Research on Microclimate Influencing Factors and Thermal Comfort Improvement Strategies in Old Residential Areas in the Post-Urbanization Stage

Abstract

China’s urbanization process has entered the stage of mid-to-late transformation and upgrading, with the urbanization and population growth rates having passed the turning point. Urban renewal has become an increasingly important issue, among which the renovation of old residential areas holds enormous potential. The improvement of the living environment is urgent, and enhancing the microclimate to improve the livability and comfort of outdoor residential spaces is a critical factor. This study presents for the first time a quantitative framework for multifactor synergistic optimization by coupling building layout closure and material albedo effects. This paper takes typical old residential areas in Fuyang as an example and uses 3D microclimate simulation software (ENVI-met Version 4.3) to establish a simulation model. It evaluates the microclimate and thermal comfort under different building layouts, green infrastructures, building envelope materials, and various surface materials. The results show that: (1) Regarding building layout, the point-cluster layout generally results in the best improvement of daily cumulative physiological equivalent temperature (PET) values, followed by row-type and enclosed layouts; (2) The optimal solutions for improving the daily average PET value are as follows: using glass as the building envelope material in the point-cluster layout; 100% tree coverage in the row-type layout; and 100% asphalt coverage as the surface material in the point-cluster layout. These three conditions reduce the daily average PET by 3.51 °C, 23.87 °C, and 2.65 °C, respectively; (3) The degree of impact on PET is ranked as: green infrastructure configuration > building layout > building envelope materials > surface materials; (4) When the building layout of the residential area is more enclosed, such as using row-type or enclosed layouts, the order of building envelope materials improving thermal comfort is: brick, concrete, and glass. When the building layout is less enclosed, such as using point-cluster layouts, the order of building envelope materials improving thermal comfort is: glass, brick, and concrete. Therefore, it is concluded that applying point-cluster layout in buildings, using glass as the building envelope material, and having 100% coverage of asphalt pavement as the surface material and 100% coverage of trees can maximize the improvement of the thermal environment of the buildings. The conclusion is applicable to old residential areas in warm temperate semi-humid monsoon climatic zones characterized by high densities (floor area ratios > 2.5) and high rates of hardening of the ground (≥80%), and is particularly instructive for medium-sized urban renewal projects with an urbanization rate between 45% and 60%.

1. Introduction

Over the past few decades, global urbanization has accelerated rapidly, with large numbers of people migrating from rural areas to cities, leading to significant urban expansion and dramatic changes in the urban landscape. In China, urbanization has achieved remarkable progress, with the urbanization rate rising from 17.9% in 1978 to 66.16% in 2023 [1]. As China enters the mid-to-late transition stage of urbanization, the focus of urban development has shifted from large-scale expansion to a dual emphasis on urban renewal and optimization. As a crucial component of urban stock space, old residential areas are playing an increasingly important role in this process.

The rapid development of urbanization is a key indicator of economic and social progress, driving economic prosperity and societal advancement. However, alongside the benefits of urbanization, its rapid pace has also led to a range of negative consequences. Among these, the urban heat island (UHI) effect has significantly deteriorated the urban living environment, causing rising temperatures, poor air quality, and a decline in thermal comfort indices [2]. In recent years, research on building microclimates has primarily focused on urban blocks [3], campuses [4,5,6], urban parks [7,8], urban canyons [9,10,11], and transportation hubs [12,13]. However, studies on the microclimate of old residential areas remain limited. As rapid urban development exposes the environmental challenges of these areas, these issues increasingly impact residents’ quality of life and pose obstacles to sustainable urban development.

Previous studies have shown that in high-density urban environments, meteorological variables around buildings, such as solar radiation, air temperature, and wind speed, are significantly influenced by urban morphology, surface materials, vegetation density and type, as well as other anthropogenic factors affecting surface energy and water balance [14]. For example, Jiang et al. [15] compared six different residential layout patterns in high-temperature and high-humidity regions and concluded that a closed layout can effectively block extreme heat, lower air temperature, and provide significantly better thermal comfort than a parallel layout. Karimimoshaver et al. [16] found that in Tabriz, a residential layout with a building orientation of 135° and a building height-to-spacing ratio of 1.5 offered the best outdoor thermal comfort. Zhao et al. [17] discovered that in arid cities, an equidistant tree arrangement provided the greatest microclimate and thermal comfort benefits, followed by cluster arrangements without crown overlap. Additionally, Lee et al. [18] highlighted that building layout and orientation play a crucial role in influencing outdoor ventilation.

To address the microclimate challenges in old residential areas, many scholars and urban planners have conducted relevant research and practical explorations. Some studies have focused on improving the microclimate by increasing green space, as urban vegetation is widely recognized as an effective strategy for sun shading, cooling, guiding airflow, capturing precipitation, and facilitating evapotranspiration [19]. Additionally, outdoor thermal environment (OTE) [20] can be enhanced by lowering air temperature through evaporative cooling and convection. Due to water’s high thermal capacity, its presence helps reduce surrounding temperatures while decreasing surface radiation emissions [21]. Other research has concentrated on energy-efficient building renovations, such as applying thermal insulation to exterior walls and replacing conventional doors and windows with energy-saving alternatives to enhance insulation performance and minimize indoor-outdoor heat transfer. However, most existing studies and practical implementations focus on improving individual factors, lacking a systematic analysis of the various microclimate influences and comprehensive optimization strategies for old residential areas.

In conclusion, conducting an in-depth study on the factors influencing microclimates and developing optimization strategies for old residential areas in the post-urbanization stage holds significant practical value. This paper aims to comprehensively and systematically analyze the various factors affecting the microclimate of old residential areas and propose targeted optimization strategies. By doing so, it seeks to provide both theoretical support and practical guidance for enhancing the living environment of residents and improving overall urban livability. To achieve this, we will conduct field research and data analysis on several representative old residential areas, thoroughly exploring the intricate relationships between microclimate factors. Additionally, we will validate and refine the proposed strategies through simulation experiments and other analytical methods. Ultimately, this research aims to serve as a valuable reference for improving microclimates and guiding urban renewal efforts in old residential areas.

