Microclimate design for open spaces: Ranking urban design effects on pedestrian thermal comfort in summer

Accepted Manuscript Title: Microclimate design for open spaces: Ranking urban design effects on pedestrian thermal comfort in summer Author: Angeliki Chatzidimitriou Simos Yannas PII: DOI: Reference:

S2210-6707(16)30083-X http://dx.doi.org/doi:10.1016/j.scs.2016.05.004 SCS 419

To appear in: Received date: Revised date: Accepted date:

8-2-2016 11-5-2016 12-5-2016

Please cite this article as: Chatzidimitriou, Angeliki., & Yannas, Simos., Microclimate design for open spaces: Ranking urban design effects on pedestrian thermal comfort in summer.Sustainable Cities and Society http://dx.doi.org/10.1016/j.scs.2016.05.004 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Microclimate design for open spaces: Ranking urban design effects on pedestrian thermal comfort in summer

Angeliki Chatzidimitrioua,* and Simos Yannasb

a

Aristotle University of Thessaloniki, School of Engineering, Thessaloniki, Greece,

[email protected] b

Architectural Association Graduate School, London, UK, [email protected]

*corresponding author: tel.: +302310824175, tel./fax.: +302310834371, e-mail address: [email protected], postal address: Katakalou 2, 54643, Thessaloniki, Greece.

Highlights 

Microclimate studies of an urban square and a courtyard in summer conditions



Simulations with selected computational tools tested against measured data



Assessment of the effects of geometry, materials, soil humidity, vegetation, water



Ranking design parameters based of their influence on pedestrian comfort indices



Significant impact of trees, soil humidity, albedo, spatial and temporal variations

Abstract The paper presents a study on the influence of urban morphology and urban design parameters such as street and building geometry, landscape elements including vegetation types, water surfaces and material properties and their effects

on pedestrian thermal comfort in cities. The data provided by the paper are based on simulations using selected computational tools (ENVI-met, RadTherm and Fluent) and performed for two typical urban spaces, a square and a courtyard. The paper focuses on summer conditions which include increasingly uncomfortable periods. It draws upon studies initiated in the city of Thessaloniki in northern Greece. However, the findings apply to many other cities with similar morphological characteristics and summer design conditions. The results are ranked according to the influence each of the design parameters considered can have on pedestrian thermal comfort. Spatial and temporal variations are highlighted. Special mention is given to the high impact of trees and soil humidity and the contrasting effects of pavement albedo. The paper provides data for use by urban designers in specifying appropriate microclimatic interventions to improve pedestrian comfort.

Keywords Urban design, microclimate, pedestrian thermal comfort, simulation of urban microclimates

1. Introduction The microclimates of cities depend significantly on their geographical location and local climate, but are also products of human activity modifying the built environment. The influence that urban design parameters can have on the microclimate of open spaces has been studied in many different urban contexts and climates. However, in most studies the effects of individual parameters are discussed separately, lacking a comparative basis to assess their relative importance as possible design options for a specific urban space. Geometric features of streets and courtyards such as aspect ratios, orientation, openings and canopies, have been examined in terms of their effects on comfort, energy fluxes and wind field in the

urban environment (Andrade and Alcoforado 2008; Ali-Toudert and Mayer 2007; Bottillo et al 2014; Chen et al 2012; Sharmin and Steemers 2013). Other studies report on the effects of building and pavement materials based on their physical properties (Santamouris 2012; Bougiatioti et al 2008; Doulos et al 2004; Asaeda and Ca 2000; Erell et al 2014). The influence of vegetation has been studied in urban parks, street trees, courtyards, roofs and walls (Shashua-Bar et al 2011; Ca, Asaeda and Abu 1998; Wong and Chen 2004; Kumakura et al 2013). The studies that consider more than one parameters influencing microclimate and comfort conditions focus mostly on geometric variations in combination with the presence of trees, “green” roofs and walls or with varying surface albedo (Berkovic, Yezioro and Bitan 2012; Lee Holst and Mayer 2013; Alexandri and Jones 2008; Emmanuel et al 2007; Shashua Bar, Tsiros and Hoffman 2012; Santamouris et al 2014). This paper provides a comparative assessment that takes account of the effects of multiple parameters on the urban microclimate in order to allow an evaluation of the relative influence of each design measure on a specific open space case. The analysis is based on simulations performed with various tools that focus on different aspects of the urban microclimate. Simulations are not expected to fully represent the complexity of the actual urban environment, which is integrated in monitoring studies; however they can provide comparable results for examining microclimate development in relation to design parameters. The choice of design parameters to consider depends on their expected effects on the environmental variables and on pedestrian comfort. There is a substantial literature on computational tools used by researchers to study microclimate conditions at different scales. CFD software has been used by Yuan and Ng 2014, Robitu et al 2006, Skote et al 2005, Santamouris et al 2012 among others. Other studies have reported using models such as SkyHelios (Matzarakis et al 2015), SOLWEIG (Lindberg, Holmer and Thorson 2008), Urban Canyon (Sanchez de la Flor and Alvarez Dominguez 2004), Town Energy Balance model (Lemonsu et al 2012),

CAT (Erell and Williamson 2006), TOWNSCOPE (Katzschner, Bosch and Rottgen 2003), as well as empirical models (Cadima 2005). The software used in this study included the microclimate simulation software ENVI-met (www.envi-met.com; Bruse 2004), the CFD software Fluent (http:/www.ansys.com/), the thermal transfer simulation model RadTherm

(http://www.thermoanalytics.com)

and

the

RayMan

model

(http:/www.urbanclimate.net/rayman/; Matzarakis, Rutz and Mayer 2007) for the calculation of thermal indices. Open spaces are vital constituents of the urban environment, hosting functions and activities that are essential to the character of a city and the quality of life it can provide to its inhabitants. The geometric proportions and materiality of open spaces affect their microclimates and functionality leading to significant social, economic and environmental outcomes. This paper presents the findings of a study that looked at the different design strategies for improving environmental conditions and pedestrian thermal comfort in outdoor spaces in summer. Trees and canopies, grass and water, spatial proportions and material properties were among the design strategies considered. The findings show that each of these design measures can have specific environmental effects that are both measurable and predictable; knowledge of these can help improve decision-making on the design of open spaces in the urban environment.

2. Research methods The paper draws upon measurements that were reported previously (Chatzidimitriou 2012; Chatzidimitriou and Yannas 2015), taken in Thessaloniki in northern Greece, a coastal city with deep street canyons and scarce open spaces. Table 1 and Figure 1 provide key data for the city’s temperate climate and dense urban environment. Data from these measurements were used for adjusting and testing the selected simulation software. The software were then used to perform the calculations reported in the present paper. The cases for which each software was

used, the output produced and the manner in which these were combined for this study are shown in Figure 2. The software selected for the study included ENVI-met for overall microclimate simulation of urban areas, Fluent for detailed characterisation of morphological parameters and air movement and RadTherm for the calculation of surface and mean radiant temperatures. The output of these tools is used as input in RayMan for the calculation of Physiological Equivalent Temperature PET (Hoppe 1999). PET values are used as the thermal indices for the assessment of pedestrian comfort in the studied open spaces, and are calculated based on environmental parameters such as air temperature and humidity, wind velocity, and short and longwave radiation fluxes summarised as the mean radiant temperature.

2.1. Base case Two typical forms of open space that are encountered in many cities were taken as generic cases for the study (Figure 3). One represents an urban square of 20m by 20m dimensions in plan, surrounded by streets 10m wide and buildings 20m high (H/W=2.0). The other represents the internal courtyard of an urban block, and is also 20m by 20m in plan and surrounded by buildings 20m high. The external surfaces of buildings and ground materials are assumed to have typical values as listed in Table 2, corresponding to materials such as medium density concrete, cement render, clay tiles, soft limestone etc (EN ISO 10456, 2007). These two urban elements were taken as representing realistic morphological configurations for Thessaloniki in Northern Greece (Figure 1). Earlier studies (Chrisomallidou Tsikaloudaki and Theodosiou 2002; Chatzidimitriou, Liveris, Bruse and Topli 2013) have reported on similar dimensions of urban areas where building heights vary between 2 and 8 storeys and most commonly in the range of 15m - 25m. Squares and courtyards can take many different shapes with areas ranging from 400m 2 to well over 2000m2. The spaces discussed here represent the smaller range in terms of floor area, but are quite typical in their geometries with common proportions and height to width ratios.

