Abstract
Background Thermal comfort significantly influences well-being, productivity, and living conditions in outdoor environments, particularly in rapidly urbanizing, warm, humid tropical climates. This study assessed the influence of 5 five common urban tree species (Cassia fistula, Tectona grandis, Plumeria obtusa, Mangifera indica, and Terminalia catappa) on outdoor thermal comfort, using the physiological equivalent temperature (PET) index in Colombo, Sri Lanka, as a case study for a tropical humid city.
Methods Field data collection encompassed measuring air and surface temperature, relative humidity, wind velocity, solar radiation, cloud cover, and sky view factor under tree canopies and adjacent exposed areas. The RayMan model was employed to estimate PET in both areas.
Results Our findings indicated that PET was consistently higher in exposed areas compared to under the tree canopy, with an average difference of 5.61 °C. Among tree parameters, sky view factor (SVF) demonstrated the most significant correlation with thermal comfort, followed by crown diameter and tree height. Furthermore, notable variations in thermal comfort were observed among tree species, with Terminalia catappa outperforming Plumeria obtusa, particularly on sunny days.
Conclusion Regression analysis highlighted the importance of integrating trees with large crowns and low SVF to create thermally comfortable outdoor spaces. Consequently, Terminalia catappa emerged as the most suitable tree species for enhancing thermal comfort in Colombo’s outdoor urban areas out of the 5 selected species. These insights will aid in selecting appropriate tree species and parameters, fostering improved outdoor thermal comfort in tropical humid cities, and facilitating sustainable urban planning and design strategies.
Introduction
A crucial step in identifying whether a particular outdoor space is suitable for livability is determining the outdoor thermal comfort. Furthermore, it is impossible to create a suitable thermal environment in urban open areas without proper determination and comprehension of human thermal comfort (Lin et al. 2010). Determining outdoor thermal comfort is therefore crucial. Due to the elevated intricacy inherent in outdoor settings when juxtaposed with indoor environments, the exploration of thermal comfort within open spaces has been relatively limited, with particular emphasis on tropical climatic conditions (Nasrollah et al. 2020). Thus, there is a lack of an exact thermal comfort index that is appropriate to assess outdoor thermal comfort due to the complexity of the urban outdoor environment and outdoor thermal comfort parameters (Johansson and Emmanuel 2006). However, physiologically equivalent temperature (PET) is considered one of the most popular and suitable outdoor thermal indices used by researchers. PET empowers individuals to juxtapose the comprehensive repercussions of intricate thermal circumstances in outdoor settings with their personal indoor encounters (Hoppe 1999). PET stands as a widely embraced bioclimatic indicator due to its assimilation of a universally recognized unit (°C) as the quantification of thermal strain. Consequently, it proves readily accessible for interpretation and comprehension, even among users unacquainted with contemporary human biometeorological lexicon, encompassing professionals such as urban designers and landscape architects (Fang et al. 2017).
Creating outdoor spaces that are thermally comfortable serves not only to enhance residents’ satisfaction but also encourages their interaction with the surrounding environment (Lin et al. 2012). A thorough understanding of human thermal comfort outdoors and the complex interaction between environmental factors and tree-related parameters in mitigating thermal stress is essential for developing urban environments that are thermally pleasant (Lai et al. 2020). This knowledge is vital for designing urban spaces that prioritize comfort and well-being, while also integrating harmoniously with the natural surroundings.
Background
Augmenting the extent of green spaces within urban ecosystems stands as a highly efficacious strategy for ameliorating the adversities of heat stress (Depietri et al. 2012). Among the constituents of these green areas, plants assume a paramount role. A proliferation in the population of trees engenders amplified shading, thereby fostering enhanced thermal comfort during daylight hours. Significantly, the shading attributes of trees have been identified as a pivotal factor significantly impacting thermal comfort (Lindberg and Grimmond 2011; Nasrollahi et al. 2020; Lin et al. 2022). This elucidates the substantial influence that trees exert in alleviating heat-related discomfort and underscores their indispensable contribution to the creation of more amenable urban microclimates.