2. Materials and Methods

2.1. The Research Subject

Fuyang is a prefecture-level city in Anhui Province, China, located between 114°52′ and 116°49′ east longitude and 32°25′ to 34°04′ north latitude. Figure 1 illustrates the regional divisions of Fuyang. The city experiences a warm temperate semi-humid monsoon climate, characterized by distinct monsoon influences, four distinct seasons, mild temperatures, and moderate rainfall. Currently, Fuyang’s urbanization rate stands at 45.16%, indicating significant potential for further urban development [22]. This study selects a typical old residential area in Fuyang as the empirical object, whose typicality is reflected in the following: (1) Climatic characteristics—a warm temperate semi-humid monsoon climate (average annual wind speed of 1.2 m/s, solar radiation of 4820 MJ/m²), with high temperature and high humidity in the summer (extreme temperature of 41.2 °C), and cold and dry composite heat stress characteristics; (2) Urbanization stage—as one of the cities with the fastest growth rate of urbanization in the central region, the distribution of its old residential areas presents “double-high” characteristics: high building density (average floor area ratio of 2.8), and high rate of hardening of the subsurface (≥85%), which reflects the contradictions of stock renewal in the post-urbanization stage. Given these considerations, this study selects Fuyang as the research subject to explore and address these urban microclimate issues.

2.2. ENVI-Met Software Simulation

Various software tools are available to simulate heat loads transferred through surfaces, such as Energy Plus, IDA ICE, TRNSYS, WUFI, and several studies have verified the accuracy of these software based on local climatic data, but there are some limitations to the application of these software. For example, Energy Plus often results in a discrepancy between the surface temperature and the actual temperature of up to 9.9~20 °C [23,24]. Trnsys modelling greatly overestimates the surface temperature of glass curtain walls [25]. Although RADI-ANCE can accurately calculate shortwave radiation, it does not adequately support dynamic exchange of longwave radiation, resulting in seasonal prediction errors of 2–6 °C [26]. In contrast, ENVI-met, as a three-dimensional computational dynamics model, can calculate a variety of variables that affect the OTE, including relative humidity (Rh), mean radiant temperature (MRT), air temperature (Ta) and wind speed (Ws) [27], and is able to provide the user with detailed data on the climate environment, which significantly improves the prediction accuracy in complex urban environments, outperforming conventional tools especially for semi-enclosed courtyard scenarios [28]. In addition, the software has built-in thermal parameters (albedo, heat capacity, emissivity) for more than 300 building materials. For the complex spatial morphology of high-density old residential areas, its three-dimensional non-uniform grid system can effectively portray the shadow casting on the building façade, which provides irreplaceable and refined modelling capabilities for the study of building energy consumption, material durability and the heat island effect.

Therefore, in this paper, a three-dimensional microclimate dynamic simulation of the microclimate environment in urban residential areas was carried out using ENVI-met software (Version 4.3). Typical materials such as concrete (albedo 0.3), asphalt (0.12) and single-layer glass (0.4) were used in this study, and the vegetation parameters were defined by combining with the measured data of local deciduous trees in Fuyang. In the simulation of warm-temperate semi-humid climatic zones (e.g., Fuyang), the error of the temperature is less than 1.0 °C, the error of Ws is less than 0.3 m/s, and the error of the humidity is controlled within 5%. Typical applications include urban heat island intensity assessment, optimal thermal comfort design and simulation of low impact development (LID) effects.

To ensure the accuracy of the ENVI-met model, this study ensures the reliability of the simulation results through a multidimensional validation strategy. Firstly, the microclimate prediction capability of the ENVI-met software has been widely validated globally, e.g., the simulation by Jiang et al. (2020) in the high-temperature and high-humidity climate zone of Wuhan showed a Ta error of less than 1.5 °C (R² > 0.85) [15]; Liu et al. (2024) reduced the normalized root mean square error by 57% in courtyard microclimate simulations by improving the Indexed View Sphere and 6-directional mean radiant temperature algorithms [29]. Secondly, the parameter settings in this study strictly follow the principle of localized calibration, and key parameters such as building surface reflectance and vegetation transpiration efficiency (Leaf Area Index = 2.5, transpiration coefficient = 0.8) refer to the measured database of typical building materials in Shanghai (e.g., concrete reflectance range 0.3–0.4). The average deviation of ENVI-met simulation results from urban meteorological station data in the high-density urban area of Hong Kong is 1.2 °C, which also validates the applicability of the model in hot and humid climate zones [30]. The above methodology has been successfully applied in studies of similar climate zones such as Nanjing, with cooling load prediction errors generally below 15% (R² = 0.91) [31]. The results demonstrate that the software is widely used to assess urban micrometeorological conditions and can accurately reflect the interactions between microclimates and local environmental factors [19].

Residential buildings vary widely, and their environments are often quite complex. To reduce this complexity and minimize the interference of extraneous factors in research, researchers typically simplify the environment during the simulation process, allowing them to focus on the primary issues at hand. For instance, Zhou et al. used ENVI-met software to investigate the coupling effects of river characteristics and riverside building forms on urban microclimates. To eliminate the adverse influences of other variables, the researchers simplified both the building and water models in their simulations, representing the buildings as cubes placed on a concrete surface [32].

2.3. Residential Area Model Construction

In this study, the ENVI-met software model was employed to investigate the multiple influencing factors in old residential areas, with necessary simplifications applied to the model. Through field research in Fuyang, three typical and representative cases of old residential areas were selected: Phase I of Tie’erchu Community, staff residential area of Fuyang Water Resources Bureau and staff residential area of Fuyang Sugar and Wine Company. The architectural space layouts of these areas were extracted and modeled separately in ENVI-met. Figure 2 illustrates the layout methods of these three areas in Fuyang. Building on this foundation, this study examines the impacts of blue and green infrastructure configurations, building envelope materials, and various surface materials on the outdoor microclimate of these old residential areas.