2.2. Study of design parameters Nine fundamental design parameters were studied as relating to morphology, material properties and the presence of trees, grass and water. The numerical values associated with each parameter are listed in Table 3. The morphological parameters are the aspect ratio (the ratio of building height H to the width W of street or open space) and the size and position of shading canopies (if any). Two additional cases of aspect ratio were considered for the square and the courtyard as variations of the base case: a wide square with H=10m and W=40m (H20W40), a deep square with H=20m and W=20m (H20W20), a wide courtyard with H=10m and W=20m (H10W20) and a deep courtyard with H=20m and W=10m (H20W10). Three canopy variants were examined as applicable to both the square and the courtyard: canopy covering of the North-West quarter of the spaces (indicated as 25% cover), canopy covering both NW and NE quarters (indicated as 50% cover), and canopy covering all four quarters (indicated as 100% cover). Canopies were considered as solid timber horizontal planes at 3m above ground level.

The material properties include the

albedo and emissivity of the pavement surface, the thermal capacity of the pavement material, and the moisture content of the soil. The presence of vegetation and water is indicated as a replacing 25%, 50% or 100% of pavement material by trees, grass or water surfaces (Figure 4). Leaf area density (LAD) profiles for trees and grass are listed in Table 4. The values assumed for material properties have been chosen to represent the range of common urban construction materials such as asphalt, concrete, bricks, marble, granite and stone.

2.3. Simulation tools assessment The software tools listed in Figure 2 were assessed against measured data taken in the summer period in open spaces in Thessaloniki (Chatzidimitriou and Yannas 2015).

Simulations of the monitored spaces revealed a good agreement

between measured and simulated results (Appendix; Chatzidimitriou 2012). For Fluent

and ENVI-met the differences between measured and simulated results did not exceed 20% for most of the variables tested. With ENVI-met the average differences were of 0.6 oC and 3.6 oC for air and surface temperatures respectively with highest differences of 2oC and 10 oC. With Fluent the average differences for air temperature and velocity were of 3.7 oC and 0.1m/s respectively, with highest differences of up to 7 oC and 0.2m/s). With RadTherm the differences were of no more than 10% for surface and globe temperatures. These correspond to averages of 1.6 oC and 0.3 oC respectively and highest differences of 7 oC and 5 oC. These comparisons underlined the

specific

advantages

and

drawbacks

of

each

tool

(Chatzidimitriou

Chrissomallidou and Yannas 2006; Chatzidimitriou 2012) thus helping to decide how best to combine these tools on the study of the parameters considered by the paper. The effects of vegetation, water and soil humidity were simulated with ENVImet v3.1 which couples atmosphere, soil and vegetation models and takes into account water surfaces. However, this model is limited in its account of morphologic complexity. Variations in surfaces and material properties were assessed with RadTherm which includes a detailed radiation model for the study of heat transfer between surfaces. This provided the best agreement with measured surface and globe temperatures. Geometric parameters and shading canopies were examined with both RadTherm and Fluent which can take into account complicated geometric models; the latter was also used to simulate airflow and obtain results for air velocity as well as temperature. The RayMan software (Matzarakis, Rutz and Mayer 2007) was applied with simulation outputs from the other software to calculate Physiological Equivalent Temperatures PET (Hoppe 1999) that provide an index which combines all the variables that have a strong measurable effect on pedestrian thermal comfort.

2.4. Simulations setup The geometric model for ENVI-met encompassed an area of 90m by 90m rising 50m in height with an additional surrounding area 19m wide (nesting grids) and a grid resolution of 1m horizontally and 2m vertically. Ground and building surface

properties are listed in Table 2 and the weather data driving the simulations are listed in Table 5. Indoor temperatures of surrounding buildings were assumed to be kept at 25oC. RadTherm simulations were run for a period of five consecutive days using hourly weather data. The model dimensions were of 80m by 80m by 20m with 1m grid resolution horizontally and vertically. Initial temperatures were set at 20 oC for all surfaces and natural convection was activated (as a heat transfer process) to take account of wind velocity. Ground surfaces were modeled as being composed of three layers with the top layer corresponding to pavement and bottom layer to soil. The software models both shortwave and longwave radiation. Radiation readings from six directions were used for the calculation of mean radiant temperatures (Matzarakis, Rutz and Mayer 2000). Hourly steady-state simulations were performed with Fluent. The basic 80m by 80m geometric model was set within a structured mesh of 600m by 1150m, 150m high providing appropriate distances for flow development (Franke et al 2004). Boundary conditions were set with velocity inlets to the north and top and a pressure outlet to the south, symmetry to east and west and adiabatic building surfaces. The input wind velocity was adjusted for the urban environment using the power law uz/uh=(z/h)a with a=0.4 (Givoni 1998). The k-ε RNG turbulence model was used and the Discrete Ordinates radiation model was combined with solar ray tracing for solar radiation calculations.

2.5. Weather data The weather data used for the simulations correspond to a very warm and sunny summer day for the city of Thessaloniki. Mean values for the selected day are given in Table 5. Similar conditions are encountered in many other cities around Southern Europe and the Mediterranean region. The paper focuses on summer conditions because clear sky and intense solar radiation in this period enhance

microclimate differentiation and urban heat island intensity creating conditions of discomfort for pedestrians. In winter, cloudiness and the overshadowing and wind reduction caused by the urban morphology lead to far more predictable environmental conditions that do not vary much outside the ambient temperature.

3. Results and Discussion The simulation studies and parametric analysis were aimed at a comparative evaluation of the effects different design parameters can have on microclimatic conditions and on thermal comfort of pedestrians in outdoor urban spaces. This was carried out by assessing the effect of each parameter on two typical forms of open urban spaces, a square and the internal courtyard of an urban block. For each of these spaces, hourly PET values were calculated at five different spots (Figure 3) representing five distinct areas of each open space. The results are examined separately for each spot and also area averaged to provide an overall PET value representing each design parameter considered. The results are listed in Tables 6, 7 and 8 according to their effects on the PET values. Results of a detailed analysis that includes daytime peaks and night-time values are summarised in Tables 6 and 7. In Table 8 PET values are expressed as averages of a daily cycle.

3.1. Urban Squares and Internal Courtyards Compared to a square, the internal courtyard of an urban block experiences more overshadowing from adjacent buildings, and thus lower MRT values which in turn translate into lower PET values. It has been reported by other studies that the MRT has a stronger effect on thermal comfort than the air temperature (Matzarakis and Mayer 2008; Emmanuel, Rosenlund and Johansson 2007). Consequently, as shown in this paper, varying the geometry of these spaces will have a stronger effect on thermal comfort in the courtyard than in the square. On the other hand, geometry also affects wind velocities and air temperatures with the result that in the square the former are higher and the air temperatures lower than in the courtyard.

3.2. Effect of aspect ratio The aspect ratio is inversely related to MRT and PET for both the square and the courtyard. Doubling of the aspect ratio (from H/W=1.0 to H/W=2.0) reduced the mean daily PET value by some 3.5 oC in the square and by close to 4.5 oC in the courtyard. At night, the relationship is reversed in the square; this is because a higher H/W ratio incurs reduced sky view and thus lower rate of heat dissipation to the night sky. For the square, a higher aspect ratio results in higher air velocity (probably due to channelling) as well as higher air temperature (due to reduced sky view), whereas reduction in aspect ratio may still result in higher air temperature owing to reduced airflow. For the courtyard, deviations from the base case aspect ratio of H/W=1.0 can be expected to reduce airflow and raise the air temperature. These results are consistent with findings by Alvarez, Sanchez and Molina (1998). Other geometrical features that may affect thermal conditions in urban squares and courtyards are the openings, recesses and extrusions of surrounding buildings including balconies and arcades.

