Mitigation of Urban Heat Risks via Shading Stations: An Experimental Comparison of Fabric and Photovoltaic Shading Authors: Liyue Jin, Xin Dong, Deo Prasad, Bao‐Jie He Published: 2026 ABSTRACT ABSTRACTExploring mitigation solutions is important for supporting heat‐resilient urban designs. Shading stations offer a practical means of blocking direct solar exposure and improving the thermal environment of the station. This study aimed to compare the heat mitigation performance of fabric and photovoltaic (PV) canopy shading in terms of microclimate regulation and outdoor thermal comfort improvement based on field experiments in subtropical Jiangsu, China. The results indicate that artificial shading could significantly reduce the universal thermal climate index (UTCI), with average daytime reductions of 3.5°C for fabric shading and 3.1°C for PV canopy shading. Although both types of shading stations were effective, they exhibited distinct mechanisms of action. Fabric shading provided more consistent cooling and radiation reduction at the ground level (0–0.5 m). During the full‐shading period (10:30–13:30), PV shading showed stronger downward longwave radiation, resulting in a weaker mitigation performance at pedestrian height (1.0–1.5 m) than that of fabric shading. Furthermore, the shading‐induced mitigation performance can be affected by underlying surfaces. Grass surfaces enhanced cooling via evapotranspiration, whereas high‐albedo hard surfaces sometimes contributed to heat accumulation. The differences in vertical temperatures verified the different mitigation mechanisms of the two types of shading stations along their height. Overall, this study provides a reference for designing artificial shading.Keywords: fabric shading, microclimatic variables, outdoor thermal comfort, photovoltaic shading, thermal quality, weather condition Introduction Over several decades, global climate change has aggravated, leading to more frequent, severe, and intense extreme weather events, including extreme heat events, with significant environmental, social, and economic consequences. Associated with urbanization, the urban heat island (UHI) effect has become a critical heat‐related challenge that intensifies thermal discomfort, reduces work efficiency, and increases cooling energy demand [1]. As such, individuals conducting sporting and leisure activities outdoors often face stronger heat stress than ever before. Prolonged exposure to intense heat significantly increases the risk of heatstroke and other heat‐related illnesses, posing a substantial threat to public health [2]. Improving the thermal quality of public spaces, which is the basis of urban livability, is important to alleviate heat stress to secure public health and safety, sustain leisure and sporting activities, and ensure economic production. The development of effective heat mitigation and adaptation strategies is imperative to protect public health and well‐being, and ensure urban sustainability and resilience [3]. Outdoor thermal comfort (OTC) is an indicator of urban thermal quality (UTQ). The universal thermal climate index (UTCI) and physiological equivalent temperature (PET), which are primarily influenced by solar radiation, air temperature, wind speed, and humidity, are comprehensively linked to public health and well‐being. Therefore, mitigation strategies must focus on the regulation of OTC and associated variables to improve thermal quality. Various cooling approaches have been developed and examined, primarily in terms of urban morphology, blue‐green infrastructure (e.g., vegetation and water bodies), and innovative surface materials [4]. These interventions work in one or several ways to reduce solar exposure, enhance ventilation, lower radiation absorption, and cool ambient air [5]. Shading facilities are effective in reducing solar heat and regulating microclimatic variables [4]. Building shading stations in outdoor exposed spaces (e.g., streets, squares, parks, and rooftops) can reduce direct solar exposure and improve thermal environments. Prior studies have demonstrated the capacity of urban shading to alleviate heat stress [6], with most studies exploring the shading cast by trees and built forms (e.g., street canyons) in urban areas. Moreover, artificial shading components are frequently viewed as flexible alternatives or complements to trees and buildings; however, few empirical studies have directly compared the mitigation performance of different types of artificial shading facilities [7]. Typically, on hot and sunny days, shading devices can effectively block incoming solar radiation and reduce air temperatures beneath the shade, thus enhancing the OTC both physiologically and psychologically [8]. A case study in Port Said, Egypt showed that black fabric shading could lower predicted mean vote values by 22% and reduce PET by more than 4°C compared to unshaded spaces [9]. Decarbonization remains a fundamental strategy for addressing climate change [10]. In contrast to the excessive reliance on carbon‐intensive fuels, which has intensified the global energy crisis and accelerated greenhouse gas emissions, carbon‐free renewable energy has been promoted and implemented in cities [11]. Solar energy offers enormous potential for decarbonization, exhibiting various scenarios and wide application feasibility in on‐site and off‐site harvesting [12]. Urban buildings and their structural elements can serve as excellent platforms for solar energy utilization. Accordingly, a novel practice has been developed to integrate solar technologies with shading facilities, that is, photovoltaic (PV) canopy shading, to achieve both decarbonization and heat mitigation in public spaces [13]. It is regarded as a promising co‐benefit pathway for increasing decarbonization capacity while improving outdoor thermal environments. Some studies have explored PV‐induced impacts on microclimate and OTC, while most of them have modeled PV canopies on building roofs, facades, and walls. In comparison, limited studies have assumed the PV canopy as a street canopy and walking route, and there is a need to investigate the impacts of PV canopy shading on urban thermal quality. A good understanding of the heat mitigation performance of PV canopy shading will be conducive to the design of heat‐resilient walkable routes. For PV canopy shading and fabric shading, the former can intercept incoming solar radiation and simultaneously generate electricity, whereas fabric shading can intercept solar radiation and ensure superior heat dissipation. It is essential to further clarify microclimate and OTC regulation performance to support decisions regarding design schemes [14]. Understanding how shading facilities interact with underlying surfaces (e.g., pervious surfaces, conventional hard surfaces, and reflective pavements) to intervene in the microclimate is also important for optimizing shading configurations. Empirical studies are necessary to reveal the thermal regulation effects of artificial shading. Although some studies have investigated the cooling benefits of artificial shading facilities, they lack effective comparisons of the impacts of fabric and PV canopy shading on urban thermal quality based on empirical measurements. To address this gap, this study conducted a field experiment in Liyang City, Jiangsu Province, China, to quantitatively assess the urban thermal quality improvement by two common artificial shading types—PV canopy and black fabric—over two types of ground surface cover (e.g., grass and high‐albedo pavement) and under different weather conditions (e.g., cloudy hot days and sunny hot days) in hot seasons. Key climatic and OTC indicators, including Ta , T mrt, and UTCI, were analyzed to explore the response to fabric and PV canopy shading. The findings provide empirical evidence and references for optimizing outdoor thermal environments by adopting shading stations. Methodology Study Area The field experiment was conducted in an academic precinct in Liyang City, Jiangsu Province, China. According to the Köppen climate classification [15], the city has a humid subtropical monsoon climate (Cwa), characterized by hot and rainy summers, cold and dry winters, and significant annual temperature variations. Based on climate data from 1991 to 2020, the average air temperature in August is 28.2°C, with an extreme maximum daily temperature reaching up to 43.0°C. The average annual precipitation is 1198.8 mm, and the total annual sunshine duration ranges from 1400 to 2200 h. The period from June to September generally has high air temperatures and solar radiation inputs, with July and August being particularly hot (Figure 1). These months are associated with the most intense thermal stress and a strong need for mitigation and adaptation measures. Considering these climatic characteristics, field measurements were scheduled for August. The measurement period spanned three days, including one cloudy day (August 22nd) and two consecutive sunny days (August 23rd–24th) (Table 1). Monthly average, maximum, and minimum air temperature, sunshine duration, and precipitation in Liyang City, Jiangsu Province, China, from 1991 to 2020. TABLE 1Combination of two shaded groups and ground surface cover for experimental investigation.Shading facilitiesCloud conditionHeat conditionDateGround surface coverPV & FabricCloudy dayHot day2024.08.22: Scenario 1Grass land surfaceCloudless dayExtreme heat days2024.08.23: Scenario 2Grass land surface2024.08.24: Scenario 3High‐albedo surface Combination of two shaded groups and ground surface cover for experimental investigation. Field Measurement Measurement location Based on application contexts, artificial shading facilities are generally categorized into three types: walkway shading, plaza (courtyard) shading, and park shading. To better mimic real‐world conditions, we designed experimental scenarios to simulate the deployment of shading facilities that are commonly encountered during daily pedestrian movements in cities. We selected a microclimatic and thermal comfort field study site on the south side of an east‐west‐oriented street, with continuous three‐story buildings on the northern side. The site receives direct sunlight for most of the day, except during early morning and late afternoon. Measurement Scenarios: Shading Types, Underground Surface and Weather Conditions The two shading structures were designed with identical geometric dimensions of 1200 mm × 2400 mm × 2500 mm (length × width × height) (Figure 2). The top surface and upper half of the south‐facing side of each structure were covered with a canopy to provide shading protection for pedestrians’ heads and their upper bodies. The PV shading station consisted of eight monocrystalline silicon canopies with dimensions of 1200 mm × 550 mm × 3 mm, offering a rigid, opaque surface with high solar blocking efficiency and structural stability. The fabric shading facility was built using a black high‐density polyethylene (HDPE) mesh fabric, which is commonly used in cities for shading applications in China. The structure had a thickness of 0.3 mm and deployed size of 2400 mm × 2400 mm. The fabric offers partial solar blockage, good breathability, and diffuse‐light transmission. An unshaded control group was used for comparison purposes. Overall, these experimental settings enabled the analysis of how the shading design affected microclimatic and OTC performance. Field measurement duration, weather