Outdoor Environmental Comfort, Public-Space Use and Perceived Sociability Among Ageing Populations: Integrating Field Surveys and Microclimatic Simulations

Abstract

Outdoor environmental comfort is increasingly relevant amid climate change and population ageing because it affects well-being, behaviour and public-space use. However, the relationship between microclimatic conditions, environmental perception, and public-space use among older adults remains underexplored. The novelty of this study lies in integrating environmental measurements, subjective thermal perception, microclimatic simulations and public-space use within a baseline assessment undertaken before the redevelopment of a real public space, providing an evidence-informed framework for age-friendly urban design. The study examines environmental conditions, subjective perception and reported public-space use among adults aged 55 years and over within the Barrendain intervention area (Beasain, Spain). A mixed-method approach combined seasonal field surveys (n = 117) with microclimatic simulations. Survey data included thermal sensation, environmental perception, adaptive behaviour, and reported public-space use and perceived sociability. PET distributions were simulated using ENVI-met/BIO-met simulations driven by meteorological data from an official weather station. The results indicate that thermal perception exhibited a marked adaptive character. Neutral temperatures ranged from 12.35 °C in the cold-period campaign to 19.87 °C in the warm-period campaign, reflecting seasonal adaptive thermal responses among older adults. Tree-covered areas presented lower PET values than exposed surfaces during warmer conditions, highlighting the role of vegetation in mitigating summer heat stress. The partial correspondence between simulation-derived PET and subjective thermal sensation further emphasises the importance of combining objective microclimatic assessment with user-centred evaluation in age-friendly public-space design. In addition, 55% of respondents reported using the intervention area primarily for physical activity, whereas 36% identified social encounters as a main motivation for visiting the public space. The findings indicate that environmental comfort should be understood as an enabling condition for public-space use and perceived sociability rather than as their sole determinant. The proposed methodological framework demonstrates how simulation-based and user-centred evidence can be combined to support climate-responsive and age-friendly urban regeneration.

1. Introduction

1.1. Ageing, Loneliness and Public Space

Population ageing, climate change and the growing recognition of public space as a determinant of healthy ageing are placing new demands on urban design. For older adults, outdoor environments are not only settings for mobility and recreation but also essential infrastructures that support everyday wellbeing, social participation and independent living. Their environmental and spatial quality may therefore influence whether older adults can comfortably access, use and remain in public spaces as part of their everyday routines.

Within this broader context, loneliness and social isolation have become major public-health concerns, particularly among older adults. Approximately one in six people globally experience loneliness, which has been associated with significant negative health outcomes and increased mortality [1]. In Europe, 13% of the population report feeling lonely always or almost always, while 35% experience loneliness at least occasionally [2]. In Spain, loneliness particularly affects older age groups, with prevalence increasing significantly after the age of 65 [3,4].

Demographic ageing further increases the relevance of these challenges. In 2024, people aged 65 years and over represented 21.6% of the European population, and this proportion is expected to continue increasing over the coming decades [5]. In this context, promoting healthy and active ageing has become a major social and urban challenge. This requires neighbourhood environments that support mobility, rest, outdoor activity, everyday encounters and continued participation in community life.

Public spaces are increasingly recognised as key forms of “social infrastructure” that support everyday interaction, inclusion and well-being. The World Health Organization and recent European policy frameworks emphasise the importance of accessible and welcoming neighbourhood environments in helping to reduce social isolation and strengthen social connection [1,6]. Consequently, urban planning and public-space design are gaining relevance as tools to support social interaction and community cohesion [7]. However, the capacity of public spaces to support these processes depends not only on their availability or accessibility, but also on whether their environmental conditions allow people to use them comfortably throughout the year.

Previous studies have explored the relationship between neighbourhood characteristics, green spaces and loneliness [8,9,10,11]. However, studies explicitly addressing loneliness and social interaction among older adults from the perspective of public space environmental conditions remain limited [7,12]. In this context, environmental comfort—particularly outdoor thermal comfort plays a key role in shaping the usability and perceived quality of urban public spaces [13]. Environmental comfort should therefore be regarded as one of several conditions that may facilitate public-space use and opportunities for encounter, rather than as a direct determinant of social interaction.

In addition to thermal conditions, previous studies have highlighted the influence of vegetation, shading, acoustic perception, spatial configuration and environmental quality on the attractiveness and sociability of public spaces, particularly among older adults [7,14,15,16,17]. These environmental and spatial attributes may affect the perceived pleasantness, accessibility and usability of a place, thereby shaping the conditions under which everyday activities and informal encounters occur.

1.2. Outdoor Thermal Comfort and Ageing Populations

Outdoor thermal comfort (OTC) strongly influences the usability and attractiveness of urban public spaces, particularly for older adults and other thermally vulnerable groups. Previous field studies have demonstrated the adaptive and context-dependent nature of outdoor thermal perception, highlighting the importance of behavioural, environmental, and psychological factors in shaping thermal responses in real urban environments. Accordingly, assessing outdoor comfort requires combining physical environmental conditions with users’ subjective responses and adaptive behaviour.

Despite the growing body of research on OTC, studies specifically focused on older adults remain relatively limited [18,19,20]. Most thermal comfort models are still based on the characteristics of a standard adult subject and do not adequately account for the physiological and behavioural differences associated with ageing [18,21,22]. These age-related differences are associated with physiological changes affecting thermoregulation, metabolic activity, and thermal sensitivity, which may alter thermal perception and adaptive capacity in outdoor environments [22,23].

Several authors have reported that older adults generally tolerate a wider range of outdoor temperatures than indoor ones [18,24], although important differences in thermal perception emerge with age. Research conducted in different climatic contexts indicates that older adults tend to prefer warmer conditions and exhibit lower thermal sensitivity than younger populations [25,26]. Studies conducted in temperate oceanic climates similar to the context of this research further indicate that ageing is commonly associated with higher clothing insulation levels, lower activity levels, and a greater reliance on adaptive behaviour [23]. Evidence specifically focused on older adults in temperate Atlantic public spaces nevertheless remains limited.

Behavioural and contextual factors also play an important role in shaping outdoor thermal perception among older adults. Evidence suggests that adaptive strategies vary according to sex, activity patterns, lifestyle, and environmental exposure [27,28]. These findings highlight the complexity of outdoor thermal perception and the importance of integrating physiological, behavioural and environmental dimensions in outdoor comfort assessments.

In the context of climate change and increasing urban heat exposure, understanding how outdoor environmental conditions may affect the everyday use and perceived usability of public spaces among older adults has become an increasingly relevant research topic [11,29,30]. Recent interdisciplinary studies have highlighted the need to integrate urban microclimate, environmental perception, and social well-being in order to support age-friendly and climate-resilient urban environments [7,12,31]. This integrated perspective is particularly relevant because thermal conditions may not correspond directly with users’ reported sensations, preferences or patterns of adaptation.

In parallel, microclimatic simulation tools such as ENVI-met have been increasingly applied to evaluate outdoor thermal comfort in urban public spaces [32,33,34]. Several works have demonstrated the influence of urban morphology, vegetation density, tree-canopy configuration, shading continuity, and surface materials on pedestrian-level thermal conditions and PET distribution [35,36]. These simulation-based approaches have been widely used to assess heat mitigation strategies, seasonal thermal variability, and the climatic performance of public spaces under different urban and environmental conditions [32,33,35]. In particular, vegetation and shaded areas have consistently been identified as key factors in reducing radiative thermal stress during warm periods [15,36,37]. However, most simulation studies focus primarily on the physical assessment of thermal conditions, while fewer studies explicitly examine how microclimatic conditions may relate to patterns of use, sociability, and social interaction among older adults [22,30,31]. Moreover, simulation outputs are not always analysed alongside season-specific field surveys and user-centred assessments capable of identifying discrepancies between modelled thermal stress and subjective thermal perception.

1.3. Research Gap, Novelty and Objectives

Despite the growing body of research on outdoor thermal comfort and ageing, three related limitations remain. First, relatively few studies focus specifically on adults aged 55 years and over across contrasting seasonal conditions. Second, subjective thermal perception, multisensory environmental appraisal and microclimatic simulation are frequently examined separately rather than within an integrated research framework. Third, the relationship between modelled thermal exposure, reported patterns of public-space use and perceived sociability remains insufficiently understood, particularly in temperate Atlantic urban contexts [22,30,31].

This study was motivated by the opportunity to integrate research into the evidence-informed redevelopment of a real urban public space through the KALELAGUN project. The planned transformation of the Barrendain intervention area (Beasain, Spain) enabled the research team to undertake a baseline assessment of environmental conditions, users’ perceptions and reported public-space use before the implementation of the proposed design. This real urban transformation process provided the opportunity to investigate how objective microclimatic conditions and subjective environmental perception jointly relate to the everyday use and perceived sociability of public spaces among older adults, while directly informing the subsequent design process.

The originality of the present study lies in the integration of four complementary components within a single case-study framework: (i) seasonal field surveys involving adults aged 55 years and over; (ii) assessment of thermal, acoustic and visual perception together with reported public-space use; (iii) ENVI-met/BIO-met simulations using elderly user profiles differentiated by sex and posture; and (iv) translation of the combined findings into design implications for an ongoing public-space redevelopment process. By integrating objective environmental assessments with users’ subjective perceptions within a real urban regeneration process, the proposed framework provides a comprehensive diagnostic approach capable of supporting evidence-informed, climate-responsive and age-friendly public-space design. The study does not seek to demonstrate a direct causal relationship between thermal comfort and social interaction; rather, it examines environmental comfort as one of the conditions that may support public-space usability and opportunities for everyday encounter.

The overall aim of this study is to characterise the baseline environmental conditions, thermal adaptation and reported public-space use among adults aged 55 years and over within the Barrendain intervention area, by integrating seasonal field surveys with ENVI-met/BIO-met simulations to support evidence-informed urban redevelopment.

The specific objectives of the study are:

• to characterise seasonal differences in thermal sensation, thermal comfort, thermal preference and multisensory environmental perception among adults aged 55 years and over;
• to analyse the association of selected personal, behavioural and contextual variables with reported thermal sensation;
• to characterise the spatial and seasonal distribution of PET within the Barrendain intervention area and identify the environmental features associated with exposed and thermally moderated areas;
• to compare simulated PET classifications with users’ reported thermal sensations and examine possible discrepancies between modelled exposure and subjective perception;
• to derive design implications for age-friendly and climate-responsive public spaces, while treating environmental comfort as an enabling condition for use and encounter rather than as a direct predictor of social interaction.

Accordingly, the study addresses the following research questions:

RQ1. How do thermal sensation, comfort and preference among adults aged 55 years and over vary across seasonal field campaigns?

RQ2. Which personal, behavioural and contextual variables are associated with reported thermal sensation within the study sample?

RQ3. What spatial and seasonal PET patterns occur within the Barrendain intervention area, and which microclimatic processes and urban features primarily explain these patterns?

RQ4. To what extent do simulated PET classifications correspond with users’ reported thermal sensations across the seasonal campaigns?

RQ5. What reported patterns of public-space use and perceived sociability emerge from the survey, and what design implications can be derived without assuming a direct causal relationship between environmental comfort and social interaction?