3.3. Effect of canopies The addition of shading canopies has a significant effect; it improved thermal comfort conditions in both spaces reducing PET values by up to 5 oC. Canopies obstruct vertical airflow and air velocity is reduced. In the square air velocity may increase under the canopies due to channelling (Figure 5), but in the enclosed courtyard airflow is more restricted (Figure 6). In areas where some of the shading is already provided by the surrounding buildings, the addition of canopies may raise the air temperature by reducing sky view. Providing openings between the canopies can help alleviate this problem.

3.4. Effect of ground surface albedo Figures 7 and 8 illustrate the effect of ground albedo on surface temperature and MRT. High albedo keeps ground surfaces cooler, but the higher fraction of solar radiation that it reflects back is affecting other surfaces thus raising daytime MRT values. These effects of albedo were found to be more significant in the square than in the courtyard; this is due to the longer exposure of its ground surface to solar radiation. The simulations showed that an increase in ground albedo from 0.5 to 0.8 led to variation in surface temperature by some 5 oC. Similar results were observed in the course of fieldwork conducted for this study (Chatzidimitriou and Yannas 2015) and more widely in the literature (Santamouris et al 2012; Akbari, Pomerantz and Taha 2001; Taha 1997; Doulos, Santamouris and Livada 2004; Alexandri and Jones 2006). Higher ground albedo resulted in increased PET values by some 2 oC. Such effect was also noted in some studies (Johansson 2006; Pearlmutter, Berliner and Shaviv 2006; Oaka 2007) and was simulated in detail by Erell et al (2014). In combination with geometric features, the albedo of pavements and building surfaces has been found to be an important factor for the urban microclimate and was also reported as contributing to climate change (Akbari, Menon and Rosenfeld 2007). Unlike ground albedo, higher reflectivity of roofs and high level terraces does not affect the radiant environment at pedestrian level and can reduce thermal loads due to incident solar radiation.

3.5. Effect of ground surface emissivity The thermal emissivities of common construction materials are mostly in the range of 0.92-0.95 (Voogt and Oke 2003). Changes in emissivity within that range had practically no effect on MRT and PET as is also reported by other studies (Doulos, Santamouris and Livada 2004; Oke et al 1991).

3.6. Effect of pavement thermal capacity A 50% increase in the volumetric heat capacity of ground cover materials from 2000kJ/m3K to 3000kJ/m3K resulted in a negligible reduction in surface temperatures, MRT and PET of less than 0.1 oC. Hourly simulation results show the effect of such increase to be positive during daytime, reducing surface and radiant temperatures and improving comfort, but turning negative at night. Clearly, given that pavements are composed of a thin layer of material, normal variations in thickness or thermal capacity will have little effect compared to that of the soil layer below.

3.7. Effect of natural soil Replacement of hard pavements by bare soil on the ground surface was shown to lead to significant improvement in air and surface temperatures reducing mean PET values by some 3.5 oC in the square and 2 oC in the courtyard. Increase in soil moisture content can lead to further improvement. Low soil albedo (0.2) reduces MRT and PET compared to hard pavement (albedo of 0.5) while increase in soil moisture content influences both the ground surface temperature and the ambient temperature.

3.8. Effect of vegetation (grass and trees) Replacing hard pavement with soil and grass resulted in PET reduction by 4 oC for the square and 2.5 oC for the courtyard (Figures 10 and 11). The addition of trees had the largest effect on both microclimate and pedestrian comfort, reducing mean daily PET values by 10 oC in the square and by 5.5 oC in the courtyard. Partial cover of the open space with trees (25% and 50% cover) showed the improvement in air temperature and PET to be concentrated below the foliages, in the shade, while changes in other areas were less significant. With full tree cover, the improvement in air temperature was higher for the courtyard, 1.2 oC compared to 0.5 oC for the

square, Figure 9. Field measurements have shown similarly large ambient and surface temperature reductions in spaces with tree cover (Chatzidimitriou and Yannas 2015).

3.9. Effect of water surfaces Replacing a hard pavement with a water surface can lower PET by up to 5 oC in the square and 3 oC in the courtyard. The microclimatic effects of water derive from its low reflectivity, which results in lower MRT, its high thermal capacity which results in lower surface temperature, and the process of evaporation which lowers the ambient temperature. The effect of water surfaces was seen to be more significant for the square than for the internal courtyard (Figures 12-15). Larger water surfaces do not result in further improvement as their influence does not extend beyond the immediate environment. It is preferable to have a number of smaller water surfaces scattered in the open space instead of a single large one. Initial analytic work was undertaken with software that did not consider evaporative processes, which is why the results do not show reductions in ambient temperature. More recent calculations with a later version of the software that includes evaporation showed area averaged air temperature reductions of around 2oC and spot reductions mostly between 2oC and 4oC (Chatzidimitriou et al 2013).

4. Ranking design parameters The assessment of design strategies is based on the differences caused by each measure in comparison to the base case. The design parameters in the simulated cases have different effects at distinct areas of the open spaces and their influence also varies during the day. In order to take into account these temporal and spatial variations ranking is provided for spot differences, considering day, night and maximum values, and mean differences in terms of area averaged and daily averaged values of thermal indices.

On Tables 6, 7 and 8 the PET values obtained for each of the design options considered for this study are compared to the PET value calculated for the base case. The difference between these PET values provides the scale for ranking the results. The calculations for measures involving trees, grass, water surfaces and soil were performed with ENVI-met, all other were assessed with RadTherm. The two tools gave slightly different PET values (of less than 1 oC) when compared for the base case daily average; hourly values revealed larger differences.

4.1. Maximum effects on thermal comfort The maximum daily PET values obtained for the central spots of the square and courtyard are listed and ranked in Table 6 according to peak difference with the base case. On the square, the largest differences with the base case occur in the afternoon, between 1400 and 1530 hours. The most effective measure in terms of comfort improvement was the full cover by trees. This reduced the PET value by more than 24oC at 1500 hours. Similar improvement was produced by shading canopies and by having a higher aspect ratio thus indicating that the improvement was the result of solar protection. The next best results came from water surfaces, wet soil or grass which gave PET reductions of some 10 oC. Reductions of some 5 oC resulted from varying ground albedo or soil humidity. Dry soil had a positive effect on PET owing to lower albedo compared to base case pavement. This option achieved its lowest PET values at 0900 hours. Departures from the base case aspect ratio led to large increases in PET. For the courtyard, the highest differences occurred between 1200 and 1400 hours which were times when direct solar radiation entered into the space. The most influential parameter was the aspect ratio. The table shows the large differences in PET resulting from variations in aspect ratio. Other results in the courtyard are of a similar nature to those obtained for the square.

4.2. Day and night effects on thermal comfort (PET) Study of the hourly PET values area averaged for all reference spots on the square and courtyard models highlighted significant differences between daytime and nightime conditions (Table 7). For daytime the most effective measure for improving thermal comfort in both square and courtyard is to have maximum cover by trees. Water surfaces, wet soil, grass and shading canopies are the next best options resulting in varying PET reductions in square and courtyard. Higher aspect ratio improves conditions in the courtyard, whereas lower aspect ratio is detrimental. After sunset the effects of all of these measures are considerably reduced and sky view becomes the critical parameter for heat dissipation. Shading canopies and high aspect ratios increase night-time PET values in the square; their effect is lesser for the courtyard which is better protected from the sun during daytime and thus less dependent on radiative cooling at night. Indicative daytime effects for distinct reference spots of each open space are shown separately in figures 19 and 20.

4.3. Mean daily effects on thermal comfort The hourly PET values for each of the five reference spots on the square and courtyard were averaged to provide mean daily values which are listed in Table 8 for 23 design options considered for this study. For each of these options Table 8 also identifies which of the five reference spots gave the largest daily differences in PET values; the locations of these five spots were identified in Figure 3. For the square, it can be seen from Table 8 that the largest improvement, a reduction by 9-10 oC compared to the base case PET, resulted from the addition of trees. The next best options were provided by shading canopies covering at least half of the area or by water surfaces; either of these resulted in reductions of 5-6 oC. Use of grass or bare soil resulted in reductions of some 4 oC. At the other end, lower aspect ratios or increase in ground albedo led to higher PET values than the base case.

