Effect of street trees shade on perceived thermal comfort in a south temperate climate: The sidewalks of Montevideo (Uruguay)

Abstract

The use of vegetation in cities is one of the most promising strategies for urban climate change adaptation and mitigation. Tree shade influences heat storage from surfaces reducing long wave radiation emission which directly affects people. People 's heat perception depends more on insolation and the temperature of surrounding objects than on air temperature itself. There is a need for analyzes that include the combined effects of physical and human variables on thermal comfort, as well as location-based studies to address its climatic and social conditions. In order to compare the effect of the trees on microenvironmental temperature and perceived thermal comfort, we measured physical parameters and performed structured interviews on three downtown streets of Montevideo, Uruguay, which had sections with and without trees on four dates during the summer. Generally, people surveyed under both treatments stated they did not feel fully comfortable due to summer heat, but the proportion of people who stated feeling in thermal comfort under tree shade was more than double than the unshaded sections.

The seasonal ARIMA analysis supported that the tree shade reduced the microenvironmental temperature by its effect on radiant temperature. By using a statistical decision tree methodology that combines all the variables in the same analysis, we found a greater impact of physical variables than personal variables on people's thermal comfort and thermal preferences. We also identified gender as a significant variable that affects people's thermal preferences, where 46.4 % of females preferred a slightly colder environment.

1. Introduction

1.1. Street trees

The use of vegetation in cities is one of the most promising strategies for urban climate change adaptation and mitigation, in addition to urban space conditioning. Cities are responsible for 60–70 % of the anthropogenic greenhouse gasses emissions. The present climate change has already been determined by past emissions and climate system inertia [1,2]. The thermal consequences may have been underestimated: bioclimatic conditions will be more stressful and these changes will be greater than the ones related to air temperatures. This whole context leads humans to look for mitigation and adaptation strategies [3,4]. Buildings, concrete, and asphalt surfaces combined with lack of vegetation cover cause higher air and higher surface temperatures in urban centers than in their surrounding suburban or rural areas. Trees can intercept up to 85 % of the solar radiation, and thus, their shadow can reduce concrete temperature up to 19 °C [[5], [6], [7]]. Tree groups provide greater cooling effects than individual trees. Street alignment design must consider basic principles to enhance its benefits [8] and provide a shade continuum. For temperate and warm regions with defined seasons, large deciduous trees are recommended as they intercept solar radiation in summer but allow it to pass through in winter, which should be considered in the design of urban tree alignments. However, studies show that they can also slightly raise the nighttime temperature of urban canyons due to the capture of outgoing longwave radiation [9]. For shading purposes, trees should be located preferably along the sidewalk on the side of the pavement and at a distance that allows their crowns to generate a continuous canopy [8,10,11]. Trees reduce heat storage from exposed surfaces reducing long wave radiation emission which directly affects people. People's heat perception depends more on both insolation and the temperature of surrounding objects (mean radiant temperature) than on air temperature itself [7,11,12].

1.2. Human thermal comfort

Human thermal comfort is a condition of mind that expresses satisfaction with the thermal environment, assessed subjectively. Physiological and psychological variations among individuals make it difficult to find an environment that satisfies them all. Air temperature, radiant temperature, wind speed, and relative humidity are primary environmental factors that characterize the thermal environment and influence people's thermal sensation. Other factors like metabolic rate and clothing are considered physiological. Thermal comfort indices such as Physiological Equivalent Temperature (PET) can be determined from the physical parameters. The air temperature that receives the greatest acceptance of comfort usually lies between 20.0 °C and 25.8 °C depending on the type of climate evaluated [[13], [14], [15], [16], [17]]. Comfort level is related to thermal sensation, since a person who feels neither cold nor hot, would be thermally comfortable. Thermal sensation is a conscious perception, usually scored into categories such as: cold, cool, slightly cool, neutral, slightly warm, warm, and hot. Although the microclimatic and physiological parameters have a great influence on this perception, they only account for approximately 50 % of the variation found among objective and subjective comfort assessments. Variables related to psychological adaptation would explain the remaining variation, such as naturalness, expectations, experiences, exposure time, perceived control, and other environmental stimuli.