2.4. Input of the ENVI-Met Parameter

The simulation was carried out using the ENVI-met V4.3 model for three-dimensional dynamic simulation (grid resolution of 2 m × 2 m × 2 m). Meteorological parameters refer to the statistical characteristics published by the China Meteorological Data Service Center. The meteorological conditions on this date are consistent with the climatic characteristics of the extreme high temperature days in the summers of 2019–2023, and the maximum temperature on the simulated day (41.2 °C) is close to the historical extreme value of Fuyang (42.3 °C), which can represent the most unfavorable working conditions of the thermal environment in the old residential areas, and the dominant wind direction (14° SW) and Ws (3.78 m/s) on that day are in line with the law of the summer monsoon, so as to avoid the interference of the simulation results caused by the abnormal weather. Specific data are shown in

Table 1. Input of the parameters used in the configuration file in the ENVI-met.

Simulation Input	Degree (Level or Extent)	Fuyang
		32°56′ N, 115°42′ E
Simulation day		20 July 2024
Runtime information
Starting time		20:00 on 19 July
Simulation cycle		28 h
Climatological data
Ta within 2 m	max	305 k
	min	300 k
Rh ≤ 2 m	max	0.89
	min	0.68
Ws of 10 m		3.78 m/s
Wind direction		14° to the south west
Material input
Wall albedo (concrete)		0.3
Roof albedo (concrete)		0.3
Soil albedo (loam)		0.0
Water albedo		0.0
Asphalt pavement albedo		0.12
Red-coated asphalt pavement		0.4
Basalt brick road albedo		0.2
Tree parameters
Tree shape		Cylindrical, medium trunked, dense
Tree species		Deciduous tree
Tree height		15 m
Crown diameter		9 m
Foliar short-wave albedo		0.18
Leaf emissivity		0.96
Grid
Nested grid number		2
Size of the z-grid cells		2 m
Number of grids (x, y, z)		98, 93, 25
Size of the x and y grid cells		2 m
Applicable climate zones		Climate zones with hot summer and cold winter
Typical cities		Fuyang, Hefei, Xinyang

.

2.5. Schematic Design

Most studies utilize four key climatic parameters of Ta, Rh, Ws, and MRT to improve microclimate. These studies often examine various aspects of urban design, including geometry, water surfaces, and vegetation [33,34]. This study includes sixty different simulation cases.

Table 2. ENVI-met model of row-type layout, infrastructure configuration and surface material.

The Row-Type Layout
Green Infrastructure Configuration
	100%	75%	50%	25%	0%
Tree
Surface material
Concrete pavement
Brick and stone pavement
Asphalt pavement

presents the row-type layout model, and based on this layout, we established configurations for green infrastructure at levels of 25%, 50%, 75%, and 100%, along with different surface materials (concrete pavement, asphalt pavement, and brick pavement). This approach allows us to simulate the effects of green infrastructure configuration and varying surface materials on the microclimate model.

Table 3. ENVI-met model of enclosed layout, infrastructure configuration and surface material.

Enclosed Layout
Blue and Green Infrastructure Configuration
	100%	75%	50%	25%	0%
Tree
Surface material
Concrete pavement
Brick and stone pavement
Asphalt pavement

and

Table 4. ENVI-met model of point-cluster layout, infrastructure configuration and surface material.

Point-Cluster Layout
Blue and Green Infrastructure Configuration
	100%	75%	50%	25%	0%
Tree
Surface material
Concrete pavement
Basalt masonry pavement
Asphalt pavement

show models based on point-cluster and enclosed layouts, with the configuration of green infrastructure and surface materials kept constant, and the building’s general height set at 24 m.

Table 5. Material model of the external epidermis of the building.

	Concrete (Filled with Block)	Single-Layer Insulation Glass	Brick
Row-type
Enclosed
Point-cluster

demonstrates the influence of building envelope materials on the microclimate by exploring different materials (concrete, glass, and brick) while keeping other factors constant across the three building layouts.

2.6. Thermal Comfort Evaluation Index

Physiological Equivalent Temperature (PET)

Physiological Equivalent Temperature (PET) is a human biological meteorological parameter that describes individual thermal perception and is used to assess thermal comfort under various climatic conditions. This index is recognized by Verein Deutscher Ingenieure (VDI) [35]. There are four reasons for using PET in this study: (i) It is the most commonly used indicator in urban thermal evaluations and serves as the standard for urban and regional planning in Germany [19,36,37]. (ii) PET encompasses the four main climatic variables: MRT, Rh, Ta, and Ws. (iii) The results are expressed in degrees Celsius, making them easier to interpret. (iv) As a thermal physiological index, PET provides a natural representation of the climate experience.

The formula for calculating PET is given by PET = 1.2 Ta − 2.2V + 0.52 (MRT − Ta), where Ta is the air temperature, V is the wind speed, and MRT is the mean radiant temperature, which reflects both the direct radiation from sunlight and the indirect radiation from building materials and pavements [38]. The model is divided into nine evaluation grades, with their corresponding thermosensation and thermal sensation indicators presented in

Table 6. PET evaluation level [39].

PET (°C)	Thermal Sensation	Hotness
>41	Very hot	Extreme heat stress
35~41	Hot	Strong heat stress
29~35	Warm	Moderate thermal stress
23~29	A little warm	Mild thermal stress
18~23	Comfortable	Comfortable
13~18	Slightly cool	Mild cold stress
8~13	Cool	Moderate cold stress
4~8	Cold	Strong cold stress
≤4	Very cold	Extreme cold stress

.

3. Results

3.1. Changes in Ta, Ws, Rh, and MRT

3.1.1. Microclimate Change Under Different Building Layouts

Figure 3 presents box plots for each indicator from 9:00 to 17:00 across different building layout simulation scenarios, allowing for an evaluation of the changes in MRT, Rh, Ws and Ta under simulated conditions. The data for these box plots represent the averages collected during the simulation period (9:00–17:00) at a height of 1.4 m above the ground.

The changes in Ta for the row-type, point-cluster, and enclosed layouts all show an increasing trend from 9:00 to 13:00, with the average temperature reaching its highest at 13:00 and slowly decreasing thereafter, and the order of the average temperatures of the three layouts at each time point is maintained as point-cluster > row-type > enclosed. The largest change in the data range occurs at 13:00 between the row-type layout and the enclosed layout. The enclosed layout reduces the direct sun area by shading the buildings from each other, which reduces the peak daytime Ta by 0.61 °C compared to the point-cluster layout. The row-type layout, on the other hand, has a larger angle (45°) between the long axis of the buildings and the dominant summer wind (14° SW), which leads to a decrease in ventilation efficiency (average Ws of 0.3 m/s), weakening the thermal convection heat dissipation capacity, and resulting in a peak Ta of 0.21 °C higher than that of the enclosed type.

