A climate-architecture modeling to explore the effect of land use change scenario on building bioclimatic design pattern in order to provide human thermal comfort

Abstract

Given the role of different land uses in climate change, the present study seeks to identify the effect of land use change on the building bioclimatic design pattern in a dry climatic type. For this purpose, Yazd station in central regions of Iran has been analysed as a representative of dry climate type. The Air Pollution Modelling (TAPM) which is a regional climate software was used to identify the effect of forest cover on climate change in regional dimensions and finally its impact on heating and cooling solutions for buildings. For this study, two scenarios were considered. One scenario was the real situation in the area and the other one was the simulation of the effect of forest cover on climate change in the studied area. Finally, the results of this study revealed that if there is a forest cover, due to the temperature increase in all months of the year, the average annual temperature will increase by 9.20 °C compared to the real conditions. On the other hand, in the forest cover simulation scenario, the relative humidity will increase by 13.4% per year. The findings also showed that, despite forest cover, the annual temperature difference and temperature fluctuations are reduced by 4 °C. Furthermore, the results of this study indicated that if there is a forest cover, the heating requirements will be reduced in total and the demand for the bioclimatic design of cooling and dehumidification will be increased.

Introduction

Today, the world’s population is increasing which has led to various changes in land use in the place of residence due to the diversification of their needs [1–3]. Some of these uses, due to land use planning, have led to community development and some of these land use changes, regardless of environmental potentials, and not taking into an account the consequences of these use changes, have had a negative effect on the natural environment and, ultimately, on human societies [3–5]. For example, the conversion of forest lands into rangelands, damming on rivers and land, or dehydration of downstream lakes, transformation of suitable agricultural land into urban settlements, etc. all of which are evidence of poor management of land use change which can in turn lead to an increase in some of the phenomena of extreme events and natural disasters in local or regional dimensions such as increased floods, desertification, dust storms, or the increase of aerosols, global warming, etc. [6–8].

However, today’s world population is heavily dependent on energy in order to advance its vast majority of activities. One of the dimensions of energy needs is to provide energy for the comfort of the people in their homes and their settlements [9–11]. In-house energy demand is heavily dependent on local and regional weather factors. The weather in each region is also severely affected by local and regional topographic complications [12, 13]. As an example, for residential areas adjacent to hot bodies of water, or hot and dry desert areas, or areas adjacent to cold and glacial heights, the type of energy demand in their habitats is completely different [14–16]. Therefore, today’s expansion of cities and land use changes can have an impact on the energy demand and living conditions of the inhabitants of their settlements [17–19]. Accordingly, the main objective of this research is to investigate the effects of forest cover on urban climate patterns and as a result changes in energy needs in urban buildings. The thermal comfort model in accordance with the ASHRAE standard 55, defines the active and passive strategies needed to achieve building’s indoor thermal comfort by implementing the climate conditions on a psychrometric chart. This model can be a suitable criterion for comparing the effect of existence/nonexistence of forest vegetation in the environment. Therefore, in this research, the results are compared on the basis of the ASHRAE standard 55 and its adaptive comfort model. Urban green spaces, especially the urban marginal forests, have the social, economic and ecological role with the benefits like treating the mental illness, being as an ideal environment for nurturing children, creating the social integrity, and maintaining comfort because of improving the quality of life space and community development. Not only from the social dimension, but also in terms of urban resilience in the face of climate change, and climate risks, such as floods, they have a significant share in the need for moderating the cooling and heating energy of homes. Trees and vegetation reduce the surface temperatures through the evaporation-transpiration. Shaded surfaces can be 11 to 25 C degrees cooler than the maximum temperature of shade-free materials [20]. Evaporation-transpiration alone or with shadows can reduce the maximum summer temperatures up 1 to 5 C degrees. Planting trees and vegetation in strategic locations around buildings, or for shadowing on pavements in parking lots and the streets, is the most useful adjustment strategy. Therefore, the use of trees, parks and urban green spaces, including urban forests, is one of the factors affecting the reduction of the urban heat island [21]. One of the worrisome issues, associated with the green coverage in balancing the surface temperature with the climate change is the drought time; when many plants, including lawns, are dried and the cooling rate resulting from their evaporation-transpiration decreases. Whereas, the consecutive dry days and long heat waves will increase during the summer with the climate change [22]. Water levels within the city, such as urban rivers, waterfronts within the parks and squares, as well as tree coverings have an important role in creating the suitable places to cope with droughts, in the event that the amount of plants, such as lawns reduces. Trees can make the levels cooler and keep them at about 15.6 degrees, to be a good alternative to other plants. Trees can also absorb carbon dioxide and release oxygen and decrease the energy consumption in buildings. Hundred 10-year-old trees can absorb 1 ton of carbon per year [23]. Zari tree of Rash species (Fagus orientalis Lipsk) in an area of one hectare, can release three thousand tons of water in the form of steam in the air during six . monthsTranspiration in trees is associated with absorption of calories، so it results in a temperature decrease [24]. Furthermore, the needle-leaved up to 40% and broad-leaved up to 20% among the tree species, are capable of taking rainwater and returning it to the air again through evaporation and thereby, increasing the relative humidity of the air. Although it seems that this effect of trees is controversial at the forest level, it is tangibly positive at the city level [24]. When vegetation is placed at urban levels, thermal equilibrium can change to cooling conditions like the rural areas. It is estimated that 1460 kg of water evaporates from a tree during a sunny summer day in average, that which requires about 860 Mg energy to evaporate it. This operation in average provides the cooling effect of air conditioning outside the building.

