., 1996. E – e modelling of turbulent air flow airborne particles by a woodland canopy. J. Appl. Ecol. 7, 441–461.
downwind of a model forest edge. Bound. Layer Meteorol. 77, 21–44. Wilson, J.D., Flesch, T.K., 1999. Wind and remnant tree sway in forest cutblocks. III. A
Livesley, S.J., McPherson, E.G., Calfapietra, C., 2016. The urban forest and ecosystem windflow model to diagnose spatial variation. Agric. Forest Meteorol. 93, 259–282.
services: impacts on urban water, heat, and pollution cycles at the tree, street, and Yang, B., Raupach, M., Shaw, R.H., Paw, U.K.T., Morse, A.P., 2006. Large eddy simulation
city scale. J. Environ. Qual. 45, 119–124. of turbulent flow across a forest edge Part I: flow statistics. Bound. Layer Meteorol.
Massman, W., 1987. A comparative study of some mathematical models of the mean wind 120, 377–412.
structure and aerodynamic drag of plant canopies. Bound. Layer Meteorol. 40, Yoshida, S., Ooka, R., Mochida, A., Murakami, S., Tominaga, Y., 2006. Development of
179–197. three dimensional plant canopy model for numerical simulation of outdoor thermal
Mochida, A., Tabata, Y., Iwata, T., Yoshino, H., 2008. Examining tree canopy models for environment. In: Proceedings of the Sixth International Conference on Urban Climate
CFD prediction of wind environment at pedestrian level. J. Wind Eng. Ind. Aerodyn. (ICUC 6). Goteborg, Sweden, June 12–16.
96, 1667–1677.

64