How to Improve the Liveability in Cities: The Effect of Urban Morphology and Greening on Outdoor Thermal Comfort

A considerable amount of literature has been published recently on the analysis of outdoor thermal comfort conditions in urban environments. These studies are presented in the following sections. In particular, this analysis was focused to identify the most used urban models and tools to evaluate the thermal comfort conditions together with the most common thermal comfort indexes.

2.1 Urban climate models and tools

In literature, there are many studies related to outdoor thermal comfort conditions and the methodologies and tools used to investigate it are quite complex and time consuming. Therefore, to offer a comprehensive overview on existing models and tools, the results of this work were synthetized in Table 1. Table 1 presents a schematic summary of published articles on the evaluation of outdoor urban thermal comfort. Studies were investigated taking into consideration: in-situ measurement campaigns; the application of various models and tools (such as ENVI-met, CitySim, Rayman, and SOLWEIG); and thermal comfort indexes. In particular, the most commonly used thermal comfort indexes are: the predicted mean vote (PMV), the predicted percentage dissatisfied (PPD), the physiologically equivalent temperature (PET), the standard effective temperature (SET), the universal thermal climate index (UTCI), the index of thermal stress (ITS). The use of thermal comfort models is often accompanied with the evaluation of local climate conditions and in-situ measurement campaigns to verify the goodness of the results obtained. In some studies, [5-9] ENVI-met, Rayman and SOLWEIG were used associated with buildings energy-use models such as CitySim or GIS-based models considering also local climate data collected with measurement campaigns [10-16]. In general, for thermal comfort analyses the most used tool is ENVI-met and in some works the goodness of the results is taken for granted [7, 13-15], while in others it is made a comparison with the results of Rayman and SOLWEIG [5, 17-27]. Furthermore, the most used thermal comfort indexes are PET, PMV, and UTCI. These, are the indexes that best represent the outdoor thermal comfort conditions at neighborhood scale.

2.2 Thermal comfort indexes

Thermal comfort is “the state of mind that expresses satisfaction with the thermal environment and is assessed by subjective evaluation” (ANSI/ASHRAE Standard 55:2020) that is influenced by physical, physiological, psychological and other variables. In this work, thermal comfort will be used to evaluate the livability of different urban environments as a function of their urban morphology, construction materials and presence of green and vegetation. Then, thermal comfort indexes consider physiological and psychological people characteristics that represent their average data [28-36].

Currently, there are several outdoor thermal comfort (OTC) tools to simulate human perception. In this work thermal indexes based on the human energy balance [37, 38] were investigated using ENVI-met, which is the most used tool to investigate outdoor thermal comfort conditions. According to the literature review [35, 36], the most used three indexes based on a human thermal balance for outdoor thermal comfort were investigated.

Table 1. Outdoor thermal comfort simulations: models, tools and indexes

* The mean radiant temperature T_mrt (in grey) is one of the main variables used to calculate thermal comfort indexes

The predicted mean vote (PMV) is based on a simple human thermal balance in controlled steady-state conditions and on seven-points thermal sensation scale. The PMV is evaluated with the heat exchanges between the human body and the environment with one-node model not considering the thermoregulatory mechanisms within the different layers of the body and it was adapted to outdoor conditions with Klima-Michel-Modell adding the short and longwave radiations fluxes and typical outdoor activities and clothing.

The Physiologically Equivalent Temperature (PET) is based on Munich Energy-balance Model for Individuals (MEMI) a two-node model with energy balance equations between the human body core, its skin and the outdoor environment in steady-state conditions. The PET is defined as the air temperature (without wind speed and solar radiation) at which the heat balance of the human body is maintained with the same core and skin temperatures under the actual environment conditions.

The Universal Thermal Climate Index (UTCI) is based on Fiala multi-node model (i.e. 15 nodes) and it can evaluate the physiological response of human body in quasi-transient conditions and in various outdoor thermal environments, including extreme weather conditions. UTCI can be defined as the air temperature causing the same response as actual conditions.

For all these three indexes PMV, PET and UTCI, one of the main variables to evaluate the thermal comfort condition is the mean radiant temperature T_mrt. The T_mrt is defined as the uniform temperature of an imaginary enclosure in which the radiant heat transfer from the human body is equal to the radiant heat transfer in the actual non-uniform enclosure (EN ISO 7726:2001). T_mrt has a strong influence on thermo-physiological comfort indices such as the PET or the PMV. Usually, the urban climate tools such as ENVI-met, SOLWEIG and CitySim simulate the T_mrt (in K) according to Eq. (1):

$T_{m r t}=\sqrt[4]{\frac{\sum_{i=1}^{6} W_{i} \cdot\left(a_{p} \cdot K_{i}+a_{l} \cdot L_{i}\right)}{a_{l} \cdot \sigma}}$     (1)

where: $W_{i}$ is a direction-dependent weighting factor, it depends on the position and orientation of the person; $K_{i}$ is the short-wave radiation fluxes received by a standard person; $L_{i}$ is the long wave radiation fluxes received by a standard person; $a_{p}$ is the absorption coefficient of a clothed human body for short-wave radiation (standard value 0.7) and $a_{l}$ is the absorption coefficient of a clothed human body for long-wave radiation (standard value 0.97); $\sigma$ is the Stefan–Boltzmann constant (5.67∙10^–8 Wm⁻²K⁻⁴).

