Comparative Study of Form and Features of Courtyards in Terms of Outdoor Thermal Comfort in Two Contrasting Climates of Iran

Courtyards as a traditional strategy were used to create a suitable microclimate for dwellers, but in contemporary architecture, inadequate knowledge of form and features of courtyards makes them deficient. This study presents practical solutions reaching optimal form and features of courtyards based on the traditional architecture of two contrasting climates of Iran. The ENVI-met.4 model was used for simulating the area percentage, water and vegetation level in a very cold climate of Ardabil and Hot-arid climate of Yazd. The simulated atmospheric parameters were imported to the Rayman1.2 to calculate PET thermal index. As the first step, 10% of the total area was advised to be considered a courtyard area for Yazd hot-arid climate, and 60% of the total area was recommended in a very cold climate of Ardabil. Next, the ratio of the water area to the total area of courtyards was simulated and the results advice, 0% of the courtyard to be considered a water area of Ardabil and 10% of the courtyard of Yazd is an optimum choice in hot-arid climate. In the final step, 20% of courtyards recommended the best condition of tree coverage in both climates. By these guidelines, designers can create a more adaptive architecture to the local climate.


Introduction
Open spaces usually cover more than two-thirds of the urban area in modern architecture; thus their microclimate dominates the urban canopy layer (UCL) climate. (Shashua-Bar & Hoffman, 2004).
The new features of the ENVI-met 4 are mentioned according to the following items: 3: Forcing (Huttner, &Bruse, 2009).
(i) Full 3D editor (ii) Advanced calculation of façade temperature and wall energy balance.
One of the most important features of ENVi-met software is its capability to generate the accurate Tmrt for the outdoor condition (Huttner et al., 2008;Chow &Brazel, 2012) and to calculate the Physiological Equivalent Temperature (PET).

Study Area
A simulation method was employed for analyzing the thermal effect of courtyards in Ardabil and Yazd, Iran. These two cities in Iran have completely different climates. Ardabil (48°18'E. Long, 38°15'N. lat) is located in the northwest of Iran and it has a very cold climate. Because of its rash, cold climate, designing the outdoor spaces in this city has some limits and specific principles. There is an impact of the cold wind and solar radiation, humidity, etc. (Jalilian & Tahbaz, 2006;Namin & Khoshvalad, 2015). Yazd (55°0'E. long, 32°0'N. lat) is situated in the center of Iran. This city has a hot and arid climate. Outdoor spaces in this region need special attention to possible ventilation and solar radiation (Hedari, 2010;Teimourtash, 2013). A very cold and long winter is one of the most significant features of the Ardabil climate. In table 1, the long-term climatic data of Ardabil are presented for the 1976 to 2010 period, and also the Physiological Equivalent Temperature (PET) is calculated.  The rough cold climate of Ardabil can be understood from table 1, so in this paper; the coldest day of the Ardabil is chosen for the simulation and the calculated data. In contrast, the hot and arid climate of Yazd -with its long-term climatic data -is introduced in

Selection of Typical Days
Based on available information about the climate of Iran, and according to the meteorological report of Ardabil, the annual mean air temperature of the city is 9.22°C and the mean relative humidity is72.33%. The East wind (90°) is a prevailing local wind with 14.975 m/s wind speed. Based on Ardabil meteorological organization reports, February, 2nd of 1989 was the coldest day in Ardabil with -33.8 °C of air temperature (Ardabilmet, 2016).
Accordingly, Yazd meteorological organization reports show that annual mean air temperature of Yazd is 19.21°C and mean relative humidity is 31.5, and also the wind velocity is 8.016m/s with northwest (315°) prevailing direction. According to the Yazd meteorological organization reports, 13 July 2012 was the hottest day in Yazd with an air temperature of 45.6°C. (Yazdmet, 2016). With this information, the thermal features of the two different climates of Iran will simulate in terms of better thermal comfort condition. The input data for simulation is presented in table4.