The Ws for the three building layouts gradually increased from 9:00 to 17:00. During this time, the ranking of Ws was point-cluster, row-type, and enclosed layouts. At 13:00, the average Ws for the row-type and enclosed layouts decreased by 0.05 and 0.1, respectively, compared to the point-cluster layout.

The Rh for the three building layouts exhibited a changing trend from 9:00 to 17:00, initially rising before reaching its lowest value at 13:00. This drop can be attributed to the increased solar radiation at that time, which elevates ground temperature and air saturation pressure. During the afternoon’s high-temperature period, the actual water vapor pressure increases relatively slowly, making it more challenging for water vapor to remain in the air. Consequently, some of the water vapor may evaporate into the atmosphere or dissipate through other means. At 13:00, the average Rh was measured at 49.60%, 48.91%, and 48.39% for the point-cluster, row-type, and enclosed layouts, respectively. Compared to the point-cluster layout, the average Rh for the row-type and enclosed layouts decreased by 0.69% and 1.41%, respectively.

In terms of MRT, the three simulations initially showed an increase before declining, with the peak occurring at 14:00. The MRT ranking was as follows: enclosed layout, row-type layout, and point-cluster layout. The MRT for the row-type layout and point-cluster layout decreased by 0.34 and 0.64, respectively.

3.1.2. Microclimate Change in External Surface Materials of Different Buildings

Figure 4 shows the average data of the simulated environment at a height of 1.4 m above the ground for different building envelope materials. Since the changes in MRT, Rh, Ws, and Ta are most significant at 13:00 during the day, a box plot for the simulation time of 13:00 is selected to represent the impact of building envelope materials on the microclimate.

The building envelope materials do not influence the variation in Ws. The differences in mean Ws are solely attributed to the layout of the buildings. In fact, the mean Ws remains consistent across the use of glass, concrete, and brick, with no discernible change.

The changes in Ta across all cases are minimal; however, the average temperatures generally follow the order of glass, concrete, and brick. For instance, in the row-type layout, the average temperatures are 37.35 °C for glass, 37.29 °C for concrete, and 37.23 °C for brick. When using glass as the building envelope material, the average Ta for concrete and brick decreased by 0.06 and 0.12, respectively.

Among the three types of building envelope materials, brick exhibits the highest average Rh, followed by concrete and glass. In the simulations, the enclosed building layout shows higher average Rh compared to the other two layouts. However, the most significant changes occur in the row-type layout, where the average Rh for brick and concrete decreased by 0.15 and 0.13, respectively.

The albedo of glass (0.4) is higher than that of brick (0.2), but its low heat capacity leads to rapid release of heat after absorption during the daytime, making its 13:00 Ta 0.12 °C higher than that of brick. In contrast, the high heat capacity of brick delays the heat release, making its daytime Ta 0.06 °C lower than that of concrete (concrete albedo 0.3). It is noteworthy that the high albedo of glass did not significantly reduce the MRT (difference < 0.5 °C), presumably its lower surface longwave emissivity (0.84) weakened the radiative heat dissipation capacity.

3.1.3. Microclimate Change in Different Surface Materials

Figure 5 presents the average data from the simulation conducted at 13:00, with measurements taken at a height of 1.4 m above ground level for different surface materials in the row-type layout. Given the consistent microclimate changes observed across the three building layouts, the row-type layout is used as a benchmark to assess the impact of the surface material on MRT, Rh, Ws, and Ta.

As the proportion of the surface material increased, the Ta of concrete and brick materials exhibited a downward trend, while asphalt showed the opposite effect. Specifically, from 25% to 100%, the average temperatures for concrete and brick decreased by 0.11 °C and 0.06 °C, respectively. In contrast, the average temperature of asphalt rose by 0.1 °C. At 100% coverage, the average temperatures for these three surface materials were 37.23 °C, 37.49 °C, and 37.28 °C, respectively, with asphalt recording the highest average temperature and concrete the lowest, resulting in a difference of 0.26 °C.

The variation in Ws is entirely dependent on the presence of vegetation. Consequently, in the absence of vegetation, changes in surface material type or the proportion of surface materials do not influence Ws.

In terms of Rh, both concrete and brick materials demonstrated an increasing trend in average Rh with a higher proportion. The average Rh of concrete increased from 48.88% at 25% coverage to 49.02%, reflecting an increase of 0.14%. Similarly, the average Rh of brick material rose from 48.87% to 48.98%, an increase of 0.11%. In contrast, asphalt exhibited a downward trend; the average Rh decreased by 0.24% when moving from 25% to 100% coverage. This indicates that the use of concrete and brick materials can enhance the Rh of the building environment, with the effect becoming more pronounced as the proportion of these materials increases.

The effect of all three modelled scenarios on the MRT showed a trend of lower MRT for higher percentages, which was more pronounced for asphalt and brick compared to the concrete material, where the low albedo of asphalt (0.12) reduces shortwave reflections but its high emissivity promotes the dissipation of longwave radiation heat, resulting in a MRT reduction of 1.08 °C compared to concrete. Despite the fact that the Ta of asphalt is 0.26 °C higher than that of concrete, its MRT reduction contributes 82% to PET improvement (2.65 °C PET reduction), suggesting that controlling the radiative load weighs more heavily on thermal comfort than temperature regulation alone in hot climate zones.

3.1.4. Microclimate Change in Different Green Infrastructure Configurations

Figure 6 presents the average data collected at 13:00 under three layout conditions, each featuring different levels of green infrastructure configuration. The simulations were conducted at a height of 1.4 m above the ground, with the objective of examining the impact of tree configuration on the microclimate of the building environment.