Role of Urban Trees in Improving Outdoor Thermal Comfort
Trees play a pivotal role in shaping the thermal environment through multifaceted mechanisms, encompassing shading, the moderation of surface temperatures, radiation interception via their canopies, and convective heat transfer from warmer locales. Moreover, trees contribute to cooling through the conversion of radiation into latent heat and subsequent evapotranspiration processes (Santamouris et al. 2014). Consequently, augmenting urban vegetation can effectively lead to a reduction in both air and surface temperatures.
In addition to the cooling effect driven by increased evapotranspiration, trees exert influence by blocking solar radiation and diminishing mean radiant temperatures within urban street canyons. A spectrum of studies has been conducted to investigate the impact of vegetation on thermal comfort, encompassing analysis ranging from individual trees to expansive urban parks (Taleghani et al. 2014). Utilizing vegetation, especially in urban spaces such as parks, offers the potential to enhance overall thermal conditions. This can be achieved through the concurrent reduction of air temperature and mean radiant temperature, alongside an increase in surrounding humidity (Weerakoon and Perera 2021). The strategic incorporation of vegetation thus emerges as a promising avenue for ameliorating the urban microclimate and fostering a more comfortable and sustainable urban environment.
Previous research suggests that a higher degree of crown diameter could cause a greater cooling sensation in relation to thermal comfort (Lin et al. 2021). Lin and Tsai (2017) indicated a strong correlation between sky view factor (SVF), average crown diameter, and PET. SVF decreased as the average crown diameter increased. Further, there is a significant correlation between SVF and PET only during sunlight and sunset (Karimi et al. 2020). The findings suggested that if the average crown diameter of tree is wider than 1.5 m, the cooling effect was especially dominated by the tree species. According to Wang et al. (2021), the increase of R/TH by 1 can reduce physiological equivalent temperature by approximately 2.5 °C (R2 = 0.8932) and 2.65 °C (R2 = 0.8424) in the morning and afternoon respectively. Trees with high trunks and wide, overlapping crowns were found to be the most suitable forms of greenery to reduce PET levels during the day (Narimani et al. 2022).
A detailed exploration into the contrast between thermal comfort levels in shaded and unshaded spaces was conducted by Lin et al. (2010). Through 12 field tests at a university campus in central Taiwan characterized by a tropical climate, the study employed RayMan software to compute the PET index, consequently evaluating the campus’ thermal conditions. The study concluded that trees and buildings could collaboratively create a thermally comfortable microclimate, proving effective during both scorching summers and mild winters in Taiwan.
Physiological Equivalent Temperature (PET)
The physiological equivalent temperature (PET) stands as a prominently employed thermal index, encompassing the entire array of meteorological parameters that impact thermal comfort, alongside considerations of physical activity and attire worn by individuals (Javan and Nasiri 2016). PET represents the equivalent temperature in a given location (indoors or outdoors), simulating the air temperature within a standard indoor setting for a human possessing core and skin temperatures analogous to those relevant to the assessed conditions. This entails maintaining the heat equilibrium of the human body under a metabolic rate of 80.0 W (light activity, supplementing basal metabolism) and a clothing heat resistance of 0.9 clo. The foundation of PET rests upon the Munich Energy-balance Model for Individuals (MEMI), a thermo-physiological heat balance model encompassing all fundamental thermoregulatory processes (Hoppe 1999; Elnabawi and Hamza 2020).
While universal thermal climate index (UTCI) and mean radiant temperature (MRT) are available to measure thermal comfort, the present study selected PET to measure thermal comfort due to its emphasis on human physiological responses. PET incorporates factors such as metabolic rate, clothing insulation, and body surface area, allowing for a more comprehensive assessment of the thermal environment as it relates to human comfort perception (Mayer and Höppe 1987). In contrast, indices like MRT primarily focus on environmental conditions without directly accounting for human physiological responses (Blazejczyk et al. 2012). Additionally, even though UTCI focuses on some physiological aspects, it primarily focuses on environmental conditions like MRT, air temperature, humidity, and wind speed. This can be less sensitive to individual variations in heat tolerance (Pompei et al. 2024). As this study aimed to understand the impact of urban trees on human thermal comfort, PET’s focus on human physiology made it the most suitable parameter for our analysis.