condition, and ground surface cover. Furthermore, two shaded groups were integrated with two types of ground surface cover to investigate the heat mitigation performance of the shading facilities (Table 1). To represent common urban environments, two types of ground surface cover were used: high‐albedo pavement (with an albedo of approximately 0.6–0.7) and a grass surface (albedo approximately 0.2–0.3). High‐albedo pavements, a widely used reflective hardscape material, are effective in reducing solar energy absorption through reflection; however, they may exacerbate human heat exposure owing to increased secondary shortwave and longwave radiation. The grass surface, characterized by lower reflectivity and higher evapotranspiration capacity, can mitigate near‐ground heat accumulation under extreme heat conditions, thereby improving the local microclimate. Three field measurement scenarios were designed according to the weather conditions and ground surface cover (Figure 2). In each scenario, three groups were tested: PV canopy shading, black fabric shading, and direct solar exposure (DSE) control, representing two shaded groups and an unshaded control group. The two shaded groups were arranged with a 10‐m separation in the east‐west direction to avoid mutual shading during measurements. All three sites were equipped with the same sensors and layouts to ensure the consistency and comparability of the data. Measurement Period To capture the temporal dynamics of the thermal environment, all measurements were segmented into four periods: A (midday/afternoon, 09:00–15:30), B (late afternoon, 15:30–18:35), C (night, 18:35–05:30) and D (early morning, 05:30–09:00). This classification was based on solar angles, diurnal heat load patterns, and conventional divisions in urban‐climate studies. Microclimatic Variables and Sensors To comprehensively investigate the mitigation performance of shading facilities, this study measured a range of key microclimatic variables that can influence outdoor heat stress and thermal comfort of the occupants. These include air temperature (Ta ), black globe temperature (Tg ), surface temperature of the shading canopy (Ts ), ground surface temperature (ST), and background microclimatic variables. Dedicated instruments and sensors were deployed to capture horizontal and vertical environmental data under different shading scenarios. The following subsections describe the setup, measurement locations, and methodological concerns for each variable (Figure 3). Experimental setup: (a) Schematic diagram of a fabric shading facilities with measuring devices and measurement points; visible images (b,c) and thermal image (d) of the underside of fabric shading facilities on high‐ albedo land surface. The colors and the values on the thermal image (d) indicate the surface temperature. Background Weather Measurement At each measurement site, five key microclimatic variables were recorded: air temperature (Ta , °C), relative humidity (RH, %), wind speed (Va , m/s), longwave and shortwave solar radiation (W/m2), and black globe temperature (Tg , °C) (Table 2). The wind speed was measured at a height of 2 m above the ground using a Testo 425 anemometer. Air temperature, black globe temperature, relative humidity, and pyranometer were measured using a TRM‐ZS2 meteorological station, with sensors positioned at a height of 1.10 m above the ground, corresponding to the average center of gravity of an adult. All microclimatic variables were logged at 1‐min intervals. Surface temperatures were captured using a Testo‐865 thermal imager, with thermal images taken approximately every 30 min throughout the experiment. TABLE 2Details of meteorological sensors.ParameterInstrumentSensorRangeAccuracy Ta Jinzhou Sunshine Weather station TRM‐ZS2‐BXPTS‐3 air temperature sensor−40 to 120°C±0.2°C RH PTS‐3 humidity sensor0–100%±2%Short and long wave radiation TBB‐2 net all‐radiation meter/ Pyranometers −2000 to 2000W/m2 ±1W/m2 Tg PTWD‐4 black globe thermometer−50 to 150°C±0.2°C Va (2m)Testo 425Thermal wire anemometer0.01–30m/s 0.03m/s+4.0% (≤20m/s) 0.5m/s+5.0% (>20m/s) Ts /ST Testo 175T‐3NTC temperature sensor−50 to 400°C(T‐type)±0.5°C (−50 to 70°C) Ta (vertical)Testo 174HNTC temperature sensor−20 to 70°C±0.5°C (−20 to 70°C) Details of meteorological sensors. TBB‐2 net all‐radiation meter/ Pyranometers 0.03m/s+4.0% (≤20m/s) 0.5m/s+5.0% (>20m/s) Tg Measurement To capture the net radiative heat load potentially experienced by the human body in shaded spaces, Tg , which accounts for the combined effects of shortwave solar radiation, longwave radiation, and air movement on the heat gain of a globe‐shaped sensor [16], was measured. Based on Tg , the mean radiant temperature (T mrt) and UTCI can be calculated. In addition to the standard measurement at 1.10 m height via a measurement station, an additional black globe thermometer was installed 25 cm below the top canopy at the center of each shading canopy. This position was selected to simulate the typical distance between a person's head and the shading surface, ensuring that the globe remained consistently shaded and well ventilated throughout the experiment. Tg was recorded at 1‐min intervals. Vertical Ta Measurement To understand the variations in vertical air temperature resulting from upward and downward energy fluxes, six Testo 174H temperature and humidity sensors were installed under each shading station in the two shaded groups. These six sensors were mounted at heights of 0.50, 1.00, 1.50, 1.75, 2.00, and 2.25 m from the ground and positioned at the center of each shading station to measure shaded air temperatures. A reference sensor was installed at the unshaded control site (without shading) at a height of 2 m to minimize ground interference and capture temperature and humidity under direct solar radiation. Each logger recorded Ta and RH data at 1‐min intervals. Ts and ST Measurement To investigate the impact of thermal inertia and surface heating of the shading canopy, it is essential to simultaneously monitor Ts and ST., as these surface thermal behaviors not only determine the re‐radiation capacity of the shading components but also affect the air temperature and heat stress levels within the shaded spaces. These measurements provide critical evidence for analyzing the underlying mechanisms of thermal regulation. For each shading station, temperature and humidity sensors (Testo‐175T3) connected to T‐type thermocouples were used. Thermocouples were affixed to the center of each top and side canopy to measure Ts , which was recorded every 1 min. In addition, T‐type thermocouples were used to monitor ST., which was measured near the central underside of the shading canopy to ensure that the measured spaces remained shaded for as long as possible. In contrast, ST under sun‐exposed conditions was recorded at a control site located 10 m away from the shaded groups, ensuring that this reference site remained unshaded throughout the measurement period (Figure 3a). T mrt, ∆ST and UTCI Calculation The radiative heat load, quantified by T mrt, is a primary contributor to the OTC during summer daytime periods [17, 18]. A study in Nanjing, China, showed a strong correlation between UTCI and T mrt under summer daytime conditions, with a UTCI threshold corresponding to T mrt values exceeding 55°C for outdoor heat stress. T mrt assessment was conducted based on Equation (1) according to ISO 7726 [19, 20]. where ε represents the emissivity of the globe (set to 0.95) and D denotes the globe diameter (set to 50 mm). The surface cooling temperature (∆ST) was calculated as the temperature difference between the sunlit and shaded surfaces. Moreover, the UTCI, which integrates Ta , RH, Va , and T mrt was calculated. Among these four input variables, Ta and RH were directly measured and input into the software programs, whereas T mrt and Va were converted based on Equations (1) and (2). Measurement stations for wind speed were set at 2 m, while Equation (2) converted it to the wind speed at 10 m [21]. Moreover, to compare the thermal quality at each site, UTCI was adopted as the primary thermal stress index [20], where the final UTCI values were interpreted using the 10‐level thermal stress classification scale [21]. where vx represents the wind speed at a specific measurement height x, m is the actual height at which the wind speed is measured (unit: m), and z0m is the surface roughness length governing momentum transfer (set to 0.01 m). Data Analytical Methods To examine the relationship between UTCI and microclimatic variables, including T mrt, Ta , Va , and Tg , linear regression analysis was applied [7]. Significance was evaluated using p values. Statistical significance is indicated with asterisks as follows: one asterisk for p‐values between 0.01 and 0.05, two asterisks for p‐values between 0.001 and 0.01, and three asterisks for p‐values below 0.001. Cases with no significant correlation were marked as “ns” [6]. Under the same environmental conditions, this study enabled direct comparisons to assess the effectiveness of PV canopy and fabric shading in improving OTC. In the analytical framework, apart from the variation in thermal parameters throughout the day and four periods (e.g., A, B, C, and D), particular attention was given to the time period of 10:30–13:30, corresponding to the peak hours of shading effectiveness. During this period, the central monitoring point beneath each shading station remained fully shaded, with no direct solar penetration. Therefore, this time window was considered optimal for evaluating the radiative shielding performance of the shading facilities and their influence on UTCI reduction under stable solar load conditions. Such targeted analysis allows the isolation of material‐driven effects by minimizing geometric variations in shadow projection. For broader comparative analyses, the dataset was grouped into daytime (05:30–18:35) and nighttime (18:35–05:30) periods based on the sunrise and sunset times during the measurement period. This dual‐layer temporal framework supports both dynamic and cumulative evaluations of the heat mitigation performance across different shading surface configurations. Analysis of Heat Mitigation Performance of Artificial Shading Stations This section analyzes the heat mitigation performance and associated mechanisms of the two types of artificial shading. Compared to the unshaded control group, the analysis revealed how the shaded stations caused differences in heat mitigation across various time periods, vertical height levels, and ground surface types. Daily Variation of Microclimatic Variables Figure 4 shows the daily variation of key microclimatic variables in the two shaded groups and one unshaded control group under the three spatial scenarios (scenario 1, 2, and 3). These variables were analyzed in the following four periods: A (midday/afternoon, 09:00–15:30), B (late afternoon, 15:30–18:35), C (night, 18:35–05:30), and D (early morning, 05:30–09:00). Daily variations of (a) (e) Air temperature, (b) (f) Relative humidity, (c) (g) Wind speed, and (d) (h) Black globe temperature under different spatial scenarios. Period A‐ Midday/Afternoon During Period A, the differences among the shaded and unshaded groups were most evident. Ta in the unshaded control group was generally higher than that in the two shaded groups, especially during the extreme heat stage, with the maximum Ta reduction reaching 4.8°C under fabric shading. RH