Trees can reduce buildings’ energy on heating and cooling demands, as well as consequent emissions of air pollutions like CO₂ by power plants, because the tree’s shadow on buildings and reduce summer air temperature and block the winter winds [25]. Green areas are actually the ecological measure to resist the problem of majority of concrete masses in urban regions. Forests are quite different from the urban built environments. The surface of every forested area has a different Bowen ratio than a mineral one, because the energy of incoming solar radiation is used for transpiration and photosynthesis by plants and thus the latent heat flux would be more than sensible heat flux. At night, the energy of inverse radiation from a green surface is produced by both the thermal and latent heat flux. Thus, the temperature around the green area is lower than that around the built environment [26]. However, trees that shade buildings in winter can increase heating needs, too. The regional climate, the size and amount of tree foliage, and locating of trees near the buildings can affect their energy conservation. Tree arrangement can save energy demands by shading on east and west walls and roofs, and protecting the building from prevailing winter winds. Reducing the wind in the summer can sometimes cause the increasing of energy use for air conditioning, but the wind and shade effects finally combine lead to reduced summer cooling energy of buildings [27, 28].

Finally, it can be said that trees can have both positive (e.g., air pollution removal) and negative functional values (e.g., trees can increase annual building energy use in certain locations) [25]. Therefore, this study seeks to explore and modelling whether this change in land use can affect the demand and energy consumption patterns of the settlements or not. Meanwhile, it is hypothesized that land use changes can affect local climate and eventually change the residential pattern of demand and energy consumption.

Materials and methods

Climate modelling and simulation scenario of the effect of forest cover in the present project

In this study, for implementing two scenarios, a climatic software model called The Air Pollution Modelling (TAPM) was used. TAPM was prepared by the Commonwealth Scientific and Industrial Research Organisation in Australia. Hurley (2008) [29] writes that this model is capable of predicting air pollution and its density for short climatic periods which are less than a year. Zawar-Reza et al. (2005, 2007) [30, 31] added that TAPM is a non-stationary three-dimensional model based on simple equations which uses a coordinate system of ground-based data. The obtained data are in accordance with analyses of climatic conditions of the area, and for reaching a conclusion, statistical methods and techniques are used. Map data required in the TAPM model are divided into three categories which are available in the system as a whole and include ground-level data or terrain such as topography and soil and water vegetation. This data with a resolution of 1 km*1 km are provided by the company which provides the model. This data are provided in all regions of the world, including Iran, but the user can replace the data with their own desired data for the model. Second is synoptic and meteorological data of the model, and the resolution of the synoptic model is 75 km which is the minimum coverage of the model, and the date of the third type depend on the user research. For example, if the research is focused on air pollution, the data of contaminants are used in this section. In fact, the third type of data depends on the researcher so that they can alter the corporate data of the model and replace it with their own data. The data which are given to the model should be of ASII type, although the model has the ability to convert various data formats [32, 33].

In this sense, it is interesting to explain that the surface information datasets provided with the TAPM are as follows [34]:

• Global terrain height data on a longitude/latitude grid at 30-S grid spacing (approximately 1 km) based on public domain data available from the US Geological Survey, Earth Resources Observation Systems (EROS) Data Center Distributed Active Archive Center (EDC DAAC).

• Australian terrain height data on a longitude/latitude grid at 9-s grid spacing (approximately 0.3 km) based on data from Geoscience Australia.

• Global land cover characterisation data on a longitude/latitude grid at 30-s grid spacing (approximately 1 km) based on public domain data available from the US Geological Survey, Earth Resources Observation Systems (EROS) Data Center Distributed Active Archive.

  Center (EDC DAAC).

• Global soil texture types on a longitude/latitude grid at 2-degree grid spacing (approximately 4 km) based on FAO/UNESCO soil classes dataset.

• Global 5-year monthly mean LAI on a longitude/latitude grid at 2-degree grid spacing (approximately 4 km) based on Boston University LAI datasets (derived from MODIS products).

• Rand’s global 10-year monthly mean sea surface temperatures on a longitude/latitude grid at 1-degree grid spacing (approximately 100 km). They are based on public domain information available from the US National Center for Atmospheric Research (NCAR).

  The synoptic scale meteorology datasets currently available are:

• Six-hourly synoptic scale analyses on a longitude/latitude grid at 0.75- or 1.0-degree grid spacing (approximately 75 km or 100 km). The database is derived from LAPS or GASP analysis data from the Australian Bureau of Meteorology, who have kindly allowed us to provide the data used by TAPM. The regions available are shown on the TAPM web site.

It should be explained that in TAPM software, there is a possibility to change the land use based on 38 different codes. However, some of these codes are related to one family of land use. For instance, one can refer to codes 35–37, which belong to the industrial use from small to large dimensions, or codes 31 to 34, which includes the application of urban usage on a different scale from small to large or even codes 29 and 30 which consider the use of the lake. Finally, code 9, which includes forest-low sparse (woodland), code 8, forest-low mid-density, code 7, forest-low dense, code 6, forest very sparse (forestland), code 5, forest sparse (forestland), code 4 under the forest middle dense, code 3 as a forest -dense, code 2, forest, tall-mid dense, and code 1, forest tall-dense can be mentioned It is worth noting that no more features and details of each of these applications are provided in the software. In this study, a land use change scenario has been used to influence the climate pattern of the surrounding areas. In this scenario, the climatic effects of forest cover, code 3, called forest -dense was modelled in the western part of Yazd (Fig. 1). The reason for choosing this forest cover in the west of Yazd is that most of the atmospheric streams for this city are in the range between southwest and northwest, so it would be better to consider land use change in western Yazd because the prevailing western trends can transfer the climatic effects of forest cover to Yazd.

Fig. 1.