Thermal Comfort Analyses of Courtyards
Courtyards are enclosed or semi-closed, open spaces housed in building environments, especially in the context of traditional architecture, which use vegetation and water to provide thermal comfort and to suit the microclimate to its inhabitants. Courtyard concept is a common method of local house designing in the Middle East, especially in Iran. (Almhafdy et al., 2013;Costello, 1977;Mirmoghtadaee, 2009;Ghaffarianhoseini et al., 2015).
The main function of courtyards in all countries is to modify the effect of climatic parameters on inhabitants and to create a suitable microclimate based on local climate. In different climates, different strategies are used to modify the rough environment. (Aldawoud, 2008). A courtyard concept with different features based on the local climate is one of the main principles of vernacular architecture (Coch, 1998). For example, the use of passive ventilation to avoid the annoying humidity in tropical and humid areas is the most significant feature of courtyards in these regions. Sadafi et al., 2011;Makaremi et al., 2012). In hot and arid regions, modification of the environment by water, vegetation, etc., is the most common solution to modify the climatic parameters. (Etzion, 1990;Attia, 2006;Al-Hemiddi et al., 2001). Cold climates also have their own features to avoid annoying wind and to reach sun rays. (Shokouhian et al., 2007;Song et al., 2015;Zhang et al., 2015).
The courtyard concept plays an important role in designing various public facilities such as education, residential areas, and healthcare facilities, but there are not enough strategies to enhance the thermal comfort of courtyards in the different climates of Iran. Uniform design of courtyards in diverse climates of Iran has caused lots of problems in terms of thermal comfort. Lack of sufficient knowledge of thermal characteristics of the courtyards is the main reason for this problem. Courtyards are mostly analyzed in terms of their airflow, solar radiation, and shadow features.
The ventilation potential of the courtyards and their effect on the thermal comfort has been evaluated by CFD studies and wind tunnels in various climates. (Mousli & Semprini, 2015;Almhafdy et al., 2015;Jamaludin et al., 2014). Shade and its effect on the thermal comfort in courtyards are reported as being an important issue in developing the comfort range and energy consumption. (Meir et al., 1995;Yaşa & Ok, 2014). Energy performance of courtyards has been investigated by many researchers (Aldawoud& Clark, 2008;Muhaisen & Gadi, 2006;Manioğlu & Yılmaz, 2008;Dunham, 1961;Al-Masri & Abu-Hijleh, 2012;Behbood et al, m2010;Safarzadeh & Bahadori, 2005). Aldawoud et al discussed the energy performance of the courtyard and of atrium-based buildings. They proved that in general, the open courtyard shows a better energy performance for lower buildings. As building height increases, however, at some point, the enclosed atrium exhibits a better energy performance than the courtyard. With atrium research, some case studies depend on other factors such as glazing and climatic parameters. (Aldawoud & Clark, 2008).
Courtyard proportions and their effect on heat gain and energy requirement were investigated in a Rome climate by Ahmed & Muhaisen. The study, consequently showed that the proportions of the building that houses the courtyard considerably influence the need for heating and cooling. (Muhaisen & Gadi, 2006). In the hot and arid climate of Iran, Behbood et al showed that courtyard method is one of the efficient strategies in terms of energy consumption (Behbood et al., 2010).
In another example of Iranian courtyard research, the passive cooling of courtyards has been investigated by the Safarzadeh et al, and they showed that a courtyard alone cannot make thermal comfort in Tehran, Iran but it can decrease the cooling energy load (Safarzadeh & Bahadori, 2005). Hedari, in his Ph.D. thesis, investigated the thermal comfort in the courtyards of Ilam, Iran. The findings of the study revealed that the people in Ilam could achieve comfort at higher indoor air temperatures compared to the recommendations by international standards like ISO 7730. The results also showed that passive systems such as the main comfort strategy could be applied to housing design in Ilam. By using the results of this study, strategies to minimize energy consumption, not only for Ilam but also for other regions, which have similar climates and cultures to Ilam, can be proposed. (Heidari, 2000).
Taban et al tried to get an optimal courtyard pattern in Dezful by local shadow analysis, and they revealed that courtyards with a length/width ratio of 1 to 1.4 (near to square form) and height/length ratio of 1.1 to 1.2 had the most proper shade on different courtyard surfaces. Results also showed that for the purpose of reducing the cooling load in summer, deep and square-shaped courtyard forms were the most preferable. The self-shading of the courtyard building acts to reduce the need for cooling by an average of about 4%. By using this proportion, the amount of shadows on the courtyard will be optimized (Taban et al., 2014). Modelling study showed that the effect of rectangular courtyard proportions on the shading and exposure conditions produced on the internal envelope of the form in four different locations in Kuala Lumpur, Cairo, Rome and Stockholm and the results showed that the shading conditions of the courtyard internal envelopes are significantly dependent on the form's proportions, location latitude, and available climatic conditions. (Muhaisen, 2006).