showed an opposite pattern to Ta . The unshaded group generally had the lowest RH, whereas fabric shading and PV canopy shading increased RH by up to approximately 12% and 8%, respectively. Va fluctuated during this period, and the differences among the three groups were less consistent than those of Ta and RH. The shaded groups tended to show slightly lower Va in some scenarios, indicating a weak wind‐sheltering effect. Tg responded most strongly to shading. In the unshaded group, Tg exceeded 50°C during extreme heat conditions, whereas the maximum Tg was approximately 48°C under PV canopy shading and no more than 47°C under fabric shading. Overall, Period A was the main stage in which shading produced clear heat mitigation, primarily by reducing radiant heat exposure and lowering Ta . Period B‐ Late Afternoon During Period B, solar radiation weakened, and the thermal difference among the groups gradually decreased. Ta in all groups tended to decline or stabilize, and the shaded–unshaded difference became smaller than that in Period A. RH increased continuously with a decrease in Ta . The RH curves of the three groups were generally close, with an overall increase of approximately 30% and a maximum inter‐group difference of only approximately 2%. Va decreased in all groups and approached a low‐speed level of approximately 0.2 m/s. The unshaded group generally maintained a relatively higher Va , followed by the PV canopy shading group, whereas fabric shading showed the lowest Va . Tg decreased rapidly as solar radiation declined, and the shading‐induced Tg reduction weakened accordingly. Taking scenario 2 as an example, the Tg difference decreased from approximately 5°C at 15:30 to nearly 0°C around 17:00. Therefore, Period B was a transition stage: the cooling effect of shading remained observable at the beginning of the period but weakened rapidly, particularly for Tg . Period C‐Night During Period C, the three groups showed similar nocturnal cooling processes. Ta continued to decrease, and the differences between the shaded and unshaded groups became smaller in all scenarios. Tg also decreased markedly after sunset and then remained similar among the three groups, indicating that the daytime shading effect on the radiant heat load was largely reduced at night. RH increased steadily throughout the night. The RH Va lues under fabric shading and in the unshaded group were generally similar, whereas PV canopy shading showed slightly lower RH during part of this period, with a difference of approximately 4% from the other two groups. Va remained weak, mostly within approximately 0.1–0.3 m/s. The fabric shading generally showed the lowest Va , but the absolute difference was limited because the overall airflow level was low. Overall, Period C was characterized by convergence in Ta and Tg , increased RH, and weak airflow, suggesting that the heat mitigation effect of shading was no longer dominant after sunset. Period D‐ Early Morning During Period D, Ta began to increase again as solar radiation increased. The divergence between the shaded and unshaded groups gradually reappeared after sunrise. Over the grass‐covered surface, differences between Ta and Tg became evident at approximately 07:30. The unshaded group showed a faster increase in Ta , while both shaded groups delayed the warming process. By 09:00, the maximum Ta difference between shaded and unshaded groups reached approximately 2.5°C. RH decreased gradually with increasing Ta , and fabric shading maintained slightly higher RH than the unshaded group, with a difference of approximately 2%. Va varied among the scenarios. In more open spaces, the unshaded group maintained a relatively higher Va , whereas fabric shading generally resulted in a lower Va due to stronger airflow obstruction. Tg increased rapidly after sunrise, and the shading effect became evident. By 09:00, the maximum Tg reduction was approximately 2°C under PV canopy shading and 5°C under fabric shading. On the high‐albedo surface, the divergence between Ta and Tg appeared approximately 0.5–1 h later than on the grass‐covered surface. Overall, Period D showed a gradual re‐establishment of the shading effect after sunrise, with a stronger response in Tg than in Ta . From a whole‐day perspective, the microclimatic differences caused by shading showed a clear diurnal pattern. The strongest differences occurred during Period A, when solar radiation was intense, and shading substantially reduced Tg and Ta while maintaining a relatively higher RH. These differences weakened rapidly during Period B as solar radiation declined and nearly disappeared during Period C, when Ta and Tg among the three groups converged under nocturnal cooling conditions. After sunrise in Period D, the differences gradually reappeared, particularly in Tg , indicating that the shading effect was closely linked to the daily variation in solar radiation. Between the two shaded groups, fabric shading produced a slightly larger maximum reduction and slightly smaller average reduction in Ta and Tg , generally maintained slightly higher RH, but it also showed a more evident airflow obstruction effect, resulting in lower Va . PV canopy shading also reduced thermal exposure, but its cooling effect was generally weaker than that of fabric shading, and its influence on airflow was relatively smaller. Therefore, the two shading stations modified the daily microclimate in different ways: fabric shading provided stronger thermal mitigation but greater wind sheltering, whereas PV canopy shading offered a more moderate cooling effect with less airflow suppression than fabric shading. Daily Variation of ST and Ts Figure 5 shows the daily variation in the ground surface temperature (ST) and shading canopy surface temperature (Ts ) under different spatial scenarios. Panels (a) and (b) compare ST among the unshaded control group, PV canopy shading, and fabric shading, whereas panels (c) and (d) compare Ts between the upper and underside surfaces of the PV and fabric canopies. Table 3 summarizes the mean values during the daytime (05:30–18:35) and nighttime (the remaining period of the day). Daily variations of (a) (b) Surface temperature at the ground level and (c) (d) Surface temperature of shading canopy under different spatial scenarios. TABLE 3Daytime and nighttime heat mitigation performance of the PV canopy and fabric shading in terms of ground surface temperature and canopy surface temperature.PeriodScenarioMean ground surface temperature (ST) (°C)Mean canopy surface temperature (Ts ) (°C)DSEPV–DSEFabric–DSEPV canopyFabricPV‐FabricDaytimeS138.3−2.8−1.439.936.13.8S238.1−1.2−0.444.440.73.7S339.51.80.946.343.82.5D‐Mean38.6−0.7−0.343.540.23.3Night‐timeS128.1−0.1−0.425.725.9−0.2S228.20.50.027.027.3−0.3S330.12.42.227.227.00.2N‐ Mean28.80.90.626.626.7−0.1 Daytime and nighttime heat mitigation performance of the PV canopy and fabric shading in terms of ground surface temperature and canopy surface temperature. Period A‐ Midday/Afternoon In Period A, ST increased rapidly in all scenarios. In the grassland scenarios, the shaded and unshaded groups showed broadly similar ST variations, although shading tended to lower ST during part of the daytime process. In the high‐albedo scenario, however, ST beneath the shaded stations remained relatively high, and PV canopy shading showed the highest ST near the end of Period A. This indicates that the effect of shading on ST was not uniform across surface types. Ts responded more clearly to solar exposure than ST. The upper surfaces of both canopies heated rapidly, whereas the underside surfaces remained lower. The PV upper surface consistently showed the highest Ts , reaching approximately 55–60°C or higher during the hottest periods. The fabric underside generally exhibited the lowest Ts . Thus, Period A was characterized by strong canopy‐surface heating, especially for the PV canopy, whereas the ground‐level ST response was more dependent on the scenario. Period B‐ Late Afternoon During Period B, ST decreased as solar radiation weakened. In the grassland scenarios, the ST curves of the three groups gradually converged. In the high‐albedo scenario, ST beneath the shaded stations, especially under PV canopy shading, remained higher at the beginning of this period but then continuously declined. Ts also decreased rapidly during Period B. The upper canopy surfaces cooled from their daytime peaks, and the temperature difference between the upper and underside surfaces decreased. Although the PV canopy remained warmer than the fabric canopy during the early part of the period, this difference decreased as solar radiation declined. Overall, Period B was a rapid surface‐cooling stage for both ST and Ts . Period C‐Night During Period C, ST remained relatively low and stable. The ST curves of the three groups were close in the grassland scenarios, whereas the shaded groups were slightly warmer than the unshaded group in the high‐albedo scenario. This suggests that nighttime ST differences were limited on grassland but were more evident over high‐albedo surfaces. Ts also converged after sunset. The upper and underside surface temperatures of both canopies decreased to a similar low range, and the daytime differences between the PV and fabric surfaces were greatly reduced. Therefore, canopy material and surface position had much weaker effects on Ts during the nighttime period. Period D‐Early Morning During Period D, ST began to increase again after sunrise. In the grassland scenarios, the three groups were close at the beginning of the period, and differences gradually appeared with increasing solar radiation. Compared with Period A, however, the ST difference remained limited. In the high‐albedo scenario, the shaded groups tended to show higher ST after solar heating intensified. Ts increased earlier and more rapidly than ST, particularly on the upper surfaces of the canopy. The PV upper surface warmed fastest and became higher than the other canopy surfaces, while the fabric upper surface remained lower. The underside surfaces of both canopies increased more slowly and stayed below their corresponding upper surfaces. Thus, Period D marked the reappearance of vertical surface temperature differences within the shading structures. Overall, the surface‐temperature results showed distinct day‐night and scenario‐dependent patterns. During daytime, the mean ST of the unshaded group was 38.6°C. PV canopy shading and fabric shading reduced the daytime mean ST by only 0.7°C and 0.3°C on average, respectively. This limited overall reduction resulted from contrasting responses among scenarios: in S1 and S2, both shading facilities reduced ST, with the PV canopy showing reductions of 2.8°C and 1.2°C and fabric shading showing reductions of 1.4°C and 0.4°C; in S3, however, ST increased by 1.8°C under PV canopy shading and 0.9°C under fabric shading. At night, the shaded groups were slightly warmer on average, with ST increases of 0.9°C under PV canopy shading and 0.6°C under fabric shading, mainly because of the larger nighttime increase in S3. For Ts , the PV canopy was clearly warmer than the fabric canopy during daytime, with a mean difference of 3.3°C. At night, this difference almost disappeared, with nighttime mean Ts values of 26.6°C for the PV canopy and 26.7°C for the fabric canopy. Consequently, the PV canopy generated stronger daytime canopy‐surface heating, whereas fabric shading helped maintain lower canopy surface