Improving Outdoor Thermal Comfort in a Steppe Climate: Effect of Water and Trees in an Urban Park

: Excess heat in urban environments is an increasing threat to human health and well-being. Furthermore, the increasingly important phenomenon of the Urban Heat Island (UHI) is exacerbating problems of livability in urban centers. Hence, there should be an increasing effort to assess the impact of heat mitigation strategies (HMSs) on outdoor thermal comfort in cities. This research has investigated how urban areas in steppe climate zones can be more thermally comfortable due to the effects of water bodies and trees, and how this might help to mitigate heat waves. Numerical simulations using the ENVI-met microclimate model have been performed for an urban park in Tabriz, Iran. In-situ measurements of air temperature (Ta) and mean radiant temperature (MRT) have been carried out in the study site and the collected data was used to validate the model (RMSE value 0.98 ◦ C for Ta and 5.85 ◦ C for MRT). Results show that water body evaporation without trees may decrease the air temperature, but on the other hand also increases the humidity, which reduces the positive impact on thermal comfort. However, the combination of water body with trees represents a better performance in the regulation of urban microclimate and thermal comfort.


Introduction
Population growth and global climate change are two critical challenges that mankind is facing today.Around 55% of the global population is living in cities and urban areas.According to projections by the Population Reference Bureau (PRB) [1], 66% of the world population will live in cities (over 85% in developed countries) by 2050.Projected warming from climate change is a significant challenge for the urban environment which is superimposed on the local phenomenon known as the urban heat island (UHI).The effect of UHI and urbanization is causing uncomfortable hot summer environments, especially for the elderly and the sick [2].Therefore, more consideration should be paid to contributions of landscape characteristics and local climate to the cooling effect [3].Climatologists and designers have provided recommendations and urban design strategies for urban heat island mitigation [4].Urban tree planting and the appropriate use of high-albedo surfaces have been suggested as strategies to control urban climate changes and to create better human thermal comfort.Numerous studies focusing on ways for improving urban microclimate and human thermal comfort by planting trees, adding water bodies, using cool surfaces and changing the urban form [5] have been conducted.A series of research studies has been conducted on the outdoor thermal comfort in tropical climates such as Dhaka, Bangladesh [6], Hong Kong [7][8][9], Selangor, Malaysia [10], Colombo, Sri Lanka [11], Hangzhou, China [12] and Singapore [13,14].However, Bartesaghi Koc et al. showed that little is known about the thermal benefits of green infrastructures in tropical and desert climates, developing countries, and southern-hemisphere regions which are experiencing significant urbanization and population growth, and are severely affected Land 2022, 11, 431 2 of 14 by heatwaves [15].This study will investigate the effectiveness of three heat mitigation strategies in the desert climate of Tabriz in the developing county of Iran, which is in the steppe climate zone: trees; pavement; and water.