Wind roses in the city of Yazd; (a) Annual (b) Winter (c) Spring (d) Summer (e) Autumn

It is also worth noting that for this scenario, the extent of simulated forest cover is estimated to be 30,000*30,000 square meters, so that the effect of this use on significant changes in the climatic pattern of Yazd city will be significant. Because firstly, in the smaller dimensions, forest cover was modelled, but no significant impact on the climatic pattern of Yazd city was noted. Hence, one of the reasons is that the effect of synoptic patterns (atmospheric systems) on dimensions smaller than the above-mentioned scale cannot be considered, and their effects cannot be applied to the climate modelling of the present study. On the other hand, the climate model used in this study is a regional climatic model and cannot be reliable in small scale simulations. It needs to be explained that in this study, the selected scenario was hypothetical and could include other uses, and there is no prospect of future use of this use of the environment. However, since the city of Yazd has a dry climate type, it was of great significance to model the effect of a forest cover on climate change and its energy consumption pattern. But in this research, as there were lots of output on one hand and the modelling was time consuming on the other hand, only outputs for two sample years were selected, one for 2003 and the other for 2006. The choice of these two periods is enough for the present study because the purpose of this study is not to determine the trend of climate change for a long period of 30 years or more. Rather, the aim of this study is to examine the effect of land use change on changing the pattern of climatic components, which is sufficient to accomplish the objectives of this study.

Therefore, the method of work in this study is that initially, for the two years of 2003 and 2006, based on real environmental conditions and climatic data, the building bioclimatic design in Yazd city was extracted and then in the next step, for these two selected years, in terms of land use change and the creation of a forest cover, the values of climate components were simulated. Then, based on these new values of simulated climatic components, a new model of indoor building bioclimatic design was extracted and the results of the real conditions and the simulated scenario were compared.

Details of the modelling of building heating and cooling strategies

Climate Consultant software is sponsored by the California Energy Commission and developed by the University of California at Los Angeles (UCLA) [35]. Climate Consultant ver. 6 was developed by Robin Liggett and Murray Milne of the UCLA Energy Design Tools Group, with technical support from Don Leeper and Carlos Gomez. This version of Climate Consultant was developed with support from the California Energy Commission PIER Program (Public Interest Energy Research) which is funded by California Utility Ratepayers.

This software is for designing and remodelling buildings that are truly climate responsive depends first on gaining a detailed accurate understanding of the local climate. Climate Consultant reads the local climate data in EPW (EnergyPlus Weather file) format and displays dozens of different graphic charts of various weather attributes. The purpose of Climate Consultant is not simply to plot climate data, but rather to organize and represent this information in an easy-to-understand new ways that reveal the subtle attributes of climate and its impact on built form. One of the best charts that the software draws is a psychrometric chart of comfort strategies. This chart is one of the most powerful design tools in Climate Consultant. It shows dry bulb temperature across the bottom and moisture content of the air up the side. Every hour in the EPW climate data file is shown as a dot on this chart.

The Design Guidelines screen shows a list of suggestions, specific to this particular climate and selected set of Design Strategies, to guide the design of buildings such as homes, shops, classrooms, and small offices. Architects call these envelopes dominated because they do not have large internal thermal loads and thus the design of the building’s envelope will have a great deal of impact on the thermal comfort of the occupants [36].

Introducing adaptive comfort model

The Adaptive Comfort model which is defined in ASHRAE Standard 55, applies in naturally ventilated spaces where people can open and close windows. Indoor conditions are acceptable when average outdoor air temperatures are between 10 °C and 33 °C, and when indoor temperatures can be held within a specified 10-degree indoor operative temperature range. Thus thermal comfort depends in part on outdoor conditions and occupants will have a wider comfort range than in buildings with mechanical HVAC systems. So this model does not apply when there are building’s heating system or air conditioning system [37]. It assumes that people will adapt their clothing to the climate (1.0 to 0.5 Clo), and that they are engage in sedentary activities (1.0 to 1.1 Met). The standard does not discuss how adaptive comfort is affected by the other building Design Strategies. Thus, on the Criteria Screen, Adaptive Comfort is defined only in terms of the zone of “Adaptive Comfort using Natural Ventilation” on a psychrometric chart.

One of the other parts of this study was the estimation and calibration of the effect of the existence and absence of forest cover on the operative comfort temperature of Yazd city. It has been observed that in constant and stable conditions, people can easily adapt themselves to the environment and provide their own thermal comfort with some passive measures. Therefore, research shows that in the temperature range of 17 °C to 31 °C, it is possible to provide adaptive comfort for the people in the way that the internal operative comfort temperature is based on a function of the outside air temperature for buildings where cooling and central heating are not required [38]:

$$T_{\mathit{oc}}=17.8+0.31T_{\mathit{\text{out}}}$$ 1

The adaptive model boundary temperatures for 90% thermal acceptability are approximately T_oc + 2.5 °C and T_oc-2.5 °C according to ASHRAE Standard 55–2013. In this equation, T_out is considered to be the monthly mean outdoor temperature. But in this study, T_out is used as the hour outdoor temperature. Therefore, based on the study data, the change in the pattern of the occurrence of Operative Comfort Temperature has been evaluated.

Introducing statistical methods to calibrate and validate the climate model

In the present study, in order to calibrate the results, the relation (2) was used:

$$\mathrm{Z}=\left({\mathrm{Sim}}_{\mathrm{i}}-{\mathrm{Ave}}_{\mathrm{sim}}/{\mathrm{Sd}}_{\mathrm{sim}}\right)\times {\mathrm{Sd}}_{\mathrm{obs}}+{\mathrm{Ave}}_{\mathrm{obs}}.$$ 2

In Eq. (2), the Z component contains calibrated values, and Sim_i is introduced as the simulated value of i^th and Ave_sim and Sd_sim are the mean and standard deviation of simulated values, respectively. Moreover, Sd_obs and Ave_obs are the mean and standard deviations of observational values, respectively.

After calibrating the simulated data, in order to evaluate their fit with real data, the Nash-Sutcliffe test was used (relationship 3).