2. Materials and Methods

The study is part of the KALELAGUN (meaning “friendly street” in Basque) research project lead by the University of the Basque Country, a multidisciplinary initiative aimed at developing evidence-based design strategies for age-friendly public spaces through the integration of environmental assessment, microclimatic simulation and participatory processes. The methodological framework is described in detail in the KALELAGUN Design Guide [12]. The framework adopts an open-innovation approach that combines environmental, spatial and social analysis with participatory co-creation.

Within this framework, the present study adopts a mixed-method case-study approach combining environmental measurements, thermal comfort assessment, microclimatic simulations, and field surveys. The objective is to examine the relationship between thermal conditions, environmental perception, and patterns of use among older adults in public space.

2.1. Study Area and Research Context

The study was conducted within the Barrendain intervention area, located in the municipality of Beasain (Gipuzkoa, Basque Country, northern Spain), one of the pilot sites of the KALELAGUN project. The project is promoted by the ADINBERRI Foundation, the Gipuzkoa Provincial Council and participating municipalities within the framework of the Gipuzkoa Strategy on Loneliness (HARIAK). It has also been recognised as a good practice within the KORALE project (Interreg Europe).

The selection of Barrendain Square emerged from an ongoing process of collaboration between the research team and the Municipality of Beasain. The team had previously collaborated with the municipality through the ETXELAGUN project and, subsequently, within the KALELAGUN project, developed in partnership with the municipalities of Beasain, Donostia-San Sebastián and Arrasate under the Gipuzkoa Strategy on Loneliness (HARIAK). During the initial stages of KALELAGUN, Beasain identified Barrendain Square as a priority site for redevelopment, with a municipal intervention already planned and funded. This provided an opportunity to integrate the research into a real urban transformation process and to generate evidence capable of informing the ongoing design proposal.

In addition, the intervention area presented several characteristics that made it particularly suitable for this research, including its central location within the municipality, its role in everyday pedestrian mobility, its frequent use by older residents, the coexistence of transit and staying activities, and the presence of contrasting microclimatic conditions associated with vegetation, paved surfaces and surrounding buildings.

For clarity, the term “Barrendain intervention area” is used throughout this paper to refer to the full pre-redevelopment urban area considered in the municipal project, including the existing playground, the roundabout, the intervening roadway, Avenida de Navarra and the adjacent pedestrian spaces. At the time of the study, these elements did not constitute a consolidated public square. The existing place commonly referred to as Barrendain Square corresponded primarily to the playground area, while most everyday pedestrian activity was concentrated along Avenida de Navarra and the surrounding routes. The surveys and the reported PET analyses therefore characterise the pre-intervention conditions of the wider redevelopment area rather than the performance of an already consolidated square.

At the time of the study, the municipal project aimed to transform the fragmented, traffic-dominated intervention area into a continuous, inclusive and accessible public space by integrating the existing playground, roundabout, roadway and adjacent pedestrian areas (Figure 1). The surveys, measurements and simulations reported in this paper characterise the pre-intervention configuration and provide baseline diagnostic evidence for the design process. The analysis was used to identify thermally exposed and protected areas, assess the role of the existing tree canopy and inform decisions concerning seating, climatic protection and spatial continuity. The proposed redevelopment itself was not simulated, and its thermal performance should therefore not be regarded as validated by the present results.

Beasain is a medium-sized municipality at an altitude of 159 m above sea level and characterised by a temperate Atlantic climate (Cfb according to the Köppen–Geiger classification) (Figure 2). This climate is defined by mild temperatures, high humidity, and abundant precipitation throughout the year. The annual mean temperature is 12.3 °C, with average minimum and maximum temperatures of 7.9 °C and 16.8 °C, respectively. These characteristics make Beasain relevant for examining seasonal thermal adaptation in a temperate Atlantic context, where outdoor use is shaped not only by summer heat exposure but also by rainfall, humidity, wind and access to winter solar radiation [38].

Beasain has a population of 13,915 inhabitants (6984 men and 6931 women), of whom 39.7% are aged 55 years or older (2635 men and 2896 women) [40]. This demographic structure is particularly relevant for the present study, given its focus on older adults and their use of public space. Moreover, 953 people over 55 years live alone in Beasain (339 men and 614 women) (EUSTAT 2022). Living alone has been associated with an increased likelihood of loneliness compared to living with others [41], highlighting the importance of social environments that support interaction and reduce isolation.

Women constitute a higher proportion of the older population, particularly from the age of 75 onwards (746 men and 1054 women). This sex imbalance is relevant for interpreting patterns of space use, perception, and social interaction among older adults (

Table 1. Population aged 55 years and older by age group in Beasain.

Age Groups	Total	Men	Women
55–64	1996	1036 (51.9%)	960 (48.0%)
65–74	1735	853 (49.1%)	882 (50.8%)
75–84	1167	520 (44.5%)	647 (55.4%)
>85	633	226 (35.7%)	407 (64.2%)
TOTAL	5531	2635 (47.6%)	2896 (52.4%)

).

The Barrendain intervention area covers approximately 5500 m² in the centre of Beasain (Figure 3). Older adults constitute one of the principal user groups considered in the KALELAGUN framework, although the planned public space is intended for the wider population.

The site occupies a strategic position within the urban fabric, connecting key mobility flows and everyday-use spaces. To the south, the train station and a high-traffic road generate a constant flow of pedestrians, while to the north it borders Avenida de Navarra, one of the municipality’s main pedestrian axes (Figure 4). The project integrates the existing roundabout with the adjacent playground, creating a continuous public space that supports both movement and stay. On the eastern side, the “Pergolas” area—frequently used by local residents, including older adults—reinforces the role of the site as a setting for everyday activity and informal encounter.

The proposed redevelopment incorporates areas for rest and leisure, children’s play, vegetation, multifunctional uses, pedestrian and cycling routes, new seating and climatic-protection measures. These design intentions are described only to establish the applied context of the research; their environmental and post-occupancy performance is outside the scope of the present baseline study.

2.2. Survey Campaign

At the time of the field campaigns, the Barrendain redevelopment had not yet been implemented and the future square did not exist as a consolidated public space. The survey location was selected pragmatically to intercept actual users of the pre-intervention area, since pedestrian activity was concentrated along Avenida de Navarra and very limited use was observed within the existing playground. This approach ensured that the survey captured the perceptions of regular users of the intervention area under pre-intervention conditions, providing the baseline information required for the subsequent redesign process.

A non-probability intercept sampling approach was adopted because the objective was to characterise environmental perception under real-use conditions rather than to estimate population prevalence. Consequently, the sampling strategy prioritised contextual diversity across seasons and weather conditions over statistical representativeness.

Only participants aged 55 years and over were included in the final sample. This age threshold was adopted because the KALELAGUN project was developed within the ADINBERRI and HARIAK strategies, which identify adults aged 55 years and over as the target population for promoting healthy ageing, preventing loneliness and supporting age-friendly environments [42,43]. Rather than focusing exclusively on older age, this approach adopts a preventive life-course perspective, recognising that physiological, behavioural and social changes relevant to everyday environmental perception may begin before the conventional threshold of 65 years. The inclusion of adults from 55 years onwards therefore enabled the study to capture a broader range of ageing trajectories, consistent with the strategic framework in which the project was developed.

Potential participants were initially approached on the basis of their apparent age and subsequently invited to participate. Eligibility (≥55 years) was confirmed before administering the questionnaire.

Fieldwork was conducted during four two-week seasonal campaigns in 2024: 29 January–11 February, 15–28 April, 1–14 July, and 23 September–6 October. Surveys were carried out on alternating days and at three different times of day: 9:00–10:00, 12:00–13:00, and 16:00–17:00, selected to capture different patterns of public-space use throughout the day.

A structured questionnaire was designed to assess environmental perception and social interaction within the study area. The instrument was organised into four main thematic sections addressing: (i) personal and behavioural characteristics; (ii) thermal perception; (iii) reported public-space use and perceived sociability; and (iv) multisensory environmental perception. Although additional socio-demographic and contextual information was collected, the present study focuses on the variables directly related to environmental perception, thermal comfort, reported public-space use and perceived sociability, as these correspond to the objectives of the study and the statistical analyses presented in the results section.

Table 2. Structure and main variables included in the questionnaire survey.

Questionnaire Section	Main Variables Collected
Personal and behavioural characteristics	Age, sex, place of birth, living situation, accompaniment, clothing insulation level, qualitative body-build category, seated/standing activity, apparent mobility and sensory conditions.
Environmental and contextual conditions	Air temperature, humidity, cloud cover, solar exposure, wind, rain, noise presence, wet surfaces, lighting conditions
Thermal perception and comfort	Thermal sensation vote (TSV), thermal comfort evaluation, thermal preference, humidity perception and preference, wind perception and preference, solar exposure perception and preference.
Multisensory environmental perception	Acoustic quality, acoustic preference, dominant sounds, visual perception of lighting conditions, colour perception, environmental pleasantness.
Public-space use and social interaction	Seasonal attendance, frequency and timing of visits, primary activities, ease of social interaction, perceived sociability, reasons for social interaction.
Spatial assessment and improvement strategies	Perceived environmental discomforts, preferred environmental characteristics, open-ended suggestions for spatial improvement.

summarises the main questionnaire sections and variables considered in the analysis. The complete questionnaire is provided in Supplementary Materials.

Questions concerning activities undertaken in the study area and perceived facilitators of social interaction allowed multiple responses. Consequently, percentages reported for these variables may exceed 100%.

The first part recorded age, sex, place of birth, clothing insulation level, qualitative body-build category, posture or activity level (seated or standing), accompaniment, and living situation. In addition, contextual environmental conditions such as sun exposure, cloudiness, wind, and noise presence were recorded during the survey.

The second part of the questionnaire focused on thermal perception using structured subjective scales. Thermal sensation vote (TSV) was assessed using a five-point scale ranging from −2 (cold) to +2 (hot), adapted from the ASHRAE framework and ISO 10551 (2019) [44] in order to improve usability and response reliability under field conditions, particularly among older adults [45,46].

Thermal comfort was evaluated using a five-point scale (1. comfortable; 2. quite comfortable; 3. slightly uncomfortable; 4. very uncomfortable; 5. extremely uncomfortable), adapted from ISO 10551 (2019), allowing respondents to assess perceived comfort under real outdoor conditions [45,46].

Thermal preference (TP) was assessed using a five-point scale (1. much warmer; 2. warmer; 3. no change; 4. cooler; 5. much cooler), based on the preference categories defined in ISO 10551 (2019) and commonly used in outdoor thermal comfort studies [47,48].

Humidity perception was evaluated using a three-point scale (1. humid; 2. neutral; 3. dry), while humidity preference was assessed through a corresponding three-point scale (1. more humid; 2. no change; 3. drier), following approaches commonly used in outdoor comfort [24,49].

Wind perception was assessed using a five-point scale (1. no wind; 2. light wind; 3. moderate wind; 4. strong wind; 5. very strong wind), consistent with descriptors commonly used in outdoor thermal comfort research [47,50].

Overall, the selected scales were adapted from international frameworks including ISO 10551 and the ASHRAE thermal sensation model, while being simplified to facilitate comprehension and usability under real outdoor survey conditions among older adults [45,51].