The T_mrt (in K) can be also calculated using a GIS-based model [11] according to Eq. (2) that is a method used in CityComfort+ [34]. The input data were elaborated using GIS tools in order to simplify the simulation using few urban variables such as the sky view factor (SVF) and the urban canyon height-to-width ratio (H/W).

$T_{m r t}=\sqrt[4]{ \frac{\left(\begin{array}{c}a_{p} \cdot E_{s o l} \cdot F_{s o l \rightarrow p}+\varepsilon_{s k y} \cdot E_{s k y} \cdot F_{s k y \rightarrow p}+ \\ +\varepsilon_{u r b} \cdot E_{u r b} \cdot F_{u r b \rightarrow p}\end{array}\right)}{\sigma}}$     (2)

where: $a_{p}$  is the absorption coefficient of a clothed human body for short-wave radiation (i.e. standard value 0.7); $\varepsilon_{s k y}$  is emissivity of the sky (0-1) elaborated knowing the dewpoint temperature; $\varepsilon_{u r b}$ is the emissivity of surface material (0-1) [37-39]; $E_{\text {sol }}$  is the direct and diffuse solar irradiance (W/m²), the direct quota was calculated knowing the shadows as a function of the solar height in each hour and the urban canyon dimensions H/W [40, 41]; $H / W$  is the long-wave radiation intensity of the sky (W/m²) elaborated according to [8]; $E_{\text {urb }}$ is the long-wave radiation intensity of urban surfaces (W/m²) elaborated according to [8]; $F_{\text {sol } \rightarrow p}$  is the view factor between the short-wave sources and a person (0-1) elaborated with GIS tools; $F_{u r b \rightarrow p}$  is the view factor between the visible sky and a person (0-1) elaborated with GIS tools; $F_{u r b \rightarrow p}$  is the view factor between urban surfaces and a person (0-1) elaborated with GIS tools; $\sigma$  is the Stefan–Boltzmann constant (5.67∙10^–8 Wm⁻²K⁻⁴).

2.2.1 The Predicted Mean Vote (PMV)

The PMV is one of the most used thermal index and it is based on the thermal balance of human body and empirical data. The PMV depends on air temperature, relative humidity, air speed, metabolic rate, and clothing insulation. It is based on the balance according to EN ISO 7730:2005 standard “Ergonomics of the thermal environment - Analytical determination and interpretation of thermal comfort using calculation of the PMV and PPD indices and local thermal comfort criteria” that reports the personal parameters, physical parameters and the main equations used to evaluate the PMV index. ENVI-met calculates the PMV according to Fangers and German standards VDI 3787-2:2008 [42], and requires the following input: (i) meteorological parameters: air temperature, mean radiant temperature, water vapour pressure, and local wind speed; and (ii) personal settings: clothing insulation, mechanical energy production of the body, and mechanical work factor. Eq. (3) shows how ENVI-met calculates the PMV:

$\begin{array}{r}

P M V=\left[0.028+0.303^{\left(-0.036 \cdot \frac{M}{A_{D u}}\right)}\right] \cdot\left(\frac{H}{A_{D u}}-E_{d}-\right. \\

\left.E_{s w}-E_{r e}-L-R-C\right)

\end{array}$     (3)

where:

•  $\frac{M}{A_{D u}} \text { and } \frac{H}{A_{D u}}$ are used to quantify the body energy production where $A_{D u}$ is the skin surface area; the first term is energy production of the body related to 1 m² of skin and depends on the persons activity, the second term is the produced energy not used for mechanical work.
• $E_{d} \text { and } E_{s w}$ are used to quantify the skin water and the vapour exchange; the first term is the amount of vapour diffusing directly through the skin; the second term is cooling effect of liquid sweat evaporating from skin.
• $E_{r e} \text { and } L$ are used to quantify the energy exchange through breathing; the first term is the energy lost by humidifying the air in the respiratory system; the second term is the energy lost and/or gained through heat exchange with the breathed air within the body.
• $R \text { and } C$ are used to quantify the energy exchange at body surface; the first term is the radiative energy balance of the body and depends on the clothing thermal insulation, the temperature of the clothing, and the mean radiative temperature; the second term is the energy exchange through convection and depends on the clothing thermal insulation, the turbulent heat transfer coefficient for heat between clothing and air, the temperature of the clothing, the air temperature, and the wind speed considering the walking velocity of the person.