temperatures. For ground‐level ST, the mitigation effect of both shading facilities was limited and strongly dependent on the underlying surface type. Vertical Temperature Profiles Under Shading Stations Figure 6 shows the vertical air temperature profiles under PV canopy and fabric shading at six measurement heights: 0.50, 1.00, 1.50, 1.75, 2.00, and 2.25 m. Panels (a) and (c) correspond to PV canopy shading, while panels (b) and (d) correspond to fabric shading. Table 4 summarizes the mean vertical air temperature during the daytime (05:30–18:35) and nighttime (the remaining period of the day). Since Figure 6 does not include the unshaded control group, this section focuses on the vertical temperature differences inside the two shading stations rather than the direct temperature reduction relative to the exposed environment. Daily variations of vertical air temperature under (a) (c) PV canopy shading, (b) (d) Fabric shading. TABLE 4Daytime and nighttime performance of the PV canopy and fabric shading in terms of vertical air temperature.PeriodScenarioVertical air temperature under PV (°C)Vertical air temperature under Fabric (°C)0.50 m1.00 m1.50 m1.75 m2.00 m2.25 m0.50 m1.00 m1.50 m1.75 m2.00 m2.25 mDay‐timeS132.632.433.033.033.033.033.032.733.133.533.233.2S235.134.835.235.135.135.135.234.835.135.135.235.0S337.537.237.937.737.737.837.537.437.837.637.735.0D‐Mean35.134.835.435.335.335.335.235.035.335.435.434.4Night‐timeS126.726.826.926.927.027.026.926.826.927.027.027.0S228.128.228.328.428.428.328.328.328.428.428.528.4S328.528.328.528.528.428.328.828.628.628.528.528.3N‐ Mean27.827.827.927.927.927.928.027.928.028.028.027.9 Daytime and nighttime performance of the PV canopy and fabric shading in terms of vertical air temperature. Under PV canopy shading, the vertical temperature differences were generally small, but a weak height‐related pattern was observed during the daytime. In S1, the daytime mean Ta ranged from 32.4°C at 1.00 m to 33.0°C at 1.50–2.25 m, with a maximum vertical difference of 0.6°C. In S2, the corresponding values ranged from 34.8°C to 35.2°C, with a maximum difference of 0.4°C. In S3, the vertical difference was slightly larger, ranging from 37.2°C at 1.00 m to 37.9°C at 1.50 m. These results indicate that PV canopy shading produced only weak vertical stratification, with the middle and upper layers being slightly warmer than the lower layer during the daytime. Under fabric shading, the vertical distribution of Ta was generally uniform, although its height‐dependent pattern differed from that under PV canopy shading. In S1, daytime mean Ta ranged from 32.7°C at 1.00 m to 33.5°C at 1.75 m, with a maximum vertical difference of 0.8°C. In S2, the vertical difference was very small, with mean Ta varying only between 34.8°C and 35.2°C. In S3, most measurement heights remained within a narrow range of 37.4–37.8°C, except for the 2.25 m height, where the mean value was lower. Therefore, fabric shading did not exhibit a consistent upper‐layer warming pattern. Its vertical temperature field was generally uniform, with only local height‐dependent deviations in the high‐albedo surface scenario. A comparison between the two shading stations showed that both facilities produced limited vertical Ta differences within the occupied space. During daytime, the mean Ta across all scenarios ranged from 34.8°C to 35.4°C under PV canopy shading and from 34.4°C to 35.4°C under fabric shading. The PV canopy showed a slightly more evident warming tendency in the middle‐to‐upper layers, especially around 1.50–2.25 m, whereas fabric shading showed a less systematic vertical pattern. This suggests that the two shading materials affected the vertical air temperature field in different ways: PV canopy shading tended to form weak upper‐layer warming, whereas fabric shading maintained a more mixed vertical temperature distribution. At night, the vertical temperature differences became much weaker under both shading stations. Under PV canopy shading, nighttime mean Ta varied within 26.7–27.0°C in S1, 28.1–28.4°C in S2, and 28.3–28.5°C in S3. Under fabric shading, the corresponding ranges were 26.8–27.0°C, 28.3‐28.5°C, and 28.3–28.8°C. The maximum vertical difference was mostly within 0.5 °C, indicating that the height‐related temperature differences almost disappeared after sunset. Therefore, the vertical thermal influence of both shading stations was mainly reflected during the daytime, whereas the nighttime air temperature was nearly vertically uniform. From the perspective of different scenarios, the overall temperature changed more clearly than the vertical gradient. S1 had the lowest Ta , S2 had a higher Ta, and S3 had the highest Ta under both shading stations. However, within each scenario, the differences among the six heights were limited. This indicates that the surface type and background thermal conditions mainly determined the overall air temperature level, whereas the shading structures only caused weak vertical redistribution of Ta within their lower spaces. Overall, the vertical temperature profiles under both shading stations were characterized by small, height‐dependent differences. PV canopy shading showed slightly stronger middle‐to‐upper layer warming, whereas fabric shading maintained a relatively uniform vertical temperature distribution. In addition to air temperature, the black globe temperature was measured at 1.10 m and 2.25 m to further examine the vertical difference in radiant thermal conditions under the shading stations. As shown in Figure 7, Tg at 1.10 m was generally higher than that at 2.25 m during the daytime under both PV canopy and fabric shading. This pattern differs from the vertical air temperature results, where the height‐related differences were generally weak, indicating that the vertical variation in radiant heat exposure was more evident than that in air temperature. The black globe temperature variations at 1.10m and 2.25m height below the (a) PV canopy shading and (b) Fabric shading at 09:00–24:00 on 2024.08.23. Under PV canopy shading, the difference between 1.10 m and 2.25 m was limited in the morning but increased in the afternoon. The maximum vertical Tg difference reached approximately 4–5°C, after which the two curves gradually converged in the evening. Under fabric shading, a similar pattern was observed, but the maximum difference was slightly smaller, generally around 3–4°C. Therefore, both shading stations showed daytime vertical differences in the black globe temperature, with the PV canopy shading producing a slightly stronger vertical Tg gradient than the fabric shading. Overall, Figure 7 indicates that the radiant thermal environment varied more clearly with height than with air temperature. Although Ta remained relatively uniform within the shaded spaces, Tg at pedestrian height was evidently higher than that near the upper layer during the daytime. This suggests that the two shading stations affected not only the air temperature field but also the vertical distribution of radiant heat. Compared with PV canopy shading, fabric shading maintained a weaker vertical Tg difference. In summary, the vertical thermal environment is not a simple cooling space but rather a stratified system with competing thermal forces. While the ambient air at pedestrian height remains cooler, the human body is simultaneously subjected to a compounded radiative burden: lingering upward longwave radiation from the underlying surface and amplified downward longwave emissions from the solar‐heated canopy overhead. This has profound implications for urban design; to optimize outdoor thermal comfort, a holistic and systems‐based approach is necessary. Shading design must be integrated with ground surface treatment (e.g., using vegetation or low‐emissivity materials) to mitigate the elevated radiant load at the occupant level. Simply installing shading facilities over conventional heat‐absorbing pavements represents an incomplete and potentially suboptimal solution for achieving thermal comfort. Microclimatic Analysis During Full‐Shading Period (10:30–13:30) During the full‐shading period from 10:30 to 13:30, the two shaded stations produced clear differences in the local microclimate compared with the unshaded control group. As shown in Figure 8a, the mean Ta at 1.10 m was 39.6 °C under unshaded conditions, whereas it decreased to 36.7°C under PV canopy shading and 36.9°C under fabric shading. This corresponded to reductions of 2.9°C and 2.7°C, respectively. The reduction in Tg was more pronounced. Figure 8b shows that Tg decreased from 47.2°C in the unshaded group to 40.6°C under PV canopy shading and 39.4°C under fabric shading, with reductions of 6.6°C and 7.8°C, respectively. This indicates that the heat mitigation effect during the full‐shading period was more strongly reflected in the radiant thermal environment than in the air temperature. Mean values of microclimatic variables at 10:30–13:30 during the observation period. (a) 1.10‐m air temperature, (b) Black globe temperature, (c) Relative humidity, (d) Wind speed, (e) Surface temperature at the ground level, (f) Surface temperature of shading canopy. RH also increased under shaded conditions. As shown in Figure 8c, RH increased from 47.6% in the unshaded group to 52.8% under PV canopy shading, and 55.1% under fabric shading. In contrast, the effect on Va differed between the two shaded stations. Figure 8d shows that Va increased from 1.01 m/s in the unshaded group to 1.11 m/s under PV canopy shading, whereas it decreased slightly to 0.96 m/s under fabric shading. This suggests that the PV canopy did not suppress airflow during the full‐shading period, whereas the fabric‐shaded station showed a weak wind‐sheltering effect. At the surface level, both shading stations reduced ST. As shown in Figure 8e, ST decreased from 42.3°C in the unshaded group to 36.7°C under PV canopy shading and 37.2°C under fabric shading. However, Ts differed clearly between the two shading materials. Figure 8f shows that the mean Ts of the PV canopy was 52.3°C, which was 6.7°C higher than that of the fabric canopy. Therefore, although both shading stations reduced Ta , Tg , and ST, the PV canopy accumulated more surface heat than the fabric canopy during the full‐shading period. These differences can be explained by the changes in the radiative components shown in Figures 9 and 10. Figure 9a shows that the total radiation intensity decreased from 1865 W/m2 in the unshaded group to 1346 W/m2 under PV canopy shading and 1341 W/m2 under fabric shading, respectively. Thus, both the shading stations reduced the total radiative load by more than 500 W/m2, and the difference between the two shaded groups was very small. The dominant contributor was the reduction in downward shortwave radiation. Compared with direct sunlight, downward shortwave radiation decreased by 541 W/m2 under PV canopy shading and 532 W/m2 under fabric shading. This large reduction explains the marked decrease in Tg and ST under both shaded conditions. Mean (a) downward shortwave (red), upward shortwave (yellow), downward longwave (grey), and upward longwave (light grey) radiation. These are the average data observed from 10:30‐13:30 during the observation period. Differences in the mean radiation (b) between under the PV shading and in direct sunlight, and (c) between under the fabric shading and in direct sunlight. Differences in the mean radiation per hour at 10:30–13:30. (a) Downward shortwave, (b) Upward