The Nash-Sutcliffe Goodness of Fit statistics [39] is computed as follows:

$$\mathit{NS}=1.0-\sum _{n=1}^{\mathit{\text{ntot}}}{\left(\mathit{OB}{S}_n-\mathit{SI}{M}_n\right)}^2/\sum _{n=1}^{\mathit{\text{ntot}}}{\left(\mathit{OB}{S}_n-\mathit{MN}\right)}^2$$ 3

After calibrating the simulated TAPM data with the real Yazd station data, in the next step, using the statistical methods presented in Table 1, the validity of the simulated data calibrated with the real station data was evaluated and the results are presented in Table 2.

Table 1.

Statistical relations used in evaluation of simulated results with real data (Ghanghermeh et al., 2015)

MAE	RMSE	MBE	CD
$\mathit{MAE}=\frac{1}{N}\sum \left|P_i-O_1\right|$	$\mathit{RME}=\left[\frac{1}{N}\sum {\left(P_i-O_i\right)}^{0.5}\right]$	$\mathit{MAE}=\frac{1}{N}\sum P_i-O_i$	$\mathit{CD}=\frac{{\sum }_i^n{\left(P_i-O_i\right)}^2}{{\sum }_i^n{\left(O_i-\overline{O}\right)}^2}$
SI	R-square	F statistic	EF
$\mathit{SI}=\left(\frac{\mathit{sd}\mathit{\text{of\hspace{0.17em}error}}}{O_i}\right)\times 100$	$R^2={\frac{\left[{\sum }_{i=1}^n\left(O_i-\overline{O}\right)\left(P_i-\overline{P}\right)\right]}{{\sum }_{i=1}^n{\left(\left(O_i-\overline{O}\right)\right)}^2{\left(\left(P_i-\overline{P}\right)\right)}^2}}^2$	$f=\frac{{\sigma }_o^2}{{\sigma }_p^2}$	$\mathit{EF}=\frac{{\sum }_i^n{\left(O_i-\overline{O}\right)}^2-{\sum }_{i=1}^n{\left(P_i-\overline{O_i}\right)}^2}{{\sum }_{i=1}^n{\left(O_i-\overline{O}\right)}^2}$

Table 2.

Validation of simulated temperature and relative humidity values using the TAPM model with real data

Years	ERORR	Air temperature	Relative humidity
	MAE	1.47	6.77
	RMSE	1.98	9.81
	MBE	0.00	0.00
2003	CD	0.01	0.11
	SI	9.64	37.24
	R2	0.96	0.67
	F	1.09	1.15
	EF	0.96	0.64
	MAE	1.36	9.36
	RMSE	1.82	13.18
	MBE	0.00	0.00
2006	CD	0.01	0.14
	SI	8.91	46.14
	R2	0.97	0.65
	F	1.19	1.46
	EF	0.99	0.86

Statistical methods, e.g. Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Bias Method (MBE), Correlation of Determination (CD), Scatter Index (SI), rate of R-Square, F-Statistic, Simulation Efficiency Method (EF), and Durbin- Watson (DW) test, which will be briefly mentioned in the test (Table 1).

In these P_i equations, the simulated rates, O_i, rates are the number of the statistical days under study (Ghanghermeh et al., 2015).

Shannon entropy Indicator

One of the objectives of this paper is to investigate the irregularity or balance of temperature and relative humidity values under the actual conditions of the topography and compare with the scenario of land-use change when applies in Yazd. Therefore, one of the strategies for achieving this goal is to use the Shannon entropy method. Entropy concept plays a major role in physics, social sciences and information theory, in such a way as to indicate the amount of uncertainty (the degree of unbalanced distribution of phenomena) in the expected information content of a message. In other words, entropy in information theory is a criterion for the amount of uncertainty (balance) expressed by a discrete probability distribution (p_i) such that this uncertainty, if distribution spreads out, is greater than when the frequency distribution is higher. This uncertainty (the degree of unbalanced distribution of phenomena) can be described by the eq. (4) (Ghanghermeh et al., 2013):

$$H_n=\sum _i^n{P}_{i}log\left(1/P_i\right)$$ 4

Where H_n is the output of Shannon entropy, and the P_i represents the probability of occurrence of temperature or relative humidity, as components of climate, for the first month to the twelfth month of each of the study periods 2003 and 2006. To implement Shannon entropy, it was necessary that daily values of temperature and relative humidity components were available for both real and simulated conditions during study periods 2003 and 2006. In order to obtain the value of the Shannon entropy, based on daily data, the monthly mean of each of actual and simulated climate variables was first extracted and based on the monthly mean of the Shannon’s entropy values was calculated. The value of the Shannon entropy varies between 0 and ln (n). In the present study, the value of 0 represents the lowest anomalies and the variability of the given climate variable, while the value of ln (n) represents the highest variability. When the value of entropy is higher than the value of ln (n), the rate of climate anomalies is more severe.

Research findings

Validation and calibration of the results of climate Modelling

As can be seen in Figs. 2a, c, and 3a, c, the simulated relative temperature and relative humidity values are displayed using TAPM software with real values without calibration. Although the TAPM software has been able to simulate fluctuations and values of relative humidity and temperature, in some days, differences between real and simulated real data are seen. But after applying relation (2), the differences between the real and the simulated values is reduced, which can be seen according to Figs. 2b, d, and 3b, d. The results of the Nash-Sutcliffe statistics show that for the temperature values in 2003 and 2006, the output of this statistic after calibration was 0.96 and 0.97, and for the relative humidity at 2003 and 2006, this value was 0.65 and 0.66, respectively which confirms the validity of the calibration results from the fit of data simulated based on real data. It is worth noting that the values above 0.5 in the Nash-Sutcliffe statistic confirm the validity of the calibration results. The results of the validation of simulated values calibrated with real data in 2003 and 2006 are given in Table 2. As can be seen, the output of all model error estimation, statistics justifies the acceptable simulation of the TAPM model for modelling the climate component of temperature and relative humidity. However, the TAPM model seems to have less error in simulating temperature values than relative humidity.