Research Aims
In the realm of urban planning, the strategic planting of street trees assumes paramount significance as a practical approach to upholding human thermal comfort. However, the cooling effect imparted by trees is contingent upon factors such as climate, tree species, and environmental conditions, introducing variability (Rahman et al. 2015; Moss et al. 2019). Consequently, a quantitative evaluation of the thermal comfort augmentation attributed to distinct tree species, accounting for shading and other relevant factors, becomes imperative. Such an assessment offers insights into the influence of tree species on the immediate environment and, consequently, on outdoor thermal comfort across diverse climatic zones (Rahman et al. 2020).
Understanding the intricate interplay between tree parameters, tree species, and outdoor thermal comfort in hot and humid climates is imperative for the development of climate-conscious urban design guidelines. The present study endeavors to elucidate the influence of factors like tree height, crown diameter, sky view factor (SVF), tree geometry, and diameter at breast height (DBH) on thermal comfort. Additionally, the study seeks to investigate the impact of 5 prevalent urban tree species—Cassia fistula (ehela), Tectona grandis (teak), Plumeria obtusa (Araliya), Mangifera indica (mango), and Terminalia catappa (Kottamba)—on outdoor thermal comfort. The PET serves as the primary thermal index for evaluating outdoor thermal comfort in this study.
Materials and Methods
Tree and environmental parameters affecting outdoor thermal comfort and the existing research gaps were identified based on the extensive literature review. Thus, a systematic methodology was developed to collect environmental measurements and tree parameters related to thermal comfort through physical data collection to assess the impacts of tree species on outdoor thermal comfort.
Study Area
The selected study area for the research is Colombo, Sri Lanka (6.9271° N, 79.8612° E). It is the commercial capital and largest city by population of Sri Lanka. Colombo is the financial center of the island and a tourist destination. The study area has mixed land uses, parks, high-density urban areas, high-rise building areas, middle-rise areas, and residential buildings, and these are present in an irregular distribution. Accordingly, considering the natural morphology in the area, 6 sites were selected. Figure 1 shows the boundary of the selected area for the study.
Study area.
Selection of Sites and Tree Species
Initially, Colombo 1 to 15 areas were considered as the sites for the preliminary data collection. Thereafter, after visiting each site, 6 study areas were selected based on the green cover percentage of the 47 wards in Colombo 1 to 15. Bloemendhal (Colombo 13) and Fort (Colombo 1) areas were selected to represent the low green cover areas. Bambalapitya (Colombo 4) and Wekanda (Colombo 2) areas were selected to represent the moderate green cover areas whereas Cinnamon Gardens (Colombo 7) and Thimbirigasyaya (Colombo 5) areas were selected to represent the high green cover areas. Table S1 shows the extent and the green-cover percentage of each selected site. The 6 sites were chosen to represent different green cover percentages, ranging from low to high, to neutralize the impact of green cover on thermal comfort.
Table S2 shows the specific roads and areas that were selected to proceed with field data collection from the selected 6 sites. The main objective of the present study is to assess the impact of 5 selected common urban tree species for outdoor thermal comfort improvement in Colombo City. The local climatic conditions may highly influence thermal comfort conditions. Therefore, the sites were carefully selected by considering the roads and areas that exhibit similar microclimatic conditions and ambient conditions.
Selection of Trees
The 6 sites and the suitable tree species for the study were selected simultaneously. During the initial field visit, the most abundant and common tree species in Colombo 1 to 15 were identified. The identified tree species were Kottamba (T. catappa), mango (M. indica), ehela (C. fistula), jak (Artocarpus heterophyllus), teak (T. grandis), Araliya (P. obtuse), na (Mesua ferrea), del (Artocarpus nobilis), Mee (Madhuca longifolia), kohomba (Azadirachta indica), wal ehela (Pterocarpus indicus), Pare mara (Albizia saman F. Muell.), Nuga (Ficus benghalensis), and Pihimbiya (Filicium decipiens).