2.2.2 The Physiological Equivalent Temperature (PET)

The PET is a thermal comfort index calculated according to the Munich Energy Balance Model for Individuals (MEMI) [43] that is the energy balance for the human body indicated in Eq. (4). It depends on skin, body and clothing temperatures, and the sweat rate; these parameters are calculated taking into account all incoming and outgoing heat fluxes between human body and outdoor urban environment [40].

$M+W+C+R+E_{d}+E_{r e}+E_{s w}+S=0$     (4)

where: $M$ is the metabolic activity, $W$ is the physical work output, C is the convective heat flow, R is the net radiation of the body, $E_{d}$ is the amount of vapour diffusing directly through the skin, $E_{r e}$ is the energy lost by humidifying the air in the respiratory system, $E_{s w}$ is cooling effect of liquid sweat evaporating from skin, and S is the storage heat flow for heating or cooling the body mass.

Knowing the skin and core temperatures, the MEMI balance can be solved to evaluate PET that is the air temperature T_a with v = 0.1 m/s, vp = 12 hPa and T_mrt = T_a. ENVI-met calculates PET according to a new model introduced by [44]. The main improvements refer to: (i) turbulent exchange coefficients for heat and water vapour fluxes calculated with the internal air velocity and the outdoor value; (ii) the sweat rate and the amount of sweat on the skin is set to zero when starting to calculate the indoor environments.

2.2.3 The Universal Thermal Climate Index (UTCI)

According to Eq. (5), the UTCI is defined as the air temperature $T_{a}$ of the reference condition causing the same model response as actual conditions, solving a 6^th order polynomial regression as function of climate conditions:

$\begin{aligned}

U T C I=f\left(T_{a} ; T_{m r t} ; v_{a} ; v_{p}\right) \\

\quad=T_{a}+\text { Off set }\left(T_{a} ; T_{m r t} ; v_{a} ; v_{p}\right)

\end{aligned}$     (5)

where: the $\text { Off set }$ is the deviation of UTCI from air temperature, $T_{a} \text { and } T_{m r t}$ are the air and mean radiant temperatures, $v_{a}$ is the wind speed, $\mathcal{v}_{p}$ is the water vapour pressure.

In ENVI-met, UTCI index leads some limits into the calculations: (i) the wind speed is limited to a range of 0.5-17 m/s at 10 m height; (ii) the height reference of wind speed is 10 m, while the outdoor thermal comfort parameters are usually calculated at 1-2 m.

Thermal comfort simulations were carried out for four typical days. In ENVI-met, with the support of Bio-met, the T_mrt and three thermal comfort indices (PMV, PET and UTCI) were calculated at 1 meter above the ground for a person with 35 years old, male, weight of 75 kg, height of 1.75 m, taking into account the following input data: clothing insulation: 0.5 clo or 0.08 m²K/W (typical summertime clothing), 1 clo or 0.015 m²K/W (typical midseason clothing), 1.5 clo or 0.23 m²K/W (typical wintertime clothing); metabolic rate: 1.9 met or 110 W/m² (walking at 0.56 m/s).

4.1 Thermal comfort analysis at a neighborhood scale

In order to investigate the local climate conditions during summer, winter and mid-season, ENVI-met was used to describe the spatial distribution of the T_mr in different urban contexts. Figure 3 shows an example of the analysis made in each neighborhood. In particular, in Mediterraneo three points were selected (see Figure 1) to evaluate the outdoor thermal comfort both in different local climate conditions (winter and summer) and in three different areas (GR is a point located in green area surrounded by buildings, SQ is a point located in an open space (without shading), UC is a point located in area with concrete pavement area surrounded by buildings).

3.png

Figure 3. Hourly values of the T_mr in ‘Mediterraneo’

In Figure 3 it is possible to observe how the T_mrt vary depending on the location in the neighborhood. GR and UC points, which were surrounded by buildings, have lower temperatures in some hours of the day (mainly in the morning and the afternoon) due to shading. In addition, the variations of T_mrt depend on the reference season and in particular the greater incidence of shading in summer, where at 3 pm the temperature drops drastically and then rises again around 5 pm. By contrast, during the winter day there is no variation.