The third part addressed patterns of social interaction, including living arrangements (living alone versus living with others), companionship at the time of the survey (alone vs. accompanied), primary activities performed in the space, seasonal attendance (winter and summer), and perceived sociability, defined as the ease of interacting with others.

The fourth part of the questionnaire addressed the multisensory perception of the space. This section included the assessment of perceived acoustic quality using a four-point scale (1. very unpleasant; 2. unpleasant; 3. pleasant; 4. very pleasant), based on a semantic differential approach consistent with ISO 10551 adapted to the acoustic domain. The use of an even-numbered scale avoids a neutral midpoint, encouraging respondents to express a clear evaluative tendency [16,52].

Acoustic preferences were measured through a four-point scale indicating desired changes in sound conditions (1. no change; 2. slightly quieter; 3. quieter; 4, much quieter). This directional preference approach is consistent with methodologies used in soundscape research and aligns conceptually with thermal preference scales, allowing the identification of desired environmental adjustments rather than static evaluations (ISO 12913-2, 2018) [53].

This section also examined visual perception of colour within the space and incorporated open-ended questions to identify users’ suggestions for spatial improvement. The inclusion of qualitative responses complements structured scales by capturing contextual, cognitive, and experiential aspects of environmental perception. This mixed-method approach is widely recommended in soundscape and environmental perception research to better understand the interaction between physical stimuli and subjective appraisal [53,54].

2.3. Environmental Measurements

Environmental data were acquired from an official meteorological station located near the study site [55]. The environmental conditions presented correspond to the meteorological data recorded during the periods in which the surveys were conducted in each season (

Table 3. Summary of the main environmental data recorded during the field surveys.

	Ta (°C)			RH (%)	Ws (m/s)	SR (W/m2)
	Mean	Max.	Min	Mean	Mean	Mean
Winter	11.27	16.0	3.3	77.47	0.96	95.47
Spring	12.36	17	3.5	72.48	0.24	329.67
Summer	20.81	28.0	14.0	77.05	0.64	501.47
Autumn	17.76	27.4	12.3	76.8	0.79	277.00

Ta = air temperature; RH = relative humidity; Ws = wind speed; SR = solar radiation.

). Mean air temperature ranged from 11.27 °C in winter to 20.81 °C in summer, with maximum and minimum values of 28.0 °C and 3.3 °C, respectively. Relative humidity showed mean values between 72.48 and 77.47 across seasons. Mean wind speed ranged from 0.24 m/s to 0.96 m/s, and solar radiation varied from 95.47 W/m² in winter to 501.47 W/m² in summer.

2.4. Thermal Perception Analysis

The relationship between air temperature and thermal sensation was analysed, with particular attention to campaign-specific neutral temperatures, the role of behavioural and contextual variables, and the limitations of simplified Ta-based approaches.

Survey data were grouped into two campaign-based periods: a cold-period campaign (January, February, and October; n = 61) and a warm-period campaign (April, July, and September; n = 56) (

Table 4. Descriptive characteristics of the two campaign datasets.

Variable	Cold-Period Campaign	Warm-Period Campaign
Field campaigns	January, February, October	April, July, September
Sample size (n)	61	56
Age (years)	73.6 ± 10.6 (55–96)	70.8 ± 6.1 (58–87)
Women/men	32/29	31/25
Air temperature (°C)	12.8 ± 5.1 (3.3–27.4)	16.0 ± 5.3 (3.5–28.0)
Relative humidity (%)	76.6 ± 14.6 (41–100)	75.6 ± 15.8 (47–100)
Solar radiation (W/m2)	134.4 ± 153.2 (1.6–632.7)	370.3 ± 218.5 (59.1–861.9)
Sun/no sun	28/33	27/29

Values are presented as mean ± standard deviation (range) where applicable.

). The cold- and warm-period groupings reflect the organisation of the survey campaigns rather than strict climatological seasons, acknowledging the influence of behavioural adaptation and seasonal expectations on outdoor thermal perception. In outdoor contexts, thermal sensation is influenced not only by instantaneous conditions but also by expectations, prior exposure, clothing habits, and everyday practices [32,33,45,49,56]. The terms “cold-period” and “warm-period” should therefore be interpreted as campaign labels rather than homogeneous thermal categories.

For each interview, the dataset included air temperature (Ta), relative humidity, solar radiation and wind speed, together with contextual descriptors such as sun exposure, cloudiness and perceived wind. Personal variables included clothing, posture, body-build category, age and sex. Thermal sensation was reported through a five-point TSV scale from −2 (very cold) to +2 (very warm).

This analysis focuses on variables consistently available across paired survey-measurement records. Body-build category was assessed qualitatively and should not be interpreted as a direct anthropometric measure. Although wind speed was recorded, it was not treated as a primary explanatory variable due to insufficiently stable distributions across the campaign datasets.

2.5. Data Processing and Statistical Analysis

To reduce variability in field TSV data, mean thermal sensation vote (MTSV) values were calculated using 1 °C air-temperature bins, a common approach in outdoor thermal comfort studies that clarifies central tendencies without removing individual variability [32,33,49,56].

For each campaign, a linear regression was fitted to the MTSV-Ta relationship to estimate neutral temperature (Tn) at MTSV = 0. Values corresponding to MTSV = −1 and +1 were also derived but were interpreted as model outputs rather than observed comfort limits.

Given the ordinal nature of TSVs and the presence of small or unbalanced subgroups, non-parametric tests were applied. Kruskal–Wallis tests were used for multiple-group comparisons, Mann–Whitney U tests for binary contrasts, and Spearman rank correlations for monotonic relationships with continuous variables. The statistical analyses were designed to identify associations between environmental variables, subjective perception and reported public-space use. Given the observational and cross-sectional design of the study, the reported relationships should be interpreted as statistical associations rather than evidence of causal relationships. Accordingly, the results were interpreted as exploratory rather than confirmatory.

2.6. Microclimatic Simulations

2.6.1. Simulation Objectives

The main objective of the simulations was to characterise the spatial and seasonal distribution of outdoor thermal conditions within the study area and to provide an independent model-based assessment of the microclimatic environment experienced by older users.

The simulations were used to complement the field surveys by enabling the comparisons between modelled environmental conditions and users’ subjective thermal responses. This combined approach made it possible to examine the extent to which thermal sensation, comfort and thermal preference were associated with the microclimatic conditions simulated for different locations within the intervention area and different periods of the year.

In addition, the simulations supported the decision-making process associated with the ongoing redesign of Barrendain Square. The spatial analysis of thermal conditions at specific locations within the intervention area helped identify areas requiring environmental improvement and informed decisions regarding the distribution of seating areas, the incorporation of vegetation, and the location of shaded and protected spaces. In this way, the simulations provided both an analytical framework for interpreting users’ perceptions and a practical basis for climate-responsive design interventions. Thermal comfort conditions were evaluated using the physiological equivalent temperature (PET) index calculated through the BIO-met module integrated within ENVI-met.

2.6.2. Model Domain and Intervention Area

The study area was analysed together with its immediate built and environmental surroundings (Figure 3). The three-dimensional urban model was developed in SketchUp 2026 through the ENVI-met/INX workflow and subsequently configured for ENVI-met 5.9 Science (Figure 5 and

Table 5. Main ENVI-met model configuration, forcing conditions and parametrisation settings.

Category	Parameter	Setting/Criterion Used in the Simulation
	Software version	ENVI-met 5.9 Science; baseline scenario S_0 representing the existing condition.
Software and scenario	Geometry workflow	SketchUp 2026 model exported through the ENVI-met/INX workflow.
	Grid size	251 × 251 × 26 cells.
	Spatial resolution	2.0 m in X, Y and Z directions; equidistant grid.
Computational domain	Physical domain	Approximately 502 m × 502 m × 52 m.
	Maximum building height	Approximately 22 m; vertical domain about 2.4 times the tallest building.
	Domain placement and rotation	Study area centred in the model with surrounding urban buffer; grid rotation 0.0°.
Meteorological forcing	Source	Measured data from the nearby Beasain meteorological station.
	Forcing variables	Air temperature, relative humidity, wind conditions and radiation/cloud conditions; dry-weather simulations without precipitation forcing.
	Simulated days	Selected field-survey dates: 7 February, 24 April, 11 July and 3 October 2024.
Simulation protocol	Simulation duration	10 h, from 08:00 to 18:00.
	Spin-up/adjustment period	08:00–10:00 treated as initial model-adjustment period and excluded from the main comparison.
	Main output times	10:00, 12:00, 14:00 and 16:00.
	Output interval	Receptor and building data every 30 min; other model outputs every 60 min.
	Upper and middle layers	0–20 cm: 20.0 °C and 65% usable field capacity; 20–50 cm: 20.0 °C and 70% usable field capacity.
Initial soil conditions	Deep and bedrock layers	50–200 cm: 19.0 °C and 75% usable field capacity; below 200 cm: 18.0 °C and 75% usable field capacity.
	Standard materials	Default ENVI-met parameters for walls, roofs, asphalt, concrete pavements, loamy soil, grass and standard vegetation elements.
Surface materials	Custom material	Shock-absorbing rubber represented as a custom material with a representative mean albedo value.
Vegetation	Tree species	Planting-plan species assigned to corresponding or closest ENVI-met plant database entries.
BioMet/PET	User profiles	Elderly female sitting/standing and elderly male sitting/standing profiles.
Analysed variables	Microclimatic outputs	Potential air temperature, relative humidity, mean radiant temperature, surface temperature, wind conditions and PET at pedestrian level.

). The model reproduced the main physical elements affecting pedestrian-level microclimatic conditions within the study area, including building volumes, paved surfaces, natural soil areas, grass surfaces, and tree vegetation. In CFD-based urban studies, domain size and the distance between the area of interest and the model boundaries are recognised as critical factors affecting numerical robustness and the interpretation of pedestrian-level conditions [34].

The computational domain was defined as an equidistant grid composed of 251 × 251 × 26 cells, with a uniform spatial resolution of 2.0 m in the X, Y, and Z directions. This resulted in an approximate physical domain of 502 m × 502 m × 52 m. The tallest building represented in the model was approximately 22 m high; therefore, the vertical domain extended to approximately 2.4 times the maximum building height, providing sufficient atmospheric space above the built volumes for near-ground microclimatic simulation.

The Barrendain intervention area was positioned in the central part of the computational domain and surrounded by an urban buffer in order to reduce the influence of lateral boundary conditions on the analysis area. The grid rotation angle was set to 0.0°. The use of an equidistant grid ensured a constant spatial resolution across the model and facilitated direct comparison among the different simulation outputs.

2.6.3. Material and Vegetation Parameterisation

Material and vegetation parameterisation was based on the correspondence between the physical elements observed in Barrendain Square and the closest available entries in the ENVI-met database (

Table 6. Material and vegetation parameterisation used in the ENVI-met model.