Existing comfort models are not fully developed to consider the effects of both environment- and human-based parameters for the assessment of outdoor thermal comfort performance [18]. Both are complementary and should be considered together in order to increase the use of outdoor spaces to reinforce social interaction [13,16,19,20].

Microclimatic parameters are commonly used to infer people's outdoor thermal comfort in open spaces. Studies that directly inquire about user's thermal perceptions and preferences in a subjective way are generally focused on public spaces with specific characteristics and purposes such as squares and parks, which are mainly used for recreational purposes. City center sidewalks are mostly functional sites to allow pedestrian traffic, where the possibilities of choosing different micro-environments to achieve thermal comfort are limited.

The specific aim of this study was to quantify and analyze the importance of street tree alignments on thermal benefit, focusing on downtown streets during the summer season. Our hypothesis was that the shade of street trees reduced the microenvironmental temperature and improved people's thermal comfort. We compared street sections with and without trees measuring physical parameters and performing thermal comfort interviews on people. In addition, to discriminate which variables best explained people's responses, we developed a multivariate analysis of the decision tree with physical data and human-based parameters.

2. Materials and methods

2.1. Climate representativeness

Montevideo, the capital city of Uruguay (−34.9055564, −56.1914125), contains 40 % of the country's population, with a density of 70 inhabitants/ha [21]. Its climate is characterized as temperate-subtropical, rainfall seasonal cycle is quite flat and there is no dry season (Cfa category of Köppen classification) [22]. The highest temperatures (maximum and mean) are in the summer months, from December to March. In the warmest month the mean temperature is above 22 °C and the mean maximum is 28 °C [23].

To verify that the study summer is within the ranges indicated by the climatology and therefore that the experiment took place under representative conditions, comparisons were made of the meteorological records of the study period with the climatological averages.

The meteorological records for the period December 2018–March 2019 were compared with the climatological averages (1981–2019) of the nearest meteorological station (Carrasco: −34.8328–56.0126) [23]. The monthly average of temperatures, wind and relative humidity, and the records for the days of the surveys (12/21/2018, 01/21/2019, 02/18/2019 and 03/28/2019) were compared.

2.2. Sites and sampling moment

“London plane” (Platanus x acerifolia) is a tree species used in many cities around the world, as it is a fast-growing tree and tolerant to air pollution and soil compaction. This deciduous tree can reach heights between 15 and 25 m and it is one of the species that provides the greatest microclimatic benefits on the streets in relation to air temperature, relative humidity, solar radiation, radiant temperature and wind speed. However, its young shoots and fruits shed small hairs that can cause allergies [[24], [25], [26]]. This species represents 48 % of the tree population in the downtown area of Montevideo, Uruguay [27]. More than 86 % of the specimens in the city have diameters greater than 40 cm. This tree-size is considered functional in relation to the benefits they provide and has been shown to adapt correctly, with a good performance in relation to other species [27,28].

Three city streets of Montevideo were selected as repetitions. Within each street a section with trees and a section without trees were identified as sampling sites (Fig. 1). The streets are in the city center (downtown) municipality, which has the second highest population density, with 128 inhabitants per hectare and one tree lined in streets for every 9 people [27]. The measurements were made in the summer season from December 2018 to March 2019.

Fig. 1.

Sampling sites in Montevideo city. Images show tree-lined sidewalks (shaded) and without trees (unshaded) sections (Google Earth Pro, 2018).

The south sidewalks were used for measurements. Although the buildings profiles were between 9 and 30 m high in these streets, the buildings around the measurement sites did not exceed the height of the trees so they did not produce any shadow on the sidewalk during the study window (11 a.m.–4 p.m.). The pavement width varied between 9 and 11 m and the sidewalks between 3 and 4 m. The pavement of the streets was asphalt and the sidewalk made of concrete tiles. The aligned trees were contemporary, adult and developed specimens of London Plane (Platanus x acerifolia). Trunk diameters ranged from 43 to 97 cm and their height was between 11 and 20 m, with good general sanitary condition. Trees were on the sidewalk, between the pedestrian pace and the pavement at an average distance of 17 m and the alignments generated a continuous canopy.