For the courtyard, as for the square, the most significant improvement was obtained from the addition of trees. The next best options were the shading canopies or a higher aspect ratio. Significant reductions in PET were also given by water surfaces or wet soil. Increase in ground albedo or lower aspect ratios than the base case were shown to have detrimental effects.

4.4. Microclimatic differences between reference spots The design of open spaces must meet a wide range of diverse and often conflicting demands and conditions. Environmental diversity can be achieved by different combinations of the design measures discussed above. An indication is provided by comparing the results obtained across the five reference spots of the square and courtyard models considered in this study (Figures 16, 17). The comparison at midday reveals that high spatial diversity of comfort is mainly caused by shading from surrounding buildings and by partial application of the design measures (canopies, vegetation and water surfaces). For the square, the largest differences in the PET values of the five spots result from configurations that have partial cover (25% and 50%) of trees and canopies (Figure 18). In the courtyard, a larger variety is observed, attributed to its aspect ratio (Figure 19). Regarding aspect ratio, the highest diversity is observed with H/W=1 in both square and courtyard; it decreases with lower ratio in the square due to total exposure of the open space and with higher aspect ratio in the courtyard due to almost full shading of the open space. When design measures are applied over the entire area of the square or courtyard spaces they will naturally reduce differentiation between the reference spots, except for the case of high albedo pavement in which the differences between exposed and shaded spots increase. At night the spatial variation of PET values across the spaces is low with differences of up to 2 oC in the square and 1 oC in the courtyard mostly where partial cover by trees and canopies is applied (Figure 20).

More diversity in thermal conditions encountered across a square or courtyard may increase pedestrians’ interest in these spaces or perhaps even the probability of comfort sensation for different people at different times of day or seasonally.

4.5. Combined effects Study of how individual parameters may affect microclimatic conditions and pedestrian comfort has provided indications of their relative importance. Figures 21 and 22 summarise the results of all the design options considered in this study and illustrates both spatial and temporal variations resulting from the application of individual or combined measures. Combinations of different values of the design parameters examined in the study may have complementary or overlapping effects on microclimatic conditions and pedestrian thermal comfort. Some indication of such effects is given by comparative examination of the simulation results summarised in the Tables and Figures. As an example, soil can complement tree shade in lowering PET, ambient and surface temperature values because of enhanced evaporation. Similar observation is reported in the literature (Shashua Bar, Pearlmutter and Erell 2009; 2011). Adding trees combines well with use of soil instead of paved surfaces and the effect is stronger than that provided by shading from canopies. The shading effect contributed by trees and canopies reduces with increasing aspect ratios as the surrounding buildings already provide some of the shading. Care is required not to obstruct airflow and night-time heat dissipation. In narrow enclosed courtyards it is desirable to have openings between the surrounding buildings to enhance airflow; these can be spaced so as not to increase solar exposure or inhibit cooling potential. Comparison between a narrow square, considered as a courtyard with openings, and an enclosed courtyard with the same aspect ratio (H/W=1) shows that large openings at the corners of the square lead to higher wind velocity but also higher ambient, surface and radiant temperatures resulting in poorer thermal comfort conditions. Finally, where the ground is shaded by canopies, trees or surrounding

buildings, varying the albedo of ground surfaces will have a negligible effect on surface temperatures and PET.

4.6. Overall effects of different design measures on thermal comfort The comparative study of the simulation results and ranking of potential design measures has provided the following observations: 

The aspect ratio has a strong effect on PET values; increase in aspect ratio leads to lower daytime PET values though this reverses at night.



The effect of aspect ratio is more significant for the courtyards than for squares.



Trees and canopies have a stronger influence on the microclimate of squares than on those of internal courtyard.



Water surfaces, wet soil and grass produce similar effects for the square and the courtyard, but at different times of the day; over the daily cycle their effect is more significant for the square probably owing to its longer exposure to solar radiation.



Changes in the properties of ground surfaces and materials have a slightly higher effect for the square over the daily cycle, but the maximum effects are similar for square and courtyard.



Natural soil on the ground has a positive effect for the open square even when the soil is dry.



The effect of surface albedo reverses at night.

4.7. Range of application This study was initiated with measurements that were taken in a particular urban environment in Northern Greece, but the urban forms and weather data used to drive the simulations reported in the paper are commonly encountered in many

other cities around Europe as well as elsewhere. As such, the evaluation of design parameters and results presented in this paper are expected to apply quite widely within the limitations of the modelling assumptions and calculation accuracy of the computational tools employed. The simulations for this study were performed for summer conditions. During winter both courtyards and squares experience a considerable amount of overshadowing which results in lesser differences between radiant and ambient temperatures and thus in smaller microclimatic variations. The winter effects of the design measures assessed above are currently subjects for further research.

5. Conclusion Urban features such as built form and street geometry, trees, vegetation, water elements and pavement materials, shape the microclimates of open spaces and influence pedestrian thermal comfort and the use of these spaces by pedestrians. The individual and combined effects of these features can be predicted and taken into account by architects and urban designers from the earliest design stages. In this paper these effects were considered for typical urban open spaces using specialised simulation tools. The results of the parametric analysis of the 35 variants considered for the square and courtyard models provide a ranking of design parameters and their likely effect on environmental conditions and pedestrian thermal comfort. The urban forms and weather data used for the simulations can be considered as representing hot summer conditions in many cities in which the combination of high air temperature, moderate wind velocity and intense solar radiation can lead to extreme levels of daytime discomfort in outdoor spaces. The simulation results for individual variable were correlated and compared with numerous studies in the literature as well as with the measurements taken as part of this study.

Appendix

Simulation software evaluation The evaluation of the simulation tools used in this study was performed by comparing monitored data at six urban open areas with simulation results for air, surface and globe temperature and wind velocity from models of the same urban areas. The evaluation was made in several stages which included adjusting the simulation setup and the input parameters for each software. Details of the study sites and simulation models as well as indicative results of the comparison are given in the following tables and figures. Table 9 presents the six sites and the days of monitoring and simulations for each one. Figure 23 presents the three geometric models used for the simulations of Site A, and figures 24-26 present simulation results in site A with notes of the measured data at four spots on the site. More detailed comparison of daily averaged and hourly data from Site A are presented in tables 10 and 11. The comparison for all sites shows a good agreement between ENVI-met results and measured data of air and surface temperatures, especially in the cases of vegetation, and between Radtherm results and measured data of surface and globe temperatures. Discrepancies between simulation results and measured values are attributed to simplifications of the geometric models, the climate data and other input parameters that differ from the actual complex urban environment. However the achieved agreement is considered adequate to allow the use of the models for comparative examination of specific parameters.

References Akbari, H., S. Menon and A. Rosenfeld (2007). Global cooling: effect of urban albedo on global temperature. 2nd PALENC Conference and 28th AIVC Conference on Building Low Energy Cooling and Advanced Ventilation Technologies in the 21st Century, September 2007, Crete, Greece.