Layer	Component	Material/Vegetation	ENVI-Met Material Code	Source/Criterion	Parameterisation
	Wall	Default wall	0100 MI	ENVI-met database	Default thermophysical parameters
Buildings	Roof	Default roof	0100 RI	ENVI-met database	Default thermophysical parameters
	Soil 1	Asphalt road	0100 ST	Existing pavement observed in the square	ENVI-met default parameters
	Soil 2	Asphalt road with red coating	0100 RA	Existing red-coated pavement	ENVI-met default or closest available surface class
Artificial pavements	Soil 3	Grey concrete pavement	0100 PG	Existing prefabricated concrete pavement	ENVI-met default parameters
	Soil 4	Dirty concrete pavement	0100 PP	Existing aged/dirty concrete pavement	ENVI-met default parameters
	Soil 5	Shock-absorbing rubber	900000	Specific/custom material layer	Custom material; a representative albedo was assigned to approximate radiative behaviour
	Soil 6	Cellar	CELLAR	Building-ground interface	ENVI-met model setting
	Soil 7	Basalt paving blocks	0100 ST	Existing dark basalt paving blocks observed in the study area	Represented using the ENVI-met default asphalt-road class adopted in the model as a proxy for the existing basalt pavement
Natural soils	Soil	Loamy soil	0100 LO	Existing permeable/green areas	ENVI-met default parameters
	Simple plant	Grass	0100 XX	Existing green areas	ENVI-met default parameters
	Tree	Prunus avium	010800	Planting plan and ENVI-met plant database	ENVI-met plant database parameters
	Tree	Acer pseudoplatanus	020033	Planting plan and ENVI-met plant database	ENVI-met plant database parameters
Vegetation	Tree	Magnolia grandiflora	021041	Planting plan and ENVI-met plant database	ENVI-met plant database parameters
	Tree	Trachycarpus fortunei	01PLDM	Planting plan and ENVI-met plant database	ENVI-met plant database parameters
	Tree	Platanus hispanica	020060	Planting plan and ENVI-met plant database	ENVI-met plant database parameters
	Tree	Ginkgo biloba	020150	Planting plan and ENVI-met plant database	ENVI-met plant database parameters

). Standard materials already included in the database, such as walls, roofs, asphalt, concrete pavements, loamy soil, grass, and standard vegetation elements, retained the default thermophysical or biophysical parameters provided by ENVI-met. This choice ensured consistency between the assigned material codes and the internal properties used by the model.

The shock-absorbing rubber surface present in the square was introduced as a custom material because it was not adequately represented by the standard ENVI-met material classes. For this type of playground safety surfacing, the available technical documentation generally focusses on impact attenuation and safety performance rather than radiative properties such as albedo or solar reflectance index (SRI). In the absence of product-specific optical measurements or declared radiative data, a representative mean albedo value was assigned on the basis of the dominant surface colour in order to approximate its radiative behaviour. This parameterisation is explicitly treated as a source of modelling uncertainty.

Tree vegetation was defined from the official planting plan for Barrendain Square and assigned to the corresponding or closest available entries in the ENVI-met plant database. The model therefore represented vegetation according to the documented botanical composition of the site, rather than as a generic green layer. Table 6 summarises the material codes, sources, and parameterisation criteria adopted in the ENVI-met model.

2.6.4. Climatic Forcing and Simulation Protocol

Meteorological input data were derived from records of the Beasain meteorological station. The station dataset was processed into a dedicated CSV file for the selected field-survey dates and imported into ENVI-met through the Advanced Meteorology module as a full-forcing meteorological file. Before running the simulations, ENVI-guide was used to verify that the forcing file covered the complete simulation period.

For each selected survey day, a separate SIMX file was generated. All simulations followed the same 10 h protocol, from 08:00 to 18:00. The first two hours, from 08:00 to 10:00, were treated as an initial model-adjustment period and were not used for the main comparison. Outputs were extracted at 10:00, 12:00, 14:00 and 16:00 to characterise the daytime evolution of the simulated conditions. The principal interpretation focused on the 12:00 output, which corresponded to the main midday survey slot.

2.6.5. Model Validation and Uncertainty

The ENVI-met simulations were developed to support a diagnostic interpretation of the existing microclimatic conditions of the study area, rather than to provide a stand-alone prediction of absolute pedestrian-level thermal values. Accordingly, the model outputs were interpreted as spatial indicators of relative microclimatic behaviour and were analysed together with the field-survey evidence on thermal perception, reported public-space use, and perceived sociability.

Measured meteorological data from the Beasain station were used to define the atmospheric forcing conditions of the simulations, including air temperature, relative humidity, wind conditions, and radiation/cloud conditions. These data ensured that the simulated scenarios were constrained by observed local meteorological conditions during the selected survey periods. In addition, the ENVI-met/INX inspection procedure was used to verify the technical consistency of the geometric model before the simulations were launched. However, station-based meteorological forcing and technical model inspection should be distinguished from full quantitative validation of pedestrian-level microclimatic fields within the study area. A full quantitative validation would require continuous on-site measurements collected within Barrendain intervention area at comparable spatial locations and time steps to the ENVI-met outputs.

This distinction between model setup verification and full quantitative validation is consistent with previous studies highlighting that urban microclimate simulations are highly sensitive to boundary conditions, material properties, vegetation parameterisation, and radiative exchanges, which can complicate their interpretation in complex urban environments [57,58,59].

For this reason, PET maps and related microclimatic outputs are not interpreted as absolute validated measurements, but as comparative spatial estimates supporting the identification of exposed and protected areas, relative PET patterns, and microclimatic tendencies within the baseline condition of the Barrendain intervention area. The simulation outputs were therefore interpreted alongside the empirical survey data on user perception and behaviour, not as a validation procedure, but as part of an integrated framework for relating modelled thermal exposure to observed patterns of use and subjective thermal responses. In this sense, the simulations were primarily intended to identify relative spatial and seasonal differences in thermal exposure rather than to provide fully calibrated predictions of absolute pedestrian-level thermal conditions. This methodological scope was considered when interpreting the simulation results and is further addressed in the discussion and limitations sections.

2.6.6. Simulation Outputs and PET Assessment

Once the geometry had been completed, the model was exported in .INX format, which is the input file required for ENVI-met simulation.

Thermal comfort was assessed through the BIO-met module using the physiological equivalent temperature (PET) index. PET was selected because it allows for user-specific thermal comfort assessment through the explicit definition of personal parameters, while BIO-met enables the post-processing of ENVI-met outputs for biometeorological evaluation [60].

To account for possible differences in thermal perception according to sex and posture, four customised elderly user profiles were defined in BIO-met: elderly female standing, elderly female sitting, elderly male standing, and elderly male sitting. The female profile was set at 75 years of age, 1.63 m in height, and 66 kg in body weight, while the male profile was set at 75 years, 1.70 m in height, and 76 kg in body weight. These values were manually assigned as representative elderly user scenarios for the simulations and were not intended to reproduce the exact anthropometric distribution of the local population [61,62].

Model outputs included air temperature, mean radiant temperature (MRT), wind speed, relative humidity, and PET distributions at pedestrian level. These variables were subsequently analysed in relation to field observations and users’ subjective thermal responses.

For the comparison between simulation-derived PET and participants’ reported thermal sensation (TSVs), four field-survey days were selected for simulation: 7 February, 24 April, 11 July, and 3 October 2024. For each of these days, PET derived from the corresponding ENVI-met/BIO-met simulation was compared exclusively with TSV responses collected on that same field-survey day. This day-matched comparison should be distinguished from the overall TSV analysis and neutral-temperature calculations, which included responses collected across the complete seasonal campaigns. Representative mean PET values were calculated using all simulated PET values within the intervention area at 12:00 and 14:00, corresponding to the main period of questionnaire administration. The resulting mean PET values and the corresponding same-day TSV responses were used for the comparison presented in Section 3.5.4.

Overall, the adopted modelling and post-processing workflow enabled the analysis of the spatial and seasonal distribution of PET within the study area, the comparison of modelled thermal conditions with users’ subjective thermal responses collected during the field surveys, and the identification of relatively exposed and thermally moderated areas within the study area to support the ongoing redesign process of Barrendain Square.

2.7. Thermal Comfort Assessment

Outdoor thermal comfort is commonly assessed through composite indices that integrate the effects of meteorological variables on the human heat balance. In the present study, thermal conditions were evaluated using the physiological equivalent temperature (PET), a widely used biometeorological index that expresses thermal conditions as an equivalent air temperature in degrees Celsius. This facilitates the interpretation and comparison in urban-climate and design-related studies [35].

PET combines the effects of air temperature, relative humidity, wind speed, and mean radiant temperature, together with personal parameters such as clothing insulation and metabolic activity. Because radiant exchange plays a major role in outdoor thermal perception, PET is particularly suitable for analysing the effects of urban morphology, shading, vegetation, and surface materials in open spaces [35,63].

PET values were interpreted according to the thermal stress classification proposed in [35]. According to this classification, PET values between 18 and 23 °C correspond to no thermal stress, values between 4–8 °C indicate extreme cold stress, 8–13 °C moderate cold stress, 13–18 °C slight cold stress, 23–29 °C indicate slight heat stress, 29–35 °C moderate heat stress, 35–41 °C strong heat stress whereas values from above 41 °C represent extreme heat stress. Conversely, PET values below 18 °C indicate different levels of cold stress, ranging from slight cold stress (13–18 °C) to extreme cold stress (<4 °C). This classification was subsequently used to compare simulated thermal conditions with participants’ reported thermal sensation (TSVs). Within the range of environmental conditions observed in this study, PET values ranged from slight cold stress during the winter and spring campaigns to no thermal stress in autumn and moderate heat stress during summer. These categories formed the basis for the comparison between simulated thermal conditions and subjective thermal sensation presented in the results section.

The use of PET in ageing populations requires careful interpretation, since thermal sensitivity and thermoregulatory responses may differ from those of younger groups [64]. Recent studies suggest that PET can be successfully applied to ageing populations, although its interpretation should account for season, local climate, and microclimatic variability [29,65].

2.8. AI Use Statement

During the preparation of this manuscript, the authors used an artificial intelligence language model (ChatGPT, version GPT-5.5, OpenAI) to assist with English language editing, text clarification, and improvement of the manuscript structure. The tool was also used to support the categorisation and description of qualitative survey responses, as well as minor enhancements to the visual clarity of some figures. No images, figures, or graphical materials were generated using artificial intelligence. All scientific content, data analysis, interpretations, and conclusions were developed and verified by the authors.

3. Results

3.1. Sample Characteristics

A total of 117 valid questionnaires were obtained (54 men and 63 women) (

Table 7. Distribution of the surveys conducted by time slot and sex.

Time Slot	Total of Responses	Men	Women
9:00–10:00	27	12	15
12:00–13:00	50	23	27
16:00–17:00	40	19	21
TOTAL	117	54	63

). Most surveys were conducted between 12:00 and 13:00, with slightly higher participation among women across all time slots. Of the total sample, 53 participants were accompanied during the survey (14 men and 39 women) while 21 reported living alone (8 men and 13 women); among them 11 were accompanied at the time of the survey (4 men and 7 women). The sample included participants across a broad range of ages, both sexes and four seasonal campaigns, providing a broad representation of older adults using the intervention area under different environmental conditions.

Most respondents (76.9%) were engaged in light standing activities (48.8% men and 51.2% women), while the remainder were seated. Seated participants accounted for approximately 25% of responses in winter and autumn, compared with around 11% in summer and spring.