Under different building layouts, the tree configuration consistently shows a trend where a higher proportion of trees results in a lower Ta. For the same proportion of trees, the enclosed layout has the lowest average temperature. With a 25% tree coverage, the average temperature is 36.96 °C, while with 100% coverage, the average temperature is 36.51 °C. Additionally, in the row-type layout, the variation caused by differences in tree coverage is the most significant. Compared to 25% tree coverage, the average temperature with 100% tree coverage is reduced by 0.58 °C.

The change in Ws caused by different tree coverage proportions is minimal, generally showing that as the proportion of trees increases, the Ws decreases. Taking the enclosed layout as an example, for every 25% increase in tree coverage, the average Ws in the building environment decreases by 0.01.

In terms of Rh and MRT, the effects of varying tree proportions are more pronounced. Rh increases with a greater number of trees, while MRT decreases. In the row-type layout at 13:00, the average Rh with 100% tree coverage is 1.85% higher than with 25% coverage, and the MRT is 3.45 °C lower. This trend is consistent across all three layout forms. Therefore, it can be concluded that increasing tree coverage significantly reduces both the temperature and MRT of the built environment, while also enhancing the Rh in the surrounding area.

3.2. Changes in the PET Index

Figure 7 presents the PET simulation results from 9:00 to 17:00 at a height of 1.4 m above the ground across all simulation cases. The findings regarding the PET index are as follows: (1) The average daily duration of extreme heat stress (PET > 41 °C) under the baseline scenario (row-type layout without optimization) is 8 h (9:00–17:00), of which the peak value of PET reaches 54.3 °C from 13:00 to 14:00, which is far beyond the threshold of human body’s tolerance (the risk of heatstroke can be triggered by the PET > 41 °C lasting for 4 h). Combined with the population density characteristics of Fuyang’s old residential areas (volume ratio > 2.5), the cumulative duration of exposure to extreme heat stress exceeded the safety threshold due to the high frequency of outdoor activities and weak heat adaptation capacity of elderly residents in high-density living environments, which may lead to a 2–3-fold increase in the incidence rate of heat-related diseases. (2) The average PET values for each simulation scenario showed an increasing trend from 9:00 to 14:00 on the simulation day and a slow decrease from 15:00 to 17:00.

Using the 14:00 average PET value of the row-type layout as the benchmark, the comparison shows that the point-cluster layout with 100% tree coverage results in the greatest improvement in PET, reducing it from 54.27 °C to 52.17 °C, a decrease of 2.1 °C. On the other hand, the enclosed layout has the opposite effect on PET, increasing it from 54.27 °C to 54.32 °C, an increase of 0.05 °C. Compared to the row-type layout, the point-cluster layout reduces PET by 0.2 °C, indicating that among the three building layouts, the point-cluster layout provides the best improvement in PET.

Using the average PET value of the row-type layout as the benchmark, Figure 8 illustrates the daily accumulation difference of the average PET value across all simulated cases. Negative values indicate an improvement in the daily average PET value due to the simulated conditions, while positive values signify a deterioration in the PET environment. The results indicate the following: (1) Among the building layouts, the daily accumulated PET value is improved the most with the point-cluster layout, followed by the row-type and enclosed layouts; (2) Regarding building envelope materials, green infrastructure configuration, and surface materials, the most effective strategies for improving the daily average PET value involve using glass as the building envelope material based on the point-cluster layout, planting 100% tree coverage based on the row-type layout, and utilizing 100% asphalt coverage as the surface material in the point-cluster layout. These three scenarios resulted in reductions of 3.51 °C, 23.87 °C, and 2.65 °C in the daily average PET value, respectively; (3) A comparison of the improvements in daily accumulated PET values reveals that the influence on PET is ranked as green infrastructure configuration, building layout, building envelope materials, and surface materials. In addition to this, the improvement of PET values by 100% tree coverage (ΔPET = −23.87 °C) was mainly due to the synergistic effect of shading effect and transpiration cooling: shading reduced MRT by 3.45 °C by reducing direct sunlight, contributing 58% of the PET reduction; transpiration contributed the remaining 42% by reducing air temperature (ΔTa = −0.58 °C) and enhancing humidity (ΔRh = +1.85%). PET reduction increased significantly from 25% to 75% coverage (4.5 °C/25%), but the rate of increase plummeted to 1.4 °C/25% beyond 75% coverage, presumably due to canopy overlap that reduced the marginal benefit of shade and saturated moisture within the canopy that inhibited transpiration efficiency. The PET improvement ability of vegetation was significantly better than that of a single material (e.g., for asphalt, ΔPET = 2.65 °C), and this result verified the irreplaceable nature of vegetation shade and transpiration, while revealing the physical bottleneck of thermal comfort optimization in the high-coverage scenario.

4. Discussion

4.1. Investigation of the Intrinsic Association Between Ta, Rh, Ws, MRT, and PET

Figure 9, Figure 10, Figure 11 and Figure 12 present color-coded maps of the Ta, Ws, Rh and MRT indices at the hottest time of the day (13:00) at a height of 1.4 m above the ground. These maps provide a clearer understanding of the trends within the region.

The output plot illustrates the numerical changes in Ta, Ws, Rh, and MRT across all conditions. Figure 10 demonstrates that the increase in planted trees correlates with a decrease in the Ta around the building. Additionally, the temperature in a specific area within the enclosed layout significantly decreased in response to changes in the sun’s angle. However, no significant changes in Ta were observed in the simulation cases involving different surface materials and building envelope materials.

Transpiration from trees releases a substantial amount of water vapor, increasing humidity. Figure 10 and Figure 11 illustrate that increased tree coverage uniformly reduces Ws in the area, impedes nearby airflow, and raises Rh, with 100% coverage resulting in the most significant change in Rh. Furthermore, the use of concrete and bricks as building envelope materials also contributes to an increase in humidity.

In cases with 25% to 100% tree coverage, the increase in tree density resulted in a significant decrease in MRT in shaded areas. Regarding the surface materials, the reduction in MRT was observed in the following order: asphalt, brick, and concrete. Except for concrete, MRT decreased with the increase in asphalt and brick coverage, while no significant changes were noted in other cases. The increase in tree coverage led to decreased Ta, MRT and Ws, and heightened Rh in the region. Higher coverage of concrete, asphalt, and masonry pavement resulted in lower Ta and MRT, while Rh increased in these areas. The use of glass as building envelope materials raised Ta and decreased Rh, in contrast to the effects of concrete and brick, which had the opposite impact. Among these factors, changes in tree coverage had the most significant influence on Ta, Ws, Rh, and MRT.