shortwave, (c) Downward longwave, and (d) Upward longwave radiation. However, the two shading stations differed in their longwave response. Downward longwave radiation increased by 50 W/m2 under PV canopy shading and by 36 W/m2 under fabric shading, whereas the changes in upward longwave radiation were small. This indicates that part of the shortwave radiation reduction was offset by additional longwave radiation emitted from the heated canopy surfaces, especially under the PV canopy. This is consistent with the higher Ts of the PV canopy shown in Figure 8f. Therefore, although the PV canopy slightly reduced ST and maintained higher Va , its hotter canopy surface enhanced downward longwave radiation, which partly weakened the radiant cooling effect. In contrast, fabric shading resulted in a lower canopy surface temperature and a smaller increase in longwave radiation, which helps explain the lower Tg during the full‐shading period. Figure 10 further shows the hourly variation in the radiative differences during the observation period. The differences in downward shortwave radiation remained much larger than those in longwave radiation throughout 10:30–13:30, confirming that shortwave radiation interception was the primary mechanism of heat mitigation for both shading stations. PV canopy shading generally showed a slightly larger shortwave radiation difference than fabric shading, whereas the longwave‐related difference was more evident under the PV canopy. This combined effect explains why the two shading stations had similar reductions in total radiation but produced different microclimatic responses: fabric shading achieved a slightly greater reduction in Tg and higher RH, whereas PV canopy shading maintained slightly lower ST and higher Va but had a warmer canopy surface. Comparison of Diurnal and Nocturnal Heat Mitigation Performance Between PV Canopy and Fabric Shading To assess the overall microclimatic regulation capacity of artificial shading stations, this subsection compares the heat mitigation performance of the PV canopy and fabric shading during both daytime (from sunrise to sunset, 05:30–18:35) and nighttime (from sunset to sunrise,18:35–05:30). The three assessment indicators, Ta and Tg , are presented in Table 5. TABLE 5Daytime and nighttime maximum and minimum heat mitigation performance of PV canopy and fabric shading in Ta , and Tg .TimeScenarioMean Ta (°C)Mean T g (°C)DSEPV‐ DSEFabric‐ DSEDSEPV‐ DSEFabric‐ DSEDay‐timeS134.5−0.3−1.041.0−4.4−4.7S236.1−1.2−1.342.0−3.4−4.2S338.7−1.5−1.344.5−2.7−3.9D‐Mean36.4−1.0−1.242.5−3.5−4.3 Night‐time S128.00.00.127.727.827.7S228.20.00.127.928.428.3S327.10.00.327.628.428.2N‐Mean27.80.00.227.728.228.1 Note: the maximum value(s) in each column is underlined and the minimum value(s) bold. Daytime and nighttime maximum and minimum heat mitigation performance of PV canopy and fabric shading in Ta , and Tg . Night‐time Note: the maximum value(s) in each column is underlined and the minimum value(s) bold. During daytime, both shading stations reduced Ta , but the magnitude of air‐temperature reduction was relatively limited. The daytime mean Ta of the unshaded group was 36.4°C. Compared with the unshaded condition, PV canopy shading reduced Ta by 1.0°C on average, whereas fabric shading reduced Ta by 1.2°C. Across the three scenarios, the Ta reduction under PV canopy shading increased from 0.3°C in S1 to 1.2°C in S2 and 1.5°C in S3. Under fabric shading, the corresponding reductions were 1.0°C, 1.3°C, and 1.3°C. This indicates that both shading stations provided a modest daytime cooling effect on Ta , with fabric shading showing a slightly larger average reduction, while PV canopy shading showed the strongest Ta reduction in S3. The daytime reduction in Tg was more pronounced than that in Ta . The daytime mean Tg of the unshaded group was 42.5°C. PV canopy shading reduced Tg by 3.5°C on average, whereas fabric shading reduced Tg by 4.3°C. In S1, Tg decreased by 4.4°C under PV canopy shading and 4.7°C under fabric shading. In S2, the reductions were 3.4 °C and 4.2 °C, respectively. In S3, the reductions were 2.7°C and 3.9°C. These results show that the shading effect was more evident in Tg than in Ta , confirming that the main daytime cooling contribution of the shaded stations was related to the mitigation of radiant heat exposure. Fabric shading consistently produced a larger Tg reduction than PV canopy shading in all three scenarios. During nighttime, the influence of the shading stations on Ta was very weak. The nighttime mean Ta of the unshaded group was 27.8°C. PV canopy shading showed almost no nighttime effect on Ta , with an average difference of 0.0°C across the three scenarios. fabric shading showed a slight warming effect, with an average increase of 0.2°C. Specifically, fabric shading increased nighttime Ta by 0.1°C in S1 and S2, and by 0.3°C in S3. Therefore, the nighttime air temperature effect of both shading stations was much weaker than their daytime effect, and fabric shading showed only a minor heat‐retention tendency. For nighttime Tg , the differences among the three groups were also small, but the shaded groups were generally slightly higher than those of the unshaded group. In S1, Tg was almost identical among the three groups, with values around 27.7–27.8°C. In S2, Tg increased from 27.9°C in the unshaded group to 28.4°C under PV canopy shading and 28.3°C under fabric shading. In S3, Tg increased from 27.6°C to 28.4°C and 28.2°C under PV canopy shading and fabric shading, respectively. On average, nighttime Tg was 27.7°C in the unshaded group, 28.2°C under PV canopy shading, and 28.1°C under fabric shading. This suggests that the shading structures had a slight nighttime warming effect on the radiant thermal environment, particularly under the PV canopy shading. Overall, the comparison between daytime and nighttime shows that the cooling effect of both shading stations was mainly concentrated during daytime. For Ta , the daytime reductions were modest, with average decreases of 1.0°C under PV canopy shading and 1.2°C under fabric shading. For Tg , the reductions were much larger, reaching 3.5°C and 4.3°C, respectively. At night, both Ta and Tg differences became very small. PV canopy shading was nearly neutral for Ta but slightly increased Tg , whereas fabric shading produced a minor warming effect on both Ta and Tg . Between the two shading stations, fabric shading provided slightly stronger daytime heat mitigation, particularly for Tg , whereas PV canopy shading had a somewhat stronger effect on maintaining Tg at night. Analysis of Outdoor Thermal Comfort Under DSE, PV and Fabric Shading Conditions Comparison of T mrt Under Shaded and Unshaded Conditions Figure 11 shows the daily variation in T mrt under unshaded conditions, PV canopy shading, and fabric shading across the three spatial scenarios. Compared with the microclimatic variables discussed above, T mrt showed a much stronger response to direct solar exposure and shading. The difference between shaded and unshaded conditions was mainly concentrated during the daytime, while the three groups became close after sunset. Daily variations of mean radiant temperature under different spatial scenarios. Period A‐Midday/Afternoon During Period A, T mrt under unshaded conditions increased rapidly and remained at a high level in all scenarios. In S1, the unshaded T mrt generally stayed around 70–80°C during the daytime heating stage. In S2 and S3, the unshaded group reached even higher levels, with peak values approaching or exceeding 80°C. In contrast, both shaded groups maintained much lower T mrt throughout most of Period A. Under PV canopy shading, T mrt was generally around 45–65°C, while under fabric shading, it was mostly lower, approximately within 40–55°C. Therefore, both shading stations substantially reduced daytime radiant heat exposure, and fabric shading generally showed a stronger reduction than PV canopy. Period B‐Late afternoon During Period B, T mrt decreased rapidly in all groups as solar radiation weakened. The unshaded group still remained higher than the shaded groups at the beginning of this period, but the difference narrowed quickly. Under PV canopy and fabric shading, T mrt also decreased, and the two shaded curves gradually approached each other. By the end of Period B, the T mrt values of the three groups were similar, indicating that the shading‐induced difference in radiant heat exposure weakened rapidly after direct solar radiation declined. Period C‐Night During Period C, T mrt remained low and stable. The curves of the unshaded group, PV canopy shading, and fabric shading almost overlapped in most nighttime intervals, generally staying around 25–30°C. This indicates that after sunset, the difference in radiant thermal conditions among the three groups became very small. Therefore, the heat mitigation effect of shading on T mrt was mainly a daytime phenomenon. Period D‐Early Morning During Period D, T mrt began to increase again after sunrise. The unshaded group increased more rapidly than the shaded groups, particularly when direct solar exposure became stronger. Under both shading stations, the morning increase in T mrt was delayed and suppressed. The difference between shaded and unshaded conditions gradually reappeared, but it was still smaller than that observed during the midday/afternoon period. The boxplot in Figure 11b further confirms the overall differences among the three groups. The unshaded condition had the widest T mrt distribution and the highest upper range, reflecting a strong daytime exposure to solar radiation. Both shading stations reduced the overall T mrt and narrowed the distribution range. Between the two shaded groups, fabric shading generally showed a slightly lower T mrt distribution than PV canopy shading, indicating a stronger ability to reduce radiant heat exposure for the former. Overall, the results show that the main contribution of shading to thermal mitigation was the reduction of the daytime radiant heat load, with fabric shading performing slightly better than PV canopy shading in terms of T mrt reduction. Comparison of UTCI Under DSE, PV Canopy, and Fabric Groups Figure 12 shows the UTCI variations in the unshaded control group (DSE), PV canopy shading group, and fabric shading group during the three consecutive summer days. Figure 13 further presents the UTCI differences between the shaded groups and the DSE group, where negative values indicate a UTCI reduction under shading. The UTCI levels were interpreted according to the heat stress categories shown in Figure 12. Comparison of UTCI levels in three consecutive summer days. UTCI difference between two shading scenarios and DSE scenario. In S1, corresponding to the grassland scenario on a hot day, the UTCI in the DSE group increased markedly during the daytime and reached a very strong heat stress level. Both PV canopy and fabric shading reduced the UTCI during most of the daytime period. As shown in Figure 13, the reduction was generally about 5–8°C before the late afternoon, with fabric shading slightly stronger than PV canopy shading for much of this period. After sunset, the three groups became close, indicating that the shading effect on the UTCI was mainly limited to the daytime. In S2, corresponding to the grassland scenario on an extreme heat day, the DSE group experienced extreme heat stress during the hottest period. Both shading stations clearly reduced the