2.3. Thermal environment measurements

Six traditional Vernon globes [29,30] were used, one in each sampling site, three in tree-lined sections and three in unshaded sections on the north facing facades of the south sidewalk. The sensors inside the globes (TagTemp-S NFC portable data logger) were programmed to record the temperature (Globe T) every hour during the study period (Fig. 2). Instantaneous measurement of surface temperature (Façade T, Sidewalk T, Street T), air temperature (Air T) and wind speed (Wind S) were recorded with an infrared thermometer and a thermo-anemometer (Fig. 2) after carrying out the pedestrian interview in the places where the globes were located.

Fig. 2.

Device technical features.

2.4. Experimental design and statistical analysis of the physical environment

A complete block design with 3 blocks was used. Each block (street) has sections shaded by trees (treatment) and the sections of reference without shadow.

It was assumed that there was no block by treatment interaction, since the species and the management received were the same, and also the construction profile and the orientation were similar. For processing and analyzing the temperature data obtained from the spheres, an autoregressive moving average model (ARIMA) [31] was used, considering the seasonal component ARIMA (2, 0, 0) (2, 0, 0)24 (appendix 7.1).

2.5. Structured interviews

Pedestrians standing (n = 95) or walking (n = 25) on the sidewalk of the three selected streets were surveyed. The total number of people interviewed in the period was 120, whose ages varied between 13 and 73 years; 57 were men and 63 women. Regarding clothing, 15 people were wearing a vest top, 74 a t-shirt, 24 a shirt, 4 a sweater and 4 a jacket; 31 were wearing short pants or a skirt and 89 were wearing long pants. The structured interviews took place during the hours of maximum temperatures (11 a.m.–4 p.m.), monthly, on clear days and with calm wind, during summer (Dec. 21, 2018, Jan. 21, Feb. 18 and March 28, 2019). Information was collected through a standardized questionnaire (appendix 7.2) and the sampling form was subjective by reasoned decision. In this type of sampling, the sample units are chosen according to some of their characteristics, such as the fact that they are walking on a shaded sidewalk. It is used when the size of the sample is very limited, and random oscillations that excessively distance the sample from the characteristics of the population want to be avoided [32]. Adults walking or standing in the selected sites at the time of sampling were interviewed without selection by any other perceived characteristic. During each selected day, five people were surveyed per site, as this was the maximum number that could be done with the interviewer available during the stipulated time, so that all interviews were done on the same day and under similar conditions.

The thermal comfort questionnaire consisted of standardized questions and answers, and it used the categories proposed in ASHRAE (2013) [13] and included aspects raised by Thorsson et al. (2007) [33], [34]. Basic sociodemographic properties and specific aspects of the individuals were surveyed in the first place: gender, age, weight, and clothing. Second, interviewees were asked about their behavior: the activity they were doing at the time the survey was started and during the previous 15 min. Finally, interviewees were asked about attitudes in relation to their thermal perception. The current thermal sensation of the individual was enquired according to a 7-degree scale: very cold, cold, slightly cold, neutral, slightly hot, hot, very hot. Thermal preferences at that time were also recorded according to a 7- degree scale: much colder, colder, slightly colder, no changes, slightly hot, hotter, much hotter, and finally, comfort degree on a 3-level scale: comfortable, acceptable, discomfort [13,33] (appendix 7.2).

The effect of the treatments on the frequency of comfortable-uncomfortable cases was studied using a contingency table and the Irwin Fisher exact test. This test is valid for a low number of observations.