The highest number of responses was collected in winter, followed by spring and autumn, while summer showed the lowest participation. Across all seasons, individuals aged 65–74 years constituted the most represented age group. Participation gradually decreased with increasing age, particularly among respondents aged 85 years and over (Figure 6).

Clothing insulation levels showed clear seasonal variation (Figure 7). In winter, most participants (94%) reported the highest insulation level (1.5 clo), with a similar distribution between men and women. In summer, lower insulation levels were more frequent, particularly among men, while women showed greater variability across age groups, reflecting greater variability in clothing behaviour. In spring, high clothing insulation levels (1.5 clo) predominated across all groups, whereas in autumn, clothing insulation became more variable, especially among women, while men tended to concentrate around moderate values.

3.2. Thermal Perception and Adaptive Responses

3.2.1. Thermal Sensation

Thermal sensation responses (TSVs) were analysed by season to examine variations across sex and age groups. Seasonal differences were observed in the distribution of TSV responses, with sex-related differences being more evident during winter and the transitional seasons than during summer (Figure 8 and Figure 9). Overall, women tended to report cooler thermal sensations than men under comparable seasonal conditions.

In winter, the strongest sex contrast was observed. Among women, 41% reported feeling cool or cold (9 out of 22), compared to 24% of men (6 out of 25). Men more frequently reported neutral or warm conditions (76%) than women (59%). Across age groups, women’s responses showed greater variability, whereas men’s responses were more strongly concentrated around neutral thermal sensations.

In spring, sex-related differences became less pronounced. A majority of men (62%) reported neutral conditions, while women’s responses were more dispersed with 50% reporting cold or cool sensations, while 43% reported neutral conditions.

In summer, responses were concentrated around neutral and warm sensations for both groups. Among women, 80% reported neutral conditions and 20% warm, while men showed a similar pattern (71% neutral, 29% warm). Differences between age groups and sexes were comparatively small during summer.

In autumn, variability increased again. Although neutral thermal sensation dominated, women showed a wider distribution of responses, with 25% of women reporting cool sensations, whereas all male respondents reported neutral conditions. This greater variability among women was observed across several age groups, while men’s responses remained highly concentrated around neutral thermal sensations.

3.2.2. Thermal Comfort Assessment

Thermal comfort responses varied seasonally, increasing from winter to summer and declining slightly in autumn (Figure 10).

In winter, 81% of responses fell within the comfortable or quite comfortable categories, while 19% reported some degree of discomfort. In spring, 70% of responses were comfortable, with 30% indicating reduced comfort. Thermal comfort peaked in summer, when 100% of responses were classified as comfortable or quite comfortable. In autumn, comfort remained high (80%), although 20% of responses reflected lower comfort levels, indicating greater variability than during summer and winter.

3.2.3. Thermal Preferences

Thermal preferences (TP) were analysed by season and sex (Figure 11), revealing marked seasonal differences and sex-related variations.

In winter, preferences were more dispersed. While 45% of respondents reported no change, 28% preferred warmer conditions and 26% preferred cooler or much cooler environments. This result is notable given the relatively low temperatures during the survey period (mean 11.27 °C, maximum 16 °C, minimum 3.3 °C). Sex-related differences were evident: 50% of women reported no change and 32% preferred warmer or much warmer conditions, whereas men showed a more dispersed distribution (40% no change) and a greater tendency to prefer cooler conditions despite the winter context.

In spring, preferences shifted towards warmer conditions, with 59% of respondents indicating warmer or much warmer environments and 41% reporting no change. No respondents preferred cooler conditions. Sex-related differences were limited. However, only women reported a preference for much warmer conditions, whereas men’s responses were confined to the warmer or no-change categories.

In summer, preferences became more homogeneous. A majority of respondents (59%) reported no change, while 41% preferred warmer or much warmer conditions. No respondents expressed a preference for cooler environments, despite relatively high temperatures (mean 20.81 °C, maximum 28 °C, minimum 14 °C). The number of surveys collected during summer was lower than in other seasons, due to reduced use of the space. Sex-related differences were minimal, with similar patterns for both groups, although women accounted for the few cases reporting a preference for much warmer conditions.

In autumn, preferences became more heterogeneous again. While 40% of respondents reported no change, 44% preferred warmer conditions and 16% cooler or much cooler environments, reflecting increased variability during the transitional season. Sex-related differences were more evident than in summer: women showed a more dispersed pattern, whereas men tended to concentrate their responses around no change and warmer conditions.

Overall, thermal preferences varied seasonally, with more heterogeneous responses during winter and autumn, a predominance of warmer preferences in spring, and a greater proportion of no-change responses during summer.

3.2.4. Air Temperature, TSVs and Neutral Temperature

To further explore the association between objective environmental conditions and subjective thermal perception, the relationship between air temperature and TSVs was analysed in order to estimate season-specific neutral temperatures. In both campaigns, regression of the 1 °C-binned MTSV values showed the expected positive relationship between Ta and TSVs (

Table 8. Regression results for the campaign-based Ta-MTSV models.

Campaign	n	Regression Equation	R2	Tn (°C)	Model-Derived Limits (°C)
Cold-period	61	MTSV = 0.086Ta − 1.060	0.61	12.35	0.70 to 24.00
Warm-period	56	MTSV = 0.055Ta − 1.098	0.39	19.87	1.77 to 37.96

The model-derived limits are shown for completeness but should not be read as directly observed comfort ranges.

). The cold-period campaign exhibited a steeper slope and better fit than the warm-period campaign, indicating a more consistent Ta signal under colder conditions.

The estimated neutral temperatures were 12.35 °C for the cold-period campaign and 19.87 °C for the warm-period campaign. These values should be interpreted as campaign-specific adaptive benchmarks rather than universal seasonal constants (Figure 12).

Additional boxplots were developed to visualise the distribution and variability of TSV responses across air-temperature intervals in both campaigns (Figure 13). The results revealed considerable dispersion in subjective responses, particularly during transitional and colder conditions, highlighting the variability of subjective thermal responses under outdoor conditions. Variability was generally lower under warmer conditions, where TSV responses tended to cluster around neutral values. Auxiliary analyses indicated that clothing was the most consistent adaptive variable in both campaigns, with significant differences in both Ta and TSVs across clothing categories. Sun exposure was associated with higher TSV values, while relative humidity showed negative monotonic relationships. In contrast, sex, age, and body-build category did not show robust effects within the present sample (

Table 9. Summary of non-parametric analyses for selected personal and contextual variables.

Variable	Cold-Period Campaign	Warm-Period Campaign	Interpretation
Clothing	TSV: Kruskal–Wallis p = 0.0089; Ta: p < 0.001	TSV: p < 0.001; Ta: p < 0.001	Strongest adaptive marker in both campaigns
Sex	Mann–Whitney p = 0.706	p = 0.563	No statistically detectable TSV difference in this sample
Age	Spearman rho = 0.091, p = 0.486	rho = −0.091, p = 0.502	No monotonic age–TSV relationship detected
Body-build category	Kruskal–Wallis p = 0.119	p = 0.135	Exploratory only; no robust effect
Sun exposure	Mann–Whitney p = 0.0039	p = 0.0423	Higher TSVs under sun exposure
Relative humidity	Spearman rho = −0.507, p < 0.001	rho = −0.339, p = 0.010	Negative association, likely reflecting broader microclimatic co-variation

).

3.3. Multisensory Perception

In addition to thermal perception, participants evaluated the acoustic and visual qualities of the space and provided suggestions for spatial improvement. These responses complement the thermal assessment by describing additional environmental perceptions and reported user preferences within the study area.

3.3.1. Acoustic Perception and Preferences

Acoustic perception and preferences were analysed by sex (Figure 14). Both groups mainly reported intermediate acoustic evaluations, although men showed a stronger preference for quieter conditions.

Among women, the predominant perception was pleasant (52%), followed by unpleasant (43%), with marginal extreme evaluations (≈5%). Among men, unpleasant evaluations predominated (50%) over pleasant ones (37%), with slightly higher extreme responses (≈13%).

Regarding acoustic preferences, the dominant response among women was no change (53%), followed by slightly quieter conditions (31%). In contrast, men showed a stronger preference for quieter environments, with 37% selecting quieter and 31% slightly quieter, while 30% reported no change.

In terms of predominant perceived sounds, traffic-related noise was the most frequently reported category among men (59%), including traffic combined with train noise (31%), traffic alone (17%) and train-only noise (11%). Human activity sounds accounted for 18% of responses, while 9% reported no dominant sound.

Among women, responses were more heterogeneous. Traffic-related sounds were also the most frequent category (31%), although a higher proportion reported no dominant sound (21%) or did not specify any particular sound (22%).

Overall, traffic-related noise constituted the predominant perceived sound source for both sexes. Men tended to evaluate the acoustic environment more negatively and showed a stronger preference for quieter conditions, whereas women displayed a more heterogeneous perception and a higher proportion of no-change responses.

3.3.2. Visual Perception of Colour

Slightly more than half of the respondents (52%) reported that there was a colour in the space that they particularly liked, while 47% answered negatively, reporting no specific preference. Among affirmative responses, references to green and vegetation-related elements clearly predominated, followed by floral colours, whereas references to built elements and the sky were comparatively infrequent. Among those respondents who answered negatively, the most common justification was the absence of any specific dominant colour (e.g., none or nothing in particular).

Seasonal differences were more pronounced than sex-related differences. The highest proportion of affirmative responses was observed in autumn (64%), while spring showed the lowest (44%), with a majority reporting no specific preferred colour (56%). In winter, responses were relatively balanced (52% yes; 48% no), and in summer a similar distribution was observed (47% yes; 47% no), with a small proportion of non-responses (6%). Overall, affirmative responses regarding preferred colours were more frequent during the autumn campaign than during the other seasonal campaigns.

Beyond visual perception, participants also identified several aspects of the space that could be improved in order to enhance comfort and usability.

3.3.3. Users’ Suggestions for Spatial Improvement

Participants were invited to identify potential improvements to the space, revealing several recurring concerns related to comfort and usability. Although responses varied between individuals, some sex-related tendencies were observed.

Among men, the most recurrent specific response was that no improvement was needed (16%), followed by requests for more seating (13%). Among women, responses were more heterogeneous, although the most frequent suggestion was the addition of more seating areas (20%), followed by weather protection measures such as covered or sheltered areas (15%).

Overall, requests for additional seating and weather protection were among the most frequently reported suggestions for improving the study area.

3.4. Reported Public-Space and Perceived Sociability

Patterns of reported public-space use and perceived sociability were analysed alongside the environmental conditions recorded during the field campaigns.

Table 10. Primary activities in the space by sex: survey results.

	Talking	Eating or Drinking	Sport	Resting	Meeting People	Sunbathing	People-Watching	Other Activities
Women	26	20	30	11	16	11	7	18
Men	16	18	34	11	15	11	8	9
TOTAL	42	38	64	22	31	22	15	27

summarises the main activities reported by respondents, providing an overview of how the intervention area was used by older adults under different seasonal conditions.

Seasonal attendance was consistently high, with nearly 95% of respondents reporting visits to the study area in winter and 93% in summer, indicating sustained year-round use of the space despite seasonal climatic differences.