The Effect of Trees
Several studies have been conducted on the effect of trees as a cooling strategy on urban microclimate and regulating human thermal comfort.For example, Yan et al. [16] reported that adding trees reduced air temperatures, and areas with greater than 55% vegetation coverage created a relatively stable thermal comfort condition for the users in a subtropical climate.Bowler et al. [17], in their research on urban greenery, concluded that trees and parks act to cool the environment on average by 1 • C compared to a nongreen area.Aboelata and Sodoudi [18] evaluated the effect of trees on UHI mitigation and their findings reported that a 50% tree cover scenario reduced air temperature (0.5 K) and enhanced thermal performance (PET).In addition, Aboelata [19] analyzed the effect of vegetation in different street orientations with aspect ratios to mitigate UHI.Their findings reported that trees are not effective at air temperature reduction in streets with an aspect ratio (H/W) 1:1 in all orientations.
There is a high correlation between solar radiation with thermal sensation [20] and it is the most important factor influencing thermal comfort indices (e.g., physiologically equivalent temperature, PET) in the urban environment [21].Shading from live trees is a very useful strategy for lowering solar radiation received at the surface [22].Adding trees can reduce the mean radiant temperature (MRT) significantly, which improves human thermal comfort.In one study trees were found to decrease PET value from 34.9 • C (hot thermal sensation) to 26.2 • C (slightly warm) on hot summer days [23].Aminipouri et al. [24] evaluated the advantages of adding trees to a street in six different climate zones during summertime and their results showed that the potential to offset MRT increased under the best scenario.Hami et al. [25] reported that the cooling potential of trees depends on their specifications including leaf area index, tree height, trunk height, crown height and crown width.But not all species of vegetation are equally effective in ameliorating temperatures in urban areas [26].On the other hand, results of a study on the radiative performance of ten planted tree species in the UK illustrated that tree species selection in the reduction of urban heat depends on the leaf size, solar altitude, and a combination of tree crown morphologies.In addition, infrared transflection towards buildings and pedestrians is considerable [27].Based on [28], different land use/land cover types play different roles in different seasons and times.In addition, the shape and size of trees can modify the effect in different climate regions [29].The evaluation of different commonly found arboreal species and their characteristics should be taken into account by professionals of the urban built environment.This would improve outdoor thermal comfort, reducing the effect of heat islands and ensuring a better quality of life for people in the urban environment [29][30][31].A few results of the research also have shown the effect of urban greening and trees on urban microclimate and thermal comfort in winter [25,[32][33][34].According to literature studies, the effect of urban vegetation on thermal comfort in summer appears more important than in winter.

The Effect of Urban Pavement
In a series of research studies that have been conducted on the performance of urban pavements, the impact of reflective surface on urban microclimate and human thermal comfort in urban areas have been investigated.For example, the wide use of materials with high solar reflectance can contribute to the mitigation of the heat island effect and the improvement of urban environmental quality [35].In other words, pavement materials with high albedo have a positive effect on urban microclimate [36].In this regard, Akbari et al. [37] showed that cool surfaces as a result of high reflectivity of solar radiation are a useful strategy for mitigating the heat island effect.In similar research, Taleghani et al. [38] reported that with increasing urban albedo, MRT and PET increase.In contrast, the water surface decreased PET value.Most of the studies investigated the relative effectiveness of different pavement materials in urban areas.However, the thermal comfort of a person is dependent on the overall energy budget, and a small reduction in air temperature might not offset the increased amount of reflected radiation from highly reflective pavements.

The Effect of Water
In parks design, water bodies are often used to provide operational, aesthetic and microclimate benefits.Urban water bodies are one of the most important components in urban greenery which may positively impact on urban microclimate, particularly in dry climates.Water is commonly designed in different forms such as ponds, rivers, lakes, and fountains.Relatively few studies have investigated the effects of the water body on microclimate and human thermal comfort in urban areas.Xu et al. [39] evaluated the effect of urban water bodies during hot days (air temperatures above 35 • C).Their results showed that increasing waterbody area decreases heat index, providing a significant improvement in human comfort in littoral areas on hot summer days.In another study, the water body had an adverse effect on thermal comfort when the water was warmer than the air temperature at night.In those cases, the water body eventually limits the cooling and thermal comfort in the surrounding city [40].Steeneveld et al. [41] concluded that water bodies increase rather than decrease the 95 percentile of the daily maximum UHI based on weather observations.Some studies have investigated the effect of moving water, including spray fountains and waterfalls, on human thermal comfort in urban areas [42][43][44][45].