A study by Taleghani et al, analyzed the energy performance and thermal comfort of the Netherlands with the light climate changes of this country. The results of this research, which analyzed monthly energy performance, comfort hours and climate scenarios, indicated that using an open courtyard from May through October and an atrium, i.e. a covered courtyard, in the rest of the year establishes an optimum balance between energy use and summer comfort for the severest climate scenario. .
Al-Masri et al, evaluated the courtyard housing in mid-rise buildings of Dubai, United Arab Emirates. In this study, the courtyards have been assessed in terms of energy consumption and daylight factors, and the result of this study shows that the energy consumption of the optimal courtyard is 11.16% less than conventional form building and about the daylight factor. In the courtyard form in both winter and summer time, the courtyard has a better thermal condition. (Al-Masri et al., 2012). The effect of a courtyard's shape and geometry on heat gain and energy efficiency in various climates has been assessed by Yaşa& Ok. In this study, the optimal shape of the courtyard was presented by CFD Fluent (Yaşa & Ok, 2014).
Climatic variables and their effect on thermal comfort are discussed by Sthapak & Bandyopadhyay. In this study, in addition to the placement of the rooms and the proportion of the openings, the degree of enclosure of the courtyards and also the breeze and shadow effect on thermal comfort -especially in light time -are scrutinized. (Sthapak & Bandyopadhyay, 2014). Because of the diverse climate of Iran, new bioclimatic research has been done by Pourvahidi & Ozdeniz. In this study, the bioclimatic conditions of Iranian cities were determined and as a result; five climates were identified in Iran.
In order to prove that this new classification is valid, the traditional architecture of these regions was juxtaposed.
It turned out that the traditional buildings have different features in these five regions. (Pourvahidi&Ozdeniz, 2013). A study of the environmental effect of the residential building courtyards in different climates was done by Taleghani et al, The findings of these studies show that the different configurations of courtyard buildings, such as natural elements, and the situation of openings in different facades are the most important findings of this review paper. (Taleghani et al., 2012).
Recently some studies have examined the comfort condition of the courtyards as a house and urban forms (Ghaffarianhoseini et al., 2015;Taleghani et al., 2015;Berkovic et al., 2012;. Comfort condition of the courtyards is simulated by ENVI-met model software using the PMV thermal index and in this study, Berkovic et al presented the design guidelines for courtyards in Israel and compared various courtyard geometries and shading using trees, openings, and galleries (Berkovic et al., 2012). Cooling strategies of Netherlands courtyard were investigated by a parametric study, and results of this paper demonstrate three heat mitigation strategies for the urban courtyard. ENVI-met was validated for the Netherlands with measurement. .
In another study of the thermal characteristic of the courtyard in a hot and humid climate of Tainan, Taiwan, by Yang et al, the results of this study showed that the integrated design approach can effectively reduce the frequency of heat stress from 79.7% to 40.5% and it is noticeable that by this study, the potential and limitations of the ENVI-met model, is applied in tropical climates (Yang et al., 2016). Thermal assessment of the Portland university courtyard is done by Taleghani et al, in this study, heat mitigation strategies in this university were examined by ENVI-met model and the results showed that the maximum PCI effect was 5.8 °C between a park and a parking lot. Vegetation and a water pond reduced 1.6 °C and 1.1 °C Ta for a bare courtyard, and finally, high albedo material increased Tmrt but reduced Ta. . The study by Ghaffarianhoseini et al, tries to represent the optimal forms of the courtyard in Malaysia. These courtyards have been examined by four parameters of orientation, vegetation, albedo, and height. Consequently, they have evaluated the optimal choice of these parameters represented by ENVI-met simulation for the Malaysian courtyards (Ghaffarianhoseini et al., 2015). Taleghani et al examined the five urban forms of the Netherlands and the final results show that the courtyard form is the best form in terms of PET, Tmrt,… in the moderate climate of the Netherlands (Taleghani et al., 2015).