Regarding primary activities, 55% of respondents visited the space for physical exercise (26% women and 29% men), 36% for social encounters (22% women and 14% men) and 32% to eat or drink (17% women and 15% men). These results showed that the intervention area supported a variety of reported activities, including physical exercise, social encounters and eating or drinking, across all survey campaigns.

Perceived sociability was notably high, with 93% of respondents indicating that interaction with other users was easy. All men (100%) reported ease of interaction, compared with 89% of women. For 64% of respondents, this perception was mainly attributed to the presence of familiar people or the social atmosphere, whereas physical characteristics of the space, such as spaciousness or environmental pleasantness, were mentioned less frequently.

Table 10 presents the distribution of the main reported activities by sex. Questions concerning activities allowed multiple responses; therefore, percentages exceed 100%.

3.5. Simulation Results

To interpret the simulated thermal behaviour of the site, the analysis focused on representative PET maps for two contrasting seasonal scenarios—winter (7 February 2024) and summer (11 July 2024)—both at 12:00, selected as a representative hour within the main survey slot (12:00–13:00), during which the largest number of questionnaires was completed (Figure 15, Figure 16, Figure 17, Figure 18, Figure 19 and Figure 20).

The figures provide a spatial comparison of contrasting winter and summer midday conditions, while the full seasonal dataset is presented in Appendix A (Figure A1, Figure A2, Figure A3, Figure A4, Figure A5, Figure A6, Figure A7 and Figure A8). The simulations revealed marked spatial differences in PET across the wider simulated domain, while the Barrendain intervention area showed clear internal microclimatic variation.

Table 11. Maximum and minimum PET values (°C) simulated within the Barrendain intervention area for four elderly user profiles across the dates and output times. Cell colours indicate the corresponding PET physiological-stress categories according to [35].

Campaign	Time	Sitting				Standing
		Men		Woman		Men		Woman
		Min	Max	Min	Max	Min	Max	Min	Max
7 February	10:00	15.00	32.5	15.43	32.75	14.89	31.84	15.32	32.09
	12:00	16.2	21.00	16.37	21.42	16.45	20.76	17.62	21.18
	14:00	15.00	17.55	15.16	18.2	15.00	17.12	15.16	17.57
	16:00	14.00	17.00	14.14	17.49	13.85	16.35	14.00	16.84
24 April	10:00	16.23	23.56	16.48	24.00	16.19	22.37	16.43	23.2
	12:00	15.1	19.54	15.27	20.00	15.04	18.98	15.21	19.26
	14:00	15.43	19.55	15.8	20.00	15.55	18.72	15.72	19.16
	16:00	16.00	22.08	16.17	22.2	16.06	21.03	16.24	21.83
11 July	10:00	31.49	59.00	31.89	59.1	31.36	58.45	31.75	58.55
	12:00	32.07	42.36	32.45	42.48	32.00	41.89	32.37	42.00
	14:00	27.92	32.81	28.29	33.25	27.99	32.5	28.37	32.93
	16:00	23.1	25.58	23.5	26.00	23.00	25.81	23.5	26.00
3 October	10:00	19.55	25.58	20.00	26.00	19.38	25.02	19.81	25.42
	12:00	18.61	23.55	19.00	24.00	18.56	22.92	18.86	23.36
	14:00	16.69	21.55	17.00	22.00	16.56	20.94	16.86	21.37
	16:00	14.83	20.54	15.00	21.00	14.7	19.96	14.86	20.4
				PET physiological stress category
				Slight cold stress 13 ≤ PET < 18 °C
				No thermal stress 18 ≤ PET < 23 °C
				Slight heat stress 23 ≤ PET < 29 °C
				Moderate heat stress 29 ≤ PET < 35 °C
				Strong heat stress 35 ≤ PET < 41 °C
				Extreme heat stress PET ≥ 41 °C

summarises the minimum and maximum PET values obtained for the four elderly user profiles—men sitting, women sitting, men standing and women standing-across the four simulated dates and output times (10:00, 12:00, 14:00 and 16:00). These values represent the spatial extrema recorded within the intervention area for each simulation scenario. Rather than describing every simulated condition individually, the following analysis focuses on the most representative thermal patterns observed at 12:00, while the 14:00 outputs are also considered to examine the subsequent daytime evolution of thermal conditions.

Seasonal, temporal and spatial variation represented the main sources of PET variability, whereas differences associated with sex and posture were generally modest in comparison. Winter and spring conditions were predominantly characterised by slight cold stress, autumn by no thermal stress, and summer by conditions ranging from slight to extreme heat stress, depending on the output time and spatial exposure.

Across the complete set of output times presented in Table 11, the simulations revealed that the highest maximum PET values were consistently recorded at 10:00 on all four simulated dates. This pattern reflected the specific interaction between solar exposure, vegetation, and urban morphology within the study area. The highest values occurred in fully exposed paved areas, where high radiative exposure contributed to the greatest levels of physiological thermal stress.

As the day progressed, changes in the shade cast by existing trees and surrounding urban elements contributed to lower PET values in parts of the intervention area. This reduction highlights the dominant role of shading in attenuating short-wave radiation and moderating pedestrian-level thermal exposure.

The minimum and maximum PET values presented in Table 11 describe the overall range of simulated thermal conditions within the intervention area but do not indicate the proportion of the study area corresponding to each physiological-stress category. Accordingly, these values were not used to quantify or isolate the influence of vegetation, pavement, or other individual urban elements. Spatial relationships between PET patterns and site characteristics were instead interpreted qualitatively from the corresponding maps and discussed in relation to the combined effects of solar exposure, shading, surface properties, vegetation, and surrounding urban morphology.

3.5.1. Winter Scenario

In the winter reference case (7 February 2024, 12:00), simulated PET values remained comparatively moderate across all elderly user profiles, ranging from 16.20 °C to 21.42 °C within the study area (Figure 15 and Figure 16). These results indicated that even under winter midday conditions the public space does not behave as a uniform thermal environment. Warmer areas tended to concentrate in the most exposed portions of the site, whereas lower PET values were generally found in more protected areas.

The simulated microclimatic fields helped explain the observed winter PET distribution (Figure 17). Mean radiant temperature (MRT) ranged from 8.46 °C to 26.34 °C, while potential air temperature varied only between 15.03 °C and 16.48 °C, relative humidity ranged between 43.73% and 51.94%, and wind speed ranged from 0.01 to 2.25 m/s across the computational domain. The substantially wider spatial variation in MRT compared with potential air temperature, together with the visual correspondence between lower-PET and lower-MRT areas, suggests that local differences in solar exposure and shading were the main factors associated with the spatial differentiation of winter PET.

3.5.2. Summer Scenario

A markedly different pattern emerged in the summer reference case (11 July 2024, 12:00). Across all user profiles, simulated PET values within the study area ranged from 32.00 °C to 42.48 °C (Figure 18 and Figure 19), indicating substantially higher levels of thermal stress than those observed during winter. The seasonal contrast is therefore substantial: the minimum PET values in summer exceeded the maximum PET values in winter, showing that the difference between the two dates is not limited to isolated hotspots but involves an upward shift of the entire thermal field.

Under summer conditions, high PET values extend across large portions of the modelled area, indicating much broader and more intense thermal stress. At the same time, the maps reveal a recurring localised mitigation effect within the analysed space. The central elongated area of the intervention area, particularly on its right-hand side, tended to maintain lower PET values than the surrounding exposed surfaces.

This behaviour was visible not only in the selected winter and summer figures, but also in the full set of seasonal simulations included in Appendix A. Although the simulated domain extends beyond the study site itself, this central tree-covered portion of the intervention area showed a consistent microclimatic signature that distinguishes it from the more exposed adjacent areas. This behaviour indicated the presence of a recurrent thermally moderated area within the study area.

This part of the square is characterised by a relatively regular and closely spaced arrangement of deciduous trees. During summer, the tree canopy generated a more continuous shaded structure that reduced radiative exposure and contributes to lower PET values. Although a similar pattern was also visible during winter, the effect was less pronounced under the corresponding seasonal radiative conditions. The visual comparison between the PET and MRT maps shows that this lower-PET area broadly coincided with a reduction in mean radiant temperature. This spatial correspondence supports the interpretation that the mitigation effect was primarily related to reduced radiative exposure generated by the tree-canopy configuration.

The simulated microclimatic fields also help explain the summer PET distribution (Figure 20). MRT ranged from 23.65 °C to 56.86 °C, while potential air temperature varied from 29.65 °C to 32.75 °C, relative humidity ranged from 52.41% to 63.73%, and wind speed ranged from 0.00 to 1.25 m/s across the computational domain. Compared with the winter scenario, the summer condition was characterised by a substantially greater radiative load, higher air temperature and relative humidity, and generally lower wind speeds. These conditions contributed to the overall upward shift in PET. However, the spatial configuration of the PET field corresponded most clearly to the distribution of MRT, indicating that radiative exposure was the principal factor associated with local PET differentiation, whereas air temperature, humidity and wind speed acted as additional modifying variables. In this sense, the lower PET values observed in the central tree-covered zone are most plausibly related to the reduction of radiative exposure produced by the canopy configuration. These relationships represent qualitative spatial associations rather than a formal statistical correlation or sensitivity analysis.

3.5.3. Spatial PET Patterns

Across the full seasonal dataset, two additional tendencies remained consistent. First, the older-woman profile systematically returned slightly higher PET values than the corresponding older-man profile. Second, the difference between sitting and standing conditions remained relatively small. These differences are detectable, but they are clearly secondary when compared with the much stronger influence of season, solar exposure, and local shade continuity. This indicated that, in the case study examined here, outdoor thermal stress is shaped more strongly by the spatial organisation of shading and exposure than by the modest differences associated with posture or sex-specific physiological parameters alone.

The simulations further indicate that spatial and temporal variations in solar exposure were more strongly associated with PET patterns than user-profile characteristics. The recurrent reduction of PET values observed in the tree-covered areas of the square indicates the capacity of vegetation to generate localised microclimatic conditions that differ substantially from those of surrounding exposed surfaces.

From a design perspective, these results highlight the importance of vegetation structure, shade continuity, and microclimatic diversity as key variables for improving outdoor thermal conditions in public spaces used by older adults. Rather than providing uniform thermal environments, public spaces may benefit from offering a range of sun-exposed and shaded areas that allow users to adapt their location according to seasonal conditions and individual thermal preferences.

3.5.4. Comparison Between Simulated PET and Subjective Thermal Responses

To explore the relationship between simulated thermal conditions and users’ subjective thermal perceptions, PET values derived from ENVI-met/BIO-met simulations were compared with the thermal sensation votes (TSVs) obtained during the field surveys (

Table 12. Comparative interpretation of simulated PET conditions and observed thermal sensation responses during the main survey period (12:00–14:00). Mean PET values correspond to the arithmetic mean of all simulated PET values within the intervention area at 12:00 and 14:00 (main survey period).