Outdoor Thermal Comfort
Many indices have been developed to estimate outdoor thermal comfort.Predicted Mean Vote (PMV) [46] is mainly considered for indoor environments while Standard Effective Temperature (SET) [47], Physiologically Equivalent Temperature (PET) [48] and Universal Thermal Climate Index (UTCI) have been typically applied outdoor.The mean radiant temperature (MRT) is the most effective factor for outdoor thermal comfort in all indexes, particularly during hot, sunny days.
In this study, physiologically equivalent temperature (PET) was used because: (i) it is a universal index and combines weather and thermo-physiological parameters (clothing and metabolic activity); (ii) it reports the result in degrees Celsius ( • C) which makes results more understandable; (iii) it has an acceptable function in both hot and colder climates; and (iv) it has a thermophysiological background and so it gives the real effect of the sensation of climate on humans.Climatic characteristics, including air temperature (Ta), relative humidity (RH), wind speed (WS), mean radiant temperature (MRT) are all used to predict PET.The thermal comfort stresses are defined over 10 scales ranging from extreme heat stress to extreme cold stress.The thermal comfort zone is believed to be between 18 • C and 23 • C for PET (Table 1) [49].The goal of this study was to investigate the effectiveness of heat mitigation strategies in a steppe climate zone.The objectives were to: (i) conduct a case study in an urban park (El Golu Park, Tabriz) in a steppe climate; (ii) investigate the effect of trees, pavement types, and water on the microclimate and human thermal comfort on the hottest day of the year through the use of ENVI-met modelling and field validation.

Study Area
This study was conducted in Tabriz (38 • 8 N; 48 • 15 E; 1350 m elevation) a metropolis in the mountain area in the northwest of Iran.Tabriz is influenced by the local steppe climate [50].In Tabriz, there is little rainfall throughout the year.El-Golu Park, as a large historic park containing an artificial lake in the southeast of Tabriz, was chosen as a case study (Figure 1) because it has a large artificial lake, which is unique to this area.

Hot (Strong heat stress)
35-41 Very hot (Extreme heat stress) >41 The goal of this study was to investigate the effectiveness of heat mitigation strategies in a steppe climate zone.The objectives were to: (i) conduct a case study in an urban park (El Golu Park, Tabriz) in a steppe climate; (ii) investigate the effect of trees, pavement types, and water on the microclimate and human thermal comfort on the hottest day of the year through the use of ENVI-met modelling and field validation.

Study Area
This study was conducted in Tabriz (38°8′ North; 48°15′ East; 1350 m elevation) a metropolis in the mountain area in the northwest of Iran.Tabriz is influenced by the local steppe climate [50].In Tabriz, there is little rainfall throughout the year.El-Golu Park, as a large historic park containing an artificial lake in the southeast of Tabriz, was chosen as a case study (Figure 1) because it has a large artificial lake, which is unique to this area.The highest percentage of district surfaces are covered in, respectively: trees (36%); water body (27%); granite pavement (sidewalk, 21%); asphalt surface (13%); and buildings (3%).
Figure 2 illustrates the monthly minimum, mean.maximum air temperatures.

Input Parameters for ENVI-Met
Overheating in summer causes a lot of problems in areas where populations are vulnerable to heat.Therefore, the hottest day (28th July) of year 2019 in terms of daily average temperature was selected as the study day.Table 2 illustrates the conditions used in the

Input Parameters for ENVI-Met
Overheating in summer causes a lot of problems in areas where populations are vulnerable to heat.Therefore, the hottest day (28th July) of year 2019 in terms of daily average temperature was selected as the study day.Table 2 illustrates the conditions used in the configuration file of the ENVI-met model and meteorological data reported in the Iranian Meteorological Organization website [52].3 This research used the example of a 35-year-old man, 175 in height and 75 kg in weight.Clothing insulation (clo) was considered to be 0.4 clo (summer clothing) with a wind speed of 0.5 m/s.

Simulation Trees Scenarios
Land 2022, 11, x FOR PEER REVIEW the trees at the site are deciduous (Poplar, Beech), and the number of coniferou (pine) at the site is quite limited.(Table 3).In this study, the physical specification of each studied tree for modeling in the met was input using the measured data (Table 3).Leaf properties have been defin cording to species (coniferous-deciduous) and tree calendar (season profile).Root etries of trees were default values.the trees at the site are deciduous (Poplar, Beech), and the number of coniferous trees (pine) at the site is quite limited.(Table 3).In this study, the physical specification of each studied tree for modeling in the ENVI met was input using the measured data (Table 3).Leaf properties have been defined ac cording to species (coniferous-deciduous) and tree calendar (season profile).Root geom etries of trees were default values.the trees at the site are deciduous (Poplar, Beech), and the number of coniferous trees (pine) at the site is quite limited.(Table 3).In this study, the physical specification of each studied tree for modeling in the ENVImet was input using the measured data (Table 3).Leaf properties have been defined according to species (coniferous-deciduous) and tree calendar (season profile).Root geometries of trees were default values.The proportions of different land-use types in the four scenarios are summarized in Table 4.The percentages show that asphalt and buildings are unchanged in all cases.