4.1.1 Mitigation strategies

One of the strategies to mitigate the urban heat island effects and improve the liveability and quality of urban environments is the use of greenings in built-up areas such as green roofs, vegetation areas and trees. There are many benefits from the use of greenery in neighborhoods: air temperature reductions, outdoor thermal comfort improvements, energy savings for space cooling, lower greenhouse gas emissions, better air quality and urban space liveability improvements, storm-water run-off reductions, and aesthetic and social benefits [46].

4-1.png

(a) Arquata

4-2.png

(b) Mediterraneo

Figure 4. Mitigation strategies: S_BAU and S_GREEN

Since urban variables significantly affect thermal comfort, two neighborhoods that had different values of urban parameters (see Table 2) have been selected to evaluate the effect of mitigation strategies in different urban environment: Mediterraneo and Arquata (Figure 4). Table 3 shows the increase (in %) of green surfaces and trees comparing the actual scenario (S_BAU) and the new one (S_GREEN) and the consequent improvements of outdoor thermal comfort conditions by evaluating PET, PMV and UTCI. The average daily variation of PET, PMV and UTCI for August 7^th, 2015 (from 8 am to 6 pm) for two points (in Arquata and in Mediterraneo UC points, see Figure 4) was shown in Table 3. It is possible to observe that greater improvement of the indexes are visible in Mediterraneo neighborhood due to more relevant mitigation action. However, the average daily variation of all the indexes improves, highlighting how urban greenery are effective in improving thermal comfort in urban environments.

Table 3. Analysis of outdoor thermal comfort condition using mitigation strategies

* In brackets the maximum difference between S_BAU and S_GREEN is indicated

Figure 5 shows how PET and UTCI change during the day by comparing the S_BAU and S_GREEN scenarios for different points in Arquata and Mediterraneo (see Figure 4). It is possible to observe an improvement of outdoor thermal comfort by using trees and green areas especially in Mediterraneo, where a large quota of green areas and roofs was inserted (see Table 3).

5-1.png

(a) Hourly values of PET (°C) at specific points (UC)

5-2.png

(b) Hourly values of UTCI (°C) at specific points (SQ)

Figure 5. Comparison between S_BAU and S_GREEN on August 7^th, 2015

6-1.png

(a) Mediterraneo: absolute difference UTCI at 1 pm

6-2.png

(b) Arquata: absolute difference PET at 5 pm

Figure 6. Comparison between S_BAU and S_GREEN on August 7^th, 2015

Figure 6 shows the absolute difference of UTCI and PET (in walking at 2 km/h) for August 7^th, 2015 in Mediterraneo and Arquata neighborhoods considering two scenarios (S_BAU and S_GREEN). According to the shape of the roofs and to the type of outdoor surfaces, green roofs, green areas and trees were used to improve the quality of urban environment. The analysis shows an improvement in external thermal comfort conditions as result of green mitigation strategies. In particular, there was a slight overall improvement throughout the neighborhood, and the main progresses were particularly visible where the new green areas have been inserted.

4.2 The effect of urban morphology on thermal comfort

The relation between outdoor thermal comfort conditions and urban morphology was investigated using two parameters: (i) the building density (BD) that is the ratio between the total volume of the buildings and the building block area; the higher the value, the denser the urban context; and (ii) the urban canyons height-to-width ratio (H/W) that is the ratio between the building height and the distance between buildings.

7.png

Figure 7. PET for August 7^th, 2015 in two courtyards (typical for the city of Turin) with H/W = 1.2 (continuous line) and H/W = 0.7 (dotted line)

In the first part of this work, the analysis of thermal comfort conditions with different type of outdoor surfaces and dimensions of urban canyon and courtyard (specific values of each point identified in the neighborhood) have been investigated. Figure 7 shows the T_mrt in two courtyards with H/W = 1.2 (continuous line) and H/W = 0.7 (dotted line) for August 7^th, 2015. With high values of H/W there was a greater shading, and PET values were slightly lower than H/W = 0.7. The type of pavement also affects outdoor thermal comfort; in fact, lower values of PET were obtained for with green areas.

8.png

Figure 8. Correlations between urban parameters (BD and H/W) and the mean radiant temperature (T_mrt)

In the second part, average values of the BD and H/W were calculated for each neighborhood, and Figure 8 shows the correlations between the urban parameters and the average T_mrt on August 7^th, 2015 for the six neighborhoods. In particular it is possible to observe that with high values of BD and H/W, which indicate low distance between buildings, the T_mrt tends to decrease, this because this is because the higher and closer the buildings are, the more shadows will be generated. This factor significantly affects mean radiant temperature and thermal comfort conditions. Furthermore, it must be noticed that Turin, compared to other high-urbanized contexts (i.e. New York or Singapore) has relatively low values of BD and H/W, therefore there is no pronounced canyon effect, which causes an increase in discomfort.