When selecting the most suitable trees, the following factors were also considered: (1) trees with overlapping crowns and trees with trunks connected to walls should be disregarded; (2) trees near a high-rise building or high wall or water body within a radius of 3 m should be excluded; and (3) trees in an extreme microclimatic condition such as coastal areas, urban parks, etc. were excluded. Moreover, the presence of all the selected tree species in the 6 selected sites was a must. Thereby, considering all these factors, Kottamba, mango, ehela, teak, and Araliya were selected as the most suitable tree species out of the observed most common species (Appendix). Table 1 shows general details of the selected 5 species.
Details of the selected urban tree species.
Experimental Design
RayMan Model
This study utilized the RayMan model (Freiburg im Breisgau, Germany), to conduct estimations of outdoor thermal comfort. Developed by the Chair for Environmental Meteorology, formerly the Chair for Meteorology and Climatology, at Albert-Ludwigs-University Freiburg, RayMan is a micro-scale model devised to compute radiation fluxes within both uncomplicated and intricate environmental contexts. Within its framework, the model computes the mean radiant temperature (Tmrt), a pivotal input parameter in the determination of thermal biometeorological indices such as the PET. Several validation studies prove that RayMan model has a good accuracy in calculating mean radiant temperature and PET (e.g., Matzarakis et al. 2007; Hwang et al. 2011). RayMan operates as a one-dimensional spatial model, signifying that all calculations are confined to a single point (Matzarakis et al. 2007; Matzarakis et al. 2010). In this study, the RayMan model was harnessed to calculate the outdoor thermal comfort index, namely the PET. This approach allowed for a comprehensive assessment of the thermal comfort conditions within the context of the specific urban environment under scrutiny.
Identification of the Parameters Affecting Thermal Comfort
Parameters that affect outdoor thermal comfort were identified through literature and referred to the parameters needed to calculate the thermal comfort by the RayMan model. The model needed only a limited number of common parameters as inputs which are discussed in detail in the forthcoming sections.
Data Collection
Various studies on thermal comfort determination have adopted distinct field measurement timeframes aligned with their specific research objectives. For instance, Tang and Ng (2016) conducted measurements between 12:30 PM and 02:00 PM, while Krüger et al. (2014) gathered data from 10:00 AM to 01:00 PM. In the present study, data were collected within the timeframe of 11:00 AM to 03:00 PM, as assessing thermal comfort requires understanding how meteorological parameter fluctuations impact occupants, especially during peak heating hours (ANSI/ASHRAE 2017). The data collection spanned from August 2021 to March 2022 and encompassed both sunny and overcast conditions. This selected time duration covered the main four climatic conditions of Colombo City. Each weather condition was subjected to 3 replicates per tree, ensuring robustness and reliability in the observations. Furthermore, data were obtained both beneath the tree canopy and in a nearby exposed area for selected trees, enabling a comparative assessment of thermal comfort disparities. Figure 2 visualizes the research framework of the current study.
Research framework.
The data collection process entailed measurements of air temperature, relative humidity, surface temperature, wind speed, and solar radiation at a 10-second sampling interval. The collected data points were subsequently averaged for analysis. The mean air temperature of Colombo was recorded as 28 to 29 °C (Jeyaraj 2022). To account for diverse material influences, surface temperature measurements encompassed pavements, green covers, and walls. Cloud cover was ascertained through visual inspection, rated out of 8 oktas, and considered overcast if cloud amounts reached 6 oktas or higher, aligning with Tang and Ng’s criteria (2016). Hourly average variations of PET and other meteorological parameters observed among tree species are depicted in Figure 3.
Hourly average variations of PET and other meteorological parameters observed among tree species.