Simulation Trees Scenarios
Four different scenarios were analyzed as simulation scenarios:  case 1: current scenario (vegetation + water body); Green area (grass and vegetation) covers about 47% of the district surface.The total number of trees in the site is 417, with three different types of trees (Figure 3).Almost all the trees at the site are deciduous (Poplar, Beech), and the number of coniferous trees (pine) at the site is quite limited (Table 3).
Land 2022, 11, x FOR PEER REVIEW 6 of 15 the trees at the site are deciduous (Poplar, Beech), and the number of coniferous trees (pine) at the site is quite limited.(Table 3).In this study, the physical specification of each studied tree for modeling in the ENVImet was input using the measured data (Table 3).Leaf properties have been defined according to species (coniferous-deciduous) and tree calendar (season profile).Root geometries of trees were default values.The proportions of different land-use types in the four scenarios are summarized in Table 4.The percentages show that asphalt and buildings are unchanged in all cases.

Simulation Trees Scenarios
Four different scenarios were analyzed as simulation scenarios:  case 1: current scenario (vegetation + water body); Albero as subcategories of the ENVI-met was used for modeling trees which needed to include three types of parameters:

•
The tree crown geometry (height and diameter of tree); • Leaf Specifications (Species, foliage albedo and tree calendar); • The root geometry (form, diameter, depth and area density of root).
In this study, the physical specification of each studied tree for modeling in the ENVImet was input using the measured data (Table 3).Leaf properties have been defined according to species (coniferous-deciduous) and tree calendar (season profile).Root geometries of trees were default values.

Numerical Simulation
The proportions of different land-use types in the four scenarios are summarized in Table 4.The percentages show that asphalt and buildings are unchanged in all cases.The simulation domain was subdivided into a three-dimensional grid of 96 × 96 × 32 cells with dimensions 4 × 4 × 4, and five nesting grids to each side.Four virtual receptors in ENVI-met simulation (A to d) as reference points for recording and receiving climatic information during the simulation period (9:00 to 17:00) is defined.Output data of ENVImet v4 were used in the RayMan model 2.1 [53] to calculate PET [48].The comparison of simulation results can have presented the impact of water and trees on urban microclimate and human thermal comfort on the site.Contributions of the water body, trees, and reflective surfaces (asphalt and granite pavement) have illustrated in four scenarios in Figure 4.The simulation domain was subdivided into a three-dimensional grid of 96*96*32 cells with dimensions 4 × 4 × 4, and five nesting grids to each side.Four virtual receptors in ENVI-met simulation (A to d) as reference points for recording and receiving climatic information during the simulation period (9:00 to 17:00) is defined.Output data of ENVImet v4 were used in the RayMan model 2.1 [53] to calculate PET [48].The comparison of simulation results can have presented the impact of water and trees on urban microclimate and human thermal comfort on the site.Contributions of the water body, trees, and reflective surfaces (asphalt and granite pavement) have illustrated in four scenarios in Figure 4.

ENVI-Met Validation
This section intends to validate the ENVI-met model using the measured and predicted air temperature (Ta) and Mean Radiant Temperature (MRT), which 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 [54].The instruments were placed 1.5 m above the ground of the study site.The specifications and technical data of the instruments are reported in (Table 5).By using a statistical method, the measured and simulated Ta and MRT of this area were analyzed.The data were measured at different times on 28 July 2019 from 09:00-17:00 in local time.

ENVI-Met Validation
This section intends to validate the ENVI-met model using the measured and predicted air temperature (Ta) and Mean Radiant Temperature (MRT), which 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 [54].The instruments were placed 1.5 m above the ground of the study site.The specifications and technical data of the instruments are reported in (Table 5).By using a statistical method, the measured and simulated Ta and MRT of this area were analyzed.The data were measured at different times on 28 July 2019 from 09:00-17:00 in local time.