Akbari, H., M. Pomerantz and H. Taha (2001). Cool surfaces and shade trees to reduce energy use and improve air quality in urban areas. Solar Energy Journal Vol. 70 No 3 pp295-310, Elsevier Science Limited. Alexandri E. and P. Jones (2008). Temperature decreases in an urban canyon due to green walls and green roofs in diverse climates. Building and Environment vol. 43, pp480-493 Alexandri, E. and P. Jones (2006). Ponds, GreenRoofs, Pergolas and High Albedo Materials; Which cooling technique for urban spaces. In proceedings of the 23rd Conference of Passive and Low Energy Architecture, Geneva, Switzerland Ali-Toudert, F. and H. Mayer (2007). Thermal comfort in an east-west oriented street canyon in Freiburg (Germany) under hot summer conditions. Theoretical and Applied Climatology vol. 87, pp223-237 Alvarez, S., F. Sanchez and J.L. Molina (1998). “Airflow Patterns at Courtyards” in Environmentally Friendly Cities, Proceedings of PLEA ’98. James & James Science Publishers Ltd, London, pp503-506. Andrade, H. and M.J. Alcoforado (2008). Microclimatic variation of thermal comfort in a district of Lisbon (Telheiras) at night. Theoretical and Applied Climatology vol. 92, pp225-237 Berkovic, S., Yezioro, A. and A. Bitan 2012 "Study of thermal comfort in courtyards in a hot arid climate", Solar Energy 86: 1173–1186. Bottillo, S. A. De Lieto Vollaro, G. Galli and A. Vallati (2014). Fluid dynamic and heat transfer parameters in an urban canyon. Solar Energy vol.99, pp1-10, http://dx.doi.org/10.1016/j.solener.2013.10.031 Bougiatioti, F., E. Evangelinos, G. Poulakos and E. Zacharopoulos (2008). The summer thermal behaviour of "skin" materials for vertical surfaces in Athens, Greece as a decive parameter for their selection. Solar Energy 83, 4, pp582-598, doi:10.1016/j.solener.2008.06.012 Bruse, M. (2004). ENVI-met v.3.0: Updated Model Overview, http://www.model.envimet.com/hg2e/doku.php?id=resources:elib Ca, V.T., T. Asaeda, and E.M. Abu (1998). Reduction in air conditioning energy caused by a nearby park. Journal of Energy and Buildings 29, pp83–92. Cadima, P. (2005). Form and Microclimate: Spatial Confinement and Heat in the Urban Patio. In proceedings of the 22nd Conference of Passive and Low Energy Architecture, Beirut, Lebanon, November 2005. Chatzidimitriou, A and S. Yannas (2015). Microclimate development in open urban spaces: The influence of form and materials. Energy and Buildings 108 (2015) 156-174. http://dx.doi.org/10.1016/j.enbuild.2015.08.048

Chatzidimitriou, A., P. Liveris, M. Bruse and L. Topli (2013). Urban redevelopment and microclimate improvement: A design project in Thessaloniki, Greece. In proceedings of the 29th PLEA Conference, Sustainable Architecture for a Renewable Future. Munich, Germany. Chatzidimitriou, A., N. Chrisomallidou and S. Yannas (2006). Ground Surface Materials and Microclimates in Urban Open Spaces. In proceedings of the 23rd PLEA conference, Clever design, affordable comfort: a challenge for low energy architecture and urban planning. Geneva, Switzerland. Chatzidimitriou, A. (2012). Evaluation of the effects of design parameters on microclimate development of urban open spaces and on thermal comfort conditions, during the summer period in Mediterranean climate. PhD Thesis (in Greek), Civil Engineering Department, Polytechnic School, Aristotle University of Thessaloniki. http://thesis.ekt.gr/thesisBookReader/id/27035#page/1/mode/2up Chen, L., E. Ng, X. An, C. Ren, M. Lee, U. Wang and Z. He (2012). Sky view factor analysis of street canyons and its implications for daytime intra-urban air temperature differentials in high-rise, high-density urban areas of Hong Kong: a GIS-based simulation approach. International Journal of Climatology Vol.32 pp121–136, doi: 10.1002/joc.2243 Chrisomallidou, N., K. Tsikaloudaki and T. Theodosiou (2002). Quality of life and open spaces: A survey of microclimate and comfort in outdoor urban areas. In proceedings of 19th PLEA Conference, Toulouse, France. Doulos, L., M. Santamouris and I. Livada (2004). Passive cooling of outdoor urban spaces. The role of materials. Solar Energy Volume 77, Issue 2, pp 231-249 Eliasson, I. (1996). Urban nocturnal temperatures, street geometry and land use. Atmospheric Environment Vol. 30(3), pp379-392 Emmanuel, R., H. Rosenlund and E. Johansson (2007). Urban shading - a design option for the tropics? A study in Colombo, Sri Lanka. International Journal of Climatology vol. 27, pp1995-2004 EN ISO 10456 (2007). Building materials and products - Hygrothermal properties Tabulated design values and procedures for determining declared and design thermal values (ISO 10456:2007). Erell, E., D. Pearlmutter, D. Boneh, P.B. Kutiel (2014). Effect of high-albedo materials on pedestrian heat stress in urban street canyons. Urban Climate vol. 10 part 2, pp 367-386, http://dx.doi.org/10.1016/j.uclim.2013.10.005 Erell, E. and T. Williamson (2006). Simulating air temperature in an urban street canyon in all weather conditions using measured data at a reference meteorological station. International Journal of Climatology Vol.26, pp1671-1694

Franke, J., C. Hirsch, A.G. Jensen, H.W. Krus, M. Schatzmann, P.S. Westbury, S.D. Miles, J.A. Wisse and N.G. Wright (2004). Recommendations on the use of CFD in predicting pedestrian wind environment. COST Action C14 "Impact of Wind and Storms on City Life and Built Environment. Working group 2 - CFD techniques. 17.05.2004. Version 1.0 Gerolympou, A., N. Kalogirou, K. Kafkoula, N. Papamichos and V. Hastaoglou (1994). Prototype project for the development and revitalisation of the historic centre of Thessaloniki. EE, Region of Central Macedonia, Technical Chamber of Greece, Organisation of Planning and Environmental Protection of Thessaloniki. Givoni, B. (1998). Climate considerations in building and urban design. John Wiley & Sons Inc. Hoppe, P. (1999). The Physiologiacal equivalent temperature - a universal indexfor the biometeorological assessment of the thermal environment. International Journal of Biometeorology Vol. 43, pp71-75 Johansson, E. (2006). Urban Design and Outdoor Thermal Comfort in Warm Climates. Studies in Fez and Colombo. PhD Thesis. Housing Development & Management, Architecture & Built Environment, Lund Institute of Technology, Sweeden. Katzshner, L., U. Bosch and M. Rottgen (2003). A methodology for bioclimatic microscale mapping of open spaces. Fifth Intrnational Conference on Urban Climate, Lodz, 2003. Proceedings, Vol. 2, pp61-65 Kumakura, E., K. Nakaohkubo, A. Hoyano (2013). Solar Shading Effects by Tree Species in an Urban Environment Using Numerical Simulation Tool. In proceedings of the 29th PLEA Conference, Sustainable Architecture for a Renewable Future. 10-12 September 2013, Munich, Germany. Lee, H., J. Holst and H. Mayer (2013). Modification of Human-Biometeorologically Significant Radiant Flux Densities by Shading as Local Method to Mitigate Heat Stress in Summer within Urban Street Canyons. Advances in Meteorology Volume 2013, Article ID 312572, http://dx.doi.org/10.1155/2013/312572 Lemonsu, A., V. Masson, L. Shashua-Bar, E. Erell and D. Pearlmutter (2012). Inclusion of vegetation in the Town Energy Balance model for modelling urban green areas. Geoscientific Model Development, 5, 1377–1393, 2012, doi:10.5194/gmd-5-1377-2012 Lindberg, F., B. Holmer and S. Thorsson (2008). SOLWEIG 1.0 - Modelling spatial variations of 3d radiant fluxes and mean radiant temperature in complex urban settings. International Journal of Biometeorology Vol.52 pp697-713