By comparing Figure 12 and Figure 13, it was found that the MRT and PET index plots produce similarities, which suggests that MRT is significantly correlated with PET, and that a reduction in MRT would play a crucial role in improving PET. The greatest reduction in PET was observed in the shade areas with 100% tree cover, and in addition, an increase in the asphalt and masonry pavement cover led to a sporadic decrease in the PET index.

The results indicate that enhancing green infrastructure configuration has the most significant effect on improving the PET value, followed by an increase in the coverage of surface materials, which also positively influences PET. Additionally, the variations in Ta and Rh resulting from different building envelope materials significantly impact the improvement of the PET index.

4.2. Mechanism of Multifactor Coupling Effect on Thermal Comfort and Optimization Suggestions

Figure 14 shows the impact and trends of different types and coverage rates of building layouts, building envelope materials, green infrastructure, and surface materials on the daily accumulated PET values. The results indicate that: (1) The influence of building envelope materials on daily accumulated PET values in enclosed and row-type layouts is ranked as follows: glass, concrete, and brick. The smallest impact occurs with glass material in the enclosed layout, which negatively affects the daily accumulated PET value, increasing it by 0.09 °C. The largest effect occurs with brick material in the row-type layout, which reduces PET by 2.02 °C. (2) As for green infrastructure configuration, a higher tree coverage rate corresponds to a stronger improvement in PET values. For example, in the point-cluster layout, the difference in the improvement of daily accumulated PET values between 25% and 100% tree coverage is 13.55 °C. (3) In terms of surface materials, as coverage increases, both asphalt and brick materials lead to improvements in daily accumulated PET values, while concrete material results in worsening PET conditions. (4) The impact of building layouts on daily accumulated PET improvement is as follows: point-cluster > row-type > enclosed.

This study reveals the critical path for optimizing the thermal environment of old residential areas by quantitatively analyzing the synergistic effects of building layouts, building envelope materials, surface materials and green infrastructure configurations on PET values. Compared with single-factor studies, the findings under the multifactor coupling perspective are more practically instructive, as discussed below:

(1)

In highly enclosed and row-type layouts, bricks improve PET better than glass (2.02 °C reduction), while in open point-cluster layout, glass materials instead show optimal thermal comfort (3.51 °C reduction). This contradiction stems from the interaction between building closure and material albedo: in a highly enclosed layout, mutual shading between buildings reduces the direct solar area, when low albedo materials (e.g., bricks) can alleviate the daytime heat buildup by absorbing the radiation and releasing the heat slowly, whereas high-albedo glass may exacerbate the intensity of the localized radiation due to multiple reflections in a dense layout [15]. In a low-enclosed point-cluster layout with a high percentage of open space, the high albedo property of glass can effectively reflect solar radiation to inactive areas, while its low heat capacity property avoids heat stagnation, which is consistent with the findings of Jiang et al. [15] in a hot and humid area.

(2)

The improvement of PET values by tree coverage (maximum reduction of 23.87 °C) was significantly higher than other factors, confirming the dual advantages of evaporative cooling and shading by vegetation. It is noteworthy that the rate of PET reduction levelled off when the green coverage was ≥75% (Figure 14b), suggesting the existence of a “threshold effect”. This phenomenon may be related to canopy closure: at low coverage (<75%), discrete canopies form localized shading zones, while continuous canopies at high coverage have limited shading efficiency, but water vapor diffusion by transpiration may be reduced due to impeded ventilation [17]. This suggests the need to balance green coverage and ventilation corridor design in practice.

(3)

Although asphalt pavements show the best PET improvement in summer (2.65 °C reduction), their high heat absorption properties may lead to reduced thermal comfort in winter. This is in line with Lai et al.’s [21] study on temperate cities, which concluded that high albedo materials (e.g., concrete) have a “winter-summer paradox” in climate regulation throughout the year. It is recommended that a dynamic strategy be used in the renovation of old residential areas: use asphalt in the main activity areas (e.g., walkways), while retaining part of the concrete in secondary areas (e.g., car parks) in order to balance winter and summer needs.

(4)

Extreme heat stress (PET > 41 °C) is a particularly significant health threat to elderly residents and outdoor workers. The optimization strategy of using point-cluster layout and 100% tree coverage can compress the duration of extreme heat stress from 8 h to 4.5 h, which provides a basis for the development of targeted policies: priority is given to the implementation of whole-area shade renovation (tree coverage ≥ 75%) in activity areas (e.g., community squares, fitness paths), combined with high emissivity paving to reduce the intensity of heat exposure; for economically disadvantaged communities, low-cost measures (e.g., rooftop sprinklers, mobile awnings) need to be adopted to compensate for the lack of greenery, avoiding worsening the coupling of the thermal health crisis with social equity issues.

4.3. Analysis of Multiple Linear Regression Models

In order to investigate whether the physiological equivalent temperature PET values produce different results depending on the variations in various indicator variable factors and the extent of the various types of influence, the data of the four influences were analyzed by multiple linear regression using SPSS 27.0.1 software. The specific data are shown in

Table 7. Comparison of each influencing factor with the corresponding PET value data.

Green Coverage	Pavement Reflectance	Pavement Coverage	Enclosure Material Reflectivity	PET
0	0.3	100	0.3	54
0	0.3	100	0.05	53.85
0	0.3	100	0.4	53.83
25	0.3	100	0.3	53.44
50	0.3	100	0.3	53.1
75	0.3	100	0.3	52.46
100	0.3	100	0.3	52.1
0	0.3	25	0.3	54
0	0.3	50	0.3	53.98
0	0.3	75	0.3	53.94
0	0.12	25	0.3	53.95
0	0.12	50	0.3	53.88
0	0.12	75	0.3	53.77
0	0.12	100	0.3	53.67
0	0.2	25	0.3	53.95
0	0.2	50	0.3	53.88
0	0.2	75	0.3	53.77
0	0.2	100	0.3	53.68

.