2.6. Physical and survey data crossing

To visualize the relationship between physical data and individuals' perceptions, the Partition platform of the JMP program was used [35]. This methodology finds groups of clusters for X values that best predict Y values. Thus, successive partitions are made, forming decision tree rules, until the desired fit is reached. The splitting criterion was "maximize significance", which means that the splits arise from the calculation of a significance value for each split; this significance value is presented as “LogWorth” and is the negative log of the adjusted p-value (LogWorth = -log10(p-value)). All the measured variables were considered to generate a decision tree with those variables that best explain people's responses.

3. Results

3.1. Climate representativeness

From the point of view of climate representativeness, the studied period was representative of the time of year, taking as a reference the climate averages of the nearby Carrasco meteorological station (Table 1). This result allows us to generalize the experiment that was developed in only one summer, despite the interannual climate variability reported for Uruguay [36,37].

Table 1.

Comparison of meteorological records (monthly mean) from December 2018 to March 2019 with climatological averages (1981–2019) and with the survey days (December 2018, January 21, 2019, February 18, 2019 and March 28, 2019).

	December				January
	1988–2019	2018	Diff.h	21st 2018	1988–2019	2019	Diff.	21st 2019
TMED (°C)a	21	21	−0.6	23	23	23	0.5	23
TXM (°C)b	26	25	−1.0	–	27	27	−0.2	–
TNM (°C)c	16	16	−0.2	–	18	19	1.2	–
TX (°C)d	40	32	−8.1	27	39	35	−4.2	29
TN (°C)e	7	8	1.2	19	10	12	2.9	16
HR (%)f	70	70	0.1	82	71	75	4.2	69
V (kn)g	9	9	−0.2	5	9	9	−0.3	7

	February				March
	1988–2019	2019	Diff.	18th 2019	1988–2019	2019	Diff.	28th 2019
TMED (°C)	22	22	−0.4	25	21	20	−0.7	20
TXM (°C)	27	27	0.1	–	25	24	−1.3	–
TNM (°C)	18	17	−0.9	–	16	16	−0.1	–
TX (°C)	38	34	−3.4	30	35	33	−2.2	25
TN (°C)	8	11	2.6	19	7	10	2.9	14
HR (%)	73	69	−4.1	71	75	76	1.6	72
V (kn)	8	7	−1.1	8	8	7	−0.6	8

^aMean air temperature.

^bMaximum air temperature mean.

^cMinimum air temperature mean.

^dAbsolute maximum air temperature.

^eAbsolute minimum air temperature.

^fRelative humidity mean.

^gWind speed mean in 24 hs

^hDifference between the climatological averages 1961–1990 and the survey days.

Mean temperatures were 21 °C in December 2018, 23 °C in January, 22 °C in February and 20 °C in March 2019. Relative humidity and mean wind speed were also representative in relation to historical records. The maximum mean temperatures also were those expected for each month of the season. Mean maximum temperatures were above 24.5 °C, which is considered the lower air temperature above which people in general prefer to be in the shade [17].

3.2. Physical microenvironment

Globe temperature records showed significant differences in the physical environment between tree-shaded sections and unshaded ones. Globe temperatures were used as a reference parameter since it is one of the variables used to estimate the mean radiant temperature, which is one of the most affected by tree shade, where lower temperatures were actually recorded [7,38].

For December, February and March, the globes located in unshaded sections, registered a difference of up to 9 °C in relation to those located in the shade (Fig. 3 a, c and d). In January, the difference reached 8 °C (Fig. 3b). Globe temperatures recorded in the shaded sections were more stable and showed lower variability than in the unshaded sites, as shown with the predicted values of the seasonal ARIMA model in Fig. 3.

Fig. 3.

Predicted values under sun (orange squares) and under tree shade (green dots) on the December 21, 2018 (a), January 21, 2019 (b), February 18, 2019 (c) and March 28, 2019 (d). Variation of the globe temperature into error bands (orange and green stripes) with a 95 % confidence interval.