Validation Results
Based on the simulated and measured results, the maximum air temperature occurred at about 13:00-16:00.The Ta difference between the results of simulated and measured is about 1.3 • C at the highest point, while the average difference is 0.5 • C. On the other hand, the maximum MRT occurred at 12:00 to 15:00 in both the simulated and measured results.The model represented very similar results between the measured and simulated data of Ta (R2 0.91, RMSE 0.98 • C) and MRT (R2 0.87, RMSE 5.85 • C).So, the performance of ENVI-met software is acceptable (Figure 5).
Based on the simulated and measured results, the maximum air temperature occurred at about 13:00-16:00.The Ta difference between the results of simulated and measured is about 1.3 °C at the highest point, while the average difference is 0.5 °C.On the other hand, the maximum MRT occurred at 12:00 to 15:00 in both the simulated and measured results.The model represented very similar results between the measured and simulated data of Ta (R2 ≃ 0.91, RMSE ≃ 0.98 °C) and MRT (R2 ≃ 0.87, ≃ 5.85 °C).So, the performance of ENVI-met software is acceptable.(Figure 5).

Investigation of Microclimate Results
LEONARDO is one of the other subcategories of the ENVI-met which shows the results of simulated variables in color-coded ranges.Distribution of air temperature at the hottest time (14:00) based on climate on the hottest day of the year (28 July 2019) for all the scenarios are shown in Figure 6.

Investigation of Microclimate Results
LEONARDO is one of the other subcategories of the ENVI-met which shows the results of simulated variables in color-coded ranges.Distribution of air temperature at the hottest time (14:00) based on climate on the hottest day of the year (28 July 2019) for all the scenarios are shown in Figure 6.The 9-h average of climate variables related to the receiving points located on the site has been calculated for Ta, RH, WS, MRT (Figure 7).According to the results, the average Ta and MRT vary depending on the presence of trees and water body in all scenarios.As shown in Figure 7, the average Ta and MRT of the current scenario (c1) during the summer day are respectively 25.9 and 37.7 °C.Therefore, the average Ta in scenarios (c2-c4) compared to the scenario c1 have increased 0.4 °C, 0.2 °C and 0.4 °C respectively.Also, shading by trees has a positive impact on the sunlight reduction.These differences have created a wide zone of "mean radiant temperatures" in different parts of the site.The average MRT compared to c1 has increased c2: 18.2 °C, c3: 1.6 °C, and c4: 22 °C, respectively.The 9-h average of climate variables related to the receiving points located on the site has been calculated for Ta, RH, WS, MRT (Figure 7).According to the results, the average Ta and MRT vary depending on the presence of trees and water body in all scenarios.As shown in Figure 7, the average Ta and MRT of the current scenario (c1) during the summer day are respectively 25.9 and 37.7 • C. Therefore, the average Ta in scenarios (c2-c4) compared to the scenario c1 have increased 0.4 • C, 0.2 • C and 0.4 • C respectively.Also, shading by trees has a positive impact on the sunlight reduction.These differences have created a wide zone of "mean radiant temperatures" in different parts of the site.The average MRT compared to c1 has increased c2: 18.2 • C, c3: 1.6 • C, and c4: 22 • C, respectively.
With regard to simulation results, scenario c2 (no vegetation scenario) provides the same results as the scenario c4, while scenario c1 and c3, have a cooling assistance benefit compared to scenarios c2 and c4.It should be noted although case 3 (no water body scenario) like case1 has improved PET, while the relative importance of the water body on air temperature reduction is quite impressive in case 1 (vegetation + water body) (Figure 8a).In detail, the daily average Ta in cases 1 and 3 are 24.6 • C and 25.1 • C respectively.Overall, it can be seen that water alone has dropped air temperature by 0.55 • C. Ta and MRT vary depending on the presence of trees and water body in all scenarios.As shown in Figure 7, the average Ta and MRT of the current scenario (c1) during the summer day are respectively 25.9 and 37.7 °C.Therefore, the average Ta in scenarios (c2-c4) compared to the scenario c1 have increased 0.4 °C, 0.2 °C and 0.4 °C respectively.Also, shading by trees has a positive impact on the sunlight reduction.These differences have created a wide zone of "mean radiant temperatures" in different parts of the site.The average MRT compared to c1 has increased c2: 18.2 °C, c3: 1.6 °C, and c4: 22 °C, respectively.With regard to simulation results, scenario c2 (no vegetation scenario) provides the same results as the scenario c4, while scenario c1 and c3, have a cooling assistance benefit compared to scenarios c2 and c4.It should be noted although case 3 (no water body scenario) like case1 has improved PET, while the relative importance of the water body on air temperature reduction is quite impressive in case 1 (vegetation + water body) (Figure 8a).In detail, the daily average Ta in cases 1 and 3 are 24.6 °C and 25.1 °C respectively.Overall, it can be seen that water alone has dropped air temperature by 0.55 ° C.    With regard to simulation results, scenario c2 (no vegetation scenario) provides the same results as the scenario c4, while scenario c1 and c3, have a cooling assistance benefit compared to scenarios c2 and c4.It should be noted although case 3 (no water body scenario) like case1 has improved PET, while the relative importance of the water body on air temperature reduction is quite impressive in case 1 (vegetation + water body) (Figure 8a).In detail, the daily average Ta in cases 1 and 3 are 24.6 °C and 25.1 °C respectively.Overall, it can be seen that water alone has dropped air temperature by 0.55 ° C.