Calculation of PET as Thermal Comfort Index
The physiological equivalent temperature, PET, is a thermal index derived from the human energy balance. It is well suited to the evaluation of the thermal component of different climates, as well as having a detailed physiological basis (Prata-Shimomura et al., 2009).
The PET index is often used in environmental comfort research for analyzing the physiological behavior of users and pedestrians, according to environmental conditions (effect of buildings and climate). (Matzarakis et al., 1999). PET as thermal comfort has been used in several studies of outdoor thermal comfort (Thorsoson et al., 2007;Andrade & Alcoforado, 2007;Oliveria & Andrade, 2007;Johansson, 2006;Emmanuel & Johansson, 2006) because the PET index has been primarily designed for outdoor use (Spagnolo & Dear RJ, 2003). PET can be calculated using free software (Rayman). This software is validated software for urban complex shading (Lin et al., 2006;Gulyas et al, 2006;Matzaraki et al., 2007). Environmental data for the PET calculation that is required in the Rayman model includes air temperature (T a ), relative humidity (RH%), wind velocity (v), mean radiant temperature (Tmrt) and vapor pressure (VP) and personal data such as human clothing and activity and local data such as the date of year, time and location. Comfort classification of PET scale is described by below table 1.

Reliability of ENVI-met4
In this step, ENVI-met model (the courtyard shape as a sample) was validated through a comparison between field measurements and simulation results. The measurements were done in a courtyard building representing a traditional house of Ardabil and Yazd. Manafzadeh house in Ardabil and German house in Yazd were selected to jsd.ccsenet.org Journal of Sustainable Development Vol. 11, No. 2;2018 assess the reliability and calibration of ENVI-met4. For this assessment, two different days in two different climates of Iran were selected. The air temperature sensor was protected by a white shield to minimize the effect of the radiation. The courtyards' environments were measured for 11 days in February for Ardabil and June for Yazd.
Among these 11 days, two days for each city were selected randomly for Envi-met simulation. February 5 th and 13 th of 2015 and June 15 th and 23 rd of 2015 were the random days. The weather data for the simulation process were taken from the local weather station. The data from simulations and measurements are compared to show the accuracy of the simulation results. To do this simulation, an Envimet Area input file and configuration files are needed. The input-file are described in table 5 and 6.  For configuration file, four extra courtyard models have been modeled to reach the correct results in terms of the neighboring environment on the courtyard affects the output data, so accordingly the surrounding vegetation, pavement, and other real parameters are included in the simulation model. The measured and simulated Ta during the mentioned days in both cities are compared in Fig3 and 4.  Comparisons of the air temperature in both cities show that correlation of measured and simulated Ta in the first day in Ardabil is 82 % (R=0.82), and also R-value for the second day is 0.88. Therefore, on average, the correlation factor is 0.71 which means that more than 88% of ENVI-met outputs are reliable in the cold climate of Ardabil. Correlation analyses of simulated and measured air temperature in Yazd represent the high correlation between simulated and measured Ta. The first-day's R-value is 0.98, and for the second day, it is 0.87. Therefore, ENVI-met outputs are completely reliable in the hot, arid climate of Yazd. The difference in the observed and simulated data could be the fact that ENVI-met does not include levels of cloudiness in its input parameters.

4.2.1Ardabil Traditional Courtyards
In this section, traditional courtyards of Ardabil were selected to be assessed in terms of physical features: (i) The ratio of the courtyard area to total area of the house (ii) The ratio of the water area to total area of the courtyard (iii) The ratio of the vegetation area to total area of the courtyard For this study, 8 traditional courtyards were selected to be investigated: (i) Khadem house.
(vii) ManafZadeh House (viii) Khalil Zadeh house.  In a comparison of the data derived from the previous tables, it can be concluded that in Yazd's hot-arid climate, due to a long and hotter summer, more water ponds were used. This is in contrast with the rough, cold climate of Ardabil where only a small area of water covers the courtyard. In terms of vegetation, a smaller area of courtyards in Yazd is covered by trees to reach optimum natural ventilation. In Ardabil's cold climate, this percentage is more than Yazd. About total area of the courtyard in these two contrasting climates, the averages of the courtyard is about 42% of the total plan and 0.26% of the total plan consider as a courtyard. In table 9, a comparison of these mentioned features is portrayed:

Results of Simulations
The study scrutinized different courtyard configurations. The sample building and its courtyard dimension were chosen by the average of the real traditional courtyard houses of Ardabil and Yazd. The 360m 2 (60*60 m) 2-story building (8 meters of height) and the other parameters of the courtyard (dimension, water amount, and the vegetation) were selected based on the minimum and maximum of the real courtyard houses of the Ardabil and Yazd.