Campaign	Mean PET (°C)	PET Category	Predominant TSV	Level of Agreement	Main Interpretation
7 February	17.6	Slight cold stress	+1/−1	Moderate	Partial correspondence between objective and subjective thermal conditions.
24 April	17.4	Slight cold stress	0/−1	Low	Neutral or slightly cool sensations despite PET values indicating slight cold stress suggest seasonal and behavioural adaptation.
11 July	34.0	Moderate heat stress	0	Very low	PET-based heat stress did not translate directly into perceived discomfort.
3 October	20.1	No thermal stress	0	High	The strongest correspondence was observed between PET classification and subjective thermal responses

). This comparison was not intended as a formal validation of the simulation outputs, but rather as an interpretative framework for analysing the degree of correspondence between modelled thermal exposure and perceived thermal conditions among older adults.

The results revealed varying levels of agreement between PET classifications and subjective responses across the different survey campaigns. The strongest correspondence was observed during the autumn campaign, when PET values fell within the no-thermal-stress category and neutral thermal sensations predominated among respondents. Winter conditions showed a moderate level of agreement, with participants reporting both slightly cool and slightly warm sensations despite PET values indicating slight cold stress.

More pronounced discrepancies emerged during spring and summer. In spring, participants predominantly reported neutral or slightly cool thermal sensations even when PET values remained within categories associated with slight cold stress. This pattern suggested the influence of seasonal adaptation and contextual factors beyond the physical thermal environment alone.

The largest divergence was observed during the summer campaign. Simulated PET values corresponded to moderate heat-stress conditions, whereas neutral thermal sensations remained the most frequently reported subjective response.

It should be noted that the observed PET–TSV discrepancy cannot be attributed specifically to the PET index, since it could reflect simulation uncertainty or other factors influencing the relationship between simulated thermal conditions and individual thermal perception.

At the same time, thermal stress indicators do not necessarily translate directly into perceived thermal discomfort. This finding is consistent with previous research highlighting the adaptive nature of outdoor thermal comfort, behavioural adjustments, expectations, previous thermal exposure, activity patterns, and psychological adaptation may all contribute to moderating subjective thermal responses.

Overall, the comparison suggests that simulated PET is effective for identifying relative differences in thermal exposure and seasonal contrasts within the study area. Nevertheless, the observed discrepancies between PET classifications and subjective responses also reflect the complexity of outdoor thermal comfort, in which objective environmental conditions interact with behavioural, psychological, physiological, and contextual adaptation processes [23,28,49]. Consequently, integrating simulation-based indicators with field-based perception data provides a more comprehensive assessment of public-space performance than either approach alone. This finding highlights the importance of complementing objective microclimatic indicators with user-centred assessments when evaluating outdoor environmental comfort among older adults.

3.5.5. Comparative Interpretation

The combined analysis of PET simulations and subjective thermal responses provided a more comprehensive understanding of outdoor comfort conditions in the pre-intervention configuration. While the simulations identified substantial seasonal and spatial variations in thermal exposure, the survey results revealed that users’ perceptions did not always correspond directly to the levels of thermal stress indicated by PET.

The simulations suggest that the wider modelled context contains markedly differentiated thermal conditions, while the pre-intervention configuration itself includes both thermally exposed and thermally protected microenvironments. The most critical conditions occurred during the warm season and in highly exposed paved areas, whereas the central tree-covered section consistently presented lower PET values. This area therefore appears to function as a local microclimatic refuge within the square, reducing radiative exposure and moderating thermal stress.

Taken together, these findings suggest that the thermal performance and usability of the pre-intervention urban area depend not only on seasonal climatic conditions but also on the availability of localised microclimatic niches generated by vegetation, shade continuity, and reduced radiative load. For older adults, these thermally moderated areas may offer more favourable conditions for longer stays and greater environmental comfort, while also creating more favourable conditions for routine social interaction and everyday use of public space. Although sociability depends on multiple environmental and social factors, the results indicated that microclimatic quality constitutes an important enabling condition for sustained public-space use throughout the year.

4. Discussion

4.1. Outdoor Thermal Comfort and Ageing

The findings indicate context-dependent relationships between microclimatic conditions, thermal perception, and reported public-space use among older adults. In the case of the Barrendain intervention area, these findings informed the ongoing redesign process aimed at improving environmental comfort and opportunities for everyday encounters.

The survey results provide insight into how older adults perceive and use the space, revealing key patterns of social interaction, seasonal attendance, and user preferences. This study supports previous evidence suggesting that older adults, similarly to the general population, tend to tolerate outdoor thermal conditions relatively well compared with indoor environments [22,24]. Despite relatively low winter temperatures, 81% of respondents classified thermal conditions as comfortable or quite comfortable, suggesting a significant degree of seasonal adaptation to outdoor environments among older adults. This finding suggests that thermal preference was not determined solely by the measured environmental conditions but was also influenced by seasonal expectations and behavioural adaptation, consistent with adaptive thermal comfort theory [47].

The results also support previous findings indicating that thermal sensitivity decreases with age [22,25]. In winter, the coexistence of preferences for both warmer and cooler conditions suggests that responses may be influenced not only by actual thermal conditions but also by seasonal expectations.

Overall, the findings are consistent with previous studies conducted in Mediterranean, continental and East Asian climates, which report that older adults generally exhibit greater behavioural adaptation and a preference for warmer outdoor conditions than younger populations [23,25,26]. However, the temperate Atlantic climate of the present study resulted in comparatively limited cold stress during winter and moderate heat stress during summer, contrasting with the more extreme seasonal conditions reported in continental and subtropical climates. These differences highlight the importance of considering local climatic context when interpreting outdoor thermal comfort and translating findings into urban design strategies.

These behavioural responses may also reflect physiological changes associated with ageing. Reduced thermoregulatory efficiency, lower metabolic activity and altered thermal sensitivity have been widely reported among older adults and may contribute to the greater variability observed in thermal preferences and adaptive strategies, particularly among women and during transitional seasons. These findings are consistent with previous studies reporting a preference for warmer outdoor environments and greater reliance on behavioural adaptation in older populations [23,66].

This is also reflected in adaptive behaviour, as reduced thermal sensitivity may be associated with higher levels of clothing insulation, particularly in winter and spring. At the same time, women appear to adopt more diverse adaptive strategies, especially in summer and autumn, adjusting clothing more flexibly to achieve thermal comfort [27].

Beyond these perceptual aspects, the analysis of thermal conditions provides further insight into how environmental factors are associated with user experience. The integrated analysis of air temperature, thermal sensation (TSVs), and adaptive responses further supports an adaptive interpretation of outdoor comfort, showing that subjective responses are shaped by both environmental conditions and behavioural factors. Air temperature was clearly relevant, but it was not sufficient on its own to account for the observed spread in subjective responses.

The difference between the cold-period and warm-period neutral temperatures (12.35 °C and 19.87 °C, respectively) is consistent with adaptive thermal comfort theory. The findings suggest that the same air temperature may be perceived differently depending on seasonal expectations, behavioural adaptation, and environmental context [32,33,47,48,49,56]. In older adults, this adaptive process may also be influenced by habits, route choice, rest patterns, and the timing of outdoor activities [18,22,65,67].

Clothing functioned as a behavioural adjustment through which participants adapted to outdoor conditions [33]. The repeated association between sun exposure and higher TSVs also reinforces the importance of radiative context in open urban spaces [15,36,37]. No statistically detectable effects of sex or age were found within the available sample.

The comparison between PET simulations and TSV responses revealed varying levels of correspondence across the different survey campaigns. The strongest agreement was observed during autumn (3 October), when PET values fell within the no-thermal-stress category and neutral thermal sensations predominated among respondents. In contrast, the weakest correspondence occurred during summer (11 July), when PET values indicated moderate heat stress while neutral thermal sensations remained the most frequently reported response. Winter and spring showed intermediate situations, with subjective responses displaying greater variability than would be expected from PET classifications alone.

These findings reinforce the adaptive interpretation of outdoor thermal comfort and suggest that thermal perception among older adults is influenced not only by the physical environment but also by behavioural adaptation, expectations, previous thermal exposure, and contextual factors. Consequently, PET appears to be particularly useful for identifying relative differences in thermal exposure and seasonal contrasts, rather than predicting subjective comfort responses directly.

The partial correspondence observed between PET and TSVs may reflect both psychological and physiological adaptation processes. Previous research has shown that thermal perception in outdoor environments is shaped not only by environmental conditions but also by expectations, previous experiences, place-related factors, and contextual appraisal [28,49]. Among older adults, age-related changes in thermoregulation and thermal sensitivity may further influence the relationship between objective thermal exposure and perceived comfort [23].

4.2. Microclimatic Performance of the Pre-Intervention Area

While a quantitative validation of the ENVI-met model was outside the scope of this study, its outputs provide useful comparative evidence for interpreting the spatial distribution of thermal exposure within the square. The PET simulations for the summer scenario identify high levels of thermal stress in the sun-exposed paved areas of the study area. This pattern is consistent with the lower survey participation recorded during the summer campaign.

Furthermore, the model’s identification of these thermally critical zones is coherent with users’ subjective feedback, where the most frequent suggestions for improvement included the addition of more seating. Therefore, the simulation should be understood as a complementary analytical tool that helps relate spatial thermal exposure to observed use patterns and perceived environmental needs.

Although these results highlight how individuals adapt to environmental conditions, it is also necessary to understand how such conditions are spatially distributed within the site. The PET simulations complement these findings by identifying spatial and temporal variations in thermal conditions, offering a more detailed understanding of microclimatic performance across the site. The simulations indicate clear spatial differences in PET distribution, with the central tree-covered areas consistently showing lower PET values than adjacent exposed surfaces, particularly during the summer scenario. In contrast, winter scenario simulations showed a much more homogeneous thermal pattern, reflecting the deciduous nature of the existing tree canopy and the greater availability of solar radiation across the site. Consequently, while the trees provide substantial shading and reduce radiant thermal loads during summer, they allow direct solar access during winter. This seasonal behaviour means that the thermally protected areas identified in the summer simulations do not correspond to permanently shaded locations throughout the year. Therefore, the partial correspondence observed between PET and subjective thermal sensation cannot be attributed simply to respondents selecting shaded areas during the winter surveys, but rather reflects the combined influence of seasonal adaptation, age-related physiological responses, and contextual factors.

During the summer scenario, these resulted in a shift from higher thermal stress in exposed areas to lower heat-stress categories beneath the tree canopy. In contrast, during the winter and spring campaigns, PET values generally remained within slight cold-stress or no-thermal-stress categories, reflecting the relatively mild Atlantic climate of the study area. These findings highlight the importance of preserving thermally diverse public spaces that simultaneously provide protection from summer heat while maintaining access to solar radiation during cooler periods.

These spatial differences in thermal conditions are particularly relevant in relation to how older adults occupy, use, and socially experience different areas of the public space.

The results also suggest that thermal heterogeneity may be advantageous in public-space design. Rather than providing uniform environmental conditions, the coexistence of sun-exposed and shaded areas allows users to select locations that best match their thermal preferences and activity patterns. This diversity may be particularly beneficial for older adults, whose thermal needs and adaptive behaviours can vary considerably across seasons and individuals.

4.3. Multisensory Perception, Social Interaction and Public-Space Use

The survey findings indicate that environmental perception extends beyond thermal comfort and includes acoustic, visual and social dimensions that together shape how older adults experience public space. The lower number of surveys collected during summer suggests that higher temperatures may have discouraged use of the space, despite generally acceptable subjective comfort responses. This finding highlights the importance of designing outdoor environments that maintain adequate thermal comfort during warmer conditions in order to support continued use, sociability, and well-being, particularly among older adults who are more vulnerable to heat stress.