Investigation of Thermal Comfort Results
Based on the physiological equivalent temperature (PET) range, the comfort zone is obtained between 18 and 23 • C (Table 1).Simulation results show that the average PET of the base model (c1) during simulation day is 23 • C (comfortable sensation).The average PET in the scenarios (c2-c4) is 35.2 • C (hot sensation), 26.7 • C (Slightly warm sensation) and 37.5 • C (hot sensation), respectively.According to PET index case 1 (vegetation + water body) is in a comfortable zone at 10:00-11:00 and at 17:00.However, the site suffers a warm sensation from 14:00-15:00 on the hottest day of the year.The results of case 2 (no vegetation scenario) showed that the presence of trees can reduce PET value (15 • C−18 • C) at the hottest time of the day around 14:00-15:00.On the other hand, case 3 (no water body scenario) has about similar results with case 1, except 13:00 (warm sensation), and from 14:00-15:00 is in hot sensation.Therefore, PET results reported that vegetation and water body in urban areas can reduce 3-7 • C at the hottest time of the day (14:00-15:00) (Table 6, Figure 8).Figure 8 represents the boxplots of PET values at the site (1.5 m agl) for all scenarios on the hottest day of the year from 09:00-17:00.

Discussion
Urban microclimate and human thermal comfort can be ameliorated through mitigation strategies.This study investigated the influence of water bodies, trees, and surface materials on the microclimate and thermal comfort in an urban park in a steppe climate.The ENVI-met model (v4) was utilized to simulate scenarios.This model uses computational fluid dynamics (CFD) to simulates surface-water-plant-air-interactions in urban areas and has been widely used in studies to the assessment of HMSs on urban microclimate [55][56][57][58] and human thermal comfort [59][60][61][62].
Thermal comfort model performance in terms of simulation Ta and MRT in the ENVImet model was validated.In most previous studies validation of the ENVI-met model was mainly focused on air temperature and MRT [59,[63][64][65].The performance of ENVI-met software was satisfactory compared with early versions [66].The reasonable performance of ENVI-met provides opportunities to researchers to test different design scenarios.
Previous studies have demonstrated that urban heat island mitigation strategies, especially the use of urban greenery, are useful for the regulation of urban microclimate, whose magnitude depend on the specific climates.On the other hand, such a strategy has a significant impact on human-biometeorology which is beneficial for improving thermophysiological indices (e.g., physiologically equivalent temperature, PET).Nevertheless, water bodies in urban areas could also impact human thermal sensation.This issue has seldom been considered in the studies.Furthermore, few studies have investigated the beneficial effect of water bodies in desert climates and developing countries.This study has tried to fill these gaps by providing an application of numerical evaluations of microclimate and outdoor thermal comfort in an urban park in Tabriz, Iran, characterized by a steppe climate.
Based on the results of the surfaces tested, the water body was more effective in reducing heat island compared with the granite pavement.The reduction was achieved because the evaporation of water decreased the surface temperature.However, it also increased the humidity, which reduced the positive impact on thermal comfort.These results confirm the modeling results by [40], where they utilized the same model to study open water surface influence on thermal sensation.However, the combination of water body with trees which reduced MRT represents better performance in the regulation of urban microclimate and thermal comfort.Our result reported that water surface in conjunction with tree cover can reduce air temperature at the pedestrian level by 0.5 • C on the hottest day of the year, which was lower than the results of previous studies [67].However, 0.5 • C is less than the RMSE value 0.98 • C for Ta so it does not necessarily represent a true reduction in temperature.Plus, a reduction that small has only a minor effect on human thermal comfort.More importantly, the MRT reduction of 22 • C is much larger than the 5.8 • C RMSE for MRT and has a substantial effect on human thermal comfort.