Fisheye images were captured at a height of 1.5 meters above the root collar beneath the tree and at the same height above the ground in the nearby exposed area. These images were oriented with the top pointing north and the right side facing east. Essential location parameters such as longitude, latitude, and altitude were collected to serve as inputs for the RayMan model. Throughout the data collection period, personal observer data—age, sex, height, and weight—remained constant, serving as reference metrics for comparative purposes.
For consistent modeling, activity levels were standardized at 80.0 W, and clothing levels were maintained at 0.9 clo—both prerequisites for PET calculations using the RayMan model. To evaluate the impact of tree parameters on outdoor thermal comfort, measurements of tree height, diameter at breast height (DBH), and crown diameter were conducted. Detailed information regarding the instrumentation employed for data collection, along with their respective specifications, is summarized in Table 2.
Specifications of the instruments used for the data collection. DBH (diameter at breast height).
Determination of Crown Diameter, Tree Geometry, DBH, and SVF
Crown diameter of the trees can be taken by measuring the shortest and the longest lengths of each crown by the distance tape. Tree geometry is constructed by considering both tree height and tree diameter; it is the ratio of tree crown radius to tree height 
, and it has a similar significance with ratio of crown diameter to tree trunk height 
, which is known as gross canopy index (GCI)(Zhang et al. 2020). The equations applied are as follows:
Diameter at breast height (DBH) is defined as the tree diameter measured at 1.37 m above the ground line on the uphill side of the tree. Sky view factor (SVF) refers to the proportion of visible sky (ranging from 0 to 1) in a hemispherical image (Zeng et al. 2018). This study calculated SVF using RayMan model.
Data Analysis
The PAleontological STatistics (PAST) tool (Hammer et al. 2001) was used to perform principal component analysis (PCA) to derive the relationship between PET with environmental parameters and tree parameters. R was used to perform k-means cluster analysis to categorize trees into clusters based on their similarities in terms of tree parameters and thermal comfort (R Core Team 2022).
Results
Data collection was carried out from August 2021 to March 2022. The environmental parameters were measured under the selected 5 trees—Araliya, ehela, mango, Kottamba, and teak—and in an adjacent site of the trees (stated as near exposed area herein). According to past literature, both sunny and overcast weather conditions were considered when collecting data as it was identified as a critical factor in outdoor thermal comfort studies (Tan and Ng 2016). Six replicate measurements were carried out for each location changing the date, time, and weather conditions. Altogether, 420 data points were collected under the trees, and 420 data points were collected in a near-exposed site of each tree. Results obtained as shown in Table 3 indicates that the average PET beneath a tree is 5.61 °C lower than the average PET in nearby exposed areas, indicating that the thermal comfort under the tree was higher than in nearby exposed areas, highlighting the need to examine the impact of different tree species and their characteristics on outdoor thermal comfort.
Summary physiological equivalent temperature (PET) values of tree species.
The observed results related to PET on both sunny and overcast days are demonstrated in Table S4. The results indicate that under each tree species, the mean PET is 4.52 °C lower on overcast days whilst the mean PET is 4.05 °C lower in the near-exposed area on overcast days. Figure S1 indicates the summary distribution of different climatic parameters under the canopy and near exposed areas for the selected tree species.
Summary Statistics of Tree Parameters
The highest mean tree height was recorded for teak whereas the lowest was recorded for Araliya. The DBH of Araliya and ehela was comparatively lower compared to the other 3 tree species. Moreover, the highest crown diameter was observed in Kottamba whilst the lowest was observed in Araliya. Contrastingly, the highest SVF was observed in Araliya, and the lowest was observed in Kottamba. The violin and boxplot graphs for each tree parameter are shown in in Figure 4.
Violin and boxplot graphs for tree parameters.
Results of the Regression Analysis
Regression analysis was conducted with the use of tree parameters to find out how one variable affects the other variable. Hence, regression analysis was carried out considering PET as the dependent variable and tree height, tree crown diameter, tree geometry, SVF, and DBH as predictor variables. The results are given in Tables S5 to S7.