Fig. 2.

Comparison of the real values of the mean daily temperature and simulated one (a) and the real values of the mean daily temperature and simulated values after calibration (b) for 2003 (c) and the real values of the mean daily temperature and simulated values after calibration (d) for 2006

Fig. 3.

Comparison of real values of relative and simulated mean humidity (a) and real values of relative and simulated mean humidity after calibration (b) for 2003 (c) and real values of relative and simulated mean humidity after calibration (d) for 2006

Analysis and comparison of climatic components for real conditions and simulated scenarios

In this step of the study, the average monthly mean values of temperature and relative humidity changes for the real conditions and simulated scenarios are compared. It needs to be explained that a general average of real data values in 2003 and 2006 has been calculated, and the same process has been performed for the Forest Zone Simulation Scenario. However, initially, the outputs showed that considering the establishment of a forest zone in the west of Yazd station, the average annual temperature will be 27.49 °C. If there is no forest zone and in real conditions, the average annual temperature with respect to the study period is 18.29 °C. But based on the simulated scenario, the minimum average temperature is 17.18 degrees Centigrade in January up to a peak of 36.52 degrees in August. While in real terms, the minimum average temperature for the January is 6.29 °C, but the maximum average temperature is 29.68 °C in July. Therefore, in the case of a simulated forest cover scenario, a time shift is occurring in the maximum monthly average temperature pattern of Yazd. Other considerations include the fact that there is no temperature decrease in any month of the year if there is a forest zone, and temperatures rise in all months. But noteworthy is that the lowest temperature rise was observed in late spring (6.33 °C in July), early summer (in June, the temperature increased by 7.78 °C) and late summer (in September, at 7.86 °C). However, the high temperatures rise from mid-autumn to early winter, respectively, for December with 11.78 °C, November at 11.51 °C and January 10.89 °C compared to real conditions. The continental and the dryness of the Yazd station cause cold and dry weather in the autumn and winter seasons. But the simulation shows that with a forest cover, the temperature increase of cold seasons is more than that of the warmer seasons. But it is necessary to explain that the dry and wet conditions of Yazd station lead to lower winter temperatures and rising summer temperatures, which results in a difference and thermal fluctuation between the monthly average of the maximum and the minimum annual temperatures. However, it is expected that in the case of land use change and forest cover, severe thermal stress in the study area will decrease, such that the existence of forest cover can be a thermal insulation of high summer temperatures and rapid loss of temperature over cold hours in winter.

Given the findings, it is clearly determined that the monthly temperature difference in the real conditions is 23.39 °C, which, according to the forest cover scenario, this difference decreased to 19.34 °C (Fig. 4).

Fig. 4.

Comparison of the average monthly temperature of the real series of climatic data and simulated time series of data

However, the findings for relative humidity variations showed that given the application of forest zone scenario, the average annual relative humidity would reach 41.45%. However, in the absence of a forest zone, and in real terms, the average annual relative humidity would be 28.11%. But based on the simulated scenario, the monthly average fluctuation of relative humidity is from a minimum of 25.44% in June to its maximum of 56.67% in January. However, based on the real conditions, the minimum and maximum monthly average relative humidity are 14.94 and 45.20%, respectively as it is the case in the simulated scenario for July and January. Therefore, it can be seen from the findings of this section of the study that the minimum and maximum relative humidity differences in the real and simulated conditions are approximately the same, so that in real terms it is 30.27% and for the simulated conditions it is 31.23%. The outputs show that although in the presence of a forest zone, humidity increases in all months, the minimum relative humidity increases with 6.8% in the late winter and early spring, and the maximum increase in relative humidity to the values 21.80 and 26.27% respectively for late summer (September) and early fall (October) (Fig. 5). As described in the previous section, forest cover can prevent thermal stress and in the next step, it can modify the ambient temperature and ultimately increase the temperature, especially in the colder months of the year. Therefore, this increase in temperature can in turn increase the reception capacity of atmospheric moisture. On the other hand, evapotranspiration from forest cover is a factor in increasing the relative humidity of the atmosphere in the study area. Pearlmutter et al. (2009) [40] and Shashua-Bar et al., (2011) [41] reported that in warm and dry urban areas such as city centres, cooling from evapotranspiration of green space and trees can dramatically improve the thermal comfort of humans.

Fig. 5.

Comparison of the average monthly relative humidity of the real series of climatic data and time series of simulated data

Monitoring the irregularity or balance of climate variables using the Shannon entropy Indicator

In this part of the study, based on the methodology described earlier, the value of Shannon entropy for simulated and real data in the two periods 2003 and 2006 was calculated and then the results were compared to each other. As seen in Fig. 6, the Shannon entropy value was for real temperature data was 0.78 in 2003, which was 0.65 for simulated data. This value was also calculated to be 0.77 and 0.67 for real and simulated data in the study period 2006, respectively. But for climate variable of the relative humidity in both periods 2003 and 2006, the calculated entropy value for data under scenario of the land-use change was less than of that based on real data so that in 2003, its value for real data were 0.87 and for data under the scenario of the land-use change (simulated) was 0.75, and in 2006 the real and simulated data values were 0.82 and 0.70, respectively. The previous discussion showed that the low entropy value caused a reduction in anomalies and variability. Therefore, the results of this section indicated that forest cover played a role in the reduction of the climate variability of temperature and relative humidity components. So, based on the scenario of the land-use change, forest cover could be led to a decrease in the extreme data domain, and thus reduced the variability of the given climate variables. Therefore, the results of the process showed a reduction in the Shannon entropy value of simulated climate values under the scenario of the presence of the water zone in comparison with the real climate data in Yazd.