According to simulation results, scenario c1, characterized by 36% trees, 27% water body, and 21% granite pavement, is an optimal scheme for summertime.The results of this case can be summarized:

•
The model presented a high correlation between measured and simulated results, with the RMSE value 0.98 • C for Ta and 5.8 • C for MRT;

•
The average air temperature and mean radiant temperature were decreased by 0.5 • C and 22 • C respectively by water body compound with trees on the hottest day of the year; • According to PET results, the cooling benefits from water body with tree cover in urban areas have a positive effect on human thermal comfort levels from "hot" to "comfortable" sensation in the hottest day of summertime; • Based on PET index, numerically expressed by 37.5 • C (c4: without all trees and water body) to 23 • C (c1: current situation including all trees and water body) which is around to the comfort zone in the hottest day.
These results will help urban designers and policymakers to think more creative and bring the concept of sustainability to using cool surfaces, especially water surface, in conjunction with planting trees which has important potential to help offset projected warming from climate change.

Conclusions
Numerous studies have demonstrated the benefits of the tree cover and cool surfaces in urban environments, through reducing solar radiation, creating a microclimate in urban spaces, mitigating the impact of urban heat islands, and overall improving the urban landscape.In this research, the effect of water body with trees on microclimate and human thermal comfort in an urban park in Tabriz have been investigated.Numerical simulations were performed by the micro-meteorological model ENVI-met, validated for Ta and MRT data, on four different scenarios.Results showed that, although water body increases the humidity it also decreases air temperature, which has positive effect on thermal comfort particularly in combination with trees.In summary, the average daily air temperature and MRT were decreased by 0.5 • C and 22 • C, respectively.Therefore, PET reached 23 • C, which is around the comfort zone on the hottest day.
Results of this study illustrate that using systematical principles from water body and tree-planting in urban areas can be an effective strategy for cooling urban temperature and improving human thermal comfort conditions, and can provide design recommendations and guidelines for urban designers and policymakers.

Figure 1 .
Figure 1.Iran map of Koppen climate classification (a) [50], location of the studied site in Tabriz map (b) [51], measurement points on the site of El-Golu park (c).

Figure 1 .
Figure 1.Iran map of Koppen climate classification (a) [50], location of the studied site in Tabriz map (b) [51], measurement points on the site of El-Golu park (c).

Figure 3 .
Figure 3.A view of three different trees planted in El-Golu park (trees exemplary dimen red).Albero as subcategories of the ENVI-met was used for modeling trees which n to include three types of parameters: 

Figure 3 .
Figure 3.A view of three different trees planted in El-Golu park (trees exemplary dimensions in red).Albero as subcategories of the ENVI-met was used for modeling trees which needed to include three types of parameters: 

Figure 3 .
Figure 3.A view of three different trees planted in El-Golu park (trees exemplary dimensions in red).Albero as subcategories of the ENVI-met was used for modeling trees which needed to include three types of parameters: 

Figure 3 .
Figure 3.A view of three different trees planted in El-Golu park (trees exemplary dimensions in red).Albero as subcategories of the ENVI-met was used for modeling trees which needed to include three types of parameters: 

Figure 3 .
Figure 3.A view of three different trees planted in El-Golu park (trees exemplary dimensions in red).