Effect of the Courtyard Area
As a first step, it is crucial to determine the courtyard area percentage based on local climate. So, in this regard, 10, 20, 40 and 60 percent of the block was considered as a courtyard and simulated for both climates.
The impact of the different areas towards influencing the microclimate features such as relative humidity, wind velocity, Tmrt, and the ambient temperature were measured recording the different factors on the values assumed by previous parameters in the center of the courtyard spaces.

Ardabil Courtyard Area
A comparison of the courtyards in terms of the area shows that the average air temperature in all four conditions is stable, and is about -11. The differences between the highest and the lowest one is 0.4 °C (10%= -11.475, 20= -11.656, 40%= -11.856 and 60%= -11.906). However in terms of Tmrt, there is a different condition. The difference between the highest and the lowest is 11.73°C. The hourly frequency of Tmrt is described in Fig 5. Based on this Fig, (in all conditions) at the beginning of the morning (7.00-8:00 am) Tmrts are in the equal range. This is also true for the end of day (16.00-18:00 pm), but in the next hours, Tmrt of 60% area grows more quickly than the others, which means that the average air temperature of the Tmrt for 60% is higher than the

Figure 5.Hourly frequency of Tmrt of Ardabil
Wind velocity is another important parameter of outdoor thermal comfort and was also investigated in these courtyards. There were not any significant differences between the conditions in terms of average wind speed. On average, there is a 0.251 m/s difference between the highest and the lowest velocity (10%=3.958m/s, 20%=3.881, 40%=3.763 and 60%=3.701m/s).
Hourly frequency of PET is described in Fig6. In FIG7, Leonardo visualization describes the Tmrt and wind speed conditions in a different configuration.

Yazd Courtyard Is
The courtyards were compared with their related Ta and the results show that there are not that many differences between the courtyards (10% = 24.89°C, 20% = 24.92, 40% = 24.94, 60% = 25°C). However, the different sky exposure means that the Tmrt values are different, and by this means there are 3.17°C differences between the maximum and the minimum Tmrt. The hourly frequency of Tmrt is described in Fig 8. Based on this Fig, in all conditions, at the beginning (7.00-8:00 am) of the morning, the mean radiant temperature values are in the equal range but as the sun rises the radiation amount increase too.

Figure 8. Hourly frequency of Tmrt of Yazd
In the next few hours, the 60% courtyards' Tmrt grows, quicker than the other courtyards'. Based on the graph, the thermal behavior of all conditions between 10:00-14:00 are equal, but the 10% courtyard loses its Tmrt faster than the others and at the end of the day (20:00) it reaches the lowest Tmrt among the courtyards. Therefore, on average, the 10% courtyard has the lowest Tmrt (10% = 52.69°C, 20% = 54.26, 40% = 55.08, 60% = 55.86°C).
The average PET value in each courtyard represents a 1.8°C difference (10% = 31.4°C, 20% = 32.2, 40% = 32.6. 60% = 33.2) between all conditions. The 10% courtyard has a better condition, and because of the closer sky condition of the 60% courtyard, it reaches less long direct and diffuse radiation. This means the R values (correlation) between PET value and Tmrt, SW.dir, and SW.diff are 0.95, 0.94, and 0.96. The hourly frequency of related PET is described by Fig9

Effect of Courtyard Water Level
The presence of water in Iranian traditional courtyards is a common principle, and water is also the main component of contemporary courtyards. This means that 0, 10, 15, and 20 are the typical percentages of water in traditional courtyard areas in both climates. In this regard, these percentages have been considered as a courtyard and simulated for both climates.
The impact that different water amounts have upon microclimate features such as relative humidity and Tmrt was considered when recording the different factors for the values assumed by previous parameters in the center of the courtyard spaces.