Matzarakis, A., D. Frohlich, M. Gangwisch, C. Ketterer and A. Peer (2015). Developments and applications of thermal indices in urban structures by RayMan and SkyHelios model. ICUC9 9th International Conference on the Urban Climate jointly with the 12th Symposium on the Urban Environment. Matzarakis, A. and H. Mayer (2008). Dependence of the thermal urban climate on morphological variables. Ber. Meteor. Inst. Univ. Freiburg 17, 129-139 Matzarakis, A., F. Rutz and H. Mayer (2007). Modelling radiation fluxes in simple and complex environments—application of the RayMan model. International Journal of Biometeorology. Vol. 51, pp323-334 Matzarakis, A., F. Rutz and H. Mayer (2000). Estimation and calculation of the mean radiant temperature within urban structures. In Biometeorology and Urban Climatology at the Turn of the Millenium (ed.by R.J. de Dear, J.D. Kalma, T.R. Oke and A. Auliciems): Selected Papers from the Conference ICB-ICUC '99, Sydney, WCASP-50, WMO/TD No 1026, 273-278 Meir, I.A., (2000). “Courtyard microclimate, a hot arid region case study”. In Architecture City Environment, Proceedings of PLEA 2000. James & James Science Publishers Ltd, London, pp 218-223. Nikolopulou, M., N.Chrisomallidou, K. Steemers, R. Compagnon, J. Kang, N. Kofoed,G. Scudo, and L. Katzschner (2004). Rediscovering the Urban Realm and Open Spaces (RUROS), 2001-2003. Key Action 4, "City of Tomorrow and Cultural Heritage", from the program "Energy, Environment and Sustainable Development", within the Fifth Framework Program of the EU. ISBN: 960-869072-2 Oke, T.R. (1987) “Boundary Layer Climates” second ed. Methuen & Co, London. Oke, T.R, G.T. Johnson, D.G. Steyn, I.D.Watson (1991). Simulation of surface urban heat islands under ‘ideal’ conditions at night––Part 2: Diagnosis and causation. Boundary Layer Meteorology vol. 56, pp339–358. Ooka, R. (2007). Recent development of assessment tools for urban climate and heat-island investigation especially based on experiences in Japan. International Journal of Climatology vol. 27, pp1919-1930 Pearlmutter, D., P. Berliner and E. Shaviv (2006). Physical modeling of pedestrian energy exchange within the urban canopy. Building and Environment Vol. 41, Isssue 6, pp783-795 Robitu, M., M. Musy, C. Inard and D. Groleau (2006). Modeling the influence of vegetation and water pond on urban microclimate. Solar Energy vol. 80 pp435-447

Sánchez de la Flor, F. and S. Alvarez Domínguez (2004). Modelling microclimate in urban environments and assessing its influence on the performance of surrounding buildings. Energy and Buildings Vol. 36, Issue 5, pp 403-413 Santamouris, M. (2014). Cooling the cities - A review of reflective and green roof mitigation technologies to fight heat island and improve comfort in urban environments. Solar Energy 103 (2014) 682-703. http://dx.doi.org/10.1016/j.solener.2012.07.003 Santamouris, M., N. Gaitani, A. Spanou, M. Saliari, K. Gianopoulou, K. Vasilakopoulou, T. Kadromateas (2012). Using cool paving materials to improve microclimate of urban areas - Design realisation and results of the Flisvos project. Building and Environment 53 pp128-136. doi:10.1016/j.buildenv.2012.01.022 Sharmin, T. and K. Steemers (2013). Effect of Canyon Geometry on Outdoor Thermal Comfort: a case study of high-density, warm-humid climate. In proceedings of the 29th PLEA Conference, Sustainable Architecture for a Renewable Future. September 2013, Munich, Germany. Shashua-Bar, L., I. X. Tsiros, M. Hoffman (2012). Passive cooling design options to ameliorate thermal comfort in urban streets of a Mediterranean climate (Athens) under hot summer conditions. Building and Environment Volume 57, November 2012, Pages 110-119 Shashua-Bar, L., D. Pearlmutter and E. Erell (2009). The cooling efficiency of urban landscape strategies in a hot dry climate. Landscape and Urban Planning 92, pp179–186. doi:10.1016/j.landurbplan.2009.04.005 Shashua-Bar, L., D. Pearlmutter and E. Erell (2011). The influence of trees and grass on outdoor thermal comfort in a hot-arid environment. Int. J. Climatol. 31: 1498– 1506. DOI: 10.1002/joc.2177 Shashua-Bar, L. and M.E. Hoffman (2004). Quantitative evaluation of passive cooling of the UCL microclimate in hot regions in summer, case study: urban streets and courtyards with trees. Building and Environment Vol. 39, Issue 9, pp10871099 Skote, M., M. Sandberg, U. Westerberg, L. Claesson and A.V. Johansson (2005). Numerical and experimental studies of wind environment in an urban morphology. Atmospheric Environment Vol. 39 pp6147-6158 Taha, H. (1997). Urban climates and heat islands: albedo, evapotranspiration and anthropogenic heat. Energy and Buildings Vol. 25, pp99-103 Voogt, J.A. and T.R. Oke (2003) Thermal remote sensing of urban climates. Remote Sensing of Environment Vol. 86 pp370-384. Wong, N.H. and Y. Chen (2004). The thermal effects of city greens on surroundings under the tropical climate. In Built Environments and Environmental Buldings,

proceedings of the 21st international conference of Passive and Low Energy Architecture, Eindhoven, The Netherlands, 10-22 September 2004 Yuan, C. and E. Ng (2014). Practical Application of CFD on environmentally sensitive architectural design at high density cities: A case study in Hong Kong. Urban Climate 8 (2014) 57-77. http://dx.doi.org/10.1016/j.uclim.2013.12.001

Figure 1. Plan of the city centre of Thessaloniki, indicating the size and form of open spaces (source: Gerolympou et al 1994).

Figure 2. Descriptive diagram of the parametric analysis process

Figure 3. Geometric form of the base case of the two types of open spaces examined in the analysis: a. square and b. courtyard

Figure 4. Indication of the location of trees in three cases of tree cover at the square (a, b, c) and the courtyard (d, e, f).

a

b

Figure 5. Vectors of wind velocity and direction in the open square at 1800hours, at 1m above ground level. a. Base case (H20W40) b. 100% cover by shading canopies (simulation results by Fluent). Flow direction is from the north (top of figure) and velocity is 1.65m/s (met. station data at 1800hours corrected for urban surface roughness using power law model)

a

b

Figure 6. Vectors of wind velocity and direction in the courtyard at noon on the N-S central cross section: a. Base case b. 100% cover by shading canopies (simulation results by Fluent). The arrow indicates the point below the slot between canopies on the south part of the courtyard.

Figure 7. Hourly surface temperature values at the central point of the square: simulation results for seven values of ground surface albedo (including base case) by Radtherm software. The examined cases include albedo values 0.8, 0.7, 0.6, 0.5, 0.4, 0.3 and 0.2 indicated as a08, a07, a06, BASE CASEa05, a04, a03 and a02 respectively.

Figure 8. Hourly mean radiant temperature (MRT) values at the central point of the square: simulation results for seven values of ground surface albedo (including base case) by Radtherm software. The examined cases include albedo values 0.8, 0.7, 0.6, 0.5, 0.4, 0.3 and 0.2 indicated as a08, a07, a06, BASE CASEa05, a04, a03 and a02 respectively.

Figure 9. Simulation results for air temperature at noon (1200 hours), in the square, 1m above ground level, in a. the base case and b. the case of 100% cover by trees (simulation results by ENVI-met).

Figure 10. Simulation results for ground surface temperature in the afternoon (1500 hours), at the square, in a. the base case and b. the case of 100% cover by grass (simulation results by ENVI-met).

Figure 11. Simulation results for ground surface temperature in the afternoon (1500 hours), at the courtyard, in a. the base case and b. the case of 100% cover by grass (simulation results by ENVI-met).

Figure 12. Comparative diagram of hourly surface temperatures at the central point of the square in the cases of pavement (BASECASE.H20W40), natural soil with relative humidity 50%, 25% and 75% (SOIL50, SOIL 25 and SOIL75 respectively), full cover by trees, grass and water surfaces (TREES100, GRASS100 and WATER100 respectively): simulation results by ENVI-met software.

Figure 13. Comparative diagram of hourly mean radiant temperatures at the central point of the square in the cases of pavement (BASECASE.H20W40), natural soil with relative humidity 50%, 25% and 75% (SOIL50, SOIL 25 and SOIL75 respectively), full cover by trees, grass and water surfaces (TREES100, GRASS100 and WATER100 respectively): simulation results by ENVI-met software.

Figure 14. Comparative diagram of hourly surface temperatures at the central point of the courtyard in the cases of pavement (BASECASE.H20W20), natural soil with relative humidity 50%, 25% and 75% (SOIL50, SOIL 25 and SOIL75 respectively), full cover by trees, grass and water surfaces (TREES100, GRASS100 and WATER100 respectively): simulation results by ENVI-met software.