3.3. Interviews

3.3.1. Comfort level

The proportion of respondents who said that they were comfortable in the shaded sections was more than double that of the unshaded section. The frequency of respondents who said that they perceived an acceptable level of comfort was similar under both treatments. All respondents who reported being uncomfortable were in unshaded places (Fig. 4).

Fig. 4.

Comfort level expressed by interviewees ‘comfortable’ (blue), ‘acceptable’ (gray), ‘discomfortable’ (red) (relative frequency) under treatments (sun/shade).

Irwin-Fisher test found statistical differences in “comfortable” and “uncomfortable” groups between treatments (sun/shade), considering error bars with a 95 % confidence (Fig. 4).

3.3.2. Comfort level and thermal sensation

Within the comfort and acceptable groups, the proportion of respondents in the neutral and slightly hot categories of thermal sensation increased greatly (Fig. 5a). Although heat discomfort was observed under both treatments, extreme heat categories had higher frequencies in unshaded places. All the subjects who reported being in discomfort responded that they were feeling some of the levels of "heat" at that time (Fig. 5a).

Fig. 5.

Thermal sensation (a) and thermal preferences («rather be») (b) expressed by interviewees (percentage frequency) inside groups of comfort level, under treatments (sun/shade).

The frequency of answers reflecting neutrality doubled under shade (Fig. 5a). Within the "comfortable" category, both in shaded and unshaded sections, the proportion of subjects who said they were feeling some level of heat was lower within “comfort” under shade (Fig. 5 a).

3.3.3. Comfort level and thermal preferences

The proportion of respondents who said that they did not prefer any thermal change at that time increased in the shaded sections (Fig. 5b). More than 75 % of the individuals who said that they did not prefer any thermal change at that time were comfortable under both treatments (Fig. 5b). Both neutral and slightly hot responses were clearly associated with comfortable responses.

3.4. Association between comfort levels and thermal preferences with other variables

Individual personal variables (gender, age, weight, clothing, activity) and physical variables of the thermal environment (Globe T, Air T, Façade T, Street T, Sidewalk T and Wind S) were analyzed together in a multivariate analysis in a decision tree (Fig. 6). The analyses identified only those variables that were significantly associated with the responses expressed by the interviewees. The first split was by façade temperature (Facade T), grouping all the discomfort answers inside the group with façade temperatures above 30 °C. Then, inside this group, sidewalk temperature (Sidewalk T) was the second variable that split the subgroup. On the other side of the decision tree, from the group with façade temperatures lower than 30 °C, the time of the day the interview was performed was the second variable to splitting that group (Time) followed by wind speed (Wind S).

Fig. 6.

Decision tree for comfort level expressed by interviewees: ‘comfortable’ (blue), ‘acceptable’ (gray), ‘uncomfortable’ (red), (P < 0,05, LogWorth>1,3).

When thermal preference was considered as a response variable, the first split was due to months, where responses for March were different from the rest of the dates. March mostly groups people who did not prefer any thermal change. For the other dates, the first split was due to treatment, that is, people under tree-shaded and unshaded sections. Here, shaded sections, grouped more people who did not prefer any change. Within the shaded group, we found slight differences between male and female interviewees (Fig. 7).

Fig. 7.

Decision tree for thermal preferences («rather be») expressed by interviewees, (P < 0,05, LogWorth>1,3).

4. Discussion

4.1. Thermal discomfort and high temperatures

There was a clear correspondence between the physical parameters and the comfort level, and the thermal discomfort was clearly associated with high temperatures. The mean air temperature registered in the meteorological station for the days of the surveys was above 20 °C, which is the air temperature reference that in general receives the best acceptance in relation to the outdoor comfort level for cities with a temperate climate such as Manchester (UK), although with lower mean temperatures for the warm season than Montevideo [17]. For the town of Nogales in Mexico, with a mean temperature of 30 °C, a “neutral” comfort temperature was observed at 25.8 °C [16]. This kind of difference is why the latest review article on this topic expressed the necessity of location-based studies to address climate and social conditions in thermal comfort studies [18].