UTCI, keeping the shaded groups mostly below the DSE curve. The UTCI reduction was generally about 4–6°C during the main daytime heating period, and fabric shading again showed a slightly larger reduction at several moments. In the evening and nighttime, the UTCI difference among the three groups decreased. In S3, corresponding to the high‐albedo surface scenario on an extreme heat day, the DSE group again reached the extreme heat stress range during the daytime. The shaded groups also experienced high UTCI, but their values remained lower than the DSE group during the main daytime period. The reduction was approximately 5–8°C from morning to early afternoon, with fabric shading generally producing slightly lower UTCI than PV canopy shading. During transition periods and nighttime, the differences among the three groups weakened. Overall, the unshaded group exhibited peak UTCI values exceeding 46°C, reaching the “extreme heat stress” category during the hottest hours of the day. In comparison, both fabric and PV canopy shading stations substantially reduced UTCI. In Scenario 3, the fabric shading achieved a greater cooling effect‐ up to 7.5°C lower during peak hours, whereas PV shading reduced UTCI by up to 6.1°C. Although shading did not fully prevent exposure to “very strong heat stress,” it effectively attenuated the intensity and duration of thermal stress. These results underscore the importance of solar control facilities in improving outdoor thermal comfort during hot periods. UTCI Reduction Performance During Full‐Shading Period (10:30–13:30) A statistical comparison of the UTCI under different shading stations during the full‐shading period was conducted. Figure 14 displays the average UTCI values and standard deviations observed in the unshaded, PV canopy shading, and fabric shading groups. In the unshaded group, the average UTCI during the midday period reached 47.5°C, which fell within the category of “extreme heat stress” (UTCI > 46°C). In comparison, the UTCI under both fabric and PV canopy shading averaged 41.2°C and 42.0°C, respectively, corresponding to reductions of 6.3°C and 5.5°C relative to unshaded spaces. Welch's t‐test confirmed that the shaded groups had significantly lower UTCI values than the unshaded group (p < 0.001). Posthoc comparisons using the Bonferroni correction further verified the statistical significance of these differences. Overall, both artificial shading facilities significantly alleviated thermal stress during midday on hot days, with fabric shading demonstrating slightly better performance than PV canopy shading. Average UTCI at 10:30–13:30 during the midday period. (a) Mean value, (b) Mean difference between DSE and PV canopy shading, and (c) Mean difference between DSE and fabric shading. In S1, fabric shading demonstrated a slightly greater cooling effect, reducing the average UTCI by 4.5°C compared to unshaded group, whereas PV canopy shading resulted in a UTCI reduction of 4.3°C. Under more intense heat conditions (S2 and S3), fabric shading outperformed PV canopy shading, with an average advantage of approximately 0.5°C to 1.5°C, indicating better effectiveness in mitigating thermal stress. The average UTCI reduction and error ranges under each condition are shown in Figure 14. The results suggest that while both shading stations achieved comparable average UTCI reductions, Fabric shading generally achieved more consistent UTCI reductions across different spatial scenarios, indicating greater performance stability under highly exposed conditions. Overall, the full‐shading period was a critical time window for assessing the cooling capacities of the different shading stations. The performance difference between the two shading stations was relatively small under the typical hot day scenario, whereas fabric shading provided greater thermal stress mitigation under extreme heat conditions. Comparison of Cooling Performance Between PV Canopy and Fabric Shading To assess the overall thermal comfort regulation capacity of the two shading stations, this subsection compares the heat mitigation performance of PV canopy shading and fabric shading during both daytime and nighttime in terms of T mrt and UTCI. Daytime was defined as 05:30–18:35, and nighttime was defined as 18:35–05:30. The mean values and shading‐induced differences are presented in Table 6. TABLE 6Daytime and nighttime maximum and minimum heat mitigation performance of PV canopy and fabric shading in T mrt, and UTCI.TimeScenarioMean T mrt (°C)Mean UTCI (°C)DSEPV‐DSEFabric‐DSEDSEPV‐DSEFabric‐DSEDay‐timeS166.8−21.1−25.241.5−4.3−4.5S265.6−16.1−21.143.2−2.8−3.2S369.1−15.5−20.144.7−2.2−2.8D‐Mean67.2−17.5−22.143.1−3.1−3.5 Night‐time S125.52.21.732.3−0.7−0.4S226.02.62.232.4−0.40.0S337.60.1−1.133.80.00.4N‐Mean29.71.70.932.9−0.40.0 Note: the maximum value(s) in each column is underlined and the minimum value(s) bold. Daytime and nighttime maximum and minimum heat mitigation performance of PV canopy and fabric shading in T mrt, and UTCI. Night‐time Note: the maximum value(s) in each column is underlined and the minimum value(s) bold. During the daytime, both shaded stations substantially reduced T mrt and UTCI compared with the unshaded condition. The daytime mean T mrt of the unshaded group was 67.2°C. PV canopy shading reduced T mrt by 17.5°C on average, whereas fabric shading produced a larger reduction of 22.1°C. Across the three scenarios, the T mrt reduction under PV canopy shading was 21.1°C in S1, 16.1°C in S2, and 15.5°C in S3. Under fabric shading, the corresponding reductions were 25.2, 21.1, and 20.1°C. Thus, fabric shading consistently exhibited stronger daytime radiant heat mitigation than PV canopy shading. A similar pattern was observed for the UTCI. The daytime mean UTCI of the unshaded group was 43.1°C. PV canopy shading reduced UTCI by 3.1°C on average, whereas fabric shading reduced it by 3.5°C. In S1, the reductions were 4.3°C under PV canopy shading and 4.5°C under fabric shading. In S2, they were 2.8°C and 3.2°C, respectively, and in S3, they were 2.2°C and 2.8°C. These results indicate that both the shading stations improved daytime thermal comfort, with fabric shading showing a slightly stronger UTCI reduction. The reduction in both T mrt and UTCI was largest in S1 and became smaller in the two extreme‐heat scenarios, indicating that shading remained effective under severe heat conditions, but its relative mitigation magnitude decreased as background heat stress intensified. During the nighttime, the effects of the shading stations were much weaker and differed from those during the daytime. For T mrt, both shading stations showed a slight average nighttime warming effect. The nighttime mean T mrt of the unshaded group was 29.7°C, whereas PV canopy shading increased T mrt by 1.7°C on average, and fabric shading increased it by 0.9°C. The warming effect was most evident in S2, with increases of 2.6°C under PV canopy shading and 2.2°C under fabric shading. In S3, however, fabric shading reduced nighttime T mrt by 1.1°C, which was different from the other nighttime cases. The differences among the three groups for nighttime UTCI were small. The nighttime mean UTCI of the unshaded group was 32.9°C. PV canopy shading reduced UTCI by 0.4°C on average, whereas fabric shading showed no average change. In S1, both shading stations slightly reduced UTCI, with reductions of 0.7°C under PV canopy shading and 0.4°C under fabric shading. In S2, PV canopy shading reduced UTCI by 0.4°C, while fabric shading showed no difference. In S3, PV canopy shading was neutral, whereas fabric shading slightly increased UTCI by 0.4°C. Therefore, the nighttime UTCI effect was weak and scenario‐dependent. Overall, the cooling performance of the two shading facilities was primarily observed during the daytime. Both PV canopy shading and fabric shading significantly reduced T mrt and UTCI, but fabric shading performed better, especially in reducing T mrt. At night, the mitigation effect largely disappeared. Shading stations tended to slightly increase T mrt, particularly under PV canopy shading, whereas their influence on UTCI was minimal. Therefore, fabric shading showed a stronger daytime thermal comfort improvement, whereas PV canopy shading exhibited a more evident nighttime radiant warming tendency.4.5 Temporal variation of thermal stress in different shading scenarios Table 7 presents the distribution of UTCI‐related thermal stress levels under unshaded, PV canopy, and fabric shading during the daytime and nighttime periods. Figure 15 further illustrates the transfer of thermal stress levels from the unshaded condition to the two shaded conditions under the same conditions. The UTCI classification includes moderate, strong, very strong, and extreme heat stress. No thermal stress was observed in any group during the observation period. TABLE 7Thermal sensation/stress classification on UTCI scale.Physiological stressUTCI Range(°C)Daytime (%)Nighttime (%)Whole day (%)DSEPVFabricDSEPVFabricDSEPVFabric No thermal stress (Level 1) +9 to +26000000000 Moderate heat stress (Level 2) +26 to +321.40038.269.154.517.229.724.2 Strong heat stress (Level 3) +32 to +3815.123.323.361.830.945.535.126.651.6 Very strong heat stress (Level 4) +38 to +4645.274.074.000025.842.222.7 Extreme heat stress (Level 5) >+4638.32.72.700021.91.51.5 Thermal sensation/stress classification on UTCI scale. No thermal stress (Level 1) Moderate heat stress (Level 2) Strong heat stress (Level 3) Very strong heat stress (Level 4) Extreme heat stress (Level 5) Transfer of thermal stress level (Levels 1–5): (a) Daytime fabric shading, (b) Daytime PV canopy shading, (c) Nighttime fabric shading, (d) Nighttime PV canopy shading. During the daytime, the unshaded condition showed the highest thermal stress level. Extreme heat stress accounted for 38.3% of the daytime period, whereas very strong heat stress accounted for 45.2%. Therefore, 83.5% of the daytime period under unshaded conditions was exposed to very strong or extreme heat stress. Under both PV canopy and fabric shading, the proportion of extreme heat stress was reduced sharply to 2.7%. Meanwhile, very strong heat stress accounted for 74.0%, and strong heat stress accounted for 23.3% under both shade conditions. This indicates that shading substantially reduced the occurrence of extreme heat stress, but most daytime periods under shade remained within the very strong heat stress category. Figure 15a,b further show that the daytime shading effect was mainly reflected in the transfer from higher stress levels to lower levels, especially from Level 5 to Level 4 or Level 3. During nighttime, the severity of heat stress decreased markedly in all groups. Extreme and very strong heat stress were not observed at night. Under unshaded conditions, strong heat stress accounted for 61.8% of the nighttime period, whereas moderate heat stress accounted for 38.2%. Under PV canopy shading, the proportion of moderate heat stress increased to 69.1%, whereas that of strong heat stress decreased to 30.9%. Under fabric shading, moderate and strong heat stress accounted for 54.5% and 45.5% of the total stress, respectively. These results indicate that both shading stations reduced nighttime thermal stress levels, with PV canopy shading showing a larger shift from strong to moderate heat stress. This transfer pattern is also reflected in Figure 15c,d. Over the whole day, the unshaded condition still showed a high proportion of severe thermal