Figure 4 .
Figure 4. Representation of different simulation scenarios on ENVI-met.

Figure 4 .
Figure 4. Representation of different simulation scenarios on ENVI-met.

15 Figure 6 .
Figure 6.Simulated air temperature values at a height of 1.5 m agl averaged for all scenarios at an urban park in Tabriz on the 28 July 2019 at 14:00 in local time.

Figure 6 .
Figure 6.Simulated air temperature values at a height of 1.5 m agl averaged for all scenarios at an urban park in Tabriz on the 28 July 2019 at 14:00 in local time.

Land 2022 , 15 Figure 7 .
Figure 7. Average air temperature values at the site (1.5 m agl) in all scenarios: (a) during on the hottest day of the year ;(b) relative humidity; (c) wind speed; and (d) mean radiant temperature.

Figure 8 .
Figure 8.Average of PET values at the site (1.5 m agl) in all scenarios, during on the hottest day of the year.

Figure
Figure7billustrates the relative humidity trend in the four scenarios.As expected,

Figure 7 .
Figure 7. Average air temperature values at the site (1.5 m agl) in all scenarios: (a) during on the hottest day of the year; (b) relative humidity; (c) wind speed; and (d) mean radiant temperature.

Figure 15 Figure 7 .
Figure7billustrates the relative humidity trend in the four scenarios.As expected, the water body in the urban park also induces a dramatic increase in relative humidity.Compared to the base model, the three other scenarios (c2-c4) provide the reduction of RH by respectively 3.93%, 6.18% and 8.87%.

Figure 8 .
Figure 8.Average of PET values at the site (1.5 m agl) in all scenarios, during on the hottest day of the year.

Figure
Figure7billustrates the relative humidity trend in the four scenarios.As expected, the water body in the urban park also induces a dramatic increase in relative humidity.Compared to the base model, the three other scenarios (c2-c4) provide the reduction of RH by respectively 3.93%, 6.18% and 8.87%.

Figure 8 .
Figure 8.Average of PET values at the site (1.5 m agl) in all scenarios, during on the hottest day of the year.

Table 1 .
Categorization PET level for different thermal perception.

Table 2 .
The conditions used in the configuration file.

Table 3 .
Profile alternatives offered to trees.

Table 3 .
Profile alternatives offered to trees.
The proportions of different land-use types in the four scenarios are summar Table 4.The percentages show that asphalt and buildings are unchanged in all cas Four different scenarios were analyzed as simulation scenarios:  case 1: current scenario (vegetation + water body); Land 2022, 11, x FOR PEER REVIEW 6 of 1

Table 3 .
Profile alternatives offered to trees.
The proportions of different land-use types in the four scenarios are summarized in Table4.The percentages show that asphalt and buildings are unchanged in all cases.Four different scenarios were analyzed as simulation scenarios:  case 1: current scenario (vegetation + water body); Land 2022, 11, x FOR PEER REVIEW 6 of 15

Table 3 .
Profile alternatives offered to trees.

Table 3 .
Profile alternatives offered to trees.

Table 4 .
Specifications, number and species of trees in each of the cases studied.

of Land Use Current Scenario (Case 1) No Vegetation Scenario (Case 2) No Water Body Scenario (Case 3) Soil and Granite Scenario (Case 4)
Four different scenarios were analyzed as simulation scenarios: • case 1: current scenario (vegetation + water body); • case 2: no vegetation scenario (water body without vegetation, replaced by bare soil); • case 3: no water body scenario (vegetation without water body, replaced by granite pavement); • case 4: soil and granite scenario (without vegetation and water body).

Table 5 .
Measuring instrument, specifications and technical data of instruments.

Table 6 .
Average PET changes in summer.