Fig. 6.

The coefficients of entropy index for the studied climate variables

Analysis of modelling results for different models of bioclimatic design

The results of modelling the new climate conditions despite forest vegetation show that air temperature and relative humidity have increased in both years. In fact, the presence of forest mass near Yazd city has led the climate from warm and dry to warm and humid. As mentioned above, significant relative humidity has reduced the extreme temperature variations in the dry climate of Yazd and increased the average annual temperature. By implementing these changes on a psychrometric chart, bioclimatic design strategies were identified based on the ASHRAE standard 55. Figure 7 shows the time-lapse data on the psychrometric data for 2003 and 2006, as well as the simulated conditions for forest near the city for these two years. Data transfer from warm and dry conditions to warm and humid conditions is well visible.

Fig. 7.

A psychrometric diagram to illustrate the strategies for bioclimatic design based on real and simulated climatic data for study courses: (a) the real conditions in 2003; (b) the simulated conditions in 2003; (c) the real conditions in 2006, and (d) simulated condition in 2006

By extracting the time required for each of the bioclimatic design strategies of the above diagrams, the differences between real and simulated conditions are well seen in Table 3. In this table, rows 2 to 8 show cooling strategies in hot year conditions. In this set of strategies, the Sun Shading of Windows, Direct Evaporation Cooling and Two-Stage Evaporative Cooling hours have increased in simulated conditions. Due to the average temperature increase in new conditions, the need to avoid sunlight also increases. Evaporative Cooling can also be effective in reducing the temperature of the air. But High Thermal Mass, High Thermal Massive Flushed, Natural Ventilation Cooling and Fan-Forced Ventilation, Cooling have been reduced. In fact, with the increase in humidity, the temperature difference between the temperature of the day and night is reduced, and there is no need for long-term storage of heat, because heat inertia is required in low-climatic buildings. Therefore, in the construction of buildings, it is better to use materials with a lower density and weight instead of using high thermal mass material. Whenever the temperature and humidity in the air increase, so that the environment passes through the threshold of the boundary of sultry situations, ventilation cannot alone provide comfort. Therefore, in such situations, the need for ventilation, especially during the warm hours of the day, has to be reduced and air conditioning systems must be used. It is necessary to explain that the moisture content of 12 g/kg is defined as the limit of human sultry conditions [42]. So, as per the psychrometric charts for real conditions, no data have exceeded the Sultry limit of the year, while in the simulated forest scenario, about one-third of the events is located beyond the sultry limit (Fig. 8). Rows 9 through 12 show heating strategies in cold weather. The hours of the Internal Heat Gain strategy have increased, but shows a decline in Passive Solar Direct Gain hours. Also, the need for Wind Protection of outdoor spaces is zero in all conditions. It is natural that with increasing air temperatures, the need for solar direct gain is reduced in winter conditions. According to bioclimatic diagrams, the temperature range requires an Internal Heat Gain, from 14 to 20 degrees Celsius and a temperature range requiring Passive Solar Direct Gain is 7 to 14 degrees Celsius. Therefore, the internal Heat Gain strategy has also increased with the average winter temperature fluctuating from 8 to 18 degrees (Fig. 7).

Table 3.

Comparison of different hours required for different strategies for bioclimatic design, based on the real series of climatic data and simulated forest cover scenarios for the two study periods

Bioclimatic Zones	Design Strategies	Real data-2003	Simulation data-2003	Real data-2006	Simulation data-2006
1	Comfort	2293	1813	2589	1519
2	Sun Shading of Windows	1372	2080	1262	1974
3	High Thermal Mass	296	16	533	16
4	High Thermal Mass Night Flushed	429	20	631	22
5	Direct Evaporation Cooling	1937	3073	1167	1201
6	Two-Stage Evaporative Cooling	1937	3879	1167	1504
7	Natural Ventilation Cooling	735	472	675	466
8	Fan-Forced Ventilation Cooling	439	262	418	171
9	Internal Heat Gain	1626	1943	2053	2199
10	Passive Solar Direct Gain Low Mass	1125	590	1138	676
11	Passive Solar Direct Gain High Mass	1326	603	1532	664
12	Wind Protection of outdoor spaces	0	0	0	0
13	Humidification only	0	0	0	0
14	Dehumidification only	0	17	0	128
15	Cooling, add Dehumidification if needed	0	1061	0	3251
16	Heating, add Humidification if needed	1542	1	1637	48

Fig. 8.

A psychrometric diagram to display the sultry limit based on real and simulated climatic data for the study periods: (a) the real conditions in 2003; (b) the simulated conditions in 2003; (c) the real conditions in 2006, and (d) simulated conditions in 2006

Rows 13 to 16 define the humidification strategies. The Humidification only strategy does not exist in any of the conditions. But dehumidification only, which does not exist in real conditions, increases in simulated conditions, because the humidity of the environment has increased. The Cooling and Dehumidification if required strategy, which is, in real condition, zero, has greatly increased under the conditions of the forest environment simulation. This is because both the average temperature and humidity have risen and there is a need for cooling coupled with dehumidification. The Heating and Humidification if needed strategy, which was considerably higher in existing conditions, has dropped to near zero in the new simulated conditions. This is due to an increase in the mean time of the average winter temperature and humidity, which reduced the need for heating combined with humidification.