Ardabil Water Amount Analyze
A comparison of the water level in the 60% area highlights the fact that air temperatures are in a roughly equal range (0% = -12.26°C, 10% = -12.33, 15% = -12.36, 20% = -11.62). It also demonstrates that owing to the cold air temperature, the relative humidity of these conditions is 100%, and, accordingly, the wind speed parameter is equal to all conditions. Radiant temperatures differ notably with different water amounts. The courtyard contains 0% water featured 10°C higher on the Tmrt scale (0% = 36.49°C, 10% = 26.24, 15% = 26.26, 20% = 26.56), and the hourly frequency of the Tmrt is described in Fig11. Figure 11. Hourly frequency of Tmrt(water amount) According to the seven, all other conditions except the 0% have a colder Tmrt. At the beginning of the referenced day (7.00-8:00 am), all conditions have equal Tmrt but by sunrise, the thermal behavior of the courtyards changes -the courtyard containing no water reaches the highest Tmrt. Then, from 9:00 to 16:00, it maintains the highest Tmrt. At the end of the day, the thermal behavior of the courtyards is similar to each other. The main atmospheric parameters of these courtyards are approximately equal, but Rlw warming has a different value for each courtyard. These values represent the air temperature change due to longwave flux divergence. This value shows higher changes compared to the others (0% = 0.65 K/h, 10% = 0.43, 15% = 0.41, 20% = 0.44 K/h), and higher air temperature variations based on long wave flux divergence will definitely cause higher Tmrt.
The results discussed in this section are represented by PET thermal index. On average, the courtyard containing 0% water in the cold climate of Ardabil has a better thermal condition than the other conditions (0% = -10.8, 10% = -13.1, 15% = -13.1, 20% = -12.4°C). The hourly frequency of PET is described in Fig 12. It also shows that the relative humidity is changed by Ta changing and water percentage. There is a 3.41% difference between the highest and lowest relative humidity of the courtyards (0 & 10% = 56.05%, 15 & 20% = 52.65%). Accordingly, there are different Tmrt in these courtyards based on different levels of water percentage. On average, the courtyard that is covered 10% by water has the better thermal condition and it is 9.05°C cooler than the hottest courtyard (0% = 44.25°C, 10% = 38.58°C, 15&20% = 47.53°C). The hourly frequency of the mean radiant temperatures of courtyards is described in Fig13.

Effect of Vegetation
Green spaces and utilization of vegetation in the courtyard will modify the different environmental parameters in open spaces and also in courtyards. Based on the traditional courtyards in both climates, four levels of greening were examined and stimulated with 0%, 20%, 40% and 60% of the courtyard being covered by vegetation. The use of different vegetation will affect the climatic features such as Ta, Tmrt, wind speed and humidity. These results were recorded using the values assumed by previous parameters in the center of the courtyard spaces.

Microclimatic Effect of Vegetation in the Courtyard of Ardabil
In this section, as a final step, the different levels of greening were simulated in the best area percentage of the courtyard in the very cold climate of Ardabil. A comparison of the data shows that the air temperature of different greening conditions varies by 1.6°C (0% = -11.475°C, 20% = -12.38°C, 40% = -12.87°C, 60% = -13.1°C). As was the case because of frozen degrees, the relative humidities of all conditions are 100%, besides the Ta, wind velocities and mean radiant temperatures are changing correspondingly by vegetation changes. On average, the wind speed of the courtyards are decreasing with the vegetation percentage (0% = 3.96, 20% = 3.23, 40% = 2.99, 60% = 2.9°C), and there is 1.06 m/s of difference between the 60% tree-covered spaces and noncovered one.
The next step is to calculate the PET thermal index. The mean radiant temperatures were used to analyze the effect of vegetation on outdoor thermal comfort. Therefore, on average the 20% covered courtyard has the highest Tmrt on the coldest day of winter (0% = 24.58, 20% = 31.49, 40% = 26.05, 60% = 24.49°C).The hourly frequency of Tmrt is described in Fig15. Ta is the air temperature (K).
Accordingly, Tmrt is the total amount of direct and diffuse radiation on a specific point. Wind speed has an inverted effect on the Tmrt by 10 8 * 0.6 near to Globe. This means that an increase in wind velocity results in a decrease in radiant temperature, so Tmrt is affected by SWdir, SWdiff and also wind speed. In the 20% covered courtyard, the sum of wind speed and the radiation amount create suitable for a combination of these two. So the PET average values (0% = -12.68, 20% = -10.60, 40% = -12.06, 60% = -12.61°C) show that the 20% courtyard is 2°C warmer on the PET scale. The hourly frequency of PET is described in Fig 16: The Tmrt were derived and compared in order to determine the coolest courtyards in terms of vegetation different percentages. The results show the 20% covered has the lowest Tmrt among all conditions. According to what has been discussed in the previous section, and based on a formula to have a lower Tmrt, the courtyard needs less direct and diffuse radiation and also high wind speed. The optimum combination of vegetation and wind speed occurs in a 20% courtyard.
In order to assess the comfort condition, the hourly frequency of PET value described in FIG18. As shown in Fig12, the 20% tree-covered courtyard has a significantly cooler PET value based on what has previously been explained. This courtyard has the best combination of wind speed and tree coverage.