stress, with extreme heat stress accounting for 21.9% and very strong heat stress accounting for 25.8%. Under both shaded conditions, the whole‐day proportion of extreme heat stress decreased to 1.5%, indicating a clear reduction in the most severe thermal exposure for the bees. However, the distribution of the remaining stress levels differed between the two stations. Under PV canopy shading, very strong heat stress accounted for 42.2%, strong heat stress for 26.6%, and moderate heat stress for 29.7%. Under fabric shading, very strong heat stress was lower at 22.7%, while strong and moderate heat stress accounted for 51.6% and 24.2%, respectively. This suggests that fabric shading was more effective in reducing the duration of very strong heat stress over the whole day, whereas PV canopy shading showed a larger proportion of moderate heat stress. Overall, both the shading stations effectively reduced the temporal duration of extreme heat stress. During the daytime, their main contribution was the reduction of Level 5 extreme heat stress, although very strong heat stress still dominated under shaded conditions. At night, the shading effect was mainly reflected in the shift from strong to moderate heat stress, particularly under PV canopy shading. Therefore, the two shading stations did not completely eliminate thermal stress, but they clearly reduced the occurrence of the most severe UTCI categories and redistributed thermal exposure toward lower‐stress levels. Correlation Between UTCI and Microclimatic Variables To identify the dominant microclimatic variables affecting UTCI during different daily periods, Table 8 presents the Pearson correlation coefficients between UTCI and five climatic variables, namely Ta , Va , RH, Tg , and T mrt, under DSE, PV, and fabric shading conditions. The analysis was conducted for four periods: early morning, midday/afternoon, late afternoon, and evening. The significance level of each correlation coefficient is indicated in Table 8. Unless otherwise specified as non‐significant, the reported correlations were statistically significant at p ≤ 0.001. TABLE 8Pearson linear correlation between UTCI and the climatic variables (Ta , Va , RH and Tg, T mrt) for each site and different period of the day during the period of measurements (2024.08.22 to 2024.08.24).PeriodSiteTaVaRHTgTmrtA Midday/AfternoonDSEr = 0.9051***r = ‐0.8344***r = ‐0.8673***r = 0.9607***r = 0.4505***PV canopyr = 0.9141***r = ‐0.6745***r = ‐0.8525***r = 0.9687***r = 0.8253***Fabricr = 0.9455***r = ‐0.3892***r = ‐0.8814***r = 0.9839***r = 0.8275***B Late afternoonDSEr = 0.9016***r = 0.7155***r = ‐0.6364***r = 0.9795***r = 0.9818***PV canopyr = 0.9389***r = 0.3575**r = ‐0.7309***r = 0.9859***r = 0.9811***Fabricr = 0.9326***r = 0.3330*r = ‐0.7186***r = 0.9877***r = 0.9803***C NightDSEr = 0.8636***r = 0.2459**r = ‐0.4978***r = 0.9150***r = 0.9201***PV canopyr = 0.9032***r = 0.1746*r = ‐0.5809***r = 0.8909***r = 0.6399***Fabricr = 0.9091***r = ‐0.3090***r = ‐0.6297***r = 0.9271***r = 0.8733***D Early morningDSEr = 0.9719***r = 0.8121***r = ‐0.9015***r = 0.9931***r = 0.9808***PV canopyr = 0.9423***r = 0.5711***r = ‐0.8840***r = 0.9573***r = 0.9581***Fabricr = 0.9055***r = ‐0.1987ns r = ‐0.8352***r = 0.9513***r = 0.9151*** Note: *p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, ns No significance. Pearson linear correlation between UTCI and the climatic variables (Ta , Va , RH and Tg, T mrt) for each site and different period of the day during the period of measurements (2024.08.22 to 2024.08.24). Note: *p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, ns No significance. During the early morning period, the UTCI was strongly correlated with most thermal variables under all three conditions. In the DSE group, Tg and T mrt showed the strongest correlations with UTCI, with correlation coefficients of 0.9931 and 0.9808, respectively. Ta was also strongly correlated with UTCI (r = 0.9719), whereas RH showed a strong negative correlation (r = ‐0.9015). Va also had a strong positive correlation with UTCI in the DSE group (r = 0.8121). Under PV canopy shading, Tg , T mrt, and Ta remained strongly correlated with UTCI, with coefficients of 0.9573, 0.9581, and 0.9423, respectively. RH showed a strong negative correlation (r = ‐0.8840), whereas Va showed a moderate positive correlation (r = 0.5711, p ≤ 0.01). Under fabric shading, Tg , T mrt, and Ta were also strongly correlated with UTCI, with coefficients of 0.9513, 0.9151, and 0.9055, respectively. RH remained negatively correlated with UTCI (r = ‐0.8352), whereas Va showed no significant correlation (r = ‐0.1987, p > 0.05). These results indicate that the early morning UTCI was mainly associated with air temperature and radiant thermal variables, whereas the role of Va differed between the two shaded stations. During the midday/afternoon period, Tg and Ta were the dominant variables associated with UTCI. In the DSE group, Tg showed the strongest positive correlation with UTCI (r = 0.9607), followed by Ta (r = 0.9051). RH showed a strong negative correlation (r = ‐0.8673), and Va was strongly correlated with UTCI (r = 0.8344). T mrt showed a weaker but still significant positive correlation (r = 0.4505). Under PV canopy shading, Tg and Ta were strongly positively correlated with UTCI, with coefficients of 0.9687 and 0.9141, respectively. RH and Va were negatively correlated with UTCI, with coefficients of −0.8525 and −0.6745, respectively, whereas T mrt showed a strong positive correlation (r = 0.8253). Under fabric shading, Tg showed the strongest correlation with UTCI (r = 0.9839), followed by Ta (r = 0.9455) and T mrt (r = 0.8275), respectively. RH showed a strong negative correlation (r = ‐0.8814), whereas Va showed a weaker negative correlation (r = ‐0.3892). Therefore, during midday/afternoon, UTCI was mainly controlled by Tg and Ta under all three conditions, whereas RH showed an inverse relationship with UTCI, and Va exhibited different correlation directions between the unshaded and shaded environments. During the late afternoon period, the correlations between UTCI and radiant thermal variables became particularly strong. In the DSE group, T mrt and Tg showed the strongest correlations with UTCI, with coefficients of 0.9818 and 0.9795, respectively. Ta was also strongly correlated with UTCI (r = 0.9016), while RH showed a moderate negative correlation (r = ‐0.6364). Va was positive correlated with UTCI (r = 0.7155, p ≤ 0.01). Under PV canopy shading, Tg and T mrt remained strongly correlated with UTCI, with coefficients of 0.9859 and 0.9811, respectively, followed by Ta (r = 0.9389). RH showed a strong negative correlation (r = −0.7309), whereas Va showed a weaker positive correlation (r = 0.3575, p ≤ 0.05). Under fabric shading, Tg , T mrt, and Ta were also strongly positively correlated with UTCI, with coefficients of 0.9877, 0.9803, and 0.9326, respectively. RH remained negatively correlated with UTCI (r = −0.7186), whereas Va showed a weak positive correlation (r = 0.3330, p ≤ 0.05). This indicates that, during the late afternoon, the UTCI was strongly associated with both radiant heat and air temperature under all three conditions, whereas the contribution of Va was relatively weaker. During the nighttime period, UTCI still showed strong positive correlations with Ta and Tg , and T mrt. In the DSE group, T mrt and Tg were strongly correlated with UTCI, with coefficients of 0.9201 and 0.9150, respectively, followed by Ta (r = 0.8636). RH showed a moderate negative correlation (r = −0.4978), whereas Va showed a weak positive correlation (r = 0.2459, p ≤ 0.05). Under PV canopy shading, Ta and Tg showed strong positive correlations with UTCI, with coefficients of 0.9032 and 0.8909, respectively. T mrt also showed a positive correlation (r = 0.6390, p ≤ 0.01), while RH showed a negative correlation (r = −0.5809). Va was weakly correlated with UTCI (r = 0.1746; no significance). Under fabric shading, Tg , Ta , and T mrt were strongly positively correlated with UTCI, with coefficients of 0.9271, 0.9091, and 0.8733, respectively. RH showed a negative correlation (r = −0.6297), while Va showed a weak negative correlation (r = −0.3090, p ≤ 0.01). These results suggest that the nighttime UTCI is mainly associated with Ta and radiant thermal variables, whereas Va has a relatively weak and unstable relationship with the UTCI. Overall, the correlation results showed that the UTCI was controlled by different combinations of variables throughout the day. Tg consistently showed the strongest or near‐strongest positive correlation with UTCI in all periods and under all three conditions, indicating the central role of globe temperature‐related radiant heat exposure. Ta was also strongly and positively correlated with UTCI throughout the day. T mrt showed particularly strong correlations during the late afternoon, night, and early morning, but its correlation was weaker during the midday/afternoon period in the DSE group. RH was consistently negatively correlated with UTCI, reflecting the inverse variation between relative humidity and thermal stress during heating processes. Va showed the most variable behavior: it was positively correlated with UTCI in the DSE group, but its correlation became weaker, negative, or non‐significant under shaded conditions. This indicates that the UTCI variation under shading stations was governed mainly by air temperature and radiant thermal conditions, while the role of airflow was secondary and depended on shading type and time period. To further examine the effects of weather conditions and underlying surface type, scenario‐specific correlation analyses were conducted for S1, S2, and S3 (Tables 9, 10, 11). Unless otherwise specified, the reported correlations were statistically significant at p ≤ 0.001. TABLE 9Pearson linear correlation between UTCI and the climatic variables (Ta , Va , RH and Tg, T mrt) for each site and different periods of the day during Scenario 1 (2024.08.22).PeriodSiteTaVaRHTgTmrtA Midday/AfternoonDSE0.9710***−0.7045***−0.9452***0.9533***−0.8237***PV canopy0.7969***−0.3423*−0.7793***0.9876***0.8277***Fabric0.8972***−0.3178ns−0.8012***0.9873***0.8311***B Late afternoonDSE0.9663***0.8439***−0.9444***0.9975***0.9818***PV canopy0.9816***0.6389**−0.9759***0.9928***0.9676***Fabric0.9301***0.4928*−0.9367***0.9867***0.9460***C NightDSE0.6946***−0.2489ns −0.3913*0.6649***0.3804*PV canopy0.3326ns −0.5306**0.0243ns 0.8036***0.4505**Fabric0.7136***−0.1516ns −0.3936*0.6862***0.3766* Note: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, ns No significance. Pearson linear correlation between UTCI and the climatic variables (Ta , Va , RH and Tg, T mrt) for each site and different periods of the day during Scenario 1 (2024.08.22). Note: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, ns No significance. TABLE 10Pearson linear correlation between UTCI and the climatic variables (Ta , Va , RH and Tg, T mrt) for each site and different periods of the day during Scenario 2 (2024.08.23).PeriodSiteTaVaRHTgTmrtA Midday/AfternoonDSE0.9268***0.1765ns −0.8761***0.9915***0.8179***PV canopy0.8007***−0.1055ns −0.7624***0.9912***0.6444***Fabric0.9314***0.2231ns −0.8907***0.9446***0.8287***B Late afternoonDSE0.9940***0.7445***−0.9591***0.9975***0.9911***PV canopy0.9846***0.0562ns −0.9436***0.9934***0.9673***Fabric0.9897***0.4500ns −0.9522***0.9964***0.9885***C NightDSE0.9895***0.1955ns −0.9726***0.9882***0.9212***PV canopy0.9915***0.1626ns −0.9663***0.9929***0.9515***Fabric0.9826***−0.8195***−0.9594***0.9824***0.9162***D Early morningDSE0.9824***0.9309***−0.9871***0.9991***0.9819***PV canopy0.9615***0.6733***−0.9623***0.9770***0.9680***Fabric0.8822***−0.6260**−0.8992***0.9907***0.8775*** Note: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, ns No significance. Pearson linear correlation between UTCI and the climatic variables (Ta , Va , RH and Tg, T mrt) for each site and different periods of the day during Scenario 2 (2024.08.23). Note: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, ns No significance. TABLE 11Pearson linear correlation between UTCI and the climatic variables (T mrt, Ta , Va , RH and Tg ) for each site and different period of the day during Scenario 3 (2024.08.24).PeriodSiteTaVaRHTgTmrtA Midday/AfternoonDSE0.6443***−0.0078ns −0.4643**0.9775***0.7646***PV canopy0.7389***0.4334**−0.7766***0.9421***0.9411***Fabric0.6905***0.4180**−0.6248***0.9708***0.8850***B Late afternoonDSE0.9796***0.7538***−0.8891***0.9951***0.9967***PV canopy0.9742***0.4647ns −0.8889***0.9974***0.9921***Fabric0.9743***0.6233**−0.8853***0.9959***0.9931***C NightDSE0.8797***0.4528***−0.4675***0.9223***0.9545***PV canopy0.9708***0.3778**−0.5559***0.8943***0.6275***Fabric0.9307***0.5842***−0.6389***0.9308***0.9048***D Early morningDSE0.9836***0.8969***−0.9842***0.9994***0.9836***PV canopy0.9474***0.6852***−0.9546***0.9801***0.9544***Fabric0.9515***0.5902**−0.9594***0.9792***0.9474*** Note: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, ns No significance Pearson linear correlation between UTCI and the climatic variables (T mrt, Ta , Va , RH and Tg ) for each site and different period of the day during Scenario 3 (2024.08.24). Note: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001, ns No significance In S1, corresponding to the grassland scenario on a hot day, the UTCI was mainly associated with Ta , Tg , and RH during the daytime. Ta and Tg were strongly positively correlated with UTCI, whereas RH showed strong negative correlations. Compared with the other scenarios, the most distinct feature of S1 was the unstable role of T mrt and Va . During midday/afternoon, T mrt was negatively correlated with UTCI in the DSE group (r = ‐0.8237), but positively correlated with PV canopy shading and fabric shading. Va also showed weak or non‐significant relationships in some shaded cases, especially under fabric shading during the midday/afternoon (r = ‐0.3178, p > 0.05). At night, several correlations became weaker or non‐significant, indicating that the UTCI‐driving factors were less stable under relatively milder hot‐day conditions. In S2, corresponding to the grassland scenario on an extreme heat day, the correlation structure became stronger and more consistent. Tg showed very strong positive correlations with UTCI in almost all periods and groups, whereas Ta and T mrt also maintained high positive correlations. RH was consistently negatively correlated with UTCI, indicating a stable inverse relationship between humidity and heat stress under extreme heat. Compared with S1, the roles of the radiant and thermal variables became more prominent and persistent. In contrast, Va remained the most unstable variable. This was non‐significant in several cases and even showed a strong negative correlation under fabric shading at night (r = −0.8195), suggesting that airflow effects were more dependent on the local shading configuration than on the overall heat level. In S3, corresponding to the high‐albedo surface scenario on an extreme heat day, UTCI showed a more radiation‐dominated correlation pattern. Tg and T mrt were strongly positively correlated with UTCI, whereas RH remained strongly negative. During midday/afternoon, the correlation between Ta and UTCI was weaker than in S1 and S2, especially in the DSE group (r = 0.6443), whereas Tg remained highly correlated with UTCI (r = 0.9775). This indicates that, over the high‐albedo surface, radiant heat exposure played a more dominant role than air temperature in determining UTCI. Therefore, compared with the grassland scenarios, S3 highlighted the enhanced influence of surface‐reflected radiation on outdoor thermal stress in the summer. These findings are consistent with previous studies on outdoor thermal comfort and urban heat mitigation. Street‐scale interventions, such as building setbacks and roadside tree planting, have been shown to modify pedestrian‐level thermal exposure [22], whereas outdoor and semi‐outdoor thermal comfort in hot climates is strongly affected by the combined effects of radiation, air temperature, humidity, and airflow [23, 24]. Previous field studies have demonstrated that fabric shade structures can improve summer thermal comfort by reducing solar exposure [25], whereas photovoltaic and reflective shade surfaces may produce different sensible heat fluxes and pedestrian thermal responses because of their material properties [26]. Vegetation‐based shading has also been widely reported as an effective strategy for mitigating daytime urban heat stress [27]. From the perspective of thermal comfort assessment, Tmrt remains one of the most critical and challenging variables because it integrates complex shortwave and longwave radiation exchanges [28]. Therefore, combining field measurements with integrated urban climate assessment tools can further improve the evaluation of radiation‐driven thermal environments [29]. These studies support the present finding that the cooling performance of shading stations should be interpreted not only by air temperature reduction, but also by radiation control, airflow modification, material heating, and underlying‐surface conditions. Conclusions This study conducted a field‐based comparison of PV canopy and fabric shading under different weather and underlying surface scenarios. The main conclusions are as follows: The heat mitigation effect of artificial shading was mainly concentrated during the daytime and was more evident in radiation‐related indicators than in air temperature. During daytime, PV canopy shading and fabric shading reduced mean Ta by 1.0°C and 1.2°C, respectively, whereas the corresponding reductions in Tg reached 3.5°C and 4.3°C. The differences between the shaded and unshaded groups became much weaker at night, with PV canopy shading showing almost no nighttime effect on Ta and fabric shading producing only a slight warming effect of 0.2°C. The vertical profiles also showed that Ta differences among different heights under the shaded stations were generally limited, whereas Tg exhibited a more evident vertical difference between 1.10 m and 2.25 m. These results indicate that the main contribution of shading is to reduce daytime radiant heat exposure rather than substantially alter the air temperature field. PV canopy and fabric shading improved the thermal environment through different mechanisms. During the full‐shading period, Tg decreased from 47.2°C under the unshaded condition to 40.6°C under PV canopy shading and 39.4°C under fabric shading. Similarly, Ta decreased from 39.6°C to 36.7°C and 36.9°C, respectively. Both shading stations substantially reduced the total radiation intensity, from 1865 W/m2 under direct sunlight to 1346 W/m2 under PV canopy shading and 1341 W/m2 under fabric shading. However, the PV canopy had a higher surface temperature than the fabric canopy and produced a larger increase in downward‐longwave radiation. As a result, fabric shading showed stronger daytime radiant‐heat mitigation, with average reductions of 22.1°C in T mrt and 3.5°C in UTCI, compared with 17.5°C and 3.1°C under PV canopy shading. Therefore, fabric shading was more effective in reducing radiant heat exposure, while PV canopy shading had the additional potential advantage of maintaining airflow and providing energy‐related functions, but with a stronger canopy‐surface heating effect. The performance of shading facilities was strongly constrained by weather conditions and the underlying surface. In the grassland scenarios, both shading stations reduced daytime ground surface temperature. PV canopy shading and fabric shading reduced mean ST by 2.8°C and 1.4°C in S1, and by 1.2°C and 0.4°C in S2, respectively. However, over the high‐albedo surface in S3, ST increased by 1.8°C under PV canopy shading and 0.9°C under fabric shading. This indicates that the thermal response of the ground surface under shading was not uniform but depended on the surface material and background heat intensity. The correlation analysis further showed that, under the high‐albedo surface scenario, UTCI was strongly associated with radiant thermal variables such as Tg and T mrt, suggesting that reflected and accumulated radiant heat played a dominant role in influencing outdoor heat stress. Both shading stations reduced the duration and intensity of thermal stress but did not completely eliminate the heat risk under extreme heat conditions. During the daytime, the unshaded group was exposed to extreme heat stress for 38.3% of the period, whereas this proportion decreased to 2.7% under both PV canopy and fabric shading. Over the entire day, the proportion of extreme heat stress decreased from 21.9% under unshaded conditions to 1.5% under both shaded conditions. However, very strong heat stress still occurred for a considerable proportion of the daytime under shaded conditions, indicating that shading mainly reduced the highest heat‐stress category rather than fully removing thermal stress from the environment. Correlation analysis showed that UTCI was mainly governed by Ta , Tg , T mrt, and RH, whereas the role of Va was weaker and more variable among the different types of shade and time periods. Overall, the findings suggest that artificial shading should be selected based on local environmental conditions and design objectives. Fabric shading is more suitable when the primary goal is radiant heat mitigation, while PV canopy shading is more appropriate for open and well‐ventilated spaces where solar‐energy utilization and airflow maintenance are also considered. In hot outdoor spaces, shading design should be combined with ground surface material control and ventilation planning to achieve more effective heat mitigation. Nomenclature Direct solar exposure Downward longwave radiation Upward longwave radiation from the ground Outdoor thermal comfort Photovoltaic Relative humidity Downward shortwave radiation Upward shortwave radiation Surface temperature at the ground level Black globe temperature Mean radiant temperature Surface temperature of shading canopy Universal Thermal Climate Index Urban heat island Wind speed Author Contributions Liyue Jin: Writing the original draft, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Xin Dong: Methodology, Investigation, Formal analysis. Deo Prasad: Writing, review & editing, Investigation, Formal analysis, Supervision. Bao‐Jie He: Writing, review & editing, Investigation, Formal analysis, Supervision, Conceptualization. Conflicts of Interest The authors declare no conflicts of interest. Acknowledgements This study was supported by the National Natural Science Foundation of China (No. W2422003, 42301339). Open access publishing facilitated by The University of Queensland, as part of the Wiley ‐ The University of Queensland agreement via the Council of Australasian University Librarians. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. References Associated Data Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.