In addition to thermal comfort, other environmental dimensions contributed to users’ overall experience of the Barrendain intervention area and provide a broader understanding of how public-space quality is associated with perceived comfort and sociability.

In relation to acoustic perception, although men tended to evaluate the sound environment more negatively and expressed a stronger preference for quieter conditions, noise was not commonly identified as a major barrier to use of the study area. Women reported a higher proportion of neutral or no-change responses, reinforcing the idea that perceived comfort is shaped not only by environmental conditions but also by social and behavioural factors.

The survey results revealed a strong preference for vegetation-related elements, particularly green and floral features, underlining the role of natural components in shaping environmental perception and potentially supporting both comfort and public-space use.

Perceived sociability was notably high, particularly among men, all of whom reported that interaction in the space is easy. In contrast, women more frequently identified the need for improvements related to comfort, suggesting a more critical perception of environmental conditions. These findings point to subtle sex-related differences in how public space is experienced and highlight the importance of incorporating inclusive design strategies. Taken together, the survey findings and the spatial distribution of PET suggest that favourable environmental conditions may facilitate opportunities for everyday use and informal social encounters. However, the present study did not quantify occupancy patterns or length of stay, and therefore these relationships should be interpreted as plausible associations rather than demonstrated causal effects.

Interestingly, respondents more frequently associated sociability with the presence of familiar people and the existing social atmosphere than with the physical characteristics of the urban area itself. Environmental quality may therefore be understood as an enabling condition that supports opportunities for everyday encounters, while existing social networks and community familiarity may play an important role in shaping perceived sociability.

Although participants frequently associated positive perceptions of the study area with familiar people and the existing social atmosphere, the questionnaire included only a limited number of items specifically addressing the social dimension. Consequently, the present study should be understood as providing an exploratory assessment of perceived sociability rather than a comprehensive analysis of social interaction.

4.4. Design Implications

Nevertheless, microclimatic performance alone is not sufficient to ensure an age-friendly and socially supportive public space. Despite its environmental advantages, the current configuration of the Barrendain intervention area reflects its origin as a set of previously disconnected spaces, including a children’s play area and a traffic roundabout separated by vehicular circulation. As a result, the site still presents spatial limitations, including fragmentation, very limited seating, interruptions caused by traffic, and the poor suitability of certain sub-areas. These factors reduce continuity, accessibility, and opportunities for everyday encounters, particularly among older users.

From this perspective, the study area can be considered environmentally favourable but only partially responsive to the social and functional needs of older adults. These findings suggest that favourable thermal conditions alone are insufficient to guarantee continued public-space use when accessibility, seating availability, spatial continuity, and opportunities for everyday encounters remain limited.

The results also indicate that age-friendly public spaces should not aim to provide uniform thermal conditions. Instead, they should incorporate a diversity of microclimatic environments, including both sun-exposed and shaded areas, allowing users to select locations that best match their thermal preferences and activity patterns throughout the year. Such environmental diversity may be particularly relevant for older adults due to individual differences in thermal sensitivity and thermoregulation.

The future Barrendain Square should therefore preserve the existing vegetation structure while strengthening spatial continuity, improving the distribution of seating areas, reducing circulation fragmentation, and introducing uses and activities that support everyday sociability, light physical activity, and climate resilience.

Practical design strategies derived from the present study include maintaining a balanced combination of shaded and sun-exposed seating areas, improving pedestrian continuity and accessibility, incorporating new vegetation that provides seasonal climatic protection, and creating flexible spaces that support both passive and active forms of everyday public-space use.

The simulations indicate that the central tree-covered zone functions as a thermally moderated area, with substantially lower PET values than surrounding exposed surfaces during summer conditions. This finding supports the design strategy of preserving and strengthening the existing vegetation structure, while strategically locating new seating areas within or near these thermally protected zones. This design implication is also consistent with users’ expressed demand for more seating and more comfortable places to stay.

Finally, the findings suggest that design decisions should not rely exclusively on objective thermal indices. Combining environmental modelling with user-centred assessment methods provides a more robust basis for designing age-friendly public spaces that are both thermally comfortable and socially supportive. Although the present study focuses on older adults, many of the proposed strategies may also improve environmental comfort, usability and opportunities for everyday encounters among the wider population, thereby contributing to broader age-friendly and climate-resilient urban design.

Beyond the Barrendain case study, the proposed framework provides a practical methodology for municipalities seeking to integrate climate adaptation, healthy ageing and social inclusion into public-space planning. Its recognition as a good practice within the KORALE (Interreg Europe) project further demonstrates its transferability and potential to support age-friendly urban policies in other European contexts facing similar demographic and climatic challenges.

Future research should examine the role of public space in supporting intergenerational interaction and addressing different forms of social isolation through longitudinal and comparative research designs.

4.5. Limitations and Future Research

Although the sample size limits broad statistical generalisation, the study provides context-specific evidence regarding thermal perception, environmental quality, and patterns of public-space use among older adults. The findings should therefore be interpreted within the exploratory and case-study nature of the research.

Another limitation concerns the survey design. A non-probability intercept sampling strategy was adopted to characterise environmental perception under real-use conditions rather than to estimate population prevalence. Although most participants (76.9%) reported that they were passing through the intervention area, the duration of outdoor exposure prior to the interview was not recorded. Consequently, the possible influence of exposure time on subjective thermal perception could not be assessed and should be considered in future research. Furthermore, the observational and cross-sectional nature of the study precludes causal interpretation of the reported associations between environmental conditions, subjective perception and reported public-space use.

Additional limitations relate to the use of in situ questionnaires conducted under real environmental conditions, which may influence participation rates, seasonal balance, and subjective responses. Perceptual evaluations are inherently context-dependent and may be affected by behavioural, social, and temporal factors that are difficult to isolate through field surveys alone.

Future research will evaluate the impact of the ongoing redesign project once construction works are completed. This evaluation will combine behavioural observation, user surveys, environmental monitoring, and updated microclimatic simulations in order to assess changes in environmental comfort, length of stay, and social interaction, while also improving the calibration and validation of the simulation framework. Particular attention will be given to the replicability of the proposed methodological approach.

Furthermore, although PET simulations provide valuable information on spatial and seasonal thermal patterns, model-based approaches cannot fully capture the complexity of subjective comfort perception and adaptive behaviour in public space. The comparison between PET and TSV responses conducted in this study illustrates the importance of combining objective environmental indicators with user-centred assessment methods when evaluating outdoor comfort among ageing populations.

The ENVI-met simulations were primarily intended to support the comparative spatial analysis of outdoor thermal conditions within the study area rather than to provide fully calibrated predictions of pedestrian-level microclimatic conditions. Although the simulations were based on meteorological forcing data obtained from the nearby official weather station, a formal quantitative validation against on-site microclimatic measurements was not conducted. Accordingly, the PET results presented here should be regarded as comparative indicators of spatial and seasonal thermal patterns rather than fully validated estimates of absolute pedestrian-level thermal exposure.

This limitation is consistent with broader challenges identified in urban microclimate modelling, where simulation outputs are sensitive to boundary conditions, material properties, vegetation representation, and the modelling of radiative exchanges [57,58,59]. Accordingly, the PET results presented here should be interpreted as comparative indicators of spatial and seasonal thermal patterns rather than as fully validated estimates of absolute pedestrian-level thermal exposure. This interpretation is consistent with previous simulation-based studies that have used urban microclimate models to identify relatively exposed and protected areas, while acknowledging uncertainties associated with local parameterisation and validation procedures.

In addition, future analyses within the KALELAGUN project will compare the findings obtained at this study site with results from two additional case studies located in different urban contexts, including a consolidated urban public space and a peripheral public space. This comparative approach is expected to provide broader insights into the relationships between thermal comfort, public-space use, accessibility, and sociability, while contributing to the refinement of the methodological framework applied in this research.

5. Conclusions

This study highlights the value of integrating seasonal field surveys, environmental perception and microclimatic simulations as a baseline assessment to support evidence-informed urban regeneration for ageing populations, while acknowledging the uncertainties inherent in microclimatic simulation.

The findings of this study demonstrate that outdoor thermal perception among older adults is shaped by the combined influence of environmental conditions, behavioural adaptation and contextual factors. Older adults reported relatively high levels of outdoor thermal comfort throughout the year despite marked seasonal differences in environmental conditions, supporting previous evidence on adaptive thermal comfort in outdoor environments.

A key methodological contribution of this study lies in integrating subjective thermal responses, field-based environmental assessment and ENVI-met/BIO-met simulations within a single diagnostic framework applied before the redevelopment of a real public space.

The combined analysis of TSVs and simulation-derived PET showed that thermal perception among older adults should not be explained solely by objective environmental indicators. Neutral temperatures differed substantially between the cold- and warm-period campaigns (12.35 °C and 19.87 °C, respectively), while the comparison between PET and TSVs revealed only a partial correspondence between modelled thermal stress and subjective thermal sensation. Together, these findings suggest the role of behavioural adaptation, seasonal expectations, and contextual factors in shaping comfort, and show the value of combining environmental modelling with user-centred assessment methods when evaluating outdoor comfort and public-space performance.

From an urban design and planning perspective, the findings indicate that age-friendly public spaces should not aim to provide uniform thermal conditions but rather a diversity of microclimatic environments capable of accommodating seasonal variation and individual thermal preferences. Vegetation, a balanced distribution of shaded and sun-exposed seating, spatial continuity and accessibility should therefore be considered complementary strategies for creating thermally comfortable and socially supportive public spaces.

The PET simulations further demonstrate that vegetation can mitigate summer heat stress while preserving solar access during cooler periods, supporting seasonally responsive urban design. The simulations also highlight the importance of preserving seasonal microclimatic diversity rather than pursuing uniform thermal conditions throughout the intervention area.

The findings also indicate that environmental quality should be understood as an enabling condition rather than the sole determinant of perceived sociability. While favourable environmental conditions may facilitate opportunities for everyday encounters, existing social networks, familiarity and spatial organisation also contribute to how older adults experience public space. Consequently, environmental, spatial and social dimensions should be addressed together when designing age-friendly urban environments.

These findings should nevertheless be interpreted as exploratory, given that the assessment of the social dimension was based on a limited number of survey items.

The present study represents the baseline phase of a broader research process. Finally, in the context of climate change and the increasing frequency of extreme temperature events, the design of resilient and inclusive public spaces becomes increasingly important. Future research will evaluate the impact of the ongoing redesign of Barrendain Square through behavioural observation, user surveys, and updated microclimatic simulations. Additional comparative analyses developed within the KALELAGUN project will also assess other public-space typologies in order to improve the replicability and methodological robustness of the proposed approach.

As an exploratory baseline case study, the proposed methodological framework demonstrates how objective environmental assessment, microclimatic simulation and users’ subjective perceptions can be integrated to support evidence-informed urban regeneration. Beyond the specific case of Barrendain, this approach provides a transferable methodology for municipalities seeking to design climate-responsive, age-friendly and socially supportive public spaces while informing urban policies for healthy and active ageing.