In combination, tree height, tree crown diameter, tree geometry, SVF, and DBH accounted for a significant 26.6% of the variability in PET, R2 = 0.266, adjusted R2 = 0.208, F (5, 69) = 4.635, P = 0.001. Therefore, a mid-level and significant relation with PET has come out when all the tree parameters are considered together; these explain 26.6% of total variance of PET. When t-test results regarding the significance of the regression coefficients are examined, it is found that only crown diameter and SVF are significant predictors of PET as only these 2 tree parameters have a significance less than 0.05 P-value. Based on the regression equation, it could be observed that within the parameters explored in this study, if PET needs to be reduced, measures such as increasing the plantation of urban trees with lower SVF and increasing the crown diameter could be identified as the variables which significantly affect the thermal comfort according to the multiple linear regression analysis.
Results of the PCA Analysis
The PCA was conducted based on the correlation matrix. The percent of variance (Appendix) explains how much variance within the construct is accounted for by that factor. According to the PCA, an eigenvalue > 1.0 is commonly used as a cut-off point to decide which principal components (PC) should be retained. Thus, in this analysis, only PC1 and PC2 should be considered as only they have eigenvalues higher than 1.0. This suggests that about 88.307% of the information (variances) contained in the data are retained by the first 2 principal components. This is an acceptably significant percentage. As decided by the eigenvalue, only PC1 and PC2 are considered. PC1 axis is the first principal direction along the samples which shows the largest variation. The PC2 axis is the second most important direction, and it is orthogonal to the PC1 axis (Figure 5).
Scatter plot of PCA.
According to the scatter plot, air temperature and surface temperature are positively correlated with PET as these 3 variables are grouped together. Furthermore, solar radiation and SVF are also positively associated with PET. However, the association with PET is not as strong as air temperature and surface temperature. The analysis has identified that PET and tree geometry is independent from one another. Tree height, DBH, crown diameter, relative humidity, and wind velocity are negatively associated with PET as these variables are positioned on opposite sides of the plot origin (opposed quadrants) compared to where PET is positioned. Moreover, the analysis suggests that wind velocity and relative humidity positively interact with each other whilst tree height and DBH positively associate with one another. The analysis also predicts that tree height, crown diameter, and DBH have a negative relationship with SVF and tree geometry.
Cluster Analysis
Clustering techniques are employed to identify subsets of data within a dataset. In this study, k-means cluster analysis was utilized to group similar observations and distinguish dissimilar ones. The goal was to explore the relationship between PET and tree parameters without treating PET as the dependent variable. The k-means method efficiently divides a dataset into k distinct groups, yielding meaningful clusters. Figure 6 shows the clusters derived from this analysis, shedding light on the interplay between tree parameters and PET, thereby enhancing the understanding of outdoor thermal comfort dynamics.
Cluster analysis.
Clustering helps to identify which observations are alike, and potentially categorize them therein. Cluster 1 (shown in red) consists of observations of Kottamba, teak, and mango whereas cluster 2 (shown in blue) consists of observations of Araliya, ehela, and mango. Hence, it can be concluded that in terms of tree parameters and thermal comfort, Araliya and ehela are alike but different compared to Kottamba and teak. However, mango is positioned under both clusters where 66.66% of the observations of mango trees are included under cluster 1; 33.33% of the observations of mango trees are clustered under the cluster 2. The results obtained by the cluster analysis could be used to explain the obtained thermal comfort difference between Araliya and Kottamba as both the trees are positioned under different clusters.
According to the cluster analysis, mango was positioned under both clusters which suggests that the mango tree possesses similarities with the other 4 trees in terms of tree parameter and thermal comfort. Out of 15 mango trees, 10 were included in the cluster 1 with Kottamba and teak while 5 mango trees were clustered with ehela and Araliya in cluster 2. If mean tree parameter values are evaluated, the average tree height of mango is 8.74 m which is less than teak and Kottamba but higher than ehela and Araliya. If the trees are ranked considering the tree geometry, mango is ranked in the middle position with a value of 3.17. Thus, in terms of tree height and tree geometry, mango trees can be categorized under both clusters. When SVF is considered, the average SVF of mango belongs to the range of ehela and Araliya, while the average DBH and crown diameter is in the range of teak and Kottamba. Moreover, obtained PET order is Araliya > teak > ehela > mango > Kottamba which indeed shows that mango is positioned between teak and Kottamba in terms of thermal comfort. Hence, considering these similarities of mango with other tree species, it is possible to comprehend the reason behind clustering mango tree under both clusters.