A multiple linear regression model was developed to describe the linear relationship between the dependent variable, physiological equivalent temperature PET, and multiple influencing factor independent variables, as shown in Equation (1):

$$y=b_0+b_1{x}_1+b_2{x}_2+\cdots +b_m{x}_m+\epsilon$$

(1)

where y is the dependent variable; $b_0$ is the constant term; $b_1\cdots b_m$ is the regression coefficient; $x_1\cdots x_m$ is each of the influencing factors, respectively; and $\epsilon$ is the random error.

The coefficients of the regression equation are shown in

Table 8. Regression coefficients and significance tests.

Ratio a
	Unstandardised Coefficient		Standardised Coefficient			Covariance Statistics
Modelling	B	Standard Error	Beta	t	Significance	Tolerances	VIF
(Constant)	53.838	0.112		481.587	0.000
Green coverage	−0.018	0.001	−0.977	−26.594	0.000	0.741	1.350
Pavement reflectance	0.934	0.252	0.132	3.704	0.003	0.784	1.276
Pavement coverage	−0.003	0.001	−0.139	−3.872	0.002	0.771	1.297
Enclosure material reflectivity	0.002	0.270	0.000	0.009	0.993	0.959	1.043

^a implecit variable: PET.

, which shows that the tolerances of the independent variables are all greater than 0.1, and the expansion factors VIF are all less than 5, which proves that there is no covariance between green coverage, paving reflectance, paving coverage, reflectance of the enclosure material and the physiological equivalent temperature PET. Green coverage had the most significant inhibitory effect on PET (standardized coefficient Beta = −0.977, p < 0.001), with PET decreasing by 0.018 units for every 1% increase in coverage, which is consistent with the cooling mechanism of vegetative transpiration shading. Notably, pavement reflectance (Beta = 0.132, p = 0.003) showed a positive correlation, with PET increasing by 0.934 units for every 0.1 increase in reflectance, and it is speculated that highly reflective materials may indirectly enhance the human radiative load by enhancing the scattering of shortwave radiation, which is more significant especially in the unshaded area. Pavement coverage, on the other hand, showed a weak negative effect (Beta = −0.139, p = 0.002), the effect of which may be related to the synergistic effect of water evaporation or shading by permeable paving. Fencing material reflectance did not pass the significance test (Beta = 0.000, p = 0.993), suggesting that the effect of vertical interface reflectance properties on open space PET is negligible. The model fit is shown in

Table 9. Model fit.

Modelling Summary b
Modelling	R	R2	Adjusted R2	Errors in Standard Estimates
1	0.993 a	0.987	0.983	0.07055

^a predictor variable: (Constant), Enclosure material reflectivity, Green coverage, Pavement reflectance, Pavement coverage. ^b implicit variable: PET.

, which was able to obtain an R-value of 0.993 for the fit of this computational model and a value of 0.983 for the adjusted R², indicating that the model fit is good.

Multiple linear regression analysis was used to normalize the data, and in the case of the reflectance of the enclosure material, this variable needs to be removed from the model; therefore, the physiological equivalent temperature PET was taken as the dependent variable, and the green coverage, paving reflectance, and paving coverage were taken as the independent variables, and according to the regression coefficients and constant terms, the coefficients were substituted into Equation (1), and the multiple linear regression model can be established as shown in Equation (2):

$$y=53.838-0.18x_1+0.934x_2-0.03x_3$$

(2)

where $y$ is the physiological equivalent temperature PET; $x_1$ is the green cover; $x_2$ is the paved reflectance; $x_3$ is the paved cover.

4.4. Strategic Feasibility Analysis

Currently, the core contradiction facing urban thermal environment optimization has shifted from pure technical feasibility to sustainability under the constraint of limited financial support. Existing studies have mostly focused on thermodynamic simulation and design strategies, but neglected a key reality, i.e., the renovation projects for old residential areas in some cities have been stalled in the middle of the project due to cost overruns or lagging benefits. Therefore, this paper combines economic and technical aspects to achieve a leapfrog path from “technically optimal” to “economically feasible” decision-making for thermal retrofit, and thus proposes a thermal comfort optimization strategy that is in line with the economic reality of Fuyang City.

This study analyzes on basis of the ‘technically optimal’ strategy, coupled with the market price of each influencing factor in Fuyang. Among them, the average market price of asphalt material in Fuyang City is 96.5 yuan/m², the market price of trees (Koelreuteria paniculata Laxm) is 71 yuan/plant, and the market price of glass is 320 yuan/m² [40,41]. The actual cost of each strategy is shown in

Table 10. Actual costs by impact factor under ‘technological optimization’.

Strategy	ΔPET (°C)	Cost (￥)	ΔPET/Cost (°C/￥)
Optimization of point-cluster layout	−1.85	0	∞
Asphalt overlay (25%)	−0.23	222,336	0.00000103
Asphalt overlay (50%)	−0.7	444,672	0.00000157
Asphalt overlay (75%)	−0.89	565,104	0.00000157
Asphalt overlay (100%)	−1.28	764,280	0.00000167
Tree overlay (25%)	−5.96	1704	0.00349765
Tree overlay (50%)	−14.26	3408	0.00418427
Tree overlay (75%)	−20.89	5112	0.00408646
Tree overlay (100%)	−23.87	6816	0.00350205
Glass curtain wall retrofit	−1.41	5,345,280	0.00000026

.

From the table, it can be seen that the economic cost-effectiveness of using asphalt as the pavement material increases with the increase in the coverage rate, while the economic cost-effectiveness of planting trees is highest when the tree coverage rate reaches 50%. In addition, the overall economic cost-effectiveness of the three strategies in the “technologically optimal” strategy follows the order of green infrastructure configuration, surface materials, and building envelope materials. Therefore, when considering the economy, 50% tree coverage is preferred as the thermal comfort improvement strategy, followed by 100% asphalt coverage as the surface material and glass as the building envelope material if the economy permits.

4.5. Limitations and Future Prospects

The effectiveness of the strategies in this study is limited by the following boundary conditions:

(1)

Climatic conditions: The simulations are based on extreme summer heat (41.2 °C) and high humidity (Rh = 72%) scenarios, and the conclusions are prioritized for the regions with hot summer and cold winter in China. For dry and hot climatic zones, the solar radiation transmission of glazing materials may lead to a deterioration of the thermal environment, and material prioritization needs to be reassessed.