When comparing "comfortable" groups under both treatments, people who said they were feeling heat (any of the 3°) represented lower proportions in shaded sections. Thus, in general, the level of comfort of the interviewees increased under shade, which emphasizes the improvement in sidewalk thermal comfort consistently with the improvement in comfort indexes based on physical variables [14,39]. This implies that the improvement in physical variables is also directly detected by the users, who found the most beneficial spots under trees, because of this thermal over-stress reduction [39]. On the other hand, no discomfort responses were found at these sites, where the microenvironments were cooler.

4.2. Thermal sensation and individual preferences

In the structured interviews, questions related to the perception of the thermal environment are related to attitudes, which can lead to imprecise and ambiguous answers and cause respondents to give normative answers [32]. For this reason, we first asked about specific aspects to later contrast consistency with the general comfort level. Responses of thermal sensation showed that the effect of trees was even greater than that reflected only by comfort responses, because satisfactory thermal sensation categories can lead to "acceptable" as well as "comfortable" comfort responses. On the other hand, this can be used to confirm that those people who answered that they were comfortable effectively perceived it. In this study, answers were generally consistent, since the groups that said that they were comfortable were composed mostly by people who did not prefer any thermal change.

Individuals who declared to be in discomfort and were feeling some heat level would indeed have preferred to be feeling cooler to some degree, which makes it clear that the discomfort is due to the heat, and they are not thermally comfortable. On the other hand, groups that said they were comfortable were, in fact, mostly made up of people who did not prefer any thermal change at that time, which shows that there is thermal satisfaction with the environment [13,16]. When comparing these groups under both treatments, the situation with the highest proportions of people who declared feeling effectively "neutral" were those who were in shaded sections.

The acceptable category could be masking other comfort levels due to the phenomenon of acquiescence: “… tendency of the interviewees to choose the answers that express agreement, to give affirmative answers rather than negative …” [32]. This would explain those people who responded as "comfortable" or "acceptable" after having expressed a "hot" or "very hot" thermal sensation. Consequently, we may interpret that those individuals who actually said to be neutral or somewhat warm, but then said that they were "acceptable", were actually “comfortable”. Taking this into consideration and analyzing thermal comfort, it can be said that the frequency of comfortable people doubled under shade, which supports the significant positive impact tree shade provides on pedestrian thermal comfort [39]. Sidewalk trees are often the subject of heated public debate due to their effect on the deterioration of sidewalks or the inconveniences that the fruiting of London planes generates. In some cases, such as the city of Montevideo, the debate reaches the point of suggesting their massive extraction without considering the positive effect on people's thermal comfort under natural shade.

4.3. Comfort responses in specific groups

Groups were split only by physical variables when comfort answers were taken as the response variable. The variables usually mentioned as the most important factors determining people's thermal comfort [13], were those that divided the comfort groups in first order. The successive divisions were based on physical environment variables, related to radiant temperature [16]. There were no associations with specific variables to the individuals. However, some aspects related to physiological and psychological adaptation become visible when analyzing the regression tree which takes “rather be” as the response variable. In this case, decisive factors involved other aspects in addition to the physical microclimatic parameters, such as month, treatment, or gender. Specifically within the individual variables, gender was a significant one in thermal preferences, with the 46,4 % of females under shade preferring a slightly colder environment, a 10 % up in relation to males (Fig. 7). Including all the variables surveyed (physical and "personal") allowed gender (sex) to be discriminated as a statistically significant variable. According to Aghamolaei et al. (2023) [18], differences in gender are one of the most important issues that should be taken into consideration in the assessment of thermal comfort in public spaces.

Different people perceive the environment in different ways and respond to stimuli depending on the information they have about the particular situation and not in direct relation to the magnitude of physical variables [20]. This involves expectations about how the thermal environment should be, rather than how it really is [20]. Because it is expected to be hot in the summer, many respondents found themselves not preferring any thermal change in the warmer months (December, January and February), despite the fact that the temperatures are considerably higher than what would be expected for an individual to feel comfortable. Thus, in those months, people are predisposed to high temperatures and are inclined to tolerate them. But at the same time, another large group would prefer to feel colder, which makes it clear that they feel hot, and a cooler environment would allow them to reach a neutrality level, particularly in unshaded sections.