Conclusion
Open spaces and their quality roughly depend on their thermal condition. Better thermal comfort condition of courtyards was proved in many types of research, and it is a common strategy in traditional architecture to modify the climatic conditions. Courtyards were mostly investigated in hot-arid and humid conditions. Uniform use of the courtyards in a different climate is a contemporary architecture issue. Therefore, a study of the physical features of traditional courtyards in a different climate in Iran could be undertaken. The results could then be compared to those found in Iranian conditions, and contextual guidelines for courtyards in fewer arid conditions could be drawn up.
Accordingly, the thermal comfort achieved during the critical day of each climate (rough clod of Ardabil, hotarid of Yazd) was selected to assess the thermal comfort condition in low contrast courtyards. Results presented new practical guidelines as to how to best adapt the contemporarily-built environment with the climate based on traditional factors to reach sustainable contextual courtyards. Reliable simulation methods present optimum solutions to create thermally comfortable courtyards in the real world. As a first step, physical features of 16 traditional courtyards of both climates were analyzed. Three main features of the courtyard were considered in analyzing: (i) The ratio of the courtyard area to the total area of the house (ii) The ratio of the water area to the total area of the courtyard (iii) The ratio of the vegetation area to the total area of the courtyard Based on the results of this section, the range of the simulation parameters was determined.
1) Initially, in order to have optimum solar radiation in both direct and diffuse conditions, the area of the courtyard was simulated. The simulation showed that the optimum courtyard size for the hot and arid climate of Yazd was 10% of the total area of the house. In contrast, to have more radiation in the rough winters of Ardabil, the optimum courtyard size is 60% of the total area of the house. Based on the PET results, there are 6.8°C differences between the 10% and 60% courtyards area in Ardabil, and by increasing the sky exposure condition the60% courtyard reaches the more SW.dir and SW.diff. Therefore, high Tmrt causes high PET values. In contrast, the area percentage of the courtyards of Yazd, due to having less solar radiation in hot summer day 10% percent courtyard area were recommended with ENVI-met4 results. By this means in average the 10% courtyard area reaches less solar radiation and it has more shaded surfaces on a hot summer day of Yazd. Pet results demonstrate the cooler environment of Yazd accordingly at 15:00 there is a 9.8°C differences between the hottest and the coolest courtyard.
2) The second step was to analyze the effect of using water as the main component in a traditional courtyard. The ratio of the water area to the total area of courtyards was simulated, and the results show that every condition except the 0% (water) courtyards produced a cooler environment in very cold winter of Ardabil (as happens in traditional courtyards). There was a 16.59°C difference between the 0% courtyards and the others at 13:00. PET values demonstrate the warmer environment of a 0% courtyard in comparison with the others by 4.7°C of PET.
Water area simulation results for Yazd's hot climate demonstrate that 10% of the water area is the optimum option in the hot and arid climate of Yazd. However, at 12:00 the 20% water courtyard has the coolest environment. During the day it has a high Tmrt temperature, but the 10% water area courtyard has a low temperature rather than the other models in average condition (during the referenced day).
3) The final step in courtyard simulation. The courtyards were simulated in 4 conditions (0%, 20%, 40%, and 60% tree covered area). At the Tmrt calculation formula present, it is summed up the direct and diffuse radiation and it is affected by wind speed inversely. The optimum conditions in which to have higher Tmrt in the presence of vegetation is a 20% tree covered area. The PET values of the show that effectively the 20% courtyard has a warmer environment in the rough, cold winter day in Ardabil.
These conditions were simulated to the hot-arid climate of Yazd, and the results show that the optimum tree covering of 20% results in less radiation and enough wind speed in the Yazd hot-arid climate. The hourly frequency of the Tmrt shows that, on average, there are 4.65°C differences between the highest and lowest Tmrt and accordingly the results of PET values demonstrate that among the different tree coverage, the 20% tree covered courtyard has better conditions than the others and, on average, has a 2.33°C cooler environment than the hottest one.