Figure 15. Comparative diagram of hourly mean radiant temperatures at the central point of the courtyard in the cases of pavement (BASECASE.H20W20), natural soil with relative humidity 50%, 25% and 75% (SOIL50, SOIL 25 and SOIL75 respectively), full cover by trees, grass and water surfaces (TREES100, GRASS100 and WATER100 respectively): simulation results by ENVI-met software.

examined case Albedo 0.8

PET values among five distinct areas (reference spots) within the square for each Tree cover 100%

H10W40

Albedo 0.2

1200 hours

Canopy 100%

Emissivity 0.70

Emissivity 0.95

Ther. Capacity 3000kJ/m3K

20

Ther. Capacity 1000kJ/m3K

simulation case

Grass cover 100%

Water cover 100%

Soil RH75%

SoilRH50%

Soil RH25%

H20W40

Grass cover 25%

Grass cover 50%

Water cover 25%

Canopy 25%

Canopy 50%

Water cover 50%

H20W20

Tree cover 25%

Tree cover 50%

max PET difference (oC) among reference spots

25 SQUARE 2300 hours

15

10

5

0

Figure 16. Diversity of comfort indices within the open spaces; highest difference of

COURTYARD 1200 hours

2300 hours

20

15

10

5

Tree cover 100%

H10W20

Water cover 50%

Water cover 100%

Canopy 100%

Tree cover 50%

Grass cover 50%

Grass cover 100%

Soil RH75%

SoilRH50%

Canopy 50%

Albedo 0.2

H20W10

Soil RH25%

Emissivity 0.70

Ther.Capacity 3000kJ/m3K

Canopy 25%

Emissivity 0.95

Ther. Capacity 1000kJ/m3K

Grass cover 25%

Water cover 25%

H20W20

Tree cover 25%

0 Albedo 0.8

max PET difference (oC) among reference spots

25

simulation case

Figure 17. Diversity of comfort indices within the open spaces; highest difference of PET values among five distinct areas (reference spots) within the courtyard for each examined case

Figure 18. PET values at noon in the square 1m above ground level, in a. the base case and b. the case of 50% cover by trees

Figure 19. PET values at noon in the courtyard 1m above ground level, in a. the base case and b. the case of 25% cover by trees

Figure 20. PET values at night in a. the square 1m above ground level, in the case of 50% cover by trees and b. the courtyard, 1m above ground level, in the case of 25% cover by trees

Figure 21. Cross sections of different cases of the square and the courtyard with indication of PET values at NW, NE and centre reference spots at noon (simulations results by ENVI-met).

Figure 22. Cross sections of different cases of the square and the courtyard with indication of PET values at NW, NE and centre reference spots at noon (simulation results by Radtherm).

a

b

c

Figure 23. Geometric models of one of the monitored sites (site a) in a. ENVI-met, b. RadTherm and c. Fluent

a Figure 24. Simulation results for air temperature at site a on the 29th of July and comparison with on site measurements at four spots with different ground surface materials; values at a.1400 hours, and b. 1700 hours

b

a

b

Figure 25. Simulation results for wind velocity at site a on the 29th of July and comparison with on site measurements at one spot (location with white marble on the pavement); values at a.1400 hours, b.1600 hours and c. 1800 hours

Figure 26. Simulation results for air temperature at site a on the 29th of July and comparison with on site measurements at four spots with different ground surface materials; values at 1600 hours

c

Table 1. Indicative winter and summer monthly average climate data for Thessaloniki (official meteor. station, http://www.hnms.gr/hnms/greek/climatology/climatology_region_diagrams_html?dr_ city=Thessaloniki_Mikra) min Air Temperature

month January July

avg Air Temperature

max Air Temperature

Relative Humidity

Wind Velocity

Wind Direction

1.3 oC

5.2 oC

9.3 oC

76.1%

2.98 m/s

NW

18.6 oC

26.6 oC

31.5 oC

53.2%

3.43 m/s

NW

Table 2. Base case material properties material properties base case

surface albedo 0.5

surface emissivity 0.9

density 2000 kg/m3

specific heat 1000 J/(kg∙K)

thermal conductivity 1 W/(m∙K)

Table 3.Analysis of design parameters Design parameter Aspect ratio (H/W ratio) Square Aspect ratio (H/W ratio) Courtyard Shading canopies Pavement surface Albedo Pavement surface Emissivity Pavement Thermal capacity (J/kgC) Soil humidity Tree cover Grass cover Water cover

base case 0.50 1.00 0% 0.50 0.90 2000 50%(pavement) 0% 0% 0%

examined cases 0.25 1.00 0.50 2.00 25% 50% 100% 0.20 0.30 0.40 0.70 0.75 0.80 1000 1500 2500 50% 25% 75% 25% 50% 100% 25% 50% 100% 25% 50% 100%

0.60 0.85 3000

0.70 0.95

0.80

Table 4. Leaf area density (LAD) profiles for trees and grass in ENVI-met vegetation type grass tree deciduous

height 0.5m 20m

LAD 1 0.30 0.00

LAD 2 0.30 0.25

LAD 3 0.30 0.50

LAD 4 0.30 1.00

LAD 5 0.30 1.50

LAD 6 0.30 2.00

LAD 7 0.30 2.00

LAD 8 0.30 2.00

LAD 9 0.30 1.75

LAD 10 0.30 0.00

Table 5. Climate data input for simulations with ENVI-met for 21st of July Initial Air Temperature 29.3oC

Relative Humidity 50%

Specific humidity 13g/kg

Wind Velocity 0.9m/s

Wind Direction SW

Solar adjustment factor 1.0

Cloud cover 0/0/0 oct.

Soil temperature 19.85oC

Table 6. Highest daily effects of design parameters on pedestrian comfort according to peak differences of PET values compared to base case scenario Square centre Simulation case Tree cover 100% H20W20 Canopy 100% Water cover 100% Water cover 50% Tree cover 50% Tree cover 25% Soil RH75% Grass cover 100% Grass cover 50% SoilRH50% Albedo 0.2 Soil RH25% Canopy 50% Canopy 25% Emissivity 0.70 Grass cover 25% Water cover 25% Thermal Capacity 3000kJ/m3K Thermal Capacity 1000kJ/m3K Emissivity 0.95 Albedo 0.8 H10W40 Courtyard centre Simulation case Tree cover 100% H20W10 Canopy 100% Water cover 100% Water cover 50% Tree cover 50% Soil RH75% Grass cover 50% Grass cover 100% SoilRH50% Soil RH25% Canopy 50% Albedo 0.2 Canopy 25% Emissivity 0.70 Tree cover 25% Water cover 25% Grass cover 25% Thermal Capacity 3000kJ/m3K Thermal Capacity 1000kJ/m3K Emissivity 0.95 Albedo 0.8 H10W20

peak PET +/- (oC) -24.4 -24.0 -23.0 -12.4 -12.3 -11.5 -9.9 -9.5 -9.0 -9.0 -8.7 -4.6 -4.3 -2.6 -2.0 -0.7 -0.6 -0.1 -0.1 +0.2 +0.3 +4.9 +23.9 peak PET +/- (oC) -23.5 -21.1 -21.0 -13.3 -13.2 -10.1 -10.1 -9.9 -9.8 -9.0 -5.6 -3.0 -1.9 -1.6 -0.5 -0.5 -0.1 -0.1 -0.1 +0.1 +0.2 +4.5 +24.6

time of peak PET +/15:00 14:00 14:00 12:00 12:00 9:00 15:00 12:00 12:00 12:00 12:00 14:00 09:00 10:30 14:00 14:00 15:00 15:00 14:00 14:00 14:00 14:00 15:30 time of peak PET +/13:30 13:00 13:00 12:00 12:00 11:00 13:30 12:00 12:00 13:30 12:00 13:00 13:30 13:00 13:30 6:00 12:00 12:00 13:30 13:30 13:30 13:00 14:00

PET (oC) 33.0oC 36.6oC 37.6oC 41.9oC 42.0oC 40.7oC 47.5oC 44.8oC 45.3oC 45.3oC 45.6oC 56.0oC 46.9oC 43.7oC 58 .6oC 59.9oC 56.7oC 57.3oC 60.5oC 60.8oC 60.9oC 65.5oC 58.0oC PET (oC) 36.3oC 31.4oC 31.5oC 44.1oC 44.2oC 46.7oC 49.7oC 47.5oC 47.6oC 50.8oC 51.8oC 49.5oC 51.3oC 50.9oC 52.7oC 21.7oC 57.3oC 57.3oC 53.1oC 53.3oC 53.4oC 57.0oC 59.5oC