In January, which is normally warmer, the trend in the proportion of individuals who answered being "comfortable" under both treatments, changed. This leads us to think that in this month, people are predisposed to feeling hot and that although the amount of "acceptable" increases under shade, perhaps the effect of tree shade on the microenvironmental temperature is not sufficient to achieve thermal neutrality and therefore elicits a “comfortable” answer. In March, on the other hand, where temperatures are lower and closer to what is expected to reach comfort, most of the individuals would not have preferred any change, and therefore we can say that they found themselves comfortable. In this month, the proportion of comfortable people in the shade triples in relation to unshaded places, which shows the more effective contribution of tree shade with less extreme temperatures.

We agree with Aghamolaei et al. (2023) [18] that the study of the combined effects of physical and human-based variables on thermal comfort in open spaces is the main challenge for researchers. We believe that the combined evaluation of the physical and human-based parameters presented in this study, through the decision tree analysis, is a contribution in that sense.

4.4. Limitations of this research

In this research, results have been obtained from widely used measurement instruments, standard surveys and rigorous statistical analyses, all of which are supported by the bibliography. However, we believe that it would be beneficial to complement the research by repeating the experiment in other years and surveying a larger number of people.

5. Conclusions

This study revealed a greater impact of physical variables than personal variables on people's thermal comfort and thermal preferences in the summer in a temperate climate. However, the gender variable resulted in a statistically significant one in thermal preferences, where females would prefer feeling cooler even under shade, a 10 % up in relation to males. Finding this result was possible by using the statistical decision tree methodology that combines all the variables in the same analysis, unlike studies that analyze the personal variables separately. At the same time, the results are supported by field work (in situ experimentation) rather than model simulations.

Our experiment showed that a great number of people surveyed under both treatments stated they did not feel fully comfortable due to summer season heat, which shows that microenvironmental temperatures recorded in the analyzed streets are above those considered thermally neutral by people. Tree shade refreshes locally, reducing the microenvironmental radiant temperature and improving people's comfort level. It also generally improves their thermal conditions of “neutrality” and eliminates the extreme “discomfort” heat sensation. The positive effect of shade on thermal comfort is still present under less extreme conditions, which means that this benefit can extend at least over the 4 months studied.

Variables that involve aspects of the global context also arise when considering preferences, including expectations in relation to the month and preferences for wooded sites. In this way, the predisposition to experience high temperatures can mask part of the direct physical effect of shade, since the sensation of heat (objective and subjective) would be at a higher level in relation to the cooling effect.

This study focused on how street trees can influence pedestrian thermal comfort, unlike the thermal comfort studies that are carried out in squares or parks. Our intention was to elucidate the effect of tree shade in places where people pass daily, not necessarily where people stay for recreational or relaxing purposes. In this sense, tree alignments constitute a powerful tool for improving the thermal microenvironment in the streets of temperate cities; therefore, they are also a tool to adapt to climate change. This is not only due to their influence on physical parameters, but also due to the improvement in perceived thermal comfort on the sidewalks, where the effect of the heat of summer is a discomfort factor. These results could help local governments to justify that urban tree alignments bring a lot more than just ornament to the cities. Trees are needed to improve people's quality of life in urban areas.

Data availability statement

Data will be made available on request.

CRediT authorship contribution statement

Emilio Terrani: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Conceptualization. Alicia Picción: Writing – review & editing, Visualization, Validation, Supervision, Methodology, Investigation, Funding acquisition, Formal analysis, Conceptualization. Oscar Bentancur: Writing – review & editing, Supervision, Software, Methodology, Formal analysis. Gabriela Cruz: Writing – review & editing, Visualization, Supervision, Resources, Project administration, Investigation, Funding acquisition, Formal analysis, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is/are the supplementary data to this article.