Thermal comfort difference between Araliya and Kottamba was significant under sunny conditions. The thermal difference under Araliya and Kottamba may result due to the differences of environmental parameters and differences of tree parameters of the 2 species. The solar radiation under the Kottamba and Araliya trees are significantly different with the solar radiation under the Araliya 2 times higher than the Kottamba. Moreover, the mean SVF of the Araliya is 0.192 whereas the mean SVF of the Kottamba tree is 0.138; the surface temperature under the Araliya is 2.84 °C higher than Kottamba. As surface temperature, solar radiation, and SVF are significant parameters which determine thermal comfort, it is evident that these 3 factors can act as the main predictors which resulted in a thermal comfort difference among Araliya and Kottamba in terms of PET. Furthermore, this can be also seen under the cluster analysis as both Kottamba and Araliya tree species are positioned under separate clusters in terms of their tree parameters and thermal comfort.
DISCUSSION
In the context of thermal comfort, higher PET values often signify discomfort or unacceptability (Yahia and Johansson 2014), while PET values ranging from 27.0 °C to 33.0 °C are deemed thermally comfortable for locales like Sri Lanka (Johansson and Emmanuel 2006). The study findings revealed an average PET of 35.05 °C under tree shade and 41.27 °C in unshaded areas. Comparative studies, such as those by Yang et al. (2013) and Zhao and Fong (2017), established the acceptable PET range for Singapore as 26 °C to 31.7 °C. Similarly, Morakinyo et al. (2017) identified the acceptable PET range for Hong Kong as 25 °C to 29 °C, while Ahamed (2003) outlined a range of 28.5 °C to 32.8 °C for Bangladesh. These tropical climates exhibit acceptable PET ranges below 33 °C. In contrast, our study observed slightly higher PET values under shaded conditions and considerably higher values in unshaded areas compared to the acceptable PET range proposed by earlier research. Dissimilarities in PET calculation and thermal comfort assessment likely contribute to this variance. Factors like sweat rate, skin temperature (Lai et al. 2020), and psychological factors such as experience and expectations (Potchter et al. 2018; Chen et al. 2018) influence individual tolerance and perception of thermal comfort. These multifaceted influences render PET estimation uncertain and reliant on external factors, rendering the established comfort zones less directly transferable to the Colombo context.
This comprehensive examination contributes to understanding the intricate relationships shaping outdoor thermal comfort dynamics. The impact of various factors on thermal comfort emerges as a complex interplay, influenced by both climatic and physiological considerations. As our study takes into account specific local conditions and nuances, it contributes valuable insights to the discourse on thermal comfort optimization in urban settings. Moreover, the PET values under the trees of Araliya, ehela, mango, Kottamba, and teak is lower by 5.51 °C, 5.49 °C, 5.61 °C, 5.84 °C, 5.61 °C respectively. In comparison with other tree species, Kottamba possesses the lowest SVF and the highest crown diameter. Thus, it indicates that SVF and crown diameter of Kottamba is favourable to provide better thermal comfort than other tree species. Therefore, this explains the reason for concluding Kottamba as the most suitable tree species to elevate thermal comfort.
The present study, while yielding valuable insights, also presents certain limitations that warrant consideration. Notably, research has indicated that the cooling effects of trees are particularly pronounced during the evening hours (Zhang et al. 2020). The study’s focus on daytime microclimatic conditions, rather than encompassing the full day, underscores the need for more extensive investigations in the future, potentially addressing this limitation.