Table 7. Day and night effects of design parameters on pedestrian comfort, according to area averaged differences of PET values compared to the base case scenario Square Simulation case Tree cover 100% Water cover 100% Soil RH75% Grass cover 100% Canopy 100% Soil RH50% H20W20 Albedo 0.2 Soil RH25% Emissivity 0.70 Ther. Capacity 1000kJ/m3K Ther. Capacity 3000kJ/m3K Emissivity 0.95 Albedo 0.8 H10W40 Courtyard Simulation case Tree cover 100% H20W10 Canopy 100% Water cover 100% Soil RH75% Grass cover 100% Soil RH50% Albedo 0.2 Emissivity 0.70 Ther. Capacity 1000kJ/m3K Emissivity 0.95 Ther. Capacity 3000kJ/m3K Soil RH25% Albedo 0.8 H10W20

area average PET +/for 1500 hours (oC) -23.5 -11.5 -9.3 -8.2 -7.5 -7.5 -1.7 -1.5 -1.2 -0.6 -0.2 +0.1 +0.2 +1.9 +12.5 area average PET+/for 1500 hours (oC) -8.2 -7.4 -5.4 -3.3 -2.9 -2.6 -2.5 -0.9 -0.3 -0.1 0.0 +0.1 +0.5 +1.0 +11.2

Square Simulation case H10W40 Tree cover 100% Soil RH75% Soil RH50% Water cover 100% Grass cover 100% Albedo 0.8 Emissivity 0.70 Ther.Capacity 1000kJ/m3K Ther. Capacity 3000kJ/m3K Emissivity 0.95 Albedo 0.2 Soil RH25% H20W20 Canopy 100% Courtyard Simulation case Tree cover 100% H10W20 Soil RH75% Grass cover 100% Soil RH50% Water cover 100% H20W10 Canopy 100% Albedo 0.8 Emissivity 0.70 Ther. Capacity 1000kJ/m3K Ther. Capacity 3000kJ/m3K Emissivity 0.95 Albedo 0.2 Soil RH25%

area average PET+/for 2300 hours (oC) -2.1 -1.9 -1.5 -1.3 -1.2 -1.2 -0.5 -0.3 -0.1 0.0 +0.1 +0.4 +0.8 +1.4 +2.0 area average PET+/for 2300 hours (oC) -2.8 -1.2 -0.7 -0.6 -0.5 -0.5 -0.1 -0.1 -0.1 -0.1 0.0 0.0 0.0 0.0 +0.3

Table 8. Mean daily effects of design parameters on pedestrian thermal comfort according to daily averaged PET values compared to the base case scenario Square Simulation case Tree cover 100% Tree cover 50% Tree cover 25% Canopy 100% Canopy 50% Water cover 100% Water cover 50% Water cover 25% Soil RH75% Grass cover 100% Grass cover 50% Grass cover 25% SoilRH50% H20W20 Canopy 25% Albedo 0.2 Soil RH25%

daily averaged PET (oC) 28.0 27.8 29.3 33.7 34.4 33.0 33.1 33.1 31.7 34.3 34.3 34.3 34.4 35.7 34.5 38.0 37.5

reference spot NW NE NW NE NE NW NW NW SW NW NW NW NW NE NW NE NW

max daily averaged PET +/- (oC) -10.2 -9.6 -8.9 -6.0 -5.3 -5.2 -5.1 -5.1 -4.1 -3.9 -3.9 -3.9 -3.8 -3.5 -3.1 -1.7 -0.7

Emissivity 0.70 Thermal Capacity 3000kJ/m3K Thermal Capacity 1000kJ/m3K Emissivity 0.95 H10W40 Albedo 0.8 Courtyard Simulation case Tree cover 100% Tree cover 50% Tree cover 25% Canopy 100% H20W10 Canopy 50% Water cover 100% Water cover 50% Canopy 25% Water cover 25% Soil RH75% Grass cover 100% Grass cover 50% Soil RH50% Grass cover 25% Albedo 0.2 Emissivity 0.70 Thermal Capacity 3000kJ/m3K Thermal Capacity 1000kJ/m3K Emissivity 0.95 Soil RH25% Albedo 0.8 H10W20

38.0 38.3 37.7 37.7 38.9 41.5 daily averaged PET (oC) 29.1 29.4 29.8 31.6 31.3 32.4 31.6 31.7 32.1 32.0 32.2 32.4 32.5 32.6 32.8 35.6 35.7 35.9 35.9 36.0 31.4 38.0 39.3

center center NW NW NW NE reference spot center NW NW NE center NE center center NW NW center center center center NW NE center center center center SW NE SE

-0.4 -0.1 +0.1 +0.1 +1.3 +1.8 max daily averaged PET +/- (oC) -5.7 -5.5 -5.1 -5.2 -4.6 -4.4 -3.2 -3.1 -3.1 -2.9 -2.6 -2.6 -2.3 -2.2 -2.1 -1.2 -0.2 0.0 0.0 +0.1 +0.8 +1.2 +6.3

Table 9 Monitored sites and simulation days for each site

Simulation software and days

Monitored sites ENVI-met

Site a 19-06-05 20-06-05 28-08-05 29-07-07

RadTherm

19-06-05 20-06-05 28-08-05 29-07-07

Fluent

29-07-07

Site b 23-06-05 26-08-05 27-08-05 22-06-07 23-06-05 26-08-05 27-08-05 27-07-07

Site c

Site d

Site e

28-06-05 25-08-05

29-06-05

29-06-05

29-06-05 29-08-05

29-06-05 29-08-05

28-06-05 25-08-05 01-08-07

Site f

03-08-07

Table 10 Indicative comparison between measured and simulated values in site a, on the 29th of July (daily average values)

-2.1%

surface temperature difference 22.4%

globe temperature difference -

-0.4%

19.5%

-

-

-4.7%

8.6%

-

-

-0.9%

21.3%

-

-

marble black

-

17.8%

3.5%

-

marble white

-

17.3%

4.4%

-

grass in shade

-

-11%

-8.8%

-

software

ground surface material

air temperature difference

ENVI-met

marble black marble white grass in shade coble stone

RadTherm

Fluent

wind velocity difference -

coble stone

-

10%

1.8%

marble black

-9.3%

11.3%

-

-

marble white

-5%

3.6%

-

3%

grass in shade

-8.9%

-23.7%

-

-

coble stone

-6.3%

3.2

-

-

Table 11 Indicative comparison between measured and simulated values in site a, on the 29th of July at 1500 hours

software

ground surface material

air temperature difference (meas.- sim.) oC

ENVI-met

marble black

-2.2 (31.1-33.3)

surface temperature difference (meas.- sim.) oC 10.6 (56.0-45.4)

globe temperature difference (meas.- sim.) oC -

marble white

-0.9 (32.3-33.2)

9.1 (48.5- 39.4)

-

-

grass in shade

-2.9 (29.9-32.8)

2.2 (26.7- 24.6)

-

-

coble stone

-1.6 (31.5-33.1)

10.6 (50.1- 39.6)

-

-

wind velocity difference (meas.- sim.) m/s -

RadTherm

Fluent

marble black

-

5.0 (56.0- 51.0)

-0.5 (39.2 - 39.7)

-

marble white

-

6.3(48.5- 42.2)

3.2 (43.9 - 40.7)

-

grass in shade

-

-2.4 (26.7- 29.1)

-7.0 (30.7 - 37.7)

-

coble stone

-

2.4 (50.1- 47.7)

-0.5 (39.7 - 40.2)

marble black

-7.2 (31.1-38.3)

-5.9 (56.0 - 61.9)

-

-

marble white

-3.5 (32.3-35.9)

-7.4 ( 48.5 - 55.9)

-

0.1 (0.9 - 0.8)

grass in shade

-5.0 (29.9-34.9)

-5.1 (26.7 - 31.8)

-

-

coble stone

-4.8 (31.5-36.4)

-8.2 ( 50.1- 58.3)

-

-