The findings showed that maintaining comfort conditions in the simulated situation are less frequent than real conditions. The percentage of the incidence of comfort was reduced from 26.6% to 20.7% for 2003 and from 29.6% to 17.3% for 2006 (Figs. 9 and 10). It can be said that in the simulated conditions, the average temperature of cold seasons has increased, but has not reached the thermal comfort threshold; on the other hand, the average temperature of the warm seasons has increased, which, in combination with increasing moisture, has caused considerable sultry conditions. So these winter and summer changes have reduced the hours of thermal comfort. A review of the results suggests that total cooling strategies will be increased throughout the year and the heating strategies will be reduced. Also, the total of strategies whose results are dehumidification increases as well. It seems that under the forest cover scenario, the winters of Yazd are more favourable than the real conditions, and the summer experiences more unfavourable conditions due to increased temperature and humidity.

Fig. 9.

Comparison of bioclimatic design strategies based on a simulated forest cover scenario with real climatic data for 2003

Fig. 10.

Comparison of bioclimatic design strategies based on the simulated forest cover scenario with real climate data for 2006

Adaptive comfort model output analysis

The results of this section indicate that for the simulated conditions of forest cover in both 2003 and 2006, the frequency of hours in the temperature range of 17 to 31 is more compared to the real condition as 4124 h have been observed in the real data in 2003 while hours has increased 4170 due to the forest cover scenario. On the other hand, for real data of 2006, there are about 4495 h and for the forest cover scenario 4651 h are in the range of 17 to 31 degrees. Therefore, as discussed in previous sections of this study, given the effect of forest cover on the moderate temperatures of cold seasons, higher number of hours in a year can be within the range of adaptive comfort model.

Changes in the number of adaptive comfort hours in simulated conditions are different in different seasons than real conditions. Studies have shown that in the spring and summer season, the number of hours with adaptive comfort has decreased, but not in autumn and winter (Fig. 11).

Fig. 11.

The number of hours with adaptive comfort capability for real conditions in 2003 and 2006 and the simulated conditions of forest cover by seasons

Figure 12 shows the range of adaptive thermal comfort using natural ventilation for the years of 2003 and 2006 in real and simulated conditions. It is noticed that the lower and upper limits of the adaptive comfort temperature have increased about 2 and 1 degrees in the simulation conditions compared to the real conditions due to an increase in average ambient temperature throughout the year. Also, due to the increase in humidity in the air and the creation of conditions similar to the sultry conditions during the hours of the year, the percentage of thermal comfort in the new situation has decreased in comparison to the current situation. In fact, the study results show that adaptive thermal comfort is more achieved by natural ventilation in warm and dry conditions and created in warm and humid conditions due to the high water vapor in natural draught while adaptive thermal comfort is provided less for the residents of the building.

Fig. 12.

A psychrometric diagram to determine the adaptive comfort range using natural ventilation based on real and simulated climatic data for study courses (a) real conditions in 2003; (b) simulated conditions in 2003; (c) real conditions in 2006; (d) simulated conditions in 2006

Discussion

In the present study, TAPM model was used to simulate the climatic components in a way that the results of validations indicated the acceptable ability of this software to simulate temperature and relative humidity components. In some studies, the performance of TAPM model has been evaluated so that the obtained results confirmed the acceptable effectiveness of this software in the simulation of climatic components [19, 43–46]. Based on this study, it was determined that due to the land use change and simulation of forest cover in western Yazd, this factor would cause climate change at this station. So, in terms of forest cover for the total average of two simulated periods in 2003 and 2006, the temperature reached 27.49 degrees, which in real terms was 18.29 degrees centigrade. On the other hand, the presence of forest cover led to an average annual fluctuation of temperature of 4 degrees centigrade.

In many studies, the effects of land use change on climate change have been analysed in the local or regional dimension, and the role of ground effects as a major component of the climate system has been confirmed in climate change. Research by Atwater, 1974 [47]; Changnon, 1981 [48]; Cotton and Pielke, 1995 [49]; Baik and Chun, 1997 [50]; Tumanov, 1999 [51] has shown that in most large and industrial cities, urbanization has dramatically changed the atmospheric parameters and features of the Earth’s surface, resulting in local and regional climate change. Using TAPM software, in their study, Khoshakhlagh, et al. (2013) [52] evaluated the effect of drying Urmia Lake on temperature changes in Maragheh city. The results of their study showed that the average annual temperature of Maragheh station increased 0.25 Degrees Centigrade due to drying of the Lake and in the warm months of the year, especially August and July, in the middle hours of the day, the average increase of about 4 degrees Celsius as well, especially in the cold months of December and January, the average temperature decreased about −3.5 degrees Celsius at night.

The first experimental researches on boreal deforestation have been conducted with energy balance models. Otterman et al. said that the surface air temperature of the northern hemisphere without trees was 1.9 °C lower than the same hemisphere fully forested of the taiga/tundra boundary [53]. Two researches showed that eliminating the boreal forests would reduce the spring surface temperature up to 2.8 °C and would delay the snowmelt time by a month [54, 55].