Leaf area index (LAI) stands as another pivotal tree attribute with implications for thermal comfort. The positive relationship between LAI and solar radiation and air temperature is noteworthy. Lower LAI values hinder the penetration of shortwave radiation, subsequently resulting in lower air and surface temperatures and, consequently, heightened thermal comfort (Rahman et al. 2020). To comprehensively understand the interplay of tree parameters and their impact on thermal comfort, future research avenues should delve into parameters like transpiration, LAI, leaf layer index (LLI), crown volume index (CVI), total canopy index (TCI), and cover plant ratio (CPR). Exploring variables such as slope of the trunk (ST), spacing (SP), and internodal length (IL) in outdoor thermal comfort studies also holds potential. Furthermore, the benefits of mixed vegetation, as demonstrated by Zhao and Fong (2017) in terms of cooler undersides, provide an intriguing avenue for enhancing thermal comfort in tropical urban settings like Colombo. By investigating the potential of mixed vegetation, future research could contribute to refining strategies for improving outdoor thermal comfort.
Conclusion
When all the measured tree variables are taken into consideration, it is established that thermal comfort is most positively impacted by the SVF of the trees followed by crown diameter and tree height respectively. There was a negligible association observed between the DBH, tree geometry, and thermal comfort. The research has inferred that the PET increases when the SVF increases. Similarly, when the crown diameter and tree height of a tree increases the PET decreases. Through the multiple linear regression model, the tree height has not been identified as a significant variable to determine PET when all the tree parameters are taken into consideration. This outcome insinuates that urban planners should take the necessary measures to reduce PET by considering trees with high crown diameter and low SVF for their urban planning development strategies.
It was noted that the average PET is higher in exposed areas, and in the areas beneath a tree, average PET is 5.61 °C lower than the exposed areas. Further, it was revealed that from the 5 tree species which were selected for this research Kottamba tree provides highest degree of thermal comfort while mango, ehela, teak, and Araliya provides thermal comfort levels in the given order.
The control site is influenced by nearby buildings and trees, man-made structures, surface material variations, potentially affecting its solar radiation, surface temperature, and wind conditions. Nevertheless, it remains the most practical reference point available. Future studies should address these limitations by conducting a comprehensive characterization of the control site, and quantifying shading caused by the building and tree using appropriate instruments. These measures aim to enhance the accuracy and reliability of future research findings.
The current investigation yields valuable insights into the pivotal role of tree species within diverse urban environments and emphasizes the necessity of leveraging trees to effectively enhance thermal comfort. Additionally, a comprehensive elucidation of the interrelation between environmental and tree parameters with thermal comfort has been delineated. Consequently, the study serves to encourage and provide a robust framework for engineers and urban planners to incorporate environmental and tree parameters into the design and operational processes of sustainable city planning, with a specific focus on regulating outdoor thermal comfort, especially within the context of Colombo. As climate change continues at an alarming rate, this endeavor becomes increasingly critical, particularly in tropical countries. As such, this study lays the groundwork for future investigations to explore the impact of diverse tree parameters and species on achieving optimal outdoor thermal comfort.
Conflicts of Interest
The authors reported no conflicts of interest.
Appendix
Boxplot of distribution of air temperature (Ta), relative humidity (RH), wind velocity, and surface temperature (Surface T) under the tree and near exposed area.
Details of the selected sites (Wickramasinghe et al. 2016).
Selected specific roads and areas from the 6 sites.
Total number of trees in each site.
Descriptive comparison of physiological equivalent temperature (PET) between sunny and overcast conditions. No. (number of trees); SE mean (standard error of the mean); StDev (standard deviation); Min (minimum PET [°C]); Max (maximum PET [°C]).
Model summary of regression analysis. R (correlation coefficient); R2 (coefficient of determination).
Analysis of variance (ANOVA) table of the regression analysis. DF (degrees of freedom); F (F-statistics); Sig. (significance level).
Coefficient table of multiple regression analysis. Std. Error (standard error); t (t-value); sig (significance).
Summary of principal component analysis (PCA) analysis. PC (principal component).
- © 2024 International Society of Arboriculture
Literature Cited
Jump to section
Related Articles
Cited By...
- No citing articles found.