(2)

Urban form: the model with a building height of 24 m and a plot ratio of 2.8 represents a typical high-density old area. In low-rise residential areas (≤6 stories), the ventilation benefits of the point-cluster layout may be reduced.

(3)

Renovation phase: the strategy targets the stock renewal scenarios in the “post-urbanization phase” (i.e., large-scale new construction stops and shifts to optimization of existing communities). If residential areas are still in a period of rapid expansion, priority needs to be given to addressing infrastructure deficiencies.

In addition, ENVI-met has three limitations in modelling the microclimate of high-density old residential areas. First, since ENVI-met requires high precision spatial and temporal data (e.g., hour-by-hour traffic flow, air-conditioning start/stop ratios) to simulate the dynamic heat sources, and the lack of such a monitoring infrastructure in the old residential areas may lead to amplification of the error when homogeneous parameters are forced to be set. Thus, this study does not take into account the interference of the dynamic anthropogenic heat sources, such as traffic and air-conditioning, on the microclimate, and may underestimate the PET values of local hotspots (e.g., windward side of street-facing buildings). The follow-up work will combine mobile monitoring and energy bill data to construct a spatial-temporal heterogeneous anthropogenic heat emission model to improve the completeness of the simulation of the thermal environment in high-density communities. Second, the case study focuses on the typical old residential areas in Fuyang, whose warm-temperate semi-humid climate and high-density built-up environment may limit the generalizability of the conclusions to the dry and cold regions. Third, due to the limitation of computational efficiency, the software has convergence constraints for long-time continuous simulation of complex three-dimensional scenarios, which needs to be replaced by segmented periodical simulation for year-round continuous computation. To address the above issues, the background heat fluxes are calibrated using historical data from Fuyang meteorological stations, and the stability of the strategy is verified by extrapolating the trend of the simulation results over multiple time periods, so as to ensure that the main conclusions are reliable within a reasonable margin of error.

Future research can be deepened in the following directions: (1) expanding the comparison of cases in multiclimatic zones and establishing the design guidelines for renewal of old residential areas based on climatic classification; (2) integrating remote sensing inversion and field monitoring data to construct a high-precision database of vegetation-surface parameters to improve the reliability of the model; (3) introducing a radiation balance model and coupled analysis of energy sources to reveal the physical mechanism of synergy effects between the architectural layout and the albedo of the materials; and developing a dynamic simulation toolkit to assess the long-term resilience of microclimate optimization strategies in the context of climate change. These explorations will help to form a scientific and regionalized thermal environment control technology system for old cities, and provide theoretical support for sustainable renewal in the global post-urbanization stage.

5. Conclusions

This paper investigates common building layouts, building envelope materials, surface materials, and green infrastructure configurations used in old residential areas. Using the ENVI-met software, a three-dimensional physical model of the architectural space in these areas was established. This study considers three building layouts: row-type, point-cluster, and enclosed layouts, as well as three types of building envelope materials: glass, concrete, and brick. It also examines three types of surface materials (concrete, asphalt, and brick) with varying coverage rates, and explores the impact of different coverage scenarios of common green infrastructure configurations on the regulation of the microclimate in building spaces. Additionally, compared to concrete and brick, point-cluster layouts with glass materials have been found to reduce outdoor thermal comfort the most. This pattern contrasts with the trends observed in enclosed and row-type layouts. It is preliminarily concluded that there is a coupling relationship between building envelope materials and building layouts in their influence on thermal comfort. The degree of building enclosure and the albedo of the materials have opposing effects on thermal comfort. Based on the current experiments, the following conclusions can be drawn:

(1)

The summer temperatures and average radiation temperatures in Fuyang exhibited an initial increase followed by a decrease from 9:00 to 17:00. Solar radiation intensity peaked at 13:00, while humidity demonstrated a trend of decreasing initially and then rising. The level of thermal comfort was at its lowest at 13:00, which could cause significant discomfort to individuals.

(2)

In evaluating the specific PET, the factors influencing the outdoor thermal environment of old residential areas in summer, in order of significance, are: green infrastructure configuration, building layout, building envelope materials, and surface materials.

(3)

Simulation analysis of four influencing factors on the architectural space of old residential areas shows that the outdoor thermal comfort of buildings improves as tree coverage increases, with the highest improvement in the daily accumulated PET value being a difference of 13.55 °C. When surface materials are asphalt or brick, the higher the tree coverage rate, the greater the improvement in PET. In contrast, concrete materials result in the opposite effect. The degree of improvement in outdoor thermal comfort based on building layout follows the order: point-cluster layout, row-type layout, and enclosed layout.

(4)

When the degree of enclosure in residential building layouts is high, as seen in row-type and enclosed layouts, the effectiveness of building envelope materials for improving thermal comfort follows the order of brick, concrete, and glass. Conversely, when the degree of enclosure is low, such as in point-cluster layout, the order of effectiveness shifts to glass, brick, and concrete for enhancing thermal comfort.

(5)

Taking all factors into account, using a point-cluster layout for the residential area, incorporating glass as the building envelope material, utilizing asphalt pavement with 100% coverage as the surface material, and planting a 100% coverage of trees can significantly enhance the thermal environment of the building.

It should be noted that when the settlement plot ratio is less than 1.5 or the urbanization rate is more than 70%, the wind protection benefits of the enclosed layout may be better than the ventilation priority strategy of this study, and dynamic adjustment in conjunction with local planning objectives is recommended.

In summary, different building layouts, building envelope materials, surface materials, and green infrastructure can all have varying degrees of impact on the microclimate and thermal comfort of buildings. Based on the results, it is suggested that in the future post-urbanization stage, the point-cluster layout should be adopted as the building layout of old residential areas, with glass as the building envelope material, asphalt as the surface material, and as many trees as possible planted as part of the green infrastructure. This approach will maximize the improvement of the microclimate and thermal comfort for the old residential areas in Fuyang, and is expected to provide support for the renovation of old residential areas in the future post-urbanization phase.