The results of a comparative study of the microclimate above a 35 m tall forest canopy in an undisturbed primary forest in Amazonia is that the cleared site average temperature range was nearly twice that of the forest. Lower night-time temperatures were found in the cleared site because of the reduced vertical mixing, and higher daytime maximum temperatures occurred, especially during dry periods, because of greater water stress, reduced evaporation, and increased convection [56]. Lee et al. investigated differences of measured temperatures between regions near the forested and non-forested areas [57]. The temperature differences are dependent on the latitude of the area, so that the forested regions would become warmer relative to the open areas when latitude increases. Researchers show that forest vegetation reduces mean annual temperature 1.5 m above ground from 0.45 to 1 °C depending upon the property of forest and locality, especially the elevation [58]. In a research done in Singapore urban environment, surface temperature reduction by grass surfacing and adding more trees is also obvious, with a reduction by up to 8 °C for grass and 10 °C for trees [59]. But in other research on the same location, the maximum air temperature reduction by planting more trees can be up to 0.7 °C when compared with the case with no vegetation. On the other hand, the mean radiant temperature reduction caused by planting more trees is notable in such urban areas. This research showed that the mean radiant temperature of the points under trees can be 40 °C lower than the points with no vegetation [60]. It should be noted that the impacts on annual mean temperatures varied between the models [61]. As seen above, there are many differences in the results of studies, depending on vegetation or latitude. However, the results of most researches have shown that decreasing the forestry temperature occurs only at low altitudes. In experimental results and simulations, increasing the temperature is shown in high levels of forest cover in relation to free fields. By artificially raising the temperature in a section of the forest around 5 °C, Melillo confirmed previous studies which showed an organic decomposition rates increase by higher temperatures and thus released more CO₂ into the atmosphere [62]. One of the reasons for increasing the temperature of forest environments in hot areas in the summer seems to be the increase in CO₂ production and the greenhouse effect on the forest floor. Bonan et al. replaced all forest vegetation north of 45^oN with bare ground or with tundra and simulated boreal deforestation by a change in the surface albedo [63]. Their results showed that the surface air temperature and atmospheric moisture always increase by creating the forest. Past deforestation has resulted in cooler temperatures by changing albedo in temperate regions [64]. So, because of the change in albedo, and the increase in receiving sunlight, forest cover can increase the temperature. Generally, forests increase temperature in the short term, but decrease the impact of heat waves in the long term, because the trees can access deep reserves of underground water and continue to live even when there is not any moisture in the upper soil levels [61]. It can be said that because of the simulation of the present study at two short periods of 2003 and 2006, the annual temperature increase in this region can be seen, while simulation results in longer periods could indicate a decrease in temperature.

Conclusion

The present study sought to identify the effect of land use change on regional climatic variation and, finally its effectiveness in the changes of in-door bioclimatic designing strategies.

Based on the results of this work, it was found that, in the real site, the minimum and maximum temperatures for the period from October, 12 to December, 10 were 21.8 and 31.4, while at the same time, the minimum and maximum temperatures above the forest were 23.9 and 29.6, respectively. But the average temperature of this time in the real site was 25.8 and above the forest was 26.5 degrees Celsius. In fact, despite the difference in temperature range, the average temperature above forest was higher than the real site. Therefore, the results of this study, which include the increase in temperature at all times of the year, can be due to the following reasons:

1. Increase in temperature at upper levels of forest mass, 2. Increase in CO₂ production, 3- Change of albedo forest range, 4. Short term effects of forest, and 5. West prevailing wind.

On base of this work, for the city of Yazd, it was determined that compared to the absence of forest cover, in the presence of forest cover, relative humidity would increase, which could be affected by an increase in the process of evapotranspiration in the study area, which resulted in an increase in relative humidity. However, given the forest zone scenario, the annual average relative humidity increased to 41.45%. In the absence of a forest zone, in real terms, the annual average relative humidity is 28.11%. In fact, the presence of forest mass near Yazd city has led the climate from warm and dry to warm and humid. In other similar studies, the existence and the absence of some of the effects of land on changes in rainfall patterns and moisture have been analysed. As in the present study, the role of land use changes in the effects of rainfall and humidity component changes has been demonstrated. For example, Ahmadi Givi et al. (2005) [65], using the RegCM regional climatological model, investigated the role of the Zagros Mountains on atmospheric systems in Iran over a quarterly period. In their research, by removing the Zagros Mountain range, it was determined that rainfall and relative humidity in the central and eastern regions of Iran increased but its value remained constant throughout the simulated range. By implementing the resulting climatic components of the forest cover scenario on the psychrometric charts, changing the pattern of data migration from warm and dry conditions to warm and humid conditions are well visible. Findings about the changes in bioclimatic strategies in real conditions and forest cover scenarios show that if there is a forest cover, the frequency of hours that require cooling strategies to apply the solutions of Sun Shading of Windows, Direct Evaporation Cooling and Two-Stage Evaporative Cooling are incremental. So that the total increase for these strategies for 2003 is 3786 h and for 2006 there are 1083 h. The findings also showed that, among the building heating solutions, only the Internal Heat Gain solution will increase by 317 h for 2003 and 146 h for 2006, depending on the forest cover scenario. Further, according to the forest cover scenario, changes in wet strategies are such that for Dehumidification only and Cooling, add Dehumidification if required, total changes of 1078 and 3379 h have an increasing trend for 2003 and 2006, respectively. In the present study, it was found that in the case of forest cover scenarios, a general change in the building climatic design would take place given the cold and hot seasons. In general, in the cold seasons, reduced heating strategies and in the warm seasons, increased cooling strategies would be required. In some studies, the effect of land use change on the change in energy consumption patterns was investigated. In these studies, as in the present study, it was concluded that changing demand for energy would change due to land use change. For example, in a study, the existence and absence of a water zone in the degree-day calories, and in particular the need for the degree-day heating and cooling were investigated. In that study, for Urmia Lake, it was found that in the absence of a lake, in other words, the dryness of the lake, the decrease in temperature in cold seasons and the increase in temperature in the warm seasons lead to an increase in cooling energy demand during the warm year and an increase in heating energy during the cold period [19]. Also, Roshan and Shakour (2018) [66], using TAPM software, simulated two drying and watering scenarios of Zarivar Lake in the northwest of Iran. However, based on the lake drying scenario, HDD levels increased for coldest months as compared to the water-lake scenario, and, on the other hand, the amount of CDD were also increased in the warmer months of the year. Also, based on the annual average, the findings indicated that the drying of Zarivar Lake would reduce the calorific value of 30 degree-day calorie in the heating sector and 111 degree-day calorie increase in the cooling section. Thus, the results obtained from this study show that land use change has a significant impact on climate change and, finally on the building bioclimatic design. Accordingly, this factor can be considered as an important factor in locating new towns and settlements.

Compliance with ethical standards

Competing interests

the authors declare that they have no competing interests